INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

A major advance in the therapy of any infectious disease would be the ability to not only cure the infection but also to confer resistance to reinfection. In the case of visceral leishmaniasis (VL), relapse after treatment may occur because of reinfection, if patients live in regions where VL is endemic, or because of multiplication of residual parasites which survive drug therapy.

Human immunodeficiency virus (HIV)-positive patients show high relapse rates after antileishmanial therapy (1, 3). In a retrospective nonrandomized open-trial study of secondary Leishmania prophylaxis in HIV-positive patients (22), annual relapse rates were 65% where there was no prophylaxis, 56% following allopurinol treatment (300 mg every 8 h), and 18% following monthly single antimonial injections (each equivalent to 850 mg of pentavalent antimony [Sb^v]). It is well known that a successful outcome of antimonial chemotherapy is dependent on an intact patient immune response (21), and since the immune response in HIV-positive patients can be discounted, the low relapse rate after antimonial prophylaxis is most likely due to the presence of drug depots at the sites of infection. However, given the short in vivo half-life of antimonials (12, 26), the prophylactic effect of single monthly antimonial injections is unexpected, although there is evidence that sodium stibogluconate (SSG) persists in tissues for prolonged periods. For example, in mice, prophylactic treatment with SSG (equivalent to 80 to 100 mg of Sb^v/kg of body weight) 6 days before infection with Leishmania donovani suppressed liver parasite burdens (11).

SSG entrapped in nonionic surfactant vesicles (NIV) is more effective than the free drug, and in BALB/c mice infected with L. donovani, treatment with a single dose of SSG-NIV gave >96% parasite suppression in the liver, spleen, and bone marrow (2). These treated mice displayed the immunological responses typical of a cured phenotype (15, 24), which indicated that SSG-NIV treatment had reversed the parasite-induced immunosuppression of VL. Other workers have used drug-abrogated infections to determine what effect limited exposure to the parasite has on immunity to a subsequent infection (4, 16).

The aim of the present study was to compare the abilities of prophylactic treatment with the free and NIV forms of SSG to protect against infection. In addition, since in VL host immunity influences treatment outcome, the effect of exposure to previous infection on any protective effect was investigated.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Materials. SSG (Pentostam) equivalent to 29.94% (wt/wt) Sb^V was obtained from the Wellcome Foundation, London, United Kingdom (UK). Mono-n-hexadecyl ether tetraethylene glycol was purchased from Chesham Chemicals Ltd., Harrow, UK. Antimony standards, dicetyl phosphate, and ash-free cholesterol were purchased from Sigma, Poole, UK. Fetal calf serum, RPMI medium, penicillin-streptomycin, and L-glutamine were purchased from Gibco BRL, Paisley, UK. All other reagents were of analytical grade.

Vesicle formation and characterization. A 150 µM concentration of surfactant-lipid, consisting of a 3:3:1 molar ratio of mono-n-hexadecyl ether tetraethylene glycol, cholesterol, and dicetyl phosphate, was melted by heating at 135°C for 2 min. The molten mixture was cooled to 70°C and hydrated with a preheated 5-ml volume of either phosphate-buffered saline (PBS) (pH 7.4) or 100-mg/ml SSG solution and homogenized with a Silverson mixer (model L4R SU; Silverson Machines, Chesham, UK) fitted with a 5/8" tubular work head (Silverson) and operated at 8,000 rpm for 15 min. Vesicle suspensions were sized with a Zetasizer 4 (Malvern Instruments Ltd., Malvern, UK).

Animals. In-house-bred Golden Syrian hamsters (Mesocricetus auratus) were used for parasite maintenance. Experimental studies used age-matched 8- to 10-week-old in-house-inbred female BALB/c mice or SCID mice with a BALB/c background (n = four or five). Maintenance and all manipulations on SCID mice were performed within an isolator.

