scieee Open visual document viewer

Comparative genomics reveals the hybrid origin of a macaque group

Zhang, Bao-Lin,Chen, Wu,Wang, Zefu,Pang, Wei,Luo, Meng-Ting,Wang, Sheng,Shao, Yong,He, Wen-Qiang,Deng, Yuan,Zhou, Long,Chen, Jiawei,Yang, Min-Min,Wu, Yajiang,Wang, Lu,Fernández-Bellón, Hugo,Molloy, Sandra,Meunier, Hélène,Wanert, Fanélie,Kuderna, Lukas F.

Abstract

© 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution NonCommercial License 4.0 (CC BY-NC).

Full text

GENETICS Compa a i e genomics e eals he hyb id o igin o a macaque g oup Bao-Lin Zhang 1 †, Wu Chen 2 †, Ze u Wang 3,4 †, Wei Pang 5 , Meng-Ting Luo 5 , Sheng Wang 1 , Yong Shao 1 , Wen-Qiang He 5 , Yuan Deng 6,7 , Long Zhou 8 , Jiawei Chen 6 , Min-Min Yang 1 , Yajiang Wu 2 , Lu Wang 9 , Hugo Fe nández-Bellon 10 , Sand a Molloy 11 , Hélène Meunie 12,13 , Fanélie Wane 14 , Lukas Kude na 15 , Tomas Ma ques-Bone 16,17,18,19 , Ch is ian Roos 20,21 , Xiao-Guang Qi 9 , Ming Li 22 , Zhijin Liu 23 , Mikkel Heide Schie up 24 , Da id N. Coope 25 , Jianquan Liu 3,26 , Yong-Tang Zheng 5,27 *, Guojie Zhang 1,8,28,30 *, Dong-Dong Wu 1,27,29,31 * Al hough species can a ise h ough hyb idiza ion, compelling e idence o hyb id specia ion has been epo ed only a ely in animals. He e, we p esen phylogenomic analyses on genomes om 12 macaque species and show ha he ascicula is g oup o igina ed om an ancien hyb idiza ion be ween he sinica and silenus g oups ~3.45 o 3.56 million yea s ago. The X ch omosomes and low- ecombina ion egions exhibi ed equal con ibu ions om each pa en al lineage, sugges ing ha hey we e less a ec ed by subsequen backc ossing and hence could ha e played an impo an ole in main aining hyb id in eg i y. We iden i ied many ep oduc ion-associ- a ed genes ha could ha e con ibu ed o he de elopmen o he mixed sexual pheno ypes cha ac e is ic o he ascicula is g oup. The phylogeny wi hin he silenus g oup was also esol ed, and unc ional expe imen a- ion con i med ha all ex an Wes e n silenus species a e suscep ible o HIV-1 in ec ion. Ou s udy p o ides no el insigh s in o macaque e olu ion and e eals a hyb id specia ion e en ha has occu ed only e y a ely in p ima es. Copy igh © 2023 The Au ho s, some igh s ese ed; exclusi e licensee Ame ican Associa ion o he Ad ancemen o Science. No claim o o iginal U.S. Go e nmen Wo ks. Dis ibu ed unde a C ea i e Commons A ibu ion NonComme cial License 4.0 (CC BY-NC). INTRODUCTION In e speci ic hyb idiza ion can acili a e species adap a ion by in o- ducing new gene ic ma e ial and no el allelic combina ions (1). In some ci cums ances, hyb idiza ion may esul in he nea -ins an a- neous o ma ion o new species, i.e., hyb id specia ion, he eby p o- mo ing biodi e si y (2). Hyb id specia ion has been an impo an mode o specia ion du ing plan e olu ion (3). Howe e , hyb id specia ion has been conside ed o occu only a ely in animals owing o he ac ha he de i ed species om in e speci ic hyb id- iza ion ypically end o be e olu iona ily less i (e.g., Dobzhansky- Mulle incompa ibili y) o a e only weakly ep oduc i ely isola ed om hei pa en al species and hence could be swamped by one o hei pa en s (4). Bo h heo e ical and empi ical s udies sugges ha hose species g oups ha ha e expe ienced apid adap i e adia ion a e p one o hyb idiza ion because he incompa ibili ies be ween hese species a e weak (1). Hyb id lineages a e mo e eadily es ablished when hyb ids ma e asso a i ely wi h each o he (5) and/o when hey became adap ed o a new en i onmen (6). An inc easing numbe o empi ical s udies ha e documen ed he es ablishmen o hyb id lineages in apidly adia ing g oups, including bu e lies, canids, baboon, bi ds, and bea s (7–11), sugges ing ha hyb id specia ion may be mo e common and hence mo e impo an han p e iously hough . Ne e heless, conc e e e idence o hyb id specia ion in animals emains sca ce and pu a i e examples a e o en highly con- en ious (12,13), impeding ou unde s anding o he genomic mechanisms unde lying hyb id specia ion. In pa icula , i is o en challenging o eliably di e en ia e be ween hyb id specia ion and pos -specia ion gene ic in og ession (4), which may be u he complica ed by he s ochas ic so ing o ances al polymo phisms among descendan lineages, a p ocess known as incomple e lineage so ing (ILS) (14). Thus, a compa a i e phylogenomic amewo k o unequi ocally demons a e hyb id specia ion has ye o be es ablished. The genus o macaques (Macaca) ep esen s an excellen model in which o s udy he in e play be ween in e speci ic hyb idiza ion and specia ion. As one o he mos success ul p ima e lineages, ma- caques now include 23 species ha a e widely dis ibu ed ac oss Sou h, Eas , and Sou heas Asia, he only excep ion being M. syl a- nus, which is con ined o he A las moun ains in No h A ica (15). Paleon ological and molecula da a sugges ha macaques o igina - ed in No h A ica ~7 million yea s (Ma) ago and hen expe ienced a bu s o specia ion in Asia du ing he pas 5 Ma (16,17). Some macaque species a e widely used as labo a o y models o s udying human disease and o accine de elopmen (18). Al hough membe s o his genus a e mo phologically and beha io ally dis- inc , in e speci ic hyb idiza ion is po en ially possible be ween any geog aphically o e lapping pai o species because hei ep o- duc i e isola ion is likely o be incomple e. This no ion is suppo ed by bo h ield obse a ions (19,20) and