Schistosomal egg antigen‐responsive CD8 T‐cell population in Schistosoma mansoni‐infected BALB/c mice

Abstract

We demonstrated here that schistosomal egg antigen (SEA) is able to stimulate an antigen‐specific, cytotoxic CD8⁺ T‐cell response in mice. Indeed, a single i.p. immunization with SEA resulted in the in vivo induction of significant cytotoxic T lymphocyte (CTL) activity in the spleen within 20 days. Effector cells were classic class I major histocompatibility complex (MHC)‐restricted CD8⁺ lymphocytes producing interferon‐γ (IFN‐γ) and interleukin‐2 (IL‐2), suggesting a type 1 response to SEA. We therefore investigated the relevance of these observations in the context of the Schistosoma mansoni parasite infection. CTL activity against SEA‐pulsed target cells was evidenced throughout the infection after in vitro stimulation of recovered splenic cells with SEA demonstrating that SEA‐specific CD8⁺ T cells with cytotoxic potentialities are present during infection. This activity was strongly increased after immunization of mice with SEA like the production of IFN‐γ in the sera. A marked reduction in the number of granulomas and of fibrosis with the presence of cells producing IFN‐γ in the liver was also observed leading to the survival of SEA‐immunized mice.

Introduction

Schistosoma mansoni infection in mice and man commonly results in a stable, long‐term disease state.¹ In both species, the disease is associated with the continual, daily production of viable eggs by intravascular worms. Many of the eggs do not reach the external environment and cause an inflammatory response when trapped in liver sinusoids, leading to extensive cell‐mediated granuloma formation and ultimately hepatic fibrosis.² This fibrosis involves the excessive deposition of newly synthesized connective tissue matrix, with collagen being the major component.³ With the help of experimental models it has been clearly established that granuloma formation is a manifestation of CD4⁺ cell‐mediated immunity directed against soluble schistosomal egg antigen (SEA) which is at a maximum 8 weeks after infection.^4,5 Granulomas are then gradually down‐regulated, largely by CD8⁺ lymphocytes,^6–8 so that by 16–20 weeks after infection these newly formed granulomas have decreased in size. The presence of CD8⁺ T cells was observed in both early and chronic murine granulomas with an increased ratio of CD8⁺ cells in the chronic phase.⁹ Adoptive transfer of splenic and lymph node cells from chronically infected mice into acutely infected mice decreased the size of nascent granulomas.¹⁰ Depletion of CD8⁺ T cells abolishes this inhibitory effect.^6,11 Furthermore, experiments with cell populations isolated from granulomas indicate that CD8⁺ T cells are required for the in vitro suppression of SEA‐induced proliferation as well as the in vivo reduction of lung granuloma size in adoptively transferred recipients.^9,11 The secretion of T helper 2 (Th2)‐related cytokines (interleukin (IL)‐4, IL‐5 and IL‐10)¹² that follows egg deposition, is vigorous 8 weeks after infection, whereas Th1‐related responses (IL‐2 and interferon‐γ (IFN‐γ)) that could limit worm survival are suppressed.¹³ Later in infection, Th2 responses are themselves down‐regulated.¹⁴ Recent studies have shown that, like naive CD4⁺ cells, precursor CD8⁺ cells can be influenced by cytokines to differentiate into either type 1 cells, which make IFN‐γ, IL‐2 and in some reports^15,16 IL‐10, or type 2 cells, which make IL‐4, IL‐5 and IL‐6·^16,17 Type 1 development is promoted by IFN‐γ and IL‐12, whilst IL‐4 provides a strong differentiation signal for CD8⁺ cells to begin secreting type 2 cytokines.^15–17 Recent work indicates that CD8⁺ T cells from schistosome‐infected mice display a type 1 cytokine profile characterized by the selective production of IFN‐γ and IL‐10·¹⁸ These type 1 CD8⁺ cells are able to persist in a strongly type 2 environment and use IL‐4 as a helper cytokine, indicating that they may be important modulators of established Th2 responses. For example, IFN‐γ is known to be a potent inhibitor of Th2 proliferation and in some cases has been shown to cause the suppressive activity of CD8 clones on Th2 responses.^19,20 Endogenous IFN‐γ has been observed to have a suppressive effect on granuloma formation in both an in vitro system²¹ and in an in vivo pulmonary egg injection model.²² Moreover, exogenous IFN‐γ has been shown to down‐regulate pulmonary granuloma size²³ and hepatic fibrosis.²⁴ Both type 1 and 2 CD8⁺ cells can be cytotoxic leading Mossmann and colleagues to term them Tc1 and Tc2.¹⁵ Because CD8⁺ cells are viewed primarily as mediators of cytotoxicity against pathogen infected cells, it has been difficult to see how they might function during infection with a large metazoan parasite. However, the recent increase in understanding of how exogenous proteins can be taken up and presented through the major histocompatibility complex (MHC) class I pathway²⁵ has made it much easier to envisage a role for CD8⁺ cells in responses to extracellular pathogens. Cytotoxic T lymphocytes (CTL) with specificity for alloantigens were found to adhere specifically to S. mansoni larvae which bear alloantigens on their surface as a result of passive acquisition from the host, but failed to damage the larvae even after prolonged periods of culture at high effector to target ratio.²⁶ However, we have recently observed that a promising vaccine candidate, the 28 000 MW glutathione S‐transferase (Sm28GST) of S. mansoni,²⁷ is able to stimulate an antigen‐specific, cytotoxic T‐cell response in infected animals and that effector cells induced in vivo were classic class I MHC‐restricted CD8⁺ lymphocytes.²⁸ Immunization with this molecule also induced a reduction in the worm burden and hepatic damage, and consequently the survival of immunized mice.^29,30 We have demonstrated a role for IFN‐γ and for the specific CD8⁺ T cells in the expression of Sm28GST‐mediated protection.³⁰ Here we attempt to characterize the CD8⁺ T‐cell population responsive to SEA, previously defined as an inducer of CD4⁺‐mediated response, after immunization with this antigen and/or during the course of the infection. We also evaluate the consequences of the immunization with SEA on the development of infection.

