Curbing gastrointestinal infections by defensin fragment modifications without harming commensal microbiota.

ARTICLE Cu bing gas oin es inal in ec ions by de ensin agmen modifica ions wi hou ha ming commensal mic obio a Louis Koeninge 1✉, Lisa Osbel 2,3, Anne Be scheid 4,5, Judi h Wendle 1, Jü gen Be ge 6, Ka ha ina Hipp 6, Till R. Leske 2, Ma ina C. Pils7, Nisa P. Malek1, Benjamin A. H. Jensen8, Heike B ö z-Oes e hel 4,5,9, Till S owig 2,10,11 & Jan Wehkamp1,9,11 The occu ence and sp ead o mul id ug- esis an pa hogens, especially bac e ia om he ESKAPE panel, inc eases he isk o succumb o un ea able in ec ions. We de eloped a no el an imic obial pep ide, Pam-3, wi h an ibac e ial and an ibiofilm p ope ies o coun e his h ea . The pep ide is based on an eigh -amino acid ca boxyl- e minal agmen o human β- de ensin 1. Pam-3 exhibi ed p ominen an imic obial ac i i y agains mul id ug- esis an ESKAPE pa hogens and addi ionally e adica ed al eady es ablished biofilms in i o, p ima ily by dis up ing memb ane in eg i y o i s a ge cell. Impo an ly, p olonged exposu e did no esul in d ug- esis ance o Pam-3. In mouse models, Pam-3 selec i ely educed acu e in es inal Salmonella and es ablished Ci obac e in ec ions, wi hou comp omising he co e mic obio a, hence displaying an added benefi o adi ional b oad-spec um an ibio ics. In conclusion, ou da a suppo he de elopmen o de ensin-de i ed an imic obial agen s as a no el app oach o figh mul id ug- esis an bac e ia, whe e Pam-3 appea s as a pa icula ly p omising mic obio a-p ese ing candida e. h ps://doi.o g/10.1038/s42003-020-01582-0 OPEN 1Depa men o In e nal Medicine I, Uni e si y Hospi al Tübingen, Tübingen, Ge many. 2Depa men o Mic obial Immune Regula ion, Helmhol z Cen e o In ec ion Resea ch, B aunschweig, Ge many. 3ESF In e na ional G adua e School on Analysis, Imaging and Modelling o Neu onal and Inflamma o y P ocesses, O o- on-Gue icke Uni e si y, Magdebu g, Ge many. 4Depa men o Mic obial Bioac i e Compounds, In e acul y Ins i u e o Mic obiology and In ec ion Medicine, Uni e si y o Tübingen, Tübingen, Ge many. 5Ge man Cen e o In ec ion Resea ch (DZIF), Pa ne Si e Tübingen, Tübingen, Ge many. 6Max-Planck Ins i u e o De elopmen al Biology, Elec on Mic oscopy, Tübingen, Ge many. 7Mouse Pa hology and His ology, Helmhol z Cen e o In ec ion Resea ch, B aunschweig, Ge many. 8No o No disk Founda ion Cen e o Basic Me abolic Resea ch, Human Genomics and Me agenomics in Me abolism, Facul y o Heal h and Medical Sciences, Uni e si y o Copenhagen, Copenhagen, Denma k. 9Clus e o Excellence - Con olling Mic obes o Figh In ec ions, Tübingen, Ge many. 10 Clus e o Excellence - Resol ing In ec ion Suscep ibili y, Hanno e , Ge many. 11 These au ho s con ibu ed equally: Till S owig, Jan Wehkamp. ✉email: [email p o ec ed] COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio 1 1234567890():,; The sp ead o an ibio ic- esis an bac e ia is an u gen public heal h h ea . Specifically, he so-called ESKAPE (En e - ococcus aecium, S aphylococcus au eus, Klebsiella pneu- moniae, Acine obac e baumannii, Pseudomonas ae uginosa, and En e obac e species) pa hogens accoun o he majo i y o nosocomial in ec ions wo ldwide being a ibu ed o ≥700,000 dea hs anually1. The Wo ld Heal h O ganiza ion (WHO) has ecen ly published a lis o 12 bac e ia agains which new an i- bio ics a e u gen ly needed, including he ESKAPE pa hogens2. While adi ional an ibio ics figh pa hogens, hey also ha e wide- anging consequences o he commensal gu mic obio a3. Adminis a ion dis up s he mic obial composi ion and can esul in a long-las ing dysbiosis, which is associa ed wi h moun ing diseases4. Dec eased di e si y and - axonomic ichness, he sp ead o an imic obial esis ance as well as inc eased coloniza- ion o oppo unis ic pa hogens, including seconda y in ec ions wi h Clos idioides di ficile, a e jus a ew o he many side-e ec s adi ional an ibio ics impose5,6. The cu en an imic obial c isis is a p oduc o he long- e m neglec ed de elopmen o new an ibio ics by pha maceu ical companies and go e nmen s7. Thus, new s a egies mo e esilien o mul id ug esis ance a e u gen ly wa an ed8. An imic obial pep ides (AMPs) a e small, ca ionic pep ides exis ing in all mul icellula o ganisms and exhibi a b oad ange o an imic obial and immunological p ope ies9. They a e con- side ed a p omising ea men op ion and ha e he po en ial o be a new gene a ion o an imic obials agains mul id ug- esis an bac e ia. Recen ly, de B eij e al. demons a ed he po en ial o no el an imic obial pep ides by de eloping SAAP-148, which showed p omising e ec s agains biofilm-associa ed skin in ec- ions in ex i o human skin and mu ine skin in i o10. Fi s p eclinical ials wi h SAAP-148 agains me hicillin- esis an S. au eus in ec ions ha e al eady been conduc ed acco ding o AdisInsigh —a da abase o d ug de elopmen 11. De ensins, he mos p ominen class o AMPs in humans, a e key e ec o molecules o inna e immuni y. These pep ides p o ec he hos om in ec ious mic obes and shape he composi ion o mic obio a a mucosal su aces12–15. To his end, he fi s iden- ified human β-de ensin, human β-de ensin 1 (hBD-1), is con- s i u i ely exp essed in su ace epi helia by monocy es, plasmacyoid dend i ic cells, and pla ele s16–18. P e iously, he an imic obial ac i i y o hBD-1 was unde es ima ed un il i was analyzed unde educed condi ions as ound in he human in es ine. Reduced hBD-1 has an inc eased an imic obial ac i i y, bu can be deg aded by in es inal p o eases19,20. We ha e ecen ly shown ha his c ea es an eigh -amino acid ca boxyl- e minal agmen (called oc apep ide) wi h e ained an imic obial ac i i y, albei low in i o s abil y21. He e, we le e aged hose findings by de eloping no el syn- he ic pep ides wi h imp o ed an imic obial ac i i y and enhanced in i o s abili y. We modified he hBD-1-de i ed oc