Leishmania-encoded orthologs of macrophage migration inhibitory factor regulate host immunity to promote parasite persistence

Abstract

Leishmania major encodes 2 orthologs of the cytokine macrophage migration inhibitory factor (MIF), whose functions in parasite growth or in the host–parasite interaction are unknown. To determine the importance of Leishmania-encoded MIF, both LmMIF genes were removed to produce an mif^−/− strain of L. major. This mutant strain replicated normally in vitro but had a 2-fold increased susceptibility to clearance by macrophages. Mice infected with mif^−/− L. major, when compared to the wild-type strain, also showed a 3-fold reduction in parasite burden. Microarray and functional analyses revealed a reduced ability of mif^−/− L. major to activate antigen-presenting cells, resulting in a 2-fold reduction in T-cell priming. In addition, there was a reduction in inflammation and effector CD4 T-cell formation in mif^−/− L. major-infected mice when compared to mice infected with wild-type L. major. Notably, effector CD4 T cells that developed during infection with mif^−/− L. major demonstrated statistically significant differences in markers of functional exhaustion, including increased expression of IFN-γ and IL-7R, reduced expression of programmed death-1, and decreased apoptosis. These data support a role for LmMIF in promoting parasite persistence by manipulating the host response to increase the exhaustion and depletion of protective CD4 T cells.—Holowka, T., Castilho, T. M., Baeza Garcia, A., Sun, T., McMahon-Pratt, D., Bucala, R. Leishmania-encoded orthologs of macrophage migration inhibitory factor regulate host immunity to promote parasite persistence.

Leishmaniasis is a parasitic infection of global significance caused by protozoans of the genus Leishmania. An estimated 50,000 deaths a year are the result of visceral involvement, which makes leishmaniasis second only to malaria in the ranking of deadly parasitic diseases (1, 2). Leishmania parasites are transmitted by the bite of an infected sand fly into the dermis, where they differentiate within the phagolysosome of myeloid cells from a flagellated promastigote into an intracellular amastigote (3–5). Mononuclear phagocytes provide an essential niche for parasites, but they also play an important role in parasite control and in establishing an effective adaptive immune response. Dendritic cells (DCs), in particular, act to present leishmanial antigens and foster a CD4 T helper (T_h) cell response (6, 7). A T_h1-type response, such as that observed in the C57BL/6 mouse model of infection, promotes IFN-γ production and NO-dependent destruction of parasites by macrophages (8, 9). However, a mixed response in which T_h2-type cytokines (IL-4 and -13) and immunosuppressive cytokines (IL-10 and TGF-β) are produced may result in progressive chronic disease, such as that observed in infected BALB/c mice (10).

To avoid destruction, Leishmania parasites produce virulence factors including specialized surface components and secreted proteins (8). Leishmania species also have been found to encode orthologs of the mammalian cytokine macrophage migration inhibitory factor (MIF). Leishmania major, which accounts for a significant disease burden, produces 2 such gene products: Lm1740MIF and Lm1750MIF (collectively referred to as the LmMIFs), and recent studies have verified the similarity of these proteins to mammalian MIF (11, 12).

Within the context of host immunity, MIF acts as a regulator of the innate response to activate CD74-dependent signal transduction and induce sustained ERK1/2 phosphorylation, resulting in enhanced expression of TLR4 and iNOS, increased production of inflammatory cytokines, and inhibition of cell-intrinsic apoptosis (13–16). MIF acts to protect the host from certain intracellular infections, but it can have a deleterious tissue-damaging effect in different inflammatory and infectious diseases (13, 14, 17–19).

Prior studies have shown that Lm1740MIF mimics mammalian MIF in its ability to activate CD74 and regulate the migration of PBMCs (11). To better understand the biologic role of LmMIFs, we employed a genetic approach to create a strain of L. major that lacks both LmMIF genes. This mif^−/− L. major strain was attenuated in its ability to persist in activated macrophages and cause disease. LmMIF further was found to promote host cell viability and inflammatory activity in a manner that attenuates the CD4 T-cell response, supporting a unique role for these cytokine orthologs in parasite persistence.

MATERIALS AND METHODS

Mice

Wild-type C57BL/6 and BALB/c mice and SCID/NCr mice were from the National Cancer Institute (Frederick, MD, USA). Cd74^−/− mice (BALB/c) were from Prof. I. Shachar (Weizmann Institute, Rehovot, Israel). Female mice were used at 8–10 wk of age. All protocols for animal use were approved by the Yale University Institutional Animal Care and Use Committee.

Parasites and cell culture

Leishmania major (MHOM/IL/79/LRC-L251) was cultivated at 23°C in Schneider's insect medium (SIM)-15: Schneider’s Insect Medium U.S. Biologic, Memphis, TN, USA) containing 15% Hyclone fetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, MA, USA) and 3.5 μg/ml gentamicin (Thermo Scientific-Gibco). mif^−/−transgenic (TG) L. major were cultivated in SIM-15 supplemented with 3 μg/ml G418 (InvivoGen, San Diego, CA, USA). Bone marrow cells were isolated from mice and bone marrow–derived macrophages (BMDMs) were cultured for 6–8 d in L929-conditioned medium (LCM): RPMI 1640 (Thermo Scientific–Gibco) containing 20% FBS, 30% L929 cell–conditioned medium, and 1% penicillin/streptomycin. Bone marrow-derived dendritic cells (BMDCs) were generated by culturing cells for 6–8 d in RPMI-10 (RPMI 1640 containing 10% FBS and 1% penicillin/streptomycin). RPMI-10 used for growing BMDCs was supplemented with 20 ng/ml granulocyte-macrophage colony-stimulating factor (GM-CSF; Biolegend, San Diego, CA, USA). The LMR7.5 T-cell hybridoma has been described (20).

PCR and cloning

All DNA primer sequences are listed in Supplemental Table 1. PCR was performed with Hi-Fidelity Platinum PCR Supermix (Thermo Scientific–Invitrogen, Carlsbad, CA, USA) using a MyCycler thermal cycler (Bio-Rad, Hercules, CA, USA) and the following program: 5 min at 95°C; 30 cycles of 1 min at 95°C, 1 min at 54°C, 1–3 min at 72°C; and 10 min at 72°C. PCR products were extracted from agarose gel fragments using the Qiaquick Gel Extraction Kit (Qiagen, Valencia, CA, USA). Restriction digest and ligation reactions were performed with enzymes from New England Biolabs (Danvers, MA, USA), and products were transformed into TOP10 Escherichia coli cells (Thermo Scientific–Invitrogen) before selection on Luria-Bertani plates.

