Unequal Rates of Y Chromosome Gene Divergence during Speciation of the Family Ursidae

Shigeki Nakagome,*1 Jill Pecon-Slattery, and Ryuichi Masuda*à *Division of Biological Science, Graduate School of Science, Hokkaido University, Sapporo, Japan; Laboratory of Genomic Diversity, National Cancer Institute-Frederick, Frederick, MD; and àDepartment of Genome Dynamics, Creative Research Initiative ‘‘Sousei’’, Hokkaido University, Sapporo, Japan

Evolution of the family Ursidae is well investigated in terms of morphological, paleontological, and genetic features. However, several phylogenetic ambiguities occur within the subfamily Ursinae (the family Ursidae excluding the and spectacled bear), which may correlate with behavioral traits of female philopatry and male-biased dispersal which form the basis of the observed matriarchal population structure in these species. In the process of bear evolution, we investigate the premise that such behavioral traits may be reflected in patterns of variation among genes with different

modes of inheritance: matrilineal mitochondrial DNA (mtDNA), patrilineal Y chromosome, biparentally inherited Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 autosomes, and the X chromosome. In the present study, we sequenced 3 Y-linked genes (3,453 bp) and 4 X-linked genes (4,960 bp) and reanalyzed previously published sequences from autosome genes (2,347 bp) in ursid species to investigate differences in evolutionary rates associated with patterns of inheritance. The results describe topological incongruence between sex-linked genes and autosome genes and between nuclear DNA and mtDNA. In more ancestral branches within the bear phylogeny, Y-linked genes evolved faster than autosome and X-linked genes, consistent with expectations based on male-driven evolution. However, this pattern changes among branches leading to each species within the lineage of Ursinae whereby the evolutionary rates of Y-linked genes have fewer than expected substitutions. This inconsistency between more recent nodes of the bear phylogeny with more ancestral nodes may reflect the influences of sex-biased dispersal as well as molecular evolutionary characteristics of the Y chromosome, and stochastic events in species natural history, and phylogeography unique to ursine .

Introduction mtDNA genes (Talbot and Shields 1996a, 1996b; Waits et al. 1999) and morphological studies (Kurten 1964; Species within the bear family Ursidae include the Mazza and Rustioni 1994). Further, the placement of world’s largest and are distributed widely in U. malayanus within Ursinae remained ambiguous among Eurasia and North and . Ursidae consists of these previously published studies. These discordant phy- 8 species: the spectacled bear ( ), sloth ornatus logenetic results underscore the inherent difficulties in con- bear ( ), ( ), Asiatic ursinus Ursus malayanus ducting evolutionary studies of species with rapid recent black bear ( ), ( Ursus thibetanus Ursus evolution and provide compelling support for the continued ), ( ), ( americanus Ursus arctos Ursus investigation of informative genomic markers. Here we as- ), and giant panda ( .The maritimus melanoleuca) sess the performance of 3 Y-linked (3,453 bp) and 4 X- progenitor of extant bears arose approximately 37–40 MYA, linked (4,960 bp) genes and conduct comparisons with mi- followed by A. melanoleuca around 12 MYA, T. ornatus at tochondrial and autosome genes to investigate differences 5–7 MYA and the progenitor of the subfamily Ursinae orig- in evolutionary rates associated with mode of inheritance. inating 4–6 MYA as indicated by the fossil record (Wayne Genes on mammalian sex chromosomes evolve differ- et al. 1991). This general pattern of Ursidae speciation is cor- ently due to patterns of inheritance. The Y chromosome is roborated by genomic data such as chromosome karyology exclusively patrilineal, whereas the mode of inheritance for (Nash and O’Brien 1987; Nash et al. 1998), protein electro- X chromosome is one-third in males and two-third in fe- phoresis (Slattery and O’Brien 1995), and molecular evolu- males (Miyata et al. 1987). With the exception of the small tionary studies (Talbot and Shields 1996b; Waits et al. 1999; pseudoautosomal region, that is, 5% of the human Y chro- Yu et al. 2004, 2007; Pages et al. 2007). mosome (Rappold 1993), the remainder, termed the nonre- Relationships between the 6 remaining species of the combining region of the Y chromosome (NRY), does not subfamily Ursinae differ between full-length mitochondrial undergo conventional recombination during male meiosis. DNA (mtDNA) (Yu et al. 2007) and nuclear genomic Consequently, genes in the NRY are thought to be under markers. For example, mtDNA phylogenetic trees depict strong selection for male-specific function or undergo deg- U. ursinus diverging first within Ursinae 4–6 MYA, fol- radation due to an accumulation of deleterious mutations lowed by a bifurcation forming 2 clades: one leading to through Muller’s ratchet (Charlesworth B and Charles- the ancestor of and . The second U. arctos U. maritimus worth D 1997), genetic hitchhiking (Charlesworth 1996), clade clearly defined the 2 species of black bear ( U. amri- background selection (Charlesworth 1996), and insertion and ) as sister taxa: a result observed canus U. thibetanus of retroposable elements (Charlesworth 1991). Most studies with autosome DNA (Yu et al. 2004), as well as other of X–Y homologs indicate that genes on the Y chromosome evolve faster than those on the X chromosome in primates, 1 Present address: Department of Integrated Biosciences, Graduate carnivores, perissodactyls, and (Haldane 1947; School of Frontier Sciences, University of Tokyo, Kashiwa, Japan. Huang et al. 1997; Pecon-Slattery and O’Brien 1998; Key words: Ursidae, sex-linked genes, male-biased dispersal, female philopatry, matriarchal structure. Makova and Li 2002; Wolfe and Li 2003; Sandstedt and Tucker 2005; Goetting-Minesky and Makova 2006) consis- E-mail: [email protected]. tent with expectations of male-driven evolution (Haldane Mol. Biol. Evol. 25(7):1344–1356. 2008 doi:10.1093/molbev/msn086 1947; Miyata et al. 1987; Shimmin et al. 1993; Makova Advance Access publication April 9, 2008 and Li 2002; Goetting-Minesky and Makova 2006).

Ó The Author 2008. 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] Sex Chromosome Evolution in Bears 1345

Like many mammalian species (see Lawson Handley and Perrin 2007), Ursinae species exhibit behavioral traits of male dispersal and female philopatry. Ecological studies demonstrate that North American (McLellan and Hovey 2001), European (Stoen et al. 2006), and Russian (Kojola et al. 2003) populations of brown bears (U. arctos) exhibit male bias in dispersal from the natal range, and thus, brown bears are spatially structured in matrilineal assemblages (Stoen et al. 2005). Similarly, the home range of males is expanded by an order of magnitude compared with fe- males of American black bears (U. americanus) (Nowak 1999), consistent with sex-biased dispersal and matriarchal

social structure (Onorato et al. 2004). In addition, these be- Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 havioral traits may be reflected in population genetic anal- yses of these species. For example, mtDNA haplotypes of brown bears (U. arctos) on the Hokkaido Island of Japan were separated clearly into 3 allopatric groups (Matsuhashi et al. 1999), indicative of female philopatry. Thus, the genetic consequences of sex-linked dispersal might be de- tected through comparisons of the level of genetic diversity

