The Cryptococcus neoformans Gene DHA1 Encodes an Antigen That Elicits a Delayed-Type Hypersensitivity Reaction in Immune Mice

Abstract

When mice are vaccinated with a culture filtrate from Cryptococcus neoformans (CneF), they mount a protective cell-mediated immune response as detected by dermal delayed-type hypersensitivity (DTH) to CneF. We have identified a gene (DHA1) whose product accounts at least in part for the DTH reactivity. Using an acapsular mutant (Cap-67) of C. neoformans strain B3501, we prepared a culture filtrate (CneF-Cap67) similar to that used for preparing the commonly used skin test antigen made with C. neoformans 184A (CneF-184A). CneF-Cap67 elicited DTH in mice immunized with CneF-184A. Deglycosylation of CneF-Cap67 did not diminish its DTH activity. Furthermore, size separation by either chromatography or differential centrifugation identified the major DTH activity of CneF-Cap67 to be present in fractions that contained proteins of approximately 19 to 20 kDa. Using N-terminal and internal amino acid sequences derived from the 20-kDa band, oligonucleotide primers were designed, two of which produced a 776-bp amplimer by reverse transcription-PCR (RT-PCR) using RNA from Cap-67 to prepare cDNA for the template. The amplimer was used as a probe to isolate clones containing the full-length DHA1 gene from a phage genomic library prepared from strain B3501. The full-length cDNA was obtained by 5′ rapid amplification of cDNA ends and RT-PCR. Analysis of DHA1 revealed a similarity between the deduced open reading frame and that of a developmentally regulated gene from Lentinus edodes (shiitake mushroom) associated with fruiting-body formation. Also, the gene product contained several amino acid sequences identical to those determined biochemically from the purified 20-kDa peptide encoded by DHA1. Recombinant DHA1 protein expressed in Escherichia coli was shown to elicit DTH reactions similar to those elicited by CneF-Cap67 in mice immunized against C. neoformans. Thus, DHA1 is the first gene to be cloned from C. neoformans whose product has been shown to possess immunologic activity.

Cryptococcus neoformans is a human fungal pathogen that typically affects patients with underlying T-cell deficiencies such as occur with AIDS or during immunosuppressing therapy (2, 10, 11). This predilection for immunosuppressed patients suggests that normal people may resist infection with C. neoformans by competent immune responses to antigenic stimulation. There is considerable experimental support for this concept from the murine model of cryptococcosis (3, 6). For example, vaccination with a C. neoformans culture filtrate antigen (CneF) in complete Freund's adjuvant (CFA) induces protection against a subsequent cryptococcal infection as well as dermal delayed-type hypersensitivity (DTH) response(s) to cryptococcal antigen (17). CneF can be used in vitro to detect anticryptococcal cell-mediated immune (CMI) responses of lymphocytes from Cryptococcus-infected patients (7).

CneF has been separated into three main components: glucuronoxylomannan, galactoxylomannan, and mannoprotein (15). The MP fraction is the primary component recognized by the anticryptococcal CMI response in mice (15). It is presumed that the MP fraction also induces the specific CMI response detected by CneF and that other components in CneF may play a modulatory role. However, little is known about the exact antigen or antigens which elicit the anticryptococcal CMI response.

Here we report that CneF prepared from an acapsular mutant of C. neoformans stimulates a DTH response in mice immunized with antigen prepared from a weakly virulent encapsulated strain of C. neoformans. Biochemical fractionation of CneF from the acapsular mutant further localized DTH reactivity to a specific fraction. Furthermore, we have cloned a gene encoding a putative protein from this fraction, and its expression product has shown DTH reactivity in mice analogous to the native antigen in CneF.

MATERIALS AND METHODS

Fungal strains.

C. neoformans strains 184A (serotype A), described by Murphy and Cozad (12), and Cap-67, an acapsular mutant derived from strain B-3501 (serotype D), described by Jacobson et al. (8), were used in this work. Cap-67 was selected for our studies because it does not produce the high-molecular-weight polysaccharide glucuronoxylomannan, thus simplifying protein purification.

Preparation of CneF from different strains.

