Host-inherent variability influences the transcriptional response of Staphylococcus aureus during in vivo infection.

ARTICLE Recei ed 4 Jul 2016 |Accep ed 14 Dec 2016 |Published 3 Feb 2017 Hos -inhe en a iabili y influences he ansc ip ional esponse o S aphylococcus au eus du ing in i o in ec ion Robe Tha ¨ne 1, Oli e Goldmann1, And eas Beineke2& E a Medina1 The ise o an ibio ic esis ance calls o al e na i e s a egies o ea bac e ial in ec ions. One a ac i e s a egy is o di ec ly a ge bac e ial i ulence ac o s wi h an i- i ulence d ugs. The exp ession o i ulence ai s by pa hogens is, howe e , no cons i u i e bu a he induced by he le el o s ess encoun e ed wi hin he hos . He e we use dual RNA sequencing (RNA-seq) o show ha in insic a iabili y in he le el o hos esis ance g ea ly a ec s he pa hogen’s ansc ip ome in i o. Th ough analysis o he ansc ip ional p ofiles o hos and pa hogen du ing S aphylococcus au eus in ec ion o wo mouse s ains, shown o be suscep- ible (A/J) o esis an (C57BL/6) o he pa hogen, we demons a e ha he exp ession o i ulence ac o s is dependen on he encoun e ed hos esis ance. We u he mo e p o ide e idence ha his dependence s ongly influences he e ficacy o an i- i ulence s a egies, highligh ing a po en ial limi a ion o he implemen a ion o hese s a egies. DOI: 10.1038/ncomms14268 OPEN 1In ec ion Immunology Resea ch G oup, Helmhol z Cen e o In ec ion Resea ch, Inho ens asse 7, 38124 B aunschweig, Ge many. 2Ins i u e o Pa hology, Uni e si y o Ve e ina y Medicine Hanno e , Bu ¨n eweg 17, 30559 Hanno e , Ge many. Co espondence and eques s o ma e ials should be add essed o E.M. (email: e a.medina@helmhol z-hzi.de). NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions 1 The ise o an imic obial esis ance is one o he mos challenging p oblems in mode n medicine, causing an inc ease in mo bidi y and mo ali y associa ed wi h common bac e ial in ec ions1. While a ailable an ibio ics a e loosing hei e ec i eness, he in oduc ion o no el bac e icidal o bac e ios a ic an ibio ics canno be conside ed a long- e m solu ion because i is e en ually ollowed by he eme gence o esis an bac e ial clones ha become inc easingly p e alen unde selec i e d ug p essu e. Consequen ly, he e is a p essing need o new an i-in ec i e agen s ha do no impose simila le els o selec ion p essu e on pa hogens as classical an ibio ics. In his ega d, al e na i e app oaches based on a enua ing bac e ial pa hogenesis by a ge ing bac e ial i ulence, he so-called ‘an i- i ulence’ s a egies, a e eme ging as p omising ools o he ea men o in ec ions2. Bac e ial pa hogens exp ess a la ge epe oi e o di e en i ulence ac o s o su i e unde he ad e se condi ions imposed by he hos en i onmen . Thus, an i- i ulence s a egies ha e been p oposed ha specifically a ge bac e ial oxins p oduced by he pa hogen o e ade hos de enses3, bac e ial ac o s media ing adhesion o he hos 4, sec e ion sys ems5as well as egula o y sys ems6and quo um-sensing signalling7. The key ea u e o an i- i ulence d ugs is he a enua ion o he pa hogen’s i ulence o aid clea ance by he hos ’s immune de enses2. These d ugs seem a ac i e, because i is belie ed ha no killing he pa hogen di ec ly exe s less selec i e p essu e o he de elopmen o esis ance2. Howe e , such an app oach will only con e he apeu ic benefi i he a ge ed i ulence ac o (s) a e ac ually exp essed by he bac e ium du ing in ec ion and i he na u al de ense mechanisms o he hos a e s ong enough o clea he pa hogen, weakened by he an i- i ulence ea men . Bac e ial pa hogenesis, on he o he hand, is s ongly influenced by he s eng h o he hos immune de ense. Fo example, a i ulen mic oo ganisms can be pa hogenic o immunocomp omised hos s, whe eas i ulen mic oo ganisms can be nonpa hogenic in immune hos s8. This si ua ion is u he complica ed by he ac ha , in addi ion o he immune s a us, inhe en cha ac e is ics o he hos , such as he gene ic backg ound, significan ly influence he capabili y o he immune sys em o o e come in ading pa hogens. Thus, he esponse o a specific pa hogen can ange om weak in suscep ible hos s, causing se e e in ec ions, o s ong in mo e esis an indi iduals, esul ing in milde diseases. These di e ences imply ha pa hogens will encoun e s onge immune p essu e in esis an han in suscep ible hos s and i ulence ac o s ha a e essen ial o coun e ac ing a weak immune esponse may no be he same as hose equi ed unde s onge immune p essu e in esis an hos s. The e o e, he dependence o i ulence ac o exp ession on hos esis ance is a po en ial limi a ion o he e ec i eness o an i- i ulence d ugs. He e we in es iga e how in insic a iabili y o hos esis ance o a pa hogen a ec s he exp ession o i ulence ac o s needed o success ully in ec he hos . We use S aphylococcus au eus,a human pa hogen ha can cause se e e in asi e in ec ions9and is no o ious o i s capaci y o de elop an ibio ic esis ances. These cha ac e is ics make S. au eus one o he mos dange ous and in ac able in ec ious pa hogens wo ldwide10. Despi e nume ous a emp s o de elop a accine ha can p e en S. au eus in ec ions, none o he accine candida es es ed in clinical ials has succeeded so a 11. This ailu e, in combina ion wi h he inc ease o an ibio ic esis ance, has lead o