Neurobiological Lessons Learned from Comparative Studies: Evolutionary Forces Shaping Brain and Behavior, 19th Annual Karger Workshop San Diego, Calif., ... 1, 2007 (Brain, Behavior and Evolution)

Neurobiological Lessons Learned from Comparative Studies: Evolutionary Forces Shaping Brain and Behavior 19th Annual Ka...

Author: Hans A. Hofmann | Caroly A. Shumway

35 downloads 538 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Neurobiological Lessons Learned from Comparative Studies: Evolutionary Forces Shaping Brain and Behavior 19th Annual Karger Workshop San Diego, Calif., November 1, 2007

Editors

Hans A. Hofmann, Austin, Tex. Caroly A. Shumway, Providence, R.I.

19 figures, 7 in color, and 4 tables, 2008

Basel • Freiburg • Paris • London • New York • Bangalore • Bangkok • Shanghai • Singapore • Tokyo • Sydney

S. Karger Medical and Scientific Publishers Basel • Freiburg • Paris • London New York • Bangalore • Bangkok Shanghai • Singapore • Tokyo • Sydney

Disclaimer The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publisher and the editor(s). The appearance of advertisements in the journal is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements. Drug Dosage The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.

Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center (see ‘General Information’). © Copyright 2008 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) Printed in Switzerland on acid-free and non-aging paper (ISO 9706) by Reinhardt Druck, Basel ISBN 978–3–8055–8999–4

Vol. 72, No. 2, 2008

Contents

89 Preface Hofmann, H.A. (Austin, Tex.); Shumway, C.A. (Providence, R.I.)

91 Gene Duplication, Co-Option and Recruitment during the Origin of the

Vertebrate Brain from the Invertebrate Chordate Brain Holland, L.Z.; Short, S. (La Jolla, Calif.) 106 Evolutionary Convergence of Higher Brain Centers Spanning the

Protostome-Deuterostome Boundary Farris, S.M. (Morgantown, W.V.) 123 Habitat Complexity, Brain, and Behavior Shumway, C.A. (Providence, R.I.) 135 Brains, Lifestyles and Cognition: Are There General Trends? Lefebvre, L. (Montréal, Qué.); Sol, D. (Barcelona) 145 Beyond Neuroanatomy: Novel Approaches to Studying Brain Evolution Pollen, A.A. (Stanford, Calif.); Hofmann, H.A. (Austin, Tex.) 159 Functional Tradeoffs in Axonal Scaling: Implications for Brain Function Wang, S.S.-H. (Princeton, N.J.) 168 Exploring the Origins of the Human Brain through Molecular Evolution Vallender, E.J. (Southborough, Mass.)

178 Author Index/Subject Index

© 2008 S. Karger AG, Basel Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Access to full text and tables of contents, including tentative ones for forthcoming issues: www.karger.com/bbe_issues

Brain Behav Evol 2008;72:89–90 DOI: 10.1159/000151469

Published online: October 7, 2008

Preface Hans A. Hofmann a Caroly A. Shumway b, c a

Section of Integrative Biology, Institute for Molecular and Cellular Biology, Institute for Neuroscience, The University of Texas at Austin, Austin, Tex., b The Nature Conservancy, and c Brown University, Department of Psychology, Providence, R.I., USA

Behavioral and neural diversity comprise one of evolution’s major outcomes. As a result, the origin and evolution of nervous systems – and particularly that of the human brain – has long attracted interest and debate [Striedter, 2004; Healy and Rowe, 2008]. While progress has been made in our understanding of how brain structure and function change over evolutionary time scales, progress also has been impeded by several obstacles. First, there is still no common agreement on appropriate brain measures [c.f. Deaner et al., 2007]. Various authors have pursued fractions, residuals, encephalization quotients, a progression index, whole brain volumes, and neocortex volumes. Secondly, our mechanistic insights into the functional consequences of observed size differences continue to be inadequate. Thirdly, we have a limited understanding of any differences beyond size. And finally, we have few behavioral assays available that allow unbiased comparisons across species. However, the field is now entering an exciting new stage where novel approaches can be utilized within a comparative context. By moving beyond neuroanatomy, the comparative method still provides the most powerful framework for understanding the evolution of brain and behavior. The 19th Karger Workshop, held in San Diego in October, 2007, highlighted the latest advances in several vertebrate and invertebrate systems, drawing broad lessons learned from these studies. The topics ranged from sensory to motor, from ecology to social aspects, from molecular to physiological to genomic approaches. The workshop was dedicated to the memory of Ted Bull© 2008 S. Karger AG, Basel 0006–8977/08/0722–0089$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/bbe

ock, whose relentless push to understand neural and behavioral complexity was a great inspiration to us and many other neurobiologists. In the first paper, Linda Holland and Stephen Short [2008] explore developmental similarities and differences between vertebrates and invertebrates. They show how two rounds of whole genome duplication and the creation of new splice forms in the vertebrate lineage led to the emergence of the neural crest. They present evidence that the genetic machinery underlying the neural crest was already present in the ancestral chordate-like Amphioxus (which lacks a neural crest). They emphasize that gene duplication allows the preservation of old functions by some duplicates and the acquisition of new ones by others. They argue that it is chiefly the increase in flexibility of old genes, rather than the evolution of entirely new genes, that allowed the evolution of new structures such as neural crest. In her review, Sarah Farris [2008] compares higherorder centers in vertebrates and invertebrates. She shows that the centralized nervous system of bilaterally symmetrical animals originated only once in evolution and exposes the fundamental groundplan consisting of a tripartite brain and a nerve cord divided into distinct antero-posterior and medio-lateral zones. The paper links comparative studies on parallelism and convergence in vertebrates that have associated evolutionary changes in brain structure and function with ecology to similar studies in invertebrates, which have independently evolved higher brain centers. Dr. Hans A. Hofmann Section of Integrative Biology The University of Texas at Austin, 1 University Station – C0930 Austin, TX 78712 (USA) Tel. +1 512 475 6754, Fax +1 512 471 3878, E-Mail [email protected]

Caroly Shumway [2008] reviews how habitat complexity influences both brain and behavior in African cichlid fishes, drawing on examples from primates and birds where appropriate. The paper demonstrates that environmental and social forces affect cichlid brains differently. She highlights the importance of quantifying complexity, addressing phylogenetic confounds, and using closely-related species and new experimental paradigms for testing the cognitive and survival value of brain and brain structure changes, both in the laboratory and in the wild. The Special Invited Guest of the workshop, Louis Lefebvre, together with his co-author, Daniel Sol, focus on convergent evolution of different types of cognitive abilities based on similar lifestyles [Lefebvre and Sol, 2008]. These authors have previously demonstrated the survival value of bigger brains. They highlight a few common principles that appear to have influenced the evolution of brains and cognition in widely divergent taxa, including the unpredictability of resources in time and space. They also emphasize the need for more work in the field if we are to understand the evolution of animal cognition. Alexander Pollen and Hans Hofmann [2008] outline four conceptual approaches that they believe will advance the field of brain evolution emerge from a historical focus on descriptive comparative neuroanatomy. They emphasize the need for (and provide examples of) reliable and efficient behavioral assays; the application of the comparative approach to developmental and physiological processes underlying species differences in brain and behavior; genome wide comparisons to identify the genetic basis for phenotypic differences; and identifying signatures of selection at the level of DNA sequence to uncover adaptive genetic changes that affect the nervous system. Finally, they also emphasize the importance of wellresolved phylogenies for comparative studies. Samuel Wang [2008] presents a biophysical approach to comparative and evolutionary neurobiology, focusing on

the costs of the construction and operation of neurons and the tradeoff between speed and energetic efficiency and volume. He applies biophysical reasoning to quantitative comparative data to identify candidate functional principles and uses cell and developmental biology to help distinguish functional principles from obligate principles. Finally, Eric Vallender [2008] describes the molecular evolution of genes involved in brain development. He explores the evolutionary forces that gave rise to the human brain by utilizing the genomic sequences of a several primates and other mammals in the search for signs of positive selection acting on DNA sequences. The review illuminates the strengths and weaknesses of these approaches and the dependence of the results on differing methodologies. He outlines a possible synthesis that would allow a more complete understanding of the genetic correlates behind the human brain and the selective events that have acted upon them. Taken together, these seven contributions to the 19th Karger Workshop provided an exciting forum for truly comparative neurobiology. Emphasizing the need for phylogenetically sound approaches and highlighting how novel methodologies can open up exciting new avenues of research created a lot of discussion and enthusiasm among participants and audience. We are confident that this volume will help pave the way towards integrative insights into the evolution of brain and behavior at many levels of biological organization. This workshop would not have been possible without the longstanding and continued support from the Karger Family. We believe we can speak for the entire community of comparative and evolutionary neurobiologists that Karger’s vision for our field has had a tremendous impact. We also want to thank the J.B. Johnston Club Program Committee for selecting this topic, and Walt Wilczynski and Blinda McClelland for expert support in both organizing the workshop and editing this volume.

References Deaner RO, Isler K, Burkhart B, van Schaik C (2007) Overall brain size, and not encephalization quotient, best predicts cognitive ability across non-human primates. Brain Behav Evol 70:115–124. Farris SM (2008) Evolutionary convergence of higher brain centers spanning the protostome-deuterostome boundary. Brain Behav Evol 72:106–122. Healy SD, Rowe C (2007) A critique of comparative studies of brain size. Proc R Soc B 274: 453–464.

90

Holland LZ, Short S (2008) Gene duplication, co-option and recruitment during the origin of the vertebrate brain from the invertebrate chordate brain. Brain Behav Evol 72:91–105. Lefebvre L, Sol D (2008) Brains, lifestyles and cognition: are there general trends? Brain Behav Evol 72:135–144. Pollen A, Hofmann HA (2008) Beyond neuroanatomy: Novel approaches to studying brain evolution. Brain Behav Evol 72:145– 158.

Brain Behav Evol 2008;72:89–90

Shumway CA (2008) Habitat complexity, brain, and behavior. Brain Behav Evol 72:123–134. Striedter G (2005) Principles of Brain Evolution. Sunderland MA: Sinauer Associates, Inc. Vallender EJ (2008) Exploring the origins of the human brain through molecular evolution. Brain Behav Evol 72:168–177. Wang SSH (2008) Functional tradeoffs in axonal scaling: implications for brain function. Brain Behav Evol 72:159–167.

Hofmann /Shumway



Gene Duplication, Co-Option and Recruitment during the Origin of the Vertebrate Brain from the Invertebrate Chordate Brain Linda Z. Holland Stephen Short Marine Biology Research Division, Scripps Institution of Oceanography, University of California at San Diego, La Jolla, Calif., USA

Key Words Lancelet ⴢ Neural crest ⴢ Midbrain/hindbrain boundary ⴢ MHB ⴢ Amphioxus ⴢ Tunicate ⴢ Alternative splicing ⴢ Genome duplication

Abstract The brain of the basal chordate amphioxus has been compared to the vertebrate diencephalic forebrain, midbrain, hindbrain and spinal cord on the basis of the cell architecture from serial electron micrographs and patterns of developmental gene expression. In addition, genes specifying the neural plate and neural plate border as well as Gbx and Otx, that position the midbrain/hindbrain boundary (MHB), are expressed in comparable patterns in amphioxus and vertebrates. However, migratory neural crest is lacking in amphioxus, and although it has homologs of the genes that specify neural crest, they are not expressed at the edges of the amphioxus neural plate. Similarly, amphioxus has the genes that specify organizer properties of the MHB, but they are not expressed at the Gbx/Otx boundary as in vertebrates. Thus, the genetic machinery that created migratory neural crest and an MHB organizer was present in the ancestral chordate, but only co-opted for these new roles in vertebrates. Analyses with the amphioxus genome project strongly support the idea of two rounds of whole genome duplication with subsequent gene losses in the vertebrate lineage. Duplicates of developmental genes were preferentially re-



tained. Although some genes apparently acquired roles in neural crest prior to these genome duplications, other key genes (e.g., FoxD3 in neural crest and Wnt1 at the MHB) were recruited into the respective gene networks after one or both genome duplications, suggesting that such an expansion of the genetic toolkit was critical for the evolution of these structures. The toolkit has also increased by alternative splicing. Contrary to the general rule, for at least one gene family with key roles in neural crest and the MHB, namely Pax genes, alternative splicing has not decreased subsequent to gene duplication. Thus, vertebrates have a much larger number of proteins available for mediating new functions in these tissues. The creation of new splice forms typically changes protein structure more than evolution of the protein after gene duplication. The functions of particular isoforms of key proteins expressed at the MHB and in neural crest have only just begun to be studied. Their roles in modulating gene networks may turn out to rival gene duplication for facilitating the evolution of structures such as neural crest and the MHB. Copyright © 2008 S. Karger AG, Basel

Introduction

Biologists have long been interested in the evolutionary origin of the vertebrates generally, and, more particularly, in the origin of vertebrate nervous systems from Linda Z. Holland Marine Biology Research Division Scripps Institution of Oceanography, University of California San Diego La Jolla, CA 92093-0202 (USA) Tel. +01 858 534 5607, Fax +01 858 534 7313, E-Mail [email protected]

dorsal nerve cord

segmental muscles notochord

Fig. 1. Adult amphioxus in side view; anterior to the left. The labeled structures have counterparts in developing vertebrates.

perforate pharynx

some ancestral nervous system in an invertebrate. During the last two centuries, numerous invertebrate phyla have been proposed as the proximate ancestors of the vertebrates [reviewed in Gee, 1996]. Recent molecular phylogenetic analyses place amphioxus basal in the chordates and tunicates as the sister group of vertebrates. Even so, it is now generally agreed that tunicates are quite divergent and that the ancestral vertebrate was an invertebrate chordate that more closely resembled modern amphioxus. The two living groups of invertebrate chordates, which split from the vertebrate lineage about half a billion years ago, are the cephalochordates (amphioxus) and the tunicates. Several recent analyses based on large numbers of nuclear genes [Blair and Hedges, 2005; Bourlat et al., 2006; Delsuc et al., 2006] indicate that cephalochordates comprise the outgroup to a tunicate + vertebrate clade. Tunicates are evolving rapidly and have lost many genes, for example, Hox7, 8, 9 and 11 in ascidians [Ikuta and Saiga, 2005] and Hox3, 5, 6, 7, 8 in appendicularians [Seo et al., 2004], and their body plans are simplified. The central nervous system has only about 330 neurons in ascidians and fewer than 150 in appendicularians [Søviknes et al., 2005, 2007; Imai and Meinertzhagen, 2007]. Therefore, even though tunicates are the sister group of vertebrates, the more slowly evolving cephalochordates appear to have preserved a more complete record of how the vertebrates evolved from their invertebrate ancestors [Schubert et al., 2006a]. The present review addresses the question of how the vertebrate brain with several million neurons to several billion (small and large mammals, respectively) [Williams and Herrup, 1988], evolved from an ancestral chordate brain similar to that of a modern amphioxus with about 20,000 neurons [Nicol and Meinertzhagen, 1991]. In particular, we discuss how the genetic toolkit has increased during the transition from invertebrate chordates to vertebrates [Holland et al., 2008; Putnam et al., 92


gonads

anus

2008; Short and Holland, 2008] and how this enlarged toolkit may have allowed the early vertebrates to invent such novelties as the midbrain-hindbrain organizer and neural crest.

The Amphioxus CNS Shares Key Domains with the Vertebrate CNS

Amphioxus shares several anatomical features with vertebrates, including a pharynx perforated with gill slits, segmentally arranged trunk muscles, a notochord, and a dorsal, hollow nerve cord (fig. 1). The anterior end of the amphioxus nerve cord is slightly swollen (this swelling is exaggerated in fig. 1) and is called the cerebral vesicle in larval amphioxus [Wicht and Lacalli, 2005]. During the nineteenth and much of the twentieth century, there was considerable debate among anatomists as to whether the amphioxus brain existed at all and, if it did, whether it had any resemblance to a vertebrate brain. Then, about 15 years ago, two different approaches began to reveal vertebrate-like features in the amphioxus brain. The first, pioneered by Holland et al. [1992], used the expression domains of developmental genes to correlate regions of the amphioxus and vertebrate brains and the second, pioneered by Lacalli et al. [1994], was the comprehensive description of the larval amphioxus brain by computerassisted, three dimensional reconstructions based on serial transmission electron microscopy. Taken together, these results indicate, that although the amphioxus CNS has relatively few neurons [Nicol and Meinertzhagen, 1991], it has likely homologs of the vertebrate diencephalon, hindbrain and spinal cord and perhaps a small midbrain as well, but it lacks a telencephalon (summarized in fig. 2). Evidence from both morphology and gene expression strongly supports the homology of most of the amphiHolland/Short

TFB MB

ERR Islet Mnx Shox Wnt3

2 3

GBX

SC

SC

Hoxb6

Hoxb4

5

Cdx1

HB

HB

Hoxb1 Hoxb2

4 Hoxb3

MB

DFB

1 DFB

GBX

AmphiCdx AmphiHox6 AmphiHox4 AmphiHox3 AmphiHox2 AmphiHox1 Wnt3 Shox Mnx Islet

ERR

Six3/6 Pax6 Otx BF1

Amphioxus

BF1 Otx2 Pax6 Six3/6

Generalized vertebrate

Fig. 2. Expression of anterior/posterior patterning genes in the CNS of amphioxus and a generalized vertebrate. Dorsal views. (TFB = Telencephalic forebrain; DFB = diencephalic forebrain; MB = midbrain; HB = hindbrain; SC = spinal cord). The numbers refer to some proposed structural homologies [Wicht and Lacalli, 2005] as follows: (1) vertebrate paired eye versus amphioxus frontal eye; (2) vertebrate epiphysis versus amphioxus lamellar body; (3) vertebrate subcomissural organ versus amphioxus infundibular organ; (4) vertebrate primary motor center versus amphioxus primary motor center; (5) vertebrate rhombomeric motoneurons versus periodically repeating dorsal compartment motoneurons of amphioxus. The anteroposterior extent of each gene expression

domain is diagrammed as a stripe (black for anterior markers, stippled for mainly hindbrain markers, and grey for Hox and Parahox genes). The expression domains for the vertebrate (at top) largely correspond in extent to those of the amphioxus (at bottom). The domains for the amphioxus genes (with references) are as follows: BF1 [Toresson et al., 1998]; Otx [Williams and Holland, 1996; Castro et al., 2006]; ERR [Bardet et al., 2005]; Islet [Jackman et al., 2000]; Mnx [Ferrier et al., 2001]; Shox [Jackman and Kimmel, 2002]; Hox genes [Schubert et al., 2006b]; Cdx [Brooke et al., 1998]; Wnt3 [Schubert et al., 2001]; Gbx [Castro et al., 2006]; Six3/6 [Kozmik et al., 2007]; Pax6 [Glardon et al., 1998]. After Shimeld and Holland [2005].

oxus cerebral vesicle with the vertebrate diencephalon. At the cellular level, the cerebral vesicle has counterparts with several diencephalic structures. Most striking perhaps is the lamellar body of amphioxus as compared to the pineal gland in the vertebrate diencephalon [Ruiz and Anadon, 1991; Lacalli et al., 1994; Olsson et al., 1994]. The lamellar body, which is a ciliary photoreceptor with the ciliary membrane modified into large whorls, strikingly resembles photoreceptor cells in the lamprey pineal complex [Meiniel, 1980; Cole and Youson, 1982]. In addition, it has been argued that the anterior photorecep-

tor of amphioxus is homologous to the vertebrate paired eyes [Lacalli, 1996] and that the amphioxus infundibular organ, which secretes an extracellular fiber that extends posteriorly in the lumen of the CNS, is homologous to the vertebrate subcommissural organ, which secretes Reissner’s fiber [Olsson et al., 1994]. Additional correspondences of gene expression support the homology of the amphioxus cerebral vesicle and vertebrate diencephalon. These include Pax6, Six3, BF1 in the forebrain and Otx in the forebrain and midbrain (fig. 2) [Benito-Gutiérrez, 2006; Kozmik et al., 2007]. In contrast, the homology of

Chordate Brain Evolution

Brain Behav 2008;72:91–105

93

Vertebrate

Pax2

Wnt1

TFB DFB SC

Gbx1

Gbx2

HB

en2 en1 Fgf8 Fgf18 Fgf17 Pax5 Pax8

MB

M DFB B HB SC

fgf8/17/18

en1/2

Pax2/5/8

Otx1/2

A

Otx2 Otx1

Amphioxus

B en

Wnt1 Otx2

fgf8

Gbx2

Pax2

Fig. 3. Expression of genes patterning the midbrain/hindbrain

boundary (MHB) in amphioxus and a generalized vertebrate. A Dorsal views. Abbreviations as in figure 2. The vertebrate para-

logs of the single amphioxus genes are expressed in patterns that overlap in space and in time. B Gene interactions at the MHB in vertebrates. Otx2 and Gbx2 mutually repress one another. The other genes activate one another and Otx2 and Gbx2. The functions of the other paralogs are not shown as they are not as wellstudied. [After Liu et al., 1999; Suda et al., 1999; Joyner et al., 2000; Ye et al., 2001; Liu et al., 2003; Raible and Brand, 2004; HidalgoSanchez et al., 2005; Nakamura et al., 2005; Islam et al., 2006].

the vertebrate midbrain with the posterior part of the amphioxus cerebral vesicle is less clear [Lacalli, 1996; Northcutt, 2003]. This region contains cells that apparently are targets for terminals from the frontal eye photoreceptor cells. Lacalli termed these cells tectal cells, to imply a possible homology with the vertebrate tectum [Lacalli, 1996]. Unfortunately, there are no known genes that are exclusive markers of the midbrain. Even so, expression of Otx and Gbx suggests that the posterior limit of the amphioxus cerebral vesicle corresponds to the vertebrate midbrain/hindbrain boundary (MHB). The Gbx2 94


and Otx2 domains abut at the vertebrate MHB and at the boundary between the cerebral vesicle and hindbrain in amphioxus [Castro et al., 2006] (fig. 3). In vertebrates, expression of Otx2 and Gbx2 positions the MHB and maintains expression of MHB markers that turn on later (for example, En2, Wnt1, Pax2 and Fgf8) and confer organizer properties on the MHB [Li and Joyner, 2001]. However, although the vertebrate MHB acts as an organizer – if transplanted into the diencephalon, it induces midbrain properties [reviewed in Nakamura et al., 2005] – there is no evidence that the comparable region in amphioxus also acts as an organizer as En2, Wnt1, Pax2 and Fgf8 are not expressed at the Otx/Gbx boundary in the amphioxus CNS (fig. 3). Although the existence of a midbrain homolog in amphioxus is uncertain, there is little doubt that amphioxus has homologs of the vertebrate hindbrain and spinal cord. The presence of a zone containing motor neurons just posterior to the cerebral vesicle is good morphological evidence for a hindbrain. There are two types of motor neurons in larval amphioxus. The ventral compartment motor neurons appear to innervate muscle fibers involved in fast swimming, and the dorsal compartment motor neurons, which are the first to differentiate, appear to be involved in slow undulatory swimming [Lacalli, 2003]. The six pairs of dorsal compartment motor neurons, which are the first to differentiate, are segmentally arranged in the amphioxus hindbrain [Lacalli and Kelly, 1999; Bardet et al., 2005]. Similarly, in vertebrates, groups of hindbrain motor neurons are segmentally arranged in pairs of rhombomeres [Guthrie and Lumsden, 1992]. Supporting the hindbrain homologies, in both amphioxus and vertebrates, these motor neurons express the ERR gene, which is expressed in few other cells in the CNS [Bardet et al., 2005]. Moreover, domains of gene expression strongly support the existence of a hindbrain in amphioxus. Even though the amphioxus hindbrain is not divided into rhombomeres like that of the vertebrate hindbrain, the anterior limits of Hox genes are staggered within it as they are within the hindbrain of lampreys and other vertebrates [Deschamps et al., 1999; Takio et al., 2004; Schubert et al., 2006b]. Numerous other hindbrain and spinal cord markers are similarly expressed in amphioxus and vertebrates. These include Gbx, Islet, Mnx, shox, Wnt3 and cdx [Castro et al., 2006] (fig. 2). Together with the presence of segmentally arranged motor neurons in amphioxus and vertebrates, the expression of these Hox genes leaves little doubt that the common ancestor of amphioxus and vertebrates had a hindbrain that was fundamentally segmented. Holland/Short

Patterning of the Neural Plate Is Comparable in Amphioxus and Vertebrates, with the Exception that Amphioxus Lacks Neural Crest and an MHB Organizer

Amphioxus has a neural plate, but lacks a clear homolog of vertebrate neural crest. Therefore, it is not surprising that although genes specifying the neural plate and the boundary between neural plate and adjacent nonneural ectoderm are expressed similarly in amphioxus and vertebrates, late neural crest specifiers and effectors are not (table 1). For example, in both vertebrates and amphioxus, down-regulation of BMPs dorsally is required for specification of the neuroectoderm [Meulemans and Bronner-Fraser, 2004; Yu et al., 2007], and in both, the neural plate specifier Sox1/2/3 is expressed in the presumptive neural plate, and Pax3/7, Dlx, Zic, Msx and Wnt6 are expressed at the edge of the neural plate and/or in adjacent non-neural ectoderm [Holland et al., 1996, 1999, 2000; Sharman et al., 1999; Schubert et al., 2001; Meulemans and Bronner-Fraser, 2004]. In contrast, homologs of most vertebrate neural crest specifiers (i.e., AP-2, Id, FoxD3 and Twist) are not expressed at the edges of the neural plate or in the dorsal neural tube in amphioxus [Yasui et al., 1998; Meulemans and Bronner-Fraser, 2002; Gostling and Shimeld, 2003; Meulemans et al., 2003] (table 1). A possible exception is snail, which is initially expressed at the edges of the neural plate in amphioxus, but later is expressed throughout the neural tube [Langeland et al., 1998]. As noted above, in both amphioxus and vertebrates, an anterior Otx domain abuts a posterior Gbx domain at the anterior end of the hindbrain [Castro et al., 2006]. This suggests that the MHB territory was already established in the ancestral chordate. Otx and Gbx mutually repress one another, and the position of the MHB is established by the posterior limit of Otx [Broccoli et al., 1999]. Organizer properties are conferred by a suite of genes including en2, Fgf8, Lmx1, Wnt1, Pax2 and their paralogs [Joyner et al., 2000; Xu et al., 2000; Nakamura et al., 2005; Guo et al., 2007; Murcia et al., 2007] (table 2, fig. 3). Although in vertebrates, ectopic expression of Otx alone can activate the expression of several of these MHB markers (i.e., en2, Wnt1, Pax2, Fgf8 and Gbx2) [Tour et al., 2002], in amphioxus, only Fgf8/17/18 among these genes is co-expressed with Otx [Meulemans and Bronner-Fraser, 2007] (fig. 3). This indicates that most of the regulatory interactions between Otx and the MHB markers that confer organizer properties on the vertebrate MHB have not been established in amphioxus. Chordate Brain Evolution

Table 1. Vertebrate neural plate and neural crest markers

Neural plate specifiers

Neural plate border specifiers

Early neural crest specifiers

Late neural crest specifiers

Sox1/2/3* Zic*

Dlx* Pax3/7* Msx* Zic*

Id c-Myc AP-2 Snail

FoxD3 SoxE Twist

* Genes expressed in comparable positions in amphioxus and vertebrates.

Table 2. Paralogs of vertebrate neural crest and midbrain/hind-

brain boundary markers Amphioxus gene

Paralogs of vertebrate neural crest markers

FoxD Twist

FoxD1 Twist1a

FoxD2 Twist2

Myc

c-myca

Id AP-2 SoxE

Id-1a AP-2␣a Sox8a

N-myc (lamprey)a Id-2a Id-3 AP2-␤a AP-2␥a a Sox9 Sox10a

FoxD3a Twist3 (chick) L-myc

FoxD4

FoxD5

Myc Id-4a AP-2␦

Paralogs of vertebrate MHB markers engrailed fgf8/17/18 pax2/5/8 Wnt1 a b

en1b fgf8b pax2b Wnt1b

en2b fgf17b pax5b

fgf18b pax8b

Paralogs expressed in presumptive neural crest. Paralogs expressed at the MHB.

Taken together, the differences in gene expression at the edges of the neural plate and at the midbrain/hindbrain boundary between amphioxus and vertebrates suggest that evolution of neural crest and organizer properties at the MHB is not due to the evolution of entirely new genes (the genes are present in the amphioxus genome) but to the expansion and modification of existing gene networks by the recruitment of old genes for new functions.

Brain Behav 2008;72:91–105

95

Expansion of the Genetic Toolkit

The answer to the question of how old genes might adopt new functions without losing the old ones may lie, at least in part, in genome duplications. The idea that two rounds of whole genome duplications occurred during the evolution of vertebrates (the 2R hypothesis) was first put forth by Ohno [1970]. This idea has been hotly debated and was particularly cast in doubt by the sequencing of the human genome, which showed only about 28,000–34,000 genes [Roest Crollius et al., 2000] compared to about 15,000–16,000 in Drosophila species [Hahn et al., 2007] and about 20,000 in species of Caenorhabditis [Gupta et al., 2007]. However, the reported death of the 2R hypothesis [Hughes and Friedman, 2003] has turned out to be premature. Analyses done in connection with the amphioxus genome project have shown that the amphioxus genome, which has about 21,600 protein-coding genes, has not undergone any whole genome duplications [Putnam et al., 2008]. Depending on the species, amphioxus (Branchiostoma) genomes have a haploid number of 18 or 19 chromosomes [Howell and Boschung, 1971; Colombera, 1974; Saotome and Ojima, 2001]. Remarkably, the amphioxus and vertebrate genomes demonstrate a very high degree of macro-synteny or correspondence at the chromosome level. Thus, comparison with vertebrate genomes allowed the amphioxus scaffolds to be grouped into seventeen linkage-groups or proto-chromosomes. For each of these linkage-groups, the bulk of the genes correspond to those on 4 or 5 human chromosomes, although gene order has been scrambled to some extent [Putnam et al., 2008]. This quadruple conserved synteny over the entire amphioxus genome and those of human and teleost fish strongly supports the 2R hypothesis and indicates that the wholegenome duplications occurred after the split between amphioxus and the lineage leading to jawed vertebrates [Putnam et al., 2008]. Although it is generally agreed that one whole genome duplication occurred at the base of the vertebrates, the available data are ambiguous as to whether the second round occurred before or after the divergence of agnathans and gnathostomes [Escriva et al., 2002; Furlong and Holland, 2002; Putnam et al., 2008]. Because the human genome only has about 30– 40% more genes than the genomes of Drosophila or Caenorhabditis, subsequent to the two rounds of whole genome duplication, there was evidently massive loss of duplicates.

96


Duplicates of Developmental Genes Were Preferentially Retained in the Vertebrate Lineage

Genome-wide comparisons between the amphioxus and vertebrate genomes established in connection with the amphioxus genome project have indicated that gene losses subsequent to whole genome duplications did not occur evenly across the genomes of early vertebrates [Putnam et al., 2008]. Instead, duplicates of transcription factor genes and those involved in signal transduction as well as other developmental genes and genes involved in neuronal activities were preferentially retained, while those involved in house-keeping functions were disproportionately lost [Putnam et al., 2008]. Thus, for the genes that might be most important for shaping an embryo, vertebrates tend to have 2–3 copies compared to one in amphioxus [Putnam et al., 2008]. This raises the question of why duplicates of developmental genes have been disproportionately retained. The answer might lie in subfunctionalization and/or neofunctionalization as proposed by Force et al. [1999]. These authors suggested that subsequent to gene duplication, subfunctionalization in which either their several expression domains are subdivided amongst the duplicates or the duplicates retain the same expression domains but partition the timing of expression, occurs more often than neofunctionalization in which some of the duplicates acquire new expression domains (diagrammed in fig. 3). However, it has been proposed that although subfunctionalization will probably peak and then decrease following gene duplication, neofunctionalization with the duplicate genes acquiring new regulatory elements will continue to increase with time and that both subfunctionalization and neofunctionalization can co-occur in a given gene family [MacCarthy and Bergman, 2007]. This model predicts a relatively high level of neofunctionalization for ancient duplicates. Neofunctionalization typically occurs by the acquisition of new regulatory elements. This can occur by cooption and perhaps also mutation of repetitive sequences/transposable elements [Dayal et al., 2004; Jurka, 2007]. For example, we found a transposable element about 1.5 kb upstream of the start codon of the amphioxus FoxD gene in one individual, which was not present in the FoxD gene in either allele of the individual used for the genome sequence [Holland, 2006]. This Alu element contains potential binding sites for numerous transcription factors, including Tcf/Lef, which mediates Wnt/␤-catenin signaling. Presumably, the presence of this element in the FoxD gene was neither advantageous nor deleterious, but one Holland/Short

can envision that the occasional incorporation of a transposon in such a location could give rise to new regulatory interactions conferring a selective advantage. An example is the RSR sequence found in a single family of sea urchins, but not others. In one genus within this family, an RSR located within the spec2a gene was shown to include functional transcription factor binding sites – an example of recent evolution of repetitive sequences to include new regulatory elements [Dayal et al., 2004]. Importantly, neofunctionalization is not limited to gene regulatory elements; the protein itself can evolve new functions. For example, amphioxus has a single retinoic acid receptor (RAR), whereas vertebrates have three duplicates (RAR␣, RAR␤, RAR␥). Of these, both the ligand-binding selectivity of RAR␤ and its embryonic expression are most like those of amphioxus RAR, suggesting that RAR␤ has conserved the ancestral functions, but that RAR␣ and RAR␥ have undergone neofunctionalization both in regard to regulatory DNA and to protein function [Escriva et al., 2006].

Comparisons of the number of gene duplicates in amphioxus and vertebrates can give an idea of when neural crest and MHB organizer genes were recruited into gene networks operating respectively at the edges of the neural plate and at the MHB. For example, if the single amphioxus gene is not expressed at the edges of the neural plate and only one of three or four vertebrate paralogs is expressed there, then the most parsimonious explanation is that the gene was co-opted into the neural crest network after the two rounds of genome duplications. Conversely, if the single amphioxus gene is not expressed at the edges of the neural plate, but three of the four vertebrate paralogs are co-expressed there, the gene was probably recruited to the neural plate boundary before gene duplications occurred. Neural crest markers show both patterns (table 1). Examples of the latter include SoxE, AP-2 and Id (table 2). Although expression of none of these genes is localized to the edges of the neural plate in amphioxus, SoxE and Id are expressed in the neural plate, and AP-2 is expressed in adjacent non-neural ectoderm. Thus, incorporation of these genes into a neural crest network may not have required major changes in the network. For example, amphioxus SoxE is initially expressed in the floor plate, and later throughout the neural tube [Meule-

mans and Bronner-Fraser, 2007], indicating that a change in timing might be all that is required for recruitment to a neural crest network. Similarly, amphioxus Id is expressed throughout the anterior neural plate, but not elsewhere in the neural plate [Meulemans et al., 2003]. In the lamprey, Id is expressed in the edges of the neural plate [Meulemans et al., 2003], suggesting that co-option of Id for a role in neural crest probably evolved before at least the second round of whole genome duplications in the vertebrate lineage. Amphioxus Ap-2, like Dlx, is not expressed in the CNS in amphioxus, but throughout the non-neural ectoderm, including the edges adjacent to the neural plate, which detach from the neuroectoderm and migrate over the neural plate as sheets of ectoderm [Holland et al., 1996; Meulemans and Bronner-Fraser, 2002]. In vertebrates, the boundary between neuroectoderm and non-neuroectoderm is not clearly delimited. Thus, because Dlx is expressed in the non-neural ectoderm adjacent to the neural plate in amphioxus and in neural crest in vertebrates, whereas Wnt6 is expressed in the edges of the neural plate in amphioxus [Schubert et al., 2001] but in the adjacent non-neural ectoderm in vertebrates [Garcia-Castro et al., 2002], neural crest might have evolved from cells on either side of this boundary [Holland et al., 1996]. Therefore, like Id and SoxE, AP-2 was probably already expressed in the ancestral chordate in more or less the right place to be co-opted into neural crest. Because in the tunicate Ciona intestinalis, the SoxE gene is broadly expressed in the CNS, whereas AP-2 is not expressed in the CNS at all [Imai et al., 2004] (photographs of in situs at http://ghost.zool.kyoto-u.ac.jp), at least these two genes were probably co-opted for neural crest at the base of the vertebrates, before the whole genome duplications. In contrast to the three genes above, Myc, FoxD and twist appear to have evolved roles in neural crest after at least one whole genome duplication in the vertebrate lineage; only one each of three twist and five FoxD genes and two of three Myc genes are expressed in vertebrate neural crest or its progenitors (table 1). In the lamprey, N-Myc, an early neural crest specifier is co-expressed with SoxB in the edges of the neural plate, while the late neural crest specifier, FoxD3, is expressed in premigratory neural crest, and the four lamprey Twist genes are expressed in late migratory (Twist A) or postmigratory (all four Twist genes) neural crest and neural crest derivatives, much later than their gnathostome homologs. In gnathostomes, FoxD3 is the only FoxD gene expressed in neural crest. In zebrafish, mutants for FoxD3 have normal numbers of premigratory neural crest cells,


Brain Behav 2008;72:91–105

Neofunctionalization and Recruitment of Genes into Gene Networks at the MHB and the Boundary of Neural- and Non-Neural Ectoderm

97

but with a reduction in snail1b expression and in the numbers of migratory trunk neural crest [Stewart et al., 2006]. Conversely, overexpression of FoxD3 in the chick promotes development of neural crest [Dottori et al., 2001]. In vertebrates, FoxD3 and twist have been intercalated at different levels into the gene network operating in premigratory neural crest. FoxD3 is downstream of Pax3, Zic1 and snail [Sato et al., 2005], and twist is also downstream of snail [Aybar et al., 2003]. This suggests a model whereby, following the whole genome duplications in vertebrates, Myc, twist and FoxD underwent neofunctionalization due to acquisition of new regulatory elements directing expression of some paralogs to neural crest progenitors. Changes in protein function have also occurred after duplication of Myc [Cole and McMahon, 1999]. However, comparative studies of the function of Twist and FoxD proteins have not been done. Expression of four amphioxus homologs of genes involved in conferring organizer properties on the vertebrate MHB (Pax2/5/8 and Fgf8/17/18, en and Wnt1) has been determined (fig. 3). None is expressed at the amphioxus MHB in a pattern comparable to that of their vertebrate homologs, suggesting that the amphioxus MHB lacks organizer properties. Pax2/5/8 is expressed throughout the length of the amphioxus hindbrain and spinal cord [Kozmik et al., 1999]. This last domain appears to be comparable to the relatively weak expression domain of vertebrate Pax2 in the hindbrain and spinal cord but not to the zone of high expression of vertebrate Pax2 and Pax5 at the MHB [Heller and Brändli, 1997, 1999]. The single amphioxus fgf8/17/18 gene is expressed throughout the length of the cerebral vesicle [Meulemans and Bronner-Fraser, 2007]. This domain appears to correspond to that of vertebrate fgf8 in the forebrain, not to the one at the MHB, which evidently arose later, possibly within the tunicate lineage (see below). en is expressed in a few cells in the amphioxus forebrain plus a few in the hindbrain, as well as in the anterior somites and a stripe in the ectoderm [Holland et al., 1997]. In vertebrates, in addition to the domain at the MHB, the two en genes are expressed in developing neurons the hindbrain as well as in muscles [Ekker et al., 1992]. Some of these domains (e.g., en2 in the amphioxus somites and in certain head muscles of vertebrates), might be comparable, but the domain at the vertebrate MHB and in the amphioxus cerebral vesicle are in the wrong position to be counterparts.Wnt1 is an exception to the general rule that duplicates of transcription factors and signaling molecules have been preferentially retained in vertebrates. In amphioxus, Wnt1, 3, 6, 9 and 10 are located on scaffold 12. In human, homologs of 98


these genes are on 4 chromosomes, showing that they have been duplicated two-fold. However, two of the four copies of Wnts 3, 9 and 10 have been lost, as have three of the four copies of Wnt1 and Wnt6. In amphioxus, Wnt1 is expressed only in the tailbud [Holland and Holland, 2000], reflecting the ancient role of Wnt1 in specification of the end of the embryo around the blastopore or, in cnidarians, the single gut opening [Holland, 2002]. Vertebrates have evidently lost this Wnt1 domain and gained the domain at the MHB. It might be that after gene duplication, one copy of Wnt1 acquired a new function at the MHB, and subsequently the copy retaining the ancestral domain around the blastopore was lost. It has been very controversial whether tunicates have an MHB. However, now that the phylogenetic positions of amphioxus and tunicates have been reversed, it appears that although some of the MHB organizer genes may have been expressed at the Otx and Gbx boundary in the common ancestor of tunicates and vertebrates, tunicates have diverged so much that it is difficult to say how much of the MHB network this ancestor had. Expression of genes patterning the tunicate CNS differs not only between two major groups of tunicates (ascidians and appendicularians) but also from one ascidian species to another. In regard to the MHB, although the anterior part of the CNS in all tunicates studied expresses Otx, the Gbx gene has been lost from tunicate genomes. Wnt1 has apparently also been lost at least from the genome of the tunicate C. intestinalis, although there are some Wnts that cannot be assigned to any particular class of Wnt genes and could represent either extremely divergent homologs of amphioxus Wnts or be independent duplicates [Hino et al., 2003]. Expression of other MHB markers differs not only between appendicularian tunicates and ascidians, but also between one species of Ciona and another. For example, Pax2/5/8 is expressed only at the anterior tip of the CNS in the appendicularian Oikopleura dioica, but posterior to the Otx region in C. intestinalis, in a zone termed the ‘neck’, between the two domains expressing the single en gene [Imai et al., 2002]. This zone has been considered homologous to the vertebrate MHB because not only is en expressed there, but it is co-expressed with Fgf8/17/18 [Imai et al., 2002]. However, a problem with this interpretation is that in C. savignyi and in the appendicularian Oikopleura dioica as well as in C. intestinalis, this posterior domain of en is within that of Hox1, a hindbrain marker [Imai et al., 2002; Jiang and Smith, 2002; Cañestro et al., 2005]. Moreover, the anterior en domain is missing in C. savignyi and O. dioica while in O. dioica, Holland/Short

Pax2/5/8 is only expressed at the anterior tip of the CNS. Thus, given the basal position of appendicularians within the tunicates (albeit with a long, rather poorly-supported branch) [Wada, 1998; Swalla et al., 2000] and the loss of Wnt1 and Gbx from tunicate genomes, it is very difficult to infer homologies of the tunicate CNS with those of vertebrates and amphioxus [discussed in Ikuta and Saiga, 2007]. It might never be determined how much of the MHB was present at the base of the vertebrates, before the genome duplicated. What can be said with reasonable certainty is that the Gbx and Otx domains, which position the MHB, were present in the ancestral chordate and that, because all three of the vertebrate Fgf8/17/18 and Pax2/5/8 homologs are expressed at the vertebrate MHB, their single homologs were probably expressed there in the ancestral vertebrate. Whether Wnt1 and/or en were expressed near this boundary in the ancestral vertebrate is uncertain because only one Wnt1 gene and two engrailed genes have persisted in the vertebrate genome subsequent to the two rounds of whole genome duplication. The new domain of Wnt1 at the Gbx/Otx boundary subsequent to duplication may have played a key role in evolution of the MHB. Wnt1 expression is required to maintain that of Fgf8, Pax2 and en2 at the MHB [Lekven et al., 2003; Canning et al., 2007]. In amphioxus, Fgf/17/18 and Pax2/5/8 are expressed in the Otx and Gbx zones respectively, and co-option of Wnt1 for a role in the midbrain might have influenced the expression of both. Several genes, including Lmx1b and Pbx, which may cooperate with en to regulate expression of Wnt1 [Tour et al., 2002; Erickson et al., 2007; Guo et al., 2007]. Expression of other vertebrate MHB markers such as Pbx and Lmx1b has not been determined in amphioxus, and therefore, the possibility remains that these genes might also have been co-opted for roles in the MHB gene network subsequent to genome duplications. Co-option of Pax2/5/8, Fgf8/17/18 and en for roles at the MHB likely occurred before gene duplication in the vertebrate lineage as all the paralogs of each of these genes – two en genes, three Pax2/5/8 genes and three Fgf8/17/18 genes – are co-expressed in overlapping spatio-temporal patterns at the vertebrate MHB (fig. 3). en1 and en2, which are transcription factors, have at least partially redundant functions at the MHB [Liu and Joyner, 2001; Murcia et al., 2007]. Pax2/5/8 genes are upstream of en2, Wnt1, and Fgf8. Overexpression of vertebrate pax5 in the chick transformed the diencephalon into a tectum-like structure and caused ectopic expression of en2, Wnt1 and Fgf8 and downregulation of Otx2 in the diencephalon and mesencephalon [Funahashi et

al., 1999]. Correspondingly, in Pax2 null mutants, en2 is downregulated [Bouchard et al., 2005]. Although gene paralogs may be co-expressed, their functions can be quite different. For example, the roles of Fgf8, Fgf17b and Fgf18, which are co-expressed in overlapping domains at the MHB, are not identical [Liu et al., 2003]. First, the expression patterns of Fgf8 and Fgf17 at the MHB, although they overlap, demonstrate both temporal and spatial differences [Xu et al., 1999]. Second, loss of Fgf17 function, but not that of Fgf18 causes the posterior midbrain to be truncated [Xu et al., 2000; Ohbayashi et al., 2002]. Finally, the Fgfs are alternatively spliced (see below), and the isoforms have different functions. For example, Fgf8b activates Gbx2 expression, but neither Fgf8a, Fgf17b nor Fgf18 can do so [Liu et al., 1999, 2003]. Thus, although Fgf8/17/18 evidently acquired a role at the MHB before the two rounds of whole genome duplication, the functions of the paralogs in this tissue have diverged by either subfunctionalization and/or neofunctionalization subsequent to the duplications.


Brain Behav 2008;72:91–105

Alternative Splicing Increases the Size of the Toolkit

Several studies have found that the degree of alternative splicing is inversely proportional to the number of gene duplicates [Kopelman et al., 2005; Su et al., 2006; Talavera et al., 2007]. However, this correlation is much stronger for recent duplicates than for ancient ones such as that which happened during the two rounds of whole genome duplication [Kopelman et al., 2005]. It has been suggested that alternative splicing and gene duplication might have comparable effects on the proteome [Kopelman et al., 2005]. For example, Shimeld et al. [2007] concluded that for Gli genes, which are regionally expressed in the CNS as well as in mesoderm and endoderm, two alternatively spliced isoforms of the single amphioxus Gli gene have a comparable function to the proteins deriving from two duplicate Gli genes in vertebrates. However, the conclusion from this study that amphioxus and vertebrates have used different strategies (alternative splicing and gene duplication) to achieve the same end is premature in the absence of a comprehensive survey of alternative splice forms and their functions for both amphioxus and vertebrate Gli genes. On the other hand, comparisons of two species of the nematode Caenorhabditis suggested that alternative splicing may have less effect on protein function at least within close relatives [Rukov et al., 2007]. In contrast, computational analysis with large data sets has indicated that in general, compared to gene 99

Fig. 4. Conservation of exon/intron orga-

nization and alternative splicing of amphioxus Pax2/5/8 and its three vertebrate paralogs. A AmphiPax2/5/8 has 11 exons (boxes) compared to 10–12 in vertebrate Pax2/5/8 homologs. Exons that correspond to one another are aligned. The length of the box shows the approximately relative length of the exons. The length of the introns, shown by black lines, does not reflect their true length. Paired domain shown by horizontal hatching; octapeptide by open box; homeodomain by crosshatching. Exon 4 split into two before the ancestral gene duplicated. Exons 6 and 10 in Pax2 are Pax2 innovations. Exon 7 has been lost from Pax2 and Pax5. B Alternative-splicing events in AmphiPax2/5/8. Vertical lines indicate sites for termination of translation in some splice forms. Two cterminal isoforms and one n-terminal one are conserved between AmphiPax2/5/8 and one or another of the vertebrate paralogs. [Amphioxus data after Short and Holland, 2008; vertebrate data after Poleev et al., 1995; Heller and Brandli, 1997; Tavassoli et al., 1997; Zwollo et al., 1997; Robichaud et al., 2004].

