AMPK/HuR-Driven IL-20 Post-Transcriptional Regulation in Psoriatic Skin

Transcription

1 ORIGINAL ARTICLE AMPK/HuR-Driven IL- Post-Trnsriptionl Regultion in Psoriti Skin Geneviève Grin, Iselle Guirud, Mtthieu Lroix, Clémene Genthon, Stéphnie Rille, Jen-Mrie Joujoux 5, Lurent Meunier, Thierry Lvre-Bertrnd, Pierre-Emmnuel Stoener,7 nd Lionel Le Glli,7 IL- is involved in the development of skin psorisis. The moleulr mehnisms underlying IL- overexpression in psoriti epidermis remin to e eluidted. We showed tht IL- ws primrily upregulted in psoriti skin t the post-trnsriptionl level. The RNA-inding protein HuR relolized to the ytoplsm of kertinoytes (KCs) of psoriti ptients, suggesting tht it stilizes numerous trnsripts, s oserved in the humn KC ell lines used to ssess IL- mrna. We hrterized epiderml HuR RNA trgets in psoriti skin using rionuleoprotein immunopreipittion nlyzed vi high-throughput sequening. Numerous trnsripts tht re upregulted in psorisis were trgeted y HuR, supporting the prtiiption of HuR in pthogeni proesses suh s morphologil hnges, innte nd dptive immune responses, nd metoli inflmmtory responses. Finlly, we identified the metoli sensor AMP-tivted protein kinse (AMPK) s eing responsile for HuR ytoplsmi reloliztion euse its tivity ws severely impired in humn psoriti epidermis, nd in vivo drug-medited AMPK inhiition in mouse epidermis promoted HuR ytoplsmi loliztion, IL- overprodution, nthosis, nd hyperkertosis. These results provide insights into the moleulr links etween metolism nd post-trnsriptionl networks during hroni inflmmtion. Journl of Investigtive Dermtology (5) 5, 77; doi:.8/jid.5.8; pulished online August 5 INTRODUCTION The IL- sufmily of ytokines (IL-9, IL-, IL-, IL-, nd IL-) (Rutz et l., ) hs n importnt role in the pthogenesis of psorisis euse of their pro-inflmmtory proprieties nd overexpression in this disese (Rutz et l., ). Indeed, IL-9 nd IL- nd, more mrkedly, IL- nd IL- hve een shown to mimi ertin lssil psoriti Dynmique des Intertions Memrnires Normles et Pthologiques (DIMNP), CNRS UMR 55, Université de Montpellier, Montpellier, Frne; Lortoire d Histologie-Emryologie-Cytogénétique, Institut des Biomolećules Mx Mousseron (IBMM), CNRS UMR 57, Fulté de Médeine Montpellier- Nîmes, Université de Montpellier, Nîmes, Frne; Institut de Reherhe en Cnérologie de Montpellier (IRCM), Institut régionl du Cner Montpellier (ICM), INSERM U89, Université de Montpellier, Montpellier, Frne; Montpellier Genomix (MGX), Institut de Génomique Fontionnelle (IGF), CNRS INSERM UMS, Université de Montpellier, Montpellier, Frne; 5 Servie d Antomopthologie, Centre Hospitlier Universitire de Nîmes, Nîmes, Frne nd Servie de Dermtologie, Centre Hospitlier Universitire de Nîmes, Nîmes, Frne Correspondene: Pierre-Emmnuel Stoener, Servie de Dermtologie, Centre Hospitlier Universitire de Nîmes, Nîmes 9, Frne or Lionel Le Glli, Shool of Mediine, University of Montpellier, NIMES Cedex 98, Frne. E-mil: pierre.stoener@hu-nimes.fr or lionel.le-glli@univ-montp.fr 7 The lst two uthors ontriuted eqully to this work. Arevitions: UTR, untrnslted region; AtD, tinomyin D; AMPK, AMP-tivted protein kinse; CompC, ompound C; IHC, immunohistohemistry; IP, immunopreipittion; KC, kertinoyte;, lesionl skin; N, non-lesionl skin; RIP-Seq, rionuleoprotein immunopreipittion nlyzed y high-throughput sequening Reeived Jnury 5; revised 8 My 5; epted June 5; epted rtile preview online 5 July 5; pulished online August 5 fetures in reonstituted humn epidermis models (S et l., 7), nd mouse models overexpressing IL-, IL-, or IL- exhiit psorisis-like pperne (Blumerg et l., ; He nd Ling ). The primry soure of IL- nd IL- is restrited to the lymphoid linege, ertin speifi tivted T-ell supopultions (T helper type, T helper type, nd T helper type 7), nd nturl killer ell susets, wheres oth immune ells nd tissue ells serete IL-9, IL-, nd IL- (St, ). IL- hs n importnt role in the lter effetor phse of psorisis pthogenesis, during whih it inhiits the terminl differentition of kertinoytes (KCs) nd ontriutes to ntimiroil ompetene nd epiderml infiltrtion y neutrophili grnuloytes (St nd Wolk, ). Ptients with psorisis exhiit n inresed level of IL- in lesionl skin () nd in the lood, where it orreltes with the Psorisis Are Severity Index sore (St nd Wolk, ). IL- inds to reeptor heterodimers tht re found on KCs ut not on immune ells. Despite their entrl role in psorisis, the moleulr mehnisms tht led to the overexpression of the IL- sufmily of ytokines hve een rrely investigted. In the present study, we show tht IL- is primrily upregulted in psoriti skin t the post-trnsriptionl level vi mrna stiliztion. We found tht the RNA-inding protein HuR (Srikntn nd Gorospe, ) inds to nd regultes IL- mrna nd relolizes to the ytosol of psoriti KCs. We further identify dditionl HuR trgets in psoriti skin using rionuleoprotein immunopreipittion nlyzed y high-throughput sequening (RIP-Seq). 7 Journl of Investigtive Dermtology (5), Volume 5 5 The Soiety for Investigtive Dermtology