Parasite preparation. L. donovani (strain MHOM/ET/67:LV82) was maintained by serial passage through hamsters as described by Carter et al. (9). To obtain a purified L. donovani amastigote preparation, the spleen of an infected hamster was removed aseptically and broken up in supplemented RPMI 1640 medium (100 µg each of penicillin and streptomycin/ml and 200 µM L-glutamine) by using a glass homogenizer. The resultant suspension was passed through a sieve to remove large debris and then pelleted by centrifugation. The pellet was resuspended in Boyle's solution (0.007 M ammonium chloride, 0.0085 M Tris [pH 7.2]) and incubated at 37°C for 10 to 20 min to lyse erythrocytes. The suspension was pelleted by centrifugation, washed twice, resuspended in 10 to 15 ml of medium, and then gently centrifuged at approximately 250 × g. The supernatant was pelleted by centrifugation and then resuspended in 10 to 15 ml of medium, and the number of amastigotes/milliliter was determined with a hemocytometer. Throughout, mice were infected by intravenous injection (tail vein, no anesthetic) of 0.5 × 10⁷ to 2 × 10⁷ L. donovani amastigotes.

In vivo efficacies of formulations. Uninfected mice were treated once intravenously with either 0.2 ml of free SSG solution (100 mg of SSG/ml) or 0.2 ml of SSG-NIV (100 mg of SSG/ml) at 4, 3, 2, or 1 week before infection. Controls were given 0.2 ml of PBS at each time point. All mice were sacrificed 14 days after infection.

Cell transfer experiments. Infected controls, infected SSG-NIV-treated mice (given a single dose of SSG-NIV [100 mg of SSG/ml] on day 7), and age-matched uninfected mice were sacrificed on day 30 or 50, and their spleens were removed and teased apart with forceps. Spleen cell suspensions from mice within the same group (n = four or five) were pooled, exposed to a nylon wool column to enrich the number of T cells present (8), and adjusted to 5 × 10⁷ cells/ml of RPMI 1640 medium. On day 0, uninfected female BALB/c mice (n = five) were given 0.2 ml of RPMI 1640 medium (controls) or 0.2 ml of T-cell suspension (10⁷ cells) prepared from one of the donor groups. On day 1 mice were infected; they were sacrificed at various times postinfection. For each mouse, plasma samples were prepared from blood collected at sacrifice and stored at 20°C until specific antibody titers were determined.

Determination of parasite burdens. Parasite burdens were determined in three sites (spleen, liver, and bone marrow) as described by Carter et al. (9). The number of Leishman-Donovan units (LDU) per organ was calculated for the liver and spleen by using the following formula: LDU = number of amastigotes per 1,000 host nuclei × the organ weight (in grams) (8). The effect of drug treatment on parasite burdens is expressed as the mean percent suppression in parasite numbers ± the standard error (SE), which was calculated by comparing each experimental value with the mean control value.

Specific antibody levels. Enzyme-linked immunosorbent assays were carried out to determine the end point titers of parasite-specific immunoglobulin G1 (IgG1) and IgG2a antibodies in the plasma of experimental mice by using the method described by Banduwardene et al. (2) and anti-mouse horseradish peroxidase conjugates at a 1:2,000 dilution.

Statistical analysis of data. The effect of drug treatment on parasite burdens was analyzed by using a one-way analysis of variance or a Student t test on the log transformed data (using LDU values for the spleen and liver data and the number of parasites/1,000 host cell nuclei for the bone marrow data). All other data were analyzed by the nonparametric Mann-Whitney U test.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Pretreatment with SSG gave no significant protection against infection with L. donovani (Fig. 1). However, pretreatment with SSG-NIV protected against infection in a time- and organ-dependent manner (Fig. 1). In mice treated with SSG-NIV 1 week before infection, significant parasite suppression was found in all three sites (spleen, P < 0.005 [Fig. 1a]; liver, P < 0.005 [Fig. 1b], bone marrow, P < 0.01 [Fig. 1c]; all P values are for comparisons to controls). If SSG-NIV treatment was given more than 2 weeks before infection, there was no significant suppression of spleen or bone marrow parasites compared to that in controls. Liver parasite burdens were still significantly lower in animals given SSG-NIV treatment 2 or 3 weeks before infection (P < 0.01). Treatment 4 weeks preinfection was not suppressive.

