Immunogenicity of a 48-Kilodalton Recombinant T-Cell-Reactive Protein of Coccidioides immitis

Abstract

The outcome of coccidioidomycosis depends to a large extent on the effectiveness of the T-cell-mediated immune (CMI) response to the fungal pathogen. For this reason, identification of Coccidioides immitis antigens which stimulate T cells is important for understanding the nature of host defense against the organism and essential for the development of an effective vaccine. Here we describe the immunogenicity of a 48-kDa T-cell-reactive protein (TCRP). The antigen is expressed by parasitic cells and localized in the cytoplasm. It stimulates the proliferative response and production of gamma interferon by T cells of mice immunized with C. immitis spherules. Specific antibody reactive with the recombinant TCRP (rTCRP) was detected in sera of patients with confirmed coccidioidal infection, and the highest titers of antibody to the recombinant protein correlated with elevated titers to the serodiagnostic complement fixation antigen of C. immitis. These results suggest that the TCRP is presented to the host during the course of infection. Immunization of BALB/c and C3H/HeN mice with the rTCRP affords a modest but significant level of protection against an intraperitoneal challenge with C. immitis. It is suggested that the rTCRP may be able to contribute to a multicomponent vaccine designed to stimulate CMI response against the parasitic cycle of C. immitis.

Coccidioides immitis is a fungal pathogen found in desert soil of the southwestern United States and northern Mexico. The incidence of coccidioidomycosis reported in the United States varies widely from year to year but can be extraordinarily high, as it was in 1991 to 1994 (16). Although the vast majority of immunocompetent patients resolve C. immitis infection spontaneously, recovery often takes weeks to months (12). Immunocompromised patients often develop disseminated disease, which can be fatal without aggressive antifungal therapy (3, 10). An effective vaccine for coccidiodomycosis would significantly improve the overall health of the people living in the areas of endemic infection.

Infection of humans with C. immitis followed by recovery from coccidioidomycosis almost always produces lifelong immunity (26). For this reason, we believe that vaccination is a feasible goal. Protection is correlated with a strong delayed-type hypersensitivity response in humans (26) and is transferred with immune T cells in mice (2). Activation of the Th1 rather than the Th2 subset of T cells is associated with spontaneous resolution of disease in mice (20, 21). Therefore, a vaccination that stimulates protective T-cell immunity is required. In experimental animals, vaccination with formalin-killed spherules (23) or spherule extracts (19) can provide protection. This suggests that T-cell-reactive antigens expressed by the parasitic phase of C. immitis could be useful components of a vaccine.

We previously described a β-galactosidase fusion protein, derived from a C. immitis cDNA expression library, that stimulated a mouse T-cell line specific for C. immitis antigens (17). The fusion protein contained a 66-amino-acid recombinant peptide expressed by the C. immitis cDNA. We previously cloned and sequenced the full length gene (28), and here we report the expression and purification of the recombinant protein. In this study, we characterized the immunogenicity of the full length recombinant T-cell-reactive protein (rTCRP) and evaluated its immunoprotective capacity against C. immitis infection in mice.

MATERIALS AND METHODS

Expression of the tcrP gene.

A 1.1-kb cDNA fragment of the tcrP gene (28), which encodes amino acids 1 to 381 of the rTCRP, was restricted from the full-length gene with NcoI and ApaI. The cDNA was cloned in frame by blunt-end ligation into the NheI and SacI sites of the pET28b plasmid vector (Novagen, Madison, Wis.). This plasmid construct is referred to as pET28b-tcrP. The pET28b-tcrP plasmid encodes a 425-amino-acid fusion protein (48-kDa rTCRP) with 23 N-terminal residues (including a 6-residue His tag) and 21 C-terminal amino acids derived from the vector.

