MyD88 and TLR9 Dependent Immune Responses Mediate Resistance to Leishmania guyanensis Infections, Irrespective of Leishmania RNA Virus Burden

Infections with Leishmania parasites of the Leishmania Viannia subgenus give rise to both localized cutaneous (CL), and metastatic leishmaniasis. Metastasizing disease forms including disseminated (DCL) and mutocutaneous (MCL) leishmaniasis result from parasitic dissemination and lesion formation at sites distal to infection and have increased inflammatory responses. The presence of Leishmania RNA virus (LRV) in L. guyanensis parasites contributes to the exacerbation of disease and impacts inflammatory responses via activation of TLR3 by the viral dsRNA. In this study we investigated other innate immune response adaptor protein modulators and demonstrated that both MyD88 and TLR9 played a crucial role in the development of Th1-dependent healing responses against L. guyanensis parasites regardless of their LRV status. The absence of MyD88- or TLR9-dependent signaling pathways resulted in increased Th2 associated cytokines (IL-4 and IL-13), which was correlated with low transcript levels of IL-12p40. The reliance of IL-12 was further confirmed in IL12AB−/− mice, which were completely susceptible to infection. Protection to L. guyanensis infection driven by MyD88- and TLR9-dependent immune responses arises independently to those induced due to high LRV burden within the parasites.


Introduction
In Latin America, infections with protozoan, obligate intracellular parasites of the Leishmania Viannia subgenus, such as L. braziliensis, L. panamensis and L. guyanensis give rise to cutaneous (CL), and metastasizing leishmaniasis, which includes disseminated (DCL) and mucocutaneous (MCL) leishmaniasis. Infections occur following transmission to humans via the bite of infected sand flies [1]. MCL is characterized by parasitic dissemination that promotes destructive secondary skin lesions. This metastasis may have tissue-specificity, spreading to the mucosa of mouth and nose in 5 to 10% of individuals with resolved CL [1,2]. MCL pathology is associated with chronic hyper-inflammation, concomitant with a modulation of T helper cytokines and cytotoxic T cell activity [3][4][5][6][7]. Studies have shown that host genetic polymorphisms (such as a single base pair substitution in the IL-6 and TNF-a promoters) [8][9][10], immune-status, and parasite derived virulence factors [11] are associated with the development of clinical MCL.
Previously, we reported that elevated pro-inflammatory immune response generation in L. guyanensis infected macrophages relies on the recognition of high levels of an endosymbiotic dsRNA virus of the Totiviridae family within metastasizing L. guyanensis promastigotes (derived from either the hamster model or human MCL isolates). The viral dsRNA genome of the Leishmania RNA virus (LRV) is recognized by the TRIF-dependent TLR3 signaling pathway, which resulted in an increased secretion of inflammatory cytokines and chemokines and exacerbated susceptibility to infection [3]. Although the absence of TLR7 in macrophages decreased LRV-induced pro-inflammatory mediators in vitro, it showed no role in progression of leishmanial disease in our mouse model [3].
In CL, a T-helper 1 (Th1) response is associated with protection, and given by high IFN-c production and efficient intracellular parasite killing by nitric oxide. Conversely, a Th2 response is associated with susceptibility to infection which is driven by high IL-4, and IL-13 and reduced intracellular parasite killing due to increased Arginase-1 activity [12]. Furthermore, immunosuppressive IL-10 also impairs Leishmania parasite disease resolution and promotes lesion persistence and high parasite burden [12]. Previous work has clearly demonstrated that, for both L. major and L. (Viannia) braziliensis, the absence of MyD88dependent signaling pathways modulated the Th1/Th2 balance following in vivo footpad infection and resulted in increased susceptibility to infection [13][14][15][16][17][18]. Further studies described that the impaired ability of MyD88-deficient (MyD88 2/2 ) mice to develop Th1-dependent immune responses was due to reduced IL-12 production [17]. This consequently elevated production of the Th2 cytokine, IL-4 and IL-13, may explain why these mice displayed reduced dendritic cell maturation and naïve T cells priming following infection [14,15,19]. Furthermore, a role for TLR9 was also implicated in these responses as both models displayed a transient increase in susceptibility, evidenced by an increased peak in footpad swelling, associated with elevated parasite levels [18,20,21].
