IRF7 Inhibition Prevents Destructive Innate Immunity—A Target for Nonantibiotic Therapy of Bacterial Infections

RESEARCH ARTICLE

INFECTIOUS DISEASE IRF7 inhibition prevents destructive innate immunity—A target for nonantibiotic therapy of bacterial infections

Manoj Puthia,1* Ines Ambite,1* Caterina Cafaro,1 Daniel Butler,1 Yujing Huang,1 Nataliya Lutay,1 Gustav Rydström,1 Birgitta Gullstrand,2 Bhairavi Swaminathan,3 Aftab Nadeem,1 Björn Nilsson,3,4 Catharina Svanborg1†

Boosting innate immunity represents an important therapeutic alternative to antibiotics. However, the molecular se- lectivity of this approach is a major concern because innate immune responses often cause collateral tissue damage. We identify the transcription factor interferon regulatory factor 7 (IRF-7), a heterodimer partner of IRF-3, as a target for non–antibiotics-based therapy of bacterial infections. We found that the efficient and self-limiting innate immune response to bacterial infection relies on a tight balance between IRF-3 and IRF-7. Deletion of Irf3 resulted in overexpression of Irf7 and led to an IRF-7–driven hyperinflammatory phenotype, which was entirely prevented if Irf7 was Downloaded from deleted. We then identified a network of strongly up-regulated, IRF-7–dependent genes in Irf3−/− mice with kidney pathology, which was absent in Irf7−/− mice. IRF-3 and IRF-7 from infected kidney cell nuclear extracts were shown to bind OAS1, CCL5,andIFNB1 promoter oligonucleotides. These data are consistent in children with low IRF7 expression in the blood: attenuating IRF7 promoter polymorphisms (rs3758650-T and rs10902179-G) negatively associated with recurrent pyelonephritis. Finally, we identified IRF-7 as a target for immunomodulatory therapy. Administering liposomal Irf7 siRNA to Irf3−/− mice suppressed mucosal IRF-7 expression, and the mice were protected against infection and renal tissue damage. These findings offer a response to the classical but unresolved question of “good versus bad http://stm.sciencemag.org/ inflammation” and identify IRF7 as a therapeutic target for protection against bacterial infection.

INTRODUCTION mice develop severe APN with urosepsis and renal abscesses due to ex- Innate immunity provides an immediate antimicrobial defense to all acerbated, aberrant inflammation. Additionally, we detected hypomor- classes of plant and animal life (1–3). Through specific recognition phic IRF3 promoter variants in humans that associate with recurrent and activation strategies, this defense eliminates pathogens while APNinchildren(4, 9). These studies define IRF-3 as a negative regu- sustaining symbiosis with commensals (4). As innate immunity relies lator of the destructive effects of TLR4 activation, providing an entry on inflammation to execute the antimicrobial defense, the innate im- point to identifying molecular determinants of tissue damage. on April 28, 2016 mune response is also a major cause of tissue dysfunction and disease. The IRFs control essential aspects of innate immunity [for reviews, This lack of specificity contributes to the pathobiology of infectious dis- see (10–13)]. IRF-3 (14) was first thought to be the prime initiator of eases and complicates the development of therapies aimed at boosting protective type I IFN responses, but with the discovery of IRF-7 (15, 16), innate immunity (5–7). It remains unclear if pathology is a necessary heterodimers of IRF-3 and IRF-7 were proposed to act in tandem as pricetopayforprotectionorifspecificmolecularstepsintheinnate transcriptional activators (17, 18). Later, it was found that IRF-7 alone immune response can distinguish “good” from “bad” inflammation. induces IFN-a expression, acting as a “master regulator” of type I IFN Transcriptional control of innate immunity is closely linked to in- responses (19). IRF-3 and IRF-7 both regulate type I IFN responses to flammation, because transcription factors determine the inflammatory viral infection. In addition, viral pathogens trigger IRF-7 phosphoryl- gene repertoire in infected hosts (8). We recently identified interferon ation, resulting in nuclear translocation and the formation of a tran- regulatory factor 3 (IRF-3) as a central regulator of mucosal innate im- scriptional complex on the IFN gene family promoter, together with, munity in the context of acute pyelonephritis (APN) (9). This tran- for example, nuclear factor kB(NF-kB), c-Jun, activating transcription scription factor is activated by uropathogenic Escherichia coli via a factor 2 (ATF-2), and p300/CREB-binding protein (20). Interactions TLR4 (Toll-like receptor 4)–dependent and CREB (3′,5′-cyclic adeno- with additional transcription factors like CREB, PU.1, and SMAD sine monophosphate response element–binding protein)–dependent add complexity and specificity (13, 15, 21). signaling pathway and a protective type I interferon (IFN)–dependent Here, we observed that IRF-3 and IRF-7 have opposing effects on the − − − − antibacterial response (4, 9). In the absence of IRF-3, infected Irf3 / innate immune response to uropathogenic E. coli.UnlikeIrf3 / mice, which developed acute, septic kidney infection and extensive tissue − − damage, Irf7 / mice rapidly cleared pathogenic E. coli without evidence 1Department of Microbiology, Immunology and Glycobiology, Institute of Laboratory 2 of tissue pathology. An Irf7-dependent gene network was transcription- Medicine, Lund University, 223 62 Lund, Sweden. Section of Rheumatology, Depart- −/− ment of Clinical Sciences, Lund University and Skåne University Hospital, 221 85 Lund, ally activated in Irf3 mice with tissue pathology, and IRF-3 and IRF-7 Sweden. 3Division of Hematology and Transfusion Medicine, Institute of Laboratory were shown to bind together and independently to promoters of these 4 Medicine, Lund University, 223 62 Lund, Sweden. Broad Institute, 7 Cambridge genes, supporting the in vivo data. The results demonstrate that IRF-3 Center, Cambridge, MA 02142, USA. *These authors contributed equally to this work. and IRF-7 control different facets of the innate immune response to kid- †Corresponding author. Email: [email protected] ney infection and the development of APN. In addition, Irf7 serves as a

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 1 RESEARCH ARTICLE

Fig. 1. IRF7 gene network in infected human kidney epithelial cells. IRF7 and IFN family gene expression in infected human kidney epithelial cells. The pathogenic strain E. coli CFT073 or the asymptomatic carrier strain E. coli 83972 was used, and the pathogen-specific response was identified by genome- wide transcriptomic analysis of total mRNA. (A)Heatmapofsignificantly regulated type I IFN genes [cut-off log2 fold change (FC), >0.585]. ABU, asymptomatic bacteriuria. (B) Type I IFN family genes and transcription factors were activated by CFT073, including IRFs and STATs (signal transducers and activators of transcription). (C) An IPA network analysis of regulated genes identified IRF7 as a transcriptional hub for genes involved in the antimicrobial defense and inflammatory responses to infection. IRF7 expression was pathogen-specific and not activated by the E. coli 83972 (created by IPA; red, activated; blue, suppressed). (D) Nuclear translocation of IRF-7 in infected cells (CFT073), visualized by confocal imaging and compared to uninfected cells [phosphate-buffered saline (PBS)]. Total cellular and nuclear staining intensities were quantified and normalized against uninfected cells (means ± SEM, four experiments; *P < 0.05, **P <0.01,unpairedt test). Downloaded from

− − putative immunotherapeutic target in Irf3 / mice, where liposomal delivery of Irf7-specific small interfering RNA (siRNA) prevented APN, to a similar extent as antibiotics. http://stm.sciencemag.org/

RESULTS

Infection activates a mucosal IRF-7 response Epithelial cells in the renal pelvic mucosa act as immune sensors of bacterial infection (22–24). Bacterial virulence factors engage host cells receptors and activate specific signaling pathways and innate immune responses (25–27). To address whether IRF-7 is activated by infection and characterized by associated gene networks, we first infected the human kidney epithelial cell line A498 with the uropathogenic E. coli on April 28, 2016 strain CFT073 (28, 29). Gene expression profiling using Illumina HumanHT-12 microarrays revealed an up-regulation of proinflammatory genes, particularly type I IFN–relatedgenes(Fig.1,AandB,andfig. S1). IRF7 expression increased in infected cells, and Ingenuity Pathway Analysis (IPA) showed that genes activated by the pathogenic strain were enriched within the same functional network, implicating IRF-7 as a key regulator of the innate immune response to uropathogenic E. coli (Fig. 1C). Nuclear translocation of IRF-7 was confirmed by confocal imaging of the infected cells. In contrast, the ABU strain E. coli 83972 did not activate IRF7 expression or trigger nuclear translocation of IRF-7 (Fig. 1D) (24).