2 Consistent with omoridity (Armstrong et l., ) nd enzymti (Hlprin nd Ohkwr, 9) studies, we demonstrte metoli ltertions in psorisis sed on the pprent derese in the tivity of the metoli sensor AMP-tivted protein kinse (AMPK) (O'Neill nd Hrdie, ) in psoriti KCs. Finlly, speifi AMPK inhiitor uses HuR ytosoli loliztion nd IL- mrna stiliztion in humn KCs. Topil pplition of the AMPK inhiitor on mouse skin uses HuR ytosoli loliztion nd IL- overprodution nd initites ertin histologil fetures of psorisis. These findings suggest tht metoli disorders ffeting AMPK tivtion in KCs re importnt in the pthophysiology of psorisis, permitting HuR-medited stiliztion of numerous mrnas, inluding IL- mrna. RESULTS IL- mrna is stilized y ellulr stresses To study IL- indution in ellulr model, we treted the HCT KC ell line with different ellulr stresses, suh s nisomyin (protein synthesis inhiition nd p8 MAP kinse tivtion) or UVB rdition (ell dmge stress) with or without tinomyin D (trnsriptionl inhiition). IL- mrna levels nd IL- gene trnsription were evluted using quntititive PCR primers tht disriminte IL- mrna nd IL- preursor mrna (pre-mrna). IL- exhiited strong mrna indution in response to nisomyin nd UVB exposure (Figures nd, respetively). Under identil onditions, IL- pre-mrna did not inrese, exept in response to mj m UVB nd 8 hours fter nisomyin tretment (Figures d nd, respetively), t whih dose nd time the inrese ws limited nd not signifint. This modest vrition in the trnsriptionl rte ould suggest tht IL- mrna stility is limiting step for IL- mrna umultion. The kinetis of mrna stility following the inhiition of trnsription with tinomyin D indited tht IL-9, IL-, nd, to greter degree, IL- were unstle in non-stressed ells ut were effiiently stilized following UVB irrdition (Figure e). Colletively, these results indited tht (i) IL- mrna ws extremely unstle in non-stressed ells, nd (ii) its indution in response to ellulr stresses ws predominntly medited y its stiliztion. Conversely, IL-9 nd IL- mrna indution ws ssoited with oth trnsriptionl nd mrna stiliztion mehnisms. Consistent with these results, post-trnsriptionl regultory elements present within the untrnslted region ( UTR) of the IL- ytokine sufmily were more numerous nd relevnt in IL- (nd IL-) thn in IL-9 nd IL- (nd IL-; Figure f). IL- mrna is stilized in psoriti skin We used immunohistohemistry (IHC) to onfirm tht KCs re the primry soure of IL- in psoriti lesions (Figure ). DEFBA, mrker of the severity of psorisis (Jnsen et l., 9), ws inresed ( -fold) t the mrna level in ompred with non-lesionl skin (N) in the entire ptient ohort (n = 7; Figure ). IL- mrna ws overexpressed in in ll tested ptients, nd IL- nd IL- mrna were inresed in of 7 ptients, IL- in 5 of 7 ptients, nd IL-9 in of 7 ptients (Figure ). These results re onsistent with previous studies (St nd Wolk, ). The IL- pre-mrna level ws not signifintly inresed in versus N nd does not ount for the upregultion of IL- mrna in psorisis. This result suggested tht the upregultion in ourred t the mrna level vi post-trnsriptionl mehnism(s) (Figure d). Similr results were otined with IL- pre-mrnas. Beuse of ptient vritions, IL- premrna ws not signifintly upregulted. In ontrst, the levels of IL-9 nd IL- pre-mrnas were signifintly inresed in ompred with N, suggesting tht the inresed gene trnsription ws mjor ontriutor to the indution of mrna in. We next estlished the lk of orreltion etween the mrna nd the pre-mrna levels in N for eh of the five ytokines (Figure e). In ontrst, the mrna levels in ppered to signifintly orrelte with the pre-mrna levels for IL-, IL-9, IL-, nd IL-. Indeed, the mehnisms unoupling the mrna levels from the pre-mrna levels in N were not present in. Beuse IL- mrna ws regulted y the mrna dey mhinery nd no inrese in IL- premrna (i.e., trnsriptionl upregultion) ws found in, IL- mrna my e (i) destilized in N nd (ii) stilized in. Comprtively, IL- mrna ws lso regulted vi n inrese in stility in, wheres IL-9, IL-, nd IL- mrnas were (i) destilized in N nd (ii) regulted vi n inrese in mrna trnsription nd stility in. HuR regultes IL- mrna stility nd relolizes to the ytoplsm of psoriti KCs Figure indites tht the pthwys involved in the stiliztion of messengers were tivted in psorisis; therefore, we foused on this spet. p8 MAP kinses, one of the est-hrterized mrna stiliztion pthwys through 5 -AUUUA- motifs in the mrna UTR (Winzen et l., 999), hve een desried to e involved in hroni inflmmtory diseses, inluding psorisis (Johnsen et l., ). IL- ontins numerous AUUUA motifs