Generation of mif^−/− and mif^−/− TG L. major

Genomic DNA was isolated from L. major using the DNeasy Blood and Tissue Kit (Qiagen), and a 900 bp region upstream of the Lm1740MIF gene was amplified with the primers MIFK5H and MIFK5ASP. This fragment was inserted into the pUC19 vector (New England Biolabs) to produce pMIF5. A 1.9 kb region downstream of the Lm1750MIF gene then was amplified with the primers MIF3SK and MIF3ASN, subcloned into a pCR2.1-TOPO vector (Thermo Scientific–Invitrogen), reamplified with primers MIF3SK and MIF3ASNE, and recloned into the pMIF5 construct to produce pMIF5/3. Separately, the pXG-HYG and pXG-PAC vectors (21, 22) were subjected to PCR with pXSX and pXASK primers to generate the hygromycin and puromycin resistance cassettes, respectively; these were cloned into pMIF5/3 to produce pMIF5/3HYG and pMIF5/3PAC.

pMIF5/3HYG was transfected into L. major using the Mouse T-cell Nucleofector kit and an Amaxa Nucleofector II (both from Lonza, Allendale, NJ, USA). Parasites were recovered in SIM-15 and spread onto solid SIM containing 1.2% agar and 15 μg/ml hygromycin. Clones were identified and grown in SIM-15 containing 30 μg/ml hygromycin. Heterozygous parasites with LmMIF genes removed from 1 chromosome were subjected to a second transfection with pMIF5/3PAC, exactly as described above, and grown on SIM-15 medium containing 12 μg/ml puromycin and 15 μg/ml hygromycin, and the homozygous mif^−/− parasites were isolated.

To reconstitute LmMIF expression, a 1.1 kb genetic fragment containing the Lm1740MIF and Lm1750MIF genes was amplified and subcloned into a pCR2.1-TOPO vector. This region was recloned into the pXG vector (21) to generate the pXGLM1 construct. This plasmid was transfected into mif^−/− parasites and resistant parasites selected on solid SIM-15 containing 3 μg/ml G418.

Real-time quantitative PCR

Measurement of RNA expression and genomic levels of Lm1740MIF and Lm1750MIF were performed as published (11). Real-time quantitative PCR (qPCR) reactions were performed with the iQ5 iCycler (Bio-Rad) and iTaq SYBR green Master Mix, with the following PCR program: 3 min at 95°C and 35 cycles of 15 s at 95°C and 30 s at 60°C. The cycle threshold (C_t) was determined and subtracted from the C_t of the housekeeping gene for rRNA 45S.

Parasite burden was determined as described elsewhere (23). Dermal lesions were excised, homogenized, and genomic DNA and RNA were extracted with the Allprep DNA/RNA/Protein Mini Kit (Qiagen). Leishmania kinetoplast DNA (LmkDNA) was amplified with published primers (23). A standard curve with genomic DNA from known a number of parasites and the relative amounts of LmkDNA in these isolates were used to quantify the number of tissue parasites based on their LmkDNA content.

RNA was extracted and quantified with the RNeasy Mini Kit and Quantitect Primer Assay (Qiagen) primers specific for Arg1, Cd86, Cxcl1, Gata3, Foxp3, Hprt, Icam1, Il1rn, Il10b, Il12b, Il17a, Il7, Il7r, Mmp13, Plau, Tbx21, Tgfb1, Tlr2, Tnf, and Tnfaip3. Expression was calculated relative to expression of the housekeeping gene Hprt.

Lm1740MIF antibody production and ELISA

A rabbit polyclonal antibody was produced against purified N-terminal 6-His-tagged Lm1740MIF (11) and anti-Lm1740MIF IgG antibodies purified with Protein A AffinityPak Columns (Thermo Fisher Scientific).

Nunc Maxisorp 96-well plates (Thermo Fisher Scientific) were coated with 10 μg/ml anti-Lm1740MIF in PBS overnight and blocked for 2 h with PBS containing 1% bovine serum albumin and 1% sucrose before addition of parasite lysates or culture medium. Recombinant native Lm1740MIF was used as a protein standard at concentrations of 4 ng/ml to 62.5 pg/ml. Protein samples were incubated 3 h and then washed from the plate before the addition of biotinylated anti-Lm1740MIF (prepared with Biotin Protein Labeling Kit; Roche, Indianapolis, IN, USA) at 1 μg/ml. After incubation and washing, avidin-conjugated horseradish peroxidase (eBioscience, San Diego, CA, USA) was added, and detection was performed with TMB ELISA detection reagent (eBioscience). Absorbance was read at 450 nm with an iMark Plate Reader (Bio-Rad), and Lm1740MIF content was calculated based on absorbance of protein standards. The sensitivity of the ELISA was 2.3 pg (2 sd from the baseline reading), the intra-assay variability was 7.4%, and the interassay variability was 17.5%. There was no cross-reactivity to murine MIF.

In vitro and in vivo infections

BMDMs were plated at 5 × 10⁴ cells per well in 4 Chamber Tissue Culture Treated Glass Slides (BD Biosciences, Franklin Lakes, NJ, USA) and allowed to adhere before infection with stationary-phase promastigote parasites at a multiplicity of infection (MOI) of 5. After 4 h, the remaining extracellular parasites were removed and LCM, with or without 100 ng/ml lipopolysaccharide (LPS; Sigma-Aldrich, St. Louis, MO, USA), was added. At 4, 24, 48, or 72 h after infection, the wells were washed and the cells were fixed with 4% paraformaldehyde before labeling with Vectashield Mounting Medium with DAPI (Vector Laboratories, Burlingame, CA, USA). Parasite nuclei and BMDM nuclei were enumerated by microscopy to quantify the number of parasites per host cell. Alternatively, BMDMs and BMDCs were plated at 5 × 10⁵ cells/well in 24-well tissue culture plates or at 10⁶ per well in 12-well tissue culture plates. The cells were infected at an MOI of 10 with stationary-phase parasites and collected for RNA extraction or for flow cytometry. NO levels were determined in medium with the Griess Reagent System (Promega, Madison, WI, USA).

For coculture studies with LMR7.5 T cells, infected BMDMs and BMDCs were cocultured with 10⁵ LMR7.5 T cells/well at the ratio of 5 LMR7.5 T cells per BMDM or 10 LMR7.5 cells per BMDC. The media were collected at 24 and 48 h after coculture and subjected to ELISA.

For in vivo infections, wild-type and mif^−/− L. major were cultured until the stationary phase over 6–7 d and metacyclic promastigotes purified on Ficoll gradients (24) (Sigma-Aldrich). In brief, parasites were centrifuged, resuspended in Schneider’s medium at 10⁸/ml and layered onto 10% Ficoll 400 (Sigma-Aldrich) in Schneider’s medium overlaying 40% Ficoll 400 in PBS. Gradients were centrifuged 10 min at 1300 g, and the interface layer was collected, washed with PBS, and centrifuged for 10 min at 1900 g. Parasites were resuspended in endotoxin-free PBS and kept on ice before injection.

Parasites (10⁶) in 20 μL were injected into the dermal layer of the skin on the dorsal side of the right foot of each mouse. Infected lesions were assessed by the width of the foot with a Pocket Dial Gage microcaliper (Starrett, Athol, MA, USA). In certain cases, mice were subjected to a second infection with 5 × 10⁶ parasites in the left foot. For histology, lesion tissue was removed from mice infected for 4 wk and stored in 10% buffered formalin. Fixed tissue was washed, transferred to 70% ethanol, embedded in paraffin, sectioned, and mounted on slides for hematoxylin and eosin staining. The sections were viewed under a Microphot FXA light microscope (Nikon Instruments Inc., Melville, NY, USA), and photographs were taken with a Spot Insight Camera and analyzed with Spot Advanced software (Diagnostic Instruments, Sterling Heights, MI, USA). The number of infiltrating myeloid cells was counted in a field at ×400 magnification.