IRF-3 and IRF-7 have opposing roles during APN − − − − To functionally characterize IRF-7, we first infected Irf3 / , Irf7 / ,and wild-type C57BL/6 mice with E. coli CFT073 by intravesical catheteri- zation and assessed the degree of inflammation, tissue damage, and bac- − − terial clearance after 24 hours or 7 days (9, 30). Unexpectedly, Irf7 / − − − − and Irf3 / mice showed radically different phenotypes. Irf3 / mice developed large abscesses, kidney edema, and purulent discharge from the renal pelvis (Fig. 2A). The pelvic mucosa and renal papilla were infil- trated by neutrophils, and urine neutrophil counts increased until day 7 (Fig. 2, B and C). Bacteria crossed the mucosal barriers in the renal pelvis and papillae (fig. S2), followed by invasion into collecting ducts and formation of abscesses (Fig. 2C and fig. S3). By contrast, kidneys from

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 2 RESEARCH ARTICLE

− − Irf7 / mice remained macroscopically and histologically unaffected (Fig. 2, A to C). Re- markably, in these mice, bacterial counts in urine and kidney tissue were lower than in − − both Irf3 / and wild-type mice. We also noted an interesting difference in the neutrophil − − response kinetics. Whereas both Irf3 / and − − Irf7 / mice showed an early peak at 24 hours, − − the neutrophils remained high in Irf3 / mice but plummeted to barely detectable levels in − − Irf7 / mice by day 3 (Fig. 2B and fig. S2C). These data suggest that IRF-7 inhibition confers a protective effect in mice, partly because of a self-limiting inflammatory response. The different gene expression profiles

in human kidney cells (Fig. 1) suggested Downloaded from that virulence is required to trigger the IRF-7 response. To address this question, − − − − we infected Irf3 / and Irf7 / mice with the ABU strain E. coli 83972 and compared the response to C57BL/6 wild-type mice (fig. S4). The mice were transiently

infected by the ABU strain, with negative http://stm.sciencemag.org/ cultures on day 7. A low inflammatory response was detected after 24 hours, quantified as urine neutrophil numbers. By 7 days, this response had subsided. There was no macroscopicevidenceoftissueinvolvement, defined by lack of edema, hyperemia, and abscesses. The results suggest that virulence is required to activate the pathogen-associated IRF-7 signaling pathway. on April 28, 2016 Gene expression analysis identifies an Irf7-dependent gene network in infected mice To understand the mechanism by which Irf7 deficiency is protective, we next re- corded gene expression profiles of infected − − − − or uninfected kidneys from Irf3 / , Irf7 / , and wild-type mice. The number of differentially expressed genes was significantly − − − − higher in Irf3 / than in Irf7 / mice Fig. 2. Opposing effects of IRF-3 and IRF-7 on UTI susceptibility and tissue pathology. Irf7−/−, Irf3−/−, (Fig.2,DandE).Consistentwiththe andC57BL/6wild-type(WT)micewereinfectedwithE. coli CFT073 by intravesical inoculation and sacrificed difference in histology, granulocyte mi- after 7 days. (A) Kidney abscesses (arrows) and pathology in CFT073-infected Irf3−/− mice. In contrast, Irf7−/− gration and activation pathways were − − mice were protected against infection-induced pathology. (B) Quantification of kidney abscess formation, strongly activated in Irf3 / mice showing neutrophil infiltration, and bacterial numbers in urine and renal tissue. Individual mice are shown (n=8to10 kidney pathology, including cytokine and micepergroup,means±SEM,twoexperiments;*P <0.05,**P <0.01,***P <0.001,unpairedt test). CFU, chemokine genes [Cxcl3, Cxcl1 (GRO-a, colony-forming units. (C) Bacterial infection was accompanied by massive neutrophil accumulation in the −/− −/− growth regulated oncogene-a), Il33,andCcl6], kidneys of Irf3 mice [visualized by immunohistochemistry and hematoxylin and eosin (H&E) staining]. Irf7 tumor necrosis factor receptor super- mice were protected from tissue damage. Bacterial and neutrophil numbers were low. C57BL/6 WT mice family genes (Tnfrsf1a, Tnfrsf1b,and showed moderate pathology, bacterial counts, and neutrophil recruitment. Arrows indicate neutrophils. Scale bars, 200, 50, and 20 mm. (D and E) Elevated number of transcriptionally regulated genes in Irf3−/− mice com- Tnfrsf11b), the interleukin-1, fMLP, and pared to Irf7−/− mice. Microarray analysis of total kidney mRNA from E. coli CFT073–infected mice relative to granulocyte colony-stimulating factor re- uninfected mice of each genotype (7 days, n = 4 mice per group). (D) Histogram showing that 956 of 1599 of ceptor genes (Il1b, Il1r1, Fpr2,andCsf3r), regulated genes were activated in Irf3−/− mice, with tissue damage compared to 199 of 599 genes in Irf7−/− adhesion molecules and ligands [Vcam1 mice (P < 0.001, c2 test). (E) Heat map of regulated genes comparing Irf3−/−, Irf7−/−, and C57BL/6 WT mice. and Icam1, Itgb1 (LFA-1), and Itgam/Itgb2

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 3 RESEARCH ARTICLE

(macrophage-1 antigen)] (fig. S5). These genes were not regulated in − − Irf7 / mice, where neutrophil recruitment was abrogated, after an initial 24-hour peak. IRF-7–dependent, prolonged, and excessive neutrophil infiltration was thus identified as a driving force for the development of − − pathology in Irf3 / mice (Fig. 2C and fig. S5). The results demonstrate that the efficient and self-limiting innate immune response in C57BL/6 relies on a tight balance between Irf3 and Irf7. Disrupting this balance by single gene deletions creates two extreme phenotypes. The Irf3 deletion and resulting overexpression of Irf7 causes a marked IRF-7–driven hyperinflammatory phenotype, which is entirely prevented if Irf7 is deleted. Further analysis identified a specific set of genes encoding proteins − − that interact with IRF-7, which itself was strongly up-regulated in Irf3 / mice. Transcription factors Junb, Stat3,andIkbkg were differentially regulated Irf7-dependent genes, providing a mechanistic basis for the exag- geratedtissueresponse(Fig.3).This gene set also overlapped with genes

induced by E. coli CFT073 in A498 human kidney cells (fig. S6). IRF-7 Downloaded from protein expression was exclusively epithelial, identifying the mucosal tissue rather than the inflammatory infiltrate as the source of IRF-7 (fig. S7).

IRF-3 and IRF-7 bind to promoters of downstream genes We subsequently examined whether both IRF-3 and IRF-7 bind to the promoters of the IRF-7–dependent genes that were up-regulated in http://stm.sciencemag.org/ vivo (Fig. 3B). DNA fragments of 187 base pairs (bp) from the OAS1 and CCL5 promoters and a fragment of 143 bp from the IFNB1 promoter were amplified and used as probes in electrophoretic mobility shift assays (EMSAs) (Fig. 4A). Each of the DNA fragments was shown to interact strongly with a nuclear protein extract from kidney cells infected with CFT073, resulting in four band shifts for the OAS1 and CCL5 promoters and five for the IFNB1 promoter (Fig. 4B and fig. S8, A and B). The probes alone formed a single low–molecular weight band, serving as a negative control. Specificity for IRF-3 and IRF-7 was confirmed by competitive inhibition, using IRF-3– or IRF-7–specific on April 28, 2016 antibodies and by depletion of IRF-3 or IRF-7, by coimmunoprecipitation with specific antibodies. Depletion was confirmed by West- ern blots, showing a marked reduction in band intensity (Fig. 4B and fig. S8C). A marked change in IRF-3 and IRF-7 band intensities was detected by EMSA after inhibition of known protein partners in the IRF-3/7 transcription complex or related proteins(forexample,enhanceosome) (Fig. 4C). The c-Jun and NF-kB regulatory kinase IKK-g was transcrip- − − tionally regulated in an IRF-7–dependent manner in Irf3 / mice (Fig. 3C). c-Jun and its cofactor ATF-2, the p65 (RELA) subunit of NF-kB, as well as CBP and p300 that acetylate IRF-3 (31) assist in IRF-3 activation and nuclear translocation (19). The results show that IRF-3 and IRF-7 bind to these promoter sequences both together and independently, suggesting that IRF-7 can regulate these genes also in the absence of IRF-3 and vice versa.

IRF-7 promoter sequence variants differ between patients – with APN and ABU Fig. 3. IRF-7 dependent gene network associated with severe infec- −/− tion. (A) Heat map identifying genes regulated by infection in susceptible Because of the protected phenotype in Irf7 mice, we hypothesized −/− Irf3 mice (7 days after infection, n = 4 mice per group). (B) Pathology- that allelic variation in IRF7 might influence the risk for APN in associated IRFs, chemokine and chemokine receptors genes, transcriptional humans. Whole-exome sequencing data from the Exome Aggregation regulators, and innate and specific immunity regulators. (C) Pathology- Consortium and the National Heart, Lung, and Blood Institute specific, IRF-7–driven gene network in Irf3−/− mice, but not in Irf7−/− or (NHLBI) Exome Sequencing Project showed that coding loss-of- C57BL/6 WT mice, based on direct and indirect interactions between the pro- function mutations in IRF7 are rare (67,209 individuals, mainly of teins encoded by those genes.

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 4 RESEARCH ARTICLE

European and African origin; tablesS1andS2).However,Bloodgene show evidence of having a protected phenotype. In these samples, we expression data predicted the existence of common promoter variants identified three polymorphic promoter SNPs (Fig. 5A and figs. S9 and that influence IRF7 expression [8086 Europeans genotyped using S10). rs3758650 and rs1092179 are tightly linked, common SNPs (mi- single-nucleotide polymorphism (SNP) microarrays (32); table S3]. nor allele frequency, 6.9%), annotated as associated with lower IRF7 ex- We therefore sequenced the IRF7 promoter in 17 Swedish children with pression in the expression quantitative trait loci (eQTL) database (Fig. recurrent APN. These children were prospectively monitored for a 5B) (32). minimum of 5 years to establish a recurrent infection pattern after an Remarkably, all of the children with APN were homozygous for the initial APN episode (9, 33). In contrast, we included 14 children who major allele rs3758650-C, which confers high IRF7 expression. By con- were diagnosed with ABU but never developed APN (9, 33)andthus trast, a proportion of the children with ABU were heterozygous for rs3758650-T, which confers lower IRF7 expression (Fisher’sexacttest,P = 0.0463). An identical association was observed for the linked rs10902179 allele, whereas the rare rs3832720 showed no disease association (Fig. 5C) (32). For validation, we genotyped an independent series of children with

APN (n = 34) or ABU (n = 43) plus a series Downloaded from of Swedish blood donors (n =41)(Fig.5D). Again, the rs3758650-T allele was more common in children with ABU as compared to children with APN or blood donors (Fisher’s exact test, P = 0.0045 and P = 0.0293, respectively; Fig. 5D). Further-

more, the association between rs3758650-T http://stm.sciencemag.org/ and reduced IRF7 expression was replicated in Swedish and Icelandic individuals (Fig. 5E and table S4). These data demonstrate the existence of common IRF7 variants that protect against recurrent APN in children. We also noted that other IRF7 variants have been associated with viral infections (34) and autoimmune conditions (35–38).

