Leptospira Interrogans Pathogen, Control the Clearance of The

TLR4- and TLR2-Mediated B Cell Responses Control the Clearance of the Bacterial Pathogen, Leptospira interrogans

This information is current as Cécilia Chassin, Mathieu Picardeau, Jean-Michel Goujon, of September 25, 2021. Pascale Bourhy, Nathalie Quellard, Sylvie Darche, Edgar Badell, Martine Fanton d'Andon, Nathalie Winter, Sonia Lacroix-Lamandé, Dominique Buzoni-Gatel, Alain Vandewalle and Catherine Werts

J Immunol 2009; 183:2669-2677; Prepublished online 27 Downloaded from July 2009; doi: 10.4049/jimmunol.0900506 http://www.jimmunol.org/content/183/4/2669 http://www.jimmunol.org/ Supplementary http://www.jimmunol.org/content/suppl/2009/07/27/jimmunol.090050 Material 6.DC1 References This article cites 31 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/183/4/2669.full#ref-list-1

Why The JI? Submit online. by guest on September 25, 2021

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2009 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

TLR4- and TLR2-Mediated B Cell Responses Control the Clearance of the Bacterial Pathogen, Leptospira interrogans1

Ce´cilia Chassin,* Mathieu Picardeau,† Jean-Michel Goujon,‡ Pascale Bourhy,† Nathalie Quellard,‡ Sylvie Darche,§ Edgar Badell,¶ Martine Fanton d’Andon,#** Nathalie Winter,¶ Sonia Lacroix-Lamande´,ʈ#** Dominique Buzoni-Gatel,ʈ Alain Vandewalle,* and Catherine Werts2#**

Leptospirosis is a widespread zoonosis caused by pathogenic Leptospira interrogans that are transmitted by asymptomatic infected rodents. Leptospiral lipoproteins and LPS have been shown to stimulate murine cells via TLRs 2 and 4. Host defense mechanisms remain obscure, although TLR4 has been shown to be involved in clearing Leptospira. In this study, we show that double (TLR2

and TLR4) knockout (DKO) mice rapidly died from severe hepatic and renal failure following Leptospira inoculation. Strikingly, Downloaded from the severe proinflammatory response detected in the liver and kidney from Leptospira-infected DKO mice appears to be independent of MyD88, the main adaptor of TLRs. Infection of chimeric mice constructed with wild-type and DKO mice, and infection of several lines of transgenic mice devoid of T and/or B lymphocytes, identified B cells as the crucial lymphocyte subset responsible for the clearance of Leptospira, through the early production of specific TLR4-dependent anti-Leptospira IgMs elicited against the leptospiral LPS. We also found a protective tissue compartmentalized TLR2/TLR4-mediated production of IFN-␥ by B and T

lymphocytes, in the liver and kidney, respectively. In contrast, the tissue inflammation observed in Leptospira-infected DKO mice http://www.jimmunol.org/ was further characterized to be mostly due to B lymphocytes in the liver and T cells in the kidney. Altogether these findings demonstrate that TLR2 and TLR4 play a key role in the early control of leptospirosis, but do not directly trigger the inflammation induced by pathogenic Leptospira. The Journal of Immunology, 2009, 183: 2669–2677.

eptospirosis is a worldwide zoonosis caused by the highly infectious disease (1). Vaccines with attenuated strains are avail- invasive Leptospira species belonging to the spirochete able in some countries, but the humoral protection is strain specific L phylum. Rodents infected by Leptospira are asymptom- and not long-lasting. Therefore, identification of mechanisms that atic carriers, and shed bacteria in their urine into water or soil, thus control the inflammatory responses triggered by pathogenic Lep- infecting human beings via skin contact. Bacteria disseminate tospira should provide a molecular basis for a better understanding by guest on September 25, 2021 through the systemic circulation to colonize the target organs: of the pathophysiology of leptospirosis. liver, lungs, and kidneys. The clinical symptoms may vary widely TLRs are pattern recognition receptors that recognize conserved from almost asymptomatic or mild manifestations, to very severe microbial components, such as LPS, lipopeptides, or hypomethy- forms with multiorgan failure (Weil’s syndrome). Antibiotic treat- lated bacterial DNA, and trigger an innate response to clear mi- ment is effective in most cases during the early stages of the in- crobes from the infected host. We have shown that leptospiral LPS fection, but untreated infections may lead to tissue damage, such as signals by TLR2 in human cells, but is recognized by both TLR4 tubulo-interstitial nephritis. Occupational or recreational outdoor and TLR2 in murine cells, due to a differential recognition of the activities resulting in close contact with animals or flood water atypical lipid A moiety of the leptospiral LPS, which is recognized increase the risk of leptospirosis in humans. Endemic and epidemic by TLR4 in murine, but not in human cells (2–4). Leptospiral outbreaks of leptospirosis are currently causing significant mor- lipoproteins, in contrast, are recognized by TLR2 in murine kidney bidity and mortality, suggesting that leptospirosis is a re-emerging epithelial cells (5). Young C3H/HeJ mice, carrying a mutation in- activating the trl4 gene, are susceptible to leptospirosis (6), and a recent study has shown that TLR4 is indeed involved in the clear- *INSERM, U773, Centre de Recherche Biome´dicale Bichat-Beaujon (CRB3), BP ance of Leptospira (7). However, the mechanisms involved in the 416, Paris France; Universite´Paris 7-Denis Diderot, Site Bichat, Paris, Paris, France; increased susceptibility to Leptospira associated with TLR4 defi- ‡Centre Hospitalier Universitaire de Poitiers, Unite´de Pathologie Ultrastructurale et ʈ ciency have not been explored. Expe´rimentale, Universite´de Poitiers, Poitiers, France; Institut National de Recher- che Agronomique, U1282, Infectiologie Animale et Sante´Publique, Nouzilly, France; Based on the studies described above, we hypothesized that double Institut Pasteur, †Unite´Postulante Biologie des Spiroche`tes, §Unite´deRe´ponses Pre´- (TLR2 and TLR4) knockout (DKO)3 mice should be more suscepti- coces aux Parasites et Immunopathologie, ¶Unite´Geńe´tique Mycobacte´rienne, #G5 Biologie et Geńe´tique de la Paroi Bacte´rienne; and **INSERM, Groupe Avenir, Paris, ble to Leptospira infection than their resistant C57BL/6J wild-type France (WT) counterparts. In this study, we investigated the consequence of Received for publication February 13, 2009. Accepted for publication June 17, 2009. leptospiral infection in adult TLR2- and/or TLR4-deficient mice and The costs of publication of this article were defrayed in part by the payment of page WT mice, taking into account several parameters including lethality, charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by Institut Pasteur and INSERM research grants. C.C. was 3 Abbreviations used in this paper used: DKO, double TLR2 and TLR4 knockout; sponsored by the French Ministe`redelaDe´fense (PhD grant DGA/MRIS). WT, wild type; BUN, blood urea nitrogen; ASAT, aspartate aminotransferase activity; 2 Address correspondence and reprint requests to Dr. C. Werts, Institut Pasteur, G5 iNOS, inducible NO synthase. Biologie et Geńe´tique de la Paroi Bacte´rienne, INSERM Groupe Avenir, 28 Rue du Dr Roux, 75724 Paris Cedex 15, France. E-mail address: [email protected] Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.0900506 2670 TLR4-DEPENDENT B CELL RESPONSES PROTECT AGAINST LEPTOSPIROSIS bacterial load, and the expression of proinflammatory mediators in the sue sections were stained with H&E to evaluate inflammatory changes target organs (liver, kidneys and lungs). We demonstrate that DKO by light microscopy. Immunohistochemical studies, using an Ab raised mice are highly susceptible to pathogenic Leptospira, and die rapidly against outer membrane lipoprotein LipL32 (1/4000), were performed using avidin-biotin coupled to peroxidase substrate kits (Vector Laboratories) from severe renal and hepatic failure, thereby constituting a model of according to the manufacturer’s instructions. Peroxidase activity was re- human leptospirosis in its acute form. Using chimeric mice and trans- vealed with diaminobenzidine (brown reaction) (Dako REAL Detection genic mice, B cells were identified as the crucial lymphocyte cell System). subset involved in the clearance of Leptospira, through TLR4-depen- Transmission electron microscopy dent mechanisms. In contrast, the severe inflammation induced by Leptospira in the organs of susceptible mice is not dependent on TLR Kidney and liver samples were fixed for 30 min in 2.5% glutaraldehyde in 0.1 M cacodylate buffer, embedded in Epon, and processed for transmis- signaling. This implies that TLR4 and TLR2 are crucial factors in host sion electron microscopy by standard procedures. protection rather than playing any direct role in driving inflammation. Real-time and reverse transcription PCR Materials and Methods Total RNA was extracted from whole organs using the RNeasy mini kit Media and reagents (Qiagen), and reverse-transcribed using Moloney murine leukemia virus reverse transcriptase (Invitrogen). cDNA was subjected to real-time PCR, Endotoxin-free PBS was from BioWhittaker. All cell culture products were using a Chromo IV sequence detector (Bio-Rad) as described (11). PCR from Invitrogen. Repurified Escherichia coli LPS and Pam3CysSK were 4 data were reported as the relative increase in mRNA transcripts vs that from Invivogen. Leptospiral LPS was purified according to Ref. 2. found in respective tissues from naive mice, corrected by the respective Mice levels of ␤-actin mRNA, used as an internal standard. The sequences of 4 primers used are depicted in supplemental Table III. Downloaded from Female C57BL/6/J mice (8- to 10-wk old) were purchased from Janvier and were used as control mice. Mice deficient for TLR2 (tlr2Ϫ/Ϫ) or TLR4 ELISA (tlr4Ϫ/Ϫ), originally given by Shizuo Akira (Osaka University, Osaka, Ja- pan), have been further backcrossed eight times into C57BL/6J mice, be- Liver or kidney pieces were weighed and homogenized (100 mg/ml) in a fore being crossed and genotyped to get double tlr2Ϫ/Ϫ/tlr4Ϫ/Ϫ deficient buffer containing 200 mM NaCl, 10 mM Tris HCl, 5 mM EDTA, 10% mice (DKO) (4). MyD88-, Rag2-, CD3-deficient mice, ␮MT mice, all with glycerol, 1 mM PMSF, with complete anti-protease mixture tablets ϫ a C57BL/6J background, were provided by Michel Chignard, James Di (Roche). After centrifugation (14000 g, 10 min), supernatants were kept