Purification of CD4 T cells and adoptive transfer

Cells were collected from popliteal lymph nodes of mice infected for 4 wk, and CD4 T cells were purified with EasySep Mouse CD4⁺ T cell Isolation Kit (Stem Cell Technologies, Vancouver, BC, Canada). The cells were subsequently incubated in 5 μM carboxyfluorescein diacetate succinimidyl ester (CFSE), washed, and resuspended at 5 × 10⁶/ml before retro-orbital intravenous injection into SCID mice.

Microarray expression analysis

Infected BMDMs were lysed at 4 or 16 h after infection, and RNA was extracted with an RNeasy Mini Kit (Qiagen). cRNA samples were further prepared for hybridization to the MouseRef-8 Gene Chip (Illumina, San Diego, CA, USA) by the Keck Biotechnology Resource Laboratory (Yale University), and chips were scanned with a BeadArray reader (Illumina). Data analysis was performed with Genomics Suite software (Partek, St. Louis, MO, USA). Relative expression profiles for cells infected with wild-type parasites vs. cells infected with mif^−/− parasites were compared, and genes were identified with significant expression differences of ≥1.5-fold and between the 2 groups (P < 0.05). Further analysis was performed to determine which molecular pathways in the MetaCore database (Thomson Reuters, New York, NY, USA) were differentially regulated in cells infected with wild-type vs. mif^−/− L. major, accounting for number of genes in a pathway and the magnitude of differential expression in each pathway. Heat maps were prepared by the Yale Keck Biotechnology Resources Laboratory. Gene expression data are available upon publication in the International MIF Consortium database (http://www.biochemmcb.rwth-aachen.de/mif_consortium_public/index.php).

Surface and intracellular staining and flow cytometry

Infected BMDMs and BMDCs were collected after 24 h and labeled in the presence of BD Mouse FC Block (BD Biosciences) with the following antibodies: FITC-conjugated anti-MHC II, PE-conjugated anti-intracellular adhesion molecule (ICAM)-1, antigen-presenting cell (APC)–conjugated anti-CD86, and PECy7-conjugated anti-CD11c or PECy7-conjugated anti-CD11b (all from eBiosciences). Cells were washed and fixed in 1% paraformaldehyde before analysis on FACSCalibur or LSR II (BD Biosciences). Lesion tissue from 4 wk-infected mice was incubated for 2 h at 37°C in 0.25% trypsin. The dermal layer was separated and minced and digested in 2.5 mg/ml type IV collagenase, 0.25 mg/ml hyaluronidase, and 10⁵ U/ml DNaseI in RPMI1640 before passage through a 70 μm cell strainer. The cells were washed before surface staining with the following antibodies: FITC-conjugated anti-F4/80, PE-conjugated anti-GR1, PerCPCy5.5-conjugated anti-CD3, PECy7-conjugated anti-CD11c, eFluor450-conjugated anti-MHC II, APC-conjugated anti-CD11b, and eFluor780-APC-conjugated anti-Ly6C (all from eBiosciences).

Popliteal lymph nodes draining the site of infection were collected 4 or 1 wk, respectively, after infection and incubated in 5 μg/ml Liberase TL (Roche) before dissociating cells over a 70 μm mesh cell strainer. Cells were surface stained with the following antibodies (from eBiosciences, except where noted): PerCPCy5.5-conjugated anti-CD4, PE-conjugated anti-PD-1 (BD Biosciences), PECy7-conjugated anti-IL7R, APC-conjugated anti-CD44, and eFluor780-conjugated anti-CD8. Fixation and permeabilization was with the Foxp3 Buffer Staining Set (eBiosciences).

Alternatively, lymph node cells were collected and incubated for 6 h in RPMI-10 with 50 ng/ml PMA (Sigma-Aldrich) and 500 ng/ml ionomycin (Sigma-Aldrich), with 1 μl BD Golgi Plug (BD Biosciences) added for the final 3 h. The cells then were surface stained with the following antibodies (from eBiosciences): PerCPCy5.5-conjugated anti-CD4, eFluor450-conjugated anti-PD-1, APC-conjugated anti-CD44, and eFluor780-conjugated anti-CD8. The cells were further subjected to intracellular/intranuclear staining, as described above, with FITC-conjugated anti-Ki67, PE-conjugated anti-IL-4 (BD Biosciences) and PECy7-conjugated anti-IFN-γ. For TUNEL staining, infected cells were collected, fixed in 1% paraformaldehyde, and permeabilized in 70% EtOH and the ApoDirect In Situ DNA Fragmentation Assay Kit (Biovision, Milpitas, CA, USA).

Restimulation and ELISA for mammalian cytokines

Soluble Leishmania antigen (SLA) was prepared from log-phase wild-type L. major promastigotes by performing 5–7 rounds of freeze/thaw/vortex followed by centrifugation and isolation of the supernatant. The cells were collected from draining lymph nodes of infected mice and cultured at 5 × 10⁵ per well with 50 μg/ml of SLA. At 48 h, the medium was collected and subjected to ELISA with Ready-SET-Go kits for detection of IFN-γ, IL-4, IL-2, and IL-10 (eBiosciences).

Statistical analysis

Parasite burdens recorded from mouse tissue were nonparametric quantities and were analyzed with the Mann-Whitney test. Parasite growth in infected BMDMs and lesion growth in infected mice involved repeated measurements that were analyzed with a 2-way ANOVA, to determine statistical significance. All other data were analyzed with the unpaired, 2-tailed Student’s t test. Probabilities were determined, and differences between groups were considered to be statistically significant at P < 0.05.

RESULTS

Double-targeted deletion of LmMIF and genetic reconstitution in L. major

Leishmania major encodes 2 MIF-like genes, Lm1740MIF and Lm1750MIF, and these were removed by using a double-targeted gene replacement strategy (Fig. 1A). Two targeting constructs were sequentially transfected into L. major, replacing both LmMIF genes with resistance cassettes to hygromycin on the first chromosome and to puromycin on the second. First, heterozygous (mif^−/+) and homozygous (mif^−/−) parasites were selected. qPCR demonstrated the absence of LmMIF genomic DNA and mRNA in the mif^−/− parasites, and gel electrophoresis of the amplified products verified the removal of the 2 LmMIF genes and the proper positional insertion of the resistance cassettes (Fig. 1A, B). Reconstitution and stable expression of Lm1740MIF and Lm1750MIF were achieved by transfecting the mif^−/− parasites with a DNA sequence from the L. major genome containing the Lm1740MIF and Lm1750MIF genes (Fig. 1A). The parasites of this LmMIF-reconstituted strain (designated mif^−/−TG L. major) produced 30–40-fold more mRNA of Lm1740MIF and Lm1750MIF than wild-type L. major (Fig. 1C).

Figure 1.

