Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Developmental Biology

http://d0012-16

n CorrE-m

albertso

PleasDev.

journal homepage: www.elsevier.com/locate/developmentalbiology

Cichlid fishes as a model to understand normal and clinical craniofacialvariation

Kara E. Powder n, R. Craig Albertson n

Department of Biology, University of Massachusetts Amherst, 221 Morrill Science Center South, 611 North Pleasant Street, Amherst, MA 01003, USA

a r t i c l e i n f o

Article history:Received 1 September 2015Received in revised form14 December 2015Accepted 21 December 2015

Keywords:CraniofacialCichlidsComplex traitMorphological variationEvolutionary mutant

x.doi.org/10.1016/j.ydbio.2015.12.01806/& 2015 Published by Elsevier Inc.

esponding authors.ail addresses: kepowder@bio.umass.edu (K.E.n@bio.umass.edu (R.C. Albertson).

e cite this article as: Powder, K.E., AlbBiol. (2015), http://dx.doi.org/10.101

a b s t r a c t

We have made great strides towards understanding the etiology of craniofacial disorders, especially for’simple’Mendelian traits. However, the facial skeleton is a complex trait, and the full spectrum of genetic,developmental, and environmental factors that contribute to its final geometry remain unresolved.Forward genetic screens are constrained with respect to complex traits due to the types of genes andalleles commonly identified, developmental pleiotropy, and limited information about the impact ofenvironmental interactions. Here, we discuss how studies in an evolutionary model – African cichlidfishes – can complement traditional approaches to understand the genetic and developmental origins ofcomplex shape. Cichlids exhibit an unparalleled range of natural craniofacial morphologies that modelnormal human variation, and in certain instances mimic human facial dysmorphologies. Moreover, theevolutionary history and genomic architecture of cichlids make them an ideal system to identify thegenetic basis of these phenotypes via quantitative trait loci (QTL) mapping and population genomics.Given the molecular conservation of developmental genes and pathways, insights from cichlids areapplicable to human facial variation and disease. We review recent work in this system, which hasidentified lbh as a novel regulator of neural crest cell migration, determined the Wnt and Hedgehogpathways mediate species-specific bone morphologies, and examined how plastic responses to dietmodulate adult facial shapes. These studies have not only revealed new roles for existing pathways incraniofacial development, but have identified new genes and mechanisms involved in shaping the cra-niofacial skeleton. In all, we suggest that combining work in traditional laboratory and evolutionarymodels offers significant potential to provide a more complete and comprehensive picture of the myriadfactors that are involved in the development of complex traits.

& 2015 Published by Elsevier Inc.

1. Craniofacial development and disease: progress andchallenges

Over 70% of all birth defects are associated with craniofacialmalformations (Hall, 2009). Although significant advances havebeen made in understanding the etiology of many Mendeliancraniofacial disorders (e.g., Treacher Collins Syndrome (Kadakiaet al., 2014), Apert Syndrome (Wilkie et al., 1995), and CrouzonSyndrome (Reardon et al., 1994)), much less is known about thefactors that contribute to complex diseases (e.g. non-syndromiccleft lip and/or palate) and normal facial variation, which resultfrom the cumulative effects of many alleles and their interactionswith the environment (Fish et al., 2014; Glazier et al., 2002;Hallgrimsson et al., 2009; Hallgrimsson et al., 2014; Hirschhornand Daly, 2005; Hochheiser et al., 2011). Even for monogenetic

Powder),

ertson, R.C., Cichlid fishes as6/j.ydbio.2015.12.018i

diseases in which the causative genes are known we still lack acomprehensive picture of the genetic, cellular, and environmentalinteractions that contribute to the severity of the disease (e.g., theimpact of genetic background (Dixon and Dixon, 2004)). The Na-tional Institute of Dental and Craniofacial Research (NIDCR) in-itiative FaceBase, a consortium aimed at advancing research infacial development and disease, states that “much research isneeded to achieve a molecular and cellular understanding of themechanisms by which genes and gene products interact to gen-erate complex phenotypes.” Thus, a significant challenge in cra-niofacial biology remains to uncover the genomic and develop-mental mechanisms that underlie the production of subtle andcomplex patterns of variation in craniofacial shape.

2. The limitations of induced mutations

Much of our knowledge of craniofacial development has beenacquired from induced mutations and loss-of-function

a model to understand normal and clinical craniofacial variation.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎2

experiments in a small handful of model organisms, particularlyMus musculus (mouse), Gallus gallus (chicken), Xenopus laevis(African clawed frog), and Danio rerio (zebrafish). These studieshave provided critical insights into the basis of early develop-mental patterning events and the etiology of craniofacial diseases,particularly “simple” Mendelian disorders. However, traditionalforward genetics screens and the use of knockout mutants are notwithout limitations. We propose that embracing the natural var-iation present in “evolutionary mutant” models (Albertson et al.,2009), including cichlid fishes, can overcome some of these lim-itations and complement our current knowledge of the geneticand developmental basis of normal and extreme, clinical cranio-facial variation. The crux of the argument is that all forward ge-netic analyses are, by definition, based on phenotype, and naturalsystems provide a vast reservoir of phenotypic variation. An un-derstanding of this natural variation will supplement findingsabout phenotypes that have been artificially generated throughmutagenesis. The limitations of traditional approaches and thecomplementary nature of evolutionary models are discussed ingreater detail below.

2.1. Screens commonly miss non-essential genes and subtlemutations

Classic forward genetic approaches screen for abnormal phe-notypes in populations of mutagenized animals. These screenstypically focus on early and often lethal phenotypes, and are wellsuited to identify genes that are essential for a particularfunction. They have a bias towards certain types of genes, such asthose that regulate a critical developmental switch or act as in-tegrators of multiple pathways. Many genes are routinely missedby such an approach, including those that have redundant func-tion, act as a genetic modifier, or contribute to variation in aprocess but are not essential for trait development. While spe-cialized (e.g., “enhancer”) screens (Kile and Hilton, 2005; Pattonand Zon, 2001) can uncover some of these factors, these are laborintensive and require base knowledge or assumptions about geneaction. In all, it is becoming increasingly appreciated (e.g., (Bolker,2012)) that new approaches and new models are needed toidentify the genes that may have been missed by classic forwardgenetic screens.

In addition to disproportionally targeting genes with early es-sential functions, screens also unequally identify certain types ofalleles and mutations. Upwards of 75% of mutant alleles identifiedin standard laboratory screens are in coding regions and result insevere disruption or complete loss of gene function (Nguyen andXu, 2008). This is perhaps unsurprising, as severe genetic lesionsare likely to have more prominent phenotypic defects that areeasier to identify by researchers, and speaks to the power andefficiency of forward genetic screens. However, complex traitssuch as normal facial variation and clinical conditions like non-syndromic cleft lip and/or palate predominantly do not result fromnull alleles with large effects, but rather from a compilation ofsmall effects from many genes that act at multiple developmentalstages or confer susceptibility to environmental effects (Fish et al.,2014; Glazier et al., 2002; Hallgrimsson et al., 2009, 2014;Hirschhorn and Daly, 2005; Hochheiser et al., 2011). In accordancewith this, the overwhelming majority (480%) of variants im-plicated in complex traits and diseases by genome wide associa-tion studies (GWAS) are predicted to be in non-coding regions(Hindorff et al., 2009; Khandelwal et al., 2013; Manolio, 2009).Further, genes with essential roles in early development are likelyto face strong stabilizing selection, and thus may not contribute tonormal patterns of variation. For instance, Shh is critical for properfacial morphogenesis (Belloni et al., 1996; Chiang et al., 1996; Huand Helms, 1999; Jeong et al., 2004; Wada et al., 2005; Young et al.,

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

2010). While subtle modulation of this pathway can produce adose-dependent phenotypic continuum during embryonic devel-opment, the effect is non-linear such that the range of “normal”variation is quite narrow and even slight changes in Shh levels canresult in major malformations such as clefting in adults (Younget al., 2010). This experiment may explain why this gene has yet tobe identified in studies of normal facial variation (Boehringer et al.,2011; Liu et al., 2012; Paternoster et al., 2012; Sherwood et al.,2011). As pointed out by Hallgrimsson and colleagues (Hall-grimsson et al., 2014), it is quite possible (and even likely) thatgenes not identified by embryonic mutagenesis screens are con-tributing to the majority of normal facial variation in humans.