Materials and methods

Host animals and parasites

Female BALB/c mice (H‐2^d), 6–8 weeks old at the beginning of the experiment, were provided by Iffa Credo, L'Arbresle, France. A Puerto Rican strain of S. mansoni was maintained in Biomphalaria glabrata snails as intermediate hosts and golden hamsters as definitive hosts. Cercariae for experimental infections were used within 1 hr after collection from 1 month‐infected snails exposed to light and temperature of 30°.

Antigens

SEA was prepared from homogenized eggs (isolated from the livers of mice infected for 8 weeks) using a previously published procedure,²¹ sterile‐filtered through 0·45 µm filter (Sartorius, Göttingen, Germany), assayed for protein content and stored at –20° until use. For cyanogen bromide (CNBr) cleavage, 10 mg SEA was reacted overnight with 20 mg of CNBr dissolved in 70% formic acid. The solvent was evaporated under a stream of nitrogen and the residue resuspended in water and lyophilized. No intact SEA remained in the CNBr cleaved preparation as verified by sodium dodecyl sulphate (SDS) gel electrophoresis.

Immunization procedure and in vitro stimulation

Mice were injected intraperitoneally with 50 µg SEA or SEACNBr mixed with aluminium hydroxide (AH) (Serva, Heidelberg, Germany) and spleens were aseptically removed 21 days after immunization. CD8⁺ T cells were separated by the magnetic cell sorting (MACS) method using colloidal super‐magnetic microbeads conjugated with monoclonal rat anti‐mouse CD8a (Miltenyi Biotec, Bergisch Gladbach, Germany) following the manufacturer's instructions. Positive fractions were collected and enumerated. CD8⁺ T cells were cultured in 24‐well plates (Nunc, Intermed S.A., Denamrk) in 1 ml of cell culture medium (ML‐10) corresponding to RPMI‐1640 (Gibco, Courbevoie, France) supplemented with 5 × 10^–5 m β‐mercaptoethanol (Merck, Darmstadt, Germany), 2 mm l‐glutamine (Merck), 1 mm sodium pyruvate (Gibco), antibiotics and 10% heat‐inactivated fetal calf serum (Gibco), at 37° in a 5% CO₂ atmosphere. Cells were stimulated with 50 µg/ml SEACNBr for 5 × 10⁵ spleen cells and 1 × 10⁶ antigen‐presenting cells (APC) per well. Stimulation with concanavalin A (Con A, Sigma, St Louis, MO) or purified hamster anti‐mouse CD3ε monoclonal antibody (mAb) 145‐2C11 (PharMingen, Hamburg, Germany) was used as positive controls. Stimulation with Con A was performed at 10 µg/ml for 3 × 10⁶ spleen cells per well (per ml), while for stimulation with anti‐mouse CD3 serial dilutions were made (starting with 20 µg/ml anti‐mouse CD3) and the optimum dose of 5 µg/ml was chosen. Twenty‐four hours after stimulation with Con A or anti‐mouse CD3 or 96 hr after antigenic stimulation, cell culture supernatants were harvested and tested for IL‐2, IL‐4, IL‐5, IL‐10 and IFN‐γ release using the enzyme‐linked immunosorbent assay (ELISA) method described in this manuscript.

Infection protocol

Animals were exposed percutaneously to 50 S. mansoni cercariae on day 0. Mice were bled through the retro‐orbital sinus at days 0, 21, 42 and 69 postinfection for determination of antibodies and cytokines in the sera. Spleens were recovered at days 21, 42 and 69 after infection for evaluation of CTL activity. Livers of day 42‐and day 69‐infected animals were fixed in Bouin's solution before being further processed for histopathological examinations and collagen measurement, or in 4% paraformaldehyde (PFA) for in situ hybridization experiments.

