Pretreatment with Recombinant Flt3 Ligand Partially Protects against Progressive Cutaneous Leishmaniasis in Susceptible BALB/c Mice

Abstract

Dendritic cells are potent antigen-presenting cells that also produce interleukin-12 (IL-12) during innate and adaptive cellular immune responses and that thereby promote the differentiation of gamma interferon (IFN-γ)-producing Th1-type CD4⁺ T lymphocytes. We hypothesized that expanded dendritic-cell populations in mice pretreated with the hematopoietic cytokine Flt3L would protect against cutaneous Leishmania major infection. Pretreatment of disease-susceptible BALB/c mice with 10 μg of recombinant Flt3L (rFlt3L) for 9 to 10 days before infection increased lymph node IL-12 p40 productive capacity 20-fold compared to that of saline-injected controls. Furthermore, 9 of 22 (40.9%) rFlt3L-pretreated BALB/c mice resolved their cutaneous infections, whereas none of the 22 control BALB/c mice healed. Healed, rFlt3L-pretreated mice did not develop disease following reinfection. Flt3L pretreatment also reduced parasite numbers 1,000-fold in the cutaneous lesions at 2 weeks after infection relative to numbers in lesions of untreated controls. However, Flt3L pretreatment did not significantly alter L. major-induced IFN-γ and IL-4 production in lymph node culture at 1, 2, and 4 weeks after infection. Despite the lack of Th immune deviation, Flt3L ligand-pretreated lymph nodes expressed up to 10-fold higher levels of IL-12 p40 and inducible (type 2) nitric oxide synthase mRNA at 7 days after infection. In contrast, treatment with rFlt3L after infection failed to protect against disease despite comparable expansions of dendritic cells and IL-12 p40 productive capacity in both infected and uninfected BALB/c mice treated with rFlt3L. We conclude that rFlt3L pretreatment before infection with L. major reduces parasite load and promotes healing of cutaneous lesions without stable cytokine deviation towards a dominant Th1 cytokine phenotype.

Different inbred strains of mice infected cutaneously with the protozoan parasite Leishmania major demonstrate highly polarized T-cell responses that mediate dissimilar disease outcomes (27). Most strains of mice, such as C57BL/6, contain infection through the expansion of antigen-specific, Th1-type CD4⁺ T-cell populations that produce gamma interferon (IFN-γ) without interleukin-4 (IL-4). In contrast, disease-susceptible BALB/c mice are strongly biased towards the development of CD4⁺ T-cell responses productive of IL-4 and IL-13. These Th2-type cytokines directly antagonize IFN-γ-dependent inducible (type 2) nitric oxide synthase (iNOS) expression and other macrophage-based mechanisms necessary for killing the intracellular parasite (20, 21). Curative immunity in BALB/c mice can be restored by interventions that prevent Th2 development in the first week of infection while promoting Th1 cytokine unipolarity, such as treatment with recombinant IL-12 (rIL-12) or anti-IL-4 antibody (3, 12). The central regulatory role of CD40-inducible IL-12 in directing protective immunity has been especially well characterized in this model. In resistant strains of mice, production of endogenous IL-12 increases in the second week of infection and is necessary for the cure of disease (10). IL-12 production is in turn dependent on activation of antigen-presenting cells (APCs) through engagement of CD40 by CD40L-expressing T cells. Genetic or antibody-mediated disruption of CD40/CD40 ligand function correspondingly disrupts Th1 T-cell development and prevents healing (13, 15).

Dendritic cells are probably the major source of IL-12 in L. major-infected mice. These APCs produce IL-12 by CD40-dependent mechanisms during the adaptive phase of cellular immunity but have also been recently shown to secrete IL-12 in direct response to cellular invasion by leishmania amastigotes (16, 34). Epidermal Langerhans cells and dermal dendritic cells are also competent IL-12-producing cells and are well situated to contribute to the early IL-12 response following cutaneous infection and to induce protective immunity when pulsed with antigen and used to vaccinate susceptible mice (8, 14). In contrast, macrophages, which are the major cellular target for parasite invasion, become unable to synthesize IL-12 when infected (2). Increased numbers of mature dendritic cells at the onset of infection with L. major might therefore be predicted to protect against disease in susceptible BALB/c mice by enhancing both presentation of antigens and production of immunoregulatory IL-12.