between patrilineal markers on the Y chromosome and ma- GenBank Accession Number for Gene Sequence trilineal mtDNA (Seielstad et al. 1998; Oota et al. 2001; Eriksson et al. 2006; Hammond et al. 2006). In general, studies designed to assess male dispersal and female philopatry are complex using models of popu- lation genetics, evolutionary genetics, and field observa- tions. Herein, we employ a phylogenetic approach using ZFY ZFX SRY SMCY SMCX PLP ALAS2 AB261807 AB261815 AB292063 AB261823 AB261831 AB292071 AB294722 intron sequences of sex-linked, single-copy Y–X homologs AB261813 AB261821 AB292069 AB261829 AB261837 AB292077 AB294727 ZFY/X (zinc finger protein on Y/X), SMCY/X (selected cDNA on Y/X), along with X-linked PLP (proteo- lipid protein) and ALAS2 (aminolevulinate, delta-, synthase 2), and coding and adjacent noncoding regions of SRY (sex-determining region on Y) to propose that evolution of extant bears is likely associated with ongoing and his- toric sex-linked social behaviors. Notably, low substitution rates in Y chromosome genes within more recent Ursinae lineages of the bear phylogeny compared with those within Source of Sample more ancestral branches may reflect the influence of male- Science, Japan biased dispersal. Japan a Materials and Methods DNA Specimens Tissue samples were obtained from male individuals from 8 species of the family Ursidae (table 1). Total DNA was extracted using the DNeasy tissue kit (Qiagen, Tokyo, Japan) or the QIAamp DNA Micro kit (Qiagen). Additional samples were included from female individuals of the brown bear (U. arctos) and Asiatic black bear (U. thibetanus) to confirm male specificity of primers de- Sun bear Hair roots Ueno Zoological Gardens, Japan AB261811 AB261819 AB292067 AB261827 AB261835 AB292075 AB294725 Polar bear Blood Tennoji Zoo, Japan AB261808 AB261816 AB292064 AB261824 AB261832 AB292072 AB294721 Liver Kobe Municipal Oji Zoo, Japan AB261812 AB261820 AB292068 AB261828 AB261836 AB292076 AB294726 Brown bear Liver Hokkaido Institute of Environmental Giant panda Fibroblasts Ueno Zoological Gardens, Japan AB261814 AB261822 AB292070 AB261830 AB261838 AB292078 AB294728

veloped for Y chromosome genes. Spectacled bear Hair roots and blood Tennoji Zoo and Yokohama Zoo, Asiatic black bear Muscle Hunting, Honshu, Japan AB261810 AB261818 AB292066 AB261826 AB261834 AB292074 AB294724 American black bear Hair roots Ikeda Zoo, Japan AB261809 AB261817 AB292065 AB261825 AB261833 AB292073 AB294723

Primer Design and Polymerase Chain Reaction Analysis of 3 Y Chromosome Genes and 4 X Chromosome Genes Polymerase chain reaction (PCR) primers were de-

signed for the final intron of ZFY/X, the fourth intron of One male sample was used for each species.

SMCY/X, the third intron of PLP, the seventh intron of a Table 1 Profiles of Samples Used in This Study Species Identification Common Name Tissue Profile Ursus arctos Ursus maritimus Ursus americanus Ursus thibetanus Ursus malayanus Ursus ursinus Tremarctos ornatus ALAS2, and the coding and noncoding regions of SRY Ailuropoda melanoleuca 1346 Nakagome et al.

(supplementary table 1, Supplementary Material online). For Material online) newly designed in this study were 5# la- ZFY/X, SMCY/X, PLP,andALAS2, the PCR primers were beled with Texas red. The cycle PCR products were se- designed from conserved exon sequences flanking the intron quenced with an automated sequencer HITACHI SQ-5500. regions, which were obtained by alignments of published cDNA sequences for the human and mouse (GenBank ac- cession numbers of the National Center for Biotechnology Phylogenetic Analysis Information : ZFY/X,NM_003411/NM_003410 in the hu- man and NM_009570/NM_011768 in the mouse; SMCY/X, Nucleotide sequences were compiled and aligned by NM_004653/NM_004187 in the human and NM_011419/ GeneWorks (IntelliGenetics, Inc., Mountain View, CA). NM_0137668inthemouse;PLP,NM_199478.1inthehuman For phylogenetic analyses, gaps (insertion/deletions [in- and NM_011123.2 in the mouse; and ALAS2, NM_000032.2 dels]) and short interspersed elements (SINEs) were re- in the human and NM_009653.2 in the mouse). moved. For comparison with sex chromosome DNA data

All PCR conditions and primer sets used in this study sets of the present study, previously published sequences Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 were described in supplementary table 2 (Supplementary of autosome genes TTR-intron 1 and IRBP-exon 1 (Yu Material online). Except for the volumes of rTaq DNA et al. 2004) were included (GenBank accession numbers polymerase (Takara Bio Inc., Otsu, Japan), all PCR mix- of previously published sequences are shown in supple- tures were same as a total volume of 50 ll including 2 ll mentary table 4, Supplementary Material online). of the DNA extract (50–100 ng/ll), 10 mM Tris–HCl Phylogenetically informative sites and variable sites (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM deoxynu- were obtained using MEGA version 3.1 (Kumar et al. cleoside triphosphate, and 0.25 lM of each prime. The vol- 2004). Phylogenetic trees were constructed using umes of rTaq DNA polymerase were used in 1.25 or 2.5 PAUP4.0b10(Swofford2002)under3differentoptimalitycri- units. For ZFY/X and SMCY/X, the intron regions of Y teria: minimum evolution (ME), maximum parsimony (MP), and X chromosomes were simultaneously amplified using and maximum likelihood (ML) methods. Under the ME crite- primers located on the exon regions (U-ZF-1F/1R and U- rion, the phylogenetic trees were reconstructed using the Ta- SMC-2F/2R). For ZFY/X, it was necessary to perform jima–Nei distance model (Tajima and Nei 1984). For the ML nested PCR using 5 ll of first-round PCR products in analysis, hierarchical likelihood ratio tests were initially the primer set of U-ZF-2F/2R because nonspecific PCR introduced to compare the goodness of fit for 56 nucleotide products where obtained in first-round PCR. In addition, substitution models using Modeltest3.7 (Posada and Crandall the PCR products of ZFY/X and SMCY/X for the brown bear 1998) for 3 individual data sets of Y-linked, X-linked, or (U. arctos) and giant panda (A. melanoleuca) were cloned autosome genes, and additional analyses of combined using the TA cloning kit (Invitrogen, Carlsbad, CA) to de- nuclear DNA data (Y-linked, X-linked, and autosome gene sign internal primers specific to each intron region of ZFY/X regions) to determine the specific model parameters. ML or SMCY/X, respectively (supplementary table 1, Supple- substitution models showed 1) the Hasegawa, Kishino, and mentary Material online). Using the combinations of the Yano (HKY) model with estimated nucleotide base frequen- exonic and internal primers, the intron regions of ZFY/X cies of A 5 0.3004, C 5 0.1730, G 5 0.2044, and and SMCY/X were separated into 2 segments to amplify T 5 0.3222; a transition:transversion ratio 5 1.7417 for the specific region of Y or X chromosome, respectively Y-linked genes; 2) the HKY þ gamma model with estimated (supplementary table 2, Supplementary Material online). nucleotide base frequencies of A 5 0.3054, C 5 0.1919, For PLP and ALAS2, the amplification of intron regions G 5 0.2239, and T 5 0.2788; a transition:transversion ratio succeeded using primer sets situated in flanking exon regions 5 2.1345; gamma 5 0.0139 for X-linked genes; 3) the HKY (supplementary table 2, Supplementary Material online). In þ gammamodelwithestimatednucleotidebasefrequenciesof order to design the sequencing primers, the TA cloning was A 5 0.2208, C 5 0.2867, G 5 0.2692, and T 5 0.2233; conducted using the PCR products of the brown bear (U. arc- a transition:transversion ratio 5 3.021; gamma 5 0.006 for- tos) and giant panda (A. melanoleuca) (supplementary table 3, autosome genes; 4) HKY þ gamma model with estimated nu- Supplementary Material online). cleotide base frequencies of A 5 0.2836, C 5 0.2084, For SRY, the nucleotide sequence data of the brown bear G 5 0.2281, and T 5 0.2800; a transition:transversion (U. arctos) (GenBank accession number: AY424666) were ratio 5 2.1748; gamma 5 0.0171 for the combined nuclear used to design 2 primer pairs for seminested PCR; U-SRY- data set. In the ME, MP, and ML methods, optimal 1F/1R and 2F/2R (supplementary tables 1 and 2, Supplemen- trees were determined by an exhaustive search. The reliability tary Material online). In the spectacled bear (T. ornatus), ofderivedphylogeneticrelationshipswasevaluatedusingboot- seminested PCR was performed using U-SRY-1F/2R or strap analyses, and clades supported by greater than 50% of U-SRY-2F/1R (supplementary table 2, Supplementary Ma- node bootstrap values were retained. For the ME and MP anal- terial online). yses,1,000iterationsofbootstrapwereperformedwithheuristic The male specificity of Y-linked gene primers was con- tree searches employing the tree-bisection-reconnection firmed by the presence of PCR product in males and the ab- (TBR) branch swapping. For the ML bootstrapping, 100 sence in females of both the brown bear (U. arctos) and iterations were implemented using these conditions. Asiatic black bear (U. thibetanus). All PCR products were A fourth method utilizing a Bayesian approach for purified with the QIAquick purification kit (Qiagen). PCR computing clade credibility values for nodes within the tree sequencing reactions used the Thermo Sequenase Primer was performed using program MrBayes (version 3.1.2) Cycle Sequencing Kit (Amersham, Piscataway, NJ). Se- (Huelsenbeck and Ronquist 2001). The Bayesian analysis quencing primers (supplementary table 3, Supplementary was also used in 3 individual data sets of Y-linked, Sex Chromosome Evolution in Bears 1347

