Comparative Analysis of Epidermal Differentiation Genes of Crocodilians Suggests New Models for the Evolutionary Origin of Avian Feather Proteins

Comparative Analysis of Epidermal Differentiation Genes of Crocodilians Suggests New Models for... The epidermis of amniotes forms a protective barrier against the environment and the differentiation program of keratino- cytes, the main cell type in the epidermis, has undergone specific alterations in the course of adaptation of amniotes to a broad variety of environments and lifestyles. The epidermal differentiation complex (EDC) is a cluster of genes expressed at late stages of keratinocyte differentiation in both sauropsids and mammals. In the present study, we identified and analyzed the crocodilian equivalent of the EDC. The gene complement of the EDC of both the American alligator and the saltwater crocodile were determined by comparative genomics, de novo gene prediction and identification of EDC transcripts in published transcriptome data. We found that crocodilians have an organization of the EDC similar to that of their closest living relatives, the birds, with which they form the clade Archosauria. Notable differences include the specific expansion of a subfamily of EDC genes in crocodilians and the loss of distinct ancestral EDC genes in birds. Identification and comparative analysis of crocodilian orthologs of avian feather proteins suggest that the latter evolved by cooption and sequence mod- ification of ancestral EDC genes, and that the amplification of an internal highly cysteine-enriched amino acid sequence motif gave rise to the feather component epidermal differentiation cysteine-rich protein in the avian lineage. Thus, sequence diversification of EDC genes contributed to the evolutionary divergence of the crocodilian and avian integuments. Key words: crocodiles, alligators, comparative genomics, integument, skin, feathers. Introduction phenotypes of their integuments have diverged significantly. Crocodilians are a clade of semiaquatic, predatory reptiles Birds have evolved feathers and beaks and only the legs are comprising 24 species of which 14 belong to the family covered with scales, whereas crocodilians have an “armored” Crocodylidae, 8 to Alligatoridae, and 2 to Gavialidae (Li skin consisting of epidermal scales, in many cases located on et al. 2007; http://www.reptile-database.org;last accessed top of dermal bony plates (osteoderms). In contrast to the September 30, 2017). The phylogenetically closest extant rel- body scales which typically develop from placodes (Musser atives of crocodilians are the birds. These two groups consti- et al. 2015; Di-Poı¨ and Milinkovitch 2016), recent studies tute the clade Archosauria and their last common ancestor have indicated that the scales present on the head of croco- lived 219–255 Ma (Shen et al. 2011; Chiari et al. 2012). After dilians are formed by physical cracking (Milinkovitch et al. the evolutionary split between crocodilians and birds, the 2013). The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 694 Genome Biol. Evol. 10(2):694–704. doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE Knapp et al. 1993; Alibardi and Toni 2006; Toni et al. 2007; Greenwold and Sawyer 2010). CBPs are conserved in reptiles and birds but are absent in mammals. Apart from the pres- ence of these proteins, very little is known about the molec- ular composition of the epidermal layers in crocodilians. The sequencing of genomes of representatives from all major clades of amniotes has facilitated a major advance in the elucidation of genes that determine the structure and function of the epidermis in mammals, reptiles and birds. In particular, it has become clear that a gene cluster originally defined as the Epidermal Differentiation Complex (EDC) in mammals is shared, with clade-specific modifications, among all amniotes investigated so far. The organization of the EDC has been previously defined for birds, lizards, snakes, and turtles (Strasser et al. 2014, 2015; Holthaus et al. 2016, 2017). EDC genes were named according to a previously established nomenclature system (Strasser et al. 2014), in which gene names begin with “Epidermal Differentiation (ED)” and, in the second part, describe either the amino acid composition or the presence of particular amino acid sequence motifs of the encoded proteins. Several individual genes within the EDC of sauropsids are expressed in differentiated epidermal keratinocytes, for exam- ple, loricrin (LOR) inlizardscales(Strasseretal. 2014), scaffol- din (SCFN) in avian claws and feathers (Strasser et al. 2015), Epidermal Differentiation Cysteine-Rich Protein (EDCRP)in feathers (Strasser et al. 2015) and Epidermal Differentiation Protein starting with a MTF motif and rich in Histidine (EDMTFH)in feathers (Alibardi et al. 2016). Moreover, there FIG.1.—Terminally differentiated epidermal keratinocytes form the barrier to the environment in crocodiles. Thin sections of the skin of a is accumulating evidence for an evolutionary origin of CBP crocodile (Crocodylus moreletii) were stained with hematoxylin and eosin genes within an ancestral EDC of sauropsids (Strasser et al. (A and B). Scale bars: 100mm(A), 20 mm(B). Epidermal keratinocyte dif- 2014). CBP genes are present in the EDCs of all sauropsids ferentiation is depicted schematically (C). Epidermal differentiation is char- investigated so far, with additional CBP genes of birds and acterized by the expression of specific differentiation-associated proteins turtles being present at loci outside of the EDC as a conse- many of which are encoded in the epidermal differentiation complex quence of gene amplification and translocation events (Ng (EDC). The final step of differentiation is cornification during which the et al. 2014; Holthaus et al. 2016). In crocodilians, CBPs have nucleus is degraded and structural proteins are cross-linked. been characterized at the gene and protein levels (Sawyer In histology, crocodile scales display several layers of living et al. 2000, 2003; Alibardi and Toni 2007; Toni et al. 2007; keratinocytes and a thick cornified layer (fig. 1). The latter Dalla Valle et al. 2009; Ye et al. 2010; Greenwold and Sawyer comprises multiple sublayers of cornified, enucleate keratino- 2013). Immuno-crossreactivity studies have suggested that cytes and resembles the stratum corneum of mammalian epi- crocodilian homologs of avian feather CBPs are expressed in dermis, although the stratum corneum of crocodiles is the subperiderm, an embryonic layer of the epidermis that is generally much thicker and remains relatively compact during shared only between crocodiles and birds (Alibardi and histological sectioning (Alibardi 2011)(fig. 1). In the terminol- Thompson 2002; Sawyer et al. 2003; Alibardi et al. 2006), ogy of herpetological skin research, the “hard” outer com- suggesting that there are evolutionary-developmental links partment of scales is referred to as the beta-layer, and the between the subperiderm and feathers (Sawyer and Knapp inner compartment of scales and the “soft” hinge regions 2003; Sawyer et al. 2003, 2005; Alibardi et al. 2006). For localized between scales are called alpha-layer of the epider- crocodilians, only CBP genes and a few isolated EDC genes, mis (Baden and Maderson 1970; Maderson 1985; Landmann identified by single gene BLAST searches (Mlitz et al. 2014), 1986; Alibardi 2005, 2011; Alibardi and Toni 2006). The have been described so far. quantitatively predominant proteins of the reptilian beta- In this study, we have identified the entire sets of EDC layer and of epidermal appendages such as claws are genes present in 2 species of crocodilians, and we have com- the corneous beta proteins (CBPs), traditionally called pared the organization of the crocodilian EDC with that of beta-keratins (Gregg and Rogers 1986; Presland et al. 1989; birds. This study has implications on the genetic control of cell Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 695 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE differentiation in crocodilian epidermis and on the evolution- breeding program. Skin tissue was sampled through a biopsy ary origin of feather genes in birds. at the flank of a 3-year-old female Morelet’s crocodile in agreement with the national laws regulating animal welfare and the guidelines of Good Veterinary Practice. For histolog- Materials and Methods ical investigation, the sample was fixed with 7.5% formalde- Genome Sequences and Gene Identification hyde, embedded in paraffin, thin-sectioned and stained with hematoxylin and eosin according to a published protocol Genes were predicted in the genome sequences of the fol- (Mlitz et al. 2014). lowing two crocodilian species (St John et al. 2012; Green et al. 2014): the American alligator (Alligator mississippiensis) and the saltwater crocodile (Crocodylus porosus). Accession RT-PCR numbers of genome sequence scaffolds corresponding to the A part of the scaled skin tissue of C. moreletii was used to EDC can be found in supplementary tables S2 and S3, prepare RNA according to a published protocol (Mlitz et al. Supplementary Material online. To predict the coding sequen- 2014; Strasser et al. 2014). The RNA was reverse-transcribed ces of EDC genes, the amino acid sequences of EDC proteins to cDNA which was subsequently amplified by PCRs with of chicken (Gallus gallus)and turtle (Chrysemys picta bellii), primers designed in C. porosus for genes of the EDC (supple- already identified in previous studies (Strasser et al. 2014; mentary table S4, Supplementary Material online). The PCR Holthaus et al. 2016), were used as queries. These queries products were purified and sequenced. The nucleotide were then used in tBLASTn searches against the nucleotide sequences of cDNAs were submitted to GenBank sequence positioned between the genes (S100A12 and (Accession numbers MG243696, MG243697, MG243698). S100A11) bordering the EDC in the investigated species. Genes coding for functional proteins were included in com- Results parative studies whereas pseudogenes belonging to EDC gene families with functional members in the same species were Identification of the Epidermal Differentiation Complex in not investigated further. In case of EDC regions with seemingly Crocodilians low gene density, the nucleotide sequence of the region in To identify the EDC gene complement of the American question was translated in silico and additional open reading alligator (Alligator mississippiensis) and the saltwater croc- frames of candidate EDC genes were identified by our already odile (Crocodylus porosus), sequences of EDC-encoded published protocol (Strasser et al. 2014). In the NCBI browser proteins of chicken and turtle (Chrysemys picta bellii) for “genomic regions, transcripts, and products” (https:// were used as queries in tBLASTn searches of the crocodil- www.ncbi.nlm.nih.gov/gene/; last accessed September 29, ian genomes (St John et al. 2012; Green et al. 2014)with 2017) information about exon coverage by RNA-seq reads a focus on the region flanked by S100A genes (fig. 2,see was consulted to check for transcribed regions in the EDC below for a detailed description of organization and gene of A. mississippiensis. Nucleotide sequences of transcribed content of the crocodilian EDC). In addition, EDC genes regions were translated in all reading frames and the possible within the latter region were predicted de novo according translation products were compared with known EDC pro- to the approach described for other sauropsids (Strasser teins of other amniotes. To test for expression of predicted et al. 2014; Holthaus et al. 2016, 2017). The existence of crocodilian EDC genes, we performed tBLASTn searches in the RNA-sequencing (RNA-seq) reads matching EDC genes of transcriptome of A. mississippiensis (St John et al. 2012; Green the American alligator (supplementary table S2, et al. 2014). The default parameters for tBLASTn searches at Supplementary Material online) validated most of the pre- the NCBI browser were used, whereby the filter for low com- dictions. Crocodilian EDC genes were named according to plexity regions in sequences was deactivated. the previously established nomenclature system (Strasser et al. 2014). Full names are listed in supplementary table Bioinformatic Analysis of Amino Acid Sequences Encoded S1, Supplementary Material online, whereas only abbre- by EDC Genes viated names are used in the text to simplify reading. When comparing our EDC predictions with the current Amino acid sequences were aligned using the program annotation of the crocodilian genome assemblies MultAlin (Corpet 1988). The ProtParam software tool at the (October 2017), we found that some EDC genes were ExPASy portal (Artimo et al. 2012) was used to calculate correctly annotated, whereas others were missed by the amino acid percentages. automatic algorithms and some gene annotations in- cluded splice sites and reading frames that were not plau- Animal Tissue and Histology sible (fig. 3A). Apparently, the short open reading frames Morelet’s crocodile (Crocodylus moreletii) was kept at the and the low sequence complexity of EDC genes did not Vienna Zoo, Vienna, Austria as part of an international allow the automatic algorithm to identify the coding 696 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE FIG.2.