The Evolutionary History of and Y- Loss

George H. Perry,* à Raul Y. Tito,* and Brian C. Verrelli* *Center for Evolutionary Functional Genomics, The Biodesign Institute, Arizona State University, Tempe; School of Life Sciences, Arizona State University, Tempe; and àSchool of and Social Change, Arizona State University, Tempe

Recent studies have suggested that gene gain and loss may contribute significantly to the divergence between and . Initial comparisons of the human and chimpanzee Y- indicate that chimpanzees have a dis- proportionate loss of Y-chromosome , which may have implications for the adaptive evolution of sex-specific as well as reproductive traits, especially because one of the genes lost in chimpanzees is critically involved in spermatogenesis in humans. Here we have characterized Y-chromosome sequences in , , and several chimpanzee for 7 chimpanzee gene–disruptive mutations. Our analyses show that 6 of these gene-disruptive mutations predate chimpan- zee–bonobo divergence at ;1.8 MYA, which indicates significant Y-chromosome change in the chimpanzee lineage Downloaded from https://academic.oup.com/mbe/article/24/3/853/1246230 by guest on 23 September 2021 relatively early in the evolutionary divergence of humans and chimpanzees.

Introduction The initial comparisons of human and chimpanzee Comparative analyses of single-nucleotide differences ( troglodytes) Y-chromosome sequences revealed that between human and chimpanzee genomes typically show although there are no lineage-specific gene-disruptive mu- estimates of approximately 1–2% divergence (Watanabe tations in the X-degenerate portion of the Y-chromosome et al. 2004; Chimpanzee Sequencing and Analysis Consor- fixed within humans, surprisingly, 4 genes, CYorf15B, tium 2005). Surprisingly, more focused comparative geno- TBL1Y, TMSB4Y, and USP9Y, are disrupted by one or more mic analyses have identified greater than 100 genes with splice site or premature stop codon mutations in chimpan- lineage-specific coding sequence disruptions in the form zees (Hughes et al. 2005; Kuroki et al. 2006; Tyler-Smith of stop codon or splice site mutations, frameshift inser- et al. 2006). Given that levels of sperm competition are tion/deletions, and gene deletions (Watanabe et al. 2004; likely greater in chimpanzees than in humans (Harcourt Chimpanzee Sequencing and Analysis Consortium 2005; et al. 1981; Dorus et al. 2004) and that the Y-chromosome Newman et al. 2005; Varki and Altheide 2005; Hahn is highly enriched for genes associated with spermatogen- and Lee 2006; Wang et al. 2006). This observation is a rad- esis, the contrast between rates of human and chimpanzee ical change from previous beliefs of the types of genetic Y-chromosome gene disruption was unanticipated. Al- changes that predominantly accompanied the divergence though the specific functions of CYorf15b, TBL1Y, and of human and chimpanzee lineages and strongly implicates TMSB4Y are not well understood (Skaletsky et al. 2003; gene structure and reorganization as important means by Yan et al. 2005), USP9Y is critical for spermatogenesis which our 2 lineages have become genetically and pheno- in humans, with gene-disruptive mutations at this locus re- typically distinct. sulting in azoospermia or the absence of sperm in semen Gene loss at any particular region of the genome can (Sun et al. 1999; Blagosklonova et al. 2000). Thus, the result in many unpredicted changes in phenotype; however, potential loss of this specific gene in the chimpanzee line- lineage-specific gene loss on the Y-chromosome is of par- age is especially puzzling. ticular interest because this chromosome is highly enriched In order to better understand the evolutionary history for genes involved in spermatogenesis (Lahn and Page 1997; and significance of chimpanzee Y-chromosome gene loss, Skaletsky et al. 2003). Therefore, studies of Y-chromosome we have characterized the nucleotide sequences involving gene loss can potentially reveal the history of evolutionary the gene-disruptive mutations in each of the CYorf15B, change between human and chimpanzee mating and fertility TBL1Y, TMSB4Y, and USP9Y genes in the gorilla (Gorilla systems. Furthermore, the Y-chromosome seems to be par- gorilla), the bonobo (Pan paniscus), and in wild-born in- ticularly prone to gene loss; most of the Y-chromosome does dividuals from several chimpanzee subspecies. This analy- not undergo meiotic recombination (Tilford et al. 2001), sis enables us to make inferences about the origin and meaning that positive or negative natural selection can have timing of these gene-disruptive mutations as well as eval- very important implications for ‘‘linked’’ variation com- uate the impact of these potentially important events that pared with that of other chromosomes. As a result, there distinguish the evolutionary histories of human and chim- has been a general trend of Y-chromosome degradation panzee Y-chromosomes. and gene loss over evolutionary time (e.g., Muller’s ratchet), which may sometimes involve the fixation of gene- Materials and Methods disruptive mutations on the background of an otherwise Although comparison of the exhaustive pseudogene adaptive haplotype (Muller 1918; Charlesworth B and catalogs for humans versus chimpanzees may provide in- Charlesworth D 2000). sight into our respective evolutionary and ecological histo- ries in general, the current draft quality of the chimpanzee Key words: pseudogene, Muller’s ratchet, sperm competition, Pan troglodytes, Pan paniscus. genome sequence (Build 1.1) precludes the reliable and comprehensive genome-wide identification of chimpanzee E-mail: [email protected]. pseudogenes without further validations, as discussed by Mol. Biol. Evol. 24(3):853–859. 2007 doi:10.1093/molbev/msm002 the Chimpanzee Sequencing and Analysis Consortium Advance Access publication January 11, 2007 (2005). However, 2 independent and high-quality

Ó The Author 2007. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: [email protected] 854 Perry et al.