2.2. Developmental pleiotropy limits the assessment of later devel-opmental stages

Mutagenesis screens are also constrained by developmentalpleiotropy, wherein a single gene acts in multiple tissues and/or atmultiple developmental stages. Defects at the earliest stage inwhich the gene has an essential function (e.g. neural crest cellmigration) often preclude examination of phenotypes at later de-velopmental stages (e.g. chondrogenesis and facial growth). Thisproblem can be circumvented using conditional knockouts ortargeted activation of alleles (Campbell et al., 2012; Chan et al.,2007; Patton and Zon, 2001), but this requires prior knowledgeabout the gene of interest. Pleiotropy is a critical concernfor facial development, as defects in neural crest cell specificationand migration have a higher likelihood of being lethal(Hall, 2009), and thus may be less applicable to normal humanvariation than alterations in later developmental stages such asosteogenesis. In order to understand the etiology of complex traitsthat manifest later in development, there is a need to supplementtraditional approaches with those that specifically address laterdevelopmental stages, including adults (Albertson and Yelick,2004).

2.3. Gene x environment (GxE) interactions account for variation inboth Mendelian diseases and complex traits

Genetic screens and genome-wide association studies (GWAS)have identified hundreds of loci that are associated withcomplex human traits and diseases. However, for most complextraits, these cumulatively have only been able to account for asmall percentage of the heritability of phenotypes. Sources of such“missing heritability” are thought to be genetic modifiers (GxG orepistasis) and environmental interactions (GxE) (Eichler et al.,2010; Gibson, 2010; Manolio et al., 2009). Even for “simple”monogenetic disorders (e.g. cystic fibrosis), we are only beginningto understand the genetic modifiers and environmentalinteractions that impact disease severity and penetrance (Nadeau,2001); in this way Mendelian phenotypes can act as complex traits(Dipple and McCabe, 2000). In addition, the influence ofthe genetic, cellular, and external environment can have sig-nificant impacts on craniofacial development. For instance, sexhormones have drastic effects on embryonic facial patterning(Cohen et al., 2014), juvenile development (Fujita et al., 2004;Marquez Hernandez et al., 2011; Verdonck et al., 1998), and boneremodeling at juvenile and adult stages (Frenkel et al., 2010; Nickset al., 2010). The hardness of diet can also have a significant impacton craniofacial geometry (Genbrugge et al., 2011; Parsons et al.,2014; Swiderski and Zelditch, 2013). The role of such sex-specificand GxE interactions represent a long-standing gap in ourknowledge of the etiology and pathophysiology of complex dis-eases that are unlikely to be answered with traditional mutagen-esis screens.

a model to understand normal and clinical craniofacial variation.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3

3. Harnessing natural variation in evolutionary mutants tounderstand complex traits

3.1. Nature as a potent “mutagenizing” force

While Charles Darwin and other 19th century biologist em-braced the astounding natural diversity of living things, the 20thcentury became the era of model organisms such as the fly andmouse (Davis, 2004). Technological advances in genome sequen-cing and molecular techniques have enabled modern biologists toonce again turn to natural variation in non-model organisms(discussed in Albertson et al. (2009); Bolker (2012); Mahar (2009);Schartl (2014)), spurring on the current “golden age” of com-parative and evolutionary genetics (Nadeau and Jiggins, 2010). Weargue that these advances also open up opportunities to useevolutionary systems to better understand the human condition.

Nature is a source of vast biodiversity, which has accumulatedover millions of years as populations are continually challenged bynatural (and sexual) selection. For craniofacial structures, trophicadaptations have resulted in morphologies exquisitely matched tofeeding niches. Sometimes these phenotypes are extreme and/ordiscontinuous (i.e., may be considered a phenotypic novelty), suchas the snout of an anteater or the cranio-dental skeleton of toothedwhales. Alternatively, adaptive radiations can produce extensivecontinuous variation in craniofacial form. Textbook examples ofthis continuous variation include Darwin's finches (Grant, 1986;Grant and Grant, 2014) and African cichlids (Kocher, 2004). Be-cause the mutations that underlie these adaptive phenotypes arepart of a well integrated genetic architecture (i.e., they are notdeleterious), such systems offer great potential to reveal new in-sights into the genes and mechanisms that underlie this importanttrait. Moreover, given that changes in gene regulation are theprimary source of both the development of complex traits in hu-mans (Hindorff et al., 2009; Hirschhorn and Daly, 2005; Khan-delwal et al.,, 2013; Korstanje and Paigen, 2002; Manolio, 2009)and morphological evolution (Carroll et al., 2005; Jacob,1977; Kingand Wilson, 1975; Stern, 2000; Wray, 2007), the study of naturalvariants should more accurately reflect the number, effect, modeof action, interaction, and environmental sensitivity of genes thatunderlie complex traits compared to induced mutagenesis screens.

3.2. Craniofacial development is conserved among vertebrates

During facial development, neural crest cells (NCCs) must co-ordinate a complex pattern of molecular and cellular signals fromthe ectoderm and the endoderm to properly induce, specify, mi-grate, proliferate, and differentiate into the bones and cartilages ofthe face (reviewed in (Brugmann et al., 2006; Santagati and Rijli,2003; Sauka-Spengler and Bronner-Fraser, 2008)). Alterations inany of these processes can produce clinical malformations orvariation on which natural selection can act.

One of the major biological findings of the late 20th centurywas that all multicellular animals share a conserved set of genesand circuits that govern development (Carroll et al., 2005). Cra-niofacial development is no exception; despite dramatic differ-ences in adult structures, the molecular and morphogenetic pro-cesses that mediate NCC and craniofacial development are highlyconserved among vertebrates (Knight and Schilling, 2006; Me-deiros and Crump, 2012; Schilling 1997). In other words, homologyoccurs at multiple levels, spanning adult skeletal elements, cellularmechanisms such as migration and fusion, and genes (Shubinet al., 2009; Wagner, 2014). This “deep homology” (Shubin, et al.,2009) allows insights into human facial development from non-mammalian systems, including both traditional models (e.g. zeb-rafish and chicken) and evolutionary models (e.g. cichlids). Forexample, mammalian and fish palates have distinct morphologies

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

and are comprised of non-homologous elements (Dougherty et al.,2013; Murray and Schutte, 2004; Swartz et al., 2011). Despite this,their development requires the action of homologous cell- andtissue-level processes including migration of cranial neural crestcells, as well as the coordinated outgrowth and fusion of multipleprocesses and structures (Dougherty et al., 2013; Murray andSchutte, 2004). Further, homologous genes regulate these cellularmechanisms. For instance, loss of irf6 or the microRNA Mirn140and Pdgf signaling network results in conspicuous orofacial cleft-ing in both humans and zebrafish (Dougherty et al., 2013; Eberhartet al., 2008; Kondo et al., 2002; Li et al., 2010; Rattanasopha et al.,2012). These insights confirm the deep homology of the vertebrateskull. While the component parts may vary across taxa, the un-derlying genetic and developmental logic that assemble thesestructures are highly conserved. This and other findings (see (Al-bertson et al., 2009)) support the idea that a common set of genesmay underlie adaptive diversification in the wild and facial dys-morphologies in the clinic. Therefore, insights gleaned from stu-dies in evolutionary models are directly relevant to human de-velopment and possibility even disease. In fact, as we illustratebelow, the same extreme phenotypic state could even be con-sidered a disease in one species/lineage and an adaptation in asecond lineage (Fig. 1).

4. The evolutionary history of cichlids makes them an idealsystem to study the genetic basis of craniofacial variation

Cichlids (family: Cichlidae) are tropical freshwater fish thathave undergone an extraordinary adaptive radiation in terms ofthe pace, number of species, and degree of morphological diversity(Fryer and Iles, 1972; Kocher, 2004; Salzburger and Meyer, 2004).In all, cichlids account for about 8% of all fish species (Kuraku andMeyer, 2008), making them one of the most species-rich verte-brate families (Salzburger and Meyer, 2004). Their main radiation(�2000 endemic species) occurred in the East African rift lakes,namely Lakes Malawi, Tanganyika, and Victoria. Within Lake Ma-lawi alone, there are upwards of 1000 species that have evolvedwithin �1–2 million years (Kornfield and Smith, 2000; Salzburgerand Meyer, 2004; Sturmbauer et al., 2001; Turner et al., 2001),with speciation events having occurred as recently as 1000 yearsago (Won et al., 2005).

4.1. Cichlids demonstrate unparalleled craniofacial variation, analo-gous to normal and clinical variation in humans

A hallmark of cichlid radiations is their incredible diversity inbehavior, color, body shape, and, critical for this synopsis, facialmorphologies (Fig. 1). Cichlids encompass an extensive variety oftrophic strategies, including algae grazing, snail crushing, sandsifting, egg predation, scale eating, and piscivorous hunting, andeach of these is associated with different morphological andfunctional demands (Fryer and Iles, 1972; Konings, 2001; Kornfieldand Smith, 2000) (Fig. 1). Despite this wide range of phenotypes,the primary axis of facial variation among all African cichlidsdistinguishes two generalized feeding strategies – biters andsuckers (Cooper et al., 2010). Fish with stout faces, small eyes,tricuspid teeth, and short, downturned jaws are adapted to bite,pick, or scrape attached algae or crush hard mollusks. On the otherhand, fish with elongated faces, larger eyes, unicuspid or bicuspidteeth, and longer, isognathus jaws have adapted to suction otherfish, small invertebrates, or algae from the water column (Albert-son et al., 2005; Cooper et al., 2010; Konings, 2001; Streelman andAlbertson, 2006). Notably, cichlid craniofacial variation hasevolved along this same axis independently in the three AfricanGreat Lakes (Lakes Malawi, Tanganyika, and Victoria) (Cooper

a model to understand normal and clinical craniofacial variation.