The pET28b-tcrP construct was used to transform Escherichia coli HMS174(DE3). Bacterial extracts were prepared by growing cells to an absorbance at 600 nm of 0.4, with or without the addition of isopropyl-β-d-thiogalactopyranoside (IPTG). The final concentration of IPTG was 1 mM. The bacteria were pelleted, sonicated, and washed repeatedly with 20 mM Tris-HCl (pH 7.9) containing 5 mM imidazole and 0.5 M NaCl (binding buffer). Inclusion bodies, which contained the recombinant protein, were isolated by successive centrifugations (20,000 × g) and washes with binding buffer and then sonicated in the same buffer, to which 6 M urea had been added. The sonicated fraction was incubated on ice for 1 h and centrifuged, and the soluble supernatant (50 ml) was passed through a 10-ml column containing His:Bind metal chelation resin (Novagen) for nickel affinity chromatographic isolation of the recombinant protein. The column was washed with 10 column volumes of binding buffer (containing 6 M urea) followed by 6 column volumes of wash buffer (20 mM Tris-HCl [pH 7.9], 20 mM imidazole, 0.5 NaCl, 6 M urea). The rTCRP was eluted from the affinity column with 500 mM imidazole and then refolded by successive dialysis steps (12 h each) at 4°C against the following buffers to yield soluble protein: (i) 40 mM Tris-HCl (pH 7.9) containing 400 mM NaCl, 4 mM dithiothreitol (DTT), and 4 M urea; (ii) 20 mM Tris-HCl (pH 7.9) containing 200 mM NaCl, 2 mM DTT, and 2 M urea; (iii) 10 mM Tris-HCl (pH 7.9) containing 100 mM NaCl, 1 mM DTT, and 1 M urea; and finally (iv) three changes of 10 mM Tris-HCl (pH 7.9) containing 100 mM NaCl. The retentate (0.1 to 0.2 mg of protein/ml) was aliquoted and stored at −70°C.

A second plasmid construct in pET11d (Novagen), which contained the initially isolated 0.18-kb cDNA (17) fragment of the tcrP gene, was previously used to transform E. coli BL21(DE3) and express a 6.5-kDa recombinant peptide (4). This recombinant peptide was used to raise the specific antibody used in this study. The fusion peptide lacked a His tag, and the expression product was purified by gel electroelution (4). The amino acid sequence of the peptide was shown to match a corresponding region of the translated tcrP gene (4, 28).

Amino acid sequence analysis of the rTCRP.

To confirm that the 48-kDa chromatographically isolated recombinant protein showed the predicted sequence of the translated tcrP gene, an amino acid sequence analysis of the putative rTCRP was conducted. A LysC digestion of the protein was performed as specified by the supplier (Promega, Madison, Wis.). Digested peptides were separated on a reducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel (10% polyacrylamide), electrotransferred to an Immobilon P membrane (Millipore, Bedford, Mass.), and stained with Coomassie brilliant blue R250 as previously described (9). A 38-kDa membrane-bound peptide was excised and subjected to N-terminal amino acid sequence analysis.

Antibody production.

Dunkin-Hartley guinea pigs (Pocono Farms, Pocono, N.Y.) were used previously to raise antisera against the purified 6.5-kDa recombinant peptide (4). In the present study, BALB/c mice were immunized subcutaneously (s.c.) with 20 μg of the purified 48-kDa rTCRP in Ribi adjuvant containing monophosphoryl lipid A and trehalose dimycolate (Ribi ImmunoChem Research, Hamilton, Mont.). The mice were immunized s.c. a second time 14 days later with the same amount of antigen alone and then exsanguinated 10 days after the boost.

Immunoblot analysis.

Protein preparations of transformed E. coli lysates were separated on an SDS-PAGE gel (10% polyacrylamide) and transferred to an Immobilon P membrane. The membrane was incubated at 4°C overnight in 3% dry milk dissolved in 10 mM Tris-HCl (pH 8.0) which contained 150 mM NaCl and 0.1% Tween 20 (TBST blocking buffer). Antiserum raised against the 6.5-kDa rTCRP peptide was used to identify the full-length tcrP gene-encoded product. The guinea pig antiserum was diluted 1:500 in TBST and then incubated with the membrane-bound proteins for 2 h at 25°C. The membrane was subsequently washed three times in TBST, incubated for 30 min with goat anti-guinea pig immunoglobulin G (IgG) (heavy and light chains) conjugated to alkaline phosphatase (1:1,000 dilution in TBST) (Kirkegaard & Perry Laboratories, Gaithersburg, Md.), washed again, and developed in substrate solution (Novagen).

Antiserum raised in BALB/c mice against the chromatographically purified 48-kDa rTCRP was used for detection of reactive proteins in SDS-PAGE separations of parasitic cell homogenates. The parasitic phase of C. immitis C735 was grown in vitro as previously described (18). Cell disruption was performed by glass bead homogenization while samples were maintained on ice (4°C). Cell wall material was separated from the homogenate by centrifugation (2,925 × g for 30 min). The supernatant, which contained primarily cell organelles plus the soluble cytosolic fraction, was centrifuged at 88,000 × g for 60 min to obtain the membrane pellet. Aliquots of the cytosolic fraction, wall pellet, and membrane pellet were further separated by SDS-PAGE and subjected to immunoblot analysis. The conditions used for immunoblotting were the same as above, except that anti-mouse IgG conjugated to horseradish peroxidase was used as the secondary antibody (1:5,000 dilution in TBST) together with the substrate provided by the manufacturer (ECL immunoblot kit; Amersham, Arlington Heights, Ill.).

