INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Listeria monocytogenes infection induces both major histocompatibility complex (MHC) class I-restricted CD8- and MHC class II-restricted CD4 T-cell responses (30, 33). Upon phagocytosis, the majority of internalized L. monocytogenes cells are digested in the phagolysosomal compartment of the macrophages while the remaining bacteria reach the cytosol via the action of the product of the hemolysin (hly) gene. L. monocytogenes-derived antigens are thus able to enter both endogenous MHC class I and exogenous MHC class II antigen processing and presentation pathways. L. monocytogenes has been used as an experimental live vector in various viral (e.g., lymphocytic choriomeningitis virus [LCMV], influenza virus, and human immunodeficiency virus [HIV]) and tumoral experimental models (16, 22, 44). The common factor of these models relies on the importance of the CD8-mediated immune response. Deliverance in vivo of a heterologous antigen for further processing along the MHC class II pathway and eventual induction of specific CD4 T lymphocytes has not yet been tested in detail in the L. monocytogenes model of vector immunization. These abilities of L. monocytogenes were evaluated only recently in an HIV peptide presentation model in vitro (19) and in vivo (28). In addition, it is only in the last few years that some targets of the Listeria-specific CD4 T cells have been defined, namely, p60 (10), listeriolysin O (38), the 3A1.1 protein (39), and the 12A4.G7 protein (4). However, a precise characterization of the dynamics of the CD4 immune response directed against a specific antigen during Listeria infection has not yet been conducted.

Numerous reports have shown that L. monocytogenes efficiently induces interleukin-12 (IL-12) secretion by macrophages, either in the noninfectious form of heat-killed bacteria (21) or as live infectious organisms (27, 40, 46). Viable L. monocytogenes induces expression of IL-12/p40 mRNA (40) and protein (27, 46) in mouse spleen macrophages in vitro as well as in vivo; the CD4 T cells induced during experimental infection secrete type 1 cytokines (20), in particular, gamma interferon (IFN-), and they mediate delayed-type hypersensitivity in all mouse strains studied (30, 33, 43). On the other hand, early neutralization of IL-12 decreases resistance to Listeria infection (42).

Experimental infection with the protozoan parasite Leishmania major results, depending on the inbred mouse strain, either in localized self-healing lesions in L. major-resistant mice or in nonhealing lesions in L. major-susceptible mice. Resistance and susceptibility have been shown to rely on the preferential expansion of Leishmania-reactive Th1 and Th2 CD4 T lymphocytes, respectively (37). Following various experimental manipulations of the immune system prior to or at the time of L. major inoculation, susceptible BALB/c mice are able to mount a Th1 immune response that results in healing. In particular, efficient immunization of susceptible BALB/c mice with soluble leishmanial antigens or purified leishmanial proteins has been obtained through coadministration of IL-12 as an adjuvant (1, 32). Among the few leishmanial molecules studied as vaccine candidates, LACK (stands for Leishmania homologue of receptors for activated C kinase) is a 36-kDa protein highly conserved among various Leishmania species and expressed at both the promastigote and amastigote stages (32) and whose functions are now under study (11). The early CD4 T-cell response in susceptible BALB/c mice is oligoclonal and reflects the expansion of a population of V4/V8 CD4 T cells specific for an I-A^d-bound LACK epitope (amino acids 158 to 173). Activation of these cells results in rapid production of IL-4, which drives the subsequent anti-L. major Th2 response, resulting in the characteristic nonhealing disease (23, 26). In vivo depletion of the V4 CD4 T cells (26) or tolerance to LACK in LACK-transgenic mice (23) leads to a healing phenotype in BALB/c mice. Experimental vaccination of mice with LACK has been found to be efficient only when the protein is coadministered with IL-12 (23), when it is administered as LACK DNA (18), or when it is coadministered with IL-12 DNA (17) for time-sustained immunity. Stimulation of the endogenous production of IL-12 at the time of immunization thus seems to be a promising way of optimizing the efficiency of a protective anti-Leishmania vaccine.