—Structure of the epidermal differentiation complex (EDC) in crocodilians. The genes of the EDC complex of the saltwater crocodile (Crocodylus porosus) and the American alligator (Alligator mississippiensis) are compared with those of a bird (chicken, Gallus gallus), aturtle(Chrysemys picta), a snake (representing squamate reptiles, Burmese python, Python bivittatus), and humans. A schematic depiction of the locus is given with the orientation of genes indicated by arrows. Simple EDC (SEDC) genes containing a single coding exon are shown as colored arrows with a black frame while other genes are shown as filled arrows. Gene clusters of more than four members are represented as boxes over which the number of genes is indicated after the symbol #. In the EDC of the snake, arrows labeled by # represent groups of genes. Orthology is shown by black vertical lines connecting genes or gene families. Discontinuity in the genome assembly, where the EDC was tentatively reconstructed through interspecies comparison, is indicated with the symbol . Note that the schemes are not drawn to scale. EDbeta genes belong to the corneous beta protein (CBP)/beta-keratin genes. In this schematic, the names of the chicken genes EDSC and EDCH5 (Strasser et al. 2014)were changed to EDQM3 and EDPQ, respectively, to indicate orthologies. CBP, corneous beta-protein (also known as beta-keratin); w, pseudogene; SFTP, S100 fused-type protein. regions of many crocodilian EDC genes. By contrast, our The EDC of Crocodilians Is Syntenic with the EDCs of Birds approach facilitated predictions of both Simple EDC and Turtles (SEDC) genes in which the coding region is confined to The overall organization of the crocodilian EDC is syn- one exon (fig. 3C) and S100 fused-type protein (SFTP) tenic with the EDC organization of birds (Archosauria) genes in which the coding sequence is present on 2 exons and also with that of the EDC in turtles (fig. 2). Only 2 (fig. 3D) (supplementary figs. S1–S3 and tables S2 and S3, SFTP genes (CRNN and SCFN) are present close to one Supplementary Material online). A single 5 -noncoding end of the EDC, and SEDC genes form the main part of exon characteristic for both SEDC (supplementary fig. the EDC in crocodilians. Comparative analysis suggests S3, Supplementary Material online) and SFTP genes was that a cluster of crocodilian CBP (beta-keratin) genes is identified for some but not all genes based on intron span- syntenic with CBPs in the EDC of other sauropsids (fig. 2). ning RNA-seq reads (fig. 3B and supplementary table S2, Two EDC genes, EDCH and EDDM, that were previously Supplementary Material online). The relative arrangement identified in birds but not in other sauropsids, have of crocodilian genome sequence scaffolds containing EDC orthologs in crocodilians (fig. 2), suggesting an evolu- genes (supplementary tables S2 and S3, Supplementary tionary origin in an ancestral archosaur. Some other fea- Material online) was predicted by alignment to the orthol- tures of the EDC, such as the presence of genes encoding ogous regions of the EDC in other amniotes (fig. 2). proline-rich proteins, are shared between crocodilians, Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 697 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE FIG.4.—RNA sequencing data suggest skin-specific expression of EDC genes in the alligator. To determine the tissue expression pattern of FIG.3.—Identification of genes in the EDC of crocodilians. (A) representative EDC genes of Alligator mississippiensis in a semiquantitative Representative view (GenBank accession number NW_017707560.1) in manner, we screened the RNA-seq data deposited in the GenBank sequence the genomic data browser (NCBI GenBank) for Alligator mississippiensis read archive (SRA). The whole amino acid sequences of the selected EDC indicating annotations, RNA-seq coverage (blue peaks), and predictions proteins were used as queries in tBLASTn searches. Transcriptome data from made in this study. (B) Alignment of the nucleotide sequence of the juvenile belly skin, liver, and heart (St John et al. 2012) were investigated. The EDCH5 gene predicted in our study with a RNA-seq read from the se- RNA-seq reads yielding 100% identical matches to the query sequence were quence read archive (SRA) (accession number SRR3208124.3838205.1, counted and plotted on the graph. Transcripts of the ubiquitously expressed determined in experiment SRX1616862). Nucleotides of coding and non- gene 5 -aminolevulinate synthase 1 (Alas1) were counted as a positive con- coding regions are indicated by red and gray fonts, respectively. The TATA trol for the analysis of each transcriptome. Accession numbers of transcrip- box is highlighted in green, the splice sites in blue and the start of trans- tome data: SRX1616878 (juvenile heart), SRX1616880 (juvenile liver), and lation in yellow. The amino acid sequence of the translated product is SRX1616862 (juvenile belly skin). shown underneath the nucleotide sequence. The analysis of EDCH6,cor- responding to LOC106738316 in panel A, is shown in supplementary figure S3, Supplementary Material online. (C and D) Schematic depiction Proteins Encoded by Crocodilian EDC Genes Are Enriched of the exon–intron structures and the positions of coding sequences (CDS) for a Small Set of Amino Acids and Sequence Repeats in SEDC (simple EDC) and S100 fused-type protein (SFTP) genes. Similar to their orthologs in other amniotes, EDC proteins birds, turtles, and squamates whereas other genes are of crocodilians contain high amounts of one or more of unique to the crocodilians (see fig. 2 for details). Yet other the amino acids glycine (G), serine (S), proline (P), lysine genes, such as EDKM, EDWM, EDQL, EDYM1, EDP3,and (K), cysteine (C), and glutamine (Q) (fig. 5). In many EDC SCFN, were confirmed to be conserved in all major saurop- proteins, the percentage of one of these amino acids sid clades, and PGLYRP3, LOR,and CRNN of crocodilians exceeds20%, asisthe case for glycine in EDQM1-2, lor- have orthologs in mammals (Strasser et al. 2014; Holthaus icrin and EDbeta1; serine in EDQM1 and loricrin; cysteine et al. 2016, 2017)(fig. 2). in EDCRPL1-3, EDPCV and several EDCH proteins; proline in EDPL, EDP2, EDPE, EDPQ1-4 and EDPCV, and glutamine in EDPQ2 and 4. Particularly, striking are the proline con- RNA-Seq Data Suggest That EDC Genes Are Expressed in tents 40% in EDPQ proteins (fig. 5). the Skin but Not in Internal Organs Many EDC proteins of crocodilians exhibit sequence To explore the tissue expression pattern of crocodilian EDC repeats rich in the above-mentioned amino acids (supple- genes, we compared the number of RNA-seq reads corre- mentary fig. S4, Supplementary Material online). Another sponding to specific EDC genes in the tissue transcriptomes characteristic of EDC proteins of crocodilians and other of the American alligator (St John et al. 2012). BLAST hits amniotes are amino- and carboxy-terminal sequence matching to EDWM, EDCH25, Beta2 (corneous beta-protein motifs containing lysine and glutamine (supplementary 2), EDPE,and EDQA (for full names, see supplementary table fig. S5, Supplementary Material online), corresponding S1, Supplementary Material online) were obtained at high to sites of Ne-(c-glutamyl)lysin protein cross-linking via numbers in the skin but not in the heart or liver (fig. 4). The transglutamination in mammalian EDC proteins (Candi housekeeping gene ALAS1 (Beer et al. 2015), used as a con- et al. 2005). Thus, the amino acid sequences of EDC pro- trol, showed a similar level of expression in all three tissues teins of crocodilians are compatible with roles as struc- (fig. 4). tural components of cornified keratinocytes. 698 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE FIG.5.—Proteins encoded by EDC genes of the American alligator are enriched for a subset of amino acid residues. The amino acid (aa) compositions (% of total aa residues) of SEDC proteins of the Alligator mississippiensis are shown. The order of the protein data corresponds to the order of their respective genes on the EDC (fig. 2). Note that not all the translation products of the two large gene clusters, that is, corneous beta proteins (CBP) and EDCH, are included here: Data for the proteins encoded by the first and the last gene of the CBP cluster and 11 out of 25 EDCH proteins are shown. The letter “p” at the end of protein names indicates that only partial sequences were available for analysis. respectively. In a previous study, two EDPQ genes were also Comparative Analysis of Crocodilian and Avian EDCs identifiedinthe turtle C. picta (Holthaus et al. 2016). By con- Suggests Lineage-Specific Gene Alterations trast, representative bird species such as chicken and ostrich Epidermal Differentiation proteins containing Cysteine do not have EDPQ genes (fig. 7). These data suggest that the Histidine motifs (EDCHs), which have a unique organization primordial EDPQ gene originated in a common ancestor of with an amino-terminal domain rich in cysteine and histidine turtles and archosaurs but was later lost in the bird lineage residues and a carboxy-terminal domain rich in cysteine and (fig. 7). Likewise, the ancestral EDP2 gene appeared to have proline residues (fig. 6A), were previously identified in birds undergone inactivation in birds (supplementary fig. S7, while no orthologs for these proteins were found in turtles Supplementary Material online). and lepidosaurs (Strasser et al. 2014; Holthaus et al. 2016, 2017). In the present study, we identified EDCH genes in both the alligator and crocodile EDC (fig. 2). The expres- The Crocodilian EDC Comprises Orthologs of Avian sion of a representative EDCH gene (EDCH25) in the skin Feather Protein Genes of the American alligator was confirmed by transcriptome data (fig. 4). By reverse-transcriptase polymerase chain re- The EDC of the chicken comprises at least three types of action (RT-PCR), we could also detect the expression of an genes that encode feather proteins: feather-CBPs (also called EDCH gene (presumably orthologous to EDCH10 of the beta-keratins) (Greenwold and Sawyer 2010; Ng et al. 2014; saltwater crocodile) in the scale skin of Morelet’s crocodile Wu et al. 2015), EDMTFH (also called histidine-rich protein, (Belize crocodile) (C. moreletii)(supplementary table S4 HRP) (Strasser et al. 2014; Alibardi et al. 2016), and EDCRP and fig. S6, Supplementary Material online). (Strasser et al. 2014, 2015). Previous studies have shown that Interestingly, 17–25 EDCH genes are present in crocodili- feather-CBPs evolved after the divergence of the avian from ans with at least 13 (C. porosus) and 21 (A. mississippien- the crocodilian lineage (Dalla Valle et al. 2009; Greenwold sis) having an intact and complete open reading frame and Sawyer 2011; Greenwold and Sawyer 2013). The char- (supplementary figs. S1 and S2, Supplementary Material acterization of crocodilian EDC genes allowed us to study the online) whereas only three to five copies are present in origin of the two other known feather proteins. birds. Considering the lengths of time between the diver- Chicken EDMTFH is characterized by a high histidine con- gence of the evolutionary lineages (fig. 6B), it appears tent which is not conserved in orthologous proteins of numer- likely that the number of EDCH genes was low in the last ous other birds (Alibardi et al. 2016). EDMTFH belongs to the common ancestor of Archosauria and increased specifi- avian EDMTF proteins which have amino acid sequences sim- cally during the evolution of crocodilians. ilar to those of crocodilian EDAA proteins (supplementary fig. Two and four EDPQ (Epidermal Differentiation proteins rich S8, Supplementary Material online). The chicken EDMTF in proline [P] and glutamine [Q]) genes were found in the genes form a subcluster within the EDC sharing synteny EDCs of the saltwater crocodile and the American alligator, with the EDAA clusters of turtles and crocodilians (fig. 2). Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 699 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE FIG.6.—The number of EDCH proteins has expanded in crocodilians. (A) Amino acid sequences of EDCH (Epidermal Differentiation protein containing cysteine histidine motifs) proteins were aligned based on the translation of EDCH genes identified in the EDCs of the chicken, the American alligator, and the saltwater crocodile. Histidine is highlighted in magenta, other colors are as indicated in figure 5.(B) Simplified model of EDCH evolution. The schematic diagram depicts a hypothesis about the change in the number of EDCH genes over time. Specific rates of gene duplication and pseudogenization were not estimated. TimeTree estimates of evolutionary divergence times are indicated (Kumar et al. 2017). Mya, million years ago. Thus, the apparent homology of avian EDMTFs, including the the dash indicates that there are repeats lacking this amino feather protein EDMTFH (Alibardi et al. 2016), and crocodilian acid sequence position (fig. 8). This sequence is absent in EDAAs suggests that, in the avian lineage, one or more an- crocodilian EDCRP-like proteins but present and amplified cestral EDAA genes were coopted for functions in feathers. (with some deviations in its carboxy-terminal residues, indi- Shared synteny and high similarity of amino acid sequences cated by X in the alignment in fig. 8A)up to >50-fold in the in the amino- and carboxy-terminal domains of the encoded central part of EDCRPs of different clades of birds (Strasser proteins identify EDCRP-like genes in crocodilians as orthologs et al. 2015)(fig. 8B). We conclude that the evolution of the of avian EDCRP (fig. 8). The orthologous EDCRP-type proteins feather protein EDCRP represents an intramolecular structural of crocodilians and birds differ by the absence or presence of innovation that has occurred in the avian lineage after its di- the repeated sequence motif CCDPCQ(K/-)(T/P)(V/-), whereby vergence from the crocodilian lineage. 700 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE FIG.7.—EDPQ is conserved in crocodilians but it has been lost in birds. (A) Schematic phylogenetic tree of the EDPQ locus in two representive bird and crocodilian species and turtles as an outgroup. Conservation of flanking genes (EDP3, EDPCV1/EDQCM, and EDDM) confirmed correct identification of the locus of species indicated. (B) Alignment of turtle and crocodile EDPQ amino acid sequences. The abbreviations of species names are shown in panel (A). among reptiles. We identified 22 complete CBP genes (in- Discussion cluding EDbetas) in A. mississippiensis and C. porosus and Implications on Epidermal Differentiation in Archosaurs 6 partial CBP genes in the C. porosus genome (supplemen- The results of the present study show, for the first time, the tary figs. S1B and S2B, Supplementary Material online), presence of an EDC in crocodilians and, by comparing its or- whereas lepidosaurs have at least 35 CBPs (Holthaus ganization with that of other amniotes, help to build a model et al. 2017), and turtles have >70 CBPs (Li et al. 2013; for the evolution of epidermal differentiation in early archo- Holthaus et al. 2016). Perhaps, the relatively small number saurs. Crocodilians are the last major clade of amniotes of CBPs is related to the limited variety of types and tex- (Strasseretal. 2014; Holthaus et al. 2016, 2017)for which tures of the scales in crocodilians (Pough et al. 2001). the structure of the EDC is reported. Therefore, the results of Comparative studies of other epidermis-associated genes the present study complete the draft inventory of diversified yielded only limited information on crocodilians (Dalla EDC structures of amniotes. Valle et al. 2011; Abbas Zadeh et al. 2017). Thus, the iden- Previous studies of crocodilian epidermal proteins had tification of multiple crocodilian EDC genes represents a focused primarily on highly abundant CBPs/beta-keratins significant advancement in the characterization of epider- (Sawyer et al. 2000; Alibardi and Toni 2007; Toni et al. mal differentiation in crocodilians. 2007; Dalla Valle et al. 2009; Ye et al. 2010; Greenwold Based on shared synteny, we propose a draft organiza- and Sawyer 2013). In agreement with previous gene tion of the EDC in the American alligator and the saltwater (Greenwold and Sawyer, 2013) and proteome (Alibardi crocodile. There is a high degree of similarity in the EDCs of and Toni 2007; Toni et al. 2007) analyses, we conclude both species, suggesting that many features have been that the number of CBP genes in crocodilians is the lowest inherited from their last common ancestor and therefore Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 701 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE FIG.8.—The avian feather protein EDCRP has evolved by the origin and expansion of an internal sequence repeat. (A) The amino acid sequences of avian Epidermal Differentiation Cysteine-Rich Protein (EDCRP) and crocodilian EDCRP-like (EDCRPL) proteins were aligned. An internal sequence repeat is present multiple times (as indicated) with variable carboxy-terminal amino acid residues (indicated by X) in birds, whereas it is absent in EDCRPLs of crocodiles. The ˆ symbols below the sequence alignment indicate positions of amino acid residues that were conserved in one or more representatives of both crocodilians and birds. Species names and the number of the respective EDCRPL isoform of the alligator are indicated in front of the sequences. EDCRP sequences of birds were published previously (Strasser et al. 2015). Species: chicken (Gallus gallus), duck (Anas platyrhynchos), pigeon (Columba livia), penguin (Pygoscelis adeliae), egret (Egretta garzetta), falcon (Falco cherrug), zebra finch (Taeniopygia guttata), ostrich (Struthio camelus), alligator (Alligator mississippiensis), and crocodile (Crocodylus porosus). (B) Schematic model of EDCRP evolution. The presence and domain organization of EDCRP-like proteins were mapped onto a schematic phylogenetic tree of archosaurs with turtles as outgroup. The origin of the internal sequence repeat motif is indicated by an asterisk. EDCRP is a component of feathers in modern birds. from the last common ancestor of all crocodilians. In line Comparison of alligator and crocodile versus chicken with this notion, a preliminary investigation of the genome shows that there is a high degree of conservation of EDC of the Indian gharial (Gavialis gangeticus)(Green et al. genes in archosaurs. Only few SEDC genes of crocodilians 2014) indicated conservation of amino acid sequences lack an ortholog in birds and vice versa. However, there are encoded by EDC genes and syntenic organization of the also important differences which have likely contributed to EDC (supplementary fig. S9, Supplementary Material on- the divergent evolution of skin phenotypes in crocodilians line, and data not shown). The current genome assemblies and birds. In fact, the epidermal stratum corneum of crocodi- of crocodilians are compatible with a continuous EDC like lians evolved into a particularly mechano-resistant and water- in the human genome. However, due to gaps in these proof component of the skin, and the beak and particularly sequence assemblies, especially in the genome of the the feathers of birds represent unique evolutionary innova- American alligator, discontinuities in the arrangement of tions that depended on modifications of the epidermal differ- EDC genes can also not be fully excluded. A rearrange- entiation process. ment of the EDC was reported for the opossum (Vanhoutteghem et al. 2008), and discrepancies exist be- Divergent Evolution of Crocodilian and Avian EDC Genes tween the chicken EDC models in the Gallus_gallus-4.0 assembly (Bellott et al. 2010; Strasser et al. 2014; Our data suggest that the EDC underwent only few changes Holthaus et al. 2016) which was used in the present study, in gene composition and arrangement in the crocodilian lin- and the Gallus_gallus-5.0 assembly (Warren et al. 2017). eage after its split from the avian lineage. These changes were As the expression of EDC genes may depend on their lo- inferred from our comparison of the EDCs of crocodilians with cation in topologically associating domains (TADs) the EDC of birds and turtles. Both the saltwater crocodile and (Poterlowicz et al. 2017), further detailed characterization the American alligator have at least 4 times as many EDCH of the EDC gene loci in crocodilians and other amniotes genes as birds, suggesting that the EDCH gene cluster ex- may yield important insights into the regulation of gene panded in the stem lineage of crocodilians (fig. 6). Other expression during epidermal differentiation. alterations of the EDC affected single genes (leading to 702 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE different numbers of EDCRPL and EDPQ genes in the saltwa- the evolutionary divergence between the crocodilian and ter crocodile vs. the American alligator, fig. 2) and have likely avian skin. occurred after the divergence of sublineages within crocodi- lians (Pough et al. 2001). Supplementary Material Importantly, the results of this study also suggest a scenario Supplementary data are available at Genome Biology and for the evolution of EDC genes in the stem lineage of birds. Evolution online. We put forward the hypothesis that the evolution of feather proteins involved the cooption of genes encoding CBP/beta- keratins, EDAAs (termed EDMTFs in birds), and a precursor of Competing Financial Interests EDCRP. The cooption of epidermal structural proteins may be The authors declare no competing financial interests. a common theme in the evolution of feathers and hair (Eckhart et al. 2008; Wagner 2014). It remains to be deter- mined which modifications in the amino acid sequences of Acknowledgments CBP/beta-keratins and EDAAs contributed to their functions as structural proteins of feathers. For the feather gene EDCRP, We thank Veronika Mlitz, Florian Ehrlich and Bahar Golabi for we identified orthologs in crocodilians (figs. 2 and 8), whereas helpful discussions and technical advice. This work was sup- no orthologs are present in turtles (Holthaus et al. 2016). The ported by the Austrian Science Fund (FWF) (grant numbers characteristic feature of EDCRP is its uniquely high number of P23801, P28004). cysteine residues which, in analogy to the numerous cysteine residues of mammalian hair proteins, have been proposed to Literature Cited serve as sites of intermolecular cross-linking via disulfide Abbas Zadeh S, et al. 2017. Phylogenetic profiling and gene expression bonds (Strasser et al. 2014). The crocodilian orthologs of studies implicate PSORS1C2 in terminal differentiation of keratino- EDCRP have a cysteine content of >20%, which is above cytes. Exp Dermatol. 