Immunolocalization.

Spherules of C. immitis C735 were prepared for microscopy studies by cryofixation with a propane-jet freezer (model Q.F. 1000; Aldrich-Erdos, Gainesville, Fla.) as previously described (18). The frozen spherules were freeze-substituted, embedded in LR White resin (25), and sectioned for immunofluorescence and immunoelectron microscopy as reported previously (18). The sections were incubated at room temperature with mouse antiserum against the purified 48-kDa rTCRP (1:400 dilution) in 0.1 M phosphate-buffered saline (PBS; pH 7.4) or with preimmune serum at the same dilution. The sections were rinsed with PBS, reacted with either fluorescein isothiocyanate (FITC)-conjugated or colloidal gold-conjugated goat anti-mouse IgG (Sigma, St. Louis, Mo.) at a 1:50 dilution in PBS for 30 min, and then washed three times with PBS. For immunofluorescence studies, sections attached to glass slides were mounted in PBS (pH 9.0)-glycerol (1:9) solution to retard the quench rate of the fluorescent stain. Sections prepared for immunoelectron microscopy were poststained with uranyl acetate and lead citrate as previously reported (25).

ELISA.

The enzyme-linked immunosorbent assay (ELISA) was performed with an indirect-screening kit (Kirkegaard & Perry Laboratories) by a previously reported method (6), except that the peroxidase substrate used was o-phenylenediamine and the intensity of the color reaction was determined by measuring the absorbance at 490 nm. The reactivity of serum samples from six patients diagnosed with coccidioidomycosis (VA Medical Center, San Diego, Calif.) was compared to that of samples from control patients (hospital admissions with no indicated fungal infection). The sera were diluted 1:400 in blocking solution (6). Assays with sera in the absence of antigen and with antigen in the absence of primary sera served as controls. All sera were tested in triplicate wells. All sera from patients with coccidioidal infection were determined to be seropositive for C. immitis antigen in the immunodiffusion-complement fixation (ID-CF) assay (14). The antibody titer in the ID assay for each patient serum was determined with the CF reference antigen previously reported (6).

C. immitis antigens.

Formalin-fixed spherules (17), the mycelial-filtrate-plus-lysate fraction (F+L) (7), the soluble conidial wall fraction (SCWF) (5), and a spherule outer wall (SOW) fraction (6) were prepared as reported previously.

Animals.

All T-cell proliferation assays and infection experiments were conducted with 10-week-old female BALB/c or C3H/HeN mice supplied by the National Cancer Institute (Bethesda, Md.). Female BALB/c mice which were used for antibody production were obtained from Charles River Breeding Laboratories (Wilmington, Mass.).

T-cell proliferation assay.

Female BALB/c mice were immunized s.c. with 10⁶ formalin-fixed spherules of C. immitis C735 in PBS (pH 7.4). The mice were given booster immunizations s.c. 14 days later with the same number of spherules in PBS, and they were given final immunizations 14 days after that with 1.5 × 10⁶ spherules in complete Freund’s adjuvant (Sigma) in both hind footpads and at the base of the tail. Control mice received complete Freund’s adjuvant alone. Ten days later, the draining lymph nodes were removed from both groups of mice. A cell suspension was made, and T cells were purified on a mouse T-cell recovery column (Cedarlane, Hornby, Ontario, Canada). The T cells (approximately 2 × 10⁵ cells) were cultured with 3 × 10⁵ irradiated, syngeneic spleen cells for 5 days with or without antigen and pulsed with [³H]thymidine for the last 18 h before the cells were harvested from the wells of the Costar plates. Details of the methods used for the lymphocyte proliferation assay were described previously (5). The reactivity of C. immitis antigens was also tested in a SCWF-specific murine T-cell line, which has been described previously (17). The cell line expresses Thy1, L3T4 (>95% of the cells), and low levels of Ly2 (<5% of the cells), as demonstrated by flow microfluorimetry. We chose to use SCWF as the selective antigen for the T-cell line because it is multicomponent and highly antigenic for C. immitis-immune lymph node cells (5). The affinity-purified rTCRP was exhaustively dialyzed against 20 mM HEPES buffer (pH 7.4) containing 0.15 M NaCl before it was used in the cellular immunoassay. The level of endotoxin contamination in the F+L, SCWF, and rTCRP preparations was measured with a Limulus amebocyte lysate QCL-1000 kit (BioWhittaker, Walkersville, Md.). All preparations had less than 675 IU (corresponding to 100 pg of E. coli J5) of endotoxin per μg of protein. Three replicate wells were used for each concentration of antigen tested.