Irf7 siRNA inhibition on April 28, 2016 attenuates APN We subsequently addressed whether IRF-7 inhibition is sufficient to mitigate destructive inflammatory responses and accentu- ate protection. Finally, we therefore tested the effect of suppressing Irf7 in APN-susceptible − − Irf3 / mice. Liposomal Irf7 siRNA was administered intravenously and intravesically 3daysbeforeandonthedayofinfection with E. coli CFT073 (Fig. 6A). Remarkably, liposomal siRNA delivery abrogated tissue destruction in infected − − Irf3 / mice. Instead, we recovered a pro- − − tected phenotype comparable to the Irf7 / − − mice (Fig. 2). The siRNA-treated Irf3 / mice showed fewer abscesses, less tissue OAS1 CCL5 IFNB1 Fig. 4. IRF-3 and IRF-7 binding to , , and promoters. (A) OAS1, CCL5, and IFNB1 damage, and lower bacterial and neutrophil promoter sequences. Binding sites for IRF-3 (in green, underlined) and IRF-7 (blue) were identified using counts than infected controls receiving scram- the JASPAR database. The indicated DNA sequences were amplified by polymerase chain reaction (PCR) bled siRNA (Fig. 6, B to D). Reduced IRF-7 andusedforEMSA.(B)EMSAwiththeOAS1 promoter oligonucleotide and the nuclear protein extract from kidney cells infected with CFT073. Four different protein-DNA complexes with IRF-3 and/or IRF-7 were de- expression was confirmed by immunohisto- tected (arrows). Specificity was confirmed by competitive inhibition with anti–IRF-3 and anti–IRF-7 antibo- chemistry in renal epithelial tissue (Fig. 6C). dies and by depletion of IRF-3 or IRF-7 through coimmunoprecipitation. (C) Binding of IRF-3 and IRF-7 was Furthermore, the transcript levels were over- competitively inhibited by antibodies to other protein partners in the transcriptional complex [anti–c-Jun, all suppressed and specifically for genes (in- anti–NF-kB p65, anti-p300, and anti–CBP (CREB-binding protein) antibodies]. cluding Stat3, S100a8, Cxcl2,andJunb)that

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 5 RESEARCH ARTICLE

Fig. 5. Human IRF7 promoter polymorphisms associate with lower IRF7 expression and protection against recurrent APN. (A) Identification of candidate markers. In a total of 31 samples from children with recurrent APN (n =17)orABU(n =14),threepolymor- phic SNPs were identified in the IRF7 promoter region (red star). Other previously reported SNPs in this region (black line) were not polymorphic in our samples (detailed in fig. S9). (B) Analysis of blood gene expression profiles from 8086 genotyped Europeans (32) predicted that the rs3758650 minor allele associ- ates with lower IRF7 expression. A similar association was observed in a set of 973 individuals from the Icelandic population (table S4). These observations are consistent with the mouse experiments. chr., chromosome. (C) For the tightly linked SNPs rs3758650 and rs10902179 (r2/d′ =1),theminorallele Downloaded from was exclusively observed among children with the protected ABU phenotype (n =14), whereas all children with the susceptible APN phenotype (n = 17) were homozygous for the major allele (Fisher’s exact test, P values). (D) Replication of the overrepresentation of the rs3758650 minor allele among 43 children http://stm.sciencemag.org/ with ABU, as compared to 34 children with APN and 41 random blood donors by pyrosequencing (c2 test, P values; primers listed in fig. S10). (E) Replication of the association between the rs3758650 minor allele genotype (pyrosequencing, n =246)andlowerIRF7 expression [quantitative reverse transcription PCR (qRT-PCR), n = 44] in Swedish blood donors (2−DDCt levels with geometric means; two- tailed t test, P values). on April 28, 2016

were identified as pathology-associated in infected murine kidney (Fig. 6E), reduced the disease score, measured as the number of abscesses and the further supporting that IRF-7 expression is suppressed by siRNA in vivo. magnitude of the neutrophil infiltrate and bacterial counts (fig. S11). We Irf7 siRNA treatment was also given 24 hours after infection to ad- further compared the therapeutic efficacy of Irf7 siRNA to the broad- dress whether Irf7 siRNA treatment represents an advance over antibi- spectrum antibiotic cefotaxime using the same starting point (fig. S12). otic therapy. This time point corresponds to the time of acute The effect of Irf7 siRNA was similar to the 100 mg/kg dose of cefotaxime symptoms, when patients are first diagnosed with febrile urinary tract but lower than 400 mg/kg, suggesting that optimization of Irf7 inhibi- infection (UTI) and treatment is initiated. siRNA therapy significantly tion therapy could result in complete clearance of infection, comparable

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 6 RESEARCH ARTICLE

− − to the protected phenotype in Irf7 / mice. This conclusion was supported by the absence of pathology-associated gene expression in Irf7 siRNA–treated mice, confirming the mechanistic role of Irf7 in this model.

DISCUSSION

Modulation of the innate immune system represents a promising, relatively under- explored alternative to failing antimicrobial therapies. Rather than targeting invading pathogens directly, the aim is to boost protective functions of host in-

nate immunity. However, the intervention Downloaded from needs to be precise, so that symptoms and tissue destruction caused by exaggerated host responses can be isolated and avoided (8). Here, we address the classical but unresolved question of “good versus bad inflammation.” In a well-defined bacterial

infection model with established human http://stm.sciencemag.org/ relevance, we identify related transcription factors that control the beneficial versus destructive arms of innate immunity and de- vise a therapeutic approach to selectively target the destructive response while en- hancing bacterial clearance. IRF-7 has been extensively studied as a determinant of type I IFN responses and host resistance to viral infection (18, 19).

− − on April 28, 2016 Irf7 / mice respond poorly to hepatitis C, and type I IFN responses are reduced (19, 39). In monocytic cells, IRF-3 and IRF-7 form heterodimers, and the ratio plays an essential role for the inducible expression of type I IFN genes (18). Phos- phorylated IRF-7, together with IRF-3, c-Jun, ATF-2, and NF-kB, binds to virus- responsive IFN-a/b promoter elements and induces low amounts of type I IFNs, Fig. 6. RNA interference therapy inhibits Irf7 and protects APN-susceptible Irf3−/− mice. (A) Irf7 siRNA which bind to IFN-a/b receptors and en- was delivered intravenously and intravesically to susceptible Irf3−/− mice, 3 days (d) before infection and on hance IFN-dependent gene expression the day of intravesical infection with E. coli CFT073. IV, intravenous. (B) Irf7 siRNA treatment reduced gross (40, 41). In view of these reported similari- kidney pathology, compared to scrambled siRNA or untreated controls (7 days after infection, abscesses ties between IRF-3 and IRF-7, we were indicated by arrows). Neg, negative. (C) Irf7 siRNA inhibited mucosal IRF-7 expression in the renal pelvic surprised to find opposing effects in vivo mucosa and reduced bacterial and neutrophil staining, compared to scrambled siRNA or untreated controls. in the kidneys of infected mice. The results Submucosal STAT3 staining was also inhibited, confirming the IRF-7 dependence of the STAT3 response as suggest that in addition to the convergent predicted by the network in Fig. 3C. Immunohistochemistry of frozen renal tissue sections 7 days after in- −/− regulation by IRF-3 and IRF-7 of antiviral fection. For staining in Irf7 mice,seefig.S7.Scalebars,100mm. (D) Irf7 siRNA reduced the frequency of type I IFN responses in monocytes, IRF-7 abscesses in Irf3−/− mice, compared to scrambled siRNA or untreated controls (n =3to8).Irf7 siRNA also regulates a facet of the antibacterial re- improved bacterial clearance, resulting in lower bacterial counts in urine and renal tissue and reduced in- sponse that causes pathology. Controlling flammation, quantified as neutrophil numbers in urine [means ± SEM; **P < 0.01, two-way analysis of var- – iance (ANOVA)]. h, hours. (E) Gene expression was suppressed after Irf7 siRNA treatment, including this IRF-7 dependent response may be proinflammatory genes and transcriptional regulators Irf7, Stat3,andJun. Tlr4 was activated, consistent with especially important to regulate the anti- a fully functional antibacterial defense. Whole genome transcriptomic analysis of total kidney RNA from Irf7 bacterial defense in the mucosa, where siRNA–treated Irf3−/− mice, compared to untreated Irf3−/− controls 7 days after infection. bacteria first contact host cells.