Table 1 Y-chromosome sequences to ensure similarity, and second, Summary of Y-Chromosome Gene-Disruptive Mutations with the homologous region on the X-chromosome to en- Similarity to sure Y-chromosome specificity. Primer sequences (all 5#–3#) Chimpanzeea used were CYorf15B forward: 5#-ACAAGTGT- # # Gene Exon Type of Mutation Gorilla Bonobo CAGCTGGTTGAGAA-3 and reverse: 5 -CAGGG- GAAAATCTGAATAAAGC-3# (fragment size 691 bp), CYorf15b 4 Stop codon No Yes # # TBL1Y 12 Splice site mutationb No Yes TBL1Y forward: 5 -TTCAACAGTTTTCTGCACTTGG-3 TMSB4Y 1 Splice site mutation No Noc and reverse: 5#-CTCAAGATGGATCAGACATTCG-3# USP9Y 8 Splice site deletion (4 bp) No Yes (917 bp), TMSB4Y forward: 5#-ACAAACCTGGTAT- 34 Splice site mutation No Yes GGCTGAGAT-3# and reverse: 5#-CCTAAACGTTCTG- 36 Frameshift deletion (4 bp) No Yes CAAGTGTACC-3# (733 bp), USP9Y exon 8 forward: 39 Splice site mutation No Yes 5#-GTTGTGTCCCCATGAACTATGA-3# and reverse: 5#- a Yes: the individual was found to have the gene-disruptive mutation as found in #

TAGCATTGTCCAAATGGTCTGA-3 (664 bp), USP9Y Downloaded from https://academic.oup.com/mbe/article/24/3/853/1246230 by guest on 23 September 2021 the 2 publicly available chimpanzee Y-chromosome sequences. No: shares the hu- exon 34 forward: 5#-AAACAATGTGCGTTTCTCCTTT-3# man sequence. # # b Mutations at exon–intron boundaries involved in intron splice site recognition. and reverse: 5 -GGTGGAAACTGAAACCATGAAT-3 c Because the bonobo did not have this mutation, this gene region was charac- (696 bp), USP9Y exon 36 forward: 5#-CATGAAATTGTTT- terized in 1 wild-born individual from each of 4 chimpanzee subspecies, all found to TAGTTTCTGTTCT-3# and reverse: 5#-CTGATGGGG- have the identical gene-disruptive mutation as in the chimpanzee reference sequences TCTTGCAATAGTT-3# (674 bp), and USP9Y exon 39 (see Results and Discussion). forward: 5#-GCAAATAAAAGCTGTTTCTGCAT-3# and reverse: 5#-GCATTCTAGAGGCACTCAAAAGA-3# (796 chimpanzee Y-chromosome sequences are publicly avail- bp). PCR reactions were performed in 25 lL reactions using able (Hughes et al. 2005; Kuroki et al. 2006), offering ex- Platinum Taq (Invitrogen, Carlsbad, California), with the fol- cellent opportunities for focused analyses on this lowing conditions: 94 °Cfor2min,followedby40cyclesof chromosome. 94 °C for 15 s, 59 °C for 30 s, and 70 °C for 30 s. We aligned the Hughes et al. (2005) and Kuroki et al. For each PCR fragment amplification, we included (2006) chimpanzee Y-chromosome sequences and the hu- DNA from both male and female human, chimpanzee, man Y-chromosome reference sequence (Build 35) for the and bonobo individuals, in order to verify that homologous CYorf15B, TBL1Y, TMSB4Y, and USP9Y genes. This align- fragments from the X-chromosome were not amplified. Fol- ment confirmed that the splice site and stop codon muta- lowing amplification, PCR products were purified with tions described by Hughes et al. (2005) are found in Shrimp Alkaline Phosphatase and Exonuclease I (USB Cor- both chimpanzee Y-chromosome sequences (and not in poration, Cleveland, Ohio), cycle sequenced with BigDye the human sequence), making it highly unlikely that these Terminator Cycle Sequencing Kit version 3.1 (Applied reflect sequencing or assembly errors. We also identified Biosystems, Foster City, California), cleaned with isopro- a 4-bp frameshift deletion in USP9Y exon 36 in the chim- panol, and analyzed by electrophoresis on an Applied Bio- panzee lineage (as inferred by comparison to human and systems 3730 capillary sequencer. Sequence data were gorilla sequences), which was not originally identified in manually aligned and analyzed using the Sequencher ver- the study conducted by Hughes et al. (2005). In total, sion 4.6 computer program (Gene Codes Corporation, Ann we characterized 7 Y-chromosome gene-disruptive muta- Arbor, Michigan). The Y-chromosome sequences gener- tions in each species (table 1). ated for this study have been deposited in GenBank with Whole blood samples from wild-born chimpanzees accession numbers EF197918–EF197935. housed at research facilities or zoological institutions were collected during regularly scheduled veterinary examina- Results and Discussion tions. DNA was isolated using a standard phenol/chloro- Origins of Chimpanzee Y-Chromosome Gene-Disruptive form extraction method (Sambrook and Russell 2001). Mutations Chimpanzee subspecies were determined from compari- sons of mitochondrial DNA and Y-chromosome sequences Table 1 displays the nucleotide sequence analysis of to those of wild-born individuals with known capture loca- each of the 7 characterized gene regions. Because the an- tion, as described by Stone et al. (2002). Bonobo, gorilla, cestral lineages separating and the common ances- and other chimpanzee DNA samples (PR00107, PR00251, tor of humans and chimpanzees likely diverged over PR00496, and PR00573) were obtained from the Integrated a relatively short period of time, ;1 Myr or less (see Biomaterials and Information Resource (IPBIR; http:// fig. 1), gene genealogies from the nuclear genome do www.ipbir.org). DNA from Clint (S006006), the captive- not consistently support the more recent common ancestor born chimpanzee who was the donor for the chimpanzee for chimpanzees and humans (Chen and Li 2001). This has genome sequence project, was obtained from the Coriell consequences for determining whether fixation events oc- Institute for Medical Research (http://www.coriell.org). curred on the human or chimpanzee lineages using the go- Tocharacterizethe7gene-disruptivemutations(table1), rilla as an outgroup sequence. However, given the smaller polymerase chain reaction (PCR) primers were designed effective population size and more recent coalescence time based on the human Y-chromosome sequence using the of the Y-chromosome (e.g., Stone et al. 2002), the gorilla Primer3 computer program (Rozen and Skaletsky 2000) nucleotide sequence can be appropriately used to polarize to amplify gene regions encompassing the 7 mutations. human and chimpanzee lineage-specific fixation events on Primer sequences were compared first with the chimpanzee this chromosome. Our analysis of the male gorilla finds Chimpanzee Y-Chromosome Gene Loss 855

