|Home | About | Journals | Submit | Contact Us | Français|
Cryptococcus neoformans causes severe, and often fatal, disease (cryptococcosis) in immunocompromised patients, particularly in those with HIV/AIDS. Although resistance to cryptococcosis requires intact T-cell immunity, a possible role for antibody/B cells in protection against natural disease has not been definitively established. Previous studies of the antibody response to the C. neoformans capsular polysaccharide glucuronoxylomannan (GXM) have demonstrated that patients who are at increased risk for cryptococcosis have lower serum levels of GXM-reactive IgM than those who are not at risk, leading to the hypothesis that IgM might contribute to resistance to cryptococcosis. To determine the influence of IgM on susceptibility to systemic cryptococcosis in a murine model, we compared the survival of mice deficient in serum IgM (secretory IgM deficient [sIgM−/−]) and C57BL/6 × 129Sv (control) mice after intraperitoneal infection with C. neoformans strain 24067 and analyzed the splenic B- and T-cell subsets by flow cytometry and the serum and splenic cytokine/chemokine and serum antibody profiles of each mouse strain. The results showed that sIgM−/− mice survived significantly longer than control mice when challenged with 105 CFU of C. neoformans 24067. Naïve sIgM−/− mice had higher levels of B-1 (CD5+) B cells, proinflammatory mediators (interleukin-6 [IL-6], IL-1β, MIP-1β, tumor necrosis factor alpha [TNF-α], and gamma interferon [IFN-γ]), and anti-inflammatory mediators (IL-10 and IL-13) and significantly higher titers of GXM-specific IgG2a 3 weeks postinfection. In addition, CD5+ splenocytes from both mouse strains had fungicidal activity against C. neoformans. Taken together, these results suggest that the inflammatory milieu in sIgM−/− mice might confer enhanced resistance to systemic cryptococcosis, stemming in part from the antifungal activity of B-1 B cells.
Cryptococcus neoformans causes life-threatening meningitis and meningoencephalitis in immunocompromised individuals. Globally, cryptococcosis, or cryptococcal disease (CD), occurs in >900,000 individuals and is responsible for >600,000 deaths annually, with the majority of cases and deaths occurring in sub-Saharan Africa (51). In addition, with the introduction of highly active antiretroviral therapy in the developing world, CD has emerged as an important and common manifestation of immune reconstitution inflammatory syndrome (IRIS) (9). CD and IRIS-associated CD are also important and emerging diseases in recipients of solid organ transplants (63).
Current dogma holds that intact cell-mediated immunity is required for host resistance to C. neoformans. However, available data suggest that antibody and B cells could also contribute to resistance to CD, including in humans (23, 56, 67, 68). Studies with murine models have established that specific human and mouse IgM can prolong survival of C. neoformans-infected mice (reviewed in reference 17). Several groups have reported that B cells contribute to resistance to murine cryptococcosis (3, 56), and several reports have demonstrated that B-1 cell-derived mononuclear phagocytes have fungicidal activity against C. neoformans (4, 26). In humans, serological studies comparing HIV-infected and HIV-uninfected subjects have demonstrated that HIV-infected individuals, a group that is highly susceptible to CD, have lower levels of glucuronoxylomannan (GXM)-reactive IgM than HIV-uninfected individuals, a group that is highly resistant to CD (reviewed in reference 68). It has also been shown that among solid organ transplant recipients, another group with increased susceptibility to C. neoformans, those who developed CD after transplantation had markedly lower serum levels of GXM-reactive IgM prior to transplant than those who did not (35). In addition, Subramaniam et al. reported that the levels of memory IgM B cells were lower in HIV-infected individuals with a history of CD and/or were lower in those who subsequently developed CD than in those with no history of CD and/or those who did not develop CD (68). Memory IgM B cells are considered the human homolog of CD5+ mouse B-1 B cells and are the source of naturally occurring and capsular-polysaccharide-reactive IgM in humans (15).
To determine the importance of natural IgM in resistance to CD, we used mice that lack serum IgM (secretory IgM-deficient [sIgM−/−] mice) (12). B cells from these mice do not produce IgM due to a secretion defect, although their ability to produce other isotypes, including IgG, is retained (12). Hence, sIgM−/− mice provide an excellent model to determine the role of IgM in resistance to infectious diseases. For example, they were used to demonstrate a critical role for IgM in protection against West Nile and influenza viruses (7, 20), borreliae (5), acute peritonitis/sepsis (13), and pneumococcal pneumonia (14). In this study, we used a systemic infection model in sIgM−/− mice to explore the role of IgM in resistance to CD in mice. Our hypothesis was that these mice would be more susceptible to death from CD because of their inability to produce IgM. However, we found the opposite; an inoculum that was lethal in 50% of control mice was significantly less lethal in sIgM−/− mice after intraperitoneal C. neoformans challenge. Compared to the levels in control mice, naïve sIgM−/− mice had higher levels of pro- and anti-inflammatory cytokine/chemokines, higher levels of serum IgG2a, and a higher level of B-1 B cells. Taken together, our data suggest that the reduced virulence of C. neoformans in these mice stems from a naïve phenotype that is more resistant to systemic fungal infection compared to that in control mice.