Fig. 1. Cichlids demonstrate extensive craniofacial variation analogous to normal and clinical variation in humans. Comparison of human (top) and cichlid (bottom) facialvariation. Most cichlid facial variation is continuous across species adapted to bite or scrape food from the substrate (left), or to suck food from the water column (right). Thisis similar to the continuous spectrum of “normal” facial variation found among human populations. Notably, some cichlid species have evolved extreme morphologies thatmimic pathological variation in humans such as micrognathia (far left) and midface hypoplasia (far right). Cichlid images courtesy of Ad Konings. All cichlids are from LakeMalawi. From left to right, including feeding mode (Konings 2001): Labeotropheus fuelleborni (biting: scraping connected algae), Tropheops sp. ‘taiwan’ (biting: nip/twist ofconnected algae), Cynotilapia zebroides (suctioning: plankton from water column or combing loose algae), Tyrannochromis nigriventer (suctioning: ambush predator of otherfish), Aulonocara rostratum (sonar hunting of invertebrates buried in sand), and Caprichromis orthognathus (head-butts mouth brooding females, then eats released eggs andlarvae). Illustrations of human skulls from different populations (from left to right: American Indian, European, and Sub-Saharan African) by Diana Marques, courtesy of theSmithsonian's National Museum of Natural History.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎4

et al., 2010; Fryer and Iles, 1972). Despite their independent evo-lution over markedly different time scales, cichlids from each lakedemonstrate remarkable convergence in craniofacial morpholo-gies (Albertson et al., 2003a; Cooper et al., 2010; Fryer and Iles,1972). These “experimental replicates” provide a powerful oppor-tunity to determine whether convergent morphologies result fromalterations of the same genes and pathways, or whether there aremultiple genetic and cellular mechanisms that result in similarphenotypes. Not only is this an important question in evolutionarytheory, but such insights into genetic architecture would alsoprovide a better understanding of which aspects of craniofacialanatomy are developmentally plastic and which are inflexible (i.e.canalized).

Craniofacial variation among most cichlid species is continuousbetween alternate short versus long face morphologies, and in thisway analogous to “normal” facial variation among human popu-lations (Fig. 1). However, certain species have evolved extrememorphologies adapted to specialized feeding methods. In theseinstances, craniofacial morphologies between cichlid lineages aremore discontinuous, mimicking pathological variation in the hu-man facial skeleton such as micrognathia, midface hypoplasia, andfacial asymmetries (Fig. 1, (Albertson et al., 2009)). As with humandysmorphologies, craniofacial differences among cichlids can bedetected as early as NCC migration (Albertson and Kocher, 2006;Powder et al., 2014), with additional shape differences accumu-lating into larval and juvenile stages (Parsons et al., 2014; Powderet al., 2015). Cichlids are thus an ideal evolutionary system tostudy the genetic and developmental origins of both subtle andextreme facial variation due to their (1) extensive morphologicalvariation, (2) mimicking of variation found in humans, and (3) si-milarities to mammals at the molecular and tissue level.

4.2. The genomic architecture of Lake Malawi cichlids is ideal forassociation and QTL mapping studies

Despite their remarkable phenotypic diversity, cichlids exhibita relatively high degree of genomic homogeneity due to their rapidevolution (�1–2 million years for Lake Malawi (Sturmbauer et al.,2001)) and ongoing hybridization. Five complete cichlid genomesare now available (Brawand et al., 2014), and among the majorinsights from this work is that approximately half of the

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

polymorphisms identified among species (which shared a com-mon ancestor �10 million years ago) exhibit incomplete lineagesorting. In other words, a significant proportion of ancestralpolymorphisms are still segregating between species via geneticrecombination and hybridization (Brawand et al., 2014; Loh et al.,2013). The result is a young species flock characterized by rela-tively low levels of nucleotide diversity (i.e. lower than that be-tween different laboratory strains of zebrafish, (Loh et al., 2008))and high levels of shared polymorphisms. Notably, these proper-ties mean that genetic/genomic comparisons between cichlidspecies (or genera) are analogous to those between humanpopulations.

A combination of high phenotypic variation and low genotypicvariation make cichlids an ideal system for population genomics.Specifically, given the high degree of shared polymorphisms(Brawand et al., 2014; Loh et al., 2013), global genetic differentia-tion is low among Lake Malawi cichlids (average FSTE0.2 (Lohet al., 2008)). Thus outlier loci that exhibit signatures of divergence(e.g. high FST values) are de facto candidates for trait differencesbetween species or genera. Essentially, neutral polymorphismshave not had time to become fixed between species, which meansthat outlier loci are likely to exist because they mediate a phe-notype that is under divergent selection. For instance, a SNP nearthe neural patterning gene irx1 was found to exhibit an outlier FSTvalue in a genome-wide comparison between cichlid lineages thathave distinct brain morphologies (Loh et al., 2008). This observa-tion suggested that irx1 might mediate differences in brain de-velopment between these fishes. Subsequent functional studiessupported this hypothesis by showing that differential expressionof irx1 is associated with changes in brain patterning that result indifferences in the relative size of the telencephalon versus thethalamus in cichlids (Sylvester et al., 2010).

With accumulating genomic resources it is now relatively easyto identify divergent loci among cichlids, however linking theseloci to a specific trait is not always straightforward because spe-cies/lineages usually differ with respect to multiple traits. Pedigreemapping of quantitative traits (e.g., QTL mapping) is a bettermethod with which to link variation in genotype with variation ina specific phenotype. Given their recent divergence, most LakeMalawi cichlid species (and many genera) produce fertile hybrids,enabling the generation of large pedigrees for QTL mapping (e.g.

a model to understand normal and clinical craniofacial variation.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 5

craniofacial traits in (Albertson et al., 2003b; Albertson et al.,2005; Cooper et al., 2011; Hu and Albertson, 2014; Parsons et al.,2012; Parsons et al., 2015; Stewart and Albertson, 2010; Streelmanand Albertson, 2006)). A key limitation of QTL mapping however isthat the identification of causative genes and/or polymorphismsfor a QTL can be challenging and expensive due the time and laborinvolved in generating sufficient numbers of recombinant animalsfor fine mapping. Thus, population genomics can efficiently iden-tify divergent loci that are under selection, but cannot link geno-type to a specific phenotype. Alternatively, QTL mapping can sta-tistically associate genetic intervals to specific phenotypes, butoften lacks the resolution to pinpoint specific causative genes andnucleotides.

Recently, investigators have begun to address these methodo-logical limitations by combining both population genomics andQTL mapping into a single experiment, thereby leveraging theadvantages while mitigating the disadvantages of each approach(Albertson et al., 2014; Barrett and Hoekstra, 2011; Parsons andAlbertson, 2013). For example, we previously used this approach toimplicate pax3a in cichlid coloration (Albertson et al., 2014). Wefirst used QTL mapping to associate a 28 cM region with levels ofred/yellow pigmentation. This region corresponds to approxi-mately 11 Mb of physical sequence and contains �315 genes,making it daunting to identify candidate genes. Population geno-mic analysis of this region identified 366 genetic polymorphismsbetween cichlid species that also differ with respect to pigmen-tation. However, of these, just 17 exhibited high levels of differ-entiation between the species, suggesting they may be driving thephenotypic differences. One was in the 5'UTR of pax3a, which hasbeen previously associated with specification of pigment cells(Minchin and Hughes, 2008). This example demonstrates howpopulation genomic analysis, which takes advantage of re-combination in the wild, can be used to effectively narrow the listof candidate genes that underlie a QTL.