gorilla gorilla and bonobo Y-chromosome could certainly address this issue at a later date. However, because of the difficulty in large-scale amplification and sequencing of these human Y-chromosome regions from genomic DNA (e.g., high X-chromosome homology), this would likely best be ac- complished with a bacterial artificial chromosome–based sequencing strategy, similar to those used by Hughes et al. (2005) and Kuroki et al. (2006) to produce their chim- panzee Y-chromosome sequences. For CYorf15B, TBL1Y, and USP9Y, the same gene- disruptive mutations present in the chimpanzee Y- chromosome sequences were also present in the bonobo se-

quence (table 1), indicating that these mutations occurred Downloaded from https://academic.oup.com/mbe/article/24/3/853/1246230 by guest on 23 September 2021 and were fixed in the common ancestor of chimpanzees bonobo and (fig. 1). There are 4 mutations that disrupt gene disruptions: the USP9Y coding region in chimpanzees, all of which were CYorf15b observed in the bonobo. Chimpanzee–bonobo Y-chromo- TBL1Y chimpanzees: some divergence has been estimated to ;1.8 MYA (Stone USP9Y western et al. 2002). Therefore, the disruptive mutations at these 3 gene disruption: TMSB4Y central genes likely occurred between ;6 and ;1.8 MYA (fig. 1). eastern In contrast to the other 3 genes, the exon–intron splice site mutation in exon 1 of the chimpanzee TMSB4Y gene is not present in our bonobo sequence. Instead, the bonobo 876543210 splice site sequence is identical to the human sequence Million years ago (table 1), suggesting that the disruptive mutation at this gene FIG. 1.—Timing of Y-chromosome gene losses during chimpanzee occurred in the chimpanzee lineage following chimpanzee– evolution. Disruptions to the coding sequences of 3 Y-chromosome genes bonobo divergence. Both of the publicly available chim- (CYorf15b, TBL1Y, and USP9Y) are estimated to have an origin in the an- cestral chimpanzee–bonobo lineage following divergence from the human panzee Y-chromosome sequences (Hughes et al. 2005; lineage, whereas the TMSB4Y coding sequence was disrupted in the chim- Kuroki et al. 2006) are from the sub- panzee lineage following chimpanzee–bonobo Y-chromosome diver- species (Pan troglodytes verus), and thus, the presence of gence, but prior to the separation of chimpanzee subspecies. the mutations in both sequences does not necessarily imply that they are fixed among chimpanzee subspecies. To ad- that, for each of the 7 gene regions, the nucleotide sequence dress this issue, we additionally obtained nucleotide se- is identical to the human sequence at the site of each of the 7 quence for this TMSB4Y gene region in 4 wild-born gene-disruptive mutations (table 1). From this, we infer that male chimpanzees representing each subspecies: 1 western each mutation occurred on the chimpanzee lineage follow- chimpanzee (Pan troglodytes verus), 1 ing divergence from the human–chimpanzee common an- (Pan troglodytes troglodytes), 1 (Pan cestor. This supported the previous conclusions of Hughes troglodytes schweinfurthii), and 1 Nigerian chimpanzee et al. (2005) that were based on comparisons of only the (Pan troglodytes vellerosus). We found that the TMSB4Y human and chimpanzee Y-chromosome sequences to the gene-disruptivemutationwaspresentinall4subspecies(table human X-chromosome sequence. 1). Although autosome and X-chromosome gene variation is In polarizing all 7 mutations to the chimpanzee line- often shared across chimpanzee subspecies because of rel- age, this simply purports that they have occurred sometime atively large effective population sizes for these loci in over the last ;6 Myr, or since the estimated divergence of chimpanzees (e.g., Kaessmann et al. 1999; Fischer et al. Pan and lineages (Kumar et al. 2005). Therefore, in 2004; Verrelli et al. 2006), this is not the case for the characterizing the male bonobo nucleotide sequence for Y-chromosome (Stone et al. 2002). Therefore, our data sug- these 7 gene regions, we can estimate the origin of the gest that the TMSB4Y mutation is fixed among chimpanzee gene-disruptive mutations along the chimpanzee lineage. subspecies and occurred after bonobo–chimpanzee diver- It should be noted here that with our alignment of chimpan- gence (;1.8 MYA) but prior to the divergence of chimpan- zee and human Y-chromosome sequences, we are only zee Y-chromosome haplogroups (;0.7 MYA; fig. 1). identifying mutations that have occurred along these 2 lin- eages and not necessarily those mutations that may have Molecular Evidence for Relaxed Functional Constraint occurred along the gorilla lineage or the bonobo lineage, following divergence from chimpanzees. Therefore, there Just as the absence of any of these Y-chromosome may be other gene-disruptive mutations along the Y- gene-disruptive mutations within any lineage does not by chromosome that are fixed between these species, in which itself imply a functional , the presence of a disruptive case any estimates of gene loss from our analysis would be mutation within a coding region does not alone imply pseu- inaccurate given the ascertainment bias inherent within our dogenization. If the gene is still transcribed, the messenger sampling design (i.e., gene regions nonrandomly chosen RNA (mRNA) could have regulatory roles and/or may still based on the presence of known gene-disruptive muta- be translated into a shorter protein that remains functional. tions). Further nucleotide sequence analysis of the entire Hughes et al. (2005) performed reverse transcriptase–PCR 856 Perry et al.

Table 2 Given that we find no clear pattern of differential func- Human–Chimpanzee dN/dS Comparisons for Disrupted Y- tional constraint between upstream and downstream Chromosome Gene Regions regions, we cannot reject the hypothesis that these 4 genes Upstream of Disruptive Downstream of Disruptive have become pseudogenes in the chimpanzee lineage. Mutation Mutation However, it is interesting to note that, in general, the whole NS NS gene (combined upstream and downstream regions) dN/dS values and levels of nucleotide divergence for these genes Gene Sites Diffs Sites Diffs d /d Sites Diffs Sites Diffs d /d N S N S are not noticeably different from those observed for CYorf15b 112 0 26 0 0.0 307 3 92 3 0.300 other, intact genes in the X-degenerate portion of the Y- TBL1Y 680 9 205 9 0.301 526 4 143 5 0.217 TMSB4Y 77 1 19 0 NA 24 1 6 0 NA chromosome (Hughes et al. 2005; Kuroki et al. 2006). USP9Ya 4032 45 1158 14 0.923 1922 14 541 4 0.985 Therefore, despite the relatively ancient age for many of the gene-disrupting mutations, the typical dN/dS values NOTE.—N, nonsynonymous; S, synonymous. Diffs, Human–chimpanzee differ-

