Lymphocytes influence Leishmania major pathogenesis in a strain-dependent manner

Cutaneous leishmaniasis (CL) is the most common form of leishmaniasis and is caused by several species of Leishmania parasite. Clinical presentation of CL varies from a self-healing infection to a chronic form of the disease determined by the virulence of infecting Leishmania species and host immune responses to the parasite. Mouse models of CL show contradictory roles of lymphocytes in pathogenesis, while acquired immune responses are responsible for host protection from diseases. To reconcile the inconclusive roles of acquired immune responses in pathogenesis, we infected mice from various genetic backgrounds with two pathogenic strains of Leishmania major, Friedlin or 5ASKH, and assessed the outcome of the infections. Our findings showed that the genetic backgrounds of L. major determine the impact of lymphocytes for pathogenesis. In the absence of lymphocytes, L. major Friedlin induced the lowest inflammatory reaction and pathology at the site of infection, while 5ASKH infection induced a strong inflammatory reaction and severe pathology. Lymphocytes ameliorated 5ASKH mediated pathology, while it exacerbated pathology during Friedlin infection. Excess inflammatory reactions, like the recruitment of macrophages, neutrophils, eosinophils and production of pro-inflammatory cytokines, together with uncontrolled parasite growth in the absence of lymphocytes during 5ASKH infection may induce severe pathology development. Taken together our study provides insight into the impact of differences in the genetic background of Leishmania on CL pathogenesis.


Introduction
Leishmaniasis is a group of diseases caused by protozoan parasites belonging to Leishmania species. Leishmaniasis threatens more than 350 million people from 98 countries in tropical and subtropical regions of the world with an estimated 0.7-1 million new cases and 20,000 to 30,000 deaths annually [1]. The disease can be categorized into three major forms: cutaneous, mucocutaneous and visceral leishmaniasis [2]. Cutaneous leishmaniasis (CL) is the most common form of the disease. Clinical manifestations of CL vary from asymptomatic to a non-healing chronic form of the disease determined by the infecting Leishmania species and by the host immune response to the parasite [3,4]. The biological processes that may help to explain why some cases are asymptomatic and others show a variety of clinical pathologies, across the spectrum of self-healing to severe forms of leishmaniasis, are ill-defined. The diverse clinical manifestations of CL demand a thorough investigation of the pathophysiology of the infection in varying host defenses and parasite virulence.
Mouse models of leishmaniasis have been extensively studied to understand and characterize the host-parasite interactions during infections [5]. In mouse models of CL, the mechanism of resistance to Leishmania major infection has been well established using genetically resistant C57BL/6 mice that reproduce the self-cure disease outcomes typically seen in humans [3]. Resistance to L. major infection is mainly mediated by the CD4 + and CD8 + T lymphocytes, represented by the induction of Th1 immune response and IFN-γ dependent killing of the parasite by macrophages [5,6]. Conversely, L. major infection drives BALB/c mice towards a Th2 response that allows for parasite replication and persistence [5].
While the role of the acquired immune response in protection against leishmaniasis has been well documented, its role in pathogenesis is not fully understood. Experimental CL studies showed the discrepant role of T/B lymphocytes in disease pathology. Some studies showed that lesion development was independent of T/B cells in CL models using L. major [7] or Leishmania amazonensis [8]. Other studies showed that lesion development caused by L. major [9,10] or L. amazonensis [11] infection was delayed and less severe in the absence of T/B cells. Soong et al. showed that CD4 + T lymphocytes are indispensable for pathogenesis during L. amazonensis infection [12]. In these studies, different species or isolates of Leishmania parasite were used. The discrepant roles of T/B cells in pathogenesis may be due to the differences in parasite species or strains. It is reasonable to hypothesize that the genetic background of the Leishmania strain is crucial to determine the variable influence of lymphocytes for pathology development.
To investigate our hypothesis, we infected mice of various background with two strains of L. major with different degrees of pathogenicity, namely strains Friedlin and 5ASKH, and measured the disease outcome. Both L. major Friedlin and 5ASKH infections were controlled in resistant C57BL/6 mice with transient lesion development and limiting parasite burden. Susceptible BALB/c mice showed progressive lesion development with uncontrolled parasite growth. In the absence of T/B lymphocytes, 5ASKH strain caused more progressive lesion development, while Friedlin strain caused limited lesion development whilst maintaining similar levels of parasite burden to 5ASKH strain. Our results indicated that lymphocytes suppressed pathogenesis during 5ASKH infection, while the presence of lymphocytes during Friedlin infection upregulated pathogenesis. We conclude that the genetic background of L. major determines the distinct influence of lymphocytes for the pathogenesis of this organism.