Evaluation of CTL activity by ⁵¹Cr‐release assays

Spleen cells (6–8 × 10⁷) removed 21 days after immunization or at various times after S. mansoni infection (days 21, 42 or 69) were stimulated in vitro with the appropriate antigen (100 µg/ml final concentration) in 10 ml of cell culture medium (ML10) at 37°, 5% CO₂. Five days later, the CTL activity was tested in a standard chromium‐release assay. P815 cells (10⁶), expressing H‐2^d were labelled with 100 µCi (= 3.7 MBq) ⁵¹Cr in Tris‐phosphate buffer, pH 7·4 at 37° for 60 min. After washing, 10^{4 51}Cr‐labelled cells were preincubated for 1 hr in either medium alone or with the appropriate antigen (100 µg/ml final concentration) for 1 hr at 37° in round‐bottom 96‐well plates. Serial dilutions of effector cells in ML10 were then added for 4 hr at 37°. The supernatants (100 µl) were harvested and radioactivity measured with a gamma counter. The percentage specific lysis was calculated as:

$$\begin{array}{l}100\times [(\text{CTL release }-\text{spontaneous release})/\\ (\text{maximum release }-\text{spontaneous release})]\end{array}$$

Maximum release was determined by the addition of 1% Triton‐X‐100. Spontaneous release was determined from target cells incubated without additional effector cells.

Granuloma number determination

Liver sections sampled at days 42 and 69 after S. mansoni infection were embedded in paraffin, and haematoxylin/eosin‐stained and, examined by an image analyser Leitz Asm 68 K (Wild Leitz, Rueil‐Malmaison, France) connected to a Leitz Diaplan microscope and the granulomas were counted. Significant differences (percentage reduction) in granuloma number were calculated between different groups and control mice using Student's t‐test.

Evaluation of fibrosis

The hepatic fibrosis, quantified by measurement of collagen and protein content,³¹ was evaluated on liver sections sampled at days 42 and 69 after S. mansoni infection. Sections (4 µm thick) were placed on slides, deparaffinized and incubated in filtered aqueous picric acid solution containing 0·1% Fast Green FCF (Sigma) which stained non‐collagenous proteins and 0·1% Sirius red F3B (Gurr BDH Chemicals Ltd, Poole, UK) which stained collagen. Sections were kept out of the light and incubated 2 hr at room temperature. They were then rinsed with distilled water until the elution fluid was completely free of colour. Each slide was covered with 1 ml of 0·1 n NaOH in absolute methanol (1 : 1, v : v) until all the colour was eluted from the section (typically a few seconds). Fluids were carefully removed and read in a spectrophotometer DU64 (Beckman Instruments Inc., Fullerton, CA). Fast green has its maximal absorbance at 630 nm and Sirius red at 540 nm. Values were obtained with the use of a previously described formula.³¹ The Student's t‐test and the coefficient of correlation were used in the statistical evaluation of the results. Data are reported as mean ± SE.