for these genes are inconsistent with expectations of neu- Downloaded from https://academic.oup.com/mbe/article/24/3/853/1246230 by guest on 23 September 2021 ences. NA (not available), could not be calculated due to a zero number of synon- ymous substitutions in the denominator. trality in the chimpanzee lineage. These seemingly incon- a The upstream and downstream regions for USP9Y were divided at the point of gruent results could be explained by the tendency for Y- the exon 34 disruptive mutation (see Results and Discussion). chromosome genes to evolve rapidly in general (Hughes et al. 2005), which may make it difficult to infer subtle dif- (RT–PCR) experiments for detection of mRNA in various ferences in functional constraint using comparisons among chimpanzee tissues. Although TMSB4Y mRNA was not subsets of these genes. This raises questions about our gen- found, detectable quantities of CYorf15b, TBL1Y, and US- eral expectations of the molecular signature associated with P9Y mRNA were present in multiple tissues. The finding of potentially gene-disrupting mutationsonthe Y-chromosome mRNA for these latter 3 genes leaves open the possibility andwhethernonneutralscenariosmayexplainthesepatterns. that they are still translated into functional . If this is Once a mutation disrupts a coding region, the gene the case, we may have certain expectations about the mo- may continue to degrade through the neutral fixation of ad- lecular evolutionary signature associated with functional ditional gene-disruptive mutations, now that functional constraint acting on protein-coding sequences. constraint on the protein-coding sequence has become re- Compared with functional genes that may be subject laxed. Therefore, we may expect that the age of the original to purifying or positive selection, pseudogenes are expected gene-disruptive mutation is correlated with the total number to follow a neutral pattern of nucleotide substitution. If a of disruptive mutations for each gene. Among the 4 chim- gene region is subject to purifying selection, the ratio of panzee Y-chromosome genes with disruptive mutations, nonsynonymous (amino acid changing) substitutions per only USP9Y contains multiple mutations that would disrupt nonsynonymous site to synonymous substitutions per syn- the coding sequence (table 1). In this respect, it is interesting onymous site (dN/dS) is expected to be less than 1, whereas to note that 4 gene-disruptive mutations occurred prior to under neutrality dN/dS will approximate 1. To evaluate chimpanzee–bonobo divergence, but that none have be- whether functional constraint may be acting on specific re- come fixed in ;1.8 Myr. The coding region of USP9Y gions of these genes, we aligned the entire human and chim- is considerably larger than the 2 other genes that were panzee coding sequences for each of the 4 genes. We disrupted in the chimpanzee–bonobo common ancestor computed and compared dN/dS for each gene for the region (CYorf15b and TBL1Y; table 2), and the absence of multiple upstream from the disruptive mutation with the region disruptive mutations in these genes may simply reflect gene downstream from the mutation (table 2). If, despite a disrup- size rather than functional constraint. Complete coding re- tive mutation, the upstream region of a gene maintains func- gion sequences for these genes in the bonobo would help to tion in chimpanzees, then upstream dN/dS may be different address this issue. For example, 21 bp downstream from the than that of the downstream region, which follows the dis- USP9Y exon 36 frameshift deletion that occurred in the ruptive mutation and is expected to be free of functional chimpanzee–bonobo common ancestor, we coincidentally constraint. found a 16-bp frameshift deletion in the bonobo that is Using Fisher’s exact tests, we find no significant dif- not present in any other primate lineage examined here ference between upstream and downstream region dN/dS, (fig. 2). Therefore, although no gene-disruptive mutations for any of the 4 genes with chimpanzee lineage gene-dis- have apparently become fixed in the chimpanzee lineage in ruptive mutations. For USP9Y, because we are not able to the last ;1.8 Myr, this gene continues to degrade in the determine which of the 4 disruptive mutations occurred first bonobo lineage. With additional sampling of the bonobo (i.e., we can infer only that they all occurred prior to chim- Y-chromosome sequence, we can better estimate the mag- panzee–bonobo divergence), we compared upstream with nitude of gene-disruptive mutations that may have recently downstream dN/dS in this fashion for each of them, and occurred in our close primate relatives. found no significant difference in any comparison. The re- Compared with CYorf15b, TBL1Y, and USP9Y, the sults for the USP9Y exon 34 gene-disruptive mutations are coding sequence of TMSB4Y was disrupted more recently shown in table 2 because this provides the most even divi- (i.e., following chimpanzee–bonobo divergence). How- sion (in terms of bp) between upstream and downstream ever, the absence of TMSB4Y mRNA from any chimpanzee regions. However, it is important to note that the small size tissue (Hughes et al. 2005) strongly suggests that this gene of some of these genes (e.g., TMSB4Y; table 2) may prohibit has become nonfunctional in the chimpanzee lineage. This us from detecting differences in functional constraint be- is an interesting observation; that is, of the 4 genes that tween upstream and downstream regions. have coding sequence–disruptive mutations, the one that Chimpanzee Y-Chromosome Gene Loss 857

AA seq E Q S D N E T A G G T K Y R L V G V L V H S G Q gorilla GAGCAGTCTGACAATGAAACTGCAGGAGGCACAAAGTACAGACTTGTAGGAGTGCTTGTACACAGTGGTCAA human GAGCAGTCTGACAATGAAACTGCAGGAGGCACAAAGTACAGACTTGTAGGAGTGCTTGTACACAGTGGTCAA chimpanzee GAGCAGTCTGACA----AACTGCAGGAGGCACAAAGTACAGACTTGTAGGAGTGCTTGTACACAGTGGTCAA bonobo GAGCAGTCTGACA----AACTGCAGGAGGCACAAAGT------GCTTGTACACAGTGGTCAA A B

