Accelerated Control of Visceral Leishmania donovani Infection in Interleukin-6-Deficient Mice

In patients with visceral leishmaniasis, increased levels ofcirculating interleukin-6 (IL-6) regularly accompany fully expressed,progressive infections (kala-azar). To experimentally test therole of IL-6, responses to an intracellular Leishmania donovaniinfection in the livers of IL-6^–/– and wild-typemice were compared. IL-6^–/– mice showed an enhancedcontrol of the infection and earlier, rapid parasite killingalong with additional evidence of a stimulated antileishmanialTh1-cell-type response: increased levels of circulating gammainterferon, accelerated granuloma assembly, and heightened responsivenessto chemotherapy. In this model of visceral leishmaniasis, IL-6appears to act in a suppressive, macrophage-deactivating fashion,which identifies it as a potential target for therapeutic blockade.

INTRODUCTION

Top

INTRODUCTION

Interleukin-6 (IL-6), a pleiotropic cytokine elaborated in responseto a wide range of inflammatory stimuli, including intracellularinfection, is ordinarily considered a proinflammatory factor(13, 18). However, in certain models, IL-6 has also been reportedto downregulate inflammatory mechanisms and impair macrophageactivation and antimicrobial effects (5, 6, 13, 33). IL-6 isexpressed in experimental cutaneous and visceral infectionscaused by Leishmania spp. (1, 11, 12, 17, 38, 45), and the spectrumof its actions has been well illustrated in these models; invitro and in vivo results indicate that IL-6 can promote (40,46), suppress (16, 37), or leave unaltered (23, 42) intracellularantileishmanial host defense. IL-6 is also regularly expressedin all forms of human leishmaniasis (cutaneous, mucosal, andvisceral) (2, 3, 8-10, 15, 19, 20, 22, 34); however, the roleof IL-6 in clinical infection remains uncertain, in part becauseof the above-mentioned disparate experimental results.

Patients who develop fully expressed visceral leishmaniasis(kala-azar) after being infected with Leishmania donovani orL. infantum (L. chagasi) show progressive disease in liver,spleen, and bone marrow. Since this progression occurs in thepresence of high levels of circulating interleukin-6 (IL-6)(2, 10, 34, 44), three possibilities appear to exist for endogenousIL-6 in kala-azar: it induces no antileishmanial action, itspotential prohost defense effect is blunted or overshadowed,or it acts in a suppressive fashion. To begin to experimentallyexamine these possibilities, responses to L. donovani infectionin the livers of IL-6-deficient mice were tested.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

Animals and visceral infection. Breeding pairs of IL-6^–/– mice (C57BL/6 background)were purchased from Jackson Laboratories (Bar Harbor, ME) andbred at Weill Cornell Medical College. Groups of three to fivefemale IL-6^–/– and C57BL/6 wild-type (WT) mice (Jackson),aged 7 to 11 weeks, were injected via the tail vein with 1.5x 10⁷ hamster spleen-derived L. donovani amastigotes (one Sudanstrain) (31). Infection was assessed microscopically using Giemsa-stainedliver imprints, and parasite burdens were measured by countingin a blind manner the number of amastigotes per 500 cell nucleiand multiplying this number by the liver's weight (mg) (Leishman-Donovanunits [LDU]) (31). Differences between mean values were analyzedby a two-tailed Student t test. These studies were reviewedand approved by the Weill Cornell Medical College's InstitutionalAnimal Care and Use Committee.

IFN- ${gamma}$ detection and tissue granuloma responses. Sera were assayed in duplicate at fourfold dilutions for detectinggamma interferon (IFN- ${gamma}$ ) activity using a murine IFN- ${gamma}$ enzyme-linkedimmunosorbent assay (BD-Pharmingen) in which the lower limitof detection was 31 pg/ml. The histologic response to the infectionwas evaluated microscopically in liver sections stained withhematoxylin and eosin. The number of granulomas (infected Kupffercells which had attracted $≥$ 5 mononuclear cells [26, 31]) wascounted in 100 consecutive 40x fields, and at 100 parasitizedfoci, the granulomatous reaction was scored as none, developing,or mature (31). Mature granulomas show a core of fused, parasitizedKupffer cells, numerous surrounding mononuclear cells, and epithelioid-cell-typechanges (28).

