INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

Pneumocystis carinii (Pc) pneumonia (PCP) is a severe and common opportunistic infection in immunocompromised hosts, such as patients undergoing chemotherapy for cancer and patients with immunodeficiencies (26). Despite widespread drug prophylaxis, PCP remains an important cause of death in AIDS patients (23). Recent advances in mucosal immunology and the partial success of anti-human immunodeficiency virus therapies suggest that new strategies for the control of opportunistic infections are feasible and necessary. Because Pc proliferates in the mucous-bathed alveoli of the lung, a better understanding of effective local mucosal immune responses might define novel immune-based measures against Pc and other pathogens which utilize the mucosa as the port of entry and/or the primary site of replication.

Both clinical and experimental data support the involvement of both T and B cells in the protection against or the recovery from PCP (27, 16). Depletion experiments demonstrated that the removal of CD4⁺ cells leads to experimental PCP and that activated specific CD4⁺ cells can protect against Pc (15). However, a role for humoral immunity is suggested by the development of specific antibody responses after recurrent episodes of PCP (6) and by the demonstration of Roth and Sidman (25) and Harmsen et al. (18) that antibodies can protect from experimental PCP in severe combined immunodeficient mice and CD4-depleted mice, respectively.

Secretory immunoglobulin A (IgA) (SIgA) is important in maintaining the immune barrier to foreign microorganisms at many surfaces lining the cavities of mammals. Although other nonspecific defense factors exist at these sites, SIgA is the predominant Ig isotype in saliva, tears, breast milk, colostrum, and secretions bathing the lamina propriae of the gastrointestinal, respiratory, and genitourinary tracts (14).

Cholera toxin (CT), the major enterotoxin produced by Vibrio cholera, consists of a toxic A subunit covalently linked to a pentamer of B subunits (CTB) which bind to the monosialoganglioside that is present on all nucleated cells (28). CT and the nontoxic CTB induce significant SIgA and serum IgG antibodies (20) and are among the few adjuvants that do not induce tolerance to unrelated proteins (7). To induce immunity to the target antigen, CTB must be administered by the same route and at the same time. Some debate exists about the capacity of CTB alone to induce memory, since commercial CTB preparations contain low amounts of CT that may act as an adjuvant without measurable symptomatology (19). The adjuvanticity and enhanced IgA response may be due to induced interleukin-1 release by macrophages, enhanced antigen uptake, enhanced major histocompatibility complex class II expression and peptide presentation, and facilitated B-cell switching to IgA-secreting cells (4, 8). CT also promotes differentiation of T cells through a Th2 subset, which increases serum and mucosal levels of IgG1 and IgA, respectively (30).

The role of secretory immunity in protection against Pc pneumonia has not been previously investigated. In the present study, we examined the feasibility of mucosally induced immunity to confer protection against experimental PCP after CD4⁺ cell depletion.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Mice. Pathogen-free, 6-week-old female BALB/c mice and athymic (nu/nu) mice (Harlan Sprague-Dawley, Inc., Indianapolis, Ind.) were used for immunization experiments and ascites production, respectively. Sera from five randomly selected animals were tested for anti-Pneumocystis antibodies by enzyme-linked immunosorbent assay (ELISA) and Western blotting to evaluate preexposure, with negative results (data not shown). Mice were housed in microfilter-topped cages and received sterile food and water. Sterilized cages were changed every week. All animals were maintained in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-approved restricted-access facility, and National Institutes of Health and AAALAC guidelines were followed.