Generation of mif^−/− L. major. A) Separate plasmid-targeting constructs were generated with regions upstream of Lm1740MIF and downstream of Lm1750MIF flanking resistance cassettes to hygromycin or puromycin. For genetic reconstitution, a 1.1 kb region of genomic DNA encompassing both Lm1740MIF and Lm1750MIF was cloned into the pXG expression vector and stably transfected into the mif^−/− strain, and mif^−/−TG parasites resistant to G418 were selected. B) PCR performed on genomic DNA of wild-type, mif^−/−, or mif^−/−TG parasites with primers directed to Lm1740MIF or Lm1750MIF. A third PCR with primers to amplify an upstream sequence of Lm1740MIF and within the hygromycin resistance cassette verified proper insertion of the targeting construct. C) qPCR performed on wild-type, mif^−/−, and mif^−/−TG parasite genomic DNA or cDNA with primers targeted to amplify Lm1740MIF and Lm1750MIF. D) Sandwich ELISA performed to detect Lm1740MIF in lysates or culture medium from wild-type, mif^−/−, and mif^−/−TG promastigote-stage parasites grown to log phase (3–4 d). C, D) Each experiment was performed 3–4 times (n = 4 for each parasite strain). *P < 0.05; ***P < 0.005.

An LmMIF-specific sandwich ELISA showed that wild-type L. major parasites contained 182 ± 11 pg LmMIF/10⁶ parasites and 588 ± 8 pg/10⁶ mif^−/−TG parasites, whereas the Lm1740MIF content of mif^−/− L. major was below the level of detection (2.3 pg/10⁶ parasites). LmMIF was released into the medium of cultured parasites to produce a concentration of 259 ± 29 pg/ml for wild-type L. major and 1432 ± 618 pg/ml for mif^−/−TG, when measured after 3–4 d of growth (Fig. 1D).

Macrophages infected with mif^−/− L. major clear parasites more rapidly and are more sensitive to activation-induced apoptosis

Both mif^−/− and mif^−/−TG L. major replicated as promastigotes in culture, infected BMDMs and persisted as intracellular amastigotes at an equivalent rate as wild-type parasites (Fig. 2A, B), suggesting that LmMIF is dispensable for the proliferation of L. major during the promastigote stage and does not impact invasion into macrophage hosts or affect parasite development in the amastigote stage.

Figure 2.

Fitness of mif^−/− L. major and persistence in host macrophages. A) Wild-type, mif^−/−, and mif^−/−TG parasites were grown in culture medium without antibiotics, and the number of parasites was determined with a hemocytometer. The experiment was performed 3 times (n = 5 for each parasite strain). B) BALB/c macrophages were infected with wild-type, mif^−/− , or mif^−/−TG L. major at an MOI of 5 and fixed and labeled with DAPI, and the parasites were recorded by enumerating by parasite nuclei vs. macrophage nuclei. The experiment was performed 3 times (n = 4 for each parasite strain). C) Wild-type macrophages stimulated with LPS 4 h after infection followed by counting of DAPI-labeled parasite nuclei in fixed cells. Each experiment was performed 5 times (n = 3–4 for each parasite strain used). ***P < 0.005 for mif^−/− vs. wild-type and mif^−/− vs. mif^−/−TG. D) BALB/c macrophages were stimulated with LPS 4 h after infection with wild-type, mif^−/−, or mif^−/−TG parasites, and nitrite content was measured in the culture medium at 48 h. The experiment was performed 3 times (n = 4 for each condition). E) Histograms of log₁₀ fluorescence of TUNEL staining in wild-type macrophages infected with wild-type, mif^−/−, or mif^−/−TG parasites 48 h after stimulation with LPS. Percentage of TUNEL⁺ macrophages determined using flow cytometry. Each experiment was performed 4 times (n = 3–4 for each condition). *P < 0.05, **P < 0.01. F) Cd74^−/− macrophages infected with wild-type, mif^−/−, or mif^−/−TG parasites 48 h after stimulation with LPS. The percentage of TUNEL⁺ macrophages determined using flow cytometry. Each experiment performed 4 times (n = 3–4 for each condition). G) Cd74^−/− macrophages stimulated with LPS 4 h after infection followed by counting of DAPI-labeled parasite nuclei in fixed cells. Each experiment was performed 5 times (n = 3–4 for each parasite strain used).

The inflammatory activation of Leishmania-infected macrophages promotes intracellular parasite destruction (25), and infected macrophages treated with LPS cleared 40% of intracellular mif^−/− L. major parasites within 48 h. By contrast, the intracellular number of wild-type or mif^−/−TG parasites increased up to 48 h under these activation conditions and then decreased by 60% at 72 h, to match the infection level of mif^−/− parasites (Fig. 2C). The more rapid clearance of mif^−/− L. major was not caused by increased NO production, which mediates parasite destruction (25, 26), as equivalent amounts of NO were produced by macrophages infected with all parasite strains (Fig. 2D).

LPS-treated macrophages infected with mif^−/− L. major exhibited increased apoptosis when compared to macrophages infected with wild-type or mif^−/−TG L. major parasites, but showed apoptosis similar to that of uninfected LPS-treated macrophages (Fig. 2E). These findings are in agreement with prior work suggesting that LmMIF, similar to mammalian MIF, inhibits macrophage activation-induced apoptosis (11) and support a role for LmMIF in preserving the host cell for parasite persistence. Naïve and infected macrophages not treated with LPS did not undergo apoptosis, regardless of parasite strain (Fig. 2E; data not shown).

The protective action of Lm1740MIF on host cell apoptosis may depend on the MIF receptor, CD74 (11). Infected macrophages from Cd74^−/− mice showed an apoptotic response to activation that was independent of parasite expression of LmMIF, confirming the requirement for CD74 in the anti-apoptotic action of LmMIF (Fig. 2F). Furthermore, Cd74^−/−macrophages cleared mif^−/− parasites at the same rate as wild-type and mif^−/−TG parasites, further supporting a mechanism in which LmMIF signals through the mammalian MIF receptor to block host cell death and promote the intracellular parasite persistence (Fig. 2G).

LmMIF regulates pathogenesis and parasite persistence in vivo

The importance of LmMIF for the growth and persistence of parasites in vivo was investigated in mouse models of infection. The development and self-resolution of cutaneous lesions in genetically resistant C57BL/6 mice was equivalent, irrespective of LmMIF expression by the infecting parasite strain (data not shown). However, in genetically susceptible BALB/c mice, lesion development was attenuated by ∼30% at 8 wk of infection in the case of mif^−/− parasites vs. wild-type L. major (Fig. 3A). Enumeration of parasites in lesional tissue by qPCR revealed a 3-fold reduction in parasite burden in mice infected with mif^−/− L. major when compared to wild-type parasites at 12 wk (Fig. 3B). Thus, the absence of LmMIF had a negative impact on parasite persistence and was associated with reduced lesion size in the later stages of infection. The reconstituted mif^−/−TG L. major strain failed to replicate the wild-type phenotype at late time points; however, parasites isolated from infected mice at those times had lost ectopic expression of LmMIF and were thus LmMIF deficient (data not shown). Because of the limitation of loss of expression, further study of the mif^−/−TG L. major strain in vivo was not pursued.

Figure 3.