molecula s udies (21,22). While some macaque species (e.g., he ascicula is g oup o ma- caques and he s ump- ailed macaque) exhibi dis inc i e mixed pheno ypes, which could, in p inciple, ha e a isen by genomic ad- mix u e (23,24), he b oade ole o hyb idiza ion and i s speci ic ou comes emain la gely unknown. He e, we pe o m mul iple genome analyses o asce ain species phylogeny in he macaque genus, assess he ole o in e speci ic hyb idiza ion in specia ion, and explo e he gene ic basis o mixed pheno ypes in macaque species. RESULTS We gene a ed 10 high-quali y macaque genome assemblies ( ables S1 o S3) using he long- ead sequencing s a egy o Nanopo e and SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 1 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 long agmen ead echnologies (s LFR and 10X Genomics), e- spec i ely. Sca old N50 sizes o hese newly assembled genomes anged om 17.7 o 33.2 Mb, while he Benchma king Uni e sal Single-Copy O hologs (BUSCO) comple eness sco es anged om ~93.3 o 94.5%. The inal anno a ed gene numbe s anged om 20,662 o 21,811 be ween di e en species using a combina- ion o ab ini io and homology-based gene p edic ion app oaches. We also included wo p e iously published genomes o M. mula a (25) and M. nemes ina, and one ou g oup species Papio hamad yas (26). Thus, ou genome assemblies co e all known macaque species g oups (Fig. 1A) (17). We i s pe o med pai wise whole-genome alignmen s agains he Chinese hesus macaque (M. mula a, heMacS) genome using he LASTZ p og am (27) and hen me ged all o hem in o mul iple genome alignmen s wi h MULTIZ (28). These genomes showed high collinea i y, wi h he aligned base pai s spanning mo e han 2.7 Gb o 90% o he e e ence assembly ( ig. S1 and able S5). To deduce he co ec b anching pa e n o he sequenced macaque species and o cha ac e ize he he e ogeneous phylogene ic signals ac oss genomes, we pa i ioned he genome alignmen s in o non- o e lapping windows o 50 kb and hen pe o med maximum like- lihood (ML) analyses o each window sequence. We ini ially ob ained he mul ispecies coalescen -based species ee (ASTRAL and STAR; ig. S3) om windows wi h high boo s ap alues (e.g., mean alue o >80%). Al hough he opologies ob ained om hese wo me hods we e mu ually consis en and la gely in acco d wi h ecen s udies (17,21), we obse ed subs an ial genealogical dis- co dance wi hin and be ween species g oups among all window ees (n= 7392 al e na i e opologies; Fig. 1, B and C, and ig. S5), mos ly pe aining o he phylogene ic posi ion o he ascicula- is g oup wi h espec o he sinica and silenus g oups. The disco d- ance o his posi ion was no a i ial consequence o he choice o window size because analyses using ei he smalle (20 kb) o la ge (100 kb) window sizes yielded b oadly simila esul s ( igs. S6 o S9). To be e unde s and he in e ela ionship o macaque g oups wi h incong uen phylogenies, we p uned he ees so as o include i e ing oup species, each ep esen ing a pa icula lineage p oposed by p e ious s udies (16,17). Two opologies we e ound o domina e: One suppo ed a sis e ela ionship be ween he ascicula is and sinica g oups (T1), whe eas he o he (T2) a o ed he ascicula is g oup as a sis e lineage o he silenus g oup. These wo ee opol- ogies we e suppo ed by a o al o 78% o windows in e lea edly dis- ibu ed ac oss he genome, wi h 47 and 31% e e ing o T1 and T2, espec i ely (Fig. 2, A and B). Speci ically, we ound ha he X ch o- mosome, whe e genomic incompa ibili ies no mally i s de elop du ing specia ion (29), exhibi ed a nea ly equal p opo ion o T1 and T2 opologies e sus he au osomes i espec i e o window size ( ig. S10). Such widesp ead mixed ances y o he au osomes and he X ch omosome led us o specula e ha he ascicula is g oup may ha e o igina ed om an ancien hyb idiza ion be ween he p ogeni o s o he sinica g oup and hose o he silenus g oup. I he hyb id o igina ion hypo hesis is co ec , we would expec o obse e an equal le el o sequence di e gence be ween he hyb id species and i s wo pa en al lineages. To es his pos ula e, we com- pa ed he sequence di e gence (D XY ) and ela i e di e gence ime be ween ascicula is e sus sinica and ascicula is e sus silenus o he au osomes and X ch omosome sepa a ely o allow o hei di - e en a es o e olu ion (30). We ound no ob ious di e ence in hese wo s a is ics calcula ed using only he sex ch omosome da a (bo h P alues > 0.05, Wilcoxon’s es ; Fig. 2, C and D). Wi h he au osomes, al hough bo h pa ame e s we e s a is ically signi ican (P< 0.001, Wilcoxon’s es ; Fig. 2, C and D), he mean D XY alue (0.1616 e sus 0.1631) and he mean es ima ed age (3.45 Ma e sus 3.56 Ma) be ween he ascicula is g oup and i s wo p ogeni o s we e only sligh ly di e en . The obse ed pa e n he e o e concu s wi h ha expec ed unde he hyb id specia ion scena io. Howe e , ex ensi e pos -specia ion gene low may also ha e p o- duced a simila signa u e (31). To es his possibili y, we pe o med D-s a is ic analyses in a sliding window (32). I ei he T1 o T2 ep- esen he ue species phylogeny, he windows leas a ec ed by gene low (absolu e D-s a is ic alues close o ze o) would o e whelming- ly suppo one o he es ed ees, whe eas hyb idiza ion would suppo bo h (33). Ou esul s suppo ed he la e p edic ion showing ha he 1% o windows wi h he lowes absolu e D-s a is ic alues suppo ed T1 and T2, espec i ely, when using T1 and T2 as 1 S a e Key Labo a o y o Gene ic Resou ces and E olu ion, Kunming Ins i u e o Zoology, Chinese Academy o Sciences, Kunming 650223, China. 