A C. neoformans filtrate from the weakly virulent encapsulated strain 184A (CneF-184A) was prepared by growing the culture in a chemically defined liquid medium for 5 days as described previously (1). After the culture was treated with 2% formalin, the C. neoformans cells were removed by filtration. The supernatant was subjected to a tangential-flow system to exclude all molecules smaller than 30 kDa (Pellicon OM-141; Millipore, Bedford, Mass.). The retentate was washed with 10 volumes of physiological saline solution, concentrated 10-fold, filter sterilized, and stored at −20°C. Each preparation of CneF-184A was standardized against the previous lot of CneF-184A for DTH reactivity in mice immunized with the previous lot. CneF-184A has been used as a standard skin test antigen for C. neoformans for the past 20 years (7, 13, 15, 16). The lot of CneF-184A used in these studies had a protein concentration of 0.171 mg/ml as determined by the bicinchoninic acid procedure (Pierce Chemical Co., Rockford, Ill.) and a carbohydrate concentration of 6.9 mg/ml as determined by the phenol-sulfuric acid procedure (4).

The CneF from Cap-67 (CneF-Cap67) was prepared in a manner similar to that for CneF-184A, except that the culture was grown in a different defined medium consisting of 7.6 mM asparagine, 2 mM MgSO₄ · 7H₂O, 22 mM KH₂PO₄, 150 mM glucose, 3 mM thiamine, 18 mM CaCl₂ · 2H₂O, and 13 mM ZnSO₄ · 7H₂O. The retentate was collected and concentrated using a Minitan apparatus (Amicon, Beverly, Mass.) with a molecular mass exclusion of 10 kDa.

Antigen fractionation.

CneF-Cap67 was lyophilized and chemically deglycosylated with anhydrous hydrogen fluoride as described by Shively (19). The resulting material was dialyzed sequentially against 100% pyridine, 10% pyridine, and finally 10% acetic acid. The retentate was then lyophilized and resuspended in water. Insoluble material was removed by centrifugation at 10,000 × g for 5 min. The supernatant from this step, referred to as HF-CneF-Cap67, was tested directly for its ability to elicit a DTH reaction in immunized mice or was purified further as indicated below. In comparison to CneF-Cap67, the carbohydrate content of HF-CneF-Cap67 was reduced by more than 95% as shown by the phenol-sulfuric assay for carbohydrates, staining of native polyacrylamide gels with periodic acid-Schiff reagent, and development of a lectin blot with concanavalin A (data not shown).

The HF-CneF-Cap67 was fractionated by chromatography on Sephadex G-75 Superfine (Pharmacia, Piscataway, N.J.). Based on analysis of these fractions as detailed in Results, larger quantities of proteins with a molecular mass range between 10 and 30 kDa were isolated by differential centrifugal ultrafiltration (Centricon 30 and Centricon 10 [Amicon]) for amino acid sequencing.

Amino acid sequencing.

Protein preparations from differential centrifugation were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and blotted to polyvinylidene difluoride, and individual bands were excised for amino acid sequencing. N-terminal sequencing and total amino acid composition measurement were performed at the University of Arizona Macromolecular Structure Facility (Tucson, Az.). Digestion with trypsin or Lys-C and internal sequencing of high-performance liquid chromotography (HPLC) fractions were performed at the Harvard University Microchemistry Facility (Cambridge, Mass.).

DNA isolation.

C. neoformans Cap-67 was grown for 48 h, and cells were recovered by centrifugation at 3,000 × g for 10 min and stored in 200-μl aliquots in screw-cap 2.0-ml microcentrifuge vials at −80°C. Prior to use, samples were thawed on ice, 1.2 ml of 0.5 mM glass beads was added, and the tubes were completely filled with lysis buffer (100 mM Tris-HCl [pH 7.5], 50 mM EDTA [pH 8.0], 1% SDS, 500 mM NaCl). The cells were disrupted by shaking the vials in a minibead beater (Biospec Products, Bartlesville, Okla.) at 4°C for three 20-s pulses at 80% of maximum speed separated by 1-min pauses and then incubated for 20 min at 68°C. The homogenate was extracted with an equal volume of phenol-chloroform-isoamyl alcohol (25:24:1). The aqueous supernatants were extracted twice with 1 volume of chloroform-isoamyl alcohol (24:1), and the DNA was precipitated with ethanol. The resulting DNA was washed with 70% ethanol, dried at room temperature, and resuspended in 100 μl of Tris-EDTA (TE) (pH 8.0).