Dendritic cells in lymphoid organs and skin can be markedly expanded by treatment with recombinant Flt3 ligand (rFlt3L) (19; I. Kremer, K. Cooper, and F. Heinzel, 60th Annu. Meet. Soc. Investig. Dermatol, J. Investig. Dermatol. 112:524). The cognate receptor for this cytokine, Flt3 (fms-like tyrosine kinase-3), is a transmembrane receptor expressed on hematopoietic progenitor cells and active in the differentiation of several bone marrow lineages. Flt3 activation results in the preferential expansion of dendritic cells in the presence of additional differentiating signals, such as c-kit ligand and granulocyte-macrophage colony-stimulating factor (24). Recombinant, soluble human forms of Flt3L are 85% homologous to murine Flt3L and capable of expanding phenotypically mature dendritic cells when administered to mice for 9 to 10 days (17). The lymphoid organs of Flt3L-treated mice demonstrate enhanced presentation of protein or allogeneic antigens (19), and treatment with Flt3L augments antitumor immunity (5, 19). Granulocyte, macrophage, and immature myeloid cell populations are also expanded, but the observed immunologic effects are attributed to dendritic cells expressing high levels of CD11c and high-to-intermediate levels of CD11b (19, 31).

Based on these observations, we hypothesized that rFlt3 ligand pretreatment to expand dendritic-cell numbers prior to L. major infection of BALB/c mice would promote unipolar Th1-type immune responses and prevent Th2-dependent progression of cutaneous disease. A secondary aim of these studies was to examine immunoregulatory effects mediated by dendritic-cell expansions peaking later in infection. IL-12 synthesis in resistant mice is normally delayed until 2 weeks after infection. This has been attributed to a specific requirement for promastigote-to-amastigote conversion for stimulation of IL-12 (28) but may instead reflect the time needed to mobilize increased numbers of lymph node dendritic cells, which are the primary source of IL-12. As support for this, both lymph node dendritic-cell numbers and IL-12 productive capacity increase in tandem during infection of resistant C57BL/6 mice but similar increases fail to occur in susceptible BALB/c mice (11). We speculated that experimental augmentation of postinfection dendritic-cell numbers might provide enough IL-12, and possibly other Th1-promoting costimulatory signals, to reverse the Th2-mediated progression of BALB/c leishmaniasis. Our findings instead show that Flt3L treatment is beneficial only when administered as pretreatment, so that dendritic-cell numbers peak at the time of cutaneous infection with L. major.

MATERIALS AND METHODS

Mice.

Four- to 6-week old female BALB/cByJ and C57BL/6 mice were purchased from Jackson Laboratories (Bar Harbor, Maine) and housed in the Case Western Reserve University animal facilities under specific-pathogen-free conditions.

Parasite cultivation and mouse infection.

L. major (World Health Organization strain WHOM/IR/-/173) was grown in M199 medium (BioWhittaker, Walkersville, Md.) containing antibiotics, supplemental glutamine, and 30% fetal calf serum (HyClone Laboratories, Logan, Utah) as described previously (30). Stationary-phase promastigotes were injected into the hind feet of recipient mice at a dose of 2 × 10⁶ organisms/footpad to initiate infection. The course of infection was monitored by measuring the thickness of footpad swelling weekly using a dial gauge caliper.

Reagents.

Recombinant human Flt3 ligand produced by CHO cells was generously provided by Elaine K. Thomas (Immunex Corp., Seattle, Wash.). rFlt3L was diluted in phosphate-buffered saline–0.1% mouse serum albumin (Sigma, St. Louis, Mo.), and 5 μg was injected in a volume of 50 μl in each hind limb of mice daily for a total of 9 or 10 injections.

Flow-cytometric analysis of accessory cell populations.

Lymph node and spleen tissue was crushed and digested in Hanks balanced salt solution (HBSS) containing collagenase IV (400 U/ml; Boerhinger Mannheim, Indianapolis, Ind.) at 37°C for 15 min, red cells were lysed using hypotonic ACK lysis buffer (150 mM ammonium chloride, 10 mM potassium carbonate, and 0.1 mM EDTA adjusted to pH 7.4), and nucleated cells were washed and resuspended in HBSS–1% fetal calf serum containing 5 mM EDTA. The unlabeled anti-FcRII/III monoclonal antibody (MAb) (10 μg of 2.4G2/ml; Pharmingen, San Diego, Calif.) was added to block nonspecific labeling due to FcR binding. Cells were then incubated with fluorescein isothiocyanate (FITC) anti-B220 (RA3-6B2), FITC anti-CD3 (2C11), phycoerythrin anti-CD11b (M1/70), and either biotinylated anti-CD40 (3/23), anti-CD11c (HL3), anti-CD8α (2.43), or anti-major histocompatibility complex class II (MHC-II) (M5/114) MAbs (Pharmingen). After a 30-min incubation at 4°C, the cells were washed three times and streptavidin-CyChrome (Pharmingen) was added for another 20-min incubation. Irrelevant biotinylated rat immunoglobulin G2b (IgG2b) (Pharmingen) was used as a negative control for streptavidin-CyChrome staining. The cells were again washed and then fixed in 1% formalin prior to analysis by FACscan (Becton Dickinson Immunocytometry Systems, Mountain View, Calif.). Cells expressing FITC B220 or CD3 were excluded by gating within the FL1 channel, and 5,000 CD19⁻ B220⁻ cells were analyzed.