2 Guangzhou Zoo and Guangzhou Wildli e Resea ch Cen e , Guangzhou 510070, China. 3 Key Labo a o y o Bio- esou ce and Eco-en i onmen o Minis y o Educa ion, College o Li e Sciences, Sichuan Uni e si y, Chengdu 610065, China. 4 Co-Inno a ion Cen e o Sus ainable Fo es y in Sou he n China, College o Biology and he En i onmen , Nanjing Fo es y Uni e si y, Nanjing 210037, China. 5 Key Labo a o y o Animal Models and Human Disease Mechanisms o he Chinese Academy o Sciences, KIZ-CUHK Join Labo a o y o Bio esou ces and Molecula Resea ch in Common Diseases, Kunming Ins i u e o Zoology, Chinese Academy o Sciences, Kunming 650223, China. 6 BGI-Shenzhen, Shen- zhen 518083, China. 7 Sec ion o Ecology and E olu ion, Depa men o Biology, Uni e si yo Copenhagen, Copenhagen DK-2100, Denma k. 8 Cen e o E olu iona y and O ganismal Biology and Women’s Hospi al a Zhejiang Uni e si y School o Medicine, Hangzhou 310058, China. 9 Shaanxi Key Labo a o y o Animal Conse a ion, College o Li e Sciences, No hwes Uni e si y, Xi’an, China. 10 Ba celona Zoo, Pa c de La Ciu adella, Ba celona 08003, Spain. 11 Dublin Zoo, Dublin 8, I eland. 12 Cen e de P ima ologie, de l’Uni e si é de S asbou g, Niede hausbe gen, F ance. 13 Labo a oi e de Neu osciences Cogni i es e Adap a i es, UMR 7364, Uni e si é de S as- bou g, S asbou g, F ance. 14 Pla e o me SILABE, Uni e si é de S asbou g, Niede hausbe gen, F ance. 15 Genome In e p e a ion Depa men , Illumina Inc., Fos e Ci y, CA, USA. 16 Ins i u e o E olu iona y Biology (UPF-CSIC), PRBB, D . Aiguade 88, Ba celona 08003, Spain. 17 Ca alan Ins i u ion o Resea ch and Ad anced S udies (ICREA), Passeig de Lluís Companys, 23, Ba celona 08010, Spain. 18 CNAG-CRG, Cen e o Genomic Regula ion (CRG), Ba celona Ins i u e o Science and Technology (BIST), Baldi i i Reixac 4, Ba celona 08028, Spain. 19 Ins i u Ca alà de Paleon ologia Miquel C usa on , Uni e si a Au ònoma de Ba celona, Edi ici ICTA-ICP, c/Columnes s/n, 08193 Ce - danyola del Vallès, Ba celona, Spain. 20 P ima e Gene ics Labo a o y, Ge man P ima e Cen e , Gö ingen, Ge many. 21 Gene Bank o P ima es, Ge man P ima e Cen e , Gö ingen, Ge many. 22 CAS Key Labo a o y o Animal Ecology and Conse a ion Biology, Ins i u e o Zoology, Chinese Academy o Sciences, Beijing 100101, China. 23 College o Li e Sciences, Capi al No mal Uni e si y, Beijing 100048, China. 24 Bioin o ma ics Resea ch Cen e, Aa hus Uni e si y, Aa hus C DK-8000, Denma k. 25 Ins i u e o Medical Gene ics, School o Medicine, Ca di Uni e si y, Ca di CF14 4XN, UK. 26 S a e Key Labo a o y o G assland Ag o-ecosys em, Ins i u e o Inno a ion Ecology and College o Li e Sciences, Lanzhou Uni e si y, Lanzhou 730000, China. 27 Na ional Resou ce Cen e o Non-Human P ima es, Kunming P ima e Resea ch Cen e and Na ional Resea ch Facili y o Pheno ypic and Gene ic Analysis o Model Animals (P ima e Facili y), Kunming Ins i u e o Zoology, Chinese Academy o Sci- ences, Kunming, Yunnan 650107, China. 28 Liangzhu Labo a o y, Zhejiang Uni e si y Medical Cen e , 1369 Wes Wenyi Road, Hangzhou 311121, China. 29 Cen e o Ex- cellence in Animal E olu ion and Gene ics, Chinese Academy o Sciences, Kunming 650223, China. 30 Villum Cen e o Biodi e si y Genomics, Sec ion o Ecology and E olu ion, Depa men o Biology, Uni e si y o Copenhagen, Copenhagen 2100, Denma k. 31 Kunming Na u al His o y Museum o Zoology, Kunming Ins i u e o Zoology, Chinese Academy o Sciences, Kunming, Yunnan 650223, China. *Co esponding au ho . Email: [email p o ec ed] (D.-D.W.); [email p o ec ed] (G.Z.); [email p o ec ed] (Y.-T.Z.) †These au ho s con ibu ed equally o his wo k. SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 2 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 he es ed opologies (Fig. 2, E and F). Fu he mo e, on he basis ha egions o lowe ecombina ion a e “cold spo s” o in og ession (29), we classi ied he ees by ecombina ion a e and ound ha T1 and T2 we e s ill dominan in he low ecombina ion a e egions o 0 o 0.1 cM/Mb ( ig. S11). This obse a ion was e en mo e e iden wi h a smalle window size (20 kb), whe e he a io o T1 and T2 was almos iden ical (T1:T2 = 32%:31%; ig. S13), p obably due o he a e aging e ec wi h he la ge window size. These indings co obo a e he iew ha hyb id specia ion, a he han pos -specia ion gene low, was esponsible o he pa e n o genomic mosaicism. Ano he unde lying assump ion o hyb id specia ion is ha all he indi iduals om he hyb id species a e uni o mly admixed. Howe e , i ancien di ec ional gene low and ILS we e o ha e oc- cu ed, i is likely ha no all indi iduals in he hyb id g oup would ha e con ained he same amoun and leng h dis ibu ion o in o- g essed alleles because subsequen ecombina ion and nega i e se- lec ion would ha e ac ed so as o pu ge he dele e ious in ude Fig. 1. Dis ibu ion map and he disco dance o phylogeny. (A) Dis ibu ion map o macaque species used in his s udy. Nomencla u e ollows Delson (16) in de ining he species g oup. We u he spli he silenus g oup in o Wes e n (W) and Eas e n (E) o ms on he basis o hei simila di e gence o o he species g oups. Macaque d awings a e copy igh , 2013, S ephen D. Nash, In e na ional Union o Conse a ion o Na u e Species Su i al Commission P ima e Specialis G oup and a e used wi h pe mission. All species excep