apep ide wi h palmi ic acid and a ious space s, such as suga s o amino acids, o c ea e lipopep ides (Pams) wi h inc eased s abili y and bac e icidal ac i i y22,23. The mos p omising pep- ide was es ed agains mul id ug- esis an pa hogens and bio- films ollowed by explo a o y sa e y assessmen . Las ly, we de e mined i s influence on he mu ine mic obio a a e o al applica ion as well as i s e ficacy in mu ine gas oin es inal in ec ion models. Resul s Design and sc eening o oc apep ide based lipopep ides iden- ifies candida e wi h imp o ed an imic obial ac i i y. Lipo- pep ides a e used as an ibio ics which a e highly ac i e agains mul id ug- esis an bac e ia and ungi24,25. P e ious s udies ha e demons a ed an enhanced ac i i y o HDPs a e a y acid modifica ion a he N- e minal end26. The eby, C 14 –C 18 long chain a y acids ha e p o en o be ideal o his pu pose27,28. Wi hin his wo k, we used a C 16 long a y acid, namely palmi ic acid oge he wi h di e en space s such as suga s o amino acids o imp o e s abili y and bac e icidal ac i i y o he ca boxyl- e minal agmen o hBD-1. We designed 5 unique lipopep ides, Fig. 1 Sc eening o di e en ly modified oc apep ides e eals pep ides wi h imp o ed an imic obial ac i i y. a Oc apep ide, he C- e minal eigh amino acids o human β-de ensin 1, was chemically modified wi h palmi ic acid and di e en space s such as suga s o amino acids o 8-amino-3.6-dioxaoc anoic acid (Ado) o gene a e lipopep ides. bAn imic obial ac i i y was measu ed by adial di usion assay. The diame e o inhibi ion zones indica es an imic obial ac i i y; a diame e o 2.5 mm (do ed line) is he diame e o an emp y well. Resul s a e means ± SEM o h ee independen expe imen s. ARTICLE COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 2COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio Pam-1 o Pam-5, based on he ecen ly disco e ed oc apep ide (Fig. 1a). This se o pep ides was sc eened o hei an imic obial ac i i y agains se e al pa hogenic bac e ia using a adial di usion assay (Fig. 1b). Pam-1, Pam-4, and Pam-5 we e gene ally inac i e agains es ed s ains, hus con as ing he po en inhibi ion o bac e ial g ow h media ed by Pam-2 and Pam-3. Bac e ial g ow h was mos s ongly inhibi ed by Pam-3, ei he on pa (S. au eus)o supe io (C. oden ium,P. ae uginosa, and S. Typhimu ium) o he oc apep ide, poin ing owa d modifica ion-specific ac i i ies. No ably, bo h Pam-2 and Pam-3 consis en ly inhibi ed S. Typhimu ium g ow h, a species he non-modified oc apep ide ailed o inhibi . Pam-3 displays bac e icidal ac i i y agains mul id ug- esis an bac e ia and slow esis ance selec ion. We used a b o h mic o- dilu ion assay o u he analyze he po en ial o ou lipopep ides o kill mul id ug esis an bac e ia belonging o he ESKAPE pa hogen panel (Supplemen a y Fig. 1 and Supplemen a y Table 1). Pam-1 and Pam-4 displayed no o low bac e icidal ac i i y, whe eas Pam-5 showed mode a e e ec s agains hese pa hogens. Simila o he esul s o he adial di usion assay, bo h Pam-2 and Pam-3 we e highly e ec i e. Rema kably, Pam-2 and Pam-3 inhibi ed he g ow h o an A. baumannii isola e (DSM30007), which is o he wise esis an o he las - eso an ibio ics, colis in and igecycline. Despi e some bac e icidal simila i ies be ween Pam-2 and Pam-3, he la e p o ed supe io o all o he Pam’s and was highly e ec i e agains hese bac e ia a concen a ions o 4.69–18.75 µM (Fig. 2a and Supplemen a y Fig. 1). We acco dingly selec ed Pam-3 o u he cha ac e iza ion as a po en ial he apeu ic agains mul id ug esis an bac e ia. As he de elopmen and selec ion o an ibio ic- esis an bac e ia in esponse o new an ibio ic candida es is a significan p oblem, we assessed he abili y o S. au eus and S. Typhimu ium o de elop esis ance agains Pam-3. When cul u ed in he p esence o sub-inhibi o y concen a ions o Pam-3 o 25 passages, no significan inc ease in he minimal inhibi ion concen a ion (MIC) was obse ed o S. au eus. In con as , he MIC o he s anda d an ibio ic, i ampicin s a ed o apidly inc ease a e fi e passages and had inc eased ≥4096- old a e 15 passages (Fig. 2b). Simila ly, al hough exposed o Pam-3 o con inuous se ial passages, no esis an S. Typhimu ium isola es eme ged, whe eas he p esence o cip ofloxacin esul ed in an inc eased MIC al eady a e 3 passages, and a ≥256- old MIC inc ease a e 19 passages (Fig. 2c). Pam-3 elimina es es ablished biofilms and causes apid killing by pe meabilizing he bac e ial memb ane. Bac e ial biofilms a e highly esis an o g ow h inhibi o s and bac e icidal ea - men egimens. Apa om he hinde ed pene a ion o an i- bac e ial agen s, ea men is u he complica ed by 10–1000 imes inc eased ole ance exhibi ed by biofilm p o ec ed bac e ia compa ed o plank onic bac e ia29. Because o ha , we assessed he abili y o Pam-3 o e adica e es ablished biofilms in a dose- depended manne . Wi hin 1 h, 300 µM o Pam-3 elimina ed P. ae uginosa in biofilms (K uskal–Wallis es , p=0.0074, Fig. 3a) and simila ly e adica ed ~99.99% o S. au eus in biofilms (K uskal–Wallis es , p=0.0026, Fig. 3a). A p ima y a ge o an imic obial pep ides is he bac e ial cell en elope. Dis u bing Fig. 2 Pam-3 kills mul id ug- esis an ESKAPE pa hogens and esis ance was no selec ed. a Suscep ibili y o mul id ug- esis an ESKAPE (E. aecium,S. au eus,K. pneumoniae,P. ae uginosa and E. coli) pa hogens o an ibio ics and Pam-3. Bac e ia suscep ible o all (blue boxes) o in e media e/ esis an o a leas one ( ed boxes) o he an ibio ics pe class. G ay boxes a e shown i he suscep ibili y o agen s in ha class was no assessed. Resul s a e exp essed as he LC 99.9 , he lowes pep ide concen a ion in mic omola ha esul ed in ≥99.9% killing. Resul s a e medians (and anges) o h ee independen expe imen s. I no ange is indica ed, hen he LC 99.9 was iden ical in all expe imen s. b,cResis ance de elopmen o S. au eus ATCC 25923 (b) and S. Typhimu