26(4):352–358. the average cysteine content of proteins encoded within the Alibardi L. 2005. Keratinization in crocodilian scales and avian epidermis: evolutionary implications for the origin of avian apteric epidermis. Belg EDC. In the evolutionary lineage leading to birds, a cysteine- J Zool. 135:9–20. rich sequence motif appeared probably by duplication and Alibardi L. 2011. Histology, ultrastructure, and pigmentation in the horny mutation of a neighboring sequence in the ancestral gene, scales of growing crocodilians. Acta Zool. 92(2):187–200. and this sequence motif was again manifold amplified. Alibardi L, et al. 2016. Immunolocalization of a histidine-rich Epidermal Consequently, the length of EDCRP increased in the avian Differentiation Protein in the chicken supports the hypothesis of an evolutionary developmental link between the embryonic subperiderm lineage, leading a much higher total number of cysteine res- and feather barbs and barbules. PLoS One 11(12):e0167789. idues per molecule than those present in its crocodilian coun- Alibardi L, Knapp LW, Sawyer RH. 2006. Beta-keratin localization in de- terparts (160 cysteine residues in chicken EDCRP vs. a veloping alligator scales and feathers in relation to the development maximum of 22 cysteine residues in crocodilian EDCRPLs). and evolution of feathers. J Subm Cytol Pathol. 38:175–192. Proteomic analysis of cornified feathers has confirmed the Alibardi L, Thompson MB. 2002. Keratinization and ultrastructure of the epidermis of late embryonic stages in the alligator (Alligator mississip- integration of EDCRP into the permanent parts of feathers piensis). J Anat. 201(1):71–84. (Strasseretal. 2014, 2015). Thus, the intramolecular modifi- Alibardi L, Toni M. 2006. Cytochemical, biochemical and molecular cation of an EDCRP-like protein has likely contributed, to- aspects of the process of keratinization in the epidermis of reptilian gether with sequence adaptations of EDAA/EDMTF proteins scales. Prog Histochem Cytochem. 40(2):173–134. and CBP/feather beta-keratins, to the evolution of the heavily Alibardi L, Toni M. 2007. Characterization of keratins and other proteins involved in the corneification of crocodilian epidermis. Tissue Cell cross-linked protein architecture of feathers. Remarkably, 39(5):311–323. feather CBPs/beta-keratins, EDMTFH, and EDCRP of the Artimo P, et al. 2012. ExPASy: sIB bioinformatics resource portal. Nucleic chicken were found to be expressed not only in feathers Acids Res. 40(Web Server issue):W597–W603. but also in the embryonic subperiderm (Strasser et al. 2015; Baden HP, Maderson PF. 1970. Morphological and biophysical identifica- Alibardi et al. 2016), a layer which has a homolog in crocodi- tion of fibrous proteins in the amniote epidermis. J Exp Zool. 174(2):225–232. lians (Alibardi and Thompson 2002; Sawyer et al. 2003). This Beer L, et al. 2015. Bioinformatics approach for choosing the correct ref- leads us to hypothesize that the primordial expression site of erence genes when studying gene expression in human keratinocytes. feather protein precursors was the subperiderm, which is evo- Exp Dermatol. 24(10):742–747. lutionarily older than feathers. Thus, it will be very interesting Bellott DW, et al. 2010. Convergent evolution of chicken Z and human X to determine the expression of crocodilian EDC genes not only chromosomes by expansion and gene acquisition. Nature 466(7306):612–616. in adult skin but also in the epidermis of embryos. Candi E, Schmidt R, Melino G. 2005. The cornified envelope: a model of In conclusion, the results of the present comparative geno- cell death in the skin. Nat Rev Mol Cell Biol. 6(4):328–340. mic analysis of the EDC in archosaurs provide a basis for study- Chiari Y, Cahais V, Galtier N, Delsuc F. 2012. Phylogenomic analyses sup- ing the specific roles of genes involved in epidermal port the position of turtles as sister group of birds and crocodiles. BMC differentiation in crocodilians and allow to further delineate Biol. 10:65. Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 703 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE Corpet F. 1988. Multiple sequence alignment with hierarchical clustering. Mlitz V, et al. 2014. Trichohyalin-like proteins have evolutionarily con- Nucleic Acids Res. 16(22):10881–10890. served roles in the morphogenesis of skin appendages. J Invest Dalla Valle L, et al. 2009. Beta-keratins of the crocodilian epidermis: com- Dermatol. 134:2682–2692. position, structure, and phylogenetic relationships. J Exp Zool B Mol Musser JM, Wagner GP, Prum RO. 2015. Nuclear b-catenin localization Dev Evol. 312(1):42–57. supports homology of feathers, avian scutate scales, and alligator Dalla Valle L, et al. 2011. Deleterious mutations of a claw keratin in mul- scales in early development. Evol Dev. 17(3):185–194. tiple taxa of reptiles. J Mol Evol. 72(3):265–273. Ng CS, et al. 2014. Genomic organization, transcriptomic analysis, and Di-Poı¨ N, Milinkovitch MC. 2016. The anatomical placode in reptile scale functional characterization of avian alpha- and beta-keratins in diverse morphogenesis indicates shared ancestry among skin appendages in feather forms. Genome Biol Evol. 6(9):2258–2273. amniotes. Sci Adv. 2:e1600708. Poterlowicz K, et al. 2017. 5C analysis of the Epidermal Differentiation Eckhart L, et al. 2008. Identification of reptilian genes encoding Complex locus reveals distinct chromatin interaction networks be- hair keratin-like proteins suggests a new scenario for the evolu- tween gene-rich and gene-poor TADs in skin epithelial cells. PloS tionary origin of hair. Proc Natl Acad Sci U S A. 105(47): Genet. 13(9):e1006966. 18419–18423. Pough FH, et al. 2001. Herpetology. 2nd ed.. Upper Saddle River (NJ): Green RE, et al. 2014. Three crocodilian genomes reveal ancestral patterns Prentice-Hall. of evolution among archosaurs. Science 346(6215):1254449. Presland RB, et al. 1989. Avian keratin genes, I. A molecular analysis of the Greenwold MJ, Sawyer RH. 2010. Genomic organization and molecular structure and expression of a group of feather keratin genes. J Mol phylogenies of the beta (beta) keratin multigene family in the chicken Biol. 209:549–560. (Gallus gallus) and zebra finch (Taeniopygia guttata): implications for Sawyer RH, et al. 2000. The expression of beta (b) keratins in the epidermal feather evolution. BMC Evol Biol. 10:148. appendages of reptiles and birds. Am Zool. 40(4):530–539. Greenwold MJ, Sawyer RH. 2011. Linking the molecular evolution of avian Sawyer RH, et al. 2003. Origin of feathers: feather beta (b)keratins are beta keratins to the evolution of feathers. J Exp Zool B Mol Dev Evol. expressed in discrete epidermal cell populations of embryonic scutate 316(8):609–616. scales. J Exp Zool B Mol Dev Evol. 295B(1):12–24. Greenwold MJ, Sawyer RH. 2013. Expression of archosaurian beta- Sawyer RH, Knapp LW. 2003. Avian skin development and the evolution- keratins: diversification and expansion of archosaurian beta-keratins ary origin of feathers. J Exp Zool B Mol Dev Evol. 298B(1):57–72. and the origin of feather beta-keratins. J Exp Zool B Mol Dev Evol. Sawyer RH, Rogers L, Washington L, Glenn TC, Knapp LW. 2005. 320(6):393–405. Evolutionary origin of the feather epidermis. Dev Dyn. Gregg K, Rogers G. 1986. Feather keratins: composition, structure and 232(2):256–267. biogenesis. In: Bereiter-Hahn J, Matoltsy AG, Sylvia-Richards K, editors. Shen XX, Liang D, Wen JZ, Zhang P. 2011. Multiple genome alignments Biology of the integument, Vol. 2 Vertebrates. Berlin-Heidelberg: facilitate development of NPCL markers: a case study of tetrapod Springer. p. 666–694. phylogeny focusing on the position of turtles. Mol Biol Evol. Holthaus KB, et al. 2016. Comparative genomics identifies epidermal pro- 28(12):3237–3252. teins associated with the evolution of the turtle shell. Mol Biol Evol. St John JA, et al. 2012. Sequencing three crocodilian genomes to illuminate 33(3):726–737. the evolution of archosaurs and amniotes. Genome Biol. 13(1):415. Holthaus KB, et al. 2017. Identification and comparative analysis of the Strasser B, et al. 2014. Evolutionary origin and diversification of epidermal epidermal differentiation complex in snakes. Sci Rep. 7:45338. barrier proteins in amniotes. Mol Biol Evol. 31(12):3194–3205. Knapp LW, Shames RB, Barnes GL, Sawyer RH. 1993. Region-specific Strasser B, Mlitz V, Hermann M, Tschachler E, Eckhart L. 2015. Convergent patterns of beta keratin expression during avian skin development. evolution of cysteine-rich proteins in feathers and hair. BMC Evol Biol. Dev Dyn. 196(4):283–290. 15:82. Kumar S, Stecher G, Suleski M, Hedges SB. 2017. TimeTree: a resource for Toni M, Dalla Valle L, Alibardi L. 2007. Review. Beta-keratins in the epi- timelines, timetrees, and divergence times. Mol Biol Evol. 34(7): dermis of reptiles: composition, sequence and molecular organization. 1812–1819. J Proteome Res. 6(9):3377–3392. Landmann L. 1986. The skin of reptiles: epidermis and dermis. In: Vanhoutteghem A, Djian P, Green H. 2008. Ancient origin of the gene Bereiter-Hahn J, Matoltsy AG, Sylvia-Richards K, editors. Biology encoding involucrin, a precursor of the cross-linked envelope of epider- of the integument, Vol. 2 Vertebrates. Berlin-Heidelberg: mis and related epithelia. Proc Natl Acad Sci U S A. Springer. p. 150–187. 105(40):15481–15486. Li Y, Wu X, Ji X, Yan P, Amato G. 2007. The complete mitochondrial Wagner GP. 2014. Homology, genes and evolutionary innovation. genome of salt-water crocodile (Crocodylus porosus) and phylogeny Princeton (NJ): Princeton University Press. of crocodilians. J Genet Genomics 34(2):119–128. Warren WC, et al. 2017. A New Chicken Genome Assembly provides Li YI, Kong L, Ponting CP, Haerty W. 2013. Rapid evolution of Beta- insight into avian genome structure. G3 (Bethesda) 7(1):109–117. keratin genes contribute to phenotypic differences that distin- Wu P, et al. 2015. Topographical mapping of a-and b-keratins on devel- guish turtles and birds from other reptiles. Genome Biol Evol. oping chicken skin integument: functional interaction and evolution- 5(5):923–933. ary perspectives. Proc Natl Acad Sci U S A. 112(49):E6770–E6779. Maderson PF. 1985. Some developmental problems of the reptilian Ye C, Wu X, Yan P, Amato G. 2010. Beta-keratins in crocodiles reveal integument. In: Gans C, Billett F, Maderson PFA, editors. Biology amino acid homology with avian keratins. Mol Biol Rep. of the reptilia, Vol 14. New York: John Wiley and Sons. p. 37(3):1169–1174. 525–598. Milinkovitch MC, et al. 2013. Crocodile head scales are not develop- mental units but emerge from physical cracking. Science 339(6115):78–81. Associate editor: Dan Graur 704 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Genome Biology and Evolution Oxford University Press

Comparative Analysis of Epidermal Differentiation Genes of Crocodilians Suggests New Models for the Evolutionary Origin of Avian Feather Proteins

Free
11 pages

Loading next page...
 
/lp/ou_press/comparative-analysis-of-epidermal-differentiation-genes-of-Vx026sUddK
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution.
ISSN
1759-6653
eISSN
1759-6653
D.O.I.
