Genetic strategies for sex-biased persistence of gut microbes across human life

A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Gene ic s a egies o sex-biased pe sis ence o gu mic obes ac oss human li e Chia a Ta acchini 1 , Giulia Alessand i 1 , Fede ico Fon ana 1,2 ,SoniaMi jamRizzo 1 , Gab iele And ea Lugli 1 , Massimiliano Gio anni Bianchi 3,4 , Leona do Mancabelli 3,4 , Giulia Longhi 1 , Chia a A gen ini 1 , Lau a Ma ia Ve gna 1 , Rosa ia Anzalone 2 , Alice Viappiani 2 , F ancesca Tu oni 1,4 ,GiuseppeTau ino 3,4 ,Ma inaChiu 3 , Sil ia A boleya 5 , Miguel Gueimonde 5 , O idio Bussola i 3,4 , Douwe an Sinde en 6 , Ch is ian Milani 1,4,7 &Ma coVen u a 1,4,7 Al hough composi ional a ia ion in he gu mic obiome du ing human de elopmen has been ex ensi ely in es iga ed, s ain- esol ed dynamic changes emain o be ully unco e ed. In he cu en s udy, sho gun me age- nomic sequencing da a o 12,415 ecal mic obiomes om heal hy indi iduals a e employed o s ain-le el acking o gu mic obio a membe s o elucida e i s e ol ing biodi e si y ac oss he human li e span. This de ailed longi udinal me a-analysis e eals hos sex- ela ed pe sis ence o s ains belonging o common, ma e nally-inhe i ed species, such as Bifidobac e ium bifidum and Bifidobac e ium longum subsp. longum. Compa a i e genome analyses, cou- pled wi h expe imen s including in ima e in e ac ion be ween mic obes and human in es inal cells, show ha specific bac e ial glycosyl hyd olases ela ed o hos -glycan me abolism may con ibu e o mo e e ficien coloniza ion in emales compa ed o males. These findings poin o an in iguing ancien sex- specific hos -mic obe coe olu ion d i ing he selec i e pe sis ence in women o key mic obial axa ha may be e ically passed on o he nex gene a ion. The in es inal mic obial communi y o an in an g adually assembles h ough a pa e ned de elopmen al p ocess ollowing bi h. In pa i- cula , he fi s h ee yea s o li e ep esen a c i ical window o oppo uni y du ing which sho - e m changes in he gu mic obio a composi ion occu in conjunc ion wi h he apid physical de elop- men o he newbo n1,2. This ini ial dynamic p ocess e en ually e ol es owa ds s able mic obe-hos in e ac ions ha a e o pa amoun impo ance o impa beneficial e ec s on hos heal h, such as me a- bolism o non-diges ible die a y ca bohyd a es and s imula ion o endogenous in es inal mucus p oduc ion, i amin syn hesis, de el- opmen and homeos asis o he immune sys em, as well as p o ec ion agains pa hogens3. In pa icula , i is known ha , a species le el, he es ablishmen o a s able gu mic obio a occu s h ough wo main die - guided s ages in ea ly childhood4,5,wi h hefi s one aking place immedia ely a e bi h when exclusi e milk- eeding begins6,7.This s age is cha ac e ized by a gu mic obial communi y which, upon being pa ly ( e ically) ansmi ed om he mo he and pa ly acqui ed h ough con ac wi h he su ounding en i onmen du ing and a e deli e y8–10, ep esen hefi s mic obial colonize s o he in an gu due o hei abili y o di ec ly o indi ec ly me abolize human milk oligosaccha ides (HMOs)11,12. Ano he ansi ion occu s du ing he weaning pe iod, ypically a ound he age o six mon hs, when in an s a e g adually in oduced o a solid and mo e a ied die , which o e s new ecological niche coloniza ion oppo uni ies13–15. Recei ed: 25 Janua y 2023 Accep ed: 30 June 2023 Check o upda es 1 Labo a o y o P obiogenomics, Depa men o Chemis y, Li e Sciences, and En i onmen al Sus ainabili y, Uni e si y o Pa ma, Pa ma, I aly. 2 GenP obio s l, Pa ma, I aly. 3 Depa men o Medicine and Su ge y, Uni e si y o Pa ma, Pa ma, I aly. 4 In e depa men al Resea ch Cen e “Mic obiome Resea ch Hub”, Uni e si y o Pa ma, Pa ma, I aly. 5 Depa men o Mic obiology and Biochemis y o Dai y P oduc s, Ins i u o de P oduc os Lác eos de As u ias, CSIC, 33300 Villa iciosa, Spain. 6 APC Mic obiome Ins i u e and School o Mic obiology, Bioscience Ins i u e, Na ional Uni e si y o I eland, T12YT20 Co k, I eland. 7 These au ho s con ibu ed equally: Ch is ian Milani, Ma co Ven u a. e-mail: ch i[email p o ec ed];[email p o ec ed] Na u e Communica ions | (2023) 14:4220 1 1234567890():,; 1234567890():,; Whole-me agenome sho gun (WMGS) sequencing ep esen s a powe ul ool o disen angle he composi ion o complex mic obial communi ies and e ie e genomes belonging o ha d- o-cul u e mic obial species16–18. Acco dingly, se e al ecen longi udinal s udies ha e in es iga ed he in an gu mic obiome composi ion, highligh ing he sequen ial age- ela ed changes a species le el19,20. Fo example, bifidobac e ial species, key mic obial axa o he in an gu mic obio a, can pe sis a a lowe le el (2–14% ela i e abundance) h oughou adul hood and can subsequen ly be passed on o he nex hos gen- e a ion by mo he - o-in an e ical ansmission21,22. Ne e heless, he impac o he hos biological sex on he assembly and main enance o he gu mic obial communi y emains poo ly in es iga ed. In he cu en s udy, a o al o 400 longi udinal ecal me agen- omes om 124 heal hy in an s (0–3 yea s) we e in es iga ed o inspec he in a-species a ia ions unde lying he assembly o he in an gu mic obiome du ing he fi s wo yea s ollowing bi h. In his con ex , we assessed i pa icula gu mic obio a membe s elici ed highe pe sis ence in emale in an s (compa ed o male coun e pa s), pe - haps o main ain c oss-gene a ional ansmission. These analyses, coupled wi h inspec ion o sho gun me agenomic da a om 12,415 heal hy subjec s (6545 emales and 5870 males) aged om a ew days o 90 yea s, allowed he iden ifica ion o wo glycosyl hyd olases, i.e., membe s o GH101 and GH136, which appea associa ed wi h pe sis- ence o B. bifidum and B. longum s ains wi h p e e en ial p esence in he emale gas oin es inal ac . Mo eo e , sex- ela ed (bifido)bac- e ial esilience was alida ed in i o by e ospec i e human clinical ial da a in ol ing he supplemen a ion o bifidobac e ial s ains dis- playing a pe sis en s. non-pe sis en geno ype. Resul s S ain dynamics o he gu -associa ed mic obio a wi hin he fi s 24 mon hs o li e Sho gun me agenomics sequencing app oaches we e applied o he mic obiomes o 11 heal hy, aginally deli e ed, ull- e m (>37 weeks o ges a ion) newbo ns, which we e longi udinally sampled a 1-, 6-, 12-, and 24-mon hs ollowing bi h (Figs. S1, S2, Supplemen a y Da a 1). Consis en wi h p e ious scien ific li e a u e ocusing on in an com- muni y s a e ypes (ICSTs)23, he species-le el axonomic classifica ion o he sequenced eads e ealed ha Bifidobac e ium longum and Esche ichia coli we e he mos p e alen mic obial in an gu compo- nen s a he p e-weaning s age, ollowed by Bifidobac