View larger version (14K): [in this window] [in a new window]   FIG. 1.   Effect of prophylactic treatment with free-SSG or SSG-NIV formulations on spleen (a), liver (b), and bone marrow (c) burdens of L. donovani-infected BALB/c mice. Uninfected mice were given a single 0.2-ml dose of SSG (100 mg of SSG/ml) or SSG-NIV (100 mg of SSG/ml) 4, 3, 2, or 1 week before infection. Controls were given 0.2 ml of PBS at each time point. On day 0 mice were infected with 1 × 107 to 2 × 107 L. donovani amastigotes; they were sacrificed 14 days later and the parasite burdens in the spleen, liver, and bone marrow were determined.

On the basis of these observations, in the experiments to determine whether drug treatment of an infection gave any protection against reinfection, reinfection was delayed until at least 30 days after drug treatment.

Compared to the results for primary controls (group A), treatment with SSG-NIV significantly suppressed parasite burdens in all three sites (group E) and was as effective on day 87 (mean reductions in parasite burdens ± SEs were as follows: spleen, 97% ± 2%; liver, 96% ± 3%; and bone marrow, 92% ± 4% [Table 1]) as it was on day 45 (mean reductions ± SEs were as follows: spleen, 95% ± 2%; liver, 98% ± 1%; and bone marrow, 99% ± 1% [Table 2]. Treatment with a similar dose of SSG solution (group D) suppressed only liver parasites, although this effect was apparent up to day 87 (Table 1 and 2).

Compared with results for secondary controls (group B), reinfection of infected mice (group C) did not increase parasite burdens in any of the three sites surveyed (Table 2).

Challenge infection of infected SSG-treated mice (group F) gave day 45 L. donovani liver burdens that were higher (P < 0.05) than in unchallenged SSG-treated mice (group D [Table 2]) but similar to those in secondary controls (group B) and mice given SSG prophylactically (group H [Table 2]). Spleen burdens of group F mice were significantly higher than those of secondary controls (group B [P < 0.0005]) and of mice given SSG prophylactically (group H [P < 0.005]) but similar to those of unchallenged SSG-treated mice (Table 2). Bone marrow parasite burdens among these groups were similar. On day 87, parasite burdens in all three sites were similar in unchallenged and challenged SSG-treated mice, mice given SSG prophylactically, and secondary controls (Table 1).

SSG-NIV treatment of infection gave no consistent protective effect against challenge infection. At day 45, liver (P < 0.05), spleen (P < 0.005), and bone marrow (P < 0.005) parasite burdens of challenged SSG-NIV-treated mice (group G) were higher than those of the unchallenged group (group E [Table 2]). Spleen and liver parasite numbers of infected SSG-NIV treated mice (group E) and secondary controls (group B) were similar (Table 2). However, bone marrow parasite burdens of challenged SSG-NIV treated mice (group G) were significantly lower (P < 0.05) than those of secondary controls (group B), and in similar experiments using different SSG-NIV preparations and different parasite challenge inocula, challenged SSG-NIV-treated mice had lower liver and/or bone marrow parasite burdens (data not shown). However, this protective effect could be explained by the presence of residual drug, since prophylactic SSG-NIV treatment also suppressed bone marrow parasite burdens compared to those of secondary controls (compare groups E and I in Tables 1 and 2). On day 87, parasite burdens of challenged SSG-NIV-treated mice (group G) were significantly higher than those of the unchallenged group in all three sites (for group E, the P value for the spleen and liver was < 0.005 and that for bone marrow was <0.0005 in comparison with results for group G [Table 1]) and not lower than those of secondary controls (group B) or mice given SSG-NIV prophylactically (group I).

