Drosophila muscles regulate the immune response against wasp infection via carbohydrate metabolism

1 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 www.na u e.com/scien i ic epo s D osophila muscles egula e he immune esponse agains wasp in ec ion ia ca bohyd a e me abolism Hai u Yang1,3 & Dan Hul ma k 1,2 We ecen ly ound ha JAK/STAT signaling in skele al muscles is impo an o he immune esponse o D osophila la ae agains wasp in ec ion, bu i was no clea how muscles could a ec he immune esponse. He e we show ha insulin signaling is equi ed in muscles, bu no in a body o hemocy es, du ing la al de elopmen o an e icien encapsula ion esponse and o he o ma ion o lamellocy es. This e ec equi es TOR signaling. We show ha muscle issue a ec s he immune esponse by ac ing as a mas e egula o o ca bohyd a e me abolism in he in ec ed animal, ia JAK/STAT and insulin signaling in he muscles, and ha he e is indi ec posi i e eedback be ween JAK/STAT and insulin signaling in he muscles. Speci ically, s imula ion o JAK/STAT signaling in he muscles can escue he de icien immune esponse when insulin signaling is supp essed. Ou esul s shed new ligh on he in e ac ion be ween me abolism, immuni y, and issue communica ion. The immune esponse in D osophila is speci ically adap ed o di e en kinds o in ec ing o ganisms. Bac e ia and ungi induce a humo al immune esponse, in which an imic obial pep ides a e induced ia Toll o Imd signal- ing1–4. La ge pa hogens, such as eggs laid by he pa asi oid wasp Lep opilina boula di, ac i a e a cellula immune esponse5–7 ha in ol es h ee hemocy e classes: plasma ocy es, c ys al cells, and lamellocy es. A ound 6–10 h a e in ec ion, plasma ocy es begin o ecognize and bind o he wasp egg8,9. Simul aneously, a new class o cells appea s in ci cula ion, he lamelloblas s, which la e di e en ia e o become ci cula ing lamellocy es. The lamel- locy es pa icipa e in o ming a capsule a ound he pa asi e, oge he wi h he plasma ocy es, which meanwhile also unde go ans o ma ion in o a lamellocy e-like s a e (lamellocy es, ype II)9. Finally, phenoloxidases om c ys al cells and lamellocy es p oduce melanin, which is deposi ed in he capsule10,11. We ha e ecen ly shown ha he encapsula ion esponse is u he modula ed by in e ac ions be ween hemocy es and o he la al issues, such as a body and muscles12–14. The cy okines Unpai ed 2 and 3 (Upd2 and Upd3), a e up egula ed in he hemocy es o wasp-in ec ed la ae, and wi hin eigh hou s a e in ec ion he skele al muscles espond by ac i a ion o he JAK/STAT (Janus kinase/signal ansduce s and ac i a o s o ansc ip ion) signaling pa hway. We ound ha a unc ional JAK/STAT signaling in he muscles was a p e equi- si e o an e icien immune esponse12. In D osophila, JAK/STAT signaling in ol es h ee cy okines (Upd1–3), one cy okine ecep o (Domeless), one y osine kinase (Hopsco ch), and one ansc ip ion ac o (S a 92E)15–17. Loss-o - unc ion mu a ions in he co esponding genes educe he encapsula ion esponse agains wasp in ec- ion12,18, while a i icial ac i a ion o his signaling igge s a cellula immune esponse, including lamellocy e o ma ion and gene a ion o melanized nodules, in a way ha is eminiscen o he encapsula ion esponse19–24. We ha e now u he s udied wha happens in he muscles o wasp-in ec ed la ae, and ound ha he in ec- ion has s ong e ec s on insulin signaling and glycogen s o age in his issue. The insulin signaling pa hway is conse ed be ween in e eb a es and e eb a es25. In D osophila, he sec e ion o eigh iden i ied insulin-like pep ides (Ilps) depends on he de elopmen al s age, ype o issue, and en i onmen al ac o s26,27. Sec e ed Ilps bind o he single insulin-like ecep o (InR) in he a ge issues, ac i a ing he downs eam componen s sequen- ially, including a phophoinosi ide 3 kinase (Pi3K92E), an AKT homolog (Ak 1), Ta ge o apamycin (To ) and 1Depa men o Molecula Biology, Umeå Uni e si y, S-901 87, Umeå, Sweden. 2Ins i u e o Biomedical Technology, Uni e si y o Tampe e, FI-33520, Tampe e, Finland. 