5. Cichlid studies can complement traditional approaches inmodel organisms

Given their morphological diversity and genomic architecture,cichlids offer a powerful evolutionary system to complement tra-ditional approaches in model organisms such as Mus musculus andGallus gallus. With the vast number of genetic tools available inlaboratory models (e.g., genome editing), the validation of candi-date genes identified in cichlids may require experiments in modelsystems. That said, functional validation of candidate genes (or atleast pathways) can also be performed in the cichlid system. Thechoice of whether to examine the mechanism of gene actionwithin cichlids themselves or within traditional model organismsis dependent on the particulars of the candidate gene or pathway.Cichlids are readily susceptible to pharmacological manipulationsthrough addition of small molecules to their water (Bloomquistet al., 2015; Fraser et al., 2008, 2013; Hu and Albertson, 2014;Parsons and Albertson, 2013; Parsons et al., 2014; Powder et al.,2015; Roberts et al., 2011; Sylvester et al., 2010, 2013). Thesechemicals have varying degrees of target specificity (i.e. may an-tagonize the entire Wnt pathway versus a specific ligand), though,and result in embryo-wide changes in gene expression. For can-didates that do not have pharmacological agents or require in-creased specificity, mechanism of gene action can be determinedusing the battery of experimental methods available in modelorganisms such as zebrafish and Xenopus, including transgenics,injections (morpholinos and/or overexpression), and transplanta-tions (Albertson et al., 2005; Powder et al., 2014). There is alsomuch promise for bringing such experiments from traditionalmodel organisms into the cichlid system. For instance, cichlids are

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

amenable to transgenesis (Fujimura and Kocher, 2011; Juntti et al.,2013), and CRISPR/Cas9-mediated genome editing (Sternberg andDoudna, 2015) is a powerful new tool that has shown great pro-mise in non-model organisms, including cichlids (Li et al., 2014).Below we briefly illustrate how cichlids, in conjunction with suchmechanistic analyses, have yielded novel insights into craniofacialdevelopment.

5.1. Lbh is a novel regulator of neural crest cell migration

Neurocristopathies (Bolande, 1974) are syndromes resultingfrom abnormal neural crest cell (NCC) development, commonlyNCC migration or proliferation (Hall, 2009). Given the wide varietyof NCC derivatives (Hall, 2009; Santagati and Rijli, 2003; Sauka-Spengler and Bronner-Fraser, 2008), neurocristopathies oftenpresent with a suite of defects in craniofacial, neural, otic, cardiac,and pigment structures or tissues. We recently identified a novelregulator of neural crest cell development in cichlids. We startedwith a QTL on linkage group 19 that contributed to variation in therelative length of the mandible (Albertson et al., 2005; Cooperet al., 2011). Notably, one of the cichlid species used to generatethis cross exhibits an extremely short jaw (“evolved micro-gnathia”) (Fig. 1, far left). Using population genomics, we were ableto narrow this region and identify limb bud and heart homolog (lbh)as a candidate gene (Powder et al., 2014). Lbh is a largely un-characterized putative transcriptional regulator (Ai et al., 2008)expressed during vertebrate limb and heart development (Briegeland Joyner, 2001), but previously not known to regulate facialdevelopment.

Using loss-of-function analyses in zebrafish and Xenopus, wedemonstrated that lbh is necessary in neural crest cells for propermigration (Powder et al., 2014). Severe depletion of Lbh resulted indramatic inhibition of migration and complete loss of NCC-derivedfacial structures, resulting in lethality. Cichlids with alternate shortand long jaw morphologies have distinct SNPs around lbh, in-cluding a non-synonymous R4Q mutation in the evolutionarilyderived fish with a short mandible. We showed that this singleamino acid change resulted in discrete shifts in neural crest cellmigration patterns. Whereas the “long jaw” variant causes moreNCCs to migrate to the mandibular arch, the “short jaw” form re-sults in fewer NCCs migrating anteriorly; this depletion of theanterior pool of NCCs may result in the diminished mandible de-rived from this population of cells. Even though lbh plays a criticalrole in NCC migration and contributes to quantitative variation incraniofacial structures, this gene had not been previously identi-fied in traditional screens for craniofacial regulators, perhaps be-cause severe abrogation of Lbh results in lethality (Powder et al.,2014).

5.2. Wnt and Hedgehog pathways regulate species-specific bonedevelopment

Contrary to our extensive understanding of how early devel-opmental events and genetic mechanisms regulate neural crestcell induction, migration, and patterning (Brugmann et al., 2006;Santagati and Rijli, 2003; Sauka-Spengler and Bronner-Fraser,2008), comparatively little is known about the mechanisms thatdetermine the shape of bones. One challenge to understandingskeletal morphologies is the re-deployment of signaling pathwaysduring multiple stages of development. For example, the Wntsignaling network plays successive critical roles during NCC in-duction (reviewed in (Sauka-Spengler and Bronner-Fraser, 2008)),proliferation and outgrowth of facial mesenchyme (Brugmannet al., 2007, 2010), fusion of facial prominences (Lan et al., 2006;Song et al., 2009), and osteogenesis (reviewed in (Hartmann,2006; Long, 2012)). Likewise, the Hedgehog developmental

a model to understand normal and clinical craniofacial variation.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎6

pathway is necessary for NCC survival (Ahlgren and Bronner-Fra-ser, 1999; Billmyre and Klingensmith, 2015), facial patterning andoutgrowth at multiple stages (Cordero et al., 2004; Heyne et al.,2015; Hu et al., 2003; Jeong et al., 2004), and chondrogenesis andosteogenesis (Mo et al., 1997; Wada et al., 2005; Yang et al., 2015).This developmental pleiotropy, the iterative usage of a commonpathway in multiple tissues or stages, can limit the assessment oflater developmental events. For example, conditional loss of acentral mediator of Wnt signaling, β-catenin, in neural crest cellsresults in early loss of these cells (Brault et al., 2001), precludingassessment of how changes in expression levels, timing, and pat-tern can result in quantitative differences in bone shape.

Recently work has combined QTL mapping, population geno-mics, and experimental manipulations in cichlids to associate theHedgehog and Wnt pathways with subtle, quantitative variation inbone shape (Hu and Albertson, 2014; Loh et al., 2008; Parsonset al., 2014; Roberts et al., 2011). In one study, we mapped QTL forthe shape of the interopercle bone and length of the retroarticularprocess of the mandible to the same genetic interval (Hu and Al-bertson, 2014; Roberts et al., 2011). Population genomics identifieda SNP in this region that is highly divergent in cichlids with al-ternate morphologies, and resides upstream of the Hedgehog re-ceptor ptch1. The evolutionarily derived ptch1 allele is associatedwith decreased expression levels of this gene and thus decreasedbone deposition, ultimately resulting in a shorter retroarticularprocess and narrower interopercle bone. These shape changeshave important functional consequences, wherein the derived al-lele confers decreased bite forces but faster jaw movements to fishthat eat by suctioning food from the water column (Hu and Al-bertson, 2014; Roberts et al., 2011). Population genomics alsoidentified a highly divergent SNP in β-catenin between cichlidswith alternate craniofacial shapes (Loh et al., 2008). This SNP isassociated with increased levels of Wnt signaling and acceleratedrates of bone deposition in the cichlid face (Parsons et al., 2014).The accelerated bone development prevents outgrowth of thepreorbital region, effectively locking into place an early larvalphenotype (i.e., paedomorphosis) (Parsons et al., 2014; Powderet al., 2015). This ultimately results in a steeper craniofacial slopemore resistant to strong bite forces in an animal that eats byshearing algae off of rocks (Cooper et al., 2011; Parsons et al.,2014). These studies illustrate how common developmentalpathways, often studied at early developmental stages, are itera-tively deployed over extended periods of development to producecomplex morphologies.

5.3. Plasticity: remodeling and GxE interactions

A critical, but under-studied, aspect of craniofacial biology isthe extent to which the environment can influence the genotype–phenotype (G–P) map and thus determine how the craniofacialskeleton develops and evolves. “Environment” is a broad term thatcould refer to either an organism's internal or external environ-ment. The internal environment includes such factors as endocrinesignaling and the biophysical interactions among cells or tissues.The external environment is the physical (or physiological) con-ditions in which the organism develops, including light, tem-perature, and mechanical stresses. Both types of environments willact to filter the set of genetic variants that contribute to pheno-typic variation (Jamniczky et al., 2010). Environmental effectscannot be accounted for in GWAS experiments and are thought tobe an important source of the “missing” heritability in complextraits, including the craniofacial skeleton (Eichler et al., 2010;Gibson, 2010; Manolio et al., 2009).