an in ensifica ion o e o s o sea ch o al e na i e ea men app oaches in ecen yea s. In his ega ds, an i- i ulence s a egies a ge ing c ucial pa hogenici y ac o s p oduced by S. au eus du ing in ec ion ha e been p oposed as an a ac i e he apeu ic op ion12,13. Howe e , since he ou come o S. au eus in ec ion is s ongly influenced by he hos ac o s such as acial o igin, age and gene ic makeup14,15, he sea ch o an i- i ulence a ge s in S. au eus has o con- side he inhe en a iabili y o he hos esponses o in ec ion. Simila o humans, a iabili y in he hos esponse o S. au eus has been also obse ed among di e en inb ed s ains o mice16. While some mouse s ains ( o example, A/J and DBA/2) a e e y suscep ible o S. au eus in ec ion, C57BL/6 mice a e highly esis an and su i e a bac e ial dose ha apidly kills mice om suscep ible s ains16. These di e ences in he capaci y o con ol S. au eus in ec ion p o ide a unique expe imen al sys em o explo e he ex en o which in insic hos a iabili y a ec s he exp ession o bac e ial i ulence ac o s du ing in ec ion. In his s udy, we use dual RNA sequencing (RNA-seq), and wo mouse s ains p e iously shown o display di e en ial suscep ibili y o S. au eus in ec ion (s ain A/J is suscep ible and s ain C57BL/6 is esis an )16,17, o in es iga e how he in insic a iabili y o hos esis ance o a pa hogen a ec s he exp ession o i ulence ac o s. Dual RNA-seq enables he simul aneous de e mina ion o he ansc ip ional esponse o he hos and pa hogen du ing in ec ion and does no equi e physical sepa a ion o p oka yo ic and euka yo ic RNA since he sequencing eads can be assigned o he hos o o he pa hogen genomes by in silico analysis18. Al hough dual RNA-seq has been success ully applied o cha ac e ize he ansc ip ional signa u e o bac e ia and hos in se e al in i o in ec ion sys ems14,15,19,20, he s udy p esen ed he e is one o he fi s using his echnology in an in i o sys em. Ou esul s demons a e he impac o he hos gene ic backg ound on he ansc ip ional esponse o S. au eus du ing in ec ion, and p o ide expe imen al e idence ha hos -dependen bac e ial exp ession o i ulence ac o s is a po en ial limi a ion o he e ficacy o an i- i ulence he apies. Resul s C57BL/6 and A/J mice di e in hei esis ance o S. au eus. Fo his s udy, we selec ed wo s ains o mice ha di e in hei suscep ibili y o S. au eus in ec ion16,17. Whe eas A/J mice a e e y ulne able o S. au eus, esul ing in a significan ly inc eased bac e ial g ow h in he kidneys (Fig. 1a) and li e (Fig. 1b) and a al in ec ion ou come (Fig. 1c), C57BL/6 mice exhibi g ea e esis ance o S. au eus and we e capable o significan ly es ic bac e ial g ow h in he kidneys (Fig. 1a) and li e (Fig. 1b) wi h all mice su i ing (Fig. 1c). The e o e, hese wo mouse s ains p o ide an excellen ool o in es iga ing he ex en o which a iabili y in he hos esponse o in ec ion a ec s he ansc ip ional esponse o S. au eus wi hin he hos . Dual RNA-seq analysis o S. au eus and in ec ed hos issue.A dual RNA-seq app oach ha enables simul aneous ansc ip ional p ofiling o bac e ia and hos issue was used o cha ac e ize he hos esponse o in ec ion and in es iga e he impac o di e en le els o hos esis ance on he ansc ip ional esponse o S. au eus. The schema ic ep esen a ion o he expe imen al design is shown in Fig. 1d. To al RNA including hos and pa hogen RNA, was isola ed om he kidneys o A/J and C57BL/ 6 mice a 48 h o in ec ion and analysed by Illumina deep sequencing. Be ween 9 and 21 millions eads we e uniquely aligned o he e e ence genome o Mus musculus assembly GRCm38.p3 (GCA_000001635.5), while be ween 32,228 and 5 millions eads we e uniquely mapped o he e ised e e ence genome o S. au eus s ain 8325-4 ( e . 21). T ansc ip ome analysis o S. au eus-in ec ed mice. Hie a chical clus e ing (Supplemen a y Fig. 1a) and p incipal componen analysis (PCA) (Supplemen a y Fig. 1b) o gene exp ession da ase s om in ec ed A/J and C57BL/6 mice, as well as o ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 2NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions issue om unin ec ed con ol mice, demons a ed high wi hin- g oup ep oducibili y and subs an ial be ween-g oup di e ences. Pai wise compa ison by pe mu a ional mul i a ia e analysis o a iance (PERMANOVA) e ealed ha he di e ences be ween all g oups we e s a is ically significan (Po0.05, Supple- men a y Table 1). Di e en ial gene exp ession analysis using DESeq2 ( e . 22) iden ified a o al o 5,540 di e en ially exp essed genes (DEGs) be ween unin ec ed and S. au eus-in ec ed A/J mice, o which 2,756 we e ound o be up- egula ed in esponse o in ec ion, while 2,784 we e down egula ed (Supplemen a y Da a 1). In C57BL/6 mice, 3,559 genes we e di e en ially exp essed be ween unin ec ed and S. au eus-in ec ed mice, o which 1,758 we e up egula ed and 1,801 we e down- egula ed in esponse o in ec ion (Supplemen a y Da a 2). The numbe s o unique and o e lapping DEGs be ween S. au eus-in ec ed A/J and C57BL/6 mice a e shown in Fig. 2a. Func ional classifica ion o he DEGs using KEGG (Kyo o Encyclopedia o Genes and Genomes) pa hway en ichmen analysis, e ealed ha a la ge g oup o hos genes wi h inc eased exp ession in esponse o in ec ion in bo h A/J and C57BL/6 mice belonged o he g