Ethics statement
The study protocol was approved by the

Animals
Six to eight weeks old female mice were used in all experiments. In this study, we included wild-type (WT) and nude BALB/c, WT and recombination activating gene 2 (Rag2) knockout (KO) C57BL/6, and NOD-SCID mice. WT BALB/c (BALB/cCrSlc) and C57BL/6 (C57BL/ 6JJmsSlc) mice were purchased from SLC, Japan. Nude BALB/c (CAnN.Cg-Foxn1 nu /CrlCrlj) and NOD-SCID (NOD.CB17-Prkdc scid/J) mice were purchased from Charles River Laboratories, Japan. Rag2 KO C57BL/6 (B6(Cg)-Rag2tm1.1Cgn/J) mice were purchased from the Jackson Laboratory. Experimental mice were maintained under specific pathogen-free conditions at the Animal Facilities of Nagasaki University. Experimental animals were fed with standard rodent pellets supplemented with grain and given water ad libitum. Throughout this study, the animals were housed in a climate-controlled (23±2˚C; relative humidity, 60%) and photoperiod controlled (12-hours light-dark cycles) animal quarters. Only healthy mice were included in the study.

Leishmania parasite and infection
L. major clone V9 (MHOM/IL/80/Friedlin) and L. major (MHOM/SU/73/5-ASKH) strain abbreviated as L. major Friedlin and L. major 5ASKH respectively were used in this study. The parasite culture was maintained in medium 199 (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (CCB, Japan), 100 U/ml penicillin and 100 μg/ml streptomycin (Gibco) at 26˚C in a biological incubator. Cultures were passaged to fresh medium at a 50-fold dilution in every 2-3 days. Stationary phase promastigotes were harvested and suspended in 1× PBS. Mice were infected with 5×10 6 stationary phase promastigotes of L. major Friedlin or 5ASKH by subcutaneous injections into their right hind footpad. Following infection, footpad thickness was measured weekly using an analog caliper. Footpad swelling was calculated by subtracting the thickness of the uninfected counter footpad from the thickness of the infected footpad. Parasite burdens in the footpads and popliteal lymph nodes (pLN) were quantified at 6 weeks post-infection. Parasite burdens in the footpads and draining lymph nodes were quantified by either limiting dilution assay (LDA) or homogenizing tissue in 10 ml of complete medium 199 and incubated in 25cm 2 culture flasks at 26˚C for the detection of parasites transformed in vitro from amastigotes to motile promastigotes. At 3 days of culture, promastigotes were counted microscopically. For LDA tissues were homogenized in 1 ml complete medium 199 following 2 fold serial dilution across 96-well culture plates. Culture plates were incubated at 26˚C. At 1-week culture plates were observed under a microscope for the presence of motile promastigotes.

Isolation of cells from the footpad of mouse
Mice were euthanized and the footpads have collected aseptically. The footpads were cut tangentially to the bone and mechanically disrupted with surgical scissors in 5 ml cold RPMI medium supplemented with 1 mg/ml collagenase (WAKO, Japan) and 1 mg/ml DNase (Roche Diagnostics, Japan). Disrupted footpads were then incubated for 30 min in a 37˚C water bath with continuous shaking. Following incubation, the reactions were stopped by adding cold RPMI medium with 2.5 mM EDTA and strained through a 70 μm cell strainer. The cell suspensions were centrifuged at 440g for 6 min at 4˚C and re-suspended in complete RPMI medium supplemented with 10% heat-inactivated fetal bovine serum, 100 U/ml penicillin and 100 μg/ml streptomycin.