in its UTR, nd the inhiition of p8 MAP kinses nerly rogted the IL- response to ellulr stress (Supplementry Figures SA nd SB online). However, it ws shown tht epiderml p8α, ut not myeloid p8α, is not essentil in skin inflmmtion in mie (Kim et l., 8). Our results supported the reltive importne of p8 MAP kinses nd/or the trget MK in the inflmmtory response of KCs, s p8 nd MK tivtion, estimted y IHC, ws limited to (promet-)metphsi KCs (Supplementry Figures SCSF online) nd is thus the onsequene of high mitoti rte in the psoriti epidermis. Another mehnism tht my ount for the mrna stiliztion is medited y the RNA-inding protein HuR (Kim nd Gorospe, 8). Indeed, the IL- UTR ontins the highest ffinity HuR heptmer-inding sites (5 -UUUUUUU- ; Leedev et l., ), nd we demonstrted tht HuR ould ind to IL- mrna (Figure ). As shown in Figure, the inrese in IL- mrna indued y nisomyin ws severely impired y smll interfering RNA-medited HuR knokdown, stressing the importne of HuR in IL- mrna overexpression. HuR mrna nd protein levels were found to e similr in ompred with N (Figures nd d). However, IHC nlysis indited tht the intrellulr 7

3 distriution of HuR in KCs ws predominntly nuler in N nd predominntly ytoplsmi in epidermis (Figure e). Beuse HuR stilizes mture mrnas in the ytoplsm (Srikntn nd Gorospe, ), these results suggested tht HuR ould stilize IL- mrna nd numerous other mrnas y interting with omptile UTRs in epidermis. HuR inds numerous trnsripts involved in the pthogenesis of psorisis To identify RNA trgets of HuR in psoriti skin, we performed RIP-Seq experiments (Srikntn nd Gorospe, ) (Supplementry Dt S online). Pre-mture (i.e., non-splied) RNAs nd mture (i.e., splied) RNAs trgeted y HuR nd identified in our RIP-Seq in N nd represented. nd 7.% of the entire trnsriptome, respetively, whih is onsistent with previous studies investigting HuR trgets using phototivtle-rionuleoside-enhned rosslinking nd immunopreipittion (Leedev et l., ; Mukherjee et l., ). Moreover,.9% of the pre-mture RNA nd 5.8% of the mture RNA trgets identified in hve een desried s HuR trgets y t lest one phototivtlerionuleoside-enhned rosslinking nd immunopreipittion study (Figure ). Mture RNAs with inresed HuR inding in psoriti skin (log.5) elong to the epithelil linege, wheres mture RNAs with the lowest fold inding (log.5) elong to other lineges (Figure ). HuR tivity ws displed from the nuleus to the ytoplsm in sed on 55.5% derese in totl pre-mture RNA reds nd 8.% inrese in totl mture RNA reds ompred with N (Figure ). The intersetion etween our RIP-Seq nd expression rry studies (Suárez-Friñs et l., ) indited n overrepresenttion of upregulted (7.%) ompred with downregulted (8.%) trnsripts (Figure 5). This overrepresenttion ws more pronouned when we onsidered only trnsripts with n inresed fold inding for HuR in (log.5): 7.% ompred with the trnsripts tht were downregulted (.%; Figure 5). To investigte the relevne of our HuR RIP-Seq, we used quntittive PCR to evlute ertin genes expressed in KCs (Figure 5). All three evluted trnsripts tht were upregulted in psorisis nd present in the RIP-Seq (CXCL-, DEFBA, nd SLCA) were downregulted y t lest 5% in HCT KCs (unstimulted or nisomyin-stimulted) following smll interfering RNA-medited HuR knokdown nd ppered to e relevnt HuR trgets. The upregulted trnsripts tht filed to ind HuR in the RIP-Seq exhiited only mild redution (o5% downregultion); however, CYP7B, whih ws exluded from the RIP-Seq for minor riteri (9 insted of RIP-Seq reds, see Supplementry Mterils nd Methods online), exhiited greter thn 5% inhiition. In ontrst, two evluted genes (CCL7 nd FAH) known to e downregulted in psorisis nd not present in HuR RIP-Seq were not ffeted y HuR knokdown. Interestingly, of trnsripts of genes tht hve een implited in psorisis y genome-wide ssoition studies (Yu et l., 8) ppered to e trgeted y HuR (Figure 5d). Numerous trgets importnt for psoriti pthophysiology ound to HuR in more undntly thn in N (e.g., CXCL-, IL7RA, IVL, HERC, HLA-DPA, HLA-E, ILG, KRT7, LIPG, NAT9, PIM, PSMA, PSORSC, RPSKB, REL, RHEB, SERPINB, or SPINK5); however, the RIP-Seq nlysis ws unle to detet IL- sufmily ytokine mrna or the mrna of other ytokines, suh s IL-8, whih is trgeted y HuR (Nors et l., ). This lk of detetion ws likely ssoited with the diffiulty enountered in immunopreipitting low-level trnsripts from skin smples nd indited tht our nlysis my hve underestimted the numer of HuR-trgeted trnsripts. Next, we explored pthwys tht were modulted y HuR in y ross-mthing HuR mture RNA trgets (exhiiting log inrese.5-fold) using the Kyoto Enylopedi of Genes nd Genomes pthwy dtse (Figure 5e), nd we identified severl pthwys tht hve een implited in the pthogenesis of psorisis (see Disussion setion). AMPK intivtion leds to HuR ytoplsmi