Montgomery Et Al11MBE Microcephaly Genes and Brains

Adaptive Evolution of Four Microcephaly Genes and theEvolution of Brain Size in Anthropoid PrimatesStephen H. Montgomery,1 Isabella Capellini,2 Chris Venditti,3 Robert A. Barton,2 andNicholas I. Mundy*,1

1Department of Zoology, University of Cambridge, Cambridge, United Kingdom2Evolutionary Anthropology Research Group, Department of Anthropology, Durham University, Durham, United Kingdom3School of Biological Sciences, University of Reading, Reading, United Kingdom

*Corresponding author: E-mail: [email protected].

Associate editor: Anne Stone

Abstract

The anatomical basis and adaptive function of the expansion in primate brain size have long been studied; however, we areonly beginning to understand the genetic basis of these evolutionary changes. Genes linked to human primarymicrocephaly have received much attention as they have accelerated evolutionary rates along lineages leading to humans.However, these studies focus narrowly on apes, and the link between microcephaly gene evolution and brain evolution isdisputed. We analyzed the molecular evolution of four genes associated with microcephaly (ASPM, CDK5RAP2, CENPJ,MCPH1) across 21 species representing all major clades of anthropoid primates. Contrary to prevailing assumptions,positive selection was not limited to or intensified along the lineage leading to humans. In fact we show that all four lociwere subject to positive selection across the anthropoid primate phylogeny. We developed clearly defined hypotheses toexplicitly test if selection on these loci was associated with the evolution of brain size. We found positive relationshipsbetween both CDK5RAP2 and ASPM and neonatal brain mass and somewhat weaker relationships between these genesand adult brain size. In contrast, there is no evidence linking CENPJ and MCPH1 to brain size evolution. The strongerassociation of ASPM and CDK5RAP2 evolution with neonatal brain size than with adult brain size is consistent with theseloci having a direct effect on prenatal neuronal proliferation. These results suggest that primate brain size may have at leasta partially conserved genetic basis. Our results contradict a previous study that linked adaptive evolution of ASPM tochanges in relative cortex size; however, our analysis indicates that this conclusion is not robust. Our finding that thecoding regions of two widely expressed loci has experienced pervasive positive selection in relation to a complex,quantitative developmental phenotype provides a notable counterexample to the commonly asserted hypothesis that cis-regulatory regions play a dominant role in phenotypic evolution.

Key words: ASPM, MCPH1, CDK5RAP2, CENPJ, brain, neurogenesis, primates.

IntroductionThe expansion of the brain, and in particular the neocortex,is a major hallmark of primate evolution (Jerison 1973;Martin 1990). After correcting for allometric scaling withbody mass, primates have larger brains than most othermammals (Martin 1990; Barton 2006b) and both absoluteand relative brain size have increased along multiple, inde-pendent primate lineages (Montgomery et al. 2010). Theadaptive significance and anatomical basis of the diversityof primate brains has long been studied using comparativemethods (for review, see Falk and Gibson 2001; Finlay et al.2001; Barton 2006a), but the investigation of the geneticbasis of primate brain expansion has only begun relativelyrecently and is currently a topic of intense interest.

The convergent evolution of increased brain size in dif-ferent lineages provides an opportunity to study whetherthe independent evolution of complex traits involves con-vergence at the molecular level (Arendt and Reznick 2007)and may provide insights into lineage-specific evolution,for example, on the human lineage. Both scans of brain-expressed genes in published primate genomes (Dorus,

Vallender, et al. 2004; Shi et al. 2006; Yu et al. 2006; Wanget al. 2007) and studies of candidate genes (e.g., Enard et al.2002; Burki and Kaessmann 2004; Wang et al. 2005) havemostly focused on identifying changes along the lineageleading to humans and have largely ignored convergentincrease in brain size in multiple primate lineages.

One group of genes of particular interest in relation tothe evolution of gross brain size is the microcephaly genes.Autosomal recessive primary microcephaly is a congenitaldisorder characterized by reduced growth of the cerebralcortex in the absence of environmental, metabolic, or cy-togenetic etiologies (Bond and Woods 2006; Cox et al.2006). In humans, it is inherited as a recessive Mendeliantrait involving at least eight loci, of which five have nowbeen identified at the molecular level: ASPM, MCPH1,CDK5RAP2, CENPJ (Jackson et al. 1998; Bond et al. 2002,2005; Thornton and Woods 2009) and the more recentlyidentified STIL (Kumar et al. 2009).

The five genes are expressed in the fetal brain duringneurogenesis (Bond et al. 2002, 2005; Jackson et al. 2002;Kouprina et al. 2005; Kumar et al. 2009). ASPM, CDK5RAP2,

© The Author 2010. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, pleasee-mail: [email protected]

Mol. Biol. Evol. 28(1):625–638. 2011 doi:10.1093/molbev/msq237 Advance Access publication October 20, 2010 625

Research

article

and CENPJ all have roles in centrosome or microtubule for-mation (Bond and Woods, 2006; Cox et al. 2006; Fish et al.2006; Buchman et al. 2010) and can affect neurogenic mi-tosis by influencing the spindle pole and astral microtubulenetwork (Fish et al. 2006; Fong et al. 2008; Cormier et al.2009; Buchman et al. 2010). MCPH1 functions in theDNA damage response pathway and apoptosis (Rickmyreet al. 2007; Wood et al. 2007, 2008) and may also affect thetiming of cell cycle progression (Brunk et al. 2007). Bothapoptosis and cell cycle length are known to have signifi-cant roles in brain development (Roth and D’Sa 2001;Calegari and Huttner 2003). The function of STIL is less wellstudied, but it localizes to the centrosome and hassome functional similarities to ASPM (Kumar et al. 2009;Thornton and Woods 2009).

The main hypotheses for how the number of neuronscould increase during brain expansion (Rakic 1988, 1995;Caviness et al. 1995; Kriegstein et al. 2006) rely on switchesbetween symmetric and asymmetric cell divisions, viachanges in spindle pole orientation, at a particular stageof neurogenesis. The functions of microcephaly genesare therefore consistent with the developmental mecha-nisms proposed to have facilitated brain expansion (Gotzand Huttner 2005; Cox et al. 2006; Kriegstein et al. 2006).Notably, the phenotypes exhibited by individuals with mi-crocephaly show that these loci affect cortical surface area,not thickness, consistent with a role in regulating the size ofthe neural progenitor pool (Desir et al. 2008). Interestingly,recent studies in humans identified single nucleotide poly-morphisms in ASPM, CDK5RAP2, and MCPH1 associatedwith total brain size or cortical surface area (Wang et al.2008; Rimol et al. 2010) but not cortical thickness, an ob-servation that again is consistent with a role in controllingthe size of the neural progenitor pool (Montgomery andMundy 2010; Rimol et al. 2010).

Previous studies of the molecular evolution of the firstfour microcephaly loci to be identified supported the hy-pothesis that they have been subject to positive selection(Zhang 2003; Evans, Anderson, Vallender, Choi, and Lahn2004; Evans, Anderson, Vallender, Gilbert, et al. 2004;Kouprina et al. 2004; Wang and Su 2004; Evans et al.2006) but provided no direct evidence that the loci wereinvolved in brain evolution as brain size was not incorpo-rated into their analyses and did not include a diverse phy-logenetic sample of species. A recent study analyzed ASPMevolution in relation to brain size in primates and con-cluded that branches with high relative telencephalon vol-ume (reported as cerebral cortex) were associated withpositive selection on ASPM (Ali and Meier 2008).

For a comprehensive understanding of the role of genesin primate brain evolution, broad comparisons across theprimate phylogeny incorporating relevant phenotypes areneeded (Carroll 2003; Goodman et al. 2005; Barton 2006a;Pollen and Hoffmann 2008; Vallender 2008). An importantissue is which aspects of brain phenotype are most salient.Measures of brain size corrected for body size (i.e., relativebrain size) are frequently used in studies investigating brainevolution as these take into account the strong correlation

between brain and body mass (Barton 2006b). However,given the implied functions of the four microcephaly genesin regulating the proliferation and survival of neurons, ab-solute brain mass may be a more relevant phenotypic mea-sure as in primates it increases linearly with the totalnumber of neurons (Herculano-Houzel et al. 2007). Inagreement with quantitative genetic analysis of brainand body size allometry (Lande 1979), it has recently beenshown that primate brain and body size differ in their evo-lutionary trajectories (Montgomery et al. 2010) suggestingthat these two traits must be developmentally and genet-ically decoupled to some extent despite their closely cor-related evolution. Crucially, because primate neocorticalneurogenesis is largely restricted to prenatal development(Rakic 1988, 2002; Bhardwaj et al. 2006) andmicrocephaly isprimarily a disorder of fetal brain growth (Cox et al. 2006),microcephaly gene evolution should be more closely re-lated to neonatal brain size than to adult brain size. Post-natal brain growth is largely driven by gliogenesis (Low andCheng 2006), axon growth (Sauvageot and Stiles 2002), andmyelination (Sowell et al. 2001) rather than by productionof new neurons. There are only two known sites in the pri-mate brain, which are small and noncortical, in which sub-stantial postnatal neurogenesis occurs (Jabes et al. 2010).Indeed, apoptosis eliminates large numbers of neurons(Buss et al. 2006). Variation in these and other nonneuro-genic processes will reduce the relationship between brainsize and neuron number as development progresses, weak-ening any association with the molecular evolution ofgenes under selection in relation to prenatal neurogenesis.Indeed, patterns of postnatal brain growth vary consider-ably across primates (Leigh 2006). Finally, if there is an as-sociation with adult brain size, given the specific effect ofmicrocephaly on the development of the cerebral cortexand their functions in cortical neurogenesis (Cox et al.2006; Thornton and Woods 2009), we might also predicta stronger association with adult neocortex size than adultwhole brain size. Unfortunately, there are insufficient com-parative data available on volumes of neonatal brainregions to test this hypothesis.