Quantitative RT-PCR
Mice footpads were homogenized with 1 ml TRIzol (Thermo Fisher Scientific) and φ1.0 stainless steel beads in the 2 ml tube using Micro Smash MS100R (TOMY, Tokyo, Japan) at 4˚C using tissue homogenizer (Tomy, Micro Smash MS-100). RNA was isolated using Trizol Reagent according to manufacturer instructions. RNA yield was measured by a spectrophotometer (Beckman Coulter DU-730). We found 30-53 μg RNA from each sample, and 2 μg of total RNA was used as the template for the synthesis of 20 μl cDNA. Synthesized cDNA was analyzed for mouse TNF-α, IL-1β, IL-6, IL-12 p40, inducible nitric oxide synthase (iNOS), and arginase 1 by reverse transcriptase real-time PCR. PCR method and primer sequences were described before [13]. Briefly real-time polymerase chain reaction (PCR) assay was carried out using 1 μl of cDNA as the template, 10 μl of SYBR Select Master Mix (Thermo Fisher Scientific) and primers on the ABI Prism 7000 Sequence Detection System (Thermo Fisher Scientific). Data were analyzed by 2−ΔΔ Ct methods and normalized by GAPDH. The thermal cycling conditions for the PCR were 94˚C for 10 min, followed by 40 cycles of 94˚C for 15 s and 60˚C for 1 min.

Statistical analysis
All statistical analysis was performed using GraphPad Prism Software Version-5. To check data distribution we had performed the Shapiro-Wilk normality test. Statistical analysis of the lesion kinetics was performed with two-way ANOVA to test differences between groups. In all cases, a P-value of less than 0.05 was considered statistically significant.

Parasite strain-dependent influence of T lymphocytes on pathogenesis in susceptible BALB/c mice
To characterize the role of T lymphocytes and L. major strains on pathogenesis, we infected susceptible BALB/c WT and nude mice with 5×10 6 stationary phase promastigotes of either L. major strain, Friedlin or 5ASKH. Following infections, the lesions were measured weekly, until 6 weeks post-infection. The parasite burden in footpads and pLN were also measured at 6 weeks post-infection, at the end of the experiment. Both L. major Friedlin and 5ASKH infection showed progressive lesion development in WT mice (Fig 1A), where three of eight mice infected with 5ASKH developed ulcerative lesions compared with one of nine for Friedlininfected mice ( Table 1). In nude mice, L. major 5ASKH infection caused more severe ulcerative lesions in all 10 mice compared to 3/8 in WT mice (Fig 1B and Table 1). Conversely, L. major Friedlin infection caused minimal lesion development with no ulcer formation in all 11 nude mice and 1/9 in WT mice (Fig 1B and Table 1). It is worth noting that both Friedlin and 5ASKH showed a similarly high level of parasite burden in the footpads and pLN not only in nude mice but also in WT mice (Figs 1C and S1). These results showed that T lymphocytes reduced immunopathology during 5ASKH infection, while T lymphocytes exacerbated lesion development during Friedlin infection.

Parasite strain-dependent influence of T/B lymphocytes on pathogenesis in resistant C57BL/6 mice
Next, we examined the L. major strain-dependent influence of lymphocytes for lesion development in resistant C57BL/6 mice that allows transient L. major infection. WT and Rag2 KO C57BL/6 mice were infected with either L. major Friedlin or 5ASKH. The WT mice showed transient non-ulcerative lesion development during infection with both L. major strains (Fig  2A and 2B) and controlled parasite growth (Fig 2C). In Rag2 KO mice, which lack both conventional T/B lymphocytes, 5ASKH infection caused delayed but progressive and non-ulcerative lesion development, whereas Friedlin infection caused delayed and lesser lesion development (Fig 2A and 2B) despite a similar level of parasite burden in Friedlin-or 5ASKH-infected footpads and pLN (Figs 2C and S2).

L. major Friedlin and 5ASKH caused minimum or severe pathology respectively in NOD-SCID mice
Lack of pathogenesis by Friedlin strain in severe immunodeficient condition was also confirmed in NOD-SCID mice. NOD-SCID mice lack T/B cells and also defects in functions of NK, NKT, macrophage, and dendritic cells. In NOD-SCID mice, there was no difference in parasite burden between Friedlin and 5ASKH infected footpad and pLN (S3C Fig). However, L. major 5ASKH showed non-ulcerative and progressive lesion development, but Friedlin showed almost no lesion development (Table 1 and S3A and S3B Fig).