oups ‘cy okine-cy okine ecep o in e ac ion’ and ‘chemokine signalling pa hway’ (Fig. 2b). Pa icula ly, genes encoding inflamma o y cy okines such as IL-6, IL-1a, IL-1band TNF-aas well as chemokines in ol ed in he chemoa ac ion o monocy es/mac ophages such as Cxcl1, Cxcl2 and Cxcl3, we e up egula ed in bo h A/J and C57BL/6 mice in esponse o S. au eus in ec ion (Supplemen a y Da a 1 and 2). Also, hos genes encoding acu e phase p o eins such as Saa1, Saa2, hap oglobin and he calcium-binding p o eins S100a8 and S100a9 we e highly induced in in ec ed A/J and C57BL/6 mice. Al hough he global analysis o he ansc ip ional da a sugges ed ha a ‘co e’ se o inflamma ion- ela ed genes was highly exp essed in bo h A/J and C57BL/6 mice in esponse o S. au eus in ec ion, he a e age old change o exp ession in his se o genes was ma kedly highe in in ec ed A/J han in in ec ed C57BL/6 mice (Fig. 2). This sugges ed ha A/J mice de eloped a mo e in ense sys emic inflamma ion han C57BL/6 mice in esponse o S. au eus in ec ion, which is indica i e o se e e sepsis leading o dea h. Besides he sys emic hype in- flamma ion, he inc eased exp ession o he gene encoding he coagula ion ac i a o issue ac o (F3) and o he gene encoding he fib inolysis inhibi o PAI-1 (Se pine1) in A/J mice (Supplemen a y Da a 1), e ealed a ne p o-coagulan s a us ha is ypical o se e e sepsis23. Al e ed coagula ion, coupled wi h mic o ascula dys unc ion occu ing du ing sepsis, dec eases issue pe usion. This leads o pe u ba ions o oxygen supply, esul ing in issue hypoxia and he ac i a ion o he hypoxia- inducible ac o alpha encoded by Hi 1a24,25. The signi- fican induc ion o Hi 1a obse ed in A/J, bu no in C57BL/6 mice, in esponse o in ec ion (Supplemen a y Tables 2 and 3), indica ed mo e se e e issue hypoxia in he kidneys o in ec ed A/J mice han in hose o C57BL/6 mice. Fu he mo e, he genes encoding a ginase 1 (A g1) and a ginase 2 (A g2) we e exp essed o a g ea e ex en in in ec ed A/J han in in ec ed C57BL/6 mice 10 9 85 4 3 7 7100 80 60 40 20 0 0246 Time o in ec ion (days) Pe cen su i al 810 A/J C57BL/6 P < 0.01 * 6 6 A/J C57BL/6 A/J C57BL/6 Resis an C57BL/6 mice 4 × 107 CFU S. au eus 4 × 107 CFU S. au eus 48 h Kidneys Dual RNA-seq Resis an hos ansc ip ome Suscep ible hos ansc ip ome Valida ion wi h isogenic mu an s S. au eus ansc ip ome in esis an hos S. au eus ansc ip ome in suscep ible hos Dual RNA-seq Suscep ible A/J mice * CFU o S. au eus in kidneys (Log10) CFU o S. au eus in li e (Log10) abc d Figu e 1 | A/J and C57BL/6 mice exhibi opposed le els o esis ance o S. au eus.Bac e ial loads in he kidneys (a) and li e (b) o A/J and C57BL/6 mice a 48 h a e in a enous inocula ion wi h 4 107CFU o S. au eus SH1000. Each symbol ep esen s he bac e ial coun s de e mined in an indi idual mouse and he ho izon al lines ep esen he a e age±s.d. o each mouse s ain. One ep esen a i e expe imen ou o h ee independen expe imen s is shown (n¼6, - es , *Po0.05). (c) Su i al cu es o A/J and C57BL/6 mice in a enously in ec ed wi h 4 107CFU o S. au eus SH1000 (n¼5, log- ank es , Po0.01). (d) Schema ic summa y o he expe imen al design o dual RNA-seq analysis. Suscep ible A/J mice and esis an C57BL/6 mice we e in ec ed in a enously wi h 4 107CFU o S. au eus SH1000, hei kidneys emo ed a 48 h a e bac e ial inocula ion and subjec ed o dual RNA-seq analysis o simul aneously de e mine he gene exp ession p ofile o he hos and pa hogen in he same sample. The genes di e en ially exp essed by S. au eus in A/J and C57BL/6 mice we e iden ified and ela ed o he in ec ion-associa ed ansc ip ional esponse o he co esponding mouse s ain. The e ec o a ge ing a i ulence ac o di e en ially exp essed by S. au eus SH1000 be ween in ec ion o A/J and C57BL/6 mice was also de e mined. NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 ARTICLE NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions 3 (Supplemen a y Da a 1). A ginase can con ibu e o endo helial cell dys unc ion by deple ing ex a-cellula L-a ginine and ni ic oxide (NO) bioa ailabili y, leading o endo helial ni ic oxide syn hase uncoupling and consequen ly o he p oduc ion o high le els o ha m ul eac i e oxygen species (ROS)26. Taken oge he , hese findings sugges ha he mic o-en i onmen in he in ec ed issue is highly di e en be ween A/J and C57BL/6 mice, which could significan ly a ec he exp ession o i ulence de e minan s by S. au eus. S. au eus ansc ip ome du ing in ec ion o A/J o C57BL/6. In pa allel, we analysed he ansc ip ome o S. au eus du ing he in ec ion o esis an C57BL/6 and suscep ible A/J mice o de e mine he impac o he di e en physiological condi ions p esen a he si e o in ec ion on he pa hogen’s ansc ip ional esponse. Hie a chical o dina ion (Supplemen a y Fig. 2a) and PCA (Supplemen a y Fig. 2b) showed high wi hin-g oup ep o- ducibili y, while PERMANOVA demons a ed ha he an- sc ip ional esponse o S. au eus in ec ing suscep ible A/J mice di e ed significan ly om ha o S. au eus in ec ing esis an C57BL/6 mice (Supplemen a y Table 2). T ansc ip abundance was de e mined by no malizing he numbe o aw eads in each da a se o gene leng h and exp essed as ansc ip s pe Kilobases pe Million (TPM) (Supplemen a y Da a 3). A o al o 85 genes we e iden