ium DSM554 (c) o Pam-3 ( ed line) and he an ibio ics i ampicin and cip ofloxacin (black line), espec i ely. Values a e old changes (in log 2 )in minimal inhibi o y concen a ion (MIC) ela i e o he MIC o he fi s passage. COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 ARTICLE COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio 3 he in eg i y and unc ion o he ou e and/o inne memb anes esul s in loss o he ba ie unc ion and dissipa ion o he memb ane po en ial30. To cla i y he mode o ac ion o Pam-3, we used a ypuA p omo o -based luci e ase epo e s ain o B. sub ilis o iden i y cy oplasmic memb ane-associa ed and cell en elope- ela ed s ess31. The ypuA p omo e was ac i a ed (2 old) by Pam-3, indica ing cell en elope impai men (Fig. 3b). Hence, o s eng hen his esul , we analyzed he influence o Pam-3 on he ansmemb ane po en ial o S. au eus NCTC8325. The p o onopho e ca bonyl cyanide m-chlo ophenyl hyd azone (CCCP) was used as a posi i e con ol o depola ize bac e ia, i.e., leading o a educ ion o hei memb ane po en ial. Upon depola iza ion, DiOC 2 (3) shi s om g een fluo escence owa d ed emission because o sel -associa ion o he dye molecules. Pam-3 ea men caused a b eakdown o he memb ane po en ial in a concen a ion-dependen manne (Fig. 3c). Bo h esul s emphasize ha Pam-3 ac s on bac e ial memb anes. Since a la ge a ie y o AMPs a ge he memb ane as po e o me s32, we nex analyzed he abili y o Pam-3 o induce memb ane lesions. To his end, we ea ed S. au eus NCTC8325wi h Pam-3 a 4x MIC and added a mix u e o Sy o9 and p opidium iodide (PI). The memb ane-pe mean Sy o9 s ains all li ing cells g een, whe eas he ed-fluo escen PI can only en e cells h ough la ge mem- b ane po es o lesions. Pam-3 led o a s ong influx o PI (Fig. 3d). To u he assess he mechanism o killing, ansmis- sion elec on mic oscopy (TEM) was pe o med o analyze changes in bac e ial mo phology o compa e bac e ial mo - phology be o e and a e ea men wi h Pam-3. In ag eemen wi h he po e o ma ion and he induc ion o cell en elope s ess and depola iza ion, Pam-3 ea men esul ed in s ong cell en elope damage wi h dis up ed memb anes and po es in mos o he cells. Fu he mo e, addi ional memb anous s uc u es could be obse ed in many cells simila o hesus macaque θ- de ensins33. Addi ionally, scanning elec on mic oscopy (SEM) was employed o obse e cell mo phological changes a e Pam-3 ea men di ec ly. Exposu e o Pam-3 esul ed in memb ane su ace dis up ion and lysed cells simila o hBD120,while con ol cells exhibi ed a b igh and smoo h su ace (Fig. 3e). Po e o ma ion was associa ed wi h as killing o hese bac e ia. We assayed bac e icidal kine ics o assess he apidness o Pam-3 media ed killing. Pam-3 killed mo e han 90% o P. ae uginosa wi hin 1 min and S. au eus wi hin 15 min (Fig. 3 ). E adica ion o he le el o de ec ion was obse ed 2 and 30 min a e Pam-3 ea men , espec i ely. Fig. 3 Pam-3 elimina es es ablished biofilms, causing apid killing by inducing la ge po es o lesions. a Bac e icidal ac i i y o Pam-3 agains es ablished biofilms o S. au eus ATCC25923 (black line) and P. ae uginosa PAO1 (g ay line). Resul s a e exp essed as he numbe o iable bac e ia (in log 10 CFU) a e 1 h ea men o 24 h old biofilms wi h Pam-3. Values a e means ± SEM o h ee eplica es om h ee independen expe imen s. S a is ics we e e alua ed by using he K uskal–Wallis es . bAc i a ion o he B. sub ilis ypuA p omo e by Pam-3 indica es cell en elope s ess. Da a a e p esen ed as mean ± SEM. Expe imen s we e ca ied ou h ee independen imes. cE ec o Pam-3 on S. au eus NCTC8325 memb ane po en ial a e 30 min o ea men . The p o onopho e CCCP was used as a posi i e con ol and 0.01% ace ic acid as a nega i e con ol. Da a a e p esen ed as mean ± SEM o h ee independen expe imen s. dFluo escence mic oscopy o S. au eus NCTC8325 ea ed wi h Pam-3 (4x MIC) e eals po e o ma ion by causing a s ong influx o ed- fluo escen p opidium iodide. Scale ba s: 1 µm. eT ansmission elec on mic oscopy o high-p essu e ozen, eeze-subs i u ed, and embedded S. au eus NCTC8325 ea ed wi h Pam-3 o 0.01% ace ic acid o 30 min o obse e memb ane dis up ion. Scale ba s, 0.2 µm. S. au eus NCTC8325 was exposed o Pam-3 o 0.01% ace ic acid o 60 min. The samples we e fixed in Ka no sky’s eagen , and mo phology was analyzed by scanning elec on mic oscopy. Scale ba s, 2 µm. Killing o S. au eus ATCC25923 (black line) and P. ae uginosa PAO1 (g ay line) a e 1–120 min exposu e o 9.38 µM (1x MIC) Pam-3. Resul s a e exp essed as he numbe o iable bac e ia (in log 10 CFU) pe millili e . Values a e means ± SEM o h ee independen expe imen s. ARTICLE COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 4COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio O ally adminis e ed Pam-3 showed good acu e ole abili y in an animal model. Po en ial side e ec s o o ally adminis e ed Pam-3 we e assessed in mice. His ological analysis and de e mi- na ion o se um ma ke s 24 h a e applica ion o wo doses o 250 µg Pam-3 did no e eal acu e oxici y. Specifically, he e we e no al e a ions in bodyweigh and no signs o sys emic oxici y o dis ess (Fig. 4a). Mo eo e , measu emen o se um le els o glu amic oxaloace ic ansaminase and c ea inine showed no significan di e ences be ween he g oups sugges ing no e ec on kidney and li e me abolism (Fig. 4b, c). Finally, his ological examina ion o gas oin es inal issues, li e , and kidney e ealed no al e a ions, excep a mino sho ening o in es inal illi he jejunum-ileum o one con ol and wo ea ed animals (Table 1). As no di e ence in hese his opa hological findings was obse ed be ween PBS and Pam-3 ea ed animals, hey we e ega ded as backg ound obse a ions (Table 1and Supplemen a y Fig. 2). Thus, we conclude ha Pam-3 ea men was no associa ed wi h acu e oxici y. Pam-3 p ese es he co e gu mic obio a. To in es iga e he e ec o Pam-3 on he in es inal