e ium pseudo- ca enula um, Bifidobac e ium b e e, Collinsella ae o aciens,and Bifidobac e ium bifidum (Supplemen a y Da a 2). Fu he mo e, as epo ed p e iously, he ela i e a e age abundances o hese species dec eased in pos -weaning (K uskal–Wallis es , p- alue < 0.01), simul aneously wi h he p og essi e coloniza ion by adul -associa ed bac e ial species, including Eubac e ium ec ale,Faecalibac e ium p ausni zii,Ru henibac e ium lac a i o mans,Akke mansia mucini- phila, and membe s o he Bac e oides genus, such as Bac e oides uni o mis (Fig. S2, Supplemen a y Da a 2)23,24. Wi h he aim o in es iga ing s ain dynamics in he de eloping in an gu ecosys em, a o al o 63 me agenomically assembled gen- omes (MAGs), co esponding o he 11 abo e-lis ed main gu - associa ed mic obial species (Supplemen a y Da a 3), we e coupled wi h conspecific publicly a ailable genome sequences in o de o build 11 species-specific da abases o e e ence s ains (Supplemen a y Da a 4). Following assessmen o comple eness (>90%) and ANI- d i en de eplica ion, hese 11 da abases we e employed o in es iga e s ain-specific pe sis ence ands abili y, i.e., he ime span du ing which longi udinal samples ha bo ed iden ical s ains (Fig. S3)25. Collec ed da a e ealed ha s ains belonging o species asso- cia ed wi h he in oduc ion o solid die s, such as E. ec ale and C. ae o aciens, assemble in o he e ogeneous s ain communi ies, appea ing ulne able o in es inal niche changes ha occu du ing he fi s wo yea s o in an li e (Fig. S3). Con e sely, specificB. longum subsp.longum,B. bifidum, B. b e e,andB. pseudoca enula um s ains es ablished s able hos -mic obe symbio ic ela ionships las ing beyond he weaning phase in 91%, 72 %, 54.5 %, and 45.4% o he inspec ed in an s, espec i ely (Table S2, Fig. S3), he eby ep esen ing he mos esilien , un a ying, and s able bac e ial communi ies o he assessed in an gu mic obiome (See supplemen a y ex o de ails). Specifically, a e accoun ing o sequencing dep h, an adjus ed a e - age numbe o 1.58 B. longum subsp. longum,1.44B. bifidum,0.82B. b e e,and0.80B. pseudoca enula um s ains we e ound o be sha ed among mul iple ime poin s wi hin he fi s bi h yea (Fig. S3), hus eme ging as significan ly mo e pe sis en han o he p incipal mem- be s o he suckling in an gu mic obiome(K uskal–Wallis wi h Dunn’s pos -hoc es , p- alues a e he Benjamini-Hochbe g co ec ion <0.05; Table S5; Fig. S4). Mo eo e , specificB. bifidum and B. longum subsp. longum s ains we e de ec ed in he in an gu up o he second yea a e bi h, coinciding wi h he las sampling ime (Supplemen a y Da a 5). Consis en ly, hese bifidobac e ial species a e known o be gene ically adap ed o colonizing he in an ’s in es ine due o pa i- cula me abolic ac i i ies and coope a i e ophic in e ac ions, i.e., c oss- eeding ac ions, owa d complex ca bohyd a es, such as HMOs and human mucin26–28. The s ain- esol ed dynamic o he decoded in an gu mic o- biome was expe imen ally alida ed using s ain-specific p ime s h ough quan i a i e eal- ime PCR (qRT-PCR) (Supplemen a y Da a 6), confi ming ha fluc ua ing and/o pe sis en pa e ns occu - ing du ing he in an gu mic obiome de elopmen a e s ic ly dependen on he mic obial species. To alida e he esul s om ou popula ion-wide me agenomic s udy and o ensu e ha he bioin o ma ic app oaches did no bias he obse ed bac e ial pe sis ence pa e ns, an independen alida ion coho was cons uc ed employing a la ge, publicly a ailable in an da ase ha includes ecal samples om mul iple ime poin s spanning he fi s yea a e bi h29 (Supplemen a y Da a 1). The s ain- esol ed mic obial communi y composi ion o his pa icula me agenomic da ase was de e mined employing he pipeline implemen ed by Mäklin e al.30. As de ailed in Table S7, he de elopmen ajec o ies o he in an gu mic obiome obse ed wi hin he alida ion coho confi med wha we no ed in ou s udy popula ion, highligh ing a longe - e m pe sis ence o bifidobac e ial s ains compa ed o hose belonging o non-bifidobac e ial species, i.e., C. ae o aciens,andE. coli (Chi-Squa ed pos -hoc es , p- alue < 0.05) (Supplemen a y Da a 7). Mo eo e , a s a is ically significan highe pe sis ence o he ea ly B. bifidum and B. longum colonizing s ains was obse ed in aginally deli e ed in an s when compa ed o hose bo n by C-sec ion (Chi- Squa ed es , p- alues < 0.05) (Supplemen a y Da a 7). These esul s expand on ou p e ious findings, sugges ing a po en ial g ea e abili y o ma e nally de i ed bifidobac e ial s ains o pe sis in he in an gu mic obiome du ing he neona al pe iod. Co ela ion be ween mic obial esilience and hos sex As we men ioned abo e, he fi s seeding o (membe s o ) he gu mic obio a is belie ed o occu du ing deli e y, in ol ing he ans e o mic obial lineages om he mo he o he espec i e newbo n22,31.In his con ex , we ques ioned i mic obial s ain pe sis ence is mo e e ec i e in emales o sus ain he in e gene a ional ansmission o ecologically well-adap ed gu mic obio a membe s. Fo his pu pose, he aginally deli e ed ac ion o he abo e-men ioned longi udinal me agenome in an da ase (72 emales and 73 males, o a o al o 145 aginally deli e ed in an s) we e in eg a ed wi h 357 addi ional, pub- licly a ailable sho gun samples om longi udinal s udies o he in an gu mic obiome (54 emales and 59 males), encompassing p e- weaning (0–6 mon hs) and pos -weaning (o e six mon hs) ime- poin s (Fig. S1, Supplemen a y Da a 1)11,32. All in an s we e heal hy, deli e ed aginally a e m, and we e no subjec ed o an ibio ic ea men (Supplemen a y Da a 1). Hence, conside ing he abo e- A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 2 desc ibed high pe sis ence o bifidobac e ial s ains h oughou he ansi ion phase o weaning, we e alua ed he sex-specific s abili y o s ain communi ies belonging o B. bifidum,B. longum subsp. longum, B. b e e,andB. pseudoca enula um be ween p e- and pos -weaning s ages by using he S ainGE ool (see Me hods). Specifically, o each in an , he bifidobac e ial s ain domina ing he gu mic obiome a 0–6 mon hs o age was compa ed wi h hose p eeminen a e he in oduc ion o complemen a y solid oods (a ound 12 mon hs). A he species le el, B. bifidum was de ec ed om lac a ion o pos -weaning in 31% o he 258 inspec ed in an s, including 36 emales and 44 males (Fig. 1a, Supplemen a y Da a 9). While no sex- ela ed di e ence in species-le el s abili y was obse ed ac oss weaning (Fishe es , p- alue = 0.349), he dominan B. bifidum e e - ence s ains de ec ed a 0–6 mon hs pe sis ed a e he weaning phase in 24 ou o 36 emale in an s (67%) and in 15 o he 44 males (34%) (Fig. 1a, Supplemen a y Da