RNA isolation.

Cultures of Cap-67 were grown under the same conditions that were used to prepare CneF-Cap67 and were stored at −80°C in 2.0-ml microcentrifuge vials prior to extraction. Aliquots were thawed on ice, and 840 μl of phenol-chloroform-isoamyl alcohol (25:24:1) and 660 μl of extraction buffer (50 mM Tris-HCl [pH 7.5], 100 mM LiCl, 5 mM EDTA [pH 8.0], 1% [wt/vol] SDS, 1% β-mercaptoethanol) were added to each vial. The tubes were then completely filled with 0.5-mm-diameter glass beads. The vials were agitated in the minibead beater at 4°C five times for 20 s at 80% of maximum speed separated by 20-s rests. After this, the vials were centrifuged for 5 min at full speed in a microcentrifuge. The supernatant was extracted with 840 μl of phenol-chloroform-isoamyl alcohol and subsequently with chloroform. The RNA was precipitated in 2 M (final concentration) LiCl, washed with 70% ethanol, air dried, and resuspended in 25 μl of diethylpyrocarbonate-treated water.

cDNA synthesis.

First-strand cDNA was synthesized using 500 ng of Cap-67 RNA, 500 ng of oligo(dT)_12–18, and the SuperScript II RNase H⁻ reverse transcriptase kit (Life Technologies, Gaithersburg, Md.) as specified by the manufacturer. After treatment with 2 U of RNase H (Life Technologies), 1 μl of first-strand cDNA was used to synthesize the second strand with Taq polymerase (Roche Molecular Biochemicals, Indianapolis, Ind.) by standard procedures (18) in a GeneAmp PCR System 2400 (Perkin-Elmer, Foster City, Calif.). Based on N-terminal and internal amino acid sequences, as described in Results, 14 degenerate oligonucleotide primers were designed and synthesized. For experiments with various combinations of these primers, 0.1 nmol of each degenerate oligonucleotide was used under the following PCR conditions: one cycle of 30 s at 94°C, 35 cycles of 20 s at 94°C, 30 s at 55°C, and 90 s at 72°C, followed by a final extension cycle of 5 min at 72°C. The PCR product from one pair (OAM51 [ACIACIGCIGCIAAYGGIGCIGC] and OAM46 [ACICCIGGIARIGCIGTRCARTCIAC]) was cloned using the TA cloning kit (Invitrogen, Carlsbad, Calif.) as specified by the manufacturer and transformed into Top10 cells (Invitrogen), generating clone pAM830.

The 5′ rapid amplification of cDNA ends system (Life Technologies) was performed as specified by the manufacturer, using 500 ng of Cap-67 RNA to obtain the 5′ end of the cDNA not present in pAM830. The oligonucleotide primers used in this procedure were as follows: the gene-specific primer 1 (GSP1) was OAM63 (ACCGGACACAAGACCAGA), GSP2 was OAM59 (AGTTGTTGCCGAGGACGATTG), and GSP3 was OAM51.

Cloning procedures and sequence analysis.

A λ phage genomic library created from C. neoformans strain B3501 (AIDS Research and Reference Reagent Program: Cryptococcus neoformans Genomic Library from Jeffrey Edman) was screened with the cDNA insert from pAM830 as a probe. DNAs from hybridizing clones were excised from the λ ZAP vector by the recommended procedure (Stratagene, La Jolla, Calif.).

Both strands of all clones were sequenced at the Laboratory of Molecular Systematics and Evolution of the University of Arizona by using the Applied Biosystems (ABI) PRISM dye terminator kit (Perkin-Elmer) and an ABI model 373 DNA sequencer. Sequence analyses were performed using the GCG sequence analysis software package (Genetics Computer Group Inc., Madison, Wis.), and protein comparisons were done through the National Center for Biotechnology Information (Bethesda, Md.) using the BLAST algorithm.

A C. neoformans strain B3501 λ phage cDNA library (AIDS Research and Reference Reagent Program: Cryptococcus neoformans cDNA Library) from Jeffrey Edman was used for some experiments as indicated.