Fluorescence-activated cell sorter (flow microfluorimetry) analysis.

Total cells were harvested from cultures incubated with antigen after 72 h in the T-cell proliferation assays described above. The cell suspensions were aliquoted and separately stained with the following antibodies to identify the responsive cell type(s): FITC-L3T4 (anti-CD4), FITC-Ly2 (anti-CD8), phycoerythrin (PE)-Thy 1.1 (all T cells), PE-anti-CD45R/B220 (all B cells), PE-anti-αβ T-cell receptor (TCR), or PE-anti-γδ TCR. All antibodies were obtained from Pharmingen (San Diego, Calif.). Responsive (blast) cells were identified by analyzing forward and side scatter, and the percentage of cells reacting with each labeled monoclonal antibody (MAb) was determined by flow microfluorimetry with a Coulter EPICS Elite flow cytometer.

IFN-γ assay.

Microtiter plate immunoassays of the levels of gamma interferon (IFN-γ) production were performed on culture supernatants of responsive T cells described above. IFN-γ was measured with an immunoassay kit supplied by Pharmingen. Rat MAb R46A1 raised against mouse IFN-γ was bound to microtiter plates and used as a capture antibody. Recombinant mouse IFN-γ was used as a standard. The immunoassays were performed as specified by the manufacturer.

Infectious challenge experiments.

Groups of eight female BALB/c or C3H/HeN mice were used for infection experiments. The mice were immunized s.c. with either SOW (10 μg) or rTCRP (4 μg) in 100 μl AdjuPrime (Pierce, Rockford, Ill.) on days 1 and 14. The animals were given a final immunization 14 days later in both hind footpads and the base of the tail with either 10 μg of SOW or 7.5 μg of rTCRP in 100 μl of CFA. Immunization with the rTCRP by this method was shown to elicit a strong response to the immunizing protein in T-cell proliferation assays (data not shown). Control mice received AdjuPrime and CFA only. At 14 days after the last immunization, the BALB/c mice were challenged with 50 viable arthroconidia of strain C735 via the intraperitoneal (i.p.) route. Although i.p. inoculation is not the natural route of infection by this fungal pathogen, this challenge mode permitted more precise control of the size of the inoculum actually delivered to host tissues and better reproducibility of levels of coccidioidal infection in the lung than the intranasal route of inoculation did. The C3H/HeN mice were challenged with 2,000 viable arthroconidia by the same route 2 weeks after the last immunization. C3H/HeN mice have been reported to show some degree of natural resistance to C. immitis infection, which is not apparent in BALB/c mice (15). Resistance to coccidioidal infection in C3H/HeN mice was compensated by the larger inoculum size. Two weeks after the challenge, the lung homogenates of surviving mice were subjected to quantitative analyses of CFU in dilution plate cultures of C. immitis as previously described (15). Necropsy and quantitative culture of the lungs of moribund mice revealed extensive pneumonia with more than 10⁶ C. immitis spherules/lung. For this reason, mice that died before day 14 postchallenge were assigned the highest value of CFU/lung found in the survivors.

Statistical analysis.

The numbers of CFU per lung were expressed on a log scale. Because these values did not fall into a normal distribution, the Mann-Whitney U test was used to compare medians in all cases.

RESULTS

Expression and purification of the tcrP.

To confirm that the tcrP gene was expressed, lysate preparations of E. coli HMS174(DE3) transformed with pET28b-tcrP and induced with IPTG were examined by SDS-PAGE and immunoblot analysis. Figure 1 shows that a 48-kDa band was detected in highest concentration within the inclusion bodies isolated from the IPTG-induced bacteria. This protein was purified by nickel affinity chromatography and identified as the rTCRP in an immunoblot with antisera previously raised in guinea pigs against a 6.5-kDa rTCRP (4). To further verify that the expressed protein was actually encoded by the C. immitis tcrP gene, the amino acid sequence of an isolated 38-kDa LysC-digested peptide of the purified recombinant protein was obtained (data not shown). The sequence of amino acids was determined to be AVASHVVRNGNITFI, which is identical to a portion of the predicted amino acid sequence of the tcrP gene (28).

