Interleukin-4 (IL-4) and IL-10 Collude in Vaccine Failure for Novel Exacerbatory Antigens in Murine Leishmania major Infection
http://www.100md.com
感染与免疫杂志 2005年第11期
Cambridge Institute for Medical Research and Department of Medicine, University of Cambridge, Cambridge CB2 2XY, United Kingdom
MRC Laboratory for Molecular Biology, Cambridge CB2 2QQ, United Kingdom
ABSTRACT
Leishmaniasis affects 12 million people but there are no vaccines in routine use. Recently, we used DNA vaccination in a susceptible BALB/c high-dose model of infection to screen 100 novel Leishmania major genes as vaccine candidates. In addition to finding novel protective antigens, we identified several antigens that reproducibly exacerbated disease. Here we examined the immune response to two of these antigens, lmd29 and 584C, that were originally identified in an expressed sequence tag cDNA sequencing project. We show that, in addition to exacerbating disease in susceptible BALB/c mice, these antigens retain a propensity to exacerbate disease in resistant C57BL/6 mice. This ability to exacerbate disease was lost when susceptible BALB/c mice were rendered resistant by disruption of the genes encoding interleukin-4 (IL-4) alone, IL-4/IL-13, or IL-4, IL-5, IL-9, and IL-13. Failure to exacerbate disease was associated with reduced IL-5 and IL-10 production in IL-4 knockout mice. Treatment of lmd29-vaccinated mice with anti-IL-10 receptor antibody prior to challenge infection converted exacerbation in wild-type BALB/c mice into highly significant antigen-specific protection. These studies demonstrate that some highly immunogenic antigens of L. major, while having an intrinsic capacity to exacerbate disease in the context of otherwise T helper 1-promoting DNA vaccine delivery, can be rendered protective by the removal of functional IL-10.
INTRODUCTION
Leishmaniasis affects 12 million people and there are 1.5 million new cases annually. There are no vaccines in routine use. Recently, we used DNA vaccination in a high-dose susceptible BALB/c mouse model of Leishmania major substrain LV39 infection to screen 100 novel L. major genes as vaccine candidates (21). In addition to finding novel protective antigens, we identified several antigens that reproducibly exacerbated disease when delivered as DNA vaccines but which usually bias towards a potentially protective Th1 response. Since gamma interferon (IFN-) knockout mice (23, 24) fail to clear the L. major, induction of a T helper 1 (Th1) response is crucial to vaccine success. However, using L. major infection in inbred mice, we (22) have demonstrated that the ability to elicit an IFN- response is not the sole predictor of vaccine success. In particular, we demonstrated (22) that interleukin-10 (IL-10) from antigen-driven CD4+CD25+ Tr1-like regulatory T cells (Tr) was responsible for vaccine failure in mice vaccinated with the Leishmania homologue of the receptor for activated C kinase (LACK) (18) in our high-dose model. Anti-IL-10R treatment in vivo rendered LACK protective in the presence of high IFN- and low IL-4/IL-5 responses.
In comparing the roles of IL-10 and IL-4 receptor signaling pathways in primary L. major infection in mice, Noben-Trauth and colleagues (20) concluded that IL-10 is as important as IL-4/IL-13 in promoting susceptibility to L. major infection. However, their relative roles in vaccine-induced immunity remain unclear. We wondered, therefore, whether induction of antigen-driven IL-10-T-cell responses is a common mechanism causing vaccine failure in other antigens that are nonprotective or disease exacerbatory in our model, or whether the classical T helper 2 (Th2) cytokines IL-4, IL-5, IL-9, and IL-13 also play a role. Here we demonstrate that two novel antigens, lmd29 and 584C, that exacerbate disease in susceptible BALB/c mice also retain a propensity to exacerbate disease in resistant C57BL/6 mice. Exacerbation of disease is lost when susceptible BALB/c mice are rendered resistant by disruption of the genes encoding IL-4 alone, IL-4 and IL-13, or IL-4, IL-5, IL-9, and IL-13. Failure to exacerbate disease in IL-4 knockout mice is associated with reduced IL-5 and IL-10 production, but antigen-vaccinated mice showed only modest enhancement of protection compared to vector-vaccinated mice. In contrast, anti-IL-10 receptor (IL-10R) antibody administered in vivo prior to challenge infection led to significant vaccine-induced protection in susceptible BALB/c mice, suggesting that IL-10 plays the major role in modulating vaccine success against L. major infection.
MATERIALS AND METHODS
Mice. Female 5- to 6-week-old BALB/c and C57BL/6 mice were purchased from Charles River Laboratories (Margate, United Kingdom) and were maintained in the Central Biomedical Services (University of Cambridge, United Kingdom) under pathogen-free conditions. IL-4 (11), IL-4/IL-13 (16), and IL-4, IL-5, IL-9, and IL-13 (7) knockout mice and wild-type controls were bred onto a BALB/c genetic background for six generations and maintained in-house. Female and male 6- to 12-week-old knockout mice were used in experiments. All breeding and procedures were carried out under United Kingdom Government Home Office guidelines.
Plasmid construction and purification. Exacerbatory (21) antigens lmd29 (accession number T67335) and 584C (accession number AI034795) and the protective (22) antigen tryparedoxin peroxidase (TRYP; clone lmf30; accession number T67356) were amplified from cDNA clones obtained from L. major substrain LV39 (MRHO/SU/59/P) cDNA libraries (1, 12). Lmd29 occurs as a single-copy expressed sequence tag (EST) hit annotated as a hypothetical protein to chromosome 18 of L. major substrain Friedlin (systematic name LmjF18.0595, www.genedb.org). The predicted properties of the protein include a molecular mass of 9.7 kDa, an isoelectric point at pH 10.8, and no signal peptide or transmembrane domains. Antigen 584C occurs as a two-gene tandem repeat on chromosome 35 and is annotated as a predicted 60S ribosomal protein L31 (systematic names LmjF35.3280 and LmjF35.3290, www.genedb.org). Predicted properties include a molecular mass of 21.2 kDa, an isoelectric point at pH 11.8, and no signal peptide or transmembrane domains. TRYP (also known as thiol-specific antioxidant) occurs as a seven-gene tandem array on chromosome 15 (systematic names LmjF15.1040/1060/1080/1100/1120/1140/1160, www.genedb.org). Predicted properties include a molecular mass of 22 kDa, an isoelectric point at pH 6.7, and no signal peptide or transmembrane domains.
All genes were inserted downstream of the cytomegalovirus promoter into a modified version (minus the neomycin resistance gene) of the expression vector pcDNA3 (Invitrogen, Paisley, United Kingdom). To check for mammalian cell expression, the three genes were also cloned into pcDNA3 containing a 3x FLAG epitope tag and transiently transfected into COS-7 cells. Expressed protein in transfected COS-7 cell lysates was detected by Western blotting using monoclonal antibody M2 (Sigma Chemical Company, Poole, Dorset, United Kingdom) that recognizes the FLAG tag. Discrete bands of 10 kDa, 21 kDa, and 22 kDa were observed, consistent with the predicted molecular masses for lmd29, 584C, and TRYP proteins, respectively. For vaccine studies, pcDNA3 was used as the vector control. Plasmid DNA was purified using Endofree plasmid maxi kits (QIAGEN Ltd., Crawley, United Kingdom), with pyrogen-free material, and the final pellet was resuspended in pyrogen-free phosphate-buffered saline.
Preparation of crude antigen. Crude freeze-thawed parasite (FTP) antigen was prepared from stationary-phase promastigotes by resuspension in 10 mM Tris-HCl (pH 8.5), 0.5 M NaCl, 1 mM phenylmethylsulfonyl fluoride, and 50 μg/ml leupeptin and freeze-thawing three times over liquid nitrogen.
Immunization. Groups of mice were injected subcutaneously into the shaven rump with two doses of 100 μg TRYP, lmd29, 584C, or vector DNA, 3 weeks apart.
Infectious challenge. L. major substrain LV39 promastigotes were cultured at 26°C in Schneider's insect medium (Sigma) supplemented with 10% fetal calf serum (Invitrogen), 2 mM L-glutamine, 50 U/ml penicillin, and 50 μg/ml streptomycin. Mice were challenged 2 weeks after the second DNA vaccine with 2 x 106 stationary phase (days 5 to 6) promastigotes into the hind footpad. Footpad depth was determined by weekly measurement with Vernier calipers. A clinical score for each mouse footpad was also recorded according to the following graded scale: 0, no evidence of inflammation; 0.5, some erythematous changes; 1.0, erythema and mild swelling; 1.5, erythema and moderate swelling; 2.0, moderate swelling and loss of soft tissue contours; 2.5, swelling with skin discoloration; 3.0, small area of ulceration; and 4.0, marked area of ulceration. A clinical score of 4 required euthanasia under United Kingdom Home Office license regulations. Our prior experience is that susceptible BALB/c mice reaching a clinical score of 4 do not recover. Hence, percent euthanasia is a surrogate for mortality. Footpad measurements and evaluation of clinical scores were determined without knowledge of the experimental group. In some experiments parasite loads were determined in draining lymph node (LN) by limiting dilution analysis, usually with 40 serial double dilutions of LN cell suspensions in liquid culture medium as above. Counts represented are the number of parasites per LN.
