Antimony-Resistant Leishmania donovani Exploits miR-466i To Deactivate Host MyD88 for Regulating IL-10/IL-12 Levels during Early Hours of Infection

The incidence of visceral leishmaniasis (VL) is increasing worldwide. The protozoan parasite Leishmania donovani, its causative agent, is the second largest parasitic killer in the world after malaria (1, 2). Sodium stibogluconate, a pentavalent antimonial popularly known as SSG, has been used for VL treatment in the Indian subcontinent for the last nine decades or so. However, recently the drug has been showing dwindling efficacy because of the emergence of antimony-resistant parasites (3, 4). It has not been in use in the Indian subcontinent for about a decade; nevertheless, 78% of the recent field isolates of L. donovani from Bihar state in India are resistant to antimonials (5). The mechanism by which such parasites are maintaining their resistant phenotype in the field remains to be elucidated. The antimony-resistant L. donovani (Sb^RLD) parasites, unlike their antimony-sensitive L. donovani (Sb^SLD) counterpart, are equipped with a number of unique attributes such as increased copy number of parasite virulence factor GP63 and surface glycoprotein amastin, and amplification of the episomal MAPK locus for better survival within the host (6, 7). They also show unique skills for manipulation of host immune response by producing disease-promoting cytokines such as IL-10 and TGF-β (8–10). There are reports that infection with Sb^RLD results in higher organ parasite number and IL-10 level in experimental infection (9), and this is also seen in kala-azar patients who harbor resistant parasites (11).

Innate immunity to Leishmania is reportedly triggered by TLRs on the APCs that detect and respond to conserved structural motifs on the invading pathogens (12). TLRs have been shown to stimulate APC function in response to defined lipid, protein, and nucleic acids present on bacterial, fungal, and viral pathogens (13). MyD88 is a universal adaptor protein used by almost all the TLRs except TLR3 to activate transcription factor NF-κB (14). It is reported that TLR2 plays a significant role in Leishmania infection (13, 15). An earlier report also suggests that MyD88 plays a vital role in TLR2-mediated clearance of Leishmania parasites (16). Previously we showed that Sb^RLD, but not Sb^SLD, expresses a unique surface glycan with terminal sugar N-acetyl galactosamine, which interacts with host cell TLR2/TLR6 to induce IL-10 in an NF-κΒ–dependent manner (9). It has been shown that exploitation of the TLR pathway is a strategy used by Leishmania parasites to establish successful infection (9, 17). Among downstream signaling molecules of TLR2, MyD88 plays the most vital role in Leishmania infection (18). MyD88-deficient mice have been shown to mount defective IL-12 responses to a number of microbial stimuli (18, 19). Importantly, IL-12 and IL-10 can inversely regulate each other to decide the parasite fate in Sb^RLD-infected host cells (20).

TLR signaling is tightly regulated by microRNAs (miRNAs) to avoid excessive inflammation during infection and tissue injury. Several miRNAs are induced by TLR activation in innate immune cells, which target the 3′-untranslated regions (UTR) of mRNAs encoding components of the TLR signaling system. The expression of most of these TLR-responsive miRNAs largely depends on NF-κB activity (21), and one of these miRNAs, miR-146, can negatively regulate the MyD88–NF-κB signaling pathway after bacterial infection (22). Recent reports also suggest that L. donovani targets cellular miRNAs to regulate host genes for its own benefit (23, 24).

During this study, we examined the possible involvement of the TLR-MyD88 pathway in experimental mice upon infection with Sb^SLD or Sb^RLD. Our study clearly shows that Sb^RLD-mediated TLR2/6 signaling in host, unlike Sb^SLD infection, is critically regulated through MyD88. We also show that MyD88-deficient mice infected with Sb^RLD display higher parasite burden, a defect associated with impaired IL-12 production. This study focuses on the unique ability of Sb^RLD to activate posttranscriptional machinery involving miR-466i, which downregulates MyD88 expression, resulting in elevated IL-10 level and aggressive pathology. We have also deciphered the detailed molecular mechanism of posttranscriptional modification of host MyD88 in Sb^RLD infection that allows the antimony-resistant parasite to regulate IL-10 and IL-12 axis during establishment of successful infection.

BALB/c mice and golden hamsters (Mesocricetus auratus) were maintained and bred under pathogen-free conditions. C57BL/6 and mice were obtained from IISc (Bangalore, India). MyD88^−/−were a kind gift from Dr. Gobardhan Das (International Centre for Genetic Engineering and Biotechnology, New Delhi, India). IL-12^−/− mice were originally obtained from Jackson Lab and maintained in National Centre for Cell Science (Pune, India); experiments were performed in the animal facility of National Centre for Cell Science. The IL-12^−/− mice and other MyD88^−/− mice were of B10 background. Use of mice and hamsters was approved by the Institutional Animal Ethics Committees of Indian Institute of Chemical Biology. All animal experiments were performed according to the National Regulatory Guidelines issued by Committee for the Purpose of Supervision of Experiments on Animals, Ministry of Environment and Forest, Government of India.

Sb^RLD (MHOM/IN/2009/BHU575/0 and MHOM/IN/2005/BHU138) and Sb^SLD (MHOM/IN/83/AG83 and MHOM/NP/03/BPK206/0) (5) maintained in golden hamsters (25) were used for this study. Amastigotes were obtained from the spleen of infected hamsters (26), transformed into promastigotes, and maintained (27).

