INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Numerous studies have demonstrated the importance of the cellular immune (Th1) response in protection against Candida albicans infection, particularly at mucosal surfaces (17, 38-40). This response appears to be important in the prevention of recurrent vulvovaginal candidiasis (RVVC) (11, 17, 21). Studies of susceptibility to this type of Candida infection indicate a local impairment of cellular immunity to C. albicans, even though this immune response to the yeast is systemically intact (18-20).

The role of antibody to Candida antigens in conferring protection against mucosal infection is less certain. Anti-Candida antibody levels may be normal or even elevated in chronic mucocutaneous candidiasis (28), even though recurrent mucosal infections occur in this disorder. In patients with vaginitis, serum anti-Candida antibody levels are normal or elevated (4, 33). However, recent studies suggest a protective role for antibodies to certain Candida antigens in preventing vaginal infection (12).

In addition to antibody, T-cell-derived antigen binding molecules (TABM) specific for the immunogen appear in the serum during a humoral (Th2) immune response (48). TABM are nonimmunoglobulin immunoproteins (7, 8, 13, 49) that bind nonprocessed antigen but share epitopes with the T-cell receptor for antigen (TCR) and, in mice, have some amino acid sequence homology to TCR C and V (9, 10). They are secreted by T cells and are present in the serum in the 10- to 50-µg/ml range. TABM are thought to have an immunoregulatory function, particularly involving suppression of cellular immunity (41, 50). They are often associated with cytokines such as transforming growth factor  (TGF-) and interleukin-10 (IL-10) and may deliver these cytokines to sites where the antigen is localized (7, 8, 29).

Animal studies have demonstrated a role for suppressor factors and suppressor cells in susceptibility to Candida infection (15). Mannan-specific lymphocytes transfer the suppression of cellular immunity to recipients (23, 27). Witkin et al. demonstrated inhibition of Candida-induced lymphocyte proliferation by sera from women with RVVC (52). Mannan-specific T-cell-suppressive activity has also been detected in patients with mucocutaneous candidiasis (16, 22).

We reasoned that TABM may be the cause of the observed suppression of cellular immunity described in the preceding paragraph and may be implicated in RVVC. Since the Candida antigen associated with suppression appears to be the polysaccharide mannan, an assay was established to detect serum TABM specific to mannan. In this study we have measured immunoglobulin G (IgG) and TABM specific to whole Candida extract (Hollister-Stier) and to both cetyltrimethylammonium bromide (CTAB) and alkaline degradation (PEAT, a method of extraction described by Peat et al. [37] and in reference 34) mannans in women susceptible to vulvovaginal candidiasis and in controls. TABM specific to CTAB mannan was purified from a patient with a high titer of serum TABM to this antigen and studied for (i) the presence of associated cytokines, (ii) cross-reactivity with other yeasts and molds, and (iii) its effect on suppression of cell-mediated immunity to Candida, as assessed by gamma interferon (IFN-) production by peripheral blood mononuclear cells (PBMC) in response to Candida extract.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Patients. Seventeen patients with RVVC were studied. The patients had a history of at least three courses of treatment for RVVC over the preceding year and a positive isolation of C. albicans from vaginal cultures within the preceding 6 months, together with persistent symptoms of vaginal itch and discharge (20). Exacerbations of RVVC were not attributable to diabetes, antibiotic treatment, human immunodeficiency virus infection, or other immune abnormalities. Their average age was 33.1 years. The average number of thrush episodes was 8 per year, and the average number of treated episodes was 6.4 per year. About half the patients reported functional gastrointestinal symptoms.

Controls. There were 22 control subjects, and the average age was 37.8 years. Because of the possibility of a shift to a Th2 type response to Candida, causing susceptibility to RVVC (38-40), care was taken to ensure that the controls had no history of disorders associated with this type of response. In particular, the controls used in the study had no history of perennial rhinitis, asthma, urticaria, eczema, persistent functional gastrointestinal symptoms, or Candida infection involving the throat, skin, or gut (38, 39). A questionnaire was used to routinely check for these disorders in both patients and controls.

