Protective T Cell Immunity against Malaria Liver Stage after Vaccination with Live Sporozoites under Chloroquine Treatment

Mice and parasites

BALB/c female mice were purchased from Harlan Laboratories (Gannat, France). WT and IFN-γRα knockout (KO) 129P2Sv/ev mice (a gift from Dr. M. Aguet, Swiss Institute for Expermental Cancer Research, Epalinges, Switzerland) (13) were bred in specific pathogen-free animal facilities. Mice used in each experiment were matched for age and sex. All experiments and procedures were performed in compliance with French Ministry of Agriculture regulations for animal experimentation (1987). Spz of the uncloned line of the 265BY strain of P. yoelii yoelii were obtained by dissection of the salivary glands of infected Anopheles stephensi mosquitoes 16–21 days after an infected blood meal. This parasite species was initially isolated from a naturally infected Thamnomys gazellae captured in the wild in Central Africa Republic by Landau and colleagues (14), and the 265 BY line has been maintained by cyclical passage between Swiss mice and Anopheles mosquitoes. The broad characteristics of the infection do not differ substantially from those of the frequently used 17X strain-derived lines, except that that the minimum numbers of sporozoites needed to obtain patent blood stage infection is higher (1000 instead of 10). In one experiment, spz irradiated at 12,000 rad were used. After this irradiation dose the spz are capable of invading hepatocytes, where development is arrested at the uninucleated trophozoite stage (15). Infected RBC (iRBC) were obtained from P. yoelii-infected mice on days 8–12 after spz injection, using heparin (Sigma-Chemie, Saint Quentin-l’Arbresles, France) as anticoagulant. These infected blood samples were then diluted 5–10 times in PBS and washed twice in PBS. The percentage of parasitemia was determined by counting the number of iRBC per 1000 RBC on Giemsa-stained blood smears, and the absolute number of iRBC was calculated using total RBC number determined with a Malassez chamber.

Immunization schedule

Mice were injected once or twice with either one or two doses of 4,000–20,000 spz. For the double immunization, spz were administered at least 10 days apart. Chloroquine hydrochloride (CQ; Sigma-Chemie) was diluted in PBS to a concentration of 8 mg/ml. Starting the same day as spz inoculation, immunized and control animals were injected i.p. with 100 μl of the CQ solution for 10 consecutive days. Mice immunized with two doses of spz received an additional 10-day regimen of CQ treatment after the second spz injection. During (every 2 days) and at the end of the CQ treatment, the absence of parasitemia in the immunized mice was confirmed by examination of Giemsa-stained slides of tail blood. The CQ regimen used for immunization derived from a previous report where it was shown that normal sporozoite injection(s) of P. berghei under this CQ treatment was able to confer protection against a live spz challenge (10). In one experiment, mice immunized with 20,000 live spz were treated with a single i.p. dose of 0.3 mg of primaquine diphosphate (PQ; Sigma-Chemie) on the day of immunization and with a daily dose of 0.8 mg of CQ for 10 days starting on the day of immunization. This PQ treatment was shown to eliminate >90% of LS parasites (data not shown). In another set of experiments, groups of mice were immunized with one dose of 25,000 irr.spz, two doses of 25,000 irr.spz 15 days apart, or one dose of 75,000 followed by two booster doses of 25,000 spz, 15 and 7 days apart.

Challenge and protection assessment

Mice immunized with live spz and treated with CQ were rested for a minimum of 15 days after the last injection of CQ to allow complete elimination of the drug. The mice were then challenged i.v. with 4000 spz or 10⁵ or 10⁶ iRBC. In one experiment, mice were challenged with the bites of four or five P. yoelii-infected mosquitoes. The presence of spz in the mosquitoes that had fed was verified by microscopic observation of their dissected salivary glands. Control mice and mice immunized with irr.spz were challenged i.v. with 4000 spz. Blood stage infection was determined by the presence of parasites in Giemsa-stained blood smears prepared daily from days 4–10 postchallenge, and parasitemia was determined by counting the number of iRBC per 1000 erythrocytes.