Peritoneal exudate cells, conveniently called macrophages (Mϕs), were harvested from BALB/cor C57BL/6 or MyD88^−/− or IL-12^−/− mice and purified as described previously (5). Most of the experiments were performed on BALB/c Mϕs; however, for those experiments where Mϕs were obtained from IL-12^−/− or MyD88^−/− mice, C57BL/6 Mϕs were used as wild type (WT) control. Mϕs were infected with stationary-phase L. donovani promastigotes at a ratio of 1:10 for 6 h, washed to remove free parasites, and incubated for desired time periods. Time point 6 h is considered as 0 h postinfection (p.i.). Supernatants obtained from Mϕs were used for cytokine analysis by ELISA. All transfection experiments were performed using RAW 264.7 cell line. Parasite soluble Leishmanial Ag (SLA) was prepared as described previously (27), and parasite lysates were prepared as described by Kapler et al. (28). The protein content of the lysate was estimated, and equal concentration of the lysate protein was directly used for the experiment without further centrifugation step.

Preinfection in serum-free medium, Mϕs were treated with 20 μM of the IKK inhibitor BAY 11-7082 for 45 min (29) and with the NF-κB inhibitor SC-3060 (30) for 1 h. Mϕs were treated with murine rIL-10 (10, 20, 200 pg/ml) or with neutralizing anti–IL-10 Ab (1, 10, 100 ng/ml) (both from BD Pharmingen), which were preincubated for 1 h preinfection. Efficacy of the anti–IL-10 Ab was tested in its ability to inhibit LPS-induced IL-10 production in uninfected Mϕs. RAW 264.7 Mϕ cells, conveniently called RAW, were treated with siRNA for MyD88 (1 μg) or control siRNA (1 μg) (Santa Cruz) for 2 h before infecting them with Sb^RLD or Sb^SLD. In some experiments, RAW cells were pretreated with 100 μM TRIF inhibitor peptide (Invivogen) 30 min preinfection. In others, Mϕs were stimulated with LPS (1 μg/ml) and used as a positive control.

The murine IL-10 promoter −17/−1576 (1.57 kb) (9), IL-12 p35 promoters −6/−1411 (1.4 kb), and IL-12 p40 promoter −103/−1165 (1.06 kb) (20) were PCR-amplified and cloned into a pGL3-Basic vector (Promega) as previously described (9). The −6009/−5008 (1 kb) region of miR-466i was cloned into pGL3-Basic vector. The +12/+802 region of murine MyD88 3′UTR was cloned in psiCHECK vector (a generous gift from Dr. Soma Banerjee, Institute of Post Graduate Medical Education and Research, Kolkata, India). The miR-466i promoter sequence [with deletion at the NF-κB binding sites (−6009/−5008), (−5862/−5852), (−5573/−5558)] and psiCHECK-MUT3′UTR (with deletion at miR-466i binding sites Δ751-757 and Δ768-775) constructs were generated from the miR-466i whole promoter construct (1 kb) and MyD88-3′UTR-psiCHECK construct, respectively, by Quick ChangeII PCR-based site-directed mutagenesis kit (Stratagene) according to the manufacturer’s protocol. The primer sequences are given in Table I. All the inserts were confirmed by sequencing. RAW264.7 cell lines were transfected with miR-466i mimics (Mission microRNA mimic; Sigma) at different concentrations ranging from 10 to 200 nM or with 25 nM synthetic miR-466i inhibitor (Mission synthetic microRNA inhibitor; Sigma). The mimics and inhibitors of miR-466i are dsRNA molecules designed to mimic or inhibit endogenous mature miR-466i when transfected into cells. RAW264.7 cell lines were transfected with pUNO-MyD88 vector (Invivogen) in some cases. In some experiments, RAW264.7 cells were transiently transfected with these constructs (2 μg) using Lipofectamine 2000 (Invitrogen), rested for 12 h, and either infected with Sb^RLD or Sb^SLD for 24 h, cotransfected with pUNO-MyD88 vector, or treated with rIL-10 or rIL-12. Luciferase activity in cell extracts, measured using the Dual Luciferase Reporter kit (Promega) according to the manufacturer’s protocols, was normalized to the level of the protein content. All transfection experiments were performed on RAW cells.

IL-10, IL-12p70, TNF-α, and IL-6 levels in the supernatant were measured using a sandwich ELISA Kit (BD Pharmingen) as per the manufacturer’s protocol. Cytokine levels were determined by measuring the OD at 450 nm using a microtiter plate reader (Multi Ex Scan; Perkin Elmer).