M. mula a and M. nemes ina we e newly sequenced in his s udy. (B) DensiT ee plo o 50-kb window ees. Blue, ed, and g een colo s ep esen he i s , second, and hi d mos common opologies, espec i ely, whe eas g ay ep esen s o he opologies. (C) Majo i y- ule consensus ee o all 50-kb window ees (n= 46,003). The numbe s abo e he b anches indica e he absolu e numbe o opologies suppo ing he spli s, whe eas he numbe s below ep esen he pe cen age alues. SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 3 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 alleles om he accep o genome du ing a long pe iod o di e gence (34). To es his p edic ion, we used HyDe (35) and LOTER (36) o quan i y he genomic con ibu ions o he wo pa en al lineages (sinica and silenus) o each o he hyb id species in he ascicula is g oup. The esul s we e consis en wi h ou expec a ion unde he hyb idiza ion hypo hesis, showing ha all h ee species om he ascicula is g oup we e uni o mly mixed and exhibi ed a gene al ex- ponen ial decay o he ac size (Fig. 3, A and B, and able S7), i - espec i e o whe he he species a e geog aphically isola ed (M. usca a) o widely dis ibu ed (M. mula a and M. ascicula is). This esul was u he co obo a ed by PhyloNe -MPL analyses, an al e na i e me hod ha allows o ILS and hyb idiza ion simul- aneously, based on he maximum pseudo-likelihood me hod (37), whe e he sec ion o he ascicula is g oup was in a iably iden i ied as a e icula e node in he scena ios allowing one and wo pas hy- b idiza ion e en s (Fig. 4A and ig. S17). Toge he , hese phylogenomic analyses concu in e ms o p o- iding consis en suppo o he hyb id o igin hypo hesis o he Fig. 2. Phylogene ic ela ionship o he i e majo macaque lineages ac oss he genome. (A) Dis ibu ion o he h ee mos common opologies by e e ence o he Chinese hesus macaque (M. mula a) genome. Colo ed bands ep esen ee opologies o each 50-kb window. Whi e in e al egions deno e missing da a. (B) Th ee mos common ees (T1 o T3) eco e ed by ML analysis and hei equency in consecu i e 50-kb windows. Values below he ee e e o he pe cen age o windows eco e ing ha opology. The ou g oup P. hamad yas is no shown. (C) Gene ic di e gence (D XY ) and (D) es ima ed di e gence imes om MCMCTREE be ween he ascicula is g oup and i s wo pu a i e pa en al lineages (sinica and silenus) based on au osomal 50-kb window sequences (le ) and he X ch omosome ( igh ), espec- i ely. P alues we e es ima ed by he Wilcoxon ank sum es . The subs i u ion a es es ima ed om MCMCTREE o he au osomes and X ch omosome we e 1.2 × 10 −9 pe si e pe yea and 0.9 × 10 −9 pe si e pe yea , espec i ely. (E) DensiT ee plo o ML ees de i ed om he 1% windows wi h he lowes absolu e D-s a is ic alues using T1, and (F) T2, as he es ed ela ionship. The alues below show he pe cen age o windows ha eco e ed he opology. SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 4 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 ascicula is g oup bu we e no in o ma i e ega ding i s di ec ion in he ini ial hyb idiza ion. Because he mi ochond ial genome and he Y ch omosome a e ep esen a i e o he ma e nal and pa e nal lineages, espec i ely, we nex pe o med phylogene ic analyses based on mi ochond ial and Y ch omosome sequences (~480 kb in leng h) o de e mine he hyb idiza ion pa e n. We ound a clea opology disc epancy be ween he mi ochond ial and he Y ch omosomal da a, wi h he ascicula is g oup ha ing a mi ochon- d ial genome simila o ha o he sinica g oup and a Y ch omo- some simila o ha o he silenus g oup (Fig. 4, B and C), sugges ing ha he ancien hyb idiza ion occu ed p edominan ly o exclusi ely be ween p o o-sinica g oup emales and p o o- silenus g oup males. Wi h his esol ed phylogeny, we u he in e ed di ec ional gene low e en s be ween macaque g oups a e he hyb id o igin o he ascicula is g oup. The esul s ob ained e ealed a complex ne wo k o ances al admix u e in Macaca (Fig. 4A and ig. S20). I is in e es ing o no e ha a e he hyb id o igin, a ious species o he ascicula is g oup main ained gene low wi h he de- scendan species o he wo pa en al g oups. Among hem, he mos no able di ec ion o gene low occu ed be ween he s ump- ailed macaque (M. a c oides) and he common ances o o mula a/ usca a. On he basis o disc epancies be ween mi ochond ial and Y ch omosomal opologies, i has been p e iously p oposed ha he s ump- ailed macaque o igina ed om hyb idiza ion be ween sinica and mula a/ usca a macaques (22). Howe e , ou analyses a e consis en wi h ano he phylogenomic s udy (38) ha indica ed he s ump- ailed macaque o be a membe o he sinica g oup wi h a low le el o in og ession om mula a/ usca a (0.05 ≤γ≤0.09; able S7). In he phylogene ic ees based on low ecombina ion a e egions and he X ch omosome ( igs. S14 and S15), he s ump- ailed macaque also clus e ed wi h he sinica g oup, sugges - ing ha mi ochond ial in og ession om mula a/ usca a ma- caques occu ed a e he ini ial specia ion. Ou phylogeny and in og ession ne wo k also e ealed a b anching pa e n wi hin he silenus g oup ha con adic s he p e- ious mo phological hypo hesis, which sugges ed ha he no h- e n- ailed macaque (M. leonina) and sou he n- ailed macaque (M. nemes ina) a e sis e species (39). Al hough hese wo species a e close ecologically and geog aphically and a e also simila in pheno ypic appea ance (Fig. 1A), all ou phylogenomic analyses wi h di e en genomic