In the interest of space we limit our discussion to the externalforces that influence the shape of the craniofacial skeleton, mainlyforaging environment that can impose distinct kinematic demands

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

on the feeding apparatus. Bone is a dynamic tissue that can senseand respond to its mechanical environment (reviewed in (Alaqeelet al., 2006; Nguyen and Jacobs, 2013; Papachroni et al., 2009;Robling and Turner, 2009)). Consistent with this, genes that par-ticipate in bone cell development and matrix deposition werefound to be up or down regulated during remodeling of the cichlidpharyngeal jaw in response to diet changes (Gunter et al., 2013;Schneider et al., 2014). However, despite an emerging under-standing of plasticity at the transcript level, we lack a compre-hensive picture of the genetic architecture of this important trait.For instance, how many genes contribute to a plastic response? Dodifferent sets of genes underlie trait development in differentenvironments? Or is plasticity regulated by the same genes actingacross environments but in which alleles are differentially sensi-tive to environmental conditions? Genetic mapping experimentsperformed in different environments can help to address theseoutstanding questions (e.g., (Kuttner et al., 2014)), and we haverecently performed such studies in cichlids (Parsons, et al., inpreparation). Specifically, F3 hybrid families were split and rearedfor 6 months in alternate foraging environments that imposeddistinct kinematic demands on the feeding apparatus, but wereotherwise nutritionally identical. The mapping results demon-strate that the foraging environment strongly influences the G–Pmap. Of 22 identified QTL for a variety of traits, only 1 was sig-nificant in both environments; in other words, distinct loci reg-ulate the response to distinct environments. This observationsuggests that cryptic genetic variation, a pool of genetic variationthat does not result in phenotypic variation under “normal” con-ditions (Gibson and Dworkin, 2004; Paaby and Rockman, 2014), isprominent for the craniofacial skeleton. We further detected in-stances of allele sensitivity to foraging conditions in that onehaplotype showed a plastic response while the alternate haplotypedid not (Parsons, et al., in preparation). While much work is nee-ded to determine the mechanistic basis of such GxE interactions,studies like this confirm that the foraging environment is a potentforce that can influence how genetic variation translates to mor-phology. Given differences in food items between distinct humanpopulations (e.g., hunter-gatherer to agrarian transitions in earlyhumans as well as modern geographical differences in diet), theseresults are directly relative to human craniofacial biology. Moregenerally, they support a shift to an eco-devo approach (Abouheifet al., 2014; Gilbert and Epel, 2008), wherein the environment isexplicitly incorporated into genetic mapping and developmentalstudies.

6. Conclusions

Despite our extensive knowledge of early craniofacial pattern-ing, we are relatively ignorant of the genetic, cellular, and en-vironmental mechanisms that combine over extended periods ofdevelopment to produce “normal” variation in humans, and un-derlie many complex craniofacial malformations. We propose thatstudying natural variation in non-model animals will be an ef-fective and fruitful approach to determine the types of genes,mutations, and environmental influences that ultimately result inthe extraordinary range of human faces. The “evolutionary mu-tant” cichlid system is ideal for this approach, given their un-paralleled array of facial variation and powerful genetic tools.Experiments described here complement traditional approaches,and will advance our understanding of the genomic and devel-opmental origins of complex morphologies.

a model to understand normal and clinical craniofacial variation.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 7

Acknowledgments

This work was supported by NSF Career IOS-1054909 (RCA) andNIH/NIDCR 1F32DE023707 (KEP). We thank an anonymous re-viewer for insightful comments and suggestions.

References

Abouheif, E., Fave, M.J., Ibarraran-Viniegra, A.S., Lesoway, M.P., Rafiqi, A.M., Raja-kumar, R., 2014. Eco-evo-devo: the time has come. Adv. Exp. Med. Biol. 781,107–125.

Ahlgren, S.C., Bronner-Fraser, M., 1999. Inhibition of sonic hedgehog signalingin vivo results in craniofacial neural crest cell death. Curr. Biol. 9 (22),1304–1314.

Ai, J., Wang, Y., Tan, K., Deng, Y., Luo, N., Yuan, W., Wang, Z., Li, Y., Wang, Y., Mo, X.,et al., 2008. A human homolog of mouse Lbh gene, hLBH, expresses in heart andactivates SRE and AP-1 mediated MAPK signaling pathway. Mol. Biol. Rep. 35(2), 179–187.

Alaqeel, S.M., Hinton, R.J., Opperman, L.A., 2006. Cellular response to force appli-cation at craniofacial sutures. Orthod. Craniofacial Res. 9 (3), 111–122.

Albertson, R.C., Cresko, W.A., Detrich 3rd, H., Postlethwait, J.H., 2009. Evolutionarymutant models for human disease. Trends Genet. 25 (2), 74–81.

Albertson, R.C., Kocher, T.D., 2006. Genetic and developmental basis of cichlidtrophic diversity. Heredity 97 (3), 211–221.

Albertson, R.C., Powder, K.E., Hu, Y., Coyle, K.P., Roberts, R.B., Parsons, K.J., 2014.Genetic basis of continuous variation in the levels and modular inheritance ofpigmentation in cichlid fishes. Mol. Ecol. 23 (21), 5135–5150.

Albertson, R.C., Streelman, J.T., Kocher, T.D., 2003a. Directional selection has shapedthe oral jaws of Lake Malawi cichlid fishes. Proc. Natl. Acad. Sci. USA 100 (9),5252–5257.

Albertson, R.C., Streelman, J.T., Kocher, T.D., 2003b. Genetic basis of adaptive shapedifferences in the cichlid head. J. Hered. 94 (4), 291–301.

Albertson, R.C., Streelman, J.T., Kocher, T.D., Yelick, P.C., 2005. Integration andevolution of the cichlid mandible: the molecular basis of alternate feedingstrategies. Proc. Natl. Acad. Sci. USA 102 (45), 16287–16292.

Albertson, R.C., Yelick, P.C., 2004. Morphogenesis of the jaw: development beyondthe embryo. Methods Cell Biol. 76, 437–454.

Barrett, R.D., Hoekstra, H.E., 2011. Molecular spandrels: tests of adaptation at thegenetic level. Nat. Rev. Genet. 12 (11), 767–780.

Belloni, E., Muenke, M., Roessler, E., Traverso, G., Siegel-Bartelt, J., Frumkin, A.,Mitchell, H.F., Donis-Keller, H., Helms, C., Hing, A.V., et al., 1996. Identification ofSonic hedgehog as a candidate gene responsible for holoprosencephaly. Nat.Genet. 14 (3), 353–356.

Billmyre, K.K., Klingensmith, J., 2015. Sonic hedgehog from pharyngeal arch 1 epi-thelium is necessary for early mandibular arch cell survival and later cartilagecondensation differentiation. Dev. Dyn. 244 (4), 564–576.

Bloomquist, R.F., Parnell, N.F., Phillips, K.A., Fowler, T.E., Yu, T.Y., Sharpe, P.T.,Streelman, J.T., 2015. Coevolutionary patterning of teeth and taste buds. Proc.Natl. Acad. Sci. USA 112 (44), E5954–E5962.

Boehringer, S., van der Lijn, F., Liu, F., Gunther, M., Sinigerova, S., Nowak, S., Ludwig,K.U., Herberz, R., Klein, S., Hofman, A., et al., 2011. Genetic determination ofhuman facial morphology: links between cleft-lips and normal variation. Eur. J.Hum. Genet. 19 (11), 1192–1197.

Bolande, R.P., 1974. The neurocristopathies: a unifying concept of disease arising inneural crest maldevelopment. Hum. Pathol. 5 (4), 409–429.

Bolker, J., 2012. Model organisms: There's more to life than rats and flies. Nature491 (7422), 31–33.

Brault, V., Moore, R., Kutsch, S., Ishibashi, M., Rowitch, D.H., McMahon, A.P., Som-mer, L., Boussadia, O., Kemler, R., 2001. Inactivation of the beta-catenin gene byWnt1-Cre-mediated deletion results in dramatic brain malformation and fail-ure of craniofacial development. Development 128 (8), 1253–1264.

Brawand, D., Wagner, C.E., Li, Y.I., Malinsky, M., Keller, I., Fan, S., Simakov, O., Ng, A.Y., Lim, Z.W., Bezault, E., et al., 2014. The genomic substrate for adaptive ra-diation in African cichlid fish. Nature 513 (7518), 375–381.

Briegel, K., Joyner, A., 2001. Identification and characterization of Lbh, a novelconserved nuclear protein expressed during early limb and heart development.Dev. Biol. 233 (2), 291–304.

Brugmann, S.A., Goodnough, L.H., Gregorieff, A., Leucht, P., ten Berge, D., Fuerer, C.,Clevers, H., Nusse, R., Helms, J.A., 2007. Wnt signaling mediates regional spe-cification in the vertebrate face. Development 134 (18), 3283–3295.

Brugmann, S.A., Powder, K.E., Young, N.M., Goodnough, L.H., Hahn, S.M., James, A.W., Helms, J.A., Lovett, M., 2010. Comparative gene expression analysis of avianembryonic facial structures reveals new candidates for human craniofacialdisorders. Hum. Mol. Genet. 19 (5), 920–930.

Brugmann, S.A., Tapadia, M.D., Helms, J.A., 2006. The molecular origins of species-specific facial pattern. Curr. Top. Dev. Biol. 73, 1–42.

Campbell, L.J., Willoughby, J.J., Jensen, A.M., 2012. Two types of Tet-On transgeniclines for doxycycline-inducible gene expression in zebrafish rod photoreceptorsand a gateway-based tet-on toolkit. PLoS ONE 7 (12), e51270.