Defective infiltration of immune cells to the site of L. major Friedlin infection
Our study showed that both parasite and host factors affected the clinical presentation of CL. Inflammatory responses are mediated by immune cells that control not only parasite growth but also contribute to the pathogenesis of CL [3,14]. To characterize the inflammation induced by either L. major Friedlin or 5ASKH infections, we isolated immune cells from Rag2 KO-infected footpads at 4 weeks post-infection and analyzed these cells by flow cytometry. It is worth noting that, L. major Friedlin infection failed to recruit immune cells to the site of infection, while a 25 times higher number of cells were recruited to the footpads infected with L. major 5ASKH. Among the cells recruited to the site of L. major 5ASKH infection, macrophages (CD11b + F4/80 + ), neutrophils (CD11b + Gr1 + ), and eosinophils (CD11b + SiglecF + ) were the major cell populations. In addition, L. major 5ASKH infected footpad also contained a higher proportion and number of activated macrophages (CD11b + F4/80 + CD86 + ) compared to that with L. major Friedlin (Figs 3A and 3B and 4A and 4B and S4).

Lower activation of host immune responses by L. major Friedlin infection
Immune cells were isolated from the footpads of Rag2 KO mice 4 weeks after infection with L. major Friedlin or 5ASKH and were cultured for 72 h with or without PMA-ionomycin stimulation. Culture supernatants were analyzed for IFN-γ, TNF-α, IL-12, and IL-10 cytokines by ELISA. L major infection augmented the production of cytokines and L. major 5ASKH induced significantly higher cytokine production than Friedlin (Fig 5). After stimulation with PMA-ionomycin, a similar trend was observed (S5 Fig).