In vivo recall. Two weeks after the second DNA vaccine, mice were injected in both hind footpads with 5 μg of FTP; 48 h later, mice were sacrificed and bled for serum, and draining LN were removed.
Cytokine assays. Pooled cells from draining LN were cultured in RPMI 1640 (Invitrogen) at 6 x 105 cells/well in U-bottomed 96-well plates and stimulated for 72 h at 37°C in 5% CO2 with or without 10 μg/ml TRYP recombinant protein or 10 μg/ml FTP. Concanavalin A (5 μg/ml) was used as a control mitogen to evaluate cell viability. For in vitro depletion of CD4+ cells, pooled LN cells were incubated with anti-CD4 (clone GK1.5) BD IMag-DM (Pharmingen) according to the manufacturer's instructions. The positive fraction was removed with a BD-IMagnet and the CD4-depleted cells were plated at 6 x 105 cells/well in U-bottomed 96-well plates as described above and stimulated or not with FTP. Supernatants were removed and cytokine levels were measured by sandwich enzyme-linked immunosorbent assay (ELISA) using antibodies from Pharmingen (IL-4, IL-5, IL-10, and IFN-) or R&D Systems (IL-13).
In vivo antibody treatment. Prior to L. major challenge, mice were injected intraperitoneally with 1 mg of anti-IL-10R monoclonal antibody (1B1.3a, DNAX) or the isotype control (GL113, DNAX).
Statistical analysis. Statistical differences (P < 0.05) between the immunization groups were determined using the unpaired, two-tailed Student's t test. Statistical differences in time to 50% euthanasia (surrogate mortality) were determined using a logrank test which is equivalent to a Mantel-Haenszel test when comparing two groups.
RESULTS
Lmd29 and 584C exacerbate disease in susceptible BALB/c and resistant C57BL/6 mice. High-dose infection of vector control BALB/c mice with virulent L. major LV39 results in a severe clinical phenotype associated with rapid footpad swelling, ulceration, and euthanasia under United Kingdom Home Office regulations (Fig. 1). Nevertheless, prior vaccination with lmd29 and 584C repeatedly resulted in a significant increase in footpad swelling (Fig. 1A and D) and clinical score (Fig. 1B and E) and earlier euthanasia (Fig. 1C and F) than with vector-vaccinated mice. In Fig. 1C and E, the median values for survival (where 50% of animals have been euthanized) were 28 days for lmd29 and 584C and 35 days for the vector control (P = 0.03).
When a particularly virulent inoculum was used, it was sometimes difficult to observe statistically significant exacerbation for all clinical features of disease against the already severe clinical outcome in vector-vaccinated mice. However, in three out four replicate experiments for each antigen we observed statistically significant exacerbation. No differences in clinical outcomes of infection were observed between vector-vaccinated and nave-DNA-infected BALB/c mice in any experiment (data not shown).
To determine whether exacerbation was an intrinsic property of lmd29 and 584C antigens, vaccination experiments were carried out in resistant C57BL/6 mice. Although a clinical score above 2.5 was never encountered on this resistant background, a significant increase in footpad swelling (lmd29 [Fig. 1G]; [584C, Fig. 1H]) and clinical score (data not shown) was observed. A trend towards high parasite counts (Fig. 1I) in the draining LN was retained for both antigens compared to counts in vector control mice (vector versus lmd29, P = 0.04; vector versus 584C, P = 0.09) when the experiment was terminated at week 8 postinfection. These data suggest that counter-protective immunity elicited by vaccination is an intrinsic property of the lmd29 and 584C antigens, although the influence of genetic background meant that C57BL/6 mice were never rendered as susceptible as BALB/c mice by prior vaccination with these antigens.
IL-10 and classical Th2 cytokines correlate with vaccine-induced exacerbation. Previous studies suggest that both IL-10 and the classical Th2 cytokine IL-4 contribute to clinical outcome in L. major-infected susceptible BALB/c mice (20). To evaluate their potential roles in vaccine-induced exacerbation of disease, we measured cytokine profiles pre- and postchallenge infection. In the absence of recombinant proteins, cytokine profiles were measured in cells harvested from draining LN and restimulated in vitro with FTP. Mice vaccinated with the protective TRYP antigen were compared with lmd29-, 584C-, and vector-vaccinated mice.
For prechallenge cytokine responses, in vivo injection of FTP was used to recall antigen-specific T cells to the LN. Prechallenge cytokine responses for the experiment presented in Fig. 1 (A to F) show enhanced IFN- responses for all three antigens compared to vector control responses (Fig. 2A), indicating that all three antigens had elicited antigen-specific cellular immune responses following DNA vaccination. Higher IFN- was observed in TRYP-vaccinated mice than in lmd29- or 584C (P < 0.05)-vaccinated mice. IL-4 measurements (Fig. 2B) were significantly higher in lmd29 (P < 0.01)- and 584C in (P < 0.05)-vaccinated than in TRYP-vaccinated or vector control mice. Antigen-driven IL-10 responses were higher in all three vaccine groups than in the vector control (Fig. 2A) but did not differ significantly from each other. Hence, the classical Th1 and Th2 cytokines IFN- and IL-4 provided better prechallenge indicators of protection versus exacerbation than absolute levels of IL-10. Since absolute levels of cytokine responses might simply reflect the amounts of antigen TRYP, lmd29, or 584C in FTP, we hypothesized that the ratio of IFN- to other cytokines might provide a better predictor of vaccine outcome. Accordingly, the ratio of IFN- to IL-10 was significantly higher in TRYP (17 ± 1.6) than in lmd29 (5 ± 1; P < 0.01) and 584C (11 ± 1.7; P = 0.05).
In a separate experiment in which significant exacerbation of disease was observed with both lmd29 and 584C (Fig. 2D inset), draining LN cytokine responses (Fig. 2C and D) were reevaluated 2 weeks postchallenge infection with L. major LV39. As expected, infected vector-vaccinated mice made responses to FTP for all cytokines measured (Fig. 2C and D). Significantly (P < 0.001 for all comparisons) higher levels of classical Th2 cytokines (IL-4, IL-5, and IL-13) were observed in vector-, lmd29-, and 584C-vaccinated mice than in TRYP-vaccinated mice (Fig. 2D), preceding the vaccine-induced protection observed in TRYP-vaccinated mice as determined by clinical score (Fig. 2D inset), footpad depth or euthanasia (data not shown). Significantly (P < 0.001 all comparisons) higher levels of IL-10 were also observed in vector-, lmd29-, and 584C-vaccinated mice than in TRYP-vaccinated mice (Fig. 2C).
IFN- responses to concanavalin A were equivalent across vector and all antigen groups (data not shown). Interestingly, however, IFN- responses to FTP were lowest in TRYP-vaccinated mice, suggesting a degree of regulation of the proinflammatory Th1 cytokine response in these protected mice, although in more than five experiments in which we have compared TRYP mice with LACK (22) or lmd29 (not shown) mice at 4 weeks postchallenge infection, we have always measured equivalent amounts of IFN-. Nevertheless, ratios of IFN- to classical Th2 cytokines IL-4, IL-5, and IL-13, as well as to IL-10, provided strong correlates of protection in TRYP-vaccinated mice compared to the severe clinical phenotypes observed in vector-, lmd-29-, or 584C-vaccinated mice (Table 1). It was not possible, however, to identify significant deviations in FTP-induced postchallenge cytokine responses in lmd29- or 584C-vaccinated from responses in vector-vaccinated mice that might be indicative of vaccine-induced responses.
Lmd29 and 584C do not exacerbate disease in IL-4, IL-4/IL-13 or IL-4, IL-5, IL-9, and IL-13 knockout mice. Although classical Th2 (IL-4, IL-5, and IL-13), as well as IL-10, responses accompanied the severe clinical phenotypes observed in lmd-29, 584C- and vector-vaccinated compared to TRYP-vaccinated mice, it was difficult to determine which cytokines were functionally important in determining disease outcome especially for antigen-driven exacerbation of disease. We therefore examined responses to lmd29 and 584C DNA vaccines in mice for which IL-4 single-knockout, IL-4/IL-13 double-knockout, and IL-4, IL-5, IL-9, and IL-13 quadruple-knockout mice had been bred in the susceptible BALB genetic background. Comparison of the clinical outcomes of infection in vector-vaccinated mice demonstrated a degree of resistance to infection in all three knockout strains, with the additional knockout of IL-5 and IL-9 providing enhanced protection over IL-4 or IL-4/IL-13 knockout (Fig. 3A).
Vaccination of matched wild-type control BALB background mice was associated with exacerbated disease for lmd29- and, to a lesser extent, 584C-vaccinated mice as determined by footpad depth (Fig. 3B) and parasite loads measured in draining LN 4 weeks after challenge infection (Fig. 3C). Importantly, neither lmd29 nor 584C was associated with exacerbation of disease in BALB/c mice rendered resistant by disruption of genes encoding classical Th2 cytokines, indicating (minimally) a role for IL-4 in determining the vaccine-induced exacerbatory phenotype. Indeed, in the absence of Th2 cytokines, a measure of antigen-induced protection was observed for both lmd29- and 584C- compared to vector-vaccinated knockout mice. Specifically, reduced footpad depths were observed at 3 weeks postinfection in IL-4 knockout mice, along with significantly lower parasite loads for within-antigen comparisons of wild-type and knockout mice, as measured in the draining LN at 4 weeks postinfection (at a time when parasite loads for vector-vaccinated knockout mice did not differ significantly from wild-type mice).