In situ hybridization

In situ hybridization was performed as previously described.³² Briefly, frozen sections were hybridized with ³⁵S‐labelled riboprobes (2 × 10⁶ c.p.m. per section) in hybridization buffer containing 15 mm DTT (Promega, Madison, WI). Sections were hybridized overnight at 45° and washed under high stringency conditions (45°, 0·1 × saline sodium citrate (SSC)). Con A‐stimulated splenic cells were used as a positive control for IFN‐γ. For negative controls, cells or sections were hybridized using sense probes for the relevant cytokine. Additional sections were treated with RNAse‐A before the prehybridization step with antisense probes. Specific hybridization was recognized as clear, dense deposits of silver grains in the emulsion overlying cells in the tissue preparations. Cells were identified as dense, discrete, well‐circumscribed areas of silver grains.

ELISA

Antibody detection

Microtitre plates (Dynatech, Denkendorf, Germany) were incubated overnight at 4° with 15 µg/ml of SEACNBr in 0·1 ml of sodium carbonate buffer (15 mm Na₂CO₃, 35 mm NaHCO₃, pH 9·6). The plates were then saturated for 2 hr with 3% phosphate‐buffered saline–bovine serum albumin (PBS–BSA). After careful washes, 0·1 ml of mouse sera diluted 1/100 for immunoglobulin G1 (IgG1) and 1/20 for IgG2a in dilution buffer (0·15 m NaCl, 0·05% Tween‐20, 10 mm PBS, pH 7·2) were dispensed into each well and incubated overnight at 4°. After additional washes, 0·1 ml peroxidase‐labelled anti‐mouse IgG1 or IgG2a (Diagnostic Pasteur, Marnes‐la‐Coquette, France) was added at a dilution of 1/1000 for anti‐IgG1 and 1/500 for IgG2a in the same buffer for 1 hr at room temperature. After a final wash, 10 mg/ml substrate (orthophenyldiamine, Sigma) in 0·1 ml sodium phosphate buffer 0·1 m, pH 5·5, containing H₂O₂ (1 ml/l) was incubated for 10 min at room temperature and the reaction stopped by the addition of 50 µl of 1 m HCl. The OD was measured using a multichannel spectrophotometer (Titertek Multiskan MCC 1340, Labsystems, Cergy‐Pontoise, France) at 492 nm. Results were expressed as the mean of duplicate wells after subtraction of the background.

Cytokine detection

IL‐4 and IFN‐γ in the sera (dilution 1/20) and IL‐2, IL‐4, IL‐5, IL‐10 and IFN‐γ in the supernatants were detected using sandwich ELISA. The antibody pairs used for the detection of these cytokines were purified rat anti‐mouse IL‐2, IL‐4, IL‐5, IL‐10 or IFN‐γ and biotin rat anti‐mouse IL‐2, IL‐4, IL‐5, IL‐10 or IFN‐γ (PharMingen, San Diego, CA). The procedure used was as recommended by the supplier. Optical densities at 492 nm were measured using a multichannel spectrophotometer (Titertek multiskan MCC 1340).