X-linked, or autosome genes and combined nuclear data nt 938–1,698 and differed by 12 variable sites between set. The nucleotide evolution models used in ML analyses the 2 species (supplementary fig. 7, Supplementary Material were also incorporated in Bayesian methods. Specific online). parameters included: a random starting tree, no phyloge- Autoapomorphic indels occurred within these genes netic constraints, 4 Markov chains run for 2,000,000 unique to a given species. A dinucleotide repeat (AT)n var- generations, empirical estimates of stability-likelihood val- ied between U. americanus and other species at nt 506–514 ues set to burn-in at 10,000 generations, and tree sampling of the SMCY-intron 4 (supplementary fig. 2, Supplementary every 100 generations. Two runs were performed to con- Material online). A 21-bp insertion was founded at nt firm the stability of posterior probability. The convergence 343–363 in the ZFY-final intron of T. ornatus and an 11-bp of Markov chains was assessed from the average standard insertion at nt 565–575 in A. melanoleuca (supplemen- deviation (,0.01) and the potential scale reduction factor tary fig. 1, Supplementary Material online). Tandem repeats (close to 1.000), as well as the log likelihood values. of TTGA motif at nt 604–615 in the ZFX-final intron varied

from 5 repeats in T. ornatus,3inU. thibetanus,and4inthe Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 other species (supplementary fig. 4, Supplementary Material Comparison of Substitution Rates online). In X-linked genes, a microsatellite (CA)n locus oc- curredatnt2,337–2,398intheALAS2-intron7(supplementary Differences in rates of substitution between nuclear fig.7,SupplementaryMaterialonline)andvariedamongbear genes located on different chromosome regions were esti- species as the following: U. arctos (14 repeats), U. maritimus mated by computing the pairwise genetic distances between (20 repeats), U. americanus (25 repeats), U. thibetanus species. Using a constraint tree generated by phylogenetic (26 repeats), U. malayanus (19 repeats), U. ursinus (25 analyses of the combined data set of sex-linked and auto- repeats), T. ornatus (20 repeats), and A. melanoleuca some genes (9,515 bp), branch lengths for each separate (11 repeats). gene category (Y-linked, X-linked, or autosome genes) The sequence for SRY consisted of the adjacent 5# were computed using the method described in Sandstedt noncoding flank (nt 1–434), the SRY gene (a single exon and Tucker (2005) and expressed as the number of substi- nt 435–1,100), and the 3# noncoding flank (nt 1,101– tution per 100 sites. The substitution rates among nuclear 1,267) (supplementary fig. 3, Supplementary Material on- genes were compared as follows: the branch of node A to line). In U. arctos, there was 1 missense mutation at nt the spectacled bear (T. ornatus), that of node A to 6 ursine 1,098 (T to C), which extended the SRY protein by an ad- species, that of node A to node B, and that of node B to 6 ditional 11 amino acids (33 bp) relative to the other species. ursine species (node A and node B as shown in fig. 3). Tests for significant differences of substitution rates among Y- linked, X-linked, and autosome genes using STATISTICA Phylogenetic Analyses version 06J (StatSoft, Tokyo, Japan) adopted 1-way anal- ysis of variance (Snedecor and Cochran 1989) and the post The reconstructions of bear phylogeny were per- hoc multiple comparison tests (’s protected least- formed on sequences in which indels and SINEs were re- significant difference [PLSD]). Significance levels of moved. The resultant alignments consisted of 1,033 bp for P , 0.05 and P , 0.01 were adopted throughout. ZFY, 848 bp for SMCY, 1,266 bp for SRY, 850 bp for ZFX, 520 bp for SMCX, 993 bp for PLP, and 1,720 bp for ALAS2. The number and the percentage of variable sites, parsimo- Results niously informative sites, and the consistency index ob- tained from MP analysis were determined (table 2). In Portions of 3 Y chromosome genes and 4 X chromo- addition, the distribution of indels being parsimoniously in- some genes were sequenced in all 8 extant species of bears. formative was estimated as diagnostic sites for the contri- These regions included 1,090 bp of the ZFY-final intron, bution to each bear lineage defined in the sex-linked and 1,096 bp of the SMCY-intron 4, and 1,267 bp of the SRY (cod- autosome trees (table 3 and supplementary table 5, Supple- ing and noncodingregions) for Y chromosome genes; as well mentary Material online). Among nuclear genes, Y-linked as 861 bp of the ZFX-final intron, 531 bp of the SMCX-intron genes have the highest numbers of substitutions and more 4, 1,004 bp of the PLP-intron 3, and 2,564 bp of ALAS2- diversity than other genes. Despite of the largest partitions intron 7 for X chromosome genes (alignments shown in sup- of sequences, X-linked genes show the lowest numbers of plementary figs. 1–7, Supplementary Material online). substitutions. Nuclear genes had very little homoplasy, with Repetitive elements were present within sex-linked Y-linked genes having the highest consistency index 5 gene sequences in bears. Within the ZFY-final intron, a 0.99 of all genes categories. On average, Y-linked genes SINE insertion occurred at nucleotide (nt) sites 366–540 were the most parsimonious informative, followed by au- shared by all 8 bear species (supplementary fig. 1, Sup- tosome genes, with X-linked genes being the most con- plementary Material online). Shared by 2 species (U. arctos served for nuclear sequence data (table 2). and U. maritimus), an identical SINE insertion was located Phylogenetic trees were reconstructed with concate- at nt 610–824 in the SMCY-intron 4 that differed only in nated sequences of ZFY-SMCY-SRY for Y chromosome total lengths of the poly-A tail (14 and 12 bp for U. arctos genes (fig. 1A) and of ZFX-SMCX-PLP-ALAS2 for X chro- and U. maritimus, respectively, in supplementary fig. 2, mosome genes (fig. 1B) using A. melanoleuca as an out- Supplementary Material online). In ALAS2-intron 7, an in- group. Both trees depicted the early divergence of sertion occurred in A. melanoleuca and T. ornatus at T. ornatus and strongly supported the monophyly of the 1348 Nakagome et al.