Footpad depth was also lower in lmd29-vaccinated double-knockout mice at 5 to 6 weeks postinfection, in 584C-vaccinated quadruple-knockout mice at 3 and 4 weeks postinfection, and in both lmd29- and 584C-vaccinated quadruple-knockout mice at 6 weeks postinfection. Nevertheless, levels of antigen-induced protection were small in knockout mice, and it was difficult to associate this with evidence for specific antigen-induced changes in cytokine profiles measured in response to FTP at 4 weeks postinfection compared to vector-vaccinated wild-type or IL-4 knockout mice (Fig. 4).
Lmd29- and 584C-vaccinated wild-type mice had similarly high IL-10 responses but significantly lower IFN- responses than vector-vaccinated wild-type mice. The most dramatic change in all groups of IL-4 knockout mice compared to their wild-type counterparts was a large decrease in IL-10 response concomitant with an increase in IFN- (Fig. 4A). Along with a reduction to undetectable levels of IL-5 in IL-4 knockout compared to wild-type mice, this clearly provided a polarized Th1 cytokine environment in which all knockout mice were rendered resistant to L. major infection in the absence of classical Th2 cytokines and low levels of the immunoregulatory cytokine IL-10. Similar results were obtained in a repeat experiment for lmd29 vaccination in IL-4 knockout mice. CD4 depletion demonstrated that all (>95% depletion throughout) of the IFN-, IL-10, and IL-5 measured in both wild-type and IL-4 knockout mice came from CD4+ T cells (data not shown).
Removal of IL-10 converts lmd29-induced exacerbation to antigen-specific protection of susceptible BALB/c mice. Examination of gene-disrupted mice demonstrated a role (minimally) for IL-4 in determining the exacerbatory phenotype induced by lmd29 or 584C vaccination in susceptible BALB/c mice. However, it was difficult to determine whether this was due to reduction of IL-5, IL-9, and IL-13 secondary to IL-4 knockout, or to reduction in antigen-driven immunoregulatory IL-10. To determine the role of IL-10 in the exacerbatory phenotype, lmd29-vaccinated mice were administered with IL-10R blocking antibody 24 h prior to challenge infection. Isotype control antibody-treated lmd29-vaccinated mice displayed an exacerbated disease phenotype compared to isotype control vector-vaccinated mice (compare Fig. 5A and B). Anti-IL-10R antibody-treated vector-vaccinated mice did not differ significantly in footpad depth over the first 3 weeks of infection compared to isotype control antibody-treated vector-vaccinated mice (Fig. 5B). However, the administration of anti-IL-10R to lmd29-vaccinated mice rendered them significantly protected compared to both isotype control antibody-treated lmd29-vaccinated (Fig. 5A) (P < 0.001) and anti-IL-10R antibody-treated vector-vaccinated (P < 0.05) (compare Fig. 5A and B) mice. Hence, blocking the regulatory activity of IL-10 reverses the intrinsic property of the lmd29 antigen to exacerbate disease. These results are consistent with lmd29-driven antigen-specific protection in the absence of a functional IL-10 response.
DISCUSSION
The results presented here demonstrate that DNA vaccination with the recently identified (21) novel L. major antigens lmd29 and 584C is associated with an intrinsic propensity to exacerbate disease in both susceptible BALB/c and resistant C57BL/6 mice challenged with a high dose of virulent L. major substrain LV39. Compared to immune responses elicited by vaccination with the protective TRYP antigen, exacerbation was associated with a bias towards higher levels of the classical Th2 cytokine IL-4 and IL-10, which may derive from Th2 or other cellular sources, including regulatory T cells, relative to the Th1 cytokine IFN-. This called into question the relative roles of the classical Th2 cytokines IL-4, IL-5, IL-9, and IL-13 versus IL-10 in determining vaccine outcome. Importantly, the exacerbatory phenotype was lost for both antigens when susceptible BALB/c mice were rendered more resistant to infection by gene disruption of all classical Th2 cytokines, and minimally by IL-4 knockout, suggesting that IL-4 at least contributes to the exacerbation of disease. However, the most prominent effect in converting the exacerbatory phenotype induced by lmd29 into highly significant vaccine-induced protection was observed when IL-10 was rendered nonfunctional by in vivo administration of anti-IL-10R antibody. The results suggest an important interplay between IL-4 and IL-10 in determining the outcome of vaccination with Leishmania antigens that is consistent with studies comparing the roles of IL-10 and IL-4 receptor signaling pathways in primary L. major infection in mice (20).
In their studies of primary infection, Noben-Trauth and colleagues (20) proposed that IL-10 is as important as IL-4/IL-13 in promoting susceptibility to L. major infection, and more specifically in determining susceptibility to the LV39 substrain of L. major that we employed here. It was of interest, therefore, that although our previous studies (14) with this strain of L. major had highlighted a role for IL-13 as well as IL-4 in determining susceptibility to infection on a BALB/c background, we found evidence here that gene disruption of the full complement of the Th2 cytokines IL-4, IL-5, IL-9, and IL-13 resulted in a more resistant phenotype than either IL-4 knockout alone or IL-4/IL-13 double knockout. Functional redundancy is highly prevalent among the classical Th2 cytokines. However, recent studies (7) of Nippostrongylus brasiliensis infection, in which Th2 cytokines play a protective role, demonstrated that only in the combined absence of IL-4, IL-5, IL-9 and IL-13 does the Th2 immune response become subverted by Th1 cytokines. The corollary here in the context of L. major infection is that only in the combined absence of all four Th2 cytokines does the full protective potential of Th1 cytokines prevail.
Interesting in relation to the interplay of Th2 and IL-10 cytokines in vaccine-induced immunity in the L. major model are the observations of Nashed and colleagues (19), who found that exacerbation of infection in resistant DBA/2 mice immunized with crude soluble leishmanial antigen in incomplete Freund's adjuvant was specifically associated with up-regulation of IL-5 and IL-10. IL-4 and IL-13, on the other hand, were upregulated early after infection in both healing and nonhealing mice. They therefore proposed that IL-5 and IL-10, rather than IL-4 and IL-13, have an important role in the early phase of infection in establishing the disease promoting Th2 response in leishmaniasis. IL-9 was also shown (19) to be expressed in BALB/c and DBA/2 mice immunized with crude soluble leishmanial antigen in incomplete Freund's adjuvant, while Gessner et al. (8) found that IL-9 was more strongly up-regulated over the first 5 to 7 days following primary L. major infection in susceptible BALB/c than in resistant C57BL/6 mice.
Recent studies (3) that produced long-lasting neutralizing anti-IL-9 antibodies in vivo also demonstrated a reduction in detrimental Th2 responses in BALB/c mice infected with L. major, highlighting IL-9 as another Th2-related susceptibility factor in infection. Hence, as with N. brasiliensis infection (7), there may be distinct spatial, temporal and hierarchical requirements for the classical Th2 cytokines in immune function, which could also explain the temporal differences in enhanced protection afforded by vaccination with lmd29 and 584C in single IL-4 knockout compared with double IL-4/IL-13 or quadruple IL-4/IL-5/IL-9/IL-13 knockout mice.
Of major importance in our vaccine-induced exacerbation model was the role of IL-10. This might be explained in part by the crucial role of IL-4 in the induction of IL-10-secreting Th2 cells (17). However, in our recent studies comparing protection afforded by TRYP vaccination with the failure of the highly immunogenic LACK antigen to protect in our model, we demonstrated that antigen-driven CD4+CD25+ Tr1-like regulatory T cells were an important source of IL-10 following vaccination (22). Tr1 cells arise from chronic stimulation of normal T cells in the presence of IL-10 (9). Certain pathogen-derived molecules, including filamentous hemagglutinin and adenylate cyclase from Bordetella pertussis and cholera toxin, promote IL-10 and inhibit IL-12 production by dendritic cells and/or macrophages and activate dendritic cells into a semi-immature phenotype that results in the induction of Tr1-type responses in vivo (15).
In relation to our specific counterprotective vaccine antigens, LACK protein has been shown previously to elicit IL-10 release from peripheral blood monocytes (2), NK cells (13) and CD45RA– CD4+ memory T cells (4) from nave subjects. LACK and the leishmanial antigen KMP11 also elicit IL-10 production in CD14 positive monocytes, as well as from CD4+ and CD8+ T cells, in peripheral blood mononuclear cells from cutaneous and mucocutaneous leishmaniasis patients (5, 6). This led Carvalho and colleagues (5) to postulate an important role for these leishmanial antigens in modulating the proinflammatory responses associated with adverse disease phenotypes like mucosal leishmaniasis and to suggest their possible application as immune therapy. In our vaccine model, it is possible that transcription of LACK, lmd29, and 584C protein at the site of vaccine inoculation could contribute to a Tr1-promoting microenvironment, either through bystander effects on the cytokine milieu and/or through direct interaction with dendritic cells. Further studies are designed to determine whether, like LACK and KMP11, lmd29 and 584C are capable of eliciting IL-10 by ligation of innate pattern recognition receptors on macrophages or dendritic cells.