da a ypes suppo he pos ula e ha he lion- ailed macaque (M. silenus) om he Sou h Indian Wes e n Gha s and M. leonina a e sis e species (Fig. 4, A o C, and igs. S14 and S15). Ou p e ious unde s anding o he phylog- eny o he silenus g oup was also con used by he no ion ha M. leonina and M. nemes ina we e he only wo Old Wo ld p ima es known o be suscep ible o HIV-1 (human immunode iciency i us ype 1) in ec ion due o he highly unusual e o ansposi ional in- se ion o a cyclophilin A2 (CypA2) gene in o he 3′un ansla ed egion o he TRIM5 locus (40,41). On he basis o he o hologous sequence, we iden i ied he same TRIM5-CypA2 usion gene in M. silenus ( ig. S25), sugges ing ha his e o ansposi ional e en e ol ed in he common ances o o he Wes e n silenus g oup (W) a leas 2.17 Ma (Fig. 4A). In i o in ec ion expe imen s using pe iphe al blood mononuclea cells (PBMCs) con i med ha M. silenus can be in ec ed wi h HIV-1 (Fig. 4D). The success o o he wise o a gi en hyb id specia ion depends upon whe he he hyb idiza ion e en s gi e ise o an es ablished, pe sis en , mo phologically and ecologically dis inc hyb id lineage (42). Today, he species o he ascicula is g oup a e well seg- ega ed om hei wo pa en al g oups by ei he ecogeog aphic ba - ie s o beha io al di e ences (43). The ex an membe s o he ascicula is g oup, howe e , display a dis inc i e mix u e o hei pa- en al species’cha ac e is ics as a esul o he ancien hyb idiza ion. Fo example, he unique blun ly bilobed and na ow penile mo - phology in males, and he b igh ed sexual skin wi h li le o no e iden swelling in emales, a e app oxima ely in e media e be ween he sinica and silenus g oups (Fig. 5D) (23,43). As hese mo phological ai s a e mos ly associa ed wi h he ep oduc i e sys em and o en ha e a polygenic basis, i is easonable o expec ha genes ela ed o hese ai s should also exhibi a mosaic pa e n wi h espec o he pu a i e pa en al lineages. To iden i y he gene ic mechanisms unde lying hese mixed pheno ypes, we used a ecen ly de eloped me hod (44) o de ec genes ha ha e been subjec o posi i e selec ion in he ascicula is Fig. 3. Genomic ances y inhe en in wo pa en al lineages. (A) Densi y plo o es ima ed γ alues ac oss 500 boo s ap eplica es in HyDe analyses. γ ep esen s he es ima ed p obabili y o inhe i ance om he ances o o he sinica g oup, whe eas 1-γ ep esen s he p obabili yo inhe i ance om he silenus g oup. The mean γ alues o he h ee hyb id species a e gi en op igh . Mmul,M. mula a;M as,M. ascicula is;M us,M. usca a. (B) Dis ibu ion o ac sizes (in 5-kb bins) o sinica g oup ances y in h ee hyb id species. SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 5 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 g oup and each o i s pa en al species. In addi ion, we applied h ee c i e ia o il e hose posi i ely selec ed genes (PSGs) o he pu pose o minimizing he po en ial bias o he small sample size in each g oup (see Ma e ials and Me hods o de ails). We iden i ied 216 PSGs in he ascicula is g oup ha we e inhe i ed om he sinica lineage ( able S14). Func ional anno a ion iden i ied 22 genes as being o e ep esen ed in se e al gene egula o y ne wo ks ela ed o ep oduc i e unc ion, such as male game e gene a ion [Gene On ology (GO):0048232] and sexual ep oduc ion (GO:0019953) (P< 0.001; able S15). O hese genes, se en (UGT1A9,ADAM20,WFDC2,FANCF,DMRTC2,SIAH1, and POC1A) exhibi sinica-de i ed nonsynonymous subs i u ions ha ha e become ixed in known unc ional domain egions (Fig. 5B and ig. S23), whe eas ano he wo (WDR48 and PLPP1) ha e sinica-de i ed mu a ions in he 1-kb ups eam egion ( able S14). O special in e es is a gene in ol ed in la onoid glucu onida ion (UGT1A9); all ou sinica-de i ed nonsynonymous mu a ions we e ound o be loca ed in he UDPGT domain egion (Fig. 5B). The UGT1A9 gene encodes UDP glucu onosyl ans e ase, which ca a- lyzes he glucu onida ion o endogenous es ogen ho mones in o wa e -soluble exc e able me aboli es (45). Because es ogen is c ucial o he de elopmen o sexual swellings (46), i is ce ainly concei able ha he sha ed subs i u ions in his gene be ween as- cicula is and sinica may ha e con ibu ed o he educed sexual Fig. 4. Hyb id o igin o he ascicula is g oup. (A) Schema o he e icula ed e olu iona y ela ionships be ween macaque species, illus a ing he hyb idiza ion and majo admix u e e en s and app oxima e di e gence imes (Ma ± 1 SD) in e ed om au osomal windows. The solid ed line deno es he in e speci ic gene low ob ained om PhyloNe analyses, whe eas he g ay dashed line deno es he gene low in e ed om Dsui e. Fo he Dsui e esul s, we show he in e sec ion gene low e en s modeled om he wo mos common ees, i.e., one suppo ing he sis e ela ionship be ween he ascicula is and sinica g oups ( ig. S5A), he o he suppo ing he sis e ela ionship be ween he ascicula is and silenus g oups ( ig. S5B). (B) Mi ochond ial and (C) Y ch omosomal ees. Numbe s a nodes e e o boo s ap alues. M. on- keana2 and M. syl anus2 a e wo addi ional sequenced male samples. De ailed sample in o ma ion is shown in able S1. (D) Plasma i al load in PBMC aliquo s om wo lion- ailed macaques (Msil-001 and Msil-002) a e 48 and 72 hou s o in ec ion wi h HIV-1 NL4-3 i us. Two no he n pig- ailed macaques (NPM-15216 and NPM-17214) we e used as posi i e con ols. SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 6 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 