Culture of lymph node cells.

Lymph node cells harvested from uninfected or infected mice were washed three times, counted, and suspended in Dulbecco modified Eagle medium (DMEM) (BioWhittaker) containing antibiotics, 2 mM glutamine, 0.1 mM nonessential amino acids, 10% fetal bovine serum (FBS), and 10 mM HEPES (pH 7.4). Cells were aliquoted into flat-bottom 96-well culture plates at 10⁶ cells per well and cultured for 48 h in DMEM–10% FBS. Stimuli included 10 μg of soluble L. major promastigote antigen (SLA)/ml. An anti-IL-4 receptor MAb (M-1; 10 μg/ml; Genzyme Corp.) was added to the culture to prevent loss of assayable IL-4 due to receptor binding (10). Conditioned media were removed at 48 h for enzyme-linked immunosorbent assay (ELISA) measurement of cytokines.

Cytokine ELISA assays.

Culture supernatants were assayed for murine cytokines using double-sandwich MAb ELISA techniques as previously described (10).

Quantitative parasite cultures.

Approximately 0.2 g of footpad tissue was minced in 2 ml of M199 medium, crushed through a 200-mesh stainless steel screen, and disrupted using a Ten-Broeck homogenizer. Footpad and lymph node suspensions were serially diluted fivefold in promastigote growth medium (M199–20% FBS) and incubated in flat-bottom 96-well plates at 26°C in humidified room air. Individual wells were examined using an inverted microscope at ×200 at 2-day intervals for the presence of motile promastigotes. Data represent the geometric mean and standard error of the last positive reciprocal dilution for each experimental group.

Reverse transcription-PCR analysis of cytokine mRNA.

Lymph node mRNA was isolated from popliteal lymph nodes using STAT-60 (Teltest, Friendswood, Tex.) in accordance with the manufacturer's instructions. Complementary DNA was produced using Superscript II reverse transcriptase and oligo(dT)-primed RNA. Comparative PCR was performed as previously described (11), with the number of amplification cycles adjusted to maintain linear kinetics of product formation. All amplifications included template-negative controls to exclude possible contamination. DNA was separated on a 1.5% agarose gel, visualized with ethidium bromide or Sybyr Gold, and scanned and quantitated using Gel-Doc software (Bio-Rad) and Optimas (Bothell, Wash.) image analysis software. Primer sequences used in these studies have been previously described (11).

Statistics.

The significance of differences in cytokine levels was assessed using the Mann-Whitney rank sum test or the Student t test. Differences in frequencies of healing or progression of disease were analyzed by Fisher's exact test.

RESULTS

Subcutaneous injection with 10 μg of Flt3L increases CD40-dependent IL-12 p40 production in the peripheral lymph nodes.

BALB/c mice were injected subcutaneously in the hind extremities with a total daily dose of 10 μg of recombinant human Flt3L for 10 days. Control BALB/c mice were injected with 50 μl of saline containing 0.1% mouse serum albumin. At day 11, analytical three-color flow cytometry confirmed nine- to sevenfold increases in CD40⁺ CD3⁻ CD19⁻ and CD11c⁺ CD3⁻ CD19⁻ cells in the lymph nodes of mice treated with Flt3L, relative to saline-injected controls (Fig. 1A). These expanded populations consisted largely of cells staining intermediate to high for CD11b and positive for CD40 (Fig. 1A), as well as for MHC-II and CD86 (not shown). Consistent with the presence of increased numbers of functional, CD40-expressing dendritic cells, the cultured lymph node cells of Flt3L-treated mice produced 20 ± 2.3 times more spontaneous and anti-CD40-induced IL-12 p40 than saline-treated mice (Fig. 1B). Serum IL-12 p40 levels were also increased in rFlt3L-treated mice compared to those in controls (14.5 ± 1.2 ng/ml compared to 3.7 ± 0.4 ng/ml, respectively) on day 10 of rFlt3L injection. These data not only confirmed the potent dendritic-cell-expanding effects of rFlt3L in vivo but also demonstrated augmentation of both local and systemic IL-12 p40 productive capacities that, if accompanied by increased IL-12 heterodimer production, was predicted to produce a bias towards curative Th1-type cellular immune responses during infection with L. major.

FIG. 1.