In this paper, we analyzed the anti-LACK CD4 immune response generated by a recombinant attenuated Listeria strain expressing the L. major LACK protein. This approach had two aims: (i) to use this foreign antigen as a comarker of the naturally processed listerial proteins and thus to monitor the specific CD4 immune response and (ii) to test the ability of L. monocytogenes, which may be considered to act both as an immunizing vector and as an adjuvant, to express in vivo this heterologous antigen and to induce a Th1-oriented immune response in vivo. We first studied in the main target organs of Listeria, i.e., the spleen and the liver, the dynamics of the CD4 LACK-specific immune response induced by L. monocytogenes after intravenous (i.v.) inoculation and characterized its ability, as a recombinant live vector, to induce a Th1-oriented CD4 immune response in both L. major-resistant B10.D2 and -susceptible BALB/c H-2^d mice. The functionality of these L. monocytogenes-induced CD4 T cells was then characterized by their ability to be recruited in the regional lymph node during the first days of a subcutaneous L. major infection and by their effect on the development of the Leishmania lesions in resistant and susceptible mice.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mice and bacterial and parasitic strains. Female BALB/c and B10.D2/ NOIaHsd mice, aged 8 to 12 weeks, were obtained from our specific-pathogen-free unit at the Institut Pasteur or from Harlan (Bicester, United Kingdom). Bacteria used were the L. monocytogenes-derived Tn917-lac actA mutant LUT12 (25), a recombinant LUT12 strain expressing the nucleoprotein of the LCMV (NP) (15), and Escherichia coli DH5 (Bethesda Research Laboratories, Bethesda, Md.). The L. major strain used was the National Institutes of Health strain 173 (MHOM/IR//173).

Construction of the recombinant LACK-L. monocytogenes. All cloning and analytical procedures were carried out according to standard protocols (2). Plasmids used were CO364, which carries the lack gene of the L. major strain MHOM/IR//173 (33), and pPG5, which carries the fragment coding for the signal sequence and the promoter and regulatory sequences of the hemolysin gene (hly) of L. monocytogenes (15). The entire lack gene without any exogenous additional sequence was excised from the CO364 plasmid with EcoRI and HindIII, and both ends were filled in with Klenow's polymerase. It was then inserted in frame with the fragment coding for the signal sequence of the hly gene at the SmaI site from the pPG5 plasmid linker, yielding the pNS3 plasmid. The sequence of the genetic construct was verified at both junctions, 3' and 5'. The pNS3 plasmid was introduced by electroporation in mutanolysin-treated (45) LUT12, yielding EMA2. Briefly, late-log-phase LUT12 cells were washed in 20 mM Tris1 mM MgCl₂ (pH 6.5), resuspended in 1/40 (vol/vol) of the same buffer supplemented with 0.5 M sucrose (TMS) and containing 50 U of mutanolysin (Sigma), and incubated for 2 min at 37°C. After one wash in TMS, the bacteria with their walls partially digested were subjected to electroporation at 200 , 25 µF, and 2.5 kV; immediately transferred to prewarmed (37°C) 0.5 M sucrose-brain heart infusion broth (Difco, Detroit, Mich.); and incubated for 2 to 3 h at 37°C before isolation of the recombinant clones in the presence of 20 µg of chloramphenicol per ml.

Detection of LACK protein expression in the recombinant L. monocytogenes. The presence of the LACK protein was detected through activation of the LACK-specific hybridoma T cells 0D12 (5 × 10⁴ cells per well) (36) using as antigen-presenting cells (APC) syngeneic bone marrow-derived macrophages (BMDM) prepared as previously described (35). The APC were either incubated with proteins of the Listeria supernatants or infected with the live recombinant Listeria. Culture was performed in a final volume of 200 µl of RPMI 1640-N-acetyl-L-alanyl-L-glutamine (Seromed, Berlin, Germany) supplemented with 10% fetal calf serum, 10 mM HEPES, 5 × 10⁵ M 2-mercaptoethanol, and 50 µg of gentamicin per ml.