Protein expression.

cDNAs carrying the complete open reading frame (ORF) were synthesized by reverse transcription-PCR (RT-PCR) amplification as described above, using oligonucleotide primers OAM65 (GGGGAATTCGCCATGTTCTCGTCCACT) and OAM66 (GGGGAATTCATTTTACAGCTGGAGAGT). The following PCR conditions were used: one cycle of 30 s at 94°C, 35 cycles of 20 s at 94°C, 30 s at 60°C, and 90 s at 72°C, and a last cycle of 10 min at 72°C. The amplified band was cloned and subsequently subcloned into the EcoRI site of the expression vector pET32a (Novagen, Madison, Wis.) to generate pAM888, which was used to transform AD494 (DE3) pLysS bacterial cells.

Protein expression was performed by standard methods after induction with 1 mM isopropyl-β-d-thiogalactopyranoside IPTG for 3 h. Protein samples were subjected to SDS-PAGE (8 and 10% polyacrylamide gels) and stained with Coomassie blue or transferred to nylon membranes (MSI, Westborough, Mass.) using a Trans-Blot SD semidry electrophoretic transfer cell (Bio-Rad, Hercules, Calif.). Fusion proteins were detected by Western blot hybridization using the S-Tag Western blot kit (Novagen).

Protein purification was performed using the His-Bind resin and purification kit (Novagen), as specified by the manufacturer, under denaturing conditions (6 M guanidine-HCl). After elution with 300 mM imidazole, proteins were concentrated using Centricon 30 and 10 concentrators (Amicon) for the fusion construct and control vector alone, respectively. The fusion protein was partially renatured by sequential washes of the protein concentrate in the Centricon column with 4, 2, 1, and 0.5 M guanidine-HCl.

Induction and elicitation of an anticryptococcal DTH response.

CBA/J mice were immunized subcutaneously with 0.1 ml of CneF-184A emulsified in an equal part of CFA at each of two sites at the base of the tail. Control mice were injected following the same procedure with sterile physiological saline emulsified in CFA (14). Six days after immunization, mouse footpads were injected with CneF-184A, other antigen preparations, or saline and footpad swelling was measured 24 h later as previously described (14). Swelling is proportional to the level of DTH reactivity to the test antigen (14). Typically, five animals per group were used for assessment of levels of DTH reactivity and for controls. Means, standard errors of the means, and unpaired Student's t test results were used to analyze the footpad data. Differences between groups with P values of 0.05 or less were taken as significant.

Nucleotide sequence accession number. Sequence data from pAM850 has been deposited with GenBank (accession number GI 266842).

RESULTS

Characterization of DTH activity in CneF from Cap-67 and its fractions.

An extract of the culture supernatant from the acapsular strain Cap67 (CneF-Cap67) was prepared in a similar manner to CneF-184A, and the two were compared for their ability to stimulate an anticryptococcal DTH response (Fig. 1). In mice previously immunized with CneF-184A emulsified in CFA, CneF-Cap67 elicited a significant level of DTH reactivity as measured by the footpad-swelling response. The magnitude of the DTH reaction to CneF-Cap67 was about half of that induced by CneF-184A. Also shown in Fig. 1, deglycosylation of CneF-Cap67 with hydrogen fluoride (HF-CneF-Cap67) did not reduce the DTH response.

FIG. 1.

DTH reactions induced by a culture filtrate antigen prepared from acapsular C. neoformans-isolated Cap-67. Mice were immunized with CneF-184A-CFA or injected with sterile physiological saline (SPSS)-CFA at two sites subcutaneously at the base of the tail 6 days before being footpad tested with CneF-184A, CneF-Cap67 (Cap 67), or HF-treated CneF-Cap67 (HF Cap 67). Five mice were used per group. The experiment was repeated twice with similar results. Vertical lines represent the standard error of the mean.

The HF-CneF-Cap67 was fractionated chromatographically by molecular weight. Following the flowthrough (fraction 1), four peaks emerged, and the fractions corresponding to each peak were pooled and concentrated. These fractions were labeled fraction 2/3 (two overlapping peaks were combined), fraction 4, and fraction 5 (Fig. 2). SDS-PAGE of fraction 4 suggested that it was composed predominantly of bands also present in the two flanking peaks, and therefore it was not analyzed further.

FIG. 2.