A

B

duplication, alternative splicing probably has considerably greater effect on protein function [Talavera et al., 2007]. There are relatively few studies of the function of isoforms of neural crest or MHB genes. The best studied of the MHB genes are fgf8 and fgf17. Seven and four splice forms of mouse and human fgf8, respectively, have been identified, as well as two each for chick and frog. There are at least two isoforms of human fgf17 and four of mouse fgf17 but only one of fgf18 is known. Because thorough analyses of alternative splicing of Fgfs have not been done, these numbers are minimum estimates. The roles of some splice forms of vertebrate Fgf8 and Fgf17 at the MHB have been studied. In the mouse, Fgf8a and Fgf8b, which differ by 33 base pairs (11 amino acids), are expressed at different levels, and their overexpression is reported to have different effects on expression of other MHB markers and either similar [Sato et al., 2001] or different [Liu et al., 1999] effects on morphology. Thus, misexpression of Fgf8a has no effect on Otx2, Gbx2, Pax2 or Pax5, but en2 and en1 expression in the diencephalon is upregulated. In contrast, misexpression of Fgf8b expands the domains of 100


Gbx2 and Pax2, induces expression of en1 and en2 expression, and eliminates Otx2 expression from the posterior diencephalon and mesencephalon [Liu et al., 1999; Sato et al., 2001]. Moreover, overexpression of Fgf8b transforms the midbrain into cerebellum whereas Fgf8a promotes midbrain identity. Comparisons of Fgf17b, 18, 8a and 8b show that only Fgf8b can induce ectopic expression of Gbx2 and strongly induce expression of the Fgf antagonist Sprouty1/2, whereas Fgf17b and 18 have the same effect as Fgf8a – upregulating midbrain genes and expanding the midbrain territory [Liu et al., 2003]. Moreover, the ability of FGF17b and FGF18 to induce en2 is greater than that of FGF8a and both FGFR17b or FGF18, but not FGF8a, could induce GBX2 at low frequency [Olsen et al., 2006]. Although it was initially reported that mouse FGF17b and FGF8b, but not FGF17a or FGF8a, have similar affinities for the various FGF receptors [Xu et al., 1999], a more recent study showed that the relative affinity of FGFs for some FGF receptors is FGF8b 1 FGF17b = FGF18 1 FGF8a [Olsen et al., 2006].

Holland/Short

For the Pax2/5/8 family, we have found that there are comparable levels of alternative-splicing for each of the three vertebrate genes as there are in the single homologous gene in amphioxus; in addition, the relative proportions of the splice forms are developmentally regulated [Short and Holland, 2008] (fig. 4). If a similar pattern is found for other genes resulting from the two rounds of whole-genome duplication, the vertebrate proteome would contain many more isoforms of genes for transcription factors and signaling molecules – the proteins that direct development – than does the amphioxus genome. As noted above, one example of the differential roles of alternative splice forms is the fgf8/17/18 family. Similar studies have not been done for Pax2/5/8 isoforms at the MHB, but it is reasonable to suppose that the isoforms are differentially expressed at the MHB. Amphioxus Pax2/5/8 has eleven exons, and comparable numbers are present in each of the vertebrate Pax2/5/8 genes (fig. 4). By doing exon to exon PCR we found twelve different splicing events in amphioxus Pax2/5/8, both in the 3ⴕ transactivation and inhibitory domains and in the 5ⴕ DNA-binding domains. These events include skipping exons 2, 7, 8, 9 and 10 individually or in various combinations plus the use of alternative splice sites for exons 4, 7, 10 and 11 [Short and Holland, 2008]. The number of combinations could well result in as many as 50–100 isoforms with varying ability to bind to DNA and/or to transactivate target sequences. Some of these splice forms are conserved with vertebrates (fig. 4), suggesting that the function of these isoforms is critical for normal development. There are, however, some isoforms that are are unique to either amphioxus or vertebrates. Such unique isoforms might well have important roles in new structures such as the MHB organizer. For amphioxus Pax2/5/8, two isoforms expressed in early development have been shown to have very different transactivation properties [Kreslova et al., 2002]. In addition, we have found that several amphioxus splice forms are temporally regulated during development [Short and Holland, 2008]. Isoforms of the vertebrate homologs of Pax2/5/8 also have differing binding and transactivation properties and are temporally regulated. To date, however, the tissue-specificity of Pax isoforms during development has not been determined. Relatively little is known about the roles of alternative splice forms of neural crest markers. Dlx5 is alternatively spliced. One isoform, which encodes a truncated protein missing the homeodomain, cannot bind to DNA or interact with MSX1, and, therefore, could regulate DLX5 function [Yang et al., 1998]. Moreover, alternative splice

forms of Xmeis1 differentially activate expression of XGli-3 and XZic3 and other neural crest markers [Maeda et al., 2001]. Given the dearth of information about the roles of alternative splice forms in development, the roles of alternative splicing in the evolution of new structures are likely to be vastly underappreciated at present. Techniques such as locked nucleic acid probes that permit in situ hybridization with isoform-specific probes have been developed [Kubota et al., 2006], but this approach is only beginning to be used for such studies.


Brain Behav 2008;72:91–105

Conclusions

Neural crest is present in lampreys and hagfish [McCauley and Bronner-Fraser, 2003; Ota et al., 2007], and it is likely that at least lampreys among the agnathans have an MHB organizer because en is expressed at the MHB [Holland et al., 1993]. Although the initial framework for both neural crest and the MHB was already built in the ancestral chordate, definitive neural crest and organizer properties at the MHB evidently evolved subsequent to at least one of the two rounds of whole genome duplications that occurred in the vertebrate lineage. Although these gene duplications were followed by gene losses, duplicates of transcription factors and signaling molecules were preferentially retained. Thus, at least one key gene for maintenance of MHB organizer properties, Wnt1, apparently acquired a new role at the MHB subsequent to one or both rounds of genome duplication as did several neural crest specifiers. Gene networks are very complex, and it is virtually certain that many more genes which acquired roles in the MHB organizer and migratory neural crest subsequent to one or both genome duplications will come to light. What does seem clear is that evolution of neither structure required the invention of new genes, except for some that are far downstream such as c-Kit, which encodes a receptor tyrosine kinase essential for migration and differentiation of neural crest-derived melanocytes [Holland et al., 2008]. For the most part, neofunctionalization of old genes most likely sufficed. Most importantly, the role of alternative splicing in the evolution of the vertebrate MHB and neural crest has only begun to be explored. Different functions for some isoforms of fgf8 at the MHB are known, but the extent of alternative splicing even of this gene has not been fully explored. Our study of alternative splicing in the Pax family has suggested that, for such ancient duplicates, the degree of alternative splicing of the single amphioxus gene is comparable to that for each of the vertebrate par101

alogs, resulting in many more tools in the vertebrate genetic toolkit. If this holds true for other ancient duplicates of developmental genes, the toolkit relevant for the evolution of new vertebrate structures and for increased complexity of old ones is likely to be much, much larger than is reflected in the comparatively modest increase in gene numbers between amphioxus and vertebrates.

Acknowledgments This work was supported by grants MCB06-20019 and IOB 0416292 from the National Science Foundation to LZH.

References Aybar MJ, Nieto MA, Mayor R (2003) Snail precedes Slug in the genetic cascade required for the specification and migration of the Xenopus neural crest. Development 130: 483– 494. Bardet PL, Schubert M, Horard B, Holland LZ, Laudet V, Holland ND, Vanacker JM (2005) Expression of estrogen-receptor related receptors in amphioxus and zebrafish: implications for the evolution of posterior brain segmentation at the invertebrate-to-vertebrate transition. Evol Dev 7:223–233. Benito-Gutiérrez E (2006) A gene catalogue of the amphioxus nervous system. Int J Biol Sci 2:149–160. Blair JE, Hedges SB (2005) Molecular phylogeny and divergence times of deuterostome animals. Mol Biol Evol 22: 2275–2284. Bouchard M, Grote D, Craven SE, Sun Q, Steinlein P, Busslinger M (2005) Identification of Pax2-regulated genes by expression profiling of the mid-hindbrain organizer region. Development 132:2633–2643. Bourlat SJ, Juliusdottir T, Lowe CJ, Freeman R, Aronowicz J, Kirschner M, Lander ES, Thorndyke M, Nakano H, Kohn AB, Heyland A, Moroz LL, Colpey RR, Telford MJ (2006) Deuterostome phylogeny reveals monophyletic chordates and the new phylum Xenoturbellida. Nature 444:85–88. Broccoli V, Boncinelli E, Wurst W (1999) The caudal limit of Otx2 expression positions the isthmic organizer. Nature 401:164–168. Brooke NM, Garcia-Fernàndez J, Holland PWH (1998) The ParaHox cluster is an evolutionary sister of the Hox gene cluster. Nature 392: 920–922. Cañestro C, Bassham S, Postelthwait JH (2005) Development of the central nervous system in the larvacean Oikopleura dioica and the evolution of the chordate brain. Dev Biol 285: 298–315. Canning CA, Lee L, Irving C, Mason I, Jones CM (2007) Sustained interactive Wnt and FGF signaling is required to maintain isthmic identity. Dev Biol 305: 276–286. Castro LF, Rasmussen SL, Holland PWH, Holland ND, Holland LZ (2006) A Gbx homeobox gene in amphioxus: insights into ancestry of the ANTP class and evolution of the midbrain/hindbrain boundary. Dev Biol 295:40–51.

102

Cole MD, McMahon SB (1999) The Myc oncoprotein: a critical evaluation of transactivation and target gene regulation. Oncogene 18:2916–2924. Cole WC, Youson JH (1982) Morphology of the pineal complex of the anadromous sea lamprey, Petromyzon marinus L. Am J Anat 165: 131–163. Colombera D (1974) Male chromosomes in two populations of Branchiostoma lanceolatum. Experientia 30:353–355. Dayal S, Kiyama T, Villinski JT, Zhang N, Liang S, Klein WH (2004) Creation of cis-regulatory elements during sea urchin evolution by co-option and optimization of a repetitive sequence adjacent to the spec2a gene. Dev Biol 273:436–453. Delsuc F, Brinkmann H, Chourrout D, Philippe H (2006) Tunicates and not cephalochordates are the closest living relatives of vertebrates. Nature 439:965–968. Deschamps J, van den Akker E, Forlani S, De Graaff W, Oosterveen T, Roelen B, Roelfsema J (1999) Initiation, establishment and maintenance of Hox gene expression patterns in the mouse. Int J Dev Biol 43: 635– 650. Dottori M, Gross MK, Labosky P, Goulding M (2001) The winged-helix transcription factor Foxd3 suppresses interneuron differentiation and promotes neural crest cell fate. Development 128:4127–4138. Ekker M, Wegner J, Akimenko M, Westerfield M (1992) Coordinate embryonic expression of three zebrafish engrailed genes. Development 116:1001–1010. Erickson T, Scholpp S, Brand M, Moens CB, Waskiewicz AJ (2007) Pbx proteins cooperate with Engrailed to pattern the midbrainhindbrain and diencephalic-mesencephalic boundaries. Dev Biol 301:504–517. Escriva H, Bertrand S, Germain P, Robinson-Rechavi M, Umbhauer M, Cartry J, Duffraisse M, Holland L, Gronemeyer H, Laudet V (2006) Neofunctionalization in vertebrates: the example of retinoic acid receptors. PLoS Genet 2(7):e102 Escriva H, Manzon L, Youson J, Laudet V (2002) Analysis of lamprey and hagfish Genes reveals a complex history of gene duplications during early vertebrate evolution. Mol Biol Evol 19:1440–1450.


Ferrier DEK, Brooke NM, Panopoulou G, Holland PWH (2001) The Mnx homeobox gene class defined by HB9, MNR2 and amphioxus AmphiMnx. Dev Genes Evol 211:103–107. Force A, Lynch M, Pickett FB, Amores A, Yan YL, Postlethwait J (1999) Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151:1531–1545. Funahashi J-i, Okafuji T, Ohuchi H, Noji S, Tanaka H, Nakamura H (1999) Role of Pax-5 in the regulation of a mid-hindbrain organizer’s activity. Dev Growth Differ 41: 59– 72. Furlong RF, Holland PWH (2002) Were vertebrates octoploid? Phil Trans R Soc B: Biol Sci 357:531–544. Garcia-Castro MI, Marcelle C, Bronner-Fraser M (2002) Ectodermal Wnt function as a neural crest Inducer. Science 297:848–851. Gee H (1996) Before the Backbone: Views on the Origin of the Vertebrates. London: Chapman and Hall. Glardon S, Holland LZ, Gehring WJ, Holland ND (1998) Isolation and developmental expression of the amphioxus Pax-6 gene (AmphiPax-6): insights into eye and photoreceptor evolution. Development 125: 2701–2710. Gostling NJ, Shimeld SM (2003) Protochordate Zic genes define primitive somite compartments and highlight molecular changes underlying neural crest evolution. Evol Dev 5: 136–144. Guo C, Qiu H-Y, Huang Y, Chen H, Yang R-Q, Chen S-D, Johnson RL, Chen Z-F, Ding Y-Q (2007) Lmx1b is essential for Fgf8 and Wnt1 expression in the isthmic organizer during tectum and cerebellum development in mice. Development 134:317–325. Gupta BP, Johnsen R, Chen N (2007) Genomics and biology of the nematode Caenorhabditis briggsae. WormBook May 3:1–16. Guthrie S, Lumsden A (1992) Motor neuron pathfinding following rhombomere reversals in the chick embryo hindbrain. Development 114:663–673. Hahn MW, Han MV, Han S-G (2007) Gene family evolution across 12 Drosophila genomes. PLoS Genet. 3:e197. Heller N, Brändli AW (1997) Xenopus Pax-2 displays multiple splice forms during embryogenesis and pronephric kidney development. Mech Dev 69:83–104.

Holland/Short

Heller N, Brändli AW (1999) Xenopus Pax-2/5/8 orthologues: Novel insights into Pax gene evolution and identification of Pax-8 as the earliest marker for otic and pronephric cell lineages. Dev Genet 24:208–219. Hidalgo-Sanchez M, Millet S, Bloch-Gallego E, Alvarado-Mallart R-M (2005) Specification of the meso-isthmo-cerebellar region: The Otx2/Gbx2 boundary. Brain Res Rev 49: 134–149. Hino K, Satou Y, Yagi K, Satoh N (2003) A genomewide survey of developmentally relevant genes in Ciona intestinalis. Dev Genes Evol 213:264–272. Holland LZ (2002) Heads or tails? Amphioxus and the evolution of anterior-posterior patterning in deuterostomes. Dev Biol 241:209– 228. Holland LZ (2006) A SINE in the genome of the cephalochordate amphioxus is an Alu element. Int J Biol Sci 2:61–65. Holland LZ, Albalat R, Azumi K, Benito-Gutiérrez E, Blow MJ, Bronner-Fraser M, Brunet F, Butts T, Candiani S, Dishaw LJ, GarciaFernàndez J, Ferrier DEK, Gibson-Brown JJ, Gissi C, Godzik A, Hallböök F, Hirose D, Hosomichi K, Ikuta T, Inoko H, Kasahara M, Kasamatsu J, Kawashima T, Kimura A, Kobayashi M, Kozmik Z, Kubokawa K, Laudet V, Litman GW, McHardy AC, Meulemans D, Nonaka M, Olinski RP, Pancer Z, Pennacchio LA, Pestarino M, Rast JP, Rigoutsos I, Roch G, Robinson-Rechavi M, Saiga H, Sasakura Y, Satake M, Satou Y, Schubert M, Sherwood N, Shiina T, Takatori N, Tello J, Vopalensky P, Wada S, Xu A, Ye Y, Yoshida K, Yoshizaki F, Yu, JK, Zhang Q, Zmasek CM, Osoegawa K, de Jong PJ, Putnam NH, Rokhsar DS, Satoh N, Holland PWH (2008) The amphioxus genome illuminates vertebrate origins and cephalochordate biology. Genome Res 18:1100–1111. Holland LZ, Holland ND (2000) Developmental expression of AmphiWnt1, an amphioxus gene in the Wnt1/wingless subfamily. Dev Genes Evol 210:522–524. Holland LZ, Kene M, Williams NA, Holland ND (1997) Sequence and embryonic expression of the amphioxus engrailed gene (AmphiEn): the metameric pattern of transcription resembles that of its segment-polarity homolog in Drosophila. Development 124: 1723– 1732. Holland LZ, Schubert M, Holland ND, Neuman T (2000) Evolutionary conservation of the presumptive neural plate markers AmphiSox1/2/3 and AmphiNeurogenin in the invertebrate chordate amphioxus. Dev Biol 226: 18–33. Holland LZ, Schubert M, Kozmik Z, Holland ND (1999) AmphiPax3/7, an amphioxus paired box gene: insights into chordate myogenesis, neurogenesis, and the possible evolutionary precursor of definitive vertebrate neural crest. Evol Dev 1:153–165.


Holland ND, Holland LZ, Honma Y, Fujii T (1993) Engrailed Expression during development of a lamprey, Lampetra japonica: A possible clue to homologies between agnathan and gnathostome muscles of the mandibular arch. Dev Growth Differ 35: 153– 160. Holland ND, Panganiban G, Henyey EL, Holland LZ (1996) Sequence and developmental expression of AmphiDll, an amphioxus Distal-less gene transcribed in the ectoderm, epidermis and nervous system: insights into evolution of craniate forebrain and neural crest. Development 122:2911–2920. Holland PWH, Holland LZ, Williams NA, Holland ND (1992) An amphioxus homeobox gene: sequence conservation, spatial expression during development and insights into vertebrate evolution. Development 116:653– 661. Howell WM, Boschung HT (1971) Chromosomes of the lancelet, Branchiostoma floridae (Order Amphioxi). Experientia 27:1495– 1496. Hughes AL, Friedman R (2003) 2R or not 2R: Testing hypotheses of genome duplication in early vertebrates. J Struct Funct Genomics 3: 85–93. Ikuta T, Saiga H (2005) Organization of Hox genes in ascidians: Present, past, and future. Dev Dyn 233:382–389. Ikuta T, Saiga H (2007) Dynamic change in the expression of developmental genes in the ascidian central nervous system: Revisit to the tripartite model and the origin of the midbrain-hindbrain boundary region. Dev Biol 312:631–643. Imai JH, Meinertzhagen IA (2007) Neurons of the ascidian larval nervous system in Ciona intestinalis: I. Central nervous system. J Comp Neurol 501:316–334. Imai KS, Hino K, Yagi K, Satoh N, Satou Y (2004) Gene expression profiles of transcription factors and signaling molecules in the ascidian embryo: towards a comprehensive understanding of gene networks. Development 131:4047–4058. Imai KS, Satoh N, Satou Y (2002) Region specific gene expressions in the central nervous system of the ascidian embryo. Mech Dev 119:S275–S277. Islam ME, Kikuta H, Inoue F, Kanai M, Kawakami A, Parvin MS, Takeda H, Yamasu K (2006) Three enhancer regions regulate gbx2 gene expression in the isthmic region during zebrafish development. Mech Dev 123: 907– 924. Jackman WR, Kimmel CB (2002) Coincident iterated gene expression in the amphioxus neural tube. Evol Dev 4:366–374. Jackman WR, Langeland JA, Kimmel CB (2000) Islet reveals segmentation in the amphioxus hindbrain homolog. Dev Biol 230: 16–26. Jiang D, Smith WC (2002) An ascidian engrailed gene. Dev Genes Evol 212:399–402.

Joyner AL, Liu A, Millet S (2000) Otx2, Gbx2 and Fgf8 interact to position and maintain a mid-hindbrain organizer. Curr Opin Cell Biol 12: 736–741. Jurka J (2008) Conserved eukaryotic transposable elements and the evolution of gene regulation. Cell Mol Life Sci: 65:201–204. Kopelman NM, Lancet D, Yanai I (2005) Alternative splicing and gene duplication are inversely correlated evolutionary mechanisms. Nat Genet 37:588–589. Kozmik Z, Holland ND, Kalousova A, Paces J, Schubert M, Holland LZ (1999) Characterization of an amphioxus paired box gene, AmphiPax2/5/8:developmental expression patterns in optic support cells, nephidium, thyroid-like structures and pharyngeal gill slits, but not in the midbrain-hindbrain boundary region. Development 126: 1295– 1304. Kozmik Z, Holland ND, Kreslova J, Oliveri D, Schubert M, Jonasova K, Holland LZ, Pestarino M, Benes V, Candiani S (2007) Pax-SixEya-Dach network during amphioxus development: Conservation in vitro but context specificity in vivo. Dev Biol 306: 143–159. Kreslova J, Holland LZ, Schubert M, Burgtorf C, Benes V, Kozmik Z (2002) Functional equivalency of amphioxus and vertebrate Pax258 transcription factors suggests that the activation of mid-hindbrain specific genes in vertebrates occurs via the recruitment of Pax regulatory elements. Gene 282:143–150. Kubota K, Ohashi A, Imachi H, Harada H (2006) Improved in situ hybridization efficiency with locked-nucleic-acid-incorporated DNA Probes. Appl Environ Microbiol 72: 5311– 5317. Lacalli TC (1996) Frontal eye circuitry, rostral sensory pathways, and brain organization in amphioxus larvae: evidence from 3D reconstructions. Phil Trans R Soc B 351:243–263. Lacalli TC (2003) Ventral neurons in the anterior nerve cord of amphioxus larvae. II. Further data on the pacemaker circuit. J Morphol 257:212–218. Lacalli TC, Holland ND, West JE (1994) Landmarks in the anterior central nervous system of amphioxus larvae. Phil Trans R Soc London B 344:165–185. Lacalli TC, Kelly SJ (1999) Somatic motoneurones in amphioxus larvae: cell types, cell position and innervation patterns. Acta Zool Stockh 80:113–124. Langeland JA, Tomsa JM, Jackman WR, Kimmel CB (1998) An amphioxus snail gene: expression in paraxial mesoderm and neural plate suggests a conserved role in patterning the embryo. Dev Genes Evol 208:569–577. Lekven AC, Buckles GR, Kostakis N, Moon RT (2003) Wnt1 and wnt10b function redundantly at the zebrafish midbrain-hindbrain boundary. Dev Biol 254:172–187. Li JYH, Joyner AL (2001) Otx2 and Gbx2 are required for refinement and not induction of mid-hindbrain gene expression. Development 128:4979–4991.

Brain Behav 2008;72:91–105

103

Liu A, Joyner A (2001) EN and GBX2 play essential roles downstream of FGF8 in patterning the mouse mid/hindbrain region. Development 128:181–191. Liu A, Li JYH, Bromleigh C, Lao Z, Niswander LA, Joyner AL (2003) FGF17b and FGF18 have different midbrain regulatory properties from FGF8b or activated FGF receptors. Development 130:6175–6185. Liu A, Losos K, Joyner A (1999) FGF8 can activate Gbx2 and transform regions of the rostral mouse brain into a hindbrain fate. Development 126:4827–4838. MacCarthy T, Bergman A (2007) The limits of subfunctionalization. BMC Evol Biol 7: 213. Maeda R, Mood K, Jones TL, Aruga J, Buchberg AM, Daar IO (2001) Xmeis1, a protooncogene involved in specifying neural crest cell fate in Xenopus embryos. Oncogene 20: 1329–1342. McCauley DW, Bronner-Fraser M (2003) Neural crest contributions to the lamprey head. Development 130:2317–2327. Meiniel A (1980) Ultrastructure of serotonincontaining cells in the pineal organ of Lampetra planeri (Petromyzontidae). Cell Tissue Res 207:407–427. Meulemans D, Bronner-Fraser M (2002) Amphioxus and lamprey AP-2 genes: implications for neural crest evolution and migration patterns. Development 129:4953–4962. Meulemans D, Bronner-Fraser M (2004) Generegulatory interactions in neural crest evolution and development. Dev Cell 7:291–299. Meulemans D, Bronner-Fraser M (2007) Insights from amphioxus into the evolution of vertebrate cartilage. PLoS ONE 2:e787. Meulemans D, McCauley D, Bronner-Fraser M (2003) Id expression in amphioxus and lamprey highlights the role of gene cooption during neural crest evolution. Dev Biol 264: 430–442. Murcia CL, Gulden FO, Cherosky NA, Herrup K (2007) A genetic study of the suppressors of the Engrailed-1 cerebellar phenotype. Brain Res 1140:170–178. Nakamura H, Katahira T, Matsunaga E, Sato T (2005) Isthmus organizer for midbrain and hindbrain development. Brain Res Rev 49: 120–126. Nicol D, Meinertzhagen IA (1991) Cell counts and maps in the larval central nervous system of the ascidian Ciona intestinalis (L.). J Comp Neurol 309:415–429. Northcutt RG (2003) Origin of the isthmus? A comparison of the brains of lancelets and vertebrates. J Comp Neurol 466:316–318. Ohbayashi N, Shibayama M, Kurotaki Y, Imanishi M, Fujimori T, Itoh N, Takada S (2002) FGF18 is required for normal cell proliferation and differentiation during osteogenesis and chondrogenesis. Genes Dev 16: 870– 879. Ohno S (1970) Evolution by Gene Duplication. Springer Verlag, New York, NY.

104

Olsen SK, Li JYH, Bromleigh C, Eliseenkova AV, Ibrahimi OA, Lao Z, Zhang F, Linhardt RJ, Joyner AL, Mohammadi M (2006) Structural basis by which alternative splicing modulates the organizer activity of FGF8 in the brain. Genes Dev 20:185–198. Olsson R, Yulis R, Rodriguez EM (1994) The infundibular organ of the lancelet (Branchiostoma lanceolatum, Acrania): an immunocytochemnical study. Cell Tissue Res 277: 107–114. Ota KG, Kuraku S, Kuratani S (2007) Hagfish embryology with reference to the evolution of the neural crest. Nature 446:672–675. Poleev A, Wendler F, Fickenscher H, Zannini MS, Yaginuma K, Abbott C, Plachov D (1995) Distinct functional properties of three human paired-box-protein, PAX8, isoforms generated by alternative splicing in thyroid, kidney and Wilms’ Tumors. Eur J Biochem 228: 899–911. Putnam NH, Ferrier DEK, Furlong RF, Hellsten U, Kawashima T, Robinson-Rechavi M, Shoguchi E, Terry A, Yu JK, Benito-Gutiérrez E, Dubchak I, Garcia-Fernàndez J, Grigoriev IV, Horton AC, de Jong PJ, Jurka J, Kapitonov V, Kohara Y, Kuroki Y, Lindquist E, Lucas S, Osoegawa K, Pennacchio LA, Salamov AA, Satou Y, Sauka-Spengler T, Schmutz J, Shin-I T, Toyoda A, Gibson-Brown JJ, Bronner-Fraser M, Fujiyama A, Holland LZ, Holland PWH, Satoh N, Rokhsar DS (2008) The amphioxus genome and the evolution of the chordate karyotype. Nature: 453:1064–1071. Raible F, Brand M (2004) Divide et impera – the midbrain-hindbrain boundary and its organizer. Trends Neurosci 27:727–734. Robichaud GA, Nardini M, Laflamme M, Cuperlovic-Culf M, Ouellette RJ (2004) Human Pax-5 c-terminal isoforms possess distinct transactivation properties and are differentially modulated in normal and malignant B cells. J Biol Chem 279: 49956–49963. Roest Crollius H, Jaillon O, Bernot A, Dasilva C, Bouneau L, Fischer C, Fizames C, Wincker P, Brottier P, Quetier F, Saurin W, Weissenbach J (2000) Estimate of human gene number provided by genome-wide analysis using Tetraodon nigroviridis DNA sequence. Nat Genet 25:235–238. Ruiz S, Anadon R (1991) The fine structure of lamellate cells in the brain of amphioxus (Branchiostoma lanceolatum, Cephalochordata). Cell Tissue Res 263:597–600. Rukov JL, Irimia M, Mørk S, Lund VK, Vinther J, Arctander P (2007). High qualitative and quantitative conservation of alternative splicing in Caenorhabditis elegans and Caenorhabditis briggsae. Mol Biol Evol 24: 909– 917. Saotome K, Ojima Y (2001) Chromosomes of the lancelet Branchiostoma belcheri Gray. Zool Sci 18:683–686. Sato T, Araki I, Nakamura H (2001) Inductive signal and tissue responsiveness defining the tectum and the cerebellum. Development 128:2461–2469.


Sato T, Sasai N, Sasai Y (2005) Neural crest determination by co-activation of Pax3 and Zic1 genes in Xenopus ectoderm. Development 132:2355–2363. Schubert M, Escriva H, Xavier-Neto J, Laudet V (2006a) Amphioxus and tunicates as model systems. TREE 21:269–277. Schubert M, Holland LZ, Stokes MD, Holland ND (2001) Three amphioxus Wnt genes (AmphiWnt3, AmphiWnt5, and AmphiWnt6) associated with the tail bud: the evolution of somitogenesis in chordates. Dev Biol 240: 262–273. Schubert M, Holland ND, Laudet V, Holland LZ (2006b) A retinoic acid-Hox hierarchy controls both anterior/posterior patterning and neuronal specification in the developing central nervous system of the cephalochordate amphioxus. Dev Biol 296: 190–202. Seo H-C, Edvardsen RB, Maeland AD, Bjordal M, Jensen MF, Hansen A, Flaat M, Weissenbach J, Lehrach H, Wincker P, Reinhardt R, Chourrout D (2004) Hox cluster disintegration with persistent anteroposterior order of expression in Oikopleura dioica. Nature 431: 67–71. Sharman AC, Shimeld SM, Holland PWH (1999) An amphioxus Msx gene expressed predominantly in the dorsal neural tube. Dev Genes Evol 209:260–263. Shimeld SM, Holland ND (2005) Amphioxus molecular biology: insights into vertebrate evolution and developmental mechanisms. Can J Zool 83: 90–100. Shimeld SM, van den Heuvel M, Dawber R, Briscoe J (2007) An amphioxus Gli gene reveals conservation of midline patterning and the evolution of hedgehog signaling diversity in chordates. PLoS ONE 2:e864. Short S, Holland LZ (2008) Alternative splicing of amphioxus Pax transcripts reveals conserved events impacting functional domains and suggests an overall expansion of isoforms following vertebrate gene duplications. J Mol Evol 67: in press. Søviknes AM, Chourrout D, Glover JC (2005) Development of putative GABAergic neurons in the appendicularian urochordate Oikopleura dioica. J Comp Neurol 490: 12– 28. Søviknes AM, Chourrout D, Glover JC (2007) Development of the caudal nerve cord, motoneurons, and muscle innervation in the appendicularian urochordate Oikopleura dioica. J Comp Neurol 503:224–243. Stewart RA, Arduini BL, Berghmans S, George RE, Kanki JP, Henion PD, Look AT (2006) Zebrafish foxd3 is selectively required for neural crest specification, migration and survival. Dev Biol 292:174–188. Su Z, Wang J, Yu J, Huang X, Gu X (2006) Evolution of alternative splicing after gene duplication. Genome Res 16:182–189. Suda Y, Nakabayashi J, Matsuo I, Aizawa S (1999) Functional equivalency between Otx2 and Otx1 in development of the rostral head. Development 126:743–757.

Holland/Short

Swalla BJ, Cameron CB, Corley LS, Garey JR (2000) Urochordates are monophyletic within the deuterostomes. Syst Biol 49: 52–64. Takio Y, Pasqualetti M, Kuraku S, Hirano S, Rijli FM, Kuratani S (2004) Evolutionary biology: lamprey Hox genes and the evolution of jaws. Nature 429:263. Talavera D, Vogel C, Orozco M, Teichmann SA, de la Cruz X (2007) The (In)dependence of alternative splicing and gene duplication. PLoS Computational Biol 3:e33. Tavassoli K, Ruger W, Horst J (1997) Alternative splicing in PAX2 generates a new reading frame and an extended conserved coding region at the carboxy terminus. Hum Genet 101:371–375. Toresson H, Martinez-Barbera JP, Beardsley A, Caubit X, Krauss S (1998) Conservation of BF-1 expression in amphioxus and zebrafish suggests evolutionary ancestry of anterior cell types that contribute to the vertebrate telencephalon. Dev Genes Evol 208: 431– 439.


Tour E, Pillemer G, Gruenbaum Y, Fainsod A (2002) Otx2 can activate the isthmic organizer genetic network in the Xenopus embryo. Mech Dev 110:3–13. Wada H (1998) Evolutionary history of freeswimming and sessile lifestyles in urochordates as deduced from 18S rDNA molecular phylogeny. Mol Biol Evol 15: 1189–1194. Wicht H, Lacalli TC (2005) The nervous system of amphioxus: structure, development, and evolutionary significance. Can J Zool 83: 122–150. Williams NH, Holland PWH (1996) Old head on young shoulders. Nature 383:490. Williams RW, Herrup K (1988) The control of neuron number. Ann Rev Neurosci 11: 423– 453. Xu J, Lawshe A, MacArthur CA, Ornitz DM (1999) Genomic structure, mapping, activity and expression of fibroblast growth factor 17. Mech Dev 83:165–178. Xu J, Liu Z, Ornitz D (2000) Temporal and spatial gradients of Fgf8 and Fgf17 regulate proliferation and differentiation of midline cerebellar structures. Development 127: 1833– 1843.

Yang L, Zhang H, Hu G, Wang H, Abate-Shen C, Shen MM (1998) An early phase of embryonic Dlx5 expression defines the rostral boundary of the neural plate. J Neurosci 18: 8322–8330. Yasui K, Zhang SC, Uemura M, Aizawa S, Ueki A (1998) Expression of a twist-related gene, Bbtwist, during the development of a lancelet species and its relation to cephalochordate anterior structures. Dev Biol 195:49–59. Ye W, Bouchard M, Stone D, Liu X, Vella F, Lee J, Nakamura H, Ang S-L, Busslinger M, Rosenthal A (2001) Distinct regulators control the expression of the mid-hindbrain organizer signal FGF8. Nature Neurosci 4: 1175–1181. Yu J-K, Satou Y, Holland ND, Shin-I T, Kohara Y, Satoh N, Bronner-Fraser M, Holland LZ (2007) Axial patterning in cephalochordates and the evolution of the organizer. Nature 445:613–617. Zwollo P, Arrieta H, Ede K, Molinder K, Desiderio S, Pollock R (1997) The Pax-5 Gene Is alternatively spliced during B-cell development. J Biol Chem 272: 10160–10168.

Brain Behav 2008;72:91–105

105



Evolutionary Convergence of Higher Brain Centers Spanning the Protostome-Deuterostome Boundary Sarah M. Farris Department of Biology, West Virginia University, Morgantown, W.V., USA

Key Words Behavioral ecology ⴢ Cerebral cortex ⴢ Feeding habits ⴢ Insects ⴢ Mushroom bodies ⴢ Social behavior ⴢ Vision

Abstract Currently available evidence supports a single origin for the centralized nervous system of bilaterally symmetrical animals. Beneath the staggering diversity of protostome and deuterostome nervous systems lies a fundamental groundplan consisting of a tripartite brain and a nerve cord divided into distinct antero-posterior and medio-lateral zones. As divergent lineages have taken independent paths towards increased encephalization, complex brain centers have arisen that serve multiple levels of sensory processing and advanced behavioral coordination and execution. Many questions arise as one surveys the distribution of these brain centers across the bilaterian phylogenetic trees. What environments did these lineages encounter that promoted the acquisition of energetically expensive brain centers composed of thousands, millions or even trillions of neurons? What novel behavioral capabilities did these brain centers in turn give rise to? Comparative studies within vertebrate clades have revealed instances of parallelism and convergence that have been instructive in associating evolutionary changes in brain structure and function with specific behavioral ecologies. The present account reviews these findings and extends them to invertebrate animals that have independently evolved higher brain centers. By expanding the

© 2008 S. Karger AG, Basel 0006–8977/08/0722–0106$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com


scope of comparative studies across phyla, it will be possible to uncover structural and functional constraints imposed by deep homology, and to better understand the environmental pressures that have given rise to brain and behavioral complexity. Copyright © 2008 S. Karger AG, Basel

Introduction

The past two decades have seen the application of a vast body of model systems research in development to questions of evolution, the so-called rise of ‘evo-devo’ [Love and Raff, 2003]. A spectacular example is found in the confirmation of Anton Dohrn’s 1875 hypothesis that the vertebrate nervous system represents a dorsoventrally inverted version of the invertebrate nervous system, a confirmation won through detailed analysis of cellular determinants and signaling pathways in species as distantly related as flies, mice, and marine annelids [Denes et al., 2007]. Deep homology of dorsoventral domains in the bilaterian nervous system is indicated by conserved roles of TGF-␤ family signaling molecules (dpp, BMPs) and their downstream effectors [Mizutani et al., 2006; Denes et al., 2007; reviewed by Hirth and Reichert, 2006; Arendt et al., 2007]. Similarly, anteroposterior domains are specified by such well-conserved patterns of gene expression that the tripartite brain is proposed to have been inherited from the bilaterian common ancestor [Hirth et Sarah M. Farris Department of Biology, West Virginia University 3139 Life Sciences Building, 53 Campus Drive Morgantown, WV 26506 (USA) Tel. +1 304 293 5201, Fax +1 304 293 6363, E-Mail [email protected]

A

Drosophila

otd/Otx2

Mouse

protocerebrum

forebrain

deutocerebrum

midbrain

tritocerebrum

hindbrain

Pax 2/5/8

subesophageal ganglion spinal cord

Hox ventral nerve cord

Pax2/5/8

Fig. 1. Conserved gene expression do-

mains during development of protostome and deuterostome nervous systems support a single origin for the bilaterian nervous system. A Conserved gene expression domains during anterior-posterior patterning of the CNS in a protostome (Drosophila) and a deuterostome (mouse). Used with permission from Lichtneckert and Reichert [2005]. B Conserved gene expression domains and cell types generated during dorsal-ventral patterning in two protostomes (the lophotrochozoan Platynereis and the ecdysozoan Drosophila) and in a deuterostome (vertebrate). Red = nk2.2, gray = pax6, cyan = msx, midline cells = black, yellow = serotonergic neurons, blue = hb9+ neurons, purple = ath+ lateral sensory neurons. Used with permission from Denes et al. [2007].

otd/Otx2

B

Hox1 orthologues unpg/Gbx2

Platynereis

Vertebrate

Drosophila

dorsal

dorsal

dorsal Dpp (Bmp2/4)

Bmp2/4

Bmp2/4 Bmp2/4

Bmp2/4

Bmp2/4

Bmp2/4 ventral

al., 2003; reviewed by Lichtneckert and Reichert, 2005; Hirth and Reichert, 2006; Arendt et al., 2007]. For example, in both a model protostome (the fruit fly, Drosophila melanogaster) and a model deuterostome (the mouse, Mus musculus), expression of the transcription factor otd/Otx defines the anterior brain (protocerebrum and deutocerebrum of the fly, forebrain and midbrain of the mouse), and unpg/Gbx specifies the posterior nervous system (tritocerebrum and ventral nerve cord of the fly, hind brain and spinal cord regions of the mouse; fig. 1A). Pax 2/5/8 expression defines a region between the insect

deutocerebrum and tritocerebrum that is equivalent to an important organizer region in the vertebrate brain, the midbrain-hindbrain boundary; and Hox gene expression further subdivides the ventral nerve cord and spinal cord posteriorly [Rhinn and Brand, 2001; Hirth et al., 2003; Lichtneckert and Reichert, 2005]. Developmental studies are not alone in suggesting a single origin for the bilaterian nervous system. Key molecular components of bilaterian nervous systems are found in modern cnidarians (jellyfish and anemones) and even in the Porifera (sponges), suggesting that these

Convergence of Higher Brain Centers


107

elements arose very early in metazoan evolution. Cells with morphological and physiological properties of neurons, including the ability to produce action potentials that are mediated by sodium and potassium currents, have long been known in cnidarian animals [Anderson and Mackie, 1977; Meech and Mackie, 1995; reviewed in Lichtneckert and Reichert, 2006]. Cnidarian neurons may be electrically coupled or form chemical synapses, and classical fast and slow neurotransmitters have been localized to these cells using immunohistochemistry and other techniques [reviewed in Kass-Simon and Pierobon, 2007]. At the ultrastructural level, cnidarian neurons contain synaptic vesicles and otherwise resemble those of bilaterians [Jha and Mackie, 1967]. Many of the molecular building blocks of the nervous system appear to have originated even earlier, however, in the sponges. A recent genomic comparison of the demosponge Amphimedon queenslandica with other species demonstrated that the sponge possesses 80% of thirty-six genes for postsynaptic density proteins [Sakarya et al., 2007]. A metabotropic GABA/glutamate-type receptor gene has been isolated in the sponge Geodia cydonium, suggesting that the molecular basis for cellular communication via these chemicals might have been established long before the first neurons arose [Perovic et al., 1999]. The preponderance of evidence for deep homology of the bilaterian nervous system has led some authors to suggest that certain populations of neuromodulatory neurons are homologous based on origination from conserved domains of gene expression in vertebrate, ecdysozoan and lophotrochozoan model systems (fig. 1B) [Denes et al., 2007; Tessmar-Raible et al., 2007]. This hypothesis that homologous neuronal populations are retained in the brains of animals 600 million years after their divergence from a common ancestor is plausible because the body of evidence for shared developmental origins is so great, and because the cell populations themselves participate in relatively simple aspects of locomotor, sensory and neurosecretory function that are shared across phyla [Denes et al., 2007; Tessmar-Raible et al., 2007].

Higher Brain Centers

The urbilaterian nervous system was centralized and exhibited complexity in its division into multiple developmental and functional domains by patterns of gene expression. It was unlikely, however, to have contained anything equivalent to the complex higher brain centers of modern lineages. In the present account, higher brain 108


centers are those that gather information from lower, unimodal processing circuits for computations relating to sensory integration, motor planning and behavioral flexibility. For example, vertebrate evolution was accompanied by an explosion in relative brain size, with the concomitant acquisition of such higher brain centers as the cerebellum, the hippocampus, and the cerebral cortex, each comprising thousands, millions or even trillions of neurons in a single individual. Concentrated masses of neural tissue are energetically expensive, and so it is only in particular ecological niches that a species might receive a favorable return on the behavioral benefits that higher brain centers provide. As any changes in brain structure and function must ultimately arise from modification of developmental patterns of gene expression and cell differentiation, the acquisition of higher brain centers in bilaterians was accompanied by developmental mechanisms for increasing cell number and generating and maintaining orderly connections among them. The functional, structural and developmental evolution of higher brain centers, particularly the mammalian cerebral cortex, remains an area of intense interest. This is largely due to the fact that our own species, Homo sapiens, is endowed with one of the largest cerebral cortices relative to body size accompanied by the largest, most complex and most flexible behavioral repertoires of any animal. But as eloquently stated by Gilles Laurent in the opening chapter of ‘23 Problems in Systems Neuroscience,’ a book in which the authors of eleven of the subsequent chapters focus on functions of the mammalian cortex: ‘When it comes to computation, integrative principles, and ‘cognitive’ issues such as perception, however, most neuroscientists act as if King Cortex appeared one bright morning out of nowhere, leaving in the mud a zoo of robotic critters…’ [Laurent, 2006]. Among the deuterostomes are lineages with staggeringly complex brains such as those of the vertebrates, but also lineages with pared-down, non-centralized nervous systems such as those observed in the echinoderms and tunicates [Brusca and Brusca, 2003]. And, although we typically regard protostomes as simple bugs and worms that have changed little from their ancestral stock, modern lineages have in fact been subject to vast expanses of evolutionary time during which their brains have adapted according to varied selective pressures, some of which have favored complex behaviors and the brain centers that support them (fig. 2). The present account will focus on what is currently understood about the structure, function and development of higher brain centers in protostome animals, Farris

building a case for the convergent evolution of these centers with those of vertebrates across a gulf of six hundred million years of independent evolution. Higher Brain Centers in Protostomes: The Mushroom Bodies The distribution of centralized nervous systems, and those lineages in which higher centers are also present is shown in figure 2A. Higher brain centers of arthropods, onychophorans and annelids are located in the protocerebral segment of the brain, which according to comparisons of developmental gene expression patterns is homologous to the vertebrate forebrain. The forebrain is the location of prominent higher centers such as the cerebral cortex and hippocampus. The mushroom bodies are the best-studied protostome protocerebral centers. These paired neuropils are observed in the brains of annelids, onychophorans and arthropods, and are best known in insects for their roles in olfactory processing and spatial, associative and context-dependent learning and memory [Erber et al., 1980; Schildberger, 1984; Li and Strausfeld, 1997, 1999; Mizunami et al., 1998; Liu et al., 1999; Roman and Davis, 2001; Perez-Orive et al., 2002; Heisenberg, 2003; Cassenaer and Laurent, 2007]. Insect mushroom bodies are composed of many hundreds to thousands of granule cell-like intrinsic neurons (Kenyon cells; fig. 3). In most species, Kenyon cells supply calyx neuropils with their dendrites and the pedunculus and lobes with parallel projecting axon-like processes, the latter of which typically bifurcate to form two or more lobes and receive further inputs from other areas of the protocerebrum while providing outputs to mushroom body efferents and recurrent neurons [Li and Strausfeld, 1999; Strausfeld, 2002]. This basic organization is shared by mushroom body-like structures in the brains of polychaete annelids, the Onychophora (lobopods), and all arthropods with the exception of the basal hexapods and the Crustacea [Strausfeld et al., 1995, 2006; Strausfeld, 1998; Heuer and Loesel, 2008]. In most representative species of these taxa, the mushroom bodies serve chemosensory processing functions as their calyces are the targets of olfactory input from primary olfactory neuropils [Strausfeld et al., 1998; Heuer and Loesel, 2008], and recent studies suggest that the insect mushroom bodies also receive gustatory input from sensory neurons on the mouthparts via neuropils in the deutocerebrum, tritocerebrum and subesophageal ganglion [Schröter and Menzel, 2003; Farris, 2008b]. Protostome mushroom bodies are not solely olfactory processing centers, however, as olfactory input is lost in aquatic insects and other Convergence of Higher Brain Centers

A

B

Fig. 2. Nervous system evolution in bilaterally symmetrical animals. A Distribution of nervous system traits across the Bilateria.

Open circle = centralized nervous system; shaded circle- cephalized nervous system; closed circle = higher brain centers present. The phylogenetic tree is based on Glenner et al. [2004], nervous system traits from Brusca and Brusca [2003]. B Distribution of mushroom bodies and mushroom-body like neuropils in the Arthropoda. Closed circle = mushroom bodies present; ? = data not available. Phylogenetic tree is based on studies by Regier et al. [2005], Glenner et al. [2006] and Carapelli et al. [2007].