Alternative hypotheses to explain the high evolutionaryrates of microcephaly genes also exist. The four loci arewidely expressed throughout the body and ASPM,CDK5RAP2 and CENPJ are particularly highly expressedin the testis (Bond et al. 2005; Kouprina et al. 2005), wheremany genes have been shown to be under sexual selectionin primates (Dorus, Evans, et al. 2004; Clark and Swanson2005; Ramm et al. 2008). However, the precise function ofthese genes in testes development and function is still un-known. For ASPM, a possible ciliary function led to the sug-gestion of a role in sperm flagellar movement that mayaffect sperm locomotion and hence be targeted by sexualselection (Ponting 2006). If the microcephaly genes do haveimportant roles in the testes or sperm, their high rates ofevolution may be associated with levels of sexual selectionand be unrelated to changes in brain size. Hence, explicittests are required before the molecular evolution of micro-cephaly loci can be linked to brain evolution.

Montgomery et al. · doi:10.1093/molbev/msq237 MBE

626

Here, we investigate the molecular evolution of ASPM,CDK5RAP2, CENPJ, and MCPH1 in relation to brain size inanthropoid primates. First, we test whether these loci areunder positive selection across anthropoids and whether ornot different anthropoid clades have experienced differentselective regimes. Second, we explore the association be-tween the rate of molecular evolution of microcephalygenes and measures of brain size, predicting a positiveand stronger association between these genes and absoluteneonatal brain size than adult brain size, a stronger asso-ciations with absolute brain size as compared with relativebrain size, and potentially a stronger relationship with neo-cortex size than whole brain size. Finally, we investigate therelationship between microcephaly genes and relative tes-tes size, a commonly used phenotypic correlate for spermcompetition and sexual selection (Harcourt et al. 1995;Ramm and Stockley 2010), to test the hypothesis that theseloci may have been under sexual selection. We find thatwhereas all four loci have been targets of selection through-out primate evolution, ASPM and CDK5RAP2 but notMCPH1 and CENPJ show positive associations with absoluteneonatal brain but not with any measure of relative brainsize or relative testis size, suggesting a role in the evolutionof total neuronal number.

Materials and Methods

Phenotypic DataData for body, brain mass, and volumes of specific brainregions were obtained from previously published data(Bauchot and Stephan 1969; Stephan et al. 1981; Zillesand Rehkemper 1988), leading to a data set of 37 primategenera including 14 catarrhines, 12 platyrrhines, one tarsierand 10 strepsirhines (supplementary table S1, SupplementaryMaterial online). Data on neonatal brain size (22 taxa) wereobtained from Capellini et al. (submitted). Data on testismass for 30 genera were taken from Harcourt et al. (1995).

To test the hypothesis that the evolution of the micro-cephaly genes should be more strongly associated with ab-solute brain mass than measures of relative brain size, wecalculated relative brain mass by performing a phylogenet-ically controlled regression analysis between log(brainmass) and log(body mass), and for neocortex, log(neocor-tex volume) was regressed against log(rest of brain volume)separately (following Barton 1998). For testis, log(testismass) was regressed against log(body mass). These analyseswere performed using a phylogenetically controlled regres-sion using phylogenetic generalized least squares (PGLS)models in BayesTraits (Pagel et al. 2004) with maximumlikelihood and 1,000 runs for each analysis. Residual valuesfrom the regression line were calculated for each taxon, andthese were used as values of relative brain size and relativetestis mass in all subsequent analyses. All phenotypic dataare provided in table S1 (Supplementary Material online),and additional phenotypic analyses are presented in thesupplementary information (Supplementary Material on-line). With PGLS, the phylogeny is converted into a vari-ance-covariance matrix, where the diagonal of the

matrix gives information on the path length from root-to-tips (the ‘‘variance’’) and the off-diagonal values ofthe matrix provide information on the shared evolutionaryhistory of any pair of species, that is the time from the rootto the last common ancestry (the ‘‘covariance’’) (Pagel1997, 1999; Freckleton et al. 2002). With PGLS regression,the variance-covariance matrix is included into the errorterm of the regression model, and the resulting estimatedregression parameters (i.e., slopes and intercepts) are ‘‘phy-logenetically controlled’’ (Pagel 1997, 1999; Freckleton et al.2002).

PhylogenyWe used a genus level composite phylogeny of primatesusing published trees. The topology is taken from Good-man et al. (2005) for haplorhine primates and Horvathet al. (2008) for strepsirhines. Proportional branch lengthswere obtained from recent studies of primate divergencedates (Purvis 1995; Page and Goodman 2001; Poux andDouzery 2004; Opazo et al. 2006) scaled to agree with datesof divergence for the deeper primate nodes estimated bySteiper and Young (2006). The tree obtained therefore hasbranch length information in time and is ultrametric (sup-plementary fig. S1, Supplementary Material online).

Laboratory MethodsGenomic DNA samples had previously been extracted fromtissue samples using Qiagen kits. Sequence data from pre-vious studies and primate genomes were collected from theonline databases, GenBank and Ensembl. From these se-quences, primers were designed using Primer3Plus (Unter-gasser et al. 2007). We sequenced exons that had previouslybeen shown to have accelerated rates of evolution or con-tained a large proportion of the coding sequence. ForASPM, we sequenced exons 3 and 18, totaling 6,235 bp(60% of the coding region). For MCPH1, three exons weresequenced: 8, 11, and 13, totaling 1,556 bp (62% of the cod-ing region). Exons 2 and 7 were sequenced for CENPJ, to-taling 1,556 bp (52% of the coding region). We sequenced 7of 38 exons of CDK5RAP2: exons 12, 20, 21, 24, 25, 32, and 33(total 2,120 bp; 37% of the coding sequence). Polymerasechain reactions and sequencing on both strands were per-formed using standard protocols (for further details andprimers, see supplementary information and table S2, Sup-plementary Material online).

Sequences were edited in SEQMAN v. 5.05 (DNASTARInc.) and aligned and checked in ClustalW in MEGA 4.0(Tamura et al. 2007). Exons of each locus were concate-nated and subsequently analyzed together; alignmentsare available on request. Sequences were obtained for 5apes, 5 Old World monkeys, and 10 New World monkeys,representing all major clades of anthropoid primates,shown in figure 1. Where phenotypic data were not avail-able for the species sequenced, we used closely related con-generic species for which data were available. Newlysequenced data have been submitted to GenBank (for ac-cession numbers, see table S3, Supplementary Material on-line). We used the Strepsirhines Microcebus murinus and

Evolution of Microcephaly Genes in Anthropoids · doi:10.1093/molbev/msq237 MBE

627

Otolemur garnettii; Ensembl IDs are shown in table S3 (Sup-plementary Material online).

Molecular EvolutionA commonmeasure used to infer selection pressures actingon coding regions of genes is the ratio of rates of nonsy-nonymous to synonymous fixed base changes. Estimationof dN/dS ratios (x) was carried out using a codon-basedmaximum likelihood method (PAML version 4; Yang2007). Several analyses were performed to test the hypoth-esis that the four loci have experienced positive selectionacross primates, in particular, in relation to brain size evo-lution. Nested models are compared using the likelihoodratio test statistic (!2[loglikelihood1 – loglikelihood2])to critical values of the chi-square distribution and degrees

of freedom as the difference in the number of parametersestimated by each model.

Site and Branch ModelsTo detect positive selection across primates, we imple-mented the site models. These allow the x to vary amongsites but not across lineages (Nielsen and Yang 1998; Yanget al. 2000). Model M1a (nearly neutral) allows sites to fallinto twocategorieswithx,1 (purifying selection) andx51 (neutral evolution), whereas model M2a (positive selec-tion) allows sites to fall into three categories with x , 1,x 5 1, and x . 1 (positive selection) (Yang et al. 2005).

In addition, we used the branch models to test whetherthe dN/dS of lineages leading to humans was significantlyhigher than nonhuman lineages and whether dN/dS

FIG. 1. Phylogeny of anthropoid primates with an indication of (a) absolute brain mass—the area of the square shows the mass of the brain asa percentage of the human brain and (b) relative brain mass (on body mass), taken from Montgomery et al. (2010). *The second species ofColobinae varied between loci (see supplementary information, Supplementary Material online).


628

significantly differed between Apes, Old World Monkeys,and New World Monkeys.

Root-to-Tip dN/dS and Gene–Phenotype AssociationsThe branch models were used to estimate the average dN/dS ratio from the ancestral anthropoid to each terminalspecies tip. These values were then set as species dataand used in a PGLS regression with measures of brain sizein BayesTraits, as explained above (Pagel 1999; Pagel et al.2004; Organ et al. 2007). Previous analyses that have testedfor correlations between phenotypes and dN/dS ratios us-ing similar methods have typically used the dN/dS of theterminal branch (e.g., Dorus, Evans, et al. 2004; Nadeau et al.2007). However, although a species’ phenotype reflects thewhole phenotypic evolution, the dN/dS of the terminalsbranch does not reflect the whole genotypic evolution.The root-to-tip dN/dS is more inclusive of the evolutionaryhistory of a locus and is a property of the species tip ratherthan the terminal branch and is therefore more suitable forregressions against phenotypic data from extant species. Inaddition, by analyzing the rate of evolution since the lastcommon ancestor of the species in our data set, allbranches are the same length and therefore not subjectto temporal effects on dN/dS (Wolf et al. 2009). One as-sumption of regression analysis is that the residuals ofthe model are normally distributed. As the residuals ofthe regression using dN/dS ratios were not normallydistributed, we used log10(dN/dS) to improve normality.Residuals of regression analysis with log-transformed dN/dS did not violate assumptions of normality and constantvariance.