Detection of parasites in the blood and quantification of parasite load in the liver of animals were performed according to the method described by Hulier et al. (16) with minor modifications. For detection of parasite in the blood, 100 μl of blood was mixed with 0.5 ml of digestion buffer (100 mM NaCl, 10 mM Tris-HCl, 25 mM EDTA, and 0.5% SDS). For LS parasite quantification, immunized mice were challenged with 30,000 spz, and 42 h later a liver biopsy was taken, frozen in liquid nitrogen, and crushed in 2 ml of digestion buffer before incubation overnight at 37°C in the presence of proteinase K (0.1 mg/ml; Roche, Paris, France). DNA used for PCR was purified by phenol/chloroform extraction. A first amplification of a β-actin gene fragment was used to verify whether all the samples contained similar amounts of mouse DNA. The DNA was then used as a template in duplicate PCR reactions using oligonucleotides specific for the small subunit ribosomal RNA gene of the parasite. A standard curve corresponding to DNA extracted from equal quantity of erythrocytes in samples containing 10-fold dilutions of a known number of P. yoelii parasite nuclei was made. The PCR products were analyzed on a 2% agarose gel and quantified by scanning densitometry. Liver parasite load units correspond to the log number of parasite nuclei per milligram of liver DNA.

Culture of malaria LSs

BALB/c hepatocytes were prepared as previously described, with minor modifications (17). Cells were isolated by collagenase perfusion (Roche) of liver fragments and were further purified over a 60% Percoll gradient (Pharmacia Biotech, Uppsala, Sweden). Hepatocyte purity and viability were >95% as assessed by trypan blue dye exclusion. Cells (60,000) were cultured in eight-chamber plastic Lab-Tek slides (Nunc, Naperville, IL) in William’s E medium (Life Technologies, Edinburgh, Scotland) supplemented with 5% FCS (Life Technologies) and 100 IU/ml penicillin-streptomycin (Life Technologies) and incubated at 37°C in 3.5% CO₂ for 24 h. After removal of medium from the culture chambers, 40,000 spz were added in 100 μl of fresh supplemented medium. Three and 24 h later, the medium was replaced by fresh complete medium. Cultures were stopped 48 h after spz infection, fixed with cold methanol, and processed for Ab titration.

Ab titration assays

Sera from immunized mice were collected 15 days after the last CQ injection. Ab titers were determined by indirect immunofluorescence assay (IFAT) using methanol-fixed sporozoites, LS, or iRBC as previously described (17).

Effects of sera on parasite development in vitro and in vivo

Sera were added to the hepatocyte cultures at a dilution of 1/10 at the time of spz addition. Cultures were then processed as described above. Schizont numbers were assessed in triplicate cultures by IFAT using hyperimmune sera recognizing P. yoelii LS. The percent inhibition was calculated by comparing the numbers of parasites in the experimental cultures with those in control wells.

Spz neutralization assay was performed by incubating 40,000 spz in a 100-μl volume of PBS containing the test sera at a final dilution of 1/10 for 15 min at 37°C. This solution was further diluted in PBS to obtain a final concentration of 40,000 spz/ml. Naive BALB/c mice were then challenged with 100 μl of this solution, corresponding to 4,000 spz. Blood stage patency was monitored during the 10 days after challenge.

Sera (200 μl) obtained from mice immunized with 20,000 spz under CQ or from control mice treated with CQ alone were passively transferred twice to groups of five recipient BALB/c mice 40 and 60 h after challenge with 4,000 spz. Parasitemia was then monitored.