mic obio a, we ea ed mice wice a an 8-h in e al wi h Pam-3 (125 o 250 µg/each dose) o PBS o ally and collec ed esh ecal samples be o e and 24 h a e applica ion. Mic obio a composi ion was analyzed using 16S RNA sequencing. Analysis o be a di e si y and calcula ion o weigh ed Uni ac Dis ances demons a ed ha changes in he mic obio a be ween he be o e and a e samples we e simila be ween Pam-3 ea ed mice and PBS ga aged con ol mice (Fig. 5a, b). Simila ly, while mino changes in he communi y s uc u e we e obse ed in bo h g oups (i.e., ea ed and un ea ed), he numbe o de ec ed species as well as he com- plexi y (Wilcoxon-Tes , Fig. 5c, d) emained compa able, hus con as ing ea men wi h adi ional an ibio ics, such as ampi- cillin (Supplemen a y Fig. 3). In line wi h he analysis o alpha and be a di e si y, Pam-3 ea men did no a ec he abundance o bac e ial gene a (Fig. 5e, ). Combined, hese esul s demon- s a e ha , wi h he applica ion egime conduc ed, Pam-3 ea - men o heal hy chow- ed mice does no a ec he o e all communi y s uc u e o di e si y o he mic obio a. Pam-3 ea men comba s acu e and es ablished gas o- in es inal in ec ions o S. Typhimu ium and Ci obac e oden ium. To assess he e ficacy o Pam-3 on acu e in es inal bac e ial in ec ions, mice we e in ec ed wi h S. Typhimu ium and ea ed o ally 6 and 22 h pos in ec ion wi h 250 µg pep ide o PBS (Fig. 6a). Pam-3 ea ed animals showed significan ly educed colony- o ming uni s (CFU) o S. Typhimu ium in cecum con en and issue (Mann–Whi ney es , p< 0.0001 and p=0.0409, Fig. 6b). Fu he mo e, Pam-3 also lowe ed he bac- e ial load in he small in es ine con en wi hou a ec ing he small in es ine issue (Mann–Whi ney es , p=0.0024 and p= 0.8621, Fig. 6c) and ends o educe weigh loss (Fig. 6d). We subsequen ly assessed he he apeu ic po en ial o Pam-3 agains an al eady es ablished in es inal in ec ion. To his end, mice we e in ec ed wi h C. oden ium and ecei ed wo doses o 250 µg Pam-3 o PBS 5 days pos in ec ion (Fig. 6a). Pam-3 ea men educed he numbe o bac e ia in he cecum con en and cecum issue (Mann–Whi ney es , p=0.0104 and p= 0.0473, Fig. 6e). T ea men also significan ly educed he numbe o iable Ci obac e in colon con en and colon issue (Mann–Whi ney es , p=0.0010 and p=0.0104, Fig. 6 ). Fig. 4 Sa e y o o ally applied Pam-3 in mice. Dose-depended o al ole ance es in mice. Animals we e ea ed wice wi h 125 µg (o ange ci cles) o 250 µg Pam-3 ( ed ci cles) o PBS (black ci cles). aWeigh change o mice (125 µg Pam-3, N=5; 250 µg Pam-3, N=6 and PBS, N=5), bglu amic oxaloace ic ansaminase (GOT) le els o ea ed animals wi h 125 µg Pam-3 (N=4) o 250 µg Pam-3 (N=6) o PBS (N=5) and cc ea inine le els (125 µg Pam-3, N=4; 250 µg Pam-3, N=6 and PBS, N=5) one day a e Pam-3 applica ion. Resul s a e p esen ed as mean ± SEM o biologically independen animals. Table 1 Sa e y o o ally applica ion o Pam-3 in mice. Gas oin es inal ac PBS Pam-3 S omach Glycogen 0/5 0/6 Lymphocy ic agg ega es 0/5 0/6 Duodenum-Jejunum Dysplasia 0/5 0/6 Inflamma ion 0/5 0/6 Jejunum-Ileum Dysplasia 1/5 2/6 Inflamma ion 0/5 0/6 Pane h cells 0/5 0/6 Cecum and Colon Dysplasia 0/5 0/6 Inflamma ion 0/5 0/6 Li e Glycogen 0/5 0/6 Lymphocy ic agg ega es 1/5 0/6 Anisoca yosis 0/5 0/6 Double nuclea ed cells 0/5 0/6 Hema opoiesis 0/5 0/6 Fa y change 0/5 0/6 Kidney Glome uli 0/4 0/6 Tubuli 3/4 3/6 Papilla 0/4 0/6 Pel is 0/4 0/6 O al ole ance es in mice. Animals we e ea ed wice wi h 250 µg Pam-3 o PBS. Resul s a e exp essed as he numbe o he o al numbe o animals wi hin he g oups ha showed signs o pa hology wi hin 24 h a e ea men . COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 ARTICLE COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio 5 Toge he hese da a co obo a e he in i o e ficacy o Pam-3 agains wo di e en en e ic pa hogens. Discussion The inc easing numbe o mul id ug- esis an pa hogens is one o he g ea es challenges o ou ime34. The Cen e s o Disease Con ol and P e en ion (CDC) es ima es ha he annual in ec- ion a e exceeds 2.8 million cases in USA alone, esul ing in mo e han 35,000 dea hs caused by mul id ug- esis an bac e ia and ungi35. Due o his ala ming de elopmen , al e na i es o con- en ional an ibio ics a e u gen ly needed36,37. He ein, we epo ha Pam-3, a palmi oleic acid-modified oc apep ide agmen om hBD-1, is e ec i e agains mul id ug- esis an ESKAPE ARTICLE COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 6COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio pa hogens in i o. Mo eo e , Pam-3 is also highly e ec i e in i o agains A. baummanii esis an o he las - eso an i- bio ics, colis in and igecycline. Biofilm-encased bac e ia a e much less suscep ible o con en ional an ibio ics han hei iden ical plank onic coun e pa s complica ing ea men s38.In con as o mos an ibio ics, ou expe imen s showed ha Pam-3 was able o e adica e es ablished S. au eus and P. ae uginosa biofilms in i o. De elopmen o an ibio ic esis ance is inc easing a an ala ming a e39. He e, we demons a ed he lack o esis ance de elopmen o Pam-3 in G am-posi i e (S. au eus) and G am- nega i e (S. Typhimu ium) bac e ia compa ed o con en ional an ibio ics, when cul u ed o 25 passages in he p esence o sub- inhibi o y concen a ions. This esul indica es ha esis ance de elopmen agains Pam-3 is a a e e en 40. One o he easons o his obse a ion could be he apid killing o bac e ia by Pam-3 and i s associa ed mode o ac ion. AMPs wi h bac e icidal e ec s o en in e ac wi h memb anes as pa o hei mode(s) o ac ion. Apa om memb ane-dis up i e mechanisms, like po e o ma ion, AMPs can kill h ough elec- os a ics and localized pe u ba ions o non-memb ane- dis up i e mechanisms, which a ge s mul iple mic obial p o- cesses and/o physiological unc ions41–43. Ou analysis e ealed ha Pam-3 causes cell en elope s ess by b eaking down he memb ane po en ial and inducing po e o - ma