First, we examine the relationship between microceph-aly gene molecular evolution and the evolution of absoluteand relative neonatal brain size. As we specifically hypoth-esize a positive association between brain size and the se-lection pressure on these loci, the significance of theregression coefficient was determined using a one-tailedt-test. A significant negative association would suggestan increase in ‘‘purifying’’ selection has acted on a locusas brain mass increased and, although interesting, wouldsuggest that the locus does not contribute to the geneticbasis of ‘‘change’’ in that phenotype and could not explainwhy the locus has evolved adaptively. Hence, we meet bothrecently suggested requirements for justifying the use ofone-tailed tests; we explain why we hypothesize an associ-ation in a particular direction, and why the opposite pat-tern can be treated the same as a nonsignificant trend inthe expected direction (Ruxton and Neuhauser 2010).

As the size of our data set for this analysis is limited bythe availability of neonatal brain size data and the imper-fect overlap between phenotype and gene sequence datasets, it is highly likely that the small sample size will result inlow power to reject the null hypothesis. To minimize thechances of Type I errors, we restrict our analyses to a smallnumber of critical tests, and we determine the specificity ofrelationships for microcephaly genes by testing for associ-ations with other genes having no known role in neurogen-esis. In particular, we test for associations with alternative

phenotypes and test for associations between the evolu-tion of genes with no known role in neurogenesis withbrain size. In addition, a Jackknife approach was takento test the robustness of the associations found and toidentify any outliers that have a dominant effect on theslope of the regression.

We subsequently explored the relationship with volumeof adult whole brain and neocortex. This analysis was per-formed to test our hypothesis that genes involved in neuro-genesis should be more strongly associated with neonatalthan adult brain mass and that if there is an associationwith adult brain size it may be stronger for neocortex sizethan whole brain size. Comparisons of nonnested modelswere performed using Akaike Information Criterion (AIC:calculated as (2 " number of parameters) # (!2 " log[li-kelihood])) to identify the best supported model, wherea lower value indicates a better fitting model, and a differ-ence between models greater than two suggests a substan-tial difference (Burnham and Anderson 2002).

In addition to the standard dN/dS ratios, we used mul-tiple regressions to investigate the association betweenphenotype and dN while controlling for dS. Here, we pre-dict a negative association between brain size and dS givenknown relationships between dS and life-history traits, suchas body size (Nikolaev et al. 2007). Conversely, a locus thatis a target of selection in relation to brain size may showa positive association with dN. Both approaches examinevariation in dN and dS relative to one another, but theymake different assumptions about the nature of the under-lying relationship. For example, a significant dN/dS–phenotype relationship suggests an association betweenphenotypic evolution and selection acting on a locusand may be obtained when both change together ina tightly correlated fashion but with one changing at a fasterrate than the other (so that the ratio correlates with theabsolute value of the changes), whereas in this case, a mul-tiple regression would show no significant correlation.Hence, differences in the results obtained may be informa-tive about the nature of the gene-phenotype correlation.

In most cases, the phenotypic data is based on a smallnumber of individuals, and the degree of interspecific var-iation is unknown. However, where interspecific variationgreatly exceeds intraspecific variation, as is expected to bethe case for brain size, results of comparative analyses arenot biased by intraspecific variation (see Nunn and Barton2001). It is also likely that error introduced by samplingsmall numbers of individuals will lead to an underestimateof correlation coefficients between two traits (Nunn andBarton 2001; Ives et al. 2007).

Finally, we used branch-site models to test for associa-tions between positive selection and brain evolution, butthis method did not produce informative results (see sup-plementary information, Supplementary Material online).

ResultsOur full data set comprises sequence data from 21 speciesfor each gene including representative species from all


629

major anthropoid clades and 11–17 newly sequenced spe-cies for each locus. The coverage of full coding sequence inthe data set comprises 60% for ASPM (from 2 exons), 37%for CDK5RAP2 (from 7 exons), 52% for CENPJ (from 2exons), and 62% for MCPH1 (from 3 exons) (see MaterialsandMethods and supplementary table S3, SupplementaryMaterial online). Previous data sets for CDK5RAP2 andCENPJ have included only four species (Evans et al.2006). With the exception of recent ASPM study by Aliand Meier (2008), which was published after the comple-tion of our data set, analyses of this gene have considered3 (Zhang 2003), 7 (Evans, Anderson, Vallender, Gilbert,et al. 2004), and 8 species (Kouprina et al. 2004), whereasstudies of MCPH1 have included 7 (Evans, Anderson,Valleder, Choi and Lahn 2004) and 13 species (Wangand Su 2004). These studies were particularly lacking ina diverse range of NewWorld Monkeys. Although we havesequenced only partial coding sequences, we have focusedon regions of the gene which contain functionally impor-tant domains or which have previously been shown tohave accelerated rates of evolution. Our data set thereforeallows us to examine how widespread selection on theseregions has been across anthropoids and to explore therelationship between this selection and phenotypicevolution.

Pervasive Adaptive Evolution in MicrocephalyGenesWe first examined whether a signal of positive selection ispresent in four microcephaly loci by performing site modeltests using a codon-based maximum likelihood method(Yang 2007; table 1). All four genes showed evidence forpositive selection across anthropoids, with estimatedomega of 2.57–5.39 at 2.28 – 6.50% of sites across the loci.The most significant results are for ASPM and CDK5RAP2that have 2.3% and 6.5% of sites having an omega of 5.39and 4.42, respectively. In CDK5RAP2, a notable feature is theclustering of sites identified as being under positive selec-tion using Bayes empirical Bayes (P. 95%) (positions 929–977 and 1683–1690), these mostly fall within an SMCdomain (Evans et al. 2006), these domains are thoughtto play a role in the chromosome segregation, regulation,and repair (Hirano 2006), but the functional significance ofthese particular sites are unknown.

Previous authors have proposed that ASPM andMCPH1have evolved at a higher rate along lineages leading fromthe last common ancestor of apes to modern humans than

on other lineages (Evans, Anderson, Vallender, Choi, andLahn 2004; Evans, Anderson, Vallender, Gilbert, et al.2004). We tested this explicitly using a branch modeland found that the evolutionary rate of change in lineagesleading to humans does not significantly differ from that inother lineages (table 2a), which is congruent with the factthat the proportional change in brain mass along this lin-eage is also not exceptional compared with the rest of pri-mate brain size evolution (Montgomery et al. 2010). Inaddition, we tested whether the evolutionary rate of thefour loci differed between apes, Old World Monkeys,and New World Monkeys and found no significant differ-ences (table 2b). Finally, we performed site models to testfor positive selection on each of the three clades separately.Although the results are influenced by differences in thenumber of sequences and the time depth for each clade,we found no instance where positive selection was found inApes and not in Old or New World Monkeys (supplemen-tary table S4, Supplementary Material online).

Associations between Gene Evolution and BrainSizeTo explicitly test the link between the molecular evolutionof microcephaly genes and brain size, we performed severaltests. We first performed a phylogenetically controlled re-gression analysis in BayesTraits (Pagel 1999; Pagel et al.2004) between the root-to-tip dN/dS, estimated usingthe branch models, and neonatal brain mass.

We found a significant association between the molec-ular evolution of CDK5RAP2 and absolute neonatal brainmass (t11 5 1.95, P5 0.039, R2 5 0.255) but no significantassociation between ASPM, CENPJ, orMCPH1 and this trait(table 3). For ASPM, however, Callithrix representeda strong outlier (fig. 2, supplementary table S5, Supplemen-tary Material online), and when this species was removed,the association between the dN/dS of ASPM and neonatalbrain mass became significant (t10 5 2.42, P5 0.018, R2 50.369). Callithrix, and the other Callitrichids, show highrates of evolution of ASPM (supplementary table S5, Sup-plementary Material online) but have the smallest brainmasses among the anthropoid primates (Stephan et al.1981); however, this is due to a secondary reduction inbrain mass in this taxon (Ford 1980; Montgomery et al.2010). The significance of the regression between ASPMand neonatal brain mass was not affected by the removalof any other species. From here on, unless otherwise stated,regressions for ASPM were performed without the

Table 1. Site Models Detecting Positively Selected Sites for Anthropoid Primatesa.

Gene NlnL

(null/M1a)lnL(positive

selection/M2a)LRT

StatisticP

ValueProportion

of Sites, v > 1 vPositively

Selected Sites

ASPM 21 220522.7 220487.0 71.4 <0.001 0.0228 5.394 195, 1407, 1437, 1558, 2016, 2185CDK5RAP2 21 27858.2 27818.7 79.0 <0.001 0.0650 4.418 853, 881, 929, 930, 948, 964, 977, 1581, 1683,

1687, 1688, 1690CENPJ 21 27695.8 27684.1 23.4 <0.001 0.0255 3.897 527, 813MCPH1 21 26375.3 26369.0 12.6 0.002 0.0558 2.574 —a Positively selected sites identified using Bayes empirical Bayes are shown in the right-hand column, numbered according to the full human coding sequence. Only siteswith P . 95% are presented; sites in bold have P . 99%. LRT, likelihood ratio test.