Immunization with LACK-L. monocytogenes. Bacterial stocks were kept at 80°C in bacterial culture medium containing 15% glycerol. Aliquots were thawed before the experiments, washed in 0.15 M NaCl, and diluted to the desired concentration in sterile apyrogen-0.15 M NaCl. i.v. injections (5 × 10⁷ bacteria) were done in a volume of 0.2 ml using a 25-gauge needle at a time interval of more than 2 weeks. The size of the inoculum was retrospectively checked by enumeration on Bacto Tryptose agar.

Preparation of mouse cell suspensions and ex vivo reactivation. Single-cell suspensions from at least three mice were prepared from spleen, liver, and popliteal lymph nodes. The isolation of the liver lymphoid cells was performed as previously described (13). Lymphoid cell reactivation was performed by incubating in triplicate wells 2 × 10⁵ to 4 × 10⁵ lymphoid cells per well in a final volume of 200 µl with either medium alone, the LACK peptide containing amino acids 158 to 173 (LACK_158-173) (from 7.5 to 30 µM) (23), the entire recombinant LACK protein (2.5 to 10 µg/ml), heat-killed L. monocytogenes (HKLM; equivalent to 10⁸ bacteria per ml), or, as a negative control, an unrelated antigen, hen egg ovalbumin (OVA; 5 µg/ml). For liver-derived lymphoid cells, irradiated (1,000 rads) syngeneic spleen cells were used as an additional source of APC (5 × 10⁴ cells per well). Cell culture supernatants were harvested at 24 h for IL-2 detection and at 48 h for IFN- and IL-4 detection and kept at 20°C before cytokine quantitation.

Immunoassays for IL-2, IL-4, and IFN-. Enzyme-linked immunosorbent assays (ELISAs) using a pair of monoclonal antibodies specific for IL-2 (JES6-1A12 and JES6-5H4; Pharmingen, San Diego, Calif.), IL-4 (11B11 and BVD6-24G2; Pharmingen), and IFN- (ATCC HB170 and AN-18.17.24 [34]) were used to quantify the amounts of IL-2, IL-4, and IFN- present in cell culture supernatants according to classic protocols (5). A standard curve for each assay was generated with known concentrations of each cytokine. IL-4 and IL-2 standards were supernatants from transformed cell lines derived from X63Ag8-653 (24), kindly provided by F. Melchers (Basel Institute, Basel, Switzerland); murine recombinant IFN- was kindly provided by G. R. Adolf (Ernst Boehringer-Institut für Arzneimittelforshung, Vienna, Austria). Standardization and quantification were done with KC4 software (Bio-Tek Instruments Inc., Winooski, Vt.).

Depletion or enrichment of T-lymphocyte subpopulations. Thy1 T cells were specifically depleted with the anti-Thy1.2 immunoglobulin M monoclonal antibody J1j (ATCC TIB 184) and complement as previously described (14). Depletion or enrichment of CD4 or CD8 T cells was performed by magnetic separation with anti-CD4- or -CD8-labeled magnetic microbeads on VS⁺ columns with a VarioMacs apparatus (Miltenyi Biotec, Bergisch Gladbach, Germany). Depletion or enrichment efficiency was checked by cytofluorimetric analysis using fluorescein isothiocyanate-conjugated anti-CD4 and phycoerythrin-conjugated anti-CD8 antibodies (Becton Dickinson, Le Pont de Claix, France) on a FACScan analyzer (Becton Dickinson).