In vivo depletions

Purified control rat Abs were purchased from Sigma-Chemie. Rat IgG2a anti-CD4 mAb (clone GK1.5, ATCC TIB 207; American Type Culture Collection, Manassas, VA) and rat IgG2b anti-CD8 mAb (clone 2.43, ATCC TIB 210; American Type Culture Collection) were used. The rat IgG2a anti-IL-12 mAb (clone C17.8.20) cell line was provided by Dr. Trinchieri (Wistar Institute, Philadelphia, PA) (18). The rat IgG1 anti-IL-6 mAb (clone MP5-20F3), the rat IgG2a anti-TNF-α (clone XT22), and the rat IgG1 anti-IFN-γ (clone XMG-1) (19) cell lines were supplied by Dr. P. Minoprio (Institut Pasteur, Paris, France). Abs were purified from culture supernatant by ammonium sulfate precipitation. Immunized mice were injected i.p. with 500 μg of anti-CD8, anti-CD4 mAbs, or control rat IgG on days 1 and 0 before challenge. More than 98% of blood CD8⁺ or CD4⁺ T cells were depleted by this procedure, as verified by cytofluorometry (FACScan; BD Biosciences, Mountain View, CA) using anti-CD4 (clone H129-19; Sigma-Chemie) and anti-CD8 (clone 53-6.7; BD PharMingen, San Diego, CA) mAbs that recognized epitopes different from those recognized by the depleting mAbs.

IFN-γ neutralization

On days 2, 1, and 0 before and days 1 and 2 after challenge, mice received a single i.p. dose of 2 mg of the anti-IFN-γ mAb XMG-1. The efficacy of the IFN-γ Abs was verified in an experiment in which anti-IFN-γ mAb treatment was able to reverse protection in mice immunized with three injections of irr.spz. In the anti-IFN-γ mAb-treated group, 80% (four of five) of the mice developed patent parasitemia compared with none (of five) in the control group.

IL-12 neutralization

At 12 h before and 3 h after challenge, mice received a single i.p. dose of 1 mg of the anti-IL-12 mAb. Efficacy was verified by testing the ability of this mAb (1 mg i.p.) to prevent mortality in mice treated with an i.v. injection of 100 μg of LPS (18). In the anti-IL-12 mAb-treated group, no mouse died (zero of five) compared with the control group, in which 60% of the mice died (three of five).

TNF-α neutralization

At the time of challenge, mice received a single i.p. dose of 1 mg of the anti-TNF-α mAb. Neutralization efficacy was verified by testing the ability of this mAb (1 mg i.p.) to prevent mortality in mice treated with an i.p. injection of 0.1 μg of LPS together with 20 mg of galactosamine (20). In the anti-TNF-α mAb-treated group, no mouse died (zero of five), whereas in the control group three of five mice died.

Inhibition of NO production

On days 0, 1, and 2 after spz challenge, mice received a single i.p. dose of 100 μg of S-methyl-thiourea (SMT; Sigma-Chemie) in 0.1 ml of PBS, a potent inhibitor of inducible NO synthase (21).

Inhibition of free radicals

On days 0, 1, and 2, mice received a single i.p. dose of 100 μg of N-acetyl-cysteine (NAC) (22).

ELISA detection of IFN-γ

Ninety-six-well plates (Nunc, Naperville, IL) coated with anti-IFN-γ (clone R4-6A2; BD PharMingen) mAb were incubated overnight at 4°C, then the sera were added, and the plates were incubated overnight at 4°C. Detection was made after adding second biotinylated Ab (clone XMG1.2; BD PharMingen), followed by incubation for 45 min at room temperature. Extravidin-phosphatase alkaline was then added, and plates were incubated for 1 h at room temperature. The phosphatase activity was measured by adding 4-methylumbelliferyl phosphate (Sigma-Chemie) as substrate and reading fluorescence at 360/460 nm using a spectrophotometer. Dilutions of known quantities of IFN-γ were used for the standard curve. The detection limit was 50 pg/ml.

Statistical analysis

To determine differences in liver parasite load between two groups or between multiple groups of mice, the Mann-Whitney U test or one-way ANOVA followed by Tukey’s multiple comparison tests were used, respectively. Calculations were performed using PRISM software version 3.02 (GraphPad, San Diego, CA).