The expression of MyD88 mRNA in Sb^RLD or Sb^SLD or LPS-treated Mϕs was analyzed. Total RNA was isolated using the total RNA isolation kit (Roche Biochemicals) following the manufacturer’s protocol. cDNA synthesis and real-time–quantitative PCR (RT-qPCR) were done as described elsewhere (5). The sequences of the forward and reverse primers are presented in Table I. Reactions were run on an Applied Biosystems 7500 Fast Real-Time PCR system. Experiments on negative controls of cDNA synthesis (i.e., without reverse transcriptase) and no-template controls (i.e., without cDNA template) were also done for each gene. All reactions were done in duplicate, and their arithmetic average threshold cycle (Ct) was used for data analysis. The fold of gene expression compared with control was calculated by 7500 Fast System SDS software, version 1.4 (Applied Biosystems, Carlsbad, CA), using the formula: Fold of expression = 2^-ΔΔCt. Real-time PCR using miRNA-specific stem-loop primers for reverse transcription and TaqMan probes for mature murine miRNA was performed in accordance with the manufacturers’ protocols. RNA was extracted from an equal amount of cells per time point using the RNeasy Mini Kit (Qiagen). cDNA synthesis was performed using SuperScript III (Invitrogen). Pri-miR-466i levels were determined using the TaqMan Pri-miRNA assays developed by Applied Biosystems. Data analysis was performed using the Fast System SDS software, version 1.4 (Applied Biosystems).

The short hairpin RNA (shRNA) constructs and self-inactivating lentiviral vector pCRI.LV (31, 32) were a kind gift from Dr. Robin Mukhopadhyay (Advanced Centre for Treatment, Research and Education in Cancer, Mumbai, India). The MyD88-specific shRNA cassette, driven by the promoter of the small nuclear RNA U6, was generated by PCR-mediated amplification of positions 915–934 of the MyD88 gene (GenBank accession number NM_010851.2), and the selection of shRNA target sequences was based on published guidelines (33). The shRNA cassette was subsequently cloned into a self-inactivating lentiviral vector pCRI.LV, and the virus produced was concentrated by ultracentrifugation as described previously (32). Virus titer was measured at 2 × 10⁶ infectious units/ml. shRNA target sequence selection and nonrelevant control vector uses were the same as reported elsewhere (34). An aliquot (50 μl) of 1000× vector concentrate of MyD88-specific shRNA construct or control vector was injected into spleen tissue in anesthetized BALB/c mice 3 d preinfection for generating MyD88 knockdown (KD) or control vector mice.

To check the efficacy of shRNA administration in spleen cells, we injected GFP-encoding control shRNA to the spleen tissue of BALB/c mice, and GFP expression in splenic Mϕs was evaluated by flow cytometry. The splenocytes were washed and Fc receptors were blocked with 5% FCS, FcgIgG (BD Pharmingen), and 0.5% BSA in PBS for 30 min. The cells were first stained with PE-conjugated anti-CD11b mAb (BD Pharmingen) at 4°C for 30 min in the dark. The false-positive populations were removed by staining with proper isotype control. Cells were centrifuged and resuspended in PBS. At least 10⁶ events were acquired on a BD FACS Aria II (BD Biosciences) for subsequent analysis using FACS Diva software (BD Biosciences).

For in vivo infections, female BALB/c mice or control vector mice or MyD88 KD mice (4–6 wk old and 20–25 g each) were injected via intracardiac route with 10⁷ cells L. donovani second-passage stationary promastigotes (suspended in 100 μl) as reported previously (35). Each group contained five mice (n = 5) that were sacrificed 4 wk p.i. The parasite burdens in the liver and spleen were enumerated and expressed in Leishman-Donovan units (36). Each experiment was performed three times.

Cytoplasmic and nuclear proteins were prepared and Western blotting was performed for p50, c-Rel, p65, MyD88 (Santa Cruz Biotechnology), β-actin, and histone (Cell Signaling Technology). Blots were probed with specific Abs. Binding of secondary HRP-labeled goat-anti rabbit or goat-anti mouse Abs (Cell Signaling Technology) was analyzed using SuperSignalR West Pico or West Dura Chemiluminescent Substrate (Pierce, Rockland, IL). Purity of nuclear and cytoplasmic extracts was checked using histone.

Total RNA (10 μg/lane) was resolved on 15% polyacrylamide-urea gels and electroblotted onto HyBond N+ membranes (GE Healthcare, Little Chalfont, U.K.). Membranes were hybridized overnight with radiolabeled antisense miRNAs in Express Hyb solution (Clontech, Mountain View, CA). After hybridization, membranes were washed three times with 2× SSC and 0.05% SDS and twice with 0.1× SSC and 0.1% SDS, exposed overnight to imaging screens, and analyzed using a Storm 860 PhosphorImager (GE Healthcare). The same blot was hybridized (on stripping in boiling 0.1% SDS) with control U6 RNA. The probe sequence for miR-466i was 5′-CACACACACACACACACACACATAC-3′.

Chromatin immunoprecipitation (ChIP) assays were performed either on uninfected Mϕs or on Sb^RLD (BHU575/BHU138)-infected Mϕs using the ChIP assay kit (Upstate) following the manufacturer’s instructions. Immunoprecipitation was done using rabbit IgG, acetylated histone (H3A), or NF-κB Abs such as αp50 and c-Rel overnight at 4°C followed by DNA extraction. PCR was performed in triplicate for each separate ChIP experiment using primers designed for regions identified as enriched in preliminary analysis. RT-qPCR was performed to amplify the −482/−645 region of IL-10 promoter using primer 5′-GCCCCACAGCACACATATCC-3′ and 5′-CCTGGGTTGAACGTCCG-3′, or specific region of miR-466i promoter using primer 5′-CTCAGGTATACGGATAGG-3′ and 5′-GACCTGGAGTTTTTAGCCTGCC-3′, or 5′-CTCAGCCTTACAATAAGC-3′ and 5′-CAGTAGGAGAAGATTAGTTG-3′, or 5′-CAACTAATCTTCTCCTACTG-3′ and 5′-GATAGGACAGTTAGGATAGC-3′. Fold enrichment method was used to calculate the signal of the promoter region for each gene as described previously (37).