RNA extraction and reverse transcription. At designated time points after infection, whole spleens were excised and total RNA was isolated using an RNeasy kit (Qiagen, Hilden, Germany) as previously described (6). Briefly, spleen cell suspensions (3 × 10⁷ cells) were homogenized with 1 ml of lysis buffer by several forced passages through a syringe of 1 ml mounted with a 25-gauge needle, and then total RNA was extracted using the columns of the kit according to the manufacturer's protocol and quantified at 260 nm. RNA was reverse transcribed using 5 µg of total RNA in a final volume of 20 µl with 200 U of Moloney murine leukemia virus reverse transcriptase (Gibco BRL) according to the manufacturer's protocol. The reverse transcription mixture consisted of 1 mM deoxynucleoside triphosphates (Pharmacia Biotech, St. Quentin Yvelines, France), 10 mM dithiothreitol, 50 mM Tris-HCl (pH 8.3), 75 mM KCl, 3 mM MgCl₂, 1 ng of hexamer (Pharmacia Biotech), and 50 U of RNAguard (Pharmacia Biotech), and the reaction was carried out over 60 min at 42°C and terminated by a final denaturing step of 5 min at 95°C.

Quantitation of IL-12/p40 transcripts. Reverse transcripts (cDNA) were quantified using a PCR method involving coamplification of an internal standard (6). Briefly, standards are generated by the addition of 2 to 4 bp to wild-type DNA molecules; this addition of nucleotides allows the exchange of a unique restriction endonuclease site present within the wild-type molecule for a new one. This nucleotide addition allows discrimination between the two amplicons after restriction endonuclease digestion. The equivalence between coamplified cDNA and standard DNA was determined following electrophoretic analysis of the digested amplicon products on an ethidium bromide-stained agarose gel.

Inoculation of mice with parasites. L. major promastigotes were obtained from amastigotes recovered from ear lesions and grown in HOSMEM II medium as previously described (3). Stationary-phase promastigotes (5 × 10⁵) were injected subcutaneously (s.c.) into the right hind footpad. Preliminary dose-effect lesion curves were established for both resistant B10.D2 and susceptible BALB/c mice (data not shown). The course of infection was monitored by measuring the footpad swelling using a dial gauge caliper. Results are expressed as increases in footpad thickness.

Statistical analysis. Results are expressed as means ± standard errors of the means (SEM). Statistical significance was calculated by Student's t test.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Characterization of in vitro LACK secretion by recombinant L. monocytogenes. The entire gene encoding the LACK protein of L. major was inserted in frame with the L. monocytogenes listeriolysin (hly) signal sequence under the control of the hly promoter in order to drive secretion of the fusion protein by L. monocytogenes. LACK secretion efficiency by L. monocytogenes was ascertained by activation of the LACK-specific CD4 T-cell hybridoma 0D12 cell line (36) in two different ways. First, in vitro secretion in the bacterial culture medium was determined. When added to MHC class II-expressing BMDM, proteins from the bacterial supernatants were able to stimulate 0D12 cells to release IL-2 in a dose-dependent manner (Fig. 1A) while control proteins from the supernatants of a recombinant L. monocytogenes expressing an unrelated protein, the NP of the LCMV (15), were inefficient. The level of LACK secretion was estimated to be 0.15 ng per 10⁶ bacteria after an overnight culture in brain heart infusion agar at 37°C. Second, BALB/c BMDM infected with LACK-L. monocytogenes were able to activate 0D12 cells (Fig. 1B), thus showing that LACK was secreted and processed efficiently during in vitro cell infection by delivering the I-A^d epitope to the MHC class II molecules on the cell surface.

View larger version (18K): [in this window] [in a new window]   FIG. 1.   Efficient expression of LACK by a recombinant L. monocytogenes was measured by activation of a LACK-specific T-cell hybridoma, either with serial twofold dilutions of TCA-precipitated (see Materials and Methods) bacterial culture supernatants from LACK-L. monocytogenes (filled circles) or from NP-L. monocytogenes as a negative control (open squares) in the presence of BALB/c BMDM as APC (A) or with LACK-L. monocytogenes- or NP-L. monocytogenes-infected BALB/c BMDM (B). IL-2 secretion was quantified by [methyl-3H]thymidine incorporation into CTLL-2 cells as described in Materials and Methods.

