Pex3 promotes formation of peroxisome-peroxisome and peroxisome-lipid droplet contact sites

Pex3 p omo es o ma ion o pe oxisome-pe oxisome and pe oxisome-lipid d ople con ac si es Lucía Amado1, Louis Pe ci ull2,3, Rico F anzkoch4,5,6,Vico Fla eme sch1, Eleni Joana B üggemann7, Olympia Eka e ini Psa haki4,5, Maya Schuldine 8, Ma ia Bohne 2,3, Ma g e H. Bülow7& Ayelén González Mon o o1,5 Pe oxisomes a e ubiqui ous o ganelles ha media e cen al me abolic unc ions, such as a y acid β- oxida ion, as well as di e se issue- and o ganism-specic p ocesses. Memb ane con ac si es, egions o close apposi ion wi h o he o ganelles o di ec communica ion, a e cen al o se e al aspec s o hei li e cycle. Pex3 is a conse ed mul i unc ional pe oxisomal ansmemb ane p o ein ha is in ol ed in he inse ion o pe oxisomal memb ane p o eins, in pexophagy, and in he o ma ion o memb ane con ac si es. He e, we show ha high Pex3 le els in Saccha omyces ce e isiae induce he o ma ion o pe oxisome clus e s su ounded by lipid d ople s, media ed by pe oxisome-pe oxisome and pe oxisome-lipid d ople con ac si es.This clus e ing occu s independen ly o Pex3 pa ne s in o he p ocesses Pex19, Inp1, and A g36. The cy osolic domain o Pex3 binds pe oxisomes, sugges ing a di ec ole in homo ypic con ac si e o ma ion. Lipid d ople -pe oxisome con ac si es equi e he lipid d ople -localized iacylglyce ol lipaseTgl4, which is en iched a his in e ace along wi h o he lipases. Pex3 o e exp ession in D osophila melanogas e simila ly al e s pe oxisome and lipid d ople mo phology and p omo es con ac si e o ma ion.Toge he , ou esul s oe no el molecula insigh s in o homo ypic pe oxisome con ac si es and pe oxisome-lipid d ople con ac si es ac oss species. Pe oxisomes a e ound in mos euka yo ic cells and a e he place o impo an me abolic eac ions, including he oxida ion o a y acids, he syn hesis o e he lipids, and he de oxica ion o hyd ogen pe oxide and glyoxyla e. While some pe oxisomal unc ions a e issue- o o ganism-specic, o he s a e highly conse ed, e.g. he oxida ion o a y acids, a p ocess ha is i ually ubiqui ous1. Impai ed pe oxisome biogenesis as well as de ec s in pe oxisomal me abolic pa hways esul in se e e human diseases collec i ely known as pe oxisomal diso de s1. Memb ane con ac si es a e s uc u ally dened egions o close o ganelle apposi ion wi hou memb ane usion2,3. A cen al unc ion o con ac si es is he exchange o ma e ial among he compa men s, including he anspo o luminal ma e ial and o memb ane lipids. Fu he mo e, he physical a achmen o he o ganelle memb anes can aec o ganelle usion, ssion, and posi ioning4. Memb ane con ac si es exis be ween i ually all pai s o o ganelles5–7 and coo dina e mul i-o ganelle p ocesses4,8. Con ac si es be ween pe oxisomes and o he o ganelles play impo an oles in die en aspec s o he pe oxisome li e cycle9. Fo example, in he yeas Saccha omyces ce e isiae, pe oxisome con ac si es wi h he endoplasmic e iculum aec pe oxisome p oli e a ion10,11 and he o ma ion o pe oxisome con ac si es wi h he cell pe iphe y de e mines he dis ibu ion o hese o ganelles among mo he and daugh e cells12. Pex3 is a pe oxisomal memb ane p o ein wi h mul iple unc ions and in e ac o s. I is in ol ed in he a ge ing o pe oxisomal memb ane p o eins, by ac ing as a docking ac o o Pex19, a cy osolic ecep o o 1Cellula Communica ion Labo a o y, Depa men o Biology/Chemis y, Osnab ück Uni e si y, Ba ba as asse 13, 49076 Osnab ück, Ge many. 2Ins i u e o Cell Dynamics and Imaging, Uni e si y o Müns e , Von-Esma ch- S . 56, 48149 Müns e , Ge many. 3Cells in Mo ion In e acul y Cen e (CiM), Uni e si y o Müns e , Müns e , Ge many. 4iBiOs-In eg a ed Bioimaging Facili y, Uni e si y o Osnab ück, Osnab ück, Ge many. 5Cen e o Cellula Nanoanaly ic Osnab ück (CellNanOs), Ba ba as asse 11, 49076 Osnab ück, Ge many. 6Di ision o Mic obiology Depa men o Biology/Chemis y, Osnab ück Uni e si y, Ba ba as asse 11, 49076 Osnab ück, Ge many. 7G oup Memb ane Con ac Si es, CURE3D Resea ch Lab, Clinic o Ca dio ascula Su ge y, Uni e si y Hospi al Düsseldo , Moo ens aße 5, 40225 Düsseldo , Ge many. 8Depa men o Molecula Gene ics, The Weizmann Ins i u e o Science, 7610001 Reho o , Is ael. email: ayelen.gonzalez.mon o [email protected] OPEN Scien ic Repo s | (2025) 15:24480 1 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s pe oxisomal memb ane p o ein p ecu so s. Ano he unc ion o Pex3 is he ec ui men o he pexophagy ecep o A g36. Bo h o hese unc ions a e conse ed in he human Pex3 p o ein13–19. Pex3 has addi ional oles in con ac si e o ma ion. I ac s as a memb ane ancho o Inp1, a e he ing p o ein in ol ed in he e en ion o pe oxisomes in he mo he cell du ing cell di ision, h ough i s in e ac ion wi h he cell co ex12,20. In he yeas Hansenula polymo pha ex ensi e acuole-pe oxisome con ac si es a e o med du ing g ow h phases ha equi e pe oxisomal g ow h. Pex3 is en iched a hese con ac si es, and o e exp ession o he p o ein esul s in con ac si e expansion du ing g ow h in glucose, sugges ing ha Pex3 migh be a e he 21. is p omp ed us o add ess he eec s o Pex3 o e exp ession in Saccha omyces ce e isiae. We nd ha cells ha o e exp ess Pex3 con ain pe oxisome clus e s su ounded by lipid d ople s, which a e o med by pe oxisome-pe oxisome and pe oxisome-lipid d ople con ac si es. We u he show ha hese con ac si es a e independen o all he known Pex3 in e ac o s. Ins ead, ecien o ma ion o pe oxisome-lipid d ople con ac si es equi es he lipid d ople -localized TAG-lipase Tgl4 bu no o he lipid d ople lipases, e en hough se e al o hem a e en iched in he in e ace. In e es ingly, simila eec s o Pex3 o e exp ession on pe oxisome mo phology and ex ended con ac wi h he lipid d ople s we e also obse ed in D osophila melanogas e , showing ha hese aspec s o Pex3 unc ion a e conse ed o me azoa. Al oge he , ou ndings expand ou unde s anding o pe oxisomal con ac si es, and o he mul i- unc ional p o ein Pex3. Resul s O e exp ession o Pex3 causes a change in he mo phology o pe oxisomes and lipid d ople s In Hansenula polymo pha, Pex3 was obse ed o be en iched in memb ane con ac si es be ween pe oxisomes and he acuole, and o e exp ession o his p o ein esul s in an ex ension o hese con ac si es21. us, we decided o es he pheno ype o Pex3 o e exp ession in Saccha omyces ce e isiae. O e exp ession om he s ong cons i u i e TEF1 p omo e esul ed in a mo phological change in pe oxisomes ma ked by mChe y di ec ed o he pe oxisomal lumen by a PTS1 signal consis ing o a se ine-lysine-leucine sequence (mCh-SKL). While con ol cells show on a e age 5.5 mCh-SKL posi i e s uc u es, cells o e exp essing Pex3 con ain mainly one s uc u e (Fig.1A and B). is s uc u e was ound in close p oximi y o he acuole in 55% o cells. We con med hese obse a ions by using o he pe oxisomal ma ke s, Pex3 i sel and Pex14, ob aining simila esul s (Supplemen al Fig.1A and B). We analyzed he dis ibu ion o his pe oxisomal s uc u e among mo he and daugh e cells and ound ha i can be ound in bo h compa men s, wi h some p e e ence o he bud (Supplemen al Fig.1C). To cha ac e ize his s uc u e we sough o desc ibe i s molecula mic oen i onmen by p oximi y bio inyla ion22, by agging Pex3 wi h Tu boID23 a i s C- e minus, which aces he cy osol. Cells exp essing Pex3-Tu boID unde he TEF1 p omo e we e incuba ed wi h bio in o 3h and he bio inyla ed p o eins we e isola ed by ani y ch oma og aphy using a s ep a idin ma ix. e bai p o ein Pex3 was among he mos highly en iched p o eins, as was i s known in e ac o Pex19 (Fig.1C). Gene On ology (GO) Te m en ichmen analysis showed he GO e m “Pe oxisome” as he mos ep esen ed anno a ion among ou en iched p o eins, as expec ed. In e es ingly, he GO e m “Lipid d ople ” was also signican ly en iched (P alue = 0.032, Fig.1C). is p omp ed us o add ess