loliztion nd IL- mrna stiliztion nd initites epiderml hnges in mouse skin Using IHC, we explored whether known HuR shuttling regultors (Kim nd Gorospe, 8) were ffeted in psoriti skin. Phosphorylted Cdk nd protein kinse Cα leling were similr in nd N (dt not shown). Phosphorylted AMPK leling ws predominntly deteted in the sl lyer of N epidermis (Figure ). Conversely, phosphorylted AMPK leling ws essentilly sent in. To investigte the potentil role of derese in AMPK tivity, we exmined the influene of ompound C (CompC), speifi AMPK inhiitor (Zhou et l., ), on HuR suellulr loliztion nd IL- mrna stiliztion. As shown in Figure, HuR ws primrily nuler in untreted HCT ells, wheres exposure to CompC triggered n inrese in ytoplsmi HuR. IL- mrna dey following tinomyin D ddition ws primrily ounterted y CompC, s % of the mrna remined fter minutes (Figure ). These oservtions suggested tht AMPK intivtion in psorisis my e responsile for the reloliztion of Figure. IL- prodution is regulted t the post-trnsriptionl level in HCT ells. () IL-9, IL-, nd IL- mrna expression in response to nisomyin ( ng ml ). () UVB dosedependent mrna expression of IL-9, IL-, nd IL- (± SEM; n = ;, ontrol). () IL-9, IL-, nd IL- pre-splied mrna expression in response to nisomyin ( ng ml ). (d) IL-9, IL-, nd IL- pre-splied mrna expression in response to inresing UVB doses. d ± SEM; n = ; Po.5, Po.5, Po., nd Po.. (e) Atinomyin D (AtD; 5 μgml )-medited termintion of trnsription in growing (open squres) versus UVB (filled squres)-irrdited ( mj m hours prior to AtD) ells for IL-9, IL-, nd IL- mrnas (one representtive is shown; n = ). (f) Post-trnsriptionl regultory elements in IL- fmily UTRs, suh s AU-rih elements (5 -AUUUA-, 5 -UAUUUAU-, nd 5 -WWWWAUUUAW WWW- ), HuR motifs (5 -UUUUUUU-, 5 -UUUAUUU-, nd 5 -UUUGUUU- ), nd mirornas with onserved mthes. 7 Journl of Investigtive Dermtology (5), Volume 5

4 HuR to the ytoplsm of KCs to promote the stiliztion of IL- nd other mrnas. To ddress this question, we used mouse skin model. Norml mouse epidermis shres moleulr fetures of humn epidermis: ytoplsmi phosphorylted AMPK leling, predominnt nuler HuR loliztion, nd miniml IL- levels (Figure d). The dily pplition of Reltive to Reltive to Reltive to d Reltive to e % Remining mrna f IL-9 5,,,,, IL- C 8 5 C 8 5 C 8 5 Anisomyin (hours) Anisomyin (hours) Anisomyin (hours) C 8 UVB (mj m ) UVB (mj m ) 8 C 8 UVB (mj m ) UVB (mj m ) 5 IL- C 8 UVB (mj m ) IL-9 pre-mrna IL- pre-mrna IL- pre-mrna C 8 5 C 8 5 C 8 5 Anisomyin (hours) Anisomyin (hours) Anisomyin (hours) C 8 IL C 8 IL- 5 C 8 UVB (mj m ) Min. fter At.D Min. fter At.D Min. fter At.D ARESore AUUUA UAUUUAU WWWWAUUUAWWWW UUU (U/A/G) UUU mirornas IL- C UVB IL-9 IL- IL / / / / / / mir-87-p; mir-, mir-58p; mir--p; mir-7 mir--5p; mir-88-p; mir-; mir-9 IL-.9 / / mir- IL-. / / 75

5 N Reltive to 5 5 N DEFBA Reltive to 5 5 N IL-9 IL- IL- IL- IL- 8,,5 5, 5 5 N N N N d Reltive to e IL-9 pre-mrna IL- pre-mrna IL- pre-mrna IL- pre-mrna IL- pre-mrna N r s =. P=.9,5 8 8, 5 N N N N IL-9 IL- r s =.8 r s =. r s =.7 P<. P=.85 P=. IL- r s =.9 P=. IL- r s =. P=. r s =.5 P=.85 IL- r s =. P =.8 5 N N N N N Figure. IL- overexpression in psorisis is onsequene of mrna stiliztion. () Immunohistohemil (IHC) stining of IL- in non-lesionl skin (N) nd lesionl skin (; one ptient, representtive of five). Sle rs = μm for originl mgnifition of nd 5 μm for originl mgnifition of. () DEFBA mrna expression in punh iopsies (± SEM; n = 7; Po.) for N (open irles) nd (filled irles). () Identil results for IL-9, IL-, IL-, IL-, nd IL- mrna expression (Po., Po.). (d) IL-9, IL-, IL-, IL-, nd IL- pre-mrna expression (Po. for N (open irles) or (filled irles)). (e) Spermn s rnk orreltion nlysis nd log plots etween pre-mrna (siss) nd mrna expression (ordinte) levels of IL-9, IL-, IL-, IL-, nd IL- in N nd (Po.5, Po.5, nd Po.). CompC on shved mouse skin (n = ; Figure d) indued histologilly fol epiderml nthosis nd hyperkertosis tht were less widespred nd pronouned thn those oserved following SDS pplition (Kim et l., 8). Moreover, CompC pplition led to derese in phosphorylted AMPK leling, HuR ytoplsmi reloliztion, nd n inrese in IL- leling in mouse epidermis. SDS-treted mie lso exhiited epiderml reloliztion of HuR nd n inrese in IL- leling ut to lesser extent. Colletively, these results suggest tht AMPK intivtion in KCs promotes the ytoplsmi reloliztion of HuR, stiliztion of IL- mrna, nd overprodution of IL-, whih in turn indues epiderml hnges. DISCUSSION Chnges in mrna undne hve dominnt role in determining the protein levels (Jovnovi et l., 5), wheres RNA proessing degrdtion ffets few genes, leit with importnt funtions (Rni et l., ). Indeed, the role of trnsriptionl regultion in psorisis hs een well doumented, e.g., the inhiition of NF-κB in