The TLR9 signaling pathway is triggered after the recognition of foreign unmethylated CpG DNA. Its position on the internal aspect of the macrophage phagosomal compartment and plasmocytoid dendritic cells makes it especially sensitive to intracellular pathogens, such as is Leishmania. All Toll-like receptors rely on the MyD88 adaptor protein for signal transduction (except TLR3, which is uniquely TRIF-dependent) and may thus explain the common susceptible phenotype of MyD88 2/2 and TLR9 2/2 mice for L. major and L. braziliensis infection.
Given the potential application for TLR-ligands such as CpG motifs to act as adjuvants in vaccination strategies [22,23] this study assesses if resistance to L. guyanensis parasite infection is mediated by MyD88, and TLR9 [22,23]. More importantly our study evaluates if host recognition of LRV1 could impair these protective immune responses, which will provide much needed information into the applicability for TLR-ligand based adjuvants being used in vaccination strategies against leishmaniasis in Latin America.

Ethical Statement
All the mice experiments were approved by the Ethics and Veterinary office regulations of the state of Vaud (SAV), Switzerland. Our laboratory has the administrative authorization numbers 2113-1 and 2113-2. Animals 8 to 10 weeks old C57BL/6 mice were purchased from Harlan Laboratories (Netherlands). MyD88 2/2 [24] and TLR9 2/2 mice [25] were obtained Prof. S. Akira (Osaka University, Japan) via P. Launois (WHO-IRTC, Lausanne, Switzerland). IL12AB 2/2 mice [26] were obtained from G. Alber (University of Leipzig, Germany). Mice were maintained under pathogen-free conditions at the animal facility of the Center of Immunity and Immunology, Lausanne (Switzerland). All mutant and deficient mice were crossed onto a C57BL/6 background for at least eight generations. Each experiment contained at least 4 mice per group.
Mouse Infection and Parasite Quantification 3610 6 parasites of L.g. LRV low (Lg17), and L.g. LRV high (M5313) were infected into the hind footpads of mice. Weekly footpad swelling measurements were taken using a Vernier caliper.
Parasite burden was quantified within the L. guyanensis infected footpads using the standard curve real time PCR quantification method based on Leishmania Kmp11 specific primers: 59-GCCTGGATGAGGAGTTCAACA-39 and 59-GTGCTCCTTCATCTCGGG-39on cDNA reverse-transcribed from total RNA extracted from footpad lysates. Briefly, footpads were homogenized using the Tissuelyzer (Qiagen), and total RNA was extracted using TRIzol reagent (Invitrogen). cDNA was synthesized using SuperScript II Reverse Transcriptase (Invitrogen), followed by purification using the QIAquick PCR purification Kit (Qiagen). Gene expression levels were analyzed using the LightCycler480 system (Roche Applied Science). Absolute parasite burden was determined using a standard curve for Leishmania Kmp11 gene content with a pre-determined number of L. guyanensis parasites.

In vitro Restimulation of Lymph Node Cells and Cytokine Quantification
Draining popliteal lymph nodes were extracted from mice and stimulated in complete DMEM (Gibco) with a final concentration of 10% heat-inactivated FBS, 1% penicillin/streptomycin, 2 mM L-glutamine, 5610 25 M of b-mercaptoethanol and 1% HEPES (Sigma-Aldrich) at 5610 6 /ml cells with 1610 6 /ml UV-irradiated L. guyanensis promastigotes. After 72 hours, levels of IFN-c, IL-4, IL-10, and IL-13 (eBioscience) and IL12p70 (BD Biosciences) were quantified by ELISA in cell-free culture supernatants. In each case the manufacturer's protocols were explicitly followed.