Ϫ ␤ http://www.jimmunol.org/ Santo, Armelle Phalipon and Claude Leclerc, respectively (Institut Pas- at 80°C until use. The production of TNF, IL-1 , and RANTES in the teur). All protocols were reviewed by the Institut Pasteur, the competent supernatants was determined using ELISA kits (R&D Systems) according authority, for compliance with the French and European regulations on to the manufacturer’s instructions. Animal Welfare and with Public Health Service recommendations. Plasma biochemical analysis Leptospira and infection experiments Blood samples were collected by retro-orbital eye puncture into tubes con- Leptospira interrogans serovar Copenhageni strain Fiocruz L1–130 was taining 50 ␮l of heparin (Choay). The samples were centrifuged (2500 rpm, maintained virulent by iterative passage in young guinea pigs as described 5 min), and the plasma was stored at Ϫ80°C. Total bilirubinemia, aspartate elsewhere (4). Just before infection, the bacteria were washed in endotoxin- aminotransferase, alanine aminotransferase, blood urea nitrogen (BUN), free PBS, counted using a Petroff-Hauser chamber and resuspended in and serum creatinine were measured in plasma samples using an Olympus

PBS. The mice were infected by the i.p. injection of Leptospira in 300 ␮l AU400 autoanalyzer. by guest on September 25, 2021 of PBS. The mice were sacrificed and bled 3 days postinfection. The kidneys, lungs, and liver of infected and naive mice were removed, and either Leukocyte purifications Ϫ frozen in liquid nitrogen and further stored at 80°C for nucleic acid Leukocytes were purified from 200 ␮l of blood. After three washes in preparations, or fixed in PFA 4% overnight for the immunohistochemical PBS-EDTA 2 mM-SVF 0.5%, RBC were eliminated by incubating sam- analyses. ples in 2 ml of red blood lysis buffer (Sigma) for 10 min at 37°C. The Bacterial loads leukocytes were then washed in PBS and counted. The leptospiral burden of infected organs was determined by real time FACS analysis DNA PCR as previously described by Ref. 8 using a light cycler Fast Start Fc receptors were blocked for 10 min on ice with anti-mouse CD16-CD32 Master SYBR green I kit (Roche Applied Science) following the manu- (1/100 mouse Fc block, clone 2.4G2 from BD Pharmingen), and then facturer’s instructions. Total DNA was extracted from 20 mg of infected stained for 15 min on ice with the following Abs from BD Pharmingen organs using the DNAeasy tissue Qiagen kit. A counted number of heat (1/200) directed against mouse differentiation markers; CD4-allophycocya- killed L. interrogans was used as a calibrator in the real time PCR. Con- nin, CD8␣-PerCP, TCR␤-PE, LY6-G-FITC, and CD11c-PE, CD19-allo- centration of target DNA was adjusted to 200 ng per reaction. Specific phycocyanin, CD3-PerCP. The cells were washed twice, and fixed in 2% primers of RNA 16S gene of Leptospira (LA (38–57) 5Ј-GGCGGC Ј Ј paraformaldehyde for subsequent FACS analysis on a FACSCalibur. Data GCGTCTAAACATG-3 and (LB (348–368) 5 -TTCCCCCCATTGAGC were analyzed using FlowJo 6.4 software (Tree Star). AAGATT-3Ј) were used (0.5 ␮M each). Primers specific for the murine housekeeping gene nidogen (NidF 5Ј-CACCCAGCTTCGGCTCAGTA-3Ј Anti-leptospira ELISA and NidR 5ЈACATTCCCCAGGCCATCGGT-3Ј) were used as an internal standard to normalize the results as described (9). Results were expressed Leptospira “Ag” was prepared as follows: 1010 Leptospira interrogans as the number of Leptospira per 200 ng of organ DNA. Fiocruz in 20 ml of EMJH medium were killed by adding 40 ␮lof37% formaldehyde for2hatroom temperature, and heating to 100°C for 30 Hematopoietic chimeric mice min. The pH was adjusted to 9.5. Maxisorb 96-wells plates (Nunc) were coated overnight at room temperature with 50 ␮lofLeptospira Ag, blocked Eight-week-old mice were irradiated (1000 rad/mouse) to deplete the bone for1hatroom temperature with 100 ␮l of PBS-10% SVF, and washed marrow progenitors, and then reconstituted on the same day by i.v. retro- with PBS-0.005% Tween 20 (Sigma-Aldrich). Successively doubled serum orbital injection of 10 ϫ 106 bone marrow white cells obtained from dilutions were added to the plates for2hatroom temperature. After rins- flushed femurs and tibias of donor mice, washed in PBS. The mice were ing, goat anti-mouse IgM or sheep whole anti-mouse Ig (1/2000, Caltag), then left for 6 wk to let the extraneous bone marrow cells replenish the both coupled to biotin, was added to the plates. After incubating for 1 h, the hematopoietic compartment before infection. Whole blood DNA was pre- plates were rinsed and streptavidin-HRP (1/200, R&D systems) was added pared with the DNAeasy Qiagen kit, and genotyped to check that proper for 20 min. Goat anti-mouse IgM or IgG both coupled to peroxydase (Cap- reconstitution had occurred. Genotyping by PCR was done using the orig- Ϫ Ϫ pel) diluted (1/1000) were alternatively used. The peroxydase activity was inal primers used by Takeuchi to genotype the tlr2 / mice (10). revealed by TMB substrate (Sigma-Aldrich), and stopped by HCL 1N. Morphological and immunohistochemical studies Reading was performed at 450 nm. The titers of the sera were expressed as Liver, kidney, and lung samples were collected, fixed in 10% formalin in PBS or in Dubosq-Brazil (for the kidneys), and embedded in paraffin. Tis- 4 The online version of the article contains supplemental materials. The Journal of Immunology 2671 Downloaded from http://www.jimmunol.org/ FIGURE 1. TLR2/4 double deficient (DKO) mice are susceptible to Leptospira interrogans Fiocruz strain. WT, tlr2Ϫ/Ϫ, tlr4Ϫ/Ϫ, or DKO mice were infected by i.p. injection of 2 ϫ 108 bacteria/mouse. A, Cumulative survival rates of (n) infected mice from three independent experiments. B, Kinetics of leptospiral load in livers and kidneys from infected WT and DKO mice (n ϭ 4 in each group). C, Leptospiral load in liver, kidneys and lungs at day 3 postinfection (n ϭ 6in each group). D, Leptospira Ag LipL32 immunolabeling in liver, kidney, and lung of a day 3 postinfected DKO mouse (bars ϭ 10 ␮m), and electron micrograph illustration of Leptospira located in the intercanalicular space of an infected DKO mouse liver (original magnification ϫ30,000). Arrows indicate the Leptospira; G indicates the glomeruli. E, Values (mean Ϯ SD, n ϭ 8 in each group) of blood urea nitrogen (BUN), serum creatinine, serum bilirubin, and aspartate .p Ͻ 0.05 compared with WT values ,ء .aminotransferase activity (ASAT) in day 3 postinfection mice