Carroll, S.B., Grenier, J.K., Weatherbee, S.D., 2005. From DNA to Diversity: MolecularGenetics and the Evolution of Animal Design, 2nd Ed. Blackwell, Oxford.

Chan, W., Costantino, N., Li, R., Lee, S.C., Su, Q., Melvin, D., Court, D.L., Liu, P., 2007. A

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

recombineering based approach for high-throughput conditional knockouttargeting vector construction. Nucleic Acids Res. 35 (8), e64.

Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H., Beachy, P.A.,1996. Cyclopia and defective axial patterning in mice lacking Sonic hedgehoggene function. Nature 383 (6599), 407–413.

Cohen, S.P., LaChappelle, A.R., Walker, B.S., Lassiter, C.S., 2014. Modulation of es-trogen causes disruption of craniofacial chondrogenesis in Danio rerio. Aquat.Toxicol. 152, 113–120.

Cooper, W.J., Parsons, K., McIntyre, A., Kern, B., McGee-Moore, A., Albertson, R.C.,2010. Bentho-pelagic divergence of cichlid feeding architecture was prodigiousand consistent during multiple adaptive radiations within African rift-lakes.PLoS One 5 (3), e9551.

Cooper, W.J., Wernle, J., Mann, K., Albertson, R.C., 2011. Functional and genetic in-tegration in the skulls of Lake Malawi Cichlids. Evol. Biol. 38 (3), 316–334.

Cordero, D., Marcucio, R., Hu, D., Gaffield, W., Tapadia, M., Helms, J.A., 2004. Tem-poral perturbations in sonic hedgehog signaling elicit the spectrum of holo-prosencephaly phenotypes. J. Clin. Investig. 114 (4), 485–494.

Davis, R.H., 2004. The age of model organisms. Nat. Rev. Genet. 5 (1), 69–76.Dipple, K.M., McCabe, E.R., 2000. Phenotypes of patients with “simple” Mendelian

disorders are complex traits: thresholds, modifiers, and systems dynamics. Am.J. Hum. Genet. 66 (6), 1729–1735.

Dixon, J., Dixon, M.J., 2004. Genetic background has a major effect on the pene-trance and severity of craniofacial defects in mice heterozygous for the geneencoding the nucleolar protein Treacle. Dev. Dyn. 229 (4), 907–914.

Dougherty, M., Kamel, G., Grimaldi, M., Gfrerer, L., Shubinets, V., Ethier, R., Hickey,G., Cornell, R.A., Liao, E.C., 2013. Distinct requirements for wnt9a and irf6 inextension and integration mechanisms during zebrafish palate morphogenesis.Development 140 (1), 76–81.

Eberhart, J.K., He, X., Swartz, M.E., Yan, Y.L., Song, H., Boling, T.C., Kunerth, A.K.,Walker, M.B., Kimmel, C.B., Postlethwait, J.H., 2008. MicroRNA Mirn140 mod-ulates Pdgf signaling during palatogenesis. Nat. Genet. 40 (3), 290–298.

Eichler, E.E., Flint, J., Gibson, G., Kong, A., Leal, S.M., Moore, J.H., Nadeau, J.H., 2010.Missing heritability and strategies for finding the underlying causes of complexdisease. Nat. Rev. Genet. 11 (6), 446–450.

Fish, J.L., Sklar, R.S., Woronowicz, K.C., Schneider, R.A., 2014. Multiple develop-mental mechanisms regulate species-specific jaw size. Development 141 (3),674–684.

Fraser, G.J., Bloomquist, R.F., Streelman, J.T., 2008. A periodic pattern generator fordental diversity. BMC Biol. 6, 32.

Fraser, G.J., Bloomquist, R.F., Streelman, J.T., 2013. Common developmental path-ways link tooth shape to regeneration. Dev. Biol. 377 (2), 399–414.

Frenkel, B., Hong, A., Baniwal, S.K., Coetzee, G.A., Ohlsson, C., Khalid, O., Gabet, Y.,2010. Regulation of adult bone turnover by sex steroids. J. Cell. Physiol. 224 (2),305–310.

Fryer, G., Iles, T.D., 1972. The Cichlid Fishes of the Great Lakes of Africa: TheirBiology and Evolution. T.F.H Publications, Neptune City, New Jersey.

Fujimura, K., Kocher, T.D., 2011. Tol2-mediated transgenesis in tilapia (Oreochromisniloticus). Aquaculture 319 (3–4), 342–346.

Fujita, T., Ohtani, J., Shigekawa, M., Kawata, T., Kaku, M., Kohno, S., Tsutsui, K., Tenjo,K., Motokawa, M., Tohma, Y., et al., 2004. Effects of sex hormone disturbanceson craniofacial growth in newborn mice. J. Dent. Res. 83 (3), 250–254.

Genbrugge, A., Heyde, A.S., Adriaens, D., Boone, M., Van Hoorebeke, L., Dirckx, J.,Aerts, P., Podos, J., Herrel, A., 2011. Ontogeny of the cranial skeleton in a Dar-win's finch (Geospiza fortis). J. Anat. 219 (2), 115–131.

Gibson, G., 2010. Hints of hidden heritability in GWAS. Nat. Genet. 42 (7), 558–560.Gibson, G., Dworkin, I., 2004. Uncovering cryptic genetic variation. Nat. Rev. Genet.

5 (9), 681–690.Gilbert, S.F., Epel, D., 2008. Ecological Developmental Biology: Integrating Epige-

netics, Medicine, and Evolution. Sinauer Associates, Sunderland, MA.Glazier, A.M., Nadeau, J.H., Aitman, T.J., 2002. Finding genes that underlie complex

traits. Science 298 (5602), 2345–2349.Grant, P.R., 1986. Ecology and Evolution of Darwin's Finches. Princeton University

Press, Princeton.Grant, P.R., Grant, B.R., 2014. 40 Years of Evolution: Darwin's Finches on Daphne

Major Island. Princeton University Press, Princeton.Gunter, H.M., Fan, S., Xiong, F., Franchini, P., Fruciano, C., Meyer, A., 2013. Shaping

development through mechanical strain: the transcriptional basis of diet-in-duced phenotypic plasticity in a cichlid fish. Mol. Ecol. 22 (17), 4516–4531.

Hall, B.K., 2009. The Neural Crest and Neural Crest Cells in Vertebrate Developmentand Evolution. Springer, New York.

Hallgrimsson, B., Jamniczky, H., Young, N.M., Rolian, C., Parsons, T.E., Boughner, J.C.,Marcucio, R.S., 2009. Deciphering the Palimpsest: Studying the RelationshipBetween Morphological Integration and Phenotypic Covariation. Evol. Biol. 36(4), 355–376.

Hallgrimsson, B., Mio, W., Marcucio, R.S., Spritz, R., 2014. Let's Face It-ComplexTraits Are Just Not That Simple. Plos. Genet. 10 (11), e1004724.

Hartmann, C., 2006. A Wnt canon orchestrating osteoblastogenesis. Trends Cell.Biol. 16 (3), 151–158.

Heyne, G.W., Melberg, C.G., Doroodchi, P., Parins, K.F., Kietzman, H.W., Everson, J.L.,Ansen-Wilson, L.J., Lipinski, R.J., 2015. Definition of critical periods for Hedge-hog pathway antagonist-induced holoprosencephaly, cleft lip, and cleft palate.Plos. One 10 (3), e0120517.

Hindorff, L.A., Sethupathy, P., Junkins, H.A., Ramos, E.M., Mehta, J.P., Collins, F.S.,Manolio, T.A., 2009. Potential etiologic and functional implications of genome-wide association loci for human diseases and traits. Proc. Natl. Acad. Sci. USA106 (23), 9362–9367.

a model to understand normal and clinical craniofacial variation.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎8

Hirschhorn, J.N., Daly, M.J., 2005. Genome-wide association studies for commondiseases and complex traits. Nat. Rev. Genet. 6 (2), 95–108.

Hochheiser, H., Aronow, B.J., Artinger, K., Beaty, T.H., Brinkley, J.F., Chai, Y., Clouthier,D., Cunningham, M.L., Dixon, M., Donahue, L.R., et al., 2011. The FaceBaseConsortium: a comprehensive program to facilitate craniofacial research. Dev.Biol. 355 (2), 175–182.

Hu, D., Helms, J.A., 1999. The role of sonic hedgehog in normal and abnormalcraniofacial morphogenesis. Development 126 (21), 4873–4884.

Hu, D., Marcucio, R.S., Helms, J.A., 2003. A zone of frontonasal ectoderm regulatespatterning and growth in the face. Development 130 (9), 1749–1758.

Hu, Y., Albertson, R.C., 2014. Hedgehog signaling mediates adaptive variation in adynamic functional system in the cichlid feeding apparatus. Proc. Natl. Acad.Sci. USA 111 (23), 8530–8534.