ified as di e en ially exp essed (p obabili y alue Z0.95) by S. au eus be ween in ec ion o A/J and in ec ion o C57BL/6 mice using NOISeq (Supplemen a y Da a 4). O hose, ansc ip s o 20 genes we e mo e abundan in S. au eus in ec ing A/J mice (Fig. 3a, Table 1, Supplemen a y Table 4), 65 genes exhibi ed g ea e exp ession in S. au eus du ing in ec ion o C57BL/6 mice (Fig. 3a, Table 2, Supplemen a y Table 5) and 594 we e exp essed a a simila le el by S. au eus in A/J and C57BL/6 mice (Fig. 3a, Supplemen a y Da a 4). One o he mos p ominen ope ons exp essed by S. au eus o a g ea e ex en du ing in ec ion o A/J mice, han du ing in ec ion o C57BL/6 mice, was he a c ope on, which encodes he a ginine Inc eased exp ession A/J A/J 1,722 1,012 Log2 old change 2 Toll-like ecep o signalling pa hway T cell ecep o signalling pa hway RIG-I-like ecep o signalling pa hway Na u al kille media ed cy o oxici y NOD-like ecep o signalling pa hway In es inal immune ne wo k o IgA p oduc ion Hema opoie ic cell lineage Cy osolic DNA-sensing pa hway Cy okine-cy okine ecep o in e ac ion Complemen and coagula ion cascades Chemokine signalling pa hway B cell ecep o pa hway A/J 0 20 40 60 80 100 Numbe o genes C57BL/6 A/J C57BL/6 3456 789 C57BL/6 C57BL/6 791 967 1,789 Dec eased exp ession a b Figu e 2 | Gene exp ession analysis o S. au eus-in ec ed kidneys om A/J and C57BL/6 mice. (a) Venn diag am showing he numbe o DEGs wi h inc eased (le ) o dec eased ( igh ) exp ession in esponse o S. au eus in ec ion ha a e unique o common be ween A/J and C57BL/6 mice. (b) Hea map o he mean log 2 - old change o gene exp ession o he significan ly up- egula ed genes de e mined be ween S. au eus-in ec ed e sus unin ec ed A/J and C57BL/6 mice wi hin he KEGG ‘immune sys em’ ca ego y (le pa ). The co esponding numbe s o significan ly up- egula ed genes in esponse o S. au eus in ec ion in A/J (blue ba s) and C57BL/6 ( ed ba s) a e shown in he igh pa o he figu e. Table 1 | DEGs be ween S. au eus in ec ing A/J and C57BL/6 mice wi h g ea e ansc ip abundance du ing in ec ion o A/J mice. Locus ag Gene symbol Desc ip ion SAOUHSC_00845 Hypo he ical SAOUHSC_02853 Hypo he ical SAOUHSC_00371 yflT Hypo he ical SAOUHSC_02964 a cR Hypo he ical SAOUHSC_01477 Hypo he ical SAOUHSC_01969 g pP Hypo he ical SAOUHSC_00101 d m Phosphopen omu ase SAOUHSC_01181 xynA Hypo he ical SAOUHSC_02967 a cD A ginine/o ni hine an ipo e SAOUHSC_01191 pmB 50S ibosomal p o ein L28 SAOUHSC_00686 Hypo he ical SAOUHSC_01803 aapA Hypo he ical SAOUHSC_02862 clpL ATP-dependen Clp p o ease, ATP-binding subuni ClpC SAOUHSC_01403 cspA Cold shock p o ein SAOUHSC_02850 cidB Hypo he ical SAOUHSC_01002 qoxB quinol oxidase AA3 subuni II SAOUHSC_01024 g aF Hypo he ical SAOUHSC_02702 Hypo he ical SAOUHSC_02697 cyC Amino acid ABC anspo e ATP-binding p o ein SAOUHSC_02665 Hypo he ical DEG, di e en ially exp essed gene. The comple e da a o he di e en ially exp essed genes wi h highe exp ession by S. au eus du ing in ec ion o he suscep ible A/J mice a e displayed in Supplemen a y Table 4. ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 4NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions deiminase (ADI) sys em (Fig. 3b, Supplemen a y Da a 4). The a c ope on comp ises he genes coding o a ginine deiminase (a cA), o ni hine ansca bamylase (a cB), ca bama e kinase (a cC), he a ginine/o ni hine an ipo e (a cD) and he ansc ip ional egula o A cR (a cR). These enzymes ca alyse he con e sion o a ginine o o ni hine, ammonia, and CO 2 , while p oducing ATP. This no only enables S. au eus o u ilize a ginine as an ene gy sou ce unde anae obic condi ions27 bu also gene a es ammonia, which con e s p o ec ion agains acidic s ess28. The exp ession o all genes o he ADI pa hway was g ea e in S. au eus du ing in ec ion o A/J han du ing in ec ion o C57BL/ 6 mice, e en hough only he di e ences in he le el o a cR and a cD exp ession eached s a is ical significance ( alse disco e y a e (FDR) o0.05, NOISeq analysis), (Fig. 3b, Supplemen a y Da a 4). O he genes, which we e di e en ially exp essed by S. au eus du ing in ec ion o A/J and C57BL/6 mice, we e hose encoding p o eins o amino acid anspo and biosyn hesis. Thus, ans- c ip s o he genes encoding a hypo he ical cys eine pe mease ( cyC), a hypo he ical lysine pe mease (lysP2), glu ama e syn hase (gl B, gl D) and homose ine dehyd ogenase (dhoM) we e mo e abundan in S. au eus du ing in ec ion o C57BL/6 mice han du ing in ec ion o A/J mice (Table 2). Because de no o amino acid syn hesis pa hways and amino acids anspo sys ems a e unde con ol o ca aboli e ep ession in he p esence o p e e ed ca bon sou ces29 in S. au eus, he inc eased exp ession o he genes encoding hese pa hways du ing in ec ion o C57BL/6 mice is consis en wi h mo e se e e nu ien limi a ion in he issue mic oen i onmen o C57BL/6 han in he issue o A/J mice. This nu ien limi a ion leads o de- ep ession o genes ha enable S. au eus o use seconda y ca bon sou ces such as amino acids. Fu he mo e, genes encoding exop o eases such as s aphopain (sspB2), se ine p o eases (sspA, splA,splE and splF), s aphos a in B (sspC), au eolysin (au ) and a cys eine p o ease (sspB) we e also exp essed by S. au eus o a g ea e ex en in C57BL/6 