ion. Taken oge he , he capaci y o induce apid killing o G am-posi i e and G am-nega i e bac e ia, a low isk o esis- ance selec ion, combined wi h biofilm e adica ion unde sco es he alue o Pam-3 o u he d ug de elopmen . The ise in an ibio ic esis ance and he u gen ly sea ch o po en ial al e na i es shi ed g ea e esea ch ocus on AMPs37,44. Mos AMPs in p e- and clinical de elopmen a e e alua ed o opical a he han o al adminis a ion, o di e en easons such as in i o e ficacy and s abili y45. We demons a ed he e ficiency o o ally adminis e ed Pam-3 in wo di e en in i o models o gas oin es inal in ec ions, namely S. Typhimu ium and C. oden ium. In bo h models, Pam-3 ea men esul ed in sig- nifican ly educed bac e ial bu den in he gas oin es inal ac . Fu he s udies, including p ope dose egimens, a e necessa y o ull e adica ion o bo h in ec ions. An ibio ic ea men has non negligible consequences. Obse - a ional, clinical, and epidemiologic s udies ha e demons a ed Fig. 5 Pam-3 ea men does no a ec mic obio a di e si y bu has influence on ce ain bac e ia phyla. Chow- ed mice we e ea ed o ally wice a an 8-h in e al wi h 125 µg Pam-3 (N=5) o 250 µg Pam-3 (N=6) o PBS (N=5) as a con ol. Feces samples we e collec ed be o e and a e ea men o obse e sho e m changes in he mic obiome. aP incipal coo dina e analysis (PCoA including g oup mean) o ecal mic obio a composi ion using Weigh ed UniF ac Dis ances be o e and a e ea men , espec i ely. bPCoA including g oup mean o ecal mic obio a composi ion using Unweigh ed UniF ac Dis ances, espec i ely be o e and a e ea men . cRichness (obse ed species) be o e and a e ea men . dFecal mic obio a was calcula ed by Shannon’s Di e si y index. The s a is ical significance was calcula ed by using Wilcoxon es . ePam-3 ea men a ec s he abundance o bac e ial gene a. Agg ega ed by genus. S a is ical analysis pe o med by he LE Se pla o m (h ps://galaxyp ojec .o g/lea n/ isualiza ion/cus om/le se/) using de aul se ings. Fig. 6 O al applica ion o Pam-3 comba s acu e and es ablished in ec ions o S. Typhimu ium and C. oden ium.aO e iew o expe imen al se up. Mice we e in ec ed wi h S. Typhimu ium ( ed ci cles) and ea ed o ally wi h ei he 250 µg Pam-3 (N=12) o PBS (N=9) a e 6 h and 22 h pos in ec ion. Mice we e in ec ed wi h C. oden ium (blue ci cles) and ea ed o ally wi h ei he 250 µg Pam-3 (N=11) o PBS (N=11) a e 5 days pos in ec ion. bCFU/ml o S. Typhimu ium in cecum con en and issue. cCFU/ml o S. Typhimu ium in small in es ine con en and issue. dBody weigh change du ing acu e S. Typhimu ium in ec ion. eCFU/g o C. oden ium in cecum con en and issue. CFU/g o C. oden ium in colon con en and issue. Resul s a e exp essed as he numbe o iable bac e ia (in log 10 CFU) in he lumen and issue and p esen ed as mean ± SEM o biologically independen animals. S a is ics we e e alua ed by using he Mann–Whi ney es . COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 ARTICLE COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio 7 ha an ibio ic ea men a ec s he gu mic obio a composi ion wi h immedia e e ec s on heal h5,46. Changes in he mic obio a composi ion, dec eased di e si y, educed axonomic ichness, and a so-called dysbiosis a e he main consequences47,48. Fu he , an ibio ics can ha e long- e m e ec s such as inc eased suscep - ibili y o in ec ions, obesi y, and obesi y-associa ed me abolic diseases5,49. In con as o con en ional an ibio ics, Pam-3 ea men had unexpec edly no app eciable e ec s on commen- sal mic obes. Al hough u he s udies a e wa an ed o igo ously add ess his assump ion, ou da a poin s owa d pa hogen specific killing by Pam-3 ea men , which would be a significan ad an age o e con en ional an ibio ics whe e dis up i e e ec on he esiden mic obio a as well as a apid d op in di e si y is commonly obse ed50. Fu he mo e, a e an ibio ic ea men , he in es ine is o en colonized by non-commensal bac e ia, which can esul in long- e m en i onmen al changes51. Ins ead o a loss o di e si y, high-dose Pam-3 ea ed mice showed an una ec ed bac e ial di e si y. Ou da a poin s owa d an inc eased numbe o obse ed species and nominally inc eased di e si y. This obse a ion is in line wi h p e ious s udies on o he agmen s om ou g oup52 sugges ing ha a ge ed a ack on ce ain ( a he high-abundan ) bac e ial s ains ees up new niches o low-abundan axa, which we e p e iously below he limi o de ec ion. Mic obio a-modula ing capabili ies o AMPs could make a di e ence in he ea men o many gu mic obio a associa ed diseases, including inflamma o y bowel disease53. In conclusion, he esul s p esen ed he e demons a e ha Pam-3 is a p omising al e na i e o figh mul id ug- esis an in ec ions in a pos -an ibio ic wo ld because o i s b oad an i- mic obial ac i i y agains G am-posi i e and G am-nega i e pa hogens and i s e ficacy agains gas oin es inal in ec ions wi hou dis up ing he esiden mic obio a. Fu u e s udies a e he e o e wa an ed o examine he ull po en ial o his and o he bios able no el pep ides. Pam-3 could open a new chap e o an e ec i e, mic obio a-sa ing ea men s a egy o bac e ial in ec ions. Me hods Mic oo ganisms and cul u e. Clinical isola es o A. baumannii DSM30007, E. aecium DSM2918, K. pneumoniae DSM30104, and P. ae uginosa DSM1117 we e p o ided by he Depa men o Labo a o y Medicine a Robe -Bosch-Hospi al S u ga , Ge many. C. oden ium DSM16636 and E. coli DSM8695 we e ob ained om he Deu sche Sammlung on Mik oo ganismen und Zellkul u GmbH B aunschweig, Ge many. Clinical isola es o A. baumannii LMG944, A. baumannii ECII, E. coli 6940, E. coli DSM682, E. aecium 11037 CHB, E. aecium 20218 CHB, K. pneumoniae 6727 and K. pneumoniae 6970 as well as S. au eus DSM20231, S. au eus ATCC25923, S. au eus ATCC33592, S. au eus ATCC43300, S. en e ica se o a Typhimu ium DSM554, P. ae uginosa ATCC27853, P. ae uginosa NRZ01677 and P. ae uginosa