An alternative source of IL-10 in the early response to vaccination with counter-protective antigens would be via recognition by NK cells or T cells of epitopes that cross-react with self or intestinal flora for which peripheral tolerance is already established. It has already been demonstrated, for example, that priming by microbial antigens from the intestinal flora determines the ability of CD4+ T cells to rapidly respond to LACK antigen in BALB/c mice infected with L. major (10). In this case, however, the T cells were memory/effector cells secreting IL-4 rather than regulatory T cells making IL-10. As a highly conserved ribosomal protein, the 584C antigen used in our study is of particular interest in relation to the cross-reactive epitope hypothesis. We found that sera from nave and vector DNA- and 584C DNA-vaccinated BALB/c mice all recognized a distinct band of 21 kDa on Western blotting of mass spectrometry-verified recombinant 584C protein made in our laboratory (unpublished observation). While this precluded use of the protein in immune assays (data not shown), it provides support for the hypothesis for preexisting recognition of this molecule in BALB/c mice, and we are currently using a bioinformatics approach to identify epitopes in both 584C and lmd29 that might cross-react with mammalian (self) proteins or sequences from intestinal flora.
In conclusion, we have shown here that two novel leishmanial antigens have an intrinsic capacity to exacerbate disease via immune mechanisms that involve both classical Th2 cytokines and IL-10. The fact that these highly immunogenic counterprotective antigens can be rendered protective in the absence of functional IL-10 encourages us to believe that a delivery system may be developed in which all antigenic proteins from this parasite could be employed in successful development of a protective vaccine for leishmaniasis. The possibility that they may also be exploited in the development of anti-inflammatory immunotherapy would be an unexpected bonus of this research.
ACKNOWLEDGMENTS
This work was supported by a fellowship to M.T.M.R. from the Addenbrooke's Hospital Trust and a program grant to J.M.B. from the British Medical Research Council.
REFERENCES
1. Almeida, A. S., B. J. Gilmartin, S. McCann, A. Norrish, A. Ivens, D. Lawson, M. P. Levick, D. F. Smith, S. D. Dyall, D. Vetrie, T. C. Freeman, R. M. Coulson, I. Sampaio, H. Schneider, and J. M. Blackwell. 2004. Expression profiling of the Leishmania life cycle: cDNA arrays identify developmentally regulated genes present but not annotated in the genome. Mol. Biochem. Parasitol. 136:87-100.
2. Antonelli, L. R., W. O. Dutra, R. P. Almeida, O. Bacellar, and K. J. Gollob. 2004. Antigen specific correlations of cellular immune responses in human leishmaniasis suggests mechanisms for immunoregulation. Clin. Exp. Immunol. 136:341-348.
3. Arendse, B., J. Van Snick, and F. Brombacher. 2005. IL-9 is a susceptibility factor in Leishmania major infection by promoting detrimental Th2/type 2 responses. J. Immunol. 174:2205-2211.
4. Bourreau, E., M. Collet, G. Prevot, G. Milon, D. Ashimoff, H. Hasagewa, C. Parra-Lopez, and P. Launois. 2002. IFN-gamma-producing CD45RA+CD8+ and IL-10-producing CD45RA-CD4+ T cells generated in response to LACK in naive subjects never exposed to Leishmania. Eur. J. Immunol. 32:510-520.
5. Carvalho, L. P., S. Passos, W. O. Dutra, M. Soto, C. Alonso, K. J. Gollob, E. M. Carvalho, and A. Ribeiro de Jesus. 2005. Effect of LACK and KMP11 on IFN- production by peripheral blood mononuclear cells from cutaneous and mucocutaneous leishmaniasis patients. Scand. J. Immunol. 61:337-342.
6. de Carvalho, L. P., M. Soto, S. Jeronimo, B. Dondji, O. Bacellar, V. Luz, G. Orge Orge, C. Alonso, A. R. Jesus, and E. M. Carvalho. 2003. Characterization of the immune response to Leishmania infantum recombinant antigens. Microbes Infect. 5:7-12.
7. Fallon, P. G., H. E. Jolin, P. Smith, C. L. Emson, M. J. Townsend, R. Fallon, and A. N. McKenzie. 2002. IL-4 induces characteristic Th2 responses even in the combined absence of IL-5, IL-9, and IL-13. Immunity 17:7-17.
8. Gessner, A., H. Blum, and M. Rollinghoff. 1993. Differential regulation of IL-9-expression after infection with Leishmania major in susceptible and resistant mice. Immunobiology 189:419-435.
9. Groux, H., A. O'Garra, M. Bigler, M. Rouleau, S. Antonenko, J. E. de Vries, and M. G. Roncarolo. 1997. A CD4+ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 389:737-742.
10. Julia, V., S. S. McSorley, L. Malherbe, J. P. Breittmayer, F. Girard-Pipau, A. Beck, and N. Glaichenhaus. 2000. Priming by microbial antigens from the intestinal flora determines the ability of CD4+ T cells to rapidly secrete IL-4 in BALB/c mice infected with Leishmania major. J. Immunol. 165:5637-5645.
11. Kuhn, R., K. Rajewsky, and W. Muller. 1991. Generation and analysis of interleukin-4-deficient mice. Science 254:707-710.
12. Levick, M. P., J. M. Blackwell, V. Connor, R. M. Coulson, A. Miles, H. E. Smith, K. L. Wan, and J. W. Ajioka. 1996. An expressed sequence tag analysis of a full-length, spliced-leader cDNA library from Leishmania major promastigotes. Mol. Biochem. Parasitol. 76:345-348.
13. Maasho, K., I. Satti, S. Nylen, G. Guzman, F. Koning, and H. Akuffo. 2000. A Leishmania homologue of receptors for activated C-kinase (LACK) induces both interferon-gamma and interleukin-10 in natural killer cells of healthy blood donors. J. Infect. Dis. 182:570-578.
14. Matthews, D. J., C. L. Emson, G. J. McKenzie, H. E. Jolin, J. M. Blackwell, and A. N. McKenzie. 2000. IL-13 is a susceptibility factor for Leishmania major infection. J. Immunol. 164:1458-1462.
15. McGuirk, P., C. McCann, and K. H. Mills. 2002. Pathogen-specific T regulatory 1 cells induced in the respiratory tract by a bacterial molecule that stimulates interleukin 10 production by dendritic cells: a novel strategy for evasion of protective T helper type 1 responses by Bordetella pertussis. J. Exp. Med. 195:221-231.
16. McKenzie, G. J., P. G. Fallon, C. L. Emson, R. K. Grencis, and A. N. McKenzie. 1999. Simultaneous disruption of interleukin (IL)-4 and IL-13 defines individual roles in T helper cell type 2-mediated responses. J. Exp. Med. 189:1565-1572.
17. Mendel, I., and E. M. Shevach. 2002. The IL-10-producing competence of Th2 cells generated in vitro is IL-4 dependent. Eur. J. Immunol. 32:3216-3224.
18. Mougneau, E., F. Altare, A. E. Wakil, S. Zheng, T. Coppola, Z.-E. Wang, R. Waldmann, R. M. Locksley, and N. Glaichenhaus. 1995. Expression cloning of a protective Leishmania antigen. Science 268:563-566.
19. Nashed, B. F., Y. Maekawa, M. Takashima, T. Zhang, K. Ishii, T. Dainichi, H. Ishikawa, T. Sakai, H. Hisaeda, and K. Himeno. 2000. Different cytokines are required for induction and maintenance of the Th2 type response in DBA/2 mice resistant to infection with Leishmania major. Microbes Infect. 2:1435-1443.
20. Noben-Trauth, N., R. Lira, H. Nagase, W. E. Paul, and D. L. Sacks. 2003. The relative contribution of IL-4 receptor signaling and IL-10 to susceptibility to Leishmania major. J. Immunol. 170:5152-5158.
21. Stober, C., U. Lange, M. T. Roberts, B. C. Gilbmartin, R. Francis, R. Almeida, C. S. Peacock, S. McCann, and J. M. Blackwell. From genome to vaccines for leishmaniasis: screening 100 novel vaccine candidates against murine Leishmania major infection. Submitted for publication.
22. Stober, C., U. Lange, M. T. M. Roberts, A. Alcami, and J. M. Blackwell. 2005. IL-10 from regulatory T cells determines vaccine efficacy in murine Leishmania major infection. J. Immunol. 175:2517-2514.
23. Swihart, K., U. Fruth, N. Messmer, K. Hug, R. Behin, S. Huang, G. Del Giudice, M. Aguet, and J. A. Louis. 1995. Mice from a genetically resistant background lacking the interferon gamma receptor are susceptible to infection with Leishmania major but mount a polarized T helper cell 1-type CD4+ T-cell response. J. Exp. Med. 181:961-971.