630

Callitrichids. Removing Callithrix does not reveal a signifi-cant association with neonatal brain mass for either CENPJ(t105 0.81, P5 0.218, R2 5 0.061) orMCPH1 (t105!0.55,P5 1.000, R2 5 0.029). The high root-to-tipMCPH1 dN/dSratio of Pan (1.39) appeared to be heavily influenced bya small number of synonymous substitutions on the termi-nal Pan branch (1 synonymous change, compared with 6synonymous changes on the Homo branch), and we noteprevious studies have found a much lower dN/dS ratio onthe terminal Pan lineage using the full coding sequence(Evans, Anderson, Vallender, Gilbert, et al. 2004; Wangand Su 2004). We therefore repeated the regressions ex-cluding the Pan data point but still found no significantresult (data not shown). No locus showed any associationwith relative neonatal brain size (table 3).

We next explored whether the association with brainmass can be observed for adult phenotypes by performingregressions of root-to-tip dN/dS with absolute and relativeneocortex and whole brain size. We found no significantassociation with any measure of absolute brain size forany locus (supplementary table S6, Supplementary Mate-rial online). To confirm this result, we performed the sameregressions but using only the species used in the neonatalregressions. Using this reduced data set, we subsequentlyfound a significant association between both absolute neo-cortex and whole brain mass with ASPM (whole brain:t10 5 2.732, P 5 0.010, R2 5 0.427; neocortex: t10 52.980, P5 0.007, R2 5 0.470) and CDK5RAP2 (whole brain:

t11 5 2.131, P 5 0.028, R2 5 0.292; neocortex: t11 5 2.00,P 5 0.035, R2 5 0.267) but not MCPH1 or CENPJ (notshown). To identify the cause of this discrepancy, we per-formed a reverse jackknife, repeating the regressions withthe full data set of adult phenotypes and removing eachspecies, one at a time (supplementary table S7, Supplemen-tary Material online). We found that for both ASPM andCDK5RAP2, the removal of Papio results in a significant as-sociation with absolute neocortex and whole brain mass(ASPM: whole brain: t15 5 2.063, P 5 0.028, R2 5 0.221:neocortex: t14 5 2.128, P5 0.0257, R2 5 0.244; CDK5RAP2:whole brain: t18 5 2.418, P5 0.013, R2 5 0.245, neocortex:t17 5 2.238, P 5 0.019, R2 5 0.228), whereas removing anyother species does not have this affect (supplementary ta-ble S5, Supplementary Material online), suggesting Papio isan outlier to a general trend.

We subsequently added the Callitrichids back into theregressions with ASPM and find that the addition of anyone Callitrichid or all three together substantially reducesor negates the association (addition of Saguinus alone:t16 5 0.612, P 5 0.274, Leontopithecus alone: t16 51.842, P 5 0.042, Callithrix alone: t16 5 0.612, P 5 0.265,all three: t185 0.302, P5 0.2383). In contrast, after removalof all three Callitrichids, the association with CDK5RAP2 isstill significant (t15 5 0.612, P5 0.042) and removal of anysingle species (removal of Saguinus: t17 5 2.010, P5 0.031,Leontopithecus: t17 5 2.359, P 5 0.015, Callithrix: t17 52.010, P 5 0.031) or any pair of Callitrichids (retention

Table 2. Branch and Site Model Tests on Separate Anthropoid Clades.

a)Model 2 Model 0

GenedN/dS lineagesleading to Homo

dN/dS ‘‘nonhuman’’lineages lnL(M2) lnL(M0)

LikelihoodRatio Statistic P value

ASPM 0.489 0.466 221351.430 221351.451 0.042 0.838CDK5RAP2 0.729 0.676 28718.412 28718.440 0.054 0.816CENPJ 0.623 0.503 28664.800 28664.986 0.373 0.542MCPH1 0.793 0.579 27301.153 27301.505 0.704 0.401

b)Model 2 Model 0

Gene dN/dS Apes dN/dS OWM dN/dS NWM lnL(M2) lnL(M0)LikelihoodRatio Statistic P value

ASPM 0.473 0.415 0.479 221350.890 221351.451 1.122 0.571CDK5RAP2 0.781 0.627 0.687 28717.931 28718.440 1.018 0.601CENPJ 0.582 0.549 0.522 28663.445 28664.986 3.081 0.214MCPH1 0.606 0.724 0.538 27300.305 27301.505 2.401 0.301

NOTE.—OWM, Old World Monkeys; NWM, New World Monkeys.

Table 3. Phylogenetically Controlled Regression Analysis between Root-to-Tip dN/dS and Brain Size in Anthropoid Primates.

Absolute Neonatal Brain Mass Relative Neonatal Brain Mass

Gene n t-Statistic P Value R2 t-Statistic P Value R2

ASPM 13 0.040 0.484 0.000 0.797 0.221 0.055ASPM (withoutCallitrichids) 12 2.420 0.018 0.369 0.967 0.178 0.086CDK5RAP2 13 1.955 0.039 0.255 20.500 1.000 0.022CENPJ 13 0.631 0.270 0.035 21.204 1.000 0.116MCPH1 13 20.697 1.000 0.042 21.126 1.000 0.103


631

of Saguinus: t16 5 1.886, P 5 0.039, Leontopithecus: t16 51.988, P5 0.032, Callithrix: t165 2.460, P5 0.012) does notsignificantly affect the association. This confirms Callitri-chids are outliers to the general trend for ASPM but notfor CDK5RAP2. It is notable that Leontopithecus has a muchsmaller effect on the ASPM result than Saguinus and Calli-thrix suggesting a complex pattern of selection on this lo-cus, more data will be required to tease out the patternwithin Callitrichids.

In addition, we repeated the analysis using relative meas-ures of adult brain and neocortex sizes. For ASPM, therewas a significant association with relative adult brain sizeon both the full data set (t19 5 1.77, P5 0.046, R2 5 0.142;without Callitrichids: t15 5 2.241, P 5 0.020, R2 5 0.239)and the reduced data set used for regressions with neonataldata (t10 5 2.59, P 5 0.013, R2 5 0.402). However, in bothcases, the result was dependent on the human data point

and when this was removed the association is no longersignificant (adult: t18 5 0.574, P5 0.287, R2 5 0.018; with-out Callitrichids: t15 5 0.923, P 5 0.185, R2 5 0.054; neo-nate: t18 5 0.880, P 5 0.200, R2 5 0.079). No othersignificant relationship was detected (supplementary tableS6, Supplementary Material online).

We next used AIC to determine the best supportedmodel for the associations between ASPM and CDK5RAP2and neonatal and adult brain mass using only the speciesfor which data are available for both phenotypes. For bothloci, the regression with neonatal brain mass has a substan-tially lower AIC than whole brain mass (ASPM: neonateAIC 5 2.44, adult AIC 5 7.39; CDK5RAP2: neonateAIC 5 12.19, adult AIC 5 14.67) suggesting a closerassociation with neonatal brain mass. Comparing adultneocortex and whole adult brain size, we found nosubstantial difference (ASPM: neocortex AIC5 7.43, whole

FIG. 2. Phylogenetically controlled regressions between root-to-tip dN/dS and absolute neonatal brain mass: for (a) ASPM, (b) CDK5RAP2, (c)CENPJ, and (d) MCPH1. Data points are raw species values, the phylogenetically controlled regression line was estimated in BayesTraits andsuperimposed on top of species data. For ASPM, two lines are displayed, the dashed line shows the regression when the outlier, Callithrix(labeled C) is included and the solid line shows the regression when it is excluded.


632

AIC 5 7.40; CDK5RAP2: neocortex AIC 5 16.29, wholeAIC 5 14.67). Hence, both the results are for ASPM andCDK5RAP2 are highly consistent.

To further explore the relationship between brain evolu-tion and themolecular evolution of themicrocephaly genes,we performed multiple regressions with neonatal brain sizeand root-to-tip dN anddS (log transformed) as independentvariables. Significant negative partial regression coefficientswere found, as predicted, for the three cytoskeletal genesand dS (one-tailed ASPM: t8 5 !2.958, P 5 0.007;CDK5RAP2: t9 5 !1.919, P 5 0.042; CENPJ: t9 5 !1.859,P 5 0.046) but interestingly not for MCPH1 (t9 5 !1.294,P 5 1.000). Neither CENPJ nor MCPH1, which show no as-sociation between brain size and dN/dS, show a significantassociation with dN (CENPJ: t9 5 0.049, P5 0.481; MCPH1:t95!2.322, P5 1.000). ForASPM, we do see an associationbetween dN and neonatal brain size (ASPM: t85 2.032, P50.035) but we do not find a significant association forCDK5RAP2 (t9 5 1.127, P 5 0.142). This suggests the asso-ciation between ASPM and dN/dS may be driven predom-inantly by an accelerated dN, whereas the association forCDK5RAP2 may have a more complex basis.

As a final analysis, we used the branch-site test to de-tect positive selection on branches along which brain sizeis estimated to have expanded greatly (see Ali and Meier2008). However, as discussed in the supplementary infor-mation (Supplementary Material online), we found thismethod was not informative as it is possible to get a pos-itive result by selecting branches at random. This is likelyto be due to positive selection acting across the phylog-eny, as demonstrated by our site model tests, and stronglyquestions the strength of the methodology used andresults obtained by Ali and Meier (2008), namely that epi-sodes of adaptive evolution of ASPM are specifically asso-ciated with expansion of the relative size of thetelencephalon.