Infection and disease progression in mice affected by LmMIF. A) BALB/c mice were infected with 10⁶ wild-type or mif^−/− L. major intradermally in the hind foot, and the lesion size was measured with a microcaliper. B) DNA extracted from lesions of mice infected for 4 or 12 wk and parasite burden determined using qPCR to measure relative quantity of Lm kDNA. Each experiment was performed 3 times (n = 8–10 for each condition). *P < 0.05; **P < 0.01; ***P < 0.005.

Reduced inflammation in lesions of mice infected with mif^−/− L. major

Histologic examination of infected skin showed the presence of a significant number of immune cells, with a greater density of these cells in mice infected with wild-type than in those with mif^−/− L. major (Fig. 4A). Flow cytometry of lesional cells revealed antigen-presenting myeloid cells, including macrophages (CD11b⁺, F4/80⁺) and DCs (CD11b⁺, CD11c^hi, MHCII^hi), as well as recruited T lymphocytes (CD3⁺), neutrophils (CD11b⁺, GR1^hi), and inflammatory monocytes (CD11b⁺Ly6C^hi) (Fig. 4B). The variation in individual cell populations was not statistically significant, but there was a trend toward a larger number of APCs in lesions of mice infected with wild-type vs. mif^−/− L. major. Analysis of lesional tissue by qPCR revealed elevated expression of mRNA for ICAM-1 and for the T_h1-related cytokines TNF-α and IL-12 in animals infected with wild-type vs. mif^−/− L. major (Fig. 4C). By contrast, genes associated with T_h2-driven alternative activation of macrophages and immunosuppression (e.g., Arg1, Tgfb1, and Il10) (27) were not differentially regulated in the absence of LmMIF (Fig. 4C).

Figure 4.

Inflammatory state at the site of infection. A) H&E staining of tissue sections from skin overlying the infected foot of naïve BALB/c mice or BALB/c mice infected for 4 wk with wild-type or mif^−/− L. major. Magnification, ×100 (top row); ×400 (bottom row).. Scale bar, 200 μm (top); 50 μm (bottom). Number of total immune cells (including myeloid cells and lymphocytes) counted per ×400 field. The experiment was performed 3 times (n = 3–4 for each condition). *P < 0.05. B) Cells were collected from infected lesion tissue and enumerated, and the phenotype was determined by flow cytometry gated for CD3⁺ T lymphocytes, CD11b^HiGR1^Hi neutrophils, CD11b⁺Ly6C^Hi monocytes, CD11b⁺F4/80^Hi macrophages, and CD11b⁺CD11c^HiMHC II^Hi DCs. The experiment was performed 3 times (n = 3–4 for each condition). ^#P < 0.1; *P < 0.05. C) RNA extracted from lesions 4 wk after infection with 10⁶ parasites and subjected to qPCR to determine gene expression. Represented as fold expression relative to that of mice infected with wild-type L. major. The experiment was performed 4 times (n = 3–5 for each gene analyzed). *P < 0.05; ***P < 0.005.

Influence of LmMIF on the gene expression profiles of infected macrophages

The impact of LmMIF on gene expression was further explored by using microarray analysis of naive macrophages and macrophages infected with wild-type or mif^−/− L. major. A comparison of expression profiles was performed with a threshold of >1.5-fold difference in relative expression (P < 0.05). In wild-type vs. mif^−/−-infected macrophages, 253 and 265 genes were differentially regulated at 4 and 16 h, respectively (Fig. 5A). Of these genes, 181 were differentially regulated at both time points. Most genes were upregulated in the presence of LmMIF, and these accounted for 66% of the differentially regulated genes at both time points.

Figure 5.

Whole-genome microarray expression analysis of infected macrophages. A) Venn diagrams of number of upregulated or downregulated by >1.5-fold in BALB/c macrophages infected with wild-type L. major vs. those infected with mif^−/− L. major, as determined by microarray analysis. P < 0.05. B) A heat map depicting relative gene expression in naïve macrophages or macrophages infected with wild-type or mif^−/− L. major for 4 or 16 h. Genes from cellular pathways of interest are displayed. Each column represents a separate replicate (n = 3 for each). C) qPCR of cDNA from macrophages infected with wild-type or mif^−/− L. major for 16 h with primers directed to amplify macrophage genes, or Leishmania kDNA. Displayed as expression relative to wild-type infected macrophages. Experiment performed 3 times (n = 3–4 for each gene analyzed). *P < 0.05; **P < 0.01.

A bioinformatic analysis employing the MetaCore database was used to identify cellular pathways that were differentially regulated in wild-type vs. mif^−/− L. major-infected macrophages. Immune response–related pathways comprised 5 of the top 7 most significantly regulated pathways at 4 h, and 6 of the top 7 at 16 h (Table 1). Based on this analysis, the genes upregulated in the presence of LmMIF were identified as being related to inflammation, innate immune signaling downstream of TLRs, and mammalian MIF-mediated activities [including p53-regulated apoptosis (28)] (Fig. 5B). In addition, pathways related to VEGF and other tissue repair responses were upregulated. The upregulation of selected genes by LmMIF was confirmed by qPCR; these included the inflammatory genes Cd86, Cxcl1 (KC), Icam1, Tlr2, and Tnf (TNF-α); the immunosuppressive genes Il1rn (IL-1 receptor antagonist) and Tnfaip3; and Plaur, which is involved in tissue repair (Fig. 5C). Among these genes, both Icam1 and Tnf were upregulated in infected lesion tissue in the presence of LmMIF (Fig. 4C).

TABLE 1.

Cellular pathways modulated in the presence of LmMIF

Enriched pathway	T	n	P	Differentially regulated genes
Wild-type vs. mif−/− (4 h)
Immune response: RAGE signaling pathway	50	9	1.150E-06	Icam1, Il1rn, Nfkb1, Nfkbia, Nfkbie, Mapk14 (p38), Pik3cd, Tlr2, Tnf
Development: VEGF signaling	65	10	1.296E-06	Ctnnb1 (beta-catenin), Map2k3 (MEK3), Nfatc1 (NFAT2), Nfkbie, Mapk14 (p38), Pik3cd, Plau, Plaur, Vegfa, Vcl (vinculin)
Development: PEDF signaling	39	8	1.648E-06	Junb, Nfkb1, Nfkbia, Nfkbie, Pik3cd, Pik3cb, Vegfa, Tnf
Immune response: histamine [1H] receptor signaling	40	8	2.022E-06	Nfkbie, Mapk14 (p38), Icam1, Nfatc1 (NFAT2) Mmp13, Nfkbia, Fos, Tnf
Immune response: HSP60 and HSP70/TLR signaling	54	9	2.266E-06	Cd86, Icam1, Map2k3 (MEK3), Mapk14 (p38), Nfkb1, Nfkbie, Tlr2, Tnf
Immune response: CD28 signaling	43	8	3.606E-06	Cd86, Mapk14 (p38), Nfatc1 (NFAT2), Nfkbie, Nfkb1, Pik3cd
Immune response, MIF: MIF-mediated glucocorticoid regulation	21	6	4.458E-06	Fos, Icam1, Nfkb1, Nfkbia, Nfkbie, Tnf
Wild-type vs. mif−/− (16 h)
Immune response: RAGE signaling pathway	43	11	5.946E-09	Icam1, Il1rn, Nfkb1, Nfkbia, Nfkbie, Mapk14 (p38), Pik3cd, Tlr2, Tnf
Immune response: HSP60 and HSP70/TLR signaling	50	10	3.793E-07	Cd86, Icam1, Map2k3 (MEK3), Mapk14 (p38), Nfkb1, Nfkbie, Tlr2, Tnf, Irak2
Immune response, MIF: MIF-induced cell adhesion, migration, angiogenesis	39	9	4.091E-07	Cd44, Cd74, Icam1, Mapk14 (p38), Mmp13, Nfkb1, Pik3cd, Src, Vegfa
Development: PEDF signaling	32	8	9.653E-07	Junb, Nfkb1, Nfkbia, Nfkbie, Pik3cd, Pik3cb, Vegfa, Tnf
Immune response: TLR5, TLR7, TLR8, and TLR9 signaling	38	8	3.945E-06	Irak2, Map2k3 (MEK3), Mapk14 (p38), Nfkb1, Nfkbia, Nfkbie, Tlr8, Tnf
Immune response, MIF: MIF-mediated glucocorticoid regulation	19	6	5.355E-06	Fos, Icam1, Nfkb1, Nfkbia, Nfkbie, Tnf
Immune response: TLR/HMGB1 signaling pathway	30	7	7.909E-06	Il1rn, Irak2, Map2k3 (MEK3), Nfkbie, Tlr2, Mapk14 (p38), Tnf