Antigens and adjuvant. Pc organisms were obtained from heavily infected lungs of dexamethasone immunosuppressed mice as previously described (1). Stained smears were also used to evaluate fungal and bacterial contamination, which were not detected. To prepare a uniform Pc antigen, a procedure that enriches small trophozoites was adapted from the improved method developed for rat Pc isolation by Merali and Clarkson (21). Pc-infected mouse lungs were homogenized in equal parts of ice-cold NKPC buffer (2.68 mM KCl, 1.47 mM KH₂PO₄, 51.1 mM Na₂HPO₄, 7.43 mM NaH₂PO₄, 62 mM NaCl, 0.05 mM CaCl₂, and 0.05 mM MgCl₂) and 100 mM dithiothreitol in water, centrifuged at 50 × g for 5 min at room temperature (RT). Pc in the supernatant were collected by centrifugation at 10,000 × g for 10 min at 4°C, resuspended in 5 ml of 0.85% NH₄Cl-NKPC, and incubated at 37°C for 5 min to lyse erythrocytes. After centrifugation (10,000 × g for 5 min at 4°C), Pc microorganisms were resuspended in NKPC with 2 U of RNase-free DNase (Boehringer-Mannheim Co., Indianapolis, Ind.)/ml and incubated at 37°C for 10 min. After three washes in NKPC, Pc were resuspended in 5 ml of the same buffer and subjected to gradient centrifugation (500 × g, 20 min, RT) over Histopaque 1.077 (Sigma Chemical Co., St. Louis, Mo.), and the upper layer was collected and washed three times in NKPC buffer. To disrupt small Pc trophozoites, tubes were freeze-thawed three times in lysis buffer (1% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}-PBS), and the lysate was dialyzed overnight against PBS and concentrated (Ultrafree; Millipore Corporation, Bedford, Mass.). Protein concentrations were determined by a microassay method (BCA Pierce Chemical Co., Rockford, Ill.). This soluble Pc antigen preparation (sPc) was aliquoted and stored in liquid nitrogen for all the immunization experiments. For Western blot analysis, Pc with (sPc) and without (Pc) CHAPS solubilization were used. CTB (List Biological Laboratories, Inc., Campbell, Calif.) was dissolved in phosphate-buffered saline (PBS) at a concentration of 1 mg/ml and used at a concentration of 5 µg with or without 25 µg of Pc soluble crude antigen protein per immunization dose, corresponding to 2 × 10⁶ trophs/dose.

Anti-mouse CD4 MAb production and treatment. The hybridoma GK1.5, which produces a rat IgG2b monoclonal antibody (MAb) against the murine CD4 receptor, was obtained from the American Type Culture Collection (Rockville, Md.). Ascites was produced after injection of 5 × 10⁶ GK1.5 hybridoma cells in pristane-primed athymic nude mice. Intraperitoneal (i.p.) biweekly injections of 300 µl of ascites (containing 900 µg of purified anti-CD4 MAb) were used for anti-CD4 depletion as described previously (2). Levels of CD4-positive cell depletion from blood and spleen were determined 4 days after anti-CD4 treatment by fluorescence-activated cell sorter analysis by using a different anti-CD4 MAb (2B6; PharMingen, San Diego, Calif.) that binds to a determinant that is distinct from the binding site of GK1.5.

Experimental groups. Mice were divided into four experimental groups with 10 animals per group and were intranasally immunized (three doses, one per week) with 25 µg of Pc soluble antigen with (group 1) or without (group 2) 5 µg of CTB (List); with 5 µg of CTB alone (group 3); or with PBS (group 4). The final volume was adjusted to 20 µl per dose. For the immunization, mice were anesthetized with ketamine cocktail (ketamine hydrochloride, 80 mg/ml; acepromazine, 1.76 mg/ml; and atropine, 0.38 mg/ml) at a dose of 15 to 20 µl/mouse in a nontraumatic fashion. This procedure permits both nasal and lung delivery of the immunogen being tested. Two weeks after the last immunization, five animals in each group were sacrificed and used to define the initial immune response to vaccination. The other five animals in each group were CD4-cell depleted with rat anti-mouse CD4 MAb. After 10 days of antibody treatment, mice were inoculated with 1.5 × 10⁶ viable Pc. After five more weeks of biweekly anti-CD4 MAb injections, the remaining animals were sacrificed for analysis of the immune response and protection.

LPR. Tracheobronchial lymph node and spleen cells were separated and incubated at a concentration of 2 × 10⁵ cells/well with optimum concentrations of Pc antigen (10 µg/ml), Con A (10 µg/ml), or medium alone [RPMI 1640, 10% fetal bovine serum, 20 mM HEPES, 50 µM 2-ME, 100 mM L-glutamine, 100 U of penicillin/ml, and 100 µg of streptomycin/ml, all from Sigma] for 5 (Pc) or 3 (Con A) days in 96-well microtiter plates (Costar, Cambridge, Mass.) following the method of Fisher et al. (9). Sixteen hours after the addition of 0.5 µCi of [³H]thymidine/well, lymphoproliferative responses (LPR) were quantified by [³H]thymidine incorporation using a liquid scintillation -counter.