IL- indution (Kumri et l., ), nd it is supported in the present study y the ehvior of IL-9, IL-, nd IL-. Here, we foused on the post-trnsriptome euse it is implited in the expression of ll IL- fmily memers nd euse overexpression of IL- ppers to e primrily onsequene of post-trnsriptionl regultion. We provide ompelling evidene impliting HuR in the stiliztion of mrnas in psoriti skin. IL- mrna is trgeted nd regulted y HuR in humn KCs, nd IHC nd RIP-Seq nlyses support the trnslotion of HuR from the nuleus to the ytoplsm in lesionl epidermis. Nuler HuR n intert with pre-mture RNAs nd ffet their proessing (Leedev et l., ; Mukherjee et l., ), nd ytoplsmi HuR trnslotion is ritil for HuR-medited stiliztion of mrnas (Wng et l., ). HuR mrna trgets re onsidered to omprise lrge portion of the trnsriptome (pproximtely 5%) nd to e importnt for ell growth, prolifertion, deth, nd immune responses (Srikntn nd Gorospe, ). HuR trnslotion hs een suspeted in hroni inflmmtory diseses, suh s sthm (Fn et l., ), nd hs een reported in ertin ners (Srikntn nd Gorospe, ). However, to the est 7 Journl of Investigtive Dermtology (5), Volume 5

6 -(d/dt) Fluoresene (75).5. Mok.5. IP HuR Temperture ( C) Temperture ( C) Reltive to HuR Atin si-neg N si-hur- si-hur- kd 8 8 Reltive to d HuR Atin 5,,,,, si-neg IL- si-hurs P P P P kd 8 8 e N Figure. HuR regultes IL- mrna nd relolizes in the ytoplsm of kertinoytes in psorisis. () DNA melting urve nlysis from IL- quntittive PCR following mok or HuR immunopreipittions (IPs). HCT ells were previously irrdited (red line) or not (lk line) for hours with UVB ( mj m ). () HCT ells were trnsfeted with ontrol smll interfering RNA (sirna) or independently with two HuR-speifi sirnas nd stimulted with nisomyin ( ng ml for 7 hours, short drk line) or not stimulted. Left: HuR protein expression. Right: IL- mrna level, verge of oth HuR-speifi sirnas (± SEM; n = for eh sirna; Po.5 nd Po.). () HuR mrna expression in non-lesionl skin (N; open irles) nd lesionl skin (; filled irles; ± SEM; n = ). (d) Western lot nlysis of smples from ptients PP. Left: N. Right:. (e) Immunohistohemistry for HuR in N nd (one ptient, representtive of five). Sle rs = μm for originl mgnifition of nd 5 μm for originl mgnifition of. of our knowledge, HuR trgets hve not een determined in ny humn disese, inluding psorisis. We found tht pthwys tht ould e ffeted y HuR ytoplsmi loliztion in psorisis inlude dherens juntion nd fol dhesion pthwys tht prtiipte in morphologil nd trnsdution hnges in psoriti KCs. Interestingly, innte nd dptive immune responses, whih re n importnt feture in psorisis pthophysiology, re lso trgeted y HuR. Indeed, KCs re n importnt omponent of the immune system, expressing FgRIII, whih is present in RIP-Seq, nd moleules tht re implited in internliztion (Gutowsk- Owsik nd Ogg, ). Finlly, the phosphtidylinositol -kinse/akt/mmmlin trget of the rpmyin (PIK/AKT/ mtor) pthwy is lso trgeted y HuR ytoplsmi reloliztion, s indited y the Kyoto Enylopedi of Genes nd Genomes ourrenes of inositol phosphte metolism, mtor, phosphtidylinositol, nd insulin signling pthwys. Positive feedk of HuR in the AKT pthwy, whih is involved in the pthogenesis of psorisis, (Mitr et l., ; Roller et l., ; Buerger et l., ), hs een reported (Singh et l., ). Finlly, AMPK tivity, whih is sensor of ellulr energy nd n ntgonist of the PIK/AKT/mTOR pthwy (Inoki et l., ), ppers to derese in psoriti lesionl epidermis. Moreover, CompC, n AMPK inhiitor, indues HuR ytoplsmi loliztion, stilizes IL- mrna in humn KCs, nd initites ertin epiderml psoriti histologil fetures. Altertions in rohydrte metolism hve een desried in psorisis for dedes nd re onsistent with our results. AMPK tivtion leds to the stimultion of energyproduing pthwys suh s glyolysis. When AMPK is intivted, ATP-onsuming pthwys suh s glyogenesis re promoted. Indeed, high level of gluose utiliztion without n inrese in glyolysis, leding to the use of gluose- 77

7 Pre-mture RNAs Mture RNAs Pre-mture RNA Uniprot tissue Count P-vlue Mture RNA Uniprot tissue Count P-vlue, Known HuR trgets Epithelium Brin 78,55.7E 5.E,89 known HuR trgets Epithelium Brin 9 9.9E 8.9E,5 Unknown HuR trgets Endothelil ell Mrophge 9.E.E,9 unknown HuR trgets Brin Skin 9.9E 9.E 5 Mukherjee et l., Mukherjee et l. 9,5,8 58 Leedev et l. 8 Leedev et l. Fold inding Uniprot tissue Count P-vlue Log.5,5 mture RNAs Log.5 9 mture RNAs Epithelium Skin Brin Skeletl musle E.E.8E 8.5E 5 Pre-mture RNA reds Mture RNA reds N 57,5, 5,9 98,7 % Figure. HuR inding to mture RNAs is inresed in psorisis. () Venn digrms ompring our iopsies for HuR RIP-Seq with pulished HuR PAR-CLIPs for ell lines (Leedev et l., ; Mukherjee et l., ). Annottion of pre-mture RNAs elonging to the UCSC introni regions nd mture RNAs elonging to the University