Enrichment analysis was used to identify signaling pathways in the MetaCore database that are differentially regulated in BMDMs infected with wild-type or mif^−/− L. major at 4 or 16 h. T, total number of genes in pathway; n, number of genes differentially expressed in pathway; P, comparison of expression of all genes in the pathway; RAGE, receptor for advanced glycation endproducts; HSP, heat shock protein; PEDF, pigment epithelium-derived factor; HMGB1, high-mobility group box 1.

LmMIF promotes activation and antigen presentation of macrophages and DCs

The gene expression profile of macrophages infected with wild-type L. major suggests a more active inflammatory phenotype than that of cells infected with mif^−/− parasites. Flow cytometry confirmed increased expression of the costimulatory molecules CD86 and ICAM-1 in macrophages infected with wild-type vs. mif^−/− L. major (Fig. 6A). Similarly, BMDCs infected with wild-type LmMIF displayed a more activated phenotype than those infected with mif^−/− L. major and expressed higher levels of CD86 and MHC II (Fig. 6A). To test whether APCs infected with wild-type rather than mif^−/− L. major induce T-cell activation more effectively, infected macrophages and DCs were cocultured with the LMR7.5 T-cell hybridoma line, which expresses a T-cell receptor specific for the LACK antigen (20). After 48 h, DCs infected with wild-type or mif^−/−TG L. major induced greater IL-2 production from LMR7.5 T cells than those infected with mif^−/− L. major (Fig. 6B). Comparatively little IL-2 was detectable after LMR7.5 T cells were cocultured with macrophages, irrespective of infecting parasite strain; this result likely reflects the limited role of the macrophages in antigen presentation when compared to Leishmania-infected DCs (7, 29).

Figure 6.

Phenotype and activity of APCs impacted by LmMIF. A) Wild-type or (C) Cd74^−/− BALB/c macrophages and DCs infected with wild-type, mif^−/− or mif^−/−TG L. major (MOI = 10) were analyzed after 24 h for surface marker expression by flow cytometry. Macrophages were gated on CD11b⁺ cells and DCs on CD11c⁺ cells. Each set of plots is representative of 4 separate experiments. B) Wild-type or (D) Cd74^−/− macrophages or DCs infected for 4 h were cocultured with LMR7.5 T cells, and media were collected after 24 or 48 h for IL-2 measurement by ELISA. Each experiment was performed 3 times (n = 5 for each condition). ^#P < 0.1; ***P < 0.001.

In contrast to what was observed in wild-type APCs, the phenotype of infected Cd74^−/− APCs, which are deficient in the MIF receptor, was not affected by the presence or absence of LmMIF. The expression of CD86 and ICAM-1 in Cd74^−/− macrophages were unchanged regardless of the infecting parasite strain, and CD86 and MHC II expression in Cd74^−/− DCs were slightly upregulated to the same degree by wild-type, mif^−/− and mif^−/−TG L. major (Fig. 6C). In addition, LMR7.5 T cells cocultured with infected Cd74^−/− macrophages or Cd74^−/− DCs produced equivalent amounts of IL-2, irrespective of the infecting L. major strain (Fig. 6D). These findings support a mechanism for LmMIF signaling through the MIF receptor on APCs to promote Leishmania-specific T-cell activation. A lower level of IL-2 production overall was evident in the Cd74^−/− experimental system, which likely reflects the globally reduced efficiency of antigen presentation in this gene-knockout background (30).

LmMIF regulates effector T-cell differentiation and function in vivo

Given the impact of LmMIF on APC function and T-cell activation in vitro, T-cell responses were investigated in infected BALB/c mice at 4 wk, a time point that would not be biased by the different parasite loads observed later in infection (Fig. 3B). Gene expression analysis of draining lymph node T cells revealed equivalent expression of Tbx21 (T-bet), Gata3, Foxp3, and Il17a, markers for T_h1, T_h2, regulatory T, and T_h17 lineages, respectively, suggesting that T-cell subset polarization is not affected by LmMIF expression in vivo (Fig. 7A).

Figure 7.

Effector T_H cell formation affected by LmMIF. A) qPCR performed on cDNA from draining lymph node cells and expression of genes quantified and represented as fold expression relative to that of mice infected with wild-type L. major. The experiment was performed 3 times (n = 5–6 for each gene analyzed). B) Cells collected from draining popliteal lymph nodes of mice infected for 4 wk with wild-type or mif^−/− L. major analyzed with flow cytometry. Cells gated on CD4⁺ T_H cells and log₁₀ fluorescence of CD44 and intranuclear Ki67 expression or (C) CD62L expression was observed. Mean fluorescence intensity of CD44 and frequency of Ki67⁺ cells were quantified. D) Lymph node cells from mice infected with wild-type or mif^−/− L. major were cultured with soluble Leishmania antigen (SLA) for 48 h and cytokine production determined by ELISA. Experiment performed 3 times (n = 5 for each condition). E) CD4⁺Ki67⁺CD44hi cells observed with flow cytometry for log₁₀ fluorescence of IFN-γ and the frequency of IFN-γ⁺ cells quantified. D, E) Experiments were performed 3 times (n = 6–8 for each condition). *P < 0.05; **P < 0.01; ***P < 0.001.