Specific IgG1 and IgG2a titers of primary controls (group A) and infected SSG-treated mice (group D) were similar on days 45 (Fig. 2) and 87, whereas those of infected SSG-NIV-treated mice (group E) were lower, although the difference was not significant. In similar experiments, IgG1 and IgG2a titers of infected SSG-NIV-treated mice were significantly lower that those of primary controls and infected SSG-treated mice (e.g., day 7 postchallenge mean end point titers ± SEs were as follows: for primary controls titers of IgG1 were 303,125,000 ± 13,268,025, while for infected SSG-NIV-treated mice they were 181,000 ± 113,537; for primary controls IgG2a titers were 11,062,500 ± 3,010,938, while for infected SSG-NIV-treated mice they were 101,000 ± 23,939). Challenge of primary controls and of mice infected and then treated with SSG or SSG-NIV did not enhance IgG1 antibody titers compared to those in the corresponding unchallenged group on day 45 (compare groups A and C, D and F, and E and G [P < 0.02] in Fig. 2b), and only reinfected SSG-NIV-treated mice had higher IgG2a levels than the unchallenged group (Fig. 2a). Antibody titers of challenged mice and those given a primary infection were significantly higher than those of secondary controls (group B) and mice given SSG (group H) or SSG-NIV (group I) prophylactically (P < 0.02 [Fig. 2]). On day 87, specific IgG1, but not IgG2a, titers of challenged SSG-treated mice were higher than those in the corresponding unchallenged group (P < 0.02 [Fig. 2]). IgG1 and IgG2a antibody titers of SSG-NIV-treated mice were higher after challenge, but the difference from those in unchallenged mice was not significant.

View larger version (38K): [in this window] [in a new window]   FIG. 2.   Specific IgG2a (a) and IgG1 (b) titers of animals whose parasite burdens are shown in Tables 1 and 2.

Recipients of T-cell-enriched suspensions prepared from the spleens of infected SSG-NIV-treated mice were not protected against L. donovani infection, since their parasite burdens were similar to control values (data not shown). In addition, transfer of cells did not confer the ability to produce enhanced antibody levels, since specific IgG1 and IgG2a titers (mean ± SE) of L. donovani-infected recipients of control cells (IgG1, 40,960 ± 15,333; IgG2a, 14,080 ± 4,695) and cells from infected mice (IgG1, 64,000 ± 23,474; IgG2a, 21,760 ± 3,833) were not significantly different. Recipients of cell suspensions from infected medium treated mice became infected, whereas no L. donovani parasites were detected in mice given cells from infected SSG-NIV-treated mice (mean parasite burdens on day 14 posttransfer ± SEs, 41 ± 21 in the spleen, 7 ± 3 in the liver, and 9 ± 5 in bone marrow). Specific IgG1 and IgG2a titers of uninfected mice given cells from L. donovani-infected mice (IgG1, 9,280 ± 4,177; IgG2a, 1,060 ± 333) were significantly lower (P < 0.02) than those of the corresponding recipients infected with L. donovani (IgG1, 64,000 ± 23,474; IgG2a, 21,760 ± 3,833). Only parasite-specific IgG1 was detected in the plasma of uninfected mice given cells from infected SSG-NIV-treated mice (2,500 ± 1,330), and the titers were significantly lower than those of uninfected mice given cells from L. donovani-infected mice.

A fully competent immune system did not appear to be a prerequisite for effective SSG-NIV treatment since the treatment was as effective against liver, spleen, and bone marrow parasites in SCID BALB/c mice as in their immunocompetent counterparts (Table 3). Treatment with SSG solution suppressed only liver parasite burdens and was less effective in SCID mice (Table 3) than in immunocompetent mice.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

Prophylactic treatment with SSG-NIV was more effective than treatment with SSG. Assuming that in all three tissue sites examined the parasiticidal concentration of SSG is the same, then the three sites could be ranked liver > spleen > bone marrow based on the ability to retain parasiticidal drug concentrations. Previous studies (13, 14) have shown that after administration of SSG-NIV more of the drug dose is directed to the liver, and higher tissue levels of antimony are obtained, than when free SSG is given at the same dose. This is not surprising, since a function of this organ is to clear particulate material from the circulation (17). A similar manipulation of drug activity at the level of particular tissues (11, 18, 27) using a variety of colloidal formulations showed that interorgan differences are formulation sensitive. The ranking of the three sites established in this study, however, probably reflects both interorgan differences in initial drug levels (delivery) and subsequent excretion rates.