PAO1 we e p o ided by he Ins i u e o Medical Mic obiology and Hygiene Tübingen, Ge many. B. sub ilis ypuA and S. au eus NCTC8325 was ob ained om he In e acul y Ins i u e o Mic obiology and In ec ion Medicine, Tübingen, Ge many. The wild- ype S. Typhimu ium s ain SL1344 ha bo ing a ch omosomally in eg a ed luxCDABE casse e, which is con- fi med by kanamycin esis ance54 and he nalidixic acid and kanamycin- esis an , bioluminescen C. oden ium s ain ICC18055 we e ob ained om Helmhol z Cen e o In ec ion Resea ch B aunschweig, Ge many. All bac e ia we e s o ed in c yo ials (Ro h) a −80 °C. Be o e each expe imen , inocula om he ozen s ocks we e g own o e nigh a 37 °C on LB o Columbia blood aga pla es (BD). Fo expe imen s, esh cul u es we e p epa ed in yp ic soy b o h (BD). Pep ides. All lipopep ides we e chemically syn hesized by EMC Mic ocollec ions GmbH (Tübingen, Ge many) and pu ified by p ecipi a ion. In de ail, pep ide syn hesis was pe o med by solid-phase Fmoc/ e -bu yl chemis y on Chlo o-(2′- chlo o) i yl polys y ene esin (Rapp Polyme e, Tübingen, Ge many) using an au oma ed pep ide syn hesize o mul iple pep ide syn hesis (Sy o, Mul iSynTech, Ge many). Side chain-p o ec ing g oups o Fmoc-amino acids we e 2,2,4,6,7- pen ame hyldihyd obenzo u an-5-sul onyl (A g), i yl (Cys) and e - bu yloxyca bonyl (Lys). Fmoc-p o ec ed amino acids including Fmoc-8-amino-3,6- dioxaoc anoic acid (Fmoc-Ado-OH, IRIS Bio ech GmbH, Ge many) we e coupled wice (double couplings, se en old mola excess o amino acids) by in si u ac i a ion using DIC/HOB and TCTU wi h DIEA. The emo al o he Fmoc- p o ec ing g oup was ca ied ou wice by ea men wi h pipe idine/DMF (1:4, / ). Resins we e washed wi h DMF (6x) a e each coupling and dep o ec ion s ep. N- e minal acyla ion o he esin bound pep ide was pe o med manually using palmi ic acid (1.2 eq), DIC/HOB (1.5 eq) wi h DIEA (2 eq) in NMP/DCM (4/1) o 16 h. A e washing wi h DMF (8x) and DCM (5x) comple eness o acyla ion was confi med by Kaise -Tes . The lipopep ide was clea ed o he esin and side-chain dep o ec ed by ea men wi h TFA/ eagen K/H2O (80/15/5, / / ) o 3 h and p ecipi a ed by adding die hyl e he . A e cen i uga ion, he pep ides we e dissol ed in e -bu yl alcohol/H2O (4:1, / ) and lyophilized. The s uc u e o he lipopep ide was confi med by RP-HPLC-ESI-MS and was >>90% (Supple- men a y Fig. 4). All pep ides we e dissol ed in 0.01% ace ic acid. An imic obial ac i i y. An imic obial ac i i y was analyzed by adial di usion assay56. Log-phase bac e ia we e cul i a ed o up o 18 h in TSB (TSB, Bec on Dickinson, USA), washed and dilu ed o 4 × 106CFU in 10 ml aga . Bac e ia we e incuba ed in 10 ml o 10 mM sodium phospha e, pH 7.4, con aining 0.3 mg/ml o TSB powde and 1% (w/ ) low EEO-aga ose (AppliChem). 1 µg o each lipopep ide was pipe ed in o punched wells in a final olume o 4 µl and di used in o he gel o 3 h a 37 °C. A e ha , a nu ien - ich gel wi h 6% TSB (w/ ) and 1% aga ose in 10 mM sodium phospha e bu e was pou ed on op o he fi s gel and incuba ed o up o 24 h a 37 °C. Then he diame e o inhibi ion zones was measu ed. Bac e icidal ac i i y. Bac e icidal ac i i y was assessed by b o h mic odilu ion assay57. Log-phase bac e ia we e collec ed by cen i uga ion (2500 pm, 10 min, 4 °C), washed wice wi h 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB and he op ical densi y a 600 nm was adjus ed o 0.1. App oxima ely 5 × 105 CFU/ml bac e ia we e incuba ed wi h se ial pep ide concen a ions (1.17–150 µM) in a final olume o 100 µl in 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB o 2 h a 37 °C. A e incuba ion, 100 µl o 6% TSB (w/ ) we e added and abso bance was measu ed a 600 nm (Tecan, Swi ze land) and moni o ed o 18 h. A e wa d, 100 µl pe well we e pla ed on LB aga o de e mine he numbe s o iable bac e ia mic obiologically. Bac e icidal ac i i y is exp essed as he LC 99.9, he lowes concen a ion ha killed ≥99.9 % o bac e ia. Fo ime-kill expe imen s, bac e ia (5 × 105CFU/ml) we e incuba ed wi h 9.38 µM Pam-3 in 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB in LoBind ubes (Eppendo ) in a o al olume o 550 µl. As an un ea ed con ol, bac e ia we e incuba ed in 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB. A e incuba ion a 37 °C and 150 pm o 1–120 min, a sample o 50 µl was aken om he suspension and added o 50 µl o a 0.05% ( / ) sodium polyane hol sul ona e (Sigma-Ald ich) solu ion, which neu alizes emaining pep ide ac i i y, and pla ed on LB aga o de e mine he numbe o iable bac e ia. Resis ance de elopmen . De elopmen o esis ance o he pep ides was assessed wi h S. au eus and S. Typhimu ium. Fo compa ison, he de elopmen o esis ance o he clinically ele an an ibio ics i ampicin and cip ofloxacin (Sigma-Ald ich) was de e mined. Bac e ia we e cul u ed o e nigh a 37 °C a 150 pm in TSB. Bac e ia we e washed wice wi h 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB. Washed bac e ia we e incuba ed wi h se ial Pam-3 o an ibio ic con- cen a ions (wi h final concen a ions o 1.17–150 µM pep ide o 0.0156–0.5 µg/ml i ampicin o cip ofloxacin) in a final olume o 100 µl in 10 mM sodium phos- pha e bu e con aining 1% (w/ ) TSB o 2 h a 37 °C. A e incuba ion, 100 µl o 6% TSB (w/ ) was added and pla es incuba ed in a humidified a mosphe e o 21 h a 37 °C and 150 pm. The MIC, he lowes concen a ion o pep ide/an ibio ic ha caused a lack o isible bac e ial g ow h, was de e mined o each bac e ial species. The ea e , 5×10 5CFU/ml o he 0.5- old MIC suspension was added o a esh medium con aining pep ides/an ibio ics and hese mix u es we e incuba ed as desc ibed abo e. This was epea ed o 25 passages. T ea men o es ablished biofilms. A log-phase cul u e o P. ae uginosa was dilu ed in BM2 medium and o S. au eus in TSB o 5 × 105CFU/ml. 100 µl o each bac e ial suspension was added o a ound-bo om polys y ene mic o i