Controls for Specificity of Gene–Phenotype AssociationsTo exclude the possibility that the gene–phenotype corre-lations reported above are coincidental, we investigatedeight loci (supplementary table S8, Supplementary Materialonline), with no known function in neurogenesis, for whichdata were already available for a reasonably large number ofspecies (n5 10–20) across the anthropoid phylogeny. Thiscontrol set includes both genes that have previously beenshown to have experienced positive selection acrossanthropoids and genes which appear to be under purifyingselection (supplementary table S9a, Supplementary Mate-rial online). We tested for an association between the root-to-tip dN/dS of these loci and absolute adult and neonatalbrain mass in the same way as described above. No locuswas found to have a significant association with either phe-notype (supplementary table S9b, Supplementary Materialonline), and neither did the removal of any one species re-sult in a significant change in the regression slope and anpositive association with brain size (data not shown). Thissuggests that the significant results presented above are un-likely to be Type I errors.

Given the strong allometric relationship between brainand body mass, we also checked whether molecular evo-lution of ASPM and CDK5RAP2 could be more strongly as-sociated with adult body mass than brain mass, due, forexample, to differences in effective population size(Nikolaev et al. 2007) or general scaling effects, using theneonate data set to allow comparisons using AIC. ForASPM, there was no significant association with adult bodymass with (t135!0.368, P5 1.000, R25 0.012) or withoutinclusion of Callithrix (t12 5 0.846, P 5 0.210, R2 5 0.067).For CDK5RAP2, there is a marginally significant associationwith body mass (t135 1.823, P5 0.048, R25 0.232) but theAIC score for an association with absolute neonatal brainmass is substantially lower (body mass AIC 5 20.515; neo-natal brain mass AIC5 12.19) showing that the associationwith brain mass is much better supported and that theweak association with body mass is likely to be due to cor-related evolution between brain and body mass.

We next tested whether the adaptive evolution of thefour loci could be better explained by sexual selection inrelation to their expression in the testes. We performedphylogenetically controlled regressions between root-to-tip dN/dS and relative testes size, a commonly used corre-late of sexual selection (see Materials and Methods). Nosignificant association was observed for ASPM, CDK5RAP2,or CENPJ. A significant association was found for MCPH1(t18 5 1.959, P 5 0.033, R2 5 0.176; table 4) but whenPan (which has an unexpectedly high dN/dS) is removed,this association is no longer significant (t17 5 0.530, P 50.301, R2 5 0.016), suggesting more data will be requiredto confirm this result.

Discussion

Molecular Evolution of Microcephaly Genes andBrain EvolutionStudying the molecular basis of convergent phenotypes hasenhanced our understanding of the evolutionary geneticsof adaptation and the constraints that act on phenotypicevolution (Arendt and Reznick 2007). Here, we show thatindependent increases in brain mass across anthropoidsmay share a common genetic basis. By sampling a substan-tial number of phylogenetically diverse species, we havedemonstrated that positive selection acted on four micro-cephaly loci across the anthropoid phylogeny and was not,as previously reported, restricted to lineages leading to hu-mans. To our knowledge, our study is the first to imple-ment robust codon-based models to test for positiveselection acting on these loci across anthropoids.

Table 4. Phylogenetically Controlled Regression Analysis betweenRoot-to-Tip dN/dS and Relative Testis Mass.

Relative Testis Mass

Gene n t-Statistic P Value R2

ASPM 20 21.396 1.000 0.097CDK5RAP2 20 20.621 1.000 0.021CENPJ 20 0.824 0.210 0.036MCPH1 20 1.96 0.032 0.176


633

(A previous study reporting such a finding [Wang and Su2004] used Model 3 in PAML which is not a robust test forpositive selection [Anisimova et al. 2002].) This is a strikingresult as pervasive positive selection is considered rare.However, as brain size has increased multiple times inde-pendently and is likely to have been under strong selectionin all major groups of anthropoids (Montgomery et al.2010), such widespread positive selection on genes in-volved in the evolution of brain size should perhaps beexpected.

We explored whether this selection is relevant to grossbrain size evolution using phylogenetically controlled re-gressions which show that the average dN/dS across ASPMand CDK5RAP2 is significantly related to absolute neonatalbrain mass and that in both cases a relatively large propor-tion of the variance is explained (R2 5 0.369 and 0.255, re-spectively). Furthermore, two key predictions of a gene’sinvolvement in prenatal neurogenesis were verified forASPM and CDK5RAP2: 1) an association with absolute brainmass as this correlates closely with total neuron number(Herculano-Houzel et al. 2007) and 2) a stronger associa-tion with neonatal than adult brain mass as cortical neuro-genesis is largely restricted to prenatal development (Rakic1988, 2002; Bhardwaj et al. 2006).

These results are not explained by a general associationwith body mass nor can they be attributed to random orgenome-wide effects because no associations were foundwith two other microcephaly genes (CENPJ and MCPH1)and eight control genes. Thus, although the results areonly marginally significant, our control tests and thehighly consistent pattern observed in the significant re-sults strongly suggest that the associations found are un-likely to be Type I errors. The results using multipleregressions suggest that while positive selection on ASPMhas brought about an increase in dN/dS mainly throughan acceleration in dN relative to dS, the pattern forCDK5RAP2 may be more complex. This suggests that cau-tion should be exercised in interpreting dN/dS ratios, andwe recommend the use of supplementary analyses such asmultiple regression to disambiguate correlations involv-ing this measure. Together with the demonstration ofpositive selection in the site analyses, these results implythat adaptive evolution on ASPM and CDK5RAP2 hasbeen involved in independent changes in brain size alongmultiple lineages during primate evolution through a rolein prenatal neurogenesis.

Although we detect a general positive association forASPM and CDK5RAP2, there are notable outliers, whichmay suggest a more complex relationship between the evo-lution of these loci and brain size. For both loci, Papio hasa much lower dN/dS than would be predicted given ourresults and the size of its the brain in this species. Assumingthe association between ASPM, CDK5RAP2 and brain evo-lution have a genuine, functional basis, the Papio dis-crepancy may indicate brain expansion can occurindependently of the evolution of these loci. However,as Papio is not represented in our neonatal data set, wecannot say whether it is an outlier due to pre- or postnatal

developmental processes. The high rate of evolution ofASPM during Callitrichid evolution is also striking. A plau-sible hypothesis is that this acceleration is related tophyletic dwarfism and the reduction of brain size in thisclade (Ford 1980; Montgomery et al. 2010). The lower rateof evolution of CDK5RAP2 in the Callitrichids raises the in-teresting possibility that different selection pressures haveacted on these loci during episodes of brain size reduction.Further work is required to address these issues.

Despite demonstrating CENPJ and MCPH1 have experi-enced pervasive positive selection during anthropoid evo-lution, we found no significant relationship between eitherlocus and any measure of absolute or relative brain mass.This potentially indicates an interesting dichotomy in evo-lutionary roles among microcephaly genes and raises theissue of whether these loci are involved in the evolutionof other traits or more specific aspects of brain phenotypethat were not considered here. It is important to emphasizethat the phenotypes on which selection for CENPJ andMCPH1 is acting in primates has not been established,and this study provides no evidence that they are involvedin the evolution of gross measures of size of the whole brainor neocortex.

Integrating Comparative Genetics andNeurobiologyOur results are consistent with models of how neuronnumber might evolve. A single additional round of prolif-erative, symmetric divisions of neuroepithelial cells in theventricular zone would double the number of neurons inthe cortex (Rakic 1988, 1995; Caviness et al. 1995). Neuro-epithelial cells have apical–basal polarity, and the switchfrom proliferative, symmetric to neurogenic, asymmetricdivisions is controlled by the orientation of the spindle poleduring mitotic division (Chenn and McConnell 1995; Gotzand Huttner 2005). An alternative, but not mutually exclu-sive, model places greater emphasis on prolonged interme-diate progenitor cell division in the subventricular zone,which may also occur by changes in spindle pole orienta-tion (Kriegstein et al. 2006). In addition, as brain size ex-pands, neural progenitors become increasingly elongated(Smart et al. 2002; Fish et al. 2008) and selection may beacting on cytoskeletal genes in response to the need tomaintain the precision of spindle orientation during mi-totic division of these highly elongated cells in largerbrained species (Zhang 2003; Kouprina et al. 2004; Fishet al. 2008). In this way, selection on these loci may bein response to the evolution of larger brains rather thancausing the change in brain size. Although it is also possibleto envisage scenarios where the change in spindle orienta-tion itself leads to the production of additional neurons,whether the role of these loci in brain size evolution is caus-ative or responsive is yet to be determined. Both scenariosare consistent with the results presented here.

Notably, there is strong agreement between the resultsof studies investigating the expression of these loci (Bondet al. 2002, 2005; Jackson et al. 2002; Kouprina et al. 2005),their function (Bond and Woods 2006; Cox et al. 2006;


634

Fish et al. 2006; Brunk et al. 2007; Wood et al. 2007; Buch-man et al. 2010), models of brain evolution (Rakic 1988;Kriegstein et al. 2006), and their molecular evolutionwhich together implicate ASPM and CDK5RAP2 as havingsignificant roles in the evolution of neuron number andbrain size.