3P esen add ess: Immunology P og am, Memo ial Sloan Ke e ing Cance Cen e (MSKCC), New Yo k, NY, 10065, USA. Co espondence and eques s o ma e ials should be add essed o D.H. (email: [email p o ec ed]) Recei ed: 9 Ma ch 2017 Accep ed: 6 No embe 2017 Published: xx xx xxxx OPEN www.na u e.com/scien i ic epo s/ 2 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 o khead box, sub-g oup O (Foxo). Finally ac i a ed insulin signaling exe s i s e ec on di e en biological p o- cesses, o ins ance g ow h, de elopmen , me abolism, beha io , li e span and immuni y28–30. Besides he classical insulin pa hway, D osophila has wo homologs o he elaxin ecep o , Lg 3 and Lg 4, o which a leas Lg 3 can se e as a ecep o o he insulin-like pep ide Ilp831–33. Ou esul s sugges ha JAK/STAT signaling in he muscles has p o ound e ec s on insulin signaling in he en i e o ganism, ha insulin in u n gi es posi i e eedback on he JAK/STAT esponse, and ha D osophila la - al skele al muscles ha e a su p ising ole in he cellula immune esponse agains wasp in ec ion by con olling ca bohyd a e me abolism and eeding beha io . Resul s Insulin signaling in muscles is equi ed o cellula immune esponse agains wasp in ec ion. To in es iga e which ac i i ies, besides JAK/STAT signaling, a e equi ed in he muscles o he immune esponses agains wasp in ec ion, we indi idually supp essed se e al well-known immune- ela ed signaling pa hways, including he p38, c-Jun N- e minal kinase (JNK), Toll, and insulin signaling pa hways. We supp essed hese pa hways by exp essing ei he RNA in e e ence o dominan -nega i e cons uc s wi h a muscle-speci ic d i e 34, Me 2-GAL4, and hen assayed encapsula ion a es and hemocy e numbe s. We ound ha supp ession o insulin signaling in he muscles, by exp essing InRRNAi o InRDN, signi ican ly educed he encapsula ion a e (Fig.1A), while supp ession o he o he es ed signaling pa hways nei he a ec ed he encapsula ion a e no he numbe o lamellocy es and plasma ocy es in ci cula ion (see Supplemen a y Fig.S1). Supp ession o insulin signaling wi h InRRNAi also educed he numbe o ci cula ing lamellocy es (Fig.1B), bu no he numbe o ci cula ing plasma ocy es (Fig.1C). Taken oge he , hese esul s indica e ha insulin signaling is equi ed in la al skel- e al muscles o an e icien cellula immune esponse. To exclude he possibili y ha he dec eased numbe o lamellocy es was due o a achmen o hese cells o o he issues, we used he msn-Che y (MSNF9mo-mChe y) s ain35 o moni o he lamellocy es in i o. The msn-Che y luo escen epo e speci ically labels lamellocy es, ce ain muscles and a ew o he issues. Using his epo e , we con i med ha he numbe o lamellocy es was dec eased in InR-supp essed la ae. The emaining msn-Che y exp ession in hese animals is essen ially limi ed o backg ound ec opic exp ession o his epo e in he eeding appa a us and ce ain acheal b anches (com- pa e Fig.1D and D’). Immunoblo ing con i med ha exp ession o InRRNAi o InRDN in muscles signi ican ly educed phospho yla ed AKT in muscles, indica ing ha hese wo cons uc s can success ully block insulin signaling in muscles (Fig.1E). To and Foxo a e wo majo downs eam a ge s o AKT signaling, To being a posi i e and Foxo a nega i e media o o insulin signaling25,36. To in es iga e which o hese wo b anches o he insulin signaling pa hway media e he e ec s o AKT in he muscles, we es ed he e ec on encapsula ion a e when we supp essed To o o e exp essed oxo in muscles. Ou esul s showed ha supp essing To in skele al muscles, by exp essing To DN, signi ican ly educed he encapsula ion a e and lamellocy e numbe s (Fig.1A,B and D”), bu we saw no e ec on he numbe o plasma ocy es in ci cula ion (Fig.1C). By con as , we did no obse e any ob ious e ec on he encapsula ion a e when we o e exp essed oxo (Fig.1A). Taken oge he , hese esul s sugges ha , besides JAK/STAT signaling, insulin signaling is also necessa y in muscles o he encapsula ion esponse, and ha his is media ed by signaling ia To . These esul s u he unde sco e he impo ance o D osophila skele al muscles in cellula immune esponse du ing wasp in ec ion. Conside ing ha insulin signaling in he a body plays