An increased proportion of Ki67⁺CD44^hi CD4 T cells in mice infected with wild-type vs. mif^−/− L. major was observed (Fig. 7B, C), which indicates an LmMIF-mediated increase in effector CD4 T cells responsive to L. major. The expression of CD62L on this T-cell population was unaffected by LmMIF, suggesting that there was no impact of LmMIF on the formation of memory CD4 T cell precursors (Fig. 7C) (31, 32). Despite an increased number of effector CD4 T cells, whole lymph node cells from mice infected with wild-type L. major, when compared to mif^−/− L. major, did not produce greater quantities of IFN-γ, IL-4, IL-10, or IL-2 when restimulated with L. major antigen in vitro (Fig. 7D). This finding indicates a potential defect in cytokine production in the presence of LmMIF, and indeed there was a reduced proportion of CD4⁺Ki67⁺CD44^hi effector CD4 T cells expressing IFN-γ or IL-4 after mitogenic restimulation from mice infected with wild-type L. major vs. those infected with mif^−/− parasites (Fig. 7E).

LmMIF promotes exhaustion and apoptosis of active effector CD4 T cells

Effector CD4 T cells are activated by antigen presentation and costimulation; however, T cells that become functionally exhausted are refractory to stimulation (33). It has been observed that a subset of effector CD4 T cells that develop in the presence of LmMIF express PD-1, a T-cell exhaustion marker (34), and these PD-1⁺ cells were present in greater frequency in mice infected with wild-type vs. mif^−/− L. major (Fig. 8A). Furthermore, this population of PD-1⁺ cells failed to express IFN-γ after mitogenic restimulation. Thus, elevated PD-1 expression in the presence of LmMIF correlates with loss of CD4 T cell functionality as assessed by reduced IFN-γ production.

Figure 8.

Increased exhaustion and reduced viability of effector T_H cells. A–C) CD4⁺Ki67⁺CD44^hi T cells observed with flow cytometry for log₁₀ fluorescence of expression of PD-1, IFN-γ (A) , and IL-7R (B). The frequency of PD-1⁺ and IL-7R^hi cells was quantified. C) qPCR was performed on cDNA from lymph node cells, and the expression of IL-7R and HPRT was quantified relative to Hprt and represented as fold expression relative to that of mice infected with wild-type L. major. Experiment performed 3 times (n = 3–5 for each gene analyzed). *P < 0.05. D) CD4⁺CD44^hi cells (IL7R^hi, IL7R^lo or total) were observed with flow cytometry for log₁₀ fluorescence of TUNEL staining, and the frequency of TUNEL⁺ cells was quantified. A, B, D) Experiments were performed 3 times (n = 6–8 for each condition). *P < 0.05; **P < 0.01; ***P < 0.005.

Exhausted PD-1⁺ T cells may express low levels of IL-7R, which is important for the long-term maintenance of effector and memory CD4 T cells, including during Leishmania infection (35). IL-7R expression was uniformly low on PD-1⁺ effector CD4 T cells responding to wild-type L. major infection, and a higher proportion of responding IL-7R^hi effector CD4 T cells was observed in mice infected with mif^−/− parasites (Fig. 8B). Furthermore, there was significantly higher overall expression of both IL-7R and IL-7 transcripts in whole lymph node cells from mice infected with mif^−/− vs. wild-type L. major (Fig. 8C).

IL-7 signaling also inhibits apoptosis (36) and TUNEL staining of CD4⁺CD44^hi cells from L. major-infected mice revealed increased apoptosis in the IL-7R^lo population of effector CD4 T cells when compared to the IL-7R^hi population (Fig. 8D). The IL-7R^lo effector CD4 T cells also showed increased apoptosis in mice infected with wild-type parasites vs. those infected with mif^−/− parasites, and the proportion of apoptotic effector CD4 T cells was elevated in the presence of LmMIF (Fig. 8D).

Long-lived CD4 T-cell protection is reduced in the presence of LmMIF

The reduced functionality and viability of effector CD4 T cells that develop in the presence of LmMIF may inhibit host parasite control. To investigate a detrimental impact of LmMIF on T-cell functionality during chronic infection, the ability of L. major-specific effector CD4 T cells to mount a response to a subsequent infection was investigated by adoptive transfer. CD4 T cells from the draining lymph nodes of mice infected with wild-type or mif^−/− L. major were purified, labeled with CFSE, and transferred into lymphocyte-deficient SCID mice before infection with wild-type parasites. After 2 wk, effector CD4 T cells were identified in the spleens, and among these were a subset of CFSE^lo cells, indicating proliferation in response to L. major infection (Fig. 9A). A higher proportion of these Leishmania-specific effector CD4 T cells was observed after transfer from wild-type parasite-infected mice, and there was a corresponding reduction in IFN-γ expression in comparison to CD4 T cells transferred from mif^−/− parasite–infected mice. These results support the conclusion that LmMIF promotes L. major-specific T-cell priming, but the T cells generated are less functional.

Figure 9.

Effector T_H-cell response to challenge infection. A) CD4 T cells purified from the draining lymph nodes of BALB/c mice 4 wk after infection with 10⁶ wild-type or mif^−/− L. major were transferred into SCID mice that were subsequently infected with wild-type L. major. Two weeks later, splenocytes were collected from SCID mice, and CD4⁺ CD44^hi cells were observed with flow cytometry for log₁₀ fluorescence of CFSE labeling and IFN-γ expression. Experiment performed 3 times (n = 5–6 for each condition). **P < 0.01; ***P < 0.005. B) BALB/c mice infected for 4 wk with 10⁶ wild-type or mif^−/− L. major in the right foot, then challenged with 5 × 10⁶ wild-type parasites in the left foot. Draining lymph node CD4⁺ T_H cells were observed for expression of Ki67 and IFN-γ.The frequency of IFN-γ and Ki67 expression on the CD4⁺ cells was quantified. The experiment was performed 3 times (n = 4 for each condition). *P < 0.05. C) Cells from draining lymph nodes at the site of the secondary challenge were cultured 48 h with SLA, and cytokine production was determined by ELISA. Experiment performed 3 times (n = 4 for each condition). N.S., no significance. *P < 0.05.

To further explore the impact of LmMIF on L. major-specific T-cell formation, mice infected for 4 wk with wild-type or mif^−/− L. major were challenged with wild-type parasites in the contralateral foot, and after 1 wk, cells from lymph nodes draining the site of secondary infection were isolated (Fig. 9B). Analysis of the CD4 T-cell population demonstrated reduced Ki67 and IFN-γ expression in mice that were infected with wild-type rather than mif^−/− L. major. Responding CD4 T cells from the challenge site produced significantly lower amounts of IFN-γ in response to parasite antigen, if mice were primarily infected with wild-type rather than with mif^−/− L. major (Fig. 9C). These data indicate that CD4 T cells conditioned in the presence of LmMIF during primary L. major infection are less able to mount a protective T_h1 type recall response to a secondary challenge.

DISCUSSION

It is notable that among more than 200 cytokines, chemokines, and growth factors in the human genome, only MIF homologous genes have been identified in Leishmania species (11). The LmMIFs therefore offer an intriguing potential pathway for manipulation of the host immune response. The findings herein suggest that LmMIFs are not involved in fundamental pathways for protozoan growth and development, nor are they required for entry into macrophages. Instead, it appears that LmMIF functions to prevent parasite destruction by host cells as mif^−/− parasites were cleared more quickly by activated macrophages than wild-type parasites. Macrophages infected with mif^−/− parasites nevertheless succumbed more readily to activation-induced apoptosis than those infected with wild-type or mif^−/−TG parasites—findings that agree with biochemical studies suggesting that Lm1740MIF inhibits apoptosis by engaging the MIF receptor CD74 (11). In further support of this conclusion, macrophages genetically deficient in CD74 failed to clear mif^−/− parasites at an accelerated rate and succumbed to apoptosis, regardless of the strain of L. major. Activation-induced, cell-intrinsic apoptosis is effective at controlling intracellular pathogens such as M. tuberculosis and L. pneumophila (37, 38) and LmMIF may signal to macrophages to prevent a similar phenomenon during Leishmania infection.