Results

Evaluation of CTL activity

Priming of CTL by immunization with SEA or SEACNBr

The capacity of SEA to induce a specific CTL response was evaluated in BALB/c (H‐2^d) mice immunized i.p. with 50 µg SEA in AH (Fig. 1a). After 21 days, recovered splenic cells stimulated 5 days in vitro with SEA (100 µg/ml) showed 26% specific lysis of SEA‐pulsed P815 (H‐2^d) target cells at an effector : target (E : T) ratio of 50 : 1. Peptidic fragments, obtained after chemical or enzymatic degradation of native antigen, have often been described as inducer of more efficient cytotoxic responses.^33,34 We observed here increased CTL activity (31% for the same E : T ratio) if splenic cells from immunized mice were stimulated in vitro with 100 µg/ml of antigenic preparation obtained after CNBr treatment of SEA (SEACNBr). In consequence, we decided to perform the same experiments with mice immunized i.p. with 50 µg of SEACNBr (Fig. 1b). Splenic cells stimulated in vitro with SEACNBr (100 µg/ml) showed more efficient CTL activity (at an E : T ratio of 50 : 1 but also at an E : T ratio of 25 : 1) against SEACNBr‐pulsed target cells than cells stimulated in vitro with SEA (100 µg/ml). These cells also expressed cytotoxic properties in the absence of in vitro stimulation (26% at an E : T ratio of 50 : 1) demonstrating an in vivo priming of CTLs by SEACNBr. In other experiments (not shown) we observed that a cytotoxic response could also be generated in T‐cell populations from unprimed mice after a 5‐day stimulation period with SEACNBr (100 µg/ml) (21% specific lysis at an E : T ratio of 50 : 1) demonstrating that CTLs can also be generated by in vitro priming with SEACNBr as previously described in other models.³³ So, considering these results, we preferred to use the CNBr‐treated SEA for the ensuing experiments.

Figure 1.

Priming of CTL by immunization with SEA or SEACNBr. BALB/c mice were immunized i.p. with 50 µg of SEA (a) or SEACNBr (b) in AH. After 21 days, spleen cells from immunized mice were cultured for 5 days in vitro without (NS) or with antigen (SEA or SEACNBr) (100 µg/ml). CTL activity was tested on P815 target cells incubated with SEA (left side) or SEACNBr (right side) (100 µg/ml for 60 min at 37°). Results are expressed as mean ± SEM.

As expected, CTLs obtained after immunization with SEACNBr were class I MHC restricted as they did not lyse SEACNBr‐pulsed EL4 (H‐2^b) target cells (12% at an E : T ratio of 50 : 1, not shown) in contrast to the pulsed histocompatible P815 cells. In addition, induction of the specific CTL response was reduced by anti‐class I MHC mAb and not by anti‐class II MHC mAb (Table 1). CD8 molecules were involved in the generation of anti‐SEACNBr CTL responses, as anti‐CD8 mAb reduced the cytotoxic activity (Table 1). Thus, effector cells induced by SEACNBr in vivo were classical class I MHC‐restricted CD8⁺ lymphocytes. As observed in Fig. 2, purified CD8⁺ splenic cells from mice immunized with SEACNBr only produced IL‐2 and IFN‐γ after specific in vitro stimulation with SEACNBr (50 µg/ml).

Table 1.

MHC class I and CD8 restriction of CTL response induction by SEACNBr

Antibody	P815 SEACNBr % specific lysis
–	37 ± 3
Anti MHC class I	18 ± 2
Anti MHC class II	30 ± 4
Anti CD8	15 ± 5

Lysis was tested at an E : T of 50 : 1 in the absence of added antibody or in the presence of saturating concentrations of the specific antibodies. The background lysis never exceeded 10%. Representative of three experiments. The dash represents collagen content per mg protein.

Figure 2.

Cytokine production after in vitro stimulation of CD8⁺ T cells from SEACNBr‐immunized mice. CD8⁺ T cells were isolated from spleens removed 21 days after immunization of BALB/c mice with 50 µg of SEACNBr. IL‐2, IL‐4, IL‐5, IL‐10 and IFN‐γ were measured by specific ELISA in cell‐culture supernatants after 24 hr of stimulation with Con A (10 µg/ml) or anti‐CD3 antibody (5 µg/ml) or 96 hr of stimulation with SEACNBr (50 µg/ml). Results are expressed as mean ± SEM.