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 7 RESEARCH ARTICLE

− − The protected phenotype in Irf7 / mice suggested that variant IRF7 uptake, tissue-specific delivery to the kidneys, and long-term safety. expression might affect human disease susceptibility. IRF7 polymorphisms Target populations should include both pediatric and adult patient have been associated with the severity of chronic hepatitis C–associated groups, especially those carrying APN-associated IRF3 promoter poly- liver cirrhosis, and IRF7 variants influence the IFN response in HIV-1 pa- morphisms, where IRF7 expression might be increased. Ongoing clin- tients (34). IRF7 genomic region polymorphisms or dysregulated IRF7 ical trials support the feasibility of RNAi-based therapies, which are expression has been described in systemic lupus erythematosus (SLE) being developed for indications including familial amyloidosis, hemo- patients (35, 36), and a strong connection between SLE and the SNP philia, and hepatitis C (45–48). rs4963128, 23-kb telomeric to IRF7 was found in a genome-wide association study (37). IRF7 genomic region polymorphisms are also associated with anticentromere autoantibody levels in patients with systemic MATERIALS AND METHODS sclerosis (38). We now identify attenuating IRF7 promoter polymorphisms with specificity for ABU patients who did not develop APN and Study design renal damage during the long-term follow-up. Their phenotypes may be This study addressed whether tissue damage is a necessary price to related to that of Irf7 siRNA–treated mice, where the disease phenotype pay for the successful elimination of pathogens by the innate immune was attenuated. defense. On the basis of the strong disease association of IRF-3 with − − The results suggested that virulence is required for bacteria to activate APN and the rapid development of kidney pathology in Irf3 / mice, – the IRF-7 driven hyperinflammatory response. The nonpathogenic we examined the closely related transcription factor IRF-7. We were Downloaded from ABU strain E. coli 83972 did not stimulate IRF-7 expression in human surprised to discover that these transcription factors control different kidney cells. Furthermore, there was no evidence of tissue pathology in facets of the host responses to bacterial infection. the kidneys of mice infected with this strain. It may be speculated that We first screened human kidney cells for gene expression in re- microbial evolution has created specific IRF-7 agonism in uropathogenic sponse to infection and showed that uropathogenic E. coli strain acti- E. coli strains to promote efficient dissemination from the site of infec- vates an inflammatory gene network with IRF7 as a hub. To address the − − tion through a hyperinflammatory response with capillary leakage. This in vivo role of IRF-7, Irf7 / mice were intravesically infected with the view is consistent with the known role of P fimbriae as pathogen-specific pathogenic strain CFT073. APN was defined by macroscopic inspection http://stm.sciencemag.org/ ligands in UTI, activating IRF-3/7 through TLR4 and a mitogen-activated followed by histology and immunohistochemistry after 24 hours and − − protein kinase–dependent downstream signaling pathway (9). Further- 7 days, and the phenotype of Irf7 / mice was compared to C57BL/6 − − more, P fimbriae are expressed by virtually all E. coli strains that cause wild-type mice or Irf3 / mice. − − urosepsis in otherwise healthy hosts (42), potentially linking them to After the surprising observation that Irf7 / mice were protected IRF-7 agonism, hyperinflammation, mucosal leakage, and the patho- from infection and kidney damage, we used genome-wide transcrip- − − genesis of systemic infection. tomic analysis to characterize genes involved in pathology in Irf3 / The results suggest that the balance between IRF-3 and IRF-7 must mice. An IRF-7–dependent gene network resembling that of infected be maintained if overreactions of the innate immune system are to be human kidney cells was identified. To explain the extreme overactiva- kept in check. Because IRF-7 and IRF-3 can work together in inducing tion of inflammatory genes in mice lacking Irf7,wereturnedtothehu-

− − − − on April 28, 2016 type I IFN responses, the opposing phenotypes in Irf3 / and Irf7 / man bladder cells to address the hypothesis that IRF-3 and IRF-7 act as mice present an apparent contradiction. In preliminary studies, we have opposing transcriptional regulators of the promoters of these − − found that Ifnb1 / mice develop severe infection and pathology with a downstream genes. Evidence was provided using EMSAs. − − tissue phenotype closely resembling that of Irf3 / mice, confirming the We also sequenced the IRF7 promoter and demonstrated the exis- importance of IFN-b to maintain the antibacterial defense and tissue tence of common IRF7 promoter variants, which were associated with integrity. However, type I IFN signaling network analysis did not reveal low IRF7 expression. This was validated in Swedish children with a any obvious difference in regulated genes between the susceptible and history of APN or ABU, and the variant was detected in the ABU group. − − the protected mice. The differences in susceptibility between Irf3 / and Finally, we tested IRF-7 as a target for immunotherapy to prevent or − − − − Irf7 / mice, therefore, do not appear to reflect direct effects on the IFN-b reverse pathology in susceptible Irf3 / mice. Using liposomal delivery signaling pathway. of Irf7 siRNA, the mice were protected against infection and tissue These findings also raise other intriguing questions. First, it is con- damage. Bacteria were also cleared from the infected kidneys of siRNA- ceivable that IRF-7 inhibition can be used to prevent and treat bacterial treated mice. − − infections other than APN. Because Irf7 / mice are susceptible to viral infections (19, 43, 44), long-term, systemic IRF-7 inhibition could have Bacterial strains side effects and should be avoided. In a treatment setting analogous to The prototype APN strain E. coli CFT073 (O6:K2:H1) (49) and the antibiotic therapy, IRF-7 inhibition would be temporary, however, ABU strain E. coli 83972 (OR:K5:H-) (24, 50–52) were cultured on making increased susceptibility to virus infections an unlikely concern. tryptic soy agar plates (16 hours, 37°C), harvested in PBS (pH 7.2) Second, the divergent molecular effects of IRF-3 and IRF-7 inhibition (1010 CFU/ml), and diluted as appropriate. Overnight static cultures await further exploration. Third, more complete genotypic profiles of E. coli CFT073 in Luria broth were used for animal inoculations. should be developed, in addition to the identified rs3758650 and rs10902179 SNPs. Susceptibility biomarkers are needed to reduce the Human kidney epithelial cells useofinvasivediagnosticproceduresand antibiotic use in children with The A498 human kidney carcinoma cell line (American Type Culture APN. Finally, specific inhibitors of IRF-7 need to be developed. For ex- Collection HTB-44) is an established model to study UTI pathogenesis ample, the RNA interference (RNAi) approach, used here in mice, could (53). A498 cells were cultured in RPMI 1640 supplemented with 1 mM be translated to humans. Concerns are the efficiency of cellular siRNA sodium pyruvate, 1 mM nonessential amino acids, gentamicin (50 mg/ml)

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 8 RESEARCH ARTICLE

(GE Healthcare), and 10% fetal bovine serum (FBS) (PAA) overnight at (Superfrost/Plus, Thermo Fisher Scientific), fixed in acetone-methanol 37°C, 90% humidity, and 5% CO2 in six-well plates (for Western blots (1:1, 10 min), dried, permeabilized (0.2% Triton X-100, 5% normal goat and RNA extraction) or eight-well chamber slides (for confocal imag- serum/PBS), and stained with primary rat antineutrophil antibody ing) (Thermo Fisher Scientific). Cells were exposed to bacteria in fresh, (NIMP-R14) (1:200; Abcam, ab2557), rabbit polyclonal anti–E. coli serum-free supplemented RPMI. antibody (1:100; Novus Biologicals, NB200-579), mouse anti-STAT3 (1:50; Abcam, ab119352), and rabbit anti–IRF-7 antibodies (1:100; Gene expression in infected cells Abcam, ab62505), followed by Alexa 488– or Alexa 568–labeled rabbit Total RNA was extracted by TRIzol (Invitrogen) followed by QIAGEN anti-rat, goat anti-mouse, and goat anti-rabbit immunoglobulin G sec- RNeasy cleanup (QIAGEN), reverse-transcribed and converted to ondary antibodies (Molecular Probes; A-21210, A-11001, and A-11011). biotin-cRNA using TargetAmp Nano-g Biotin-aRNA Labeling kit Nuclei were counterstained with DAPI (4′,6-diamidino-2-phenylindole) (Epicentre Biotechnologies), hybridized onto an Illumina HumanHT- (0.05 mM; Sigma-Aldrich). Slides were examined by fluorescence mi- 12 Expression BeadChip for 16 hours at 58°C, washed, stained (Illumina croscopy (AX60, Olympus Optical). Richard-Allan Scientific Signature Wash Protocol), and scanned (BeadArray Scanner 500GX). Normal- Series Hematoxylin 7211 and Eosin-Y 7111 (Thermo Scientific) were ized human kidney cell data were of high quality with replicate corre- used to counterstain the tissue sections. For staining and isotype controls, lation >0.98 and no systematic bias. Genes with log2 fold changes see fig. S13. greater than 0.585-fold over buffer control were analyzed by IPA soft-