CQ treatment does not inhibit LS development

CQ, a potent inhibitor of blood stage parasites (23), does not prevent the emergence of blood stage parasites when applied only during the PE stages (24, 25). It is thus considered to have no effect on LS parasites, although no formal demonstration of this was obtained. We therefore first confirmed that CQ treatment during the 45-h LS (on the day on spz inoculation and the next day) does not inhibit P. yoelii LS development in vivo after a spz challenge of normal mice (Fig. 1⇓). As expected, mice (n = 20) treated for 10 days with CQ starting on the day of sporozoite inoculation did not develop patent parasitemia, whereas all nontreated control mice (n = 20) did.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Chloroquine does not inhibit parasite development in the liver. Naive mice were infected with 30,000 live sporozoites, and in one group (n = 5) mice were treated with 0.8 mg of CQ on days 0 and 1 after injection, whereas control group mice (n = 5) received PBS alone. The amount of parasite ribosomal DNA in the liver was measured by quantitative PCR 42 h postinfection. Results are expressed as the mean parasite liver load ± SEM of five mice. The difference between the two groups is not statistically significant.

Protective immunity is conferred by spz immunization under CQ in BALB/c mice

We next assessed whether immunization with infective spz in the presence of CQ (henceforth abbreviated spz/CQ) was able to induce a protective immune response against an spz challenge made after CQ had been completely eliminated from the blood. Different regimens of spz/CQ (dose and frequency) were performed in BALB/c mice (Table I⇓). For normal spz, the acquisition of strong protective immunity was observed when 10,000 or more spz were used (as further confirmed by a negative parasite-specific PCR on blood samples taken 6 days after challenge). By contrast, even inoculation with two doses of 25,000 irr.spz failed to protect against spz challenge; sterile immunity was only obtained at substantially higher irr.spz doses (Table I⇓). For both types of immunization, protection did not wane substantially with time, as determined by rechallenge 2–4 mo after the first challenge (Table I⇓ and data not shown). Furthermore, the protection conferred by spz/CQ immunization of BALB/c mice was equally effective when the challenge was performed naturally, i.e., by the bite of infected mosquitoes. None of the mice (zero of five) immunized with 20,000 spz/CQ developed patent parasitemia after challenge by four or five P. yoelii-infected mosquitoes, whereas all control mice (five of five) receiving CQ alone did.

Development of sporozoite into LS parasite is needed for protection

When PQ, an antimalarial drug that eliminates LS parasites (26), was administered concurrently with spz/CQ immunization, none of the BALB/c mice was protected from a subsequent spz challenge (Table II⇓). The abrogation of protection by PQ treatment demonstrated that LS parasites are essential for the induction of protective immunity.

Protective immunity conferred by spz/CQ immunization in BALB/c mice is mainly directed against LS parasites

There are three possible targets for the LS-dependent protection elicited by spz/CQ immunization: the invading spz, the intracellular LS itself, and the blood stage parasites. Analysis of the Abs induced by our immunization protocol did not allow us to distinguish between these possibilities. Indeed, sera from the immunized animals reacted by IFAT with spz, LS, and blood stage parasites, and titers were correlated with the immunizing spz dose (Table III⇓). This contrasts with the IFAT reactivity of sera obtained from irr.spz-immunized mice, in which spz and, to a lesser extent, LS parasites were labeled, but not blood stage parasites.

The sera from spz/CQ-immunized mice were then tested in three different functional assays. A strong inhibition of spz invasion and development in hepatocytes in vitro was obtained with sera from irr.spz-immunized mice; however, sera from the spz/CQ-immunized mice failed to do so; on the contrary, they produced an enhancement of spz invasion and development in some groups (Table III⇑). The lack of an Ab effect on spz was confirmed by the spz infectivity neutralization in vivo assay, as sera from all spz/CQ-immunized mice failed to prevent infection (Table III⇑). Finally, passive transfer of sera from mice immunized with 20,000 spz/CQ or mice receiving CQ alone to groups of five mice challenged with 4,000 spz at the end of LS development (40 h post-spz challenge) and at the beginning of blood stage development (60-h spz challenge) had no effect on the course of the subsequent blood stage infection.