a 9), sugges ing g ea e s ain-le el s abili y in he la ge in es ine o emale in an s (Fishe es , p- alue = 0.007). Simila ly, 49.6% o all assessed in an s (67 emales and 61 males) exhibi ed c oss-weaning pe sis ence o he B. longum subsp. longum species, whose dominan e e ence s ains iden ified a 0–6mon hs we e also ound a 12–24 mon hs in 45% o he emales, being sig- nifican ly highe han wha was obse ed o male in an s (21 %) (Fishe es ,p- alue = 0.005) (Fig. 1a, Supplemen a y Da a 9). Specifically, males appea o unde go highe fluc ua ions in (bifido)bac e ial s ain composi ion compa ed o emale in an s, whose ea ly-eng a ed dominan B. bifidum and B. longum subsp. longum s ain was main- ained h ough he weaning phase wi h a highe equency. In con as , B. b e e and B. pseudoca enula um species we e de ec able ac oss ime-poin s (p e- and pos -weaning) only in 14.3% and 9.6% o in an s, espec i ely (Fig. 1a), showing no significan sex- associa ed di e ence in s ain pe sis ence (Fishe es p- alue > 0.05) (Fig. 1a, Supplemen a y Da a 9). As B. longum subsp. longum exhibi ed highe sex-biased s abili y/pe sis ence in he in an gu , we ex ended he abo e-desc ibed analysis o dominan s ains by explo ing he whole B. longum subsp. longum s ain composi ion h oughou he weaning phase by using di e en bioin o ma ic app oaches. Fo his pu pose, we ob ained 377 non- edundan B. longum subsp. longum genomes by assembling me agenome-de i ed da a om he in an longi udinal da ase s (258 in an s, 126 emales, and 132 males). This genome collec ion was in eg a ed wi h publicly a ailable B. longum subsp. longum ch omosomal sequences and, ollowing comple eness assessmen sandde eplica ion,wasemployedasagenomeda abase o he inS ain ool33. Conside ing he in an s de ec ed by inS ain wi h c oss-weaning pe sis ence o B. longum subsp. longum a he species le el (71 emales and 65 males), we obse ed ha a leas one s ain ound in he p e-weaning age was main ained in he pos -weaning phase in 52 emales (76%) and 31 males (47%) (Fishe es p- alue = 0.003) (Fig. 1b), hus co obo a ing he findings epo ed abo e. Mo eo e , a e a pe sis ence e en (n= 83), he ea ly-eng a ed B. longum s ains eached an a e age ela i e abundance o 71% ± 21% in he o e all pos -weaning B. longum subsp. longum s ain communi ies (Fig. 1c). These figu es, al hough simila be ween in an males and emales, we e significan ly highe han hose calcula ed o he B. longum subsp. longum s ains no in ol ed in pe sis ence episodes (n= 171, a e age ela i e abundance o 29% ± 29%, Mann–Whi ney es , p- alue = 0.001) (Fig. 1c). These findings imply ha when a pe sis ence e en occu s, i in ol es s ains ecologically a o ed o colonize he in an gas oin es inal ac a highe ela i e abundance, hus dom- ina ing he conspecific s ain popula ion. No ably, analysis o he o e all B. longum subsp. longum s ain popula ion showed ha pe sis ence o specific s ains seems o occu a a s a is ically significan highe a e in emale in an s compa ed o male in an s, also highligh ing a supe io coloniza ion capabili y o he pe sis en B. longum subsp. longum membe s compa ed o he coex- is ing non-pe sis en s ains. Based on cu en scien ific da a, one may a gue ha ollowing ini ial inocula ion by he ma e nal ecal bac e ia, die ing edien s such as hos -de i ed glycans play a ole in selec ing pe sis en (bifido)bac- e ial s ains, such as B. bifidum and B. longum subsp. longum mem- be s, based on hei abili y o me abolize hese amino suga s34,35.Such bifidobac e ial s ains fi s o age on lac ose, HMOs,and pe haps o he milk-associa ed glycans o glycop o eins, subsequen ly aking ad an- age o in es inal mucin glycans bo h as binding si es and as a ca bon sou ce in a s ain-specific manne , leading o hei long-las ing pe sis ence35. Iden i ying genes associa ed wi h sex-specificmic obial pe sis ence As B. bifidum and B. longum subsp. longum s ains exhibi ed sex- specific esilience ac oss in an weaning, we pe o med a compa a i e genomic analysis o explo e he po en ial gene ic ea u es dis in- guishing hese bifidobac e ial species om B. b e e and B. pseudoca- enula um, which displayed no o poo pe sis en beha io o sex- ela ed di e ences (Fig. S1). To gain an accu a e o e iew, he analysis encompassed all comple e and well-anno a ed genome assemblies a ailable om public eposi o ies o a o al o 119 genomes belonging o hese bifidobac e ial axa. F om hese su eys, 14 p o ein amilies ( e e ed o as PDC) we e iden ified wi hin he co e/accesso y gene epe oi e unique o B. bifi- dum and B. longum subsp. longum, while hey we e absen in B. b e e and B. pseudoca enula um ch omosomes (See Supplemen a y ex o de ails) (Fig. 1b, Supplemen a y Da a 10). Among hese, we ound wo GH-encoding genes ha we e selec ed o u he explo a ion con- side ing he known impo ance o hos glycans in modula ing hos - mic obe in e ac ions, i.e., he p edic ed ex acellula memb ane- ancho ed mucin-deg ading glycosyl hyd olase amily 101 (GH101)34, and he glycosyl hyd olase amily 136 (GH136) which is p edic ed o ac as an ex acellula lac o-N-biosidase36 (Fig. S5). Specifically, genes encoding GH101 and GH136 enzymes we e ound in all sc eened B. bifidum genomes, while 89% and only 38% o he ch omosomes belonging o B. longum subsp. longum showed he p esence o genes speci ying GH101 and GH136, espec i ely (Fig. 1b; Fig. S6; Supple- men a y Da a 10). The in ol emen o GH101 and GH136 ac i i ies in hos glycan me abolism was e ified h ough a ansc ip omic su ey on B. bifidum PRL2010 cul i a ed unde in i o condi ions, showing he up- egula ion o hese GHs when mucin was used as he sole ca - bon sou ce a he han glucose (MRS-based medium) (Supplemen a y Da a 11) (See supplemen a y ex o de ails). These findings sugges ha he GH101 and GH136 enzymes, ac ing on mucin glycan co e s uc u es, a e in ol ed in he obse ed long- e m coloniza ion in he hos gu o B. bifidum and B. longum subsp. longum s ains by p o- iding an endogenous sou ce o nu ien s in he absence o die a y glycans. Sex-specific gu pe sis ence o B. longum and B. bifidum s ains om bi h o la e s ages o hos li e To alida e he hypo hesis ha B. bifidum and B. longum subsp. longum s ains can achie e highe coloniza ion in he human ( emale) gu ac oss human li espan, a o al o 12,415 c oss-sec ional me agenomic ecal samples (o which 6545 o 54% we e o emale o igin and 5870 o 46% we e de i ed om males) om oughly 3000 in an s (0–4yea s old, 1456 emale and 1541 males), 918 child en (5–18 yea s old, 434 emales and 484 males), 6147 adul s (19–55 yea s old, 3379 emales and 2768 males), and 2353 elde ly (56–90 yea s old, 1276 