Serum samples. Institutional ethics approval was obtained for the study. Informed consent from the blood donors was obtained. Blood (10 ml) was collected by venipuncture into Vacutainer tubes, allowed to clot at room temperature, and centrifuged to recover the serum. The serum was frozen at 20°C in multiple aliquots of 0.5 to 1 ml. A fresh aliquot of serum was used for each assay.

Candida antigens. Two different preparations of mannan derived from C. albicans were used. The CTAB method involves formation of a complex with the mannan, which is subsequently isolated. Purification of mannan by the PEAT method involves degrading under alkaline conditions and precipitation with Fehling's solution (37). The mannan produced by the CTAB method is believed to be a more native product (34). The mannans are highly branched glycoproteins containing essentially mannose, and less than 10% of the molecular weight is attributable to protein. We also used the Hollister-Stier (Spokane, Wash.) whole C. albicans extract, which is used for skin testing. This extract is prepared by growing the yeast on defined medium, harvesting mycelia, and then defatting, drying, and extracting in glycerin. Protein and polysaccharide fractions are retained in the C. albicans extract. A similar method was used to prepare the fungal extracts, which were also purchased from Hollister-Stier.

Monoclonal antibody to human TABM. The mouse monoclonal antibody (MG3C9-1A12) specifically recognizes human TABM (30). This antibody was prepared by the immunization of BALB/c mice with M_r 33,000 TABM isolated from Cohn fraction III serum proteins by (NH₄)₂SO₄ precipitation, molecular sieving, and immunoabsorption. The immunogen did not contain immunoglobulins or albumin but was recognized by antibodies specific for TCR C and human TABM. It was purified from culture supernatant (25-fold; 700 µg/ml) using protein G-Sepharose (Pharmacia) and stored in phosphate-buffered saline (PBS) with sodium azide at 4°C. This IgG monoclonal antibody binds to TABM from serum and a lysate from a T-cell line but does not bind to IgG, IgM, IgA, human serum albumin, lysates from a B-cell line, or TGF- (30).

ELISA. Extensive enzyme-linked immunosorbent assay (ELISA) experiments were carried out to determine the optimum conditions for the measurement of TABM and IgG levels to each of the Candida antigens. The standard for each ELISA consisted of a pool of equal volumes of sera from six patients with recurrent thrush. It was aliquoted in 0.2-ml volumes and stored frozen until required. The serum pool was serially diluted twofold in PBS-Tween-1% gelatin (PTG) for IgG and PBS (pH 7.2) for TABM. For each ELISA, a standard curve was generated after plotting the optical density against arbitrary Units per milliliter, with each dilution of the standard being assigned the appropriate number of units per milliliter. The units of activity for each serum sample were determined after plotting standard curves using the Beckman Immunofit EIA/RIA analysis program (version 3.0). All dilutions of the standard and samples were tested in duplicate.

ELISA for detection of human IgG specific to CTAB and PEAT mannans and whole Candida extract (Hollister-Stier). The mannan preparations were diluted in PBS (pH 7.4) and coated overnight at 37°C on Falcon (Becton Dickinson, Paramus, N.J.) Pro-Bind plates. After being washed with PBS-0.05% Tween 20 (PBS-T), the plates were blocked with 1% gelatin in carbonate buffer (pH 9.6) for 90 min at 37°C and rewashed. The standard serum pool was assigned 10⁶ units/ml and serially diluted from 1/500 to 1/128,000 to produce a standard curve. Serum samples were diluted 1/5,000 in PTG, 100 µl was added per well in duplicate, and the plates were incubated for 90 min at 37°C. The plates were washed, 100 µl of affinity-purified peroxidase-conjugated sheep anti-human IgG (Silenus) was added (1/10,000), and the plates were incubated for 90 min at 37°C. Following washing, 100 µl of 3,3',5,5'-tetramethylbenzidine hydrochloride (TMB) substrate (Kirkegaard and Perry Laboratories, Gaithersburg, Md.) was added, and the reaction allowed to proceed at room temperature until an optical density of approximately 2 was obtained. The reaction was stopped with 2 M H₂SO₄, and the plates were read at 450 nm.