10.1093/gbe/evy035
Publisher site
See Article on Publisher Site

Abstract

The epidermis of amniotes forms a protective barrier against the environment and the differentiation program of keratino- cytes, the main cell type in the epidermis, has undergone specific alterations in the course of adaptation of amniotes to a broad variety of environments and lifestyles. The epidermal differentiation complex (EDC) is a cluster of genes expressed at late stages of keratinocyte differentiation in both sauropsids and mammals. In the present study, we identified and analyzed the crocodilian equivalent of the EDC. The gene complement of the EDC of both the American alligator and the saltwater crocodile were determined by comparative genomics, de novo gene prediction and identification of EDC transcripts in published transcriptome data. We found that crocodilians have an organization of the EDC similar to that of their closest living relatives, the birds, with which they form the clade Archosauria. Notable differences include the specific expansion of a subfamily of EDC genes in crocodilians and the loss of distinct ancestral EDC genes in birds. Identification and comparative analysis of crocodilian orthologs of avian feather proteins suggest that the latter evolved by cooption and sequence mod- ification of ancestral EDC genes, and that the amplification of an internal highly cysteine-enriched amino acid sequence motif gave rise to the feather component epidermal differentiation cysteine-rich protein in the avian lineage. Thus, sequence diversification of EDC genes contributed to the evolutionary divergence of the crocodilian and avian integuments. Key words: crocodiles, alligators, comparative genomics, integument, skin, feathers. Introduction phenotypes of their integuments have diverged significantly. Crocodilians are a clade of semiaquatic, predatory reptiles Birds have evolved feathers and beaks and only the legs are comprising 24 species of which 14 belong to the family covered with scales, whereas crocodilians have an “armored” Crocodylidae, 8 to Alligatoridae, and 2 to Gavialidae (Li skin consisting of epidermal scales, in many cases located on et al. 2007; http://www.reptile-database.org;last accessed top of dermal bony plates (osteoderms). In contrast to the September 30, 2017). The phylogenetically closest extant rel- body scales which typically develop from placodes (Musser atives of crocodilians are the birds. These two groups consti- et al. 2015; Di-Poı¨ and Milinkovitch 2016), recent studies tute the clade Archosauria and their last common ancestor have indicated that the scales present on the head of croco- lived 219–255 Ma (Shen et al. 2011; Chiari et al. 2012). After dilians are formed by physical cracking (Milinkovitch et al. the evolutionary split between crocodilians and birds, the 2013). The Author(s) 2018. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 694 Genome Biol. Evol. 10(2):694–704. doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE Knapp et al. 1993; Alibardi and Toni 2006; Toni et al. 2007; Greenwold and Sawyer 2010). CBPs are conserved in reptiles and birds but are absent in mammals. Apart from the pres- ence of these proteins, very little is known about the molec- ular composition of the epidermal layers in crocodilians. The sequencing of genomes of representatives from all major clades of amniotes has facilitated a major advance in the elucidation of genes that determine the structure and function of the epidermis in mammals, reptiles and birds. In particular, it has become clear that a gene cluster originally defined as the Epidermal Differentiation Complex (EDC) in mammals is shared, with clade-specific modifications, among all amniotes investigated so far. The organization of the EDC has been previously defined for birds, lizards, snakes, and turtles (Strasser et al. 2014, 2015; Holthaus et al. 2016, 2017). EDC genes were named according to a previously established nomenclature system (Strasser et al. 2014), in which gene names begin with “Epidermal Differentiation (ED)” and, in the second part, describe either the amino acid composition or the presence of particular amino acid sequence motifs of the encoded proteins. Several individual genes within the EDC of sauropsids are expressed in differentiated epidermal keratinocytes, for exam- ple, loricrin (LOR) inlizardscales(Strasseretal. 2014), scaffol- din (SCFN) in avian claws and feathers (Strasser et al. 2015), Epidermal Differentiation Cysteine-Rich Protein (EDCRP)in feathers (Strasser et al. 2015) and Epidermal Differentiation Protein starting with a MTF motif and rich in Histidine (EDMTFH)in feathers (Alibardi et al. 2016). Moreover, there FIG.1.—Terminally differentiated epidermal keratinocytes form the barrier to the environment in crocodiles. Thin sections of the skin of a is accumulating evidence for an evolutionary origin of CBP crocodile (Crocodylus moreletii) were stained with hematoxylin and eosin genes within an ancestral EDC of sauropsids (Strasser et al. (A and B). Scale bars: 100mm(A), 20 mm(B). Epidermal keratinocyte dif- 2014). CBP genes are present in the EDCs of all sauropsids ferentiation is depicted schematically (C). Epidermal differentiation is char- investigated so far, with additional CBP genes of birds and acterized by the expression of specific differentiation-associated proteins turtles being present at loci outside of the EDC as a conse- many of which are encoded in the epidermal differentiation complex quence of gene amplification and translocation events (Ng (EDC). The final step of differentiation is cornification during which the et al. 2014; Holthaus et al. 2016). In crocodilians, CBPs have nucleus is degraded and structural proteins are cross-linked. been characterized at the gene and protein levels (Sawyer In histology, crocodile scales display several layers of living et al. 2000, 2003; Alibardi and Toni 2007; Toni et al. 2007; keratinocytes and a thick cornified layer (fig. 1). The latter Dalla Valle et al. 2009; Ye et al. 2010; Greenwold and Sawyer comprises multiple sublayers of cornified, enucleate keratino- 2013). Immuno-crossreactivity studies have suggested that cytes and resembles the stratum corneum of mammalian epi- crocodilian homologs of avian feather CBPs are expressed in dermis, although the stratum corneum of crocodiles is the subperiderm, an embryonic layer of the epidermis that is generally much thicker and remains relatively compact during shared only between crocodiles and birds (Alibardi and histological sectioning (Alibardi 2011)(fig. 1). In the terminol- Thompson 2002; Sawyer et al. 2003; Alibardi et al. 2006), ogy of herpetological skin research, the “hard” outer com- suggesting that there are evolutionary-developmental links partment of scales is referred to as the beta-layer, and the between the subperiderm and feathers (Sawyer and Knapp inner compartment of scales and the “soft” hinge regions 2003; Sawyer et al. 2003, 2005; Alibardi et al. 2006). For localized between scales are called alpha-layer of the epider- crocodilians, only CBP genes and a few isolated EDC genes, mis (Baden and Maderson 1970; Maderson 1985; Landmann identified by single gene BLAST searches (Mlitz et al. 2014), 1986; Alibardi 2005, 2011; Alibardi and Toni 2006). The have been described so far. quantitatively predominant proteins of the reptilian beta- In this study, we have identified the entire sets of EDC layer and of epidermal appendages such as claws are genes present in 2 species of crocodilians, and we have com- the corneous beta proteins (CBPs), traditionally called pared the organization of the crocodilian EDC with that of beta-keratins (Gregg and Rogers 1986; Presland et al. 1989; birds. This study has implications on the genetic control of cell Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 695 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE differentiation in crocodilian epidermis and on the evolution- breeding program. Skin tissue was sampled through a biopsy ary origin of feather genes in birds. at the flank of a 3-year-old female Morelet’s crocodile in agreement with the national laws regulating animal welfare and the guidelines of Good Veterinary Practice. For histolog- Materials and Methods ical investigation, the sample was fixed with 7.5% formalde- Genome Sequences and Gene Identification hyde, embedded in paraffin, thin-sectioned and stained with hematoxylin and eosin according to a published protocol Genes were predicted in the genome sequences of the fol- (Mlitz et al. 2014). lowing two crocodilian species (St John et al. 2012; Green et al. 2014): the American alligator (Alligator mississippiensis) and the saltwater crocodile (Crocodylus porosus). Accession RT-PCR numbers of genome sequence scaffolds corresponding to the A part of the scaled skin tissue of C. moreletii was used to EDC can be found in supplementary tables S2 and S3, prepare RNA according to a published protocol (Mlitz et al. Supplementary Material online. To predict the coding sequen- 2014; Strasser et al. 2014). The RNA was reverse-transcribed ces of EDC genes, the amino acid sequences of EDC proteins to cDNA which was subsequently amplified by PCRs with of chicken (Gallus gallus)and turtle (Chrysemys picta bellii), primers designed in C. porosus for genes of the EDC (supple- already identified in previous studies (Strasser et al. 2014; mentary table S4, Supplementary Material online). The PCR Holthaus et al. 2016), were used as queries. These queries products were purified and sequenced. The nucleotide were then used in tBLASTn searches against the nucleotide sequences of cDNAs were submitted to GenBank sequence positioned between the genes (S100A12 and (Accession numbers MG243696, MG243697, MG243698). S100A11) bordering the EDC in the investigated species. Genes coding for functional proteins were included in com- Results parative studies whereas pseudogenes belonging to EDC gene families with functional members in the same species were Identification of the Epidermal Differentiation Complex in not investigated further. In case of EDC regions with seemingly Crocodilians low gene density, the nucleotide sequence of the region in To identify the EDC gene complement of the American question was translated in silico and additional open reading alligator (Alligator mississippiensis) and the saltwater croc- frames of candidate EDC genes were identified by our already odile (Crocodylus porosus), sequences of EDC-encoded published protocol (Strasser et al. 2014). In the NCBI browser proteins of chicken and turtle (Chrysemys picta bellii) for “genomic regions, transcripts, and products” (https:// were used as queries in tBLASTn searches of the crocodil- www.ncbi.nlm.nih.gov/gene/; last accessed September 29, ian genomes (St John et al. 2012; Green et al. 2014)with 2017) information about exon coverage by RNA-seq reads a focus on the region flanked by S100A genes (fig. 2,see was consulted to check for transcribed regions in the EDC below for a detailed description of organization and gene of A. mississippiensis. Nucleotide sequences of transcribed content of the crocodilian EDC). In addition, EDC genes regions were translated in all reading frames and the possible within the latter region were predicted de novo according translation products were compared with known EDC pro- to the approach described for other sauropsids (Strasser teins of other amniotes. To test for expression of predicted et al. 2014; Holthaus et al. 2016, 2017). The existence of crocodilian EDC genes, we performed tBLASTn searches in the RNA-sequencing (RNA-seq) reads matching EDC genes of transcriptome of A. mississippiensis (St John et al. 2012; Green the American alligator (supplementary table S2, et al. 2014). The default parameters for tBLASTn searches at Supplementary Material online) validated most of the pre- the NCBI browser were used, whereby the filter for low com- dictions. Crocodilian EDC genes were named according to plexity regions in sequences was deactivated. the previously established nomenclature system (Strasser et al. 2014). Full names are listed in supplementary table Bioinformatic Analysis of Amino Acid Sequences Encoded S1, Supplementary Material online, whereas only abbre- by EDC Genes viated names are used in the text to simplify reading. When comparing our EDC predictions with the current Amino acid sequences were aligned using the program annotation of the crocodilian genome assemblies MultAlin (Corpet 1988). The ProtParam software tool at the (October 2017), we found that some EDC genes were ExPASy portal (Artimo et al. 2012) was used to calculate correctly annotated, whereas others were missed by the amino acid percentages. automatic algorithms and some gene annotations in- cluded splice sites and reading frames that were not plau- Animal Tissue and Histology sible (fig. 3A). Apparently, the short open reading frames Morelet’s crocodile (Crocodylus moreletii) was kept at the and the low sequence complexity of EDC genes did not Vienna Zoo, Vienna, Austria as part of an international allow the automatic algorithm to identify the coding 696 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE FIG.2.