Table 2 Phylogenetic Parameters of Y Chromosome and X Chromosome Genes in Ursidae No. of %of No. of %of Parsimonious Parsimonious Gene Gene Sequence Variable Variable Informative Informative Consistency Regions Segments Length (bp) Sites Sites Sites Sites Index Y chromosome ZFY-final intron 1,090 60 5.50 10 0.92 genes SMCY-intron 4 881 (1,094–1,096)a 50 4.56 6 0.55 SRY 5#-noncoding 434 14 3.23 3 0.69 region SRY 3#-noncoding 167 8 4.79 1 0.60 region SRY-coding region 666 34 4.86 3 0.43 Combined 3,453 166 4.76 23 0.66 0.994 X chromosome ZFY-final intron 861 21 2.44 2 0.23 Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 genes SMCY-intron 4 531 23 4.33 4 0.75 PLP-intron 3 1,004 31 3.09 6 0.60 ALAS2-intron 7 1,803 (2,562–2,564)b 53 2.07 3 0.12 Combined 4,960 128 2.58 15 0.30 0.949 Autosome genes IRBP-exon 1 1,280 37 2.89 4 0.31 TTR-intron 1 1,067 55 5.15 15 1.41 Combined 2,347 92 3.92 19 0.81 0.945

a 1,094 bp, the polor bear with SINE insertion; 1,096 bp, the brown bear with the SINE insertion. b 2,562 bp, the giant panda with 759 insertion; 2,564 bp, the spectacled bear with 761-bp insertion.

6 ursine species that subsequently split to form 2 lineages: Based on lineages within Ursinae defined in each phy- one composed of U. arctos, U. maritimus, and U. ameri- logeny of Y-linked, X-linked (fig. 1A and B), or autosome canus and the other consisting of a less resolved clade of genes (fig. 2), we examined the distribution of diagnostic U. ursinus, U. malayanus, and U. thibetanus. Both Y sites for the lineages and investigated the contribution of and X chromosome phylogenetic trees, and the shared eachchromosomalgenetothe specifictopology ofcombined SINE insertion of SMCY-intron 4, supported the common nuclear phylogeny (table 3 and supplementary table 5, ancestry of U. arctos and U. maritimus. The divergence of Supplementary Material online). The monophyly of Ur- U. ursinus, U. malayanus, and U. thibetanus existed as a tri- sinae was strongly supported in all genes, especially Y chotomy with X-linked genes and was weakly supported by chromosome genes (N 5 18). For the first divergence of only the ME analyses of Y-linked genes. 6 species into 2 Asian species and the other (fig. 3), the di- Our reanalysis of previously published autosome agnostic sites were shown only in autosome genes (N 5 2 genes of TTR-intron 1 and IRBP-exon 1 recapitulated the in the former lineage and N 5 4 in the latter lineage). On original findings (Talbot and Shields 1996b; Waits et al. the other hand, the clade of U. arctos, U. maritimus, and 1999; Slattery et al. 2000; Yu et al. 2004, 2007). The phy- U. americanus was specific to sex-linked genes (N 5 5inY- logenetic trees (fig. 2) indicated a close association between linked genes and N 5 4 in X-linked genes). Weak support U. arctos and U. maritimus observed with sex-linked genes for the linage of U. thibetanus, U. ursinus, and U. malaya- but differ by uniting U. americanus with U. thibetanus. Au- nus or U. thibetanus and U. americanus was obtained in tosome genes further supported the clade of U. malayanus both X-linked (N 5 2) and autosome genes (N 5 1) or on- and U. ursinus as sister taxa. Thus, the phylogeny based on ly autosome genes (N 5 1). The monophyly of U. arctos autosome genes defined 2 clades that separated the 6 ursine and U. maritimus was supported consistently across all nu- species into a lineage composed of the 2 species U. malaya- clear genes within Ursinae (N 5 3 in Y-linked, N 5 2in nus and U. ursinus and with the remainder of U. arctos, U. X-linked, and N 5 2 in autosome genes). Overall, nodal maritimus, U. americanus, and U. thibetanus formed the support within the combined nuclear phylogeny varied second lineage (fig. 2). from moderate to high and the pattern of substitution A combined analysis of concatenated nuclear genes was diagnostic and parsimoniously informative (consis- (9,511 bp) resulted in a well-supported consensus phylog- tency index 5 0.96) and exhibited little homoplasy. eny for bear speciation (fig. 3). The early divergence of T. ornatus and the monophyly of Ursinae were recovered, consistent with results of the separate analyses of sex-linked Comparison of Rates of Substitution among Genome (fig. 1A and B) and autosome genes (fig. 2). Within Ursinae, Regions 2 lineages were established, which is consistent with the phylogeny of autosome genes (fig. 2). The first was com- Nucleotide substitution rates were compared between posed of Asian species of U. malayanus and U. ursinus as Y-linked, X-linked, and autosome genes using the topology sister species. The second clade, composed of the remaining derived from the combined data analysis as a constraint tree 4 Ursinae positioned both species of black bears (U. amer- to compute genetic distances estimated from branch lengths icanus and U. thibetanus) as more basal to the sister taxa of (table 4). Overall, the substitution rate of Y chromosome U. arctos and U. maritimus. genes was the highest, intermediate for autosome genes, Downloaded fromhttps://academic.oup.com/mbe/article/25/7/1344/1046167bygueston23September2021

Table 3 The Distribution of Diagnostic Sites Located in Y-Linked, X-Linked, and Autosome Genes That Define Each of the Bear Lineages Y Chromosome Genes X Chromosome Genes Autosome Genes ZFY-Final SMCY-Intron ZFX-Final SMCX-Intron PLP-Intron ALAS2-Intron IRBP-Exon IRBP-Exon Lineages Intron 4 SRY Y Chromosome Intron 4 3 7 X Chromosome 1 1 Autosomes Substitution Substitution Monophyly of Ursus (1)a (1)a arctos and Ursus Deletion Deletion Substitution Substitution maritimus (1)a SINE (1)a None N 5 3 None None None (1)a N 5 2 (1)a (1)a N 5 2 Monophyly of Ursus ursinus and Ursus Substitution malayanus None None None N 5 0 None None None None N 5 0 None (2)a N 5 2 Monophyly of Ursus thibetanus and Ursus Substitution americanus None None None N 5 0 None None None None N 5 0 None (1)a N 5 1 Substitution U. americanus, (1)a U. maritimus, and Substitution Substitution Substitution Deletion Substitution U. arctos group (2)a (2)a (1)a N 5 5 None (2)a (1)a None N 5 4 None None N 5 0 U. thibetanus, U. ursinus, and U. Substitution Substitution Substitution malayanus group None None None N 5 0 None (1)a (1)a None N 5 2 None (1)a N 5 1 U. arctos, U. Substitution maritimus, U. (3)a amcericanus, and Deletion U. thibetanus None None None N 5 0 None None None N 5 0 None (1)a N 5 4 e hoooeEouini er 1349 Bears in Evolution Chromosome Sex Substitution Substitution Substitution Substitution (6)a (4)a (2)a (2)a Deletion Deletion Substitution Substitution Substitution Deletion Substitution Deletion Subfamily Ursinae (1)a (1)a (6)a N 5 18 None (1)a (3)a (2)a N 5 8 (2)a (1)a N 5 5

NOTE.—N, number of sites (see supplementary table 5 and figs. 1–7 of alignments files, Supplementary Material online). a Parentheses are the number of sites showing substitutions, deletions, or SINE insertions compared with other lineage. 1350 Nakagome et al.