emales and 1077 males) we e subjec ed o s ain-le el analyses employing he same app oach and in ol ing he abo e desc ibed B. longum subsp. longum and B. bifidum e e ence genome da abases in o de o analyze he longi udinal in an da a (Fig. S1). As expec ed, suckling in an s showed he highes p e alence o B. longum subsp. longum and B. bifidum species, exceeding 70% and 40%, espec i ely (Fig. 2a, Supplemen a y A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 3 Da a 10). In e es ingly, when we explo ed adul popula ions, we obse ed ha sex ma kedly impac ed he p e alence o hese bifido- bac e ial axa (PERMANOVA R2= 0.0627 and 0.0558; F= 494.18 and 420.03, bo h p- alues = 0.0099; Supplemen a y Da a 12). Specifically, 41% and 28% o he adul emale subjec s ha bo ed membe s o B. longum subsp. longum and B. bifidum species, espec i ely, which, con e sely, colonized only 26% and 11% o he age-ma ched male indi iduals (Fishe es , p- alues < 0.001) (Fig. 2a, Supplemen a y Da a 10). In con as , elde ly subjec s showed he lowes p e alence o bifidobac e ial species, anging om an a e age o 26% o B. longum subsp. longum o an a e age o 9% o B. bifidum species, wi h simila alues be ween emale and male indi iduals (Fig. 2a, Supplemen a y B. pseudoca enula um 0% 20% 40% 60% 0% 20% 40% 60% 80% a) d) P e alence in B. bi idum and B. longum subsp. longum Bi idobac e ium longum subsp. longum Bi idobac e ium bi idum Bi idobac e ium longum subsp. longum Bi idobac e ium bi idum Bi idobac e ium b e e Bi idobac e ium pseudoca enula um 4'-phosphopan e heinyl ans e ase supe amily p o ein ABC anspo e ATP-binding p o ein Amino ans e ase class I/II- old py idoxal phospha e-dependen enzyme FHA domain-con aining p o ein Zincin-like me allop o ease Py idoxal phospha e-dependen amino ans e ase Hypo he ical p o ein Endo-alpha-N-ace ylgalac osaminidase (GH101) Ace yl ans e ase YPDG domain-con aining p o ein (GH136) YihY/ i ulence ac o B kB amily p o ein GCN5- ela ed N-ace yl ans e ase Se ine/ h eonine-p o ein kinase PknA DUF1846 domain-con aining p o ein 14 sha ed genes 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 82% 82% 64% 64% 100% 100% 100% 96% 84% 81% 79% 89% 79% 38% 97% 56% 100% 70% Pe cen age o in an s ( emales and males) wi h species s abili y also showing s ain pe sis ence o e ime Pe cen ages o in an s showing pos -weaning s abili y o bi idobac e ial species B. pseudoca enula um B. bi idum B. b e e Males Females STRAIN-LEVEL ** p- alue=0.005 ** p- alue=0.006 B.bi idum B. b e e 36 61 67 20 17 15 10 44 SPECIES-LEVEL B. longum subsp. longum Numbe o B. longum subsp. longum s ains Numbe o B. longum subsp. longum s ains b) Numbe o B. longum subsp. longum s ains speci ic o p e-weaning phase Numbe o B. longum subsp. longum s ains speci ic o pos -weaning phase Numbe o B. longum subsp. longum s ains sha ed be ween p e- and pos -weaning phases c) Es ima ed ela i e abundance o B. longum subsp. longum s ains Pe sis en B. longum subsp. longum s ains Non pe sis en B. longum subsp. longum s ains 71% 29% A e age ela i e abundance B. longum subsp. longum a b p g 0 1 2 3 4 5 0 1 2 3 4 5 6 A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 4 Da a 10). Possible con ounding ac o s such as dai y ood consump- ion and lac ase pe sis ence we e accoun ed o he di e ences in B. longum subsp. longum and B. bifidum p e alence be ween adul emales and males. No ably, since his ype o in o ma ion was no a ailable in public da ase s, we used geog aphic egions as p oxy a iables (sou h/no h Eu ope o lac ase pe sis ence and Eu ope/ China o dai y ood consump ion), e idencing no significan asso- cia ion be ween geog aphic egion and sex-dependen p e alence o he a ge (bifido)bac e ial species (Fishe es p- alues > 0.05) (Sup- plemen a y Supplemen a y Da a 13). These findings suppo he in iguing no ion ha B. bifidum and B. longum subsp. longum cans ablycolonize hehumangu omin ancy o adul hood wi h an appa en p e e ence in women du ing hei ep oduc i e age, possibly as po en ial ese oi s o mic obial ans- mission o new gene a ions. The genomes o B. bifidum and B. longum subsp. longum wi h comple eness equal o o g ea e han 90% de ec ed wi hin he in an , child, adul , and elde ly gu mic obiomes we e subjec ed o a genome- wide sc eening o assess he occu ence o genes encoding p edic ed GH101 and GH136 enzymes and hus o deduc hei p e alence ac oss popula ions (Fig. 2b, Supplemen a y Da a 10). Acco ding o he su ey esul s, he GH101 o B. bifidum and B. longum subsp. longum we e de ec ed be ween 88% and 100% o he 12,415 ecal me agenomic da ase s, which, being g ea e han expec ed om public genome sc eening, suppo s hei po en ial key ole in coloniza ion and su i- al o he human gas oin es inal ac ac oss he en i e hos li espan (Supplemen a y Da a 10). Ins ead, he GH136 was ound in an a e age o 45% o me age- nomic samples, wi h he highes equency ound in adul women (63%), which is 50 % highe han ha obse ed in age-ma ched males (42%) (Fishe es , p- alue < 0.05) (Supplemen a y Da a 10). Howe e , such sex-specific di e ences we e no e iden in ea ly in ancy and olde adul hood. In iguingly, he highe coloniza ion pe o mances obse ed in emales compa ed wi h males seem o disappea in indi iduals olde han 50 yea s, which is conco dan wi h menopause age (Supplemen a y Da a 10). Recen ly, sex ho mones ha e been ega ded as po en d i e s o sexual dimo phism a he gu mic o- biome le el, being associa ed wi h composi ional di e ences be ween sexes and p o ound changes in he gu mic obial communi y o p egnan women37–40. I has been p oposed ha , besides i s ole in he modula ion o he immune sys em and bile acid sec e ion, which may hen egula e he gu mic obio a, s e oid sex ho mones can be me abolized by specific gu -associa ed bac e ial enzymes, hus di ec ly impac ing mic obial me abolism and g ow h37,41,42. Al hough li le is ye known abou he impac o sex on he human gu mic o- bio a and e en less abou he unde lying mechanisms, i has been epo ed ha a highe le el o sialyla ion cha ac e izes he in es inal mucus o he emale gu compa ed o ha o he male popula ion43. Fu he mo e, i has also been a gued ha he emale sex ho mone es adiol may up egula e he exp ession and glycosyla ion o human mucins44–46, possibly shaping he emale gu mic obio a in a o o mucin-u ilizing (bifido)bac e ia. In e es ingly, he es adiol le el is highe in emales han males, e en in he p epube al phase47.Indeed, he hypo halamic-pi ui a y-gonadal axis, which is in ol ed in he de elopmen and egula ion o he ep oduc i e sys em, unde goes ansien ac i a ion du ing he fi s six mon hs o li e in