an impo an ole o he con ol o gene al me ab- olism, and ha he a body is also an immune esponsi e issue2,37,38, we nex es ed he ole o insulin sign- aling in he a body on he encapsula ion esponse. Howe e , silencing o o e exp essing InR in he a body wi h a a body-speci ic d i e , FB-GAL439, did no signi ican ly a ec he encapsula ion a e (Fig.1F), indica ing ha insulin signaling in he a body is no essen ial o he immune esponse agains wasp in ec ion. Simila ly, since hemocy es a e main playe s in he encapsula ion esponse, we also in es iga ed whe he insulin signal- ing in hemocy es a ec his esponse. Fo his pu pose, we exp essed InRRNAi in hemocy es wi h a combina ion (“HH-GAL4”)13 o wo hemocy e d i e s, He-GAL4 and HmlΔ-GAL4, bu again we could no obse e any e ec on he encapsula ion a e (Fig.1G). Howe e , o e exp ession o InR in hemocy es sligh ly imp o ed he encapsu- la ion a e, om 71% o 81% (Fig.1G), bu al oge he hese esul s sugges ha insulin signaling is nei he essen- ial in a body no in hemocy es o he encapsula ion esponse. We also es ed he e ec o supp essing p38, JNK and Toll signaling in a body, bu nei he o hese ea men s signi ican ly a ec ed he encapsula ion a e, o he numbe o hemocy es (see Supplemen a y Fig.S2). Wasp in ec ion a ec s bo h insulin and JAK/STAT signaling. Since we ound ha an in ac insulin signaling pa hway was equi ed in he skele al muscles o a success ul encapsula ion esponse, we nex in es i- ga ed how wasp in ec ion a ec s insulin signaling ac i i y in muscles and elsewhe e. Using immunoblo ing, we ound ha phospho yla ed AKT was signi ican ly educed in muscles 50 hou s a e wasp in ec ion, bu we saw no signi ican di e ence a he 27-hou ime poin (Fig.2A and B), indica ing ha a longe pe iod o wasp in ec ion leads o educed insulin signaling ac i i y in skele al muscles. Fu he mo e, we es ed whe he wasp in ec ion also a ec s insulin signaling ac i i y in he a body. Fo his pu pose, we used he GPH cons uc as a epo e o insulin signaling. This cons uc encodes g een luo escen p o ein (GFP) used o a plecks in homology (PH) domain. In esponse o insulin signaling he GFP-PH usion p o ein becomes edi ec ed o he cell mem- b ane, due o he a ini y o he PH domain o phospha idylinosi ol (3,4,5)- isphospha e. The localiza ion o his epo e is no dis inc in muscles, bu i wo ks well in he a body40. We ound ha GFP was localized a he cell memb ane, bo h in unin ec ed la ae and 27 hou s a e wasp in ec ion. Howe e , 50 hou s a e wasp in ec ion GFP became en iched in he cy oplasm, sugges ing ha , jus like in he muscles, insulin ac i i y is educed in he a body a e a longe pe iod o wasp in ec ion (Fig.2C). www.na u e.com/scien i ic epo s/ 3 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 Figu e 1. PI3K-AKT/TOR signaling in muscles is equi ed o a cellula immune esponse agains wasp in ec ion. (A) Encapsula ion a e when a ec ing insulin signaling, TOR signaling, and Foxo signaling in skele al muscles wi h indica ed geno ypes. (B,C) Numbe o lamellocy es (B) and plasma ocy es (C) pe la a a e 12 h wasp in ec ion, when supp essing PI3K signaling o TOR signaling wi h he indica ed gene ic cons uc s. (D,D”) msn-Che y – labeled lamellocy es in la ae a e 12 h wasp in ec ion, when supp essing insulin o TOR signaling wi h he indica ed gene ic cons uc s. (E) Immunos aining o phospho-AKT om la al skele al muscles, when supp essing insulin signaling by exp essing InRRNAi o InRDN wi h Me 2-GAL4. Quan i ica ion o he immunos ained bands is shown in he lowe panel. (F) Encapsula ion a es when supp essing o ac i a ing PI3K signaling in he a body wi h he indica ed gene ic cons uc s. (G) Encapsula ion a e when www.na u e.com/scien i ic epo s/ 4 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 P e iously we ound ha JAK/STAT signaling ac i i y is induced in skele al muscles as ea ly as 8 hou s a e wasp in ec ion and is main ained a a high ac i i y un il 27 hou s a e wasp in ec ion12. We ha e now u he moni o ed JAK/STAT ac i i y wi h he 10xS a -GFP epo e up o 50 hou s a e wasp in ec ion. As shown in Fig.2D and