109

Fig. 3. Generalized schematic diagram of the mushroom bodies

of one hemisphere of the insect brain. Olfactory input (blue line) from the antennal lobes (ant lo) is the primary source of input to the mushroom body calyces (ca) in most insects. Olfactory projection neurons synapse on the dendrites of mushroom body intrinsic neurons, the Kenyon cells (KC; shades of green). Kenyon cells provide a parallel system of axon-like processes that bifurcate to form the lobes where they synapse onto the Purkinje celllike dendrites of mushroom body efferent neurons (red). Efferents transmit this information to various areas of the protocerebrum (pr) for further processing. Lo = Lobula of the optic lobes; me = medulla of the optic lobes.

species that are secondarily anosmic, resulting in a loss of the calyx while the parallel fiber system of local circuits in the lobes is retained [Strausfeld et al., 1998; Farris, Brown, Sinakevitch and Strausfeld, unpublished data]. In the higher Hymenoptera (ants, bees and wasps), the mushroom body calyces receive both olfactory input from the antennal lobes as well as significant visual input from the optic lobes [Gronenberg and Hölldobler, 1999; Gronenberg, 2001; Ehmer and Gronenberg, 2002]. In spiders, the mushroom bodies are also well developed and receive the preponderance of their inputs from primary visual neuropils [Strausfeld and Barth, 1993]. The mushroom bodies of protostomes all share key architectural features, particularly the parallel fiber system formed by a multitude of tiny granule cell-like intrinsic neurons. Given their patchy distribution across the 110


protostome phylogenetic tree, however, it seems likely that despite their similarities, these structures are not homologous. The distribution of mushroom bodies in the arthropods alone makes a good case for at least two independent origins of these higher brain centers. Mushroom bodies are present in chelicerates, myriapods (likely a paraphyletic grouping containing the centipedes and millipedes) and all except the most basal order of insects (fig. 2B). The most recent molecular phylogenies support a single origin for the hexapods, including the insects, from within the Crustacea [Regier et al., 2005; Glenner et al., 2006; Carapelli et al., 2007]. All crustaceans lack mushroom bodies [Strausfeld, 1998], as do basal hexapods such as the Collembola, Protura and Archaeognatha [although there is some evidence for mushroom bodies in the Diplura that remains to be substantiated using modern histological and imaging methods; Hanström, 1940]. The distribution of mushroom bodies across the arthropod phylogenetic tree can therefore best be interpreted to suggest that they arose at least twice independently; once very early in the evolution of this taxon, and once much later, within the insect lineage. Insect mushroom bodies share many features of structure, function, connectivity and development, strongly supporting their homology within this group [Farris, 2005]. The mushroom bodies are the best-studied higher center in the protostome brain, in part due to the regular and easily recognizable organization of their intrinsic neurons. As previously noted, however, mushroom bodies have been acquired in just four protostome phyla (fig. 2). This means that for the majority of protostomes with centralized nervous systems, the site of multimodal sensory integration, higher computations involved in learning and other forms of plasticity, and in motor planning and execution is the so-called ‘diffuse’ neuropil of the protocerebrum. Even in protostomes such as the insects, the mushroom bodies receive few dedicated inputs in that most afferents also provide collaterals to regions in the surrounding protocerebrum [Strausfeld and Li, 1999; Strausfeld, 2002; Kirschner et al., 2006; Jefferis et al., 2007]. This suggests that mushroom bodies, when present, do not replace the ancestral functions of protocerebral circuits, but rather provide novel processing functions in parallel to or in sequence with protocerebral circuits. So the looming question is: why do some species have mushroom bodies, whereas others do not? What selective pressures drove the acquisition of mushroom bodies? What do mushroom bodies provide, functionally, that general protocerebral processing circuits lack? Answering these questions is beyond the scope of this review Farris

article, but will be critical to understanding the ‘what’, ‘how’ and ‘why’ of higher brain center evolution in protostomes, and might even contribute some insight to the same questions that are being asked of the vertebrate cortex, hippocampus and cerebellum. Higher Brain Centers in Deuterostomes: The Vertebrate Cortex When investigating patterns of evolution within higher brain regions, instances of parallel evolution provide striking examples of how similar selective pressures generated by shared behavioral ecologies can drive the repeated acquisition of the same architectural (and presumably functional) solutions. The examples below are considered parallelisms because selection has acted on homologous brain centers that are clearly derived from a common ancestor, for example, the cerebral cortices of mammals or the mushroom bodies of insects. In the case of mammals, the point of divergence among the species being compared is approximately 100 million years ago [Kaas, 2007], while the insect lineages have undergone nearly 300 million years of independent evolution [Grimaldi and Engel, 2005]. The mammalian lineage arose approximately 230 million years ago with a small neocortex in the dorsal forebrain, dwarfed by the ventral olfactory cortex perhaps similar to what is observed in basal extant eutherians such as tenrecs and hedgehogs [Kaas, 1995, 2007]. The neocortex is uniquely organized as an array of repeating, six-layered columns that expanded and diminished in number with remarkable plasticity during mammalian evolution. The neocortex of basal mammals has approximately 20 functional areas, each composed of numerous columns that receive input from a particular sensory modality or other ascending or horizontal source. Computations are performed within and across columns, and produce output through other areas of cortex or descending brain centers. Based on studies of these basal species, as well as comparisons of conserved areas across more derived lineages, the ancestral array of cortical areas appears to have been mostly sensory: two or more somatosensory areas (S1 and S2), an auditory area (A1), and at least one visual area (V1). The cortex has undergone independent expansions in size in several mammalian lineages, which in each case has been associated with an increase in modularity and the acquisition of novel functional areas [Kaas, 1995, 2007]. Increasing cortical size and neuron number creates problems with connectivity, as individual neurons must contact more neighbors at greater distances in order

to maintain the same circuits that were present in a smaller cortex. The solution, in many cases, appears to have been the subdivision of cortex into locally connected modules, operating as functional units and decreasing the total number of long-distance connections needed for communication among them [Kaas, 2000, 2007]. Cortical modules produced in such a way might provide additional power to the ancestral computation as parallel processing streams; or may have been free to undergo evolutionary modifications and acquisitions of novel computational functions. The latter is abundantly demonstrated by the visual cortex, which is proposed to have comprised a single cortical area in ancestral mammals, but can contain 30 or more areas in macaques and other derived primates [Felleman and Van Essen, 1991]. Neurons in primate visual cortical areas respond to stimuli ranging from the relatively simple (orientation in V1 and motion in MT) to complex combinations of visual orientations and contours, the latter presumably for threedimensional shape recognition [Orban, 2008]. This profusion of processing modules that break down the visual scene into fine components is likely to be adaptive for species with behavioral ecologies that depend on both extracting critical details from the background and synthesizing other features into relevant percepts, such as shapes. As areas of the cortex have undergone evolutionary enlargement and become more modular due to connectivity constraints, these modules have in some cases acquired novel, adaptive functions. It should therefore not be surprising to see in species with similar behavioral ecologies which have evolved under similar selective pressures, that parallel instances of cortex expansion and functional area acquisition have occurred. An excellent example of this is seen when comparing the visual cortices of three distantly related arboreal animals, the tree shrews, squirrels and old world monkeys, with that of a ground-dwelling mammal, the rat (fig. 4). Physiological, anatomical and cladistic analyses suggest that V1 and V2 are homologous across all four species, but the tree-dwelling lineages have each independently acquired a profusion of additional areas [Kaas, 2002, 2004, 2007]. In squirrels and primates, these extrastriate areas respond to various aspects of motion such as speed and optic flow, depth perception, and other features of object orientation in three dimensions [van Hooser and Nelson, 2006; Orban, 2008]. The behavioral and sensory ecology of arboreal mammals has driven the acquisition of these high-powered visual processing capabilities: they possess large eyes with wide binocular visual fields and high visual acuity,



111

o Fo

Olfactory Bulb

w Pa

M2 M1

w Pa

t o Fo

t

p

Upper Face Lip PVLimbs

PR

A

SC V2

Face & Vibrissae

Face

Li

V1

S1

Dys.

Face

S2

L i m bs

Aud R h in a

l su

lcu

s

D

B

A, B

C

E

Fig. 4. Parallel evolution of visual cortex in arboreal mammals. A Visual cortex in most mammals, represented here by the rat, is

composed of two functional areas (V1 and V2, dark blue and medium blue, respectively). B–D Light blue shading indicates additional independently acquired visual cortical areas in the squirrel

allowing these agile animals to accurately navigate as they leap from perch to perch as well as imparting the ability to spot potential predators approaching from above [Kaas, 2002; van Hooser and Nelson, 2006]. A preponderance of cone photoreceptors in the retina suggests 112

D

C


(B), tree shrew (C) and owl monkey (D). E Phylogeny of the mammals with location of taxa represented in A–D labeled. A and D used with permission from Kaas [2004]; B, C and E used with permission from Kaas [2002].

that color vision is also important for these species [van Hooser and Nelson, 2006]. The additional expansion of visual cortex observed in primates is proposed to have been driven by other visually dependent behaviors in the primate lineage such as frugivory and sociality [Lefebvre Farris

et al., 2004]. Frugivorous animals must be able to recognize many potential food sources and make judgments about their quality based on ripeness using visual cues, as well as possessing the capacity for visual-spatial learning so that food sources can be revisited as new crops of fruit ripen. Sociality also requires extensive use of visual cues for the recognition of conspecifics within the social group. All of these behaviors might plausibly drive the evolution of enhanced visual processing circuits that are capable of extracting key features for use in decision making in the context of food acquisition and social interaction. Structural Convergence of Mushroom Bodies with Vertebrate Cortex In some vertebrates, particularly those species in which the cortical sheet has undergone an evolutionary expansion in size, the cortex has acquired characteristic folds and fissures termed gyri and sulci, respectively [Welker, 1990]. Gyrencephalic species such as humans possess a greatly enlarged and folded cortex, whereas lissencephalic species like the rat possess a smaller, smooth cortex. Similar morphologies are acquired by the mushroom body calyces, which are formed by the dendrites of intrinsic neurons called Kenyon cells and that receive input from primary sensory centers in the brain. The calyces range in morphology from a single fused sphere to a doubled cup shape (fig. 5A, B). It has long been recognized that this latter ‘gyrencephalic’ phenotype is associated with the largest mushroom bodies, for example the honey bee Apis mellifera with 170,000 Kenyon cells [Witthöft, 1967] and the cockroach Periplaneta americana with 175,000 Kenyon cells per mushroom body [Neder, 1959]. In contrast, only 2500 Kenyon cells form the ‘lissencephalic’ single calyx of Drosophila melanogaster [Hinke, 1961; Balling et al., 1987]. In scarab beetles (Coleoptera: Scarabaeidae), where species might possess either single or double calyces, double calyces are associated with a three-fold increase in mushroom body volume relative to central brain volume and a nine-fold increase in Kenyon cell number (fig. 5C, D) [Farris and Roberts, 2005]. Expansion of the mammalian cortex is associated with the acquisition of new functional areas, some of which serve novel processing functions. Is this also the case for evolutionarily enlarged insect mushroom bodies? Farris and Roberts [2005] noted that the doubled calyces of scarab beetles were divided into subcompartments that were not present in single calyces and, based on the diurnal foraging activity of these species, suggested that the nov-

el subcompartments received input from visual centers. In a subsequent study, fluorescent dextrans were used to trace outputs from the optic lobes, which are the primary visual neuropils of the brain. As predicted, optic lobe output tracts produced collaterals innervating the newly acquired subcompartments of doubled calyces, but did not innervate any portion of single calyces [Farris, 2008a]. Double calyces also received olfactory inputs from the antennal lobes, which are the primary sensory input to the mushroom bodies of scarabs with single calyces and for most insects. In insects, large mushroom bodies with gyrencephalic calyces are observed in three divergent, distantly related taxa, the scarab beetles, the cockroaches, and the social Hymenoptera, and are thus likely to have been acquired independently in these lineages (fig. 5E). Have the large mushroom bodies in all of these taxa been subpartitioned into novel functional units as has been observed in scarabs and in the visual cortices of arboreal mammals? Such subpartitioning of large mushroom bodies appears to be the general rule, as the evidence for structural and functional subcompartmentalization of the doubled calyces of the social Hymenoptera (ants, bees and wasps) is well documented. The deeply cup-shaped calyces of these insects are subpartitioned into three concentric zones: the lip, collar and basal ring [Mobbs, 1982, 1984]. The lip receives olfactory input from the antennal lobes, the collar receives primarily visual input from the medulla and lobula of the optic lobes but possesses a small dorsal subcompartment receiving gustatory input from the subesophaeal ganglion, and the basal ring is further subdivided into three zones that receive visual, olfactory and gustatory input (fig. 6) [Mobbs 1982, 1984; Gronenberg and Hölldobler, 1999; Gronenberg, 2001; Ehmer and Gronenberg, 2002; Strausfeld, 2002; Schröter and Menzel, 2003; Lòpez-Riquelme and Gronenberg, 2004]. The dendrites of morphologically and chemically identified subpopulations of Kenyon cells are mostly restricted to within a single calyx subcompartment, and their axons are grouped into distinct laminae at characteristic levels within the lobes [Strausfeld et al., 2000; Strausfeld, 2002. The mushroom bodies of cockroaches are best known from studies in the blattid cockroach Periplaneta americana. The calyces are completely doubled and as deeply cup-shaped as those of the social Hymenoptera, and functional subpartitioning is clearly indicated by segregation of afferent inputs into four concentric zones [Strausfeld and Li, 1999]. However, the subcompartments



113

A

B

0.09

114

MB volume:Brain volume

room bodies and double calyces in scarab beetles. A Mushroom body in one hemisphere of the brain of the dung beetle Phanaeus vindex, a feeding specialist. As in all specialist scarabs included in the study by Farris and Roberts [2005], the calyx (Ca) forms a single ovoid neuropil. B Mushroom body of the fig beetle Cotinus mutabilis, a feeding generalist. As in other generalist scarabs, this species has doubled calyces with distinct partitioning into dorsal and ventral subcompartments (arrows). C Comparison of mushroom body volume versus total central brain volume in six scarab species with double calyces and five species with single calyces. D Comparison of Kenyon cell number in species with double calyces versus single calyces. E Distribution of large mushroom bodies with doubled calyces across the insect phylogenetic tree suggests the independent evolution of this neuroarchitecture in three separate lineages. Tree adapted from Grimaldi and Engel [2005]. C , D measurements made as described by Farris and Roberts [2005]. Scale bar in A = 20 ␮m, B = 50 ␮m. Kc = Kenyon cell bodies; Pe = pedunculus.

0.08

C

120,000

D

100,000

0.07 Kenyon cell number

Fig. 5. Evolution of large mush-

0.06 0.05 0.04 0.03

80,000 60,000 40,000

0.02 20,000 0.01 0

0

Double Single Calyx morphology

Double Single Calyx morphology

Archaeognatha Zygentoma Paleoptera Orthoptera Dictyoptera Paraneoptera Hymenoptera Coleoptera Neuropterida Diptera Lepidoptera


E

Farris

A

B

C

Fig. 6. Structural and functional subcompartments in the calyx of the honey bee mushroom bodies. A The mushroom body in one hemisphere of the honey bee brain, showing the location of the source of olfactory input (antennal lobe, al) and visual input (optic lobe lobula, lo and medulla, me). Box indicates the area shown in B. B High magnification view of the calyx showing the lip (lp), collar (co) and basal ring (br) subcompartments. C Fluorescent

dextran tracing of afferent input from the optic lobes (green) and antennal lobes (red) reveals mostly non-overlapping input to the collar and lip, respectively, as well as the basal ring. A few overlapping optic lobe afferent terminals are observed in the ventral lip and dorsal basal ring (arrows). ca = Calyx; cb = Kenyon cell bodies; mca = medial calyx; d = dorsal; l = lateral; ml = medial lobe. Figures modified with permission from Gronenberg [2001].

of the cockroach calyces differ from those in the social Hymenoptera in two ways. First, visual input is carried by a single output neuron from the optic lobe medulla and accordingly defines only a tiny subcompartment in the dorsal calyx [Strausfeld and Li, 1999]. This is not surprising considering that cockroaches, unlike most social Hymenoptera such as honey bees, are not active in daylight and thus are unlikely to rely heavily on visual cues for mushroom body-associated functions such as learning and memory [however see Mizunami et al., 1993 and Kwon et al., 2004 for demonstrations that Periplaneta americana are capable of visual spatial learning]. The remaining subcompartments of the Periplaneta calyx are targeted by different populations of olfactory and mechanosensory afferents from the antennal lobes [Strausfeld and Li, 1999]. Again, these modalities are likely to be of particular importance to an insect that is active in the dark or in low-light conditions. The second difference in the organization of calyx subcompartments in Periplaneta is that Kenyon cell dendrites are only roughly re-

stricted to more dorsal or more ventral compartments, rather than being contained within a single compartment. This suggests that relative to the Hymenoptera, Kenyon cell dendrites in Periplaneta play a more substantial role in integrating input to the calyx, although the Kenyon cell axons are still organized into precise layers that are sampled by mushroom body efferents as seen in the honey bee [Li and Strausfeld, 1997, 1999; Strausfeld and Li, 1999; Strausfeld, 2002]. In summary, insect mushroom bodies, like vertebrate cerebral cortices, have independently undergone evolutionary expansions in multiple lineages. Both structures have acquired additional subcompartments, perhaps due to physical limitations in maintaining connectivity as neuronal number has increased. Although ancestral subcompartments and their functions have been maintained during this expansion, new subcompartments have been free to acquire new functions processing the same or different sensory modalities.



115

Did Shared Behavioral Ecologies Drive the Expansion of the Mushroom Bodies and the Cerebral Cortices? As alluded to above, the type and preponderance of sensory inputs to the insect mushroom bodies are linked to the relative importance of those sensory modalities in the behavioral ecology of a particular insect species. Significant visual input to the mushroom bodies is rare, occurring only in the social Hymenoptera and, as recently shown, in some species of scarab beetles. In one group of social Hymenoptera, the ants, the amount of calyx devoted to receiving visual input is tightly correlated with the size of the optic lobes that provide those inputs and with the size of the eyes themselves [Gronenberg and Hölldobler, 1999]. Ants possessing large eyes typically also have large optic lobes that provide input to a large portion of the calyces, similar to what is observed in bees and which is likely to represent the ancestral state for social Hymenoptera (fig. 6). Foragers of ant species with large eyes navigate between food sources and a central nest using visual landmarks, and in the case of the ant Cataglyphis fortis, employ polarized light cues for path integration [Müller and Wehner, 1988; Gronenberg and Hölldobler, 1999; Collett et al., 2001]. Other ant species with large eyes and large visual representations in the calyx are visual predators [Gronenberg and Hölldobler, 1999]. In contrast, ants that employ primarily olfactory cues for navigation outside of the colony, or that do not leave the colony at all as adults, have eyes and optic lobes that are reduced or absent, and visual representations in the calyx that are correspondingly diminished or entirely lacking [Gronenberg and Hölldobler, 1999; Gronenberg, 2001]. Bees and wasps, which also possess large eyes and optic lobes with extensive output to the collar region of the calyx, are similarly well known for their spatial learning capacity and the use of visual cues for navigation while foraging [Capaldi and Dyer, 1999; Menzel et al., 2005]. Honey bees also employ visual cues while foraging to learn the location of high-quality floral nectar sources [Giurfa et al., 1995; Giurfa and Lehrer, 2001]. Honey bees may also be trained in complex visual learning tasks that depend upon shape and color recognition and the grouping of these cues based on similarity [Zhang et al., 1999, 2004], and in delayed-matching-to-sample tasks, the latter of which had previously been demonstrated only in primates [Giurfa et al., 2001]. Complex visual learning capabilities have recently been demonstrated in the wasp Polistes fuscatus, in which face recognition is employed to assess dominance relationships between nestmates [Tibbetts, 2002]. 116


This relationship among behavioral ecology, the sensory periphery and sensory representations in the mushroom bodies is strikingly convergent with observations of somatosensory specializations and their representations in the sensory cortices of mammals. For example, star-nosed moles are fossorial insectivores with greatly reduced eyes and a large star-shaped tactile organ on the snout, the eponymous ‘star.’ The mole uses the appendages of the star to explore its environment via saccadelike movements and to rapidly and accurately detect food items such as earthworms [Catania and Remple, 2004]. Sensory input from the star tactile organ and its eleven appendages forms three separate representations within a huge swath of somatosensory cortex [S1, S2 and S3; Catania, 2005]. Such detailed topographic maps have yet to be demonstrated in the insect mushroom bodies, although a rough map of the dorsal to ventral axis of the optic lobe medulla, and thus the ventral to dorsal axis of the eye (due to a chiasm between the two) is observed in honey bees [Ehmer and Gronenberg, 2002]. Projection neurons from antennal lobe glomeruli also form a rough map as they terminate in the mushroom body calyx and the lateral horn of the protocerebrum, and these representations might carry information about particular classes of odorant molecules such as those relating to food versus pheromones [Jefferis et al., 2007; Lin et al., 2007]. Nevertheless, the expansion of somatosensory cortex in star-nosed moles parallels the expansion of visual representation in the mushroom bodies of social Hymenoptera so that in both cases more processing power in a higher brain center is devoted to a sensory modality of unique importance to the behavioral ecology of the animal. What behavioral ecologies drive the overall expansion of higher brain centers? In both insects and vertebrates, increases in the size of a brain region during the life of an animal might occur during times when that animal is particularly reliant on behaviors that are mediated by that brain region, demonstrating a general relationship between the size of the neural substrate and its behavioral output. In songbirds, song control nuclei in the telencephalon undergo seasonal increases and decreases in size associated with the onset of reproductive behavior and song production [reviewed in Wilbrecht and Kirn, 2004]. Seasonal changes in the size of the hippocampus, a higher brain center that plays a particularly important role in spatial learning, have been reported in food-caching birds such as chickadees [Smulders et al., 1995; but see Hoshooley et al., 2007 for a study that did not find this relationship]. In insects, the worker honey bee mushFarris

room bodies undergo an expansion in volume during adulthood that has both experience-expectant and experience-dependent phases [Withers et al., 1993, 1995; Durst et al., 1994; Fahrbach et al., 1998]. The beginning of the experience-expectant phase is coincident with the onset of orientation flights, short flights in which the soon-to-be-forager circles the hive and presumably makes note of visual landmarks in the surroundings [Capaldi et al., 2000]. Evidence suggesting that the increase in mushroom body volume is driven by sensory experience acquired from orientation and/or foraging flights comes from two very different experiments. First, colony manipulations in which bees are induced to forage precociously, at approximately one week of age rather than three weeks, result in a correspondingly precocious expansion of the mushroom bodies [Withers et al., 1993, 1995]. Second, treating young adult bees with the muscarinic agonist pilocarpine, which mimics cholinergic input to the calyces from olfactory and visual afferents, also results in a premature expansion of mushroom body volume [Ismail et al., 2006]. Together, these studies present a robust case for a role for the honey bee mushroom bodies in forager-specific behaviors that may encompass a panoply of complex behaviors associated with navigation between the food sources and the hive as well as learning features of specific food sources. The above examples present the case for a relationship between brain region size and the need an animal has for behaviors mediated by that brain region. This relationship can be extended to evolutionary comparisons as well. Returning to birds, chickadees and other food-storing Paridae, as well as the crows and jays (Corvidae), display evolutionarily enlarged hippocampi when compared with non-food-caching species [Krebs et al., 1989; Lucas et al., 2004]. A large telencephalon encompassing association areas such as the hyperstriatum ventrale and neostriatum is observed in lineages of birds in which species demonstrate a greater capacity for novel and innovative behaviors, many of which are related to food acquisition [Lefebvre et al., 2004; Sol et al., 2005a, b]. Similar associations between the enlargement of the neocortex and behavioral innovations such as tool use are also reported for primates [Lefebvre et al., 2004]. In addition to the ability to respond to a problem by initiating a completely novel behavior, these lineages are also characterized by overall flexibility in feeding ecology [Lefebvre et al., 2004; Sol et al., 2005a, b]. Feeding generalists, whether insects, birds or mammals, must use multiple sensory cues to assess the quality of each food source that is available, perhaps in the context of time (for example, fruit ripeness),

and remember these associations to optimize future encounters. Although this feeding ecology imparts more flexibility in food choice, it also places increased demand on sensory and cognitive systems than would a feeding ecology that is more restricted [Bernays and Wcislo, 1994; Bernays, 1998, 2001]. These constraints might be even more consequential to insects in which the nervous system is many orders of magnitude smaller than that of most vertebrates and thus possesses far less processing power. Furthermore, many insects live for only a short time as adults, and the costs of investing in a nervous system that can support the flexibility required by a generalist feeding ecology might be prohibitive, as well as unnecessary if an individual is unlikely to live long enough to encounter a given food source more than once. Yet some insects are feeding generalists, some live for weeks or months as adults, and some have large mushroom bodies. Is there any relationship among these factors? In scarab beetles, only species belonging to subfamilies with generalist feeding ecologies possess enlarged mushroom bodies with doubled calyces receiving visual input from the optic lobes [Farris and Roberts, 2005; Farris, 2008a]. Scarab species with small mushroom bodies and single calyces that received only olfactory input were specialist dung feeders belonging to the subfamilies Scarabaeinae and Aphodiinae. One generalist species that was included in the 2005 study, the ruteline Popillia japonica, is a serious pest of horticultural plantings in the eastern United States and has been observed to feed on the leaves, flowers and fruits of over 300 taxonomically distributed species of plants [Potter and Held, 2002]. Japanese beetle adults live for approximately four weeks (a long lifetime for an insect) during which time the females leave the host plant to oviposit in the soil approximately once every three days. This presents an evolutionary scenario during which learning both the identifying characteristics and the locations of suitable food sources would be beneficial for the individual, thus representing a selective pressure for the acquisition of these capabilities and the neural substrates that support them. Although field studies of the foraging choices of individual beetles over time have not been carried out, results of a recent study of the effects of applying non-host odors to otherwise palatable host plants support the hypothesis that Japanese beetles make repeated visits to a food source once it is deemed suitable [Held et al., 2003]. The use of companion plants and the application of non-host masking odors is a pest management strategy based on the finding that volatiles produced by non-host plants might



117

‘mask’ or prevent insects from detecting host plant volatiles, sometimes even producing a repellent effect on the pest species [Thiery and Visser, 1986, 1987]. In these studies of the specialist Colorado potato beetle, a specialist pest of plants belonging to the subfamily Solanaceae, blocking the insect’s detection of host-specific olfactory cues impaired host finding and reduced feeding. In the generalist Japanese beetle, however, the opposite result was observed: over test periods of three to four days, suitable host plants that were associated with companion plants and non-host masking odors recruited more beetles than control plants [Held et al., 2003]. Because the experimental odors and companion plants are not naturally associated with palatable hosts, it is unlikely that their presence was naturally attractive to Japanese beetles. Furthermore, the low rate of accumulation of beetles at control plants suggested that feeding-induced volatiles and/or aggregation pheromones played only a minor role in increasing the attractiveness of a plant. Another potential explanation for the dramatic increase in the number of Japanese beetles appearing at the masking odor and companion plant associated host plants is that their location was learned by beetles that returned multiple times during the duration of the experiment. Although the behavior of individual beetles was not recorded in this study, it is possible that once a beetle made it past the noxious non-host odors associated with an otherwise palatable host, it was able to learn these odors as cues signaling the location of a good place to eat. Over time this would result in a pattern of apparently increased recruitment to the host, as beetles that had associated the masking odor with food returned in increasing numbers. The comparisons of mushroom body morphology and behavioral ecology in scarab beetles suggest that the evolution of novel features characteristic of generalist scarab mushroom bodies were driven by the need for increased cognitive capabilities imparted by a generalist feeding ecology. Is feeding ecology a universal driver for large mushroom body evolution, or do other factors play a role? The social Hymenoptera are an excellent example of a taxon in which just such a complicating factor must be considered: the evolution of sociality. Has Eusociality Driven the Evolutionary Enlargement of Higher Brain Centers in Insects? Eusociality, defined in insects as the presence of cooperative brood care, overlap of generations and reproductive division of labor, is found in species belonging to four families of the Hymenoptera. With the exception of the Formicidae (ants), all of which are eusocial [Hölldobler 118


and Wilson, 1990], the Vespidae, Sphecidae and Apidae contain species that display many intermediate forms of social organization, including some species that are solitary [Hunt, 1999; Cameron, 2004; Brady et al., 2006]. In these latter families it is also likely that eusociality arose more than once independently [Brady et al., 2006]. To date, all studies of mushroom body morphology in these four families have shown these neuropils to be greatly enlarged with doubled, deeply cup-shaped calyces similar to those observed in the honey bee (fig. 6). These species range in social organization from solitary to eusocial and include members of all four families in which eusociality has arisen [Vowles, 1955; Jawlowski, 1959; Gronenberg and Hölldobler, 1999; Ehmer and Hoy, 2000; Gronenberg, 2001; Mares et al., 2005; Molina and O’Donnell, 2007; O’Donnell et al., 2007]. Furthermore, large mushroom bodies might be present in lineages of aculeate Hymenoptera that have not produced eusocial species, such as the Chrysidoidea, Pompilidae and Mutilidae [S.M. Farris, unpublished data; N.J. Strausfeld, personal communication]. The apparent universality of large mushroom bodies across four families of Hymenoptera in which social organization is highly variable provides strong evidence against sociality as a driving force in the initial acquisition of large mushroom bodies in these insects. Feeding ecology varies widely across the Hymenoptera, and although feeding generalism is characteristic of some hymenopteran taxa, specialization does occur in some species. Yellowjacket wasps (Vespula germanica) are extreme generalists that hunt other insects, collect nectar and pollen, and scavenge from a variety of humanproduced food sources [Wood et al., 2006]. The Apidae (bees) feed entirely on the nectar, pollen or oils produced by plants, and species can be classified as polylectic (feeding on many plants), oligolectic (feeding on a few closely related species) or even monolectic (feeding on a single species) [Minkley et al., 1994]. There are no studies to date in which mushroom body size has been quantified and compared across species with different feeding ecologies or social organizations, but based on the available data, it is unlikely that any of crown Hymenoptera will lack the characteristic large mushroom bodies with doubled, cup-shaped, subcompartmentalized calyces [SM Farris, unpublished data]. The one remaining complex behavioral feature that links all these species is the extensive use of olfactory and visual cues for navigation to and from a central colony or, in the case of solitary species, a burrow that is provisioned with food for the larva. Whether place-centered foraging and the requisite spatial learnFarris

ing capabilities drove the evolution of large mushroom bodies in the ancestors of extant ants, bees and wasps awaits further studies employing more species encompassing additional branches of the Hymenoptera family tree. The last group of insects in which large mushroom bodies have evolved is the cockroaches (Dictyoptera). Cockroaches such as Periplaneta americana are famously generalist detritivores that live for up to a year as adults [Bell et al., 2007]. They are gregarious [some species are semisocial and even provide parental care and cooperative brood care for their young; Costa, 2006] and may carry food to a central hiding place in a manner reminiscent of the Hymenoptera that forage for food from a central nest [Bell et al., 2007; S.M. Farris, personal observation]. The cockroach Blattella germanica has been demonstrated to form spatial memories of food locations when tested in a laboratory setting [Durier and Rivault, 2001]; Periplaneta also demonstrates the ability to learn a specific angle between two visual cues in association with a food-related odor [Kwon et al., 2004]. The ability to perform all of these behaviors might have driven the evolution of large mushroom bodies in cockroaches, or alternately, some of these behaviors may have emerged as a result of the acquisition of large mushroom bodies due to other forces. One provocative possibility is suggested by studies of termites, which comprise the other major eusocial insect lineage and which, similar to the social Hymenoptera, possess greatly enlarged mushroom bodies yet arose from within a taxon in which large mushroom bodies were most likely ancestral [in the case of termites, the cockroaches; Farris and Strausfeld, 2003; Inward et al., 2007a, b]. For this latter reason both the termites and the social Hymenoptera are less than ideal models for identifying the factors that initiated the acquisition of enlarged mushroom bodies. However, they do represent excellent opportunities for investigating the effects of different social organizations and foraging strategies on other features of mushroom body morphology, with the benefit of a two-way comparison afforded by the parallel evolution of sociality in these distantly related lineages [Grimaldi and Engel, 2005].

Conclusions

Evidence supporting a single origin of the bilaterian nervous system suggests that higher brain centers in animals as distantly related as annelids, insects and vertebrates share common structural and functional elements. Convergence of Higher Brain Centers

Adaptation of this common groundplan to similar environments has resulted in the convergent evolution of higher brain centers that spans the protostome-deuterostome boundary. This is evident when comparing representations of the sensory periphery in higher brain centers, in which inputs from sensory modalities that are of particular importance to the animal occupy a greater portion of the mushroom body calyx (insects) or cerebral cortex (mammals). Behavioral ecologies that place stronger demands on flexible behaviors, often related to food acquisition, appear to drive the evolutionary enlargement of higher brain centers in both vertebrates and insects. As these brain centers become larger, physical constraints favor the formation of additional functional subcompartments that might take on novel processing roles. These new functions could in turn prove adaptive to the behavioral ecology of the animal, as suggested by the acquisition of visual input to the calyces in diurnally foraging scarab beetles and honey bees, and of novel visual processing areas in the visual cortex of arboreal mammals. Taken together, the above-cited studies support the presence of uniformly applicable principles of brain and behavioral evolution in animals. The field is wide open for studies that further detail and solidify these relationships, and that explore the developmental mechanisms underlying the evolution of neuroarchitectural and behavioral diversity.

Acknowledgements The author thanks Dr. Hans Hofmann and Dr. Caroly Shumway for organizing the 2007 Karger Workshop and for inviting the author to present her research. The author would also like to thank an anonymous reviewer whose comments greatly improved the quality of this manuscript.

References

Anderson PAV, Mackie GO (1977) Electrically coupled, photosensitive neurons control swimming in a jellyfish. Science 197: 186– 188. Arendt D, Denes AS, Jékely G, Tessmar-Raible K (2007) The evolution of nervous system centralization. Phil Trans R Soc B Epub ahead of print:1–6. Balling A, Technau GM, Heisenberg M (1987) Are the structural changes in adult Drosophila mushroom bodies memory traces? Studies on biochemical learning mutants. J Neurogenet 4:65–73. Bell WJ, Roth LM, Nalepa CA (2007) Cockroaches: Ecology, Behavior and Natural History. Baltimore, MD: Johns Hopkins University Press.


119

Bernays EA (1998) The value of being a resource specialist: behavioral support for a neural hypothesis. Am Nat 151:451–464. Bernays EA (2001) Neural limitations in phytophagous insects: Implications for diet breadth and evolution of host affiliation. Annu Rev Entomol 46: 703–727. Bernays EA, Wcislo WT (1994) Sensory capabilities, information processing, and resource specialization. Q Rev Biol 69: 187–204. Brady SG, Sipes S, Pearson A, Danforth BN (2006) Recent and simultaneous origins of eusociality in halictid bees. Proc Biol Sci 273: 1643–1649. Brusca RC, Brusca GJ (2003) Invertebrates, 2nd ed. Sunderland, MA: Sinauer Associates. Cameron SA (2004) Phylogeny and biology of neotropical orchid bees (Euglossini). Annu Rev Entomol 49:377–404. Capaldi EA, Dyer FC (1999) The role of orientation flights on homing performance in honeybees. J Exp Biol 202: 1655–1666. Capaldi EA, Smith AD, Osborne JL, Fahrbach SE, Farris SM, Reynolds DR, Edwards AS, Martin A, Robinson GE, Poppy GM, Riley JR (2000) Ontogeny of orientation flight in the honeybee revealed by harmonic radar. Nature 403:537–540. Carapelli A, Liò P, Nardi F, van der Wath E, Frati F (2007) Phylogenetic analysis of mitochondrial protein coding genes confirms the reciprocal paraphyly of Hexapoda and Crustacea. BMC Evol Biol 7 Suppl 2:S8. Cassenaer S, Laurent G (2007) Hebbian STDP in mushroom bodies facilitates the synchronous flow of olfactory information in locusts. Nature 448:709–713. Catania KC (2005) Evolution of sensory specialization in insectivores. Anat Rec A Discov Mol Cell Evol Biol 287: 1038–1050. Catania KC, Remple FE (2004) Tactile foveation in the star-nosed mole. Brain Behav Evol 63: 1–12. Collett TS, Collett M, Wehner R (2001) The guidance of desert ants by extended landmarks. J Exp Biol 204:1635–1639. Costa JT (2006) The Other Insect Societies. Cambridge, MA: Belknap Press. Denes AS, Jékely G, Steinmetz PRH, Raible F, Snyman H, Prud’homme B, Ferrier DEK, Balavoine G, Arendt D (2007) Molecular architecture of annelid nerve cord supports common origin of nervous system centralization in Bilateria. Cell 129:277–288. Durier V, Rivault C (2001) Effects of spatial knowledge and feeding experience on foraging choices in German cockroaches. Anim Behav 62:681–688. Durst C, Eichmuller S, Menzel R (1994) Development and experience lead to increased volume of subcompartments of the honeybee mushroom body. Behav Neural Biol 62:259– 263. Ehmer B, Gronenberg W (2002) Segregation of visual input to the mushroom bodies in the honeybee (Apis mellifera). J Comp Neurol 451:362–373.

120

Ehmer B, Hoy R (2000) Mushroom bodies of vespid wasps. J Comp Neurol 416:93–100. Erber J, Masuhr T, Menzel R (1980) Localization of short-term memory in the brain of the bee, Apis mellifera. Physiol Entomol 5:343–358. Fahrbach SE, Moore D, Capaldi EA, Farris SM, Robinson GE (1998) Experience-expectant plasticity in the mushroom bodies of the honey bee. Learn Mem 5: 115–123. Farris SM (2005) Evolution of insect mushroom bodies: Old clues, new insights. Arthropod Struct Dev 34:211–234. Farris SM (2008a) Structural, functional and developmental convergence of the insect mushroom bodies with higher brain centers of vertebrates. Brain Behav Evol, in press. Farris SM (2008b) Tritocerebral tract input to the insect mushroom bodies. Arthro Struc Dev, in press. Farris SM, Roberts NS (2005) Coevolution of generalist feeding ecologies and gyrencephalic mushroom bodies in insects. Proc Nat Acad Sci USA 102:17394–17399. Farris SM, Strausfeld NJ (2003) A unique mushroom body substructure common to both basal cockroaches and to termites. J Comp Neurol 456:305–320. Felleman DJ, Van Essen DC (1991) Distributed hierarchical processing in the primate cerebral cortex. Cereb Cortex 1:1–47. Giurfa M, Lehrer M (2001) Honeybee vision and floral displays: from detection to close-up. In: Cognitive Ecology of Pollination: Animal Behavior and Floral Evolution (Chittka L, Thomson JD, eds), pp 61–82. Cambridge, UK: Cambridge University Press. Giurfa M, Núnez J, Chittka L, Menzel R (1995) Colour preferences of flower-naive honeybees. J Comp Physiol A 177:247–259. Giurfa M, Zhang S, Jenett A, Menzel R, Srinivasan MV (2001) The concepts of ‘sameness’ and ‘difference’ in an insect. Nature 410: 930–933. Glenner H, Hansen AJ, Sørensen MV, Ronquist R, Huelsenbeck JP, Willerslev E (2004) Bayesian inference of the metazoan phylogeny: a combined molecular and morphological approach. Curr Biol 14: 1644–1649. Glenner H, Thomsen PF, Hebsgaard MB, Sørensen MV, Willerslev E (2006) The origin of insects. Science 314:1883–1884. Grimaldi D, Engel M (2005) Evolution of the Insects. New York: Cambridge University Press. Gronenberg W (2001) Subdivisions of hymenopteran mushroom body calyces by their afferent supply. J Comp Neurol 436: 474– 489. Gronenberg W, Hölldobler B (1999) Morphologic representation of visual and antennal information in the ant brain. J Comp Neurol 412:229–240. Hanström B (1940) Inkretorische Organe, Sinnesorgane und Nervensystem des Kopfes einiger niederer Insektenordnungen. Kungl Svenska Vetenskaps Akademiens Handlingar 18:1–266.


Heisenberg M (2003) Mushroom body memoirs: from maps to models. Nature Neurosci Rev 4:266–275. Held DW, Gonsiska P, Potter DA (2003) Evaluating companion planting and non-host masking odors for protecting roses from the Japanese beetle (Coleoptera: Scarabaeidae). J Econ Entomol 96:81–87. Heuer CM, Loesel R (2008) Immunofluorescence analysis of the internal brain anatomy of Nereis diversicolor (Polychaeta, Annelida). Cell Tissue Res 331:713–724. Hinke W (1961) Das relative postembryonale Wachstum der Hirnteile von Culex pipiens, Drosophila melanogaster und Drosphila Mutanten. Z Morphol Okol Tiere 50:81–118. Hirth F, Reichert H (2006) Basic nervous system types: one or many? In: Evolution of Nervous Systems (Kaas JH, ed), vol 1, pp 55–72. Oxford, UK: Academic Press. Hirth F, Kammermeier L, Frei E, Walldorf U, Noll M, Reichert H (2003) An urbilaterian origin of the tripartite brain: developmental genetic insights from Drosophila. Development 130:2365–2373. Hölldobler B, Wilson EO (1990) The Ants. Berlin: Springer-Verlag. Hoshooley JS, Phillmore LS, Sherry DF, Macdougall-Shackleton SA (2007) Annual cycle of the black-capped chickadee: seasonality of food-storing and the hippocampus. Brain Behav Evol 69:161–168. Hunt JH (1999) Trait mapping and salience in the evolution of eusocial vespid wasps. Evolution 53:225–237. Inward D, Beccaloni G, Eggleton P (2007a) Death of an order: a comprehensive molecular phylogenetic study confirms that termites are eusocial cockroaches. Biol Lett 3: 331–335. Inward DJG, Vogler AP, Eggleton P (2007b) A comprehensive phylogenetic analysis of termites (Isoptera) illuminates key aspects of their evolutionary biology. Mol Phylogenet Evol 44:953–967. Ismail N, Robinson GE, Fahrbach SE (2006) Stimulation of muscarinic receptors mimics experience-dependent plasticity in the honey bee brain. Proc Natl Acad Sci USA 103: 207–211 . Jawlowski H (1959) The structure of corpora pedunculata in Aculeata (Hymenoptera). Fol Biol 7:61–70. Jefferis GSXE, Potter CJ, Chan AM, Marin EC, Rohlfing T, Mauer CRJ, Luo L (2007) Comprehensive maps of Drosophila higher olfactory centers: spatially segregated fruit and pheromone representation. Cell 128: 1187– 1203. Jha RK, Mackie GO (1967) The recognition, distribution and ultrastructure of hydrozoan nerve elements. J Morphol 123:43–61. Kaas JH (1995) The evolution of isocortex. Brain Behav Evol 46:187–196. Kaas JH (2000) Why is brain size so important: design problems and solutions as neocortex gets bigger or smaller. Brain Mind 1:7–23.

Farris

Kaas JH (2002) Convergences in the modular and areal organization of the forebrain of mammals: implications for the reconstruction of forebrain evolution. Brain Behav Evol 59:262–272. Kaas JH (2004) Evolution of somatosensory and motor cortex in primates. Anat Rec 281A: 1148–1156. Kaas JH (2007) Reconstructing the organization of neocortex of the first mammals and subsequent modifications. In: Evolution of Nervous Systems (Kaas JH, Striedter GF, Bullock TH, Preuss TM, Rubenstein J, Krubitzer LA, eds), vol 3, pp 27–48. New York: Academic Press. Kass-Simon G, Pierobon P (2007) Cnidarian chemical neurotransmission, an updated overview. J Comp Biochem Physiol A Mol Integr Physiol 146:9–25. Kirschner S, Kleineidam C, Zube C, Rybak J, Grünewald B, Rössler W (2006) Dual olfactory pathway in the honeybee, Apis mellifera. J Comp Neurol 499:933–952. Krebs JR, Sherry DF, Healy SD, Perry VH, Vaccarino AL (1989) Hippocampal specialization of food-storing birds. Proc Nat Acad Sci USA 86:1388–1392. Kwon HW, Lent DD, Strausfeld NJ (2004) Spatial learning in the restrained American cockroach Periplaneta americana. J Exp Biol 207: 377–383. Laurent G (2006) Shall we even understand the fly’s brain? In: 23 Problems in Systems Neuroscience (van Hemmen JL, Sejnowski TJ, eds), pp 3–21. New York: Oxford University Press. Lefebvre L, Reader SM, Sol D (2004) Brains, innovations and evolution in birds and primates. Brain Behav Evol 63: 233–246. Li YS, Strausfeld NJ (1997) Morphology and sensory modality of mushroom body extrinsic neurons in the brain of the cockroach, Periplaneta americana. J Comp Neurol 387:631– 650. Li YS, Strausfeld NJ (1999) Multimodal efferent and recurrent neurons in the medial lobes of cockroach mushroom bodies. J Comp Neurol 409:647–663. Lichtneckert R, Reichert H (2005) Insights into the urbilaterian brain: conserved genetic patterning mechanisms in insect and vertebrate brain development. Heredity 94: 465– 477. Lichtneckert R, Reichert H (2006) Origin and evolution of the first nervous system. In: Evolution of Nervous Systems (Kaas JH, ed), vol 1, pp 290–315. Oxford, UK: Academic Press. Lin H, Lai J, Chin A, Chen Y, Chiang A (2007) A map of olfactory representation in the Drosophila mushroom body. Cell 128: 1205– 1217. Liu L, Wolf R, Ernst R, Heisenberg M (1999) Context generalization in Drosophila visual learning requires the mushroom bodies. Nature 400:753–756.


Lòpez-Riquelme GO, Gronenberg W (2004) Multisensory convergence in the mushroom bodies of ants and bees. Acta Biol Hung 55: 31–37. Love A, Raff R (2003) Knowing your ancestors: themes in the history of evo-devo. Evol Dev 5:327–330. Lucas JR, Brodin A, de Kort SR, Clayton NS (2004) Does hippocampal size correlate with the degree of caching specialization? Proc Biol Sci 271: 2423–2429. Mares S, Ash L, Gronenberg W (2005) Brain allometry in bumblebee and honey bee workers. Brain Behav Evol 66:50–61. Meech RW, Mackie GO (1995) Ionic currents in giant motor axons of the jellyfish, Aglantha digitale. J Neurophysiol 69: 884–893. Menzel R, Greggers U, Smith A, Berger S, Brandt R, Brunke S, Bundrock G, Hülse S, Plümpe T, Schaupp F, Schüttler E, Stach S, Stindt J, Stollhoff N, Watzl S (2005) Honey bees navigate according to a map-like spatial memory. Proc Natl Acad Sci USA 102:3040–3045. Minkley RL, Wcislo WT, Yanega D (1994) Behavior and phenology of a specialist bee (Dieunomia) and sunflower (Helianthus) pollen availability. Ecology 75:1406–1419. Mizunami M, Weibrecht JM, Strausfeld NJ (1993) A new role for the insect mushroom bodies: Place memory and motor control. In: Biological Neural Networks in Invertebrate Neuroethology and Robotics (Beer RD, Ritzman RE, McKenna T, eds), pp 199–225. New York: Academic Press. Mizunami M, Weibrecht J, Stausfeld N (1998) Mushroom bodies of the cockroach: their participation in place memory. J Comp Neurol 402:520–537. Mizutani CM, Meyer N, Roelink H, Bier E (2006) Threshold-dependent BMP-mediated repression: a model for a conserved mechanism that patterns the neuroectoderm. PLoS Biol 4:1777–1788. Mobbs PG (1982) The brain of the honeybee Apis mellifera L. The connections and spatial organization of the mushroom bodies. Phil Trans R Soc Lond B 298:309–354. Mobbs PG (1984) Neural networks in the mushroom bodies of the honeybee. J Insect Physiol 30:43–58. Molina Y, O’Donnell S (2007) Mushroom body volume is related to social aggression and ovary development in the paper wasp Polistes instabilis. Brain Behav Evol 70:137–144. Müller I, Wehner R (1988) Path integration in desert ants, Cataglyphis fortis. Proc Nat Acad Sci USA 85: 5287–5290. Neder R (1959) Allometrisches Wachstum von Hirnteilen bei drei verschieden grossen Schabenarten. Zool Jahrb Anat 77:411–464. O’Donnell S, Donlan N, Jones T (2007) Developmental and dominance-associated differences in mushroom body structure in the paper wasp Misocyttarus mastigophorus. Dev Neurobiol 67: 39–46.

Orban GA (2008) Higher order visual processing in macaque extrastriate cortex. Physiol Rev 88:59–89. Perez-Orive J, Mazor O, Turner GC, Cassenaer S, Wilson RI, Laurent G (2002) Oscillations and sparsening of odor representation in the mushroom body. Science 297: 359–365. Perovic S, Krasko A, Prokic I, Müller IM, Müller WE (1999) Origin of neuronal-like receptors in Metazoa: cloning of a metabotropic glutamate/GABA-like receptor from the marine sponge Geodia cydonium. Cell Tissue Res 296:395–404. Potter DA, Held DW (2002) Biology and management of the Japanese beetle. Annu Rev Entomol 47:175–205. Regier JC, Shultz JW, Kambic RE (2005) Pancrustacean phylogeny: hexapods are terrestrial crustaceans and maxillopods are not monophyletic. Proc Biol Sci 272:395–401. Rhinn M, Brand M (2001) The midbrain-hindbrain boundary organizer. Curr Opin Neurobiol 11:34–42. Roman G, Davis R (2001) Molecular biology and anatomy of Drosophila olfactory associative learning. Bioessays 23: 571–581. Sakarya O, Armstron KA, Adamska M, Adamski M, Wang I-F, Tidor B, Degnan BM, Oakley TH, Kosik KS (2007) A post-synaptic scaffold at the origin of the animal kingdom. PLoS One 2:e506. Schildberger K (1984) Multimodal interneurons in the cricket brain: properties of identified extrinsic mushroom body cells. J Comp Physiol 154:71–79. Schröter U, Menzel R (2003) A new ascending sensory tract to the calyces of the honeybee mushroom body, the subesophageal-calycal tract. J Comp Neurol 465:168–178. Smulders TV, Sasson AD, DeVoogd TJ (1995) Seasonal variation in hippocampal volume in a food-storing bird, the black-capped chickadee. J Neurobiol 27:15–25. Sol D, Duncan RP, Blackburn TM, P. C, Lefebvre L (2005a) Big brains, enhanced cognition, and response of birds to novel environments. Proc Natl Acad Sci USA 102: 5460– 5465. Sol D, Lefebvre L, Rodriguez-Teijeiro JD (2005b) Brain size, innovative propensity and migratory behavior in temperate Palearctic birds. Proc R Soc Lond B 272:1433–1441. Strausfeld NJ (1998) Crustacean- insect relationships: the use of brain characters to drive phylogeny amongst segmented invertebrates. Brain Behav Evol 52:186–206. Strausfeld NJ (2002) Organization of the honey bee mushroom body: representation of the calyx within the vertical and gamma lobes. J Comp Neurol 450:4–33. Strausfeld NJ, Barth FG (1993) Two visual systems in one brain: Neuropils serving the secondary eyes of the spider Cupiennius salei. J Comp Neurol 328:55–63.