Pretreatment with rFlt3L for 10 days increases lymph node dendritic-cell numbers and productive capacity for spontaneous and CD40-inducible IL-12 p40. (A) Fluorescence-activated cell sorter analysis of non-T- and non-B-cell popliteal lymph node populations. Cells staining positively for FITC anti-B220 and FITC anti-CD3 were excluded from analysis. Vertical axis, labeling intensity of phycoerythrin-conjugated anti-CD11b antibody for the remaining cells; horizontal axis, reactivity with CyChrome-labeled isotype IgG isotype control antibody and anti-CD40 and anti-CD11c antibodies. CD11c⁺ and CD40⁺ cells that also stained high or intermediate for CD11b⁺ were analyzed within the indicted gate, with the data shown as percentages of total analyzed cells. Similar findings were obtained in a study of the spleen cell populations. These data are representative of three experiments. (B) Lymph node cells from control and rFlt3L-treated BALB/c mice (n = 3) were suspended in DMEM–10% FBS in the presence of 5 μg of control rat IgG antibody/ml or rat anti-mouse CD40 MAb FGK45. After 48 h of culture, the conditioned media were assayed for mouse IL-12 p40 by ELISA. Shown are the concentrations (± standard errors of the means) of IL-12 p40 detected for control and rFlt3L-treated mice. The increases in spontaneous and anti-CD40-inducible IL-12 p40 after rFlt3L treatment were 24- and 20-fold, respectively (P < 0.05).

Pretreatment with Flt3L partially protects against progressive cutaneous leishmaniasis.

To test the hypothesis that pretreatment with Flt3L partially protects against progressive cutaneous leishmaniasis, disease-susceptible BALB/c mice were pretreated subcutaneously in the hind limbs with 10 μg of Flt3L for 10 days before infection with 2 × 10⁶ L. major promastigotes in the hind feet. Other groups of BALB/c mice were pretreated for 10 days with endotoxin-free saline containing 0.1% mouse serum albumin. We observed a pronounced delay in the development of footpad swelling in mice pretreated with Flt3L but not in mice treated with saline (Fig. 2). Starting at 5 weeks after infection, two of five Flt3L-pretreated mice developed slowly progressive footpad thickening, whereas three resolved footpad thickening. This finding was reproduced in four long-term experiments lasting from 4 to 17 weeks, with a total of 9 of 22 (40.9%) Flt3L-pretreated mice either resolving or failing to develop progressive footpad thickening (final measurement was less than 3 mm), compared to none of the 22 saline-treated controls (Table 1). This difference was statistically significant (P ≤ 0.001; two-sided Fisher's exact test). The approximately 60% of Flt3-pretreated mice that did not heal were partially protected, as demonstrated by an average 3- to 4-week delay in the progression of footpad thickening relative to controls.

FIG. 2.

Pretreatment of BALB/c mice with rFlt3L protects against progression of cutaneous L. major infection and leads to long-term resistance to reinfection. (A) Groups of five BALB/c mice were pretreated for 9 days with saline (control) or 10 μg of rFlt3L given as subcutaneous injections split between both hindlimbs. On day 10, mice were infected with L. major in the hind feet. Shown are the weekly mean (± standard error of the mean [SEM]) footpad thicknesses from two separate studies (expt 1 and expt 2). Footpad data for rFlt3L-pretreated mice are shown individually for each animal to indicate the dichotomous effects of pretreatment on outcome. Control mice were euthanized at 6 and 7 weeks due to progressive ulceration and necrosis. (B) Surviving rFlt3L-pretreated mice (n = 3) were reinfected at 13 weeks after infection and compared to a new group of control BALB/c mice.

TABLE 1.

Effect of rFlt3L pre- and posttreatment on disease outcomes in L. major-infected BALB/c mice

Expt	Treatmentc	Days of rFlt3L injectiona	No. of mice cured/totalb
1	Saline	−10–0	0/5
	Flt3 ligand	−10–0	3/5
2	Saline	−10–0	0/5
	Flt3 ligand	−10–0	2/5
	Flt3 ligand	0–+10	0/5
3	Saline	−9–0	0/8
	Flt3 ligand	−9–0	4/8
4	Saline	−10–0	0/4
	Flt3 ligand	−10–0	0/4

^aRecombinant human Flt3L was injected into both hind feet at 5 μg/foot daily for the indicated days relative to the start of infection (day 0). 

^bCure was defined by a footpad thickness of <3 mm at the end of the experiment. Mice with nonprogressive disease were monitored for 12 weeks in experiments 1 and 2 and for 4 and 5 weeks for experiments 3 and 4, respectively. 

^cTreatment with Flt3 ligand consisted of 10 μg given subcutaneously. 