of Cliforni Snt Cruz (UCSC) trnsript. () UniProt tissue nlysis of mture RNAs with inresed (log.5) or deresed (log -.5) HuR inding in involved versus uninvolved skin. () Totl pre-mture RNA nd mture RNA reds in the RIP-Seq for N nd., lesionl skin; N, non-lesionl skin; PAR-CLIP, phototivtle-rionuleoside-enhned rosslinking nd immunopreipittion; RIP-Seq, rionuleoprotein immunopreipittion nlyzed y high-throughput sequening. phosphte vi the pentose phosphte pthwy (Herdenstm 9; Hlprin nd Ohkwr, 9) nd to glyogen umultion (Hrmon nd Phizkerley, 98), hs een desried in the psoriti epidermis. Moreover, numerous ommon drugs used to tret psorisis re onsidered to enhne AMPK tivity, suh s vitmin D (Hdd et l., 8) or retinoi id (Lee et l., 8). Furthermore, methotrexte indues the umultion of minoimidzole roximide rionuleotide (Bekers et l., ), whih is potent AMPK tivtor. Colletively, these findings relted to metolism, moleulr intertions, nd drug therpy reinfore the importne of AMPK inhiition in psorisis. Notly, UVB is known to derese AMPK tivity nd relolize HuR to the ytosol (Zhng nd Bowden, 8), nd we demonstrted tht UVB promotes IL- indution in KCs, pprently in ontrdition with the effiy of UVB tretment for psorisis. However, UVB therpy involves mny other mehnisms nd ell lineges s for exmple the UVB-indued immune suppression medited y Treg ells following mphiregulin stimultion y sophils (Meulenroeks et l., 5). In summry, we provide ompelling evidene for the existene of pthogeni sde linking AMPK intivtion, HuR ytoplsmi reloliztion, nd inresed IL- stiliztion in the psoriti epidermis. Our findings further support the importnt post-trnsriptionl regultory role of HuR in the pthogenesis of psorisis nd implite HuR in the psoriti KC response to metoli stress. Understnding the ontriution of this potentil pthwy to psorisis will e enhned y future studies tht will inlude oth therpeuti trgeting of AMPK nd HuR nd relevnt genetilly modified niml models. MATERIA AND METHODS Humn tissue smples We used tumor nk filities offered y the University Hospitl Crémeu, Nîmes (Frne). This ws n open study tht ws pproved y the ethis ommittee of the University Hospitl of Crémeu, Nîmes (Frne), in ordne with Frenh lws nd regultions, nd thus ll ptients provided written onsent prior to study prtiiption. The inluded ptients with hroni plque-type psorisis hd not reeived topil nti-psoriti therpy during the previous weeks or systemi therpy during the previous weeks. Kertome iopsies, whih were used for IHC, quntittive rel-time reverse trnsriptsepcr, nd RIP- Seq, were hrvested from nd N (t distne of t lest 5 m from lesionl plque) in ptients with plque-type psorisis. Histology For humn smples, formlin-fixed prffin-emedded tissue setions (7 μm) were used for immunoperoxidse stining. The setions were ounterstined with Hrris hemtoxylin (Sigm-Aldrih, St. Louis, MO). Seril setions ( μm) of formlin-fixed prffinemedded mouse k skin were stined with hemtoxylin-eosinsfrn or used for immunoperoxidse stining nd ounterstined with Hrris hemtoxylin. For humn nd mouse immunoperoxidse stining, negtive ontrols were proessed with mthed ontrol ntiody. All setions were oserved with n Olympus AX7 mirosope (Olympus, Tokyo, Jpn). 78 Journl of Investigtive Dermtology (5), Volume 5

8 RIP-Seq Down in psorisis Up in psorisis.9 % 8. % 7. % % Mture RNAs Pre-nd mture RNAs Pre-mture RNA 8.7 %. % 7. % Log( involved ).5 uninvolved Reltive to CXCL- DEFBA SLCA CCNE si-neg si-hurs si-neg si-hurs si-neg si-hurs si-neg si-hurs Reltive to CYP7B LCN CCL7 FAH si-neg si-hurs si-neg si-hurs si-neg si-hurs si-neg si-hurs d Gene COG FBXL9 HLA-E ILB ILR NFKBIA NOS PRSS5 PSMA REL RNF SDC SLC9A8 SPATA STAT TNIP TRAFIP TSC TYK Lous q p. p. 5q. p. q. 7q. q. 7q. p. q. q. p. p. 5q. 5q. 5q. 9p. 9p. Pre-mRNA mrna e KEGG_PATHWAY Lysosome Endoytosis F gmm R-medited phgoytosis Inositol phosphte metolism mtor signling pthwy Neurotrophin signling pthwy Adherens juntion Fol dhesion Phosphtidylinositol signling system Pthwys in ner Insulin signling pthwy Count P-vlue.7E 5 7.E 5.E.9E.E.7E.8E.9E.8E.E.E Figure 5. HuR inds numerous trnsripts involved in psorisis pthophysiology. () Comprison of rionuleoprotein immunopreipittion nlyzed y highthroughput sequening (RIP-Seq) HuR mrna trgets (Supplementry Dt S online, exoni sheet) with up- nd downregulted genes found in pulished psorisis trnsriptomes (Suárez-Friñs et l., ). () The sme nlysis with RIP-Seq HuR mrna trgets exhiiting n inresed fold inding (log.5) to HuR in lesionl skin () versus non-lesionl skin (N). () Smll interfering RNA (sirna) knokdown of HuR. HCT ells were trnsfeted nd stimulted s desried in Figure (± SEM; n = for eh sirna; Po.5, Po.5, nd Po.5), nd the expression of ertin trnsripts ws determined y qpcr. (d) Tle of genes suggested to e implited in psorisis y genome-wide ssoition studies (Yu et l., 8) tht ppered to e HuR trgets in our RIP-Seq nlysis (log-fold inding in ). (e) Kyoto Enylopedi of Genes nd Genomes (KEGG) nlysis of HuR mture RNA trgets exhiiting log.5-fold inresed inding in using DAVID ioinformtis resoures. Pthwys with Po. re inluded in the tle. Counts represent the numer of mthing genes for eh KEGG pthwy. 79