Chemotherapy. WT and IL-6^–/– mice were treated intraperitoneally(i.p.) starting 1 week after infection with (i) a single day+7 injection of low-dose (50-mg/kg of body weight) pentavalentantimony (Sb) (sodium stibogluconate [Pentostam]; Wellcome FoundationLtd., London, United Kingdom) or (ii) three injections on days+7, +9, and +11 of low-dose (1-mg/kg) amphotericin B (totaldose, 3 mg/kg; Gensia Laboratories Ltd., Irvine, CA) (27, 30,31). In parallel, separate groups of WT mice received eitheroptimal-dose Sb (500 mg/kg, once on day +7) or optimal-doseamphotericin B (5 mg/kg on days +7, +9, and +11; total dose,15 mg/kg) (27, 30, 31). All mice, including untreated controls,were sacrificed on day +14. Day +14 LDU were compared to day+7 LDU to determine percent parasite killing.

RESULTS AND DISCUSSION

Top

RESULTS AND DISCUSSION

One week after infection, liver parasite burdens were similarin IL-6^–/– and WT animals (Fig. 1). Thereafter,however, the kinetics of intracellular infection diverged, asIL-6^–/– mice showed an enhanced control of L. donovanireplication, earlier onset of parasite killing, and overallaccelerated resolution of hepatic infection. As anticipatedwith the C57BL/6 background (32), WT animals also proceededto convert initial susceptibility to the self-healing phenotype.In this model, the latter is dependent on a Th1-cell-type responseand is mediated by IFN- ${gamma}$ -stimulated mononuclear-cell influx,macrophage activation, and induction of intracellular parasitekilling (25, 26, 32, 39).

View larger version (16K):
[in this window]
[in a new window]

FIG. 1. Course of L. donovani liver infection in WT versus IL-6^–/– mice. Results are from three experiments and indicate mean values ± SEM for 9 to 12 mice per group per time point. *, P < 0.05.

Enhanced antileishmanial activity in IL-6^–/– micewas accompanied by increased levels of circulating IFN- ${gamma}$ (Table1), along with two additional expressions of IFN- ${gamma}$ -dependentdefense against L. donovani—accelerated granuloma assemblyand a heightened responsiveness to chemotherapy (27) (Fig. 2and 3). Serum IFN- ${gamma}$ was not detected in uninfected mice but wasmeasureable earlier and at higher levels in infected IL-6^–/–animals than in infected WT animals. Measuring IFN- ${gamma}$ in serumis a useful marker of a Th1-cell-type response (31); however,the physiologic implication of activity in serum (versus insitu IFN- ${gamma}$ expression in infected tissue) is unknown.

View this table:
[in this window]
[in a new window]

TABLE 1. Serum IFN- ${gamma}$ responses in mice^a

View larger version (168K):
[in this window]
[in a new window]

FIG. 2. Liver histologic reaction 2 weeks after L. donovani challenge. (A) Although well infected (Fig. 1), few parasitized foci in WT mice show a mononuclear-cell response; there is only a single developing granuloma in this representative field. (B) In contrast, IL-6^–/– mice show a response at virtually all infected foci, and granulomas, scored as developing to mature (arrow), are numerous. Original magnification, x400.