ELISA for detection of human TABM specific to CTAB and PEAT mannans and whole Candida extract (Hollister-Stier). C. albicans mannans were bound to Falcon ELISA plates as described above. The plates were blocked with 1% human serum albumin (HSA) in PBS and washed. The standard serum pool was assigned 1,000 U/ml and serially diluted from 1/20 to 1/320 to produce a standard curve. Serum samples were tested at a 1/50 dilution in PBS, 100 µl was added per well, and the plates were incubated. Protein-G purified mouse monoclonal anti-human TABM antibody was diluted 1/1,000 in PTG, 100 µl was added to the wells, and the plates were incubated and washed. Affinity-purified peroxidase-conjugated sheep anti-mouse immunoglobulins (Silenus) (100 µl) diluted 1/500 in PTG was added and incubated. The reaction was completed as described above.

Statistical analysis. Two-tailed statistical analysis of the data comparing patients and controls was performed using the Mann-Whitney U test.

Purification of TABM specific for Candida CTAB mannan. Serum was obtained from the patient with the highest titer of TABM specific to CTAB mannan. She had a history of recurrent vaginal thrush for 10 years, with nine episodes reported in the previous year. She had not taken antibiotics for several years. She also had a history of symptoms consistent with an irritable bowel syndrome and fibromyalgia. To 41 ml of serum, saturated ammonium sulfate was added to 43% with continuous mixing; the mixture was centrifuged at 9,000 rpm in an SS-34 rotor for 20 min at 4°C. The pellet was dissolved in PBS in approximately half the starting volume and dialyzed against PBS (pH 7.2) overnight at 4°C. The sample was then filtered through a 0.45-µm-pore-size filter. TABM recognized by the monoclonal anti-human TABM antibody (MG3C9-1A12) immobilized on Sepharose (Pharmacia) were isolated by affinity chromatography. The "total" TABM bound by the antibody were eluted with glycine-HCl (pH 2.8) and neutralized with Tris. Candida CTAB mannan immobilized on nitrocellulose discs (34) was rehydrated with PBS, washed, and blocked with 2% HSA in PBS for 2 h at 37°C. The discs were then washed five times with 50 ml of PBS (pH 7.2). The "total" TABM were then mixed with the mannan-nitrocellulose discs and incubated for 2.5 h at room temperature. The discs were thoroughly washed five times with 50 ml of PBS. TABM was eluted with 10 ml of glycine-HCl (pH 2.8) for 10 min at room temperature. The eluate CTAB-TABM was then neutralized and dialyzed overnight against 0.1 M Tris-0.15 M NaCl (pH 7.2) at 4°C. Octylglucopyranoside (OG; ICN, Irvine, Calif.) was added to 30 mM, and the sample was stored at 4°C in Tris-NaCl-OG (TNO). The protein concentration of purified CTAB-TABM was 32.9 µg/ml, using the Bio-Rad dye reagent and bovine serum albumin as the standard. The amount of CTAB-TABM recovered was 329 µg.

Antigen specificity of CTAB-TABM. The CTAB-TABM was tested for antigen-specific binding to CTAB mannan and the Hollister-Stier Candida extract after serial dilution in PBS or TNO. This was followed by addition of the mouse anti-human TABM antibody and a peroxidase-labeled sheep anti-mouse IgG antibody.

SDS-PAGE. CTAB-TABM was diluted 1:1 to 50 µg/ml in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer and reduced by the addition of dithiothreitol to 50 mM. The mixture was boiled for 5 min, and then iodoacetamide (Sigma) was added to 100 mM. Approximately 500 ng of CTAB-TABM was resolved in 8 to 25% polyacrylamide PHAST gels and resolved in a PHAST system (Pharmacia). Proteins in the gel were stained with a silver stain.