Immunization with LACK-L. monocytogenes induces LACK-reactive CD4 Th1 T lymphocytes. Immunization with LACK-L. monocytogenes was carried out in the H-2^d BALB/c and B10.D2 mice, susceptible and resistant, respectively, to L. major infection. Expansion of LACK-reactive T lymphocytes was studied in spleen at various time points after a single i.v. injection of LACK-L. monocytogenes. The LACK-dependent T-lymphocyte stimulation was very brief, since in vivo expression of LACK was short-lived due to rapid elimination of the recombinant plasmid: only 1% of the injected bacteria still harbored the plasmid after 24 h (data not shown). This allowed us to mimic the physiological short-lived presentation of the LACK antigen to the immune system during L. major infection as previously reported (7, 36).

View larger version (23K): [in this window] [in a new window]   FIG. 2.   Kinetics of the LACK-specific immune response in B10.D2 mice (straight lines) and BALB/c mice (broken lines) after a single i.v. inoculation of LACK-L. monocytogenes (filled symbols) or NP-L. monocytogenes as a negative control (open symbols). Ex vivo reactivation of spleen (A) or liver (B) lymphoid cells (4 × 105 cells per well) was carried out in the presence of the LACK158-173 peptide (7.5 µM), the LACK protein (2.5 µg/ml), or HKLM (108 organisms/ml). IL-2 and IFN- secretion levels in the supernatants were quantified by ELISA as described in Materials and Methods. Results are expressed as means of results of triplicate experiments ± SEM. aa, amino acids.

View this table: [in this window] [in a new window]   TABLE 1.   Effects of Thy1 or CD4 depletion and CD4 or CD8 enrichment on IFN- secretion by B10.D2 splenocytes on day 5 after LACK-L. monocytogenes inoculationa

Early cytokine profile of LACK-reactive T cells from LACK-L. monocytogenes-immunized mice recovered from draining lymph nodes during L. major infection. The ability of primed LACK-reactive T cells generated by the Listeria immunization to be recruited within the draining lymph node was tested during the early phase of a leishmanial infection. Two weeks after immunization with LACK-L. monocytogenes, BALB/c and B10.D2 mice were infected s.c. with L. major and on day 3 the presence of LACK-reactive CD4 T lymphocytes in the lymph node draining the infected site, through cytokine secretion after specific in vitro reactivation, was characterized. When cells from the contralateral lymph node were used as negative controls, no IL-2, IL-4, or IFN- secretion could be detected in either resistant or susceptible mice (data not shown).

View this table: [in this window] [in a new window]   TABLE 3.   Effect of Thy1 or CD4 depletion on IFN- and IL-4 secretion by BALB/c lymph node cells isolated on day 3 after L. major footpad inoculation (5 × 105 promastigotes)a

Protective function of the LACK-reactive T cells generated by LACK-L. monocytogenes immunization. The potential efficiency of LACK-L. monocytogenes immunization on the control of an L. major infection was tested in both resistant and susceptible mice. In resistant B10.D2 mice (Fig. 3A), control of lesion progression could be observed only after two injections of LACK-L. monocytogenes; lesion development was slower than in the control groups and reached a lower plateau, and this plateau was maintained throughout the whole course of the clinical process. In contrast, the characteristic progressive disease in susceptible BALB/c mice could not be restrained (Fig. 3B). Even infection with a lower dose of parasites (3 × 10⁴) close to the last inoculation of LACK-L. monocytogenes (day 8) or reinjections of LACK-L. monocytogenes when footpad thickness increases reached 2 to 3 mm in order to boost the anti-LACK immune response did not modify the progression of the lesions (data not shown).