121

Strausfeld NJ, Li YS (1999) Organization of olfactory and multimodal afferent neurons supplying the calyx and pedunculus of the cockroach mushroom bodies. J Comp Neurol 409:603–625. Strausfeld NJ, Buschbeck EK, Gomez RS (1995) The arthropod mushroom body: Its functional roles, evolutionary enigmas and mistaken identities. In: The Nervous Systems of Invertebrates: An Evolutionary and Comparative Approach (Briedbach O, Kutsch W, eds), pp 349–381. Basel: Birkhäuser Verlag. Strausfeld NJ, Hansen L, Li Y, Gomez RS, Ito K (1998) Evolution, discovery, and interpretations of arthropod mushroom bodies. Learn Mem 5:11–37. Strausfeld NJ, Homberg U, Kloppenburg P (2000) Parallel organization in honey bee mushroom bodies by peptidergic Kenyon cells. J Comp Neurol 424:179–195. Strausfeld NJ, Strausfeld CM, Stowe S, Rowell D, Loesel R (2006) The organization and evolutionary implications of neuropils and their neurons in the brain of the onychophoran Euperipatoides rowelli. Arthropod Struct Dev 35:169–196.

122

Tessmar-Raible K, Raible F, Christodoulou F, Guy F, Rembold M, Hausen H, Arendt D (2007) Conserved sensory-neurosecretory cell types in annelid and fish forebrain: insight into hypothalamus evolution. Cell 129: 1389–1400. Thiery D, Visser JH (1986) Masking of host odour in the olfactory orientation of the Colorado potato beetle. Entomol Exp Appl 41: 165–172. Thiery D, Visser JH (1987) Misleading the Colorado potato beetle with an odor blend. J Chem Ecol 13:1139–1146. Tibbetts E (2002) Visual signals of individual identity in the wasp Polistes fuscatus. Proc Biol Sci 269:1423–1428. van Hooser SD, Nelson SB (2006) The squirrel as a rodent model of the human visual system. Vis Neurosci 23:765–778. Vowles DM (1955) The structure and connexions of the corpora pedunculata in bees and ants. Quart J Micro Sci 96:239–255. Welker W (1990) Why does cerebral cortex fissure and fold? A review of determinants of gyri and sulci. In: Cerebral Cortex. Vol. 8B: Comparative Structure and Evolution of the Cerebral Cortex Part II (Jones EG, Peters A, eds), pp 3–136. New York: Plenum Press.


Wilbrecht L, Kirn JR (2004) Neuron addition and loss in the song system: regulation and function. Ann NY Acad Sci 1016:659–683. Withers GS, Fahrbach SE, Robinson GE (1993) Selective neuroanatomical plasticity and division of labour in the honeybee. Nature 364: 238–240. Withers GS, Fahrbach SE, Robinson GE (1995) Effects of experience and juvenile hormone on the organization of the mushroom bodies of honey bees. J Neurobiol 26: 130–144. Witthöft W (1967) Absolute Anzahl und Verteilung der Zellen im Hirn der Honigbiene. Zeitschr Morphol Tiere 61:160–164. Wood GM, Hopkins DC, Schellhorn NA (2006) Preference by Vespula germanica (Hymenoptera: Vespidae) for processed meats: implications for toxic baiting. J Econ Entomol 99:263–267. Zhang S, Lehrer M, Srinivasan MV (1999) Honeybee memory: navigation by associative grouping and recall of visual stimuli. Neurobiol Learn Mem 72:180–201. Zhang S, Srinivasan MV, Zhu H, Wong J (2004) Grouping of visual objects by honeybees. J Exp Biol 207:3289–3298.

Farris



Habitat Complexity, Brain, and Behavior Caroly A. Shumway a, b a

The Nature Conservancy, and b Brown University, Department of Psychology, Providence, R.I., USA

Key Words Habitat complexity ⴢ Fish ⴢ Cichlid ⴢ Evolution ⴢ Brain ⴢ Behavior ⴢ Ecology

Abstract More complex brains and behaviors have arisen repeatedly throughout both vertebrate and invertebrate evolution. The challenge is to tease apart the forces underlying such change. In this review, I show how habitat complexity influences both brain and behavior in African cichlid fishes, drawing on examples from primates and birds where appropriate. These species groups share a number of similarities. They exhibit a considerable range of brain to body weight within their group. Often highly visual, the species show a diversity of habitat types, social systems, and cognitive abilities. Phylogenies are well established. In closely-related cichlid fishes from the monophyletic Ectodine clade of Lake Tanganyika, habitat complexity is directly correlated with social variables, including species richness, diversity, and abundance. Total brain size, telencephalic and cerebellar size are positively correlated with habitat complexity. Visual acuity and spatial memory are also enhanced in cichlids living in more complex environments. I speculate that species-specific neural effects of environmental complexity could be the consequence of the corresponding social changes. However, environmental and social forces affect brains differently. Environmental forces exert a broader effect on brain struc-



tures than social ones, suggesting either allometric expansion of the brain structures in concert with brain size and/or co-evolution of these structures. To advance our understanding of the mechanism by which habitat complexity affects brain and behavior will require the use of closely-related species, quantification of complexity, hypothesis testing restricting analysis to a single variable and path analyses to explore the order of importance of such variables. We will also need new experimental paradigms exploring the cognitive and survival value of brain and brain structure changes both in the laboratory and in the wild. Copyright © 2008 S. Karger AG, Basel

Introduction ‘The natural emphasis on discovering commonalities, homologies and analogies that simplify has distracted us from the main result of evolution, which is to create differences’. Ted Bullock, 2002

More complex brains and behaviors have arisen repeatedly throughout both vertebrate and invertebrate evolution. The challenge is to tease apart the forces underlying such change. Both mosaic evolution, also known as the adaptationist model, which hypothesizes that selection can act on functionally distinct brain regions controlling particular Dr. Caroly A. Shumway The Nature Conservancy 159 Waterman Street Providence, RI 02906 (USA) Tel. +1 401 331 7110 ext. 13, Fax +1 401 273 4902, E-Mail [email protected]

behaviors [e.g., Barton and Harvey, 2000], and developmental constraints, the hypothesis that developmental processes regulate brain change as a whole [e.g., Finlay and Darlington, 1995] are thought to play a role in driving brain/behavior changes [Striedter, 2005]. However, fishes are thought to be freer from developmental constraints on brain expansion than other vertebrates due to their indeterminate growth. As reviewed by Kotrschal et al. [1998], skull volume is considered unlikely to be a constraint to brain size due to the fact that the brain is considerably smaller than the skull cavity in most fishes. Brain size might be constrained only in the smallest fish: the smaller the fish, the proportionately larger its brain. The reduced developmental constraints, coupled with whole genome duplication that occurred twice in the teleost lineage [Ohno, 1970; Amores et al., 1998] as well as the tremendous diversity of fish habitats might explain the extraordinary diversity of fish brains we see today. In my lab, we work to understand how complex brains and behaviors are shaped over different timescales – from development to evolutionary, using closely-related African cichlid fishes. Given the intricacy of the neural substrates underlying complex behaviors, comparative studies of closely related species can be a particularly powerful approach [e.g., Healy et al., 2005; Farris, 2008; Lefebvre and Sol, 2008]. Although we have known for years that higher order brain regions show considerable diversity in size across vertebrates and invertebrates, we have had a limited understanding as to what caused such changes [Bullock, 2002]. In recent years, progress in understanding the evolution of brain and behavior has occurred due to the following three advances: 1. Quantification of complexity: Although earlier papers had demonstrated correlations of brain size with different habitat features, the environmental features were not quantitatively assessed. For example, Bauchot et al. [1977, 1989] qualitatively associated a larger telencephalon in marine fish families that inhabit a coral reef environment. Similarly, Huber et al. [1997] had demonstrated significant differences in overall brain size of African cichlids from all three of the African rift lakes (Lakes Victoria, Malawi, and Tanganyika) as well as the size of the telencephalon relative to qualitative categories of physical environments. These papers, however, were useful in their broad comparisons across families and their identification of the importance of the environment on the size of the brain and brain structures. 2. Increased use of phylogenetically sound comparisons and phylogenetic controls: Earlier papers, such as 124


the study of 113 teleost families by Bauchot et al. [1979] were confounded by variations in phylogenetic distance. We now recognize that because of the problem of statistical nonindependence in correlational studies of species [Felsenstein, 1985; Harvey and Purvis, 1991], it is necessary to control for varying relatedness [but see Rheindt et al., 2004, for discussion of pitfalls in phylogenetic testing]. 3. Experimental manipulations and hypothesis-testing: Correlations obviously say nothing about causality. For us to move beyond ‘just-so’ stories speculating on the connection among environmental forces, brain and behavior, hypotheses on these connections must be specifically tested [Gould and Lewontin, 1979]. Recent papers on fishes have demonstrated experimentally that habitat complexity causes greater species diversity and richness [e.g., Lingo et al., 2006] and that limiting the visual field influences home range size, mating and feeding, showing the influence of visual stimuli on how an animal uses its space [Imre et al., 2002; Rilov et al., 2007]. Most significantly, Sol et al. [2007] tested the hypothesis that big brains confer an evolutionary advantage, an assumption not tested by previous researchers. By comparing natural mortalities in 200 bird species with brain size, they showed for the first time that big-brained animals have greater survival levels than their smaller-brained counterparts. The result holds with or without independent contrasts. In this review, I present our work exploring the influence of habitat complexity on both brain and behavior in African cichlids, drawing on other examples where appropriate. The main conclusion of this paper is that for us to advance our understanding of habitat complexity on brain and behavior, we must consider neural and behavioral evolution of a given animal in context with how that animal influences and is influenced by the environmental and social world around them.

Cichlids Are a Model System for Exploring Links between Brain and Behavioral Complexity

My lab is interested in understanding the forces underlying telencephalic expansion: environmental, social, or both. Understanding which forces are most important in telencephalic expansion and the relationship among these forces is of great importance in primate, bird, and teleost evolution [Bshary et al., 2002; Dunbar and Shultz, 2007; Lefebvre and Sol, 2008]. These species groups share a number of similarities. They exhibit a considerable Shumway

range of brain to body weight amongst their group. The relative telencephalic size (log telencephalic volume/log body volume) of African cichlids, for example, varies thirteen to twenty-five fold among species [unpublished observations]. Often highly visual, the species show a diversity of habitat types, social systems, and cognitive abilities. Phylogenies are well established. African cichlids provide an unparalleled opportunity to understand how environmental and social pressures influence neural evolution due to the extremely fine ecological, behavioral, and neuronal comparisons that can be made. They live in habitats of varying complexity (ranging from sand to large fusiform rocks), exhibit diverse feeding strategies (ranging from detritivores to scale-eaters) and differ in social behaviors (ranging from polygamy to monogamy), yet share close genetic similarity. The Ectodini clade in Lake Tanganyika, for example, comprises 35 species, enabling one to judiciously select species varying either by habitat (sand/intermediate habitats/rock) or by social type (monogamy/polygamy), while keeping feeding behavior constant. Relative telencephalic volume can vary as much as two-fold among species in this clade [unpublished observations]. Recent phylogenies suggest that sand-dwelling is the ancestral trait, and that rock-dwelling independently evolved twice in the clade [Koblmüller et al., 2004]. Further, cichlids are highly visual, similar to birds and primates, and their visual behavior is well known [cf., Fernald, 1982; Keenleyside, 1991; Carleton and Kocher, 2001]. Comparison of visual and spatial abilities in fishes, the dominant vertebrate radiation, is likely to yield useful comparisons with mammals and birds and aid our understanding of how animals use visual information as a representation of their spatial world.

Habitat Complexity and Ecology

phy), a standard measure used in coral reef biology, and a new video-based technique for assaying habitat complexity in aquatic ecosystems called optical intensity [Shumway et al., 2007]. The new method is a purely visual technique that is both quantitative and scale independent. Essentially, the method measures the variation in optical intensity obtained from video images. [For detailed methods, see Shumway et al., 2007]. We tested the technique on 38 quadrats (5 ! 5 m) in the lake to determine if three freshwater habitats (sand, intermediate, and rock) were quantitatively different. The method generated significant differences among these habitats. A comparison of the values obtained from optical intensity with rugosity showed that the measures were positively correlated. Both methods have their advantages and disadvantages. Rugosity can be measured more quickly in the field and requires less time for analysis. However, one must know the scale of interest in advance, as the index can vary with variations in the chain length and link size. Scale matters, because the importance of any given visual feature depends on its size relative to that animal [Schmidt-Nielsen, 1984]. Further, rugosity cannot discriminate the shape or size of different habitats, relative to scale, nor is it capable of distinguishing fine textural differences [Roberts and Ormond, 1987; McCormick, 1994]. The advantage of the optical intensity method is that it enables one to adjust the scale after the field measures are made; the measure can be used at depths and in locations inaccessible to SCUBA; and the method enables one to pursue a nested analysis of scale [Shumway et al., 2007]. Combining both continuous measures alleviates the disadvantages of each method, and might ensure a more robust analysis, as each variable is influenced by different aspects of the surface topography. Numerous authors have previously recommended multiple approaches for quantifying habitat complexity [e.g., McCormick, 1994].

Quantification of Complexity Quantification of habitat complexity is essential in the search for factors underlying speciation and the development of behavioral and neural complexity. Without quantification and the use of a continuous variable, one is unable to discern whether there are corresponding quantitative or qualitative shifts in brain and behavior as an animal evolves in a new environment. Further, comparisons are not possible among researchers or across different habitats. We used two measures to quantify habitat complexity in Lake Tanganyika: rugosity (aka surface topogra-

Habitat Complexity and the Social Milieu Habitat complexity has clearly been shown to affect an animal’s survival, influencing reproduction, foraging, and predator avoidance. In fishes alone, habitat complexity has been correlated with juvenile survival [e.g., Johnson, 2007], predation and predator avoidance strategies [Brown and Warburton, 1997], alternative male reproductive strategies [Lukasik et al., 2006], and home range size [Imre et al., 2002]. In thinking about how habitat complexity might influence brain and behavior, one must consider the cor-

Habitat Complexity, Brain, and Behavior


125

Number of species

Number of species

25 20 15 10 5 p = 0.0012 0 –1.5 –1.0 –0.5

A

0 0.5 1.0 1.5 2.0 Rugosity

B

70 50 30 p = 0.0732

Fig. 1. Correlation between habitat com-

plexity and social variables for Tanganyikan cichlids. Species richness (top); abundance (middle) and Simpson’s Index of Diversity (bottom) as a function of rugosity and optical intensity. All correlations are significant, with the exception of abundance as a function of rugosity, which shows a trend. From Shumway et al. [2007], with permission.

Simpson’s index of diversity

C 1.0 0.8 0.6 0.4 0.2 0 –1.5 –1.0 –0.5

E

responding social changes. A spatially complex habitat, such as a coral reef or rock habitat, provides more niche space than a simple one [Caley and St. John, 1996]. Accordingly, habitat complexity has been correlated with social complexity, as measured by species richness, diversity or abundance. Figure 1 shows the relationship between habitat complexity and social variables among the Ectodini clade in Lake Tanganyika, as measured by rugosity (aka surface topography) and optical intensity [Shumway et al., 2007]. Both complexity measures were directly correlated with species richness (fig. 1A, B), and diversity (fig. 1E, F); optical intensity alone correlated with abundance (fig. 1D). A similar correlation between habitat complexity and species diversity has previously been shown for coral reef fishes [Luckhurst and Luckhurst, 1978] and intertidal rock pool fishes [Griffiths et al., 2006]. However, on coral reefs, correlations between 126

0 0.5 1.0 1.5 2.0 Rugosity


p = 0.0068 0 0.5 1.0 1.5 2.0 Rugosity

Number of individuals

90

10 –1.5 –1.0 –0.5

20 15 10 5 p = 0.0079 0 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 Optical intensity 110 90 70 50 30

p = 0.0125 10 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 Optical intensity D

Simpson’s index of diversity

Number of individuals

110

25

1.0 0.8 0.6 0.4 p = 0.0417

0.2 –1.5 –1.0 –0.5 0 0.5 1.0 1.5 2.0 2.5 Optical intensity F

abundance and complexity measures have varied, depending on the fish family, fish size, and location [Risk, 1972; Luckhurst and Luckhurst, 1978; Roberts and Ormond, 1987].

Habitat Complexity and the Brain

Habitat Complexity Correlates with Brain and Brain Structure Size Several papers on fishes have qualitatively associated ecological factors with size and/or other morphological differences in sensory structures and integration centers such as the telencephalon and cerebellum [elasmobranchs and pelagic teleosts: Lisney and Collin, 2006; chondrichthyans: Yopak et al., 2007; deep-sea fishes: Wagner, 2002; cyprinids: Kotrschal and Junger, 1988; Antarctic dragonShumway

Monogamous

Polygamous

abc

c

ab

a

abc

0.5

ab a

a

0.5

+

+

+

Ro

C

Fig. 2. Habitat complexity and the brain. Residual measures (mean 8 SE) of the telencephalon (A) and cerebellum (B) in 7 Ectodine

species showing their relationship to socio-ecological variables. The bars indicate preferred habitats [sand (white); intermediate (gray); rock (black)]. Polygamous species are to the left of the hatched vertical line, monogamous species to the right. Letters denote homogeneous subsets at p ! 0.05, as determined by ANOVA followed by Tukey post-hoc tests. Sample sizes for all volumetric measures are as follows: Xenotilapia ochrogenys (14); Enantiopus melanogenys (17); Xenotilapia bathyphila (3); Xenotilapia flavipinnis (9); Xenotilapia boulengeri (8); Xenotilapia spiloptera (6); and Asprotilapia leptura (11). C Color-coded covariance matrices showing the statistical strength and direction of the associations

A.

le

pt

pt

ur a

er a

ri ge

ilo sp X.

X.

bo

fla

ul

vi

pi

yp

X.

ba

en

nn

is

la hi

ys

th

og an el

+

+ +

+

* * *

*

* * ** *

+

+

ck Ru siz go e # s S o of Sp it y ci in ec al di ie or vid s ga u ni als za tio n

+

Ro

* * *

+


+

E.

+

*

Independent contrasts

* ** Telencephalon Cerebellum Midbrain +

*

Hypothalamus Brain Medulla Olfactory bulb


+

Residuals

Ro

*+

X.

X.

Fractions

m

oc

A.

hr

le

og

pt

en

en

ys

ur a

er a pt

X.

sp

ul

ilo

en

ge

ri

is nn

bo X.

vi

yp

pi

hi fla X.

th

an

ba

og

og

el m

B

E.

X.

en

en

ys

ys

la

0

hr oc X.

bc

ab

1.0

0

A

d

cd Cerebellum residuals

Telencephalon residuals

bc

1.0

Monogamous

Polygamous

1.5 d

between socio-ecological variables and neural measures. Results are shown for fractions (left), residuals (middle) and independent contrasts (right). The cladograms show the results of hierarchical clustering along both axes. The color gradient runs from dark blue (strong negative association: p ! 0.01) to dark red (strong positive association: p ! 0.01). Light colors indicate weaker relationships. Significant associations are indicated (* p ! 0.05; ** p ! 0.01), as are trends (+ p ! 0.10). For Social Organization (the only dichotomous variable), the results are based on t tests: red indicates that a brain region is larger in monogamous species, and blue that it is larger in polygamous species. Adapted from Pollen et al. [2007], with permission.

fishes: Eastman and Lannoo, 2003]. The size differences were interpreted, but not systematically tested, as being linked with habitat type, feeding specializations, or locomotory activity.

Only a few studies, however, have directly compared habitat complexity to the size of the brain or brain structures alone. Ratcliffe et al. [2006] showed that bats from more complex environments had significantly larger relative neocortices than those from simpler, open space en-



127

vironments. These results were obtained before independent contrasts; a trend was seen after independent contrasts. Safi and Dechmann [2005] used wing area as a surrogate for habitat complexity in bats, and found positive correlations with the size of the hippocampus and the inferior colliculus. Budeau and Verts [1986] demonstrated that brain size differed among chipmunks relative to the qualititatively determined complexity of their arboreal habitat. Bernard and Nurton [1993] showed that arboreal rodents have bigger brains than fossorial species. We compared relative brain and brain structure size, habitat complexity, and social organization among seven closely-related Ectodine cichlid species [Pollen et al., 2007]. The confound of foraging behavior was eliminated as much as possible; all species are benthic feeders. The sand-dwellers are benthic sifters, feeding on insect larvae and other invertebrates in the sand; the rock-dweller, A. leptura, feeds on algae on rock surfaces [Takamura, 1984; Konings, 1998]. Note that all of the species compared have an elongated fusiform body and similar head shape. We used three approaches to compare brain structures and habitat complexity. Volumes of brains and brain structures were obtained from gross brain measures, using an ellipsoid model ground-truthed with volumetric measures from sections. To explore allometric relationships and developmental constraints, we compared fractions, i.e., the absolute size of brain structures to the size of the rest of the brain [Finlay et al., 2001]. We also calculated residuals (from absolute size comparisons) and fractions. The results from fractions as well as independent contrasts (combining the results from both fractions and residuals) show a significant positive correlation with the telencephalon and habitat measures (fig. 2A, B). Brain size and cerebellar size are positively correlated with species number (which is correlated with habitat complexity); the medulla and olfactory bulb are negatively correlated with habitat measures. These results within the monophyletic Ectodini clade confirm, at a finer scale, the work by Huber et al. [1997], showing lake-specific differences across cichlid families in overall brain size and the size of various brain structures, including the telencephalon, relative to qualitative categories of physical environments. One of the most surprising findings of this work was that environmental and social factors differentially affect the brain, with environmental factors showing a broader effect on a range of brain structures compared to social factors. Although five out of seven of the brain measures showed a relationship with habitat measures, only two brain structures, the telencephalon and hypothalamus, are correlated with social factors (fig. 2C, right). 128


Habitat Complexity and Behavior

Linking Form to Function A number of papers have demonstrated correlations between flexibility and/or innovative feeding behavior and brain size or brain structure size. Farris [2008; Farris and Roberts, 2005] showed a correlation between generalist feeding and mushroom body size in scarab beetles; Sol, Lefebvre and colleagues [Sol et al., 2005; Lefebvre and Sol, 2008] demonstrated a relationship between innovative feeding behavior and brain size in birds; as did Reader and Laland [2002] for innovative tool use and brain size in primates. Ratcliffe et al. [2006] found a correlation with flexible foraging behavior and brain size in bats. Others have demonstrated correlations with presumed complexity of spatial tasks and brain structure size [e.g., bower-building and brain size: Madden, 2001; cerebellar size: Day et al., 2005; food-storing in birds and rodents and hippocampal size: Healy et al., 2005; Lucas et al., 2004; home-range and hippocampal size: Healy et al., 2005]. Similarly, mammalian studies found that the relative brain size of mammalian fruit and seed-eaters is larger than that of leaf-eaters, with the assumption made, but not tested, that the task of obtaining fruit and seeds is more spatially challenging than obtaining leaves, due to the spatial and temporal patchiness of fruits and seeds [Jolicoeur et al., 1984; Bernard and Nurton, 1993]. Visual Behavior Given the importance of vision to cichlids, we began by exploring the relationship between habitat complexity and visual behavior. In the Ectodini clade, there are only two (independent) transitions from a sand to a rock environment. How might the visual system need to evolve during this transition? There may be differences in light levels, contrast, and color sensitivity. There might need to be trade-offs between temporal resolution and spatial resolution. A rock-dweller could need enhanced object recognition and edge detection skills. Spatial memory might need to be enhanced, possibly at the loss of lateral line capabilities. Social tasks, too, may differ. As the number of species and number of individuals is increased in a more complex environment, there might be more complex predator-prey interactions as well as more complex competitive and cooperative social interactions [Caley and St. John, 1996; Gray et al., 2000]. To date, we have found differences in light levels, spatial resolution, and spatial memory between sand and rock-dwelling species and environments, as described below. Shumway

100

X. flavipinnis X. spiloptera A. leptura

Average time to landmarks (s)

Average response (%)

80

3,500

60

40

20

0 0.01

A

3,000 2,500 2,000 1,500 1,000 500 0

0.1

1

Spatial frequency (cycles/degree)

10

1

B

2

3

4

Landmarks

Fig. 3. Habitat complexity and behavior. A Effect of habitat complexity on visual acuity. Comparison of the average response at each spatial frequency. Error bars represent standard error. The dashed line represents the 50% response level. At this level, the rock-dweller A. leptura can detect a significantly higher spatial frequency than the two species from the less complex environ-

ments. (sand: X. flavipinnis; intermediate habitats: X. spiloptera). [From Dobberfuhl et al., 2005; adapted with permission]. B Comparison of spatial memory in a sequential maze. The rockdweller A. leptura is significantly faster (mean 8 SE) at reaching each of the four visual landmarks than the sand-dweller X. flavipinnis.

By selecting three Ectodine species and controlling for confounding social variables (social organization), we found that habitat complexity is associated with differences in visual acuity [Dobberfuhl et al., 2005]. Testing for optomotor/optokinetic responses and controlling for lens size, we compared three cichlids differing in habitat preference (fig. 3A), sand-dweller: X. flavipinnis, intermediate habitats: X. spiloptera, rock-dweller: A. leptura. All are monogamous/biparental species. At the 50% response point, the rock-dwelling species perceived a spatial frequency 4! higher and minimum separable angle (MSA) and one-third that of species from less complex habitats. We speculate that rock-dwelling favors adaptation for enhanced spatial resolution, used in spatial navigation.

sample task than non food-storing juncos [Hampton and Shettleworth, 1996]; and preferentially prefer spatial cues over local color and pattern cues [Brodbeck, 1994]. We compared the spatial abilities of the sand- and rock-dwelling species Xenotilapia flavipinnis and Asprotilapia leptura, respectively, in a sequential maze [Shumway et al., 2006]. Each maze contained four local landmarks and two extramaze cues. The rock-dwelling fish completed the task more quickly, and reached each landmark significantly faster than the sand-dwelling species (fig. 3B). Although the rate of learning appeared the same, the rock-dwelling fish also had fewer wrong turns. These behavioral results correspond with preliminary studies showing that the hippocampal homologue, Dl, is 20–36% larger in two rock-dwelling species compared to two sand-dwelling species [Shumway et al., 2004, 2006]. Dl is known to be involved in spatial memory in fishes [Salas et al., 2003]. Studies in mammals and birds found similar variation (10–40%) in hippocampal volume across species or sex [Jacobs et al., 1990; Reboreda et al., 1996; Safi and Dechmann, 2005].

Spatial Learning Only a few studies have directly tested the relationship between physical complexity, brain differences, and cognitive ability across closely-related species or populations. In fishes, Girvan and Braithwaite [1998] found differences in spatial learning ability and strategies between pond and river populations of sticklebacks. In birds, food-storing species place more importance on spatial cues than visual landmarks in a spatial memory task: non foodstorers do not [Clayton and Krebs, 1994]. Food-storing chickadees perform better in a spatial non-matching to Habitat Complexity, Brain, and Behavior

Plasticity Correlations are obviously not causal. To understand how adaptation occurs, we must tease apart the contribution of both environmental and genetic components of Brain Behav Evol 2008;72:123–134

129

brain-behavior differences. Many studies have shown experience-dependent neuronal plasticity in a variety of sensory systems for developing animals and adults [e.g., visual: Daw, 2004; somatosensory: Shumway et al., 1999, 2005]. Fish might be more plastic than other vertebrate classes. Neurogenesis occurs throughout life in numerous brain regions, including area Dl, other areas of the dorsal telencephalon, the ventral telencephalon, primary sensory areas, and the cerebellum [Zupanc et al., 2005; Adolf et al., 2006]. The widespread neurogenesis might well constitute part of the functional mechanism underlying experience-dependent plasticity in fishes, in addition to Hebbian synaptic changes. Kihslinger et al. [2006a, b] recently reported volumetric differences in the salmon telencephalon hatchery-reared in simple or complex environmental conditions. Lema et al. [2005] found environmental influences on the proliferation rate of neurogenesis in salmon. Cichlids show considerable phenotypic plasticity, with plasticity having been demonstrated for visual and social behavior, neural structure/function, and feeding morphology [Liem and Osse, 1975; Meyer, 1990, 1993; Insel and Fernald, 2004]. We explored the role of environmental manipulations on visual behavior by raising the sand-dweller X. flavipinnis fry in both a simple (sand) and enriched (rock) environment, under normal social conditions (in a school), and testing visual acuity at various developmental stages. Visual acuity has previously been shown to be experience-dependent in both fishes [Wahl et al., 1993] and mammals [Prusky et al., 2003]. Pilot experiments showed a significant difference in visual acuity after 12 months, as measured by minimal separable angle [Sandercock, pers. comm.]. We also tested the effect of rearing the fry in two different social conditions (isolated versus a school), under normal habitat conditions (sand). Social manipulations altered visual acuity as well, but the effects were seen only within a critical, and earlier, time period. Significant results were seen only at 4 months; by 12 months the difference between the two groups was negligible [Sandercock, pers. comm.].

Our next steps are three-fold: First, we want to identify the most significant ecological cues underlying habitat complexity’s effect in cichlid fishes. To date, we have little understanding of what visual features really matter to animals in a natural, complex environment, or how species-specific this might be [i.e., their ‘umwelt’ or subjective world; Uexküll, 1985]. To identify important visual cues, we will use the sequential maze to test the importance of different visual cues in spatial learning; spatial novelty experiments that compare the reaction of the sand- and rock-dweller to form, color, and position; and eye-tracking of a natural visual scene to determine what the fish considers important, as measured by its length of gaze. Secondly, we want to further explore the developmental influence of habitat complexity. By exploring both enrichment (e.g., raising sand species in a rock environment) and impoverishment (raising rock species in a sand environment), we can determine how much plasticity contributes to the species-specific differences found. Behavioral drive is thought to be an important evolutionary force underlying cichlid evolution [Seehausen, 2000]. Is it possible that the capacity for plasticity, then, distinguishes those clades or families that are most speciose? Cichlids are a superb system for exploring these types of integrative questions. Plasticity experiments could also tease apart the importance of the complexity of the visual scene versus the complexity of the spatial task. In other words, if the visual environment differs but the behavioral requirements do not, would brain changes be seen? Fish could be raised in a two-dimensional complex visual scene (rock) versus a 2-D sand scene, but with no complexity in 3-D space. Finally, we want to use modeling to predict and deepen our understanding of how complex brains and behaviors evolve. Research suggests that brains are constrained by information theoretic principles [Bialek, 1987; Laughlin et al., 1998], limiting the variety of imaginable architectures and opening the door toward not just descriptive, but predictive modeling [e.g., Linsker 1988].

Evolution in Context Future Directions ‘Out of the entire spectrum of physical stimuli that each particular habitat offers, only a fraction is actually used by any one species or individual, which must be proficient in many aspects, including finding prey, avoiding predators, or recognizing mates.’ Kotrschal, van Staaden and Huber, 1998

130


An animal does not exist or evolve in a vacuum (fig. 4). Although this review has focused on how a given animal is affected by its physical and social environment, the genetic, neural and behavioral evolution of an animal must be considered in context with how that animal both influences and is influenced by the environment and the Shumway

Environmental forces

Genes

Gene networks

Social forces

Neurons

Neuronal Behavior networks

Hormones

Fig. 4. Evolution in context. Neural and behavioral evolution of an animal must be considered in context with how that animal both influences and is influenced by the environment and the social world around them. The influence of both forces can be seen at the level of genes, neurons, and behavior. The animal’s behavior, in turn, influences its social milieu. Environmental conditions can also influence social organization. The converse is also true; species diversity and abundance can alter environmental conditions.

social world around it. For example, the animal’s behavior can influence the social milieu of conspecifics [Insel and Fernald, 2004]. Social dominance can even control the level of neurogenesis in the telencephalon in subordinate fish [Sørensen et al., 2007]. Environmental conditions (such as food scarcity) can also influence social organization, as shown for numerous species of mammals [Lott, 1991]. The converse is also true. Species diversity and abundance can alter environmental conditions, particularly for those animals considered ‘ecosystem engineers’ [Moore, 2006]. We must continue to view this broader picture and recognize that multiple selective pressures drive the evolution of a larger brain and enhanced cognition. The debate continues on whether environmental or social forces exert a greater influence on cortical and brain expansion [Dunbar and Shultz, 2007]. The two are not mutually exclusive. Reader and Laland [2002] showed that innovative tool use and social learning covaried in primates, and both behaviors correlated with increased brain size. The paper by Lefebvre and Sol in this volume [2008] attempts a synthesis of the two hypotheses, and argues that both social and physical variables might be driven by the same ecological factors, such as environments with unpredictable resources. In cichlid fishes, we find that habitat complexity is directly correlated with social variables, including species Habitat Complexity, Brain, and Behavior

richness, diversity and abundance. Although both forces shape the brain, social forces appear to influence visual behavior at an earlier stage than environmental ones. Taken together, the field data and the plasticity data suggest that species-specific neural effects of environmental complexity could be the consequence of the corresponding social correlates. However, environmental and social forces affect brains differently. Environmental forces exert a much broader effect on brain structures, suggesting either allometric expansion of the brain structures in concert with brain size and/or co-evolution of these structures. Co-evolution of the telencephalon and cerebellum has been shown in primates [MacLeod et al., 2003]; although the two structures appear to evolve at different rates [Finlay and Darlington, 1995] and can vary in their extent of developmental plasticity [Zupanc, 2001; Shumway et al., 2005]. In contrast, social forces, specifically abundance and social organization, are more selective, correlating only with telencephalic and hypothalamic expansion. Cichlids could also help us better understand the difference between these forces. Differences in the timing and scope of change among primary and integrative brain structures for environmental and social forces might help provide clues regarding the mechanism of change.

Conclusions

In this review, I have shown that habitat complexity correlates with both behavioral and brain changes in a number of species. The influence of the environment is both innate and experientially induced. The ecological change hypothesis, or the notion that the requirements for ecological problem-solving can drive brain expansion, is not dead. To advance our understanding of the mechanism by which habitat complexity affects brain and behavior will require the use of closely-related species, quantification of complexity, hypothesis testing restricting analysis to a single variable (e.g., foraging behavior, habitat complexity, spatial task complexity, social complexity) and path analyses to explore the order of importance of such variables. Plasticity experiments might also help resolve the order of timing and influence of socio-ecological variables on the brain. Finally, we will need new experimental paradigms exploring the cognitive and survival value of brain and brain structure changes both in the laboratory and in the wild.


131

Acknowledgements This review is dedicated to the memory of Ted Bullock, whose relentless push to understand brain and behavioral complexity was a great inspiration. We thank our principal collaborator, Hans Hofmann, who collaborated on the quantification of habitat complexity and the relationship between ecological and social variables, and whose laboratory led the comparison of gross brain volumetric analysis. This work is the culmination of many students and technicians in the Shumway lab, with special thanks to Adam Dobberfuhl, Justin Scace, Jeremy Ullmann, Margaret Bell, Liz Higgins, and Maria Sandercock. We thank Robert Wakafumbe and Mathias Msaferi Igulu (Tanzania Fisheries Research Institute, TAFIRI), Alex Pollen, Suzy Renn, Adam Dobberfuhl,

George and Wilbrod Kazumbe, and Sarah Bahan for assistance in the field. Many thanks to TAFIRI, the Tanzania Commission on Science and Technology (COSTECH) and Alfeo Nikundiwe (University of Dar Es Salaam) for their kind support of our research. Finally, we extend sincere thanks to Andy Cohen and Ellinor Michel from the Nyanza Project as well as Saskia Marijnissen and the Vaitha brothers for providing materials and support to our field work. We are grateful to the anonymous reviewers for comments on the manuscript. This research was supported by and NSF grants IBN-02180005 to Caroly Shumway (CAS) and IBN021795 to Hans Hofmann (HAH), a German-American ResearchNetworking Program grant to CAS and HAH, the New England Aquarium to CAS, and the Bauer Center for Genomics Research to HAH.

References 1 Adolf B, Chapouton P, Lam CS, Topp S, Tannhäuser B, Strähle U, Götz M, Bally-Cuif L (2006) Conserved and acquired features of adult neurogenesis in the zebrafish telencephalon. Dev Biol 295: 278–293. 2 Amores A, Force A, Yan Y-L, Joly L, Amemiya C, Fritz A, Ho RK, Langeland J, Prince V, Wang Y-L, Westerfield M, Ekker M, Postlethwait JH (1998) Zebrafish hox clusters and vertebrate genome evolution. Science 282:1711–1714. 3 Barton RA, Harvey PH (2000) Mosaic evolution of brain structure in mammals. Nature 405:1055–1058. 4 Bauchot R, Bauchot ML, Platel R, Ridet JM (1977) The brains of Hawaiian tropical fishes: brain size and evolution. Copeia 1977: 42–46. 5 Bauchot R, Ridet JM, Bauchot ML (1979) Encephalization and evolutive level in aquatic vertebrates. Vie Mileu (A Biol. Mar., B Oceanogr.) 28–29:253–265. 6 Bauchot R, Randall JE, Ridet J-M, Bauchot M-L (1989) Encephalization in tropical teleost fishes and comparison with their mode of life. J Hirnforsch 30:645–669. 7 Bernard RTF, Nurton J (1993) Ecological correlates of relative brain size in some South African rodents. S Afr J Zool 28:95–98. 8 Bialek W (1987) Physical limits to sensation and perception. Ann Rev Biophys Chem 16: 455–478. 9 Budeau DA, Verts BJ (1986) Relative brain size and structural complexity of habitats of chipmunks. J Mamm 67:579–581. 10 Brodbeck DR (1994) Memory for spatial and local cues: a comparison of a storing and a nonstoring species. Anim Learning Behav 22:119–133. 11 Brown C, Warburton K (1997) Predator recognition and anti-predator responses in the rainbowfish Melanotaenia eachamensis. Behav Ecol Sociobiol 41: 61–68.

132

12 Bshary R, Wickler W, Fricke H (2002) Fish cognition: a primate’s eye view. Anim Cogn 5:1–13. 13 Bullock TH (2002) Grades in neural complexity: How large is the span? Integr Comp Biol 42: 757–761. 14 Caley MJ, St. John J (1996) Refuge availability structures assemblages of tropical reef fishes. J Anim Ecol 65: 414–428. 15 Carleton KL, Kocher TD (2001) Cone opsin genes of African cichlid fishes: tuning spectral sensitivity by differential gene expression. Mol Biol Evol 18: 1540–1550. 16 Clayton NS, Krebs JR (1994) Memory for spatial and object-specific cues in food-storing and non-storing birds. J Comp Psychol 174:371–379. 17 Daw NW (2004) Mechanisms of plasticity in the visual cortex. In: The Visual Neurosciences (Chalupa LM, Werner JS, eds), pp 126– 145. Cambridge MA: MIT Press. 18 Day LB, Wetcott DA, Olster DH (2005) Evolution of bower complexity and cerebellum size in bowerbirds. Brain Behav Evol 66:62– 72. 19 Dobberfuhl AP, Ullmann J, Shumway CA (2005) Visual acuity, environmental complexity, and social organization in African cichlid fishes. Behav Neurosci 119: 1648– 1655. 20 Dunbar RIM, Shultz S (2007) Evolution in the social brain. Science 317:1344–1347. 21 Eastman JT, Lannoo MJ (2003) Diversification of brain and sense organ morphology in Antarctic dragonfishes (Perciformes: Notothenioidei: Bathydraconidae). J Morphol 258:130–150. 22 Farris SF, Roberts NS (2005) Coevolution of generalist feeding ecologies and gyrencephalic mushroom bodies in insects. Proc Natl Acad Sci USA 102:17394–18399.


23 Farris S (2008) How social and ecological forces affect insect mushroom bodies. Brain Behav Evol 72:106–122. 24 Felsenstein J (1985) Phylogenies and the comparative method. Am Nat 125:1–15. 25 Fernald RD (1982) Retinal projections in the African cichlid fish, Haplochromis burtoni. J Comp Neurol 206:379–389. 26 Finlay BL, Darlington RB (1995) Linked regularities in the development and evolution of mammalian brains. Science 268:1578–1584. 27 Finlay BL, Darlington RB, Nicastro N (2001) Developmental structure in brain evolution. Behav Brain Sci 24:263–308. 28 Girvan JR, Braithwaite VA (1998) Population differences in spatial learning in threespined sticklebacks. Proc R Soc Lond B 265: 913–918. 29 Gray SJ, Jensen SP, Hurst JL (2000) Structural complexity of territories: preference, use of space and defense in commensal house mice, Mus domesticus. Anim Behav 60:765–772. 30 Griffiths SP, Davis AR, West RJ (2006) Role of habitat complexity in structuring temperate rockpool ichthyofaunas. Mar Ecol Prog Ser 313:227–239. 31 Gould SJ, Lewontin RC (1979) The spandrels of San Marco and the panglossian paradigm: a critique of the adaptationist programme. Proc R Soc Lond B 205:581–598. 32 Hampton RR, Shettleworth SJ (1996) Hippocampus and memory in a food-storing and in a non-storing bird species. Behav Neurosci 110:946–964. 33 Harvey PH, Purvis A (1991) Comparative methods for explaining adaptations. Nature 351:619–624. 34 Healy SD, de Kort SR, Clayton NS (2005) The hippocampus, spatial memory and food hoarding: a puzzle revisited. Trends Ecol Evol 20:17–22.

Shumway

35 Huber R, van Staaden MJ, Kaufman LS, Liem KF (1997) Microhabitat use, trophic patterns, and the evolution of brain structure in African cichlids. Brain Behav Evol 50: 167– 182. 36 Imre I, Grant JWA, Keeley ER (2002) The effect of visual isolation on territory size and population density of juvenile rainbow trout (Oncorhynchus mykiss). Can J Fish Aqu Sci 59:303–309. 37 Insel TR, Fernald RD (2004) How the brain processes social information: searching for the social brain. Ann Rev Neurosci 27: 697– 722. 38 Jacobs LF, Gaulin SJC, Sherry DF, Hoffman GE (1990) Evolution of spatial cognition: Sex-specific patterns of spatial behavior predict hippocampal size. Proc Natl Acad Sci USA 87:6349–6352. 39 Johnson DW (2007) Habitat complexity modifies post-settlement mortality and recruitment dynamics of a marine fish. Ecology 88:1716–1725. 40 Jolicoeur P, Pirlot P, Baron G, Stephan H (1984) Brain structure and correlation patterns in Insectivora, Chiroptera, and Primates. Syst Zool 33:14–29. 41 Keenleyside M (1991) Parental care. In: Cichlid Fishes: Behavior, Ecology and Evolution (Keenleyside M, ed) pp 191–208. London: Chapman and Hall. 42 Kihslinger RL, Nevitt GA (2006a) Early rearing environment impacts cerebellar growth in juvenile salmon. J Exp Biol 209: 504–509. 43 Kihslinger RL, Lema SC, Nevitt GA (2006b) Environmental rearing conditions produce forebrain differences in wild Chinook salmon Oncorhynchus tshawytscha. Comp Biochem Phys A 145:145–151. 44 Koblmüller S, Salzburger W, Sturmbauer C (2004) Evolutionary relationships in the sand-dwelling cichlid lineage of Lake Tanganyika suggest multiple colonization of rocky habitats and convergent origin of biparental mouthbrooding. J Mol Evol 58: 79– 96. 45 Konings A (1998) Tanganyika Cichlids in their Natural Habitat. Miami, FL: Cichlid Press. 46 Kotrschal K, Junger J (1988) Patterns of brain morphology in mid-European Cyprinidae (Pisces, Teleostei): A quantitative histological study. J Hirnforsch 3:341–352. 47 Kotrschal K, Van Staaden MJ, Huber R (1998) Fish brains: evolution and environmental relationships. Rev Fish Biol Fisheries 8: 373– 408. 48 Laughlin SB, de Ruyter van Steveninck RR, Anderson JC (1998) The metabolic cost of neural information. Nat Neurosci 1:36–41. 49 Lefebvre L, Sol D (2008) Brains, lifestyles, and cognition: are there general trends? Brain Behav Evol 72:135–144.


50 Lema SC, Hodges MJ, Marchetti MP, Nevitt GA (2005) Proliferation zones in the salmon telencephalon and evidence for environmental influence on proliferation rate. Comp Biochem Physiol A 141:327–335. 51 Liem KF, Osse JWM (1975) Biological versatility, evolution, and food resource exploitation in African cichlid fishes. Am Zool 15: 427–454. 52 Lingo ME, Szedlmayer ST (2006) The influence of habitat complexity on reef fish communities in the northeastern Gulf of Mexico. Env Biol Fishes 76: 71–80. 53 Linsker R (1988) Self-organization in a perceptual network. Computer 21:105–117. 54 Lisney TJ, Collin SP (2006) Brain morphology in large pelagic fishes: a comparison between sharks and teleosts. J Fish Biol 68:532– 554. 55 Lott DF (1991) Intraspecific Variation in the Social Systems of Wild Vertebrates. Cambridge, UK: Cambridge University Press. 56 Lucas JR, Brodin A, Selvino RDK, Clayton NS (2004) Does hippocampal size correlate with the degree of caching specialization? Proc R Soc Lond B 271:2423–2429. 57 Luckhurst BE, Luckhurst K (1978) Analysis of the influence of substrate variables on coral reef fish communities. Mar Biol 49: 317– 323. 58 Lukasik P, Radwan J, Tomkins JL (2006) Structural complexity of the environment affects the survival of alternative male reproductive tactics. Evolution 60:399–403. 59 MacLeod CE, Zilles K, Schleicher A, Rilling JK, Gibson KR (2003) Expansion of the neocerebellum in Hominoidea. J Hum Evol 44: 401–429. 60 Madden J (2001) Sex, bowers and brains. Proc R Soc Lond B 268:833–838. 61 McCormick M (1994) Comparison of field methodologies for measuring surface topography and their associations with a tropical reef fish assemblage. Mar Ecol Prog Ser 112: 87–96. 62 Meyer A (1993) Phylogenetic relationships and evolutionary processes in East African cichlid fishes. Ecol Evol 8:279–284. 63 Meyer A (1990) Ecological and evolutionary consequences of the trophic polymorphism in Cichlasoma sitrinellum (Pisces: Cichlidae). Biol J Linn Soc 39: 279–300. 64 Moore JW (2006) Animal ecosystem engineers in streams. BioSci 56: 237–246. 65 Ohno S (1970) Evolution by Gene Duplication. Heidelberg: Springer-Verlag. 66 Pollen AA, Dobberfuhl AP, Igulu MM, Scace J, Renn SCP, Shumway CA, Hofmann HA (2007) Environmental complexity and social organization sculpt the brain in Lake Tanganyikan cichlid fish. Brain Behav Evol 70: 21–39. 67 Prusky GT, Douglas RM (2003) Developmental plasticity of mouse visual acuity. Eur J Neurosci 17:167–173.

68 Ratcliff JM, Fenton MB, Shettleworth SJ (2006) Behavioral flexibility positively correlated with relative brain volume in predatory rats. Brain Behav Evol 67:165–176. 69 Reader SM, Laland KN (2002) Social intelligence, innovation, and enhanced brain size in primates. Proc Natl Acad Sci USA 99: 4436–4441. 70 Reboreda JC, Clayton NS, Kacelnik A (1996) Species and sex differences in hippocampal size in parasitic and non-parasitic cowbirds. NeuroReport 7:505-508. 71 Rheindt FE, Grafe TU, Abouheif E (2004) Rapidly evolving traits and the comparative method: how important is testing for phylogenetic signal? Evol Ecol Res 6:377–396. 72 Rilov G, Figueira WF, Lyman SJ, Crowder LB (2006) Complex habitats may not always benefit prey: linking visual field with reef fish behavior and distribution. Mar Ecol Prog Ser 329:225–238. 73 Risk MJ (1972) Fish diversity on a coral reef in the Virgin Islands. Atoll Res Bull 193: 1– 6. 74 Roberts CM, Ormond RFG (1987) Habitat complexity and coral reef fish diversity and abundance on red sea fringing reefs. Mar Ecol Prog Ser 41:1–8. 75 Safi K, Dechmann DK (2005) Adaptation of brain regions to habitat complexity: a comparative analysis in bats (Chiroptera). Proc R Soc Lond B 272:179–186. 76 Salas C, Broglio C, Rodriguez F (2003) Evolution of forebrain and spatial cognition in vertebrates: conservation across diversity. Brain Behav Evol 62:72–82. 77 Schmidt-Nielsen K (1984) Scaling: Why Is Animal Size so Important? Cambridge, MA: Cambridge University Press. 78 Seehausen O (2000) Explosive speciation rates and unusual species richness in haplochromine cichlid fishes: Effects of sexual selection. Adv Ecolog Res 31:237–274. 79 Shumway CA, Morissette J, Gruen P, Bower JM (1999) Plasticity in cerebellar tactile maps in adult rats. J Comp Neurol 413:583–592. 80 Shumway CA, Dobberfuhl AP, Scace JG (2004) Habitat and social differences affect visual abilities and forebrain regions in cichlids. Int Congr Neuroeth #187, Denmark. 81 Shumway CA, Morrisette J, Bower JM (2005) Mechanisms underlying reorganization of fractured tactile cerebellar maps following deafferentation in adult rats. J Neurophysiol 94:2630–2643. 82 Shumway CA, Scace JG, Bell M (2006) Cichlid fish differing in habitat complexity show spatial ability differences in a maze. Soc Neurosci Abstr #817.10. San Diego, CA. 83 Shumway CA, Hofmann HA, Dobberfuhl AP (2007) Quantifying habitat complexity in aquatic ecosystems. Freshwater Biol 52: 1065–1076.