A Species 86/64/75 Ursus arctos Brown bear 0.975 93/98/96 1.00 Ursus maritimus Polar bear 67/-/- - Ursus americanus American black bear

94/-/- - Ursus ursinus Sloth bear 100/100/100 1.00 Ursus malayanus Sun bear

Ursus thibetanus Asiatic black bear Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021

Tremarctos ornatus Spectacled bear

Ailuropoda melanoleuca Giant panda

0.005

B Species 72/61/60 Ursus arctos Brown bear 1.00 85/87/83 1.00 Ursus maritimus Polar bear

99/98/98 Ursus americanus American black bear 1.00 Ursus thibetanus Asiatic black bear 70/-/- - Ursus ursinus Sloth bear 77/88/83 1.00 Ursus malayanus Sun bear

Tremarctos ornatus Spectacled bear

Ailuropoda melanoleuca Giant panda

0.002

FIG. 1.—Phylogenetic trees for combined data from ZFY-SMCY-SRY (3,238 bp) of Y chromosome genes (A) and ZFX-SMCX-PLP-ALAS2 (4,083 bp) of X chromosome genes (B) for 8 species of Ursidae. Nearly identical topologies were obtained among ME, MP, ML, and Bayesian analysis. Shown as representative are the ME trees using Tajima–Nei distance model. Under ME, phylogenetic analysis resulted in a single tree for Y chromosome or X chromosome genes, respectively (tree score was 0.05437 for Y chromosome genes or 0.02948 for X chromosome genes). Phylogenetic analysis using MP criteria recovered 2 equivalent trees in Y chromosome genes and 1 equivalent tree in X chromosome genes. For ML, ln likelihood scores were 5327.98091 in Y chromosome genes and 6430.86294 in X chromosome genes. Above numbers in italics adjacent to nodes represent bootstrap values as ME (1,000 iterations)/MP (1,000 iterations)/ML (100 iterations), and below the numbers show posterior probability values. A solid asterisk indicates a SINE insertion of SMCY-intron 4 in a common ancestor to Ursus arctos and Ursus maritimus. and the lowest for those genes located on the X chromosome relative to autosome genes, and roughly equivalent to the (node A to T. ornatus, and each ursine species in table 4 estimates from X-linked gene regions (table 4 and fig. 4). and supplementary fig. 8, Supplementary Material on- line). These relative differences in substitution rates are not consistent across the bear phylogeny (table 4 and fig. 4). Discussion In particular, estimates of the relative differences between Y-linked, X-linked, and autosome substitution rates from Patterns of evolution for genes located on the sex chro- node A to node B were discordant with those based on mosomes are assessed within the bear family Ursidae. The node B to each ursine species (table 4 and fig. 4). In the male-determining gene SRY is sequenced in entirety along former, Y-linked genes had significantly higher values with intron segments from X–Y homologs ZFX/ZFY and compared with those from the X-linked and autosome SMCX/SMCY and X-linked genes PLP and ALAS2. Phylo- genes (P , 0.01), which themselves evolved at the roughly genetic analyses of the sex-linked genes indicate low levels same rate. In contrast, substitution rates computed from of homoplasy, consistent with previous studies of specia- node B to each ursine species were not significantly ele- tion in the family (Pecon-Slattery and O’Brien vated, but rather were appreciably lower in Y-linked genes 1998; Pecon-Slattery et al. 2004; Johnson et al. 2006; King Sex Chromosome Evolution in Bears 1351

Species

69/74/78 Ursus arctos Brown bear 1.00 95/90/87 Ursus maritimus Polar bear 1.00 Ursus americanus American black bear 90/69/73 0.98 53/55/57 0.90 Ursus thibetanus Asiatic black bear

Ursus malayanus Sun bear

97/90/88 1.00 Ursus ursinus Sloth bear Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 Tremarctos ornatus Spectacled bear

Ailuropoda melanoleuca Giant panda

0.001

FIG. 2.—The ME trees reconstructed using previously published autosome concatenated sequences IRBP-Exon 1 and TTR-Intron 1 (2,281 bp) of 8 Ursidae species. Genetic distances were calculated using Tajima–Nei distance model. Nearly identical topologies were obtained from MP, ML, and Bayesian analysis. Shown as representative is the ME tree using Tajima–Nei distance model. Under ME, phylogenetic analysis resulted in a single tree (tree score was 0.04203). Phylogenetic analysis using MP criteria recovered 1 equivalent tree. For ML, ln likelihood score was 3804.52089. Numbers in italics above internal branches show the bootstrap values for ME (1,000 iterations)/MP (1,000 iterations)/ML (100 iterations), and below numbers are posterior probability values by the Bayesian analysis. et al. 2007), but exhibit unusual patterns of substitution (Yu et al. 2007) and autosome genes (Yu et al. 2004). Fur- within the evolution of the bear family. ther, ZFY and SMCY contain SINE elements not found in their respective X homologs that are phylogenetically infor- mative in bear speciation. First described in a subset of bear species (Slattery et al. 2000), the present study reveals that Phylogenetic Assessment of Y Chromosome and X the SINE insertion in the -final intron occurs in all bear Chromosome Genes ZFY species and the SINE in SMCY offers further confirmation Both Y-linked and X-linked genes affirm the early di- of the close association between the polar bear (U. mariti- vergence of the spectacled bear (T. ornatus) and the mono- mus) and brown bear (U. arctos). Thus, the ZFY SINE is phyletic lineage of the 6 species of the subfamily Ursinae ancestral to all bears and inserted into the Y chromosome observed with phylogenetic analyses of complete mtDNA prior to the divergence of the 8 extant species at least 12

Species 92/92/95 1.00 Ursus arctos Brown bear 100/99/100 1.00 Ursus maritimus Polar bear 81/52/57 0.91 Ursus americanus American black bear

100/100/100 1.00 B Ursus thibetanus Asiatic black bear

Ursus malayanus Sun bear A 97/89/89 1.00 Ursus ursinus Sloth bear

Tremarctos ornatus Spectacled bear

Ailuropoda melanoleuca Giant panda

0.002

FIG.3.—Phylogenetic tree of combined nuclear genes data (9,511 bp) from Y chromosome genes (ZFY-SMCY-SRY), X chromosome genes (ZFX- SMCX-PLP-ALAS2), and autosome genes (IRBP-TTR) for 8 species of Ursidae. Nearly identical topologies were obtained among ME, MP, ML, and Bayesian analysis. Shown as representative is the ME tree using Tajima–Nei distance model. Under ME, phylogenetic analysis resulted in a single tree (tree score was 0.04081). Phylogenetic analysis using MP criteria recovered 2 equivalent trees. The MP tree has a consistency index of 0.962. For ML, ln likelihood score was 15721.31571. Above numbers in italics adjacent to nodes represent bootstrap values as ME (1,000 iterations)/MP (1,000 iterations)/ML (100 iterations), and below the numbers show posterior probability values. Node A shows the ancestral node of Ursidae excluding the giant panda, and node B indicates the ancestral node of the subfamily Ursinae. 1352 Nakagome et al.

a MYA (Wayne et al. 1991), and the SMCY SINE is more recent, supporting the view that the polar bear evolved from brown bear populations isolated in northernmost areas of Asia during the last glaciations and rapidly adapted to ex- treme environmental conditions (Kurten 1964; Kurten - 1968; Talbot and Shields 1996a). This close association be-

TTR tween U. arctos and U. maritimus is clearly supported not Intron 1 Autosomes only by patterns of substitution of sex-linked genes exam-

- ined in this study but also in previous molecular genetic studies as well (Talbot and Shields 1996a, 1996b; Waits IRBP Exon 1 et al. 1999; Yu et al. 2004, 2007).

a Within the subfamily Ursinae, 2 lineages are recov-

ered, the first, well-supported by both Y- and X-linked Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021

X genes and comprised of sister taxa of the brown bear (U. 0.042 0.118 0.535 0.302 arctos) and polar bear (U. maritimus) with the American Chromosome black bear (U. americanus) having a relatively basal posi- tion within this clade. The second lineage is less resolved

c and is collapsed into a trichotomy of the sun bear (U. ma- layanus), sloth bear (U. ursinus), and Asiatic black bear (U. -Intron 7 0.001 thibetanus). These findings differ from those based on au- tosome genes that separate Ursinae into 2 clades: one com- ALAS2 posed of the 2 Asian species and the other formed by the

- remaining 4 species (fig. 2). In addition, previously pub- lished phylogenetic trees based on mtDNA (Talbot and PLP 0.024 Intron 3 Shields 1996a; Waits et al. 1999; Yu et al. 2007) and au-

c tosome genes (Yu et al. 2004), as well as morphological - characters (Allen 1938) and the fossil record (Kurten 0.002

SMCX 1964) explicitly define the monophyly for the Asiatic black Intron 4 bear (U. thibetanus) and American black bear (U. ameri- canus) as sister taxa.