E, we ound ha GFP exp ession in skele al muscles was educed by 50 hou s a e wasp in ec ion when compa ed wi h he 27-hou ime poin , hough i is s ill highe han in he unin ec ed con ol, indica ing ha a longe pe iod o wasp in ec ion educes he wasp in ec ion-induced ac i a ion o JAK/STAT signaling in skele al muscles. JAK/STAT signaling in muscles egula es insulin signaling sys emically. Supp ession o insulin (Fig.1A–D) o JAK/STAT signaling12 in D osophila la al skele al muscles gene a ed simila pheno ypes, wi h educed lamellocy e p oduc ion and a comp omised encapsula ion esponse agains wasp in ec ion. The e o e, we suspec ed ha hese wo signaling pa hways migh in e ac wi h each o he . Fi s , we in es iga ed whe he JAK/STAT signaling a ec s insulin signaling in he skele al muscles o unin ec ed la ae. Fo his pu pose, we supp essed JAK/STAT signaling in muscles by exp essing he dominan -nega i e S a 92EDN o domeDN cons uc s wi h he Me 2-GAL4 d i e . As shown in Fig.3A and B, his educed he le el o phospho yla ed AKT in he mus- cles, sugges ing a educed le el o insulin signaling ac i i y. On he o he hand, ac i a ion o JAK/STAT signaling by o e exp ession o S a 92E wi h he Me 2-GAL4 d i e did no u he inc ease AKT phospho yla ion (Fig.3C and D). Taken oge he , hese esul s sugges ha JAK/STAT signaling is equi ed in skele al muscles o a no mal le el o insulin ac i i y. This aised he ques ion whe he JAK/STAT signaling a ec s he insulin esponse only locally, in he skele al muscles, o i insulin signaling is sys emically a ec ed also in o he insulin-sensi i e issues, o example he a body. To es his, we supp essed JAK/STAT signaling in he skele al muscles by he same s a egies as abo e, and moni o ed insulin ac i i y in he a body wi h he GPH epo e . We ound ha exp ession o S a 92EDN o domeDN in muscles abolished GFP localiza ion o he cell memb ane in he a body (Fig.3E,F, and G), indica ing ha insulin signaling is also educed in he a body when JAK/STAT signaling is supp essed in skele al muscles. Besides, insulin signaling was also educed in he a body when we supp essed insulin signaling in muscles (Fig.3H). Th ee insulin-like pep ides, Ilp2, Ilp3 and Ilp5, a e p oduced in a clus e o neu osec e o y cells in he b ain, and hey a e belie ed o be pa icula ly impo an in he egula ion o me abolism in D osophila26. We ound ha ilp3 exp ession was subs an ially educed in he b ain when we supp essed JAK/STAT signaling in he muscles, while ilp2 and ilp5 exp ession was inc eased (Fig.3I). This obse a ion is in ag eemen wi h he ecen inding ha ilp3 is speci ically dedica ed o he sys emic con ol o ci cula ing suga s, while ilp2 esponds o amino acids41. Su p isingly, al hough he pheno ypic e ec s o supp essed insulin signaling in he muscles o he wise closely mimic hose o JAK/STAT supp ession, InR supp ession in muscles had no signi ican e ec on he exp ession o any o he h ee insulin-like pep ides in he b ain (Fig.3I). I is possible ha he sys emic e ec s o his geno ype a e media ed by o he insulin-like pep ides elsewhe e. To con ol he speci ici y o ou qPCR assays, we epea ed hem wi h a second se o p ime pai s o each o he h ee ilp genes, wi h i ually iden ical esul s. Figu e3I is based on he mean alues o he wo assays o he indi idual samples. Taken oge he , hese esul s sugges ha supp ession o JAK/STAT signaling in he muscles has a sys emic e ec on insulin signaling, p obably in he en i e o ganism, and ha in e e ence wi h insulin signaling in he muscles has a simila eedback e ec . Nex , we in es iga ed he mechanism by which JAK/STAT signaling may a ec insulin ac i i y. I has p e i- ously been epo ed ha insulin signaling in D osophila la al muscles a ec s body size, p obably by egula ing eeding beha io 42. Thus, we suspec ed ha he sys emically educed insulin ac i i y a e supp essing JAK/STAT signaling in la al skele al muscles was due o dec eased eeding ac i i y. To es his idea, we assayed he la al ood in ake by measu ing he in ake o ood dye, B illian Blue FCF. In ag eemen wi h he published epo s, we ound ha supp ession o