males and he fi s wo yea s in emales48,49.Thise en ,calledminipube y,o “endoc ine pube y”, induces es os e one p oduc ion in males and es adiol in emales50–52. In addi ion o hese ho mone-d i en e ec s, i should be no ed ha complex mechanisms o DNA me hyla ion pa e ns in colon issue-specific genes (excluding loci loca ed in he X and Y ch omo- somes) may con ibu e o he es ablishmen o sex- and age- ela ed di e ences in he in es inal en i onmen s53,54. Consis en ly, a iabili y in me hyla ion signa u es was obse ed in subjec s o a ious ages, including bo h newbo ns and adul s, when compa ing emales o males53,55. Al oge he , hese findings sugges ha , compa ed o males, he emale-specific in es inal su oundings, including glycan s uc u es, may be s uc u ally mo e p edisposed o c ea ing a sui able en i on- men o ce ain ea ly colonizing mic obial species, such as B. bifidum and B. longum. In pa icula , when he gene encoding he GH136 enzyme is p esen , i could play a pa in enhancing (bifido)bac e ial pe sis ence in women o ep oduc i e age, possibly due o a sex- specific mucus s uc u e. The ole o GH136 in bac e ial e ical ansmission om mo he o newbo n B. bifidum and B. longum ep esen wo o he main species ha can be ma e nally inhe i ed by means o e ical ansmission56. In pa icula , coloniza ion o he in an gu om ma e nal gu -associa ed s ic ly anae obic species, such as Bifidobac e ium, is belie ed o occu h ough di ec con ac wi h ma e nal gu mic obio a du ing bi h and/ o in ol e an en e o-mamma y pa hway, which may ans e ecologi- cally well-adap ed bac e ia ia b eas milk o in an s57–60. S a ing om his scien ific e idence and exploi ing he non-ubiqui ous p esence o GH136-encoding genes in B. longum s ains, we decided o es ima e he in ol emen o GH136 in mo he - o-in an e ical ansmission e en s h ough inspec ion o publicly a ailable me agenomic ecal samples om 132 ull- e m aginally deli e ed heal hy newbo ns and hei co esponding mo he s5,61 (Fig. S1). No ably, ecal samples collec ed om mo he s a childbi h and newbo ns a one mon h we e subjec ed o B. bifidum and B. longum subsp. longum s ain acking analyses, and he eads we e hen mapped o he GH136 gene sequence. Ou esul s indica ed ha specifics ainso B. longum subsp. longum and B. bifidum we e p esen in he samples om mo he -in an dyads in 36.4% and 29.5% o he sc eened cases, espec i ely, sug- ges ing pe ina al e ical ansmission e en s. In e es ingly, 72.9% o he B. longum subsp. longum s ains de ec ed as e ically ansmi ed om mo he s o newbo ns ha bo he GH136-encoding gene, while Fig. 1 | Gende -specific pe sis ence o B. bifidum,B. longum subsp. longum,B. b e e,andB. pseudoca enula um species and he 14 genes sha ed be ween he emale-associa ed pe sis en bifidobac e ial species. In panel (a), he ba cha on he op displays he species-le el pe sis ence om p e- (1–6 mon hs) o pos - weaning s ages (12–24 mon hs) o B. bifidum,B. longum subsp. longum, B. b e e,and B. pseudoca enula um in he in an popula ion. Below, ba plo s ep esen he 12–24-mon hs gende -specific s abili y o he bifidobac e ial s ains, exp essed as he pe cen age o (in an ) emales and males showing a e -weaning pe sis ence o hesamebifidobac e ial s ains iden ified a 0–6 mon hs. S a is ically significan gende - ela ed di e ences we e highligh ed wi h an as e isk on he op o he columns (Fishe es , p- alue = 0.006 and p- alue = 0.005). Panel (b) e e s o he inspec ion o he whole B. longum subsp. longum s ain communi ies in emale and male in an s. Each pai o ba plo s shows he numbe o s ains iden ified in he p e-weaning (le ) and pos -weaning phase ( igh ). Di e en colo s highligh he numbe o B. longum subsp. longum s ains ound only in p e-weaning age (ligh blue), only in he pos -weaning phase (ligh g een), and sha ed be ween p e- and pos -weaning ime-poin (pink). Panel (c) depic s he s a is ically significan di - e ence in an es ima ed ela i e abundance (ob ained by no malizing he genome co e age on he co esponding genome leng h) o pe sis en (n= 83, ligh blue) and non-pe sis en (n=171,o ange)B. longum subsp. longum s ains in he in an pos -weaning ime poin s (Mann–Whi ney es , p- alue = 0.001). The boxes a e de e mined by he 25 h and 75 h pe cen iles. The whiske s a e de e mined by 1.5 in e qua ile ange (IQR). The line in he boxes ep esen s he median, while he c oss ma ke (X) ep esen s he a e age. In panel (d), he Venn diag am on he le highligh s he 14 genes sha ed by B. longum subsp. longum and B. bifidum,whileba cha s on he igh -side epo he p e alence o each PDC in publicly a ailable comple e genomes B. longum and B. bifidum. A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 5 jus 28.6% o he B. longum subsp. longum s ains ha did no appea o be in ol ed in e ical ansmission e en s we e shown o possess he GH136 gene (Fische Tes p- alue < 0.001, Supplemen a y Da a 14). Acco dingly, sex- ela ed gene ic and epigene ic hos ac o s seem o be associa ed wi h specific mic obial genomic ea u es o ensu e pe - sis ence and, he e o e, he es ablishmen o selec ed conso ia o bifidobac e ial s ains ha may be ma e nally ansmi ed h ough deli e y o he o sp ing. 0% 10% 20% 30% 40% 50% 60% 70% 80% In an s (0-4 yea s) Adul s (19-55 yea s) Elde ly (56-90 yea s) 0% 10% 20% 30% 40% 50% Bi idobac e ium bi idum Bi idobac e ium longum subsp. longum FHA domain-con aining p o ein Zincin-like me allop o ease Py idoxal phospha e-dependen amino ans e ase Hypo he ical p o ein Endo-alpha-N-ace ylgalac osaminidase (GH101) Ace yl ans e ase YPDG domain-con aining p o ein (GH136) YihY/ i ulence ac o B kB amily p o ein GCN5- ela ed N-ace yl ans e ase Se ine/ h eonine-p o ein kinase PknA DUF1846 domain-con aining p o ein 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 90% 85% 100% 54% 100% 100% 100% 100% 89% 91% 88% 90% 82% 41% 98% 37% 100% 92% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 98% 86% 100% 53% 100% 100% 100% 98% 84% 99% 98% 99% 90% 53% 99% 24% 99% 95% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 88% 84% 100% 59% 100% 100% 100% 100% 92% 100% 100% 100% 86% 43% 99% 29% 100% 95% P e alence P e alence o bi idobac e ial species in human popula ions ac oss di e en li e s ages Bi idobac e ium longum subsp. longum Bi idobac e ium bi idum Bi idobac e ium longum subsp. longum Bi idobac e ium bi idum 42%63% a) b) Amino ans e ase py idoxal phospha e- dependen enzyme 4'-phosphopan e heinyl ans e ase supe amily p o ein ABC anspo e ATP-binding p o ein Fig. 2 | Pe sis ence o B. bifidum and B. longum subsp. longum ac oss human li e s ages. In panel (a), he e ical ba g aph depic s he pe