-Final These phylogenetic discrepancies between sex-linked

Intron genes and autosome genes, as well as mtDNA genomes (Yu ZFX et al. 2007), may be due to an insufficient accumulation of a informative substitutions in the subfamily Ursinae. Com- bined nuclear data (9,515 bp) provides a more resolved

Y phylogeny (fig. 3) marked by high nodal support and little 0.190 0.178 0.186 0.112homoplasy 0.116but 0.137 still 0.118 does 0.687 notrecover 0.368 the expected sister– Chromosome taxa relationship of the 2 species of black bear. Rather,

c the observed association between the Asiatic black bear - (U. thibetanus) with the sloth bear (U. ursinus) and sun bear 0.005 0.140.305 0.059 0.322 0.178 0.058 0.169 0.219 0.293 0.211 0.091 0.627 0.326 SRY Coding

(U. malayanus) is supported mostly by diagnostic changes within X-linked genes (N 5 2) (fig. 1B, table 3 and supple-

c c mentary table 5, Supplementary Material online). To inves- - tigate the possible influence of incongruent phylogenetic 0.001 0.003 SRY information of X chromosome, we reconstructed the phy- Noncoding logeny of nuclear genes without the X-linked data (supple- Y Chromosome Genes X Chromosome Genes Autosome Genes c - mentary fig. 9, Supplementary Material online). The resultant topology is completely consistent with that based 0.008 SMCY Intron 4 on total combined data indicating that the split of the clade of 2 black bears is not attributable to incongruent X-linked partitions but instead reflects the strong association between -Final the American black bear (U. americanus) with the sister 0.8990.6020.297 0.812 0.536 0.277 0.647 0.507 0.140 0.633 0.458 0.175 0.770 0.534 0.235 0.216 0.059 0.158 0.456 0.222 0.234 0.506 0.304 0.202 0.234 0.118 0.116 0.324 0.163 0.161 0.352 0.195 0.158 0.865 0.166 0.699 0.577 0.181 0.395 Intron

ZFY taxa clade of the brown bear (U. arctos) and polar bear (U. maritimus) (figs. 1A and B; table 3 and supplementary table 5, Supplementary Material online). Interestingly, the identical topology is recovered in a parallel study (Pages et al. 2007) based on combined data from 14 nuclear genes b b including 3 Y-linked genes of ZFY, SRY, and UBE1Y. Branches The average substitution rates from node to each Ursine species. Concatenating 4 (Y chromosome), 4 (X chromosome), andThe 2 negative (autosomes) values genes. of branch lengths. Therefore, there is no indication of chromosome bias in lin- B a b c eage definitions and phylogenetic information present Brown bearPolar bear 0.290 0.290 0.236 0.236 0.339 0.169 0.149 0.255 0.059 0.186 0.350 0.175 0.194 0.277 0.722 0.473 Sloth bear 0.065 0.358 0.167 0.304 0.214 0.176 0.366 0.288 0.058 0.177 0.144 0.917 0.482 American black bearAsiatic black bear 0.551 0.091 0.480 0.169 0.148 0.373 0.297 0.498 0.221 0.057 0.203 0.198 0.707 0.422 Sun bear 0.495 0.357 to the spectacled bearto Ursine 1.293 0.905 0.677 0.914 0.989 0.256 0.222 0.749 0.233 0.362 0.798 0.763 0.784 to to Ursine to each species A Table 4 Substitution Rates (numbers per 100 sites) of Nuclear Genes (Y chromosome, X chromosome, and autosomes) A A B B within Y-linked, X-linked, or autosome genes is reflected Sex Chromosome Evolution in Bears 1353

1

0.8

** Y>X, Y>A 0.6 * A>X

0.4 Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021

0.2

0 A to B U. arctos U. maritimus U. americanus U. thibetanus U. malayanus U. ursinus All ursine bears Node B to each ursine species

FIG. 4.—Substitution rates of Y chromosome (black), X chromosome (white), and autosome (gray) genes in the branch of node A to B and node B to each species of ursine (fig. 3) are constructed using the values shown in table 4. Substitution rate values indicate the number of substitutions per 100 sites. In the graphs, the comparisons between chromosomes (Y . X, Y . A, or A . X) above the bars show the significant differences of the substitution rates (double asterisks are P , 0.01, and single asterisk is P , 0.05) between chromosomes (Y: Y chromosome, X: X chromosome, and A: autosomes). equivalently in the nuclear gene tree (fig. 3; table 3 and sup- Low Divergence of Y Chromosome Genes within plementary table 5, Supplementary Material online). Ursinae We explore 3 possible explanations for the discordant Differential Rates of Substitution between Y evolutionary rates observed with the Y chromosome in bear evolution (table 4 and fig. 4) namely, 1) biogeographical Chromosome, X Chromosome, and Autosome Genes effects, 2) evolutionary characteristics of the Y chromo- Consistent with expectations under the hypothesis of some, and 3) behavioral traits in the natural history of male-driven evolution (Haldane 1947; Miyata et al. 1987), the subfamily Ursinae. Modern species of Ursinae evolved the overall average substitution rates of Y chromosome genes recently and rapidly during the and are the highest among nuclear genes (table 4 and supplemen- (Yu et al. 2007) and likely experienced dramatic climatic taryfig.8,SupplementaryMaterialonline)andobservedother changes linked with glaciations. For example, the brown mammalian taxasuchasprimates(Huangetal. 1997; Makova bear (U. arctos) is distributed throughout North America, and Li 2002), carnivores (Pecon-Slattery and O’Brien 1998), Europe, and Asia yet has unique phylogeographic patterns perissodactyls (Goetting-Minesky and Makova 2006), and of extirpation and recolonization linked with patterns of rodents (Sandstedt and Tucker 2005). These results also dem- glaciations (Hofreiter et al. 2004; Miller et al. 2006; Stoen onstrate that the X-linked genes have the lowest and autosome et al. 2006). Stochastic changes brought about by popula- genes intermediate rates of substitution (table 4 and supple- tion expansion, migration, and contraction during this time mentary fig. 8, Supplementary Material online). A notable ex- may have facilitated selective sweeps in the Y chromosome ception to the expected rates of change with Y . XandY. A that led to less than expected substitution rates in Ursinae occursinevolutionofthesubfamilyUrsinae(table4andfig.4). observed today. The effective population size of Y chromo- Across the entire bear family, the ratio of Y/X 5 1.933 (95% some is smaller than autosomes (Lawson Handley and confidence interval[CI]: 1.262–2.605)(theequation:V(Y) 5 Perrin 2007). Therefore, if a favored mutation occurred Y(1 Y)/[L(1 4Y/3)2], V(X) 5 X(1 X)/[L(1 4X/3)]2, on the Y chromosome in ancestral populations undergoing V(Y/X) 5 V(Y)/E(X)2 þ E(Y)2V(X)/E(X)4, and Y/X 5 frequent structural changes, then the allele might spread Y/X 1.96s and Y/Xþ 5 Y/X þ 1.96s in accordance more rapidly than expected. The replacement of a circulat- with Sandstedt and Tucker [2005]) and a 5 3.624 (95% ing Y chromosome with another within ancestral popula- CI: 1.451–13.176) (the equation: Y/X 5 3 a/(2 þ a)shown tions of Ursinae would result in the appearance of low in Miyata et al. [1987]) changes within Ursinae whereby levels of divergence between species observed here. Y/X 5 1.313 (95% CI: 0.680–1.945) and a 5 1.556 (95% In contrast to the fixation of favorable mutations re- CI: 0.587–3.687). Thus, possible male-driven evolution is sulting in selective sweeps within a species, reduced genetic highly supported in the deeper nodes of the phylogeny but diversity on the Y chromosome is influenced also by the not in the more recent Ursinae. lack of conventional recombination during meiosis. A 1354 Nakagome et al. positive correlation exists between recombination rate and chromosome, and autosomes. In addition, Hoelzer (1997) polymorphism within genomes (Innan and Stephan 2003). and Hoelzer et al. (1998) showed that nodes within mito- Consequently, due to the lack of recombination needed to chondrial gene trees were deeper than those in nuclear gene remove harmful mutations, the fate of genes on the Y chro- trees if the female migration rate was low. Lastly, the effect mosome would depend on the inexorable accumulation of behavioral traits on the effective population size of au- of mutations (of which some will be deleterious), due to tosomes may not be associated with differences in dispersal Muller’s ratchet (Charlesworth B and Charlesworth D between sexes but rather male polygyny (Chesser and 1997). Advantageous mutations on the Y chromosome Baker 1996). Therefore, we suggest low divergence of could cause the fixation of all deleterious mutations present Y-linked genes observed here compared with high diver- on the chromosome, and successive ‘‘selective sweeps’’ of gence of mtDNA for the subfamily Ursinae (Yu et al. this kind would cause the fixation of deleterious alleles 2007) may be due, in part, to male migration, female phil- at many Y-linked loci (genetic hitchhiking). Background opatry, and polygynandry during ursine bear speciation.