FIG. 2.—Continued degradation of the USP9Y coding region in bonobos. The USP9Y exon 36 gene region shown for several primate species with the 4-bp frameshift deletion (A) that occurred in the chimpanzee–bonobo ancestral lineage, by comparison with the gorilla and human sequences. A second example of more recent gene degradation is shown by a 16-bp frameshift deletion (B) that is unique to the bonobo lineage. produces no detectable mRNA is the one that has occurred selective sweeps on the (e.g., Filatov et al. much more recently. Similar RT–PCR experiments in bo- 2000; Bachtrog 2004). nobo tissue may help to determine whether TMSB4Y tran- It is also possible that one or more of the gene- Downloaded from https://academic.oup.com/mbe/article/24/3/853/1246230 by guest on 23 September 2021 scription was either abolished prior to chimpanzee–bonobo disruptive mutations themselves may have been adaptive. divergence (i.e., following a regulatory region mutation) Olson (1999) has proposed the ‘‘less-is-more’’ hypothesis, leading to relaxed functional constraint on the amino acid which states that losses of gene function during hominin sequence or, alternatively, if this gene remains functional evolution may in some cases have conferred a fitness ben- provided we find no other disruptive mutations in this gene efit. However, as of yet, few examples conclusively support in bonobos. this hypothesis. For example, although Stedman et al. (2004) proposed that a frameshift deletion in the myosin Mating Systems and Y-Chromosome Evolution gene MYH16 led to masticatory gracilization and brain-size expansion in the genus Homo, this interpretation has sub- Hughes et al. (2005) originally proposed that sequently been called into question (Perry et al. 2005; CYorf15b, TBL1Y, TMSB4Y, and USP9Y may have been McCollum et al. 2006). More recently, 2 studies have evolutionary ‘‘casualties’’ of strong positive selection else- shown that a premature stop codon mutation in the human where on the Y-chromosome. By this action, low-frequency CASPASE12 gene was likely swept toward fixation by pos- disruptive mutations may become fixed because a linked itive selection (Wang et al. 2006; Xue et al. 2006), possibly advantageous mutation elsewhere on the Y-chromosome because loss of CASPASE12 gene function increases resis- is highly advantageous. Although this is less likely to hap- tance to the system-wide response to infection, or sepsis pen on X-chromosomal and autosomal haplotypes due to (Saleh et al. 2004, 2006). Therefore, it will be interesting recombination, the completely linked nonrecombining na- to determine how many examples of gene loss in fact fit ture of the Y-chromosome provides for such a scenario a picture of adaptive evolution when examining differences (Rice 1987; Bachtrog 2004). One possible adaptive sce- between humans and other . nario that could cause differential rates of gene loss on For example, given the high levels of sperm compe- the Y-chromosome may involve the difference in mating tition in chimpanzees, it is difficult to reconcile how chim- systems between humans and chimpanzees. panzee USP9Y-disruptive mutations could have been Chimpanzees and bonobos both have multi-male/ neutral because loss of function of this gene in humans multi-female mating systems (Nishida 1968; Kano 1982; leads to the absence of sperm in semen (Sun et al. 1999; Goodall 1986) with presumably high levels of sperm com- Blagosklonova et al. 2000). Interestingly, Gerrard and Fi- petition relative to humans. Evolutionary consequences of latov (2005) identified 2 different disruptive mutations in this difference may include a relatively greater testis to a small segment of the USP9Y coding region from the black body size ratio in chimpanzees than humans (Harcourt spider (Ateles geoffroyi). Spider monkeys, like et al. 1981) and significantly rapid evolution for genes in- chimpanzees and bonobos, have a multi-male/multi-female volved in sperm development and function in the chimpan- mating system (Eisenberger 1973; Cant 1978; Chapman zee and bonobo lineages (e.g., Dorus et al. 2004). et al. 1993, 1995), raising the possibility that knocking Therefore, given the enrichment for genes involved in sper- out USP9Y gene function may have been advantageous matogenesis on the Y-chromosome, it seems reasonable for primates with high levels of sperm competition. How- that there may have been many selective sweeps during ever, we cannot exclude the possibility that other genes the evolution of the chimpanzee Y-chromosome as a result compensate for the loss of USP9Y function in nonhuman of strong sexual selection. It is also possible that some of primates but not in humans, or that USP9Y gained new these selective episodes led to the fixation of gene-disruptive function in the human lineage, such that its disruption is mutations elsewhere on the Y-chromosome. In this study, less deleterious in nonhuman primates. To more fully test we have shown that 3 of the 4 analyzed Y-chromosome these hypotheses will require Y- chromosome sequences genes were disrupted prior to the divergence of chimpanzees from additional primate species, including multiple exam- and bonobos, implying that many sex-specific (i.e., Y- ples of each mating system. chromosomal) fixation events in the chimpanzee lineage were relatively ancient. These initial interspecific analyses Conclusion shed light on the historical impact that gene-disruptive muta- tions may have on fixation rates (i.e., dN/dS analyses), The discussion of gene gain and loss has been of great whereas population genetic analyses can eventually be interest and debate in understanding how humans and our used to test hypotheses about more recent events including primate relatives diverged (e.g., Olson 1999; Gilad et al. 858 Perry et al.