the dilution corresponding to twice the background level of the ELISA. esis that mice doubly deficient for TLR2 and TLR4 might exhibit by guest on September 25, 2021 Alternatively, same dilutions of the different sera were compared by mea- increased susceptibility to infection. To do this, adult (6–8 wk) surements of the OD at 450 nm. Ϫ Ϫ Ϫ Ϫ WT, TLR2 (tlr2 / ), TLR4 (tlr4 / ), and DKO mice with a Total IgM ELISA C56BL6/J background were infected via the i.p. route with 2 ϫ 8 Maxisorb 96-well plates were coated overnight at room temperature with 10 L. interrogans strain Fiocruz (Fig. 1A). Strikingly, the DKO 50 ␮l of rabbit anti-mouse IgG (HϩL) (2 ␮g/ml, MP Biomedicals) in mice were very susceptible to infection and died 4–6 days post carbonate buffer 0.1M (pH 9), blocked for 1 h with 100 ␮lofPBSϪ1% infection, while all the WT mice and tlr2Ϫ/Ϫ mice survived. BSA (Sigma-Aldrich), and washed with PBS-0.005% Tween 20. Succes- Tlr4Ϫ/Ϫ mice exhibited an intermediate phenotype with a lethality sively doubled serum dilutions in PBS/0.1% BSA were added for2hat ϭ room temperature. Purified IgM was provided by Farida Nato (Institut Pas- curve statistically different from that of DKO mice ( p 0.0032). teur), and used as a standard. After washing, goat anti-mouse IgM coupled Experiments performed with lower infecting doses (supplemental to alkaline phosphatase (1/10000, Sigma-Aldrich) was added to the plate. Table I) indicated that the lethality in DKO mice was dose-depen- After incubating for 1 h, the plates were rinsed and alkaline phosphatase dent. For all subsequent experiments, we therefore chose to use the activity was revealed at 405 nm by adding p-nitrophenylphosphate sub- 8 strate (1 mg/ml; Sigma-Aldrich) for 1 h. concentration of 2 ϫ 10 bacteria, which was likely to highlight the potential differences between infected WT and TLR-deficient ␮ Rescue of MT mice with serum mice that we wanted to investigate. Naive, day 3, and day 20 postinfected WT mice, as well as day 3 postin- The bacterial load progressively increased in the liver and kid- fected DKO mice were bled. After coagulation overnight at 4°C, the sera neys from DKO mice over the first 3 days following infection, were collected, heat-inactivated at 56°C for 30 min, and then frozen. Twenty microliters of pooled sera from day 3 or 20 postinfection or from whereas it rapidly declined in tissues from infected WT mice (Fig. naive mice was injected i.p. to ␮MT mice 4 h before leptospiral infection. 1B), suggesting that the clearance of Leptospira was severely impaired in DKO mice. At day 3 postinfection, the leptospiral load Statistical analysis was significantly greater in the liver, and to a lesser extent in kid- Results are expressed as means Ϯ SD. Statistical differences between WT neys and lungs in infected DKO mice than in WT mice (Fig. 1C). and other mice were analyzed using the unpaired, two-tailed Student’s t Ͻ Strikingly, the bacterial load was comparable in the infected livers test. p values 0.05 were considered to be significant. Survival curves Ϫ/Ϫ were compared using the log rank (Mantel Cox test) analysis using Graph- from tlr4 mice and DKO mice, while in kidneys, the bacterial Ϫ Ϫ Ϫ Ϫ Pad Prism software. load remained significantly lower in tlr2 / and tlr4 / than in DKO mice (Fig. 1C). These findings suggest that tissue differences Results in TLR signaling may exist, and that both TLR2 and TLR4 control DKO mice as a murine model of acute leptospirosis the bacterial growth in the kidney. Immunohistochemical studies Both TLR2 and TLR4 have been implicated in the detection of using an Ab directed against LipL32, one of the major lipoproteins Leptospira in mouse cells. We therefore set out to test the hypoth- of Leptospira species (12), confirmed the presence of leptospiral 2672 TLR4-DEPENDENT B CELL RESPONSES PROTECT AGAINST LEPTOSPIROSIS

Ags within the organs. LipL32 labeling revealed discrete Lepto- spira filaments within the liver, kidneys, and lungs from DKO mice on day 3 postinfection (Fig. 1D). Most of the helicoidal- shaped bacteria were found in bile canaliculi and intercanalicular spaces of the liver, in the renal interstitium, and surrounding the alveolar cells in the lungs (Fig. 1D). The high bacterial load measured day 3 postinfection in DKO mice, just before they died, was associated with marked deteriorations in liver and kidney function. Infected DKO mice exhibited significant higher levels of BUN, serum creatinine, serum bilirubin, and aspartate aminotransferase activity (ASAT) (Fig. 1E). Levels of serum bilirubin and serum creatinine were also significantly higher in infected tlr2Ϫ/Ϫ and tlr4Ϫ/Ϫ than in infected WT mice. These findings suggest that both TLR2 and TLR4 are involved in the pathological process elicited by Leptospira.

Defective expression of IFN-␥ and iNOS mRNA in Leptospira-infected DKO mice Downloaded from To better assess the roles of TLR2 and TLR4 in the induction of the inﬂammatory response caused by Leptospira, the levels of messenger RNA (mRNA), and the expression of cytokines and chemokines were measured in the liver, kidneys and lungs (Fig. 2 and supplemental Fig. 1) from WT, tlr2Ϫ/Ϫ, tlr4Ϫ/Ϫ, and DKO mice 3 days after bacterial inoculation. As equivalent levels of cytokine messengers were found in organs from WT, tlr2Ϫ/Ϫ, http://www.jimmunol.org/ tlr4Ϫ/Ϫ, and DKO naive mice (supplemental Table II), all subsequent analyses of mRNA levels in organs from infected mice were done in comparison to those of naive WT mice. IFN-␥ is a crucial effector of innate and adaptive immunity; it primes the macrophages and enhances their phagocytic activity. Inducible NO synthase (iNOS) is involved in both the host’s defenses and pathogenesis, as it produced NO, which can kill invad-