Gel filtration with Sephadex G-75 of HF-treated CneF from the acapsular strain Cap-67. Bars indicate total protein for each fraction. The regression line was calculated from molecular weight markers (molecular weight is shown in thousands on the right-hand axis).

Fractions 1, 2/3, and 5 were tested for their abilities to elicit a DTH response in mice immunized with CneF-184A in CFA. As shown in Fig. 3, the flowthrough (fraction 1) did not elicit an anticryptococcal DTH reaction, fraction 2/3 stimulated a marginal level of DTH reactivity, and fraction 5 elicited footpad swelling which was nearly half of that elicited by the reference CneF-184A (Fig. 3). The level of DTH elicited by fraction 5 was comparable to that elicited by unfractionated HF-CneF-Cap67 (Fig. 1).

FIG. 3.

DTH reactions elicited in CneF-CFA-immunized mice with fractions 1, 2/3, or 5 obtained from the fractionation of HF-CneF-Cap67 on a Sephadex G-75 column. Mice were immunized or injected with saline-CFA as indicated in the legend to Fig. 1. Footpad testing was done with CneF-184A as a control or with the designated antigen fraction.

Molecular mass standards, which were run in the same column, indicated that fraction 5 contained proteins ranging from 25 to less than 13 kDa. An analogous fraction restricted to the molecular mass range between 10 and 30 kDa was prepared by differential centrifugation separation of HF-CneF-Cap67 (see Materials and Methods). When this fraction, referred to as fraction B, was injected into the hind footpads of immune mice at 1 μg of protein per pad, it elicited a marginal response, but when it was injected at 3 μg of protein per pad, it elicited a response equivalent to the response elicited by CneF-Cap67 or HF-CneF-Cap67 (Fig. 4).

FIG. 4.

DTH reactions elicited in CneF-CFA-immunized mice with fractions derived from HF-CneF-Cap67 by differential centrifugation using Centricon 10 and Centricon 30 centrifugal filter devices. Mice were immunized or injected with saline-CFA as indicated in the legend to Fig. 1. Footpad testing was done with CneF-184A as a control or with the 10-30 kDa component at two different protein concentrations.

Protein sequencing.

Fraction 5 from HF-CneF-Cap67 contained a prominent protein doublet migrating at approximately 20 kDa in SDS-PAGE. The two bands were excised from SDS-PAGE gels for detailed analysis. The total amino acid compositions of the proteins in the two bands were very similar (data not shown), and the first 20 amino acids of their N-terminal sequences (DTPYLGSVLTTAANGAASVS) were identical. The N-terminal amino acid was not methionine, suggesting that the full-length gene product may be a signal peptide or other leader sequence that was absent in the SDS-PAGE-purified protein.

The more abundant and slower-migrating protein of the pair was subjected to proteolytic digestion, and the HPLC-purified fragments were submitted for amino acid sequencing. Digestion with trypsin yielded several small fragments, none of which resulted in sequences longer than 4 amino acids. However, digestion with Lys-C produced longer peptides. Four of the HPLC peaks from the Lys-C digestion (identified as peaks 56, 62, 64, and 65) ranged from 6 to 18 amino acids, and their sequences could be superimposed. Also, two other HPLC peaks (identified as peaks 11 and 16) were very similar to each other in their amino acid sequences. The biochemically determined sequences from the N terminus, peak 62, peak 11, and an additional HPLC peak (peak 15) are shown in Fig. 5 in relation to the deduced sequence from the DHA1 gene, which was cloned as described below.

FIG. 5.

Complete cDNA with flanking genomic sequences and amino acid sequence of DHA1. The boxed sequence indicates a putative signal peptide, and the inverted triangles indicate introns of 60, 55, 61, and 53 bp from 5′ to 3′, respectively. Downstream of the stop codon is a polyadenylation signal (as indicated). The arrows indicate the beginning and the end of the RT-PCR clone, and the sequences upstream and downstream were obtained from genomic clones. Underlined are the sequences of the two oligonucleotides used for the RT-PCR. Underlined peptide sequences match those obtained from sequencing the amino terminus and Lys-C peptides. X indicates an amino acid for which biochemical analysis could not distinguish between S, T, V, and G. The dash indicates that no amino acid was identified by biochemical sequencing, and the bold N indicates a glycosylated asparagine, both by biochemically determined molecular weight and by the consensus sequence NASE.