Flt3L-induced resolution of cutaneous lesions is associated with reduced parasite load and resistance to reinfection.

Reductions in footpad thickening were correlated with approximately 1,000-fold reduced numbers of viable parasites present within the footpad tissue, as determined by quantitative culture of homogenized tissues (Table 2). This reduction was first apparent at 2 weeks after infection and persisted in rFlt3L-pretreated mice demonstrating small footpad lesions after 4 weeks of infection. Footpad thickening in rFlt3L-pretreated mice whose lesions were not restricted by 4 weeks after infection was, in contrast, associated with a parasite load not significantly different from that of controls, validating the change in lesion size as an index for rFlt3L-altered parasite burden in these studies.

TABLE 2.

Pretreatment with rFlt3L reduces parasite burden at the site of L. major infection^a

Wk after infection of tissue harvest	Treatment	n	Mean log10 parasites/g of tissue ± SEM	Pb
2	Saline (control)	5	6.85 ± 0.40	NA
	rFlt3L	5	3.80 ± 1.21	0.015
4	Saline (control)	8	6.61 ± 0.51	NA
	rFlt3L Tx	4 (nonhealing)	5.89 ± 0.97	0.12
		4 (healing)	3.62 ± 0.62	<0.01

^aGroups of eight mice were treated with saline or 5 μg of rFlt3L subcutaneously in both footpads for 9 days before infection. Footpad tissues were diced and crushed in M199-30% FBS culture media and sequentially diluted fivefold for limiting dilution analysis. 

^bCompared to saline control. NA, testing is not applicable. 

Long-term control of lesions during primary infection also led to complete resistance to reinfection. Three rFlt3L-pretreated BALB/c mice that had resolved cutaneous disease for a period of 14 weeks after infection were reinfected with 2 × 10⁶ L. major promastigotes (Fig. 2). A group of normal BALB/c mice were similarly infected to serve as a normally susceptible control group. Over a period of 4 more weeks, the previously cured mice did not demonstrate increased footpad swelling, whereas footpads in the infected control mice increased by 2.5 mm in thickness and developed ulcers.

Antigen-specific cytokine response of the draining lymph nodes during infection in Flt3L-pre-treated mice.

The draining lymph nodes of L. major-infected mice were examined at 1, 2, and 4 weeks after infection to determine if protective therapy with rFlt3L increased parasite-specific production of IFN-γ or blocked the Th2-dominant cytokine response normally observed in infected BALB/c mice (Fig. 3). These responses mediate resistance or susceptibility to progressive disease, respectively. However, antigen-specific IFN-γ responses of control and rFlt3L-treated mice during primary infection were not significantly different. Pretreatment did not increase IFN-γ productive capacity at 2 and 4 weeks, when leishmania-induced immune responses are typically maximal during infection. Treatment with rFlt3L also failed to induce IFN-γ synthesis at 1 week after infection, when this response is normally absent in control infected mice. Furthermore, IL-4 production was not significantly or stably reduced in rFlt3L-pretreated mice during infection. Specifically, IL-4 levels in culture were not significantly different from those of control mice at 1 and 4 weeks after infection and IL-4 production was transiently decreased only twofold relative to that in control mice at 2 weeks after infection (P = 0.08; Mann-Whitney U test). In contrast, resistant C57BL/6 mice produced approximately three times as much antigen-specific IFN-γ and fourfold less IL-4 at 2 weeks after infection (P < 0.05 for both). Finally, comparisons between rFlt3L-treated mice with healing or progressive cutaneous lesions at 4 weeks after infection failed to correlate cure with either reduced IL-4 production or increased IFN-γ production.

FIG. 3.

Effects of rFlt3L pretreatment and L. major infection on the antigen-specific cytokine response of draining lymph node cells. BALB/c mice were pretreated with Flt3L for 10 days and infected with L. major. Popliteal lymph nodes were harvested at the indicated times after infection. Cell suspensions were incubated with media or 20 μg of SLA/ml. Conditioned media were assayed for IFN-γ and IL-4 after 48 h of culture. Lymph nodes from concurrently infected, disease-resistant C57BL/6 mice were included in the 2-week analysis. At 4 weeks after infection, rFlt3L-treated mice showing progressively decreasing footpad sizes (Flt3L healer) or increasing sizes (Flt3L nonhealer) were separately analyzed for cytokine response. Also included (14 wk reinfection) are lymph node responses 4 weeks after concurrent L. major infection of naive BALB/c mice and BALB/c mice that had been treated with rFlt3L and cured 14 weeks previously. Naive lymph node cultures produced less than 0.1 and 0.04 ng of SLA-stimulated IFN-γ and IL-4/ml, respectively (not shown). Control BALB/c mice infected for 4 weeks produced similar amounts of IFN-γ and IL-4 spontaneously or in response to the L. major antigen. Spontaneous cytokine release was otherwise at least fourfold less than that induced by the antigen alone.