—Structure of the epidermal differentiation complex (EDC) in crocodilians. The genes of the EDC complex of the saltwater crocodile (Crocodylus porosus) and the American alligator (Alligator mississippiensis) are compared with those of a bird (chicken, Gallus gallus), aturtle(Chrysemys picta), a snake (representing squamate reptiles, Burmese python, Python bivittatus), and humans. A schematic depiction of the locus is given with the orientation of genes indicated by arrows. Simple EDC (SEDC) genes containing a single coding exon are shown as colored arrows with a black frame while other genes are shown as filled arrows. Gene clusters of more than four members are represented as boxes over which the number of genes is indicated after the symbol #. In the EDC of the snake, arrows labeled by # represent groups of genes. Orthology is shown by black vertical lines connecting genes or gene families. Discontinuity in the genome assembly, where the EDC was tentatively reconstructed through interspecies comparison, is indicated with the symbol . Note that the schemes are not drawn to scale. EDbeta genes belong to the corneous beta protein (CBP)/beta-keratin genes. In this schematic, the names of the chicken genes EDSC and EDCH5 (Strasser et al. 2014)were changed to EDQM3 and EDPQ, respectively, to indicate orthologies. CBP, corneous beta-protein (also known as beta-keratin); w, pseudogene; SFTP, S100 fused-type protein. regions of many crocodilian EDC genes. By contrast, our The EDC of Crocodilians Is Syntenic with the EDCs of Birds approach facilitated predictions of both Simple EDC and Turtles (SEDC) genes in which the coding region is confined to The overall organization of the crocodilian EDC is syn- one exon (fig. 3C) and S100 fused-type protein (SFTP) tenic with the EDC organization of birds (Archosauria) genes in which the coding sequence is present on 2 exons and also with that of the EDC in turtles (fig. 2). Only 2 (fig. 3D) (supplementary figs. S1–S3 and tables S2 and S3, SFTP genes (CRNN and SCFN) are present close to one Supplementary Material online). A single 5 -noncoding end of the EDC, and SEDC genes form the main part of exon characteristic for both SEDC (supplementary fig. the EDC in crocodilians. Comparative analysis suggests S3, Supplementary Material online) and SFTP genes was that a cluster of crocodilian CBP (beta-keratin) genes is identified for some but not all genes based on intron span- syntenic with CBPs in the EDC of other sauropsids (fig. 2). ning RNA-seq reads (fig. 3B and supplementary table S2, Two EDC genes, EDCH and EDDM, that were previously Supplementary Material online). The relative arrangement identified in birds but not in other sauropsids, have of crocodilian genome sequence scaffolds containing EDC orthologs in crocodilians (fig. 2), suggesting an evolu- genes (supplementary tables S2 and S3, Supplementary tionary origin in an ancestral archosaur. Some other fea- Material online) was predicted by alignment to the orthol- tures of the EDC, such as the presence of genes encoding ogous regions of the EDC in other amniotes (fig. 2). proline-rich proteins, are shared between crocodilians, Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 697 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE FIG.4.—RNA sequencing data suggest skin-specific expression of EDC genes in the alligator. To determine the tissue expression pattern of FIG.3.—Identification of genes in the EDC of crocodilians. (A) representative EDC genes of Alligator mississippiensis in a semiquantitative Representative view (GenBank accession number NW_017707560.1) in manner, we screened the RNA-seq data deposited in the GenBank sequence the genomic data browser (NCBI GenBank) for Alligator mississippiensis read archive (SRA). The whole amino acid sequences of the selected EDC indicating annotations, RNA-seq coverage (blue peaks), and predictions proteins were used as queries in tBLASTn searches. Transcriptome data from made in this study. (B) Alignment of the nucleotide sequence of the juvenile belly skin, liver, and heart (St John et al. 2012) were investigated. The EDCH5 gene predicted in our study with a RNA-seq read from the se- RNA-seq reads yielding 100% identical matches to the query sequence were quence read archive (SRA) (accession number SRR3208124.3838205.1, counted and plotted on the graph. Transcripts of the ubiquitously expressed determined in experiment SRX1616862). Nucleotides of coding and non- gene 5 -aminolevulinate synthase 1 (Alas1) were counted as a positive con- coding regions are indicated by red and gray fonts, respectively. The TATA trol for the analysis of each transcriptome. Accession numbers of transcrip- box is highlighted in green, the splice sites in blue and the start of trans- tome data: SRX1616878 (juvenile heart), SRX1616880 (juvenile liver), and lation in yellow. The amino acid sequence of the translated product is SRX1616862 (juvenile belly skin). shown underneath the nucleotide sequence. The analysis of EDCH6,cor- responding to LOC106738316 in panel A, is shown in supplementary figure S3, Supplementary Material online. (C and D) Schematic depiction Proteins Encoded by Crocodilian EDC Genes Are Enriched of the exon–intron structures and the positions of coding sequences (CDS) for a Small Set of Amino Acids and Sequence Repeats in SEDC (simple EDC) and S100 fused-type protein (SFTP) genes. Similar to their orthologs in other amniotes, EDC proteins birds, turtles, and squamates whereas other genes are of crocodilians contain high amounts of one or more of unique to the crocodilians (see fig. 2 for details). Yet other the amino acids glycine (G), serine (S), proline (P), lysine genes, such as EDKM, EDWM, EDQL, EDYM1, EDP3,and (K), cysteine (C), and glutamine (Q) (fig. 5). In many EDC SCFN, were confirmed to be conserved in all major saurop- proteins, the percentage of one of these amino acids sid clades, and PGLYRP3, LOR,and CRNN of crocodilians exceeds20%, asisthe case for glycine in EDQM1-2, lor- have orthologs in mammals (Strasser et al. 2014; Holthaus icrin and EDbeta1; serine in EDQM1 and loricrin; cysteine et al. 2016, 2017)(fig. 2). in EDCRPL1-3, EDPCV and several EDCH proteins; proline in EDPL, EDP2, EDPE, EDPQ1-4 and EDPCV, and glutamine in EDPQ2 and 4. Particularly, striking are the proline con- RNA-Seq Data Suggest That EDC Genes Are Expressed in tents 40% in EDPQ proteins (fig. 5). the Skin but Not in Internal Organs Many EDC proteins of crocodilians exhibit sequence To explore the tissue expression pattern of crocodilian EDC repeats rich in the above-mentioned amino acids (supple- genes, we compared the number of RNA-seq reads corre- mentary fig. S4, Supplementary Material online). Another sponding to specific EDC genes in the tissue transcriptomes characteristic of EDC proteins of crocodilians and other of the American alligator (St John et al. 2012). BLAST hits amniotes are amino- and carboxy-terminal sequence matching to EDWM, EDCH25, Beta2 (corneous beta-protein motifs containing lysine and glutamine (supplementary 2), EDPE,and EDQA (for full names, see supplementary table fig. S5, Supplementary Material online), corresponding S1, Supplementary Material online) were obtained at high to sites of Ne-(c-glutamyl)lysin protein cross-linking via numbers in the skin but not in the heart or liver (fig. 4). The transglutamination in mammalian EDC proteins (Candi housekeeping gene ALAS1 (Beer et al. 2015), used as a con- et al. 2005). Thus, the amino acid sequences of EDC pro- trol, showed a similar level of expression in all three tissues teins of crocodilians are compatible with roles as struc- (fig. 4). tural components of cornified keratinocytes. 698 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE FIG.5.—Proteins encoded by EDC genes of the American alligator are enriched for a subset of amino acid residues. The amino acid (aa) compositions (% of total aa residues) of SEDC proteins of the Alligator mississippiensis are shown. The order of the protein data corresponds to the order of their respective genes on the EDC (fig. 2). Note that not all the translation products of the two large gene clusters, that is, corneous beta proteins (CBP) and EDCH, are included here: Data for the proteins encoded by the first and the last gene of the CBP cluster and 11 out of 25 EDCH proteins are shown. The letter “p” at the end of protein names indicates that only partial sequences were available for analysis. respectively. In a previous study, two EDPQ genes were also Comparative Analysis of Crocodilian and Avian EDCs identifiedinthe turtle C. picta (Holthaus et al. 2016). By con- Suggests Lineage-Specific Gene Alterations trast, representative bird species such as chicken and ostrich Epidermal Differentiation proteins containing Cysteine do not have EDPQ genes (fig. 7). These data suggest that the Histidine motifs (EDCHs), which have a unique organization primordial EDPQ gene originated in a common ancestor of with an amino-terminal domain rich in cysteine and histidine turtles and archosaurs but was later lost in the bird lineage residues and a carboxy-terminal domain rich in cysteine and (fig. 7). Likewise, the ancestral EDP2 gene appeared to have proline residues (fig. 6A), were previously identified in birds undergone inactivation in birds (supplementary fig. S7, while no orthologs for these proteins were found in turtles Supplementary Material online). and lepidosaurs (Strasser et al. 2014; Holthaus et al. 2016, 2017). In the present study, we identified EDCH genes in both the alligator and crocodile EDC (fig. 2). The expres- The Crocodilian EDC Comprises Orthologs of Avian sion of a representative EDCH gene (EDCH25) in the skin Feather Protein Genes of the American alligator was confirmed by transcriptome data (fig. 4). By reverse-transcriptase polymerase chain re- The EDC of the chicken comprises at least three types of action (RT-PCR), we could also detect the expression of an genes that encode feather proteins: feather-CBPs (also called EDCH gene (presumably orthologous to EDCH10 of the beta-keratins) (Greenwold and Sawyer 2010; Ng et al. 2014; saltwater crocodile) in the scale skin of Morelet’s crocodile Wu et al. 2015), EDMTFH (also called histidine-rich protein, (Belize crocodile) (C. moreletii)(supplementary table S4 HRP) (Strasser et al. 2014; Alibardi et al. 2016), and EDCRP and fig. S6, Supplementary Material online). (Strasser et al. 2014, 2015). Previous studies have shown that Interestingly, 17–25 EDCH genes are present in crocodili- feather-CBPs evolved after the divergence of the avian from ans with at least 13 (C. porosus) and 21 (A. mississippien- the crocodilian lineage (Dalla Valle et al. 2009; Greenwold sis) having an intact and complete open reading frame and Sawyer 2011; Greenwold and Sawyer 2013). The char- (supplementary figs. S1 and S2, Supplementary Material acterization of crocodilian EDC genes allowed us to study the online) whereas only three to five copies are present in origin of the two other known feather proteins. birds. Considering the lengths of time between the diver- Chicken EDMTFH is characterized by a high histidine con- gence of the evolutionary lineages (fig. 6B), it appears tent which is not conserved in orthologous proteins of numer- likely that the number of EDCH genes was low in the last ous other birds (Alibardi et al. 2016). EDMTFH belongs to the common ancestor of Archosauria and increased specifi- avian EDMTF proteins which have amino acid sequences sim- cally during the evolution of crocodilians. ilar to those of crocodilian EDAA proteins (supplementary fig. Two and four EDPQ (Epidermal Differentiation proteins rich S8, Supplementary Material online). The chicken EDMTF in proline [P] and glutamine [Q]) genes were found in the genes form a subcluster within the EDC sharing synteny EDCs of the saltwater crocodile and the American alligator, with the EDAA clusters of turtles and crocodilians (fig. 2). Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 699 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE FIG.6.—The number of EDCH proteins has expanded in crocodilians. (A) Amino acid sequences of EDCH (Epidermal Differentiation protein containing cysteine histidine motifs) proteins were aligned based on the translation of EDCH genes identified in the EDCs of the chicken, the American alligator, and the saltwater crocodile. Histidine is highlighted in magenta, other colors are as indicated in figure 5.(B) Simplified model of EDCH evolution. The schematic diagram depicts a hypothesis about the change in the number of EDCH genes over time. Specific rates of gene duplication and pseudogenization were not estimated. TimeTree estimates of evolutionary divergence times are indicated (Kumar et al. 2017). Mya, million years ago. Thus, the apparent homology of avian EDMTFs, including the the dash indicates that there are repeats lacking this amino feather protein EDMTFH (Alibardi et al. 2016), and crocodilian acid sequence position (fig. 8). This sequence is absent in EDAAs suggests that, in the avian lineage, one or more an- crocodilian EDCRP-like proteins but present and amplified cestral EDAA genes were coopted for functions in feathers. (with some deviations in its carboxy-terminal residues, indi- Shared synteny and high similarity of amino acid sequences cated by X in the alignment in fig. 8A)up to >50-fold in the in the amino- and carboxy-terminal domains of the encoded central part of EDCRPs of different clades of birds (Strasser proteins identify EDCRP-like genes in crocodilians as orthologs et al. 2015)(fig. 8B). We conclude that the evolution of the of avian EDCRP (fig. 8). The orthologous EDCRP-type proteins feather protein EDCRP represents an intramolecular structural of crocodilians and birds differ by the absence or presence of innovation that has occurred in the avian lineage after its di- the repeated sequence motif CCDPCQ(K/-)(T/P)(V/-), whereby vergence from the crocodilian lineage. 700 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE FIG.7.—EDPQ is conserved in crocodilians but it has been lost in birds. (A) Schematic phylogenetic tree of the EDPQ locus in two representive bird and crocodilian species and turtles as an outgroup. Conservation of flanking genes (EDP3, EDPCV1/EDQCM, and EDDM) confirmed correct identification of the locus of species indicated. (B) Alignment of turtle and crocodile EDPQ amino acid sequences. The abbreviations of species names are shown in panel (A). among reptiles. We identified 22 complete CBP genes (in- Discussion cluding EDbetas) in A. mississippiensis and C. porosus and Implications on Epidermal Differentiation in Archosaurs 6 partial CBP genes in the C. porosus genome (supplemen- The results of the present study show, for the first time, the tary figs. S1B and S2B, Supplementary Material online), presence of an EDC in crocodilians and, by comparing its or- whereas lepidosaurs have at least 35 CBPs (Holthaus ganization with that of other amniotes, help to build a model et al. 2017), and turtles have >70 CBPs (Li et al. 2013; for the evolution of epidermal differentiation in early archo- Holthaus et al. 2016). Perhaps, the relatively small number saurs. Crocodilians are the last major clade of amniotes of CBPs is related to the limited variety of types and tex- (Strasseretal. 2014; Holthaus et al. 2016, 2017)for which tures of the scales in crocodilians (Pough et al. 2001). the structure of the EDC is reported. Therefore, the results of Comparative studies of other epidermis-associated genes the present study complete the draft inventory of diversified yielded only limited information on crocodilians (Dalla EDC structures of amniotes. Valle et al. 2011; Abbas Zadeh et al. 2017). Thus, the iden- Previous studies of crocodilian epidermal proteins had tification of multiple crocodilian EDC genes represents a focused primarily on highly abundant CBPs/beta-keratins significant advancement in the characterization of epider- (Sawyer et al. 2000; Alibardi and Toni 2007; Toni et al. mal differentiation in crocodilians. 2007; Dalla Valle et al. 2009; Ye et al. 2010; Greenwold Based on shared synteny, we propose a draft organiza- and Sawyer 2013). In agreement with previous gene tion of the EDC in the American alligator and the saltwater (Greenwold and Sawyer, 2013) and proteome (Alibardi crocodile. There is a high degree of similarity in the EDCs of and Toni 2007; Toni et al. 2007) analyses, we conclude both species, suggesting that many features have been that the number of CBP genes in crocodilians is the lowest inherited from their last common ancestor and therefore Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 701 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE FIG.8.—The avian feather protein EDCRP has evolved by the origin and expansion of an internal sequence repeat. (A) The amino acid sequences of avian Epidermal Differentiation Cysteine-Rich Protein (EDCRP) and crocodilian EDCRP-like (EDCRPL) proteins were aligned. An internal sequence repeat is present multiple times (as indicated) with variable carboxy-terminal amino acid residues (indicated by X) in birds, whereas it is absent in EDCRPLs of crocodiles. The ˆ symbols below the sequence alignment indicate positions of amino acid residues that were conserved in one or more representatives of both crocodilians and birds. Species names and the number of the respective EDCRPL isoform of the alligator are indicated in front of the sequences. EDCRP sequences of birds were published previously (Strasser et al. 2015). Species: chicken (Gallus gallus), duck (Anas platyrhynchos), pigeon (Columba livia), penguin (Pygoscelis adeliae), egret (Egretta garzetta), falcon (Falco cherrug), zebra finch (Taeniopygia guttata), ostrich (Struthio camelus), alligator (Alligator mississippiensis), and crocodile (Crocodylus porosus). (B) Schematic model of EDCRP evolution. The presence and domain organization of EDCRP-like proteins were mapped onto a schematic phylogenetic tree of archosaurs with turtles as outgroup. The origin of the internal sequence repeat motif is indicated by an asterisk. EDCRP is a component of feathers in modern birds. from the last common ancestor of all crocodilians. In line Comparison of alligator and crocodile versus chicken with this notion, a preliminary investigation of the genome shows that there is a high degree of conservation of EDC of the Indian gharial (Gavialis gangeticus)(Green et al. genes in archosaurs. Only few SEDC genes of crocodilians 2014) indicated conservation of amino acid sequences lack an ortholog in birds and vice versa. However, there are encoded by EDC genes and syntenic organization of the also important differences which have likely contributed to EDC (supplementary fig. S9, Supplementary Material on- the divergent evolution of skin phenotypes in crocodilians line, and data not shown). The current genome assemblies and birds. In fact, the epidermal stratum corneum of crocodi- of crocodilians are compatible with a continuous EDC like lians evolved into a particularly mechano-resistant and water- in the human genome. However, due to gaps in these proof component of the skin, and the beak and particularly sequence assemblies, especially in the genome of the the feathers of birds represent unique evolutionary innova- American alligator, discontinuities in the arrangement of tions that depended on modifications of the epidermal differ- EDC genes can also not be fully excluded. A rearrange- entiation process. ment of the EDC was reported for the opossum (Vanhoutteghem et al. 2008), and discrepancies exist be- Divergent Evolution of Crocodilian and Avian EDC Genes tween the chicken EDC models in the Gallus_gallus-4.0 assembly (Bellott et al. 2010; Strasser et al. 2014; Our data suggest that the EDC underwent only few changes Holthaus et al. 2016) which was used in the present study, in gene composition and arrangement in the crocodilian lin- and the Gallus_gallus-5.0 assembly (Warren et al. 2017). eage after its split from the avian lineage. These changes were As the expression of EDC genes may depend on their lo- inferred from our comparison of the EDCs of crocodilians with cation in topologically associating domains (TADs) the EDC of birds and turtles. Both the saltwater crocodile and (Poterlowicz et al. 2017), further detailed characterization the American alligator have at least 4 times as many EDCH of the EDC gene loci in crocodilians and other amniotes genes as birds, suggesting that the EDCH gene cluster ex- may yield important insights into the regulation of gene panded in the stem lineage of crocodilians (fig. 6). Other expression during epidermal differentiation. alterations of the EDC affected single genes (leading to 702 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Comparative Analysis of Epidermal Differentiation Genes of Crocodilians GBE different numbers of EDCRPL and EDPQ genes in the saltwa- the evolutionary divergence between the crocodilian and ter crocodile vs. the American alligator, fig. 2) and have likely avian skin. occurred after the divergence of sublineages within crocodi- lians (Pough et al. 2001). Supplementary Material Importantly, the results of this study also suggest a scenario Supplementary data are available at Genome Biology and for the evolution of EDC genes in the stem lineage of birds. Evolution online. We put forward the hypothesis that the evolution of feather proteins involved the cooption of genes encoding CBP/beta- keratins, EDAAs (termed EDMTFs in birds), and a precursor of Competing Financial Interests EDCRP. The cooption of epidermal structural proteins may be The authors declare no competing financial interests. a common theme in the evolution of feathers and hair (Eckhart et al. 2008; Wagner 2014). It remains to be deter- mined which modifications in the amino acid sequences of Acknowledgments CBP/beta-keratins and EDAAs contributed to their functions as structural proteins of feathers. For the feather gene EDCRP, We thank Veronika Mlitz, Florian Ehrlich and Bahar Golabi for we identified orthologs in crocodilians (figs. 2 and 8), whereas helpful discussions and technical advice. This work was sup- no orthologs are present in turtles (Holthaus et al. 2016). The ported by the Austrian Science Fund (FWF) (grant numbers characteristic feature of EDCRP is its uniquely high number of P23801, P28004). cysteine residues which, in analogy to the numerous cysteine residues of mammalian hair proteins, have been proposed to Literature Cited serve as sites of intermolecular cross-linking via disulfide Abbas Zadeh S, et al. 2017. Phylogenetic profiling and gene expression bonds (Strasser et al. 2014). The crocodilian orthologs of studies implicate PSORS1C2 in terminal differentiation of keratino- EDCRP have a cysteine content of >20%, which is above cytes. Exp Dermatol. 