ing bacteria, but also contributes to the cell injury and organ dam- by guest on September 25, 2021 age they produce. The levels of IFN-␥ and iNOS mRNA, when detectable in DKO mice remained dramatically low in liver, kidneys, and lungs from day 3 postinfection compared with those in WT mice (Fig. 2A). In the kidneys, TLR2 or TLR4 deficiency alone did not impair the production of IFN-␥ mRNA, whereas in the liver, deficiency of TLR2 and TLR4 resulted in very low IFN-␥ mRNA expression. These findings suggest that TLR2 and TLR4 FIGURE 2. Levels of cytokines and chemokine expression in organs of have overlapping roles in the kidney, whereas they act synergis- Ϫ/Ϫ Ϫ/Ϫ ␥ infected WT, tlr2 , tlr4 , and DKO mice. Organs were collected three tically in the liver to produce IFN- . days after Leptospira infection with 2 ϫ 108 bacteria/mouse. A, Relative mRNA expression levels of IFN-␥ and iNOS quantified by real-time PCR Leptospira-induced inflammation is TLR-independent in liver, kidneys and lungs of mice. Data are mean Ϯ SD from duplicate measurements from two independent experiments (n ϭ 8 mice in each Contrasting with the low levels of IFN-␥ and iNOS mRNA in the group). B, The relative mRNA expression levels of proinflammatory me- organs of infected DKO mice (Fig. 2A), the levels of mRNA for diators were quantified by real-time PCR in the liver and kidneys of in- proinflammatory cytokines (IL-6, TNF) and chemokines (RANTES, fected mice. Data are mean Ϯ SD from duplicate measurements from two p Ͻ 0.005 vs WT ,ء .(MIP-2) were higher in tissues of day 3 postinfected DKO than in independent experiments (n ϭ 8 mice in each group those of WT mice (Fig. 2B and supplemental Fig. 1). When de- values. C, IL-1␤, TNF and RANTES measured by ELISA in liver and tectable, greater levels of IL1-␤ and chemokine CXCL1 mRNA kidneys from naive and day 3 postinfected mice. Values (mean Ϯ SD, n ϭ were also observed in the organs of infected DKO mice (Fig. 2B 3 in each group) are representative of two independent experiments. and supplemental Fig. 1). These findings indicate that the colonization of target tissues by pathogenic Leptospira leads to significant TLR2- and TLR4-independent inflammatory responses. At MyD88 is the adaptor molecule of the IL-1 receptor as well as day 3 postinfection, the lungs of DKO infected mice were less of most TLRs, except TLR3, which is the receptor of viral double colonized by Leptospira than their liver or kidneys, and expressed strand RNA. MyD88-deficient (MyD88Ϫ/Ϫ) mice were infected by very low levels of the mRNAs of the proinflammatory cytokines Leptospira to find out whether the TLR2- and TLR4-independent IL-6 and TNF. We therefore conducted all the subsequent exper- inflammation observed in the infected organs could be due to other iments in this study on the liver and kidneys only. Consistent with TLRs, such as TLR5 or TLR9. Survival curve ( p ϭ 0.61), bacte- the overall increase in mRNA expression of proinflammatory me- rial loads in kidneys and liver, and proinflammatory cytokine diators, the liver and kidneys from DKO mice day 3 postinfection mRNA profiles obtained after infecting MyD88Ϫ/Ϫ mice were produced greater amounts of IL1-␤ and RANTES, and TNF and equivalent to those observed in DKO mice (Fig. 3), suggesting that RANTES, respectively, than those from WT mice (Fig. 2C). only TLR2 and TLR4 are needed to prevent leptospirosis, but that The Journal of Immunology 2673

FIGURE 4. Bone marrow-derived cells expressing TLR2 and TLR4 Downloaded from FIGURE 3. Leptospira induce MyD88-independent inflammatory re- control Leptospira infection. A, Chimeric mice checked by PCR genotyp- Ϫ/Ϫ 8 sponses in liver and kidneys. MyD88 mice were infected with 2 ϫ 10 ing of WT or mutated tlr2 allele found in blood. All naive chimeric mice bacteria/mouse. A, Comparison of cumulative survival rates of (n) infected were checked individually. Chimeric mice were infected with 2 ϫ 108 Ϫ/Ϫ MyD88 and DKO mice from two independent experiments. B, Lepto- bacteria/mouse. B, Leptospiral load in liver and kidneys from day 3 postin- Ϫ/Ϫ spiral load in livers and kidneys from day 3 postinfected MyD88 and fected chimeric mice. C, q-RT-PCR profiles day 3 postinfection. DKO mice. C, Comparison of the relative levels of mediators mRNA ex- DKO3WT stands for irradiated WT mice replenished with DKO bone http://www.jimmunol.org/ pression in liver (L) and kidneys (K) of day 3 postinfection. Data are marrow cells, WT3DKO, irradiated DKO mice replenished with WT mean Ϯ SD from duplicate measurements from two independent experi- bone marrow cells, control chimeric mice with autologous reconstitutions ments (n ϭ 6 mice in each group). are denoted WT3WT and DKO3DKO. Data are mean Ϯ SD from one experiment with n ϭ 3 chimeric mice per group, and are representative of .p Ͻ 0.05 vs WT mice ,ء .two independent experiments mechanisms independent of the TLRs (and the IL1-receptor) are involved in the inflammation observed in the organs. in tissues from chimeric mice further indicate that deficient ex- Bone marrow-derived cells expressing TLR2 and TLR4 control pression of both TLR2 and TLR4 in cells from the hematopoietic Leptospira clearance compartment is not only responsible for the bacterial load, but also by guest on September 25, 2021 Reciprocal chimeras of WT and DKO mice were then constructed for the subsequent inflammatory response. Interestingly, the serum to determine which cell compartment controls susceptibility or re- parameters (i.e., BUN, creatinine, bilirubin and ASAT) paralleled sistance toward pathogenic Leptospira (Fig. 4A). Reconstitution of the amount of Leptospira colonizing the tissues of chimeric mice irradiated WT mice with DKO bone marrow cells (DKO3WT) (supplemental Fig. 2), suggesting that the intensity of the inflam- resulted in significantly higher bacterial colonization in livers and matory reaction is proportional to the leptospiral burden. kidneys of infected mice, as compared with the low level of bacteria detected in tissues from infected irradiated DKO mice recon- B lymphocytes play a crucial role in clearing leptospiral stituted with WT bone marrow cells (WT3DKO) (Fig. 4B). These infection data suggest that cells from the hematopoietic compartment ex- Leptospiral infection induced a rise in polymorphonuclear neutro- pressing TLR2 and TLR4 play a major role in controlling bacterial phils, dendritic cells and in T cells expressing TCR␥/␦ in periph- clearance. Bacterial loads in tissues from day 3 postinfected con- eral blood from DKO mice day 3 postinfection (supplemental Fig. trol (WT3WT and DKO3DKO) chimeric mice were compara- 3). These observations indicate that a deficiency of both TLR2 and ble to those from infected WT or DKO mice, respectively, indi- TLR4 does not impair the recruitment of these cells. In contrast, cating that the reconstitution was functional (Fig. 4, A and B). infected DKO mice exhibited a slight decrease in T cells express- However, the difference in bacterial load observed between ing CD8 and CD4, and a drop in B lymphocytes expressing CD19 DKO3DKO and DKO3WT chimeric mice suggests that the (supplementary Fig. 3), suggesting that these lymphocyte subsets nonhematopoietic compartment (i.e., mucosal cells) also contrib- might play a role in the development of leptospirosis. To test this ute to the clearance of bacteria (Fig. 4B). Most of the increase in hypothesis, RagϪ/Ϫ deficient mice devoid of both T and B cells IFN-␥ mRNA expression after infection with Leptospira, origi- were infected with 2 ϫ 108 pathogenic Leptospira. All RagϪ/Ϫ nates from the hematopoietic compartment, as illustrated by the mice died from the leptospiral infection (Fig. 5A), which indicates absence of the transcript of this cytokine in the liver of DKO3WT that B cells and/or T cell subsets make a major contribution to the chimeric mice (Fig. 4C). In contrast, the expression of IFN-␥ and defense against Leptospira infection. We next analyzed the con- iNOS mRNAs detected in the kidneys of DKO3WT chimeric sequences of inoculating ␮MT mice deficient only for mature B mice is equivalent to that of WT3WT mice and suggests that cells, and CD3Ϫ/Ϫ mice deficient only for T lymphocytes (supple- parenchymal renal cells express IFN-␥ and iNOS. Consistent with mental Fig. 4A) with 2 ϫ 108 Leptospira. Unexpectedly, lethality the data from infected WT and DKO mice (Fig. 2A), our results was similar in the ␮MT and in RagϪ/Ϫ mice ( p ϭ 0.45), demon- further indicate that the increased expression of IFN-␥ and iNOS strating that B cells must make a major contribution to murine mRNA by intrinsic renal cells in response to Leptospira infection resistance to acute leptospirosis (Fig. 5A). Furthermore, the bac- is dependent on both TLR2 and TLR4 signaling (Fig. 4C). The terial load was significantly greater in the liver and kidneys from mRNA expression profiles (Fig. 4C) for IL-6, TNF and RANTES day 3 postinfected ␮MT and RagϪ/Ϫ mice than in those from WT 2674 TLR4-DEPENDENT B CELL RESPONSES PROTECT AGAINST LEPTOSPIROSIS