he subcellula localiza ion o lipid d ople s when Pex3 is o e exp essed by uo escence mic oscopy. is analysis e ealed ha unde hese condi ions, lipid d ople s a e s ongly ec ui ed o he pe oxisomal s uc u e (Fig.1D). We obse ed and quan ied he p esence o LDs in close p oximi y o pe oxisomal s uc u es in 90 cells om h ee independen expe imen s. Upon Pex3 o e exp ession, 95% o he pe oxisomal s uc u es we e in close p oximi y o o comple ely su ounded by lipid d ople s, whe eas he emaining 5% o pe oxisomal s uc u es did no p esen any lipid d ople in hei icini y (Fig.1D). e size o lipid d ople s is also sligh ly inc eased unde hese condi ions (Supplemen al Fig.1D). We conclude ha o e exp ession o Pex3 causes a change in he mo phology o pe oxisomes and lipid d ople s, esul ing in he obse a ion o a single pe oxisomal s uc u e pe cell, which is in close p oximi y o lipid d ople s. Pex3 o e exp ession induces a clus e o pe oxisomes su ounded by lipid d ople s, which includes pe oxisome-pe oxisome and pe oxisome-lipid d ople con ac si es To unde s and he cha ac e is ics o he s uc u e o med by lipid d ople s and pe oxisomes upon Pex3 o e exp ession, we sough o enla ge he s uc u e o gain spa ial esolu ion. is was achie ed by dele ing PEX11, which esul s in enla ged pe oxisomes24,25. Addi ionally, we cul u ed he cells in he p esence o olea e, which esul ed in enla ged lipid d ople s. ese condi ions allowed sucien spa ial esolu ion o dis inguish he pe oxisomal ma ix om he memb ane, as pe oxisomal memb ane p o eins (Pex13 and Pex3) we e obse ed o su ound BFP di ec ed o he pe oxisomal lumen by a C- e minal PTS1 signal (BFP-SKL) (Fig.2A and B, Supplemen al Fig.1E–G). Simila o wha we obse ed be o e, Pex3 o e exp ession caused clus e ing o he pe oxisomal signal. e inc eased esolu ion p o ided by hese condi ions allowed us o obse e ha hese clus e s con ain mul iple indi idual maxima o he BFP-SKL signal (Supplemen al Fig.1E–G). Figu e2A and B show examples o hese clus e s, as well as line p oles ac oss hem, illus a ing ha he pe oxisomal memb ane p o eins show peaks be ween he indi idual maxima o he BFP-SKL signal (Fig.2A and B). is indica es ha he pe oxisomal s uc u e o med upon Pex3 o e exp ession co esponds o a clus e o mul iple pe oxisomes. e lipid d ople s ma ked by E g6-2xmKa e2 we e obse ed o su ound he clus e o pe oxisomes. Gi en he close and specic p oximi y obse ed be ween hese o ganelles upon o e exp ession o Pex3, we easoned ha such an o ganelle clus e would likely be o med by pe oxisome-pe oxisome and pe oxisome- lipid d ople con ac si es. To es his, we pe o med on-sec ion CLEM omog aphy o cells o e exp essing Pex3-mKa e2, using he mKa e2 signal o loca e he s uc u es. is app oach e ealed clus e s o pe oxisomes su ounded by lipid d ople s ha included pe oxisome-pe oxisome and pe oxisome-lipid d ople con ac si es (Fig.2C–D). Figu e2C shows a on-sec ion co ela i e uo escence mic oscopy and TEM image om a cell Scien ic Repo s | (2025) 15:24480 2 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ o e exp essing Pex3-mKa e2, used o iden i y he ele an egion o TEM omog aphy. Figu e2D shows imaging o his egion by TEM omog aphy, o e layed wi h a 3D model econs uc ing he obse ed o ganelles. Figu e2E and F and Supplemen al ideos 1 and 2 show example egions o he omog am, con aining pe oxisome-lipid d ople and pe oxisome-pe oxisome con ac si es, espec i ely. We asked i he o ma ion o pe oxisome-pe oxisome con ac si es and pe oxisome-lipid d ople con ac si es occu ed independen ly, o whe he he clus e ing o he wo ypes o o ganelles was in e linked. To es his, Fig. 1. O e exp ession o Pex3 causes he o ma ion o a single pe oxisomal s uc u e su ounded by lipid d ople s. (A–B) O e exp ession o Pex3 p oduces he collapse o all pe oxisomal signal in o one s uc u e. Panel A shows ep esen a i e pic u es o a s ain exp essing mChe y-SKL cons uc o isualize he lumen o he pe oxisomes, ei he wi h Pex3 a endogenous le els (Con ol) o o e exp essed (TEF1p -PEX3) and he acuola lumen s ained wi h CMAC. Cell ou lines a e shown as whi e dashed lines. Scale ba : 2μm. Panel B shows he quan ica ion o he amoun o pe oxisomal s uc u es pe cell.  ee independen expe imen s we e pe o med and 30 cells we e analyzed o each expe imen and condi ion. Small diamonds co espond o indi idual cells, bigge diamonds co espond o he a e age o independen expe imen s. e die en s ains we e compa ed using an unpai ed wo- ailed S uden ’s - es . *** P < 0.001. (C) Tu bo ID o o e exp essed Pex3-Tu boID en iches pe oxisomal and lipid d ople s p o eins. Volcano plo showing ela i e p o ein in ensi y in a pull-down o bio inyla ed p o eins be ween a s ain o e exp essing Pex3 agged c- e minally wi h he Tu boID p o ein (TEF1p -PEX3-TID) and a wild ype con ol s ain (w ). Pe oxisomal p o eins a e ma ked in magen a and lipid d ople s p o eins in g een. GO e m en ichmen analysis o he g oup o p o eins signican ly en iched in he Pex3-TID pull-down showed an en ichmen o he GO Te ms “pe oxisomes” and “lipid d ople s”. (D) e o med pe oxisomal s uc u e is su ounded by lipid d ople s. Rep esen a i e pic u es o a s ain exp essing mChe y-SKL cons uc o isualize he lumen o he pe oxisomes, ei he wi h Pex3 a endogenous le els (Con ol) o o e exp essed (TEF1p -PEX3), he acuola lumen s ained wi h CMAC and lipid d ople s s ained wi h Bodipy. Scale ba : 2μm. e zoomed in egion shows a pe oxisomal s uc u e su ounded by lipid d ople s wi h a scale ba o 1μm. Cell ou lines a e shown as whi e dashed lines. Scien ic Repo s | (2025) 15:24480 3 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ we o e exp essed Pex3 in a s ain ha is de oid o lipid d ople s because i lacks he syn hases o iglyce ides and s e ol es e s, namely Dga1, L o1, A e1 and A e2 ( om he e on e med ΔLDs)26. O e exp ession o Pex3 caused he accumula ion o pe oxisomal signal in o a single s uc u e i espec i e o he absence o lipid d ople s, showing ha lipid d ople s a e no equi ed o he o ma ion o he pe oxisome-pe oxisome con ac si es (Fig.2G and H). The cy osolic domain o Pex3 in e ac s wi h pe oxisomes Pex3 is ancho ed o he pe oxisomal memb ane by a single ansmemb ane domain a i s N- e minus, which is sucien o cause a ge ing o he pe oxisomes, and con ains a globula cy osolic C- e minal domain27–29 (Fig.3A). To add ess a possible di ec in ol emen o Pex3 in con ac si e o ma ion, we exp essed he cy osolic domain (CD – amino acids 40–441) used o GFP a i s N- e minus and lacking he ansmemb ane egion (Fig.3A). We obse ed ha his cons uc localized a pe oxisomes, ma ked by Pex14-HaloTag (Fig.3B). e cons uc did no deco a e lipid d ople s ma ked by E g6-2xmKa e2, and was only obse ed en iched in hese s uc u es when pe oxisomal signal was also p esen (Fig.3B, see zoomed in o ganelles). Consis en ly, calcula ion o he Mande ´s coecien s M1 and M2 o GFP-Pex3(CD) wi h Pex14-HaloTag showed high alues, while he coecien s o o e lap wi h he E g6-2xmKa e2 signal we e much lowe and compa able o he ones obse ed be ween Pex14-HaloTag and E g6-2xmKa e2 (Fig.3C). Scien ic Repo s | (2025) 15:24480 4 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ To es i he in e ac ion o he cy osolic domain o Pex3 wi h pe oxisomes is s ong enough o enable o ganelle e he ing, we a icially di ec ed his domain o mi ochond ia. is was achie ed by agging i wi h he uo escen p o ein mKa e2 and an N- e minal Al a ag30, In addi ion, he mi ochond ial ou e memb ane ecep o Tom70 was agged wi h a nanobody ha ecognizes he Al a ag30, so i should di ec he Pex3(CD) o mi ochond ia (Fig.3D). Indeed, Al a-mKa e2-Pex3(CD) deco a es he mi ochond ial ne wo k unde hese condi ions, as con med by co-localiza ion wi h he mi ochond ial ma ke Ci 1-HaloTag (Fig.3E). We quan ied he minimum dis ance o each indi idual pe oxisome o he