24. Wang, Z. E., S. L. Reiner, S. Zheng, D. K. Dalton, and R. M. Locksley. 1994. CD4+ effector cells default to the Th2 pathway in interferon gamma-deficient mice infected with Leishmania major. J. Exp. Med. 179:1367-1371.(Mark T. M. Roberts, Carme)
MRC Laboratory for Molecular Biology, Cambridge CB2 2QQ, United Kingdom
ABSTRACT
Leishmaniasis affects 12 million people but there are no vaccines in routine use. Recently, we used DNA vaccination in a susceptible BALB/c high-dose model of infection to screen 100 novel Leishmania major genes as vaccine candidates. In addition to finding novel protective antigens, we identified several antigens that reproducibly exacerbated disease. Here we examined the immune response to two of these antigens, lmd29 and 584C, that were originally identified in an expressed sequence tag cDNA sequencing project. We show that, in addition to exacerbating disease in susceptible BALB/c mice, these antigens retain a propensity to exacerbate disease in resistant C57BL/6 mice. This ability to exacerbate disease was lost when susceptible BALB/c mice were rendered resistant by disruption of the genes encoding interleukin-4 (IL-4) alone, IL-4/IL-13, or IL-4, IL-5, IL-9, and IL-13. Failure to exacerbate disease was associated with reduced IL-5 and IL-10 production in IL-4 knockout mice. Treatment of lmd29-vaccinated mice with anti-IL-10 receptor antibody prior to challenge infection converted exacerbation in wild-type BALB/c mice into highly significant antigen-specific protection. These studies demonstrate that some highly immunogenic antigens of L. major, while having an intrinsic capacity to exacerbate disease in the context of otherwise T helper 1-promoting DNA vaccine delivery, can be rendered protective by the removal of functional IL-10.
INTRODUCTION
Leishmaniasis affects 12 million people and there are 1.5 million new cases annually. There are no vaccines in routine use. Recently, we used DNA vaccination in a high-dose susceptible BALB/c mouse model of Leishmania major substrain LV39 infection to screen 100 novel L. major genes as vaccine candidates (21). In addition to finding novel protective antigens, we identified several antigens that reproducibly exacerbated disease when delivered as DNA vaccines but which usually bias towards a potentially protective Th1 response. Since gamma interferon (IFN-) knockout mice (23, 24) fail to clear the L. major, induction of a T helper 1 (Th1) response is crucial to vaccine success. However, using L. major infection in inbred mice, we (22) have demonstrated that the ability to elicit an IFN- response is not the sole predictor of vaccine success. In particular, we demonstrated (22) that interleukin-10 (IL-10) from antigen-driven CD4+CD25+ Tr1-like regulatory T cells (Tr) was responsible for vaccine failure in mice vaccinated with the Leishmania homologue of the receptor for activated C kinase (LACK) (18) in our high-dose model. Anti-IL-10R treatment in vivo rendered LACK protective in the presence of high IFN- and low IL-4/IL-5 responses.
In comparing the roles of IL-10 and IL-4 receptor signaling pathways in primary L. major infection in mice, Noben-Trauth and colleagues (20) concluded that IL-10 is as important as IL-4/IL-13 in promoting susceptibility to L. major infection. However, their relative roles in vaccine-induced immunity remain unclear. We wondered, therefore, whether induction of antigen-driven IL-10-T-cell responses is a common mechanism causing vaccine failure in other antigens that are nonprotective or disease exacerbatory in our model, or whether the classical T helper 2 (Th2) cytokines IL-4, IL-5, IL-9, and IL-13 also play a role. Here we demonstrate that two novel antigens, lmd29 and 584C, that exacerbate disease in susceptible BALB/c mice also retain a propensity to exacerbate disease in resistant C57BL/6 mice. Exacerbation of disease is lost when susceptible BALB/c mice are rendered resistant by disruption of the genes encoding IL-4 alone, IL-4 and IL-13, or IL-4, IL-5, IL-9, and IL-13. Failure to exacerbate disease in IL-4 knockout mice is associated with reduced IL-5 and IL-10 production, but antigen-vaccinated mice showed only modest enhancement of protection compared to vector-vaccinated mice. In contrast, anti-IL-10 receptor (IL-10R) antibody administered in vivo prior to challenge infection led to significant vaccine-induced protection in susceptible BALB/c mice, suggesting that IL-10 plays the major role in modulating vaccine success against L. major infection.
MATERIALS AND METHODS
Mice. Female 5- to 6-week-old BALB/c and C57BL/6 mice were purchased from Charles River Laboratories (Margate, United Kingdom) and were maintained in the Central Biomedical Services (University of Cambridge, United Kingdom) under pathogen-free conditions. IL-4 (11), IL-4/IL-13 (16), and IL-4, IL-5, IL-9, and IL-13 (7) knockout mice and wild-type controls were bred onto a BALB/c genetic background for six generations and maintained in-house. Female and male 6- to 12-week-old knockout mice were used in experiments. All breeding and procedures were carried out under United Kingdom Government Home Office guidelines.
Plasmid construction and purification. Exacerbatory (21) antigens lmd29 (accession number T67335) and 584C (accession number AI034795) and the protective (22) antigen tryparedoxin peroxidase (TRYP; clone lmf30; accession number T67356) were amplified from cDNA clones obtained from L. major substrain LV39 (MRHO/SU/59/P) cDNA libraries (1, 12). Lmd29 occurs as a single-copy expressed sequence tag (EST) hit annotated as a hypothetical protein to chromosome 18 of L. major substrain Friedlin (systematic name LmjF18.0595, www.genedb.org). The predicted properties of the protein include a molecular mass of 9.7 kDa, an isoelectric point at pH 10.8, and no signal peptide or transmembrane domains. Antigen 584C occurs as a two-gene tandem repeat on chromosome 35 and is annotated as a predicted 60S ribosomal protein L31 (systematic names LmjF35.3280 and LmjF35.3290, www.genedb.org). Predicted properties include a molecular mass of 21.2 kDa, an isoelectric point at pH 11.8, and no signal peptide or transmembrane domains. TRYP (also known as thiol-specific antioxidant) occurs as a seven-gene tandem array on chromosome 15 (systematic names LmjF15.1040/1060/1080/1100/1120/1140/1160, www.genedb.org). Predicted properties include a molecular mass of 22 kDa, an isoelectric point at pH 6.7, and no signal peptide or transmembrane domains.
All genes were inserted downstream of the cytomegalovirus promoter into a modified version (minus the neomycin resistance gene) of the expression vector pcDNA3 (Invitrogen, Paisley, United Kingdom). To check for mammalian cell expression, the three genes were also cloned into pcDNA3 containing a 3x FLAG epitope tag and transiently transfected into COS-7 cells. Expressed protein in transfected COS-7 cell lysates was detected by Western blotting using monoclonal antibody M2 (Sigma Chemical Company, Poole, Dorset, United Kingdom) that recognizes the FLAG tag. Discrete bands of 10 kDa, 21 kDa, and 22 kDa were observed, consistent with the predicted molecular masses for lmd29, 584C, and TRYP proteins, respectively. For vaccine studies, pcDNA3 was used as the vector control. Plasmid DNA was purified using Endofree plasmid maxi kits (QIAGEN Ltd., Crawley, United Kingdom), with pyrogen-free material, and the final pellet was resuspended in pyrogen-free phosphate-buffered saline.
Preparation of crude antigen. Crude freeze-thawed parasite (FTP) antigen was prepared from stationary-phase promastigotes by resuspension in 10 mM Tris-HCl (pH 8.5), 0.5 M NaCl, 1 mM phenylmethylsulfonyl fluoride, and 50 μg/ml leupeptin and freeze-thawing three times over liquid nitrogen.
Immunization. Groups of mice were injected subcutaneously into the shaven rump with two doses of 100 μg TRYP, lmd29, 584C, or vector DNA, 3 weeks apart.
Infectious challenge. L. major substrain LV39 promastigotes were cultured at 26°C in Schneider's insect medium (Sigma) supplemented with 10% fetal calf serum (Invitrogen), 2 mM L-glutamine, 50 U/ml penicillin, and 50 μg/ml streptomycin. Mice were challenged 2 weeks after the second DNA vaccine with 2 x 106 stationary phase (days 5 to 6) promastigotes into the hind footpad. Footpad depth was determined by weekly measurement with Vernier calipers. A clinical score for each mouse footpad was also recorded according to the following graded scale: 0, no evidence of inflammation; 0.5, some erythematous changes; 1.0, erythema and mild swelling; 1.5, erythema and moderate swelling; 2.0, moderate swelling and loss of soft tissue contours; 2.5, swelling with skin discoloration; 3.0, small area of ulceration; and 4.0, marked area of ulceration. A clinical score of 4 required euthanasia under United Kingdom Home Office license regulations. Our prior experience is that susceptible BALB/c mice reaching a clinical score of 4 do not recover. Hence, percent euthanasia is a surrogate for mortality. Footpad measurements and evaluation of clinical scores were determined without knowledge of the experimental group. In some experiments parasite loads were determined in draining lymph node (LN) by limiting dilution analysis, usually with 40 serial double dilutions of LN cell suspensions in liquid culture medium as above. Counts represented are the number of parasites per LN.