2003; Fortna et al. 2004; Hurles 2004; Wang et al. 2006). Fischer A, Wiebe V, Paabo S, Przeworski M. 2004. Evidence for Although we find that many of the Y-chromosome gene a complex demographic history of chimpanzees. Mol Biol disruptions in the chimpanzee lineage are relatively ancient Evol. 21:799–808. in origin, others have found that 3 of these genes are still Fortna A, Kim Y, MacLaren E, et al. (16 co-authors). 2004. Lin- transcribed in chimpanzees, and in general we find little ev- eage-specific gene duplication and loss in human and great evolution. PLoS Biol. 2:E207. idence for relaxed functional constraint relative to other Y- Gerrard DT, Filatov DA. 2005. Positive and negative selection on chromosome genes. Therefore, the full pseudogene status of mammalian Y chromosomes. Mol Biol Evol. 22:1423–1432. these genes warrants additional scrutiny. With a greater sam- Gilad Y, Man O, Paabo S, Lancet D. 2003. Human specific loss of pling of gene disruption events throughout the human and olfactory receptor genes. Proc Natl Acad Sci USA. 100:3324– chimpanzee genomes, it will be possible to determine 3327. whether the higher rate of gene disruption in chimpanzees Goodall J. 1986. The chimpanzees of Gombe: patterns of behav- is unique in comparison to other chromosomes. In light of ior. Cambridge: Belknap Press of Harvard University Press. Hahn Y, Lee B. 2006. Human-specific nonsense mutations iden-

the results found here, it will also be of interest to determine Downloaded from https://academic.oup.com/mbe/article/24/3/853/1246230 by guest on 23 September 2021 whether genes on other chromosomes that play a role in sper- tified by genome-sequence comparisons. Hum Genet. matogenesis or fertility also show different patterns of gene 119:169–178. Harcourt AH, Harvey PH, Larson SG, Short RV. 1981. Testis disruption between the human and chimpanzee genomes. weight, body weight and breeding system in primates. Nature. 293:55–57. Acknowledgments Hughes JF, Skaletsky H, Pyntikova T, Minx PJ, Graves T, Rozen S, Wilson RK, Page DC. 2005. Conservation of Y-linked genes We thank Phil Hedrick for encouraging conversation during human evolution revealed by comparative sequencing and Anne Stone, the New Iberia Research Center, Primate in chimpanzee. Nature. 437:100–103. Foundation of Arizona, IPBIR, and the Coriell Institute for Hurles M. 2004. Gene duplication: the genomic trade in spare Medical Research for the primate samples used in this parts. PLoS Biol. 2:E206. study. This work was supported by funding to B.C.V. from Kaessmann H, Wiebe V, Paabo S. 1999. Extensive nuclear DNA the Center for Evolutionary Functional Genomics in The sequence diversity among chimpanzees. Science. 286:1159– 1162. Biodesign Institute at the Arizona State University. Kano T. 1982. The social group of pygmy chimpanzees (Pan pan- iscus) of Wamba. Primates. 23:171–188. Literature Cited Kumar S, Filipski A, Swarna V, Walker A, Hedges SB. 2005. Placing confidence limits on the molecular age of the hu- Bachtrog D. 2004. Evidence that positive selection drives man-chimpanzee divergence. Proc Natl Acad Sci USA. Y-chromosome degeneration in Drosophila miranda. Nat 102:18842–18847. Genet. 