133

84 Sol D, Duncan RP, Blackburn TM, Cassey P, Lefebvre L (2005) Big brains, enhanced cognition, and response of birds to novel environments. Proc Natl Acad Sci USA 102: 5460–5465. 85 Sol D, Székely T, Liker A, Lefebvre L (2007) Big-brained birds survive better in nature. Proc R Soc Lond B 274:763–769. 86 Sørensen C, Øverli Ø, Summers CH, Nilsson GE (2007) Social regulation of neurogenesis in teleosts. Brain Behav Evol 70:239–246. 87 Striedter G (2005) Principles of Brain Evolution. Sunderland, MA: Sinauer.

134

88 Takamura K (1984) Interspecific relationships of aufwuch-eating fishes in Lake Tanganyika. Env Biol Fishes 4: 225–241. 89 Uexküll J von (1985) Environment [Umwelt] and inner world of animals. In: Foundations of Comparative Ethology (Burghardt GM, ed), pp 222–245. New York: Van Nostrand Reinhold. 90 Wagner HJ (2002) Sensory brain areas in three families of deep-sea fish (slickheads, eels and grenadiers): comparison of mesopelagic and demersal fishes. Mar Biol 141:807– 817. 91 Wahl CM, Mills EL, Mcfarland WM, DeGisi JS (1993) Ontogenetic changes in prey selection and visual acuity of the Yellow Perch, Perca flavescens. Can J Fish Aquat Sci 50: 743–749.


92 Yopak KE, Lisney TJ, Collin SP, Montgomery JC (2007) Variation in brain organization and cerebellar foliation in Chondrichthyans: sharks and holocephalans. Brain Behav Evol 69:280–300. 93 Zupanc GKH (2001) A comparative approach towards the understanding of adult neurogenesis. Brain Behav Evol 58:246–249. 94 Zupanc GKH, Hinsch K, Gage FH (2005) Proliferation, migration, neuronal differentiation, and long-term survival of new cells in the adult zebrafish brain. J Comp Neurol 488:290–319.

Shumway



Brains, Lifestyles and Cognition: Are There General Trends? Louis Lefebvre a Daniel Sol b a

Department of Biology, McGill University, Montréal, Qué., Canada; b CREAF (Research Center for Forestry Applications), Autonomous University of Barcelona, Barcelona, Catalonia, Spain

Key Words Brain ⴢ Evolution ⴢ Neuroecology ⴢ Cognition

Abstract Comparative and experimental approaches to cognition in different animal taxa suggest some degree of convergent evolution. Similar cognitive trends associated with similar lifestyles (sociality, generalism, new habitats) are seen in taxa that are phylogenetically distant and possess remarkably different brains. Many cognitive measures show positive intercorrelations at the inter-individual and inter-taxon level, suggesting some degree of general intelligence. Ecological principles like the unpredictability of resources in space and time may drive different types of cognition (e.g., social and non-social) in the same direction. Taxa that rank high on comparative counts of cognition in the field are usually the ones that succeed well in experimental tests, with the exception of avian imitation. From apes to birds, fish and beetles, a few common principles appear to have influenced the evolution of brains and cognition in widely divergent taxa. Copyright © 2008 S. Karger AG, Basel

Introduction

Why do some animal taxa have larger brains than others? Body size allometry and grade shifts account for most of the variance in brain size, but a substantial pro© 2008 S. Karger AG, Basel 0006–8977/08/0722–0135$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com


portion remains when these variables are factored out. What is behind this remaining variance? Because the brain is an organ that processes, stores and integrates sensory, motor and information, the most obvious hypothesis is that there are cognitive advantages in affording larger brains [Jerison, 1973]. Important advances have been made since Jerison’s pioneering work, but several unresolved issues still stimulate debates among contemporary researchers. Are cognitive activities selected as independent modules or can they be part of more general processes that cause some animals to consistently outrank others in several cognitive measures? What are the evolutionary forces that select for larger brains? Are there convergent principles that govern the evolution of brain size in different taxa? Do whole brain size differences, as opposed to finer measures, mean anything in terms of cognitive performance? Several research strategies have been used to answer these questions. Some use principles derived from work on humans, others focus on specialized behaviors (e.g., food storing) seen only in a few non-human taxa. In this paper, we review an approach initially developed with birds and based on field observations of innovative behaviors. We first discuss a series of operational measures of cognition, and ask whether variation in the measures suggests correlated or independent evolution. Next, we test whether these cognitive measures co-vary with the size of neural structures. Finally, we discuss some ecological contexts in which the cognitive measures associated with size of neural structures might be Louis Lefebvre Department of Biology, McGill University 1205 avenue Docteur Penfield Montréal, Qué., H3A 1B1 (Canada) Tel. +1 514 398 6457, Fax +1 514 398 5069, E-Mail [email protected]

selected and ask whether the same contexts might explain the evolution of large brains in independent lineages.

Operational Measures of Cognition

Recent comparative work on correlates of brain size has focused on continuous operational measures of cognition. Although many of the pioneering studies of the 1980’s used categorical life history and ecological variables [e.g., diet, Harvey et al., 1980; development mode, Bennett and Harvey, 1985; Iwaniuk and Nelson, 2003], the focus since the late 1990’s has been on frequency counts of naturally-occurring behaviors in the wild. These measures include the propensity for social deception in primates [Byrne, 1993; Byrne and Corp, 2004], innovation in birds [Lefebvre et al., 1997a] and primates [Reader and Laland, 2002], social learning in primates [Reader and Laland, 2002] and tool use in birds [Lefebvre et al., 2002] and primates [Reader and Laland, 2002]. Given that animal cognition is extremely difficult to define, the strict operational nature of these comparative indices is an advantage; the indices simply measure variation on a continuum, without specifying thresholds beyond which behaviors can be considered ‘intelligent’ or ‘complex’ or which specific cognitive mechanisms are involved [Giraldeau et al., 2007]. Moreover, because the counts are based on field observations, these measures of cognition are biologically meaningful and comparable among species or higher taxonomic levels. Falsifiable predictions can then be made on the indices, for example that they are all positively correlated or that the size of certain brain areas is larger in taxa that have higher counts of innovation, tool use, deception and/or social learning. Because the counts are based on case reports, these measures of cognition are vulnerable to a number of biases [reviewed in Lefebvre et al., 2004]. The importance of these biases needs to be rigorously evaluated before using the cognitive counts in analyses. In birds, for example, nine possible biases have been checked and found not to account for the positive correlation between innovation rate and residual brain, forebrain or mesopallium size [Nicolakakis and Lefebvre, 2000; Lefebvre et al., 2001]. The most important bias is that associated with the effort devoted by scientists to observing different species, but this effect can be mitigated by including a measure of research effort in multivariate models [Reader and Laland, 2002]. 136


Because the above operational measures are based on anecdotal reports, they also need to be complemented by experiments. In general, frequency count and experimental approaches yield similar results. Taxa that rank high on innovation, tool use, deception and social learning counts tend to be those that succeed in experimental tests of sophisticated cognitive processes. For example, an anecdotal report on use of tools in New Caledonian crows [Orenstein, 1972] has been followed by extensive field work [Hunt, 1996; Hunt and Gray, 2004], as well as laboratory experiments [Weir et al., 2002; Kenward et al., 2005; Taylor et al., 2007] that have yielded detailed understanding of the most complex form of tool manufacture known in non-human animals. If New Caledonian crows are unique in their tool use, they are also part of the most innovative genus of the entire class Aves [Lefebvre et al., 1997a], as well as the genus that has the highest tool use count [Lefebvre et al., 2002]. In the New World, Quiscalus is the second most innovative passerine genus after Corvus. In line with this, Q. lugubris is, among the opportunistic birds of Barbados, the species that does best at innovative problem-solving [Webster and Lefebvre, 2001]; in field and aviary experiments, Q. lugubris also shows imitation [Lefebvre et al., 1997b], backward conditioning of alarm calls [Griffin and Galef, 2005], as well as flexible use of tool-like processing of dry food [Morand-Ferron et al., 2004, 2006, 2007a, b]. In primates, Pan troglodytes, the species that tops the comparative data bases on deception [Byrne and Whiten, 1988], innovation, tool use and social learning rate [Reader and Laland, 2002] is also the one that shows the most sophisticated behavior in experiments on cultural transmission [Bonnie et al., 2007], imitation [Whiten et al., 1996], and cooperation [Melis et al., 2006]. In the New World, the genus Cebus ranks high on all comparative counts [Reader and Laland, 2002], and uses rocks to dig up tubers [Moura and Lee, 2004] and crack open nuts [Fragaszy et al., 2004]; like chimpanzees [Lefebvre, 1982; Brosnan et al., 2008], they will barter with humans in captivity [Drapier et al., 2005]. Interspecifc comparisons on birds and primates show positive correlations between reversal learning, an experimental measure taken in captivity, and innovation rate, a frequency count taken in the field [Lefebvre et al., 2004]. For the moment, the major discrepancy between the experimental and count approaches to cognition seems to be in the field of avian imitation. In primates, the high rank of apes on frequency count scales is consistent with the fact that, in experiments, they show more cognitively demanding forms of social learning than do monkeys [e.g., imitation: Visalberghi and Fragaszy, 1990]. In birds, however, imitation Lefebvre /Sol

has been demonstrated in species, e.g., pigeons and quail, that have both small brains and rank very low on innovation and tool use counts [Zentall, 2004].

Brain Areas Involved in Cognition

One of the questions that has long worried psychologists is the modular or general-process nature of cognition. One purely empirical way of addressing this issue is to see if performance on different cognitive tasks across individuals or taxa shows positive, negative or zero correlations. In the first case, this means that there might be a common general process behind the positively correlated activities, or that the modular processes controlling each activity are selected together. Negative correlations would suggest trade-offs, such that enhancement of one type of cognition and/or memory is costly and requires a decrease in other types of cognition [Sherry and Schacter, 1987]. A zero correlation would imply that the systems are independent. A surprising number of cognitive measures show positive correlations. For example, Reader and Laland [2002] showed that social learning, innovation and tool use counts are all positively correlated across primate species. Deaner et al. [2006] found a similar trend of positive correlations over primate genera for up to 30 different cognitive tests conducted in captivity. In birds, innovation rate is positively correlated with tool use [Lefebvre et al., 2002] and with performance on tests of problem-solving and reversal learning [Lefebvre et al., 2004]. For the moment, the negative relationship between food-storing and innovation in corvids and parids [Lefebvre and Bolhuis, 2003] stands out as an exception, even though it is based on a small sample. At the individual level, Bouchard et al. [2007] have found that latency of social learning is positively correlated with innovative problem-solving in pigeons, even after the common correlate of neophobia has been removed from the two measures. Positive correlations across individuals are a common feature of cognitive test batteries run on humans as well as outbred rodents [Plomin, 2001]. All these lines of evidence suggest that some of the variation in cognition between individuals and species reflects a general process. Finally, there seems to be strong agreement in the ranking of taxa in field and laboratory measures of cognition. This is reassuring, as field-based measures can be criticized for their lack of controls and lab-based measures for their lack of ecological relevance, as well as the confounding effects of differential response to captivity and testing by humans.

In order to better understand the predicted inter-taxon relationship between cognitive measures and neural center efficiency, it would be useful to know which neural measure we should focus on: size of whole brains or of restricted areas? Volumes, neuron numbers or receptor densities? Recent work on humans, birds, rodents and non-human primates suggests possible directions. When neuroscientists look at the control centers that are involved in different cognitive tasks, the result they most often come up with is that each task involves a network of localized centers distributed in many parts of the brain. Lesion and neuronal recording techniques are useful in identifying precise areas crucial to the correct functioning of a given cognitive system, but they are incapable of mapping the whole set of areas that are active. In contrast, techniques like MRI, immediate early genes and receptor site mapping can inform us about how broad or localized should be our search for brain correlates of cognition. The techniques routinely compare neural activation during a particular cognitive task [e.g., imitation of observed movement, Iacoboni et al., 1999] to that of the closest control (e.g., movement or observation only), and thus underestimate the total number of brain areas active during cognitive processing. Bearing this in mind, the most frequent result of brain imaging studies is that a number of different localized centers distributed all over the brain are involved in each cognitive activity. For example, a meta-analysis of 64 MRI studies in humans reveals a very broad distribution of areas active in different types of human tool use situations [Lewis, 2006]. When the mapping of the areas is restricted to those that are reported in at least four of the 64 studies, eight areas in the cortex, plus areas in the cerebellum and basal ganglia appear to be involved. During macaque tool use, 10 areas show MRI activity, from different parts of the right and left cerebellum to parts of the basal ganglia and cortical areas such as the precuneus and inferior temporal cortex [Obayashi et al., 2001]. During cooperative interactions in a prisoner’s dilemma game, at least five cortical and subcortical areas are active in humans [Rilling et al., 2002]. Unfortunately, brain imaging studies are usually conducted on single species and almost never compare animals that show lifestyle differences likely to affect cognition. A remarkable exception is the work of Goodson and co-authors on the neural basis of avian sociality. The work from this group uses state-of-the-art neuroscience techniques to, among other things, map the receptor density of neuropeptides involved in sociality; the sample of

Brains, Lifestyles and Cognition


Variation among Animals

137

species is remarkably broad for this kind of study (five), and the authors distinguish cases of independent and phylogenetically-correlated evolution. The studies identify neurohormone receptor site differences that correlate with sociality differences in 13 different brain centers ranging from the sub-pallial septum to the stria terminalis, the hypothalamus and the hippocampus [see table 1 in Goodson et al., 2006]. In several other brain areas (e.g., the medial nidopallium, the ventral pallidum), high densities of neuropeptide binding sites are also found, even though no interspecific differences seem to be linked to significant differences in their density.

Enhanced Performance of Enlarged Brain Areas

Brains vary in whole size, size of their parts, density of neurons and glia, as well as density of neurotransmitter receptors. Each of these features has been suggested to reflect, with different levels of accuracy, the performance of the brain in cognitive tasks. The size of comparative databases on each of these measures is inversely proportional to the difficulty of obtaining data on them. Corrected endocast estimates of whole brain size from museum specimens are the easiest data to gather, whereas binding densities and neuron numbers are the most workintensive. Gross, easy-to-obtain measures are usually assumed to be temporary ‘best of a bad job’ estimates of finer, hard-to-obtain data. This assumption can easily be tested by calculating the proportion of variance in the fine structure that can be predicted from the gross measure. In birds, for example, residual size of the whole brain (i.e., the residuals of log-brain size regressed against log-body size) predicts 95% of the variance in residual size of the mesopallium and nidopalllium, while residual size of the telencephalon predicts 99% [Timmermans et al., 2000]. Trends in primates are similar: 97% of the variance in residual size of the cortex is predicted by residual brain size, while 99% is predicted by residual telencephalon size [data from Stephan et al., 1981]. It is only when measures such as executive brain ratio (volume of pallial areas/brainstem) are used that the similarity between primates and birds breaks down: interspecific variance in executive ratio is well predicted by absolute size of the brain (89%), but not by residual size (21%). Herculano-Houzel and colleagues have recently contributed a crucial methodological test of assumptions by estimating the relationship between brain size and neuron numbers across six species of rodents [HerculanoHouzel et al., 2006; Herculano-Houzel, 2007] and six spe138


cies of primates [Herculano-Houzel et al., 2007]. They found a very tight relationship between residuals of neuron numbers and brain size, after taking out the common effects of body size. Interestingly, the slope of this relationship differs between primates and rodents, with neuron numbers increasing at a faster rate with increasing brain size in the former than in the latter. This finding is consistent with the fact that primates have better cognitive abilities than do rodents of a similar size. It would be interesting if Goodson’s comparative research program on sociality and fine level neuronal measures could address the assumptions of brain size research, for example by looking at the relationship between neuropeptide receptor density and size of structures involved in sociality such as the amygdala, the septal complex, the hypothalamus and the hippocampus.

Relationship between the Size of Neural Centers and Cognitive Measures

Selection favoring enhanced performance on a particular cognitive process is expected to drive increases in the brain areas involved. We can thus predict an association among the neuron numbers, binding density or size of a brain region and the performance on the cognitive tasks the region controls. Most of the neuroecological tests on brain evolution deal with lifestyles that are assumed to require enhanced cognitive performance, not the cognitive processes themselves. For example, the robust finding that relative size of the hippocampus is positively associated with interspecies [Lucas et al., 2004] and inter-family [Sherry et al., 1989; Krebs et al., 1989] differences in food-storing assumes a link between enhanced needs for remembering the position of stored food and enhanced neuronal performance of a larger hippocampus. Overall, comparative tests of spatial memory are in the direction predicted by the assumption, but results are not as robust and clear as those linking storing itself with hippocampal size. The literature linking song repertoire size and relative size of song production nuclei (e.g., HVC and RA) in oscines is on firmer ground. Not only is the association between repertoire and nucleus size clearly established at the inter- [DeVoogd et al., 1993] and intra-specific [Garamszegi and Eens, 2004] levels, it also appears to characterize gender differences: species in which only males sing have dimorphic song nuclei, whereas species where both sexes sing do not [Brenowitz, 1997]. In addition, evidence is accumulating that song learning coincides with neuronal growth in HVC [Zeng Lefebvre /Sol

et al., 2007], that HVC and RA control different levels of song organization [Yu and Margoliash, 1996], that larger learned repertoires are reflected in increased dendritic spine density in HVC [Airey et al., 2000a], that females select for males with both a larger HVC and a larger repertoire [Airey and DeVoogd, 2000], that HVC size and song complexity predict male quality [Pfaff et al., 2007] and that both repertoire size and HVC size are inherited [Airey et al., 2000b]. Eventually, the impressive advances that characterize the song learning literature need to be imitated in other areas of neuroecology. Research on general (as opposed to adaptively specialized) cognitive abilities has also concentrated on the link between brains and lifestyles, assuming that more complex lifestyles demand more complex cognition that demand in turn larger brains. Diet [Harvey et al., 1980; Reader and MacDonald, 2003; Ratcliffe et al., 2006] and sociality [Dunbar, 1998] have been the main lifestyle variables studied. Direct tests on brains and cognition are recent. Interspecific differences in social deception in primates are positively correlated with cortex size [Byrne, 1993; Byrne and Corp, 2004], as are differences in rates of social learning, innovation and tool use [Reader and Laland, 2002]. A similar relationship has been found for birds between residual size of pallial areas and rate of innovation [Timmermans et al., 2000] and tool use [Lefebvre et al., 2002]. Paradoxically, cognition is often unrelated with lifestyles that co-vary with brain size. Social deception rate is not significantly associated with group size [Byrne and Corp, 2004], despite the fact that both variables are associated with cortex size. More work of the type conducted on song is clearly needed on both specialized (e.g., spatial) and general cognition to understand the three-way link between lifestyle, cognitive abilities and neural control areas.

Size: Whole Brains or Brain Areas? Absolute or Allometrically Corrected?

The brain is not a homogeneous organ, but a network of pathways and centers which can vary in the degree of specialization. Its size is also strongly dependent on taxonomic grade shifts and allometry. Over the years, researchers have debated whether brains should be compared in terms of absolute size (i.e., as a whole, as well as uncorrected for allometry) or in terms of allometrically corrected size of their component parts. For the moment, studies on primates, birds and insects seem to yield divergent results on this issue. A recent analysis by Deaner Brains, Lifestyles and Cognition

et al. [2007] suggests that absolute brain size, uncorrected for allometry, is the best predictor of comparative cognitive performance in primates. Absolute size of the cortex is also a good predictor of rates of social deception [Byrne and Corp, 2004], innovation, tool use and social learning [Reader and Laland, 2002]. In insects, new insights have recently been proposed by Sarah Farris [Farris, 2005, 2008; Farris and Roberts, 2005]. The key neuroecological difference in insect brains appears to lie in the degree of gyrification of the mushroom bodies. Gyrencephalic mushroom bodies with multiple calyces characterize insects with generalist diets (e.g., Scarites subterraneus), opportunistic invasive lifestyles (e.g., cockroaches, Japanese beetles) or complex societies (e.g., honeybees, termites). In contrast, specialized insects such as flesh flies, ladybirds and dung beetles have lissencephalic mushroom bodies with single calyces. The number of Kenyon cells is also much higher in the gyrencephalic mushroom bodies of generalist beetles [Farris and Roberts, 2005]. In birds, the mesopallium and nidopallium are thought to be equivalent to the cortex of mammals. Residual size of the mesopallium is marginally better than that of other telencephalic structures in predicting innovation rate in birds [Timmermans et al., 2000], whereas that of the nidopallium is the best predictor of true tool use rate [Lefebvre et al., 2002]. Differences in r2 are very small, however, and do not provide very robust evidence that a focus on restricted brain areas better accounts for inter-taxon variance in cognition than does a focus on the whole brain. Data on mesopallium, nidopallium, telencephalon and whole brain size are currently available on 67 avian species [Iwaniuk and Hurd, 2005]. Table 1 shows that the average partial correlation with tool use and innovation rate is not higher when pallial components are used than when the anatomical level chosen is the whole brain. Table 1 also shows that, contrary to the situation in primates, absolute size of the brain or of pallial areas is unrelated to differences in innovation and tool use rates. Finally, different kinds of allometric corrections all seem to produce similar results in birds, whether they be residuals of regressions against log body size, executive brain ratios (volume of the mesopallium plus nidopallium divided by volume of the brainstem) or EQ [using the equation of Martin, 1981]. For the moment, the data on insects, primates and birds thus do not yield a clear consensus on the neural level that best predicts cognitive differences within each group.


139

Genetic Variation in Brain Size

Table 1. Mean proportion of variance in innovation and tool use

rate explained by neural size measures

If variation in the size, gyrification or neuron numbers of brain areas is a key correlate of adaptive cognitive differences between taxa, we should be able to identify genetic differences that lead to neural changes. Recent work on mice and humans provides interesting examples of such genetic effects. In mice, transgenic individuals that express an abnormal form of B-catenin in neural precursors show large increases in brain size due to extension of the surface area of the cortex leading to strong gyrification [Chen and Walsh, 2002]. In humans, single genes that in their abnormal form cause primary microcephaly (ASPM and MCPH1) show allelic variation suggestive of positive selection in contemporary human populations [Mekel-Bobrov et al., 2005; Evans et al., 2005; Vallender, 2008]. The phylogenetic history of ASPM and microcephalin genes shows faster rates of positive selection in the branches that separate humans and chimpanzees, as well as the branches that separate apes from Old World monkeys [Kouprina et al., 2004]. The implications of this genetic variation for cognition are for the moment obscure. Inter-population variation in ASPM and MCPH1 alleles shows no consistent relationship with performance on IQ tests [Mekel-Bobrov et al., 2007], but does seem to show an intriguing degree of correspondence with tonal versus atonal features of different languages [Dediu and Ladd, 2007].

Selective Pressures Favoring Larger Brains

Many studies on ecological correlates of whole brain size assume that natural, not sexual, selection drives the co-evolution. Instead, sexual selection is usually thought to produce changes in restricted brain areas involved in modular specializations [e.g., song nuclei, the hippocampus of brood parasites and polygynous rodents; see Jacobs, 1996 for a review]. Sexual selection often leads to reduction in whole brain size, as animals invest instead in costly traits that allow them to attract mates (bright colors, intricate displays, exaggerated structures) or outcompete individuals of their own sex (large size, large fertilization organs, combat structures). For example, there appears to be a trade-off in bats between relative size of the brain and of the testes [Pitnick et al., 2006]. In primates, species with large size dimorphism tend to have smaller brains than those that are more monomorphic [Schillaci, 2006]. Selection on brain areas also appears to be sex-specific in this order; brain evolution in males ap140


Residual whole brain against body weight Residual telencephalon against body weight Residual mesopallium against body weight Residual nidopallium against body weight Executive ratio (nidopallium + mesopallium/brainstem) EQ [equation from Martin, 1981] Absolute brain size Absolute mesopallium and nidopalium volume

0.37 0.38 0.37 0.37 0.35 0.36 0.01 0.01

Based on raw data from Iwaniuk and Hurd [2005]. All specieslevel (n = 67) neural and body size measures first averaged at the level of the parvorder (n = 22). Values in the right column are means of the r2 of the regressions of each neural measure against innovation rate, on the one hand, and tool use rate on the other [data from Lefebvre et al., 2004].

pear to have been driven by physical conflict, whereas socio-cognitive skills seem to have been the main driver in females [Lindefors et al., 2007]. Exceptions to this generally negative relationship between brain size and sexual selection are bower birds and zebra finches. In the family Ptilonorhynchidae, species that build more complex bowers have larger brains [Madden, 2001; see however Day et al., 2005] and/or cerebella [Day et al., 2005] than those that either build simpler bowers or no bowers at all. Zebra finches that sing more complex songs have both a larger HVC and telencephalon than those that sing simpler songs; song complexity and size of both the HVC and the whole telencephalon are heritable in this species [Airey et al., 2000b]. Whether these trends occur in other species needs to be determined. Bowerbird females prefer males that build more complicated bowers (and bowers that incorporate colors from local foods), and female zebra finches prefer males that sing more complex songs. As more complex zebra finch songs are associated both with a larger HVC and a larger telencephalon [Airey et al., 2000b], and more complex bowers associated with a larger brain [Madden, 2001] or cerebellum [Day et al., 2005], female choice is sexually selecting for the size of neural structure size. In polygynous meadow voles, females prefer males with better spatial ability; spatial ability in this species is associated with a larger hippocampus, again resulting in sexual selection for neural structure size [Spritzer et al., 2005]. Paradoxically, Spritzer et al. [2005] found that males with better spatial abilities attract more mates, but they were not able to show that these males father more offspring [see Sherry, 2006 for a discussion]. Lefebvre /Sol

The natural selection forces that have been suggested to shape brain size evolution mostly emphasize the need for enhanced cognition in three main contexts, foraging, parental care and social relationships. The pioneeering work in this area originated in the late 1970’s and early 1980’s, centered around Paul Harvey, Tim Clutton-Brock and Peter Bennett, and focused on categorical measures of ecological correlates [e.g., diet, Harvey et al., 1980] of brain size. Continuous variables were then added, most notably size of the social unit [Sawaguchi and Kudo, 1990; Perez-Barberia et al., 2007]. In all cases, the association with brain size assumes an untested cognitive intermediary: larger neural centers allow some animals to be more flexible than others in their foraging techniques [Reader and Laland, 2002; Ricklefs, 2004; Ratcliffe et al., 2006], better able to monitor spatial and temporal variation of patchy food sources [e.g., fruit, Clutton-Brock and Harvey, 1980], process information about multiple food types [Reader and MacDonald, 2003], or better interact with more social partners [Dunbar, 1998], detect and rapidly capture evasive prey [Gittleman, 1986], extract rare hidden foods [Parker and Gibson, 1977] and/or program a larger repertoire of foraging patterns [Changizi, 2003]. These presumed functions are not mutually exclusive and several or all of them might have contributed to brain size evolution in a given taxon. Pollen et al. [2007] showed for example that physical and social variables affecting cichlid brain evolution are highly correlated. In fact, physical and social variables might be driven by common ecological factors. The more a food is clumped and predictable in time and space, the more an animal can specialize on it and exploit it alone, driving others away. The solitary-group living continuum and the specialist-generalist continuum may thus vary together, as shown in a recent model [Overington et al., in press], and ecological and social theories of cognition might thus have more in common than is often assumed. The fitness benefits of large brains have never been directly quantified within any population, but recent comparative analyses in birds and mammals suggest that large brain size might enhance survival in nature. In birds, mean adult mortality is lower in populations, species and families with larger brains, after the effect of other factors known to influence mortality (the most important being body size) have been taken out [Sol et al., 2007]. In birds [Sol et al., 2005] and mammals [Sol et al., in press], but not in fish [Drake, 2007], the long-term survival of populations introduced by humans outside their natural range is higher in species with larger brains, an association that is largely independent of other factors

that influence the invasion process. Results on avian and mammalian introductions suggest that environmental change might be a key factor in the evolution of enlarged brains.



Convergent Brain Size Evolution among Taxa

In birds and primates, there appears to be a remarkable degree of convergence in the relationship between pallial area size, innovation rate, tool use and reversal learning [Lefebvre et al., 2004]. Does this imply that the evolution of larger brains and brain areas have been driven by similar selective pressures? The finding that brain size facilitates survival in novel regions in birds [Sol et al., 2005] as well as mammals [Sol et al., in press] suggests some convergence in the adaptive function of the brain. For social variables, the relationship also appears to be convergent, but shows some complications. In primates, social group size, social learning and social deception are all positively correlated with size of the pallial area. In birds, the relationship between brain size and social group size appears to be curvilinear rather than linear [Emery et al., 2007]; beyond group sizes of 70, animal numbers seem to be too large to allow the individual recognition and relationships that would put a premium on social computing powers of the brain. The relationship between social learning and the brain is also difficult to establish in birds because the number of reported cases of social learning in the wild is both small and strongly skewed towards a single taxon: oscines [Lefebvre and Bouchard, 2003]. The idea that a broader diet might be one of the main drivers of brain and cognitive evolution is supported in birds [Overington et al., in press], in primates [Reader and MacDonald, 2003], in bats [Ratcliffe et al., 2006] and in insects [Farris and Roberts, 2005; Farris, 2005, 2008]. In fish, brain size differences between cichlid species correlate with mating system, habitat complexity and interspecific competition [Pollen et al., 2007; Shumway, 2008], suggesting that in this taxon also, lifestyles that demand more information-processing seem to select for larger brains and brain areas.

Conclusion

Comparative and experimental approaches to animal cognition show largely consistent results. Similar trends associated with similar lifestyles are also seen in taxa that are phylogenetically distant and possess remarkably dif141

ferent brains. Many cognitive measures show positive intercorrelations at the inter-individual and inter-taxon level; ecological principles such as the unpredictability of resources in space and time might drive different types of cognition (e.g., social and non-social) in the same direction. If different cognitive activities involve networks of localized centers distributed in many areas of the brain and if most of these cognitive activities are positively correlated across individuals and species, then we need to incorporate this into our thinking about brain evolution. Imagine two species that differ on four cognitive measures, each one of which involves in turn eight distributed brain centers. Among New World monkeys, for example, marmosets and capuchins show clear differences in rates of innovation, tool use, social learning, as well as reversal learning speed. In all four cases, capuchins outrank marmosets, as they do on measures of absolute and relative size of the whole brain and cortex. In the 20 to 25 million years since the marmoset and capuchin lineages split, how did divergent evolution of their brains and cognition occur? By separate changes in the genes affecting each of the 32 brain centers involved in the four cognitive processes? By a limited number of genetic changes that had broad consequences all over the brain? To more efficiently tackle these questions, we need more comparative data on fine neural differences in animals that show divergent lifestyles, of the type that Goodson and his group have gathered on avian sociality. We need to iden-

tify more gene changes that are associated with neural differences, be they changes that could affect the whole brain [Kouprina et al., 2004; Evans et al., 2005; MekelBobrov et al., 2005, 2007; Vallender, 2008] or cortex [Chen and Walsh, 2002] or localized centers. We need brain imaging data on batteries of cognitive processes that are either positively (e.g., reversal learning, social learning, and tool use) or negatively (spatial memory) correlated across taxa. Finally, we need to identify the evolutionary processes that drive increases in brain size, integrating retrospective approaches where present-day brain-ecology associations are linked to historical events, as well as prospective approaches, asking how present ecological conditions select for enhanced cognitive skills and larger brains. Only when we assemble enough evidence for all these issues we will be able to construct a general theory for the evolution of large brains. Acknowledgements This paper is dedicated to the memory of Georg Baron (who died in 2007), co-author of the most widely used database on primate [Stephan et al., 1981], insectivore and bat brain component size, and professor (to L.L., among others) of animal behavior at the Université de Montréal. We thank Andrew Iwaniuk for providing access to the raw data used in Iwaniuk and Hurd [2005]. D.S. is supported by a Ramón y Cajal fellowship and a Proyecto de Investigación (CGL2007–66257) from the Ministerio de Educación y Ciencia (Spain). L.L. is supported by NSERC (Canada).

References Airey DC, DeVoogd TJ (2000) Variation in song complexity and HBC volume are significantly related in zebra finches. NeuroReport 10: 2339–2344. Airey DC, Kroodsma DE, DeVoogd TJ (2000a) Differences in the complexity of song tutoring cause differences in the amount learned and in dendritic spine density in a songbird telencephalic song control nucleus. Neurobiol Learn Mem 73:274–281. Airey DC, Castillo-Juarez H, Casella G, Pollak EJ, DeVoogd TJ (2000b) Variation in the volume of zebra finch song control nuclei is heritable: developmental and evolutionary implications. Proc R Soc Lond B 267: 2099– 2104. Bennett PM, Harvey PH (1985) Relative brain size and ecology in birds. J Zool 207: 151– 169. Bonnie KE, Horner V, Whiten A, de Waal FBM (2007) Spread of arbitrary conventions among chimpanzees: a controlled experiment. Proc R Soc Lond B 274:367–372.

142

Bouchard J, Goodyer W, Lefebvre L (2007) Innovation and social learning are positively correlated in pigeons. An Cog 10:259–266. Brenowitz EA (1997) Comparative approaches to the avian song system. J Neurobiol 33:517– 531. Brosnan SF, Grady MF, Lambeth SP, Schapiro SJ, Beran MJ (2008) Chimpanzee autarky. PloS One 3:e1518. Byrne RW, Corp N (2004) Neocortex size predicts deception rate in primates. Proc R Soc Lond B 271:1693–1699. Byrne RW, Whiten A (eds) (1988) Machiavellian Intelligence: Social Expertise and the Evolution of Intellect in Monkeys, Apes and Humans. Oxford, UK: Oxford University Press. Byrne RW (1993) Do larger brains mean greater intelligence? Behav Brain Sci 16:696–697. Changizi MA (2003) Relationship between number of muscles, behavioral repertoire size, and encephalization in mammals. J Theor Biol 220:157–168.


Chen A, Walsh CA (2002) Regulation of cerebral cortical size by control of cell cycle exit in neural precursors. Science 297:365–369. Clutton-Brock TH, Harvey PH (1980) Primates, brain and ecology. J Zool Lond 190: 309– 323. Day LB, Westcott DA, Olster DH (2005) Evolution of bower complexity and cerebellum size in bowerbirds. Brain Behav Evol 66:62– 72. Deaner, Ro, van Scahik, CP, Johnson V (2006) Do some taxa have better domaon-general cognition than others? A meta-analysis of non-human primate studies. Evol Psych 4: 149–196. Deaner RO, Isler K, Burkart J, van Schaik C (2007) Overall brain size, and not encephalization quotient, best predicts cognitive ability across non-human primates. Brain Behav Evol 70:115–124.

Lefebvre /Sol

Dediu D, Ladd DR (2007) Linguistic tone is related to the population frequency of the adaptive haplogroups of two brain size genes, ASPM and microcephalin. Proc Natl Acad Sci USA 104:10944–10949. DeVoogd TJ, Krebs JR, Healy SD, Purvis A (1993) Relations between song repertoire size and the volume of brain nuclei related to songcomparative evolutionary analyses amongst oscine birds. Proc R Soc Lond B 254:75–82. Drake JM (2007) Parental investment and fecundity, but not brain size, are associated with establishment success in introduced fishes. Funct Ecol 21:963–968. Drapier M, Chauvin C, Dufour V, Uhlrich P, Thierry B (2005) Food-exchange with humans in brown capuchin monkeys. Primates 46:241–248. Dunbar RIM (1998) The social brain hypothesis. Evol Anthropol 6:178–190. Emery NJ, Seed AM, von Bayern AMP, Clayton NS (2007) Cognitive adaptations of social bonding in birds. Phil Trans R Soc B 362: 489–505. Evans PD, Gilbert SL, Mekel-Bobrov N, Vallender EJ, Anderson JR, Vaez-Azizi LM, Tishkoff SA, Hudson RR, Lahn BT (2005) Microcephalin, a gene regulating brain size, continues to evolve adaptively in humans. Science 309:1717–1720. Farris SM (2005) Evolution of insect mushroom bodies: old clues, new insights. Arthr Struc Dev 34:211–234. Farris SM (2008) Evolutionary convergence of higher brain centers spanning the protostome-deuterostome boundary. Brain Behav Evol 72:106–122. Farris SM, Roberts NS (2005) Coevolution of generalist feeding ecologies and gyrencephalic mushroom bodies in insects. Proc Natl Acad Sci USA 102:17394–17399. Fragaszy D, Izar P, Visalberghi E, Ottoni EB, De Oliveira MG (2004) Wild capuchin monkeys (Cebus libidinosus) use anvils and stone pounding tools. Am J Primatol 64: 359–366. Garamszegi LZ, Eens M (2004) Brain space for a learned task: strong intraspecific evidence for neural correlates of singing behavior in songbirds. Brain Res Rev 44:187–193. Giraldeau LA, Lefebvre L, Morand-Ferron J (2007) Can restrictive definitions lead to biases and tautologies? Behav Brain Sci 30: 411–412. Gittleman JL (1986) Carnivore brain size, behavioural ecology, and phylogeny. J Mammal 67: 23–36. Goodson JL, Evans AK, Wang Y (2006) Neuropeptide binding reflects convergent and divergent evolution in species-typical group sizes. Horm Behav 50:223–236. Griffin AS, Galef BG Jr (2005) Social learning about predators: does timing matter? Anim Behav 69:669–678. Harvey PH, Clutton-Brock TH, Mace GM (1980) Brain size and ecology in small mammals and primates. Proc Natl Acad of Sci USA 77: 4387–4389.


Herculano-Houzel S (2007) Encephalization, neuronal excess, and neuronal index in rodents. Anat Rec 290:1280–1287. Herculano-Houzel S, Mota B, Lent R (2006) Cellular scaling rules for rodent brains. Proc Natl Acad Sci USA 103:12138–12143. Herculano-Houzel S, Collins CE, Wong PY, Kaas JH (2007) Cellular scaling rules for primate brains. Proc Natl Acad Sci USA 104: 3562–3567. Hunt GR (1996) Manufacture and use of hooktools by New Caledonian crows. Nature 379: 249–251. Hunt GR, Gray RD (2004) Direct observations of pandanus-tool manufacture and use by a New Caledonian crow (Corvus moneduloides). Anim Cogn 7:114–120. Iacoboni M, Woods RP, Brass M, Bekkering H, Mazziotta JC, Rizzolatti G (1999) Cortical mechanisms of human imitation. Science 286:2526–2528. Iwaniuk AN, Hurd PL (2005) The evolution of cerebrotypes in birds. Brain Behav Evol 65: 215–230. Iwaniuk AN, Nelson JE (2003) Developmental differences are correlated with relative brain size in birds: a comparative analysis. Can J Zool 81:1913–1928. Jacobs LF (1996) Sexual selection and the brain. Trends Ecol Evol 11:A82–A86. Jerison HJ (1973) Evolution of the Brain and Intelligence. New York: Academic Press. Kenward B, Weir AAS, Rutz C, Kacelnik A (2005) Tool manufacture by naïve juvenile crows. Nature 433:121–121. Kouprina N, Pavlicek A, Solomon G, Gersch W, Yoon YH, Collura R, Ruvolo M, Barrett JC, Woods CG, Walsh CA, Jurka J, Larionov V (2004) Accelerated evolution of the ASPM gene controlling brain size begins prior to human brain expansion. PLoS Biol 2: 653– 663. Krebs JR, Sherry DF, Healy SD, Perry H, Vaccarino AL (1989) Hippocampal specialization in food storing birds. Proc Natl Acad Sci USA 86:1388–1392. Lefebvre L, Bouchard J (2003) Social learning about food in birds. In: The Biology of Traditions (Perry S, Fragaszy DM, eds), pp 94–126. Cambridge, UK: Cambridge University Press. Lefebvre L (1982) Food exchange strategies in an infant chimpanzee. J Human Evol 11: 195– 204. Lefebvre L, Bolhuis JJ (2003) Positive and negative correlates of feeding innovations in birds: evidence for limited modularity. In: Animal Innovation (Reader SM, Laland KN, eds), pp 39–62. Oxford, UK: Oxford University Press. Lefebvre L, Juretic N, Timmermans S, Nicolakakis N (2001) Is the link between innovation rate and forebrain size caused by confounding variables? A test on North American and Australian birds. Anim Cogn 4:91–97. Lefebvre L, Nicolakakis N, Boire D (2002) Tools and brains in birds. Behaviour 139:939–973.

Lefebvre L, Reader SM, Sol D (2004) Brains, innovations and evolution in birds and primates. Brain Behav Evol 63: 233–246. Lefebvre L, Whittle P, Lascaris E, Finkelstein A (1997a) Feeding innovations and forebrain size in birds. Anim Behav 53:549–560. Lefebvre L, Templeton J, Brown K, Koelle M (1997b) Carib grackles imitate conspecific and Zenaida dove tutors. Behaviour 134: 1003–1017. Lewis JW (2006) Cortical networks related to human use of tools. Neuroscience 2: 211– 231. Lindefors P, Nunn CL, Barton RA (2007) Primate brain architecture and selection in relation to sex. BMC Biol 5: 20. Lucas JR, Brodin A, de Kort SR, Clayton NS (2004) Does hippocampal size correlate with the degree of caching specialization? Proc R Soc Lond B 271:2423–2429. Madden J (2001) Sex, bowers and brains. Proc R Soc Lond B 268:833–838. Martin RD (1981) Relative brain size and basal metabolic-rate in terrestrial vertebrates Nature 293:57–60. Mekel-Bobrov N, Gilbert SL, Evans PD, Vallender EJ, Anderson JR, Hudson RR, Tishkoff SA, Lahn BT (2005) Ongoing adaptive evolution of ASPM, a brain size determinant in Homo sapiens. Science 309:1720–1722. Mekel-Bobrov N, Posthuma D, Gilbert SL, Lind P, Gosso MF, Luciano M, Harris SE, Bates TC, Polderman TJC, Whalley LJ, Fox H, Starr JM, Evans PD, Montgomery GW, Fernandes C, Heutink P, Martin NG, Boomsma DI, Deary IJ, Wright MJ, de Geus EJC, Lahn BT (2007) The ongoing adaptive evolution of ASPM and Microcephalin is not explained by increased intelligence. Hum Mol Genet 16:600–608. Melis AP, Hare B, Tomasello M (2006) Chimpanzees recruit the best collaborators. Science 311:1297–1300. Morand-Ferron J, Lefebvre L, Reader SM, Sol D, Elvin S (2004) Dunking behaviour in Carib grackles. Anim Behav 68:1267–1274. Morand-Ferron J, Veillette M, Lefebvre L (2006) Stealing of dunked food in Carib grackles (Quiscalus lugubris). Behav Proc 73: 342– 347. Morand-Ferron J, Giraldeau LA, Lefebvre L (2007a) Wild carib grackles play a producerscrounger game. Behav Ecol 18:916–921. Morand-Ferron J, Lefebvre L (2007b) Flexible expression of a food-processing behaviour: determinants of dunking rates in wild Carib grackles of Barbados. Behav Proc 76: 218– 221. Moura ACD, Lee PC (2004) Capuchin stone tool use in Caatinga dry forest. Science 306:1909– 1909. Nicolakakis N, Lefebvre L (2000) Innovation rate and forebrain size in birds of western Europe: feeding, nesting and confounding variables. Behaviour 137:1415–1429.


143

Obayashi S, Suhara T, Kawabe K, Okauchi T, Maeda J, Akine Y, Onoe H, Iriki A (2001) Functional brain mapping of monkey tool use. Neuroimage 14:853–861. Orenstein RI (1972) Tool-use by Caledonian crow (Corvus moneduloides). Auk 89: 674– 676. Overington SE, Dubois F, Lefebvre L (2008) Food unpredictability drives both generalism and social foraging: a game theoretical model. Behav Ecol 19:836–841. Parker ST, Gibson KR (1977) Object manipulation, tool use and sensorimotor intelligence as feeding adaptations in cebus monkeys and great apes. J Hum Evol 6:623–641. Perez-Barberia FJ, Shultz S, Dunbar RIM (2007) Evidence for coevolution of sociality and relative brain size in three orders of mammals. Evolution 61:2811–2821. Pfaff JA, Zanette L, MacDougall-Shackleton SA, MacDougall-Shackleton EA (2007) Song repertoire size varies with HVC volume and is indicative of male quality in song sparrows (Melospiza melodia). Proc R Soc Lond B 274: 2035–2040. Pitnick S, Jones KE, Wilkinson GS (2006) Mating system and brain size in bats. Proc R Soc Lond B 273:719–724. Plomin R (2001) The genetics of g in human and mouse. Nature Rev Neurosci 2:136–141. Pollen AA, Dobberfuhl AP, Scace J, Igulu MM, Renn SCP, Shumway CA, Hofmann HA (2007) Environmental complexity and social organization sculpt the brain in Lake Tanganyikan cichlid fish. Brain Behav Evol 70: 21–39. Ratcliffe JM, Fenton MB, Shettleworth SJ (2006) Behavioral flexibility is positively correlated with relative brain volume in predatory bats. Brain Behav Evol 67:165–176.

144

Reader SM, Laland KN (2002) Social intelligence, innovation and enhanced brain size in primates. Proc Natl Acad Sci USA 99: 4436–4441. Reader SM, MacDonald K (2003) Environmental variability and primate behavioural flexibility. In: Animal Innovation (Reader SM, Laland KN, eds), pp 83–116. Oxford, UK: Oxford University Press. Ricklefs RE (2004) The cognitive face of avian life histories – The 2003 Margaret Morse Nice Lecture. Wilson Bull 116:119–133. Rilling JK, Gutman DA, Zeh TR, Pagnoni G, Berns GS, Kilts CD (2002) A neural basis for social cooperation. Neuron 35: 395–405. Sawaguchi T, Kudo H (1990) Neocortical development and social-structure in primates. Primates 31:283–289. Schillaci MA (2006) Sexual selection and the evolution of brain size in primates. PloS One 1:e62. Sherry DF (2006) Neuroecology. Ann Rev Psychol 57:167–197. Sherry DF, Schacter DL (1987) The evolution of multiple memory-systems. Psychol Rev 94: 439–454. Sherry DF, Vaccarino AL, Buckenham K Herz RS (1989) The hippocampal complex of foodstoring birds. Brain Behav Evol 34:308–317. Shumway C (2008) Habitat complexity, brain, and behavior. Brain Behav Evol 72:123–134. Sol D, Bacher S, Reader SM, Lefebvre L (in press) Brain size predicts the success of mammal species introduced into novel environments. Am Nat (in press). Sol D, Duncan RP, Blackburn TM, Cassey P, Lefebvre L (2005) Big brains, enhanced cognition, and response of birds to novel environments. Proc Natl Acad Sci USA 102: 5460–5465. Sol D, Szekely T, Liker A, Lefebvre L (2007) Big brained birds survive better in nature. Proc R Soc Lond B 274:763–769.


Spritzer MD, Meikle DB, Solomon NG (2005) Female choice based on male spatial ability and aggressiveness among meadow voles Anim Behav 69:1121–1130. Stephan H, Frahm H, Baron G (1981) New and revised data on volumes of brain structures in insectivores and primates. Folia Primatol 35:1–29. Taylor AH, Hunt GR, Holzhaider JC, Gray RD (2007) Spontaneous metatool use by New Caledonian crows. Current Biol 17: 1504– 1507. Timmermans S, Lefebvre L, Boire D, Basu P (2000) Relative size of the hyperstriatum ventrale is the best predictor of innovation rate in birds. Brain Behav Evol 56:196–203. Vallender EJ (2008) Exploring the Origins of the Human Brain through Molecular Evolution. Brain Behav Evol 72:168–177. Visalberghi E, Fragaszy DM. (1990) Do monkeys ape? In: Language and Intelligence in Monkeys and Apes (Parker ST, Gibson KR, eds), pp 247–273. Cambridge, UK: Cambridge University Press. Webster S, Lefebvre L (2001) Problem-solving and neophobia in a Passeriforme Columbiforme assemblage in Barbados. Anim Behav 62:23–32. Weir AAS, Chappell J, Kacelnik A (2002) Shaping of hooks in new Caledonian crows. Science 297:981. Whiten A, Custance DM, Gomez JC, Teixidor P, Bard KA (1996) Imitative learning of artificial fruit processing in children (Homo sapiens) and chimpanzees (Pan troglodytes). J Comp Psychol 110:3–14. Zeng SJ, Song K, Xu N, Zhang XW, Zuo MX (2007) Sex difference in cellular proliferation within the telencephalic ventricle zone of Bengalese finch. Neurosci Res 58: 207– 214. Zentall TR (2004) Action imitation in birds. Learn Behav 32:15–23.