Cloning and expression of the putative DTH-eliciting antigen.

From the N-terminal sequence and the two groups of similar internal sequences (HPLC peaks 56, 62, 64, and 65 and HPLC peaks 11 and 16), degenerate oligonucleotides were designed and used as primers for RT-PCR amplification. Only the combination of oligonucleotides OAM51 (from the N-terminal sequence) and OAM46 (from peaks 56, 62, 64, and 65) produced a product of 776 bp. This was cloned to generate plasmid pAM830. Sequencing of this clone confirmed that it contained the coding sequences for both primers. In addition, the deduced amino acid sequence of pAM830 contained a segment of amino acids which corresponded to the peptides of HPLC peaks 11 and 62 (Fig. 5). A C. neoformans strain B3501 λ phage genomic library was screened with pAM830, and 12 positive clones were obtained. From one of these clones, a HindIII fragment, of approximately 4 kb, was subcloned to form plasmid pAM850. The fragment was completely sequenced and was found to contain the entire gene, which we have termed DHA1 for “delayed hypersensitivity antigen 1.”

Subsequent studies were carried out to isolate the 5′ end of the DHA1 cDNA, which was missing from the RT-PCR product. Attempts to obtain clones from an available C. neoformans λ phage cDNA library prepared from strain B-3501 were unsuccessful. Therefore, 5′ RACE-PCR was performed using RNA isolated from Cap-67, and this provided the missing 5′ end of the transcript. With this information, a cDNA containing the complete DHA1 ORF was synthesized by RT-PCR with oligonucleotides 5′ of the ATG and 3′ of the stop codon of the putative ORF from genomic clone pAM850 (Fig. 5).

Comparison of the genomic and cDNA sequences showed the ORF to encode a protein of 327 amino acids with a mass of about 34 kDa. It has similarity to the product of a developmentally regulated gene, priA, from Lentinus edodes (shiitake mushroom), associated with fruiting-body formation (GenBank accession number X60956) (9). Of special note, there is extensive similarity between a “zinc cluster”-like motif located at the C terminus of PRIA and a 79-amino-acid sequence near the C terminus of DHA1 (Fig. 6). Additional sequence analyses indicate that there is a putative signal peptide sequence, four introns of 60, 55, 61, and 53 bp from 5′ to 3′, and a putative Kozak sequence. The deduced amino acid sequence contains a single motif for N-glycosylation (NASE) at amino acid 276 of the full-length protein. Furthermore, all peptide sequences obtained by either tryptic or Lys-C digestion can be found within the cDNA sequence, four of which are represented in Fig. 5. The matches between the biochemically sequenced and the deduced amino acid sequences demonstrate that the protein that had been biochemically purified from CneF-Cap67 is encoded by the plasmid pAM850.

FIG. 6.

Alignment of similar amino acid sequences from PRIA of L. edodes and DHA1 of C. neoformans. White lettering on a black background indicates identity; + indicates conserved amino acid substitutions.

DTH response in mice elicited by the expression product of the gene.

To test whether the protein encoded by DHA1 could elicit DTH, the cloned cDNA from pAM850 was subcloned into the bacterial expression vector pET32a to generate pAM890 and the recombinant fusion protein was purified from bacterial cultures as indicated in Materials and Methods. When the fusion protein was tested for DTH stimulation (Fig. 7), it elicited a DTH response analogous to that shown by CneF-Cap67 (Fig. 1) or HF-CneF-Cap67 (Fig. 1 and 4).

FIG. 7.

DTH reactions elicited in CneF-CFA-immunized mice with the DHA1 fusion protein (rDHA1). Mice were immunized or injected with saline-CFA as indicated in the legend to Fig. 1. Footpad testing was done with CneF-184A, with the expression of vector without the DHA1 insert (Vector) or with rDHA1.

DISCUSSION

Vaccination of mice with extracts of C. neoformans results in a CMI response that can be detected by subsequent antigenic stimulation to elicit DTH responses, as measured by footpad swelling (13). In the studies reported here, we have identified one gene whose product contributes to this process.