e pla e and incuba ed o 24 h a 37 °C in a humidified a mosphe e. Then, plank onic bac e ia we e emo ed by wo wash s eps wi h PBS. Nex , biofilms we e exposed o se ial pep ide dilu ions (9.38–300 µM) in a final olume o 100 µl in 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB o 1 h a 37 °C in a humidified a mosphe e. As a con ol, bac e ia we e exposed o 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB wi hou pep ide. A e wa d, adhe en bac e ia in each well we e esuspended, and he numbe o iable bac e ia was de e mined mic obiologically. To isualize he da a on a loga i hmic scale, a alue o 1 CFU was assigned when no g ow h occu ed. The biofilm deg ada ion assay was pe - o med in ag eemen wi h he o iginal epo desc ibing his me hod fi s 58. In e ac ion wi h he bac e ial memb ane. A specific bac e ial epo e s ain wi h he gene ic backg ound o Bacillus subi ilis 1S34, ca ying he p omo e o he ypuA gene, used o he fi efly luci e ase epo e gene, was used o iden i y cell en elope- ela ed damage caused by ea men wi h an imic obial compounds31. The assay was ca ied ou in ag eemen wi h o me epo s desc ibing his me hod20,59. ARTICLE COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 8COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio Bac e ia we e cul u ed o an OD 600 o 0.9 in LB b o h wi h 5 µg/ml e y h omycin a 37 °C and dilu ed o an OD 600 o 0.02. Se ial pep ide dilu ions (0.15–150 µM) we e p epa ed in a mic o i e pla e and incuba ed wi h he adjus ed bac e ial suspension a 37 °C o 1 h. Subsequen ly, ci a e bu e (0.1 M, pH 5) con aining 2 mM luci e in (I is Bio ech, Ge many) was added and luminescence was measu ed using a mic opla e eade (Tecan, Swi ze land). Bac e ial memb ane po en ial. Fo de e mina ion o memb ane po en ial chan- ges, S. au eus NCTC8325 was g own o log-phase in LB +0.1% glucose, ha es ed and he op ical densi y a 600 nm (OD 600 ) was adjus ed o 0.5. Bac e ia we e incuba ed wi h 30 µM 3,3′-die hyloxaca bocyanine iodide (DiOC 2 (3), In i ogen™) o 15 min in he da k and ea ed wi h se ial pep ide concen a ions o 30 min. The p o onopho e ca bonyl cyanide m-chlo ophenyl hyd azone (CCCP, Sigma Ald ich) was used as a posi i e con ol and DMSO o 0.01% ace ic acid as nega i e con ols. Fluo escence was measu ed a an exci a ion wa eleng h o 485 nm and wo emission wa eleng hs, 530 nm (g een) and 630 (nm) ed, using a mic opla e eade (Tecan, Swi ze land). Bac e ial po e o ma ion. Po e o ma ion was moni o ed using he Li e/Dead BacLigh bac e ial iabili y ki (Molecula P obes)60.S. au eus was g own in LB a 37 °C o log-phase and, 100 µl aliquo s we e ea ed wi h 37.5 µM Pam-3 (4x MIC) o le un ea ed as a con ol. Samples we e aken a e 10 min o pep ide ea men , hen 0.2 µl o a 1:1 mix u e o SYTO9 and p opidium iodide (PI) was added and u he incuba ed o 15 min a RT in he da k. Fluo escence mic oscopy was ca ied ou using a Zeiss Axio Obse e Z1 au oma ed mic oscope. Images we e acqui ed wi h an O ca Flash 4.0 V2 came a (Hamama su), C Plan-Apo 63x/1.4 Oil DIC and alpha Plan-Apoch oma 100x/1.46 Oil Ph3 objec i es (Zeiss) and p o- cessed using he Zen so wa e package (Zeiss). T ansmission and scanning elec on mic oscopy. Mo phologic analysis o bac e ia was cha ac e ized by elec on mic oscopy21. App oxima ely 1.2 × 109 CFU/ml bac e ia we e incuba ed wi h 37.5 µM Pam-3 (4x MIC) in 10 mM sodium phospha e bu e con aining 1% (w/ ) TSB b o h o 30 o 120 min a 37 °C. As a con ol, bac e ia we e exposed o 0.01% ace ic acid. A e wa d bac e ia we e fixed in Ka no sky’s eagen . Fo ansmission elec on mic oscopy (TEM), bac e ia we e high-p essu e ozen (HPF Compac 03, Enginee ing O fice M. Wohlwend GmbH) in capilla ies, eeze-subs i u ed (AFS2, Leica Mic osys ems) wi h 2% OsO 4 and 0.4% u anyl ace a e in ace one as subs i u ion medium and embedded in EPON. Ul a hin sec ions we e s ained wi h u anyl ace a e and lead ci a e and analyzed wi h a Tecnai Spi i (The mo Fishe Scien ific) ope a ed a 120 kV. Fo scanning elec on mic oscopy (SEM), bac e ia we e washed in PBS and finally fixed wi h 1% OsO 4 on ice o 1 h. Nex , samples we e p epa ed on polylysin-coa ed co e slips, dehyd a ed in a g adua ed se ies o 100% e hanol and c i ical poin d ied (Pola on) wi h CO 2 . Finally, samples we e spu e -coa ed wi h a 3 nm hick laye o pla inum (Sa ema ic CCu-010) and examined wi h a Hi achi Regulus 8230 field emission scanning elec on mic oscope (Hi achi) a an accele a ing ol age o 5 kV. Mice. C57BL/6N mice we e gene a ed and main ained (including b eeding and housing) a he animal acili ies o he Helmhol z Cen e o In ec ion Resea ch (HZI) unde enhanced specific pa hogen- ee (SPF) condi ions61. Animals used in he expe imen s we e gende and age ma ched. Female and male mice wi h an age o 8–12 weeks we e used. S e ilized ood and wa e ad libi um was p o ided. Mice we e kep unde a s ic 12-h ligh cycle (ligh s on a 7:00 am and o a 7:00 pm) and housed in g oups o wo o six mice pe cage. All mice we e eu hanized by asphyxia ion wi h CO 2 and ce ical disloca ion. All animal expe imen s we e pe o med in ag eemen wi h he local go e nmen o Lowe Saxony, Ge many (app o ed pe mission No. 33.19-42502-04-18/2499). Sa e y. Age- and gende -ma ched mice ecei ed wo doses o pep ides (0, 125 µg o 250 µg) sol ed in 100 µl PBS o ally pe day. Bodyweigh and appea ance we e eco ded. The ollowing day, mice we e sac ificed and s omach, kidney, spleen, li e , small in es ine, cecum, and colon we e emo ed o his ological sco ing. A ound 1 ml o blood was aken om he hea o measu e inflamma o y ma ke s, including c ea inine in he kidney and he enzyme le els o glu ama e-oxalace a e- ansaminase (GOT) in he li e . In ec ion and ea men o mice Salmonella Typhimu ium in ec ion. Fo S. Typhimu ium in ec ions expe imen s, age- and sex-ma ched mice be ween 10 and 14 weeks o age we e used. Bo h- emale and