sis ence o B. bifidum and B. longum subsp. longum ac oss he human li e span. On he op, ba plo s show he di e ences in he pe sis ence o hese bifidobac e ial species be ween emale and male indi iduals aged 5–55 yea s. In panel (b), ho izon al ba cha s ep esen he p e alence o he 14 genes sha ed by B. bifidum and B. longum subsp. longum ac oss popula ions. S a is ically significan gende -based di e ences in he occu ence o GH136 a e highligh ed wi h a ed ame, and he espec i e p e alence (pe cen age) a e epo ed alongside (Fishe es , p- alue < 0.05). A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 6 Re ospec i e clinical s udies suppo he GH136-d i en, long- e m gu pe sis ence o B. longum subsp.longumin women In o de o assess he ole o he accesso y GH136 in he coloniza ion and pe sis ence o B. longum subsp. longum in he human gu , we analyzed da a om published e ospec i e clinical s udies in which heal hy human pa icipan s ecei ed daily o al doses o iable cells o wo gene ically di e en B. longum subsp. longum s ains62,63 (Fig. S1). Specifically, a o al o 21 heal hy indi iduals (52% emales) ecei ed a ea men consis ing o a daily dose o 1010 iable cells o B. longum subsp. longum AH1206, a s ain ha ha bo s genes encoding he GH136 and GH101 enzymes (Supplemen a y Da a 10). A second g oup o 10 indi iduals (40% emales) consumed he same daily dosage (1010 iable cells) o a B. longum subsp. longum s ain named AG1, posses- sing GH101 bu lacking he GH136 gene (Supplemen a y Da a 15). Sho gun me agenomic da a o s ool samples collec ed be o e he in e en ion (baseline), a e 21–28 days o o al bac e ial adminis a- ion ( ea men ), and a e he ollow-up pe iod (pe sis ence) we e exploi ed o e alua e he pe sis ence o B. longum subsp. longum AH1206 and AG1 s ains (Fig. 3a). Fecal me agenomic eads om each subjec we e mapped agains he adminis e ed bifidobac e ial s ain genome sequences (AH1206 o AG1), conside ing only >99% homology ma ches. The esul s showed ha bo h AH1206 and AG1 s ains we e de ec able a he end o he 28- day ea men pe iod and hen dec eased a e he e mina ion o mic obial supplemen a ion in a s ain-specificmanne (Fig.3b, Sup- plemen a y Da a 15). Rema kably, he ela i e abundance o AH1206 mapped eads emained significan ly highe compa ed wi h he p e- ea men baseline e en 200 days a e comple ion o ea men (Wilcoxon es , p- alue < 0.01) (Fig. 3b, Supplemen a y Da a 15). In con as , he ela i e a e age abundance o AG1 mapped eads was no significan ly highe when compa ed o he si ua ion be o e he mic obial in e en ion as ea ly as 28 days a e in e up ion o con- sump ion (Wilcoxon es , p- alue > 0.05), indica ing ha he le el o AG1 one mon h a e ea men was e e ed o ha obse ed in he p e- ea ed mic obiomes. In iguingly, when assessing he deg ee o long- e m AH1206 coloniza ion among emale and male olun ee s, we iden ified gende - ela ed di e ences in he le el o s ain pe sis- ence. Indeed, a he pe sis ence es ime-poin (a ound 200 days o ollow-up), he a e age ela i e abundance o AH1206- ela ed mapped eads om emales was ound o be significan ly highe compa ed wi h hei own baseline (Wilcoxon es , p- alue < 0.01), while male pa ici- pan s did no exhibi such long- e m pe sis ence o AH1206 (Wilcoxon es o a e age mapped eads a 200 days s. baseline, p- alue > 0.05) (Fig. 3c, Supplemen a y Da a 15). O e all, hese findings possibly indica e ha GH136 posi i ely impac s he ( emale hos -associa ed) pe sis ence o B. longum subsp. longum s ains. E alua ion o he molecula in e ac ion o pe sis en and non- pe sis en B. longum s ains wi h human in es inal cells h ough ansc ip omics analyses To assess he ole o he non-ubiqui ously p esen B. longum-encoded GH136 in hos -mic obe c oss- alk and o in es iga e molecula in e - ac ions be ween pe sis en and non-pe sis en B. longum subsp. longum s ains and he hos cells, we applied an in i o app oach in ol ing human cell lines placed in con ac wi h bac e ial cells (Fig.S1).Specifically, we cul i a ed Caco2/HT29-MTX cell monolaye s in di ec physical con ac wi h B. longum subsp. longum PRL2022 o 1898B s ains, which had been selec ed based on he p esence o absence o he gene encoding he accesso y GH136, espec i ely. Subsequen ly, he ansc ip omes om bo h bac e ial and human cell lines we e in es iga ed h ough RNA-Seq expe imen s aimed a e alua ing he di e en ially exp essed genes (DEGs) be ween each ea men (PRL2022- and 1898B-Caco2/HT29-MTX con ac ) and he espec i e con ol condi ions (absence o con ac ), conside ing s a is ically significan a old-change ≥2a ap- alue ≤0.05 a e co - ec ion o mul iple compa isons using he False Disco e y Ra e (FDR) p ocedu e (Figs. S7, S8, Table S16, Supplemen a y Da a 17). Following hos cell con ac , a o al o 334 and 492 bac e ial genes om PRL2022 and 1898B s ains, espec i ely, we e classified as DEGs when compa ed o con ol samples (Supplemen a y Da a 16). Among he s a is ically significan up egula ed ansc ip omes (~50% o o al DEGs) (Fig. 4a, Fig. S6), we ound ansc ip s co esponding o genes encoding p iming and p ocessing glycosyl ans e ases in ol ed in exopoly- saccha ide (EPS) p oduc ion, se e al ca bohyd a e and amino acid modi ying enzymes, and a ious p o ein kinases (Supplemen a y Da a 16). Focusing on (bifido)bac e ial hos -glycan and ca bohyd a e me abolism genes, i was ound ha bo h PRL2022- and 1898B- ela ed ansc ip omes showed significan ly highe exp ession o he GH101- co esponding gene (JL750_RS06690, BLSL_RS08555) when compa ed o con ol samples, in addi ion o genes p edic ed o encompass a numbe o 16–19 ABC anspo e s in ol ed in ca bohyd a e up ake, including ha iden ified abo e among he 14 genes sha ed be ween B. longum and B. bifidum (Fig. 4a, b, d, e; Fig. S7). Mo eo e , he GH136 gene (JL750_RS09800), p esen only in B. longum subsp. longum PRL2022, was shown o be up egula ed (Fig. 4b, Supplemen a y Da a 16). As a esul , in es inal bac e ial coloniza ion appea s o be a mul- i ac o ial p ocess ha in ol es a ious ca bohyd a e-modi ying enzymes and mic obial su ace componen s. These appea o co o- bo a e he in ol emen o he (bifido)bac e ial mucin-deg ading enzymes GH101 and GH136 in hos -mic obe in e play and mucosal su ace coloniza ion. Howe e , as mammalian ho mones may a ec bac e ial gene exp ession64–66, we e alua ed whe he he endoc ine milieu con ained in he e al bo ine se um (FBS, used in combina ion wi h DMEM o human cell cul u e, see he “Me hods”sec ion) could impac he exp ession o he GH136 gene. Specifically, we di e en ially g ew PRL2022 on he cul u e medium DMEM wi h and wi hou FBS, ob aining no significan