Discussion
The distinct impact of lymphocytes for the pathogenesis of CL was confirmed to be attributed to the genetic background of L. major. Our results clarified the contradictory findings of previous independent studies, in which lesion development by CL was either similar [7,8], delayed or less pronounced [9,10,12] in the absence of T/B lymphocyte-mediated immune responses compared to wild-type mice. The L. major 5ASKH infection model was consistent with reports that lesion development by either L. major (MHOM/UZ/91/PM2) or L. amazonensis (MPRO/ BR/72/M1845) was similar or less in the presence of lymphocytes [7,8]. The L. major Friedlin infection model also explains the previous report that delayed or no lesion development in the absence of T/B lymphocytes by L. major (WHOM/IR/-/173) [9,10] or L. amazonensis (MHOM/BR/77/LTB0016) [12] infection. Thus, the impact of lymphocytes for pathogenesis can be determined by the genetic background of L. major. In this study, the genetic background of L. major strains was not analyzed, which is one of the limitations of the study. We compared only two strains of L. major we had no comparison with other strains of L. major or different species of Leishmania. Further immunological analyses are needed to inquire into the differential lymphocyte response towards different Leishmania strains and species. Variable influence of lymphocytes for pathogenesis by Leishmania In Rag2-deficient C57BL/6 mice, 5ASKH infection caused progressive lesion development. In the early phase of its infection, the degree of footpad swelling was delayed in the absence of T/B lymphocytes indicating that even during 5ASKH infection, lymphocytes could be attributed to lesion formation to some extent. In fact, 5ASKH-infected footpads of WT mice showed a trend of higher expression of pro-inflammatory molecules, such as IL-6, TNF-α, IL-1β, IL-12p40 and iNOS compared with Friedlin-infected footpads (S6 Fig). The preceding studies also independently showed delayed lesion development after L. major (WHO strain WHOM/ IR/-/173) infection in C.B-17 SCID mice compared with BALB/c mice [9,10].
Pathogenesis can be caused by either the direct tissue-damaging effect of the parasites, parasite-induced immune-mediated skin inflammation or their combined effects. Belkaid et al. showed that lesion development starts after the parasite burden reaches its peak and that it coincides with leukocyte trafficking in resistant C57BL/6 mice [15], which emphasized the importance of parasite-induced inflammatory reactions for pathogenesis. We showed that L. major 5ASKH infection induced higher inflammatory reactions both in the presence and absence of T/B lymphocytes. Massive recruitment of immune cells, mainly macrophages, neutrophils and eosinophils, was observed at the site of 5ASKH infection even in the absence of lymphocytes. It is consistent with the prior study which showed that lesion development was associated with acute infiltration of leukocytes, primarily macrophages, neutrophils and eosinophils [15]. Immune cells recruited to the site of 5ASKH infection were also highly augmented with a larger number of CD86 + -activated macrophages and a higher level of cytokine production was observed compared with Friedlin infection. The inflammatory cytokines are able to activate the macrophages for intracellular killing of the parasite. In wild-type mice, Tlymphocytes are known to be the major source of IFN-γ which is central for the protective immunity. In the absence of T/B lymphocytes, mice failed to control parasites. It indicates that T/B lymphocytes are critical for the control of the parasites, mainly through the production of sufficient amount of critical cytokines, like IFN-γ and TNF-α. In the absence of T/B lymphocytes, macrophages, dendritic cells, NK cells, and group 1 innate lymphoid cells may be one of the sources of IFN-γ, which is insufficient for the protection (Fig 5). It may explain why the immunodeficient mice infected with both L. major strains showed no difference in parasite burden. However, augmented inflammatory responses after 5ASKH infection may have caused the severe pathology observed. Strong inflammatory reactions at the site of L. major 5ASKH infection may explain the severe pathology observed even in the absence of T/B Variable influence of lymphocytes for pathogenesis by Leishmania lymphocytes. L. major Friedlin infection failed to induce inflammatory reactions, like recruitment of immune cells and pro-inflammatory cytokines production at the site of infection in the absence of T/B lymphocytes, which may explain the minimum pathology observed. The outcome of Friedlin infection in the immunodeficient mice was consistent with the reported no pathology associated with less infiltration of monocytes at the site of L. amazonensis infection in mice lacking T cells [12]. Although we compared inflammatory responses only at 4 weeks after L. major Friedlin/5ASKH infection, it is also important to compare early inflammatory responses to address the mechanisms responsible for the different disease outcomes. Different Leishmania parasites may have diverse molecules or strategies determining the impact of lymphocytes for immuno-pathogenesis. Comparative analysis showed a significant number of virulence factors or molecules differentially expressed by Leishmania isolates [16][17][18]. We also analyzed only a few cytokines or host molecules after L. major infection. Further studies are required to identify the molecules and pathways involved.
CL has a wide spectrum of clinical manifestations depending on the host immunity and the infecting Leishmania species. CL is caused by several species of Leishmania parasite. Variable influence of lymphocytes for pathogenesis by Leishmania L. killicki, L. aethiopica, L. arabica, L. infantum and L. major are responsible for CL [19,20]. Many studies have been reported on CL, but few studies focused on the diversity of the clinical presentations. We extensively studied infection outcomes caused by two different strains of L. major in animal models with different genetic backgrounds and varying immune competency. L. major Friedlin and 5ASKH were originally isolated from patients with CL in the old world, Israel and Turkmenistan respectively, then adapted to the conditions in the Laboratory and established as the strains [21]. In immune-competent mice, we observed different infection outcomes in different host genetic backgrounds. Friedlin infection caused almost no ulceration in any genetic background studied, whereas 5ASKH infection caused ulceration in BALB/c mice and non-ulcerative lesions in C57BL/6 mice. 5ASKH infection caused more ulcerative lesions in T lymphocyte-deficient BALB/c mice compared with WT mice. Terabe et al. showed that T/B lymphocytes are a prerequisite for ulcer formation during L. amazonensis infection in BALB/c mice [8]. Therefore, particular Leishmania species may have distinct immunopathologies of ulcer formation.

Conclusion
Different species or strains of Leishmania were associated with a wide spectrum of diseases in human [19,20,22,23] and mouse models. Previous studies showing that L. major LV39 caused a non-healing infection [24][25][26] whereas other L. major strain (MHOM/IL/81/FEBNI) caused healing infections in IL-4 deficient BALB/c mice [25][26]. Furthermore, it now appears that strains of L. major can cause variant pathogenesis even in the absence of lymphocytes. Lymphocytes differentially regulate pathogenesis during L. major Friedlin/5ASKH infection. Leishmania has mechanisms to evade host immune protection, thereby aiding survival inside the host [27]. L. major Friedlin may have a superior immune evasion mechanism compared with strain 5ASKH that allows it to survive within the host by eliciting minimum immune-mediated pathogenesis. Understanding the basis for the diversity in pathogenesis is important in determining how to apply our knowledge for the development of new approaches for the treatment of leishmaniasis.