Quantification of Cytokine Transcripts in the Draining Lymph Nodes of Infected Mice
RNA was extracted from draining popliteal lymph node cell suspensions using TRIZol reagent (Life technologies) and then clean and concentrated using appropriate columns from Zymoresearch. cDNA was reverse-transcribed and purified as described previously [3]. Relative quantification at the transcript level was performed by qRT-PCR using the Lightcycler480 system (Roche applied sciences) and results were calculated using the 2 2DDCT method with Tata binding protein (Tbp) as a reference gene and normalized to give the average expression level of the WT C57BL/6 a value of 1. Gene specific primers used were the following: Tbp-59-CCGTGAATCTTGGCTGTAAAC-39, 39-TCCAGAACTGAAAATCAACGC-59; Il4-59-CGGAGATG-GATGTGCCAAAC-39,

Results
MyD88 and TLR9 are Involved in the Development of Resistance to L. guyanensis Infection in Mice, Irrespective of LRV Presence The ability of TLR9 and MyD88 to impact immune responses against L.g. LRV high and L.g. LRV low parasites was investigated using subcutaneous footpad infection in a murine model. These two isolates were derived from the M5313 strain of L. guyanensis based on reproducible disseminated lesion development in the Golden hamster model of infection [27]. The ability of these isolates to induce secondary lesion development could be reliant on their high LRV viral burden [3].
Following infection, we could ascertain that L. guyanensis isolates, regardless of their LRV burden, displayed similar disease evolution phenotypes in MyD88 2/2 and TLR9 2/2 mice. More specifically, we observed that L. guyanensis infection in MyD88 2/2 mice induced non-healing, progressive lesions that were associated with increased parasite load (Fig. 1). However, only the MyD88 2/2 infected with the L.g. LRV high developed sever ulcers at the footpad injection site (Fig. 1A). This increase in susceptibility was correlated to an increased parasite burden in the footpad (Fig. 1C-D). On the another hand, the TLR9 2/2 mice displayed a transient increased peak in footpad swelling (Fig. 1A-B) that correlated with a transient increased in parasitemia at 4 weeks post infection (Fig. 1), and this after being infected either with L.g. LRV low or L.g. LRV high parasites. Interestingly, the L.g. LRV high infected TLR9 2/2 mice with resolved lesions retained a higher degree of parasitemia as compared to wild-type C57BL/6 mice at week 10 (Fig. 1C).
Additionally, we also analyzed the role of other MyD88dependent TLRs and from our results we could exclude a likely role for TLR2, and TLR4 signaling pathway in resistance to L. guyanensis as these mice display similar footpad swelling profiles as wild-type mice following infection with either L.g. LRV high or L.g. LRV low parasites (Fig. S1).
Increased Susceptibility of MyD88 2/2 and TLR9 2/2 Mice to L. guyanensis Infection is Associated with the Development of a Th2 Response Given the clear increase in susceptibility of both the MyD88 2/2 and the TLR9 2/2 mice infected with both strains of L. guyanensis parasites, the immune response within the draining lymph nodes (LN) of the mice was investigated. Analysis of cytokine production in supernatants of L.g. LRV high stimulated LN cells revealed that MyD88 2/2 or TLR9 2/2 mice mounted a typical Th2 cell response with high levels of IL-4, IL-13 associated with a decrease in IFN-c production comparison with control mice (at week 4 and 10) (Fig. 2A). L.g. LRV low infection in MyD88 2/2 and TLR9 2/2 mice, also induced Th2 immune responses as shown by high IL-13 and IL-4 production nevertheless this response was less pronounced than that of the same mice infected with L.g. LRV high (Fig. 2B). Interestingly, TLR9 2/2 mice infected with L.g. LRV low produced significantly more IFN-c at week 10 and this could be to counteract the earlier Th2 initiated response (Fig. 2B). Similar results were obtained by quantitative real time PCR (qRT-PCR) Figure 1. MyD88 2/2 and TLR9 2/2 mice are more susceptible to infection with L. guyanensis. Wild type C57BL/6, MyD88 2/2 and TLR9 2/2 mice (n$4) were infected into the hind footpads. Footpad swelling of mice infected with either L.g. LRV high (A) or L.g. LRV low (B) was measured weekly over 10 weeks using a Vernier caliper. Parasite burden in footpads of L.g. LRV high (C) or L.g. LRV low (D) infected mice was determined at indicated time points by quantitative-real time PCR using parasite-specific KMP11 primers. Results are expressed as the mean6 SEM. Significance determined at *p# 0.05, **p#0.01, ***p#0.005. doi:10.1371/journal.pone.0096766.g001 analysis on the draining lymph node (dLN) cells for Th1 and Th2 associated cytokines on all experimental mice groups (Fig. S2). This clearly demonstrates that MyD88 and TLR9 deficient mice have impaired immune response generation after infection with L.guyanensis, regardless of the presence of LRV virions.