SEACNBr‐specific CTL response during the course of S. mansoni infection

CTL activity against SEACNBr‐pulsed target cells was detected in BALB/c mice infected by S. mansoni if recovered splenic cells were stimulated in vitro for 5 days with SEACNBr (100 µg/ml)(Fig. 3). The best activity (32%) was observed at an E : T ratio of 50 : 1 (represented here) and 42 days after infection. But if mice were injected with 50 µg of SEACNBr in AH, one day before being infected by parasite larvae, we evidenced without in vitro stimulation a cytotoxic response 21 days and especially 42 days after infection (respectively 23% and 38%) demonstrating that a single immunization with SEACNBr was sufficient to elicit a significant CTL response for 6 weeks in S. mansoni‐infected mice. This CTL activity was increased by additional in vitro culture of recovered splenic cells for 5 days with SEACNBr (100 µg/ml) (respectively to 38 and to 48%). After 69 days, the cytotoxic response dramatically decreased but remained greater than that observed in non‐immunized infected mice.

Figure 3.

SEACNBr‐specific CTL response during the course of S. mansoni infection. BALB/c mice were immunized (I) or not i.p. with 50 µg of SEACNBr 1 day before being infected by S. mansoni (Inf). Spleen cells recovered 21, 42 or 69 days after infection were cultured for 5 days in vitro without (NS) or with SEACNBr at a dose of 100 µg/ml (SEACNBr). CTL activity was tested on P815 cells incubated with SEACNBr (100 µg/ml for 60 min at 37°). Results are expressed at a E : T ratio of 50 : 1 as mean ± SEM.

Cytokine and antibody response during the course of S. mansoni infection

The production of cytokines and of SEACNBr‐specific antibodies was evaluated during the course of infection in the mice immunized with SEACNBr (Fig. 4). We observed that immunization with SEACNBr increased especially the presence of IFN‐γ in the sera of infected mice in particular at day 69 after infection. The weaker production of IL‐4 observed during infection is also up‐regulated but neither IFN‐γ nor IL‐4 were detected in the sera of non‐infected immunized‐mice. We showed an important production of SEACNBr‐specific IgG1 Abs since 21 days after immunization which was maintained when immunized mice were also infected by the parasite one day after SEACNBr administration. In contrast, the weaker induction of SEACNBr‐specific IgG2a antibodies diminished if immunized mice were infected by S. mansoni.

Figure 4.

Cytokine and specific antibody responses during the course of S. mansoni infection in mice immunized with SEACNBr. Mice immunized i.p with 50 µg of SEACNBr (I) or not (NI) were infected by S. mansoni one day later (Inf) and bled 21, 42 and 69 days after infection. For each time point, sera were pooled and tested by specific ELISA for cytokines (IFN‐γ and IL‐4) or antibodies (IgG1 and IgG2a). Results are expressed as mean ± SEM.

Effect of SEA or SEACNBr administration on hepatic pathology of S. mansoni‐infected mice

In SEA or SEACNBr‐immunized mice, marked reductions in the number of hepatic granulomas were found, particularly at day 42 after infection (31% for SEA and 64% for SEACNBr) compared with control animals (Table 2). Immunization also induced marked diminutions in collagen content/mg protein (ranging from 25 to 20% for SEA and from 50 to 30% for SEACNBr depending on the time of infection) compared with the control group (Table 2). Results were expressed after subtraction of collagen content/mg protein in normal mice 42 or 69 days after the beginning of the experiment. Thus, the level of collagen in the liver of SEACNBr‐immunized mice was close to that observed in normal mice.

Table 2.

Reduction of hepatic damage after immunization with SEA or SEACNBr

	Number of hepatic granulomas (% reduction) on day		Amount of collagen (µg)/mg of protein (% reduction) on day
Immunization status	42	69	42	69
None	117 ± 05	188 ± 11	11.67 ± 1.8	26.52 ± 1.4
SEA	81 ± 02 (31%)	135 ± 07 (28%)	8.77 ± 0.2 (25%)	21.03 ± 1.4 (20%)
SEACNBr	41 ± 15 (64%)	101 ± 10 (46%)	5.80 ± 0.1 (50%)	18.50 ± 0.8 (30%)

Immunizations were performed with 50 µg of SEA or SEACNBr, and mice were challenged 1 day later with 50 cercariae of S. mansoni. Analysis was performed on liver sections sampled on days 42 and 69 of infection. The number of granulomas present and the collagen and protein content in the sections were determined. The protection was expressed as the percent reduction of the mean of the result in immunized mice compared with control mice.