or purifying selection acting against these harmful muta- Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 tions could reduce the number of Y chromosome variants Conclusions within a species, leading to decreased diversity of the Y chromosome in a finite population. This work constitutes the first direct comparisons of Y As described above, the biogeographical effects expe- chromosome, X chromosome, autosomes, and mtDNA rienced by ursine species combined with the nonrecombining phylogenies from the bear family Ursidae for the purpose characteristics of the Y chromosome could be the basis for of evaluating the influence of social structure into species low divergence of Y-linked genes in Ursinae. However, if divergence. Although the Y chromosome has the highest these 2 factors alone regulate Y chromosome divergence, substitution rates in the deeper nodes of the bear phylogeny, then other mammalian species within these same biogeo- low rates are shown in evolution of the subfamily Ursinae. graphical regions should show similar patterns because these We suggest 3 possibilities for these results: biogeographical processes are not specific to bears. For example, the diver- effects such as iterative glaciations of the recent past, mo- gence of the 7 species within the domestic cat lineage and the lecular evolutionary characters of the Y chromosome, and 5 species of the cat lineage of Felidae occurred behavioral traits of sex-biased dispersal and reproductive roughly at the same time as Ursinae and within similar bio- system. Although all these explanations combined are sig- geographic zones (Johnson et al. 2006). However, these felid nificant, we propose that dispersal patterns and mating sys- lineages do not show reduced diversity on the Y chromo- tems should be strongly considered when reconstructing some but instead are defined by unique diagnostic substitu- molecular evolutionary history of , especially in tions (Pecon-Slattery et al. 2004) correlated with increased those with the matriarchal social structures. rates of substitution relative to genes located on autosomes and the X chromosome (Johnson et al. 2006). Supplementary Material We propose that social behavior may be a significant factor in the discordant patterns of genome evolution in Supplementary figures 1–9 and tables 1–5 are avail- bears. The Ursinae, like most large mammals, exhibit able at Molecular Biology and Evolution online (http:// male-biased dispersal and female philopatry (Matsuhashi www.mbe.oxfordjournals.org/). et al. 1999; Nowak 1999; McLellan and Hovey 2001; Kojola et al. 2003; Onorato et al. 2004; Stoen et al. Acknowledgments 2005, 2006). Migration caused by male-biased dispersal al- lows gene flow and is a potent force in homogenizing ge- We would like to thank Dr T. Mano (Hokkaido Insti- netic divergence among subpopulations. In addition, the tute of Environmental Science), K. Ito and M. Kasahara reproductive system of bears is polygynandrous by which (Ueno Zoological Gardens), J. Morita (Ikeda Zoo), Dr K. a female may mate with 2 or more males, who themselves Murata (Nihon University), K. Takami (Tennoji Zoo), may pair with several different females (Nowak 1991). M. Ueda (Yokohama Zoo), and S. Dakemoto for supplying Therefore, if the social structure of female philopatry, bear samples. We also appreciate sample preparations by male-biased dispersal, and polygynandry was present in an- C. Nishida (Hokkaido University). Our thanks go to Dr S. A. cestral populations of Ursinae, relatively low number of Sandstedt (Michigan University), Dr Y. Ishibashi (Forest male breeders and high male migration rates would result and Forest Products Research Institute), and Dr H. Tsuruga in low effective population size and reduction of genetic (Hokkaido Institute of Environmental Science) for helpful diversity for the Y chromosome among breeding groups suggestions. This study was supported by Grants-in-Aid for (Chesser and Baker 1996). Furthermore, female philopatry Scientific Research from the Japan Society for the Promo- has been shown to increase the effective population size of tion of Science and by the 21st Century Center of Excel- mtDNA by one-half relative to autosomes and almost 6 lence Program ‘‘Neo-Science of Natural History’’ at times relative to the Y chromosome (Chesser and Baker Hokkaido University financed from the Ministry of Educa- 1996). Thus, genetic diversity of the Y chromosome among tion, Culture, Sports, Science, and Technology, Japan. ancestral populations would be lost at a faster rate than mtDNA and autosomes. These results were also supported in Laporte and Charlesworth (2002) showing that under Literature Cited predominantly male migration, high genetic differentiation Allen GM. 1938. The mammals of China and Mongolia. New was obtained in mtDNA, followed by Y chromosome, X York: American Museum of Natural History. Sex Chromosome Evolution in Bears 1355