Whole-genome microarray analysis revealed at least 125 genes to be more highly expressed during infection with wild-type than with mif^−/− L. major, whereas only 56 genes showed reduced expression. Many differentially regulated genes are in pathways for MIF/CD74 signaling, innate immunity, and inflammation; these include the MIF-dependent cytokines and chemokines Tnf, Cxcl1 (KC), and Cxcl2 (MIP-2) (14, 39, 40). Genes necessary for antigen presentation, such as Cd86, Icam1, and H2-Ab1 (major histocompatibility class II), also were found to be upregulated by LmMIF. A functional impairment in the ability of mif^−/− L. major-infected DCs to present parasite antigen was confirmed by reduced activation of cocultured Leishmania-specific T cells. As expected, the effect of LmMIF on antigen presentation required CD74, and both Cd74 and its associated coreceptor, Cd44, were upregulated by LmMIF, suggesting a positive feedback mechanism for augmenting macrophage survival and increasing antigen presentation.

In accordance with microarray findings, the lesions of mice infected with wild-type vs. mif^−/− L. major showed an increase in APC numbers and enhanced proinflammatory gene expression (Icam1, Tnf, and Il12b). Based on the ability of LmMIF to regulate cell survival and apoptosis, the increased number of cells in wild-type L. major infection may be caused by increased recruitment or reduced death of phagocytic host cells or both. Increased phagocyte activity in the presence of LmMIF also had a surprising impact on the corresponding CD4 T-cell response, promoting enhanced differentiation into activated Ki67⁺CD44^hi effectors. However, these effector CD4 T cells produced lower levels of IFN-γ, supporting a model in which LmMIF drives the formation of effector CD4 T cells that have reduced functionality.

Nonfunctional T cells typically arise from the exhaustion of previously activated effectors, which can occur during chronic infections where antigen load is high and inflammatory conditions prevail (33, 41). Functional T-cell exhaustion also has been shown to occur during infection by protozoans such as Leishmania (42–45). The population of responding effector CD4 T cells that expressed the exhaustion marker PD-1 was increased in mice infected with wild-type vs. mif^−/− parasites. These PD-1⁺ CD4 T cells did not express IFN-γ, which supports a previous report that exhausted, nonfunctional T cells fail to block persisting L. major (34). Exhausted T cells ultimately undergo apoptosis (33), and indeed there was an increase in apoptosis of effector CD4 T-cell populations in the presence of LmMIF. Reduced viability of exhausted T cells may be caused by a reduction in IL-7 survival signals (33, 46, 47). The overall expression of Il7 and Il7r were reduced in lymph nodes of mice infected with wild-type but not mif^−/− L. major, and there was a greater proportion of IL-7R^lo effector CD4 T cells in mice infected with wild-type vs. mif^−/− parasites. IL-7R is expressed on both memory and effector CD4 T cells (35, 48), and both populations may contribute to control of infection and immunity to reinfection by Leishmania (32, 35).

The CD4 T cells transferred into SCID mice behaved differently, depending on their origin from mice infected with wild-type vs. mif^−/− L. major parasites. A larger proportion of L. major-specific T cells proliferated in response to infection after transfer from wild-type vs. mif^−/−-infected mice but this population also showed reduced expression of IFN-γ. In addition, in mice infected with wild-type L. major vs. those infected with mif^−/− L. major there was a diminished T-cell response to secondary infection characterized by a reduced T-cell IFN-γ response but preserved IL-4 and -10 production. In this setting, it appears that LmMIF specifically limits the establishment of a protective T_h1-type response. It should be noted that in the case of both the BALB/c infection model and in adoptively transferred SCID mice, host mice were not protected from a subsequent infection, in accordance with prior studies, most likely because, irrespective of the presence or absence of LmMIF, a mixed T_H1/T_H2 CD4 T-cell response is maintained that ensures susceptibility to reinfection (10, 49, 50).

The present data support a model in which a sustained proinflammatory action of LmMIF leads to enhanced exhaustion of CD4 T cells characterized by reduced IFN-γ production and IL-7R expression, and increased effector CD4 T-cell apoptosis (Fig. 10). The resulting reduction in protective T cells may explain the reduced pathogenicity of the mif^−/− L. major strain when compared to the wild-type, as evidenced by reduced lesion size and parasite burden after 8 wk of infection. This relatively late effect suggests enhanced development of a Leishmania-specific adaptive response and improved control of infection in the LmMIF-deficient condition.

Figure 10.

Summary of the overall impact of LmMIF during Leishmania infection. LmMIF is released by L. major and signals through the MIF receptor (CD74) on host macrophages to block apoptosis and prevent clearance of internalized parasites. Genes related to immune activation and inflammation are also upregulated by LmMIF, leading to enhanced antigen presentation by DCs. Effector T-cell priming and proliferation are accelerated; however these cells are nonfunctional and more readily succumb to exhaustion and apoptosis. The formation of long-lived effector T_h cells is consequently inhibited, resulting in impaired control of chronic infection with L. major.

The ability of a parasite-encoded MIF to negatively affect T-cell responses was first suggested by Sun et al. (51), who demonstrated that Plasmodium MIF induces a depletion of memory CD4 T_h1 cells and a reduced immunity to malaria. The present work supports a similar phenomenon in chronic leishmaniasis and suggests that parasite-encoded MIFs serve a general purpose in suppressing the development of a protective T-cell response to evade immune destruction or facilitate reinfection. Thus, it may be speculated that Leishmania-encoded MIFs benefit both the parasite and host by facilitating long-term parasitism and the induction of a permissive but nontissue-destructive adaptive T-cell response. In light of previous studies that have successfully targeted parasite MIF orthologs in treating and preventing disease (52, 53), LmMIF could serve as a potential vaccine target to augment the development of protective T-cell responses in Leishmania-endemic settings.

Glossary

APC

antigen-presenting cell

BMDC

bone marrow–derived DC

BMDM

BMD macrophage

CFSE

carboxyfluorescein diacetate succinimidyl ester

C_t

cycle threshold

DC

dendritic cell

FBS

fetal bovine serum

ICAM

intracellular adhesion molecule

LCM

L929-conditioned medium

LmkDNA

Leishmania kinetoplast DNA

LmMIF

Leishmania major-encoded MIF

LPS

lipopolysaccharide

MIF

macrophage migration inhibitory factor

MOI

multiplicity of infection

PBMC

peripheral blood mononuclear cell

PD

programmed death

qPCR

quantitative PCR

SIM

Schneider's insect medium

SLA

soluble Leishmania antigen

TG

transgenic

T_h

T helper