swelling in emales o bo h g oups. In simila ein, he compa ison in ol ing ascicula is-silenus e sus he sinica g oup iden i ied 195 PSGs in he ascicula is g oup ha we e inhe i ed om he silenus lineage ( able S16). Func ional anno a ion e ealed 10 o be ela ed o human ep oduc i e pheno ypes (HP0000134: emale hypogo- nadism; HP:0000137: abno mali y o he o a ies, o a ian disease, P< 0.005; able S17). Fou o hem (PRLR,PTPN11,KCNU1, and MSH6) ha bo silenus-de i ed nonsynonymous mu a ions, ixed in unc ional domain egions (Fig. 5C and ig. S24), whe eas h ee (IZUMO4,BAZ1B, and SPIDR) ha e silenus-de i ed mu a ions all occu ing in 1-kb ups eam egions ( able S16). Toge he , hese esul s p o ide addi ional gene ic e idence o suppo he mosaic model o ep oduc i e mo phologies obse ed in he ascicula- is g oup. DISCUSSION The o igin o a new species om hyb idiza ion be ween wo p eex- is ing species is one o he mos spec acula modes o specia ion and has consequen ly a ac ed he a en ion o e olu iona y biologis s o decades (47,48). Al hough adi ionally conside ed o be mo e p e alen in plan s, hyb id specia ion has in ecen yea s been ecognized as being mo e common in animals han p e iously hough , including examples om bo h in e eb a es and e e- b a es (49). Howe e , con incing e idence o hyb id specia ion in animals emains sca ce, and ou unde s anding o he unde lying genomic mechanisms is s ill qui e limi ed. He e, ou phylogenomic analyses using di e en da a ypes and me hods as well as mo pho- logical e idence, ha e p o ided consis en suppo o he hypo h- esis ha he ascicula is g oup o macaques o igina ed om an ancien hyb idiza ion be ween he sinica and silenus g oups, he eby p o iding us wi h an unpa alleled oppo uni y o Fig. 5. The gene ic basis unde lying he mix u e o pheno ypes in he ascicula is g oup macaques. (A) Genomic dis ibu ion o PSGs iden i ied in he ascicula is g oup. Only PSGs associa ed wi h ep oduc i e unc ions a e depic ed. (Band C) a e wo examples o ep oduc ion- ela ed PSGs wi h he la ges numbe o ixed non- synonymous mu a ions ha we e inhe i ed om he sinica and silenus g oups, espec i ely. The posi ion o he mu a ion and p edic ed unc ional domain a e bo h shown. (D) Diag am showing he h ee-pole mo phocline in macaque (23,24). Species o he ascicula is g oup gene ally exhibi a mix u e o glans penis and sexual skin mo phology om hei pa en al species g oups (silenus and sinica), al hough he e is a ia ion wi hin he g oup. Examples o species a e M. nemes ina (silenus g oup), M. mula a ( ascicula is g oup), and M. hibe ana (sinica g oup). Macaque d awings by J. Shi. SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 7 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 in es iga e he ac o s d i ing he o igin and main enance o hyb id species in highe animals. On he basis o he age es ima es de i ed om he pu a i ely neu al sequences o au osomal windows (Fig. 4A), ou esul s sugges ha he ancien hyb id o ma ion o he ascicula is g oup occu ed ~3.45 o 3.56 Ma, soon a e he ini ial sepa a ion o he wo pa en al lineages (p o o-sinica and p o o-silenus) ~3.86 Ma. The di e gence ime o he wo pa en al lineages and subsequen hyb idiza ion coincided wi h he apid glacio-eus a ic luc ua ions in he ea ly-middle Pliocene (50). As a e y dynamic geog aphical egion in Asia, he land b idge (Is hmus o K a) be ween he Malay peninsula and he Sunda egion was epea edly a ec ed by glacio- eus a ic luc ua ions du ing he ea ly-middle Pliocene. We he e o e specula e ha he ea ly sea le el highs and (when he sea le el was abo e he edge o he con inen al shel ) may ha e led o he ini ial sepa a ion o p o o-sinica and p o o-silenus, while he subsequen lowe ing o he sea le el acili a ed he seconda y con ac equi ed o hyb idiza ion. Howe e , because no ob ious pos zygo ic isola- ion mechanisms we e obse ed among macaque species, he mos insu moun able di icul ies would be how hyb id in eg i y was main ained and ein o ced a e he ini ial hyb idiza ion. P ezygo ic ba ie s, such as ecological di e gence o /and geog aphic isola ion, a e hus likely o ha e been o g ea e impo ance in he es ablish- men o hyb id axa. We p opose ha geog aphic isola ion may no be a necessa y equi emen o hyb id specia ion because o he con inual o e lapping dis ibu ion be ween membe s o he ascicu- la is g oup o macaques (e.g., M. mula a and M. ascicula is) and hei pa en al g oup species (Fig. 1A). Ne e heless, he ascicula is g oup macaques a e now well seg ega ed om hei wo pa en al species by ecological ba ie s and/o beha io al di e ences (43, 51). In pa icula , he ascicula is g oup macaques show dis inc i e mixed sexual pheno ypes, and he oo p in s o na u al selec ion ha e been de ec ed in hese cha ac e is ics, sugges ing ha asso a- i e ma ing beha io may ha e played an impo an ole in he o igin and main enance o he ascicula is g oup as a hyb id species. In p ima es, p ima y and seconda y sexual cha ac e is ics a e o en he a ge o male- emale ma e choice/sexual selec ion (52,53). Such no el combina ions o sexually selec ed ai s in he ascicula is g oup may ha e led o he eme gence o no el ma e p e - e ences, which could ha e p omo ed he es ablishmen o a dis- inc lineage. We u he examine he genomic makeups om he o igina ion o a new species as he consequence o hyb idiza ion and how he subsequen genomic changes in he hyb id lineage acili a ed he es- ablishmen