Ig measurements. ELISAs were developed to monitor specific IgA, IgM, and IgG in serum and bronchoalveolar lavage (BAL). ELISA microtiter plates (Sigma) were coated with 1 µg of soluble Pc antigen in carbonate buffer (pH 9.6)/well for 16 h at 4°C, blocked for 1 h with 3% nonfat dry milk in Tris-buffered saline (TBS) buffer (pH 7.6), washed, and incubated overnight with optimal dilutions of serum (1:100) or BAL (1:2) in TBS. After washing, alkaline phosphatase (AP)-conjugated anti-mouse IgG, IgM, or IgA (Sigma) secondary antibodies were added and incubated for 1 h, and the reaction was developed by adding p-nitrophenyl phosphate (Sigma Fast pNPP tablets; Sigma) substrate. Quantification was obtained by measuring absorbance at 405 nm with an automatic ELISA reader.

Ig subclass measurements. ELISAs were developed to monitor specific IgG subclasses in sera from infected and control mice. ELISA microtiter plates (Corning high binding; Fisher, Pittsburgh, Pa.) were coated with 1 µg of soluble Pc antigen in carbonate buffer (pH 9.6)/well for 16 h at 4°C, blocked for 1 h with 3% nonfat dry milk in TBS buffer (pH 7.6), washed, and incubated overnight with optimal dilutions of sera (1:100) or BALs (1:2) in TBS. After washing, rabbit anti-mouse IgG1 or IgG2a (Rockland, Gilbertsville, Pa.) secondary antibodies were added followed by goat anti-rabbit conjugated with AP (Sigma Chemical Co.). The reaction was developed by adding p-nitrophenyl phosphate (Sigma Fast pNPP tablets; Sigma Chemical Co.) substrate. Quantification was obtained by measuring absorbance at 405 nm with an automatic ELISA reader.

SDS-PAGE and Western blot analysis. After reducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), mouse Pc antigen (with and without CHAPS solubilization) and normal mouse lung antigen were transferred to membranes (Immobilon-P; Millipore), blocked with 3% nonfat dry milk in TBS, and incubated overnight with sera (diluted 1:100) or BALs (diluted 1:2) from the different groups. After washing three times, membranes were incubated with peroxidase-conjugated anti-mouse IgG for sera or AP-conjugated anti-mouse IgA for BALs for 1 h at RT. Immune reactivities were detected by using chemiluminescence with either Phototope-horseradish peroxidase (New England Biolabs, Inc., Beverly, Mass.) for peroxidase or Immun-Star (Bio-Rad Laboratories, Hercules, Calif.) for AP.

Assessment of Pc infection. Levels of infection were independently evaluated by two examiners in a blinded protocol by microscopic examination of Giemsa-stained lung impression smears (1). Examiners' infection score mean values for each animal were calculated, and results for each experimental group were analyzed. The counting system was based on the number of microorganisms per field (magnification, ×1,000) as follows: greater than 100 organisms, 5+; 11 to 100, 4+; 1 to 10, 3+; 1 to 9 in 10 fields, 2+; or 1 organism in up to 30 fields, 1+. The score for no organisms in 50 fields was 0.

PCR amplification of Pc rRNA mitochondrial genes. DNA was extracted from paraffin-embedded blocks from the different experimental groups using the QIAamp tissue protocol (QIAGEN, Valencia, Calif.). DNA amplification was performed in a 25-µl volume by using Ready-to-go PCR beads (Amersham Pharmacia Biotech, Piscataway, N.J.) and mitochondrial rRNA primers as described by Harmsen et al. (18). Positive amplification was evidenced by electrophoresis in ethidium bromide-stained agarose gels and visualization by UV transillumination.