ware [Ingenuity Systems, QIAGEN (www.qiagen.com/ingenuity)]. Heat Global gene expression in infected kidneys Downloaded from maps were constructed using the free Gitools 2.1 software. Total kidney RNA was extracted from infected and uninfected mice of each genetic background. After physical disruption in lysis buffer using Confocal microscopy Precellys Lysing kits (Bertin Technologies) in a tissue homogenizer Kidney epithelial cells were stained with rabbit anti–IRF-7 primary an- (TissueLyser LT, QIAGEN), RNA was extracted with the RNeasy tibody (Abcam, ab62505) (1:50 in 5% FBS/PBS, overnight at 4°C), Mini Kit (QIAGEN), amplified using GeneChip 3’IVT Express Kit followed by secondary goat anti-rabbit Alexa Fluor 488–conjugated an- (Affymetrix), hybridized onto Mouse Genome 430 PM array strips tibody (Molecular Probes, A-11034) (1:200 in 5% FBS and 0.025% Triton (Affymetrix) (16 hours at 45°C), washed, stained, and scanned in- http://stm.sciencemag.org/ X-100/PBS, 1 hour at room temperature). Cells were counterstained with house using the GeneAtlas System (Affymetrix). Data were normal- DRAQ5 (Abcam) and examined in a LSM 510 META laser scanning ized using Robust Multi Average implemented in the Partek Express confocal microscope (Carl Zeiss). Fluorescence was quantified using Software (54, 55). Significantly altered genes were sorted by relative Photoshop CS5. For staining controls, see fig. S13. expression [two-way ANOVA model using method of moments, P < 0.05, absolute fold change > 1.41 (56)]. Experimental UTI Mice were bred at the Microbiology, Immunology and Glycobiology Nuclear protein extraction (MIG) animal facility, Lund, Sweden,andhousedinspecificpathogen– Kidney epithelial cells were cultured in 75-cm2 flasks (31.6 × 105 cells/ free individually ventilated cages at a constant temperature of 23°C on 75-cm2 flask). For cell fractionation, the NE-PER Nuclear and Cyto- a 12-hour light-dark cycle with lights on at 7:00 a.m., with food and plasmic Extraction Reagents (Thermo Fisher Scientific) were used to on April 28, 2016 − − − − water ad libitum. Female C57BL/6 wild-type, Irf3 / ,andIrf7 / mice separate the cytosolic and nuclear fractions according to the manufac- − − wereusedat9to15weeksofage.Irf3 / mice on a C57BL/6 background turer’sinstructions. − − were provided by T. Taniguchi, University of Tokyo, Japan (18). Irf7 / mice were from the Riken BioResource Center, National BioResource Coimmunoprecipitation and Western blotting Project of Ministry of Education, Culture, Sports, Science and Tech- Nuclear fractions were incubated with rabbit anti–IRF-3 (5 mg/ml; nology, Japan, with permission from T. Taniguchi (18). For the number Santa Cruz, sc-9082) or rabbit anti–IRF-7 (5 mg/ml; Santa Cruz, of mice used in each experiment, see table S5. All animal experiments sc-74471) antibody overnight, and complexes were collected with followed the ARRIVE (Animal Research: Reporting of In Vivo Exper- magnetic Dynabeads Protein G (Life Technologies). Total or de- iments) guidelines for reporting animal preclinical studies. pleted nuclear fractions were run on SDS–polyacrylamide gel elec- Anesthetized mice (isoflurane) were infected by intravesical inocu- trophoresis (4 to 12% bis-tris gels; Invitrogen), blotted onto lation with E. coli CFT073 or E. coli 83972 (108 CFU in 0.1 ml) (30)and polyvinylidene difluoride membranes (GE Healthcare), blocked with sacrificed under anesthesia. Kidneys were aseptically removed, and 5% nonfat dry milk (NFDM), and incubated with rabbit anti–IRF-3 macroscopic pathology was documented by photography. One kidney (1:500 in 5% NFDM) and mouse anti–IRF-7 (1:200 in 5% NFDM). was divided for RNA extraction (on dry ice), histology and immuno- The blots were washed with PBS Tween 0.1% (PBST) and incubated staining [in OCT (optimal cutting temperature) compound; VWR], or with secondary antibodies in 5% NFDM [1:4000, goat anti-rabbit protein extraction (frozen at −80°C). Infection was quantified in the sec- horseradish peroxidase (HRP) (Cell Signaling) or rabbit anti-mouse ond kidney by viable counts of homogenate in 5 ml of PBS. Urine HRP (DAKO)]. The blots were washed with PBST and developed with samples were collected before and at regular times after infection and ECL Plus detection reagent (GE Healthcare). Blots were imaged using quantitatively cultured. Neutrophil numbers were quantified in uncen- the Bio-Rad ChemiDoc System. For unedited Western blot, see fig. S14. trifuged urine by use of a hemocytometer. PCR analysis Histology and immunohistochemistry OAS1, CCL5,andIFNB1 promoter fragments (187, 187, and 143 bp, Tissues embedded and frozen in OCT were cryosectioned (8 mm; respectively) containing predicted IRF-3 and IRF-7 binding sites were Leica microtome), collected on positively charged microscope slides amplified by PCR using 15 ng of total A498 genomic DNA. Forward

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 9 RESEARCH ARTICLE and reverse primer sequences were designed as follows: OAS1,5′- nearby genes, we analyzed SNP and gene expression microarray data GTTCAGAGAAAGGCTGGGCTG-3′ (forward) and 5′-GCTGACT- generated from peripheral blood samples[eightdatasetstotaling973 GAGCTTGGACTGC-3′ (reverse); CCL5,5′-AGATGAGAGAG- individuals from the Icelandic population (57) and 8086 individuals CAGTGAGGGAG-3′ (forward) and 5′-CCTCTGCAGCTCAGG- of other European populations (32)]. Gene expression in the Icelandic CTGGC-3′ (reverse); IFNB1,5′-GATAGGAGCT TAAATAAAGAG- data set was quantified as the mean log10 expression ratio compared to 3′ (forward) and 5′-CCTTTCTCCATGGGTATGGCC-3′ (reverse). pooled reference RNA samples and regressed against the number of risk Thermal cycling conditions were initial denaturation (95°C, 3 min), alleles carried, age, gender, relatedness, and differential white blood cell followed by 35 cycles comprising a denaturation (95°C, 30 s), an anneal- counts. Quantification and regression of gene expression in the other ing (59°C, 30 s), and extension (72°C, 30 s) steps, and a final extension data sets were performed as described (32, 58, 59). (72°C, 6 min). Combined IRF7 sequence and gene expression analysis. Peripheral blood samples were obtained from Swedish healthy blood Electrophoretic mobility shift assay donors (Clinical Immunology and Transfusion Medicine, Lund, Sweden), Amplified DNA sequences from the OAS1, CCL5,andIFNB1 promo- and mononuclear cells were extracted using Lymphoprep (Axis-Shield). ters were used as probes and labeled with GelGreen (Biotium). Each DNA and RNA were isolated using standard kits (QIAGEN and Macherey- reaction contained 1 mgofDNAprobeand2mg of nuclear extract from Nagel, respectively). Genotyping for the IRF7 promoter SNP rs3758650 infected A498 cells in binding buffer [100 mM tris, 500 mM NaCl, and was by pyrosequencing (n = 246).

10 mM DTT (pH 7.0)]. For the band shift/competition assay, 2 mgof RNA was reverse-transcribed using SuperScript III First-Strand Syn- Downloaded from rabbit anti–IRF-3 (Santa Cruz, sc-9082), anti-p300 (Santa Cruz, sc-584 X), thesis System for RT-PCR (Invitrogen). Gene expression was quantified mouse anti–IRF-7 (Santa Cruz, sc-74471), anti–NF-kBp65(SantaCruz, using QuantiTect Primer Assays for IRF7 and GAPDH and SsoAdvanced sc-8008), anti-CBP (Santa Cruz, sc-7300 X), or goat anti–phosphorylated Universal SYBR Green Supermix (Bio-Rad) on a Rotor-Gene Q (QIAGEN). c-Jun (Santa Cruz, sc-16312) were used. Binding reactions were incubated qRT-PCR reactions were run in technical duplicates, and the DCt values at 15°C for 30 min, loaded onto a 6% nondenaturing, nonreducing were obtained after normalizing the average Ct values for IRF7 to the − DDCt polyacrylamide gel, and ran in a 50 mM tris (pH 7.0), 0.38 M glycine, average Ct values for GAPDH.2 was calculated with respect to and2mMEDTAbufferat100Vfor2to3hours.Gelswereimaged themeanofthereferencegenotype. http://stm.sciencemag.org/ using the Bio-Rad ChemiDoc System. For unedited gels, see figs. S15 and S16. Liposomal Irf7 RNAi delivery in vivo − − Irf3 / mice were subjected to siRNA treatment at a dose of 5 mg/kg. A IRF-7 and IRF-3 promoter genotyping dose (300 ml) of Silencer Select Predesigned Irf7 siRNA (Life Technol- IRF7 sequence analysis. The Exome Aggregation Consortium ogies, 4404010 #s79411) was injected with Invivofectamine reagent (Cambridge, MA; http://exac.broadinstitute.org, February 2015) in- (Life Technologies, 1377501) into the tail vein (200 ml) and intravesically cludes exome sequencing data from 60,706 unrelated individuals ex- (100 ml). For preventive reduction of Irf7 expression, Irf7 siRNA was cluding those affected by severe pediatric disease. The Exome Variant injected 3 days before infection and on the day of infection with E. coli Server (NHLBI GO Exome Sequencing Project, Seattle, WA; http://evs. CFT073. For therapeutic Irf7 inhibition, Irf7 siRNA was injected 24 hours gs.washington.edu/EVS/, February 2015) comprises 2203 African- and 3 days after infection. Ambion In Vivo Negative Control siRNA on April 28, 2016 Americans and 4300 European-Americans unrelated individuals, total- was used as a control (Life Technologies, 4457287). For transcriptomic ing 6503 samples. Genomic DNA was extracted from heparinized analysis, total RNA was isolated from the kidneys of Irf7 siRNA–treated − − peripheralbloodofpatientswithAPN,ABU,orhealthycontrols,using Irf3 / mice and control mice using mirVana microRNA isolation kit the QIAamp DNA Blood Midi Kit (QIAGEN) and TrueStart Hot (Ambion by Life Technologies) followed by organic extraction using StartTaq mixture (Thermo Fisher Scientific). acid phenol/chloroform, and analyzed using Mouse Genome 430 PM Study population 1. Children in southern Sweden, with a consistent array strips in a GeneAtlas (Affymetrix). UTI pattern over several years, were identified after at least 5 years of follow-up (33). Two groups were selected; one group had recurrent Antibiotic treatment in vivo − − APN (n = 17), and the second group had ABU, with no prior sympto- After infection with E. coli CFT073, Irf3 / mice were treated with the matic infection (n = 14). Sequenced products (GATC Biotech) were third-generation cephalosporin broad-spectrum antibiotic, cefotaxime analyzed by DNAClub (by X. Chen)andBioEdit7.0(byT.Hall). (STRAGEN Nordic A/S) at a dose of 100 or 400 mg/kg body weight. Study population 2. UTI-pronepatientswereidentifiedafterapro- Cefotaxime was given by intraperitoneal injection twice a day for 6 days spective, long-term follow-up of about 30 years (33). Controls were starting 24 hours after infection. adult healthy blood donors from the same geographic area. The IRF7 promoter SNP rs3758650 was genotyped in patients with ABU (n =43), Statistical analysis APN (n = 34), and controls (n = 41) by pyrosequencing on a PyroMark For confocal staining, integrated density values were analyzed with un- 96MA instrument (QIAGEN). For promoter sequences and PCR pri- paired two-tailed t test. Pathology, bacterial numbers, and neutrophil mers, see figs. S9 and S10. responses were compared by unpaired two-tailed t test or two-way Expression quantitative trait loci determination. The blood eQTL ANOVA. Unpaired two-tailed Student’s t test for homoscedastic browser (32) comprises nontransformed peripheral blood samples from variances was used to evaluate qRT-PCR data using InStat software 5311 individuals and replication analysis in another 2775 samples, to- (version 3.06, GraphPad). Two-tailed c2 test was used to evaluate the taling 8086. genotyping data. Significance was accepted at *P < 0.05, **P < 0.01, or Association of the risk alleles with gene expression. To test for ***P < 0.001. For animal numbers, see table S5. Data were examined associations between identified risk variants and the expression of using Prism (version 6.02, GraphPad).