Our initial observations indicated that CneF from an acapsular mutant (Cap-67) of C. neoformans was able to stimulate a DTH response in mice vaccinated with CneF from the encapsulated strain, 184A. Deglycosylation of CneF-Cap67 with anhydrous hydrogen fluoride did not affect its ability to elicit a positive DTH response in mice immunized against C. neoformans. We included this deglycosylation step because it had been very helpful in separating protein antigens in extracts from another fungal pathogen, Coccidioides immitis (5). Although most of the carbohydrate is removed by anhydrous hydrogen fluoride, residual monosaccharides would be expected to be present after this treatment (19). By biochemical fractionation, the predominant DTH activity was localized to a pair of protein bands approximately 20 kDa in size. Since all biochemical analysis was carried out after deglycosylation, we are not sure of the apparent molecular weight of DHA1 prior to this step, and with glycosylation, it could be considerably larger. The total amino acid composition and N-terminal sequence determinations showed the two proteins to be virtually identical, and we presume that one is a partial degradation product of the other and that both are derived from the same gene. Using biochemically determined N-terminal and internal amino acid sequences, the gene corresponding to the protein, DHA1, has been cloned. The deduced sequence, excluding the signal peptide, is 305 amino acids and has a mass approximately 10 kDa greater than that estimated for the biochemically purified protein. Although we did not determine the C-terminal amino acid sequences of the biochemically purified protein, most if not all of the mature protein appears to be intact in the 20-kDa protein since the internal fragment peak 15 is located 18 amino acids from the C terminus (Fig. 5). Another amino acid sequence (peak 11) also corresponds to a portion of the deduced protein, further indicating that DHA1 encodes the protein associated with DTH activity in fraction 5 of CneF-Cap67.

There is a striking similarity between the deduced C. neoformans DHA1 polypeptide and that of another basidiomycete fungus, the PRIA protein of L. edodes (9). PRIA is a developmentally regulated protein expressed most prominently during early fruiting-body development in the shiitake mushroom. The DHA1 polypeptide is cysteine rich, with 22 Cys residues in the 305-residue mature protein (7.2%). Fifteen of these residues are positionally conserved between DHA1 and PRIA. The two polypeptides have a high degree of similarity in their C termini, with 37% identical residues and an additional 19% conservative changes over a 79-amino-acid region (Fig. 6). This region resembles the zinc cluster domain of metallothionines responsible for binding zinc and other heavy metals (20). Evidence suggesting that the PRIA polypeptide interacts with zinc comes from experiments in which it caused sensitivity to zinc and other heavy metals when expressed in Escherichia coli. Although the DHA1 protein is found in the C. neoformans extracellular filtrate, its biological function is unknown. The PRIA protein has not been localized yet. The structural similarities between DHA1 and PRIA suggest possible avenues for future work to define the role of DHA1 in C. neoformans.

Our findings indicate that the fusion protein resulting from expression of DHA1 does not induce footpad reactions in saline-CFA-injected control mice but does elicit DTH reactions in mice immunized with CneF-184A. In previous studies, mice vaccinated with CneF-184A in CFA not only responded to subsequent CneF-184A footpad challenge with a positive DTH response but also were protected against infection with C. neoformans (13). These observations raise the possibility that vaccination with the recombinant antigen of DHA1 may afford protection against a cryptococcal infection. However, it is not yet known whether the stimulus for DTH is the same which evokes protection. Clearly, this is an important question to pursue in future studies.

Our findings do not preclude the possibility that one or more other proteins may exist in encapsulated strains of C. neoformans which possess DTH-stimulating activity. Of note in our work, the magnitude of the DTH response elicited by either the purified or the recombinant purified protein was consistently no more than half that elicited by CneF-184A. In work by Murphy et al., the MP fraction of CneF-184A was identified as predominantly associated with the murine CMI response (15). SDS-PAGE analysis of CneF-184A indicates that the MP fraction is considerably larger than that identified in CneF-Cap67. Although our immunological experiments suggest that DHA1 exists in strain 184A, this has not yet been demonstrated. At least one glycosylation site exists in the ORF of DHA1. Glycosylation of the protein would increase its molecular weight and result in lower mobility in SDS-PAGE. It is also possible that other proteins or glycoproteins which are present in CneF-184A and that are unrelated to the antigen from DHA1 contribute to the anticryptococcal DTH reactivity. Further work is needed to clarify this issue as well.