FIGURE 6. Protective role of T cells in the kidney infected by Lepto- spira. A, Comparison of the relative levels of mediators mRNA expression Downloaded from in liver and kidneys of day 3 postinfected CD3Ϫ/Ϫ, ␮MT, DKO and WT mice. For CD3Ϫ/Ϫ mice data are mean Ϯ SD from n ϭ 4 mice. Values for WT, DKO, and ␮MT mice are from Fig. 5. B, Illustrations of nodular FIGURE 5. B Lymphocytes are crucial for clearing leptospiral infec- infiltrates (arrows) in fixed day 3 postinfected kidneys. Bars represent the tion. A, Comparison of cumulative survival rates of (n) mice infected with number (mean Ϯ SD) of interstitial nodular infiltrates per surface areas 2 ϫ 108 bacteria/mouse from 2 independent experiments. B, Leptospiral (104 squared micrometers) from five different tissue sections in each group. http://www.jimmunol.org/ load in day 3 postinfected liver and kidneys. C, Comparison of the relative levels of mediators mRNA expression in the liver and kidneys of day 3 postinfected mice. Data are mean Ϯ SD from two independent experiments pression induced by Leptospira are mainly attributable to lympho- (n ϭ 6 mice in each group). Values for DKO mice are the same as in Fig. p Ͻ 0.05 compared with WT values. cytes. In contrast, the expression of RANTES mRNA was not ,ء .3 altered in the liver or kidneys of infected RagϪ/Ϫ mice, indicating that its expression did not depend on lymphocytes (Fig. 5C). mice (Fig. 5B). Interestingly, infections with lower doses (2 ϫ 106 Therefore, the TLR2/4-independent expression of RANTES mRNA and 2 ϫ 104 Leptospira/mouse) resulted in the death of all infected identified in chimeric mice (Fig. 4C) appears to be due to another, ␮MT mice, although this occurred later than in ␮MT mice infected not yet identified, subset of cells from the hematopoietic compart- by guest on September 25, 2021 with 2 ϫ 108 bacteria (data not shown). Taken together, these ment. Similarly, B lymphocytes appear to be responsible for the findings provide strong evidence that B cells are essential to the TLR2/4-dependent IFN-␥ mRNA expression in the liver, whereas early control and clearance of the infection. In contrast to ␮MT T cells are involved in the expression of TLR2/4-dependent IFN-␥ mice, all the infected CD3Ϫ/Ϫ mice survived the infection (n ϭ mRNA in the kidneys (Fig. 5C). These results were confirmed 6, data not shown). Besides, leptospiral loads in the liver and using CD3Ϫ/Ϫ mice that lack T cells (Fig. 6A). To further inves- kidneys of day 3 postinfected CD3Ϫ/Ϫ mice were the same as tigate the role of T cells in the kidney, inflammation was evaluated those of infected WT mice (data not shown). This indicates that by H&E staining of kidney sections from day 3 postinfected T cells are not essential to control leptospiral infection. Con- CD3Ϫ/Ϫ, ␮MT and WT mice (Fig. 6B). Despite similar low bac- sistent with a putative role of B cells in the protection of in- terial burden in the kidneys of CD3Ϫ/Ϫ and WT mice, only fected liver, the day 3 postinfected ␮MT mice exhibited severe CD3Ϫ/Ϫ mice exhibited a noticeable interstitial inflammation and liver failure. The fact that the serum levels of BUN and creat- nodular infiltrates (Fig. 6B). In contrast, almost no nodular infil- inine of infected RagϪ/Ϫ mice but not of ␮MT mice, were sig- trates were detected in ␮MT mice, although these mice present a nificantly higher than those of infected WT mice along with a higher bacterial load in the kidney than WT mice (Fig. 5B). These significant higher level of serum creatinine in day 3 postin- data suggest that T cells, through their TLR2/4-dependent produc- fected CD3Ϫ/Ϫ than in WT mice (supplemental Fig. 4B), also tion of IFN-␥, play a key role in the protection of the kidney upon suggests that T but not B cells might play a selective protective Leptospira infection. role in infected kidneys. Early TLR4-dependent anti-Leptospira IgM are important for Lymphocytes are involved in TLR2/4 independent inflammation leptospiral clearance ␥ and in the TLR2/4 dependent IFN- response To distinguish between a cellular or humoral role of B cells, heat- Despite the high levels of Leptospira present, very low levels of inactivated serum obtained from infected WT and DKO mice was IL-6 mRNA were detected in kidneys and liver of both RagϪ/Ϫ injected i.p. in ␮MT mice 4 h before they were infected with a low and ␮MT mice (Fig. 5C). This suggests that B lymphocytes are dose (2 ϫ 105 bacteria/mouse) of Leptospira (Fig. 7A). Two to involved in the TLR2 and TLR4 (TLR2/4)-independent, expres- three weeks after infection, all the control ␮MT (data not shown) sion of proinflammatory IL-6 mRNA in kidneys and liver. The and ␮MT mice treated with the serum from naive or day 3 postin- much lower levels of TNF mRNA detected in the infected kidneys fected mice died, whereas all the mice that had been injected with of RagϪ/Ϫ mice than in those of ␮MT mice also raised the pos- the serum from day 20 postinfected WT mice survived the lepto- sibility that T cells may be responsible for TNF mRNA expression spiral infection (Fig. 7A). Heat inactivation of the serum excluded in kidneys (Fig. 5C). Taken together these findings suggest that the the possible involvement of the complement from the donor mice MyD88-independent IL-6 and TNF mRNA proinflammatory ex- in the rescue of the ␮MT animals, suggesting that specific humoral The Journal of Immunology 2675

To better understand the TLR contribution to the Ab production, the levels of specific Abs were next measured in the naive and day 3 postinfected TLR-deficient mice. Much lower levels of Lepto- spira specific total Ig or IgM Abs were found in the sera of infected DKO and tlr4Ϫ/Ϫ mice than in those of WT or tlr2Ϫ/Ϫ mice (Fig. 7C). However, total IgM levels were equivalent in the different strains of mice (Fig. 7C), excluding any global defect of IgM production in tlr4Ϫ/Ϫ or DKO mice following infection. Equivalent titers of specific Leptospira IgM were also found in day 3 postinfected CD3Ϫ/Ϫ mice (260 Ϯ 137, n ϭ 4) and WT mice (228 Ϯ 147, n ϭ 6). These findings demonstrate that the early TLR4-dependent production of anti-Leptospira IgM does not require T cells and suggests that LPS could be the Ag responsible for this early production of IgM. Indeed, immunization with purified leptospiral LPS leads to deficient production of anti-Leptospira IgM in both tlr4Ϫ/Ϫ and DKO mice, indicating that TLR4 is required for the specific production of IgM in response to leptospiral LPS (Fig. 7D). Taken together, these in vivo data are evidence of

the existence of a host defense mechanism, which pays an impor- Downloaded from tant part in the clearance of Leptospira and involves the early TLR4-dependent production of speciﬁc IgM Abs by B cells, triggered mainly by leptospiral LPS. To further investigate the role of TLR2 and TLR4 in the production of IgG Abs, which seems important for the clearance of