In vivo recall. Two weeks after the second DNA vaccine, mice were injected in both hind footpads with 5 μg of FTP; 48 h later, mice were sacrificed and bled for serum, and draining LN were removed.
Cytokine assays. Pooled cells from draining LN were cultured in RPMI 1640 (Invitrogen) at 6 x 105 cells/well in U-bottomed 96-well plates and stimulated for 72 h at 37°C in 5% CO2 with or without 10 μg/ml TRYP recombinant protein or 10 μg/ml FTP. Concanavalin A (5 μg/ml) was used as a control mitogen to evaluate cell viability. For in vitro depletion of CD4+ cells, pooled LN cells were incubated with anti-CD4 (clone GK1.5) BD IMag-DM (Pharmingen) according to the manufacturer's instructions. The positive fraction was removed with a BD-IMagnet and the CD4-depleted cells were plated at 6 x 105 cells/well in U-bottomed 96-well plates as described above and stimulated or not with FTP. Supernatants were removed and cytokine levels were measured by sandwich enzyme-linked immunosorbent assay (ELISA) using antibodies from Pharmingen (IL-4, IL-5, IL-10, and IFN-) or R&D Systems (IL-13).
In vivo antibody treatment. Prior to L. major challenge, mice were injected intraperitoneally with 1 mg of anti-IL-10R monoclonal antibody (1B1.3a, DNAX) or the isotype control (GL113, DNAX).
Statistical analysis. Statistical differences (P < 0.05) between the immunization groups were determined using the unpaired, two-tailed Student's t test. Statistical differences in time to 50% euthanasia (surrogate mortality) were determined using a logrank test which is equivalent to a Mantel-Haenszel test when comparing two groups.
RESULTS
Lmd29 and 584C exacerbate disease in susceptible BALB/c and resistant C57BL/6 mice. High-dose infection of vector control BALB/c mice with virulent L. major LV39 results in a severe clinical phenotype associated with rapid footpad swelling, ulceration, and euthanasia under United Kingdom Home Office regulations (Fig. 1). Nevertheless, prior vaccination with lmd29 and 584C repeatedly resulted in a significant increase in footpad swelling (Fig. 1A and D) and clinical score (Fig. 1B and E) and earlier euthanasia (Fig. 1C and F) than with vector-vaccinated mice. In Fig. 1C and E, the median values for survival (where 50% of animals have been euthanized) were 28 days for lmd29 and 584C and 35 days for the vector control (P = 0.03).
When a particularly virulent inoculum was used, it was sometimes difficult to observe statistically significant exacerbation for all clinical features of disease against the already severe clinical outcome in vector-vaccinated mice. However, in three out four replicate experiments for each antigen we observed statistically significant exacerbation. No differences in clinical outcomes of infection were observed between vector-vaccinated and nave-DNA-infected BALB/c mice in any experiment (data not shown).
To determine whether exacerbation was an intrinsic property of lmd29 and 584C antigens, vaccination experiments were carried out in resistant C57BL/6 mice. Although a clinical score above 2.5 was never encountered on this resistant background, a significant increase in footpad swelling (lmd29 [Fig. 1G]; [584C, Fig. 1H]) and clinical score (data not shown) was observed. A trend towards high parasite counts (Fig. 1I) in the draining LN was retained for both antigens compared to counts in vector control mice (vector versus lmd29, P = 0.04; vector versus 584C, P = 0.09) when the experiment was terminated at week 8 postinfection. These data suggest that counter-protective immunity elicited by vaccination is an intrinsic property of the lmd29 and 584C antigens, although the influence of genetic background meant that C57BL/6 mice were never rendered as susceptible as BALB/c mice by prior vaccination with these antigens.
IL-10 and classical Th2 cytokines correlate with vaccine-induced exacerbation. Previous studies suggest that both IL-10 and the classical Th2 cytokine IL-4 contribute to clinical outcome in L. major-infected susceptible BALB/c mice (20). To evaluate their potential roles in vaccine-induced exacerbation of disease, we measured cytokine profiles pre- and postchallenge infection. In the absence of recombinant proteins, cytokine profiles were measured in cells harvested from draining LN and restimulated in vitro with FTP. Mice vaccinated with the protective TRYP antigen were compared with lmd29-, 584C-, and vector-vaccinated mice.
For prechallenge cytokine responses, in vivo injection of FTP was used to recall antigen-specific T cells to the LN. Prechallenge cytokine responses for the experiment presented in Fig. 1 (A to F) show enhanced IFN- responses for all three antigens compared to vector control responses (Fig. 2A), indicating that all three antigens had elicited antigen-specific cellular immune responses following DNA vaccination. Higher IFN- was observed in TRYP-vaccinated mice than in lmd29- or 584C (P < 0.05)-vaccinated mice. IL-4 measurements (Fig. 2B) were significantly higher in lmd29 (P < 0.01)- and 584C in (P < 0.05)-vaccinated than in TRYP-vaccinated or vector control mice. Antigen-driven IL-10 responses were higher in all three vaccine groups than in the vector control (Fig. 2A) but did not differ significantly from each other. Hence, the classical Th1 and Th2 cytokines IFN- and IL-4 provided better prechallenge indicators of protection versus exacerbation than absolute levels of IL-10. Since absolute levels of cytokine responses might simply reflect the amounts of antigen TRYP, lmd29, or 584C in FTP, we hypothesized that the ratio of IFN- to other cytokines might provide a better predictor of vaccine outcome. Accordingly, the ratio of IFN- to IL-10 was significantly higher in TRYP (17 ± 1.6) than in lmd29 (5 ± 1; P < 0.01) and 584C (11 ± 1.7; P = 0.05).
In a separate experiment in which significant exacerbation of disease was observed with both lmd29 and 584C (Fig. 2D inset), draining LN cytokine responses (Fig. 2C and D) were reevaluated 2 weeks postchallenge infection with L. major LV39. As expected, infected vector-vaccinated mice made responses to FTP for all cytokines measured (Fig. 2C and D). Significantly (P < 0.001 for all comparisons) higher levels of classical Th2 cytokines (IL-4, IL-5, and IL-13) were observed in vector-, lmd29-, and 584C-vaccinated mice than in TRYP-vaccinated mice (Fig. 2D), preceding the vaccine-induced protection observed in TRYP-vaccinated mice as determined by clinical score (Fig. 2D inset), footpad depth or euthanasia (data not shown). Significantly (P < 0.001 all comparisons) higher levels of IL-10 were also observed in vector-, lmd29-, and 584C-vaccinated mice than in TRYP-vaccinated mice (Fig. 2C).
IFN- responses to concanavalin A were equivalent across vector and all antigen groups (data not shown). Interestingly, however, IFN- responses to FTP were lowest in TRYP-vaccinated mice, suggesting a degree of regulation of the proinflammatory Th1 cytokine response in these protected mice, although in more than five experiments in which we have compared TRYP mice with LACK (22) or lmd29 (not shown) mice at 4 weeks postchallenge infection, we have always measured equivalent amounts of IFN-. Nevertheless, ratios of IFN- to classical Th2 cytokines IL-4, IL-5, and IL-13, as well as to IL-10, provided strong correlates of protection in TRYP-vaccinated mice compared to the severe clinical phenotypes observed in vector-, lmd-29-, or 584C-vaccinated mice (Table 1). It was not possible, however, to identify significant deviations in FTP-induced postchallenge cytokine responses in lmd29- or 584C-vaccinated from responses in vector-vaccinated mice that might be indicative of vaccine-induced responses.
Lmd29 and 584C do not exacerbate disease in IL-4, IL-4/IL-13 or IL-4, IL-5, IL-9, and IL-13 knockout mice. Although classical Th2 (IL-4, IL-5, and IL-13), as well as IL-10, responses accompanied the severe clinical phenotypes observed in lmd-29, 584C- and vector-vaccinated compared to TRYP-vaccinated mice, it was difficult to determine which cytokines were functionally important in determining disease outcome especially for antigen-driven exacerbation of disease. We therefore examined responses to lmd29 and 584C DNA vaccines in mice for which IL-4 single-knockout, IL-4/IL-13 double-knockout, and IL-4, IL-5, IL-9, and IL-13 quadruple-knockout mice had been bred in the susceptible BALB genetic background. Comparison of the clinical outcomes of infection in vector-vaccinated mice demonstrated a degree of resistance to infection in all three knockout strains, with the additional knockout of IL-5 and IL-9 providing enhanced protection over IL-4 or IL-4/IL-13 knockout (Fig. 3A).
Vaccination of matched wild-type control BALB background mice was associated with exacerbated disease for lmd29- and, to a lesser extent, 584C-vaccinated mice as determined by footpad depth (Fig. 3B) and parasite loads measured in draining LN 4 weeks after challenge infection (Fig. 3C). Importantly, neither lmd29 nor 584C was associated with exacerbation of disease in BALB/c mice rendered resistant by disruption of genes encoding classical Th2 cytokines, indicating (minimally) a role for IL-4 in determining the vaccine-induced exacerbatory phenotype. Indeed, in the absence of Th2 cytokines, a measure of antigen-induced protection was observed for both lmd29- and 584C- compared to vector-vaccinated knockout mice. Specifically, reduced footpad depths were observed at 3 weeks postinfection in IL-4 knockout mice, along with significantly lower parasite loads for within-antigen comparisons of wild-type and knockout mice, as measured in the draining LN at 4 weeks postinfection (at a time when parasite loads for vector-vaccinated knockout mice did not differ significantly from wild-type mice).