mi ochond ial ne wo k, o cells ei he exp essing o no exp essing he Al a-mKa e2-Pex3(CD) cons uc . e exp ession o he cons uc caused a s ong shi o smalle dis ances (Fig.3F). Figu e3E shows a ep esen a i e image o he expe imen , and addi ional examples a e shown in Supplemen al Fig.2. is da a indica es ha he in e ac ion o Pex3(CD) wi h pe oxisomes is s ong enough o induce o ganelle e he ing. Based on hese esul s we p opose ha he homo ypic pe oxisomal con ac si es a e o med by Pex3 being ancho ed o he pe oxisomal memb ane ia i s ansmemb ane domain, and addi ionally in e ac ing wi h o he pe oxisomes h ough i s cy osolic domain. In con as , he pe oxisome-lipid d ople con ac si e is likely no di ec ly media ed by Pex3, and only indi ec ly induced by i s o e exp ession. Known in e ac o s o Pex3 a e no in ol ed in o ming he pe oxisome-lipid d ople clus e Pex3 is a mul i unc ional p o ein in ol ed in die en p ocesses ela ed o he li e-cycle o pe oxisomes, including he a ge ing o pe oxisomal memb ane p o eins, pexophagy, and a ge ing o pe oxisomes o he co ex, which s ongly inuences pe oxisome inhe i ance12–17,19,29. ese unc ions a e media ed by he di ec in e ac ion o Pex3 wi h die en binding pa ne s (Fig.4A). Ta ge ing o pe oxisomal memb ane p o eins o he pe oxisomal memb ane in ol es i s in e ac ion wi h he cy osolic ecep o Pex1914,17 while i s ole in au ophagy is media ed by i s in e ac ion wi h he pexophagy ecep o A g3615. Finally, he e he ing o pe oxisomes o he cell co ex is media ed by Pex3 in e ac ing wi h Inp112,20 (Fig.4A). Nex , we es ed i any o he known in e ac o s o Pex3 is in ol ed in he o ma ion o he pe oxisome-lipid d ople clus e . To do his, we o e exp essed Pex3 in s ains lacking ei he ATG36 o INP1. e ep esen a i e mic oscopy images and quan ica ions shown in Fig.4B–G show ha he pheno ype caused by Pex3 o e exp ession in hese backg ounds does no die om he con ol cells, indica ing ha nei he Inp1 no A g36 a e equi ed o he o ma ion o his s uc u e. e ac ha A g36 is no equi ed o he o ma ion o his s uc u e sugges s ha i is no an in e media y in he p ocess o pexophagy. Howe e , we decided o es i pexophagy is induced by Pex3 o e exp ession, since his has been epo ed in mammalian cell lines19 In yeas , GFP is esis an o acuola deg ada ion, and hus he gene a ion o a ee GFP band has been used as a eadou o acuola deg ada ion o die en p o eins and o ganelles, including pe oxisomes31–33. us, we compa ed he appea ance o a ee GFP band in cells o e exp essing Pex3 wi h cells g own unde known pexophagy-inducing condi ions, namely g ow h in olea e medium ollowed by ni ogen s a a ion33. Figu e4H and Supplemen al Fig.3A, B and C show ha in cells exp essing Pex14-GFP, a ee GFP band appea s unde pexophagy-inducing condi ions (P.I.C), indica ing acuola deg ada ion o pe oxisomes. In con as , he lack o a ee GFP band in he s ain o e exp essing Pex3 when g own o loga i hmic phase in glucose (L.P.G.) shows ha o e exp ession o his p o ein alone does no induce pexophagy. In e es ingly, pexophagy can s ill be induced in his s ain, sugges ing ha i is also no blocked by he o ma ion o he pe oxisomal clus e s. To es he equi emen o PEX19, we could no dele e his gene, as his esul s in he absence o unc ional pe oxisomes34. We hus used an al e na i e s a egy, by add essing whe he o e exp ession o a mu an e sion o Pex3 ha does no in e ac wi h Pex19 s ill p omo es he o ma ion o he pe oxisome-lipid d ople clus e . Fig. 2. O e exp ession o Pex3 induces an accumula ion o pe oxisomes su ounded by LDs ha con ains Pex- Pex and Pex-LD con ac si es. (A–B) Enla ged pe oxisomes and LDs e eal ha he s uc u es con ain se e al maxima o pe oxisome lumen signal, wi h pe oxisomal memb ane be ween hem. Rep esen a i e pic u es o a s ain wi h o e exp essed Pex3 (TEF1p -PEX3), exp essing he BFP-SKL cons uc o isualize he lumen o he pe oxisomes, E g6-2xmKa e2 ma king he lipid d ople monolaye , and ei he Pex3 (A) o Pex14 (B) agged wi h mNeonG een as ma ke s o he pe oxisomal memb ane. To p oduce enla ged pe oxisomes and lipid d ople s, he cells con ain a dele ion o PEX11, and we e g own wi h olea e as he sole ca bon sou ce o 20hs. Cell ou lines a e shown as whi e dashed lines. Scale ba s: 1μm. Each g aph shows he signal o BFP- SKL and he co esponding pe oxisomal memb ane p o ein o e a line ac oss he s uc u e, depic ed in he me ged image. (C–F) On-sec ion CLEM omog aphy con ms ha he s uc u e in ol es Pex-Pex and Pex-LD con ac si es. Panel C shows a ep esen a i e image o on-sec ion CLEM done on a s ain wi h o e exp essed Pex3 (TEF1p -PEX3) and agged wi h 2xmKa e2. Scale ba : 600nm. Panel D shows a omog aphy image o he same sec ion o e layed wi h he 3D model ec ea ed om he images, showing pe oxisomes in magen a and lipid d ople s in yellow. Scale ba : 50nm. Panels E and F show zoomed in egions o he omog am as examples o he pe oxisome-lipid d ople (E) and pe oxisome-pe oxisome (F) con ac si es. Scale ba s: 20nm. (G–H) LDs a e no necessa y o he o ma ion o he clus e o pe oxisomes. Panel G shows ep esen a i e pic u es o s ains wi h o e exp essed Pex3 (TEF1p -PEX3) in con ol cells and in s ains ha canno p oduce lipid d ople s (ΔLDs), exp essing Pex14-2xmKa e2 o isualize he pe oxisomes, lipid d ople s we e s ained wi h Bodipy and he acuola lumen was s ained wi h CMAC. Cell ou lines a e shown as whi e dashed lines. Scale ba : 2μm. Panel H shows he quan ica ion o he amoun o pe oxisomal s uc u es pe cell.  ee independen expe imen s we e pe o med and 30 cells we e analyzed o each expe imen and condi ion. Small diamonds co espond o indi idual cells, bigge diamonds co espond o he a e age o independen expe imen s. e die en s ains we e compa ed using an unpai ed wo- ailed S uden ’s - es . n.s., no signican . ◂ Scien ic Repo s | (2025) 15:24480 5 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ e c ys al s uc u e o human Pex3 in complex wi h a agmen o human Pex19 has p e iously been sol ed. I was ound ha he in e ac ion wi h Pex19 is media ed by a egion cen e ed a ound HsPex3-T p104, which also includes Leu10716,39. Alignmen o he AlphaFold-gene a ed s uc u e p edic ion o ScPex3 wi h he s uc u e o HsPex3 indica ed ha his egion is highly conse ed and ha he equi alen esidues in ScPex3 a e T p128 and Leu131 (Fig.4I). We hus in oduced he mu a ions T p128Lys and Leu131Lys in ScPex3 (Pex3Mu ) and es ed he abili y o his mu an o in e ac wi h Pex19, suppo pe oxisome biogenesis, and o m he pe oxisome- lipid d ople s uc u es upon o e exp ession. Unlike w Pex3, Pex3Mu was no co-pu ied wi h msGFP2-Pex19, indica ing ha he mu a ions dis up his in e ac ion (Fig.4J and Supplemen al Fig.3 D, E and F). e Pex19- Pex3 in e ac ion has been epo ed o be equi ed o impo PTS1 con aining pe oxisomal ma ix p o eins14. Consis en ly, we obse ed ha exp ession o Pex3Mu in a s ain lacking endogenous Pex3 does no escue he impo o BFP-SKL in o pe oxisomes, con ming ha his mu a ion dis up s he in e ac ion (Fig.4K). Figu e4L– N con ain ep esen a i e mic oscopy images and he co esponding quan ica ion showing ha o e exp ession o his mu an in a backg ound con aining endogenous le els o w Pex3 o ha e unc ional pe oxisomes, causes he same mo phological pheno ype as he o e exp ession o w Pex3. us, he in e ac ion wi h Pex19 is no equi ed o Pex3 o induce clus e ing o pe oxisomes and lipid d ople s. The iacylglyce ol lipaseTgl4 is in ol ed in he o ma ion o he pe oxisome-lipid d ople con ac si es We sough o iden i y addi ional molecula playe s in ol ed in he o ma ion o con ac si es upon Pex3 o e exp ession. Since many con ac si e e he p o eins a e in ol ed in he o ma ion o mo e han one con ac si e35–41, we es ed he in ol emen o known e he p o eins o lipid d ople s o pe oxisomes wi h o he o ganelles. Scien ic Repo s | (2025) 15:24480 6 