Age-matched sIgM−/− mice (provided by Marianne Boes, Harvard Medical School) (12) and C57BL/6 × 129Sv (control) mice (Albert Einstein College of Medicine) were used for this study. C57BL/6 × 129Sv mice (henceforth designated C57×129Sv) were used as controls because the sIgM−/− strain was created with 129 ES cells in C57BL/6 blastocysts (M. Boes, personal communication). Since previous studies with sIgM−/− mice included mixed background controls (12, 13, 55) and the mice were not completely backcrossed when we began our studies, we used C57×129Sv mice as controls. C57×129Sv mice were also used in studies of cryptococcal pathogenesis in inducible nitric oxide synthase (iNOS) knockout mice (55). However, to establish that the virulence of C. neoformans was similar in C57×129Sv mice and C57BL/6 mice, we compared the survival of C57BL/6 (National Cancer Institute, Bethesda, MD) and C57×129Sv mice after infection with C. neoformans and found that the survival of the two strains similar statistically by the log rank test (data not shown). Therefore, subsequent experiments were done using C57×129Sv mice as controls, with an inoculum of 3 × 105 to 5 × 105 CFU. All mice were maintained in the Institute for Animal Studies (IAS) of the Albert Einstein College of Medicine (AECOM) and given unrestricted access to food and water, and all mouse experiments were conducted with prior approval from the Animal Care and Use Committee of AECOM by following established guidelines.
A serotype D strain of Cryptococcus neoformans ATCC 24067 (American Type Culture Collection, Manassas, VA) was used. This strain has been used extensively to study antibody immunity to C. neoformans (16, 19, 22, 24, 42, 46). The organism was grown for 52 to 56 h at 37°C with shaking in Difco Sabouraud dextrose broth (Becton Dickinson, Franklin Lakes, NJ), washed twice in sterile, endotoxin-free phosphate-buffered saline (PBS) at pH 7.4 (Mediatech, Herndon, VA), counted in a hemocytometer using trypan blue for viability, and diluted to the desired concentration in PBS. Mice were infected intraperitoneally (i.p.) with the following two different inocula: (i) 1 × 107 CFU of C. neoformans 24067 per mouse (a dose that resulted in 100% mortality within 30 days in control mice) and (ii) 3 × 105 to 5 × 105 CFU of C. neoformans 24067 per mouse (a dose that resulted in 50% mortality within 140 days in control mice). Mice were bled retro-orbitally at 24 h postinfection and every week thereafter for 4 weeks to collect serum samples for various assays described hereafter. For survival studies, infected mice were observed at least once daily.
GXM concentrations in the sera of infected mice were determined by an antigen-capture enzyme-linked immunosorbent assay (ELISA), using a previously described protocol (19). Briefly, 96-well microtiter plates were coated for 1 h at 37°C with a human IgM monoclonal antibody (MAb) to GXM, G19 (42), at 5 μg/ml. The plates were blocked overnight with 1% bovine serum albumin (BSA) (Sigma, St. Louis, MO) in PBS, pH 7.2 to 7.4 (1% BSA-PBS), and washed with PBS containing 0.05% Tween 20 (Sigma) prior to use; all microtiter plate washes were done three times using a SkanWasher 400 (Molecular Devices, Sunnyvale, CA). Serum samples were diluted 1:100 in PBS, incubated with 200 μg of proteinase K (Sigma) per ml for 4 h at 37°C, and then boiled for 30 min in a water bath in order to inactivate the proteinase K. GXM purified from ATCC 24067 by following a previously described protocol (19) was diluted to 10 μg/ml in PBS and similarly treated. Serum samples were applied to the plates in duplicate and then serially diluted 1:3 in 1% BSA-PBS. Next, the plates were washed and reincubated with 5 μg/ml of a mouse IgG1 MAb to GXM, 2H1 (provided by A. Casadevall, AECOM). Bound 2H1 was detected by incubating the plates with a 1:5,000 dilution of alkaline phosphatase (AP)-conjugated goat anti-mouse IgG1 (Southern Biotechnology, Birmingham, AL). Plates were then developed with p-nitrophenyl phosphate (Sigma) in bicarbonate buffer (pH 9.8), and absorbances were measured by an ELISA reader (Tecan, Austria) at 405 nm. Wells with no sera were included as a negative control to determine the background absorbance of the reagents.