Statistical analysis. Comparisons between groups and levels of significance were carried out by using computer-based statistical programs (StatView, Apple). Values of P  than 0.05 were considered significant.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

CD4⁺ cell depletion. Pilot experiments indicated that 4 days after a single i.p. injection of 300 µg of anti-CD4 MAb, less than 1% CD4⁺ cells remained in the blood or spleen of anti-CD4-treated animals (data not shown). In the present study, to achieve severe CD4⁺ cell depletion, 2 weeks after the last immunization, all animals received an excess of anti-CD4 MAb (900 µg) two times per week that continued throughout the remaining 7 weeks of the experiment.

Cellular immune responses. As expected, no statistically significant differences between controls and immunized animals in the response to the control mitogen Con A were observed (data not shown). However, statistically significant differences were observed in lymph node cell LPR from the group immunized with Pc-CTB (cpm, 7,603 ± 1,741) relative to LPR from the animals immunized with Pc (206 ± 26; P < 0.002), CTB (321 ± 134; P < 0.01), or PBS (121 ± 24; P < 0.002). LPR observed with spleen cells in the Pc-CTB group (703 ± 52) were 10 times weaker than lymph node cell LPR (7,603 ± 1,741). However, smaller but statistically significantly higher responses (P < 0.01) were also observed for spleen cell LPR from Pc-CTB-immunized group (703 ± 52) relative to the PBS control group (209 ± 78).

Humoral immune responses after immunization. After immunization, significantly higher specific IgM and IgG serum antibody levels (P < 0.01 and P < 0.001, respectively) were obtained from the Pc-CTB-immunized animals compared with those of the PBS-immunized control group (Table 1). Moreover, significantly higher levels of specific IgA were observed in BALs from Pc-CTB-immunized animals (P < 0.01) (see Fig. 2C).

Antibody subclasses in BAL. At the end of the experiment, we measured the levels of specific IgA, IgG, IgG1, and IgG2a antibodies in BALs of PBS- and Pc-CTB-immunized animals. Compared to control animals, statistically significant higher levels of IgA (P < 0.001), IgG (P < 0.0001), IgG1 (P < 0.001), and IgG2a (P < 0.001) were observed in BALs of Pc-CTB-immunized animals (Fig. 1).

View larger version (30K): [in this window] [in a new window]   FIG. 1.   Humoral immune responses of IgG, IgA, IgG1, and IgG2a from BALs to mouse Pc antigen at the end of the experiment. After immunization, animals were CD4 cell depleted with rat anti-mouse CD4 MAb. After 10 days of antibody treatment, mice were inoculated with 1.5 × 106 viable Pc. After five more weeks of biweekly anti-CD4 MAb injections, the remaining animals were sacrificed for analysis of specific IgG, IgA, IgG1, and IgG2a in BALs by ELISA, as described in Materials and Methods. A 405 nm, absorbance at 405 nm.

SDS-PAGE and Western blot analysis. Our vaccine soluble antigen contained most peptides found in the crude antigen preparation (data not shown), except for slightly lower amounts of the 130-kDa major surface glycoprotein (Fig. 2A). Western blot analysis using BAL IgA from Pc-CTB-immunized animals showed strong reactivity to a Pc antigen with a relative mobility of 55 to 60 kDa and weak reactivities to antigens of 32 and 25 kDa (Fig. 2B). Sera from Pc-CTB-immunized mice demonstrated strong IgG antibody reactivity to Pc antigens with relative mobilities of 130, 55 to 60, 28, and 25 kDa (Fig. 2C1). No reactivity was observed when using normal mouse lung tissue as antigen (data not shown), and serum IgG reactivities disappeared when Pc antigen was treated for 1 h with proteinase K before separation by SDS-PAGE (Fig. 2C2).

View larger version (86K): [in this window] [in a new window]   FIG. 2.   SDS-PAGE and Western blot analysis of IgG and IgA antibodies from Pc-CTB-immunized mice to different Pc antigens. (A) Results of SDS-PAGE showing molecular weight markers (MWM) and sPc antigen. (B) Results of IgA Western blot analysis using 1:2 diluted BAL from Pc-CTB-immunized animals and sPc antigen. (C) Results of IgG Western blot analysis using 1:100 diluted serum from Pc-CTB-immunized animals and sPc antigen. Lanes: 1, sPc (no treatment); 2, proteinase K treatment. The arrow indicates a 55- to 60-kDa antigen.