The Increased Susceptibility of Mice Infected with L. guyanensis Parasites is due to the Impaired Production IL-12 Given that previous reports have demonstrated that MyD88 2/2 mice following infection with different Leishmania species have an impaired ability to produce IL-12, we tested whether this could also be the case in LRV harboring L. guyanensis infected mice. Firstly, we analyzed the transcript levels of IL-12p40 and IL-12p35 in the draining LN of L.g. LRV high or L.g. LRV low infected MyD88 2/2 and TLR9 2/2 mice as compared to controls. We observed that both the MyD88 2/2 and the TLR9 2/2 mice infected with L.g. LRV high parasites had significantly diminished transcripts of IL-12p40 and IL-12p35 except for the latter in MyD88 2/2 mice at 10 weeks post infection (Fig. 3A). In L.g. LRV low infected mice, there was no consistent modulation of IL-12p35 as compared to WT mice at both time points analyzed (Fig. 3A), whilst IL-12p40 transcripts were diminished in MyD88 2/2 and TLR9 2/2 mice only at 4 weeks post infection (Fig. 3A). As modulation at the RNA level does not necessarily reflect those at the protein level, we attempted to quantify the amount of IL-12p40 protein in the culture supernatant by ELISA but the levels were below the detection limit (data not shown).
Finally, the essential role of IL-12 in the development of resistance to infection was tested by analyzing the disease evolution profile and development of a Th1 immune response in IL-12AB 2/ 2 mice (that do not express functional IL-12) in infections with L.g. LRV high parasites. As with infections with other Leishmania strains, IL12AB 2/2 mice were completely susceptible to infection, resulting in progressive, ulcerating lesion development and increased parasite burden (Fig. 3B and 3C). This necessitated the sacrifice of these mice at week 8 and confirmed that IL-12 was sufficient for the development of resistance to infection in C57BL/ 6 mice infected with L. guyanensis parasites, even when carrying a high burden of LRV.

Discussion
The purpose of this study was to elucidate the role of MyD88 and TLR9 in disease progression and immune response generation in mice infected with L. guyanensis, and to determine if a high Leishmania RNA virus (LRV) burden within promastigotes could impair these protective immune responses.
Our experimental data clearly supported a role for the innate immune system and MyD88-dependent signaling pathways in the development of resistance to infection with L. guyanensis, and also implied a partial role for the MyD88 dependent, TLR9 signaling pathway. The results show that protective immune responses elicited via MyD88 and TLR9 arise independently to those mediated by TLR3 and viral dsRNA recognition. As evidenced by the fact that infection with either L.g LRV high or L.g LRV low parasites into the MyD88 2/2 or TLR9 2/2 mice were more susceptible than wild type controls regardless of the viral burden. However, presence of a high LRV burden does influence overall disease severity, where all mice infected with L.g. LRV high showed an increased footpad swelling peak as compared to L.g. LRV low . This increase in susceptibility is reliant on TLR3 recognition of viral dsRNA that seems to modulate Th1/Th2 immune response polarization or to impaired parasite killing mechanisms.