36:518–522. Kuroki Y, Toyoda A, Noguchi H, et al. (19 co-authors). 2006. Com- Blagosklonova O, Fellmann F, Clavequin MC, Roux C, Bresson parative analysisof chimpanzee and human Ychromosomes un- JL. 2000. AZFa deletions in Sertoli cell-only syndrome: a ret- veils complex evolutionary pathway. Nat Genet. 38:158–167. rospective study. Mol Hum Reprod. 6:795–799. Lahn BT, Page DC. 1997. Functional coherence of the human Cant JGH. 1978. Population survey of the spider monkey, Ateles Y chromosome. Science. 278:675–680. geoffroyi, at Tikal, Guatemala. Primates. 19:525–535. McCollum MA, Sherwood CC, Vinyard CJ, Lovejoy CO, Scha- Chapman CA, White FJ, Wrangham RW. 1993. Defining sub- chat F. 2006. Of muscle-bound crania and human brain evo- group size in fission-fusion societies. Folia Primatol (Basel). lution: the story behind the MYH16 headlines. J Hum Evol. 61:31–34. 50:232–236. Chapman CA, Wrangham RW, Chapman LJ. 1995. Ecological Muller HJ. 1918. Genetic variability, twin hybrids and constant constraints on group size—an analysis of spider hybrids, in a case of balanced lethal factors. Genetics. monkey and chimpanzee subgroups. Behav Ecol Sociobiol. 3:422–499. 36:59–70. Newman TL, Tuzun E, Morrison VA, Hayden KE, Ventura M, Charlesworth B, Charlesworth D. 2000. The degeneration of McGrath SD, Rocchi M, Eichler EE. 2005. A genome-wide Y chromosomes. Philos Trans R Soc Lond B Biol Sci. survey of structural variation between human and chimpanzee. 355:1563–1572. Genome Res. 15:1344–1356. Chen FC, Li WH. 2001. Genomic divergences between humans Nishida T. 1968. The social group of wild chimpanzees in the and other hominoids and the effective population size of the Mahale Mountains. Primates. 9:167–224. common ancestor of humans and chimpanzees. Am J Hum Olson MV. 1999. When less is more: gene loss as an engine of Genet. 68:444–456. evolutionary change. Am J Hum Genet. 64:18–23. Chimpanzee Sequencing and Analysis Consortium. 2005. Initial Perry GH, Verrelli BC, Stone AC. 2005. Comparative analyses sequence of the chimpanzee genome and comparison with the reveal a complex history of molecular evolution for human . Nature. 437:69–87. MYH16. Mol Biol Evol. 22:379–382. Dorus S, Evans PD, Wyckoff GJ, Choi SS, Lahn BT. 2004. Rate Rice WR. 1987. Genetic hitchhiking and the evolution of reduced of molecular evolution of the seminal protein gene SEMG2 genetic activity of the Y sex chromosome. Genetics. 116: correlates with levels of female promiscuity. Nat Genet. 36: 161–167. 1326–1329. Rozen S, Skaletsky HJ. 2000. Primer3 on the WWW for Eisenberger JF. 1973. Reproduction in two species of spider mon- general users and for biologist programmers. In: Krawetz S, keys, Ateles fusciceps and Ateles geoffroyi. J . Misener S, editors. Bioinformatics methods and protocols: 54:955–957. methods in molecular biology. Totowa (NJ): Humana Press. Filatov DA, Moneger F, Negrutiu I, Charlesworth D. 2000. Low p. 365–386. variability in a Y-linked plant gene and its implications for Saleh M, Mathison JC, Wolinski MK, Bensinger SJ, Fitzgerald P, Y-chromosome evolution. Nature. 404:388–390. Droin N, Ulevitch RJ, Green DR, Nicholson DW. 2006. Chimpanzee Y-Chromosome Gene Loss 859