View larger version (34K):
[in this window]
[in a new window]

FIG. 3. Response to chemotherapy. One week after infection (day +7), WT mice (open bars) and IL-6^–/– mice (filled bars) received either a single i.p. injection of a suboptimal concentration of Sb (50 mg/kg) or i.p. injections of a suboptimal concentration of amphotericin B (AmB) (1 mg/kg on days +7, +9, and +11; total dose, 3 mg/kg). Other WT mice (hatched bars) received optimal doses of Sb (500 mg/kg once on day +7) or AmB (5 mg/kg on days +7, +9, and +11; total dose, 15 mg/kg). Liver parasite burdens were measured on day +14 and compared to day +7 burdens to determine the percent parasite killing. The results (means ± SEM) are for 9 to 10 mice per group from two experiments. *, P was <0.05 versus values for similarly treated WT mice.

The inflammatory response at liver foci parasitized by L. donovaniculminates in granuloma assembly (26, 31). In IL-6^–/–mice, mononuclear-cell influx with the encircling of infectedKupffer cells was rapid and considerably more prominent thanin WT mice (Fig. 2). Early on, at week 2, 91% ± 4% ofparasitized foci in IL-6^–/– animals had attractedsufficient mononuclear cells to be scored at least as developinggranulomas (31), yielding 502 ± 42 granulomas per 10040x microscopic fields versus 30% ± 5% and 158 ±24, respectively, in WT mice (three experiments; means ±standard errors of the means [SEM]; P < 0.05). In addition,at week 2, mature-appearing granulomas, the structural tissuecorrelate of the antileishmanial Th1 response (28), had alreadybegun to appear in IL-6^–/– mice (Fig. 2). By week4 (not shown), IL-6^–/– and WT mice demonstratedsimilar histologic reactions complete with numerous mature granulomas;however, in IL-6^–/– mice, 62% ± 7% of week4 granulomas were devoid of visible amastigotes versus 28% ±5% of those in WT mice (P < 0.05). By week 8 (not shown),most granulomas had disappeared or were involuting in both WTand IL-6^–/– mice, indicating no runaway inflammationas well as no requirement for IL-6 to terminate the tissue inflammatoryresponse.

In L. donovani-parasitized liver, IFN- ${gamma}$ and mononuclear-cellrecruitment also regulate the intracellular leishmanicidal effectsof conventional pentavalent-Sb chemotherapy (26, 27). In viewof the responses in IL-6^–/– animals noted above,infected mice were treated with Sb, with the expectation ofenhanced responsiveness. As shown in Fig. 3, low-dose Sb (50mg/kg) was threefold more active in IL-6^–/– micethan in WT mice, and liver parasite killing in IL-6^–/–animals approached that induced in WT mice injected with 10-fold-moredrug (500 mg/kg, optimal dose [31]). IL-6^–/– micewere also treated with amphotericin B. Although this agent'sleishmanicidal effect does not require host T cells, IFN- ${gamma}$ , orgranuloma formation, its efficacy is nonetheless enhanced byactivated endogenous Th1-type mechanisms (30). As shown in Fig.3, heightened responsiveness to chemotherapy in IL-6^–/–mice also extended to amphotericin B; low-dose therapy (3 mg/kg,total dose) was appreciably more active and produced the high-levelparasite killing achieved by optimal treatment in WT animalsgiven fivefold-more drug (15 mg/kg, total dose [30]).