changes in RT-qPCR-based GH136 gene exp ession esul s ( - es p- alue > 0.05). Fu he mo e, he ansc ip ome o human cell lines placed in con ac wi h B. longum subsp. longum was assessed and compa ed wi h ha achie ed in he absence o bac e ial cells. No ably, a o al o 1253 (874 up egula ed) and 1404 (892 up egula ed) hos cell- ela ed genes we e iden ified as DEGs in he PRL2022-exposed and 1898B-exposed g oups, espec i ely (Fig. S8, Supplemen a y Da a 17). Among he B. longum-induced up egula ed hos ansc ip s, we ound genes encoding pa e n ecogni ion ecep o s, which eac o bac e ia (e.g., Toll-like ecep o s)67,68, and cell signaling molecules ha aid cell communica ion in immune esponses (e.g., cy okines)69 (Supplemen a y Da a 17). Focusing on he exp ession o human mucin genes, ansc ip ome analysis o Caco2/HT29-MTX cells upon exposu e o B. longum subsp. longum PRL2022 (which ha bo s a GH136-encoding gene) e ealed up- egula ion o genes encoding mucin5B (mucus/gel- o ming), mucin3A, and mucin17 (cell su ace-associa ed)70, as well as glucosaminyl (N- ace yl) ans e ase 3, which ca alyzes he o ma ion o co e 2 and co e 4 O-glycans on mucin- ype glycop o eins71,72 (Fig. 4c, Supplemen a y Da a 17). Rema kably, only mucin3A and mucin5B we e exp essed a significan ly highe le els in he non-pe sis en B. longum subsp. longum 1898B-exposed hos -de i ed ansc ip ome compa ed wi h con ol samples (Fig. 4 , Supplemen a y Da a 17). O e all, hese findings sugges ha B. longum subsp. longum s ains significan ly influence he ansc ip ome o human in es inal cells, modula ing he exp ession o genes in ol ed in he syn hesis o mucus laye componen s. Rema kably, B. longum subsp. longum PRL2022 (encoding he GH136 enzyme) appea ed able o s imula e he exp ession o hos mucin and mucin- ela ed genes o a g ea e ex en (100% inc ease in he numbe o up egula ed hos mucin- ela ed genes) han he conspecific 1898B s ain lacking he GH136 gene, implying a possible compe i i e ad an age and enhanced pe sis ence in he gu o ( emale) human hos s. A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 7 Discussion Se e al s udies ha e demons a ed ha e ical ansmission om mo he o in an s is a pi o al ou e o he ea ly es ablishmen o (ce - ain) membe s o he gu mic obio a, which can pe sis ac oss sub- sequen s ages o human li e, al hough wi h a educed abundance8,28. In his con ex , i is pa icula ly impo an o he human emale hos o sus ain he pe sis ence o hose ea ly colonize s ha may la e be ma e nally ansmi ed onewgene a ions. In he cu en s udy, h ough mul i-omics app oaches, we e ealed ha B. bifidum and specifics ainso B. longum subsp. longum a) Baseline Mic obial in e en ion Follow up Baseline Mic obial in e en ion Follow up AH1206 AG1 Fecal sample 28 0 Day Day b) c) A e age ela i e abundance o mapped eads Rela i e abundance o mapped eads AG1 AH1206 3% 2% 1.5% 1% 0.5% 0% 2.5% AH1206 T ea men wi h AG1 T ea men wi h AH1206 Pe sis ence a 200 days pos - ea men wi h AH1206 Pe sis ence a 200 days pos - ea men wi h AH1206 P e- ea men (baseline) ela i e abundance o AH1206 mapped eads Rela i e abundance o AH1206 mapped eads * 028 56 228 AG1 * 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 3% 2% 1% 0% 4% A e age ela i e abundance o mapped eads * * * * Fig. 3 | Analysis o da a om human e ospec i e clinical s udies based on he supplemen a ion o B. longum subsp. longum AH1206 and AG1 s ains. Panel (a) shows he expe imen al ou line o he human e ospec i e ials conside ed in his s udy. Panel (b) epo s he a e age ela i e abundance o mapped eads o B. longum subsp. longum AG1206 and AG1 in he me agenomic samples du ing he bac e ial supplemen a ion and ollow-up. The significan di e ence in he a e age abundance o AH1206 and AG1 mapped eads be ween ollow-up (pe sis ence) and he co esponding baseline is indica ed by an as e isk in he Box and Whiske plo (**p- alue < 0.01; Wilcoxon signed- ank es ,p- alue = 0.009). The boxes a e de e mined by he 25 h and 75 h pe cen iles. The whiske s a e de e mined by 1.5 in e qua ile ange (IQR). The line in he boxes ep esen s he median, while he c oss ma ke (X) ep esen s he a e age. In panel (c), he end o AG1206 and AG1 mapped eads ( ela i e abundance) ac oss he co esponding in e en ional s u- dies (n=22andn= 10, espec i ely) is shown o each sample. Di e ences in he pe sis ence o AH1206 s ains be ween emale and male popula ions (n=21,11 emales and 10 males) a e de ailed in he ba plo s depic ing he a e age ela i e abundance o AH1206 mapped eads de ec ed a baseline (n= 21, 11 emales and 10 males) and a e he e mina ion o ea men (n= 21, 11 emales and 10 males) (**p- alue < 0.01; Wilcoxon signed- ank es ,p- alue = 0.039). A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 8 can es ablish a long-las ing coloniza ion beha io p e e en ially in he in es inal en i onmen o emales compa ed wi h male indi iduals. No ably, hese axa a e enowned o hei highes le el o e ical ansmission among bifidobac e ial species22,73. In e es ingly, sc een- ing o he gene ic de e minan s possibly in ol ed in ( emale) human gu pe sis ence led o he iden ifica ion o wo mucin-deg ading GH amilies, GH101 and GH136, p esen in all B. bifidum and pa icula B. longum subsp. longum s ains. Specifically, he non-ubiqui ous dis ibu ion o he GH136 gene wi hin B. longum subsp. longum axon highligh s he in ol emen o his gene in mo he - o-in an e ical ansmission and led o i s iden ifica ion as a s ain-specific gene ic key de e minan o s able GH101 T ansc ip o B. longum subsp. longum 1898B when g own in con ac wi h Caco2/HT29-MTX T ansc ip o B. longum subsp. longum 1898B om con ol Bac e ial coun e pa B. longum 1898B- Caco2/HT29 con ac e) 19 10 5 49 1 118 ABC anspo e glycosyl ans e ases amino acid ABC anspo e hypo he ical p o ein se ine/ h eonine p o ein kinase O he s a) d) glucosaminyl (N-ace yl) ans e ase 3, mucin ype mucin 3A, cell su ace associa ed mucin 5B, oligome ic mucus/gel- o ming 16 82 13 1 167 ABC anspo e glycosyl ans e ases amino acid ABC anspo e hypo he ical p o ein se ine/ h eonine p o ein kinase O he s T ansc ip o euka yo ic hos om B. longum subsp. longum PRL2022-con ac condi ion T ansc ip o euka yo ic hos om con ol T ansc ip o euka yo ic hos om B. longum subsp. longum 1898B-con ac condi ion T ansc ip o euka yo ic hos om con ol Euka yo ic Hos coun e pa in B. longum PRL2022-Caco2/HT29 con ac Euka yo ic Hos coun e pa in B. longum 1898B-Caco2/HT29 con ac GH101 GH136 T ansc ip o B. longum subsp. longum PRL2022 when g own in con ac wi h Caco2/HT29-MTX T ansc ip o B. longum subsp. longum PRL2022 om con ol Bac e ial coun e pa in B. longum PRL2022- Caco2/HT29 con ac b) c) ) No malized coun eads (TMM) No malized coun eads (TMM) No malized coun eads (TMM) mucin 3A, cell su ace associa ed mucin 17, cell su ace associa ed mucin 5B, oligome ic mucus/gel- o ming ENSG00000169876.14 ENSG00000169894.18 ENSG00000140297.13 ENSG00000117983.17 ENSG00000169894.18 ENSG00000117983.17 JL750_RS9800 JL750_RS06690 N o m a lize d c o un e a d s ( T MM) 0 0 100 200 300 400 0 0 200 400 600 800 1000 BLSL_RS08555 N o m a l ize d c o un e a d s ( T MM) 9 E E E E p e e e n i B l P R P L 2022 C 2/H T 29 0 100 200 300 p g N S e n m 050100150 A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 9 80. Lugli, G. A. e al. Isola ion o no el gu bifidobac e ia using a combina ion o me agenomic and cul i a ion app oaches. Gen- ome Biol. 20,96(2019). 