Detection of TGF- associated with CATB-TABM. (i) Direct ELISA. Microtiter trays were coated with 1 µg of CTAB-TABM per well, and the wells were blocked with 200 µl of 0.1% gelatin. Then 100 µl of 1:400 normal rabbit serum and rabbit anti-TGF-1 or anti-TGF-2 IgG (Santa Cruz Biotechnology, Santa Cruz, Calif.) were added to the wells, and the trays were incubated at 37°C for 90 min, after which the trays were washed five times, and 100 µl of alkaline phosphatase-conjugated goat anti-rabbit IgG (Sigma) was added. After another 90 min, the trays were washed five times, and p-nitrophenyl phosphate substrate was added. The color was read at 410 nm with a Dynatech microplate reader.

(ii) Capture ELISA. CTAB-TABM (200 ng) in 100 µl of RD 51 (R&D Systems) diluent was mixed with 100 µl of RD51 diluent with or without different amounts of CTAB mannan or Hollister-Stier Candida extract. Some CTAB-TABM was mixed and acid activated with HCl as described by R&D Systems. The mixtures were held at room temperature, and after 15 min HCl-containing samples were neutralized with NaOH and the entire 200-µl sample and TGF-2 standards were added to R&D Systems TGF-2 (capture) ELISA trays. The trays were incubated and developed as described by the manufacturer.

PBMC cultured with whole Candida extract. In preliminary experiments, it was found that cellular proliferation in response to whole Candida extract (Hollister-Stier) was optimal at a final dilution of 1/1,000 and 6 days of culture as measured by [³H]thymidine uptake (data not shown). Although culture for 3 days is usual when phytohemagglutinin is used as a stimulus for cell proliferation, earlier studies (11) have shown that longer periods of culture are required if Candida extract is used. Peak production of IFN- by PBMC also occurred with a 1/1,000 dilution of Candida extract and 6 days of culture at 37°C. The blood was anticoagulated with lithium heparin, and PBMC were isolated using Ficoll (Pharmacia, Uppsala, Sweden). PBMC were suspended at 1.0 × 10⁶ cells/ml in AIM V serum-free medium (Gibco, Life Technologies, Melbourne, Australia). Mercaptoethanol was present at 5 × 10⁵ M in all cultures. The Hollister-Stier Candida extract was used in PBMC culture after dialysis in PBS and dilution in AIM V (11). It was stored frozen until required.

(i) PBMC proliferation and IFN- production. PBMC were isolated from the blood of healthy females. Then 0.5 × 10⁶ cells/ml were cultured in round-bottom Falcon tissue culture tubes (Becton Dickinson) in a volume of 0.5 ml. Candida antigen in AIM V medium was added to give a final dilution of 1/1,000 of the original extract. Cultures were performed without and with the addition of sterilized TGF-1, (R&D Systems). Concentrations of TGF-1 in culture ranged from 7.8 to 2,000 pg/ml. After 6 days of culture at 37°C in a 5% CO₂ incubator, 1 µCi of [³H]thymidine was added for 6 h. Cells were then harvested, and the incorporated radioactivity was counted. For IFN- production, PBMC were also cultured as described above and the supernatant was collected and frozen until the IFN- levels were determined.