9 G Grin et l. C + CompC % Remining mrna N C + CompC 8 Min. fter t. D d H.E.S p-ampk HuR IL- C CompC SDS Figure. AMP-tivted protein kinse (AMPK) intivtion leds to HuR ytoplsmi loliztion nd IL- mrna stiliztion nd initites psoriti fetures. () Immunohistohemistry (IHC) for phosphorylted AMPK (p-ampk) in non-lesionl skin (N) nd lesionl skin (; one ptient, representtive of five). Sle rs = μm for originl mgnifition of nd 5 μm for originl mgnifition of. () Confol nlysis of HuR loliztion in HCT ells tht were treted or not (C) with Compound C (CompC; μm for 5 hours). Sle r = μm. () IL- mrna stility in HCT ells tht were treted (filled squres) or not (open squres) with CompC ( μm) for hours (one representtive; n = ) (d) Dorsl mouse skin tht ws treted dily with ontrol C (5% ethnol), SDS (5% PBS), or CompC ( μm, 5% ethnol) for weeks. Hemtoxylin-eosin-sfrn (H.E.S.) stining (green line, nthosis); n =. IHC stining for p-ampk, HuR, nd IL-; n =. Sle rs = μm for originl mgnifition of nd 5 μm for originl mgnifition of. RNA purifition nd quntittive rel-time PCR Skin inflmmtion models Totl RNA ws prepred using TRIzol (Invitrogen, Crlsd, CA) for ellulr smples or the MgNA Pure Compt RNA Isoltion Kit (Rohe Applied Siene, Mnnheim, Germny) for frozen tissue smples ( 8 C). RNAs were heked for DNA ontmintion with the pproprite primers. Reverse trnsription ws onduted using reverse-trnsried murine Moloney leukemi virus reverse trnsriptse (Invitrogen) with oligo(dt) primers (Invitrogen) for mrna quntifition nd rndom primers (Promeg, Mdison, WI) for premrna quntifition. C57BL/ mie (Chrles River Lortories, Chtillon-sur-Chlronne, Frne) were treted dily on shved k skin with either 5% SDS in phosphte-uffered sline, CompC-HCl (Merk, Drmstdt, Germny) in 5% ethnol, or the ontrol (5% ethnol). Two weeks lter, the mie were killed, nd the skin ws exised. Two independent experiments onsisting of t lest three mie in eh group were performed. The Ethis Committee on Animl Reserh of Lnguedo-Roussillon, Frne, hs pproved this study (CEEALR-9). Rionuleoprotein immunopreipittion nlyzed y high-throughput sequening Sttistil nlyses To limit individul geneti vrition, we pooled five nd five N smples for eh HuR or mok IP. Frozen smples ( 8 C) were disrupted using the MgNA Pure Instrument (Rohe Applied Siene). IPs of endogenous RNA-protein omplexes from nd N were evluted s previously desried (Peritz et l., ) using Protein A-Sephrose (Sigm-Aldrih, St Louis, MO) tht hd een pre-oted with μg of nti-hur or isotype ontrol (mok) ntiodies. RNA-Seq is desried in the Supplementry Mterils nd Methods online. 7 Journl of Investigtive Dermtology (5), Volume 5 The vlues re expressed s the mens ± SEM. The dt were nlyzed using n unpired two-tiled Student s t-test, or pired Wiloxon test, or Spermn s rnk orreltion. Chnges were identified s signifint if the P-vlue ws o.5. P-vlues ssoited with RIP-Seq funtionl nnottions were derived using DAVID ioinformtis resoures (Hung et l., 9). CONFLICT OF INTEREST The uthors stte no onflit of interest.

10 ACKNOWLEDGMENTS This projet ws supported y the Soiété Frnçise de Dermtologie. We thnk Lurent Henry for providing omments on the mnusript; Ghysline Tueh, Ptrii Boissin, Amndine Lekhell, Mrie-Jenne Roux, Asm Znti, Smir Hnin, nd An Gámez for their outstnding tehnil ssistne; Dr Jen-Mr Psussi, Dr Jen-Alin Fehrentz, Dr Jen-Louis Bnères, nd Céline M'Kdmi for their kind provision of vrious regents; Mikël Lefevre for use of the onfol lser mirosope; Professor Psl Roger for tumor nk ess; nd the Réseu d Histologie Expérimentle de Montpellier histology fility for proessing our niml tissues. SUPPLEMENTARY MATERIAL Supplementry mteril is linked to the online version of the pper t REFERENCES Armstrong AW, Hrskmp CT, Armstrong EJ () Psorisis nd metoli syndrome: systemti review nd met-nlysis of oservtionl studies. JAmAdDermtol8:5 Bekers A, Orgne S, Timmermns L et l. () Methotrexte enhnes the ntinoli nd ntiprolifertive effets of 5-minoimidzole-- roxmide rioside. Mol Cner Ther 5:7 Blumerg H, Conklin D, Xu WF et l. () Interleukin : disovery, reeptor identifition, nd role in epiderml funtion. Cell :99 Buerger C, Mlisiewiz B, Eiser A et l. () mtor nd its downstrem signlling omponents re tivted in psoriti skin. Br J Dermtol. 