Footpad depth was also lower in lmd29-vaccinated double-knockout mice at 5 to 6 weeks postinfection, in 584C-vaccinated quadruple-knockout mice at 3 and 4 weeks postinfection, and in both lmd29- and 584C-vaccinated quadruple-knockout mice at 6 weeks postinfection. Nevertheless, levels of antigen-induced protection were small in knockout mice, and it was difficult to associate this with evidence for specific antigen-induced changes in cytokine profiles measured in response to FTP at 4 weeks postinfection compared to vector-vaccinated wild-type or IL-4 knockout mice (Fig. 4).
Lmd29- and 584C-vaccinated wild-type mice had similarly high IL-10 responses but significantly lower IFN- responses than vector-vaccinated wild-type mice. The most dramatic change in all groups of IL-4 knockout mice compared to their wild-type counterparts was a large decrease in IL-10 response concomitant with an increase in IFN- (Fig. 4A). Along with a reduction to undetectable levels of IL-5 in IL-4 knockout compared to wild-type mice, this clearly provided a polarized Th1 cytokine environment in which all knockout mice were rendered resistant to L. major infection in the absence of classical Th2 cytokines and low levels of the immunoregulatory cytokine IL-10. Similar results were obtained in a repeat experiment for lmd29 vaccination in IL-4 knockout mice. CD4 depletion demonstrated that all (>95% depletion throughout) of the IFN-, IL-10, and IL-5 measured in both wild-type and IL-4 knockout mice came from CD4+ T cells (data not shown).
Removal of IL-10 converts lmd29-induced exacerbation to antigen-specific protection of susceptible BALB/c mice. Examination of gene-disrupted mice demonstrated a role (minimally) for IL-4 in determining the exacerbatory phenotype induced by lmd29 or 584C vaccination in susceptible BALB/c mice. However, it was difficult to determine whether this was due to reduction of IL-5, IL-9, and IL-13 secondary to IL-4 knockout, or to reduction in antigen-driven immunoregulatory IL-10. To determine the role of IL-10 in the exacerbatory phenotype, lmd29-vaccinated mice were administered with IL-10R blocking antibody 24 h prior to challenge infection. Isotype control antibody-treated lmd29-vaccinated mice displayed an exacerbated disease phenotype compared to isotype control vector-vaccinated mice (compare Fig. 5A and B). Anti-IL-10R antibody-treated vector-vaccinated mice did not differ significantly in footpad depth over the first 3 weeks of infection compared to isotype control antibody-treated vector-vaccinated mice (Fig. 5B). However, the administration of anti-IL-10R to lmd29-vaccinated mice rendered them significantly protected compared to both isotype control antibody-treated lmd29-vaccinated (Fig. 5A) (P < 0.001) and anti-IL-10R antibody-treated vector-vaccinated (P < 0.05) (compare Fig. 5A and B) mice. Hence, blocking the regulatory activity of IL-10 reverses the intrinsic property of the lmd29 antigen to exacerbate disease. These results are consistent with lmd29-driven antigen-specific protection in the absence of a functional IL-10 response.
DISCUSSION
The results presented here demonstrate that DNA vaccination with the recently identified (21) novel L. major antigens lmd29 and 584C is associated with an intrinsic propensity to exacerbate disease in both susceptible BALB/c and resistant C57BL/6 mice challenged with a high dose of virulent L. major substrain LV39. Compared to immune responses elicited by vaccination with the protective TRYP antigen, exacerbation was associated with a bias towards higher levels of the classical Th2 cytokine IL-4 and IL-10, which may derive from Th2 or other cellular sources, including regulatory T cells, relative to the Th1 cytokine IFN-. This called into question the relative roles of the classical Th2 cytokines IL-4, IL-5, IL-9, and IL-13 versus IL-10 in determining vaccine outcome. Importantly, the exacerbatory phenotype was lost for both antigens when susceptible BALB/c mice were rendered more resistant to infection by gene disruption of all classical Th2 cytokines, and minimally by IL-4 knockout, suggesting that IL-4 at least contributes to the exacerbation of disease. However, the most prominent effect in converting the exacerbatory phenotype induced by lmd29 into highly significant vaccine-induced protection was observed when IL-10 was rendered nonfunctional by in vivo administration of anti-IL-10R antibody. The results suggest an important interplay between IL-4 and IL-10 in determining the outcome of vaccination with Leishmania antigens that is consistent with studies comparing the roles of IL-10 and IL-4 receptor signaling pathways in primary L. major infection in mice (20).
In their studies of primary infection, Noben-Trauth and colleagues (20) proposed that IL-10 is as important as IL-4/IL-13 in promoting susceptibility to L. major infection, and more specifically in determining susceptibility to the LV39 substrain of L. major that we employed here. It was of interest, therefore, that although our previous studies (14) with this strain of L. major had highlighted a role for IL-13 as well as IL-4 in determining susceptibility to infection on a BALB/c background, we found evidence here that gene disruption of the full complement of the Th2 cytokines IL-4, IL-5, IL-9, and IL-13 resulted in a more resistant phenotype than either IL-4 knockout alone or IL-4/IL-13 double knockout. Functional redundancy is highly prevalent among the classical Th2 cytokines. However, recent studies (7) of Nippostrongylus brasiliensis infection, in which Th2 cytokines play a protective role, demonstrated that only in the combined absence of IL-4, IL-5, IL-9 and IL-13 does the Th2 immune response become subverted by Th1 cytokines. The corollary here in the context of L. major infection is that only in the combined absence of all four Th2 cytokines does the full protective potential of Th1 cytokines prevail.
Interesting in relation to the interplay of Th2 and IL-10 cytokines in vaccine-induced immunity in the L. major model are the observations of Nashed and colleagues (19), who found that exacerbation of infection in resistant DBA/2 mice immunized with crude soluble leishmanial antigen in incomplete Freund's adjuvant was specifically associated with up-regulation of IL-5 and IL-10. IL-4 and IL-13, on the other hand, were upregulated early after infection in both healing and nonhealing mice. They therefore proposed that IL-5 and IL-10, rather than IL-4 and IL-13, have an important role in the early phase of infection in establishing the disease promoting Th2 response in leishmaniasis. IL-9 was also shown (19) to be expressed in BALB/c and DBA/2 mice immunized with crude soluble leishmanial antigen in incomplete Freund's adjuvant, while Gessner et al. (8) found that IL-9 was more strongly up-regulated over the first 5 to 7 days following primary L. major infection in susceptible BALB/c than in resistant C57BL/6 mice.
Recent studies (3) that produced long-lasting neutralizing anti-IL-9 antibodies in vivo also demonstrated a reduction in detrimental Th2 responses in BALB/c mice infected with L. major, highlighting IL-9 as another Th2-related susceptibility factor in infection. Hence, as with N. brasiliensis infection (7), there may be distinct spatial, temporal and hierarchical requirements for the classical Th2 cytokines in immune function, which could also explain the temporal differences in enhanced protection afforded by vaccination with lmd29 and 584C in single IL-4 knockout compared with double IL-4/IL-13 or quadruple IL-4/IL-5/IL-9/IL-13 knockout mice.
Of major importance in our vaccine-induced exacerbation model was the role of IL-10. This might be explained in part by the crucial role of IL-4 in the induction of IL-10-secreting Th2 cells (17). However, in our recent studies comparing protection afforded by TRYP vaccination with the failure of the highly immunogenic LACK antigen to protect in our model, we demonstrated that antigen-driven CD4+CD25+ Tr1-like regulatory T cells were an important source of IL-10 following vaccination (22). Tr1 cells arise from chronic stimulation of normal T cells in the presence of IL-10 (9). Certain pathogen-derived molecules, including filamentous hemagglutinin and adenylate cyclase from Bordetella pertussis and cholera toxin, promote IL-10 and inhibit IL-12 production by dendritic cells and/or macrophages and activate dendritic cells into a semi-immature phenotype that results in the induction of Tr1-type responses in vivo (15).
In relation to our specific counterprotective vaccine antigens, LACK protein has been shown previously to elicit IL-10 release from peripheral blood monocytes (2), NK cells (13) and CD45RA– CD4+ memory T cells (4) from nave subjects. LACK and the leishmanial antigen KMP11 also elicit IL-10 production in CD14 positive monocytes, as well as from CD4+ and CD8+ T cells, in peripheral blood mononuclear cells from cutaneous and mucocutaneous leishmaniasis patients (5, 6). This led Carvalho and colleagues (5) to postulate an important role for these leishmanial antigens in modulating the proinflammatory responses associated with adverse disease phenotypes like mucosal leishmaniasis and to suggest their possible application as immune therapy. In our vaccine model, it is possible that transcription of LACK, lmd29, and 584C protein at the site of vaccine inoculation could contribute to a Tr1-promoting microenvironment, either through bystander effects on the cytokine milieu and/or through direct interaction with dendritic cells. Further studies are designed to determine whether, like LACK and KMP11, lmd29 and 584C are capable of eliciting IL-10 by ligation of innate pattern recognition receptors on macrophages or dendritic cells.