However, reinfection of healed, rFlt3L-pretreated BALB/c mice did not result in disease and was instead associated with preserved Th1 responses unopposed by Th2 activity. Specifically, rFlt3L-cured mice produced similar amounts of IFN-γ and fivefold less IL-4 (P < 0.05) than control infected BALB/c mice (Fig. 3). This is consistent with immune deviation as a mechanism for secondary resistance to infection. A similar pattern of reduced IL-4 recall responsiveness and preserved IFN-γ synthesis characterizes reinfection of BALB/c mice previously cured of infection by early treatment with rIL-12 (12).

Increased lymph node expression of IL-12 p40 and iNOS mRNA in rFlt3L-treated mice without altered IFN-γ and IL-4 expression.

Levels of IFN-γ and IL-4 mRNA expression were comparable in control and rFlt3L-pretreated BALB/c mice at 1 and 2 weeks after infection, with differences not exceeding twofold (Fig. 4). However, levels of both IL-12 p40 and iNOS mRNAs were markedly increased in the lymph nodes of rFlt3L-pretreated mice at 7 days after infection compared to those in control infected mice. Whereas IL-12 p40 expression decreased in control mice following infection, IL-12 p40 mRNA increased in rFlt3L-treated mice to levels that were eightfold greater than that in concurrently infected control tissues. Although rFlt3L-treated mice did not exhibit changes in IFN-γ or IL-4 relative to control mice, expression of iNOS mRNA was 10-fold greater than that in control mice at 7 days after infection. No iNOS mRNA was present in uninfected tissues. At 14 days after infection, both IL-12 p40 and iNOS persisted at levels twofold greater than control values. This pattern of 10-fold-increased IL-12 p40 and iNOS mRNA expression in infected, rFlt3L-pretreated mice was observed in a second study.

FIG. 4.

Effect of rFlt3L pretreatment on cytokine and iNOS mRNA expression in the draining lymph nodes of L. major-infected BALB/c mice. BALB/c mice were injected daily for 10 days with saline or 10 μg of rFlt3L split between both hindlimbs and then infected with L. major. Popliteal lymph node tissue RNA was obtained from pooled popliteal lymph nodes on days 7 and 14 after infection for analysis by comparative reverse transcription-PCR (RT-PCR). Data represent densitometric values for the indicated PCR products normalized for expression of the housekeeping hypoxanthine phosphoribosyltransferase (HPRT) gene. RNA was reverse transcribed using oligo(dT), and PCR was performed as indicated in Materials and Methods. Control lymph node tissue is from uninfected, untreated BALB/c mice. These data were reproducible following repeated reverse transcription and PCR analysis.

Delayed rFlt3L treatment does not alter progressive cutaneous disease in BALB/c mice.

Because pretreatment with rFlt3L cured L. major-infected BALB/c mice, we next studied the issue of whether delayed treatment with rFlt3L would have similar effects on disease and immunologic outcomes. Surprisingly, 10 days of rFlt3L injections starting at the time of infection provided no more protection against the development of cutaneous pathology than did saline (Fig. 5). In contrast, pretreatment with rFlt3L cured half of the infected BALB/c mice and profoundly reduced the rate of disease progression in the others. Although L. major infection may have antagonized the dendritic-cell-inducing effects of rFlt3L, we confirmed that dendritic cells were present in similar numbers in the draining lymph nodes of both infected and uninfected mice treated for 10 days with 10 μg of rFlt3L (data not shown). Infection also had no effect on the spontaneous and CD40-inducible IL-12-productive capacity of rFlt3L-treated mice (Fig. 6). Delayed Flt3L treatment again did not enhance IFN-γ production relative to that of control lymph node cells but significantly increased IL-4 production in infected mice relative to that in saline-treated, infected BALB/c mice. Despite the similar expansion of IL-12 p40 production by rFlt3L-induced APC populations, the major effect of delayed rFlt3L therapy in L. major infection was to exaggerate the inherent BALB/c bias towards parasite-specific IL-4 responses.

FIG. 5.