FIG. 1.

SDS-PAGE analysis of the 48-kDa recombinant protein expressed by E. coli transformed with a 1.1-kb cDNA insert of the C. immitis tcrP gene in the pET28b-tcrP construct. Bacterial extracts separated by SDS-PAGE were obtained from strain HMS174(DE3), which was transformed but not induced (−IPTG; 5 μg of total lysate) or transformed and IPTG-induced (+IPTG). The protein contents of the lysate (Lys.; 5 μg) from the transformed and induced bacteria, together with the supernatant (Supn.; 2 μg) and inclusion bodies (Inc.; 5 μg) derived from the same lysate were compared. The purified nickel affinity-bound recombinant protein (R.P.; 2 μg) is shown in a separate lane. The protein content of each fraction was visualized by staining with Coomassie brilliant blue R-250. Antiserum previously raised against a 6.5-kDa TCRP was used to identify the 48-kDa rTCRP in an immunoblot (Ib.) of the total lysate of the transformed and induced E. coli strain.

Expression of the native TCRP in the parasitic phase of C. immitis.

Antiserum raised in mice against the purified 48-kDa rTCRP was used to screen parasitic cell homogenates in an attempt to identify the native TCRP. SDS-PAGE gel separations of the spherule cytosolic fraction, wall pellet, and cell membrane pellet were subjected to immunoblot analysis (Fig. 2). The antiserum recognized a 50.4-kDa band only in the cytosolic fraction. Antiserum raised in guinea pigs to the 6.5-kDa fusion protein (4) recognized the same cytosolic polypeptide (data not shown). The predicted molecular size of the TCRP encoded by the full-length tcrP gene (1,197-bp open reading frame, which translates 399 amino acids) has been reported to be 45.2 kDa.

FIG. 2.

SDS-PAGE and immunoblot (Ib.) analyses of spherule (Sph.) protein fractions derived from cell homogenates. Protein separations (left panels) of the cytosolic fraction (Cyt.; 5 μg), wall pellet (W.P.; 5 μg), and membrane pellet (M.P.; 5 μg) were reacted on immunoblots (matching right panels) with antiserum raised against the 48-kDa rTCRP. The purified 48-kDa recombinant protein (R.P.; 5 μg) and immunoblot are shown in separate lanes.

To confirm that the TCRP was present in the cytoplasm of in vitro-grown spherules, the same murine anti-rTCRP serum was used for immunolocalization studies of sectioned cells. When the antiserum was reacted with cryofixed, resin-embedded sections of parasitic cells, the labeled antibody was localized to the cytoplasm (Fig. 3A and C). Endospores and presegmented spherules (8) showed a high density of cytoplasmic label. The preimmune murine control serum did not react with the tissue sections (Fig. 3B and D). The guinea pig anti-6.5-kDa antibody was also shown to bind to the cytoplasm of parasitic cells but with lower affinity than the mouse antibody raised to the full-length rTCRP.

FIG. 3.

Immunolocalization of the TCRP in parasitic cells of C. immitis by using the anti-48-kDa rTCRP primary antibody and secondary antibody conjugated with FITC (A) or colloidal gold (C) for immunofluorescence or immunoelectron microscopy, respectively. Control sections (B and D) were reacted with preimmune serum and conjugated secondary antibody. Bars, 30 μm (A) and 1 μm (C). Abbreviations: en, endospores; pm, plasma membrane; sw, spherule wall.

ELISA.

Serum from all patients tested with confirmed coccidioidal infection showed antibody binding to the rTCRP in the indirect ELISA (Table 1). The selected patient sera revealed a range of antibody titers to the serodiagnostic, CF antigen based on ID-CF assays. The high and low titers to the CF antigen correlated with high and low antibody reactivity with the rTCRP based on optical densities in the ELISA. Control patient sera showed no reactivity with the rTCRP in the ELISA.

TABLE 1.

Human serum reactivity with the rTCRP in ELISA

Patient no.	CF antibody titera	Mean ODb ± SD
1	1:512	0.943 ± 0.053
2	1:256	0.731 ± 0.088
14	1:256	0.687 ± 0.054
21	1:128	0.280 ± 0.014
5	1:64	0.177 ± 0.062
12	1:16	0.143 ± 0.077
26c	0	0.037 ± 0.005
27c	0	0.061 ± 0.008

^aDetermined by ID-CF assays (14) with a standard C. immitis CF reference antigen (6). 