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ Dele ion o he pe oxisomal-mi ochond ial e he s Pex34 o Fzo11 did no aec o ma ion o Pex3-dependen pe oxisome and lipid d ople clus e s (Supplemen al Fig.4A), indica ing ha hey a e no equi ed o o ma ion o he con ac si es. We also es ed he in ol emen o he he splicing-gene a ed pai o p o eins Ldo16-Ldo45, which e he lipid d ople s o he acuole42,43. Dele ion o hese genes did no aec he clus e ing o pe oxisomes o lipid d ople s, no he p oximi y o he clus e ed s uc u e o he acuole (Supplemen al Fig.4B). Finally, i was ecen ly shown ha he human p o ein M1 Spas in e he s lipid d ople s o pe oxisomes44. We dele ed he yeas homolog Sap1, bu obse ed no al e a ion o he Pex3-dependen pheno ype (Supplemen al Fig.4C). Pe oxisomes and lipid d ople s a e also linked by he ole o he endoplasmic e iculum du ing hei biogenesis. Bo h lipid d ople s and p e-pe oxisomal esicles bud om a subdomain o he endoplasmic e iculum ma ked by he e iculon-like p o ein Pex3045. I was epo ed ha he same domain can bind simul aneously o a lipid d ople and a pe oxisome, causing a close associa ion be ween hem46. Dele ion o PEX30 did no aec he o ma ion o he pe oxisomal and lipid d ople clus e s (Supplemen al Fig.4D), indica ing ha his p o ein is no di ec ly in ol ed in o ming hem. We nex pe o med a genome-wide mic oscopy-based sc een, o iden i y ac o s aec ing he o ma ion o his clus e . We c ossed a s ain ca ying TEF1p -Pex3 and he pe oxisomal ma ke Pex14-mKa e2 wi h a genome-wide collec ion o dele ion mu an s o non-essen ial genes47 and hypomo phic DAmP allele mu an s o essen ial genes48 using an au oma ed ma ing and spo ula ion p ocedu e49,50. e esul ing mu an collec ion con ains TEF1p -Pex3, Pex14-2xmKa e2 and each gene dele ed o deple ed (Fig.5A). We analyzed his collec ion by au oma ed mic oscopy ae labeling cells wi h Bodipy and CMAC, o s ain lipid d ople s and acuoles espec i ely. e esul ing images we e analyzed manually sea ching o s ains in which he pheno ype was dis up ed. We ound a single hi in which he pheno ype was signican ly dis up ed, which ca ied he dele ions o he gene encoding o he lipase Tgl451. is s ain was manually e-cons uc ed o con m ha he dis up ion o he pheno ype did no depend on he gene ic backg ound used o he sc een and e-analyzed by mic oscopy (Fig.5B). In cells ha o e exp essed Pex3, 95% o he pe oxisomal s uc u es we e p oximal o lipid d ople s, while his numbe d opped o 51% in cells ha in addi ion lacked Tgl4 (Fig.5C). Re-inse ion o he TGL4 ORF wi h i s endogenous p omo e in a plasmid ully eco e ed he in e ac ion be ween pe oxisomes and lipid d ople s, indica ing ha he eec is specic o he lack o he gene, and no a seconda y eec o he genomic modica ion (Supplemen al Fig.5A and B). We es ed he specici y o he pheno ype by dele ing o he lipid d ople -localized TAG lipases. Nei he dele ion o he genes encoding o he lipases homologous o Tgl4, namely Tgl3 and Tgl552,53 (Fig.5B and C) no o he ones, Tgl1, Ldh1 o Yeh154–56 (Supplemen al Fig.5C and D) caused a dis up ion o he pheno ype. e homologous lipases also did no cause a u he dis up ion when combined wi h he dele ion o TGL4 (Supplemen al Fig.5E and F). is sugges s ha i is he physical p esence o he p o ein Tgl4 ha aec s he o ma ion o he con ac si e and no i s ac i i y as a lipase. To es his hypo hesis, we assessed he eec o mu a ion S315G, which dis up s i s ac i e si e57, and obse ed no educ ion in he associa ion o lipid d ople s o he pe oxisomal clus e (Fig.5D and E). In addi ion, we analyzed he localiza ion o Tgl3, 4, and 5 on he lipid d ople su ace upon induc ion o hese con ac si es. Again, we used he s ain lacking PEX11 and g ew he cells in he p esence o olea e o inc ease he size o pe oxisomes and lipid d ople s and gain spa ial esolu ion. We obse ed ha all h ee lipases Fig. 3. e cy osolic domain o Pex3 binds pe oxisomes and is able o e he hem o ano he o ganelle. (A–C) e cy osolic domain o Pex3 binds o pe oxisomes. Panel A shows a diag am o he GFP-Pex3(CD) cons uc . is cons uc con ains he cy osolic domain o Pex3 (aa40-441) used o GFP ag a he N- e minus ins ead o he ansmemb ane domain as in he ull leng h Pex3. Panel B shows ep esen a i e pic u es o he colocaliza ion expe imen o GFP-Pex3(CD) wi h pe oxisomes (Pex14-HaloTag) and lipid d ople s (E g6- 2xmKa e2). Cell ou lines a e shown as whi e dashed lines. Scale ba : 2μm and 0.8μm. Panel C shows he co-localiza ion analysis o he expe imen in B using Mande ´s coecien s M1 and M2 o he o e lap o : Pex14-HaloTag and GFP-Pex3(CD) (cyan diamonds), E g6-2xmKa e2 and GFP-Pex3(CD) (g een diamonds) o Pex14-HaloTag and E g6-2xmKa e2 (black diamonds).  ee independen expe imen s we e pe o med and 30 cells we e analyzed o each expe imen . Each small diamond ep esen s a single cell, and he bigge ones ep esen he a e age o each o h ee independen expe imen s. (D) Diag am o he s a egy used o ec ui he cy osolic domain o Pex3 o mi ochond ia a icially. e ou e mi ochond ial memb ane p o ein Tom70 was agged in he C- e minus wi h a Nanobody ha ecognizes he Al aTag (Tom70-NB-Al a). e cy osolic domain o Pex3 was agged wi h an Al aTag and an mKa e2 uo escen p o ein in he N- e minus (Al aTag- mKa e2-Pex3(CD) cons uc ). is causes he ec ui men o Al aTag-mKa e2-Pex3(CD) o mi ochond ia. (E–F) Ta ge ing he cy osolic domain o Pex3 o mi ochond ia e he s pe oxisomes o his o ganelle. Panel E shows ep esen a i e images o he localiza ion o pe oxisomes (GFP-SKL) and mi ochond ia (Ci 1-HaloTag) in he p esence o absence o Al aTag-Pex3(CD), wi h Tom70 used o Al a Nanobody in he backg ound. A maximum in ensi y p ojec ion o he Z-s acks is shown o each image. Cell ou lines a e shown as whi e dashed lines. Scale ba : 2μm. Panel F shows he measu emen s o dis ances be ween pe oxisomes (GFP- SKL) and mi ochond ia (Ci 1-HaloTag) in he p esence o absence o he Al aTag-Pex3(CD) cons uc as desc ibed.  ee independen expe imen s we e pe o med and 30 cells we e analyzed o each expe imen . Each small diamond ep esen s a single pe oxisome, and he bigge ones ep esen he a e age o each o h ee independen expe imen s. e dis ibu ion o dis ances o mi ochond ia be ween he wo s ains was compa ed using a Kolmogo o -Smi no es , **** P < 0.0001. Compa ison o he means o each expe imen wi h a wo- ailed unpai ed S uden ´s - es esul s in a P alue < 0.05. ◂ Scien ic Repo s | (2025) 15:24480 7 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ we e en iched in he egion o he lipid d ople ha was in con ac wi h he pe oxisomes in some cells, as can be app ecia ed in he example images and line p oles (Fig.5F and G). ese en ichmen s, howe e , we e no equally equen o all lipases: Tgl4 and Tgl5 we e en iched mo e equen ly han Tgl3. is can be obse ed by he esul ing peaks o med by a e aging many cells in he line p oles (Fig.5G). To explo e he equi emen s o he o ma ion o Tgl4 oci a lipid d ople -pe oxisome in e aces, we pe o med a mic oscopy-based sc een (Supplemen al Fig.6A). A PEX3 o e exp ession allele, genes o lipid d ople and pe oxisome isualiza ion (E g6-mCh and BFP-SKL), and Tgl4-GFP we e in oduced in o he genome-wide dele ion and DAmP lib a ies47,48 by an au oma ed ma ing app oach49,50. Cells we e cul u ed in he p esence o olea e o expand lipid d ople s and analyzed by au oma ed mic oscopy. We iden ied a o al o 86 mu an s in which he accumula ion o Tgl4-GFP oci a lipid d ople -pe oxisome in e aces was ully o pa ially blocked (Supplemen al Table 5, example mic oscopy images in Supplemen al Fig.6B). We analyzed he common unc ions among he genes iden ied by he sc een (Supplemen al Fig.6C). e bigges g oup o genes was ela ed o ene gy me abolism. is is also e idenced by he en ichmen o he GO Te ms “mi ochond ion o ganiza ion”, “mi ochond ial espi a o y chain complex assembly” and “mi ochond ion” in ou hi lis wi h adjus ed p- alues o 0.0009, 0.001, and 0.008, espec i ely. O he g oups o hi s co esponded o hypoxia signaling, au ophagy, and lipid homeos asis. Taken oge he , his sugges s ha he p esence o Tgl4 a his o ganelle in e ace is egula ed by he me abolic s a e o he cell. Addi ionally, we iden ied eigh genes om he memb ane con ac si e da abase, ep esen ing a h ee old en ichmen o he expec ed amoun gi en he ac ion o he genome anno a ed, likely eec ing he igh in e ela ions wi hin he cellula con ac si e ne wo k. Scien ic Repo s | (2025) 15:24480 8 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ The pheno ype o o e exp ession o Pex3 is conse ed in me azoans Since Pex3 is a conse ed p o ein, and many o i s unc ions, like he inco po a ion o pe oxisomal memb ane p o eins and he ole in pexophagy a e conse ed in me azoans, we decided o es i he pheno ype o induc ion o con ac si es is also conse ed. us, we o e exp essed Pex3-HA in D osophila melanogas e la ae in he midgu using he GAL/UAS sys em58, and we obse ed by uo escence mic oscopy cells om he midgu , compa ing hem o cells exp essing only GFP-SKL, o obse e pe oxisomes when Pex3 is no o e exp essed (Fig.6A and B). In he la al midgu , we obse ed ha o e exp ession o Pex3 caused a educ ion o he numbe o pe oxisomes, as well as an inc ease in hei size, whe eas he numbe o lipid d ople s was unaec ed (Fig.6B– D). is pheno ype was exace ba ed du ing de elopmen , wi h a d as ic inc ease in pe oxisomal and lipid d ople size in he adul midgu (Fig.6E–H). We exp essed Pex3-HA wi h ano he pe oxisomal ma ke , YFP- SKL. Using Ai yscan con ocal mic oscopy, we ound ha Pex3-HA clea ly labeled he su ace o he enla ged pe oxisome, while YFP was impo ed in o he pe oxisomal lumen by he pe oxisomal a ge ing sequence, as expec ed (Fig.6I). Fu he mo e, hese enla ged pe oxisomes we e closely associa ed wi h lipid d ople s as can be obse ed by he shape de o ma ion o he o ganelles when hey a e nex o each o he . us, he pheno ype closely esembles he one obse ed in yeas . Discussion In his wo k, we ha e desc ibed ha Pex3 o e exp ession induces he o ma ion o pe oxisome-pe oxisome and pe oxisome-lipid d ople con ac si es. I we make he app oxima ion ha Pex3 exp essed unde he con ol o he TEF1 p omo e would ha e he le els o he Te 1 p o ein, and based on he in eg a ed high h oughpu da a in he p o ein abundance da abase PaxDB 5.059, he o e exp ession sys em ha we used would gene a e app oxima ely a 200- old inc ease in p o ein le els. We do no hink o his as mimicking a physiological scena io, bu a he as a ool o disco e y, which has in he pas p o en use ul o exace ba e one o he unc ions o a mul i unc ional p o ein o e he o he s. Fo example, o he p o eins Vps39 and C m1, o e exp ession allowed iden ica ion o hei ole as e he s o he acuole-mi ochond ia con ac si e60,61. is ole could la e be con med a endogenous le els o he p o eins ei he by add essing he dele ion o sepa a ion-o - unc ion mu an s61,62. Ou mic oscopy-based sc eens illus a e how use ul his ool can be, since i al eady allowed he Fig. 4. Fo ma ion o he s uc u e is independen o known in e ac o s o Pex3. (A) Diag am o Pex3 known in e ac o s and hei unc ions. (B, E) Rep esen a i e images o s ains o e exp essing Pex3 (TEF1p -PEX3) in con ol cells and s ains lacking Inp1 (B) o A g36 (E). All s ains exp ess Pex14 used o 2xmKa e2 o isualize he pe oxisomes, lipid d ople s we e s ained wi h Bodipy and he acuola lumen was s ained wi h CMAC. Cell ou lines a e shown as whi e dashed lines. Scale ba s: 2 μm. (C, F) Quan ica ion o he numbe o pe oxisomal s uc u es pe cell in he mic oscopy expe imen s desc ibed be o e. Small diamonds co espond o indi idual cells, bigge diamonds co espond o he a e age o independen expe imen s.  ee independen expe imen s we e pe o med and 30 cells we e analyzed o each expe imen and condi ion. e die en s ains we e compa ed using an unpai ed wo- ailed S uden ’s - es . n.s., no signican . (D, G) Quan ica ion o he ac ion o pe oxisomal s uc u es wi h accumula ions o lipid d ople s nex o hem in he mic oscopy expe imen s desc ibed abo e.  ee independen expe imen s we e pe o med and 30 cells we e analyzed o each expe imen and condi ion. e die en s ains we e compa ed using an unpai ed wo- ailed S uden ’s - es . n.s., no signican . (H) Pex3 o e exp ession does no induce pexophagy. Whole cell lysa es o s ains exp essing Pex14-GFP wi h o wi hou Pex3 o e exp ession we e analyzed by Wes e n blo . e cells we e ei he g own in media con aining glucose o loga i hmic phase (L.P.G) o g own in media con aining olea e and shied o ni ogen s a a ion medium o induce pexophagy (P.I.C). e p esence o a ee GFP band is indica i e o acuola deg ada ion o pe oxisomes. e whole Wes e n blo memb ane as well as a loading con ol is shown in Supplemen al Fig.3 A. (I) e amino acids in ol ed in hsPex3 in e ac ion wi h hsPex19 a e conse ed in yeas . To he le, he s uc u e ob ained o Homo sapiens Pex3 (cyan) in e ac ing wi h a pep ide o hsPex19 (da k blue)17. Amino acids W104 and L107 o hsPex3 a e in ol ed in he in e ac ion wi h hsPex19. To he igh , he s uc u e p edic ed by AlphaFold o Saccha omyces ce e isiae Pex3 (ligh blue) shows ha i con ains a s uc u ally conse ed W and L in he same posi ions (W128 and L131 in scPex3). (J) Pex3(W128K, L131K) canno in e ac wi h Pex19. Ani y pu ica ion o msGFP2-Pex19 co-pu ies Pex3-mKa e2-Al aTag bu no Pex3(W128K, L131K)-mKa e2-Al aTag. e comple e Wes e n blo memb anes a e shown in Supplemen al Fig.3 B and C. (K) Pex3(W128K, L131K) does no suppo BFP-SKL impo in o pe oxisomes. In a s ain ha exp esses BFP-SKL, endogenous Pex3 was dele ed and ei he Pex3w o Pex3(W128K, L128K) we e e-in oduced in a plasmid. Quan ica ion o he numbe o BFP-SKL punc a pe cell is shown o he igh . (L–N) O e exp ession o Pex3(W128K, L131K), which canno in e ac wi h Pex19, s ill causes agg ega ion o pe oxisomes and ec ui men o lipid d ople s. Panel K shows ep esen a i e images o s ains o e exp essing Pex3 om a plasmid (TEF1p -PEX3-mKa e2) con aining ei he Pex3 w o Pex3 mu an . Lipid d ople s we e s ained wi h Bodipy and he acuola lumen was s ained wi h CMAC. e s ain con ains Pex3 w in he backg ound, exp essed om i s genomic locus, in o de o ha e no mal pe oxisomes. Cell ou lines a e shown as whi e dashed lines. Scale ba s: 2μm. Panel L shows he quan ica ion o he numbe o pe oxisomal s uc u es pe cell in he mic oscopy expe imen s desc ibed be o e. Small diamonds co espond o indi idual cells, bigge diamonds co espond o he a e age o independen expe imen s. Panel M shows he quan ica ion o he p opo ion o pe oxisomal s uc u es wi h accumula ions o lipid d ople s nex o hem in he mic oscopy expe imen s desc ibed abo e.  ee independen expe imen s we e pe o med and 30 cells we e analyzed o each expe imen and condi ion. e die en s ains we e compa ed using an unpai ed wo- ailed S uden ’s - es . n.s., no signican . ◂ Scien ic Repo s | (2025) 15:24480 9 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ De ec ion o pexophagy ia whole cell lysa e and Wes e n blo All cells we e g own o e nigh in SDC media as p ecul u es. e loga i hmic phase glucose (L.P.G.) samples, we e dilu ed in SDC media and g own o 20hs wi hou exceeding an OD600 = 0.3 and ha es ed. e pexophagy- inducing condi ions samples (P.I.C) we e dilu ed in syn he ic media con aining olea e as he ca bon sou ce (0.2% Olea e, 0.1% Tween-80) and g own o 20hs, hen shied o ni ogen s a a ion medium (wi hou ni ogen sou ce o amino acids) o 22hs and ha es ed. Whole cell lysa es we e gene a ed by mechanical dis up ion using glass beads in lysis bue (3M u ea, 1.875mM EDTA-KOH p.H. 8.0, 1.2% (w/ ) SDS, 37.5mM T is–HCl p.H. 6.8, 1.5% ( / ) glyce ol. B omophenol blue was added o he samples o a nal concen a ion o 