Lefebvre /Sol



Beyond Neuroanatomy: Novel Approaches to Studying Brain Evolution Alexander A. Pollen a Hans A. Hofmann b a

Program in Neuroscience, Stanford University, Stanford, Calif., and b Section of Integrative Biology, Institute for Molecular and Cellular Biology, Institute for Neuroscience, The University of Texas at Austin, Austin, Tex., USA

Key Words Behavior ⴢ Brain development ⴢ Neuroethology ⴢ Genomics ⴢ Molecular evolution ⴢ Phylogenetics

Abstract The study of the evolution of brain structure and function, although fascinating, has been contentious, largely due to the correlative nature of neuroanatomical comparisons and the often ill-defined categorizations of habitat and behavior. We outline four conceptual approaches that will help the field of brain evolution emerge from a historical focus on descriptive comparative neuroanatomy. First, reliable, efficient and unbiased behavioral assays must be developed to characterize relevant cross-species differences in addition to focused studies of neuroanatomy. Second, developmental and physiological processes underlying neuroanatomical and behavioral differences can be analyzed using the comparative approach. Third, genome-wide comparisons including genome-wide linkage mapping, transcriptional profiling, and direct sequence comparisons, can be applied to identify the genetic basis for phenotypic differences. Finally, signatures of selection in DNA sequence can provide clues about adaptive genetic changes that affect the nervous system. These four approaches, which all depend on well-resolved phylogenies, will build on detailed neuroanatomical studies to provide a richer understanding of mechanistic and selective factors underlying brain evolution. Copyright © 2008 S. Karger AG, Basel



Introduction

Remarkable differences in brain and behavior can be found even among closely related species. By studying these differences, the field of brain evolution can highlight relationships between structure and function, reveal constraints and selective pressures, and address questions about the evolution of our own brains. However, the field has been criticized for focusing too heavily on size measures of the whole brain or of functionally heterogeneous structures. A recent review argues that more measurements and correlations involving brain structure size will not further our understanding of the function or evolution of the nervous system [Healy and Rowe, 2007]. Thus a major challenge for the field is to develop experimental approaches within a comparative framework that allow the functional analysis of neural phenotypes in the context of development, physiology and behavior. In the following, we present four comparative approaches for studying variability beyond neuroanatomy (see fig. 1 for a conceptual overview), and we discuss examples of the functional insights – from neurobiology and other areas – yielded by these approaches (table 1). First, we emphasize the crucial, yet often under-appreciated, role behavioral testing has to play in any comparative analysis. Second, in line with much recent attention given to the burgeoning field of ‘evo-devo’, we highlight how careful analyses of neural development and physiology across species, given recent molecular advances, can provide important and often unexpected insights into Dr. Hans A. Hofmann Section of Integrative Biology The University of Texas at Austin, 1 University Station – C0930 Austin, TX 78712 (USA) Tel. +1 512 475 6754, Fax +1 512 471 3878, E-Mail [email protected]

Neuroanatomy Development Physiology

Fig. 1. Concept map illustrating four ap-

proaches to brain evolution studies, in addition to neuroanatomy. Evolutionary changes at the level of genes and regulatory elements can affect developmental and physiological processes underlying differences in brain structure and behavior. The four approaches, in the context of well-resolved phylogenies, may be used to study brain evolution at all levels of biological organization.

Genes and regulatory elements

Molecular evolution

Behavior

1. ‘Fairly’ compare brain and behavior across species

2. Analyze mechanisms underlying structural and behavioral changes

3. Determine genetic/ genomic changes aﬀecting phenotype

4. Search for signatures of positive selection

All analyses depend on well-resolved phylogenies

Table 1. Model systems, which have advanced our understanding of brain evolution Model system

Neuroanatomy

Behavioral assays

Food storing in birds/ rodents

Size of hippocampus

Spatial memory

V1a receptor and social affiliation in voles

AVP and V1a receptor distribution

Preference test

Expansion of cortex

Allometric analyses

Ion channel variation in electric organ of electric fish

Neural circuitry

Jaw morphology and dentition in cichlid fish Pelvic spines, plate armor, pigmentation in stickleback Beak morphology in Darwin’s finches

Development and physiology

Genetics and genomics

Signatures of molecular evolution

Behavioral pharmacology

V1a receptor transgenics

Ka/Ks analysis of V1a receptor

Developmental series using molecular markers of proliferation

Brain transcriptome analysis across species

Ka/Ks analysis of aspm, mcph1 and other genes

Electric organ discharge

Voltage clamp analysis of ion currents

Feeding behavior

Electromyographic recordings in jaw muscles

n/a

Ka/Ks analysis of Na+ channel genes Expression of candidate genes, QTL mapping QTL mapping, transgenics

Feeding specialization

Developmental series

allelic variation of ectodysplasin gene

Expression of candidate genes, transgenics in chicken

These systems have applied one or more of the four approaches discussed in this review in addition to comparative neuroanatomy.

functional relationships as well as constraints. Third, exploiting modern techniques in genetics and genomics allows us to identify genetic and molecular factors underlying species differences in behavior and/or brain structure or function. Fourth, the advent of genome-scale sequence repositories enables us to identify regions of the genome that have experienced increased selective pressure and might contribute to adaptations of the nervous system. For all of these approaches, we underscore the vital importance of strong phylogenetic hypotheses; only if the 146


evolutionary relationships between the species under study are taken into account, can meaningful insights emerge. Exemplary systems such as Microtus voles and electric fish have been analyzed by multiple approaches, and we return to these systems in several sections (table 1). By integrating these approaches, comparative studies can describe additional levels of variability across species and examine the functional role of these differences. We do not mean to minimize the importance of additional fine-scale neuroanatomical approaches. Indeed, Pollen /Hofmann

many studies at the anatomical level have focused on specific structures that are likely relevant to ecological correlates measured [see Shumway, 2008], and have highlighted functional relationships, including between the avian high vocal center and song complexity [Devoogd et al., 1993; Spencer et al., 2005], the avian auditory midbrain nucleus and auditory localization [Iwaniuk et al., 2006], the hippocampus and spatial learning [Krebs et al., 1989; Clayton and Krebs, 1995; Reboreda et al., 1996], and many others [e.g., Barton, 1998, 2004, 2006; Rilling et al., 2008]. Additionally, anatomical studies, coupled with phylogenetic analysis, can use gene expression, neurochemistry, and hodology to identify homologous circuits and fields across distantly related species [Maler and Hincke, 1999; Reiner et al., 2005; Castro et al., 2006; Harvey-Girard et al., 2007; Northcutt, 2008]. However, this review focuses on approaches beyond neuroanatomy that can contribute to the study of vertebrate brain evolution [see Holland and Short, 2008 for a discussion of the transition from the invertebrate to the vertebrate brain in chordate evolution; and Farris, 2008 for an in-depth review of the evolution of higher brain centers at the protostome-deuterostome boundary], and we argue that comparisons of behavior and of the underlying developmental, physiological and genetic factors are now ripe for highlighting mechanistic and selective factors underlying brain evolution.

The astonishing diversity of brain structures across species stands in direct relationship to the equally impressive diversity of animal behavior. It has long been assumed that social behavior, foraging behavior, anti-predator behavior and cognition (to name just a few behavioral traits) underlie much of the diversity in brain structure and function [Roth and Wullimann, 2000]. Although most studies that have pursued these hypotheses have focused on comparing neuroanatomical features in a quantitative and standardized fashion, very few have employed standardized behavioral assays across species using the comparative approach and related the results to brain structure and function. It is often difficult to design an experimental paradigm that is ‘fair’, i.e., not biased towards any one of the species under investigation, as species differences that are not directly relevant to the behavior under study might interfere in non-obvious ways. For example, in studies on spatial learning using a food reward, one species might simply be more motivated

by the food reward used, and yet would appear to be superior at spatial learning [Odling-Smee and Braithwaite, 2003]. It is therefore one of the greatest challenges in understanding brain evolution to devise ways of examining social, cognitive and other behaviors in a way that is efficient, robust and relevant given the differences in natural history encountered by many species. Studies of food-storing birds provide a classic example of the comparative method as applied to neuroanatomy and behavior. These studies have been reviewed extensively elsewhere [e.g., Clayton, 1998; Bolhuis and Macphail, 2001; Shettleworth, 2003; Emery, 2006], but below we discuss the insights and controversies that emerge from these studies as examples of opportunities and challenges for the comparative study of behavior. Early studies showed that hippocampal size is greater in food-storing than in non-storing species [Krebs et al., 1989; Sherry et al., 1989], and that hippocampal size within the corvid and parid families correlates with the intensity of food-storing behavior [Healy and Krebs, 1992, 1996; Hampton et al., 1995]. These initial comparative studies of neuroanatomy separated food-storing ecology into only a few categories, but subsequent laboratory studies of behavior have found some evidence that spatial learning performance on other tasks also correlates with food-storing and neuroanatomy [e.g., Basil et al., 1996; Biegler et al., 2001]. The insights from these studies are functional – that the hippocampus likely plays a role in spatial memory – and evolutionary – that selective pressures for improved spatial learning might have driven a size increase in the hippocampus. The functional prediction complements experiments on the function of the hippocampus in standard model systems, and the selective hypothesis has been explored by analyzing hippocampus size in species that may have faced other ecological pressures, such as dispersal, brood parasitism, and migration, for improved spatial memory [reviewed in Clayton, 1998]. Although the correlation between hippocampus size and food-storing ecology appears robust [Lucas et al., 2004], the correlation of these traits with spatial learning abilities is more controversial. Macphail and Bolhuis [2001] review over thirty laboratory studies and note that even though there is a trend for superior performance in species with higher food-storing demands, a few studies show the opposite result and many studies do not show significant differences. These imperfect correlations are likely due in part to limitations of the comparative method. For example, regressions examining functional hypotheses about brain structure size (e.g., correlations of

Novel Approaches to Brain Evolution


Approach 1: Behavioral Assays

147

forebrain and social group size) also cannot explain all data points, even when the relationship is highly significant [Shettleworth, 2003]. Nonetheless, studies of spatial learning compare fewer species than studies of neuroanatomy, making counter-examples more difficult to interpret and highlighting the need for efficient behavioral assays that can be repeated many times. Conflicting results could also stem from the use of different experimental paradigms across studies, which might involve different behavioral capacities (e.g., memory retention over short vs. long timescales) and different contextual variables (e.g., size and type of reward, type of stimuli used, history of animals), and could thus be more suitable or biased towards some species. One way to address this issue of fairness, discussed by Shettleworth [2003], has been to ask a slightly different question: does food-storing ecology correlate with an increased preference for spatial cues relative to color cues in memory tasks? In a sense, measuring this within-species preference normalizes for contextual variables in the paradigm that may contribute to absolute differences across species, and results suggest that spatial cues are in fact more important to food-storing species [reviewed in Shettleworth, 2003]. In another review, Bolhuis and Macphail [2001] advance an even more fundamental critique of comparative studies. They argue that comparative studies address only levels of ultimate causation – the functional role of a phenotype and the selective forces that favor the phenotype – and cannot provide insights into the mechanistic basis of behavior. However, comparative studies in several systems have been extremely successful at revealing mechanistic factors underlying behavioral evolution. For example, the fast-start escape behavior of fishes is an excellent comparative model system for analyzing the neural circuitry and musculoskeletal function underlying this vital behavior. In an elegant series of experiments Hale et al. [2002] carefully described the electromyogram (EMG) features and kinematics of rapid escape behavior in four fish species at key positions in the vertebrate phylogeny and showed that several of the control features associated with this behavior exhibit a mosaic pattern of ancestral and derived traits. The mechanistic insights from these studies depended on the development of robust, ‘fair’ (i.e., unbiased), and efficient behavioral assays. A similar case can be made for Microtus voles, where species comparisons beyond neuroanatomy have come to include developmental, physiological and genetic differences (which we discuss in subsequent sections). Microtus vole species vary in their mating and parental care system. Some species are monogamous and bipa148


rental, whereas others are polygynous with maternalonly care. By housing monogamous prairie vole males with a female in the laboratory, it is possible to induce a partner preference, which can be quantitatively measured as ‘social affiliation’ in the partner association test. Using a two-choice paradigm, males are placed between the familiar female and an unfamiliar female, and the amount of time spent near each is measured. Importantly, this assay can also be applied to polygynous species and thus overcomes the issue of ‘fairness’ (or species bias). At the same time, it is sufficiently robust to characterize subtle within-species differences [Hammock and Young, 2005], and efficient enough to achieve reasonable sample sizes. The partner association test, in combination with comparisons beyond neuroanatomy, has made it possible to examine which neurobiological differences in vole species contribute to divergence in social behavior. The neuropeptide arginine vasopressin was implicated as a candidate for behavioral differences based on studies of vasopressin in other systems and on differences in the distribution of the neuropeptide and its receptor across vole species. By injecting the arginine vasopressin or a receptor antagonist, the partner association test demonstrated that vasopressin is necessary and sufficient for the formation of partner preference [Winslow et al., 1993]. This example and many subsequent experiments in Microtus [reviewed in Lim and Young, 2006] illustrate the potential for comparative studies of behavior to provide mechanistic insights. In this case, early behavioral comparisons categorized Microtus species as monogamous or non-monogamous, and the subsequent development of an unbiased, robust, and efficient assay in prairie voles made it possible to examine candidate proximate factors derived from other comparative studies. Although less complex than affiliative behavior, feeding behavior can often be objectively compared across species by using the underlying morphology as a proxy for behavior. Two well-known examples that illustrate this point particularly well are the association between beak shape and trophic niche in Darwin’s finches [Grant and Grant, 2006], and the movement of mouthparts in relation to feeding specialization in cichlid fishes [Liem, 1979; Wainwright et al., 2001]. However, no attempts have been made to correlate these morphological structures with brain structure or function. An exceptionally elegant system for comparing behavior across species in an unbiased (‘fair’) manner – and one that comes with much neurobiological insight – is available for weakly electric fishes. These animals can Pollen /Hofmann

sense changes to the electric field around them and produce electrical emissions for communication and electrolocation, which can vary dramatically in terms of duration and shape [see, for example, Hopkins, 1999]. Because active electrosensing and communication is conducted already in the currency of the nervous system – i.e., changes in the membrane potential of neurons, receptors and electrocytes – a recording electrode is all that is needed for acquiring comparable behavior data. Few other systems exist that allow such elegant comparisons, but there is an increasing need to overcome this obstacle independent of taxonomic group and the type of behavior under investigation, especially given the opportunities that are arising from other approaches. In the next three sections, we discuss these comparative approaches beyond neuroanatomy that can be used to identify proximate factors that may contribute to brain and behavioral evolution.

Comparative approaches to development and physiology have yielded mechanistic insights about factors underlying differences in brain structure and behavior, and there is great potential for future work. Because the neocortex is the anatomical location underlying many higher human cognitive functions, there is great interest in understanding how and why the human cortex expanded. Many studies have compared the size of the cortex across mammals and evaluated whether ecological factors or developmental constraints are correlated with changes in size [e.g., Harvey and Krebs, 1990; Finlay and Darlington, 1995; Reader and Laland, 2002]. However, by combining the comparative method with advances in our understanding of the molecular and developmental basis of cortical development, recent studies have highlighted mechanistic factors that may contribute to cortical expansion. Two general models have been advanced to explain the evolutionary expansion of the cortex. Both models are based on modifications to one of the three phases of cell division that produce cortical excitatory neurons. In phase 1, progenitors along the neuroepithelium divide symmetrically and ultimately produce radial glia cells. The Radial Unit Hypothesis [Caviness et al., 1995; Rakic, 1995] predicts that simply altering phase 1 can scale the size of the cortex. During phase 2, radial glia cells divide asymmetrically producing a single daughter neuron and

a radial glia cell. In phase 3, radial glia cells produce intermediate progenitor cells, which divide symmetrically in the subventricular zone producing at least two daughter neurons. A radial glia cell that produces an intermediate progenitor cell will thus ultimately produce at least twice the number of neurons per a cell division, compared with directly producing a single neuron. The Intermediate Progenitor Hypothesis [Kriegstein et al., 2006] posits that proportionately more neurogenesis occurs during phase 3 in the evolutionary expansion of the cortex. These models would be impossible to reconcile by focusing experiments on a single model system, such as mouse. In fact, mutations in mice can increase the size of the cortex according to either model. Over-expression of a stabilized beta-catenin transgene increases founder cell division in phase 1 [Chenn and Walsh, 2003], whereas knocking out the gene Cux2 increases the proportion of neurogenesis in phase 3 while leaving phase 1 unchanged [Cubelos et al., 2007]. Nonetheless, by applying the comparative method to developmental processes, several studies have evaluated the predictions of each of these models. The Radial Unit Hypothesis predicts that the adult cortex size should correlate with the size of the embryonic ventricular zone. The Intermediate Progenitor Hypothesis, on the other hand, predicts that the size of the adult cortex should correlate with the size of the subventricular zone (SVZ), which contains intermediate progenitor cells. Additionally, because SVZ cells produce predominantly upper layer neurons [Wu et al., 2005], the proportion of upper layer neurons should be greater in animals with increased relative cortex size. At this point, the Intermediate Progenitor Hypothesis appears to have stronger support, as it has been shown that SVZ size correlates with cortex size in turtle, rats, ferrets, macaques and humans [Smart et al., 2002; Martínez-Cerdeno et al., 2006; Bayatti et al., 2008]. Furthermore, primates and humans in particular have a greater proportion and diversity of upper-layer neurons than do other mammals [Cajal, 1909; Hill and Walsh, 2005]. Despite these insights, there is still an urgent need for more comparative experiments. For example, the Intermediate Progenitor Hypothesis predicts that local changes in SVZ thickness might contribute to gyri and sulci formation, but this has only been observed indirectly in human and macaque [Kriegstein et al., 2006]. Do changes in SVZ thickness predict gyri and sulci formation in orders that independently evolved folded brains? Similarly, does the proportion of upper-layer neurons also increase in non-primate orders with an expanded cortex? At a mo-



Approach 2: Development and Physiology as Functionally Relevant Traits

149

lecular level, do changes in the regulation or sequence of genes, such as Cux2, that control the division of intermediate progenitors, correlate with changes in neocortex size or gyrification? Birds and – to an even larger extent – teleost fishes show a remarkable diversity in forebrain sizes just as seen in mammals [Huber et al., 1997; Iwaniuk and Hurd, 2005; Pollen et al., 2007; Lefebvre and Sol, 2008]. Do changes in patterns of proliferation also underlie these differences? Recent developmental studies in chick suggest that a SVZ is also present in the avian striatum and dorsal ventricular ridge, opening the door to comparisons of SVZ size across avian species [Striedter and Keefer, 2000; Charvet et al., 2007; Cheung et al., 2007]. Meanwhile, recent work describes a much more prominent role for adult neurogenesis in the forebrain of fishes than in mammals, indicating a possible mechanism in addition to embryonic neurogenesis by which patterns of proliferation could affect forebrain structure [Ekström et al., 2001; Zupanc et al., 2006]. Evolutionary modifications of developmental processes can have obvious effects on brain structure, but evolutionary changes in physiological processes can also affect brain structure and behavior [see also Wang, 2008]. As mentioned previously, morphological changes that affect physiology such as in the cichlid jaw apparatus or the beak of Darwin’s finches have been related to feeding behavior, and might also correlate with changes in brain structure. With respect to neurophysiology, studies of communication in weakly electric fish provide an excellent example of insights into mechanistic neurobiology that physiological comparisons combined with other approaches can produce. Below, we discuss studies of electric organ physiology in electric fish that incorporate direct sequence comparisons and signatures of adaptive selection (further discussed in Approach 3: Genetic and Genomic Analyses, and Approach 4: Signatures of Selection in DNA Sequence, respectively). Weakly electric fish use electrical emissions for communication and electrolocation, yet the duration of signals can vary 100-fold across species, and the waveform is also variable [Hopkins, 1999]. Although communication is generally difficult to study at the genetic level, ion channels are known to be a key component in electrical discharges. Na+ channels were previously implicated in the discharges of one species of electric fish [Ferrari et al., 1995], and mutations at functional sites in Na+ channels are known to affect muscle and neuron firing rate in many human clinical syndromes [George, 2005], making Na+ channels good candidate genes for a role in electric communication. However, identifying whether di150


vergence in the physiology of Na+ channels of electric fish underlies species differences requires comparative studies. Electric communication has evolved independently in two groups of teleosts, the mormyriforms in Africa and the gymnotiforms in South America. Zakon et al. [2006, 2008] compared the expression and coding sequence of two candidate genes encoding sodium channels, NaV1.4a and NaV1.4b, in three gymnotiforms and one mormyrid with four related non-electric fish. They observed that NaV1.4a expression had been independently lost from muscle and gained in the electric organ in gymnotiforms and mormyriforms. By integrating their physiological comparison with additional approaches, the authors asked whether changes in NaV1.4a were functional (see Approach 3: Genetic and Genomic Analyses) and adaptive (see Approach 4: Signatures of Selection in DNA Sequence). Direct sequence comparisons revealed that in both gymnotiforms and mormyriforms, many highly conserved amino acids have been replaced, particularly in the domain of the channel responsible for the final steps of fast inactivation. In fact, some of these amino acid substitutions overlap with mutations underlying human channelopathies, and are consistent with the rapid electric pulses observed in these species, whereas replacements at other sites may provide new insights into natural mechanisms for regulating the function of Na+ channels. The loss of muscle expression and gain of electric organ expression also coincided with an increased rate of change in the coding sequence of NaV1.4a in both lineages suggesting that changes in the protein were in fact adaptive. Thus comparative analysis of the physiology underlying electric emissions, combined with additional approaches, not only illuminates the neurobiological mechanism, but it also suggests that the number of ways to produce rapid electric pulses might be limited, as changes in the same ion channel gene appear to be relevant in convergent examples. Again, comparative studies can address many more questions. In this case, only four species of electric fish were compared, and correlations between the structure of Na+ channels and the properties of the electric discharge were not explicitly examined. Future studies that include species spanning the great diversity of signal forms in electric fish may reveal further unknown properties of Na+ channels and additional convergent events. The quantitative nature of electric organ discharges also lends itself to behavioral comparisons. Moreover, Na+ channels probably act in concert with other ion channels that can also be examined. In fact, the regulation of K+ Pollen /Hofmann

channels has also been implicated within species in sex and life history differences in electric organ discharge patterns [Stoddard et al., 2006]. More generally, the comparative study of physiology can be used to examine the functional basis for many species-specific phenotypic differences, to identify examples of convergent evolution, and in some cases to infer the sequence of evolutionary changes [e.g., Berenbrink et al., 2005; Bridgham et al., 2006]. Thus comparative studies of developmental and physiological processes underlying natural diversity can contribute to a mechanistic and evolutionary understanding of the nervous system.

The ability to compare genomic variability across populations and species complements the mechanistic comparisons of Approach 2 (Development and Physiology). In many cases, developmental and physiological pathways underlying a derived trait are either poorly understood or involve too many genes to narrow down top candidates underlying evolutionary change. By contrast, genome-wide comparisons can provide unbiased measures of how well genomic regions, expression patterns, and sequence variations associate with derived traits. Two major techniques, genome-wide linkage mapping and transcriptional profiling have been used to study the genetic basis for some morphological differences in a diverse range of species. A third technique of direct sequence comparisons can also be used to highlight genetic changes that may be functional. These techniques, particularly when combined with functional assays, are powerful ways to determine how derived traits are produced, and are also likely to illuminate novel roles for genes in producing phenotypes of interest. Genome-wide linkage mapping involves hybridizing phenotypically divergent individuals from related populations or species and allows for the identification of chromosomal markers that segregate with traits of interest. The power of this technique is that the actual genomic regions responsible for evolved traits are revealed, and their effect sizes and level of dominance can be estimated. Indeed, quantitative trait loci have been identified that control substantial variation in fruitfly mating behavior [Moehring and Mackay, 2004; Gleason and Ritchie, 2004], jaw and tooth specializations in cichlid fish [Streelman et al., 2003; Albertson et al., 2005; Streelman and Albertson, 2006], the reduction of pelvic spines and armor plates in stickleback fish [Shapiro et al., 2004; Co-

losimo et al., 2004], albinism in cavefish populations [Protas et al., 2006], and morphological traits in many other vertebrates. These studies provide insights into fundamental evolutionary questions about the type and magnitude of genetic changes underlying recent phenotypic diversification [Kocher, 2004; Peichel, 2005]. However, the application of genome-wide linkage mapping to brain and behavior remains a challenge for several reasons. First, analysis often requires over 100 individuals, and efficient and standardized comparisons of behavior can be difficult. Second, the plasticity characterizing many neural phenotypes makes mapping difficult because phenotypes might not reliably correspond to genotypes. Finally, mapping studies identify chromosomal windows, but often lack the resolution to identify individual genes. Therefore, identifying the key gene within a large window of sequence could be easier for a trait such as pigmentation, whose genetic basis is well understood [e.g., Protas et al., 2006; Miller et al., 2007], than for behavioral traits, where the genetic basis is much less clear. Despite these challenges, neural phenotypes are amenable to genome-wide linkage mapping. Indeed, recent linkage studies have identified loci controlling anxiety in mice strains [Talbot et al., 1999; Yalcin et al., 2004] and anti-predator behavior in zebrafish [Wright et al., 2006]. Similarly, loci affecting brain structure size and neuron number in inbred mouse strains have been identified [Williams et al., 1998; Dong et al., 2007], and endophenotypes such as neural gene expression, and levels of hormones and other metabolites can be measured efficiently [e.g., Freimer et al., 2007]. Transcriptional profiling can also be used to highlight the genetic basis for phenotypic changes [Hofmann, 2003]. This approach involves comparing gene expression in specific tissue types and developmental stages thought to underlie phenotypic differences. For a targeted set of genes, in situ hybridization can achieve the histological specificity to detect qualitative changes in expression. To reveal quantitative changes, transcript abundance can be compared across the genome using microarrays or massively parallel sequencing technologies [Renn et al., 2004; Hoheisel, 2006; Renn et al., 2008; Torres et al., 2008; Vera et al., 2008]. Compared to genomewide linkage mapping, transcriptional profiling has some limitations for identifying the specific genetic changes responsible for phenotypic changes: Regulatory changes do not necessarily underlie phenotypic differences; cis and trans factors controlling gene expression are difficult to distinguish [Osada et al., 2006; Genissel et al., 2008]; and further experiments are required to determine



Approach 3: Genetic and Genomic Analyses

151

whether expression differences are causes or effects of phenotypic differences. However, transcriptional profiling also has clear advantages over genome-wide linkage mapping: it requires fewer individuals, can be applied across species in which genetic crosses are impossible, and can provide more global information about changes in networks of gene expression. In fact, transcriptional profiling of candidate genes has been applied to a canonical example of adaptive specialization, the beaks of Darwin’s finches, to identify a regulatory change contributing to morphological evolution. By comparing beak development in six species of Darwin’s finches and the chick, Abzhanov et al. [2004] identified the developing mesenchyme of the beak prominence as the tissue in which phenotypes likely diverge. In a targeted in situ screen of growth factor genes, the authors identified striking changes of Bmp4 expression localized to this tissue, and then applied functional tests (discussed below) to confirm a role for Bmp4 in beak morphology. Transcriptional profiling has also been applied to compare human and chimpanzee gene expression in adult post-mortem tissues, including many brain regions [Cáceres et al., 2003; Khaitovich et al., 2004, 2006]. These initial findings provide a broad overview of differences between species and also between brain regions, but are difficult to interpret because of hybridization differences to human arrays, different cellular compositions, and the challenge of distinguishing neutral and adaptive gene dosage changes. Nonetheless, follow-up studies have confirmed cortical gene expression changes in synaptogenic thrombospondins [Cáceres et al., 2007], and multi-species arrays [Gilad et al., 2005], ‘next generation’ sequencing technologies [Torres et al., 2008; Vera et al., 2008], and new methods of analysis [Khaitovich et al., 2005] can overcome many of these limitations. Finally, direct sequence comparisons have been used to find specific nucleotide changes that may contribute to novel phenotypes. Most frequently, this approach has been applied to candidate genes. As mentioned earlier, a number of studies in Microtus voles demonstrated that changes in the pattern of arginine vasopressin receptor expression may underlie species differences in pair-bonding behaviors [for review see Lim and Young, 2006]. Direct sequence comparisons of promoter regions have helped to elucidate the genetic basis for changes in gene expression by suggesting that an expanded microsatellite repeat in the 5ⴕ promoter region of the monogamous but not the polygynous vole might be responsible for the observed regulatory changes [Lim et al., 2004]. However, it 152


should be noted that the presence or absence of this expanded microsatellite repeat does not correlate with the mating systems in other Microtus vole species [Fink et al., 2006]. In addition, highly conserved non-coding elements acting as long-range enhancers can also play a major role in gene regulation [Pennachio et al., 2006]. Aiming at identifying such enhancers, Sasaki et al. [2008] studied the distribution of the AmnSINE1 retrotransposon family across mammalian genomes and found that members of the AmnSINE1 family are highly conserved within mammals, suggesting that the expansion of this retrotransposon family early in the evolution of mammals might have affected the expression of many genes, including those associated with mammalian-specific forebrain development. Of course, hypotheses based on direct sequence comparisons must be tested in functional assays. Functional analyses are extremely valuable for evaluating the results of genome-wide linkage mapping, transcriptional profiling and especially direct sequence comparisons. For example, the expanded microsatellite version of the prairie vole arginine vasopressin promoter has been shown to modulate expression of a reporter gene in cell culture [Hammock and Young, 2004], and the overall promoter region is sufficient to drive expression of a transgene in a prairie vole-like pattern in the mouse forebrain [Young et al., 1999]. Similarly, two of the duplicated AmnSINE1 sequences, near the gene Fgf8 [which affects cortical patterning: Fukuchi-Shimogori and Grove, 2001] and near the gene Satb2 [which specifies upper-layer neuron identity: Alcamo et al., 2008; Britanova et al., 2008], were shown to drive expression of a reporter gene in forebrain domains matching components of the endogenous expression pattern of these genes [Sasaki et al., 2008]. The fact that these elements act as forebrain enhancers suggests that other members of the AmnSINE1 family might also drive neural gene expression. These enhancer elements could be particularly relevant to mammal brain evolution, because the expansion and subsequent conservation of this retrotransposon family corresponds to the expansion of the dorsal cortex from three to six layers and other structural changes in the forebrain [Sasaki et al., 2008]. Thus direct sequence comparisons coupled with expression assays can highlight sequence changes affecting gene regulation across many phylogenetic levels. In addition to expression assays, functional tests can also use transgenic and knock-in approaches to determine whether identified genes are sufficient to drive phenotypic changes. For example, genome-wide linkage mapping in stickleback fish suggested that an allele for Pollen /Hofmann

ectodysplasin (EDA) was responsible for loss of armor plates. Transgenic analysis in freshwater stickleback fish showing extreme plate reduction confirmed that a mouse ortholog of the gene was sufficient to drive the formation of additional plates [Colosimo et al., 2005]. Similarly, based on the striking changes in BMP4 expression in Darwin’s finches, Abzhanov et al. [2004] used a viral vector to drive BMP4 in the mesenchyme of the chick beak prominence, producing deep and broad beaks resembling those of ground finches. In polygynous montane voles, viral mediated expression of arginine vasopressin receptor in the ventral pallidum, but not the caudate, was shown to increase affiliative behaviors reminiscent of pairbonding [Pitkow et al., 2001]. Knock-in experiments are a more precise form of analysis in which the sequence of a particular locus is actually replaced with another allele. This technique most closely recapitulates evolutionary changes, and will be extremely useful for evaluating the function of evolving protein domains and enhancer elements. For example, replacing a forelimb enhancer in mouse with a bat version that had accumulated sequence changes, increased expression of the nearby developmental gene Prx1 in the forelimb, and quantitatively increased mouse forelimb length during development [Cretekos et al., 2008]. These morphological and neurobiological examples highlight the potential of genetic and genomic analyses to generate testable hypotheses about the mechanistic basis of phenotypic evolution in the nervous system, and to complement approaches that compare developmental and physiological processes across species.

Approach 4: Signatures of Selection in DNA Sequence

Analyses of molecular evolution can provide a shortcut to identifying genetic changes that contribute to adaptation. Whereas direct sequence comparisons can be used to highlight putative functional changes (Approach 3: Genetic and Genomic Analyses), sequence comparisons in the context of molecular evolution models might also show that adaptive selection is the best explanation for a disproportionate number of sequence changes being fixed in a given region. One general signature of adaptive evolution useful for interspecies comparisons is to identify conserved regions of the genome that have accumulated an unusually large number of functional nucleotide changes in a particular lineage. Although most studies have focused on protein-coding genes, new techniques are being developed to study adaptive changes in nonNovel Approaches to Brain Evolution

coding regions of the genome. By correlating signatures of adaptive selection in specific genes or elements with phenotypic changes that these genes might affect, studies of molecular evolution, as in other genetic and genomic analyses, can generate testable hypotheses about the genetic basis for novel phenotypes. How can we estimate the strength and nature of selection acting on genes? In protein-coding sequences, signatures of selection may be examined by comparing the ratio of amino acid replacement substitutions (Ka) to synonymous substitutions (Ks). Nucleotide substitutions that replace amino acids comprise putative functional changes that are subject to the forces of natural selection. In contrast, nucleotide substitutions at degenerate positions of a codon or in nearby non-functional sequences approximate neutral changes that are fixed only at a rate determined by genetic drift. If amino acid replacements are deleterious, purifying selection will reduce the Ka/Ks ratio. Indeed, in most proteins, the Ka/Ks ratio is low (⬃0.15), highlighting constraints on protein structure. In principle, the Ka/Ks ratio should exceed one only when adaptive selection fixes replacement substitutions more quickly than drift can fix synonymous substitutions. By comparing the sequence of genes affecting neural development across mammals, several studies have identified correlations between Ka/Ks ratios above one and changes in brain size. For example, Evans et al. [2004a, b] analyzed the sequences of ASPM and MCPH1, two genes underlying human primary microcephaly. In both genes, very high Ka/Ks values were observed at historical branches of the primate phylogeny, prior to the divergence of great apes, and also in the lineage leading to humans, and low values were observed in non-primate mammalian lineages. Additional comparative studies have identified other correlations between Ka/Ks values in neurodevelopmental genes and mammal brain size [see Vallender, 2008]. Although it is tempting to speculate that adaptive changes in these genes might affect brain size, genes involved in neural development are frequently expressed in many other tissues, and selection could be acting on non-brain phenotypes [Ponting, 2006]. Nonetheless, these mechanistic hypotheses about the role of amino acid substitutions in brain development are testable in cell culture and in standard model systems. Additionally, evidence for historical episodes of adaptive selection, as observed for ASPM and MCPH1 in early primates, may provide a window into understanding selective pressures and constraints facing ancestral species [Messier and Stewart, 1997].


153

Because Ka/Ks comparisons apply only to changes in protein sequences, it has been difficult to study the evolution of non-coding sequences across species. However, the majority of functional DNA in the genome, as measured by sequence conservation, appears to be non-coding [Mouse Genome Sequencing Consortium, Waterston et al., 2002], and it has been suggested that regulatory and not coding changes underlie many phenotypic differences that arise in recent evolution [Britten and Davidson, 1971; King and Wilson, 1975; Carroll et al., 2001; but see Hoekstra and Coyne, 2007]. Thus, the ability to detect not just protein coding, but also non-coding loci that have been evolving under adaptive selection will be extremely valuable to the study of brain evolution. Several recent studies have analyzed the human genome for signatures of selection in non-coding sequences [Pollard et al., 2006; Prabhakar et al., 2006]. As in studies applying the Ka/Ks ratio, these studies identify genomic regions with a high proportion of potentially functionaltering nucleotide substitutions. However, function-altering substitutions are harder to identify in non-coding sequence than in protein-coding sequence, where the amino acid code is well understood. Nonetheless, substitutions in non-coding sequence can be considered potentially function-altering if they occur in highly conserved sequence. This is because it appears likely that highly conserved non-coding sequence has been under purifying selection for functional reasons to maintain the same sequence characteristics over great phylogenetic distances. In fact, many non-coding elements that are highly conserved between humans and chick or mouse act as transcriptional enhancers for nearby genes [Pennacchio et al., 2006] or encode RNA genes and could contribute to the evolution of gene regulation. To study non-coding changes in the lineage leading to humans, Pollard et al. [2006] identified elements that are highly conserved between rodents and chimpanzee, but that have undergone rapid sequence changes, significant against a neutral model, in the human lineage. In the most exceptional case, the authors identified HAR1F, a novel RNA gene with an element that shares 116 out of 118 bases between chimpanzee and chicken, but has undergone 18 substitutions on the human lineage. Expression studies in macaque and human indicate that HAR1F is expressed in Cajal-Retzius neurons of the subpial granular layer during stages of cortical development in midgestation, as well as several other brain regions and tissues. Given that the subpial granular layer is particularly prominent in humans [Sidman and Rakic, 1973; Zecevic and Rakic, 2001], HAR1F sequence changes might con154


tribute to unique aspects of the human cortex. Again, future laboratory work can test this mechanistic hypothesis. These examples illustrate the potential for molecular evolution approaches to highlight genetic changes that could contribute to adaptive phenotypes in a given lineage. In some cases, studies can suggest adaptive roles and functionally important domains of candidate genes, such as ASPM and MCPH1 in primate brain evolution. In other cases, studies can identify novel elements or genes, such as HAR1F. In all cases, functional tests are required to evaluate whether genetic changes on a lineage contribute to phenotypic changes on the lineage. Nonetheless, results of molecular evolution studies are strengthened when they intersect with other approaches, such as a high Ka/Ks in physiologically relevant NaV1.4a channels of electric fish (Approach 2: Development and Physiology), or results from other genome scale comparisons (Approach 3: Genetic and Genomic Analyses). Given improved techniques for detecting selection in non-coding sequences, rapidly increasing sequence data from diverse species, and improved techniques for functional tests, we expect studies of molecular evolution to contribute to our understanding of the mechanistic basis of brain evolution.

Overcoming ‘Just-So Stories’ and Phylogenetic Confounds

Evolutionary change in the structure of coding regions and regulatory elements often alters developmental and physiological processes and may ultimately result in neuroanatomical and behavioral differences across species. Although correlation studies on brain evolution have been useful for determining large scale patterns of evolutionary change, there are serious shortcomings. Chief among them, of course, is the causation problem – correlations with ecology or other factors do not prove those factors drove the change [see Shumway, 2008]. The danger is therefore, that correlative results are explained with plausible yet unproven ‘just-so stories’ and that other (including non-adaptive) alternative explanations are ignored [see Gould and Lewontin, 1979]. Another problem for interpreting these kinds of datasets lies in the complexity of life histories animals display and the habitats in which they live. The common focus on associations between brain structure and single behavioral or ecological traits may be inconclusive, as properties of the environment or social system often correlate with Pollen /Hofmann

each other [see also Healy and Rowe, 2007]. For example, in birds and primates, correlations exist between the size of forebrain regions and social group size [Reader and Laland, 2002; Burish et al., 2004; Lefebvre and Sol, 2008], yet the size of forebrain regions has also been associated with many other ecological factors [see Shumway, 2008] as well as developmental constraints with the rest of the brain [Finlay and Darlington, 1995; Finlay et al., 2001]. Until recently, genomic and mechanistic layers of brain evolution were inaccessible, and the field focused on correlating structural differences between taxa within an evolutionary context, often unaware of phylogenetic relationships. However, due to shared ancestry, comparative data sets at any phenotypic level often violate statistical assumptions of independence [Felsenstein, 1985; Harvey and Pagel, 1991]. Thus, in order to draw conclusions from the covariation of traits across taxa, one needs to take into account this phylogenetic non-independence. Assuming the phylogenetic relationships between the species studied are known, one generally accepted method to overcome the effect of shared evolutionary history is to calculate differences in (extant and ancestral) trait values between sister taxa [Felsenstein, 1985; Garland et al., 1992]. Two traits are considered evolutionarily correlated (i.e., change in one trait has been accompanied by change in the other) if these (standardized) differences – or phylogenetically independent contrasts – in one trait significantly covary with contrasts in the other trait. This approach has become very common, but the more basic problem is that a well-resolved phylogeny often does not exist. Even in groups that have been relatively well studied, there are often alternative hypotheses for the

phylogenetic relationships. It is therefore important to conduct comparative analyses for the different phylogenetic hypotheses if a consensus has not yet been reached. Thus comparative studies involving the four approaches presented here must continue to avoid ‘just-so’ stories and correct for phylogenetic confounds.

Conclusion

The novel experimental approaches discussed here allow the integrative study of behavioral, physiological, genetic and genomic changes underlying brain function within an evolutionary framework. By utilizing these approaches together with robust phylogenies, the field of brain evolution can re-invent itself to study all levels of biological organization and ultimately uncover the driving forces, constraining factors and proximate mechanisms that have resulted in the diversity of brains and behaviors as we find them everywhere in the natural world.

Acknowledgements We thank Susan Renn and Caroly Shumway for many fruitful discussions over the years and Abraham Bassan, Kim Hoke, David Kingsley, Ron Oldfield, Phil Reno, Anna Sessa, and an anonymous reviewer for critically reading earlier versions of the manuscript. Supported by NSF grant IBN-021795 and the Institute for Cellular and Molecular Biology (H.A.H.) and an NDSEG fellowship (A.A.P.).

References Abzhanov A, Protas M, Grant BR, Grant PR, Tabin CJ (2004) Bmp4 and morphological variation of beaks in Darwin’s finches. Science 305:1462–1465. Albertson RC, Streelman JT, Kocher TD, Yelick PC (2005) Integration and evolution of the cichlid mandible: the molecular basis of alternate feeding strategies. Proc Natl Acad Sci USA 102:16287–16292. Alcamo EA, Chirivella L, Dautzenberg M, Dobreva G, Fariñas I, Grosschedl R, McConnell SK (2008) Satb2 regulates callosal projection neuron identity in the developing cerebral cortex. Neuron 57:364–377. Barton RA (1998) Visual specialization and brain evolution in primates. Proc R Soc Lond B 265:1933–1937.


Barton RA (2004) Binocularity and brain evolution in primates. Proc Natl Acad Sci USA101: 10113–10115. Barton RA (2006) Olfactory evolution and behavioral ecology in primates. Am J Primatol 68:545–558. Basil JA, Kamil AC, Balda RP, Fite KV (1996) Differences in hippocampal volume among food-storing corvids. Behav Brain Evol 47: 156–164. Bayatti N, Moss JA, Sun L, Ambrose P, Ward JF, Lindsay S, Clowry GJ (2008) Molecular neuroanatomical study of the developing human neocortex from 8 to 17 postconceptional weeks revealing the early differentiation of the subplate and subventricular zone. Cereb Cortex 18:1536–1548.

Berenbrink M, Koldkjaer P, Kepp O, Cossins AR (2005) Evolution of oxygen secretion in fishes and the emergence of a complex physiological system. Science 307:1752–1757. Biegler R, McGregor A, Krebs JR, Healy SD (2001) A larger hippocampus is associated with longer-lasting spatial memory. Proc Natl Acad Sci USA 98:6941–6944. Bolhuis JJ, Macphail EM (2001) A critique of the neuroecology of learning and memory. Trends Cogn Sci 5:426–433. Bridgham JT, Carroll SM, Thornton JW (2006) Evolution of hormone-receptor complexity by molecular exploitation. Science 312: 97– 101. Britanova O, de Juan Romero C, Cheung A, Kwan KY, Schwark M, Gyorgy A, Vogel T,


155

Akopov S, Mitkovski M, Agoston D, Sestan N, Molnár Z, Tarabykin V (2008) Satb2 is a postmitotic determinant for upper-layer neuron specification in the neocortex. Neuron 57:378–392. Britten RJ, Davidson EH (1971) Repetitive and non-repetitive DNA sequences and a speculation on the origins of evolutionary novelty. Q Rev Biol 46: 111–138. Burish MJ, Kueh HY, Wang SH (2004) Brain architecture and social complexity in modern and ancient birds. Brain Behav Evol 63: 107– 124. Cáceres M, Lachuer J, Zapala MA, Redmond JC, Kudo L, Geschwind DH, Lockhart DJ, Preuss TM, Barlow C (2003) Elevated gene expression levels distinguish human from non-human primate brains. Proc Natl Acad Sci USA 100:13030–13035. Cáceres M, Suwyn C, Maddox M, Thomas JW, Preuss TM (2007) Increased cortical expression of two synaptogenic thrombospondins in human brain evolution. Cereb Cortex 17: 2312–2321. Cajal SR (1909–1911) Histologie du Système Nerveux de l’Homme et des Vertébrés, 2 Vols. (trans. L. Azoulay). Madrid: Instituto Ramón y Cajal de C.S.I.C.: 1952–1955. Carroll SB, Grenier JK, Weatherbee SD (2001) From DNA to Diversity: Molecular Genetics and the Evolution of Animal Design. Malden, MA: Blackwell Science. Castro LF, Rasmussen SL, Holland PW, Holland ND, Holland LZ (2006) A Gbx homeobox gene in amphioxus: insights into ancestry of the ANTP class and evolution of the midbrain/ hindbrain boundary. Dev Biol 295:40–51. Caviness VS, Takahashi T, Nowakowski RS (1995) Numbers, time and migration and neocortical neuronogenesis: a general developmental and evolutionary model. Trends Neurosci 18:379–383. Charvet C, Perez K, Striedter GF (2007) How do you make a larger brain? Delayed neurogenesis in parakeets versus bobwhite quail. Soc for Neurosci Abstracts E6:347. Chenn A, Walsh CA (2003) Increased neuronal production, enlarged forebrains and cytoarchitectural distortions in beta-catenin overexpressing transgenic mice. Cereb Cortex 13:599–606. Cheung AF, Pollen AA, Tavare A, DeProto J, Molnár Z (2007) Comparative aspects of cortical neurogenesis in vertebrates. J Anat 211:164–176. Clayton NS (1998) Memory and the hippocampus in food-storing birds: a comparative approach. Neuropharmacology 37: 441–452. Clayton NS, Krebs JR (1995) Memory in foodstoring birds: from behaviour to brain. Curr Opin Neurobiol 5: 149–154. Colosimo PF, Hosemann KE, Balabhadra S, Villarreal G Jr, Dickson M, Grimwood J, Schmutz J, Myers RM, Schluter D, Kingsley DM (2005) Widespread parallel evolution in sticklebacks by repeated fixation of Ectodysplasin alleles. Science 307:1928–1933.