To evaluate the role of MyD88 in an in vivo long-term model of Sb^RLD infection, we used female BALB/c mice or control vector mice or MyD88 KD mice (10 mice/group) infected with Sb^RLD to study survival kinetics as described previously (38).

Livers were fixed in 10% formalin (Merck) and embedded in paraffin. Tissue sections (5 mm) were stained with H&E to study their microarchitecture by light microscopy. Photomicrographs were taken with a Nikon Eclipse E200 microscope.

Each experiment was performed three times, and the data represent mean of three independent experiments with ±SD unless otherwise mentioned. Statistical significance between means of various groups was determined using a two-tailed Student t test. Only p values <0.05 were considered to be statistically significant. Values were considered extremely significant if ***p < 0.001, very significant if ≤ **p < 0.01, significant if *p ≤ 0.05, and not significant when p > 0.05. Error bars indicate the mean ± SD. Data were analyzed using Prism 5.0 (GraphPad, San Diego, CA).

To determine the role of MyD88 in Sb^RLD-mediated infection, we infected Mϕs from C57BL/6 (WT) and MyD88^−/− mice with either Sb^SLD (AG83/BPK206) or Sb^RLD (BHU575/BHU138), and intracellular amastigotes/100 Mϕs and IL-10 in the culture supernatant were measured. We observed that infection of MyD88^−/− Mϕs with Sb^RLD increased significantly the intracellular parasite number (Fig. 1A) and raised the IL-10 level in the culture supernatant as compared with the Sb^RLD-infected WT counterpart (Fig. 1B). Furthermore, the IL-10 reporter assay in RAW cells transfected with MyD88 siRNA, but not with TRIF inhibitory peptide, showed significant upregulation of luciferase activity when infection was done with Sb^RLD, but not with Sb^SLD (Supplemental Fig. 1A). Because IL-12 plays an antagonistic role in IL-10 induction, we studied the IL-12 status in Sb^RLD-infected MyD88^−/− Mϕs and also in its WT counterpart. Sb^RLD infection resulted in significantly reduced IL-12 production in the culture supernatant in MyD88^−/− Mϕs as compared with infection with WT C57BL/6 Mϕs (Fig. 1C). Such differences in IL-12 and IL-10 production were absent in the case of Sb^SLD infection in MyD88^−/−and WT Mϕs (Fig. 1). Sb^RLD infection resulted in significantly reduced TNF-α production in the culture supernatant of MyD88^−/− Mϕs as compared with WT C57BL/6 Mϕs (Supplemental Fig. 1B), whereas IL-6 productions were comparable regardless of the nature of input parasites (Supplemental Fig. 1C). Because MyD88-mediated regulation of IL-10 or IL-12 was not observed in the case of Sb^SLD infection, all further experiments were performed with Sb^RLD parasite unless otherwise mentioned.

To determine whether increased IL-10 production from Sb^RLD (BHU575)-infected MyD88^−/− Mϕs has any influence on IL-12 production, we treated infected MyD88^−/− Mϕs with three different dilutions of IL-10 neutralizing Ab (1, 10, 100 ng/ml), where appropriate isotype was used as internal control. It was observed that treatment with anti-IL-10 Ab failed to restore IL-12 level in Sb^RLD-infected MyD88^−/−Mϕs significantly (Supplemental Fig. 1D). Efficacy of the anti–IL-10 Ab was tested in its ability to inhibit LPS-induced IL-10 production in uninfected Mϕs, and the resulting IL-10 production was analyzed by Western blotting. Our result showed that anti–IL-10 Ab is effective because it neutralizes LPS-driven IL-10 production (Supplemental Fig. 1E). To determine whether reduced IL-12 in Sb^RLD-infected MyD88^−/−Mϕs was due to a higher number of intracellular parasites, we cocultured MyD88^−/− and C57BL/6 Mϕs with the lysate of Sb^RLD normalized against the number of parasites and also in terms of protein content, and resulting IL-12 production was measured (Supplemental Fig. 1F). Incubation of both MyD88^−/− and C57BL/6 Mϕs with parasite lysates resulted in a comparable pattern of IL-12 production as infection with live parasites, indicating that IL-12 production in Sb^RLD-infected MyD88^−/−Mϕs is independent of number of intracellular parasites.

Because the absence of MyD88 dictates elevated IL-10 and concomitant decrease in IL-12 in Sb^RLD infection, the role of MyD88 in regulating IL-10 and IL-12 reciprocity in Sb^RLD-mediated infection was further investigated. RAW Mϕs were infected with Sb^RLD, with or without MyD88 overexpressing pUNO-MyD88 expression vector, and IL-10 and IL-12 levels were measured in the culture supernatant. There was a significant decrease in IL-10 and increase in IL-12 production in the culture supernatant of Sb^RLD-infected pUNO-MyD88–transfected cells (Fig. 1D) and a significant downregulation of IL-10 reporter and upregulation of IL-12 reporter as compared with untransfected Sb^RLD-infected cells (Fig. 1E). Similar experiment with pGL3-IL-10 and pGL3-IL-12 reporter constructs resulted in comparable observations. Thus, there is a clear association between cytokine assay in the supernatant and the reporter assays on IL-10 and IL-12.