Implications for Evolutionary Genetics ofAdaptationThe inferred role of ASPM and CDK5RAP2 in the evolutionof primate brain mass have implications for our under-standing of the evolutionary genetics of adaptation. First,they provide evidence that a complex, polygenic quantita-tive phenotype evolved by convergence at the molecularlevel (Cresko et al. 2004; Mundy 2005; Arendt and Reznick2007). Second, our results go against a commonly assertedhypothesis that evolution of form occurs primarily throughchanges in cis-regulatory sequences (King and Wilson 1975;Carroll 2005, 2008; Wray 2007).

Changes in cis-regulatory sequences are proposed to bemore important for phenotypic evolution as their modularnature limits pleiotropic effects (Carroll 2005; but seeHoekstra and Coyne 2007; Lynch and Wagner 2008; Sternand Orgogozo 2008). In fact the evolution of brain mass hasbeen singled out as an example where regulatory evolutionis likely to be the predominant evolutionary mechanism:‘‘the evolution of complex traits such as brain size . . .musthave a highly polymorphic and largely regulatory basis’’(Carroll 2005; emphasis ours). It is therefore interesting thatthe microcephaly genes are expressed throughout the body(Bond et al. 2005; Kouprina et al. 2005) but are subject topositive selection. If ASPM and CDK5RAP2 are in fact in-volved in brain evolution, how are pleiotropic effectsavoided?

The first point to note is that the pathology of primarymicrocephaly itself shows that pleiotropic effects of micro-cephaly gene disruption can be limited to the brain (Bondand Woods 2006). Both ASPM and CDK5RAP2 are alterna-tively spliced (Kouprina et al. 2005; Buchman et al. 2010),which may reduce pleiotropic effects (Hughes 2006;Hoekstra and Coyne 2007; Lynch and Wagner 2008). Alter-natively, the evolution of ASPM may not affect nonneuralcells either due to the elongated cell morphology of neuro-epithelial cells (Fish et al. 2006) or cell-dependent recruit-ment factors (Van der Voet et al. 2009). These explanationsare not mutually exclusive and provide plausible mecha-nisms to reduce the pleiotropic effects of protein evolution(Hoekstra and Coyne 2007; Lynch and Wagner 2008). Morebroadly, research is showing a diversity of mechanisms forthe possible genetic basis of aspects of primate brain evo-lution, including coding sequence evolution (e.g., thisstudy; Wang et al. 2005; Vallender and Lahn 2006; Uddinet al. 2008), gene duplication (Burki and Kaessmann 2004;Marques-Bonet et al. 2009), noncoding RNA evolution(Zhang et al. 2008), and changes in regulatory sequencesand gene expression (Khaitovich et al. 2005, 2006; Rockmanet al. 2005; Gilad et al. 2006; Prabhakar et al. 2006; Haygoodet al. 2007; Somel et al. 2009). Indeed, the mosaic nature of

brain structure evolution in mammals indicates a complexgenetic basis to changes in brain size (Barton and Harvey2000).

ConclusionsWe have presented evidence implicating ASPM andCDK5RAP2 in the evolution of brain size across anthropoidprimates, a result which is consistent with an effect of thetwo loci on neurogenesis via mitotic spindle orientation.Despite showing that CENPJ andMCPH1 have been subjectto positive selection, we find no evidence to link these locito the evolution of gross brain size, and the mechanism ofselection acting on these loci is therefore unresolved. Thisstudy demonstrates the importance of including pheno-typic data and a phylogenetically broad range of specieswhen attempting to associate the evolution of genes withbrain size evolution (Carroll 2003; Goodman et al. 2005;Barton 2006a; Pollen and Hoffmann 2008; Vallender2008). This point is especially pertinent to the literatureon human genetic evolution where claims are often basedon differences between humans and chimpanzees or a smallnumber of nonhuman primates. The results also clearlyhighlight the importance of including measures of neonatalbrain size in studies of primate brain evolution. Finally, ourresults suggest a conserved genetic basis for brain evolutionin primates, providing an important example where geneticbasis of a complex developmental phenotype has involvedcoding sequence evolution.

Supplementary MaterialSupplementary information, figure S1, and tables S1–S9 areavailable at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).

AcknowledgmentsWe are grateful to Andrew Kitchener and Drew Bain(National Museums Scotland), Mike Bruford (ZoologicalSociety of London), and Leona Chemnick (Center for Re-production of Endangered Species, San Diego Zoo) for pro-viding tissue samples and Barry Keverne, Chris Jiggins, theMundy and Jiggins Labs (University of Cambridge), andAdrian Friday for helpful discussions and Jennifer Fishfor correspondence on the function of ASPM and Mark Pa-gel for helpful comments on the phylogenetic analyses.S.H.M. and N.I.M. thank Biotechnology and Biological Sci-ences Research Council, the Leverhulme Trust, and MurrayEdwards College for financial support. I.C. and R.A.B. thankBiotechnology and Biological Sciences Research Council/Natural Environment Research Council for financial sup-port (grant number BB/E014593/1). C.V. thanks the Lever-hulme Trust for financial support (grant number: ECF/2009/0029).

ReferencesAli F, Meier R. 2008. Positive selection in ASPM is correlated with

cerebral cortex evolution across primates but not with whole-brain size. Mol Biol Evol. 25(11):2247–2250.


635

S9

http://www.mbe.oxfordjournals.org/

http://www.mbe.oxfordjournals.org/

Anisimova M, Bielawski JP, Yang Z. 2002. Accuracy and power ofBayes prediction of amino acid sites under positive selection.Mol Biol Evol. 19(6):950–958.

Arendt J, Reznick D. 2007. Covergence and parallelism reconsidered:what have we learned about the genetics of adaptation? TrendsEcol Evol. 23(1):26–32.

Barton RA. 1998. Visual specialization and brain evolution inprimates. Proc R Soc B Biol Sci. 265:1933–1937.

Barton RA. 2006a. Chimp comparisons won’t explain humanevolution. Nature 443:632.

Barton RA. 2006b. Primate brain evolution: integrating comparative,neurophysiological and ethological data. Evol Anthropol.15(5):224–236.

Barton RA, Harvey PH. 2000. Mosaic evolution of brain structure inmammals. Nature 405:1055–1058.

Bauchot R, Stephan H. 1969. Encephalisation et niveau evolutif chezles Simiens. Mammalia 33:225–275.

Bhardwaj RD, Curtis MA, Spalding KL, et al. (11 co-authors). 2006.Neocortical neurogenesis in humans is restricted to develop-ment. Proc Natl Acad Sci U S A. 103(33):12564–12568.

Bond J, Roberts E, Mochida GH, et al. (12 co-authors). 2002. ASPM isa major determinant of cerebral cortical size. Nat Genet.32:316–320.

Bond J, Roberts E, Springell K, et al. (20 co-authors). 2005. Acentrosomal mechanism involving CDK5RAP2 and CENPJcontrols brain size. Nat Genet. 37(4):353–355.

Bond J, Woods CG. 2006. Cytoskeletal genes regulating brain size.Curr Opin Cell Biol. 18:95–101.

Brunk K, Vernay B, Griffith E, Reynolds NL, Strutt D, Ingham PW,Jackson AP. 2007. Microcephalin coordinates mitosis in thesyncytial Drosophila embryo. J Cell Sci. 120:3578–3588.

Buchman JJ, Tseng HC, Zhou Y, Frank CL, Xie Z, Tsai LH. 2010.CDK5RAP2 interacts with pericentrin to maintain the neuralprogenitor pool in the developing neocortex. Neuron66:386–402.

Burki F, Kaessmann H. 2004. Birth and adaptive evolution ofa hominoid gene that supports high neurotransmitter flux. NatGenet. 35:1061–1063.

Burnham KP, Anderson DR. 2002. Model selection and multimodelinference: a practical information-theoretic approach, 2nd ed.New York: Springer.

Buss RR, Sun W, Oppenheim RW. 2006. Adaptive roles ofprogrammed cell death during nervous system development.Annu Rev Neurosci. 29:1–35.

Calegari F, Huttner WB. 2003. An inhibition of cyclin-dependentkinases that lengthens, but does not arrest, neuroepithelialcell cycle induces premature neurogenesis. J Cell Sci.116:4947–4955.

Capellini I, Venditti C, Barton RA. 2010. Phylogeny and metabolicscaling in mammals. Ecology 91:2783–2793.

Carroll SB. 2003. Genetics and the making of Homo sapiens. Nature422:849–857.

Carroll SB. 2005. Evolution at two levels: on genes and form. PLoSBiol. 3:1160–1166.

Carroll SB. 2008. Evo-Devo and an expanding evolutionary synthesis:a genetic theory of morphological evolution. Cell 134:25–36.

Caviness VS, Takahashi T, Nowakowski RS. 1995. Numbers, time andneocortical neurogenesis: a general developmental and evolu-tionary model. Trends Neurosci. 18(9):279–282.

Chenn A, McConnell SK. 1995. Cleavage orientation and theasymmetric inheritance of Notch 1 immunoreactivity inmammalian neurogenesis. Cell 82:631–641.

Clark NL, Swanson WJ. 2005. Pervasive adaptive evolution inprimate seminal proteins. PLoS Genet. 1(2):e35.

Cormier A, Clemont MJ, Knossow M, Lachkar S, Savarin P, Toma F,Sobel A, Gigant B, Curmi PA. 2009. The PN2-3 domain of CPAP

implements a novel mechanism for tubulin sequestration. J BiolChem. 284:6909–6917.