^bOD, optical density at 490 nm. 

^cControls. 

Antigenicity of the rTCRP in vitro.

The rTCRP was initially identified as a β-galactosidase fusion peptide which stimulated a murine T-cell line specific for the SCWF of C. immitis (17). The 48-kDa rTCRP was also able to stimulate this T-cell line (Fig. 4). The response to rTCRP was slightly higher than that to the crude mycelial culture F+L fraction but not as high as that to SCWF.

FIG. 4.

Proliferative response of the C. immitis antigen-specific T-cell line after incubation with the multicomponent mycelial culture F+L antigen, the purified rTCRP, and the SCWF against which the cell line was selected. Each antigen was tested at 10 μg (dry weight)/ml of culture medium. The data are expressed as the mean counts per minute above background. The mean values and standard deviations (SD) of three separate determinations are shown. The background (cpm in the absence of antigen) was 2,532 cpm (SD, ±541).

To determine whether the rTCRP could stimulate lymph node T cells from C. immitis-immune animals, BALB/c mice were immunized three times with formalin-fixed spherules or with adjuvant alone as described in Materials and Methods. The results of a typical T-cell proliferation assay are shown in Fig. 5. The rTCRP is antigenic for immune but not control T cells. Analysis of the phenotype of cells obtained from mice immunized with formalin-fixed spherules and stimulated with the rTCRP antigen was conducted by flow microfluorimetry analysis. Cells were harvested from cultures after 72 h of incubation with rTCRP and then stained with MAbs raised to specific cell surface markers. As expected, essentially none of the blast cells expressed B-cell markers. Almost all the rTCRP-stimulated blast cells expressed Thy1 (98%) and the αβTCR (89%), while 66% expressed CD4 and 45% expressed CD8.

FIG. 5.

Reactivity of lymph node cells derived from BALB/c mice immunized with formalin-fixed spherules after incubation with increasing concentrations of the purified rTCRP (○), compared to the response of cells from nonimmune mice incubated under the same conditions (•). The mean values above background and SD of three separate determinations are shown. The background for T cells from immune mice was 811 cpm (SD, ±505); the background for nonimmune mice was 295 cpm (SD, ±198).

The culture supernatants of lymph node-derived cells of immune mice (Fig. 5) which had proliferated in vitro either in response to C. immitis antigens or to a control mitogen (concanavalin A) were assayed for levels of IFN-γ production. Supernatants from cultures of immune cells not exposed to immunogens were also tested. The results (Table 2) reveal that the purified rTCRP stimulated immune cells to produce IFN-γ at levels which were slightly above that of T cells incubated with the crude C. immitis antigen (F+L). The negative and positive controls yielded the predicted results.

TABLE 2.

IFN-γ production by blast cells in response to antigen^a

Antigen	Amt of IFN-γ (pg/ml) on:
	Day 2	Day 4
None	<50	<50
F + L	739 (±30)	2,004 (±42)
rTCRP	885 (±43)	2,476 (±99)
ConA	3,897 (±350)	NDb

^aThe concentration of each antigen was 10 μg/ml of blast cell medium, while the concentration of concanavalin A (ConA) was 5 μg/ml. 

^bND, not done. 

Vaccination experiments.

To determine the effect of rTCRP on the course of coccidioidomycosis in BALB/c mice, three immunization and challenge experiments were conducted. A recombinant protein immunization method that elicited a strong T-cell proliferative response to the rTCRP was used (data not shown). The results of a representative experiment are shown in Fig. 6. BALB/c mice were immunized with the rTCRP and challenged with 50 viable arthroconidia by the i.p. route as described in Materials and Methods. At 14 days postchallenge, the surviving mice were sacrificed and the number of CFU/lung was determined by quantitative culture. The median number of CFU/lung was 3.7 (log₁₀) in mice immunized with the rTCRP compared to a median of 5.1 (log₁₀) in the controls; the difference is statistically significant (P = 0.037, Mann-Whitney U test). Similar results were seen in a second experiment in BALB/c mice; the median number of organisms in the lungs of rTCRP-immune mice was 3.45 (log₁₀) compared to 5.18 (log₁₀) in the controls. The results of the second experiment did not achieve statistical significance (P = 0.14, Mann-Whitney U test). The level of protection established by immunization with the crude SOW antigen was more striking. Efforts to identify the T-cell-reactive components of SOW are under way (27).

FIG. 6.