Pretreatment, but not delayed treatment, with rFlt3L protects BALB/c mice against progression of cutaneous L. major infection. Groups of four BALB/c mice were treated for 10 days with saline (control) or 10 μg of rFlt3L (Flt3 pretreatment) given daily by subcutaneous injection split between both hindlimbs. On day 11, all mice were infected with 2 × 10⁶ L. major promastigotes. A separate group of mice (posttreatment) were treated with a 10-day course of rFlt3L starting on the day of infection. Shown are weekly mean footpad thicknesses (± standard errors of the means) measured weekly as a marker for disease progression. Control and rFlt3L posttreatment mice were euthanized at week 6 due to progressive ulceration and necrosis in all animals. Starting at week 8, two of four rFlt3L pretreatment mice with resolving footpad thicknesses (healers) are shown separately from the remaining mice, which demonstrated slowly progressive disease.

FIG. 6.

Effects of L. major infection on cellular and cytokine responses following delayed rFlt3L therapy. Groups of five BALB/c mice were infected in both hind feet with L. major and treated for the subsequent 10 days with saline or 10 μg of rFlt3L administered as half doses in each hindlimb. A separate group of uninfected BALB/c mice was similarly treated with rFlt3L. The popliteal lymph nodes were harvested from all groups on day 11 after infection and studied for IFN-γ and IL-4 production in response to 10 μg of SLA/ml or for IL-12 p40 production in response to 10 μg of anti-CD40 MAb/ml. Solid bars, unstimulated controls. Shown are the mean (± standard error of the mean) cytokine levels present after 48 h of culture. IL-4 production in response to the antigen was significantly greater in rFlt3L-posttreated mice than in infected controls (P < 0.05; Mann-Whitney U test).

DISCUSSION

The central finding of these studies is that pretreatment with recombinant human Flt3L prevents progressive cutaneous disease in approximately 40% of L. major-infected BALB/c mice. This is the first report showing an anti-infective function for this hematopoietic cytokine. Pretreated mice not only had reduced lesion sizes but also demonstrated 1,000-fold reductions in parasite load starting as early as 2 weeks after infection. Flt3-assisted cure also resulted in long-lasting immunity against reinfection. However, the mechanism by which Flt3L treatment exerts its protective effects during primary infection was not associated with redirected T-cell differentiation towards unipolar Th1 cytokine responses that are normally required for cure. In particular, lymph node IL-4 production in vitro and IL-4 mRNA expression in vivo were not suppressed significantly or durably in rFlt3L-pretreated mice. This lack of stable immune deviation in rFlt3L-treated mice during primary infection is unusual compared to other forms of curative immunotherapy previously described in the model of murine leishmaniasis. For instance, BALB/c mice treated with activating CD40 antibodies, rIL-12, CTLA4-Ig, anti-IL-4 MAb, or cytotoxic anti-CD4⁺ MAb in the first week after L. major infection develop unopposed Th1-type responses by the third or fourth week after infection; this altered cytokine phenotype mediates the observed change in disease outcome (3, 4, 6, 12, 29). The failure of rFlt3L treatment to similarly bias towards unipolar Th1 cytokine responses was especially surprising in that treatment promoted strong regional and systemic IL-12 p40 synthesis. Flt3L-treated mice sustained 8-fold increased levels of IL-12 p40 mRNA in the infected lymph node at 7 days after infection, 20-fold increases in IL-12 p40 protein in cultured lymph node cells, and significantly greater levels of circulating IL-12 p40. If indicative of IL-12 heterodimer synthesis, this should have predicted enhanced, early development of IFN-γ-producing cells and suppressed expansion of deleterious IL-4 and IL-13-producing cells (12). However, preliminary studies have not confirmed increased IL-12 p70 levels in lymph node or splenic cultures as a result of rFlt3L pretreatment (data not shown). Until these major mechanistic issues are resolved by more detailed analysis of direct and indirect effects of rFlt3L on iNOS activation and cytokine production by different APC populations, our findings most clearly indicate that rFlt3L treatment mediates recovery from primary infection through mechanisms that are not associated with T-cell deviation towards a protective Th1 phenotype.