Leptospira, WT and TLR-deficient mice were infected with 2 ϫ http://www.jimmunol.org/ 108 of an attenuated Fiocruz strain. This strain has been passaged many times in vitro and has lost part of its virulence because only FIGURE 7. Early anti-Leptospira IgM participation in clearance of Lep- 50% of the infected DKO and no tlr4Ϫ/Ϫ died upon infection (data tospira. A, Heat inactivated serum from day 20, (D-20) postinfected WT, not shown). The specific anti-Leptospira IgG were measured in the but not naive serum, nor serum from day 3 (D-3) postinfected WT or DKO serum 20 days postinfection (Fig. 7E). Although reduced in mice, protects the ␮MT mice from death after leptospiral infection with Ϫ Ϫ Ϫ Ϫ tlr4 / mice and increased in tlr2 / mice, the specific IgG re- 2 ϫ 105 bacteria/mouse. Results are from one experiment (n ϭ 3 mice per group) and representative of two independent infections. B, Anti-Lepto- sponses were not statistically different from that of WT mice (Fig. spira IgM and IgG quantification in sera from WT and DKO mice infected 7E). However, the anti-Leptospira IgG titer was markedly de- by guest on September 25, 2021 with 2 ϫ 108 Leptospira and used in the ␮MT protection experiment. Data creased in the DKO mice, indicating that the presence of either (mean Ϯ SD) are expressed as the OD at 450 nm measured in sera (n ϭ 3 TLR2 or TLR4 is required to obtain an efficient anti-Leptospira p Ͻ 0.005 compared with the values IgG production (Fig. 7E). These data suggest that both specific ,ء .per group) at the dilution 1/180 from naive WT. C, Titers of anti Leptospira total Ig (Ig), anti Leptospira early IgM production and adaptive IgG responses are impaired in IgM and total IgM were measured in the serum from day 3 postinfected DKO mice, which certainly contribute to the exquisite sensitivity Ͻ ء mice. , p 0.005 vs WT values. D, Anti-Leptospira IgM obtained three of DKO mice toward leptospiral infection. days after intraperitoneal immunization with 20 ␮g of leptospiral purified Ϯ LPS. Data (mean SD) are expressed as arbitrary titers measured in sera Discussion p Ͻ 0.005 vs WT ,ء .n ϭ 6 in each group) from two independent injections) values. E, IgG titers of D-20 sera of WT and TLR deficient mice infected In this in vivo study, we highlight several mechanisms involved with 2 ϫ 108 attenuated Leptospira/mouse. Data (mean Ϯ SD) are ex- in murine resistance to leptospirosis. We took advantage of the pressed as the OD at 450 nm measured in sera (n ϭ 3 per group) at the fact that C57BL/6J mice are resistant to this infection to exam- p Ͻ 0.005 vs D-20 WT values. ine the roles of both TLR2 and TLR4 in the host response, and ,ء .dilution 1/360 to demonstrate that L. interrogans-infected DKO mice repro- duce the symptoms and features of acute human leptospirosis, immunity is crucial for bacterial clearance. Interestingly, the serum with severe hepatic and renal failure, associated with pulmonary from day 3 postinfected WT mice significantly delayed the death inflammation. Collectively, our findings indicate that defective of the ␮MT mice, when compared with the lethality observed after control by TLR2 and TLR4 greatly enhances the colonization of treatment by the serum from day 3 postinfected DKO mice ( p ϭ the liver, kidneys, and lungs of DKO mice by Leptospira, and is 0,022). These results indicate that some Abs from the WT serum, associated with impaired production of IFN-␥ and iNOS mRNA, which are missing in the DKO serum, exert protective effects. In- two important effectors of innate immunity involved in the killing deed, specific anti-Leptospira IgM Abs, which are the first specific of bacteria. Abs to be produced in serum, were absent in the serum from day-3 Once established in the organs, pathogenic Leptospira induced postinfected DKO mice (Fig. 7B). Serum from day 3 postinfected a marked and hitherto unknown MyD88-independent inflamma- WT mice was able to delay, but not to fully protect against lethal tory response that obviously did not contribute to efficient clear- Leptospira infection, despite the same IgM titers found in the sera ance of the bacteria. We had not expected this TLR2/4-indepen- from day 3 and 20 postinfected WT mice. This result suggests that dent inflammation, as previous in vitro studies conducted by our the specific IgM triggered in infected WT mice, although contrib- group and by others had shown that proinflammatory cytokines uting, are not essential to clear the Leptospira. Indeed, the rescue and chemokine responses by macrophages or epithelial cells to of the mice is probably due to the presence of IgG Abs in the whole Leptospira or their purified components were dependent on serum from day 20 postinfection, that were not detected at day 3 TLR2 and TLR4 (2, 4, 7, 13). Using chimeric and transgenic mice, postinfection (Fig. 7B). we further demonstrated that leukocytes, and B and T lymphocytes 2676 TLR4-DEPENDENT B CELL RESPONSES PROTECT AGAINST LEPTOSPIROSIS in particular, are responsible for this MyD88-independent inflam- early specific IgM titers equivalent to those found in WT mice. We mation. These results are at odds with the current generally held also show that the production of specific anti-Leptospira IgM de- view that TLR stimulation initiates a proinflammatory response pends on the recognition of leptospiral LPS by TLR4, strongly required for the recruitment of leukocytes to sites of infection, and suggesting that upon infection the LPS of Leptospira induces the subsequent killing of invading pathogens. Interestingly, infec- TLR4-dependent IgM production. Our results are in line with two tion by another spirochete, Borrelia burgdorferi, also results in a pioneering studies from the late 1970s. In the first of these studies, MyD88-independent release of inflammatory mediators, through Adler and Faine (22, 23) showed that mice treated with cyclophos- binding to the host integrin ␣3␤1 (14). Innate receptors from the phamide, a selective suppressor of B cell activity, died from in- recently described Nod-like receptor family could be potential can- fection with Leptospira, and that early circulating Abs protected didates for eliciting the TLR-independent inflammation induced by the host against Leptospira. In the second study, it was demon- the invasive pathogenic Leptospira. Nod1 and Nod2 proteins are strated that a challenge with LPS from conventional Gram-nega- intracellular receptors involved in the recognition of peptidogly- tive bacteria induced a peak of IgM Abs on day 8 postinfection in can, an essential component of the bacterial cell wall (15). Acti- C57BL/6J mice, which did not require the presence of T cells, and vation of Nod proteins has recently been shown to induce the se- was not found in tlr4-deficient C3H/HeJ mice (24). However, as cretion of the RANTES chemokine independently of TLR adaptors early emphasized by Adler and Faine (25), the Ab response elicited MyD88 or TRIF (16). In this study, increased levels of RANTES against leptospiral LPS, that we characterize as an early protective and other chemokines were detected in the organs of infected DKO anti-Leptospira IgM response, seems not sufficient to fully protect mice. However, the contribution of Nod proteins to Leptospira- the mice from the lethal leptospirosis, which can however be induced inflammation is not straightforward to demonstrate, be- avoided by pretreatment of the mice with a day 20 postinfection Downloaded from cause Nod1- and Nod2-deficient mice express both TLR2 and serum containing IgG. The early, TLR4-dependent IgM response, TLR4, and can therefore efficiently clear the Leptospira (data not which helped to clear Leptospira, could possibly be due to spe- shown). Ϫ/Ϫ cialized “innate” B cells, namely peritoneal B1 cells and splenic Although the tlr2 mice did not die from leptospirosis, TLR2 marginal zone B cells, which have been shown to cooperate to deficiency certainly contributes to the early lethality of DKO mice, Ϫ/Ϫ respond to blood-borne T-independent Ags by inducing an IgM as the death of tlr4 mice infected with Leptospira occurred wave in the first 3 days postinfection (26). Another study showed http://www.jimmunol.org/ later than that of DKO mice ( p ϭ 0.0032). Indeed, we found a that the resolution of Borrelia hermsii-induced relapsing fever did compensatory role of TLR2 and TLR4 in eliciting specific IgG require IgM and was concurrent with B1b cell expansion, which against Leptospira. Moreover TLR2 and TLR4 appear to play a further suggested that B1 cells play a key role in the host’s de- role in the control of IFN-␥ and iNOS mRNAs expression. Con- fenses against spirochetes (27). Moreover, mice lacking TLR2, or sistent with their role in host defense, the low levels of IFN-␥ and the TLR adaptor MyD88 did generate anti-Borrelia hermsii IgM, iNOS mRNA correlated well with the high bacterial loads found in but the kinetic pattern was delayed, and they suffered more severe the infected kidneys and liver of DKO mice suggesting that both episodes of bacteremia (28, 29). This demonstrates that the spiro- TLR2 and TLR4 are involved in the clearance of bacteria from the