156

Colosimo PF, Peichel CL, Nereng K, Blackman BK, Shapiro MD, Schluter D, Kingsley DM (2004) The genetic architecture of parallel armor plate reduction in threespine sticklebacks. PLoS Biol 2:E109. Cretekos CJ, Wang Y, Green ED, Martin JF, Rasweiler JJ 4th, Behringer RR (2008) Regulatory divergence modifies limb length between mammals. Genes Dev 22:141–151. Cubelos B, Sebastián-Serrano A, Kim S, Moreno-Ortiz C, Redondo JM, Walsh CA, Nieto M (2007) Cux-2 controls the proliferation of neuronal intermediate precursors of the cortical subventricular zone. Cereb Cortex doi:10.1093/cercor/bhm199. Devoogd TJ, Krebs JR, Healy SD, Purvis A (1993) Relations between song repertoire size and the volume of brain nuclei related to song: comparative evolutionary analyses amongst oscine birds. Proc Biol Sci USA 254: 75–82. Dong H, Martin MV, Colvin J, Ali Z, Wang L, Lu L, Williams RW, Rosen GD, Csernansky JG, Cheverud JM (2007) Quantitative trait loci linked to thalamus and cortex gray matter volumes in BXD recombinant inbred mice. Heredity 99:62–69. Ekström P, Johnsson CM, Ohlin LM (2001) Ventricular proliferation zones in the brain of an adult teleost fish and their relation to neuromeres and migration (secondary matrix) zones. J Comp Neurol 436:92–110. Emery NJ (2006) Cognitive ornithology: the evolution of avian intelligence. Phil Trans R Soc Lond B 361:23–43. Evans PD, Anderson JR, Vallender EJ, Gilbert SL, Malcom CM, Dorus S, Lahn BT (2004a) Adaptive evolution of ASPM, a major determinant of cerebral cortical size in humans. Hum Mol Genet 13:489–494. Evans PD, Anderson JR, Vallender EJ, Choi SS, Lahn BT (2004b) Reconstructing the evolutionary history of microcephalin, a gene controlling human brain size. Hum Mol Genet 13:1139–1145. Farris SM (2008) Evolutionary convergence of higher brain centers spanning the protostome-deuterostome boundary. Brain Behav Evol 72:106–122. Felsenstein J (1985) Phylogenies and the comparative method. Am Nat 125:1–15. Ferrari MB, McAnelly ML, Zakon HH (1995) Individual variation in and androgen-modulation of the sodium current in electric organ. J Neurosci 15:4023–4032. Fink S, Excoffier L, Heckel G (2006) Mammalian monogamy is not controlled by a single gene. Proc Natl Acad Sci USA 103:10956–10960. Finlay BL, Darlington RB (1995) Linked regularities in the development and evolution of mammalian brains. Science 268:1578–1584. Finlay BL, Darlington RB, Nicastro N (2001) Developmental structure in brain evolution. Behav Brain Sci 24:263–308. Freimer NB, Service SK, Ophoff RA, Jasinska AJ, McKee K, Villeneuve A, Belisle A, Bailey JN, Breidenthal SE, Jorgensen MJ, Mann JJ, Cantor RM, Dewar K, Fairbanks LA (2007) A


quantitative trait locus for variation in dopamine metabolism mapped in a primate model using reference sequences from related species. Proc Natl Acad Sci USA 104: 15811– 15816. Fukuchi-Shimogori T, Grove EA (2001) Neocortex patterning by the secreted signaling molecule FGF8. Science 294:1071–1074. Garland T, Harvey PH, Ives AR (1992) Procedures for the analysis of comparative data using phylogenetically independent contrasts. Syst Biol 41: 18–32. Genissel A, McIntyre LM, Wayne ML, Nuzhdin SV (2008) Cis and trans regulatory effects contribute to natural variation in transcriptome of Drosophila melanogaster. Mol Biol Evol 2008 25:101–110. George AL (2005) Inherited disorders of voltagegated sodium channels. J Clin Invest 115: 1990–1999. Gilad Y, Rifkin SA, Bertone P, Gerstein M, White KP (2005) Multi-species microarrays reveal the effect of sequence divergence on gene expression profiles. Genome Res 15: 674–680. Gleason JM, Ritchie MG (2004) Do quantitative trait loci (QTL) for a courtship song difference between Drosophila simulans and D. sechellia coincide with candidate genes and intraspecific QTL? Genetics 166: 1303– 1311. Gould SJ, Lewontin RC (1979) The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme. Proc R Soc Lond B 205:581–598. Grant PR, Grant BR (2006) Evolution of character displacement in Darwin’s finches. Science 313:224–226. Hale ME, Long JH Jr, McHenry MJ, Westneat MW (2002) Evolution of behavior and neural control of the fast-start escape response. Evolution 56:993–1007. Hammock EA, Young LJ (2004) Functional microsatellite polymorphism associated with divergent social structure in vole species. Mol Biol Evol 21: 1057–1063. Hammock EA, Young LJ (2005) Microsatellite instability generates diversity in brain and sociobehavioral traits. Science 308: 1630– 1634. Hampton RR, Sherry DF, Shettleworth SJ, Khurgle M, Ivy G (1995) Hippocampal volume and food-storing behavior are related in parids. Brain Behav Evol 45:54–61. Harvey PH, Krebs JR (1990) Comparing brains. Science 249:140–146. Harvey PH, Pagel MD (1991) The Comparative Method in Evolutionary Biology. Oxford, UK: Oxford University Press. Harvey-Girard E, Dunn RJ, Maler L (2007) Regulated expression of N-methyl-D-aspartate receptors and associated proteins in teleost electrosensory system and telencephalon. J Comp Neurol 505:644–668. Healy SD, Krebs JR (1992) Food-storing and the hippocampus in corvids: amount and volume are correlated. Proc R Soc Lond B 248: 241–245.

Pollen /Hofmann

Healy SD, Krebs JR (1996) Food storing and the hippocampus in Paridae. Brain Behav Evol 47:195–199. Healy SD, Rowe C (2007) A critique of comparative studies of brain size. Proc Biol Sci 274: 453–464. Hill RS, Walsh CA (2005) Molecular insights into human brain evolution. Nature 437:64–67. Hoekstra HE, Coyne JA (2007) The locus of evolution: evo devo and the genetics of adaptation. Evolution 61:995–1016. Hofmann HA (2003) Functional genomics of neural and behavioral plasticity. J Neurobiol 54:272–282. Hoheisel JD (2006) Microarray technology: beyond transcript profiling and genotype analysis. Nat Rev Genet 7:200–210. Holland LZ, Short S (2008) Gene duplication, co-option and recruitment during the origin of the vertebrate brain from the invertebrate chordate brain. Brain Behav Evol 72:91–105. Hopkins CD (1999) Design features for electric communication. J Exp Biol 202: 1217–1228. Huber R, van Staaden MJ, Kaufman LS, Liem KF (1997) Microhabitat use, trophic patterns, and the evolution of brain structure in African cichlids. Brain Behav Evol 50: 167–182 Iwaniuk AN, Hurd PL (2005) The evolution of cerebrotypes in birds. Brain Behav Evol 65: 215–230. Iwaniuk AN, Clayton DH, Wylie DR (2006) Echolocation, vocal learning, auditory localization and the relative size of the avian auditory midbrain nucleus (MLd). Behav Brain Res 167:305–317. Khaitovich P, Enard W, Lachmann M, Pääbo S (2006) Evolution of primate gene expression. Nat Rev Genet 7:693–702. Khaitovich P, Muetzel B, She X, Lachmann M, Hellmann I, Dietzsch J, Steigele S, Do HH, Weiss G, Enard W, Heissig F, Arendt T, Nieselt-Struwe K, Eichler EE, Pääbo S (2004) Regional patterns of gene expression in human and chimpanzee brains. Genome Res 14: 1462–1473. Khaitovich P, Pääbo S, Weiss G (2005) Toward a neutral evolutionary model of gene expression. Genetics 170:929–939. King MC, Wilson AC (1975) Evolution at two levels in humans and chimpanzees. Science 188:107–116. Kocher TD (2004) Adaptive evolution and explosive speciation: the cichlid fish model. Nat Rev Genet 5:288–298. Krebs JR, Sherry DF, Healy SD, Perry VH, Vaccarino AL (1989) Hippocampal specialization of food-storing birds. Proc Natl Acad Sci USA 86:1388–1392. Kriegstein AR, Noctor S, Martínez-Cerdeno V (2006) Patterns of neural stem and progenitor cell division may underlie evolutionary cortical expansion. Nat Rev Neurosci 7:883–890. Lefebvre L, Sol D (2008) Brains, lifestyles and cognition: are there general trends? Brain Behav Evol 72:135–144.


Liem KF (1979) Modularity multiplicity in the feeding mechanism in cichlid fishes, as exemplified by the invertebrate pickers of Lake Tanganyika. J Zool Lond 189:93–125. Lim MM, Young LJ (2006) Neuropeptidergic regulation of affiliative behavior and social bonding in animals. Horm Behav 50: 506– 517. Lim MM, Wang Z, Olazábal DE, Ren X, Terwilliger EF, Young LJ (2004) Enhanced partner preference in a promiscuous species by manipulating the expression of a single gene. Nature 429:754–757. Lucas JR, Brodin A, de Kort SR, Clayton NS (2004) Does hippocampal size correlate with the degree of caching specialization? Proc R Soc Lond B 271:2423–2429. Macphail EM, Bolhuis JJ (2001) The evolution of intelligence: adaptive specializations versus general process. Biol Rev Camb Philos Soc 76:341–364. Maler L, Hincke MT (1999) Distribution of calcium/calmodulin-dependent kinase 2 in the brain of Apteronotus leptorhynchus. J Comp Neurol 408:177–203. Martínez-Cerdeno V, Noctor SC, Kriegstein AR (2006) The role of intermediate progenitor cells in the evolutionary expansion of the cerebral cortex. Cereb Cortex 16(suppl 1):152– 161. Messier W, Stewart CB (1997) Episodic adaptive evolution of primate lysozymes. Nature 385: 151–154. Miller CT, Beleza S, Pollen AA, Schluter D, Kittles RA, Shriver MD, Kingsley DM (2007) cis-Regulatory changes in Kit ligand expression and parallel evolution of pigmentation in sticklebacks and humans. Cell 131: 1179– 1189. Moehring AJ, Mackay TF (2004) The quantitative genetic basis of male mating behavior in Drosophila melanogaster. Genetics 167: 1249–1263. Mouse Genome Sequencing Consortium, Waterston RH, Lindblad-Toh K, Birney E, Rogers J, Abril JF, Agarwal P, Agarwala R, Ainscough R, Alexandersson M, An P, Antonarakis SE, Attwood J, Baertsch R, Bailey J, Barlow K, Beck S, Berry E, Birren B, Bloom T, Bork P, Botcherby M, Bray N, Brent MR, Brown DG, Brown SD, Bult C, Burton J, Butler J, Campbell RD, Carninci P, Cawley S, Chiaromonte F, Chinwalla AT, Church DM, Clamp M, Clee C, Collins FS, Cook LL, Copley RR, Coulson A, Couronne O, Cuff J, Curwen V, Cutts T, Daly M, David R, Davies J, Delehaunty KD, Deri J, Dermitzakis ET, Dewey C, Dickens NJ, Diekhans M, Dodge S, Dubchak I, Dunn DM, Eddy SR, Elnitski L, Emes RD, Eswara P, Eyras E, Felsenfeld A, Fewell GA, Flicek P, Foley K, Frankel WN, Fulton LA, Fulton RS, Furey TS, Gage D, Gibbs RA, Glusman G, Gnerre S, Goldman N, Goodstadt L, Grafham D, Graves TA, Green ED, Gregory S, Guigó R, Guyer M, Hardison RC, Haussler D, Hayashizaki Y, Hillier LW, Hinrichs A, Hlavina W, Holzer

T, Hsu F, Hua A, Hubbard T, Hunt A, Jackson I, Jaffe DB, Johnson LS, Jones M, Jones TA, Joy A, Kamal M, Karlsson EK, Karolchik D, Kasprzyk A, Kawai J, Keibler E, Kells C, Kent WJ, Kirby A, Kolbe DL, Korf I, Kucherlapati RS, Kulbokas EJ, Kulp D, Landers T, Leger JP, Leonard S, Letunic I, Levine R, Li J, Li M, Lloyd C, Lucas S, Ma B, Maglott DR, Mardis ER, Matthews L, Mauceli E, Mayer JH, McCarthy M, McCombie WR, McLaren S, McLay K, McPherson JD, Meldrim J, Meredith B, Mesirov JP, Miller W, Miner TL, Mongin E, Montgomery KT, Morgan M, Mott R, Mullikin JC, Muzny DM, Nash WE, Nelson JO, Nhan MN, Nicol R, Ning Z, Nusbaum C, O’Connor MJ, Okazaki Y, Oliver K, Overton-Larty E, Pachter L, Parra G, Pepin KH, Peterson J, Pevzner P, Plumb R, Pohl CS, Poliakov A, Ponce TC, Ponting CP, Potter S, Quail M, Reymond A, Roe BA, Roskin KM, Rubin EM, Rust AG, Santos R, Sapojnikov V, Schultz B, Schultz J, Schwartz MS, Schwartz S, Scott C, Seaman S, Searle S, Sharpe T, Sheridan A, Shownkeen R, Sims S, Singer JB, Slater G, Smit A, Smith DR, Spencer B, Stabenau A, Stange-Thomann N, Sugnet C, Suyama M, Tesler G, Thompson J, Torrents D, Trevaskis E, Tromp J, Ucla C, Ureta-Vidal A, Vinson JP, Von Niederhausern AC, Wade CM, Wall M, Weber RJ, Weiss RB, Wendl MC, West AP, Wetterstrand K, Wheeler R, Whelan S, Wierzbowski J, Willey D, Williams S, Wilson RK, Winter E, Worley KC, Wyman D, Yang S, Yang SP, Zdobnov EM, Zody MC, Lander ES (2002) Initial sequencing and comparative analysis of the mouse genome. Nature 420:520–562. Northcutt RG (2008) Forebrain evolution in bony fishes. Brain Res Bull 75:191–205. Odling-Smee L, Braithwaite V (2003) The influence of habitat stability on landmark use during spatial learning in the three-spined stickleback. Anim Behav 65: 701–707. Osada N, Kohn MH, Wu CI (2006) Genomic inferences of the cis-regulatory nucleotide polymorphisms underlying gene expression differences between Drosophila melanogaster mating races. Mol Biol Evol 23:1585–1591. Peichel CL (2005) Fishing for the secrets of vertebrate evolution in threespine sticklebacks. Dev Dyn 234:815–823. Pennacchio LA, Ahituv N, Moses AM, Prabhakar S, Nobrega MA, Shoukry M, Minovitsky S, Dubchak I, Holt A, Lewis KD, PlajzerFrick I, Akiyama J, De Val S, Afzal V, Black BL, Couronne O, Eisen MB, Visel A, Rubin EM (2006) In vivo enhancer analysis of human conserved non-coding sequences. Nature 444:499–502. Pitkow LJ, Sharer CA, Ren X, Insel TR, Terwilliger EF, Young LJ (2001) Facilitation of affiliation and pair-bond formation by vasopressin receptor gene transfer into the ventral forebrain of a monogamous vole. J Neurosci 21:7392–7396. Pollard KS, Salama SR, Lambert N, Lambot MA, Coppens S, Pedersen JS, Katzman S, King B,


157

Onodera C, Siepel A, Kern AD, Dehay C, Igel H, Ares M Jr, Vanderhaeghen P, Haussler D (2006) An RNA gene expressed during cortical development evolved rapidly in humans. Nature 443:167–172. Pollen AA, Dobberfuhl AP, Scace J, Igulu MM, Renn SC, Shumway CA, Hofmann HA (2007) Environmental complexity and social organization sculpt the brain in Lake Tanganyikan cichlid fish. Brain Behav Evol 70: 21–39. Ponting CP (2006) A novel domain suggests a ciliary function for ASPM, a brain size determining gene. Bioinformatics 22: 1031–1035. Prabhakar S, Noonan JP, Pääbo S, Rubin EM (2006) Accelerated evolution of conserved noncoding sequences in humans. Science 314:786. Protas ME, Hersey C, Kochanek D, Zhou Y, Wilkens H, Jeffery WR, Zon LI, Borowsky R, Tabin CJ (2006) Genetic analysis of cavefish reveals molecular convergence in the evolution of albinism. Nat Genet 38:107–111. Rakic P (1995) A small step for the cell, a giant leap for mankind: a hypothesis of neocortical expansion during evolution. Trends Neurosci 18:383–388. Reader SM, Laland KN (2002) Social intelligence, innovation, and enhanced brain size in primates. Proc Natl Acad Sci USA 99: 4436–4441. Reboreda JC, Clayton NS, Kacelnik A (1996) Species and sex differences in hippocampus size in parasitic and non-parasitic cowbirds. Neuroreport 7:505–508. Reiner A, Yamamoto K, Karten HJ (2005) Organization and evolution of the avian forebrain. Anat Rec A Discov Mol Cell Evol Biol 287:1080–1102. Renn SC, Aubin-Horth N, Hofmann HA (2004) Biologically meaningful expression profiling across species using heterologous hybridization to a cDNA microarray. BMC Genomics 5:42. Renn SC, Aubin-Horth N, Hofmann HA (2008) Fish and Chips: Functional genomics of social plasticity in an African cichlid fish. J Exp Biol 211:3041–3056. Rilling JK, Glasser MF, Preuss TM, Ma X, Zhao T, Hu X, Behrens TE (2008) The evolution of the arcuate fasciculus revealed with comparative DTI. Nat Neurosci 11:426–428. Roth G, Wullimann MF (2000) Brain Evolution and Cognition. New York: John Wiley and Sons, Inc. Sasaki T, Nishihara H, Hirakawa M, Fujimura K, Tanaka M, Kokubo N, Kimura-Yoshida C, Matsuo I, Sumiyama K, Saitou N, Shimogori T, Okada N (2008) Possible involvement of SINEs in mammalian-specific brain formation. Proc Natl Acad Sci USA 105: 4220– 4225.

158

Shapiro MD, Marks ME, Peichel CL, Blackman BK, Nereng KS, Jónsson B, Schluter D, Kingsley DM (2004) Genetic and developmental basis of evolutionary pelvic reduction in threespine sticklebacks. Nature 428: 717–723. Sherry DF, Vaccarino AL, Buckenham K, Herz RS (1989) The hippocampal complex of foodstoring birds. Brain Behav Evol 34:308–317. Shettleworth SJ (2003) Memory and hippocampal specialization in food-storing birds: challenges for research on comparative cognition. Brain Behav Evol 62:108–116. Shumway CA (2008) Habitat complexity, brain and behavior. Brain Behav Evol 72:123–134. Sidman RL, Rakic P (1973) Neuronal migration, with special reference to developing human brain: a review. Brain Res 62:1–35. Smart IH, Dehay C, Giroud P, Berland M, Kennedy H (2002) Unique morphological features of the proliferative zones and postmitotic compartments of the neural epithelium giving rise to striate and extrastriate cortex in the monkey. Cereb Cortex 12:37–53. Spencer KA, Buchanan KL, Leitner S, Goldsmith AR, Catchpole CK (2005) Parasites affect song complexity and neural development in a songbird. Proc Biol Sci 272: 2037–2043. Stoddard PK, Zakon HH, Markham MR, McAnelly L (2006) Regulation and modulation of electric waveforms in gymnotiform electric fish. J Comp Physiol A 192: 613– 624. Streelman JT, Webb JF, Albertson RC, Kocher TD (2003) The cusp of evolution and development: a model of cichlid tooth shape diversity. Evol Dev 5: 600–608. Streelman JT, Albertson RC (2006) Evolution of novelty in the cichlid dentition. J Exp Zool B Mol Dev Evol 306:216–226. Striedter GF, Keefer BP (2000) Cell migration and aggregation in the developing telencephalon: pulse-labeling chick embryos with bromodeoxyuridine. J Neurosci 20: 8021– 8030. Talbot CJ, Nicod A, Cherny SS, Fulker DW, Collins AC, Flint J (1999) High-resolution mapping of quantitative trait loci in outbred mice. Nat Genet 21:305–308. Torres TT, Metta M, Ottenwälder B, Schlötterer C (2008) Gene expression profiling by massively parallel sequencing. Genome Res 18: 172–177. Vallender E (2008) Exploring the origins of the human brain through molecular evolution. Brain Behav Evol 72:168–177.


Vera JC, Wheat CW, Fescemyer HW, Frilander MJ, Crawford DL, Hanski I, Marden JH (2008) Rapid transcriptome characterization for a nonmodel organism using 454 pyrosequencing. Mol Ecol 17:1636–1647. Wainwright PC, Ferry-Graham LA, Waltzek TB, Carroll AM, Hulsey CD, Grubich JR (2001) Evaluating the use of ram and suction during prey capture by cichlid fishes. J Exp Biol 204: 3039–3051. Wang SS-H (2008) Functional tradeoffs in axonal scaling: implications for brain function. Brain Behav Evol 72:159–167. Williams RW, Strom RC, Goldowitz D (1998) Natural variation in neuron number in mice is linked to a major quantitative trait locus on Chr 11. J Neurosci 18:138–146. Winslow JT, Hastings N, Carter CS, Harbaugh CR, Insel TR (1993) A role for central vasopressin in pair bonding in monogamous prairie voles. Nature 365: 545–548. Wright D, Nakamichi R, Krause J, Butlin RK (2006) QTL analysis of behavioral and morphological differentiation between wild and laboratory zebrafish (Danio rerio). Behav Genet 36:271–284. Wu SX, Goebbels S, Nakamura K, Nakamura K, Kometani K, Minato N, Kaneko T, Nave KA, Tamamaki N (2005) Pyramidal neurons of upper cortical layers generated by NEX-positive progenitor cells in the subventricular zone. Proc Natl Acad Sci USA 102: 17172– 17177. Yalcin B, Willis-Owen SA, Fullerton J, Meesaq A, Deacon RM, Rawlins JN, Copley RR, Morris AP, Flint J, Mott R (2004) Genetic dissection of a behavioral quantitative trait locus shows that Rgs2 modulates anxiety in mice. Nat Genet 36:1197–1202. Young LJ, Nilsen R, Waymire KG, MacGregor GR, Insel TR (1999) Increased affiliative response to vasopressin in mice expressing the V1a receptor from a monogamous vole. Nature 400:766–768. Zakon HH, Lu Y, Zwickl DJ, Hillis DM (2006) Sodium channel genes and the evolution of diversity in communication signals of electric fishes: convergent molecular evolution. Proc Natl Acad Sci USA103:3675–3680. Zakon HH, Zwickl DJ, Lu Y, Hillis DM (2008) Molecular evolution of communication signals in electric fish. J Exp Biol 211: 1814– 1818. Zecevic N, Rakic P (2001) Development of layer I neurons in the primate cerebral cortex. J Neurosci 21:5607–5619. Zupanc GK (2006) Neurogenesis and neuronal regeneration in the adult fish brain. J Comp Physiol A 192:649–670.

Pollen /Hofmann



Functional Tradeoffs in Axonal Scaling: Implications for Brain Function Samuel S.-H. Wang Department of Molecular Biology and Princeton Neuroscience Institute, Princeton University, Princeton, N.J., USA

Key Words Optimization ⴢ Conduction ⴢ Evolution ⴢ Spindle cells

Abstract Like electrical wires, axons carry signals from place to place. However, unlike wires, because of the electrochemical mechanisms for generating and propagating action potentials, the performance of an axon is strongly linked to the costs of its construction and operation. As a consequence, the architecture of brain wiring is biophysically constrained to trade off speed and energetic efficiency against volume. Because the biophysics of axonal conduction is well studied, this tradeoff is amenable to quantitative analysis. In this framework, an examination of axon tract composition can yield insights into neural circuit function in regard to energetics, processing speed, spike timing precision, and average rates of neural activity. Copyright © 2008 S. Karger AG, Basel

Introduction

An axon conveys action potentials from its parent neuron to its targets by a combination of passive spread of depolarization and active re-amplification by voltagedependent ion channels. The resulting conduction speed © 2008 S. Karger AG, Basel 0006–8977/08/0722–0159$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com


can range from tens of centimeters per second to several hundred meters per second. Consequently, the time taken for an action potential traveling from its point of origin to the axon’s terminal arborization is variable, and can be enough to contribute substantially to delays in neural processing. Conduction speed is a function of the biophysical properties of axons. Decades of work have led to a detailed quantitative understanding of axonal conduction in terms of biophysics [Hodgkin, 1971] and bioenergetics [Ritchie, 1995]. This history makes axons a particularly favorable target for an analysis of the tradeoffs among speed and construction, length, and energy costs [Ringo et al., 1994; Harrison et al., 2002; Aboitiz et al., 2003; Wen and Chklovskii, 2005; Wang et al., 2008]. Vertebrate axons can respond to selection pressure for speed with two adaptations that increase conduction velocity (fig. 1): myelination and increased diameter. Myelin, found almost exclusively in vertebrate axons, reduces capacitance and leak, with two consequences for conduction speed. First, conduction is sped up by reduced leak across the membrane, which allows depolarization to spread further. Relative reduced leak across the membrane also underlies increases in conduction speed that occur with increasing axon diameter. Second, the reduction in capacitance decreases the amount of time needed to depolarize the membrane to threshold. As a result, a Dr. Samuel Wang Department of Molecular Biology, Princeton University Lewis Thomas Laboratory, Washington Road Princeton, NJ 08544 (USA) Tel. +1 (609) 258 0388, Fax +1 (609) 258 1028, E-Mail [email protected]

Unmyelinated axons 0.5 μm wide fast-conducting low capacitance large volume

1 μm 10 μm

meter diameter in myelinated axons [Hursh, 1939; Ritchie, 1995], but axon volume is proportional to the square of the diameter. Indiscriminate scaling-up of all axons is thus impractical as a means of maximizing overall processing speed. Instead, the diameters of axons are likely to represent a tradeoff of functional role, which might impose speed-of-processing requirements, against the costs of construction and/or generating action potential activity. One consequence of long-distance conduction is potential imprecision in the arrival time of action potentials at the end of the axon. Delays in spike times, and therefore potential timing imprecision, should increase with transmission distance and be the most severe in unmyelinated axons because of their slow conduction velocities. Transmission speeds can increase or decrease by up to 20% if an axon has fired recently [Swadlow, 2000], presumably because of changes in the activation/inactivation state of voltage-gated channels or accumulation of intracellular sodium. An action potential traveling in an unmyelinated axon at 0.3 m/s across a 15-cm human brain would take half a second to arrive at its destination, with variation in arrival time of up to 100 ms. Similar reasoning applies to another unmyelinated pathway, C fibers that extend from cutaneous receptors to transmit painful stimuli. In this case the path length is quite large, essentially ruling out the possibility of transmitting information in the form of spike timing. Mechanisms that would be difficult for unmyelinated axons to implement over cross-brain distances include spike timing-dependent plasticity and detection of synchronous action potentials arriving from multiple sources. Thus unmyelinated long-distance axons might encode signals in the form of firing rates and brief bursts rather than precisely Wang

timed spikes, and represent a case in which a premium is placed on volume minimization with the accompanying penalties of higher metabolic rate and slow, imprecise transmission. In contrast, arrival times in myelinated axons would vary at most by a few milliseconds. Myelinated axons are likely to play functional roles requiring speed or precise timing. Their low capacitance and intrinsically precise timing allows them to transmit information with much less energy expenditure than unmyelinated axons. Communication over rapidly conducting axons might be necessary to synchronize activity between distant brain areas [Bush and Sejnowski, 1996; Swadlow, 2000; Varela et al., 2001].

Specialized Axon Populations in Mammalian Nervous Systems

The strong linkage between axonal size, speed, and energetics suggests the possibility that the microstructure of an axon population can provide clues about the axons’ functional role. I will illustrate this possibility using examples drawn from the literature and from recent work in my own laboratory. I will focus in particular on how axons vary among mammals of different sizes. Particular attention will be paid to scaling with respect to brain and body size. Increases in these parameters correspond to longer conduction distances and may provide selection pressure to make certain axons wider. Example 1: Cerebellar Parallel Fibers and Sparse Coding Cerebellar parallel fibers (PFs) are among the thinnest known vertebrate axons and represent an extreme anatomical adaptation. A systematic examination of their properties across mammalian species from mouse to macaque [Wyatt et al., 2005] shows that the great majority of parallel fibers are unmyelinated with diameters of 0.2– 0.3 ␮m, consistent with the idea that they are evolutionarily selected for compactness. Indeed, PFs are quite closely packed: each PF forms hundreds of synapses on Purkinje neurons and interneurons, and each Purkinje cell receives approximately 200,000 PF synapses. This massive divergence and convergence motivated Marr [1969] and Albus [1971] to propose that the cerebellar cortex acts as a pattern detector in which the number of Purkinje cell inputs is very high as part of a sparse coding scheme. In this ‘perceptron’ model, maximizing the number of PFs would maximize combinatorial possibilities. Functional Tradeoffs in Axon Scaling

Therefore PFs may be narrow in order to fit large numbers of axons into a limited space. The packing density of unmyelinated PFs creates a problem of high metabolic cost both because of the high numerical density of PFs and because of their high peraxon capacitance. However, the cerebellum does not use more energy than other brain regions [Sokoloff, 1996], suggesting that unmyelinated PFs fire sparsely and at low average rates. This would be consistent with the suggestion that sensory input triggers activity in a sparse population of PFs [Marr, 1969; Albus, 1971]. Consistent with this idea, low spontaneous firing rates and burstlike sensory responses in cerebellar granule cells have been observed in vivo in decerebrated animals [Eccles et al., 1966] and under anesthesia [Shambes et al., 1978; Chadderton et al., 2004; Jörntell and Ekerot, 2006]. PFs are very slow-conducting, with a measured conduction speed of 0.3 m/s [Eccles et al., 1967]. This conduction velocity can be converted to conduction times in combination with PF lengths, which may scale up weakly with brain size [from 2.8 mm in mouse, Soha et al., 1997, to 5.7–6.0 mm in cat and macaque, Brand et al., 1976; Mugnaini, 1983]. Halving the PF length to account for the fact that conduction begins at the T-junction gives maximum conduction times of 5–10 ms. A long conduction time also implies a certain degree of variability. In PFs, conduction speed varies by as much as 20% from trial to trial [Gardner-Medwin, 1972; Llinás, 1982]. Thus any processing operations requiring submillisecond precision would be likely to take place at an earlier stage of processing. In addition to the preponderance of unmyelinated PFs, myelinated parallel fibers have been found in marmoset, cat and macaque in the deep regions nearest the Purkinje cell layer. These axons are 0.4–1.1 ␮m wide, have expected conduction times of 0.5–1.0 ms, and may convey fast feedfoward inhibition via basket cells to Purkinje cells. Alternately, myelination of PFs might not have a special functional role. Because PF diameters scale up weakly with brain size, a small subpopulation of PFs might simply exceed a size threshold for attracting myelination. Example 2: Afferent Axons to Neocortical Area MT: Precise Timing Differences for Motion Detection? In area MT, which is thought to play an important role in the processing of moving visual stimuli, the axonal projection from V1 is unusually rich in very large axons up to 3 ␮m in diameter with conduction times of about 1.3 ms [Movshon and Newsome, 1996]. Rapid conduction might reduce variability in conduction time, a potentially important component of a motion detection mechaBrain Behav Evol 2008;72:159–167

161

Widest optic nerve axon (μm)

2

10 9 8 7 6 5 4 3

2

0.1

1

10 Brain weight (g)

100

1,000

Fig. 2. Increase in widest optic nerve axon diameter with brain size. The widest reported axon for a variety of species plotted against brain size. The fitted log-log slope is +0.18 8 0.03 [adapted from Wang et al., 2008].

nism that relies on precise spike timing. The Reichardt detector, a delay-based model for responding to moving visual stimuli, relies on a disynaptic connection in which two locations in the visual world project to a common target, and information from one location arrives after a fixed delay. A detection mechanism in the target that responds to the coincidence of the two inputs can then report whether the undelayed line is activated after the other by the delay interval. In area MT, axons from area V1 terminate in layer 4 with a collateral termination in layer 6 of nearby regions. Because layer 6 neurons project to layer 4, a given layer 4 cell might therefore receive a monosynaptic connection from one location and disynaptic connections from nearby locations in visual space [Rockland, 2002]. The difference in timing along the two paths could provide a delay line. In such an anatomical arrangement, wide axons in the V1-MT projection would allow delays along the two paths to be precise relative to one another. For the timing of visual events to be preserved in MT, stages of processing earlier than V1 also need to be precisely timed. V1-MT axons come from Meynert cells [Tigges et al., 1981; Lia et al., 2003], which may receive input from magnocellular layers of the lateral geniculate nucleus. Magnocellular layers in turn receive input from M-cells, large ganglion cells of the retina. Notably, the widest observed axons of the optic nerve scale up in diameter with path length from the eye to the brain (fig. 2) 162


so that the shortest estimated eye-to-brain conduction times are considerably less than 1 ms, even in horses and bottlenose dolphins [Wang et al., 2008]. If M-cells gave rise to the widest optic nerve axons, they could constitute the first step of a rapid, precisely timed pathway reaching from the retina to area MT. In the future, this speculation is testable by applying trans-synaptic tracing methods such as viral transneuronal tracing [Ekstrand et al., 2008] to large-brained mammals such as macaque. After infection of retinal M-cells or magnocellular thalamic neurons, a specialized motion detection pathway should be visible in the form of expression in MT-projecting neurons in area V1. An additional question requiring investigation in this model is the role of synaptic delays, which in the retina can reach tens of milliseconds.

Tradeoffs in a Mixed Axon Population: Neocortical White Matter

Within a given species, metabolic rates at different neocortical locations in white matter are similar. The range of observed values is approximately 1.2-fold in mouse [Nowaczyk and Des Rosiers, 1981], in rat [Sokoloff et al., 1977; Collins et al., 1987], and in macaque [Kennedy et al., 1978; Shapiro et al., 1978]. In addition, brain metabolic rates are also remarkably independent of waking or cognitive state [reviewed in Sokoloff, 1996]. Metabolic rate may be limited by the energy-supplying capacity of the circulatory network [West et al., 1997]. In this way, neocortical white matter architecture would have a limited energy budget within which to optimize both volume [Hsu et al., 1998; Chklovskii, 2004] and conduction time [Ringo et al., 1994]. Among different species, neocortical white matter brain metabolic rate is more variable. In general, tissuespecific metabolic rates decrease according to orderly power laws with respect to body size for up to five orders of magnitude [West et al., 1997]. Such a trend is observed for the per-gram neocortical white matter metabolic rate. Total white matter volume also scales relative to gray matter volume [Hofman, 1988], in this case with an increasing, supralinear power law. These macroscopic scaling relationships suggest that white matter architecture might be constrained by parameters that scale with body or brain size [Changizi, 2001; Zhang and Sejnowski, 2001; Harrison et al., 2002]. Because white matter is composed largely of densepacked axons and the per-axon metabolic cost of spike Wang

firing is strongly linked to axon type and diameter, knowing the abundance and size distribution of unmyelinated and myelinated axons allows an estimate to be made of the maximum supportable firing rate in white matter. We have examined the corpus callosum, which is composed entirely of long-distance axons that are aligned in parallel and therefore easily measured (fig. 3a). Past measurements, done under nonuniform fixation conditions, have reported widely diverging values for the fraction of myelinated axons even within a single species [reviewed in Olivares et al., 2001]. A likely source of error in many surveys is the use of brains preserved by postmortem immersion in fixative, which can lead to tissue degradation, especially in small, unmyelinated fibers. To resolve this problem, we made ultrastructural measurements in 6 species from tissue prepared by uniform perfusion and tissue preparation methods; and additional measurements in 8 more species using experimental and analytical methods that were relatively resistant to technical error and misinterpretation. Our measurements suggest two properties that may be invariant across species with respect to brain size: shortest cross-brain conduction time and maximum supportable average firing rate. If animals are selected for fast cross-brain impulse conduction, then callosal myelination should be more prevalent in larger-brained mammals. In larger brains, we have found a clear tendency for more callosal axons to be myelinated, ranging from least shrew (10% myelinated axons) and mouse (33%) to macaque (70%). The average fraction of myelinated axons increases monotonically with brain diameter (fig. 3b), supporting the idea that longer conduction distances are associated with the adaptation of increased myelination. Even considering myelinated axons alone, size distributions differ systematically from smaller to larger brains. Callosal axons wider than 2 ␮m are a feature of brains heavier than approximately 100 g, including marmoset, cat, macaque, cetaceans, and great apes, including humans. The widest callosal axons scale almost linearly with brain diameter, in contrast to the mean diameters of unmyelinated axons and myelinated axons overall, which vary little across species. Because conduction speed among myelinated axons is linearly proportional to axon diameter [Hursh, 1939], the shortest cross-brain conduction times are therefore similar across species, less than 5 ms in all species examined (fig. 3c). These short conduction times suggest that, for a specialized population of large myelinated axons, a major selection factor has been a need for fast conduction.

Axon size distributions can also be used to estimate the energetic costs of generating electrical activity. Electrically speaking, each action potential consists of sufficient sodium entry to discharge the membrane capacitance, followed by a repolarization current. Thus the amount of charge required to fire a spike can be estimated by calculating the axonal capacitance. This can then be converted to a metabolic rate by calculating how much ATP, and therefore glucose, is needed to restore the supporting ionic gradients via sodium and potassium transport. White matter is an advantageous substrate for such a calculation because, unlike gray matter, it lacks other signaling processes such as neurotransmitter release and handling, the generation of synaptic potentials and summation, and second messenger signaling [Attwell and Laughlin, 2001]. A particularly interesting parameter comes from dividing the observed per-gram energy consumption with the predicted per-gram, per-action-potential cost. The ratio of these quantities is approximately 14 spikes/s across a variety of mammals ranging in brain size from shrew to macaque (fig. 3d). However, not all metabolic energy is converted to action potential firing. A substantial fraction of the brain’s metabolism continues in the absence of electrical activity; the rate of glucose consumption drops by no more than half under conditions such as coma, barbiturate anesthesia, and ouabain blockade of sodium pump activity [reviewed in Sokoloff, 1996]. A more conservative assumption, that half the metabolic energy in white matter is available for generating action potentials, indicates that the maximum supportable firing rate, averaged across all neurons, is 7 8 2 Hz. This estimate is useful because, at present, the distribution of firing rates among neocortical neurons is not known [Shoham et al., 2006]. The estimate should be directly testable by future cellular-level physiological investigations in awake, behaving animals [Dombeck et al., 2007]. Finally, trends in myelinated axon size and abundance have corresponding trends at the level of macroscopic white matter volume. As expected, the density of axons decreases with increasing brain size (fig. 3e). In conjunction with estimates of the total number of neurons in neocortex, the density of axons can account quantitatively for the total amount of white matter. A key consequence of the scaling-up of brains is a disproportionate increase in white matter volume relative to gray matter volume (fig. 3f). Callosal axons present the converse of the case of parallel fibers, in which the predominance of unmyelinated axons maintains compactness and is well suited to sup-

Functional Tradeoffs in Axon Scaling


163

port low-frequency and/or sparse firing. More generally, because of metabolic constraints, the mean firing rate should be strongly linked to the distribution of axon sizes. I suggest that the average attainable firing rate of axon tracts in the brain, and therefore of the neurons that give rise to them, should be correlated with the abundance of myelinated axons and anticorrelated with the abundance of unmyelinated axons.

Future Directions: Cellular and Functional Correlates of Axon Diameter

To place the study of axon scaling in a functional context it is will be necessary to trace axons to their origins and targets, as a means of placing them in identified circuits. Attractive first targets for tracing are the widest axons, which are specialized for fast processing. Axons have the dual role of transmitting signals and sending material down their length. Thus it would be expected that a fast-conducting, wide axon would be able to support a more extensive terminal arborization than a slow-conducting, narrow axon. This suggests the possibility that wider axons might synapse with more neuronal targets and have a role in long-distance communication disproportionate to their number. The classical case is the squid giant axon, which innervates the mantle to facilitate rapid jetting, and is so large that it arises from the fusion of many large neurons in the giant fiber lobe [Young, 1939]. In mammals, where axons arise exclusively from single neurons, large axonal arborizations may be more prominent in large brains for several reasons. First, very large axons have been observed in several cases, including the optic nerve and the corpus callosum, as previously described. In addition, generally speaking, postmitotic cell types, including neurons, tend to be larger in mammals of increasing body size. If axon diameter is limited by the size of the parent neuron, the maximum supportable axon diameter – and terminal arborization – might also vary across species. Therefore the largest axons in big-brained animals may be ‘superhighways’ in which impulses are transmitted and the information they convey is distributed to a large number of targets. Three unusually large neuron types are known in the neocortex of big-brained primates: Betz, Meynert, and spindle cells. The regional specificity with which these cell types appear is consistent with the possibility that particular neocortical functions require high transmission speed. Betz cells of primary motor cortex (Brodmann area 4) and Meynert cells of primary visual cortex (Brodmann area 164


17) have attracted particular attention due to their large cellular volume, unique dendritic arborization patterns, distinctive connections, and thick myelinated axons [Chan-Palay et al., 1974; Scheibel and Scheibel, 1978; Rivara et al., 2003]. Although large neurons have been described in other large-brained mammals, the exceptionally large size and unusual architecture of these cells in primates [Le Gros Clark, 1942; Kaas, 2000] suggests the possibility that these neuronal subtypes constitute cellular substrates for specialized sensorimotor capacities such as nimble digital movement and vision [Le Gros Clark, 1959; Martin, 1990]. Betz cells are involved in setting muscle tone prior to fine motor output [Heffner and Masterton, 1983; Rivara et al., 2003] and Meynert cells participate in the processing of visual motion [Fries et al., 1985; Lia et al., 2003]. A cellular specialization for which a functional role has not yet been proposed is the spindle cell, a particular type of projection neuron that is characterized by a vertical, fusiform morphology and very large size [Nimchinsky et al., 1999]. The projection patterns of axons emanating from spindle cells are not yet known. An area ripe for future investigation is the examination of the correspondence between axonal structure and functional role. The largest axons provide a particu-

Fig. 3. Long-distance axons in the mammalian neocortex. a Representative micrographs of callosal tissue sectioned in sagittal view. b The fraction of axons that are myelinated increases with brain diameter (slope = 1.04 8 0.14% myelination per cm brain diameter). Each symbol represents pooled values from one animal. c Estimated cross-brain conduction times in axons of the corpus callosum, plotted for all myelinated axons (average values; open triangles) and for the widest axons (box plots). d Scaling of per-action potential metabolic costs. The estimated metabolic cost per unit weight of white matter of generating an action potential in all fibers, GAP, decreases as the –0.31 8 0.03 power of body weight (open circles). Direct measurement of white matter metabolic rates per unit time, Gt, measured by autoradiographic methods (filled circles), gives a power law slope of –0.32 8 0.06. The GAP and Gt axes are on logarithmic scales and are aligned with one another by a factor of 14 spikes/s. e Density of callosal axons. Box plots indicate median, 25th and 75th percentiles, and dots indicate measurements outside this range. f Contribution of large axons to neocortical white matter volume. The estimated total volume of white matter axons (filled squares) can account for the empirically observed supralinear growth in white matter volume (open circles) relative to gray matter. Replacing the overall axon distribution with that of mouse (open squares) reduces the axon volume to an approximately linear relationship [adapted from Wang et al., 2008].

Wang

shrew

mouse

marmoset 100

cat

Myelination (%)

80

macaque

60

cat

macaque

mouse 40 marmoset rat

20

shrew

0 0

b

2 4 6 Brain diameter (cm)

a 1 μm

least shrew mouse rat

0

c

harbor porpoise gorilla striped dolphin human bottlenose dolphin humpback whale

5

•

1

shrew mouse rat

0.1

3 2

1

0.1

d

0.3 0

e

1 2 3 4 5 Brain diameter (cm)


6

ferret marmoset rabbit macaque cat dog

White matter volume (mm3)

Density (axons/μm2)

6 5 4

Measured metabolic rate Gt (nmol glucose/g/s)

Widest axons

Estimated metabolic cost GAP (nmol glucose/g/spike)

•

All myelinated axons 10

marmoset cat macaque orangutan chimpanzee

Estimated cross-brain conduction times (ms)

15

f

10 1 Brain weight (g)

10

1

100

105 104 103

slope = 1

102 10

102 103 104 105 Gray matter volume (mm3)


165

larly convenient substrate for future studies because of the relative ease of identifying and tracing them. However, the wide range of observed axon diameters, especially in large-brained animals, suggests that axons might be quite functionally diverse. Therefore a more

general treatment of brain function will eventually come to include an understanding of how axon diameter and its correlates fit into the processing strategies used by neural circuits.

References Aboitiz F, López J, Montiel J (2003) Long distance communication in the human brain: timing constraints for inter-hemispheric synchrony and the origin of brain lateralization. Biol Res 36:89–99. Albus JS (1971) A theory of cerebellar function. Math Biosci 10:25–61. Attwell D, Laughlin SB (2001) An energy budget for signaling in the grey matter of the brain. J Cereb Blood Flow Metab 21:1133–1145. Brand S, Dahl AL, Mugnaini E (1976) The length of parallel fibers in the cat cerebellar cortex. An experimental light and electron microscopic study. Exp Brain Res 26:39–58. Bush P, Sejnowski TJ (1996) Inhibition synchronizes sparsely connected cortical neurons within and between columns in realistic network models. J Comput Neurosci 3:91–110. Chadderton P, Margrie TW, Häusser M (2004) Integration of quanta in cerebellar granule cells during sensory processing. Nature 428: 856– 860. Changizi MA (2001) Principles underlying mammalian neocortical scaling. Biol Cybem 84: 207–215. Chan-Palay V, Palay SL, Billings-Gagliardi SM (1974) Meynert cells in the primate visual cortex. J Neurocytol 3:631–658. Chklovskii DB (2004) Synaptic connectivity and neuronal morphology: two sides of the same coin. Neuron 43:609–617. Collins RC, McCandless DW, Wagman IL (1987) Cerebral glucose utilization: comparison of [14C]deoxyglucose and [6–14C]glucose quantitative autoradiography. J Neurochem 49: 1564–1570. Dombeck DA, Khabbaz AN, Collman F, Adelman TL, Tank DW (2007) Imaging large-scale neural activity with cellular resolution in awake, mobile mice. Neuron 56:43–57. Eccles JC, Ito M, Szentágothai J (1967) The Cerebellum as a Neuronal Machine. 343 pp. New York: Springer-Verlag. Eccles JC, Llinás R, Sasaki K (1966) The mossy fibre-granule cell relay of the cerebellum and its inhibitory control by Golgi cells. Exp Brain Res 1:82–101. Ekstrand MI, Enquist LW, Pomeranz LE (2008) The alpha-herpesviruses: molecular pathfinders in nervous system circuits. Trends Molec Med: In press. Fries W, Keizer K, Kuypers HG (1985) Large layer VI cells in macaque striate cortex (Meynert cells) project to both superior colliculus and prestriate visual area V5. Exp Brain Res 58: 613–616.

166

Gardner-Medwin AR (1972) An extreme supernormal period in cerebellar parallel fibres. J Physiol (Lond) 222:357–371. Harrison KH, Hof PR, Wang SS-H (2002) Scaling laws in the mammalian neocortex: does form provide clues to function? J Neurocytol 31: 289–298. Heffner RS, Masterton RB (1983) The role of the corticospinal tract in the evolution of human digital dexterity. Brain Behav Evol 23: 165– 183. Hodgkin, AL (1971) The Conduction of the Nervous Impulse. 108 pp. Liverpool: Liverpool University Press. Hoffmeister B, Jänig W, Lisney SJ (1991) A proposed relationship between circumference and conduction velocity of unmyelinated axons from normal and regenerated cat hindlimb cutaneous nerves. Neuroscience 42: 603–611. Hofman MA (1988) Size and shape of the cerebral cortex in mammals. II. The cortical volume. Brain Behav Evol 32:17–26. Hsu A, Tsukamoto Y, Smith RG, Sterling P (1998) Functional architecture of primate cone and rod axons. Vision Res 38:2539–2549. Hursh JB (1939) Conduction velocity and diameter of nerve fibers. Am J Physiol 127: 131– 139. Jörntell H, Ekerot CF (2006) Properties of somatosensory synaptic integration in cerebellar granule cells in vivo. J Neurosci 26:11786– 11797. Kaas JH (2000) Why is brain size so important: Design problems and solutions as neocortex gets bigger or smaller. Brain Mind 1:7–23. Kennedy C, Sakurada O, Shinohara M, Jehle J, Sokoloff L (1978) Local cerebral glucose utilization in the normal conscious macaque monkey. Ann Neurol 4:293–301. Le Gros Clark WE (1942) The cells of Meynert in the visual cortex of the monkey. J Anat 74: 369–376. Le Gros Clark WE (1959) The Antecedents of Man. Edinburgh: Edinburgh University Press. Lia H, Fukuda M, Tanifuji M, Rockland KS (2003) Intrinsic collaterals of layer 6 Meynert cells and functional columns in primate V1. Neuroscience 120:1061–1069. Llinás RR (1982) General discussion: radial connectivity in the cerebellar cortex: a novel view regarding the functional organization of the molecular layer. In: The Cerebellum (Palay SL, Chan-Palay V, eds) New Vistas Exp Brain Res (Suppl) Vol 6, pp 189–194. New York: Springer-Verlag.


Marr D (1969) A theory of cerebellar cortex. J Physiol 202:437–470. Martin RD (1990) Primate Origins and Evolution. Princeton, NJ: Princeton University Press. Movshon JA, Newsome WT (1996) Visual response properties of striate cortical neurons projecting to area MT in macaque monkeys. J Comp Neurol 16:7733–7741. Mugnaini E (1983) The length of cerebellar parallel fibers in chicken and rhesus monkey. J Comp Neurol 220:7–15. Nimchinsky EA, Gilissen E, Allman JM, Perl DP, Erwin JM, Hof PR (1999) A neuronal morphologic type unique to humans and great apes. Proc Natl Acad Sci USA 96:5268–5273. Nowaczyk T, Des Rosiers MH (1981) Application of the 2-deoxy-D-[14C]-glucose method to the mouse for measuring local cerebral glucose utilization. Eur Neurol 20:169–172. Olivares R, Montiel J, Aboitiz F (2001) Species differences and similarities in the fine structure of the mammalian corpus callosum. Brain Behav Evol 57:98–105. Ringo JL, Doty RW, Demeter S, Simard PY (1994) Time is of the essence: a conjecture that hemispheric specialization arises from interhemispheric conduction delay. Cereb Cortex 4: 331–343. Ritchie JM (1995) Physiology of axons. In: The Axon: Structure, Function and Pathophysiology (Waxman SG, Kocsis JD, Stys PK, eds), pp 68–96. New York: Oxford University Press. Rivara CB, Sherwood CC, Bouras C, Hof PR (2003) Stereologic characterization and spatial distribution patterns of Betz cells in human primary motor cortex. Anat Rec 270A:137–151. Rockland KS (2002) Non-uniformity of extrinsic connections and columnar organization. J Neurocytol 31:247–253. Rushton WAH (1951) A theory of the effects of fibre size in medullated nerve. J Physiol 115: 101–122. Scheibel ME, Scheibel AB (1978) The dendritic structure of the human Betz cell. In: Architectonics of the Cerebral Cortex (Brazier MAB, Pets H, eds), pp 43–57. New York: Raven Press. Shambes GM, Gibson JM, Welker W (1978) Fractured somatotopy in granule cell tactile areas of rat cerebellar hemispheres revealed by micromapping. Brain Behav Evol 15:94–140. Shapiro HM, Greenberg JH, Reivich M, Ashmead G, Sokoloff L (1978) Local cerebral glucose uptake in awake and halothane-anesthetized primates. Anesthesiology 48:97–103.