Cox J, Jackson AP, Bond J, Woods CG. 2006. What primarymicrocephaly can tell us about brain growth. Trends Mol Med.12(8):358–366.

Cresko WA, Amores A, Wilson C, Murphy J, Currey M, Phillips P,Bell MA, Kimmel CB, Postlethwait JH. 2004. Parallel genetic basisfor repeated evolution of armour loss in Alaskan threespinestickleback populations. Proc Natl Acad Sci U S A.101(16):6050–6055.

Desir J, Cassart M, David P, Van Bogaert P, Abramowicz M. 2008.Primary microcephaly with ASPM mutation shows simplifiedcortical gyration with antero-posterior gradient pre- and post-natally. Am J Med Genet A. 146A:1439–1443.

Dorus S, Evans PD, Wyckoff GJ, Choi SS, Lahn BT. 2004. Rate ofmolecular evolution of the seminal protein gene SEMG2correlates with levels of female promiscuity. Nat Genet.36:1326–1329.

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 ofHomo sapiens. Cell 119:1027–1040.

Enard W, Przeworski M, Fisher SE, Lai CSL, Wiebe V, Kitano T,Monaco AP, Paabo S. 2002. Molecular evolution of FOXP2,a gene involved in speech and language. Nature 418:869–871.

Evans PD, Anderson JR, Vallender EJ, Choi SS, Lahn BT. 2004.Reconstructing the evolutionary history of Microcephalin, a genecontrolling human brain size. Hum Mol Genet. 13(5):1139–1145.

Evans PD, Anderson JR, Vallender EJ, Gilbert SL, Malcom CM,Dorus S, Lahn BT. 2004. Adaptive evolution of ASPM, a majordeterminant of cerebral cortical size in humans. Hum Mol Genet.13(5):489–494.

Evans PD, Vallender EJ, Lahn BT. 2006. Molecular evolution of thebrain size regulator genes CDK5RAP2 and CENPJ. Gene375:75–79.

Falk D, Gibson KR, editors 2001. Evolutionary anatomy of theprimate cerebral cortex, 1st ed. Cambridge: CambridgeUniversity Press.

Finlay BL, Darlington RB, Nicastro N. 2001. Developmental structurein brain evolution. Behav Brain Sci. 24:298–304.

Fish JL, Dehay C, Kennedy H, Huttner WB. 2008. Making briggerbrains—the evolution of neural-progenitor-cell division. J CellSci. 121:2783–2793.

Fish JL, Kosodo Y, Enard W, Paabo S, Huttner WB. 2006. ASPMspecifically maintains symmetric proliferative divisions of neuro-epithelial cells. Proc Natl Acad Sci U S A. 103(27):10438–10443.

Fong KW, Choi YK, Rattner JB, Qi RZ. 2008. CDK5RAP2 isa pericentriolar protein that functions in centrosomal attach-ment of the c-tubulin ring complex. Mol Biol Cell.19(1):115–125.

Ford S. 1980. Callitrichids as phyletic dwarfs, and the place of theCallitrichidae in Platyrrhini. Primates 21(1):21–43.

Freckleton RP, Harvey PH, Pagel M. 2002. Phylogenetic analysis andcomparative data: a test and review of evidence. Am Nat.160(6):712–726.

Gilad Y, Oshlack A, Smyth GK, Speed TP, White KP. 2006. Expressionprofiling in primates reveals a rapid evolution of humantranscription factors. Nature 440:242–245.

Goodman M, Grossman LI, Wildman DE. 2005. Moving primategenomics beyond the chimpanzee genome. Trends Genet.21(9):511–517.

Gotz M, Huttner WB. 2005. The cell biology of neurogenesis. NatRev Mol Cell Biol. 6:777–778.

Harcourt AH, Purvis A, Liles L. 1995. Sperm competition: matingsystem, not breeding season, affects testes size of primates. FunctEcol. 9(3):468–476.


636

Haygood R, Fedrigo O, Hanson B, Yokoyama KD, Wray GA. 2007.Promoter regions of many neural- and nutrition-related geneshave experienced positive selection during human evolution.Nat Genet. 39:1140–1144.

Herculano-Houzel S, Collins CE, Wong P, Kaas JH. 2007. Cellularscaling rules for primate brains. Proc Natl Acad Sci U S A.204(9):3562–3567.

Hirano T. 2006. At the heart of the chromosome: SMC proteins inaction. Nat Rev Mol Cell Biol. 7:311–322.

Hoekstra HE, Coyne JA. 2007. The locus of evolution: evo devo andthe genetics of adaptation. Evolution 61(5):995–1016.

Horvath JE, Weisrock DW, Embry SL, Fiorentino I, Balhoff JP,Kappeler P, Wray GA, Willard HF, Yoder AD. 2008. Developmentand application of a phylogenomic toolkit: resolving theevolutionary history of Madagascar’s lemurs. Genome Res.18:489–499.

Hughes TA. 2006. Regulation of gene expression by alternativeuntranslated regions. Trends Genet. 22(3):119–122.

Ives AR, Midford PE, Garland T Jr. 2007. Within-species variation andmeasurement error in phylogenetic comparative methods. SystBiol. 56(2):252–270.

Jabes A, Lavenex PB, Amaral DG, et al. (4 co-authors). 2010.Quantitative analysis of postnatal neurogenesis and neuronnumber in the macaque monkey dentate gyrus. Eur J Neurosci.31:273–285.

Jackson AP, Eastwood H, Bell SM, et al. (16 co-authors). 2002.Identification of microcephalin, a protein implicated in de-termining the size of the human brain. Am J Hum Genet.71(1):136–142.

Jackson AP, McHale DP, Campbell DA, et al. (13 co-authors). 1998.Primary autosomal recessive microcephaly (MCPH1) maps tochromosome 8p22-pter. Am J Hum Genet. 63(2):541–546.

Jerison HJ. 1973. Evolution of the brain and intelligence, 1st ed.. NewYork: Academic Press.

Khaitovich P, Enard W, Lachmann M, Paabo S. 2006. Evolution ofprimate gene expression. Nat Rev Genet. 7:693–702.

Khaitovich P, Hellmann I, Enard W, Nowick K, Leinweber M,Franz H, Weiss G, Lachmann M, Paabo S. 2005. Parallel patternsof evolution in the genomes and transcriptomes of humans andchimpanzees. Science. 309(5742):1850–1854.

King MC, Wilson AC. 1975. Evolution at two levels in humans andchimpanzees. Science 188(4184):107–116.

Kouprina N, Pavlicek A, Collins NK, et al. (18 co-authors). 2005. Themicrocephaly ASPM gene is expressed in proliferating tissuesand encodes for a mitotic spindle protein. Hum Mol Genet.14(15):2155–2165.

Kouprina N, Pavlicek P, Mochida GH, et al. (13 co-authors). 2004.Accelerated evolution of ASPM gene controlling brain sizebegins prior to human brain expansion. PLoS Biol. 2(5):653–662.

Kriegstein A, Noctor S, Martınez-Cerdeno V. 2006. Patterns of neuralstem and progenitor cell division may underlie evolutionarycortical expansion. Nat Rev Neurosci. 7:883–890.

Kumar A, Girimaji SC, Duvvari MR, Blanton SH. 2009. Mutations inSTIL, encoding a pericentriolar and centrosomal protein causeprimary microcephaly. Am J Hum Genet. 84(2):286–290.

Lande R. 1979. Quantitative genetic analysis of multivariateevolution, applied to brain: body size allometry. Evolution33(1):402–416.

Leigh SR. 2006. Brain growth, life history, and cognition in primateand human evolution. Am J Primatol. 62:139–164.

Low LK, Cheng HJ. 2006. Axon pruning: an essential step underlyingthe developmental plasticity of neuronal connections. PhilosTrans R Soc Lond B Biol Sci. 361:1531–1544.

Lynch VJ, Wagner GP. 2008. Resurrecting the role of transcriptionfactor change in developmental evolution. Evolution62(9):2131–2154.

Marques-Bonet T, Kidd JM, Ventura M, et al. (23 co-authors). 2009.A burst of segmental duplications in the genome of the Africangreat ape ancestor. Nature 457:877–881.

Martin RD. 1990. Primate origins and evolution, 1st ed. Princeton(NJ): Princeton University Press.

Montgomery SH, Capellini I, Barton RA, Mundy NI. 2010.Reconstructing the ups and downs of primate brain evolution:implications for adaptive hypotheses and Homo floresiensis. BMCBiol. 8(9).

Montgomery SH, Mundy NI. 2010. Brain evolution: microcephalygenes weigh in. Curr Biol. 20(5):R244–R246.

Mundy NI. 2005. A window on the genetics of evolution: MC1R andplumage colouration in birds. Proc R Soc B Biol Sci.272:1633–1640.

Nadeau NJ, Burke T, Mundy NI. 2007. Evolution of an avianpigmentation gene correlates with a measure of sexual selection.Proc R Soc B Biol Sci. 274:1807–1813.

Nielsen R, Yang Z. 1998. Likelihood models for detecting positivelyselected amino acid sites and applications to the HIV-1 envelopegene. Genetics 148:929–936.

Nikolaev SI, Montoya-Burgos JI, Popadin K, Parand L, Margulies EH.National Institutes of Health Intramural Sequencing CenterComparative Sequencing Program, Antonarakis SE. 2007. Life-history traits drive the evolutionary rates of mammalian codingand noncoding genomic elements. Proc Natl Acad Sci U S A.104(51):20443–20448.