To understand the role of MyD88 in Sb^RLD-mediated infection in an in vivo scenario, we transfected BALB/c mice with MyD88-lentiviral shRNA cassette to generate MyD88 KD mice; lentiviral-shRNA empty vector transfected mice were considered as control. To show that the Mϕs were effectively infected by shRNA lentiviral particles, GFP-encoding control shRNA was injected to the spleen tissue of BALB/c mice, and GFP expression in splenic Mϕs was evaluated by flow cytometry where CD11b⁺ Mϕs were gated from the whole splenocytes (Fig. 2AaI). The percentage of Mϕs in the splenocytes containing the GFP-encoding lentiviral cassette construct in normal untransfected (WT) mice (Fig. 2AaII), control empty cassette transfected mice splenic Mϕs (Fig. 2AaIII), and MyD88-lentiviral shRNA cassette transfected mice splenic Mϕs (Fig. 2AaIV) were also enumerated by flow cytometry. Expression of GFP in splenic CD11b⁺ cells was found to be 58.2% in control mice and 56.4% in MyD88 KD mice, respectively, at 4 wk posttransfer (Fig. 2Aa). Percentage of GFP⁺ cells for control group varied between 54 and 59% and for MyD88 KD group between 51 and 56% (Fig. 2Ab). Further, 47% decrease of MyD88 expression in the spleen tissue of MyD88 KD as compared with control mice was determined from Western blot analysis, which confirmed the generation of MyD88 KD mice (Fig. 2B).The hepatic and splenic parasite burdens at 4 wk p.i. were then determined p.i. with Sb^RLD (BHU575) parasites. It was observed that these were significantly higher in Sb^RLD-infected MyD88 KD mice (p < 0.001) as compared with control shRNA-treated mice (Fig. 2C, 2D, respectively). Splenocytes from individual mice from each of the groups (Fig. 2C, 2D) were stimulated with SLA and resulting cytokine levels in the culture supernatant were determined; there was a significant increase in IL-10 level, whereas IL-12 and TNF-α levels were severely decreased in Sb^RLD-infected MyD88 KD mice as compared with WT or control shRNA-treated mice (Fig. 2E–G, respectively). However, IL-6 production was relatively low and comparable among all three groups (Fig. 2H).

Because MyD88 was playing a critical role in regulating both IL-10 and IL-12 (Figs. 1, 2), it was important to find out how MyD88 was regulated in Sb^RLD infection. Determination of fold expression of MyD88 mRNA level in Mϕs infected with Sb^S or Sb^RLD revealed that the level remained essentially unaltered as compared with uninfected Mϕs (Fig. 3A), whereas control LPS treatment resulted in significant upregulation. However, Western blot analysis revealed that there was hardly any detectable expression of MyD88 protein in Mϕs infected with Sb^RLD, but not with Sb^SLD (Fig. 3B), and this was not restored in the presence of IL-10 neutralizing Ab at different dilutions (1, 10, 100 ng/ml) and using suitable isotype control (Fig. 3B). Moreover, addition of exogenous rIL-10 (10, 20, 200 pg/ml) in Sb^SLD (Ag83)-infected Mϕs failed to abrogate MyD88 expression (Supplemental Fig. 2A), indicating a possible involvement of host posttranscriptional machinery in regulating MyD88 level during Sb^RLD infection.

Microarray analysis was performed to estimate the expression of different miRNAs at different time points in Sb^RLD- or Sb^SLD-infected Mϕs against uninfected Mϕs (accession no. GSE57042, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE57042; R. Mukhopadhyay, S. Chowdhury, B. Mukherjee, S. Mukherjee, T. Mitra, J.C. Dujardin, S. Sundar, H.K. Majumder, and S. Roy, unpublished observations). Among a large number of differentially regulated miRNAs, miR-466i was screened because it showed complementarity at positions 751–757 and 768–775 with 3′UTR of MyD88 from TargetScanVer 5.1 (Supplemental Fig. 3) and significant upregulation of miR-466i in Sb^RLD-infected Mϕs (accession no. GSE57042, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE57042). Further, real-time analysis revealed that the expression of pri-miR-466i was ∼8-fold higher in Sb^RLD-infected Mϕs at 24 h p.i. (Fig. 3C). Finally, Northern blot and densitometry analysis of three independent experiments also confirmed the overexpression of miR-466i in Sb^RLD-infected Mϕs at 24 h p.i. (Fig. 3Da, b, respectively), where U6 RNA level was used as the loading control.

It was observed that pri-miR-466i expression was ∼8-fold higher in Sb^RLD-infected Mϕs at around 4 h p.i., which was maintained up to the time point the experiment was performed, that is, until 24 h (Fig. 4A). Checking the MyD88 protein status simultaneously by Western blot demonstrated that there was a significant decrease in MyD88 from 4 to 12 h p.i., and there was hardly any detectable MyD88 thereafter (Fig. 4B). Transfection of RAW264.7 cells with miR-466i mimic abrogated MyD88 expression in a dose-dependent manner (10–200 nM) as compared with the nonspecific control miR in RAW cells (Fig. 4C). Again, transfection of RAW cells with miR-466i inhibitor restored MyD88 expression in Sb^RLD-infected RAW cells 24 h p.i. (Fig. 4D). The specificity of miR-466i binding to 3′UTR of MyD88 was further confirmed from the luciferase activity in psiCHECK–3′UTRMyD88, psiCHECK–MUT3′UTRMyD88 (ΔMUT 751–757), and psiCHECK–MUT3′UTRMyD88 (ΔMUT 768–775) constructs. Infection with Sb^RLD or treatment with miR-466i mimics for 24 h resulted in significant reduction (2-fold, p < 0.001) of luciferase activity in psiCHECK–3′UTRMyD88 as compared with normal control. In contrast, only partial attenuation of luciferase activity was seen in the Sb^RLD-infected or miR-466i–treated MUT constructs, whereas no significant attenuation resulted upon deletion of both sites (ΔMUT 751–775 and ΔMUT 768–775; Fig. 4E).