IFN‐γ has been described as a potent inhibitor of collagen synthesis both in vitro and in vivo.²⁴ We decided to examine the production of this cytokine in the liver of SEACNBr‐immunized mice by in situ hybridization. As shown in Fig. 5, IFN‐γ expression was detected in immunized mice using an IFN‐γ antisense RNA probe (Fig. 5c). This production of IFN‐γ was conserved if SEACNBr‐immunized mice were infected by S. mansoni one day after immunization (Fig. 5e). We observed that positive cells were located in all the hepatic tissue and not necessarily around the granulomas. No transcripts were detected in the livers of non immunized infected mice (Fig. 5a). Con A‐stimulated splenic cells were used as a positive control for IFN‐γ (Fig. 5b) and no signals were detected when liver sections were hybridized with the sense IFN‐γ riboprobes (Fig. 5d, f).

Figure 5.

Expression of IFN‐γ in the liver of SEACNBr‐immunized mice. Cryosections of liver from mice immunized or not i.p. with 50 µg of SEACNBr 1 day before being infected or not by S. mansoni were hybridized in situ with mouse IFN‐γ RNA probes. Livers of day 42‐infected mice (a), SEACNBr‐immunized mice (c) and SEACNBr‐immunized infected mice (e) were hybridized with an IFN‐γ antisense probe. SEACNBr‐immunized mice (d) and SEACNBr‐immunized infected mice (f) were hybridized with an IFN‐γ sense probe. Con A‐stimulated splenic cells (b) hybridized with an IFN‐γ antisense probe were used as positive control for IFN‐γ.

Effect of SEA or SEACNBr administration on survival of S. mansoni‐infected mice

The immunization with SEA (not shown) or SEACNBr (Fig. 6) also durably protected the mice, as all the immunized animals survived 2 months longer than control mice before being killed. As expected, the mice from the control group died rapidly around 90–100 days of infection by S. mansoni.

Figure 6.

Survival of SEACNBr‐immunized mice infected by S. mansoni. Mice immunized or not i.p. with 50 µg of SEACNBr were exposed percutaneously to 50 cercariae of S. mansoni. Results are representative of three independent experiments.