Jacob, F., 1977. Evolution and tinkering. Science 196 (4295), 1161–1166.Jamniczky, H.A., Boughner, J.C., Rolian, C., Gonzalez, P.N., Powell, C.D., Schmidt, E.J.,

Parsons, T.E., Bookstein, F.L., Hallgrimsson, B., 2010. Rediscovering Waddingtonin the post-genomic age: Operationalising Waddington's epigenetics revealsnew ways to investigate the generation and modulation of phenotypic varia-tion. Bioessays 32 (7), 553–558.

Jeong, J., Mao, J., Tenzen, T., Kottmann, A.H., McMahon, A.P., 2004. Hedgehog sig-naling in the neural crest cells regulates the patterning and growth of facialprimordia. Genes Dev. 18 (8), 937–951.

Juntti, S.A., Hu, C.K., Fernald, R.D., 2013. Tol2-mediated generation of a transgenichaplochromine cichlid, Astatotilapia burtoni. PLoS One 8 (10), e77647.

Kadakia, S., Helman, S.N., Badhey, A.K., Saman, M., Ducic, Y., 2014. Treacher CollinsSyndrome: the genetics of a craniofacial disease. Int. J. Pediatr. Otorhinolar-yngol. 78 (6), 893–898.

Khandelwal, K.D., van Bokhoven, H., Roscioli, T., Carels, C.E., Zhou, H., 2013. Geno-mic approaches for studying craniofacial disorders. Am. J. Med. Genet. C Semin.Med. Genet. 163C (4), 218–231.

Kile, B.T., Hilton, D.J., 2005. The art and design of genetic screens: mouse. Nat. Rev.Genet. 6 (7), 557–567.

King, M.C., Wilson, A.C., 1975. Evolution at two levels in humans and chimpanzees.Science 188 (4184), 107–116.

Knight, R.D., Schilling, T.F., 2006. Cranial neural crest and development of the headskeleton. Adv. Exp. Med. Biol. 589, 120–133.

Kocher, T.D., 2004. Adaptive evolution and explosive speciation: the cichlid fishmodel. Nat. Rev. Genet. 5 (4), 288–298.

Kondo, S., Schutte, B.C., Richardson, R.J., Bjork, B.C., Knight, A.S., Watanabe, Y., Ho-ward, E., de Lima, R.L., Daack-Hirsch, S., Sander, A., et al., 2002. Mutations inIRF6 cause Van der Woude and popliteal pterygium syndromes. Nat. Genet. 32(2), 285–289.

Konings, A., 2001. Malawi cichlids in their natural habitat. Cichlid Press, El Paso, TX.Kornfield, I., Smith, P.F., 2000. African Cichlid Fishes: Model Systems for Evolu-

tionary Biology. Annu. Rev. Ecol. Syst. 31, 163–196.Korstanje, R., Paigen, B., 2002. From QTL to gene: the harvest begins. Nat. Genet. 31

(3), 235–236.Kuraku, S., Meyer, A., 2008. Genomic analysis of cichlid fish ‘natural mutants. Curr.

Opin. Genet. Dev. 18 (6), 551–558.Kuttner, E., Parsons, K.J., Easton, A.A., Skulason, S., Danzmann, R.G., Ferguson, M.M.,

2014. Hidden genetic variation evolves with ecological specialization: the ge-netic basis of phenotypic plasticity in Arctic charr ecomorphs. Evol. Dev. 16 (4),247–257.

Lan, Y., Ryan, R.C., Zhang, Z., Bullard, S.A., Bush, J.O., Maltby, K.M., Lidral, A.C., Jiang,R., 2006. Expression of Wnt9b and activation of canonical Wnt signaling duringmidfacial morphogenesis in mice. Dev. Dyn. 235 (5), 1448–1454.

Li, L., Meng, T., Jia, Z., Zhu, G., Shi, B., 2010. Single nucleotide polymorphism asso-ciated with nonsyndromic cleft palate influences the processing of miR-140.Am. J. Med. Genet. A 152A (4), 856–862.

Li, M., Yang, H., Zhao, J., Fang, L., Shi, H., Li, M., Sun, Y., Zhang, X., Jiang, D., Zhou, L.,et al., 2014. Efficient and heritable gene targeting in tilapia by CRISPR/Cas9.Genetics 197 (2), 591–599.

Liu, F., van der Lijn, F., Schurmann, C., Zhu, G., Chakravarty, M.M., Hysi, P.G., Woll-stein, A., Lao, O., de Bruijne, M., Ikram, M.A., et al., 2012. A genome-wide as-sociation study identifies five loci influencing facial morphology in Europeans.PLoS Genet. 8 (9), e1002932.

Loh, Y.H., Bezault, E., Muenzel, F.M., Roberts, R.B., Swofford, R., Barluenga, M., Kidd,C.E., Howe, A.E., Di Palma, F., Lindblad-Toh, K., et al., 2013. Origins of sharedgenetic variation in African cichlids. Mol. Biol. Evol. 30 (4), 906–917.

Loh, Y.H., Katz, L., Mims, M., Kocher, T.D., Yi, S.V., Streelman, J.T., 2008. Comparativeanalysis reveals signatures of differentiation amid genomic polymorphism inLake Malawi cichlids. Genome Biol. 8 (7), R113.

Long, F., 2012. Building strong bones: molecular regulation of the osteoblast line-age. Nat. Rev. Mol. Cell. Biol. 13 (1), 27–38.

Mahar, B., 2009. Evolution: Biology's next top model? Nature 458, 695–698.Manolio, T.A., 2009. Cohort studies and the genetics of complex disease. Nat. Genet.

41 (1), 5–6.Manolio, T.A., Collins, F.S., Cox, N.J., Goldstein, D.B., Hindorff, L.A., Hunter, D.J.,

McCarthy, M.I., Ramos, E.M., Cardon, L.R., Chakravarti, A., et al., 2009. Findingthe missing heritability of complex diseases. Nature 461 (7265), 747–753.

Marquez Hernandez, R.A., Ohtani, J., Fujita, T., Sunagawa, H., Ishikawa, E., Tsuba-moto, N., Kawata, T., Kaku, M., Motokawa, M., Tanne, K., 2011. Mandibular andfemoral growth alteration after sex hormone disruption in growing mice. Or-thod. Craniofacial Res. 14 (2), 63–69.

Medeiros, D.M., Crump, J.G., 2012. New perspectives on pharyngeal dorsoventralpatterning in development and evolution of the vertebrate jaw. Dev. Biol. 371

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

(2), 121–135.Minchin, J.E., Hughes, S.M., 2008. Sequential actions of Pax3 and Pax7 drive xan-

thophore development in zebrafish neural crest. Dev. Biol. 317 (2), 508–522.Mo, R., Freer, A.M., Zinyk, D.L., Crackower, M.A., Michaud, J., Heng, H.H., Chik, K.W.,

Shi, X.M., Tsui, L.C., Cheng, S.H., et al., 1997. Specific and redundant functions ofGli2 and Gli3 zinc finger genes in skeletal patterning and development. De-velopment 124 (1), 113–123.

Murray, J.C., Schutte, B.C., 2004. Cleft palate: players, pathways, and pursuits. J. Clin.Invest 113 (12), 1676–1678.

Nadeau, J.H., 2001. Modifier genes in mice and humans. Nat. Rev. Genet. 2 (3),165–174.

Nadeau, N.J., Jiggins, C.D., 2010. A golden age for evolutionary genetics? Genomicstudies of adaptation in natural populations. Trends Genet. 26 (11), 484–492.

Nguyen, A.M., Jacobs, C.R., 2013. Emerging role of primary cilia as mechanosensorsin osteocytes. Bone 54 (2), 196–204.

Nguyen, D., Xu, T., 2008. The expanding role of mouse genetics for understandinghuman biology and disease. Dis. Model. Mech. 1 (1), 56–66.

Nicks, K.M., Fowler, T.W., Gaddy, D., 2010. Reproductive hormones and bone. Curr.Osteoporos. Rep. 8 (2), 60–67.

Paaby, A.B., Rockman, M.V., 2014. Cryptic genetic variation: evolution's hiddensubstrate. Nat. Rev. Genet. 15 (4), 247–258.

Papachroni, K.K., Karatzas, D.N., Papavassiliou, K.A., Basdra, E.K., Papavassiliou, A.G.,2009. Mechanotransduction in osteoblast regulation and bone disease. TrendsMol. Med. 15 (5), 208–216.

Parsons, K.J., Albertson, R.C., 2013. Unifying and generalizing the two strands ofevo-devo. Trends Ecol. Evol. 28 (10), 584–591.