Enhanced bacterial clearance and sepsis resistance in caspase- Tyler-Smith C, Howe K, Santos FR. 2006. The rise and fall of the 12-deficient mice. Nature. 440:1064–1068. ape Y chromosome? Nat Genet. 38:141–143. Saleh M, Vaillancourt JP, Graham RK, et al. (16 co-authors). Varki A, Altheide TK. 2005. Comparing the human and chimpan- 2004. Differential modulation of endotoxin responsiveness zee genomes: searching for needles in a haystack. Genome by human caspase-12 polymorphisms. Nature. 429:75–79. Res. 15:1746–1758. Sambrook J, Russell DW. 2001. Molecular cloning: a laboratory Verrelli BC, Tishkoff SA, Stone AC, Touchman JW. 2006. Con- manual. Cold Spring Harbor (NY): Cold Spring Harbor Lab- trasting histories of G6PD molecular evolution and malarial oratory Press. resistance in humans and chimpanzees. Mol Biol Evol. Skaletsky H, Kuroda-Kawaguchi T, Minx PJ, et al. (40 co-authors). 23:1592–1601. 2003. The male-specific region of the human Y chromosome is Wang X, Grus WE, Zhang J. 2006. Gene losses during human a mosaic of discrete sequence classes. Nature. 423:825–837. origins. PLoS Biol. 4:e52. Stedman HH, Kozyak BW, Nelson A, Thesier DM, Su LT, Low Watanabe H, Fujiyama A, Hattori M, et al. (45 co-authors). 2004. DNA sequence and comparative analysis of chimpanzee chro- DW, Bridges CR, Shrager JB, Minugh-Purvis N, Mitchell MA. mosome 22. Nature. 429:382–388. 2004. Myosin gene mutation correlates with anatomical Downloaded from https://academic.oup.com/mbe/article/24/3/853/1246230 by guest on 23 September 2021 Xue Y, Daly A, Yngvadottir B, et al. (14 co-authors). 2006. Spread changes in the human lineage. Nature. 428:415–418. of an inactive form of caspase-12 in humans is due to recent Stone AC, Griffiths RC, Zegura SL, Hammer MF. 2002. High lev- positive selection. Am J Hum Genet. 78:659–670. els of Y-chromosome nucleotide diversity in the genus Pan. Yan HT, Shinka T, Kinoshita K, et al. (11 co-authors). 2005. Mo- Proc Natl Acad Sci USA. 99:43–48. lecular analysis of TBL1Y, a Y-linked homologue of TBL1X Sun C, Skaletsky H, Birren B, Devon K, Tang Z, Silber S, Oates R, related with X-linked late-onset sensorineural deafness. J Page DC. 1999. An azoospermic man with a de novo point Hum Genet. 50:175–181. mutation in the Y-chromosomal gene USP9Y. Nat Genet. 23:429–432. Tilford CA, Kuroda-Kawaguchi T, Skaletsky H, et al. (12 co- Yoko Satta, Associate Editor authors). 2001. A physical map of the human Y chromosome. Nature. 409:943–945. Accepted January 3, 2007