insulin signaling in la al muscles educed he in ake o blue dye, indica ing educed eeding ac i i y (Fig.3J). Mo e impo an ly, we obse ed a educed eeding beha io also when we supp essed JAK/STAT signaling in skele al muscles (Fig.3J). In D osophila, i has been shown ha dec eased nu ien a ail- abili y educes insulin signaling36. Al oge he , hese esul s indica e ha blocking JAK/STAT signaling in skele al muscles may ha e a ec ed insulin signaling sys emically, by a ec ing eeding ac i i y and ilp3 exp ession. Since wasp in ec ion educes insulin signaling in bo h muscles and a body, we hypo hesized ha wasp in ec ion migh also educe eeding ac i i y. We measu ed he ood in ake 50 hou s a e wasp in ec ion, and as expec ed i was signi ican ly educed (Fig.3K). This could explain he educed insulin ac i i y du ing wasp in ec ion. Insulin signaling in D osophila skele al muscles egula es JAK/STAT signaling. Nex , we in es- iga ed whe he insulin signaling also egula es JAK/STAT signaling in muscles. We a i icially ac i a ed o sup- p essed insulin signaling in muscles by exp essing InRRNAi o wild ype InR, and moni o ed he esul ing JAK/ STAT ac i i y wi h he 10xS a -GFP epo e . In e es ingly, we ound ha knockdown o InR in muscles educed GFP exp ession om his epo e (compa e Fig.4A and B; quan i ied in Fig.4G), indica ing a educed baseline supp essing o ac i a ing PI3K signaling in hemocy es wi h he indica ed gene ic cons uc s. Da a in o ma ion: Encapsula ion a es we e de e mined in a leas h ee independen expe imen s, and in o al a leas 100 la ae we e analyzed. Fo hemocy e coun s, a leas eigh la ae we e analyzed o each geno ype. Ba s show a e ages and s anda d de ia ions. Fo immunos aining quan i ica ion, a leas h ee independen expe imen s we e analyzed. Ba s show he a e age amoun s o phospho-AKT, no malized o o al AKT, and s anda d de ia ions. The P- alues (unpai ed - es , unequal a iance) a e indica ed, NS: no signi ican . www.na u e.com/scien i ic epo s/ 5 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 Figu e 2. Wasp in ec ion a ec s insulin and JAK/STAT signaling. (A) Wes e n blo , showing he amoun o phospho-AKT in la al skele al muscles a e 27 o 50 h wasp in ec ion. (B) Quan i ica ion o immunos ained bands a e 27 o 50 h wasp in ec ion. (C) Insulin signaling ac i i y in la al a body de ec ed by GPH epo e a e 27 o 50 h wasp in ec ion. (D) JAK/STAT signaling ac i i y in la al skele al muscles de ec ed by he 10XS a -GFP epo e a e 27 o 50 h wasp in ec ion. (E) Quan i ica ion o GFP signaling in muscles. Da a in o ma ion: Fo immunos aining quan i ica ion, a leas h ee independen expe imen s we e analyzed. Ba s show he a e age amoun s o phospho-AKT no malized o o al AKT, and s anda d de ia ions. Fo GFP quan i ica ion, a leas 10 la ae we e quan i ied. Ba s show a e ages and s anda d de ia ions. The P- alues (unpai ed - es , unequal a iance) a e indica ed, NS: no signi ican . www.na u e.com/scien i ic epo s/ 6 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 Figu e 3. JAK/STAT signaling egula es insulin signaling sys emically. (A,B) Amoun o phospho-AKT in la al skele al muscles when supp essing JAK/STAT signaling by exp essing S a 92EDN o domeDN, assayed by wes e n blo ing. (C,D) Amoun o phospho-AKT in la al skele al muscles when ac i a ing JAK/STAT signaling by o e exp essing wild- ype S a 92E, assayed by wes e n blo ing. (E–H) Insulin signaling ac i i y in la al a body isualized by GPH epo e in con ol (E) o a e supp essing JAK/STAT (F,G) o insulin signaling (H) in muscles wi h he indica ed gene ic cons uc s. Localiza ion o GFP o he cell memb ane indica es a high le el o insulin signaling ac i i y. (I) ilp2, ilp3, and ilp5 ansc ip s in la al b ain when www.na u e.com/scien i ic epo s/ 7 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 JAK/STAT ac i i y when insulin signaling is supp essed in muscles. Con e sely, o e exp ession o wild ype InR in muscles led o inc eased JAK/STAT ac i i y (compa e Fig.4C and D; quan i ied in Fig.4H). These esul s sug- ges ha insulin signaling posi i ely egula es