Charlesworth B. 1991. The evolution of sex chromosomes. Makova KD, Li WH. 2002. Strong male-driven evolution of Science. 251:1030–1033. DNA sequences in humans and apes. Nature. 416:624–626. Charlesworth B. 1996. The evolution of chromosomal sex Matsuhashi T, Masuda R, Mano T, Yoshida MC. 1999. determination and dosage compensation. Curr Biol. 6:149–162. Microevolution of the mitochondrial DNA control region in Charlesworth B, Charlesworth D. 1997. Rapid fixation of the Japanese brown bear (Ursus arctos) population. Mol Biol deleterious alleles can be caused by Muller’s ratchet. Evol. 16:676–684. Res. 70:63–73. Mazza P, Rustioni M. 1994. On the phylogeny of Eurasian bears. Chesser RK, Baker RJ. 1996. Effective sizes and dynamics of Palaeontographica Abt A. 230:1–38. uniparentally and diparentally inherited genes. Genetics. McLellan BN, Hovey FW. 2001. Natal dispersal of grizzly bears. 144:1225–1235. Can J Zool. 79:838–844. Eriksson J, Siedel H, Lukas D, Kayser M, Erler A, Hashimoto C, Miller CR, Waits LP, Joyce P. 2006. Phylogeography and Hohmann G, Boesch C, Vigilant L. 2006. Y-chromosome mitochondrial diversity of extirpated brown bear (Ursus analysis confirms highly sex-biased dispersal and suggests arctos) populations in the contiguous United States and a low male effective population size in bonobos (Pan Mexico. Mol Ecol. 15:4477–4485. Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 paniscus). Mol Ecol. 15:939–949. Miyata T, Hayashida H, Kuma K, Mitsuyasu K, Yasunaga T. Goetting-Minesky MP, Makova KD. 2006. Mammalian male 1987. Male-driven molecular evolution: a model and mutation bias: impacts of generation time and regional nucleotide sequence analysis. Cold Spring Harb Symp Quant variation in substitution rates. J Mol Evol. 63:537–544. Biol. 52:863–867. Haldane JBS. 1947. The mutation rate of the gene for Nash WG, O’Brien SJ. 1987. A comparative chromosome haemophilia and its segregation ratios in males and females. banding analysis of the Ursidae and their relationship to other Ann Eugen. 13:158–163. carnivores. Cytogenet Cell Genet. 45:206–212. Hammond RL, Handley LJ, Winney BJ, Bruford MW, Perrin N. Nash WG, Wienberg J, Ferguson-Smith MA, Menninger JC, 2006. Genetic evidence for female-biased dispersal and gene O’Brien SJ. 1998. Comparative genomics: tracking chromo- flow in a polygynous primate. Proc Biol Sci. 273:479–484. some evolution in the family Ursidae using reciprocal Hoelzer GA. 1997. Inferring phylogenies from mtDNA variation: chromosome painting. Cytogenet Cell Genet. 83:182–192. mitochondrial-gene trees versus nuclear-gene trees revisited. Nowak RM. 1991. Walker’s mammals of the world, 5th ed. Evolution. 51:622–626. Baltimore (MD): Johns Hopkins Press. Hoelzer GA, Wallman J, Melnick DJ. 1998. The effects of social Nowak RM. 1999. Walker’s mammals of the world, 6th ed. structure, geographical structure, and population size on the Baltimore (MD): Johns Hopkins. evolution of mitochondrial DNA: II. Molecular clocks and the Onorato DP, Hellgren EC, Bussche RAVD, Raymond Skiles JJ. lineage sorting period. J Mol Evol. 47:21–31. 2004. Paternity and relatedness of American black bears Hofreiter M, Serre D, Rohland N, Rabeder G, Nagel D, recolonizing a desert montane island. Can J Zool. Conard N, Munzel S, Paabo S. 2004. Lack of phylogeography 82:1201–1210. in European mammals before the last glaciation. Proc Natl Oota H, Settheetham-Ishida W, Tiwawech D, Ishida T, Acad Sci USA. 101:12963–12968. Stoneking M. 2001. Human mtDNA and Y-chromosome Huang W, Chang BH, Gu X, Hewett-Emmett D, Li W. 1997. Sex variation is correlated with matrilocal versus patrilocal differences in mutation rate in higher primates estimated from residence. Nat Genet. 29:20–21. AMG intron sequences. J Mol Evol. 44:463–465. Pages M, Sebastien C, Catherine K, Mathilde P, Sandrine H, Huelsenbeck JP, Ronquist F. 2001. MrBayes: Bayesian inference Catherine H. 2007. Combined analysis of fourteen nuclear of phylogenetic trees. Bioinformatics. 17:754–755. genes refines the Ursidae phylogeny. Mol Phylogenet Evol. Innan H, Stephan W. 2003. Distinguishing the hitchhiking and doi: 10.1016/j.ympev.2007.10.019. background selection models. Genetics. 165:2307–2312. Pecon-Slattery J, O’Brien SJ. 1998. Patterns of Y and X Johnson WE, Eizirik E, Pecon-Slattery J, Murphy WJ, chromosome DNA sequence divergence during the Felidae Antunes A, Teeling E, O’Brien SJ. 2006. The late Miocene radiation. Genetics. 148:1245–1255. radiation of modern Felidae: a genetic assessment. Science. Pecon-Slattery J, Pearks Wilkerson AJ, Murphy WJ, O’Brien SJ. 311:73–77. 2004. Phylogenetic assessment of introns and SINEs within King V, Goodfellow PN, Pearks Wilkerson AJ, Johnson WE, the Y chromosome using the cat family Felidae as a species O’Brien SJ, Pecon-Slattery J. 2007. Evolution of the male- tree. Mol Biol Evol. 21:2299–2309. determining gene SRY within the cat family Felidae. Posada D, Crandall KA. 1998. MODELTEST: testing the model Genetics. 175:1855–1867. of DNA substitution. Bioinformatics. 14:817–818. Kojola I, Danilov PI, Hanna-Mari L, Vladimir B, Andrei Y. Rappold GA. 1993. The pseudoautosomal regions of the human 2003. Brown bear population structure in core and periphery: sex chromosomes. Hum Genet. 92:315–324. an analysis of human statistics from Russian Karelia and Sandstedt SA, Tucker PK. 2005. Male-driven evolution in Finland. Ursus. 14:17–20. closely related species of the mouse Mus. J Mol Evol. Kumar S, Tamura K, Nei M. 2004. MEGA3: integrated software 61:138–144. for molecular evolutionary genetics analysis and sequence Seielstad MT, Minch E, Cavalli-Sforza LL. 1998. Genetic alignment. Brief Bioinform. 5:150–163. evidence for a higher female migration rate in humans. Nat Kurten B. 1964. The evolution of the polar bear, Ursus maritimus Genet. 20:278–280. Phipps. Acta Zool Fenn. 108:1–30. Shimmin LC, Chang BH, Li WH. 1993. Male-driven evolution of Kurten B. 1968. Pleistocene mammals of Europe. Chicago (IL): DNA sequences. Nature. 362:745–747. Aldine. Slattery JP, Murphy WJ, O’Brien SJ. 2000. Patterns of diversity Laporte V, Charlesworth B. 2002. Effective population size and among SINE elements isolated from three Y-chromosome population subdivision in demographically structured pop- genes in carnivores. Mol Biol Evol. 17:825–829. ulations. Genetics. 162:501–519. Slattery JP, O’Brien SJ. 1995. Molecular phylogeny of the red Lawson Handley LJ, Perrin N. 2007. Advances in our un- panda (Ailurus fulgens). J Hered. 86:413–422. derstanding of mammalian sex-biased dispersal. Mol Ecol. Snedecor GW, Cochran WG. 1989. Statistical methods. Ames 16:1559–1578. (IA): Iowa State University Press. 1356 Nakagome et al.

Stoen O-G, Bellemain E, Saebo S, Swenson JE. 2005. Kin- phylogenetic estimation from multiple fragments of mtDNA. related spatial structure in brown bears Ursus arctos. Behav Mol Phylogenet Evol. 13:82–92. Ecol Sociobiol. 59:191–197. Wayne RK, Van Valkenburgh B, O’Brien SJ. 1991. Molecular Stoen OG, Zedrosser A, Saebo S, Swenson JE. 2006. Inversely distance and divergence time in carnivores and primates. Mol density-dependent natal dispersal in brown bears Ursus Biol Evol. 8:297–319. arctos. Oecologia. 148:356–364. Wolfe KH, Li WH. 2003. Molecular evolution meets the Swofford DL. 2002. PAUP*: phylogenetic analysis using genomics revolution. Nat Genet. 33(Suppl):255–265. parsimony (* and other methods). Version 4.0b10. Sunderland Yu L, Li QW, Ryder OA, Zhang YP. 2004. Phylogeny of the (MA): Sinauer. bears (Ursidae) based on nuclear and mitochondrial genes. Tajima F, Nei M. 1984. Estimation of evolutionary distance Mol Phylogenet Evol. 32:480–494. between nucleotide sequences. Mol Biol Evol. 1:269–285. Yu L, Li YW, Ryder OA, Zhang YP. 2007. Analysis of complete Talbot SL, Shields GF. 1996a. A phylogeny of the bears mitochondrial genome sequences increases phylogenetic (Ursidae) inferred from complete sequences of three mito- resolution of bears (Ursidae), a mammalian family that chondrial genes. Mol Phylogenet Evol. 5:567–575. experienced rapid speciation. BMC Evol Biol. 7:198. Downloaded from https://academic.oup.com/mbe/article/25/7/1344/1046167 by guest on 23 September 2021 Talbot SL, Shields GF. 1996b. Phylogeography of brown bears (Ursus arctos) of Alaska and paraphyly within the Ursidae. Mol Phylogenet Evol. 5:477–494. John H McDonald, Associate Editor Waits LP, Sullivan J, O’Brien SJ, Ward RH. 1999. Rapid radiation events in the family Ursidae indicated by likelihood Accepted March 26, 2008