Discussion

In the present work, we demonstrated that SEA, defined as the inducer of the CD4⁺‐mediated granulomatous response, is also able to stimulate an antigen‐specific, cytotoxic CD8⁺ T‐cell response in mice after immunization or during the course of the infection by the parasite S. mansoni. So, a single i.p. injection of SEA resulted in the in vivo induction of a significant CTL activity in the spleen within 20 days, demonstrating the possibility of generating a cytotoxic activity with a soluble antigen such as SEA. Moreover, not only could the intact molecule induce an efficient specific CTL response but also its CNBr‐treated form. The high activity obtained after stimulation of splenic cells with the CNBr‐treated SEA may be caused by the generation of peptidic fragments after treatment, inducing of a more efficient CTL activity. We also observed a decrease in the E : T ratio necessary for significant cytotoxic response in the sense of an increase of specific CTLs after immunization with SEACNBr. Consequently, we preferred to use the CNBr‐cleaved form of the SEA for the ensuing experiments. This work is the second demonstration of induction of cytotoxic activity by an antigen from the extracellular pathogen S. mansoni and, as for Sm28GST,²⁸ effector cells induced by SEACNBr in vivo were classic class I MHC‐restricted CD8⁺ lymphocytes. These cells only produced IL‐2 and IFN‐γ after in vitro stimulation with SEACNBr showing a type 1 (Tc1) profile of cytokine secretion for SEACNBr‐specific CD8⁺ T cells. At this stage of our work, it seemed important to study the relevance of these observations in the context of the S. mansoni infection. In contrast to that observed with Sm28GST,²⁸ CTL activity against SEACNBr‐pulsed target cells was measured during the course of the infection with a maximal expression 42 days after infection if recovered splenic cells were stimulated in vitro with SEACNBr. This suggests that SEA‐specific CD8⁺ T cells with cytotoxic potentialities are present during infection. Moreover, mice injected with SEACNBr before infection showed without in vitro stimulation a significant cytotoxic response 21 days and especially 42 days after infection, demonstrating that a single immunization with SEACNBr was sufficient to elicit a significant CTL response for 6 weeks in S. mansoni‐infected mice. This CTL activity was strongly increased by additional in vitro culture with SEACNBr. After 69 days, the specific activity dramatically decreased but remained greater than that observed in non‐immunized infected mice. So, infection did not have a suppressive effect on the cytotoxic response to SEACNBr contrary to that observed in the case of immunization with Sm28GST²⁸ or with vaccinia virus expressing the human immunodeficiency virus envelope protein gp160.³⁵ SEA or SEACNBr‐immunized‐mice seemed durably protected as they survived more than two months after the death of control animals. The role of the immunization‐induced CD8⁺ CTLs in this protection and in the reduction of the hepatic pathology observed during the same time period remains to be elucidated. SEA‐specific cells may act directly by reducing the number of S. mansoni eggs. So, after immunization with a single dose of SEA but especially of SEACNBr, we observed a marked reduction in the number of granulomas in the liver of S. mansoni‐infected mice. This has already been observed by other groups, but only after multiple injections of small doses of SEA, and the levels of IL‐4 were lower in SEA‐treated groups than in infected controls.³⁶ We also observed a reduction of hepatic fibrosis as quantified by measurement of collagen content in the liver of infected mice after immunization with SEA or SEACNBr. In this case also, SEACNBr seemed more effective to induce a protection. We hypothesized that treatment of SEA with CNBr probably allowed epitope(s) with protective potentialities to be exposed, and the characterization of such promising candidate(s) in the SEACNBr preparation is currently underway. This important reduction of hepatic damage is associated with the presence in the liver of SEACNBr‐immunized mice of cells producing IFN‐γ, known to be a potent inhibitor of fibrogenesis. These cells were located throughout the hepatic tissue and not necessarily around the granulomas. This observation was of considerable interest because such production was not evidenced normally in the liver of infected mice and could represent a first explanation for the mode of action of SEACNBr. Moreover the liver, a major organ affected by schistosome infection, is one of identified sites for disposal of postactivated CD8⁺ T cells³⁷ and we envisage the production of IFN‐γ by these cells. Chensue and colleagues have studied the local granuloma (GR) response to S. mansoni eggs and observed that all GR cultures contained IFN‐γ, and that IL‐10 production was accelerated at 8 weeks and abrogated at 20 weeks, consistent with the expansion and abatement of Th2 activity.³⁸ Using MHC class I and/or MHC class II knockout mice, we and other groups^39,40 have confirmed that CD8⁺ T cells play little direct role in the pathogenesis of schistosomiasis.⁴¹ However, a role for the CD8⁺ T cells induced after immunization by SEA in the perturbation of the CD4⁺ T‐cell response (cytokine production or help of antibody production) to infection by schistosomes could not be excluded. We observed the production of SEACNBr‐specific IgG1 antibody, with a maximum 69 days after immunization with this antigen that was maintained when immunized mice were infected by the parasite. In contrast, we observed only a weak induction of SEACNBr‐specific IgG2a antibody after administration of SEACNBr, which diminished if immunized mice were infected by S. mansoni but remained greater than that observed in non‐immunized infected mice. Moreover, immunization with SEACNBr especially potentiated the production of IFN‐γ, suggesting the coexistence of type 1 and type 2 responses in the sera of infected mice. Our results were in agreement with recent work showing that CD8⁺ T cells from schistosome‐infected mice display a type 1 cytokine profile characterized by the selective production of IFN‐γ and IL‐10.¹⁸ A possible explanation for the presence of this type of response is that during infection the initial expansion of CD8⁺ T cells occurs prior to week 6, when parasite egg production begins. During this prepatent period, IL‐4 levels are low and a type 1 response prevails.

In conclusion, our work provides another example that an antigen from the extracellular parasite S. mansoni, notable for its ability to induce CD4⁺ T‐cell response, also stimulates a marked CD8⁺ T‐cell response and protect the mice from infection.

Glossary

Abbreviations

Ab

antibody

AH

aluminium hydroxide

CNBr

cyanogen bromide

CTL

cytotoxic T lymphocyte

E : T

effector : target

SEA

schistosomal egg antigen