It seems that drug uptake and persistence in certain tissue locations, rather than an immune effect, could explain the protection observed in response to challenge of infected SSG-NIV-treated mice. The fact that the level of protection was similar to that obtained by prophylactic SSG-NIV treatment supports this hypothesis. The greater prophylactic activity of SSG-NIV treatment than of free-SSG treatment reflects the former's ability to direct a large proportion of the injected drug dose to tissues (14). The high efficacy of SSG-NIV depends on the quantity of drug entrapped (27); interexperimental variability could reflect NIV preparation-dependent drug entrapment efficiency, which in turn influences delivery to different tissues.

Protective immunity was not transferred with spleen cell suspensions prepared from SSG-NIV-treated mice, since on infection, cell recipients had parasite burdens which were similar to those of controls. Failure to transfer immunity may be a consequence of the innate susceptibility of BALB/c mice to L. donovani infection (5-7) or of the limited exposure of cell donors to L. donovani infection before drug treatment.

It has been suggested that L. donovani-infected mice are resistant to reinfection (19, 20, 25). The results of this study confirmed that challenge of L. donovani-infected mice does not result in higher parasite burdens than those obtained in unchallenged animals. However, challenge of infected mice which had been treated with either free SSG or SSG-NIV raised parasite burdens in the liver (day 45 in free-SSG-treated mice and days 45 and 87 in SSG-NIV-treated mice), spleen (days 45 and 87 in SSG-NIV-treated mice), and bone marrow (day 87 in SSG-NIV-treated mice) compared to levels in unchallenged mice. It may be no coincidence that the raised burdens occurred in sites where parasite numbers had been lowered by drug treatment. The presence of an upper limit on parasite load in chronically infected mice would explain why parasite burdens of primary and secondary controls were similar by day 87. Surprisingly, challenge of L. donovani-infected animals with a second infection did not result in enhanced IgG1 or IgG2a antibody titers compared to those of primary controls. Perhaps the high antibody titers of primary controls (>1:100,000) meant that on challenge the mice could not produce any more pathogen-specific antibody since the total specific B-cell population had been stimulated. This could explain why by day 87 antibody levels in primary and secondary controls were similar.

Stern et al. (25) found that in nude mice it was possible to transfer immunity against L. donovani with unfractionated T cells from euthymic L. donovani-infected mice. The failure to transfer immunity by using cells from infected mice in this study may be due to differences in experimental protocols. In this study, cells transferred to euthymic murine recipients were collected earlier postinfection (after 30 to 50 days instead of 16 to 24 weeks), and parasite burdens in recipients were determined earlier postinfection (day 14 instead of week 4 or 8). The ability to produce specific antibody was, however, transferred with the spleen cells, since specific IgG1 (cells from both infected controls and infected SSG-NIV-treated mice) and IgG2a (cells from infected controls only) was detected in the plasma of uninfected recipients. Previous studies have shown that spleen cells from SSG-NIV-cured, L. donovani-infected mice respond to specific stimulation in vitro (2) by day 24 posttreatment (day 31 postinfection), which suggests that potentially protective memory lymphocytes should have been present.

At the same dose, the superiority of vesicular over free SSG was clearly demonstrated in this study. It elicited greater parasite suppression in the spleen, liver, and bone marrow, and free SSG did not have the prophylactic activity of SSG-NIV, which was as effective in immunocompetent and immunocompromised animals. The activity of SSG-NIV in the immunocompromised host may give this formation a significant advantage over currently available antileishmanial formulations, since high relapse rates occur in AIDS patients with VL after treatment with antimonial (1) or amphotericin B (3, 23) drug formulations. Studies to develop an SSG-NIV formulation for possible clinical use are under way.