o ep oduc i e ba ie s owa d bo h pa en al species. We ound ha he sex ch omosomes (X) and egions o low ecom- bina ion display mo e conspicuous pa e ns o mosaicism. Gene al- ly, hese egions a e conside ed o be less pe meable o in og ession as hey a e less pe missi e o o eign genes due o lowe hyb id i ness (54). Howe e , species ha ha e expe ienced apid adap i e adia ion p o ide jus such an oppo uni y because hese incompa - ible egions in he descendan lineages a e likely o be less gene ical- ly di e gen om each o he , allowing hem o hyb idize and p oduce iable o sp ing (29). In simila ein, hese less pe meable egions could also se e as a s ong ba ie agains u he pa en al in og ession in hyb ids and hence would be expec ed o exhibi mo e mosaicism han o he au osomal egions once hyb idiza ion occu ed (55,56). Las , we ha e esol ed se e al long-s anding e olu iona y co- nund ums du ing he apid specia ion o macaques, in ol ing con- o e sies su ounding phylogeny and complex ances al hyb idiza ions. Fo example, ou analyses con i m ha he s ump- ailed macaque is a membe o he sinica g oup wi h a low le el o in og ession om mula a/ usca a, and ha he lion- ailed macaque (M. silenus) and M. leonina a e sis e species, bo h o which can be in ec ed by HIV. Conside ing he high species di e - si y o macaques and he s ong lineage s uc u e wi hin some species, genome sequencing da a om all macaque species and di - e en subspecies will undoub edly help o u he elucida e he complex e olu iona y his o y and biogeog aphy o his genus. We a e, howe e , con iden ha ou conclusions ega ding he hyb id specia ion o he ascicula is g oup will no change wi h he acqui- si ion o new sequence da a. In summa y, ou s udy epo s he occu ence o an unusual ancien hyb idiza ion e en in p ima es and illus a es how specia- ion h ough na u al hyb idiza ion can a ise ia he eshu ling o s anding gene ic a ia ion and how hyb id species may main ain hei gene ic in eg i y h ough he ac ion o selec ion. Ou s udy p o ides bo h a s a egy and a pipeline o genome analyses o iden- i y hyb id specia ion, which should pa e he way o he iden i ica- ion and explo a ion o u he such e en s in he u u e. MATERIALS AND METHODS Sample in o ma ion and e hics s a emen A o al o 12 macaque blood o DNA samples we e ob ained o his p ojec . Ten we e used o ull-genome assembly, whe eas samples om he o he wo males, om M. onkeana and M. syl anus, e- spec i ely, we e used o whole-genome sho gun esequencing. These samples we e collec ed om mul iple sou ces in China and Ge many. De ailed in o ma ion is gi en in able S1. All samples we e collec ed legally and in acco dance wi h he policy o he Animal Ca e and Use E hics o Kunming Ins i u e o Zoology (ap- p o al ID: SMKX-20180701-01 and SMKX-2021-01-002), which con o ms o he egula o y s anda ds o he human ca e and ea - men o animals in esea ch. Genome sequencing, assembly, and anno a ion The 10 newly assembled macaque genomes we e sequenced on di - e en pla o ms due o he p o ac ed p ocess o collec ing issue samples (see able S1 o de ails). We used Supe no a ( e sion 2.0.0, 10X Genomics Inc., Pleasan on, CA, USA) wi h de aul pa- ame e s o assemble he eads gene a ed om he s LFR and 10X Genomics Ch omium eads pla o ms. Fo he Nanopo e long eads, we i s pe o med sel -e o co ec ion o all Nanopo e long eads using Nex Deno o so wa e ( 2.4.0) and hen assembled hem in o con igs using w dbg-1.2.8 (57). The aw assemblies we e u he polished by Illumina sho eads in Pilon 1.22 (58) h ee imes unde he de aul se ings. The comple eness o he new macaque genomes was assessed by BUSCO ( 3.0.2) (59) based on he mammal-speci ic se o 4104 single-copy o hologs (mammalia_odb9). Repea elemen s we e p edic ed by Repea Maske 4.0.6 (60). Fo p o ein-coding gene anno a ion, we i s ob ained p o ein se- quences om i e well-anno a ed mammalian species, namely, human (Homo sapiens, GCA_000001405.28), chimpanzee (Pan oglody es, GCA_002880755.3), go illa (Go illa go illa, SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 8 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024 GCA_900006655.3), o angu an (Pongo abelii, GCA_002880775.3), and mouse (Mus musculus, GCA_000001635.8). These p o ein se- quences we e mapped o each de no o genome using TBLASTN 2.2.26 (61) wi h an E alue cu o o 1 × 10 −5 . P o eins wi h mul- iple adjacen hi s we e connec ed o each o he using genBlas A 1.0.4 (62). We il e ed ou hose candida e loci wi h homologous block leng hs sho e han 30% o he leng h o he que y p o ein. Fo he ab ini io p edic ion, we used Augus us 3.0.3 (63) wi h op- imized pa ame e s ained om 1000 andomly selec ed homolo- gous genes. Las , we in eg a ed all gene se s o o m a comp ehensi e and non edundan gene se using in-house Pe l sc ip s. Whole-genome alignmen s We i s pe o med pai wise whole-genome alignmen s agains Chinese hesus macaque (M. mula a, heMacS) using he LASTZ p og am (27) ( 1.04.03) unde he ollowing pa ame e s: K= 4500, l= 3000, Y= 15,000, E= 150, H= 2000, O= 600, T= 2. The o iginal alignmen s we e hen p ocessed by he chainNe package (64) o gene a e he ecip ocal bes ne alignmen (de aul pa ame e s, excep o ax Chain whe e we used “-minSco e = 5,000 -linea Gap = medium”). Subsequen ly, we used he “ma -swap”(65) command (LAST so wa e) o so he alignmen esul s and ob ain he op imal pai wise syn eny blocks be ween genomes. Las , we