Parsons, K.J., Marquez, E., Albertson, R.C., 2012. Constraint and opportunity: thegenetic basis and evolution of modularity in the cichlid mandible. Am. Nat. 179(1), 64–78.

Parsons, K.J., Trent Taylor, A., Powder, K.E., Albertson, R.C., 2014. Wnt signallingunderlies the evolution of new phenotypes and craniofacial variability in LakeMalawi cichlids. Nat. Commun. 5, 3629.

Parsons, K.J., Wang, J., Anderson, G., Albertson, R.C., 2015. Nested levels of adaptivedivergence: the genetic basis of craniofacial divergence and ecological sexualdimorphism. G3 5 (8), 1613–1624.

Paternoster, L., Zhurov, A.I., Toma, A.M., Kemp St, J.P., Pourcain, B., Timpson, N.J.,McMahon, G., McArdle, W., Ring, S.M., Smith, G.D., et al., 2012. Genome-wideassociation study of three-dimensional facial morphology identifies a variant inPAX3 associated with nasion position. Am. J. Hum. Genet. 90 (3), 478–485.

Patton, E.E., Zon, L.I., 2001. The art and design of genetic screens: zebrafish. Nat.Rev. Genet. 2 (12), 956–966.

Powder, K.E., Cousin, H., McLinden, G.P., Craig Albertson, R., 2014. A nonsynon-ymous mutation in the transcriptional regulator lbh is associated with cichlidcraniofacial adaptation and neural crest cell development. Mol. Biol. Evol. 31(12), 3113–3124.

Powder, K.E., Milch, K., Asselin, G., Albertson, R.C., 2015. Constraint and diversifi-cation of developmental trajectories in cichlid facial morphologies. Evodevo 6,25.

Rattanasopha, S., Tongkobpetch, S., Srichomthong, C., Siriwan, P., Suphapeetiporn,K., Shotelersuk, V., 2012. PDGFRa mutations in humans with isolated cleft pa-late. Eur. J. Hum. Genet. 20 (10), 1058–1062.

Reardon, W., Winter, R.M., Rutland, P., Pulleyn, L.J., Jones, B.M., Malcolm, S., 1994.Mutations in the fibroblast growth factor receptor 2 gene cause Crouzon syn-drome. Nat. Genet. 8 (1), 98–103.

Roberts, R.B., Hu, Y., Albertson, R.C., Kocher, T.D., 2011. Craniofacial divergence andongoing adaptation via the hedgehog pathway. Proc. Natl. Acad. Sci. USA 108(32), 13194–13199.

Robling, A.G., Turner, C.H., 2009. Mechanical signaling for bone modeling and re-modeling. Crit. Rev. Eukaryot. Gene Expr. 19 (4), 319–338.

Salzburger, W., Meyer, A., 2004. The species flocks of East African cichlid fishes:recent advances in molecular phylogenetics and population genetics. Nat-urwissenschaften 91 (6), 277–290.

Santagati, F., Rijli, F.M., 2003. Cranial neural crest and the building of the vertebratehead. Nat. Rev. Neurosci. 4 (10), 806–818.

Sauka-Spengler, T., Bronner-Fraser, M., 2008. A gene regulatory network orches-trates neural crest formation. Nat. Rev. Mol. Cell. Biol. 9 (7), 557–568.

Schartl, M., 2014. Beyond the zebrafish: diverse fish species for modeling humandisease. Dis. Model. Mech. 7 (2), 181–192.

Schilling, T.F., 1997. Genetic analysis of craniofacial development in the vertebrateembryo. Bioessays 19 (6), 459–468.

Schneider, R.F., Li, Y., Meyer, A., Gunter, H.M., 2014. Regulatory gene networks thatshape the development of adaptive phenotypic plasticity in a cichlid fish. Mol.Ecol. 23 (18), 4511–4526.

Sherwood, R.J., Duren, D.L., Mahaney, M.C., Blangero, J., Dyer, T.D., Cole, S.A., Czer-winski, S.A., Chumlea, W.C., Siervogel, R.M., Choh, A.C., et al., 2011. A genome-wide linkage scan for quantitative trait loci influencing the craniofacial complexin humans (Homo sapiens sapiens). Anat. Rec. 294 (4), 664–675.

Shubin, N., Tabin, C., Carroll, S., 2009. Deep homology and the origins of evolu-tionary novelty. Nature 457 (7231), 818–823.

Song, L., Li, Y., Wang, K., Wang, Y.Z., Molotkov, A., Gao, L., Zhao, T., Yamagami, T.,Wang, Y., Gan, Q., et al., 2009. Lrp6-mediated canonical Wnt signaling is re-quired for lip formation and fusion. Development 136 (18), 3161–3171.

Stern, D.L., 2000. Evolutionary developmental biology and the problem of variation.Evolution 54 (4), 1079–1091.

Sternberg, S.H., Doudna, J.A., 2015. Expanding the Biologist's Toolkit with CRISPR-Cas9. Mol. Cell. 58 (4), 568–574.

a model to understand normal and clinical craniofacial variation.

K.E. Powder, R.C. Albertson / Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 9

Stewart, T.A., Albertson, R.C., 2010. Evolution of a unique predatory feeding appa-ratus: functional anatomy, development and a genetic locus for jaw laterality inLake Tanganyika scale-eating cichlids. BMC Biol. 8, 8.

Streelman, J.T., Albertson, R.C., 2006. Evolution of novelty in the cichlid dentition. J.Exp. Zool. B Mol. Dev. Evol. 306 (3), 216–226.

Sturmbauer, C., Baric, S., Salzburger, W., Ruber, L., Verheyen, E., 2001. Lake levelfluctuations synchronize genetic divergences of cichlid fishes in African lakes.Mol. Biol. Evol. 18 (2), 144–154.

Swartz, M.E., Sheehan-Rooney, K., Dixon, M.J., Eberhart, J.K., 2011. Examination of apalatogenic gene program in zebrafish. Dev. Dyn. 240 (9), 2204–2220.

Swiderski, D.L., Zelditch, M.L., 2013. The complex ontogenetic trajectory of man-dibular shape in a laboratory mouse. J. Anat. 223 (6), 568–580.

Sylvester, J.B., Rich, C.A., Loh, Y.H., van Staaden, M.J., Fraser, G.J., Streelman, J.T.,2010. Brain diversity evolves via differences in patterning. Proc. Natl. Acad. Sci.USA 107 (21), 9718–9723.

Sylvester, J.B., Rich, C.A., Yi, C., Peres, J.N., Houart, C., Streelman, J.T., 2013. Com-peting signals drive telencephalon diversity. Nat. Commun. 4, 1745.

Turner, G.F., Seehausen, O., Knight, M.E., Allender, C.J., Robinson, R.L., 2001. Howmany species of cichlid fishes are there in African lakes? Mol. Ecol. 10 (3),793–806.

Verdonck, A., De Ridder, L., Kuhn, R., Darras, V., Carels, C., de Zegher, F., 1998. Effectof testosterone replacement after neonatal castration on craniofacial growth in

Please cite this article as: Powder, K.E., Albertson, R.C., Cichlid fishes asDev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.12.018i

rats. Arch. Oral. Biol. 43 (7), 551–557.Wada, N., Javidan, Y., Nelson, S., Carney, T.J., Kelsh, R.N., Schilling, T.F., 2005.

Hedgehog signaling is required for cranial neural crest morphogenesis andchondrogenesis at the midline in the zebrafish skull. Development 132 (17),3977–3988.

Wagner, G.P., 2014. Homology, Genes, and Evolutionary Innovation. PrincetonUniversity Press, Princeton, NJ.

Wilkie, A.O., Slaney, S.F., Oldridge, M., Poole, M.D., Ashworth, G.J., Hockley, A.D.,Hayward, R.D., David, D.J., Pulleyn, L.J., Rutland, P., et al., 1995. Apert syndromeresults from localized mutations of FGFR2 and is allelic with Crouzon syn-drome. Nat. Genet. 9 (2), 165–172.

Won, Y.J., Sivasundar, A., Wang, Y., Hey, J., 2005. On the origin of Lake Malawi ci-chlid species: a population genetic analysis of divergence. Proc. Natl. Acad. Sci.USA 102 (Suppl. 1), 6581–6586.

Wray, G.A., 2007. The evolutionary significance of cis-regulatory mutations. Nat.Rev. Genet. 8 (3), 206–216.

Yang, J., Andre, P., Ye, L., Yang, Y.Z., 2015. The Hedgehog signalling pathway in boneformation. Int. J. Oral. Sci. 7 (2), 73–79.

Young, N.M., Chong, H.J., Hu, D., Hallgrimsson, B., Marcucio, R.S., 2010. Quantitativeanalyses link modulation of sonic hedgehog signaling to continuous variationin facial growth and shape. Development 137 (20), 3405–3409.

a model to understand normal and clinical craniofacial variation.