Because it was evident that Sb^RLD-mediated activation of miR-466i plays a critical role in MyD88 deactivation, it was important to determine the mode of its activation. Use of the promoter prediction software suggested the miR-466i promoter to be present ∼5 kb upstream of the miR-466i gene with an accurate prediction score of 1.000. Promoter scanning of miR-466i revealed the presence of three NF-κB binding sites: site I (−5862/−5847), site II (−5537/−5558), and site III (−5516/−5502) (Fig. 5A). Real-time analysis at 4 h p.i. revealed ∼2.5-fold decrease in pri-miR-466i in Mϕs preincubated with an inhibitor of IKK (BAY-117082, inhibitor I) or NF-κB (SC-3060, inhibitor II) before Sb^RLD infection (Fig. 5B). Deletion of site I resulted in significant abrogation of luciferase activity (2-fold; p < 0.01) as compared with the whole promoter construct, whereas that of both sites I and II resulted in further decrease in activity (2.2-fold; p < 0.01). Deletion of all the three sites resulted in complete abolition of luciferase activity (Fig. 5C) in Sb^RLD-infected RAW cells, suggesting that all three NF-κB binding sites are important for Sb^RLD-mediated miR-466i overexpression at 4 h p.i. Further, Western blot analysis performed to determine possible NF-κB subunits in the nuclear extract of Sb^RLD-infected Mϕs at 4 h p.i. (Fig. 5D) showed that detectable amount of p50/c-Rel, but not p65, was present in the nuclear extract at 4 h post Sb^RLD infection.

Previously we showed that p50/c-Rel–dependent IL-10 surge starts from 12 h post Sb^RLD infection (9). A ChIP-qPCR analysis of all three NF-κB binding sites of miR-466i promoter showed significant histone acetylation with subsequent recruitment of p50/c-Rel in all three sites of miR-466i promoter in WT Mϕs, 4 h post Sb^RLD infection (Fig. 6A). Interestingly, there was no recruitment of p50/c-Rel in IL-10 promoter at this early time point of Sb^RLD infection as observed previously (Fig. 6B) (9). However, similar ChIP-qPCR analysis of IL-10 promoter in MyD88^−/−and IL-12^−/−Mϕs revealed significant recruitment of p50/c-Rel 4 h post Sb^RLD infection (Fig. 6B). Time kinetics of IL-10 production in Sb^RLD (BHU575)-infected MyD88^−/−and IL-12^−/−Mϕs also revealed significant upsurge of IL-10 level as early as 4 h p.i., but in the case of BHU575-infected C57BL/6 Mϕs, this was observed 12 h p.i. (Fig. 6C). Identical time kinetics of IL-10 was also observed in BHU138-infected C57BL/6 Mϕs, MyD88^−/−,and IL-12^−/−Mϕs (data not shown). Kinetics of IL-12 expression condition revealed elevated IL-12 level in Sb^RLD-infected C57BL/6 Mϕ 2 h p.i., which gradually diminished thereafter and remained essentially low from 8 h onward until 24 h p.i. However, in the case of Sb^RLD-infected MyD88^−/− Mϕs, IL-12 remained at low level throughout the course of infection (Fig. 7A). Our results thus suggest that the absence of MyD88/IL-12 might result in early activation of IL-10 in Sb^RLD-infected host as compared with their WT counterpart.

Survival kinetics of Sb^RLD (BHU575)-infected BALB/c, MyD88 KD, and control shRNA-treated mice was studied. It was observed that MyD88 KD mice started dying around 45 d p.i. and all of the MyD88 KD mice were dead by 60 d p.i. Infected BALB/c and control shRNA-treated mice remained alive up to the termination of the experiment, that is, 60 d p.i. (Fig. 7B). Histological study of the hepatic tissue of the above group of mice showed significantly higher infected Kupffer cells (KC) with a definite focused cellular response. Most of the granulomas present were at an immature stage of development, with very few mature granulomas (MGs) in the case of Sb^RLD-infected MyD88 KD mice (Fig. 7C, upper panel, 7D). In contrast, most of the infected KCs in Sb^RLD-infected BALB/c mice had a focal cellular response, and the granulomas formed had progressed to full maturity (MG), with organized cuffing of infiltrating cells (Fig. 7C, lower panel, 7D). Moreover, it was also evident that many of the granulomas formed in Sb^RLD-infected BALB/c mice were without L. donovani parasites, that is, they were sterile granulomas, and such granulomas were absent in MyD88 KD–infected mice.