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 10 RESEARCH ARTICLE

Study approval 11. T. Fujita, J. Sakakibara, Y. Sudo, M. Miyamoto, Y. Kimura, T. Taniguchi, Evidence for a nu- Experiments were approved by the Malmö/Lund Animal Experimental clear factor(s), IRF-1, mediating induction and silencing properties to human IFN-b gene regulatory elements. EMBO J. 7, 3397–3405 (1988). Ethics Committee at the Lund District Court, Sweden (#M104-10 and 12. H. Nguyen, J. Hiscott, P. M. Pitha, The growing family of interferon regulatory factors. Cytokine #M44-13). Procedures followed institutional, national, and European Growth Factor Rev. 8,293–312 (1997). Union guidelines. The studies of human UTI were approved by the 13. Y. Mamane, C. Heylbroeck, P. Génin, M. Algarté, M. J. Servant, C. LePage, C. Deluca, H. Kwon, Ethics Committee of the Medical Faculty, Lund University, Sweden R. Lin, J. Hiscott, Interferon regulatory factors: The next generation. Gene 237,1–14 (1999). 14. W. C. Au, P. A. Moore, W. Lowther, Y. T. Juang, P. M. Pitha, Identification of a member of the (LU106-02, LU236-99, Dnr 298/2006; 463/2010). Patients gave their interferon regulatory factor family that binds to the interferon-stimulated response ele- written informed consent. Blood donor samples were collected subject ment and activates expression of interferon-induced genes. Proc. Natl. Acad. Sci. U.S.A. to ethical approval (Lund University Ethical Review Board, 2013/54; 92, 11657–11661 (1995). Icelandic Data Protection Authority, 2001010157; National Bioethics 15. L. Zhang, J. S. Pagano, IRF-7, a new interferon regulatory factor associated with Epstein- – Committee, 01/015) and informed consent. No individuals were ap- Barr virus latency. Mol. Cell. Biol. 17, 5748 5757 (1997). 16. K. Honda, A. Takaoka, T. Taniguchi, Type I interferon gene induction by the interferon proached solely for the purpose of this study. regulatory factor family of transcription factors. Immunity 25, 349–360 (2006). 17. M. G. Wathelet, C. H. Lin, B. S. Parekh, L. V. Ronco, P. M. Howley, T. Maniatis, Virus infection induces the assembly of coordinately activated transcription factors on the IFN-b enhancer SUPPLEMENTARY MATERIALS in vivo. Mol. Cell 1, 507–518 (1998). 18. M. Sato, H. Suemori, N. Hata, M. Asagiri, K. Ogasawara, K. Nakao, T. Nakaya, M. Katsuki, S. Noguchi, www.sciencetranslationalmedicine.org/cgi/content/full/8/336/336ra59/DC1 N. Tanaka, T. Taniguchi, Distinct and essential roles of transcription factors IRF-3 and IRF-7 in Fig. S1. CFT073-specific gene expression in infected human kidney cells. response to viruses for IFN-a/b gene induction. Immunity 13,539–548 (2000). Downloaded from −/− −/− Fig. S2. Bacterial persistence in Irf3 , Irf7 , and C57BL/6 mice infected with CFT073. 19. K. Honda, H. Yanai, H. Negishi, M. Asagiri, M. Sato, T. Mizutani, N. Shimada, Y. Ohba, A. Takaoka, −/− −/− Fig. S3. Neutrophil recruitment in Irf3 , Irf7 , and C57BL/6 mice infected with CFT073. N. Yoshida, T. Taniguchi, IRF-7 is the master regulator of type-I interferon-dependent immune – Fig. S4. The ABU strain E. coli 83972 does not activate the IRF-7 driven hyperinflammatory responses. Nature 434,772–777 (2005). response in mice. 20. A. Takaoka, T. Tamura, T. Taniguchi, Interferon regulatory factor family of transcription −/− Fig. S5. Pathology-associated activation of the granulocyte adhesion pathway genes in Irf3 factors and regulation of oncogenesis. Cancer Sci. 99, 467–478 (2008). mice. 21. T. Taniguchi, N. Tanaka, K. Ogasawara, S. Taki, M. Sato, A. Takaoka, Transcription factor IRF-1 and −/− Fig. S6. Activation of the CFT073-specific gene network in infected Irf3 mice. its family members in the regulation of host defense. Cold Spring Harb. Symp. Quant. Biol. 64, Fig. S7. Control of mucosal expression of pathology-associated molecules in infected mice. 465–472 (1999). http://stm.sciencemag.org/ Fig. S8. IRF-3 and IRF-7 binding to CCL5 and IFNB1 promoter oligonucleotides. 22. C. Svanborg-Edén, L. A. Hanson, U. Jodal, U. Lindberg, A. S. Ǻkerlund, Variable adherence Fig. S9. IRF7 promoter sequence. to normal human urinary-tract epithelial cells of Escherichia coli strains associated with − Fig. S10. Primers used for amplification of the IRF7 promoter SNP rs3758650 ( 886). various forms of urinary-tract infection. Lancet 2, 490–492 (1976). −/− Fig. S11. Efficacy of Irf7 siRNA treatment in APN-susceptible Irf3 mice. 23. H. Leffler, C. Svanborg-Edén, Chemical-identification of a glycosphingolipid receptor for −/− Fig. S12. Antibiotic treatment of APN-susceptible Irf3 mice. Escherichia coli attaching to human urinary tract epithelial cells and agglutinating human Fig. S13. Staining and isotype controls. erythrocytes. FEMS Microbiol. Lett. 8, 127–134 (1980). Fig. S14. Unedited Western blot for fig. S8. 24. W. W. Agace, S. R. Hedges, M. Ceska, C. Svanborg, Interleukin-8 and the neutrophil re- Fig. S15. Unedited gels for Fig. 4. sponse to mucosal gram-negative infection. J. Clin. Invest. 92, 780–785 (1993). Fig. S16. Unedited gels for fig. S8. 25. J. E. Fowler Jr., T. A. Stamey, Studies of introital colonization in women with recurrent Table S1. Exome Aggregation Consortium IRF7. urinary infections. VII. The role of bacterial adherence. J. Urol. 117, 472–476 (1977). Table S2. Exome Variant Server variant IRF7. 26. H. Lomberg, L. Ǻ. Hanson, B. Jacobsson, U. Jodal, H. Leffler, C. Svanborg-Edén, Correlation Table S3. Westra cis-eQTL IRF7.