81. Hya , D. e al. P odigal: p oka yo ic gene ecogni ion and ans- la ion ini ia ion si e iden ifica ion. BMC Bioin o ma. 11,119(2010). 82. Lugli,G.A.,Milani,C.,Mancabelli,L.,VanSinde en,D.&Ven u a,M. MEGAnno a o : a use - iendly pipeline o mic obial genomes assembly and anno a ion. FEMS Mic obiol. Le . 363, nw049(2016). 83. Lugli, G. A. e al. MEGAnno a o 2: a pipeline o he assembly and anno a ion o mic obial genomes. Mic obiome Res. Rep. 2, 15 (2023). 84. an Dijk, L. R. e al. S ainGE: a oolki o ack and cha ac e ize low- abundance s ains in complex mic obial communi ies. Genome Biol. 23, 74 (2022). 85. Olm, M. R., B own, C. T., B ooks, B. & Banfield, J. F. dRep: a ool o as and accu a e genomic compa isons ha enables imp o ed genome eco e y om me agenomes h ough de- eplica ion. ISME J. 11, 2864–2868 (2017). 86. Pa ks,D.H.,Imel o ,M.,Skenne on,C.T.,Hugenhol z,P.& Tyson, G. W. CheckM: assessing he quali y o mic obial genomes eco e ed om isola es, single cells, and me agenomes. Genome Res. 25,1043–1055 (2015). 87. Lomba d, V., Golaconda Ramulu, H., D ula, E., Cou inho, P. M. & Hen issa , B. The ca bohyd a e-ac i e enzymes da abase (CAZy) in 2013. Nucleic Acids Res. 42, D490–5(2014). 88. Zhao, Y. e al. PGAP: pan-genomes analysis pipeline. Bioin o - ma ics 28,416–418 (2012). 89. En igh , A. J., Van Dongen, S. & Ouzounis, C. A. An e ficien algo i hm o la ge-scale de ec ion o p o ein amilies. Nucleic Acids Res. 30,1575–1584 (2002). 90. Le unic, I., Khedka , S. & Bo k, P. SMART: ecen upda es, new de elopmen s and s a us in 2020. Nucleic Acids Res. 49, D458–D460 (2021). 91. Wa e house, A. e al. SWISS-MODEL: homology modelling o p o ein s uc u es and complexes. Nucleic Acids Res. 46, W296–W303 (2018). 92. Ta acchini, C. e al. Assessing he genomic a iabili y o ga d- ne ella aginalis h ough compa a i e genomic analyses: e olu- iona y and ecological implica ions. Appl. En i on. Mic obiol. 87, 1–16 (2020). 93. Langmead, B. & Salzbe g, S. L. Fas gapped- ead alignmen wi h Bow ie 2. Na . Me hods 9,357–359 (2012). 94. Ande s,S.,Pyl,P.T.&Hube ,W.HTSeq—a Py hon amewo k o wo k wi h high- h oughpu sequencing da a. Bioin o ma ics 31, 166 (2015). 95. Li, H. & Du bin, R. Fas and accu a e long- ead alignmen wi h Bu ows-Wheele ans o m. Bioin o ma ics 26,589–595 (2010). 96. Lo e,M.I.,Hube ,W.&Ande s,S.Mode a edes ima iono old change and dispe sion o RNA-seq da a wi h DESeq2. Genome Biol. 15, 550 (2014). 97. Mo aza i, A., Williams, B. A., McCue, K., Schae e , L. & Wold, B. Mapping and quan i ying mammalian ansc ip omes by RNA- Seq. Na . Me hods 5,621–628 (2008). 98. Bianchi, M. G. e al. Ca echin and p ocyanidin B 2 modula e he exp ession o igh junc ion p o eins bu do no p o ec om inflamma ion-induced changes in pe meabili y in human in es - inal cell monolaye s. Nu ien s 11, 2271 (2019). 99. Se afini, F. e al. E alua ion o adhesion p ope ies and an i- bac e ial ac i i ies o he in an gu commensal Bifidobac e ium bifidum PRL2010. Anae obe 21,9–17 (2013). 100. Dobin, A. e al. STAR: ul a as uni e sal RNA-seq aligne . Bioin- o ma ics 29,15–21 (2013). 101. Robinson,M.D.,McCa hy,D.J.&Smy h,G.K.edgeR:aBio- conduc o package o di e en ial exp ession analysis o digi al gene exp ession da a. Bioin o ma ics 26,139–140 (2010). Acknowledgemen s We hank GenP obio S l o he financial suppo o he Labo a o y o P obiogenomics. Pa o his esea ch is conduc ed using he High- Pe o mance Compu ing (HPC) acili y o he Uni e si y o Pa ma. This esea ch has financially been suppo ed by he P og amme “FIL-Quo a Incen i an e”o Uni e si y o Pa ma and co-sponso ed by Fondazione Ca ipa ma”. D. .S. is a membe o APC Mic obiome I eland unded by Science Founda ion I eland (SFI), h ough he I ish Go e nmen ’s Na ional De elopmen Plan (G an no. SFI/12/RC/2273-P1 and SFI/12/RC/ 2273-P2). G.T. has been suppo ed by “Fondazione Ca ipa ma”in he amewo k o he p ojec en i led “Pa ma Mic obio a”.LMVhasbeen suppo ed by by “P og amma Ope a i o Nazionale 2014–2020 o he I alian Minis y o Uni e si y and Resea ch. The unding om P ojec AGL2017-83653R (Spanish “Minis e io de Ciencia, Inno ación y Uni- e sidades (MCIU)”,“Agencia Es a al de In es igación (AEI)”and FEDER) is also acknowledged. Au ho con ibu ions Concep ualiza ion, supe ising, coo dina ion, and e ising o manu- sc ip by F.T., O.V., D. .S., C.M., and M.V. RNAseq and bac e ial DNA sequencing expe imen s by G.A., G.L., R.A., and A.V. Collec ion da a and genomic, me agenomic, and ansc ip omic da a p ocessing and ana- lysis by C.T., G.A., S.M.R., M.B., C.A., L.M.V., S.A., G.T., M.C., and M.G. Assis ing in compu a ional amewo k de elopmen , implemen a ion, and compu a ional analyses by F.F., G.A.L., and L.M. D a ing o manu- sc ip by C.T., G.A., and C.M. All au ho s c i ically e ised and app o ed he manusc ip . Compe ing in e es s The au ho s decla e no compe ing in e es s. Addi ional in o ma ion Supplemen a y in o ma ion The online e sion con ains supplemen a y ma e ial a ailable a h ps://doi.o g/10.1038/s41467-023-39931-2 Co espondence and eques s o ma e ials should be add essed o Ch is ian Milani o Ma co Ven u a. Pee e iew in o ma ion Na u e Communica ions hanks he anon- ymous e iewe s o hei con ibu ion o he pee e iew o his wo k. A pee e iew file is a ailable. Rep in s and pe missions in o ma ion is a ailable a h p://www.na u e.com/ ep in s Publishe ’sno eSp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a filia ions. Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons license, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons license and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To iew a copy o his license, isi h p://c ea i ecommons.o g/ licenses/by/4.0/. © The Au ho (s) 2023 A icle h ps://doi.o g/10.1038/s41467-023-39931-2 Na u e Communica ions | (2023) 14:4220 16