Previously we have shown that Sb^RLD, but not Sb^SLD, interacts with cell-surface TLR2/TLR6 and produces IL-10 (9). Among the downstream signaling molecules of TLRs, absence of MyD88-adapter protein has been previously linked to high susceptibility to Leishmania major infection coupled with a polarized Th2 response (18). In this study, we show that infection of Mϕ with Sb^RLD results in higher number of amastigotes/100 Mϕσ and increased IL-10 production in MyD88^−/−Mϕs as compared with WT C57BL/6 Mϕs (Fig. 1A, 1B). MyD88 may play a specific role because inhibition of MyD88-independent TRIF pathway seems to have no effect on Sb^RLD-mediated IL-10 induction (Supplemental Fig. 1A). Sb^RLD infection also resulted in reduced IL-12 production in MyD88^−/−Mϕs (Fig. 1C), and this was not due to increased IL-10 because neutralizing Ab against IL-10 failed to restore the IL-12 level (Supplemental Fig. 1D). Moreover, incubation of MyD88^−/− Mϕs with parasite lysate also resulted in reduced level of IL-12 as compared with WT Mϕs, suggesting that reduction in IL-12 level in MyD88^−/− is independent of intracellular replicating parasites (Supplemental Fig. 1F). Furthermore, overexpression of MyD88 in Sb^RLD-infected RAW cells resulted in reduced IL-10 and increased IL-12 production (Fig. 1D, 1E), showing that Sb^RLD may modulate host MyD88 during infection. There is a previous report that also indicates that MyD88 plays a critical role in maintaining IL-12 level during infection with a number of intracellular pathogens (19). However, the earlier effects were not observed in the case of Sb^SLD infection (Fig. 1), pointing out that MyD88-dependent modulation might be specific to Sb^RLD. To validate the role of MyD88 in in vivo condition, we infected MyD88 KD BALB/c mice with Sb^RLD, which resulted in increased hepatic and splenic parasite burden as compared with their WT counterparts. SLA stimulation of whole splenocytes of MyD88 KD mice also resulted in reduced IL-12 and increased IL-10 (Fig. 2) levels, suggesting that MyD88 is important in regulating IL-10 and IL-12 levels in host during Sb^RLD infection, as reported previously in the case of Toxoplasma infection (39).

Interestingly, although MyD88 appeared to be playing a critical role in regulating IL-10 and IL-12 reciprocity in Sb^RLD-infected Mϕs, there was no significant difference in the level of MyD88 mRNA between Sb^RLD and Sb^SLD-infected Mϕs. However, there was hardly any detectable MyD88 protein in the case of Sb^RLD infection, unlike Sb^SLD infection (Fig. 3B).There is a previous report that reveals that IL-10 negatively regulates MyD88-dependent signaling in response to LPS stimulation (40). Because infection with Sb^RLD results in increased IL-10 production in host (9), it is possible that increased IL-10 level might regulate MyD88-dependent IL-12 level in Sb^RLD-infected host. The notion that downregulation of MyD88 in Sb^RLD infection is independent of increased IL-10 stems from the following observations: 1) Ab to IL-10 failed to restore MyD88 in Sb^RLD-infected cells (Fig. 3B), and 2) exogenous rIL-10 treatment failed to show any effect over MyD88 expression in Sb^SLD-infected Mϕs (Supplemental Fig. 2A). These observations further indicate that IL-10 surge may have no role in regulating MyD88 expression during L. donovani infection.

We then investigated the role of posttranscriptional machinery that might be involved in regulating MyD88 function during Sb^RLD infection. A recent report has also characterized the expression of a large number of miRNAs in Leishmania-infected human phagocytes (23). Sb^RLD infection leads to upregulation or downregulation of a large number of miRNAs in host (accession no. GSE57042, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE57042). Among these, miR-466i, which has been previously reported to increase the stability of IL-10 mRNA, leading to its overproduction (41), was also observed to be upregulated in Sb^RLD-infected host (accession no. GSE57042, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE57042). Interestingly, target scan analysis identified two binding sites for miR-466i at 3′UTR of MyD88 (Supplemental Fig. 3). From real-time analysis and Northern blot it was evident that miR-466i is overexpressed in Sb^RLD-infected Mϕs, but not when infection is with the sensitive counterpart (Fig. 3C, 3D).

To determine the role of miR-466i, if any, in Sb^RLD-infected Mϕs, kinetics of miR-466i expression was studied in corroboration with the MyD88 level (Fig. 4A, 4B). Our study revealed upregulation of miR-466i expression 4 h p.i. in Sb^RLD-infected Mϕs, which was associated with initiation of abrogation MyD88 expression. The importance of miR-466i in the posttranscriptional regulation of MyD88 was further substantiated by using 466i mimics and inhibitors. Reduced level of MyD88 in the presence of 466i mimics, as opposed to stability of MyD88 in the presence of 466i inhibitor, lends credence in favor of miR-466i as downregulator of MyD88 (Fig. 4C, 4D). Moreover, deletion of miR-466i binding site in 3′UTR of MyD88 resulted in restoration of luciferase activity of the 3′UTR reporter construct, suggesting that miR-466i indeed binds with 3′UTR of MyD88 and regulates its level posttranscriptionally during Sb^RLD infection (Fig. 4E). It should be mentioned that although miR-466i seems to play a critical role in regulating MyD88 expression in Sb^RLD-infected Mϕs, this observation needs to be validated in an in vivo model in the future to determine the definite role of miR-466i during Sb^RLD infection.