As with other studies, we observed that the susceptible disease phenotype was more severe in MyD88 2/2 mice than in TLR9 2/ 2 , and implies that other TLRs (that utilize MyD88) could participate in the disease resolution and elimination of parasites. Currently, we can exclude a role for TLR7 and the cell membrane localized TLR4 and TLR2, as mice deficient for these receptors do not display any discernable difference in disease phenotype in either the L.g LRV high or L.g LRV low infected mice [3].
In this study, we specifically demonstrated that the loss of MyD88 and TLR9 signaling pathways resulted in a modulation of the Th1/Th2 balance, with a significant increase in Th2 associated cytokines (IL-4 and IL-13) in the draining lymph nodes of both MyD88 2/2 and TLR9 2/2 mice. This Th2-polarized immune response could also be correlated with a diminished production of IL-12p40 and IL-12p35 transcripts ex vivo, following infection, which we also confirmed was essential for healing and elimination of parasites. Although other studies have demonstrated a role for both MyD88 and TLR9 in the generation of IL-12 and IFN-c-dependent protective immune responses murine models of infection with other Leishmania parasites (specifically L. major, L. donovani and L.(V) braziliensis [13][14][15]17,[19][20][21]) this study demonstrates that the viral immune response driven by LRV does not impair protection conferred by MyD88 and TLR9 pathways nor those elicited due to IL-12.
Although there is currently no vaccine available against leishmaniasis, the past has shown us that protection is, indeed, possible. Leishmanization -the injection of live parasites into an individual to induce protection-was a common practice in the Middle East but adverse events such as uncontrolled skin lesions, HIV-co-infections, and variability in protection limited it potential application on a large scale. Thus, several different vaccine strategies have been investigated in the pursuit of the anti-Leishmania vaccine, including whole organism formulations using live-attenuated or killed organisms, and defined vaccines targeting Leishmania specific motifs such as the LACK antigen or DNAbased vaccines. Unfortunately, however, poor immunogenicity and other factors have limited their success.
Given this limitation, CpG based motifs (that are recognized by TLR9) and other MyD88 stimulating ligands have been investigated for their potential as adjuvants to boost the development of protective immune responses against Leishmania parasites [28,29]. Studies have shown that regardless of the vaccine antigen used CpG-oligodeoxynucleotides (CpG-ODN) enhanced the protection of mice subsequently challenged with the parasites. Specifically, vaccination of susceptible BALB/c mice with soluble Leishmania antigen (SLA) and CpG provided enhanced protection in challenged mice that persisted for over 6 months [30]. This protection was linked with the development of memory CD4 and CD8 T cell immune responses [31]. Given that TLR9 responses remain protective regardless of the LRV burden in infecting parasites, CpG-based adjuvants stand as a promising addition for vaccines serving regions that are endemic to parasites of mixed LRV-content.
It is important to note, however, that TLR-stimulation can result in potent inflammatory responses that could exacerbate severity of Leishmania infections, as we previously reported for the pathogenic TLR3 stimulation by high viral dsRNA burden in L. guyanensis [3]. Thus, further experimentation is required to gauge the effects of all TLR-based adjuvants on infections with parasite species to better tailor their therapeutic use in leishmaniasis.

Supporting Information
Figure S1 TLR2 2/2 and TLR4 2/2 mice display no difference in footpad swelling following L. guyanensis infection. Mice (n$5) were infected into the hind footpads with either L.g. LRV high or L.g. LRV low , and footpad swelling was measured weekly over 10 weeks using a vernier caliper. (TIF) Figure S2 Absence of MyD88 and TLR9 signaling pathways increases transcripts of Th2 cytokines. At specific time points RNA from draining popliteal lymph node cells of wild-type, TLR9 2/2 and MyD88 2/2 mice infected by L.g LRV high (A) or L.g LRV low (B) were reverse-transcribed into cDNA and relative transcript levels were determined using gene specific primers by quantitative real time PCR. Results were expressed as mean6 SEM of the individual mice analyzed per group with the average value for the C57BL/6 mice given a value of 1. Tbp was used as a reference gene. A.U.: arbitrary units corresponding to fold changes in expression. Significance determined at *p#0.05, **p#0.01, ***p#0.005. (TIF)