chete species Leptospira and Borrelia both induce Myd88-inde- by guest on September 25, 2021 target organs. We further show that in the infected kidney, TLR2/ pendent inflammatory responses, and a MyD88-dependent early 4-dependent IFN-␥ production emanates from both T cells and IgM response. These bacteria are very different in terms of their parenchymal cells. This finding is consistent with previous studies biology and pathogenesis, but share a helicoidal morphology, showing that leukocytes and resident renal parenchymal cells, in particular tubular cells, produced IFN-␥ (17), and that TLR4 ex- which allows them to be motile and to enter the bloodstream. The pression in both parenchymal and hematopoietic cells is involved very early TLR2- or TLR4-dependent IgM response that they en- in innate resistance to Escherichia coli (18). Our results are also counter therefore seems to be a crucial event in the host’s defense consistent with previous in vitro data which emphasized the role of against blood-borne spirochetes. TLR2 in epithelial renal cells in response to leptospiral lipopro- Our results obtained in mice with a C57BL/6J background are teins (5). In contrast, the present findings suggest that B cells are consistent with the previously reported susceptibility toward lep- mainly responsible for TLR2/4-dependent IFN-␥ production in the tospirosis of young C3H/HeJ mice, which are naturally defective liver. Although this cell subset is not usually viewed as producing for TLR4 (6, 7, 30) and C3H/SCID mice, which are deficient for IFN-␥, some of us recently reported that B cells can indeed pro- T, B, and NK cells (30). The present study also reveals that both ␥ duce this cytokine (19). Further studies are required to clarify IFN- and T cells are involved in the host’s defenses against lep- Ϫ/Ϫ whether this IFN-␥ production results from direct stimulation of B tospirosis. However, CD3 mice, which lack T cells, did not die cells. from leptospirosis. Similarly, transgenic C57BL/6J mice, which We also demonstrated that the protective role of B cells relies are deficient for IFN-␥, also did not die following infection with partly on early, TLR4- dependent production of IgM specifically the Fiocruz strain of Leptospira (31). We now understand that the directed against Leptospira. This finding is consistent with the sim- effective contribution of these mediators was masked by the pres- ilar lethality curves of infected tlr4Ϫ/Ϫ and ␮MT mice. Circulating ence of efficient B cell responses that cleared the bacteria. IgMs, whether natural or Ag-induced, are high-avidity, pentameric One compelling point raised by our study is that the liver lesions Abs with a marked ability both to activate the complement path- induced by the pathogenic Leptospira seem to contribute more to way and to opsonize bacteria, making them more vulnerable to the induced lethality than the kidney injury, because ␮MT, which fatal phagocytosis. Specific Abs have been shown to play an im- lack B cells but do have protective T cells, are dying despite a very portant role in leptospirosis. The predominant and neutralizing Abs limited renal lesion. When B cells or TLR4 is missing, the lepto- present in the serum of infected patients include IgM directed spiral load is higher in the liver than in the kidney, which may be against LPS, and IgG directed against heat shock proteins and reflecting a better protection of the kidney against Leptospira in- lipoproteins (20, 21). Strikingly, a profound alteration in specific fection. Indeed we found a double level of protection in the kidney IgM titers was found in the tlr4Ϫ/Ϫ and DKO-infected mice. LPS because both T cells and parenchymal cells produce IFN-␥, is known to be a T-independent Ag. Accordingly, Leptospira in- whereas in the liver only B cells are providing this protection. fection of CD3Ϫ/Ϫ mice, which are deficient in T cells only, led to Moreover the IFN-␥ production in the kidney only requires the The Journal of Immunology 2677 presence of either TLR2 or TLR4, although in the liver the pro- rogans serovar icterohaemorrhagiae infection and contributes to in vivo control tective effect of TLR4 and TLR2 is synergistic. Finally, only pa- of leptospiral burden. Infect. Immun. 74: 887–895. 8. Merien, F., D. Portnoi, P. Bourhy, F. Charavay, A. Berlioz-Arthaud, and renchymal renal cells are expressing iNOS, which can contribute G. Baranton. 2005. A rapid and quantitative method for the detection of Lepto- also to the bacterial clearance. Such highly sophisticated defense spira species in human leptospirosis. FEMS Microb. Lett. 249: 139–147. 9. Morrison, T. B., Y. Ma, J. H. Weis, and J. J. Weis. 1999. Rapid and sensitive mechanisms may be explaining why Leptospira can chronically quantification of Borrelia burgdorferi-infected mouse tissues by continuous flu- colonize the rodent’s kidney without major lesions. orescent monitoring of PCR. J. Clin. Microbiol. 37: 987–992. By using a combination of different methods, the present study 10. Takeuchi, O., K. Hoshino, and S. Akira. 2000. Cutting edge: TLR2-deficient and MyD88-deficient mice are highly susceptible to Staphylococcus aureus infection. provides a more comprehensive understanding of the relative roles J. Immunol. 165: 5392–5396. of lymphocytes and TLR activation, and identifies a hitherto-un- 11. Chassin, C., J. M. Goujon, S. Darche, L. du Merle, M. Bens, F. Cluzeaud, known link between TLRs and B cell responses in host defenses C. Werts, E. Ogier-Denis, C. Le Bouguenec, D. Buzoni-Gatel, and A. Vandewalle. 2006. Renal collecting duct epithelial cells react to pyelonephri- against Leptospira. To sum up, this in vivo study highlights a tis-associated Escherichia coli by activating distinct TLR4-dependent and -inde- compartmentalized host response and reveals that B cells play a pendent inflammatory pathways. J. Immunol. 177: 4773–4784. 12. Haake, D. A., G. Chao, R. L. Zuerner, J. K. Barnett, D. Barnett, M. Mazel, dual role in the early clearance of Leptospira both through TLR4- J. Matsunaga, P. N. Levett, and C. A. Bolin. 2000. The leptospiral major outer dependent IgM production elicited against leptospiral LPS in the membrane protein LipL32 is a lipoprotein expressed during mammalian infec- bloodstream, and the TLR2- and TLR4-dependent production of tion. Infect. Immun. 68: 2276–2285. ␥ 13. Hung, C. C., C. T. Chang, K. H. Chen, Y. C. Tian, M. S. Wu, M. J. Pan, IFN- in the liver. In the kidney, protection relies on the comple- A. Vandewalle, and C. W. Yang. 2006. Upregulation of chemokine CXCL1/KC mentary roles of parenchymal cells and T cells, expressing iNOS by leptospiral membrane lipoprotein preparation in renal tubule epithelial cells. and IFN-␥ that involve activation of both TLR2 and TLR4. When Kidney Int. 69: 1814–1822. 14. Behera, A. K., E. Hildebrand, S. Uematsu, S. Akira, J. Coburn, and L. T. Hu. these protective mechanisms are missing, B cells in the liver and 2006. Identification of a TLR-independent pathway for Borrelia burgdorferi- Downloaded from T cells in the kidney are mainly responsible for the deleterious, induced expression of matrix metalloproteinases and inflammatory mediators through binding to integrin ␣ 3 ␤ 1. J. Immunol. 177: 657–664. TLR-independent inflammation induced by Leptospira. Because 15. Werts, C., S. E. Girardin, and D. J. Philpott. 2006. TIR, CARD and PYRIN: three human cells have impaired ability to detect leptospiral atypical domains for an antimicrobial triad. Cell Death Differ. 13: 798–815. lipid A through TLR4 (4), the present findings may explain why 16. Werts, C., L. le Bourhis, J. Liu, J. G. Magalhaes, L. A. Carneiro, J. H. Fritz, S. Stockinger, V. Balloy, M. Chignard, T. Decker, et al. 2007. Nod1 and Nod2 human beings are more susceptible to leptospirosis than mice. induce CCL5/RANTES through the NF-␬B pathway. Eur. J. Immunol. 37:

2499–2508. http://www.jimmunol.org/ 17. Timoshanko, J. R., S. R. Holdsworth, A. R. Kitching, and P. G. Tipping. 2002. Acknowledgments IFN-␥ production by intrinsic renal cells and bone marrow-derived cells is re- We thank Dana Philpott, Stephen Girardin (University of Toronto, Can- quired for full expression of crescentic glomerulonephritis in mice. J. Immunol. ada), and Ivo Boneca (Institut Pasteur) for critical reading of the manu- 168: 4135–4141. script. We thank Shizuo Akira (Osaka University, Japan), Michel Chig- 18. Schilling, J. D., S. M. Martin, C. S. Hung, R. G. Lorenz, and S. J. Hultgren. 2003. Toll-like receptor 4 on stromal and hematopoietic cells mediates innate resistance nard, Claude Leclerc, James Di Santo, and Armelle Phalipon (Institut to uropathogenic Escherichia coli. Proc. Natl. Acad. Sci. USA 100: 4203–4208. Pasteur) for providing transgenic mice. We thank Roger Peronet (Institut 19. Menard, L. C., L. A. Minns, S. Darche, D. W. Mielcarz, D. M. Foureau, D. Roos, Pasteur) for i.v. injections. We thank David Haake (UCLA, California) and F. Dzierszinski, L. H. Kasper, and D. Buzoni-Gatel. 2007. B cells amplify IFN-␥ ␣ J. Immunol. Farida Nato (Institut Pasteur) for the kind gifts of LipL32 antiserum and production by T cells via a TNF- -mediated mechanism. 179: 4857–4866. purified IgM, respectively. 20. Adler, B., and S. Faine. 1978. The antibodies involved in the human immune by guest on September 25, 2021 response to leptospiral infection. J. Med. Microbiol. 11: 387–400. 21. Guerreiro, H., J. Croda, B. Flannery, M. Mazel, J. Matsunaga, M. Galvao Reis, Disclosures P. N. Levett, A. I. Ko, and D. A. Haake. 2001. Leptospiral proteins recognized The authors have no financial conflict of interest. during the humoral immune response to leptospirosis in humans. Infect. Immun. 69: 4958–4968. 22. Adler, B., and S. Faine. 1976. Susceptibility of mice treated with cyclophosph- References amide to lethal infection with Leptospira interrogans serovar pomona. Infect. 1. Bharti, A. R., J. E. Nally, J. N. Ricaldi, M. A. Matthias, M. M. Diaz, Immun. 14: 703–708. M. A. Lovett, P. N. Levett, R. H. Gilman, M. R. Willig, E. Gotuzzo, and 23. Adler, B., and S. Faine. 1977. Host immunological mechanisms in the resistance J. M. Vinetz. 2003. Leptospirosis: a zoonotic disease of global importance. Lan- of mice to leptospiral infections. Infect. Immun. 17: 67–72. cet Infect. Dis. 3: 757–771. 24. Izui, S., R. A. Eisenberg, and F. J. Dixon. 1981. Subclass-restricted IgG poly- 2. Werts, C., R. I. Tapping, J. C. Mathison, T. H. Chuang, V. Kravchenko, clonal antibody production in mice injected with lipid A-rich lipopolysaccha- I. Saint Girons, D. A. Haake, P. J. Godowski, F. Hayashi, A. Ozinsky, et al. 2001. rides. J. Exp. Med. 153: 324–338. Leptospiral lipopolysaccharide activates cells through a TLR2-dependent mech- 25. Adler, B., and S. Faine. 1982. Evidence that leptospiral lipopolysaccharide is not anism. Nat. Immunol. 2: 346–352. an important protective antigen. J. Med. Microbiol. 15: 259–262. 3. Que-Gewirth, N. L., A. A. Ribeiro, S. R. Kalb, R. J. Cotter, D. M. Bulach, 26. Martin, F., A. M. Oliver, and J. F. Kearney. 2001. Marginal zone and B1 B cells B. Adler, I. S. Girons, C. Werts, and C. R. Raetz. 2004. A methylated phosphate unite in the early response against T-independent blood-borne particulate anti- group and four amide-linked acyl chains in leptospira interrogans lipid A: the gens. Immunity 14: 617–629. membrane anchor of an unusual lipopolysaccharide that activates TLR2. J. Biol. 27. Alugupalli, K. R., R. M. Gerstein, J. Chen, E. Szomolanyi-Tsuda, Chem. 279: 25420–25429. R. T. Woodland, and J. M. Leong. 2003. The resolution of relapsing fever bor- 4. Nahori, M. A., E. Fournie-Amazouz, N. S. Que-Gewirth, V. Balloy, M. Chignard, reliosis requires IgM and is concurrent with expansion of B1b lymphocytes. C. R. Raetz, I. Saint Girons, and C. Werts. 2005. Differential TLR recognition of J. Immunol. 170: 3819–3827. leptospiral lipid A and lipopolysaccharide in murine and human cells. J. Immu- 28. Bolz, D. D., R. S. Sundsbak, Y. Ma, S. Akira, J. H. Weis, T. G. Schwan, and nol. 175: 6022–6031. J. J. Weis. 2006. Dual role of MyD88 in rapid clearance of relapsing fever Bor- 5. Yang, C. W., C. C. Hung, M. S. Wu, Y. C. Tian, C. T. Chang, M. J. Pan, and relia spp. Infect. Immun. 74: 6750–6760. A. Vandewalle. 2006. Toll-like receptor 2 mediates early inflammation by lep- 29. Alugupalli, K. R., S. Akira, E. Lien, and J. M. Leong. 2007. MyD88- and Bru- tospiral outer membrane proteins in proximal tubule cells. Kidney Int. 69: ton’s tyrosine kinase-mediated signals are essential for T cell-independent patho- 815–822. gen-specific IgM responses. J. Immunol. 178: 3740–3749. 6. Pereira, M. M., J. Andrade, R. S. Marchevsky, and R. Ribeiro dos Santos. 1998. 30. Nally, J. E., M. C. Fishbein, D. R. Blanco, and M. A. Lovett. 2005. Lethal Morphological characterization of lung and kidney lesions in C3H/HeJ mice in- infection of C3H/HeJ and C3H/SCID mice with an isolate of Leptospira inter- fected with Leptospira interrogans serovar icterohaemorrhagiae: defect of CD4ϩ rogans serovar copenhageni. Infect. Immun. 73: 7014–7017. and CD8ϩ T-cells are prognosticators of the disease progression. Exp. Toxicol. 31. Athanazio, D. A., C. S. Santos, A. C. Santos, F. W. McBride, and M. G. Reis. Pathol. 50: 191–198. 2008. Experimental infection in tumor necrosis factor ␣ receptor, interferon ␥ and 7. Viriyakosol, S., M. A. Matthias, M. A. Swancutt, T. N. Kirkland, and interleukin 4 deficient mice by pathogenic Leptospira interrogans. Acta Trop. J. M. Vinetz. 2006. Toll-like receptor 4 protects against lethal Leptospira inter- 105: 95–98.