Taken together, these results suggest that in L. donovani infectionof the liver, IL-6 exerts an early, suppressive effect on hostdefense and likely accomplishes this action by restraining Th1-cell-typeantileishmanial responses, including those dependent on IFN- ${gamma}$ .Since these observations were limited to the liver, differentresults might have been seen in spleen or bone marrow, organsalso targeted by L. donovani. In addition, under quite differentconditions (testing dendritic-cell transfer in L. donovani infectionof the spleen), IL-6 was not suppressive but instead mediatedthe leishmanistatic effect of this treatment approach (40).Therefore, under certain circumstances and within the same basicmodel, IL-6 appears to show different and perhaps organ-specificeffects, reemphasizing the complexity of its variable actions.While this analysis focused on IFN- ${gamma}$ , essential in virtuallyall aspects of defense against L. donovani, including macrophageactivation and responsiveness to chemotherapy (27, 32), furtherstudy is needed to fully address how IL-6 suppresses antileishmanialdefenses in vivo. IFN- ${gamma}$ , for example, is secreted by multiplecell types via more than one pathway, and although largely IL-12induced, it can also be provoked by other cytokines (32). IL-6may target one or more of these components in L. donovani-infectedhosts. Other potentially relevant effects reported for IL-6are its inhibition of Th1-cell differentiation and/or its promotionof suppressive Th2-type responses (13, 35) and its direct impairmentof macrophage responsiveness to the activating cytokines IFN- ${gamma}$ and tumor necrosis factor (5, 6, 16, 33). The last effect ofIL-6 extends to the inhibition of IFN- ${gamma}$ -responsive genes, includingchemokine and transcription factor gene expression (33).

As judged by results from cytokine- or cytokine receptor-deficientmice and/or WT mice treated with cytokine antagonists, IL-6(this report), IL-10, IL-27, IL-13, and transforming growthfactor β (TGF-β) (but not IL-4) have all been shownto promote experimental L. donovani infection of the liver (24,29, 31, 36). The suppressive effects of IL-13 and TGF-βare comparatively modest (31). However, IL-10, IL-6, and IL-27receptor signaling overtly influence L. donovani liver infection,and in their absence very similar tripartite phenotypes develop:enhanced Th1-type responses, accelerated tissue inflammation,and rapid parasite killing (29, 36). Reports that IL-6 participatesin IL-27 and IL-10 secretions and that IL-27 can induce IL-10(14, 21, 41) raise the possibility of a single, multistep pathwayto IL-10, thought previously to be the key suppressive cytokinein the L. donovani model (24, 29). However, IL-10 is expressedin both L. donovani-infected IL-6^–/– mice (40) andIL-27 receptor-deficient mice (36). The latter findings, then,leave open the possibility that IL-6, IL-27, and IL-10 may induceseparate deactivating effects, unless IL-6 and/or IL-27, actingalong with TGF-β (4, 21), enables or is required for thefull expression of IL-10's suppressive action.

Finally, it is worth pointing out that L. donovani-infectedC57BL/6 (as well as BALB/c) WT mice eventually develop a primarilyTh1-cell-type response which, via IL-12, IFN- ${gamma}$ , and tumor necrosisfactor (27, 32, 39, 43), comes to overshadow the early effectsof all of the preceding suppressively acting cytokines. In L.donovani-parasitized WT liver, the net result is acquired resistanceand near resolution of the infection (Fig. 1) (31, 32, 43).Nevertheless, even once under way, the expression of the self-healingphenotype can still be rebalanced and accelerated by immunointervention,for example, by injecting anti-IL-10 receptor antibody. In WTmice, such treatment markedly enhances IFN- ${gamma}$ secretion, granulomaassembly, and liver parasite killing and acts synergisticallywith chemotherapy (29). Preliminary data indicate that L. donovaniinfection induces IL-6 in liver tissue in C57BL/6 WT mice (semiquantitativePCR [31]; H. Murray, unpublished); thus, targeting endogenousIL-6 for inhibition may represent a new form of antileishmanialtherapy which might be used alone or with chemotherapy. To testthis possibility, future experiments with WT mice with establishedinfections will be directed at IL-6 receptor blockade, a nowclinically feasible treatment (7).

ACKNOWLEDGMENTS

This work was supported by NIH grant 2R56-AI16963-24A1.

FOOTNOTES

* Mailing address: 1300 York Avenue, New York, NY 10065. Phone: (212) 746-6330. Fax: (212) 746-6332. E-mail: hwmurray{at}med.cornell.edu

Published ahead of print on 23 June 2008.

Editor: W. A. Petri, Jr.

REFERENCES

Top