(ii) IFN- assay. A sandwich ELISA was developed to measure IFN-. Briefly, a monoclonal anti-human IFN- antibody (Serotec, Oxford, United Kingdom) was coated (1 µg/ml) onto an ELISA plate (Costar) in carbonate buffer overnight at 4°C. The plate was washed and blocked with 1% HSA in PBS for 90 min at 4°C. The sample or standard (Pharmingen, San Diego, Calif.) IFN- (2,000 to 3.96 pg/ml) was diluted in AIM V medium and added to the plate, which was incubated for 90 min at 37°C. After the plate was washed, a polyclonal rabbit anti-human IFN- antibody (Peprotech) was applied (1 µg/ml), followed by a peroxidase-labeled sheep anti-rabbit IgG (Silenus). The color reaction was developed as described above.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Serum IgG and TABM levels to C. albicans antigens. A comparison of levels of IgG to C. albicans CTAB and PEAT mannans between patients and controls showed the levels to be significantly higher in the patient group (P = 0.046 and 0.030 for CTAB and PEAT mannans, respectively). There was no significant difference between patients and controls in IgG levels to the whole Candida extract (Hollister-Stier) (P = 0.079) (Fig. 1). Levels of TABM to the CTAB mannan preparation were also significantly higher in the patient group (P = 0.029). There was no significant difference in levels of TABM to the PEAT mannan (P = 0.072) or to whole Candida extract (Hollister-Stier) (P = 0.217) between patients and controls (Fig. 2).

View larger version (11K): [in this window] [in a new window]   FIG. 1.   The levels of antigen-specific IgG specific to C. albicans CTAB (A) and PEAT (B) mannans and to the whole Candida extract (Hollister-Stier) (C) were determined in a control group and a patient group. P values were obtained by the Mann-Whitney U test. The bars represent the median values.

View larger version (12K): [in this window] [in a new window]   FIG. 2.   The levels of antigen-specific TABM to C. albicans CTAB (A) and PEAT (B) mannans and to the whole Candida extract (Hollister-Stier) (C) were determined in a control group and a patient group. P values were obtained by the Mann-Whitney U test. The bars represent the median values.

Characteristics of CTAB-specific TABM. A 1:1,000 dilution of the monoclonal antibody MG3C9-1A12 reacted with 15 to 500 ng of purified CTAB-TABM, while monoclonal antibodies to kappa and lambda light chains, immunoglobulin gamma, and immunoglobulin mu chains did not react with 15 to 500 ng of CTAB-TABM but did react with 15 ng of IgG or IgM (data not shown). CTAB-TABM was diluted in TNO or PBS and tested by ELISA for binding to CTAB mannan and whole Candida extract (Hollister-Stier). It was found that as long as the diluent was fresh, similar titration curves were obtained with the two diluents (data not shown). CTAB-TABM bound not only to the Candida CTAB mannan but also to other fungal extracts, including Pityrosporum, Cladosporium, Trichophyton, Penicillium, Aspergillus, and Alternaria (Fig. 3). Binding with a negative control antigen (-lactoglobulin) was very weak.

View larger version (31K): [in this window] [in a new window]   FIG. 3.   A panel of fungal extracts were coated onto ELISA plates. 1, 500 ng/well; 2, 1/2,000 dilution of the Hollister-Stier extracts. The CTAB-TABM was diluted in PBS and tested for reactivity with the different extracts. Antigen-specific TABM was detected with the MG3C9-1A12 anti-human TABM antibody. OD, optical density.

View larger version (28K): [in this window] [in a new window]   FIG. 4.   CTAB-TABM (1 µg) was reduced with 50 mM dithiothreitol and alkylated with 50 mM iodoacetamide. Reduced, alkylated CTAB-TABM (200 ng) was resolved on an 8 to 25% polyacrylamide gradient gel. Resolved proteins were stained by silver stain. Molecular weights were determined by comparison with the mobilities of prestained molecular weight standard proteins (Mr 200,000 to 18,000).

Presence of cytokines in the CTAB-TABM. The purified CTAB-TABM was tested for the presence of TGF- and IL-10 by ELISA. TGF-2 was detected by direct ELISA (Fig. 5A), but there was little or no TGF-1. IL-10 was not detected (data not shown).