JAK/STAT signaling in muscles. In addi ion, in line wi h he abo e esul s, we ound ha exp ession o InRRNAi in muscles blocked he wasp-induced JAK/STAT ac i a ion (compa e Fig.4E and F; quan i ied in Fig.4H). Al oge he , we conclude ha insulin signaling can posi i ely egula e JAK/ STAT signaling in D osophila skele al muscles. Recip ocal in e ac ion be ween JAK/STAT and insulin signaling in he immune esponse. Since we had ound a ecip ocal posi i e in e ac ion be ween he JAK/STAT and insulin signaling pa hways in D osophila muscles, we es ed whe he we could escue he immune esponse when we had supp essed one o hese wo signaling pa hways by s imula ing he o he pa hway. We ound ha o e exp ession o wild ype InR in he backg ound o S a 92EDN o domeDN exp ession in muscles signi ican ly imp o ed he encapsula ion esponse and lamellocy e o ma ion, bu had no ob ious e ec on he numbe o ci cula ing plasma ocy es (Fig.5A–C). Impo an ly, while s imula ing JAK/STAT signaling in muscles does no no mally enhance he immune esponse12, o e exp ession o wild ype S a 92E can la gely escue he encapsula ion esponse and lamellocy e numbe s when InR is supp essed, again wi hou ob ious e ec on he numbe o plasma ocy es in ci cula ion (Fig.5A–C). Thus, o he immune esponse, loss o one o hese wo signaling pa hways in he muscles can be compensa ed by inc eased ac i i y in he o he pa hway, u he unde sco ing he ac ha he JAK/STAT and insulin signaling pa hways in e ac and mu ually a ec each o he . Glycogen s o age in skele al muscles is equi ed o cellula immune esponse. The s ong in e - ac ion be ween JAK/STAT and insulin esponses in muscles, and he sys emic e ec s o hese esponses on insulin signaling in o he issues, aised he ques ion i gene al me abolism we e also a ec ed unde hese condi ions. We i s in es iga ed lipid me abolism in whole animals when we supp essed JAK/STAT o insulin signaling in skel- e al muscles wi h UAS-domeDN, UAS-S a 92EDN, UAS-InRDN, o UAS-InRRNAi. Howe e , nei he ee iglyce ide no o al iglyce ide was ob iously a ec ed (see Supplemen a y Fig.S3). Nex , we in es iga ed i ca bohyd a e le els we e a ec ed, ei he in hemolymph o in skele al muscles. We saw no e ec on ee glucose, bu we ound ha ehalose concen a ion in hemolymph was signi ican ly educed when we supp essed JAK/STAT o insulin signaling in he muscles (see Supplemen a y Fig.S4). This indica es ha JAK/STAT and insulin signaling in muscles a ec ca bohyd a e me abolism in he en i e o ganism. In D osophila la ae he ca bohyd a e s o age o m, glycogen, is p ima ily ound in he skele al muscula u e43,44. We he e o e also assayed he e ec o hese wo signaling pa hways on he glycogen s o es in he muscles. In e es ingly, we ound ha glycogen le els we e d as ically educed in he muscles when we supp essed insulin o JAK/STAT signaling (Fig.6A). Thus, JAK/STAT and insulin signaling in muscles bo h a ec gene al ca bohyd a e me abolism. In ag eemen wi h he indings o Bajga e al.45, he glycogen s o es we e also a ec ed in wasp-in ec ed la ae (Fig.6B), bu we saw no e ec on ee glucose o ehalose (see Supplemen a y Fig.S4). We nex in es iga ed o wha ex en glycogen s o age in muscles a ec s he cellula immune esponse agains wasp in ec ion. We a i icially dec eased o inc eased glycogen s o age in muscles by supp essing glycogen syn- hase o glycogen phospho ylase by exp essing UAS-GlySRNAi o UAS-GlyPRNAi, espec i ely wi h he Me 2-GAL4 d i e (Fig.6C). We ound ha dec eased glycogen s o age in la al skele al muscles d ama ically educed he encapsula ion a e om a ound 50% in con ols o 0% (Fig.6D). Howe e , he e was no ob ious e ec on he encapsula ion a e when glycogen s o age was inc eased (Fig.6D). Fu he mo e, we obse ed a dec eased num- be o lamellocy es when he glycogen s o es we e deple ed in he muscles, bu he e was no ob ious e ec on he numbe o plasma ocy es in ci cula ion (Fig.6E and F). Taken oge he , hese esul s show ha he glycogen ese es in he muscles a e necessa y o an e icien encapsula ion esponse. To es whe he he comp omised immune esponse was due o a dec eased