The mechanism of Sb^RLD-mediated miR-466i activation was thereafter investigated. Promoter prediction tool identified a probable promoter region with three potential NF-κB binding sites 5 kb upstream of the miR-466i gene (Fig. 5A). Pretreatment of Mϕs with NF-kB inhibitors resulted in significant abrogation of Sb^RLD-mediated miR-466i upregulation (Fig. 5B). From the site-directed mutagenesis promoter deletion analysis it was evident that all of these three NF-κB binding sites are required for optimal activation of miR-466i in Sb^RLD-infected RAW cells (Fig. 5C), proving that miR-466i activation in Sb^RLD is NF-κB dependent. There is a previous report on the NF-κB p65-dependent transactivation of miRNA genes post Cryptosporidium pavum infection in epithelial cells (42). Our results revealed nuclear translocation of p50/c-Rel, but not of p65, even at 4 h p.i., which corresponds with the increased expression of miR-466i in Sb^RLD-infected Mϕs (Fig. 5D).

ChIP-qPCR analysis revealed recruitment of p50/c-Rel at all three NF-κB binding sites of miR-466i promoter, but not at the NF-κB binding site of IL-10 promoter 4 h post Sb^RLD infection (Fig. 6A, 6B). Previously we have shown recruitment of p50/c-Rel at the specific site of IL-10 promoter in Sb^RLD-infected Mϕs at 12 h p.i (9). However, in the case of MyD88^−/−Mϕs or IL-12^−/−Mϕs, p50/c-Rel was recruited at the specific site of IL-10 promoter even at 4 h p.i. (Fig. 6B). This observation also goes hand in hand with an early surge in IL-10 level at 4 h p.i. in Sb^RLD-infected MyD88^−/− or IL-12^−/− Mϕs, whereas Sb^RLD-mediated IL-10 surge occurred at 12 h p.i. in WT Mϕs (Fig. 6C), as previously reported by our group (9). This observation reveals that absence of IL-12 and MyD88 might result in early recruitment of p50/c-Rel in the IL-10 promoter of Sb^RLD-infected Mϕs.

Previously we have reported a low level of IL-12 in Sb^RLD-infected WT Mϕs at 24 h p.i (9). However, time kinetics of IL-12 production in Sb^RLD-infected Mϕs revealed an initial increase in IL-12 level from 0 to 2 h p.i., which started decreasing significantly from 4 h p.i. and remained essentially low thereafter up to 24 h p.i. as observed previously (9). In MyD88^−/− Mϕs, in contrast, IL-12 level remained low throughout the course of Sb^RLD infection (Fig. 7A), indicating that MyD88 is essential for maintaining the IL-12 level during Sb^RLD infection. It is therefore tempting to speculate that increased IL-12 level might prevent binding of p50/c-Rel and subsequent activation of IL-10 in the early hours of Sb^RLD infection. There are previous reports that also suggest that presence of IL-12 might hinder activation of IL-10 promoter (20, 43). There is a previous report that during the early point of infection, toxoplasma stimulates a potent IL-12 response that leads to IFN-γ–dependent control of its replication (44). In turn, a series of in vitro and in vivo studies has demonstrated that dendritic cells, as well as neutrophils and IFN-γ–primed Mϕs, produce IL-12 in response to Toxoplasma gondii stimulation (45, 46). Further, survival kinetics revealed decreased survival of MyD88 KD mice in response to Sb^RLD infection as compared with WT BALB/c mice (Fig. 7B). From the histological study it was evident that there was qualitative difference in the nature of granuloma formation between Sb^RLD-infected BALB/c and MyD88 KD mice (Fig. 7C, 7D). Although there was increased sinusoidal dilation, numerous parasitized KC within immature granuloma in Sb^RLD-infected MyD88 KD mice, Sb^RLD infection of WT BALB/c mice resulted in numerous capsulated MGs with few L. donovani bodies, pointing to a nonpersistent infection. Previously, MG formation has been associated with clearance of Leishmania infection (36). An earlier report also suggested that the level of IL-12 might act as a critical factor in determining the nature of granuloma formation during experimental infection with VL (47). Because a parasite would not like to kill its host for the sake of its own survival, it is tempting to speculate that early activation of IL-12 and delayed activation of IL-10 in Sb^RLD infection might be instrumental for survival of resistant parasites in the infected host. A previous report also found decreased survival rate of MyD88^−/− mice in response to T. gondii infection (48). Taken together, our results suggest that at an early time point of Sb^RLD infection, MyD88 plays a vital role in maintaining IL-12 level in the host. This IL-12 might not allow activation of IL-10 promoter at an early time point of Sb^RLD infection (20). However, with subsequent increase in miR-466i level there is a degradation of MyD88, coupled with reduction in IL-12 level. Eventually, reduced IL-12 level might result in subsequent activation of IL-10 promoter, leading to an IL-10 surge in the host (Fig. 8). In this context it may be mentioned that although miRNA mediated control of TLR signaling molecules, either through mRNA decay or by translational inhibition, it might not be as rapid as proteasomal degradation, but this might provide a selective advantage during infection (21). Thus, restoration of MyD88 might act as a critical target in Sb^RLD parasite-mediated leishmaniasis.

The authors have no financial conflicts of interest.