han in A/J mice (Fig. 4a). These p o eases can gene a e pep ides in he ex acellula en i onmen ha can be impo ed by he bac e ium ia specialized anspo sys ems and used o e ie e amino acids30. Besides hei ole in me abolism, ex a- cellula p o eases a e po en i ulence ac o s ha help S. au eus o e ade he hos immune de enses31,32. O he genes encoding impo an i ulence ac o s such as he immunodominan s aphylococcal an igen B (isaB), he ex acellula fib inogen binding p o ein E b (e b), which is in ol ed in inhibi ion o phagocy osis33, he po e- o ming cy olysin alpha- oxin (hla), and he amphipa hic a-helical phenol-soluble modulins psma1, psma2 and psma3, which can kill hos cells by damaging he plasma memb ane34, we e also exp essed by S. au eus o a highe ex en du ing in ec ion o C57BL/6 mice (Fig. 4b). These findings indica e ha S. au eus exp ess g ea e le els o i ulence ac o s du ing in ec ion o C57BL/6 han du ing in ec ion o A/J mice, which is mos p obably d i en by he di e en g ow h phase o he bac e ia in he wo mouse s ains. The exp ession o hese i ulence ac o s is con olled by egula o y elemen s such as wo-componen egula o y sys ems (TCRS) and ansc ip ional egula o y sys ems in esponse o en i onmen al cues encoun e ed by he bac e ium du ing in ec ion35. The s aphylococcal quo um-sensing sys em accesso y gene egula o (ag ) was ound highly exp essed by S. au eus du ing in ec ion o bo h A/J and C57BL/6 mice. The ag sys em comp ises wo di e gen ansc ip s, RNAII and RNAIII, which a e unde he con ol o wo dis inc p omo e s, P2 and P3, espec i ely36. RNA II encodes he quo um-sensing elemen s Ag B, Ag D, Ag C and Ag A ha ep esen an au oca aly ic senso y ansduc ion sys em. RNAIII encodes del a-hemolysin (hld)37 and is a majo egula o o i ulence ac o s in S. au eus, inducing he ansc ip ion o a ious ex acellula p o eases and oxins38. Al hough he genes encoding he ag P2 ope on SAO in A/J mice 500 400 300 200 100 0 500 400 300 200 100 0 a cA a cB a cD a cC a cR a cA a cB a cD a cC a cR TPM TPM 500 400 300 200 100 0 TPM 500 400 300 200 100 0 TPM 600 800 1,000 * * 400 200 0 TPM SAO in C57BL/6 mice SAO in A/J 20 65 594 SAO in C57BL/6 a b Figu e 3 | Gene exp ession analysis o S. au eus du ing in ec ion o A/J and C57BL/6 mice. (a) Venn diag am showing he numbe o unique and common exp essed genes be ween S. au eus in ec ing A/J and C57BL/6 mice based on di e en ial gene exp ession analysis de e mined wi h NOISeq. (b) Gene composi ion and o ganiza ion o he genes o he ADI ope on and hei le el o exp ession in S. au eus du ing in ec ion o A/J (blue ba s) o C57BL/6 ( ed ba s) mice. Each ba ep esen s he mean o TPM±s.d. o iplica es. NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 ARTICLE NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions 5 (ag BDCA) we e exp essed by S. au eus o a simila le el in bo h mouse s ains, he exp ession le el o RNAIII/hld was g ea e du ing in ec ion o C57BL/6 mice (Fig. 4c, Supplemen a y Da a 4). The gene encoding he ansc ip ional egula o Sa R was also up egula ed by S. au eus in ec ing C57BL/6 mice (Fig. 4c, Supplemen a y Da a 4). This could explain he highe le el o ansc ip s encoding p o eases and oxins de ec ed in S. au eus du ing in ec ion o C57BL/6 mice, since bo h RNAIII ( e . 38) and Sa R39 ac i a e hei ansc ip ion. An imic obial pep ides (AMPs) a e an impo an pa o he hos inna e immune de ense agains S. au eus by di ec ly impai ing he in eg i y o he bac e ial cell wall40. The bac e ial gene encoding phospha idylglyce ol lysyl ans e ase (mp F), which is pa o he cell wall s ess s imulon and media es Table 2 | DEGs be ween S. au eus in ec ing A/J and C57BL/6 mice wi h g ea e ansc ip abundance du ing in ec ion o C57BL/6 mice Locus ag Gene symbol Desc ip ion SAOUHSC_02260 hld Del a-hemolysin SAOUHSC_00411.1 psma1 Alpha phenol soluble modulin SAOUHSC_02566 sa R Hypo he ical SAOUHSC_02971 au Zinc me allop o einase au eolysin SAOUHSC_00435 gl B Glu ama e syn hase la ge subuni SAOUHSC_01788 h S Th eonyl- RNA syn he ase SAOUHSC_00987 sspB Cys eine p o ease SAOUHSC_00248 ly M Pep idoglycan hyd olase SAOUHSC_02571 ssaA Sec e o y an igen SAOUHSC_00427 sle1 Au olysin SAOUHSC_02941 n dG Hypo he ical SAOUHSC_01001 qoxA Quinol oxidase subuni I SAOUHSC_00964 Hypo he ical SAOUHSC_00401 Hypo he ical SAOUHSC_00717 saeP Hypo he ical SAOUHSC_00741 n dI Ribonucleo ide educ ase s imula o y p o ein SAOUHSC_01942 splA Se ine p o ease SplA SAOUHSC_00083 sbnI Hypo he ical SAOUHSC_00348 psF 30S ibosomal p o ein S6 SAOUHSC_00436 gl D Glu ama e syn hase subuni be a SAOUHSC_00051 plc 1-phospha idylinosi ol phosphodies e ase SAOUHSC_00411.2 psma2 Alpha phenol soluble modulin SAOUHSC_01121 hla Alpha-hemolysin SAOUHSC_00272 Hypo he ical SAOUHSC_00801 secG P ep o ein anslocase subuni SecG SAOUHSC_01935 splF Se ine p o ease SplF SAOUHSC_02369 poE DNA-di ec ed RNA polyme ase subuni del a SAOUHSC_00268 Hypo he ical SAOUHSC_01110 e b Fib inogen-binding p o ein-like p o ein SAOUHSC_01688 lepA GTP-binding p o ein LepA SAOUHSC_02855 amiD2 LysM domain-con aining p o ein SAOUHSC_02762 Hypo he ical SAOUHSC_02114 dagK Pu a i e lipid kinase SAOUHSC_02372 Hypo he ical SAOUHSC_02430 h sA ABC anspo e pe iplasmic binding p o ein SAOUHSC_01320 dhoM