Despite the lack of strong Th1 cytokine responses in the lymph nodes draining the cutaneous site of infection, iNOS expression was strongly increased in rFlt3L-treated mice at 7 days. This preceded the decrease in tissue parasite load and represents the most likely mechanism for the partial protection observed. Nitric oxide is a critical leishmanicidal agent in the protective murine host response, and the earlier and greater expression of iNOS in resistant C57BL/6 mice, compared to that in BALB/c mice, correlates with rapid parasite clearance (32). Consequently, rFlt3L-induced iNOS activity in the first week of infection may have been sufficient to reduce the initial parasite load to levels that were suboptimal for sustained infection. A threshold infectious inoculum of L. major has been defined previously for BALB/c mice (1). Progressive reductions in numbers of infecting parasites across this threshold result in increasingly delayed rates of lesion development. At very low doses, lesions fail to appear and mice develop long-lasting resistance to disease associated with the emergence of detectable unipolar Th1 immunity only at the time of reinfection. The appearance of strongly Th1-polarized responses only after reinfection of rFlt3L-cured mice is consistent with prior control of suboptimal inocula as a potential mechanism for cure. However, these findings do not indicate how rFlt3L treatment activates the expression of iNOS. Because iNOS expression is typically induced in response to IFN-γ, transient generation of this cytokine at times not assayed in these studies cannot be ruled out. However, the acute induction of IFN-γ mRNA expression normally observed at 1 to 2 days after infection was not enhanced in rFlt3L-pretreated mice (data not shown). Alternatively, rFlt3L-directed myeloid- and dendritic-cell differentiation might instead enhance the sensitivity of APCs to IFN-γ-dependent expression of iNOS. Future studies are also needed to determine if the anti-infective properties of rFlt3L in murine leishmaniasis are IFN-γ dependent or mediated by iNOS or other leishmanicidal responses directly induced by rFlt3L exposure at the site of infection.

Another unexpected finding was the failure of rFlt3L pretreatment to expand cytokine recall responses of lymph node T cells to leishmanial antigen, which was not predicted by other studies showing enhanced in vivo APC function in rFlt3L-treated mice. In other experimental systems, Flt3L treatment increased the magnitude of T-cell responses to test antigens (26, 33, 35) and potentiated antitumor cellular immunity mediated by NK cells and cytolytic CD8⁺ T cells (5, 7). The lack of accelerated or enhanced CD4⁺ T-cell responsiveness in our studies might instead reflect the unique biology of live Leishmania infection compared to responses engendered by soluble antigen. For instance, Leishmania-infected macrophages exhibit deficient APC function and effectively sequester antigens from MHC-II processing and presentation during the early phase of infection (9, 22, 23, 36). Increased numbers of APCs induced by rFlt3L therefore may not necessarily recruit increased T-cell responses if access to parasite antigens is limited during early infection. Dendritic-cell numbers also decline to normal levels 7 days after stopping rFlt3L therapy, suggesting that subsequent increases in antigenic load may not coincide with peak APC function in vivo. Although we attempted to circumvent this by starting rFlt3L treatment at the time of infection, this served only to markedly increase L. major-specific IL-4 responses without altering IFN-γ synthesis. These mice were also fully susceptible to progressive disease. These data suggest that the protective effect of Flt3L is mediated before 10 days of infection and may be distinct from effects mediated by enhanced APC function, which did not expand Th1 development in vivo and which instead amplified the intrinsic bias towards Th2 development present after the first week of BALB/c infection.

In summary, pretreatment with rFlt3L before L. major infection cures about 40% of disease-susceptible BALB/c mice despite the persistence of IL-4-dominated cytokine responses indistinguishable from those of control infected BALB/c mice. The presence of increased IL-12 p40 production and iNOS expression in vivo indicates that rFlt3L-induced macrophage and dendritic-cell populations are activated, and we speculate that this may be related to early parasite killing. Otherwise, the lack of enhanced Th1 antigen-specific responses consistent with increased APC function in the presence of IL-12 p40 is unexplained. Leishmania-induced sequestration of antigens within parasitized macrophages may be contributory, in which case these studies do not argue against the potential use of rFlt3L as a vaccine adjuvant. Future studies to identify possible antagonisms due to the mixed expansion of distinct myeloid and lymphoid dendritic cells with different immunoregulatory functions are indicated (18, 25). The present study does not exclude increased production of immunosuppressive cytokines from rFlt3L-expanded cells, although no differences in IL-10 mRNA expression were observed (data not shown). We also cannot exclude a defect specific to synthesis of bioactive IL-12 heterodimer, but not IL-12 p40, in rFlt3L-pretreated mice. Because the protective effect of rFlt3L was rapidly lost when dendritic-cell expansions peaked after the first week of infection, increased numbers of dendritic cells in the draining lymph node appear to be insufficient to reverse the Th2 phenotype in BALB/c mice with established infection, thus confirming a dominant role for CD4⁺ T-cell-intrinsic mechanisms in mediating susceptibility. We conclude that rFlt3L pretreatment protects against infectious disease caused by a well-characterized intracellular protozoan parasite. Since in vivo microbicidal responses were apparently triggered in the absence of Th1-biased adaptive cellular immunity, the anti-infective properties of rFlt3L may be induced directly by this hematopoietic cytokine or through activation of innate cellular immunity. Further study of the microbicidal effects of rFlt3L may prove relevant to understanding the basic biology of cellular immunity against intracellular parasitism and may suggest clinical applications of this recombinant cytokine in settings where the adaptive cellular immune response is deficient or inappropriate to specific infectious threats.