An alternative source of IL-10 in the early response to vaccination with counter-protective antigens would be via recognition by NK cells or T cells of epitopes that cross-react with self or intestinal flora for which peripheral tolerance is already established. It has already been demonstrated, for example, that priming by microbial antigens from the intestinal flora determines the ability of CD4+ T cells to rapidly respond to LACK antigen in BALB/c mice infected with L. major (10). In this case, however, the T cells were memory/effector cells secreting IL-4 rather than regulatory T cells making IL-10. As a highly conserved ribosomal protein, the 584C antigen used in our study is of particular interest in relation to the cross-reactive epitope hypothesis. We found that sera from nave and vector DNA- and 584C DNA-vaccinated BALB/c mice all recognized a distinct band of 21 kDa on Western blotting of mass spectrometry-verified recombinant 584C protein made in our laboratory (unpublished observation). While this precluded use of the protein in immune assays (data not shown), it provides support for the hypothesis for preexisting recognition of this molecule in BALB/c mice, and we are currently using a bioinformatics approach to identify epitopes in both 584C and lmd29 that might cross-react with mammalian (self) proteins or sequences from intestinal flora.
In conclusion, we have shown here that two novel leishmanial antigens have an intrinsic capacity to exacerbate disease via immune mechanisms that involve both classical Th2 cytokines and IL-10. The fact that these highly immunogenic counterprotective antigens can be rendered protective in the absence of functional IL-10 encourages us to believe that a delivery system may be developed in which all antigenic proteins from this parasite could be employed in successful development of a protective vaccine for leishmaniasis. The possibility that they may also be exploited in the development of anti-inflammatory immunotherapy would be an unexpected bonus of this research.
ACKNOWLEDGMENTS
This work was supported by a fellowship to M.T.M.R. from the Addenbrooke's Hospital Trust and a program grant to J.M.B. from the British Medical Research Council.
REFERENCES
1. Almeida, A. S., B. J. Gilmartin, S. McCann, A. Norrish, A. Ivens, D. Lawson, M. P. Levick, D. F. Smith, S. D. Dyall, D. Vetrie, T. C. Freeman, R. M. Coulson, I. Sampaio, H. Schneider, and J. M. Blackwell. 2004. Expression profiling of the Leishmania life cycle: cDNA arrays identify developmentally regulated genes present but not annotated in the genome. Mol. Biochem. Parasitol. 136:87-100.
2. Antonelli, L. R., W. O. Dutra, R. P. Almeida, O. Bacellar, and K. J. Gollob. 2004. Antigen specific correlations of cellular immune responses in human leishmaniasis suggests mechanisms for immunoregulation. Clin. Exp. Immunol. 136:341-348.
3. Arendse, B., J. Van Snick, and F. Brombacher. 2005. IL-9 is a susceptibility factor in Leishmania major infection by promoting detrimental Th2/type 2 responses. J. Immunol. 174:2205-2211.
4. Bourreau, E., M. Collet, G. Prevot, G. Milon, D. Ashimoff, H. Hasagewa, C. Parra-Lopez, and P. Launois. 2002. IFN-gamma-producing CD45RA+CD8+ and IL-10-producing CD45RA-CD4+ T cells generated in response to LACK in naive subjects never exposed to Leishmania. Eur. J. Immunol. 32:510-520.
5. Carvalho, L. P., S. Passos, W. O. Dutra, M. Soto, C. Alonso, K. J. Gollob, E. M. Carvalho, and A. Ribeiro de Jesus. 2005. Effect of LACK and KMP11 on IFN- production by peripheral blood mononuclear cells from cutaneous and mucocutaneous leishmaniasis patients. Scand. J. Immunol. 61:337-342.
6. de Carvalho, L. P., M. Soto, S. Jeronimo, B. Dondji, O. Bacellar, V. Luz, G. Orge Orge, C. Alonso, A. R. Jesus, and E. M. Carvalho. 2003. Characterization of the immune response to Leishmania infantum recombinant antigens. Microbes Infect. 5:7-12.
7. Fallon, P. G., H. E. Jolin, P. Smith, C. L. Emson, M. J. Townsend, R. Fallon, and A. N. McKenzie. 2002. IL-4 induces characteristic Th2 responses even in the combined absence of IL-5, IL-9, and IL-13. Immunity 17:7-17.
8. Gessner, A., H. Blum, and M. Rollinghoff. 1993. Differential regulation of IL-9-expression after infection with Leishmania major in susceptible and resistant mice. Immunobiology 189:419-435.
9. Groux, H., A. O'Garra, M. Bigler, M. Rouleau, S. Antonenko, J. E. de Vries, and M. G. Roncarolo. 1997. A CD4+ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 389:737-742.
10. Julia, V., S. S. McSorley, L. Malherbe, J. P. Breittmayer, F. Girard-Pipau, A. Beck, and N. Glaichenhaus. 2000. Priming by microbial antigens from the intestinal flora determines the ability of CD4+ T cells to rapidly secrete IL-4 in BALB/c mice infected with Leishmania major. J. Immunol. 165:5637-5645.
11. Kuhn, R., K. Rajewsky, and W. Muller. 1991. Generation and analysis of interleukin-4-deficient mice. Science 254:707-710.
12. Levick, M. P., J. M. Blackwell, V. Connor, R. M. Coulson, A. Miles, H. E. Smith, K. L. Wan, and J. W. Ajioka. 1996. An expressed sequence tag analysis of a full-length, spliced-leader cDNA library from Leishmania major promastigotes. Mol. Biochem. Parasitol. 76:345-348.
13. Maasho, K., I. Satti, S. Nylen, G. Guzman, F. Koning, and H. Akuffo. 2000. A Leishmania homologue of receptors for activated C-kinase (LACK) induces both interferon-gamma and interleukin-10 in natural killer cells of healthy blood donors. J. Infect. Dis. 182:570-578.
14. Matthews, D. J., C. L. Emson, G. J. McKenzie, H. E. Jolin, J. M. Blackwell, and A. N. McKenzie. 2000. IL-13 is a susceptibility factor for Leishmania major infection. J. Immunol. 164:1458-1462.
15. McGuirk, P., C. McCann, and K. H. Mills. 2002. Pathogen-specific T regulatory 1 cells induced in the respiratory tract by a bacterial molecule that stimulates interleukin 10 production by dendritic cells: a novel strategy for evasion of protective T helper type 1 responses by Bordetella pertussis. J. Exp. Med. 195:221-231.
16. McKenzie, G. J., P. G. Fallon, C. L. Emson, R. K. Grencis, and A. N. McKenzie. 1999. Simultaneous disruption of interleukin (IL)-4 and IL-13 defines individual roles in T helper cell type 2-mediated responses. J. Exp. Med. 189:1565-1572.
17. Mendel, I., and E. M. Shevach. 2002. The IL-10-producing competence of Th2 cells generated in vitro is IL-4 dependent. Eur. J. Immunol. 32:3216-3224.
18. Mougneau, E., F. Altare, A. E. Wakil, S. Zheng, T. Coppola, Z.-E. Wang, R. Waldmann, R. M. Locksley, and N. Glaichenhaus. 1995. Expression cloning of a protective Leishmania antigen. Science 268:563-566.
19. Nashed, B. F., Y. Maekawa, M. Takashima, T. Zhang, K. Ishii, T. Dainichi, H. Ishikawa, T. Sakai, H. Hisaeda, and K. Himeno. 2000. Different cytokines are required for induction and maintenance of the Th2 type response in DBA/2 mice resistant to infection with Leishmania major. Microbes Infect. 2:1435-1443.
20. Noben-Trauth, N., R. Lira, H. Nagase, W. E. Paul, and D. L. Sacks. 2003. The relative contribution of IL-4 receptor signaling and IL-10 to susceptibility to Leishmania major. J. Immunol. 170:5152-5158.
21. Stober, C., U. Lange, M. T. Roberts, B. C. Gilbmartin, R. Francis, R. Almeida, C. S. Peacock, S. McCann, and J. M. Blackwell. From genome to vaccines for leishmaniasis: screening 100 novel vaccine candidates against murine Leishmania major infection. Submitted for publication.
22. Stober, C., U. Lange, M. T. M. Roberts, A. Alcami, and J. M. Blackwell. 2005. IL-10 from regulatory T cells determines vaccine efficacy in murine Leishmania major infection. J. Immunol. 175:2517-2514.
23. Swihart, K., U. Fruth, N. Messmer, K. Hug, R. Behin, S. Huang, G. Del Giudice, M. Aguet, and J. A. Louis. 1995. Mice from a genetically resistant background lacking the interferon gamma receptor are susceptible to infection with Leishmania major but mount a polarized T helper cell 1-type CD4+ T-cell response. J. Exp. Med. 181:961-971.
24. Wang, Z. E., S. L. Reiner, S. Zheng, D. K. Dalton, and R. M. Locksley. 1994. CD4+ effector cells default to the Th2 pathway in interferon gamma-deficient mice infected with Leishmania major. J. Exp. Med. 179:1367-1371.(Mark T. M. Roberts, Carme)