Homose ine dehyd ogenase SAOUHSC_00986 sspC Cys eine p o ease SAOUHSC_00411.3 psma3 Alpha phenol soluble modulin SAOUHSC_02112 Hypo he ical SAOUHSC_02972 isaB Immunodominan an igen B SAOUHSC_02885 Hypo he ical SAOUHSC_01326 lysP2 Hypo he ical SAOUHSC_02127 sspB2 S aphopain hiol p o einase SAOUHSC_01936 splE Se ine p o ease SplE SAOUHSC_00728 l aS Hypo he ical SAOUHSC_00625 mnhA Pu a i e mono alen ca ion/H þan ipo e subuni A SAOUHSC_02763 opp-1F Pep ide ABC anspo e ATP-binding p o ein SAOUHSC_00988 sspA Glu amyl endopep idase SAOUHSC_00711 Hypo he ical SAOUHSC_00561 aX Hypo he ical SAOUHSC_02550 FdhD Fo ma e dehyd ogenase accesso y p o ein SAOUHSC_00875 ndh2 Hypo he ical SAOUHSC_01359 mp F Hypo he ical SAOUHSC_01192 A Hypo he ical SAOUHSC_02887 isaA Immunodominan an igen A SAOUHSC_02254 g oEL chape onin G oEL SAOUHSC_02485 poA DNA-di ec ed RNA polyme ase subuni alpha SAOUHSC_01462 gpsB Hypo he ical SAOUHSC_00367 cyP Hypo he ical SAOUHSC_01062 Hypo he ical SAOUHSC_00893 namA FMN oxido educ ase SAOUHSC_00144 ausA Hypo he ical SAOUHSC_00652 huA I on compound ABC anspo e ATP-binding p o ein SAOUHSC_02883 ssaA LysM domain-con aining p o ein SAOUHSC_01431 ms B Me hionine sul oxide educ ase B DEG, di e en ially exp essed gene. The comple e da a o he di e en ially exp essed genes wi h highe exp ession by S. au eus du ing in ec ion o he esis an C57BL/6 mice a e displayed in Supplemen a y Table 5. ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 6NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions esis ance o ca ionic AMPs, by educing he nega i e cha ge o he memb ane su ace41, was exp essed by S. au eus o a la ge ex en du ing in ec ion o esis an C57BL/6 han du ing in ec ion o A/J mice. Simila ly, o he genes o he cell wall s ess s imulon including aX, au olysin encoding genes (ssaA, ly M, amiD2 and isaA), genes in ol ed in cell sepa a ion (gpsB and sle1), g oEL encoding a chape onin, and he lipo eichoic acid syn hase encoding gene l aS exhibi ed g ea e exp ession le els in S. au eus du ing in ec ion o C57BL/6 mice han du ing in ec ion o A/J mice. The highe induc ion o hese genes migh eflec an adap i e esponse o S. au eus o keep cell wall in eg i y when con on ed wi h high le els o cell wall damaging agen s in he issue o C57BL/6 mice. No ably, genes ela ed o cen al me abolism, i on acquisi ion, gene al and oxida i e s ess esponses we e exp essed by S. au eus o a simila ex en du ing in ec ion o A/J and C57BL/6 mice (Supplemen a y Da a 4), indica ing ha hei exp ession was no influenced by he le els o hos esis ance o in ec ion. The RNA-seq esul s we e alida ed in a selec ed se o genes om hos and pa hogen by quan i a i e e e se ansc ip ion– PCR (RT–PCR; Supplemen a y Fig. 3) and a he p o ein le el by ELISA (Supplemen a y Fig. 4). E ec o dec eased hos esis ance on S. au eus ansc ip ion. Nex , we de e mined i dep essing specific mechanisms o hos esis ance in C57BL/6 mice a ec ed he ansc ip ional esponse o S. au eus du ing in ec ion. Fi s , we hypo hesized ha apoli- pop o ein B (ApoB), he majo s uc u al p o ein o e y low- densi y (VLDL) and low-densi y lipop o eins (LDL), con ibu es o he esis ance o C57BL/6 mice o S. au eus in ou in ec ion model, because he gene encoding Apo B (Apob) was exp essed o a g ea e ex en in he kidneys o C57BL/6 han in he kidneys o A/J mice a bo h gene exp ession (Fig. 5a) and p o ein le el (Fig. 5b), be o e and a e in ec ion (Fig. 5c). Fu he mo e, ApoB has been shown o con ibu e o he hos de ense agains S. au eus in expe imen al models o skin42 and espi a o y43 in ec ion by an agonizing he ag quo um sensing sys em o S. au eus. To alida e his hypo hesis, A/J and C57BL/6 mice we e ea ed wi h 4-aminopy azolopy imidine (4-APP), a d ug ha impai s low-densi y lipop o ein sec e ion44, and subsequen ly in ec ed in a enously wi h S. au eus. Whe eas inhibi ion o ApoB ende ed C57BL/6 mice mo e suscep ible o S. au eus, demons a ed by he significan ly highe bac e ial loads in kidneys a 48 h in ec ion (P- alueo0.05, - es ), (Fig. 5d), i did no a ec he le el o suscep ibili y o A/J mice (Fig. 5d). Nex , we in es iga ed i he educed esis ance in C57BL/6 mice a e a enua ion o ApoB sec e ion impac ed he ansc ip ional esponse o S. au eus du ing in ec ion. Fo his pu pose, we compa ed he exp ession o a se o i ulence- ela ed genes in S. au eus in ec ing 4-APP- ea ed C57BL/6 mice wi h he exp ession o he same se o genes in S. au eus in ec ing C57BL/6 mice ea ed wi h ehicle alone. The esul s show ha he le el o exp ession o all genes es ed hld/RNAIII, sa R,hla, sspA,au , aX and gl B was dec eased by 4-APP- ea men , while he exp ession o he hypo he ical alanine pe mease (aapA) was highe in S. au eus in ec ing 4-APP- ea ed han in ehicle- ea ed C57BL/6 mice (Table 3). These genes we e also ound o be di e en ially exp essed be ween S. au eus in ec ing C57BL/6 mice and S. au eus in ec ing C57BL/6 mice deficien in he exp ession o MyD88, an adap o molecule ha is essen ial o he signalling o IL-1R/TLR amily (Table 3). Since MyD88-deficien mice a e mo e suscep ible o S. au eus han wild ype C57BL/6 mice45, hese obse a ions u he demons a ed he ema kable influence o he le el o hos esis ance on he ansc ip ional esponse o S. au