Level of Pc infection in the lungs of immunized animals. After 5 weeks of CD4⁺ cell depletion, no Pc organisms were observed in Giemsa-stained lung smears of Pc-CTB-immunized mice during microscopic evaluation. However, PCR amplifications using Pc-specific mitochondrial rRNA gene primers were positive in all groups. This means that DNA and/or microscopically undetectable levels of Pc organisms persist in the lungs of Pc-CTB-immunized animals. Other immunization groups showed progressively higher levels of infection. Specifically, levels of infection (scored on the 0-to-5 scale described in Materials and Methods) were as follows (score ± standard error of the mean [95% confidence interval]): Pc antigen-immunized animals, 2.4 ± 0.4 (1.9-2.9); CTB-immunized animals, 3.1 ± 0.6 (2.4-3.8); and PBS-immunized animals, 4.3 ± 0.4 (3.7-4.9). The lower levels of infection observed in the CTB-immunized group suggests some nonspecific immune-mediated level of protection. Additionally, Pc antigen alone was able to evoke lower but statistically significant (P < 0.01) levels of protection compared with the PBS control group.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

This study provides evidence that intranasal immunization with nonviable inocula and CTB can induce protection against Pc pneumonia that persists after CD4⁺ cell depletion. Immunization elicited specific local cellular immunity (lymph node cell lymphoproliferative responses), systemic (serum IgG), and local mucosal (IgA in BALs) humoral immune responses. Based on correlations between antibody levels and levels of infection, we hypothesize that the protection against Pc is mediated to some extent by specific humoral immune responses. By Western blot analysis, both local BAL IgA and systemic IgG antibody strongly detected a 55- to 60-kDa mouse Pc antigen. Interestingly, we and others reported that a 55-kDa antigen from rat Pc elicited strong cellular (29) and humoral (15, 24) immune responses with cells and sera from experimentally infected, convalescent rats, respectively. The 130-kDa major surface protein of mouse Pc is scarcely present in our small trophozoite antigen preparation (Fig. 2A); it did react with serum IgG (Fig. 2C1) but not with BAL IgA antibodies (Fig. 2B). In agreement with our results, two recent publications from Gigliotti et al. (11, 12) stressed the importance of a 50- to 65-kDa antigen. They demonstrated that the majority of the antibodies produced by local B cells from mice recovering from PCP recognized a Pc 50- to 65-kDa antigen (11), and that the Pc gpA (130-kDa) antigen, despite its immunogenicity, was not associated with protection (12).

Induced and constitutive cell-mediated immune responses are traditionally associated with protection against more opportunistic infections, but antibodies were recently implicated in the clearance of Candida, Cryptococcus, and P. carinii from mucosal and alveolar surfaces (5). Non-CD4⁺ immune cells including NK and CD8⁺ cells can cooperate in the clearance of Pc microorganisms from the lung by nonspecific and specific immune mechanisms, respectively (3, 24). Consistent with this hypothesis, general inflammatory responses in the lung induced by bacteria (17) and/or cellular mediators can also activate macrophages and reduce Pc burden in the alveoli (10).

Our results demonstrate the feasibility of mucosal immunization against Pc using a nonviable immunogen. They also suggest that protection in CD4-depleted animals probably requires the participation of specific antibody responses. As shown in Fig. 1, higher levels of IgA and IgG in BALs were observed in the group of animals immunized with Pc-CTB compared with the control (PBS) group. In this experiment, we did not know the origin of the observed specific antibodies. It is possible that during Pc infection, inflamed basement membranes permit the leakage of serum immunoglobulins into the alveolar space. We recently observed that passive transfer of serum antibodies from Pc-CTB-immunized animals partially protected dexamethasone immunosuppressed animals from developing Pc lung infection (data not shown). This results suggest that serum IgG could participate in reducing Pc load in alveolar spaces by different humoral mechanisms. Since Pc is an extracellular pathogen, passive immunization by systemic and/or intranasal instillation of specific antibodies may also be feasible as a immunoprophylactic or therapeutic method. The detailed nature of the developed and retained protective immunity against Pc and the protective antigen(s) remains to be elucidated. However, we suggest that promotion of specific humoral immunity by mucosal intranasal vaccination may provide an additional strategy for protection against Pc and other opportunistic infections.