View larger version (19K): [in this window] [in a new window]   FIG. 3.   Protective status of B10.D2 mice and BALB/c mice after LACK-L. monocytogenes inoculation. (A) B10.D2 mice received one (see insert) or two (on days 0 and 22) i.v. injections of LACK-L. monocytogenes or NP-L. monocytogenes. Two weeks later, 5 × 105 to 1 × 106 stationary-phase promastigotes were inoculated s.c. into the right hind footpad and the resulting footpad swelling was monitored. (B) BALB/c mice received three injections of LACK-L. monocytogenes or NP-L. monocytogenes at days 0, 16, and 46. Five weeks later, 4.5 × 105 stationary-phase NIH 173 promastigotes were inoculated s.c. into the right hind footpad and the resulting footpad swelling was monitored. Results are expressed as means ± SEM (five or six mice per group). LM, L. monocytogenes.

View larger version (21K): [in this window] [in a new window]   FIG. 4.   Quantitative analysis of IL-12/p40 mRNA levels in spleens of LACK-L. monocytogenes- and NP-L. monocytogenes-inoculated BALB/c mice. Splenic total RNAs were extracted at various time points after the inoculation of 5 × 107 LACK-L. monocytogenes or NP-L. monocytogenes cells, and quantitative reverse transcription-PCR was carried out as described in Materials and Methods. Results are expressed as numbers of IL-12/p40 transcripts per 106 copies of actin mRNA and are the sums of results of three separate experiments. Variation in the actin mRNA quantifications in the various samples was less than 3% (0.99 ± 0.03 × 106 copies/ng of total RNA; means ± SEM; n = 14). LM, L. monocytogenes.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this paper, we analyzed the CD4 immune response generated by a recombinant attenuated Listeria strain expressing a parasite antigen, the LACK protein of L. major. Our system enabled us to use this foreign antigen and its well-defined I-A^d epitope as a comarker of the naturally processed listerial proteins and to trace in vivo the CD4 immune response generated in the course of a short-lived listerial infection. Our study showed that the attenuated L. monocytogenes actA mutant (25) was efficient in delivering in vivo the LACK protein for further processing along the MHC class II pathway. Evaluation of the ability of L. monocytogenes as a live recombinant vector to induce CD4 T cells reactive to the expressed foreign antigen has been studied only recently in an HIV peptide presentation model in vitro (19) and in vivo (28). Other live bacterial vectors expressing LACK, namely, Salmonella enterica serovar Typhimurium and Mycobacterium bovis BCG, have also been studied in vivo with the LACK model. In these model systems, LACK-reactive T cells could not be detected after immunization (N. Glaichenhaus, unpublished results).

Reports have shown that LACK is presented in vitro by MHC class II-expressing mononuclear phagocytes only during the first hours following infection with stationary-phase L. major promastigotes (7, 36). Furthermore, macrophages infected with metacyclic promastigotes or amastigotes only weakly restimulated or could not restimulate the LACK-specific-T-cell hybridoma cell line 0D12, respectively (7, 36). Those reports suggest that the I-A^d LACK peptide may be accessible to the immune system only at a critical early period of Leishmania infection. The use of a plasmid construct enabled us to mimic the short-lived physiological presentation of the LACK I-A^d peptide to the immune system during L. major infection and to test its ability to induce CD4 T cells of the Th1 phenotype, which could exert a protective function against parasite infection.

The peak of IL-2 and IFN- secretion detected on day 5 in spleen may reflect the in situ expansion of LACK-reactive T cells or their recruitment in infected foci, since the spleen is at the same time a lymphoid organ and a target of Listeria infection. The former phenomenon may be considered predominant, if one assumes that the dynamics observed in the liver, which is a nonlymphoid organ, reflect the importance of the recruitment observed in the infectious sites, whatever the organ considered. The anti-LACK immune response faltered on day 7 in spleen. This decrease in the measured T-cell response is most probably related to the release of the LACK-reactive T cells in the vascular compartment, where they could be later recruited in the periphery either as activated or memory T cells, or to apoptosis after interaction and activation with the infected APC. The presence of LACK-reactive T cells was transient in liver and can probably be related to the transient release of the LACK protein; clearance of L. monocytogenes actually occurred in 5 days in this organ.