ood in ake, which he e o e led o malnu i ion, we es ed he la al eeding beha io when deple ing he glycogen s o age in muscles. We ound ha ood in ake was signi ican ly educed when he glycogen s o es we e deple ed in he muscles, indica ing ha a su icien supply o glycogen, as an ene gy sou ce o he skele al muscles, was equi ed o no mal eeding ac i i y (Fig.6G). Since wasp in ec ion gene a es a majo shi in he ca bohyd a e me abolism, and since a educed ood in ake also comp omises he immune esponse agains wasp in ec ion, we specula ed ha s a a ion migh on he o he hand a ec he immune esponse45. To es his, we in ec ed s a ed la ae, and ound ha s a a ion signi ican ly educed he encapsula ion a es compa ed wi h he no mal die con ol (Fig.6H), sugges ing ha nu i ion is an impo an ac o o D osophila encapsula ion esponse. Lamellocy es a e a e in heal hy unin ec ed la ae. Howe e , s a a ion alone could induce o ma ion o a small numbe o lamellocy es, bu wasp in ec ion did no u he inc ease he numbe o lamellocy es in he s a ed la ae (Fig.6I), and he numbe s emained much lowe han in ed, in ec ed la ae (Figs1B and 6D). Fu he mo e, s a a ion alone educed he numbe o ci cu- la ing plasma ocy es, compa ed o no mal die con ols (compa e Fig.6J o Figs1C and 6F). These esul s u he unde sco e ha a well- ed condi ion is e y impo an o an e icien cellula immune esponse agains wasp in ec ion. supp essing JAK/STAT o insulin signaling in muscles. (J,K) Feeding ac i i y when supp essing insulin and JAK/STAT signaling (J), o a e 50 h wasp in ec ion (K). Da a in o ma ion: Fo immunos aining quan i ica ion, a leas h ee independen expe imen s we e analyzed. Ba s show he a e age amoun s o phospho-AKT, no malized o o al AKT, and s anda d de ia ions. Fo eeding beha io , a leas h ee independen expe imen s we e pe o med. Ba s show a e ages and s anda d de ia ions. The P- alues (unpai ed - es , unequal a iance) a e indica ed, NS: no signi ican . www.na u e.com/scien i ic epo s/ 8 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 Discussion We obse ed e y simila physiological e ec s when we supp essed JAK/STAT o insulin signaling in he muscles o o he wise heal hy D osophila la ae. In bo h cases, we obse ed supp essed eeding beha io and ca bohyd a e accumula ion in he muscles, and bo h manipula ions in e e ed wi h he abili y o he la ae o cope wi h a wasp in ec ion. The e ec s we e also connec ed in he sense ha supp essed JAK/STAT signaling led o educed insulin signaling, and ice e sa. We conclude ha he basal JAK/STAT ac i i y in he muscles is impo an o he ca - bohyd a e me abolism, no only in he muscles bu in he en i e o ganism. The ole o he inc eased JAK/STAT signaling ha was seen a e wasp in ec ion is s ill unce ain, bu i should p obably be in e p e ed in his con ex . 27 13 10 1214 20 Figu e 4. Insulin signaling in muscles egula es JAK/STAT signaling. (A–F) JAK/STAT signaling ac i i y in la al skele al muscles, de ec ed by he 10xS a -GFP epo e , when supp essing o ac i a ing insulin signaling by exp essing InRRNAi o wild ype InR, espec i ely (A–D), o a e 27 h wasp in ec ion, when supp essing insulin signaling by exp essing InRRNAi (E,F). (G,H) Quan i ica ion o GFP signaling in muscles. Da a in o ma ion: Fo GFP quan i ica ion, a leas 10 la ae we e quan i ied. Ba s show a e ages and s anda d de ia ions. The P- alues (unpai ed - es , unequal a iance) a e indica ed, NS: no signi ican . www.na u e.com/scien i ic epo s/ 9 Scien i ic RepoRTS | 7: 15713 | DOI:10.1038/s41598-017-15940-2 Glucose me abolism as a sou ce o ene gy is c ucial o he g ow h, su i al, p oli e a ion, and di e en ia ion o ac i a ed immune cells in mammals46. In D osophila, Bajga e al. ha e shown ha wi hin 18 hou s a e wasp in ec ion, up o one hi d o he ecen ly inges ed glucose was ealloca ed om o he issues o he hemocy es45. Conside ing ha hemocy es ep esen a e y small ac ion o he la al biomass, hei suga consump ion mus be eno mous. Unde hese condi ions, he wasp-induced JAK/STAT esponse in he muscles may boos insulin signaling and help o edi ec nu ien s owa ds