0.005% and 2-me cap oe hanol oanal concen a ion o 0.5%. Equi alen amoun s o each sample we e subjec ed o SDS- PAGE, ans e ed o a ni ocellulose memb ane and p ocessed o Wes e n blo as desc ibed in he p e ious sec ion. e seconda y an ibody was coupled o ho se- adish pe oxidase, and he signal was de ec ed using Sigma Lumi-Ligh Plus Wes e n Blo ing Subs a e in an Azu e 600 imaging Sys em. e whole unc opped wes e n blo memb anes is shown in Supplemen al Fig.3A, and addi ional epe i ions in Supplemen al Fig.3B and C. In he wo o he epe i ions, he p e-cul u es we e included in he expe imen (g own in glucose o he L.P.G condi ion and in olea e o he P.I.C. condi ion). Fly husband y Flies we e ea ed on s anda d co nmeal ood (130g ya n aga , 248g Bake ’s yeas , 1223g Co nmeal and 1.5l suga bee sy up in 20l dis illed wa e ) and kep in a 25°C incuba o wi h ligh –da k-cycle. Fly lines used in his s udy we e mex-Gal4 (kindly p o ided by he lab o I ene Miguel-Aliaga), UAS-Pex3-GFP, UAS-Pex3- HA (kindly p o ided by he lab o Reinha d Baue ) and UAS-GFP-SKL (Blooming on D osophila s ock cen e #28882). Fo s udies in he la al gu , he ollowing geno ypes we e used: w; mex-Gal4; UAS-GFP-SKL and w; mex-Gal4; UAS-GFP-SKL/UAS-Pex3-HA. Fo s udies in he adul gu , he ollowing geno ypes we e used: w; mex-Gal4; UAS-GFP-SKL, w; mex-Gal4; UAS-GFP-SKL/UAS-Pex3-HA and w; mex-Gal4; UAS-YFP-PTS1/ UAS-Pex3-HA. Animals we e ea ed on s anda d die and analyzed as 3 d ins a la ae o 5day old adul s (male and emale), espec i ely. Imaging o D osophila gu s An ibodies used in his s udy we e α-GFP (San a C uz Bio echnology) and α-HA (In i ogen), α-TOMM20 (Sigma-Ald ich). Fo immunohis ochemis y, gu s om 3 d ins a la ae o adul ies we e dissec ed in PBS and xed o 1h in 0.5% PBS-Tween20 and 4% o maldehyde. Tissue was washed wi h 0.5% PBS-Tween20 and blocked wi h donkey se um be o e incuba ion wi h he p ima y an ibody (o e nigh a 4°C). e issue was washed in 0.1% PBS-Tween20 be o e incuba ion wi h BODIPY 581/591 (e mo Scien ic) and he seconda y an ibody a oom empe a u e in he da k o 1h. Seconda y an ibodies coupled o Alexa o Cyanine dyes we e om Molecula p obes. e issue was washed and incuba ed o 5min wi h DAPI (4',6-diamidino-2- phenylindole). Fo imaging, we used a Zeiss LSM 710 wi h a 25 × wa e lens (Plan-Neoua , Zeiss), 40 × wa e lens (C-Apoch oma , Zeiss), and 63 × wa e lens (Plan-Apoch oma , Zeiss) and a Zeiss LSM 880 wi h Ai yscan de ec o . We used ImageJ o quan i y pe oxisome and lipid d ople numbe and a ea om a leas 3 indi idual cells om die en expe imen s. Lib a y gene a ion and high- h oughpu mic oscopy All yeas manipula ions we e pe o med in high-densi y o ma (384–1,536 s ains pe pla e) using a RoToR bench- op colony a ay ins umen (Singe Ins umen s). In o de o nd key p o eins in ol ed in he o ma ion o he pe oxisome-pe oxisome o pe oxisome-lipid d ople con ac si es (Fig.5A), s ain AGMY1303, wi h an o e exp ession o Pex3 (TEF1p -Pex3) and a pe oxisomal ma ke (Pex14-2xmKa e2) was c ossed wi h a genome-wide lib a y o dele ion47 and hypomo phic allele48 s ains, by he syn he ic gene ic a ay me hod49,50. Fo analysis o Tgl4 localiza ion (Supplemen al Fig.6), s ain yMB1326 (TEF2p -Pex3 E g6-mCh BFP-SKL Tgl4-GFP) was c ossed wi h he same mu an collec ions. Cells we e ma ed on ich medium pla es, diploids we e selec ed and spo ula ion was induced by incuba ing he cells o  e o eigh days on ni ogen s a a ion media pla es. Haploid cells we e selec ed on 50mg/L Cana anine and 50mg/L ialysine. Finally, haploid cells con aining he combina ion o all desi ed manipula ions we e selec ed. A subse o s ains was e ied by mic oscopy and con med by PCR. Fo he au oma ed imaging o he ob ained lib a ies, cells we e  s ans e ed om aga pla es in o 384 well pla es o g ow h in liquid medium. Fo he sc een o genes in ol ed in o ma ion o lipid d ople -pe oxisome con ac s (Fig.5A), cul u es we e g own o e nigh a 30°C in SDC. A JANUS liquid handle (Pe kinElme ) connec ed o he incuba o was used o dilu e he s ains o an OD600 o ∼0.2, and pla es we e incuba ed a 30°C o 4h. Cells we e washed and esh SDC media was added con aining 20µM 7-amino-4-chlo ome hylcouma in (CMAC) dye and 1µg/ml o Bodipy dye, ollowed by hal an hou incuba ion and ano he wash s ep. Fo he Tgl4-GFP sc een (Supplemen al Fig.6), liquid handling was pe o med using a Mic oP o 300 liquid handle . Following he same o e nigh cul u e as he p e ious sc een, cells we e g own in he p esence o 0.2% olea e in syn he ic medium o 24h o lipid d ople enla gemen . S ains we e hen ans e ed by a liquid handle in o glass-bo om 384-well mic oscope pla es (Ma ical Bioscience) coa ed wi h concana alin A (Sigma-Ald ich) and incuba ed o 20min o allow adhesion o cells o he bo om o he pla es. Ae wa ds, wells we e washed wice wi h SDC medium (SC medium o olea e ea ed cells) o emo e non-adhe en cells lea ing a cell monolaye . Pla es we e hen ans e ed o an Olympus au oma ed in e ed uo escence mic oscope sys em. In he sc een o lipid d ople -pe oxisome con ac si e p o eins (Fig.5A), cells we e imaged in SDC a 18–20°C using a 60 × ai lens (NA 0.9) and wi h an ORCA- ER cha ge-coupled de ice came a (Hamama su), using ScanR sowa e. In he Tgl4-GFP sc een (Supplemen al Fig.6), cells we e imaged in SC medium a 18–20°C using a 60 × ai lens (NA 0.9) and wi h an ORCA-ash4.0 Scien ic Repo s | (2025) 15:24480 16 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ came a (Hamama su), using ScanR sowa e. Ae acquisi ion, images we e manually e iewed using ImageJ (Na ional Ins i u es o Heal h). Da a a ailabili y e p o eomics da ase gene a ed in his s udy is a ailable ia P o eomeXchange wi h iden ie PXD063514. O he da ase s gene a ed du ing he cu en s udy a e a ailable om he co esponding au ho on easonable eques . Recei ed: 13 Decembe 2024; Accep ed: 18 June 2025 Re e ences 1. Wande s, R. J. A., Baes, M., Ribei o, D., Fe dinandusse, S. & Wa e ham, H. R. e physiological unc ions o human pe oxisomes. Physiol. Re . 103, 957–1024 (2023). 2. Sco ano, L. e al. Coming oge he o dene memb ane con ac si es. Na . Commun. 10, 1–11 (2019). 3. Eisenbe g-Bo d, M., Shai, N., Schuldine , M. & Bohne , M. A e he is a e he is a e he : Te he ing a memb ane con ac si es. De . Cell 39, 395–409 (2016). 4. P inz, W. A., Toulmay, A. & Balla, T. e unc ional uni e se o memb ane con ac si es. Na . Re . Mol. Cell Biol. 21, 7–24 (2020). 5. Valm, A. M. e al. Applying sys ems-le el spec al imaging and analysis o e eal he o ganelle in e ac ome. Na u e 546, 162–167 (2017). 6. Shai, N. e al. Sys ema ic mapping o con ac si es e eals e he s and a unc ion o he pe oxisome-mi ochond ia con ac . Na . Commun. 9 (2018). 7. Kakimo o, Y. e al. Visualizing mul iple in e -o ganelle con ac si es using he o ganelle-Ta ge ed spli -GFP sys em. Sci. Rep. 8 (2018). 8. Voel z, G. K., Sawye , E. M., Hajnóczky, G. & P inz, W. A. Making he connec ion: How memb ane con ac si es ha e changed ou iew o o ganelle biology. Cell 187, 257–270 (2024). 9. Shai, N., Schuldine , M. & Zalck a , E. No pe oxisome is an island—Pe oxisome con ac si es. Biochim. Biophys. Ac a Mol. Cell Res. 1863, 1061–1069 (2016). 10. Da id, C. e al. A combined app oach o quan i a i e in e ac ion p o eomics and li e-cell imaging e eals a egula o y ole o endoplasmic e iculum (ER) Re iculon Homology P o eins in pe oxisome biogenesis. Mol. Cell. P o eomics 12, 2408–2425 (2013). 11. Yan, M., Rachubinski, D. A., Joshi, S., Rachubinski, R. A. & Sub amani, S. Dys e lin domain-con aining p o eins, Pex30p and Pex31p, localized o wo compa men s, con ol he numbe and size o olea e-induced pe oxisomes in pichia pas o is. Mol. Biol. Cell 19, 885–898 (2008). 12. Knoblach, B. e al. An ER-pe oxisome e he exe s pe oxisome popula ion con ol in yeas . EMBO J. 32, 2439–2453 (2013). 