male mice we e used in expe imen s. Wa e and ood we e wi hd awn o 4 h be o e mice we e ea ed wi h 20 mg/mouse o s ep omycin by o al ga age. A e wa d, mice we e supplied wi h wa e and ood ad libi um. 20 h a e s ep- omycin ea men , wa e and ood we e wi hd awn again, 4 h be o e he mice we e o ally in ec ed wi h 105CFU o S. Typhimu ium in 200 μl PBS. D inking wa e ad libi um was supplied immedia ely and ood 2 h pos in ec ion (p.i.). A e 6 and 22 h p.i. mice ecei ed 250 µg pep ide sol ed in 100 µl PBS o only PBS o ally. 48 h a e in ec ion, mice we e sac ificed, and in es inal o gans we e emo ed o assess he bac e ial bu den in he lumen and issues. Mice we e weighed e e y day o eco d po en ial body-weigh loss. Ci obac e oden ium in ec ion. Bioluminescence exp essing C. oden ium s ain ICC180 was used o all in ec ion expe imen s55.C. oden ium inocula we e p e- pa ed by cul u ing bac e ia o e nigh a 37 °C in LB b o h wi h 50 µg/ml kana- mycin. Subsequen ly, he cul u e was dilu ed 1:100 in esh medium, and sub- cul u ed o 4 h a 37 °C in LB b o h62. Bac e ia we e washed wice in phospha e- bu e ed saline (PBS). Mice we e o ally inocula ed wi h 108CFU o C. oden ium dilu ed in 200 µl PBS. Weigh o he mice was moni o ed and eces we e collec ed o measu emen s o he pa hogen bu den. A e 5 days pos in ec ion mice ecei ed wice 250 µg pep ide sol ed in 100 µl PBS o only PBS o ally. The ol- lowing day, mice we e sac ificed o assess bac e ial bu den in he lumen and issues o he cecum and he colon. Analysis o bac e ial loads in eces. F esh ecal samples we e collec ed and weigh ed. Samples we e homogenized in 1 ml LB media by bead-bea ing wi h 1 mm zi conia/silica beads wice o 25 s using a Mini-Beadbea e -96 (BioSpec). To de e mine CFUs, dilu ions o homogenized samples we e pla ed on LB pla es wi h 50 μg/ml kanamycin. Analysis o bac e ial loads in in es inal con en and sys emic o gans. All mice we e eu hanized by asphyxia ion wi h CO 2 a indica ed ime poin s. In es inal is- sues (small in es ine, cecum, colon) we e emo ed asep ically. To collec in es inal con en , o gans we e flushed wi h PBS. O gans we e opened longi udinally, cleaned ho oughly wi h PBS and weigh ed. O gans and con en we e homogenized in PBS using a Poly on homogenize (Kinem a ica). Dilu ions o homogenized samples we e pla ed on LB pla es con aining 50 μg/ml kanamycin o de e mine CFUs. Mic obio a analysis. Feces samples we e collec ed a di e en ime poin s (be o e and a e in ec ion), and bac e ial DNA was ex ac ed om snap- ozen eces using a phenol-chlo o o m-based me hod p e iously desc ibed63. In b ie , 500 µl o ex ac ion bu e (200 mM T is (Ro h), 20 mM EDTA (Ro h), 200 mM NaCl (Ro h), pH 8.0), 200 µl o 20% SDS (AppliChem), 500 µl o phenol:chlo o o m: isoamyl alcohol (PCI) (24:24:1) (Ro h) and 100 µl o zi conia/silica beads (0.1 mm diame e ) (Ro h) we e added pe eces sample. Lysis o bac e ia was pe o med by mechanical dis up ion using a Mini-BeadBea e -96 (BioSpec) o wo imes 2 min. A e cen i uga ion, he aqueous phase was p ocessed by ano he phenol:chlo o- o m:isoamyl alcohol ex ac ion be o e p ecipi a ion o DNA using 500 µl iso- p opanol (J.T. Bake ) and 0.1 olume o 3 M sodium ace a e (Applichem). Samples we e incuba ed a −20 °C o a leas se e al hou s o o e nigh and cen i uged a 4 °C a maximum speed o 20 min. The esul ing DNA pelle was washed, d ied using a speed acuum and esuspended in TE Bu e (Applichem) wi h 100 µg/ml RNase I (Applichem). C ude DNA was column pu ified (BioBasic Inc.) o emo e PCR inhibi o s. 16S RNA gene amplifica ion o he V4 egion (F515/R806) was pe o med acco ding o an es ablished p o ocol p e iously desc ibed64. B iefly, DNA was no malized o 25 ng/µl and used o sequencing PCR wi h unique 12-base Gola y ba codes inco po a ed ia specific p ime s (ob ained om Sigma). PCR was pe o med using Q5 polyme ase (NewEnglandBiolabs) in iplica es o each sample, using PCR condi ions o ini ial dena u a ion o 30 s a 98 °C, ollowed by 25 cycles (10 s a 98 °C, 20 s a 55 °C, and 20 s a 72 °C). A e pooling and no maliza ion o 10 nM, PCR amplicons we e sequenced on an Illumina MiSeq pla o m ia 250 bp pai ed-end sequencing (PE250). Using Usea ch8.1 so wa e package (h p://www.d i e5.com/usea ch/) he esul ing eads we e assembled, fil e ed and clus e ed. Sequences we e fil e ed o low-quali y eads and binned based on sample-specific ba codes using QIIME 1.8.065. Me ging was pe o med using - as q_me gepai s—wi h as q_maxdi s 30. Quali y fil e ing was conduc ed wi h as q_fil e (- as q_maxee 1), using a minimum ead leng h o 250 bp and a minimum numbe o eads pe sample =1000. Reads we e clus e ed in o 97% ID OTUs by open- e e ence OTU picking and ep esen a i e sequences we e de e mined by use o UPARSE algo i hm66. Abundance fil e ing (OTUs clus e >0.5%) and axonomic classifica ion we e pe o med using he RDP Classifie execu ed a 80% boo s ap confidence cu o 67. Sequences wi hou ma ching e e ence da ase we e assembled as de no o using UCLUST. Phylogene ic ela ionships be ween OTUs we e de e mined using Fas T ee o he PyNAST alignmen 68. Resul ing OTU absolu e abundance able and mapping file we e used o s a is ical analyses and da a isualiza ion in he R s a is ical p og amming en i onmen package phyloseq69. S a is ics and ep oducibili y. Expe imen al esul s we e analyzed o s a is ical significance using G aphPad P ism 8.2 (G aphPad So wa e Inc.). Apa om mic obio a analysis, all da a we e analyzed by Mann–Whi ney es o K uskal–Wallis es . He e, di e ences we e analyzed by Wilcoxon signed- ank es o K uskal–Wallis es compa ison o o als be ween g oups70. OTUs wi h K uskal–Wallis es <0.05 we e conside ed o analysis. COMMUNICATIONS BIOLOGY | h ps://doi.o g/10.1038/s42003-020-01582-0 ARTICLE COMMUNICATIONS BIOLOGY | (2021) 4:47 | h ps://doi.o g/10.1038/s42003-020-01582-0 | www.na u e.com/commsbio 9