View larger version (79K): [in this window] [in a new window]   FIG. 5.   (A) CTAB-TABM was coated to microtiter trays at 1 µg/well. A 100-µl volume of a 1:400 dilution of normal rabbit serum (NRS) control and rabbit anti-TGF-1 and anti-TGF-2 IgG was added. After incubation and washing, alkaline phosphatase-conjugated goat anti-rabbit IgG was added. Binding of anti-rabbit IgG was detected by a 10-min incubation with p-nitophenyl phosphate substrate. The optical density (OD) was determined at 410 nm. (B and C) Binding of CTAB-TABM by mannan (B) or Candida extract (C) activated the associated TGF-2. CTAB-TABM (200 ng) was incubated with or without CTAB-mannan (B) or Candida extract (C) for 2 h at room temperature or 15 min with HCl (B). After incubation, the mixture was added to microtiter trays coated with anti-TGF-2, and bound TGF-2 was detected by ELISA. Picogram levels of TGF-2 detected in the samples were determined by comparison of the optical density obtained with the sample to that of TGF-2 standards.

Regulation of the PBMC proliferation and IFN- production induced by whole Candida extract. Because TABM have immunoregulatory activity (7, 8, 29), we determined the effect of TGF- on Candida extract-induced proliferation by PBMC (Fig. 6A) and of both TGF- (Fig. 6B) and CTAB-TABM (Fig. 7) on Candida extract-induced IFN- production by PBMC. The addition of recombinant human TGF- to PBMC cultures inhibited both the cell proliferation (Fig. 6A) and the amount of IFN- produced (Fig. 6B). This response was considered representative, since similar results were found in two other normal individuals (data not shown). The amount of IFN- produced increased from 256 pg/ml at a 1/8,000 dilution of Candida extract to 689 pg/ml at a 1/1,000 dilution, and the addition of CTAB-TABM to cultures of PBMC with added whole Candida extract reduced the Candida-induced IFN- production by 30 to 60% (Fig. 7). These results are representative of the data obtained by combining different doses of Candida extract and CTAB-TABM. The suppressive effect appeared to be greater at higher dilutions of Candida extract (Fig. 7).

View larger version (17K): [in this window] [in a new window]   FIG. 6.   Whole Candida extract (1/1,000) and TGF-1 (0, 7.8 to 2,000 pg/ml) were added to normal PBMC and cultured for 6 days. The amounts of cell proliferation (A) and IFN- produced (B) were measured. The first bar represents PBMC background activity without antigen (Ag). The next bar represents PBMC plus Candida extract (+Ag) but without TGF-.

View larger version (43K): [in this window] [in a new window]   FIG. 7.   Effect of CTAB-TABM on whole Candida extract stimulation of IFN- production by PBMC. Different concentrations of CTAB-TABM (0, 1.25, 2.5, and 5.0 µg/ml) were added to cultures with medium or Candida antigen (1/1,000, 1/4,000, and 1/8,000). The levels of IFN- were determined after 6 days of culture.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

TABM are circulating antigen-specific immunoproteins derived from T cells. Human and murine TABM share epitopes associated with the constant region of TCR C chains (10, 30). Also, murine TABM have an amino acid sequence homology to TCR C (10) and V chains (9). However, although TABM are antigenically and structurally similar to TCR C chains, they are not identical to them, and it is unlikely that TABM are shed TCR. TABM, unlike TCR, ligate non-major histocompatibility complex-associated antigens (13, 48). TABM are thought to have an immunoregulatory function, particularly the suppression of cell-mediated immunity with immune deviation (8, 41, 50). Accordingly, the presence of circulating TABM specific to an antigen can provide an indication of the nature (Th1 or Th2) of the immune response.

An important feature of this work is the measurement of serum TABM to Candida antigens with a monoclonal antibody (MG3C9-1A12) that binds to epitopes unique to human TABM. The availability of this antibody has removed many of the difficulties previously encountered in measuring TABM levels with polyclonal antisera: the low concentrations and hydrophobicity of TABM and the possible cross-reactivity of the polyclonal antibodies between IgG and TABM (6, 8).