he needs o he immune cells. While muscle-speci ic s imula ion o JAK/STAT signaling could no u he enhance he immune esponse in well- ed la ae12, we ha e he e shown ha i escues he immunode iciency caused by supp essed insulin signaling. The in e ac ions be ween in ec ion, insulin signaling, eeding beha io and immune esponses in ol e ene gy and nu ien s. The e is e idence, om D osophila and o he sys ems, ha nu i ion, mic obiome, me abolism, and immuni y a e all in e connec ed45,47,48. P e ious s udies ha e al eady shown ha he ene gy supply is one o he mos impo an ac o s a ec ing pa hogen g ow h47. Indeed, he p esence o a pa asi e c ea es a compe i ion o a ailable nu ien s and a s a a ion-like s a e in he la a. Ou indings sugges ha a wasp-induced educ ion o ood in ake is de imen al o he cellula immune esponses o wasp in ec ion. I seems ha du ing wasp in ec- ion, he bene i s o an adequa e ood in ake ou weigh hose o a educed ood in ake. Se e al s udies ha e shown ha he con ol o immuni y and me abolism a e mu ually in e connec ed, and ha insulin signaling plays an impo an ole in his in e ac ion ( e iewed in e s28,29). The in e ac ions be ween pa hogens and insulin signaling a e complica ed and, depending on he na u e o he pa hogen, he e ec s may go in opposi e di ec ions. Some s udies ha e shown ha in ec ions educe insulin signaling ac i i y49–51. Howe e , o he s udies show ha some in ec ions can ac i a e insulin signaling, a leas in D osophila52. Ou esul s show ha a e a longe pe iod o wasp in ec ion, insulin signaling ac i i y is educed in bo h D osophila la al skele al muscles and a body. These indings indica e ha modula ion o insulin signaling in esponse o in ec ion is a gene al phenomenon in bo h e eb a es and in e eb a es. On he o he hand, insulin signaling also egula es immune esponses. Inac i a ion o PI3K/AKT signaling gene ally leads o educed in lamma o y and immune esponses, as demons a ed in di e en animal models53. In D osophila, nume ous s udies ha e demons a ed he impo ance o nu ien a ailabili y and insulin/TOR signaling o o ganismal g ow h in gene al and o hema opoiesis in pa icula , a leas in he lymph gland54–58. He e, we ha e ocused on he special ole o muscles and he nu ien s a us in his issue, in he con ex o wasp in ec ion. We ha e shown ha supp ession o insulin signaling in la al skele al muscles, bu no in he a body and hemocy es, blocks he encapsula ion esponse almos en i ely and educes he numbe o lamellocy es, sug- ges ing ha insulin signaling is essen ial in he immune esponse o wasp in ec ion. A he same ime he numbe o ci cula ing plasma ocy es is no mal, showing ha basic la al hema opoiesis is la gely una ec ed. A mu ually posi i e in e ac ion be ween JAK/STAT and insulin signaling has also been obse ed in o he sys- ems. In mammals, he e is e idence sugges ing ha JAK/STAT signaling can di ec ly eed in o he insulin signal- ing pa hway. Fo example, when JAK2 is ac i a ed by ho mones, such as g ow h ho mone o lep in, i can media e phospho yla ion o he insulin ecep o subs a e and he eby ac i a e insulin signaling59,60. Con e sely, p e ious s udies in mammals epo ed ha insulin signaling posi i ely egula es S a 3 and S a 5 in di e en con ex s61–63. Howe e , he li e a u e ega ding hese wo signaling in e ac ions is complex and pa ly con lic ing. Fo ins ance, i was also epo ed ha insulin signaling nega i ely egula es S a 3 ansc ip ion in human melanoma cells64. Figu e 5. Mu ual escue by he JAK/STAT and insulin signaling pa hways. (A–C) Encapsula ion a es (A), numbe o lamellocy es (B) and numbe o plasma ocy es (C) pe la a, when ac i a ing insulin signaling and simul aneously supp essing JAK/STAT signaling, o ice e sa, wi h he indica ed gene ic cons uc s. Da a in o ma ion: Encapsula ion a es we e de e mined in a leas h ee independen expe imen s, and in o al a leas 100 la ae we e analyzed. Fo hemocy e coun s, a leas eigh la ae we e analyzed o each geno ype. Ba s show a e ages and s anda d de ia ions. The P- alues (unpai ed - es , unequal a iance) a e indica ed, NS: no signi ican .