9: 59 Fn J, Ishmel FT, Fng X et l. () Chemokine trnsripts s trgets of the RNA-inding protein HuR in humn irwy epithelium. JImmunol8: 89 Gutowsk-Owsik D, Ogg GS () The epidermis s n djuvnt. JInvest Dermtol :98 Hdd SM, Fleming S, Thompson AM (8) Trgeting AMPK: new therpeuti opportunity in rest ner. Crit Rev Onol Hemtol 7:7 Hlprin KM, Ohkwr A (9) Crohydrte metolism in psorisis: n enzymti study. JInvestDermtol:59 Hrmon CS, Phizkerley PJ (98) Glyogen metolism in psoriti epidermis nd in regenerting epidermis. Clin Si 7:98 He M, Ling P () IL- trnsgeni mie: in vivo evidene of overlpping funtions for IL-, IL-, nd IL- in the epidermis. JImmunol8: 798 Herdenstm CG (9) On the in vitro metolism of leled gluose in norml nd psoriti skin slies. At Derm Venereol Suppl (Stokh) (Suppl 7): Hung DW, Shermn BT, Lempiki RA (9) Bioinformtis enrihment tools: pths towrd the omprehensive funtionl nlysis of lrge gene lists. Nulei Aids Res 7: Inoki K, Kim J, Gun K-L () AMPK nd mtor in ellulr energy homeostsis nd drug trgets. Annu Rev Phrmol Toxiol 5:8 Jnsen PAM, Rodijk-Olthuis D, Hollox EJ et l. (9) β-defensin- protein is serum iomrker for disese tivity in psorisis nd rehes iologilly relevnt onentrtions in lesionl skin. PLoS One :e75 Johnsen C, Funding AT, Otkjær K et l. () Protein expression of TNF-lph in psoriti skin is regulted t posttrnsriptionl level y MAPKtivted protein kinse. JImmunol7:8 Jovnovi M, Rooney MS, Mertins P et l. (5) Immunogenetis. Dynmi profiling of the protein life yle in response to pthogens. Siene 7: 598 Kim C, Sno Y, Todorov K et l. (8) The kinse p8α serves ell type speifi inflmmtory funtions in skin injury nd oordintes pro- nd nti-inflmmtory gene expression. Nt Immunol 9:97 Kim HH, Gorospe M (8) Phosphorylted HuR shuttles in yles. Cell Cyle 7: Kumri S, Bonnet MC, Ulvmr MH et l. () Tumor nerosis ftor reeptor signling in kertinoytes triggers interleukin--dependent psorisis-like skin inflmmtion in mie. Immunity 9:8999 Leedev S, Jens M, Theil K et l. () Trnsriptome-wide nlysis of regultory intertions of the RNA-inding protein HuR. Mol Cell :5 Lee YM, Lee JO, Jung J-H et l. (8) Retinoi id leds to ytoskeletl rerrngement through AMPK-R nd stimultes gluose uptke through AMPK-p8 MAPK in skeletl musle ells. JBiolChem8:997 Meulenroeks C, vn Weelden H, Shwrtz C et l. (5) Bsophil-derived mphiregulin is essentil for UVB irrdition-indued immune suppression. J Invest Dermtol 5:8 Mitr A, Ryhudhuri SK, Ryhudhuri SP () IL- indued ell prolifertion is regulted y PIK/Akt/mTOR signling sde. Cytokine :8 Mukherjee N, Cororn DL, Nusum JD et l. () Integrtive regultory mpping indites tht the RNA-inding protein HuR ouples pre-mrna proessing nd mrna stility. Mol Cell :79 Nors LB, Gillespie GY, Hrkins L et l. () HuR, RNA stility ftor, is expressed in mlignnt rin tumors nd inds to denine- nd uridinerih elements within the untrnslted regions of ytokine nd ngiogeni ftor mrnas. Cner Res :5 O'Neill LAJ, Hrdie DG () Metolism of inflmmtion limited y AMPK nd pseudo-strvtion. Nture 9:55 Peritz T, Zeng F, Knnnykl TJ et l. () Immunopreipittion of mrnaprotein omplexes. Nt Proto :5778 Rni M, Ryhowdhury R, Jovnovi M et l. () High-resolution sequening nd modeling identifies distint dynmi RNA regultory strtegies. Cell 59:987 Roller A, Perino A, Dpvo P et l. () Blokde of phosphtidylinositol -kinse PIKδ or PIKγ redues IL-7 nd meliortes imiquimodindued psorisis-like dermtitis. JImmunol89: Rutz S, Wng X, Ouyng W () The IL- sufmily of ytokines from host defene to tissue homeostsis. Nt Rev Immunol :7895 S SM, Vldez PA, Wu J et l. (7) The effets of IL- sufmily ytokines on reonstituted humn epidermis suggest potentil roles in utneous innte defense nd pthogeni dptive immunity in psorisis. JImmunol78: 9 St R () IL- fmily of ytokines. Cytokine Growth Ftor Rev :5 St R, Wolk K () Reserh in prtie: IL- nd IL-: signifine for epithelil homeostsis nd psorisis pthogenesis. J Dtsh Dermtol Ges 9: 58 Singh M, Mrtinez AR, Govindrju S et l. () HuR inhiits poptosis y mplifying Akt signling through positive feedk loop. JCellPhysiol 8:89 Srikntn S, Gorospe M () UneCLIPsing HuR nuler funtion. Mol Cell : 9 Srikntn S, Gorospe M () HuR funtion in disese. Front Biosi 7:895 Suárez-Friñs M, Lowes MA, Z LC et l. () Evlution of the psorisis trnsriptome ross different studies y gene set enrihment nlysis (GSEA). PLoS ONE 5:e7 Wng W, Fn J, Yng X et l. () AMP-tivted kinse regultes ytoplsmi HuR. Mol Cell Biol :5 Winzen R, Krht M, Ritter B et l. (999) The p8 MAP kinse pthwy signls for ytokine-indued mrna stiliztion vi MAP kinse-tivted protein kinse nd n AU-rih region-trgeted mehnism. EMBO J 8: 998 Yu W, Gwinn M, Clyne M et l. (8) A nvigtor for humn genome epidemiology. Nt Genet :5 Zhng J, Bowden GT (8) UVB irrdition regultes Cox- mrna stility through AMPK nd HuR in humn kertinoytes. Mol Crinog 7: 978 Zhou G, Myers R, Li Y et l. () Role of AMP-tivted protein kinse in mehnism of metformin tion. JClinInvest8:77 7