used MULTIZ ( 11.2) (28) o me ge all pai wise alignmen s in o mul i- ple genome alignmen s using heMacS as he e e ence. Phylogene ic analysis Sliding window ee and signal o con lic On he basis o he mul iple genome alignmen , we used sc ip s p o- ided in (66) o pa i ion he alignmen (20 au osomes and one X ch omosome) in o nono e lapping windows o 50 kb wi hou con- side ing he p o ein-coding con en . Meanwhile, we excluded windows ha con ained mo e han 10% gaps o ha d-masked epea sequences. Fo each window o he alignmen , an ML ee was cons uc ed using RAxML 8.1.15 (67) wi h he GTRGAMMA model and apid boo s apping o 100 eplica es and speci ying P. hamad yas as he ou g oup. The numbe o al e na i e ee opolo- gies and hei ela i e equency we e ca ego ized by PhyBin 0.3 p og am (68). Con lic s be ween he gene ees and hei equen- cies we e summa ized using he CONSENSE p og am in PHYLIP 3.697 package (69). The coalescen species- ee was es ima ed om he a o emen ioned sliding window ees by wo summa y coales- cen -based me hods, ASTRAL (70) and STAR (71). Because he majo disag eemen among all opologies conce ned he ela i e placemen o he ascicula is g oup (Fig. 1, B and C), o he sake o simplici y, we p uned he ees so as o include only i e ing oup species: M. syl anus,M. hibe ana,M. silenus,M. nig a, and M. mula a. These i e species ep esen ed he syl anus g oup, sinica g oup, silenus g oup (E), silenus g oup (W), and as- cicula is g oup, espec i ely, and we e selec ed on he basis o hei longes sca old N50 size. Simila ly, we used PhyBin 0.3 (68) o es- ima e he ca ego ies o al e na i e opologies and hei ela i e e- quencies. We also e alua ed smalle (20 kb) and la ge (100 kb) window sizes, bu his did no change ou esul s ma e ially. We he e o e used he 50-kb block size o all subsequen analyses. Mi ochond ial genome ee The mi ochond ial genome sequence o each species was ob ained om Illumina sho eads using NOVOplas y 2.4 (72). K-me was se o 33, and he mi ogenome o M. syl anus (AJ309865) down- loaded om he Na ional Cen e o Bio echnology In o ma ion (NCBI) da abase was used as a s a ing e e ence. The eliabili y o mi ochond ial con ig assemblies was u he alida ed ia BLAST sea ches agains he e e ence mi ogenome (AJ309865). The D-loop and all ans e RNA ( RNA) genes we e no used owing o hei high mu a ion a e and high a e o loss. P o ein- coding genes we e ansla ed in o amino acid sequences o ensu e open eading ames and o a oid NUMTs (nuclea mi ochond ial DNA segmen s). An addi ional 18 macaque mi ochond ial genome sequences om NCBI we e added o expand ou da ase s and o u he alida e ou sample iden i ica ion. The aligned genomes we e pa i ioned in o p o ein-coding genes and noncoding ag- men s, and he p o ein-coding genes we e u he pa i ioned in o i s , second, and hi d codon posi ions. Pa i ionFinde 2.1.6 (73) was used o e alua e he bes pa i ioning scheme unde he Baye- sian in o ma ion c i e ion. The ML analyses we e ca ied ou using RAxML 8.1.15 wi h 1000 boo s ap eplica ions unde he bes pa - i ion scheme and he GTRGAMMA model. Y ch omosome phylogeny We con ined ou analyses o i e Y-linked genes (TSPY10,SRY,ZFY, USP9Y, and RPS4Y1) owing o he subs an ial echnical challenges p esen ed by he sequence alignmen o epe i i e sequence egions o he Y ch omosome. Se en o ou genome assemblies we e om males. To expand ou da ase , we gene a ed whole-genome sho gun sequences o wo male samples om M. onkeana and M. syl anus, and u he downloaded one male sample om M. nemes ina (SRR5947292). The Illumina sho eads o each species we e mapped o he Chinese hesus macaque genome ( heMacS) using BWA-MEM 0.7.12 (74) wi h he de aul se ings. A e ob aining he bam iles, we used ANGSD (75) o ob ain he consensus se- quence o hese genes wi h he ollowing il e ing pa ame e s: -b bam iles. x -minQ 20 -minMapQ 20 - emo e_bads -uniqueOnly - egion. x -dohaplocall 1 -doCoun s 1. The conca ena ion o he abo emen ioned i e genes comp ised a o al o 480,333 base pai s (bp). Again, we pe o med he phylogene ic analyses using RAxML .8.1.15 unde he GTRGAMMA model wi h 1000 boo - s ap eplica ions. Phylogene ic signal ela i e o he local ecombina ion a e No ecombina ion map is cu en ly a ailable o he e e ence genome o he Chinese hesus macaque ( heMacS). To cons uc his, we used he li o e ool (76) o con e he la es ine-scale linkage map o Indian hesus macaque ( heMac8) (77). Fi s , he genome o heMac8 was aligned o ha o heMacS using he LASTZ p og am 1.04.03 (27) wi h he same pa ame e s men ioned o he genome alignmen s. A e ob aining he chain ile, which eco ds he links o he ecip ocal bes o hologous egions o he genome, we ans e ed he coo dina es o he ecombina ion map o heMac8 o heMacS coo dina es using he p og am li o e (76). Only he success ully de i ed li o e posi ions (~96.3%) we e used in downs eam analyses. Hyb idiza ion and in og ession analyses Al hough ILS ce ainly unde pins some aspec s o phylogene ic dis- co dance wi hin and be ween clades, i is unlikely o be he only ex- plana ion o he no able di e ences we obse ed ac oss all window ees. The e o e, we used se e al di e en me hods o es o hy- b idiza ion and in og ession e en s in he p esence o ILS. We SCIENCE ADVANCES |RESEARCH ARTICLE Zhang e al.,Sci. Ad . 9, eadd3580 (2023) 1 June 2023 9 o 13 Downloaded om h ps://www.science.o g a Cen o de Biologia Molecula Se e o Ochoa on Feb ua y 21, 2024