In the patient group there were raised levels of TABM to the CTAB mannan but not to the PEAT mannan or whole Candida extract. The CTAB mannan is thought to be a more native antigen, since the extraction procedure is less harsh than for the PEAT method (34). The mannan component of Candida has been linked with suppression of the cellular immune response (14, 16, 22, 23, 27), and TABM specific to mannan may be associated with suppressed cellular immunity to the yeast. However, the monoclonal antibody may detect other TABM, not associated with immunosuppression, which bind to antigens in the whole Candida extract. TABM are heterogeneous, and previous studies (30) have shown that the monoclonal antibody may detect more than one type. Also, if most of the TABM are specific for CTAB mannan and there is less mannan in the whole Candida extract, the TABM titer against mannan will be greater than that against extract. This may explain why TABM levels to whole Candida extract were not raised in the patient group.

Because TABM have been linked to regulation of cell-mediated immunity (8, 29, 50), TABM specific to CTAB mannan (CTAB-TABM) were isolated from the serum of the patient with the highest titer. This was done by using the MG3C9-Sepharose column to purify TABM from the serum, followed by using mannan linked to nitrocellulose discs to isolate mannan-specific TABM. SDS-PAGE of CTAB-TABM demonstrated the presence of a molecular species at M_r 86,000 and minor bands at M_r 43,000 and 22,000. This CTAB-TABM bound specifically to CTAB mannan and did not contain immunoglobulins. It is likely that the M_r 86,000 and 43,000 molecular species are multimers of an M_r 22,000 protomer, because these hydrophobic TABM have a strong tendency to polymerize (6, 7). The M_r 22,000 molecular species may be a "protomeric" TABM or TGF- associated with CTAB-TABM.

TGF-2 was found to be associated with CTAB-TABM at concentrations of up to 1,800 pg/µg of TABM. Most of the TGF-2 was detectable only with the binding of the CTAB-TABM to mannan or whole Candida extract. At optimal ratios of CTAB-TABM to mannan, there was a 37-fold increase in the detectable levels of TGF-2 and less detectable TGF-2 with high and very low concentrations of mannan.

Both TGF- and CTAB-TABM were shown to inhibit the cellular immune response to C. albicans, as assessed by IFN- production by Candida-stimulated PBMCs. TGF- is thought to prevent the action of IL-12 in inducing IFN- production (3, 24, 43). The inhibitory effect of CTAB-TABM on IFN- production may be attributable to the associated TGF-2. Although it may be anticipated that anti-TGF- would block this inhibition, in preliminary experiments (data not shown) we found that anti-TGF- antibody per se inhibited IFN- production by Candida-stimulated PBMCs. TGF- promotes the differentiation of Th1 cells in culture (42, 46) but not in vivo (2), which may explain this observation. Other workers (31) have encountered similar problems when studying the effects of anti-TGF- in cell cultures. In view of the pleotropic properties of TGF-, extensive studies may be necessary to clarify the issue.

If TGF-2 associated with CTAB-TABM mediates suppression of IFN- production, this effect may depend on the concentrations of both mannan extract and CTAB-TABM, since the level of detectable TGF-2 varies with the TABM/antigen ratio. This effect of the ratio between TABM and antigen on detectable TGF- has been observed with other purified TABM. We have previously used the MG3C9-1A12 monoclonal antibody to isolate TABM specific to benzoic acid conjugated to HSA (BA-TABM) in a patient sensitive to the solvent toluene (25). In this preparation, the concentration of detectable TGF- on binding of the TABM to antigen at optimum ratios was three- to four-fold higher than in the CTAB-TABM. Perhaps the large amounts of activated TGF- in BA-TABM suppress the Th1 response; the smaller amounts found in CTAB-TABM may modulate the Th1/Th2 balance. Some species of TABM are also associated with other cytokines (e.g., IL-10), which may synergize with TGF- in the suppression of the cellular immune response (8, 15, 32).

Antibodies to Candida mannan are thought to be cross-reactive with the mannans of other yeasts and molds (1). For this reason, we examined the CTAB-TABM for similar cross-reactivity. It was found to bind strongly to baker's yeast (Saccharomyces cerevisiae), Pityrosporum, Cladosporium, Trichophyton, Penicillium, Aspergillus, and Alternaria. There was only weak binding to the PEAT mannan, indicating significant antigenic differences between the two mannans.