Representative box plot of CFU of C. immitis detected in dilution plate cultures per lung homogenate obtained from infected BALB/c mice immunized with antigen (rTCRP or SOW) or immunized with adjuvant alone (CONTROL). The boxes indicate the 25th and 75th percentiles, and the bars show the 5th and 95th percentiles. The line within the box indicates the median. Each symbol represents a single animal; the solid symbols represent dead mice.

Protection experiments were also conducted using C3H/HeN mice. The immunization protocol and route of inoculation were the same as above, but the inoculum dose was increased to 2000 viable arthroconidia. The results of a representative protection experiment with this mouse strain are shown in Fig. 7. The reduction of the fungal burden in lungs of rTCRP-immunized C3H/HeN mice compared to the controls is statistically significant (P = 0.033, Mann-Whitney U test), and the data obtained from these studies are comparable to the results of protection experiments with the BALB/c strain.

FIG. 7.

Representative box plot of CFU of C. immitis detected in dilution plate cultures per lung homogenate obtained from infected C3H mice immunized with antigen (rTCRP or SOW) or immunized with adjuvant alone (CONTROL). The boxes indicate the 25th and 75th percentiles, and the bars show the 5th and 95th percentiles. The line within the box indicates the median. Each symbol represents a single animal; the solid symbols represent dead mice.

DISCUSSION

We have described the T-cell reactivity of a C. immitis recombinant protein which was originally identified as a 6.5-kDa fusion peptide in a cDNA expression library of the saprobic phase of the pathogen. The fusion peptide tested positive in a T-cell immunoblot analysis (4, 17). Antibody raised to the purified recombinant peptide was used in immunofluorescence studies, which showed that the antigen was localized primarily in the cell wall of arthroconidia (4). In previous studies, a 59-kDa band identified in SDS-PAGE separations of the mycelial culture filtrate was also shown by immunoblot analysis to react with the same antibody (4). However, an internal amino acid sequence of the purified polypeptide failed to match the predicted amino acid sequence of the TCRP (28), suggesting that the antibody bound nonspecifically to the mycelial culture filtrate component.

To identify the full-length gene coding for the TCRP, the 0.18-kb partial cDNA clone was used to screen a genomic library of C. immitis derived from strain C735 for isolation of the full-length gene (28). The predicted 45.2-kDa protein has a calculated pI of 5.1, is generally hydrophilic, and contains no export signal sequence or membrane-spanning domain. The predicted amino acid sequence, when compared to the protein database, revealed 50% identity and 70% similarity to a mammalian 4-hydroxyphenylpyruvate dioxygenase (HPPD) protein (24). HPPD converts 4-hydroxyphenylpyruvate to homogenisate, a step in the major pathway of degradation of phenylalanine and tyrosine. The C. immitis protein sequence also showed similarity (approximately 60%) to bacterial homologs of HPPD. The Shewanella cowelliana protein was suggested to be involved in pigment formation, since expression of the bacterial gene in E. coli led to secretion of a dark brown pigment in vitro (11). When the full-length C. immitis gene was originally expressed in E. coli BL21(DE3) transformed with pET-21d-tcrP and grown overnight in Luria broth at 37°C, a deep brown pigment was formed (28). The chemical nature of the pigment was not characterized. The original pET-21d-tcrP construct lacked a His tag, and the recombinant protein was isolated by SDS-PAGE gel separation and electroelution of the specific band (28). Expression of the 1.1-kb partial tcrP gene in the present study, using E. coli HMS174(DE3) transformed with pET 28b-tcrP, showed no visible brown pigment released by the bacterial culture. The C. immitis protein expressed in pET 28b-tcrP differed from the full-length gene product by deletion of 18 residues at the C terminus. The truncation of tcrP was engineered for convenience in subcloning the gene into the pET28b vector, which expressed a His tag upstream from the N terminus that was used in affinity isolation of the recombinant protein. Preliminary immunogenicity studies of the full-length gene product expressed by pET21d-tcrP showed no difference in the levels of T-cell reactivity in cellular immunoassays compared to the truncated gene product.

The TCRP expressed during the parasitic cycle of C. immitis is localized in the cytoplasm. However, patients with coccidioidomycosis have antibody to the protein and patients with disseminated disease show the highest titers of antibody to the TCRP. These data suggest that the native TCRP is presented to the host during infection and the host makes an antibody response. TCRP is apparently released from the parasitic cells. Although the protein lacks a signal peptide (28), the native 50.4-kDa TCRP has been detected by immunoblot analysis of SDS-PAGE separations of the culture filtrate of endosporulating spherules (not shown). The antigen may be present in the residual cytoplasm of mature spherules, which is not compartmentalized during endospore differentiation, and released at the time of spherule rupture (8).