The major consequence of spz/CQ immunization of BALB/c mice was observed when the hepatic parasite load was quantified by PCR after challenge with 30,000 spz (Fig. 2⇓A). The reduction in LS parasite load increased with the immunizing dose and reached 99% for the most stringent immunization schedule (two injections of 20,000 spz), nearly equating the total elimination of LS parasites obtained in mice immunized with three injections of irr.spz (Fig. 2A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Vaccination with live sporozoites followed by CQ treatment inhibits parasite development in the liver. A, Groups of BALB/c mice were either immunized with different doses of live spz, followed by a 10-day CQ treatment starting the day of the spz inoculation or were treated with CQ alone. Other groups were immunized with different doses of irradiated spz as described in Materials and Methods. They were challenged with 30,000 spz at least 25 days after the last immunizing dose of spz, and parasite development in the liver was quantified as mentioned above. Numbers in brackets represent the percentage of LS inhibition. BLD, below the level of detection. ∗, p < 0.05; ∗∗, p < 0.01 (vs control mice, by ANOVA, followed by Tukey’s test). Blood stage parasitemia of immunized animals challenged with P. yoelii sporozoites after a spz (B) or a blood stage (C) challenge is shown. Groups of BALB/c mice were either controls or immunized once with an i.v. injection of 20,000 live spz. Both groups were treated with CQ for 10 days starting the day of spz inoculation. Mice were challenged with 4,000 spz or 10⁶ iRBC 25 days after injection of the immunizing dose of spz. For each animal, parasitemia was monitored by examination of blood films stained with Giemsa reagent. Only points at which parasitemia was patent are plotted. The number of mice with patent parasitemia relative to the total number of mice in the group is shown in the top right corner of each panel.

A minor effect on blood stages of spz immunization during CQ treatment could additionally be inferred. In groups of mice immunized with 20,000 spz/CQ, LS parasite load reduction was not complete (48–77%; Fig. 2⇑A). However, in groups immunized in the same way, only a few (10%) of the spz-challenged mice developed patent parasitemia (Table I⇑). Two sets of experiments were conducted to explore the occurrence of blood stage immunity. Groups of five BALB/c mice immunized with 20,000 live spz/CQ were challenged with 4,000 spz or with two different doses of iRBC. Sterile protection was observed in four of five spz-challenged mice, and only one of five developed parasitemia that peaked at lower levels and cleared faster than controls (Fig. 2⇑B). Sterile protection was observed only in one of five or two of five spz/CQ-immunized mice after challenge with 10⁶ or 10⁵ iRBC, respectively (Table IV⇓). The course of infection in the mice challenged with 10⁶ iRBC was monitored for 3 wk postchallenge. Peak parasitemias and time to clearance were only lower in some of the immunized mice (Fig. 2⇑C). These observations indicate that a minor anti-blood stage component operates after the major elimination of LS parasites resulting from spz/CQ immunization.

Protective immunity conferred by spz/CQ immunization can be induced in mice with a different genetic background