The higher levels of serum TABM to CTAB mannan in the patient group may be associated with a shift from a Th1 to a Th2 response. The high levels of IgG to both CTAB and PEAT mannans in the patients may reflect an increased Th2 response to mannan. However, it is possible that raised antibody levels to mannan play a direct role in increasing susceptibility to Candida infection. Higher levels of anti-mannan antibodies have been found in patients with mucocutaneous candidiasis (28). Although recent reports indicate that antibody to the mannoprotein complex may protect against vaginal Candida infections (12), it is quite possible these antibodies are not specific for the mannan component. Different methods of extracting the mannan result in different protein contents (34), and it is possible that the protective antibody is specific to the protein moiety (14).

Considerable care was taken to ensure that the controls had no history of allergy which may indicate a general shift toward a Th2 response, including to C. albicans. There are reports implicating raised levels of Candida-specific IgG and IgE in asthma, eczema, and RVVC (38, 39). Several patients had functional gastrointestinal symptoms. It is possible that the presence of C. albicans as a commensal in the gut may cause local effects on gut function in some patients, perhaps by binding to CTAB-TABM. For example, we have demonstrated that TGF- enhances the release of neuropeptides from sensory nerve endings, which could alter gut motility (25).

Although there are reports of TGF- playing a protective role in Candida infection, they are limited to the situation in naive animals (45). TGF- may increase the susceptibility to vaginal infection by suppressing cellular immunity (24, 43). It is proposed that circulating Candida-specific TABM interact with Candida antigen (mannan), increasing the level of detectable (i.e., active) TGF- associated with the TABM. If C. albicans is present in the vaginal tract, there may be a local increase in the level of detectable TGF-, associated with TABM at that site. This could explain the local suppression of cellular immunity to C. albicans in patients with recurrent vulvovaginal candidiasis, even though this immune response may be systemically intact (17). Such a process would enable circulating TABM to regulate the local immune response to C. albicans at any site where the yeast proliferates. It has been shown (5) that antifungal therapy can induce a shift from a Th2 to a Th1 response to Candida antigens, presumably by reducing the antigen load. This observation suggests a systemic rather than local regulation of the immune response. To date, no studies have been performed to measure TABM in human vaginal secretions, but T-cell numbers are not increased in the vaginal wall in experimental Candida infections in animals, according to preliminary studies (18). The mannan-specific TABM we have purified may be the previously described suppressor factors found in patients with recurrent mucosal Candida infections (16, 22, 52).

Both cellular and humoral immune responses occur to C. albicans. Since this yeast is a vaginal commensal organism, a balance between a potentially damaging cellular immune response and a susceptibility to vaginal infection needs to be maintained (18). CTAB-TABM, by delivering cytokines such as TGF- to the site of the organism, may be important in regulating this balance, with the level of detectable TGF- depending on the ratio between CTAB-TABM and mannan. TGF- is thought to be pivotal in the Th1-Th2 balance in a number of other infectious diseases such as malaria (35, 36, 47), leishmaniasis (51), and perhaps filariasis (26). Using the monoclonal antibody, we have detected elevated levels of TABM to filariasis antigens in patients who are chronic carriers of this parasite (30). Also, a shift from a Th2 to a Th1 response has been observed in filariasis treated with ivermectin (44), suggesting systemic regulation of this immune response, perhaps by TABM.

It is possible that measurement of the serum level of TABM specific to CTAB mannan may be of assistance in identifying patients susceptible to RVVC. However, the role of TABM may depend on other factors which increase the numbers of C. albicans in the vaginal tract, such as the use of antibiotics, pregnancy, cyclic variation of hormone levels, and the use of oral contraceptives. Effective antifungal treatment, by reducing yeast numbers, may shift the immune response toward a Th1 type, as reported in animal work (5). If so, there may be a fall in the titer of mannan-specific TABM. Further studies are required to determine the role of TABM in Candida infection.