of P blood group phenotype, vesicoureteral reflux and bacterial attachment in patients on April 28, 2016 Table S4. Association between the identified variants and IRF7 expression in the Icelandic with recurrent pyelonephritis. N. Engl. J. Med. 308, 1189–1192 (1983). population. 27. A. Stapleton, E. Nudelman, H. Clausen, S. Hakomori, W. E. Stamm, Binding of uropatho- Table S5. Number of mice used in inoculation experiments. genic Escherichia coli R45 to glycolipids extracted from vaginal epithelial cells is dependent on histo-blood group secretor status. J. Clin. Invest. 90, 965–972 (1992). 28. H.L.Mobley,D.M.Green,A.L.Trifillis,D.E.Johnson,G.R.Chippendale,C.V.Lockatell,B.D.Jones, REFERENCES AND NOTES J. W. Warren, Pyelonephritogenic Escherichia coli and killing of cultured human renal proximal tubular epithelial cells: Role of hemolysin in some strains. Infect. Immun. 58, 1281–1289 (1990). 1. B. Beutler, Innate immunity: An overview. Mol. Immunol. 40, 845–859 (2004). 29. R. A. Welch, V. Burland, G. Plunkett III, P. Redford, P. Roesch, D. Rasko, E. L. Buckles, S.-R. Liou, 2. D. Ferrandon, J.-L. Imler, C. Hetru, J. A. Hoffmann, The Drosophila systemic immune re- A. Boutin, J. Hackett, D. Stroud, G. F. Mayhew, D. J. Rose, S. Zhou, D. C. Schwartz, N. T. Perna, sponse: Sensing and signalling during bacterial and fungal infections. Nat. Rev. Immunol. H. L. T. Mobley, M. S. Donnenberg, F. R. Blattner, Extensive mosaic structure revealed by the 7, 862–874 (2007). complete genome sequence of uropathogenic Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 99, 3. H. Kumar, T. Kawai, S. Akira, Pathogen recognition by the innate immune system. Int. Rev. 17020–17024 (2002). Immunol. 30,16–34 (2011). 30. C. Svanborg Edén, R. Freter, L. Hagberg, R. Hull, S. Hull, H. Leffler, G. Schoolnik, Inhibition of 4. B. Ragnarsdóttir, N. Lutay, J. Grönberg-Hernandez, B. Köves, C. Svanborg, Genetics of in- experimental ascending urinary tract infection by an epithelial cell-surface receptor ana- nate immunity and UTI susceptibility. Nat. Rev. Urol. 8, 449–468 (2011). logue. Nature 298, 560–562 (1982). 5. C. A. Dinarello, A. Simon, J. W. M. van Der Meer, Treating inflammation by blocking interleukin-1 in 31. W. Suhara, M. Yoneyama, T. Iwamura, S. Yoshimura, K. Tamura, H. Namiki, S. Aimoto, T. Fujita, a broad spectrum of diseases. Nat. Rev. Drug Discov. 11, 633–652 (2012). Analyses of virus-induced homomeric and heteromeric protein associations between IRF-3 6. M. Feldmann, R. N. Maini, Anti-TNF therapy, from rationale to standard of care: What and coactivator CBP/p300. J. Biochem. 128, 301–307 (2000). lessons has it taught us? J. Immunol. 185, 791–794 (2010). 32. H.-J. Westra, M. J. Peters, T. Esko, H. Yaghootkar, C. Schurmann, J. Kettunen, M. W. Christiansen, 7. S. Kang, T. Tanaka, T. Kishimoto, Therapeutic uses of anti-interleukin-6 receptor antibody. B. P. Fairfax, K. Schramm, J. E. Powell, A. Zhernakova, D. V. Zhernakova, J. H. Veldink, L. H. Van Int. Immunol. 27,21–29 (2015). Den Berg, J. Karjalainen, S. Withoff, A. G. Uitterlinden, A. Hofman, F. Rivadeneira, P. A. C t’Hoen, 8. R. Medzhitov, T. Horng, Transcriptional control of the inflammatory response. Nat. Rev. E.Reinmaa,K.Fischer,M.Nelis,L.Milani,D.Melzer,L.Ferrucci,A.B.Singleton,D.G.Hernandez, Immunol. 9, 692–703 (2009). M. A. Nalls, G. Homuth, M. Nauck, D. Radke, U. Völker, M. Perola, V. Salomaa, J. Brody, A. Suchy-Dicey, 9. H. Fischer, N. Lutay, B. Ragnarsdóttir, M. Yadav, K. Jönsson, A. Urbano, A. Al Hadad, S. Rämisch, S. A. Gharib, D. A. Enquobahrie, T. Lumley, G. W. Montgomery, S. Makino, H. Prokisch, C. Herder, P. Storm, U. Dobrindt, E. Salvador, D. Karpman, U. Jodal, C. Svanborg, Pathogen specific, IRF3- M. Roden, H. Grallert, T. Meitinger, K. Strauch, Y. Li, R. C. Jansen, P. M. Visscher, J. C. Knight, dependent signaling and innate resistance to human kidney infection. PLOS Pathog. 6, B. M. Psaty, S. Ripatti, A. Teumer, T. M. Frayling, A. Metspalu, J. B. van Meurs, L. Franke, e1001109 (2010). Systematic identification of trans eQTLs as putative drivers of known disease associations. 10. M. Miyamoto, T. Fujita, Y. Kimura, M. Maruyama, H. Harada, Y. Sudo, T. Miyata, T. Taniguchi, Nat. Genet. 45, 1238–1243 (2013). Regulated expression of a gene encoding a nuclear factor, IRF-1, that specifically binds to 33. B.Ragnarsdóttir,K.Jönsson,A.Urbano,J.Grönberg-Hernandez,N.Lutay,M.Tammi,M.Gustafsson, IFN-b gene regulatory elements. Cell 54, 903–913 (1988). A.-C. Lundstedt, I. Leijonhufvud, D. Karpman, B. Wullt, L. Truedsson, U. Jodal, B. Andersson,

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 11 RESEARCH ARTICLE

C. Svanborg, Toll-like receptor 4 promoter polymorphisms: Common TLR4 variants may 51. F. Sundén, L. Håkansson, E. Ljunggren, B. Wullt, Escherichia coli 83972 bacteriuria protects protect against severe urinary tract infection. PLOS One 5, e10734 (2010). against recurrent lower urinary tract infections in patients with incomplete bladder emp- 34. J. Chang, R. J. Lindsay, S. Kulkarni, J. D. Lifson, M. Carrington, M. Altfeld, Polymorphisms in tying. J. Urol. 184, 179–185 (2010). interferon regulatory factor 7 reduce interferon-a responses of plasmacytoid dendritic 52. J. Zdziarski, E. Brzuszkiewicz, B. Wullt, H. Liesegang, D. Biran, B. Voigt, J. Grönberg-Hernandez, cells to HIV-1. AIDS 25, 715–717 (2011). B. Ragnarsdottir, M. Hecker, E. Z. Ron, R. Daniel, G. Gottschalk, J. Hacker, C. Svanborg, U. Dobrindt, 35. A. Kawasaki, H. Furukawa, Y. Kondo, S. Ito, T. Hayashi, M. Kusaoi, I. Matsumoto, S. Tohma, Y. Takasaki, Host imprints on bacterial genomes—Rapid, divergent evolution in individual patients. H. Hashimoto, T. Sumida, N. Tsuchiya, Association of PHRF1-IRF7 region polymorphism with PLOS Pathog. 6, e1001078 (2010). clinical manifestations of systemic lupus erythematosus in a Japanese population. Lupus 21, 53. M. Hedlund, M. Svensson, A. Nilsson, R. D. Duan, C. Svanborg, Role of the ceramide- 890–895 (2012). signaling pathway in cytokine responses to P-fimbriated Escherichia coli. J. Exp. Med. 36. L.-H. Lin, P. Ling, M.-F. Liu, The potential role of interferon-regulatory factor 7 among Tai- 183, 1037–1044 (1996). wanese patients with systemic lupus erythematosus. J. Rheumatol. 38, 1914–1919 (2011). 54. R. A. Irizarry, B. M. Bolstad, F. Collin, L. M. Cope, B. Hobbs, T. P. Speed, Summaries of Affy- 37. The International Consortium for Systemic Lupus Erythematosus Genetics (SLEGEN), J. B. Harley, metrix GeneChip probe level data. Nucleic Acids Res. 31, e15 (2003). M. E. Alarcón-Riquelme, L. A. Criswell, C. O. Jacob, R. P. Kimberly, K. L. Moser, B. P. Tsao, T. J. Vyse, 55. B. M. Bolstad, R. A. Irizarry, M. Ǻstrand, T. P. Speed, A comparison of normalization methods C. D. Langefeld, S. K. Nath, J. M. Guthridge, B. L. Cobb, D. B. Mirel, M. C. Marion, A. H. Williams, for high density oligonucleotide array data based on variance and bias. Bioinformatics 19, J. Divers, W. Wang, S. G. Frank, B. Namjou, S. B. Gabriel, A. T. Lee, P. K. Gregersen, T. W. Behrens, 185–193 (2003). K. E. Taylor, M. Fernando, R. Zidovetzki, P. M. Gaffney, J. C. Edberg, J. D. Rioux, J. O. Ojwang, 56. C. Eisenhart, The assumptions underlying the analysis of variance. Biometrics 3,1–21 (1947). J. A. James, J. T. Merrill, G. S. Gilkeson, M. F. Seldin, H. Yin, E. C. Baechler, Q. Z. Li, E. K. Wakeland, 57. V. Emilsson, G. Thorleifsson, B. Zhang, A. S. Leonardson, F. Zink, J. Zhu, S. Carlson, A. Helgason, G.R.Bruner,K.M.Kaufman,J.A.Kelly,Genome-wideassociationscaninwomenwithsystemic G. B. Walters, S. Gunnarsdottir, M. Mouy, V. Steinthorsdottir, G. H. Eiriksdottir, G. Bjornsdottir, lupus erythematosus identifies susceptibility variants in ITGAM, PXK, KIAA1542 and other loci. I. Reynisdottir, D. Gudbjartsson, A. Helgadottir, A. Jonasdottir, A. Jonasdottir, U. Styrkarsdottir, Nat. Genet. 40, 204–210 (2008). S.Gretarsdottir,K.P.Magnusson,H.Stefansson,R.Fossdal,K.Kristjansson,H.G.Gislason,T.Stefansson,