The apparent molecular mass of the native TCRP is 50.4 kDa which is approximately 5,000 Da larger than the mass predicted by the deduced amino acid sequence (28). Although the translated tcrP gene contains one predicted N-glycosylation site in the N-terminal region, localization of the protein to the cytosol would suggest that it is not glycosylated. We have no explanation for the difference in molecular size of the predicted and native TCRP based on the data at hand.

The SCWF-specific T-cell line (17) proliferated in response to rTCRP, though at a lower level than in response to whole SCWF. These data suggest that TCRP is only one of several T-cell-stimulatory components in the SCWF mixture. BALB/c mice immunized with formalin-fixed spherules show a good cellular response to the rTCRP in the lymph node proliferation assays. The phenotype analysis of the responsive cells indicated that they are primarily CD4 T cells. We measured IFN-γ production by the proliferating cells in the T-cell assays. The rTCRP stimulated production of the cytokine at a level comparable to that in cells exposed to the crude, multicomponent C. immitis antigen. We conclude that the TCRP is an immunogenic component of C. immitis spherules which elicits a Th1 response in the immunized host.

The immunoprotective capacity of the rTCRP was determined by using the same immunization schedule that elicited a strong T-cell proliferation response. The protocol was slightly different from that used to immunize mice with formalin-fixed spherules for T-cell proliferation assays. Since spherules are particulate, no adjuvant was used for the first two spherule immunizations but CFA was used for the final immunization. Mice which demonstrate genetic susceptibility or intermediate resistance to C. immitis infection were used to stimulate the range of susceptibility and resistance within the human population. An inoculum which was known to cause lethal disease in the respective strains of mice was used in each case (15). We used the i.p. route of infection because we have extensive experience with this method and have shown that it is highly reproducible. Furthermore, mice do not begin to die from i.p. challenge with arthroconidia for at least 10 to 14 days postinoculation, which should permit the specific T-cell immune response to control the spread of the pathogen. On the other hand, i.p. inoculation is not as rigorous a test of potential vaccine candidates as intranasal challenge (23). In our experiments we found that immunization of BALB/c and C3H/HeN mice with the rTCRP had a modest protective effect on the course of coccidioidal infection. In three separate experiments, the immunized mice had about 10-fold fewer organisms in their lungs than did the control animals. Immunization with the SOW multicomponent antigen (6) resulted in a 100-fold decrease in the number of CFU/lung, and immunization with formalin-fixed spherules (22) led to sterile lungs. Therefore, the immunoprotective capacity of the rTCRP appears to be limited. In addition, since C. immitis TCRP is 50% identical to a mammalian homolog (25), there is a theoretical possibility of eliciting a potentially harmful autoimmune response by TCRP immunization. However, there are several caveats. We immunized mice in a way that elicited a good T-cell proliferative response, but it may not have been optimal. Variables such as antigen dose, timing of immunization, and adjuvants used can have profound effects on the success or failure of immunization (13). Adjuvants such as recombinant interleukin-12 (1) may significantly improve the protective effect of immunization with the rTCRP.

We maintain that development of a vaccine against coccidioidomycosis is a feasible objective. In addition to the support of this goal in the literature (19, 23), the experiments shown here prove that immunization with the rTCRP can confer some protection to genetically susceptible mice. Cytokine profile studies of coccidioidomycosis-susceptible BALB/c and resistant DBA2/J mice have supported the earlier contention that Th1 responses primarily account for immunoprotection against disease (20). It is essential, therefore, to ensure that the potential vaccine candidates elicit a Th1-type response. However, development of an effective vaccine is a formidable task for several reasons. First, since the protective immune response in coccidioidomycosis is T-cell mediated, a more elaborate screening assay than would be required for the evaluation of a humoral response is necessary. Second, it is impossible to predict which parasitic cell proteins are most likely to elicit a protective response on the basis of the results of in vitro T-cell assays, since the latter do not necessarily correlate with the outcome of immunoprotection experiments. Finally, it is not clear whether a single protein or multiple proteins will be required for a protective immune response. The ideal vaccine against coccidioidal infections may be a pan-Coccidioides vaccine including several molecules expressed during different stages of the parasitic cycle and conserved among different strains. Characterization of the protective activity of the rTCRP is a promising first step in the development of such a multicomponent recombinant vaccine for coccidioidomycosis.