Nunn CL, Barton RA. 2001. Comparative methods for studyingprimate adaptation and allometry. Evol Anthropol. 10:81–98.

Opazo JC, Wildman DE, Prychitko T, Johnson RM, Goodman M.2006. Phylogenetic relationships and divergence times amongNew World monkeys (Platyrrhini, Primates). Mol PhylogenetEvol. 40:274–280.

Organ CL, Shedlock AM, Meade A, Pagel M, Edwards SV. 2007.Origin of avian genome size and structure in nonaviandinosaurs. Nature 446:180–184.

Page SL, Goodman M. 2001. Catarrhine phylogeny: noncoding DNAevidence for a diphyletic origin of the mangabeys and fora human-chimpanzee clade. Mol Phylogenet Evol. 18(1):14–25.

Pagel M. 1997. Inferring evolutionary processes from phylogenies.Zool Scr. 26(4):331–348.

Pagel M. 1999. Inferring the historical patterns of biologicalevolution. Nature 401:877–884.

Pagel M, Meade A, Barker D. 2004. Bayesian estimation of ancestralcharacter states on phylogenies. Syst Biol. 53:673–684.

Pollen AA, Hoffmann HA. 2008. Beyond neuroanatomy: novelapproaches to studyingbrain evolution.BrainBehav Evol. 72:145–158.

Ponting C. 2006. A novel domain suggests a ciliary function for ASPM,a brain size determining gene. Bioinformatics 22(9):1031–1035.

Poux C, Douzery EJP. 2004. Primate phylogeny, evolutionary ratevariations, and divergence times: a contribution from thenuclear gene IRBP. Am J Phys Anthropol. 124(1):1–16.

Prabhakar S, Noonan JP, Paabo S, Rubin EM. 2006. Acceleratedevolution of conserved noncoding sequences in humans. Science314:786.

Purvis A. 1995. A composite estimate of primate phylogeny. PhilosTrans R Soc Lond B Biol Sci. 348(1326):405–421.

Rakic P. 1988. Specification of cerebral cortical areas. Science241(4862):170–176.

Rakic P. 1995. A small step for the cell, a giant leap for mankind:a hypothesis of neocortical expansion during evolution. TrendsNeurosci. 18:383–388.

Rakic P. 2002. Neurogenesis in adult primate neocortex: anevaluation of the evidence. Nat Rev Neurosci. 3:65–71.

Ramm SA, Oliver PL, Ponting CP, Stockley P, Emes RD. 2008. Sexualselection and the adaptive evolution of mammalian ejaculateproteins. Mol Biol Evol. 25(1):207–219.


637

Ramm SA, Stockley P. 2010. Sperm competition and sperm lengthinfluence the rate of mammalian spermatogenesis. Biol Lett.6:219–221.

Rickmyre JL, DasGupta S, Ooi DL, Keel J, Lee E, Kirschner MW,Wassell S, Lee LA. 2007. The Drosophila homolog of MCPH1,a human microcephaly gene, is required for genomic stability inthe early embryo. J Cell Sci. 120:3565–3577.

Rimol LM, Agartz I, Djurovic S, et al. (16 co-authors). 2010. Sex-dependent association of common variants of microcephalygenes with brain structure. Proc Natl Acad Sci U S A.107:384–388.

Rockman MV, Hahn MW, Soranzo N, Zimprich F, Goldstein DB,Wray GA. 2005. Ancient and recent positive selection trans-formed opioid cis-regulation in humans. PLoS Biol. 3(12):e387.

Roth KA, D’Sa C. 2001. Apoptosis and brain development. MentRetard Dev Disabil Res Rev. 7(4):261–266.

Ruxton GD, Neuhauser M. 2010. When should we use one-tailedhypothesis testing? Meth Ecol Evol. 1:114–117.

Sauvageot CM, Stiles CD. 2002. Molecular mechanisms controllingcortical gliogenesis. Curr Opin Neurobiol. 12:244–249.

Shi P, Bakewell MA, Zhang J. 2006. Did brain-specific genes evolvefaster in humans than in chimpanzees? Trends Genet.22(11):608–613.

Smart IHM, Dehay C, Giroud P, Berland M, Kennedy H. 2002.Unique morphological features of the proliferative zones ofpostmitotic compartments of the neural epithelium giving riseto striate and extrastriate cortex in the monkey. Neuron12:37–53.

Somel M, Franz H, Yan Z, et al. (15 co-authors). 2009.Transcriptional neoteny in the human brain. Proc Natl AcadSci U S A. 106(14):5743–5748.

Sowell ER, Delis D, Stiles J, Jernigan TL. 2001. Improved memoryfunctioning and frontal lobe maturation between childhood andadolescence: a structural MRI study. J Int Neuropsychol Soc.7:312–322.

Steiper ME, Young NM. 2006. Primate molecular divergence dates.Mol Phylogenet Evol. 41(2):384–394.

Stephan H, Frahm H, Baron R. 1981. New and revised data onvolume of brain structures in insectivores and primates. FoliaPrimatol. 35:1–29.

Stern DL, Orgogozo V. 2008. The loci of evolution: how predictableis genetic evolution? Evolution 62(9):2155–2177.

Tamura K, Dudley J, Nei M, Kumar S. 2007. MEGA4: molecularevolutionary genetics analysis (MEGA) software version 4.0. MolBiol Evol. 24(8):1596–1599.

Thornton GK, Woods CG. 2009. Primary microcephaly: do all roadslead to Rome? Trends Genet. 25(11):501–510.

Uddin M, Opazo JC, Wildman DE, Sherwood CC, Hof PR,Goodman M, Grossman LI. 2008. Molecular evolution of thecytochrome c oxidase subunit 5A gene in primates. BMC Evol Biol.8:8.

Untergasser A, Nijveen H, Rao X, Bisseling T, Geurts R,Leunissen JAM. 2007. Primer3Plus, an enhanced web interfaceto Primer3. Nucleic Acids Res. 35:W71–W74.

Vallender EJ. 2008. Exploring the origins of the human brain throughmolecular evolution. Brain Behav Evol. 72:168–177.

Vallender EJ, Lahn BT. 2006. A primate-specific acceleration in theevolution of the caspase-dependent apoptosis pathway. HumMol Genet. 15(20):3034–3040.

van der Voet M, Berends CWH, Perreault A, Nguyen-Ngoc T,Gonczy P, Vidal M, Boxem M, van den Heuvel S. 2009. NuMA-related LIN-5, ASPM-1, calmodulin and dynein promote meioticspindle rotation independently of cortical LIN-5/GPR/G. Nat CellBiol. 11:269–277.

Wang HY, Chien HC, Osada N, Hashimoto K, Sugano S, Gojobori T,Chou CK, Tsai SF, Wu CI, Shen CK. 2007. Rate of evolution inbrain expressed genes in humans and other primates. PLoS Biol.5(2):e13.

Wang JK, Li Y, Su B. 2008. A common SNP of MCPH1 is associatedwith cranial volume variation in Chinese population. Hum MolGenet. 17(9):1329–1335.

Wang Y, Su B. 2004. Molecular evolution of Microcephalin, a genedetermining human brain size. Hum Mol Genet.13(11):1131–1137.

Wang YQ, Qian YP, Yang S, et al. (13 co-authors). 2005. Acceleratedevolution of the pituitary adenylate cyclase-activating poly-peptide precursor gene during human origin. Genetics170:801–806.

Wolf JBW, Kunstner A, Nam K, Jakobsson M, Ellegren H. 2009. Non-linear dynamics of non-synonymous (dN) and synonymous (dS)substitution rates affects inference of selection. Genome BiolEvol. 1(1):308–319.

Wood JL, Liang Y, Chen J. 2008. Microcephalin/MCPH1 associateswith the condensin II complex to function in homologousrecombination repair. J Biol Chem. 283(43):29586–29592.

Wood JL, Singh N, Mer G, Chen J. 2007. MCPH1 functions in anH2AX-dependent but MDC1-independent pathway in responseto DNA damage. J Biol Chem. 282(48):35416–35423.

Wray GA. 2007. The evolutionary significance of cis-regulatorymutations. Nat Rev Genet. 8:206–216.

Yang Z. 2007. PAML 4: phylogenetic analysis by maximumlikelihood. Mol Biol Evol. 24:1586–1591.

Yang Z, Nielsen R, Goldman N, Pedersen AMK. 2000. Codon-substitution models for heterogeneous selection pressure atamino acid sites. Genetics 155:431–449.

Yang Z, Wong WSW, Nielsen R. 2005. Bayes empirical Bayesinference of amino acid sites under positive selection. Mol BiolEvol. 22:1107–1118.

Yu XJ, Zheng HK, Wang J, Wang W, Su B. 2006. Detecting lineage-specific adaptive evolution of brain-expressed genes in humanusing rhesus macaque as outgroup. Genomics 88(6):745–751.

Zhang J. 2003. Evolution of the human ASPM gene, a majordeterminant of brain size. Genetics 165:2063–2070.

Zhang R, Wang YQ, Su B. 2008. Molecular evolution of a primate-specific microRNA family. Mol Biol Evol. 25(7):1493–1502.

Zilles K, Rehkemper G. 1988. The brain with special reference to thetelencephalon. In: Schwartz JH, editor. Orang-utan biology.Oxford: Oxford University Press. p. 157–176.


638

Documents

Montgomery Et Al11MBE Microcephaly Genes and Brains