38. F. D. Carmona, R. Gutala, C. P. Simeón, P. Carreira, N. Ortego-Centeno, E. Vicente-Rabaneda, B. G. Leifsson, U. Thorsteinsdottir, J. R. Lamb, J. R. Gulcher, M. L. Reitman, A. Kong, E. E. Schadt, Downloaded from F. J. Garcia-Hernández, P. García de la Peña, M. Fernández-Castro, L. Martínez-Estupiñán, K. Stefansson, Genetics of gene expression and its effect on disease. Nature 452,423–428 M. V. Egurbide, B. P. Tsao, P. Gourh, S. K. Agarwal, S. Assassi, M. D. Mayes, F. C. Arnett, F. K. Tan, (2008). J. Martín, Spanish Scleroderma Group, Novel identification of the IRF7 region as an anticen- 58. The International HapMap 3 Consortium, Integrating common and rare genetic variation tromere autoantibody propensity locus in systemic sclerosis. Ann. Rheum. Dis. 71,114–119 in diverse human populations. Nature 467,52–58 (2010). (2012). 59.T.Lappalainen,M.Sammeth,M.R.Friedländer,P.A.C.t’Hoen,J.Monlong,M.A.Rivas, 39.M.Dorner,J.A.Horwitz,B.M.Donovan,R.N.Labitt,W.C.Budell,T.Friling,A.Vogt,M.T.Catanese, M. Gonzàlez-Porta, N. Kurbatova, T. Griebel, P. G. Ferreira, M. Barann, T. Wieland, L. Greger, T. Satoh, T. Kawai, S. Akira, M. Law, C. M. Rice, A. Ploss, Completion of the entire hepatitis C M. van Iterson, J. Almlöf, P. Ribeca, I. Pulyakhina, D. Esser, T. Giger, A. Tikhonov, M. Sultan, virus life cycle in genetically humanized mice. Nature 501, 237–241 (2013). G. Bertier, D. G. MacArthur, M. Lek, E. Lizano, H. P. J. Buermans, I. Padioleau, T. Schwarzmayr, http://stm.sciencemag.org/ 40. K. Honda, T. Taniguchi, Toll-like receptor signaling and IRF transcription factors. IUBMB Life O. Karlberg, H. Ongen, H. Kilpinen, S. Beltran, M. Gut, K. Kahlem, V. Amstislavskiy, O. Stegle, 58, 290–295 (2006). M. Pirinen, S. B. Montgomery, P. Donnelly, M. I. McCarthy, P. Flicek, T. M. Strom, The Geudavis 41. S. Ning, J. S. Pagano, G. N. Barber, IRF7: Activation, regulation, modification and function. Consortium, H. Lehrach, S. Schreiber, R. Sudbrak, Á. Carracedo, S. E. Antonarakis, R. Häsler, Genes Immun. 12, 399–414 (2011). A.-C.Syvänen,G.-J.vanOmmen,A.Brazma,T.Meitinger,P.Rosenstiel,R.Guigó,I.G.Gut,X.Estivill, 42. J. R. Johnson, P. L. Roberts, W. E. Stamm, P fimbriae and other virulence factors in Escherichia E. T. Dermitzakis, Transcriptome and genome sequencing uncovers functional variation in humans. coli urosepsis: Association with patients’ characteristics. J. Infect. Dis. 156, 225–229 (1987). Nature 501,506–511 (2013). 43. W. Li, M. J. Hofer, A. L. Noçon, P. Manders, I. L. Campbell, Interferon regulatory factor 7 (IRF7) is required for the optimal initial control but not subsequent clearance of lympho- Acknowledgments: We thank U. Thorsteinsdottir, G. Thorleifsson, and K. Stefansson (deCODE ge- cytic choriomeningitis virus infection in mice. Virology 439, 152–162 (2013). netics) for the data associating risk alleles with gene expression in the Icelandic population. Funding: 44. H.-W. Chen, K. King, J. Tu, M. Sanchez, A. D. Luster, S. Shresta, The roles of IRF-3 and IRF-7 in We gratefully acknowledge the support of the Swedish Research Council, Swedish Cancer Society, innate antiviral immunity against dengue virus. J. Immunol. 191, 4194–4201 (2013). Söderberg Foundation, Knut and Alice Wallenberg Foundation, Österlund and Sharon D. Lund Foun-

45. J. Lieberman, P. A. Sharp, Harnessing RNA interference for therapy: The silent treatment. dation, Anna-Lisa and Sven-Erik Lundgren Foundation, Maggie Stephens Foundation, H. J. Forssman on April 28, 2016 JAMA 313, 1207–1208 (2015). Foundation, Medical Faculty of Lund University, Royal Physiographic Society, Network of Excellence, 46. T. Coelho, D. Adams, A. Silva, P. Lozeron, P. N. Hawkins, T. Mant, J. Perez, J. Chiesa, S. Warrington, Infect-ERA, and ALF (Avtal om Läkarutbildning och Forskning) grants from the Medical Faculty and E. Tranter, M. Munisamy, R. Falzone, J. Harrop, J. Cehelsky, B. R. Bettencourt, M. Geissler, J. S. Butler, Regional Laboratories (Labmedicin Skåne). Author contributions: I.A., M.P., C.C., D.B., Y.H., N.L., G.R., A.Sehgal,R.E.Meyers,Q.Chen,T.Borland,R.M.Hutabarat,V.A.Clausen,R.Alvarez,K.Fitzgerald, B.G., B.S., A.N., B.N., and C.S. designed the study, performed the experiments, analyzed the data, and C.Gamba-Vitalo,S.V.Nochur,A.K.Vaishnaw,D.W.Y.Sah,J.A.Gollob,O.B.Suhr,Safetyand wrote the paper. Specifically, M.P. performed the animal experiments and performed immunohisto- efficacy of RNAi therapy for transthyretin amyloidosis. N.Engl.J.Med.369, 819–829 (2013). chemistry together with D.B. I.A. and B.S. performed the bioinformatics analyses, and Y.H., B.G., C.C., 47.J.K.Nair,J.L.S.Willoughby,A.Chan,K.Charisse,M.R.Alam,Q.Wang,M.Hoekstra,P.Kandasamy, and B.N. performed the human genetic analysis. C.C. and A.N. performed the chromatin interaction A. V. Kel’in, S. Milstein, N. Taneja, J. O’shea, S. Shaikh, L. Zhang, R. J. van der Sluis, M. E. Jung, experiment. Statistical analyses were performed by M.P. (animal data), I.A. (transcriptomics), and Y.H. A. Akinc, R. Hutabarat, S. Kuchimanchi, K. Fitzgerald, T. Zimmermann, T. J. C. van Berkel, M. A. Maier, (genotyping). Competing interests: The authors declare that they have no competing interests. A K. G. Rajeev, M. Manoharan, Multivalent N-acetylgalactosamine-conjugated siRNA localizes in patent has been filed to protect the use of IRF-7 siRNA–based therapy. hepatocytes and elicits robust RNAi-mediated gene silencing. J. Am. Chem. Soc. 136, 16958–16961 (2014). Submitted 23 December 2015 48. H. L. A. Janssen, H. W. Reesink, E. J. Lawitz, S. Zeuzem, M. Rodriguez-Torres, K. Patel, A. J. van der Meer, Accepted 8 March 2016 A. K. Patick, A. Chen, Y. Zhou, R. Persson, B. D. King, S. Kauppinen, A. A. Levin, M. R. Hodges, Published 27 April 2016 Treatment of HCV infection by targeting microRNA. N. Engl. J. Med. 368, 1685–1694 (2013). 10.1126/scitranslmed.aaf1156 49. G. R. Nielubowicz, H. L. T. Mobley, Host–pathogen interactions in urinary tract infection. Nat. Rev. Urol. 7, 430–441 (2010). Citation: M. Puthia, I. Ambite, C. Cafaro, D. Butler, Y. Huang, N. Lutay, G. Rydström, B. Gullstrand, 50. G. Bergsten, M. Samuelsson, B. Wullt, I. Leijonhufvud, H. Fischer, C. Svanborg, PapG- B. Swaminathan, A. Nadeem, B. Nilsson, C. Svanborg, IRF7 inhibition prevents destructive innate dependent adherence breaks mucosal inertia and triggers the innate host response. immunity—A target for nonantibiotic therapy of bacterial infections. Sci. Transl. Med. 8, 336ra59 J. Infect. Dis. 189, 1734–1742 (2004). (2016).

www.ScienceTranslationalMedicine.org 27 April 2016 Vol 8 Issue 336 336ra59 12 IRF7 inhibition prevents destructive innate immunity−−A target for nonantibiotic therapy of bacterial infections Manoj Puthia, Ines Ambite, Caterina Cafaro, Daniel Butler, Yujing Huang, Nataliya Lutay, Gustav Rydström, Birgitta Gullstrand, Bhairavi Swaminathan, Aftab Nadeem, Björn Nilsson and Catharina Svanborg (April 27, 2016) Science Translational Medicine 8 (336), 336ra59. [doi: 10.1126/scitranslmed.aaf1156] Editor's Summary

Beyond antibiotics Antibiotics are the cornerstone of antibacterial therapy; however, the increasing development of resistant bacterial strains stresses the need for alternate approaches. Now, Puthia et al. target bacterial infection by boosting innate immunity. They found that a tight balance between the transcription factor interferon regulatory factor 7 (IRF-7) and its heterodimeric partner IRF-3 is critical for an efficient and self-limiting innate immune response to bacterial infection. Dysregulation of this balance contributed to kidney pathology in infected mice, and IRF7 attenuating polymorphisms associate with recurrent Downloaded from pyelonephritis in human children. Indeed, targeting IRF-7 protected mice against infection and renal tissue damage, suggesting that IRF7 could be a therapeutic target for protection against bacterial infection.

http://stm.sciencemag.org/

The following resources related to this article are available online at http://stm.sciencemag.org. This information is current as of April 28, 2016.

Article Tools Visit the online version of this article to access the personalization and on April 28, 2016 article tools: http://stm.sciencemag.org/content/8/336/336ra59 Supplemental "Supplementary Materials" Materials http://stm.sciencemag.org/content/suppl/2016/04/25/8.336.336ra59.DC1 Related Content The editors suggest related resources on Science's sites: http://stm.sciencemag.org/content/scitransmed/6/235/235ra61.full http://stm.sciencemag.org/content/scitransmed/7/307/307ra154.full Permissions Obtain information about reproducing this article: http://www.sciencemag.org/about/permissions.dtl

Science Translational Medicine (print ISSN 1946-6234; online ISSN 1946-6242) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Copyright 2016 by the American Association for the Advancement of Science; all rights reserved. The title Science Translational Medicine is a registered trademark of AAAS.