129P2Sv/ev mice have a different genetic background from BALB/c mice and express a different MHC haplotype, H2^b rather than H2^d. Preliminary experiments established that 129P2Sv/ev mice are less receptive to spz from P. yoelii 265 BY, as patent blood parasitemia could only be obtained reproducibly after challenge with a minimum of 7,500 spz (n = 15 129P2Sv/ev mice/group in three experiments) compared with 4,000 spz for BALB/c mice. A strong protection (60–100%) was induced against spz challenge by immunization with two injections of 10,000 or 20,000 spz/CQ or one injection of 20,000 spz/CQ (Table V⇓). This protection was also characterized by a statistically significant reduction in the LS parasite load (Fig. 3⇓C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Protection induced by vaccination with live sporozoites followed by CQ treatment is mediated by T cells and IFN-γ. A, Groups of BALB/c mice were injected with 20,000 spz and treated with CQ from days 0–10 starting the day of spz inoculation. They were challenged i.v. with 4,000 spz at least 15 days after the last CQ injection. Immunized mice were injected with the following rat Abs: control IgG, anti-CD8, or anti-CD4 on days 1 and 0 before challenge. Data are cumulative results of two experiments with four or five mice per experiment. All naive mice (n = 9) developed patent blood stage parasitemia after spz challenge. Protection is defined as the number of mice that did not develop patent blood stage over the number of spz-injected animals. On day 10, as calculated by Fisher’s exact test: ∗, p = 0.009 between immunized mice treated with rat IgG and immunized mice treated with anti-CD8 Abs; ∗, p = 0.029 between immunized mice treated with rat IgG and immunized mice treated with anti-CD4 Abs. Numbers in brackets represent the percentage of protection. B, Groups of BALB/c mice were either immunized with 20,000 live spz, followed by a 10-day CQ treatment starting the day of the spz inoculation. Control groups received CQ alone. Immunized mice were injected with the control IgG, anti-CD8, or anti-CD4 Abs as described above before challenge. Mice were challenged with 35,000 live spz after the injection of the immunizing dose of spz, and parasite development in the liver was quantified as in Fig. 1⇑. Results are expressed as the mean ± SEM of five mice. ∗, p < 0.05 vs control mice (by Tukey’s test). Results are expressed as the mean liver parasite load units ± SEM of five mice. Numbers in brackets represent the percentage of LS inhibition. C, Concentrations of IFN-γ in the sera of control mice or mice immunized once with an i.v. injection of 20,000 live spz and treated with CQ for 10 days starting the day of spz inoculation. Sera were collected 42 h after challenge with 35,000 spz. Results are expressed as the mean ± SEM of five mice. ∗, p < 0.01 vs control mice (by Mann-Whitney U test). D, Groups of wild-type (WT) and IFN-γR KO 129P2Sv/ev mice were either controls or immunized once with an i.v. injection of 20,000 live spz. All groups were treated with CQ for 10 days starting on the day of spz inoculation. Mice were challenged with 35,000 spz at least 25 days after the immunizing dose of spz, and parasite development in the liver was quantified as described in Fig. 1⇑. Results are expressed as the mean liver parasite load units ± SEM of five mice. ∗, p < 0.05 vs control mice (by Tukey’s test). Representative results from two independent experiments are shown.

CD4⁺ and CD8⁺ T cells are involved in protection induced by spz/CQ immunization against LS parasites in BALB/c mice

Having established that Abs did not seem to be implicated in the protection induced by spz/CQ immunization against LS parasites, we evaluated the role of CD4⁺ and CD8⁺ T cells in protection. BALB/c mice immunized with 20,000 spz/CQ were then treated with depleting anti-CD4 or anti-CD8 Abs before spz challenge. A partial reversal of protection against infection was obtained by depletion of CD8⁺ or CD4⁺ T cells (Fig. 3⇑A). This reversal was mirrored by a reduced in vivo elimination of LS parasites in the immunized depleted mice (Fig. 3⇑B).

Role of IFN-γ in the protection induced by spz/CQ immunization against LS parasites

Different cytokines have been shown to inhibit LS development in vivo and in vitro (27). The roles of these cytokines were tested by Ab neutralization just before and during spz challenge of BALB/c mice immunized with 20,000 spz/CQ. No reversal of protection could be observed when neutralizing Abs against IL-6, IL-12, and TNF-α were used, whereas injection of anti-IFN-γ Abs partially reversed the induced protection (Table VI⇓). The importance of IFN-γ was underlined by the presence of high levels of this cytokine in the serum of immunized mice, but not in that of spz-challenged naive mice (Fig. 3⇑C).

To confirm these results, 129P2Sv/ev mice deficient (KO) for the α-chain of the IFN-γR were used. A significant abrogation of protection was observed in the KO mice (Table V⇑), and this was associated with a reversal of a reduction in the LS parasite load (Fig. 3⇑D).

Protection is dependent on NO in BALB/c mice

Finally, we evaluated the role of toxic mediators in the protection induced against LS parasites. Treatment of spz-immunized mice under CQ with SMT, a potent and selective inhibitor of inducible NO synthase (21), at the time and 1 day after spz challenge completely abrogated protection (Table VI⇑). By contrast, treatment with NAC, a potent inhibitor of oxygen radicals (22), had no effect on protection (Table VI⇑).