The CD4 component of the anti-LACK immune response detected in spleen and liver was clearly with a Th1 cytokine secretion profile in mice of both strains, with secretion of IFN- occurring without any IL-4 after ex vivo LACK reactivation. This result is in accordance with the fact that wild-type L. monocytogenes is a well-known inducer of IL-12 secretion by macrophages (21, 27, 40, 42, 46). When LACK is injected as an immunogen into BALB/c mice, either as protein alone (18) or in incomplete (23) or complete Freund's adjuvant (29), significant IL-4 and IL-5 secretion is observed in the draining lymph node cell populations 1 to 2 weeks afterward. The fact that we did not observe any detectable amount of IL-4 after LACK-L. monocytogenes immunization suggests that our Listeria vector had at least transiently prevented the Th2 differentiation of LACK-reactive T lymphocytes or, alternatively, induced the deletion of the type 2, already differentiated cells.

These LACK-reactive T cells induced through L. monocytogenes immunization could be recruited efficiently in the lymph node draining an L. major-infected site, where they could be detected on day 3 after parasite inoculation and still exhibited a Th1 secretion pattern, characterized by a strong level of IFN- secretion. This strong level of IFN- secretion observed in the L. major-infected lymph node on day 3 might then favor a bias towards Th1 differentiation of the L. major-driven T-cell immune response, especially in susceptible BALB/c mice, and result in eventual protection in mice of both resistant and susceptible strains. Effectively, a more efficient control of the leishmanial infection was observed in the resistant B10.D2 mice but needed two immunizing inoculations of LACK-L. monocytogenes. However, in the susceptible BALB/c mice, the progression of the disease went on unabated. LACK-L. monocytogenes-immunizing boosts during infection did not modify the evolution of the lesions. This point suggests that IL-12 secretion after the LACK-L. monocytogenes boosts may have occurred too late during L. major infection to curb disease progression, as previously observed (41).

Why could the short-term immune response driven by L. monocytogenes, which is a powerful known IL-12 secretion inducer, not block the Th2 differentiation of the L. major-specific immune response under these experimental conditions? Some reports have shown that persistence of IL-12 production after immunization is important for time-sustained immunity (17). In our work, the attenuated Listeria persisted only a few days, 3 days in spleen and 5 days in liver (data not shown). Nevertheless, the recombinant attenuated L. monocytogenes vectors were able to significantly increase IL-12/p40 mRNA transcription in vivo. The duration of in vivo IL-12 secretion achieved under these experimental conditions might, however, be too short to ensure optimal efficacy. One way to trigger a more sustained IL-12 production would be to test a wild-type virulent L. monocytogenes which persists longer in its host. On the other hand, in vivo expression of LACK and its further use as a source of immunogenic peptides are very short-lived and, even if the stimulation of the immune system is sufficient to induce protection in B10.D2 mice, stronger antigenic stimulation may be needed to produce the same protective effect in BALB/c mice. In preliminary experiments, we have indeed observed that immunization with a chromosomal construct of a LACK-expressing actA attenuated L. monocytogenes strain which expressed LACK during the whole period of infection was able to delay L. major lesion onset in BALB/c mice. More precise characterization of the effect of immunization with LACK chromosomal constructs in attenuated or wild-type L. monocytogenes on L. major lesion control is under way.

Viewed together, our results show that L. monocytogenes, as a live recombinant vector, in addition to being able to generate effector CD8 T lymphocytes in infection and tumor models, is able to generate in vivo CD4 T cells with a Th1 phenotype that can exert their antiparasite function, at least in L. major-resistant mice.