26(4):352–358. the average cysteine content of proteins encoded within the Alibardi L. 2005. Keratinization in crocodilian scales and avian epidermis: evolutionary implications for the origin of avian apteric epidermis. Belg EDC. In the evolutionary lineage leading to birds, a cysteine- J Zool. 135:9–20. rich sequence motif appeared probably by duplication and Alibardi L. 2011. Histology, ultrastructure, and pigmentation in the horny mutation of a neighboring sequence in the ancestral gene, scales of growing crocodilians. Acta Zool. 92(2):187–200. and this sequence motif was again manifold amplified. Alibardi L, et al. 2016. Immunolocalization of a histidine-rich Epidermal Consequently, the length of EDCRP increased in the avian Differentiation Protein in the chicken supports the hypothesis of an evolutionary developmental link between the embryonic subperiderm lineage, leading a much higher total number of cysteine res- and feather barbs and barbules. PLoS One 11(12):e0167789. idues per molecule than those present in its crocodilian coun- Alibardi L, Knapp LW, Sawyer RH. 2006. Beta-keratin localization in de- terparts (160 cysteine residues in chicken EDCRP vs. a veloping alligator scales and feathers in relation to the development maximum of 22 cysteine residues in crocodilian EDCRPLs). and evolution of feathers. J Subm Cytol Pathol. 38:175–192. Proteomic analysis of cornified feathers has confirmed the Alibardi L, Thompson MB. 2002. Keratinization and ultrastructure of the epidermis of late embryonic stages in the alligator (Alligator mississip- integration of EDCRP into the permanent parts of feathers piensis). J Anat. 201(1):71–84. (Strasseretal. 2014, 2015). Thus, the intramolecular modifi- Alibardi L, Toni M. 2006. Cytochemical, biochemical and molecular cation of an EDCRP-like protein has likely contributed, to- aspects of the process of keratinization in the epidermis of reptilian gether with sequence adaptations of EDAA/EDMTF proteins scales. Prog Histochem Cytochem. 40(2):173–134. and CBP/feather beta-keratins, to the evolution of the heavily Alibardi L, Toni M. 2007. Characterization of keratins and other proteins involved in the corneification of crocodilian epidermis. Tissue Cell cross-linked protein architecture of feathers. Remarkably, 39(5):311–323. feather CBPs/beta-keratins, EDMTFH, and EDCRP of the Artimo P, et al. 2012. ExPASy: sIB bioinformatics resource portal. Nucleic chicken were found to be expressed not only in feathers Acids Res. 40(Web Server issue):W597–W603. but also in the embryonic subperiderm (Strasser et al. 2015; Baden HP, Maderson PF. 1970. Morphological and biophysical identifica- Alibardi et al. 2016), a layer which has a homolog in crocodi- tion of fibrous proteins in the amniote epidermis. J Exp Zool. 174(2):225–232. lians (Alibardi and Thompson 2002; Sawyer et al. 2003). This Beer L, et al. 2015. Bioinformatics approach for choosing the correct ref- leads us to hypothesize that the primordial expression site of erence genes when studying gene expression in human keratinocytes. feather protein precursors was the subperiderm, which is evo- Exp Dermatol. 24(10):742–747. lutionarily older than feathers. Thus, it will be very interesting Bellott DW, et al. 2010. Convergent evolution of chicken Z and human X to determine the expression of crocodilian EDC genes not only chromosomes by expansion and gene acquisition. Nature 466(7306):612–616. in adult skin but also in the epidermis of embryos. Candi E, Schmidt R, Melino G. 2005. The cornified envelope: a model of In conclusion, the results of the present comparative geno- cell death in the skin. Nat Rev Mol Cell Biol. 6(4):328–340. mic analysis of the EDC in archosaurs provide a basis for study- Chiari Y, Cahais V, Galtier N, Delsuc F. 2012. Phylogenomic analyses sup- ing the specific roles of genes involved in epidermal port the position of turtles as sister group of birds and crocodiles. BMC differentiation in crocodilians and allow to further delineate Biol. 10:65. Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 703 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Holthaus et al. GBE Corpet F. 1988. Multiple sequence alignment with hierarchical clustering. Mlitz V, et al. 2014. Trichohyalin-like proteins have evolutionarily con- Nucleic Acids Res. 16(22):10881–10890. served roles in the morphogenesis of skin appendages. J Invest Dalla Valle L, et al. 2009. Beta-keratins of the crocodilian epidermis: com- Dermatol. 134:2682–2692. position, structure, and phylogenetic relationships. J Exp Zool B Mol Musser JM, Wagner GP, Prum RO. 2015. Nuclear b-catenin localization Dev Evol. 312(1):42–57. supports homology of feathers, avian scutate scales, and alligator Dalla Valle L, et al. 2011. Deleterious mutations of a claw keratin in mul- scales in early development. Evol Dev. 17(3):185–194. tiple taxa of reptiles. J Mol Evol. 72(3):265–273. Ng CS, et al. 2014. Genomic organization, transcriptomic analysis, and Di-Poı¨ N, Milinkovitch MC. 2016. The anatomical placode in reptile scale functional characterization of avian alpha- and beta-keratins in diverse morphogenesis indicates shared ancestry among skin appendages in feather forms. Genome Biol Evol. 6(9):2258–2273. amniotes. Sci Adv. 2:e1600708. Poterlowicz K, et al. 2017. 5C analysis of the Epidermal Differentiation Eckhart L, et al. 2008. Identification of reptilian genes encoding Complex locus reveals distinct chromatin interaction networks be- hair keratin-like proteins suggests a new scenario for the evolu- tween gene-rich and gene-poor TADs in skin epithelial cells. PloS tionary origin of hair. Proc Natl Acad Sci U S A. 105(47): Genet. 13(9):e1006966. 18419–18423. Pough FH, et al. 2001. Herpetology. 2nd ed.. Upper Saddle River (NJ): Green RE, et al. 2014. Three crocodilian genomes reveal ancestral patterns Prentice-Hall. of evolution among archosaurs. Science 346(6215):1254449. Presland RB, et al. 1989. Avian keratin genes, I. A molecular analysis of the Greenwold MJ, Sawyer RH. 2010. Genomic organization and molecular structure and expression of a group of feather keratin genes. J Mol phylogenies of the beta (beta) keratin multigene family in the chicken Biol. 209:549–560. (Gallus gallus) and zebra finch (Taeniopygia guttata): implications for Sawyer RH, et al. 2000. The expression of beta (b) keratins in the epidermal feather evolution. BMC Evol Biol. 10:148. appendages of reptiles and birds. Am Zool. 40(4):530–539. Greenwold MJ, Sawyer RH. 2011. Linking the molecular evolution of avian Sawyer RH, et al. 2003. Origin of feathers: feather beta (b)keratins are beta keratins to the evolution of feathers. J Exp Zool B Mol Dev Evol. expressed in discrete epidermal cell populations of embryonic scutate 316(8):609–616. scales. J Exp Zool B Mol Dev Evol. 295B(1):12–24. Greenwold MJ, Sawyer RH. 2013. Expression of archosaurian beta- Sawyer RH, Knapp LW. 2003. Avian skin development and the evolution- keratins: diversification and expansion of archosaurian beta-keratins ary origin of feathers. J Exp Zool B Mol Dev Evol. 298B(1):57–72. and the origin of feather beta-keratins. J Exp Zool B Mol Dev Evol. Sawyer RH, Rogers L, Washington L, Glenn TC, Knapp LW. 2005. 320(6):393–405. Evolutionary origin of the feather epidermis. Dev Dyn. Gregg K, Rogers G. 1986. Feather keratins: composition, structure and 232(2):256–267. biogenesis. In: Bereiter-Hahn J, Matoltsy AG, Sylvia-Richards K, editors. Shen XX, Liang D, Wen JZ, Zhang P. 2011. Multiple genome alignments Biology of the integument, Vol. 2 Vertebrates. Berlin-Heidelberg: facilitate development of NPCL markers: a case study of tetrapod Springer. p. 666–694. phylogeny focusing on the position of turtles. Mol Biol Evol. Holthaus KB, et al. 2016. Comparative genomics identifies epidermal pro- 28(12):3237–3252. teins associated with the evolution of the turtle shell. Mol Biol Evol. St John JA, et al. 2012. Sequencing three crocodilian genomes to illuminate 33(3):726–737. the evolution of archosaurs and amniotes. Genome Biol. 13(1):415. Holthaus KB, et al. 2017. Identification and comparative analysis of the Strasser B, et al. 2014. Evolutionary origin and diversification of epidermal epidermal differentiation complex in snakes. Sci Rep. 7:45338. barrier proteins in amniotes. Mol Biol Evol. 31(12):3194–3205. Knapp LW, Shames RB, Barnes GL, Sawyer RH. 1993. Region-specific Strasser B, Mlitz V, Hermann M, Tschachler E, Eckhart L. 2015. Convergent patterns of beta keratin expression during avian skin development. evolution of cysteine-rich proteins in feathers and hair. BMC Evol Biol. Dev Dyn. 196(4):283–290. 15:82. Kumar S, Stecher G, Suleski M, Hedges SB. 2017. TimeTree: a resource for Toni M, Dalla Valle L, Alibardi L. 2007. Review. Beta-keratins in the epi- timelines, timetrees, and divergence times. Mol Biol Evol. 34(7): dermis of reptiles: composition, sequence and molecular organization. 1812–1819. J Proteome Res. 6(9):3377–3392. Landmann L. 1986. The skin of reptiles: epidermis and dermis. In: Vanhoutteghem A, Djian P, Green H. 2008. Ancient origin of the gene Bereiter-Hahn J, Matoltsy AG, Sylvia-Richards K, editors. Biology encoding involucrin, a precursor of the cross-linked envelope of epider- of the integument, Vol. 2 Vertebrates. Berlin-Heidelberg: mis and related epithelia. Proc Natl Acad Sci U S A. Springer. p. 150–187. 105(40):15481–15486. Li Y, Wu X, Ji X, Yan P, Amato G. 2007. The complete mitochondrial Wagner GP. 2014. Homology, genes and evolutionary innovation. genome of salt-water crocodile (Crocodylus porosus) and phylogeny Princeton (NJ): Princeton University Press. of crocodilians. J Genet Genomics 34(2):119–128. Warren WC, et al. 2017. A New Chicken Genome Assembly provides Li YI, Kong L, Ponting CP, Haerty W. 2013. Rapid evolution of Beta- insight into avian genome structure. G3 (Bethesda) 7(1):109–117. keratin genes contribute to phenotypic differences that distin- Wu P, et al. 2015. Topographical mapping of a-and b-keratins on devel- guish turtles and birds from other reptiles. Genome Biol Evol. oping chicken skin integument: functional interaction and evolution- 5(5):923–933. ary perspectives. Proc Natl Acad Sci U S A. 112(49):E6770–E6779. Maderson PF. 1985. Some developmental problems of the reptilian Ye C, Wu X, Yan P, Amato G. 2010. Beta-keratins in crocodiles reveal integument. In: Gans C, Billett F, Maderson PFA, editors. Biology amino acid homology with avian keratins. Mol Biol Rep. of the reptilia, Vol 14. New York: John Wiley and Sons. p. 37(3):1169–1174. 525–598. Milinkovitch MC, et al. 2013. Crocodile head scales are not develop- mental units but emerge from physical cracking. Science 339(6115):78–81. Associate editor: Dan Graur 704 Genome Biol. Evol. 10(2):694–704 doi:10.1093/gbe/evy035 Advance Access publication February 12, 2018 Downloaded from https://academic.oup.com/gbe/article-abstract/10/2/694/4852790 by Ed 'DeepDyve' Gillespie user on 16 March 2018

Journal

Genome Biology and EvolutionOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from Google Scholar, PubMed
Create lists to organize your research
Export lists, citations
Access to DeepDyve database
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off