13. Bu ne , S. F., Fa é, J. C., Naza ko, T. Y. & Sub amani, S. Pe oxisomal Pex3 ac i a es selec i e au ophagy o pe oxisomes ia in e ac ion wi h he pexophagy ecep o A g30. J. Biol. Chem. 290, 8623–8631 (2015). 14. Fang, Y., Mo ell, J. C., Jones, J. M. & Gould, S. J. PEX3 unc ions as a PEX19 docking ac o in he impo o class I pe oxisomal memb ane p o eins. J. Cell Biol. 164, 863–875 (2004). 15. Mo ley, A. M., Nu all, J. M. & He ema, E. H. Pex3-ancho ed A g36 ags pe oxisomes o deg ada ion in Saccha omyces ce e isiae. EMBO J. 31, 2852–2868 (2012). 16. Sa o, Y. e al. Cha ac e iza ion o he in e ac ion be ween ecombinan human pe oxin Pex3p and Pex19p: Iden ica ion o TRP- 104 in Pex3p as a c i ical esidue o he in e ac ion. J. Biol. Chem. 283, 6136–6144 (2008). 17. Sa o, Y. e al. S uc u al basis o docking o pe oxisomal memb ane p o ein ca ie Pex19p on o i s ecep o Pex3p. EMBO J. 29, 4083–4093 (2010). 18. Schmid , F. e al. e ole o conse ed PEX3 egions in PEX19-binding and pe oxisome biogenesis. T affic 13, 1244–1260 (2012). 19. Yamashi a, S. I., Abe, K., Ta emichi, Y. & Fujiki, Y. e memb ane pe oxin PEX3 induces pe oxisome-ubiqui ina ion-linked pexophagy. Au ophagy 10, 1549–1564 (2014). 20. Hulmes, G. E. e al. e Pex3-Inp1 complex e he s yeas pe oxisomes o he plasma memb ane. J. Cell Biol. 219 (2020). 21. Wu, H. e al. Pe oxisome de elopmen in yeas is associa ed wi h he o ma ion o Pex3-dependen pe oxisome- acuole con ac si es. Biochim. Biophys. Ac a Mol. Cell Res. 1866, 349–359 (2019). 22. Fe nández-Suá ez, M., Sco Chen, T. & Ting, A. Y. P o ein-p o ein in e ac ion de ec ion in i o and in cells by p oximi y bio inyla ion. NIH Public Access 23, 1–7 (2008). 23. B anon, T. C. e al. Ecien p oximi y labeling in li ing cells and o ganisms wi h Tu boID. Na . Bio echnol. 176, 139–148 (2018). 24. E dmann, R. & Blobel, G. Gian pe oxisomes in oleic acid-induced Saccha omyces ce e isiae lacking he pe oxisomal memb ane p o ein Pmp27p. J. Cell Biol. 128, 509–523 (1995). 25. Yi ach, E. e al. Sys ema ic mul i-le el analysis o an o ganelle p o eome e eals new pe oxisomal unc ions. Mol. Sys . Biol. 18, 1–21 (2022). 26. Sandage , L. e al. S o age lipid syn hesis is non-essen ial in yeas . J. Biol. Chem. 277, 6478–6482 (2002). 27. Höh eld, J., Veenhuis, M. & Kunau, W. H. PAS3, a Saccha omyces ce e isiae gene encoding a pe oxisomal in eg al memb ane p o ein essen ial o pe oxisome biogenesis. J. Cell Biol. 114, 1167–1178 (1991). 28. Soukupo aa, M., Sp enge , C., Go gas, K., Kunau, W.-H. & Dod l, G. Iden ica ion and Cha ac e iza ion o he Human Pe oxin PEX3. Eu . J. Cell Biol. 78. h p://www.u bansche .de/jou nals/ejcb (1999). 29. Schmid , F. e al. Insigh s in o pe oxisome unc ion om he s uc u e o PEX3 in complex wi h a soluble agmen o PEX19. J. Biol. Chem. 285, 25410–25417 (2010). 30. Gö zke, H. e al. e ALFA- ag is a highly e sa ile ool o nanobody-based bioscience applica ions. Na . Commun. 10, 1–12 (2019). 31. To ggle , R., Papinski, D. & K a, C. Assays o moni o au ophagy in Saccha omyces ce e isiae. Cells h p s : / / d o i . o g / 1 0 .3 3 9 0 / c e l l s 6 0 3 0 0 2 3 (2017). 32. Koli sida, P. e al. e py u a e dehyd ogenase complex egula es mi ophagic acking and p o ein phospho yla ion. Li e Sci. Alliance 9 (2023). 33. Yu, J. e al. Au ophagy- ela ed p o ein U A g7 con ibu es o mycelial g ow h, i ulence, asexual ep oduc ion and cell s ess esponse in ice alse smu ungus Us ilaginoidea i ens. Fungal Gene ics Biol. 159 (2022). 34. He ema, E. H., Gi zalsky, W., Van Den Be g, M., E dmann, R. & Dis el, B. Saccha omyces ce e isiae Pex3p and Pex19p a e equi ed o p ope localiza ion and s abili y o pe oxisomal memb ane p o eins. EMBO J. 19, 223–233 (2000). 35. Le ine, T. P. & Mun o, S. Dual a ge ing o Osh1p, a yeas homologue o oxys e ol-binding p o ein, o bo h he Golgi and he nucleus- acuole junc ion. Mol. Biol. Cell 12, 1633–1644 (2001). 36. Loewen, C. J. R., Roy, A. & Le ine, T. P. A conse ed ER a ge ing mo i in h ee amilies o lipid binding p o eins and in Opi1p binds VAP. EMBO J. 22, 2025–2035 (2003). Scien ic Repo s | (2025) 15:24480 17 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ 37. K am, E. & Gold a b, D. S. N j1p is he ou e -nuclea -memb ane ecep o o oxys e ol-binding p o ein homolog Osh1p in Saccha omyces ce e isiae. J. Cell Sci. 117, 4959–4968 (2004). 38. Elbaz-Alon, Y. e al. Lam6 egula es he ex en o con ac s be ween o ganelles. Cell Rep. 12, 7–14 (2015). 39. Mu ley, A. e al. L c1 is an ER-localized s e ol anspo e and a componen o ER-mi ochond ia and ER- acuole con ac s. J. Cell Biol. 209, 539–548 (2015). 40. Toulmay, A. & P inz, W. A. A conse ed memb ane-binding domain a ge s p o eins o o ganelle con ac si es. J. Cell Sci. 125, 49–58 (2012). 41. Liu, L. K., Choudha y, V., Toulmay, A. & P inz, W. A. An inducible ER-Golgi e he acili a es ce amide anspo o alle ia e lipo oxici y. J. Cell Biol. 216, 131–147 (2017). 42. Diep, D. T. V. e al. A me abolically con olled con ac si e be ween acuoles and lipid d ople s in yeas . De . Cell 59, 740-758.e10 (2024). 43. Ál a ez-Gue a, I. e al. LDO p o eins and Vac8 o m a acuole-lipid d ople con ac si e o enable s a a ion-induced lipophagy in yeas . De . Cell 59, 759-775.e5 (2024). 44. Chang, C. L. e al. Spas in e he s lipid d ople s o pe oxisomes and di ec s a y acid acking h ough ESCRT-III. J. Cell Biol. 218, 2583–2599 (2019). 45. Joshi, A. S. e al. A amily o memb ane-shaping p o eins a ER subdomains egula es p e-pe oxisomal esicle biogenesis. J. Cell Biol. 215, 515–529 (2016). 46. Joshi, A. S. e al. Lipid d ople and pe oxisome biogenesis occu a he same ER subdomains. Na . Commun. 9, 1–12 (2018). 47. Giae e , G. e al. Func ional P ofiling o he Saccha omyces Ce e isiae Genome. h p://www.kegg.com (2002). 48. B eslow, D. K. e al. A Comp ehensi e S a egy Enabling High-Resolu ion Func ional Analysis o he Yeas Genome. 49. Cohen, Y. & Schuldine , M. Ad anced me hods o high- h oughpu mic oscopy sc eening o gene ically modied yeas lib a ies. In Me hods in Molecula Biology Vol. 781 127–159 (Humana P ess Inc., 2011). 50. Hin, A., Tong, Y. & Boone, C. Syn he ic gene ic a ay analysis in Saccha omyces Ce e isiae. F om: Me hods in Molecula Biology ol. 313 (2006). 51. Rajakuma i, S., Rajasekha an, R. & Daum, G. T iacylglyce ol lipolysis is linked o sphingolipid and phospholipid me abolism o he yeas Saccha omyces ce e isiae. Biochim. Biophys. Ac a Mol. Cell Biol. Lipids 1801, 1314–1322 (2010). 52. A hens aed , K. & Daum, G. Tgl4p and Tgl5p, wo iacylglyce ol lipases o he yeas Saccha omyces ce e isiae a e localized o lipid pa icles. J. Biol. Chem. 280, 37301–37309 (2005). 53. A hens aed , K. & Daum, G. YMR313c/TGL3 encodes a no el iacylglyce ol lipase loca ed in lipid pa icles o Saccha omyces ce e isiae. J. Biol. Chem. 278, 23317–23323 (2003). 54. Jand osi z, A. e al. e lipid d ople enzyme Tgl1p hyd olyzes bo h s e yl es e s and iglyce ides in he yeas , Saccha omyces ce e isiae. Biochim. Biophys. Ac a Mol. Cell Biol. Lipids 1735, 50–58 (2005). 55. oms, S., Debelyy, M. O., Conne h, M., Daum, G. & E dmann, R. e pu a i e Saccha omyces ce e isiae hyd olase Ldh1p is localized o lipid d ople s. Euka yo . Cell 10, 770–775 (2011). 56. Köel, R., Tiwa i, R., Falque , L. & Schnei e , R. e Saccha omyces ce e isiae YLL012/YEH1, YLR020/YEH2, and TGL1 Genes Encode a No el Family o Memb ane-Ancho ed Lipases a A e Requi ed o S e yl Es e Hyd olysis. Mol. Cell Biol. 25, 1655– 1668 (2005). 57. Ku a , C. F. e al. Obese yeas : T iglyce ide lipolysis is unc ionally conse ed om mammals o yeas . J. Biol. Chem. 281, 491–500 (2006). 58. B and, H. & Pe imon, N. Ta ge ed Gene Exp ession as a Means o Al e ing Cell Fa es and Gene a ing Pheno ypes (1993). 59. Huang, Q., Szkla czyk, D., Wang, M., Simono ic, M. & Me ing, C. on. PaxDb 5.0: Cu a ed p o ein quan ica ion da a sugges s adap i e p o eome changes in yeas s. Mol. Cell. P o eom. 22 (2023). 