eus du ing in ec ion. Ta ge exp ession a ec s e ficacy o an i- i ulence app oaches. A e ha ing demons a ed he influence o he le els o hos esis ance on bo h he quali y and quan i y o S. au eus an- sc ip ional esponse, we sough o de e mine he consequences o his dependence on he e ec i eness o an i- i ulence s a egies. Fo his pu pose, we assessed he e ec o neu alizing a i ulence ac o ha di e ed in exp ession be ween S. au eus in ec ing A/J and C57BL/6 mice on he bac e ia fi ness du ing in ec ion. We chose au eolysin, which has been shown o be impo an o ull i ulence o S. au eus in expe imen al in ec ion models46,47 and was exp essed o a significan ly g ea e ex en by S. au eus du ing in ec ion o C57BL/6 han du ing in ec ion o A/J mice in ou s udy (FDRo0.05, NOISeq analysis). C57BL/6 and A/J mice we e simul aneously challenged wi h wild ype and an au eolysin- deficien s ain o S. au eus (wild ype, Dau ) and he amoun o each bac e ial s ain was de e mined in he kidneys o in ec ed mice a 48h o in ec ion. The o e all amoun o S. au eus bac e ia coun ing bo h wild ype and Dau s ains was significan ly g ea e (P¼0.0138, - es ) in he kidneys o A/J (1.3 108±4.3 107) han in he kidneys o C57BL/6 mice (1.03 107±5.4 106). Mo eo e , while a lowe amoun o Dau han wild ype S. au eus was eco e ed om C57BL/6 mice, he amoun o Dau eco e ed om A/J mice was compa able o ha o wild ype s ain (Fig. 6a). Thus, he Dau had a compe i i e disad an age when co- adminis e ed wi h wild ype S. au eus in C57BL/6 mice (mean compe i i e index o Dau /wild ype S. au eus o 0.013), while Dau and wild ype S. au eus we e equally compe i i e a e co-adminis a ion in A/J mice (mean compe i i e index o Dau /wild ype S. au eus o 1.595) (Fig. 6b). Taken oge he , hese esul s indica e ha he e ficacy o a ge ing a i ulence ac o by an i- i ulence s a egies will s ongly depend on i s le el o exp ession by he pa hogen du ing in ec ion, which in u n is highly influenced by he in insic le els o hos esis ance o in ec ion. Discussion An i- i ulence s a egies based on a enua ion o bac e ial pa hogenesis by he specific inhibi ion o i ulence ac o s essen ial o he pa hogen’s su i al du ing in ec ion48, ha e ecei ed inc easing a en ion as no el ea men op ions o in ec ions caused by an ibio ic- esis an pa hogens48. The concep o an i- i ulence he apy is s ill e y much in i s in ancy and he e o e mo e esea ch is needed o explo e i s p ac icabili y. One impo an aspec ha should be conside ed ca e ully when designing an i- i ulence s a egies is ha he exp ession o i ulence ai s by he pa hogens is no cons i u i e bu a he influenced by he specific en i onmen encoun e ed du ing in ec ion. Consequen ly, absen exp ession o he a ge ed i ulence ac o s could ende an i- i ulence s a egies comple ely ine ec i e. The e o e, i is essen ial o unde s and he impac o he wide- anging, in e -indi idual a ia ion o he hos esponse on he pa hogen’s exp ession o i ulence de e minan s du ing in ec ion. Ou s udy suppo s he idea ha he hos gene ic backg ound a ec s he ansc ip ional esponse o S. au eus du ing in ec ion. The limi ed capabili y o he immune de ense mechanisms o suscep ible A/J mice o con ol S. au eus g ow h led o he de elopmen o an in ense inflamma o y esponse, appa en by he disp opo iona e exp ession o inflamma o y cy okines and damage-associa ed molecula pa e ns. T ansc ip ional ep o- g amming o he euka yo ic cells, possibly esul ing om he concomi an accumula ion o acidic p oduc s and he lowe ed oxygen ension (hypoxia) in he in ec ed issue, in ol ed he induc ion o Hi 1a ha encodes he cen al media o o ansc ip ional esponses o hypoxia HIF-1a49,50. The e o e, he NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 ARTICLE NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions 7 majo challenge aced by S. au eus in suscep ible A/J mice seems o be he adap a ion o he ad e se condi ions imposed by he hype inflamma o y esponse and hypoxic mic oen i onmen s. To su i e in he sep ic A/J mice, S. au eus inc eased exp ession o he ADI ope on (a cABDCR), which is gene ally induced unde anae obic condi ions27 and is impo an o ene gy gene a ion, bu also p o ec s S. au eus agains acidic s ess28. The supe io esis ance mechanisms o C57BL/6 mice agains S. au eus enabled a be e con ol o bac e ial mul iplica ion. The e o e, S. au eus aced he main challenge o coun e ac ing he powe ul hos de ense mechanisms o he esis an C57BL/6 mice. The ansc ip ion da a indica ed ha S. au eus esponded o he ad e se en i onmen encoun e ed wi hin C57BL/6 mice by inc easing he exp ession o cy o oxins and ex acellula p o eases. Cy o oxins such as alpha-hemolysin (hla) and PSMs (psma1-3 and hld) help S. au eus o a oid phagocy ic killing by inducing po es in he memb ane o hos cells, leading o cell dea h51 and can p omo e bac e ial sp eading by dis up ing he 2,000 20,000 10,000 8,000 6,000 4,000 2,000 SAO in A/J mice SAO in C57BL/6 mice 0 40,000 1,000 3,000 Sa R –+ + Sa R * 2,500 2,000 1,500 1,000 500 0 1,000 2,000 3,000 4,000 5,000 6,000 0 800 700 600 * * * * * * * ** * * TPM TPM TPM TPM TPM TPM 500 400 300 200 100 0 sspB2 sspB sspA sspC AIP Ag B Ag D Ag A Ag C P P P3 P2 + RNAIII RNAII ag A ag A ag C au 600 0 25,000 50,000 75,000 100,000 125,000 150,000 175,000 500 400 300 200 * * 100 0au ag C ag B ag B ag D ag D hld hld hld AIP O SM I F D CT S Y O SM I F D CT S Y spIA spIE spIF psma1 psma2 psma3 hla ab c Figu e 4 | Exp ession o selec ed genes by S. au eus du ing in ec ion o C57BL/6 o A/J mice. (a) Exp ession le els o genes encoding p o eases. (b) Exp ession le els o genes encoding oxins. (c) Schema ic ep esen a ion and le el o exp ession o genes encoding he global egula o Ag and Sa R. Red ba s show exp ession alues o he genes exp essed by S. au eus in C57BL/6 mice and blue ba s show exp ession alues o he genes exp essed by S. au eus in A/J mice. Each ba ep esen s he mean TPM±s.d. o iplica es. ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 8NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions epi helial ba ie 52. The s ong induc ion o genes encoding key exop o eases, including he me allop o einase au eolysin (au ), se ine p o eases (splA,sspA,splE and splF), s aphos a in B (sspC) and cys eine p o ease (sspB), u he accen ua es he impo ance o immune e asion o S. au eus su i al in C57BL/6 mice. These p o eases can clea e and deg ade componen s o he complemen sys em53 and can inhibi neu ophil chemo axis54. The exp ession o hese i ulence de e minan s is o ches a ed h ough ans- c ip ional and pos - ansc ip ional egula ion by egula o y sys ems. The exp ession o RNAIII, a majo egula o o hese ac o s38, was highe exp essed by S. au eus du ing in ec ion o C57BL/6 mice han in A/J mice, while he exp ession o he au oca aly ic senso y ansduc ion sys em did no di e be ween bo h mouse s ains. This appa en disc epancy could, howe e , be explained by g ea e ansc ip abundance o Sa R in S. au eus in ec ing C57BL/6 mice, which mi iga es he exp ession o he ag P2 ope on, while ha ing no appa en e ec on he exp ession o RNAIII ( e . 55). Mo eo e , Sa R enhances he exp ession o genes encoding se e al ex acellula p o eases egula ed by RNAIII ( e . 39) and, he e o e, may ac in syne gy wi h RNAIII o boos he exp ession o i ulence ac o s, equi ed o su i al unde he s ong immune p essu e in C57BL/6 mice. The s ong immune p essu e in he esis an mice could also explain he g ea e exp ession o se e al genes o he cell wall s ess s imulon by S. au eus in ec ing C57BL/6 mice. I has been shown ha he magni ude o cell wall s imulon induc ion s ongly 600 Unin ec ed A/J C57BL/6 A/J C57BL/6 *** 10 8 6 4 2 Vehicle Vehicle 4-APP 4-APP *** ** Fold change in Apob exp ession (C57BL/6 ela i e o A/J) Fold change in Apob exp ession (C57BL/6 ela i e o A/J) CFU S.au eus in kidneys (Log10) 400 200 0 RNA-Seq RNA-Seq 25 20 15 10 5 0 S. au eus RT-PCR RT-PCR ab cd Figu e 5 | Apolipop o ein B (ApoB) con ibu es o esis ance agains S. au eus bloods eam in ec ion in C57BL/6 mice. (a) Rela i e old change o Apob exp ession alues in he kidneys o unin ec ed C57BL/6 espec o hose in he kidneys o unin ec ed A/J mice de e mined by ei he RNA-Seq (whi e ba s) o qRT–PCR (black ba s). Each ba ep esen s he mean ela i e old change±s.d. o iplica es. (b) Exp ession o ApoB in kidneys o A/J (le ) and C57BL/ 6 ( igh ) mice de e mined by immunos aining o kidney issue using specific an ibodies agains ApoB. Magnifica ion X40. (c) Rela i e old change o Apob exp ession alues in he kidneys o S. au eus-in ec ed C57BL/6 in compa ison o hose in he kidneys o S. au eus-in ec ed A/J mice a 48 h o in ec ion de e mined by ei he RNA-Seq (whi e ba s) o qRT-PCR (black ba s). Each ba ep esen s he mean ela i e old change±s.d. o iplica es. (d) Bac e ial loads in he kidneys o A/J (blue symbols) and C57BL/6 ( ed symbols) mice ea ed wi h 4-Aminopy azolo[3,4-d]py imidine (4-APP) (open symbols) o wi h ehicle alone (solid symbols) a 48 a e in a enous inocula ion wi h 2 107CFU o S. au eus s ain SH1000. Each symbol ep esen s he bac e ial coun s de e mined in an indi idual mouse and he ho izon al lines ep esen he a e age±s.d. o each mouse s ain (n¼6, - es , **Po0.01, ***Po0.001). Table 3 | S. au eus gene exp ession du ing in ec ion o 4-APP- ea ed and MyD88-deficien C57BL/6 mice. Locus ag Gene symbol Desc ip ion Rela i e old change in 4-APP- ea ed o ehicle- ea ed C57BL/6 mice (mean (s.e.)) Rela i e old change in MyD88- deficien o wild ype C57BL/6 mice (mean (s.e.)) SAOUHSC_02265 ag A Accesso y gene egula o p o ein A 1.60 (0.22) 1.62 (0.5) SAOUHSC_02566 sa R Accesso y egula o R 1.83 (0.22) 1.23 (0.09) SAOUHSC_01121 hla Alpha-hemolysin 3.32 (1.56) 18.06 (0.16) SAOUHSC_02971 au Zinc me allop o einase au eolysin 2.31 (0.96) 2.59 (0.75) SAOUHSC_00988 sspA Glu amyl endopep idase 2.19 (0.01) 2.32 (0.46) SAOUHSC_00561 aX P o ein V aX 3.91 (2.76) 1.11 (0.04) SAOUHSC_00436 gl D Glu ama e syn hase subuni be a 2.41 (1.11) 6.03 (1.66) SAOUHSC_01803 aapA D-se ine/D-alanine/glycine anspo e 1.21 (2.32) 1.69 (0.29) RNAIII Regula o y RNA 1.80 (0.53) 2.52 (0.08) Le el o exp ession o a subse o genes by S. au eus du ing in ec ion o 4-APP- ea ed compa ed wi h ehicle- ea ed C57BL/6, espec i ely MyD88-deficien compa ed wi h wild- ype C57BL/6 mice de e mined by quan i a i e e e se ansc ip ion–PCR (qRT–PCR). NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14268 ARTICLE NATURE COMMUNICATIONS | 8:14268 | DOI: 10.1038/ncomms14268 | www.na u e.com/na u ecommunica ions 9