Wang

Shoham S, O’Connor DH, Segev R (2006) How silent is the brain: is there a ‘dark matter’ problem in neuroscience? J Comp Physiol A 192:777–784. Soha JM, Kim S, Crandall JE, Vogel MW (1997) Rapid growth of parallel fibers in the cerebella of normal and staggerer mutant mice. J Comp Neurol 389:642–654. Sokoloff L (1996) Cerebral metabolism and visualization of cerebral activity. In: Comprehensive Human Physiology Vol 1 (Greger R, Windhorst U, eds), pp 579–602. Berlin: Springer-Verlag. Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada O, Shinohara M (1977) The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897–916.


Swadlow HA (2000) Information flow along neocortical axons. In: Time and the Brain. Conceptual Advances in Brain Research (Miller R, ed), pp 131–155. Amsterdam: Harwood Academic Publishers. Tigges J, Tigges M, Anschel S, Cross NA, Letbetter WD, McBride RL (1981) Areal and laminar distribution of neurons interconnecting the central visual cortical areas 17, 18, 19, and MT in squirrel monkey (Saimiri). J Comp Neurol 202:539–560. Varela F, Lachaux JP, Rodriguez E, Martinerie J (2001) The brainweb: phase synchronization and large-scale integration. Nat Rev Neurosci 2:229–239. Wang SS-H, Shultz JR, Burish MJ, Harrison KH, Hof PR, Towns LC, Wagers MW, Wyatt KD (2008). Shaping of white matter composition by biophysical scaling constraints. J Neurosci 28:4047–4056.

Wen Q, Chklovskii DB (2005) Segregation of the brain into gray and white matter: A design minimizing conduction delays. PLoS Comput Biol 1(7):e78. West GB, Brown JH, Enquist BJ (1997) A general model for the origin of allometric scaling laws in biology. Science 276:122–126. Wyatt KD, Tanapat P, Wang SS-H (2005) Speed limits in the cerebellum: constraints from myelinated and unmyelinated parallel fibers. Eur J Neurosci 21:2285–2290. Young JZ (1939) Fused neurons and synaptic contacts in the giant nerve fibres of cephalopods. Phil Trans R Soc London Biol 229:465–503. Zhang K, Sejnowski TJ (2000) A universal scaling law between gray matter and white matter of cerebral cortex. Proc Natl Acad Sci USA 97: 5621–5626.


167



Exploring the Origins of the Human Brain through Molecular Evolution Eric J. Vallender Division of Neurochemistry, New England Primate Research Center, Harvard Medical School, Southborough, Mass., USA

Key Words Hominid ⴢ Primate ⴢ Human evolution ⴢ Brain evolution ⴢ Neurogenetics ⴢ Molecular evolution

Abstract The emergence of the human brain is one of evolution’s most compelling mysteries. With its singular importance and astounding complexity, understanding the forces that gave rise to the human brain is a major undertaking. Recently, the identification and publication of the complete genomic sequence of humans, mice, chimpanzees, and macaques has allowed for large-scale studies looking for the genic substrates of this natural selection. These investigations into positive selection, however, have generally produced incongruous results. Here we consider some of these studies and their differences in methodologies with an eye towards how they affect the results. We also clarify the strengths and weaknesses of each of these approaches and discuss how these can be synthesized to develop a more complete understanding of the genetic correlates behind the human brain and the selective events that have acted upon them. Copyright © 2008 S. Karger AG, Basel

Introduction

The human brain is wondrously large and complex. It, perhaps more than any other feature, defines what it means to be human. Understanding how the human © 2008 S. Karger AG, Basel 0006–8977/08/0722–0168$24.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com


brain evolved and the genetic differences underlying this evolution offers not only insight into the unique biology and neurological diseases affecting humans, but also scratches the philosophical itch of what makes us human. In recent years advances in sequencing technology have offered scientists the opportunity to begin thorough investigations of the question. With complete genomes now available for human, mouse, rat, chimpanzee, dog, and rhesus macaque and lower coverage genomes for additional species, it has become possible to consider these questions on a genomic scale. There have been many attempts to do just that: to use genomic data to identify the genetic signatures of positive selection on the human genome, some broadly and some specifically on the brain. Although some advances have been made in looking at regulatory changes or changes in transcriptomics, and without a doubt this will prove to be an important component to the brain evolution story, the vast majority of studies have focused on protein coding changes. This has been the case for several reasons. First, protein coding changes are simple to identify. Regulatory regions are evolutionarily labile, at least more so than protein coding regions, and still cannot be easily notated in genomic sequence. This also ties into the second major point; functionality is more easily assessed bioinformatically for proteins than for regulatory sequence. Our understanding of the physiochemical properties of the amino acids and the structure and function of proteins allows for protein changes to be placed into context. By way of contrast, changes even in known regulatory regions alEric J. Vallender Division of Neurochemistry, New England Primate Research Center Harvard Medical School, One Pine Hill Drive Southborough, MA 01772 (USA) Tel. +1 508 624 8194, Fax +1 508 786 3317, E-Mail [email protected]

Species-pair relative rate comparisons

Relative rate of amino acid fixation Polymorphism-based approaches

Humans Australopithecus Homo Homo sapiens

Chimpanzees Gorillas Orangutans Old World Monkeys New World Monkeys

Fig. 1. Simplified primate phylogeny.

most always require in vitro or in vivo studies for any functionality to be assessed. Finally, there exist well established metrics, such as K A/KS (or dN/dS or ␻) for the detection of positive selection in protein sequences. K A/KS has long been used to identify selection from divergence data. KS is the ratio of synonymous mutations per synonymous site and is generally thought to be representative of the neutral rate of mutation at least in mammals where synonymous sites are generally understood to be under no selective constraint. K A is representative of the number of amino acid replacement substitutions per possible replacement site. Taken together, K A/KS values equal to one indicate neutrality, whereas K A/KS values significantly different from one reflect the effects of selection. Most genes show K A/KS values significantly less than one (often around 0.2) reflecting the general selective constraint on established proteins. K A/KS values greater than one are much less common and reflect the effects of positive selection in driving amino acid changing mutations to fixation at rates more rapidly than expected by neutrality. More recent positive selection can also be identified elsewhere in the genome through polymorphism-based methods (fig. 1). Some of these methods rely on deviations from the neutral expectation of the allele frequency spectrum which reflects the relative proportions of single nucleotide polymorphisms occurring at various frequencies, such as Tajima’s D [Tajima, 1989] and Fay and Wu’s H [Fay and Wu, 2000], and others compare rates of polymorphism to rates of divergence, such as the McDonaldKreitman [McDonald and Kreitman, 1991] and HudsonKreitman-Aguade test [Hudson et al., 1987]. Increasingly

sophisticated tests continue to be developed with bases elsewhere, such as the sizes of linkage disequilibrium blocks. These methods, however, focus on relatively recent selection; signatures of ancient selective events are eliminated over evolutionary time. Recent positive selection can and has been identified in regulatory regions as well as protein-coding regions through a dearth of high frequency segregating sites (or balancing selection through an excess of these sites), but there exists little in the way of statistical methodology to identify selective events once the equilibrium they disturb has been reestablished. Other methodologies using different metrics obtained from polymorphism data suffer from the same fate. Yet these same issues that have so far stymied investigations into regulatory sequence evolution still play a role in our understanding of protein evolution underlying the emergence of the human brain. Many large gene or genomic studies have been undertaken to look for positive selection in human brain evolution. Although some general patterns have emerged, such as the widespread loss or change of olfactory genes, often these studies have returned different results. Most find evidence of selection somewhere, but the lists of genes often do not overlap. Some of the reasons for these differences are methodological or result from different starting data sets or levels of quality control. What is important to note, however, is that some of the differences that are observed do not simply represent incongruities among studies, but actually represent different answers to biologically different questions.

Exploring the Origins of the Human Brain


169

Here we will consider the breadth of human brain evolution covered by these studies, focusing on the different questions they ask and assumptions they make. This will hopefully offer insight into the diverse results found in the studies and lead to a more complete understanding not only of human brain evolution, but also the investigations themselves.

Recent Human Selection

There have been a significant number of studies over the past several years focusing on recent human selection, meaning selection operating over the last two hundred thousand years and roughly correlating to the emergence of anatomically modern Homo sapiens from Africa. This has been fueled by two major advancements in the field. First, the publication of the results of the HapMap survey made available large amounts of polymorphism data to the community [The International HapMap Consortium, 2005; Frazer et al., 2007]. Second, the emergence of new methodologies and statistics for the study of recent positive selection has allowed researchers new and novel ways to interpret the data [Sabeti et al., 2006; Thornton et al., 2007]. There have been several instances of candidate genes involved in brain development and function showing evidence for positive selection in humans, specifically the microcephaly-associated genes MCPH1 [Evans et al., 2005] and ASPM [Mekel-Bobrov et al., 2005]. These genes, whose evolution includes evidence of more ancient positive selection discussed further below, were initially of interest because mutations therein caused primary microcephaly, a brain developmental disorder resulting in smaller than normal brain sizes without concomitant physical abnormalities. Regions of high frequency extended haplotypes were identified for these genes which were unable to be modeled through neutral evolution following standard demographic assumptions. This remains an outstanding question, however, as other studies have suggested that these haplotype patterns might not be out of the ordinary in human genomes [Currat et al., 2006; Yu et al., 2007]. The patterns of haplotypic diversity seen in ASPM and MCPH1 do not seem to be unique to these genes. Although formally possible that numerous genes have undergone significant positive selection akin to these two, it seems more likely that there exists some less understood and underappreciated demographic scenario that accounts for the findings. What is certain is that there is some underlying mechanism that we do not fully understand. 170


Relevant to these discussions, however, is the role of assumptions in coloring our perceptions of the studies. As stated above, these two genes were originally targeted for study because of their association with the microcephaly brain development disorder, and although it is tempting to associate selective events with these phenotypes, it is not a given. Indeed, numerous recent studies have since shown that the haplotypes identified in these genes are not associated with IQ or brain size [Woods et al., 2006; Dobson-Stone et al., 2007; Mekel-Bobrov et al., 2007; Rushton et al., 2007; Timpson et al., 2007]. This, in and of itself, does not invalidate findings of positive selection, and in fact it is possible that positive selection is acting on another phenotypic character, but it casts a long shadow over any and all interpretations. The finding itself, though, is independent of phenotypic understanding. This fact must be kept at the forefront when evaluating the conclusions drawn from the data. Apart from these candidate gene studies, numerous whole genome scans for positive selection have been undertaken [ The International HapMap Consortium, 2005; Carlson et al., 2005; Kelley et al., 2006; Voight et al., 2006; Sabeti et al., 2007; Williamson et al., 2007]. Although these studies do find similar results among genes with very strong selective signatures (the lactase gene and genes involved in skin pigmentation for instance), there remains substantial variability. Only one of these studies identifies nervous system development genes as particularly well represented [Williamson et al., 2007] and several specifically exclude the candidate genes ASPM and MCPH1. So what information can be taken from this? First, there are the general issues that exist within the field. Demography is problematic, especially in humans where population bottlenecks exist not only in the species as a whole, but also within sub-populations. The improvement of methodologies and increasing access to genomic levels of variation will allow for greater empirical comparisons between genes, but the pervasiveness of positive selection remains elusive and will affect our understanding of what constitutes an outlier relative to the neutral expectation. This has been illustrated particularly in the case of ASPM and MCPH1. Ascertainment bias is also a potential problem. Many of the current SNPs available for the whole genome studies were previously identified in specific subpopulations which might affect the interpretation of results. Further, many rare variants could be missing from these studies, affecting tests that focus on deviations from the allele frequency spectrum. Additionally, variation in recombination rates across the genome might render certain gene-gene comparisons inVallender

valid. The effects of recombination not only on variation, but on the tests themselves, have yet to be fully explored. But apart from the methodological difficulties that will eventually be resolved, there remains variability because the nature of the tests themselves is different. Although some test statistics focus on completed sweeps, others focus on those sweeps which are ongoing, necessarily identifying different subsets of genes. Within those focusing on completed sweeps, some identify sweeps from the more ancient past (two hundred thousand years) whereas others identify more recent events (fifty thousand years). Although these differences might sound semantic, the evolutionary milieu was greatly variable among the time periods, and the genes and phenotypes under selection were likely very different. This is particularly relevant as it relates to brain evolution. As stated above, anatomically modern humans are believed to have emerged between 100,000 and 200,000 years ago; the ‘human’ brain was already established at this point. It is possible that additional genetic changes might have occurred imparting greater cognitive or linguistic abilities, but the large-scale anatomical changes that are the hallmark of the human brain has been static since then. Indeed the brain size of Homo heidelbergensis, who lived five hundred thousand years ago and is thought to be the direct ancestor of Homo sapiens, was only slightly smaller than that of an average human living today [Neill, 2007]. This, coupled with drastic changes between Australopithecus and early Homo approximately two million years ago, suggests that perhaps the most salient genetic sweeps of positive selection affecting the emergence of the modern human brain are outside the scope of detection by polymorphism based approaches.

Divergence from Primates

Because of the inability of polymorphism scans to detect the selective sweeps of five hundred thousand years ago or more that led to the human brain, much work has focused on the differences between humans and chimpanzees. The role of protein coding changes in the human-chimpanzee divergence began to take a back seat to regulatory changes long ago [King and Wilson, 1975], but recently this has been revisited as a number of studies have focused on amino acid altering mutations in this terminal human lineage. When the chimpanzee genome was published, only the human, mouse, and rat genomes were publicly availExploring the Origins of the Human Brain

able with the dog following shortly thereafter. Because of this, early studies focusing on the differences between humans and chimpanzees used one or more of these distantly related species as an outgroup [Clark et al., 2003; The Chimpanzee Sequencing and Analysis Consortium, 2005; Bustamante et al., 2005; Khaitovich et al., 2005; Nielsen et al., 2005; Arbiza et al., 2006]. The evolutionary time separating human and chimpanzee (roughly 5 million years) is at least one fifteenth that separating the other, non-primate, mammalian species. Because of this, questions arose as to the identification of orthologs between the species as well as improper alignments and multiple mutations at the same position. It was because of these potential confounds that additional studies were undertaken when the rhesus genome was published [Gibbs et al., 2007]. These early studies produced interesting lists of genes possibly evolving more rapidly in humans, but they largely failed to overlap. Certain categories and genes, particularly those expected to be under very strong positive selection, such as spermatogenesis genes and immune response genes, do show up regularly, but the overlap is nevertheless much less than perhaps is anticipated. The reasons for this can be legion but most seem focused on methodological rather than biological differences. In particular, studies have varied in whether they consider the human-chimpanzee branch as a whole or attempt to partition changes between the two hominoid terminal branches. In the case of the latter, variation in methods for ancestral sequence reconstruction might play a role. Another source of variation among the studies is in which chimpanzee sequence is used and how strict quality control standards are applied. Because of the short evolutionary time, a few genotyping errors, or even slightly deleterious polymorphisms appearing as fixed differences, can have major effects. As we have seen, studies hoping to identify positive selection on the human genome by K A/KS values greater than one (indicative of positive selection) on the human lineage since the divergence of chimpanzees or on the combined terminal lineages of humans and chimpanzees have largely been ineffectual in identifying genes related to brain differences, although not entirely. One study that focused on brain-expressed genes found a correlation between brain expression and higher evolutionary rates in the human terminal branch compared to the chimpanzee terminal branch [Yu et al., 2006]. This study suggests that some of the disparity observed among the previous studies is the result of improper reconstruction of ancestral sequences as a consequence of long divergent outgroups Brain Behav Evol 2008;72:168–177

171

such as rodents. This remains unproven, however, and it remains to be determined which is the best or correct method. Although there are many differences among the studies, there are also some similarities. Genes involved in sensory perception, and more specifically chemosensory perception, consistently appear to be overrepresented in these studies of positive selection. In particular, it is well established that large changes in olfactory genes are widespread in humans, chimpanzees, and more broadly across primates [Gimelbrant et al., 2004; Gilad et al., 2005]. Nevertheless, the failure to identify large numbers of brain genes under positive selection between humans and chimpanzees has raised questions. The phenotypic difference in the brains of the two species is unmistakable, so what possibilities exist to explain the genetic findings? Three primary explanations immediately present themselves. The first is that the phenotypic differences in the brains of human and chimpanzees are the result not of many changes, but rather of a very few, yet quite significant, differences. In this scenario only a handful of genes need be positively selected whereas the remainder of the brain-associated genes remain largely unchanged. This is unfulfilling for several reasons not the least of which is the seemingly impossible task of a single or few genes accomplishing changes of this magnitude. Nevertheless, it remains a possibility, although unlikely. The second possibility is that in fact protein-coding changes are of greatly diminished importance relative to regulatory changes. As mentioned previously, this has long been considered the case, largely because the magnitude of the changes was thought to be incongruous with the high levels of similarity between the two genomes. More recently this understanding has been altered to emphasize the complexity of the system and epistatic protein networks. The general argument is that even small protein changes have such far reaching consequences that it is unlikely that they could be selected for without significantly disturbing the brain as a whole. The final possibility is simply that current methods lack the power to detect positive selection between humans and chimpanzees. Indeed it might not even be methodological deficits, but rather a simple sample-sizerelated statistical power failure. The crux of this argument is that so little evolutionary time has elapsed since the divergence of humans and chimpanzees that there are simply not enough mutations to separate the signal from the noise in these studies. Mutation rates are such that it is not unlikely for human-chimpanzee orthologs to have 172


no differences even at synonymous sites. The random and stochastic nature of genetic drift means that the difference between zero, one, or two changes is not particularly great, and yet with so few changes expected between humans and chimpanzees they take on greater importance. Because of this, a gene evolving under neutral, or even significantly negative, selection might result in a K A/KS value greater than one simply because stochastic chance has resulted in a KS value that is zero or close to it. At the same time, a gene under positive selection might appear to have a K A/KS value at or less than one simply because stochastic chance has over-inflated the KS value. In short, the problem with the human-chimpanzee comparison, and other short lineage comparisons, is that the effect of stochastic noise is disproportionately large relative to the desired signal. One way in which this failure of the human-chimpanzee comparison can be overcome is to use more divergent lineages. In practice, this has been done by comparing humans to old world monkeys, in particular the rhesus macaque [Gibbs et al., 2007]. The power issue is overcome as the evolutionary time is increased to a degree such that signal reasserts itself over noise; but rather than identify genes important since the human-chimpanzee divergence, this comparison identifies genes under positive selection as far back as the ape-old world monkey divergence. There have been relatively few studies that incorporate the rhesus genome alongside the human and chimpanzee genomes and a non-primate outgroup. Those that have been done, however, tend to show a greater number of genes under positive selection in the lineage leading from the catarrhine ancestor to the hominoid ancestor and include a number of brain genes of interest. This is perhaps unsurprising as there are substantial phenotypic differences between the brain of an old world monkey and that of an ape. In fact, the importance of the monkey to ape transition and its relevance to human evolution is largely understated. Nevertheless, the genes thus identified are not necessarily those that lead to the uniquely human phenotype; they are rather the genes (some of the genes) responsible for the ape phenotype. These are important when considering non-human primate models of disease, which generally tend to be old- or new-world monkeys, and might provide interesting candidate genes, but the evidence does not provide proof that these genes lead to uniquely human characters.

Vallender

Differences across Species in Levels of Selection

There is one additional general failure of the K A/KS metric in assessing positive selection which we have neglected. That is its ability to be reduced to less than one, despite bouts of positive selection, due to the diluting effects of negative selection either by time or across the gene itself. Periods of evolutionary time in which a gene is under negative selection or neutrality can overwhelm bouts of positive selection. Indeed this is likely to be the case in the human-rhesus comparison for those genes that have undergone positive selection since the humanchimpanzee divergence. Although K A is elevated as a result of selective forces for only this short time, the KS denominator remains constant. The net effect is a perhaps slightly higher K A/KS value than would have been observed had there been no positive selection event, but still a value less than one is indicative of overall negative selection across the complete time frame studied. The same effect can be seen if only one part of a protein is under positive selection while the rest remains under strong selective constraint. In this case the selected region is overwhelmed by the rest of the protein and the net result again is an overall K A/KS value less than one and the appearance of negative selection. It is not hard to imagine scenarios for which this situation might be plausible; for instance, positive selection acting on ligand binding moieties while the rest of the protein remains fixed. Indeed this observation has been made for the major histocompatibility complex (MHC), which is generally regarded as one of the prime examples of positive selection. In the MHC the antigen binding regions are under extreme diversifying selection whereas other structural components remain more stable. In the case of the MHC, the areas under positive selection overwhelm those under constraining selection, but the MHC is unusual in this regard. Similar patterns of selective differences within a gene are also seen in membrane-bound receptors wherein transmembrane domains tend to be evolving at slower rates than either intracellular or extracellular regions. Regardless of the mechanism, the result is the same: a K A/KS value which is slightly greater than it would have been had the positive selection event not occurred, but still not great enough to be detected as relevant in the absence of other data. What can be done to ameliorate these problems? To address the dilution effects of time, shorter intervals can be used. This nevertheless suffers from either the loss of statistical power as time decreases, as the human-chimpanzee examples illustrate, or a complete Exploring the Origins of the Human Brain

inability to determine if an intermediate sequence is unavailable. For instance in the divergence of apes from old world monkeys there is no intermediary sequence information between the old world monkeys and the divergence of the gibbons and siamangs. This time period is thus irreducible. More can be done to address heterogeneity across physical space. Rather than focus on entire genes as the unit of study, predefined domains can be used. Alternatively, primary sequence sliding windows or tertiary sequence defined regions can be used assuming that multiple testing is appropriately accounted for. This suffers from the same loss of statistical power as the size under study decreases. Whether there are fewer mutational events because there is not enough time or because there are not enough physical locations, the paucity of mutations results in a lack of power. One approach that can be taken is to look for differences in K A/KS values between species pairs. The null expectation here is that levels of selection reflected by K A/KS are the same between the species pairs under study. Deviations from the null can be tested for and used to identify differences in selective regimes. One outstanding difficulty with this approach, however, is that the nature of this difference is ambiguous. A difference could be the result of one species pair experiencing increased negative selection, one species pair experiencing positive selection, or one species pair undergoing a relaxation of selective constraint. The last category is particularly important as relaxation of selective constraint is not affected solely by selective pressures, but also is influenced by demographic factors including effective population size. Previous studies have attempted to control for these demographic effects by attempting to identify a baseline level of relaxation of constraint from either a control set of genes or, more broadly, from the whole genome. The earliest study, predating the publication of the rhesus genome, focused on the differences in KA/KS values between humans-macaques and mouse-rat [Dorus et al., 2004]. This study found an increase in the KA/KS values of brain genes that was not seen in a control set of housekeeping genes. Moreover, the brain genes, when subdivided by functional class, showed a significant acceleration among those genes involved in developmental processes. (In these cases, ‘acceleration’ refers to an increase in Ka/Ks values or the fixation rate of amino acid mutations scaled to that of synonymous mutations). The difference in K A/KS values between the brain set and the control set ruled out the effects of population size and demographics in generating the pattern, but it remained Brain Behav Evol 2008;72:168–177

173

formally possible simply that rodent brains had increased purifying selection or that primate brains had undergone selective relaxation. Although not excluded by the genetic data, these interpretations seem unlikely given our current understanding of primate and rodent neurology. This study garnered two particularly relevant criticisms. The first involved the identification of ‘brain genes.’ Although this list contained genes generally believed to be relevant to brain function and development, it was by no means exhaustive. Similarly, there was no bias in the selection of brain genes, but nevertheless it remains possible that these genes are not representative of brain genes as a whole and that the results should not be generalized. The second criticism goes hand in hand with the first, suggesting that the control genes used are not generally appropriate and that different or additional genes would have produced more certain results. Both of these criticisms are appropriate and reflect limitations of the study at the time it was undertaken. They both can be ameliorated somewhat by the introduction of more recently published genomic data. Similar studies were undertaken during the preparation of the rhesus and chimpanzee genomes for publication. In the rhesus, genes with higher K A/KS values in primates compared to rodents were overrepresented in categories of taste and smell sensory perception as well as the broad category of transcription factors [Gibbs et al., 2007]. This forms perhaps the most direct comparison as the species pairs used are the same. The same caveat that was noted above, however, applies to these studies as well. Although gaining statistical power through the use of a more distantly related species (old world monkey), genes that are identified might have undergone positive selection at some point during a longer evolutionary time frame. As mentioned before, it seems notable that many individual examples seem to indicate that these genes are particularly enhanced for positive selection during the lineage leading from monkeys to apes rather than the uniquely human terminal lineage since the chimpanzee divergence. Although not directly comparable to the earlier study, this confound was removed by studies using the chimpanzee genome [ The Chimpanzee Sequencing and Analysis Consortium, 2005; Khaitovich et al., 2005]. Using the human-chimpanzee comparison against the mouse-rat, these studies found that brain genes as a whole showed an acceleration in hominoids relative to rodents. This was not seen for genes representative of other organs including heart, kidney, liver, and even testis. Although not quite significantly faster than genes as a whole (p = 0.08), 174


this was a significant acceleration compared to any other organ (p ! 0.05). More work is obviously needed and the caveats regarding power to detect selection in short lineages still apply, but this is perhaps the most promising indication that protein changes on a large scale might have played a role in the emergence of the human brain. Other studies, however, have reached opposite conclusions from the same basic premise. One study found no evidence of a human acceleration in brain-specific genes when compared to other genes in the genome [Shi et al., 2006]. This study compared the rate of evolution of genes on the human terminal branch to the chimpanzee terminal branch using the rhesus macaque sequence as an outgroup to parse changes. This comparison raised two important points for these comparisons. The first is the conceptually simple, but practically difficult, issue of defining ‘brain genes.’ This difficulty is pervasive and beyond the scope of this discussion. The second, however, is apropos and relates to the relative importance of the chimpanzee error rate in any study using the data. The chimpanzee genome appears to have sequencing error rates approaching 0.07% [Taudien et al., 2006], 50–100 times higher than the human genome and only an order of magnitude less than anticipated divergence levels between humans and chimpanzees. The differences in results thus are perhaps suggestive of the strong effect of these error rates and how they are treated by, or not treated by, the various studies. An additional study which used a separately derived macaque sequence, in this case the sister species of rhesus, Macaca fascicularis, also failed to find an acceleration in brain-expressed genes compared to other genes in the genome [Wang et al., 2007]. The authors offer several reasons for the discordant findings including unintended biases towards slowly evolving, highly expressed genes or particularly slow evolving categories of genes, but discount these possibilities after additional examinations of the data. They also raise the specter of definitions, arguing possible differences between ‘brain-expressed’ and ‘brain-specific’ genes. Although they fail to find a difference in their data between the two, the points raised are valid and are likely to impact other studies. So what can be made of these discrepancies? Perhaps the clearest interpretation is that from the outset defining both the experimental data set, the ‘brain genes’, and the control data set must be defined clearly. Differing definitions of brain genes are likely to cause differences in results. This problem is difficult to address as consensus is unlikely to be reached and the concept is simple enough to appear intuitive. The importance of this problem is Vallender

highlighted in studies that found highly discordant lists of ‘brain genes’ depending on how the term is defined [Shi et al., 2006]. Control data sets contain not only these problems, but also problems in defining which set of data is itself is best. Other tissue-specific genes might better reflect the constraint experienced simply by being a tissue-specific gene, but selective forces are hardly uniform and if we have learned nothing else, it is that selection might be acting on genes and traits in ways of which we are largely unaware. Perhaps the most appropriate way to address control gene sets is simply to use multiple sets and compare the differing results. One final point needs be made about these studies. As stated at the outset, the basic premise of the studies is that positive selection will cause an elevation in K A/KS from the usual selective pressure for that gene. Because we have no way of knowing what the natural selective pressure for that gene is, we depend upon other species to offer direction, and control gene sets to inform us regarding the relevance of those species. This is all well and good, and might indeed represent the best that can be achieved at this time, but the underlying assumption is still a fairly large one and one that needs be examined before drawing conclusions that might be too definitive.

Exceptions That Prove the Rule?

Previously we have discussed some of the numerous genome-wide or otherwise large scale studies that have purported to search for positive selection in the human genome. The picture painted is a muddied one that offers no clear conclusions especially as it relates to genes involved with the brain. Are there many genes whose protein changes are responsible for the human phenotype? Is the evolution of the human brain represented in widespread signatures of selection in the genome? We still do not know. What is becoming clear, however, is that for a subset of genes at least, evidence for selection exists. And although the exact phenotypic traits upon which selection is acting is still unproven, for many of these genes a role in the evolution of the human brain remains a likely possibility. Above we touched upon studies of the microcephalyassociated genes MCPH1 and ASPM in regards to ongoing selection in humans, but there is strong evidence for evolution during the primate history of these genes as well [Zhang, 2003; Evans et al., 2004b; Kouprina et al., 2004]. ASPM shows K A/KS values greater than one, indicative of positive selection in both the lineage separatExploring the Origins of the Human Brain

ing the lesser apes, gibbons and siamangs, from the great apes as well as in the human terminal branch since divergence from the chimpanzee. These values are significantly different from other species pairs and tend to cluster discretely within the gene. MCPH1 shows a pattern that has become more familiar to studies of primate brain gene evolution [Evans et al., 2004a; Wang and Su, 2004]. As with ASPM, primate lineages leading to human (from the catarrhine ancestor) are evolving at rates greater than other species. Also similar to ASPM, the protein changing mutations tend to be clustered heterogeneously within the protein. Unlike ASPM, however, a K A/KS value greater than one is not seen in the human terminal branch. Indeed, the human terminal K A/KS value is not significantly different from the chimpanzee, gorilla, orangutan, or gibbon terminal branch. Rather, the bout of positive selection in MCPH1 appears to have occurred in the lineage leading from the catarrhine ancestor to the great ape ancestor, perhaps reflecting the emergence of the ape brain rather than specifically that of the human. GLUD2, which, in primates, encodes the brain-specific isoform of glutamate dehydrogenase (GDH), a protein responsible for the catabolism of the neurotransmitter glutamate, is another example of a brain-specific gene that shows a signature of positive selection [Burki and Kaessmann, 2004]. GLUD2 arose from a duplication event of GLUD1 after the divergence of the lineage that would become the apes from the catarrhine ancestor. Shortly after its emergence the new gene underwent a bout of positive selection (as newly emerged duplicates are wont to do). As with MCPH1, this period of positive selection was confined to the lineage between its birth at roughly the time of the catarrhine ancestor and the divergence of the great apes. It thus again seems likely that it represents a gene responsible for the ape brain phenotype. The gene that started it all was FOXP2. Originally identified as the source of a linguistic disorder, the molecular evolution of FOXP2 was found to show evidence for strong positive selection in the human terminal branch [Enard et al., 2002; Zhang et al., 2002]. This gene is a good example of both the shortcomings of the human-chimpanzee comparison as well as how they can be overcome. Despite the apparent effects of selection, the human terminal branch does not offer a K A/KS value significantly greater than one. This is due solely to the lack of statistical power as this branch contains two amino acid changes and zero synonymous changes. Nevertheless, a diagnosis of positive selection can be made because this is such an extreme departure from the amino acid Brain Behav Evol 2008;72:168–177

175

mutation rate seen in other species (only one amino acid difference separates chimpanzees from mice). From these and other examples, it is clear that there exist genes for which positive selection can unequivocally be deduced and who likely have played some role in the emergence of the human brain. What remains undetermined is the pervasiveness of these effects and their relative importance compared to other mechanisms shaping the emergence of the human brain, including the evolution of non-coding and regulatory regions. It is further unclear exactly what the different phenotypes imparted by the changes are and whether they exist in isolation or only in the context of other changes occurring at the same time. Further studies will certainly be forthcoming and with them a greater understanding of how the human brain has emerged.

A Footnote on Anthropocentrism

In the course of these studies, it becomes inevitable that there is talk of anthropocentrism. Several factors should be made clear: First, there is no scientific reason

to think that the lineage leading to humans is privileged or otherwise different from the lineages leading to other species. It might be the case that the mechanisms driving the emergence of the human phenotype vary somewhat from other species, but this is probably not true and no evidence has been presented indicating that this is the case. Second, that positive selection was at work on the human brain should not come as a surprise or otherwise set it apart from other phenotypes. Indeed, we are interested in the brain because, as humans, it is such a major part of who we are. Behaviors, psychiatric disorders, emotions, language, all of these intrigue us and warrant the study of the brain. Other traits unique to humans, such as the changes in body hair and sweat glands related to a novel thermoregulatory strategy, are equally important and warrant study. Finally, the same studies can, and likely will, be done for any species. We can legitimately ask what makes a mouse so ‘mousy’ or a cat so ‘catty.’ The methodologies will be largely similar and we will expect to see the same sorts of results. That these studies generally take a back seat in visibility to those in humans does not reflect on the science itself, but rather on the priorities of our human society.

References Arbiza L, Dopazo J, Dopazo H (2006) Positive selection, relaxation, and acceleration in the evolution of the human and chimp genome. PLoS Comput Biol 2:e38. Burki F, Kaessmann H (2004) Birth and adaptive evolution of a hominoid gene that supports high neurotransmitter flux. Nat Genet 36: 1061–1063. Bustamante CD, Fledel-Alon A, Williamson S, Nielsen R, Hubisz MT, Glanowski S, Tanenbaum DM, White TJ, Sninsky JJ, Hernandez RD, Civello D, Adams MD, Cargill M, Clark AG (2005) Natural selection on protein-coding genes in the human genome. Nature 437: 1153–1157. Carlson CS, Thomas DJ, Eberle MA, Swanson JE, Livingston RJ, Rieder MJ, Nickerson DA (2005) Genomic regions exhibiting positive selection identified from dense genotype data. Genome Res 15:1553–1565. Clark AG, Glanowski S, Nielsen R, Thomas PD, Kejariwal A, Todd MA, Tanenbaum DM, Civello D, Lu F, Murphy B, Ferriera S, Wang G, Zheng X, White TJ, Sninsky JJ, Adams MD, Cargill M (2003) Inferring nonneutral evolution from human-chimp-mouse orthologous gene trios. Science 302: 1960– 1963.

176

Currat M, Excoffier L, Maddison W, Otto SP, Ray N, Whitlock MC, Yeaman S (2006) Comment on ‘Ongoing adaptive evolution of ASPM, a brain size determinant in Homo sapiens’ and ‘Microcephalin, a gene regulating brain size, continues to evolve adaptively in humans’. Science 313:172; author reply 172. Dobson-Stone C, Gatt JM, Kuan SA, Grieve SM, Gordon E, Williams LM, Schofield PR (2007) Investigation of MCPH1 G37995C and ASPM A44871G polymorphisms and brain size in a healthy cohort. Neuroimage 37:394– 400. Dorus S, Vallender EJ, Evans PD, Anderson JR, Gilbert SL, Mahowald M, Wyckoff GJ, Malcom CM, Lahn BT (2004) Accelerated evolution of nervous system genes in the origin of Homo sapiens. Cell 119:1027–1040. Enard W, Przeworski M, Fisher SE, Lai CS, Wiebe V, Kitano T, Monaco AP, Paabo S (2002) Molecular evolution of FOXP2, a gene involved in speech and language. Nature 418: 869–872. Evans PD, Anderson JR, Vallender EJ, Choi SS, Lahn BT (2004a) Reconstructing the evolutionary history of microcephalin, a gene controlling human brain size. Hum Mol Genet 13:1139–1145.


Evans PD, Anderson JR, Vallender EJ, Gilbert SL, Malcom CM, Dorus S, Lahn BT (2004b) Adaptive evolution of ASPM, a major determinant of cerebral cortical size in humans. Hum Mol Genet 13:489–494. Evans PD, Gilbert SL, Mekel-Bobrov N, Vallender EJ, Anderson JR, Vaez-Azizi LM, Tishkoff SA, Hudson RR, Lahn BT (2005) Microcephalin, a gene regulating brain size, continues to evolve adaptively in humans. Science 309:1717–1720. Fay JC, Wu CI (2000) Hitchhiking under positive Darwinian selection. Genetics 155: 1405–1413. Frazer KA, Ballinger DG, Cox DR, et al (2007) A second generation human haplotype map of over 3.1 million SNPs. Nature 449:851–861. Gibbs RA, Rogers J, Katze MG, et al (2007) Evolutionary and biomedical insights from the rhesus macaque genome. Science 316: 222– 234. Gilad Y, Man O, Glusman G (2005) A comparison of the human and chimpanzee olfactory receptor gene repertoires. Genome Res 15: 224–230. Gimelbrant AA, Skaletsky H, Chess A (2004) Selective pressures on the olfactory receptor repertoire since the human-chimpanzee divergence. Proc Natl Acad Sci USA 101:9019– 9022.

Vallender

Hudson RR, Kreitman M, Aguade M (1987) A test of neutral molecular evolution based on nucleotide data. Genetics 116:153–159. Kelley JL, Madeoy J, Calhoun JC, Swanson W, Akey JM (2006) Genomic signatures of positive selection in humans and the limits of outlier approaches. Genome Res 16: 980– 989. Khaitovich P, Hellmann I, Enard W, Nowick K, Leinweber M, Franz H, Weiss G, Lachmann M, Paabo S (2005) Parallel patterns of evolution in the genomes and transcriptomes of humans and chimpanzees. Science 309: 1850–1854. King MC, Wilson AC (1975) Evolution at two levels in humans and chimpanzees. Science 188:107–116. Kouprina N, Pavlicek A, Mochida GH, Solomon G, Gersch W, Yoon YH, Collura R, Ruvolo M, Barrett JC, Woods CG, Walsh CA, Jurka J, Larionov V (2004) Accelerated evolution of the ASPM gene controlling brain size begins prior to human brain expansion. PLoS Biol 2:E126. McDonald JH, Kreitman M (1991) Adaptive protein evolution at the Adh locus in Drosophila. Nature 351:652–654. Mekel-Bobrov N, Gilbert SL, Evans PD, Vallender EJ, Anderson JR, Hudson RR, Tishkoff SA, Lahn BT (2005) Ongoing adaptive evolution of ASPM, a brain size determinant in Homo sapiens. Science 309:1720–1722. Mekel-Bobrov N, Posthuma D, Gilbert SL, Lind P, Gosso MF, Luciano M, Harris SE, Bates TC, Polderman TJ, Whalley LJ, Fox H, Starr JM, Evans PD, Montgomery GW, Fernandes C, Heutink P, Martin NG, Boomsma DI, Deary IJ, Wright MJ, de Geus EJ, Lahn BT (2007) The ongoing adaptive evolution of ASPM and Microcephalin is not explained by increased intelligence. Hum Mol Genet 16:600–608.

Exploring the Origins of the Human Brain

Neill D (2007) Cortical evolution and human behaviour. Brain Res Bull 74:191–205. Nielsen R, Bustamante C, Clark AG, Glanowski S, Sackton TB, Hubisz MJ, Fledel-Alon A, Tanenbaum DM, Civello D, White TJ, J JS, Adams MD, Cargill M (2005) A scan for positively selected genes in the genomes of humans and chimpanzees. PLoS Biol 3:e170. Rushton JP, Vernon PA, Bons TA (2007) No evidence that polymorphisms of brain regulator genes Microcephalin and ASPM are associated with general mental ability, head circumference or altruism. Biol Lett 3: 157– 160. Sabeti PC, Schaffner SF, Fry B, Lohmueller J, Varilly P, Shamovsky O, Palma A, Mikkelsen TS, Altshuler D, Lander ES (2006) Positive natural selection in the human lineage. Science 312:1614–1620. Sabeti PC, Varilly P, Fry B, et al (2007) Genomewide detection and characterization of positive selection in human populations. Nature 449:913–918. Shi P, Bakewell MA, Zhang J (2006) Did brainspecific genes evolve faster in humans than in chimpanzees? Trends Genet 22:608–613. Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123:585–595. Taudien S, Ebersberger I, Glockner G, Platzer M (2006) Should the draft chimpanzee sequence be finished? Trends Genet 22: 122– 125. The Chimpanzee Sequencing and Analysis Consortium (2005) Initial sequence of the chimpanzee genome and comparison with the human genome. Nature 437:69–87. The International HapMap Consortium (2005) A haplotype map of the human genome. Nature 437:1299–1320. Thornton KR, Jensen JD, Becquet C, Andolfatto P (2007) Progress and prospects in mapping recent selection in the genome. Heredity 98: 340–348.

Timpson N, Heron J, Smith GD, Enard W (2007) Comment on papers by Evans et al., and Mekel-Bobrov et al, on Evidence for Positive Selection of MCPH1 and ASPM. Science 317: 1036; author reply 1036. Voight BF, Kudaravalli S, Wen X, Pritchard JK (2006) A map of recent positive selection in the human genome. PLoS Biol 4:e72. Wang HY, Chien HC, Osada N, Hashimoto K, Sugano S, Gojobori T, Chou CK, Tsai SF, Wu CI, Shen CK (2007) Rate of evolution in brain-expressed genes in humans and other primates. PLoS Biol 5:e13. Wang YQ, Su B (2004) Molecular evolution of microcephalin, a gene determining human brain size. Hum Mol Genet 13:1131–1137. Williamson SH, Hubisz MJ, Clark AG, Payseur BA, Bustamante CD, Nielsen R (2007) Localizing recent adaptive evolution in the human genome. PLoS Genet 3:e90. Woods RP, Freimer NB, De Young JA, Fears SC, Sicotte NL, Service SK, Valentino DJ, Toga AW, Mazziotta JC (2006) Normal variants of Microcephalin and ASPM do not account for brain size variability. Hum Mol Genet 15: 2025–2029. Yu F, Hill RS, Schaffner SF, Sabeti PC, Wang ET, Mignault AA, Ferland RJ, Moyzis RK, Walsh CA, Reich D (2007) Comment on ‘Ongoing adaptive evolution of ASPM, a brain size determinant in Homo sapiens’. Science 316: 370. Yu XJ, Zheng HK, Wang J, Wang W, Su B (2006) Detecting lineage-specific adaptive evolution of brain-expressed genes in human using rhesus macaque as outgroup. Genomics 88:745–751. Zhang J (2003) Evolution of the human ASPM gene, a major determinant of brain size. Genetics 165: 2063–2070. Zhang J, Webb DM, Podlaha O (2002) Accelerated protein evolution and origins of human-specific features: Foxp2 as an example. Genetics 162:1825–1835.


177

Author Index Vol. 72, No. 2, 2008

Farris, S.M. 106 Hofmann, H.A. 89, 145 Holland, L.Z. 91 Lefebvre, L. 135 Pollen, A.A. 145

Short, S. 91 Shumway, C.A. 89, 123 Sol, D. 135 Vallender, E.J. 168 Wang, S.S.-H. 159

Subject Index Vol. 72, No. 2, 2008

Alternative splicing 91 Amphioxus 91 Behavior 123, 145 Behavioral ecology 106 Brain 123, 135 – development 145 – evolution 168 Cerebral cortex 106 Cichlid 123 Cognition 135 Conduction 159 Ecology 123 Evolution 123, 135, 159 Feeding habits 106 Fish 123 Genome duplication 91 Genomics 145 Habitat complexity 123 Hominid 168



Human evolution 168 Insects 106 Lancelet 91 MHB 91 Midbrain/hindbrain boundary 91 Molecular evolution 145, 168 Mushroom bodies 106 Neural crest 91 Neuroecology 135 Neuroethology 145 Neurogenetics 168 Optimization 159 Phylogenetics 145 Primate 168 Social behavior 106 Spindle cells 159 Tunicate 91 Vision 106

Neurobiological Lessons Learned from Comparative Studies: Evolutionary Forces Shaping Brain and Behavior, 19th Annual Karger Workshop San Diego, Calif., ... 1, 2007 (Brain, Behavior and Evolution)

Brain, Behavior and Epigenetics

Hormones, Brain and Behavior

Dementia 3 volumes (Brain, Behavior, and Evolution)

Learning and Memory: From Brain to Behavior

Introduction to Brain and Behavior

DHEA and the Brain (Nutrition, Brain, and Behavior)

Coffee, Tea, Chocolate, and the Brain (Nutrition, Brain and Behavior)

Coffee, Tea, Chocolate, and the Brain (Nutrition, Brain and Behavior)

The Neurobiology of Parental Behavior (Hormones, Brain, and Behavior)

Evolution of the Primate Brain: From Neuron to Behavior

Nutritional Neuroscience (Nutrition, Brain and Behavior)

An Introduction to Brain and Behavior

Evolution and Human Behavior

Shaping Primate Evolution: Form, Function, and Behavior (Cambridge Studies in Biological and Evolutionary Anthropology)

Microeconomics: Behavior, Institutions, and Evolution

Behavior Analysis And Behavior Modification

Emotional Cognition: From Brain to Behavior (Advances in Consciousness Research)

Reproductive Ecology and Human Evolution (Evolutionary Foundations of Human Behavior)

Meaningful Information: The Bridge Between Biology, Brain, and Behavior

Infectious Behavior: Brain-Immune Connections in Autism, Schizophrenia, and Depression

Handbook of molecular-genetic techniques for brain and behavior research

Human Behavior, Learning, and the Developing Brain: Atypical Development

Infectious Behavior: Brain-Immune Connections in Autism, Schizophrenia, and Depression

Human Behavior, Learning, and the Developing Brain: Typical Development

Cognitive Biology: Evolutionary and Developmental Perspectives on Mind, Brain, and Behavior (Vienna Series in Theoretical Biology)

The Brain and Behavior: An Introduction to Behavioral Neuroanatomy

Hormone Behavior Relations of Clinical Importance: Endocrine Systems Interacting with Brain and Behavior

The Neurophysics of Human Behavior: Explorations at the Interface of the Brain, Mind, Behavior, and Information

Neurobiological Lessons Learned from Comparative Studies: Evolutionary Forces Shaping Brain and Behavior, 19th Annual Karger Workshop San Diego, Calif., ... 1, 2007 (Brain, Behavior and Evolution)

Brain, Behavior and Epigenetics

Hormones, Brain and Behavior

Dementia 3 volumes (Brain, Behavior, and Evolution)

Learning and Memory: From Brain to Behavior

Introduction to Brain and Behavior

DHEA and the Brain (Nutrition, Brain, and Behavior)

Coffee, Tea, Chocolate, and the Brain (Nutrition, Brain and Behavior)

Coffee, Tea, Chocolate, and the Brain (Nutrition, Brain and Behavior)

The Neurobiology of Parental Behavior (Hormones, Brain, and Behavior)

Evolution of the Primate Brain: From Neuron to Behavior

Nutritional Neuroscience (Nutrition, Brain and Behavior)

An Introduction to Brain and Behavior

Evolution and Human Behavior

Shaping Primate Evolution: Form, Function, and Behavior (Cambridge Studies in Biological and Evolutionary Anthropology)

Microeconomics: Behavior, Institutions, and Evolution

Behavior Analysis And Behavior Modification

Emotional Cognition: From Brain to Behavior (Advances in Consciousness Research)

Reproductive Ecology and Human Evolution (Evolutionary Foundations of Human Behavior)

Meaningful Information: The Bridge Between Biology, Brain, and Behavior

Infectious Behavior: Brain-Immune Connections in Autism, Schizophrenia, and Depression

Handbook of molecular-genetic techniques for brain and behavior research

Human Behavior, Learning, and the Developing Brain: Atypical Development

Infectious Behavior: Brain-Immune Connections in Autism, Schizophrenia, and Depression

Human Behavior, Learning, and the Developing Brain: Typical Development

Cognitive Biology: Evolutionary and Developmental Perspectives on Mind, Brain, and Behavior (Vienna Series in Theoretical Biology)

The Brain and Behavior: An Introduction to Behavioral Neuroanatomy

Hormone Behavior Relations of Clinical Importance: Endocrine Systems Interacting with Brain and Behavior

The Neurophysics of Human Behavior: Explorations at the Interface of the Brain, Mind, Behavior, and Information

Recommend Documents