60. Hönsche , C. e al. Cellula me abolism egula es con ac si es be ween acuoles and mi ochond ia. De . Cell 30, 86–94 (2014). 61. Bisinski, D. D. e al. C m1 is a componen o mul iple acuola con ac si es equi ed o sphingolipid homeos asis. J. Cell Biol. 221 (2022). 62. González Mon o o, A. e al. Vps39 in e ac s wi h Tom40 o es ablish one o wo unc ionally dis inc acuole-mi ochond ia con ac si es. De . Cell 45, 621-636.e7 (2018). 63. Gu e ich, V. V. & Gu e ich, E. V. Analyzing he oles o mul i- unc ional p o eins in cells: e case o a es ins and GRKs. C i . Re . Biochem. Mol. Biol. 50, 440–452. h ps://doi.o g/10.3109/10409238.2015.1067185 (2015). 64. Lubin, J. W., Rao, T., Mandel, E. K., Wu ke, D. S. & Lundblad, V. Dissec ing p o ein unc ion: An ecien p o ocol o iden i ying sepa a ion-o - unc ion mu a ions ha encode s uc u ally s able p o eins. Gene ics 193, 715–725 (2013). 65. Sch ade , M., King, S. J., S oh, T. A. & Sch oe , T. A. Real ime imaging e eals a pe oxisomal e iculum in li ing cells. 3671, 3663–3671 (2000). 66. Bonekamp, N. A., Sampaio, P., de Ab eu, F. V., Lüe s, G. H. & Sch ade , M. T ansien complex in e ac ions o mammalian pe oxisomes wi hou exchange o ma ix o memb ane ma ke p o eins. T affic 13, 960–978 (2012). 67. Go gas, K. & Zaa , K. Ana omy and emb yology pe oxisomes in sebaceous glands III. Mo phological simila i ies o pe oxisomes wi h smoo h endoplasmic e iculum and Golgi s acks in he ci cumanal gland o he dog* **. Ana . Emb yol. 169 (1984). 68. Mas , F. D., Rachubinski, R. A. & Ai chison, J. D. Pe oxisome p ognos ica ions: Explo ing he bi h, li e, and dea h o an o ganelle. J. Cell Biol. 219, 1–13 (2020). 69. Pan, X. e al. Nucleus- acuole junc ions in Saccha omyces ce e isiae a e o med h ough he di ec in e ac ion o Vac8p wi h N j1p. Mol. Biol. Cell 11, 2445–2457 (2000). 70. Filadi, R. e al. TOM70 sus ains cell bioene ge ics by p omo ing IP3R3-media ed ER o mi ochond ia Ca2+ ans e . Cu . Biol. 28, 369-382.e6 (2018). 71. Hollens ein, D. M. e al. Vac8 spa ially connes au ophagosome o ma ion a he acuole in S. Ce e isiae. J. Cell Sci. 132 (2019). 72. Vei , M., Laage, R., Die ich, L., Wang, L. & Unge mann, C. Vac8p elease om he SNARE complex and i s palmi oyla ion a e coupled and essen ial o acuole usion. EMBO J. 20, 3145–3155 (2001). 73. Wang, Y. X., Ca le , N. L. & Weisman, L. S. Vac8p, a acuola p o ein wi h a madillo epea s, unc ions in bo h acuole inhe i ance and p o ein a ge ing om he cy oplasm o acuole. J. Cell Biol. 140, 1063–1074 (1998). 74. Schmid , O., P anne , N. & Meisinge , C. Mi ochond ial p o ein impo : F om p o eomics o unc ional mechanisms. Na . Re . Mol. Cell Biol. 11, 655–667. h ps://doi.o g/10.1038/n m2959 (2010). 75. González Mon o o, A. e al. Subuni exchange among endolysosomal e he ing complexes is linked o con ac si e o ma ion a he acuole. Mol. Biol. Cell 32, b 14 (2021). 76. Binns, D. e al. An in ima e collabo a ion be ween pe oxisomes and lipid bodies. J. Cell Biol. 173, 719–731 (2006). 77. Janke, C. e al. A e sa ile oolbox o PCR-based agging o yeas genes: New uo escen p o eins, mo e ma ke s and p omo e subs i u ion casse es. Yeas 21, 947–962 (2004). 78. Gene oso, W. C., Go a di, M., O eb, M. & Boles, E. Simplied CRISPR-Cas genome edi ing o Saccha omyces ce e isiae. J. Mic obiol. Me hods 127, 203–205 (2016). 79. Schä e , A., Ke ssen, D., Veenhuis, M., Kunau, W.-H. & Schliebs, W. Func ional Simila i y be ween he Pe oxisomal PTS2 Recep o Binding P o ein Pex18p and he N-Te minal Hal o he PTS1 Recep o Pex5p. Mol. Cell Biol. 24, 8895–8906 (2004). 80. G imm, J. B. e al. A gene al me hod o imp o e uo opho es using deu e a ed auxoch omes. JACS Au 1, 690–696 (2021). 81. Day, K. J., Casle , J. C. & Glick, B. S. Budding yeas has a minimal endomemb ane sys em. De . Cell 44, 56-72.e4 (2018). 82. Bol e, S. & Co delie es, F. P. A guided ou in o subcellula colocaliza ion analysis in ligh mic oscopy. J. Mic osc. 224 (2006). Scien ic Repo s | (2025) 15:24480 18 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/ 83. Ollion, J., Cochennec, J., Loll, F., Escudé, C. & Boudie , T. TANGO: A gene ic ool o high- h oughpu 3D image analysis o s udying nuclea o ganiza ion. Bioin o ma ics 29, 1840–1841 (2013). 84. Olsen, J. V. e al. Highe -ene gy C- ap dissocia ion o pep ide modica ion analysis. Na . Me hods 4, 709–712 (2007). 85. Cox, J. & Mann, M. MaxQuan enables high pep ide iden ica ion a es, indi idualized p.p.b.- ange mass accu acies and p o eome- wide p o ein quan ica ion. Na . Bio echnol. 26, 1367–1372 (2008). 86. Cox, J. e al. And omeda: A pep ide sea ch engine in eg a ed in o he MaxQuan en i onmen . J. P o eome Res. 10, 1794–1805 (2011). 87. Tyano a, S., Temu, T. & Cox, J. e MaxQuan compu a ional pla o m o mass spec ome y-based sho gun p o eomics. Na . P o oc. 11, 2301–2319 (2016). 88. Ade , N. R. & Kukulski, W. iCLEM: Combining high-p ecision, oom empe a u e CLEM wi h c yo-uo escence mic oscopy o iden i y e y a e e en s. Me hods Cell Biol. 140, 303–320 (2017). 89. K eme , J. R., Mas ona de, D. N. & Mcin osh, J. R. Compu e Visualiza ion o  ee-Dimensional Image Da a Using IMOD. (1996). Acknowledgemen s We hank Ch is ian Wingen and Fa mi e Bujupi o gene a ing he UAS-Pex3-GFP and UAS-Pex3-HA y lines. We hank Hada Meye and Yeyni As a o hei assis ance wi h he high- h oughpu sc eens. Au ho con ibu ions L.A. pe o med mos yeas expe imen s, analyzed he da a, and p epa ed gu es. L.P. Pe o med and analyzed he Tgl4 localiza ion mic oscopy based sc een, and some addi ional yeas expe ime ns. R.F. Pe o med and p ocessed he on-sec ion CLEM TEM Tomog aphy. V. F. helped wi h yeas expe imen s. E.J.B. helped wi h y expe imen s. O.E.P. supe ised elec on mic oscopy expe imen s and da a p ocessing. M.S. and M.Bohne . supe ised mic oscopy-based genome-wide sc een expe imen s. M.Bülow pe o med and supe ised y expe - imen s. A.G.M supe ised he whole p ojec , p epa ed gu es and w o e he ini ial d a o he manusc ip . All au ho s e ised and edi ed he manusc ip . Funding Open Access unding enabled and o ganized by P ojek DEAL. is p ojec was unded h ough a Deu sche Fo schungsgemeinscha (DFG) indi idual esea ch g an o Ayelén González Mon o o (GO3313/1-1)wi h ad- di ional suppo om SFB944 and SFB1557. Wo k pe o med in he Bohne lab was suppo ed by he DFG, p ojec s SFB1557 P3 (p ojec ID 467522186), SFB1348 A13 (p ojec ID 386797833) and FOR5815 P6 (p ojec ID 538651361). Wo k in he Bülow lab was suppo ed by DFG g an s 417982926 and 535112684. Wo k on pe oxi- somes in he Schuldine lab is suppo ed by an Is ael Science Founda ion g an ISF 914/22. e obo ic sys em o he Schuldine lab was pu chased h ough he kind suppo o he Bly he B enden-Mann Founda ion. MS is an Incumben o he D . Gilbe Omenn and Ma ha Da ling P o esso ial Chai in Molecula Gene ics Decla a ions Compe ing in e es s e 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 e online e sion con ains supplemen a y ma e ial a ailable a h p s : / / d o i . o g / 1 0 . 1 0 3 8 / s 4 1 5 9 8 - 0 2 5 - 0 7 9 3 4 - 2 . Co espondence and eques s o ma e ials should be add essed o A.G.M. Rep in s and pe missions in o ma ion is a ailable a www.na u e.com/ ep in s. Publishe ’s no e Sp 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lia ions. Open Access is 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 licence, and indica e i changes we e made. e 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 licence, 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 licence 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 licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/. © e Au ho (s) 2025 Scien ic Repo s | (2025) 15:24480 19 | h ps://doi.o g/10.1038/s41598-025-07934-2 www.na u e.com/scien i ic epo s/