The levels of splenic B and T cells were determined by flow cytometry. Briefly, mice were infected with 105 C. neoformans cells i.p. (or treated with PBS as controls) and killed at specific times postinfection. Spleens from both infected and PBS-treated mice were removed and macerated using frosted microscope slides (Fisher Scientific, Pittsburgh, PA) in 3 to 5 ml of 1% BSA-PBS. The cell suspension was passed through a 70-μm filter (BD Falcon, Bedford, MA) to remove gross tissue debris and isolate the splenocytes. The splenocytes were pelleted at 1,200 rpm at 4°C for 10 min, and the erythrocytes were lysed in 5 ml of ACK buffer [0.15 M NH4Cl, 10 mM KHCO3, and 0.1 mM EDTA adjusted to a pH of 7.2 to 7.4 with 1 N HCl (all reagents from Sigma)] at room temperature for 5 min. Following the lysis, the buffer was diluted in 3 volumes of 1% BSA-PBS, and cells were pelleted, washed twice, and passed through a 40-μm filter (BD Falcon) to remove fat cells and further cellular debris. The cells were suspended in a final volume of 12 ml of 1% BSA-PBS and counted on a hemocytometer, and an aliquot was appropriately diluted in order to a get the final desired concentration of 107 total cells per ml. A total of 100 μl of the cell suspension (106 cells) was removed and stained using fluorescein isothiocyanate (FITC)-conjugated anti-mouse CD19 (BD Biosciences, San Jose, CA), phycoerythrin-cyanine 7 (PE-Cy7)-conjugated anti-mouse CD5 (eBioscience, San Diego, CA), Cy5-conjugated anti-mouse IgM, Fc5μ fragment-specific antibody (Jackson ImmunoResearch, West Grove, PA), PE-Cy5-conjugated anti-mouse CD4, allophycocyanin (APC)-conjugated anti-mouse CD8, and FITC-conjugated anti-mouse F4-80 (all reagents from eBioscience) for 1 h at 4°C. Cells were then washed and fixed with 2% paraformaldehyde. Single-color controls and fluorescence minus one controls were also included to ensure proper gating. Data was collected on a FACSCalibur (Becton Dickinson, Sunnyvale, CA) interfaced to CellQuest software (Macintosh version; Becton Dickinson, Sunnyvale, CA). Fifty thousand events per sample were analyzed. All analyses were performed using FlowJo version 7.2.2 for Windows software (Ashland, OR). B-1a B cells were characterized as CD19+ CD5+, and B-1b B cells were characterized as CD19+ CD5− (30).
Total and GXM-specific serum IgG and IgG subtype levels were determined by ELISA. Briefly, 96-well microtiter plates were coated with 10 μg/ml of goat anti-mouse IgG (Southern Biotechnology) for 1 h at 37°C for total IgG determination and with 10 μg/ml of purified GXM in PBS for 3 h at room temperature for GXM-specific IgG determination. The plates were blocked overnight with 1% BSA-PBS and washed with PBS-Tween 20 prior to use. Samples were obtained from naïve, 1-week postinfection, and 3-week postinfection states. A mouse IgG standard (Sigma) was applied to the plates at a starting concentration of 10 μg/ml. Serum samples were diluted 1:1,000 in PBS for total IgG and 1:60 for GXM-specific IgG determination, applied to the plates in duplicate, and then serially diluted 1:3 in 1% BSA-PBS, followed by incubation at 37°C for 1 h. Plates were then detected with a 1:2,500 dilution of AP-conjugated goat anti-mouse IgG and/or AP-conjugated goat anti-mouse IgG1, IgG2a, IgG2b, and IgG3 (all from Southern Biotechnology) at 37°C for 1 h and washed and developed with p-nitrophenyl phosphate (Sigma) in bicarbonate buffer (pH 9.8), and absorbances were measured by an ELISA reader (Tecan) at a wavelength of 405 nm. Wells with no sera were included as a negative control to determine the background absorbance of the reagents. The results were plotted on a semilog scale after subtraction of the background optical density (OD), and a titration curve was generated. The titer was defined as the point at which the titration curve crossed an OD of 0.1.
Serum IgE levels were determined by ELISA. In brief, 96-well microtiter plates were coated with 2 μg/ml of goat anti-mouse IgE (BD Pharmingen) for 1 h at 37°C. The plates were blocked overnight with 1% BSA-PBS and washed with PBS-Tween 20 prior to use. Serum samples were prepared from the naïve and 3-week-postinfection blood samples, diluted 1:20, applied to the plates in duplicate, and then serially diluted 1:3 in 1% BSA-PBS, followed by incubation at 37°C for 1 h. An IgE standard (BD Pharmingen) was also applied to the plates at 1 μg/ml. The plates were then incubated with a 1:1,000 dilution of AP-conjugated goat anti-mouse IgE (Southern Biotechnology) at 37°C for 1 h, washed, and developed with p-nitrophenyl phosphate (Sigma) in bicarbonate buffer (pH 9.8). Absorbances were measured by an ELISA reader (Tecan) at a wavelength of 405 nm. Wells with no sera were included as a negative control to determine the background absorbance of the reagents. The results were plotted on a semilog scale after subtraction of the background OD, and a titration curve was generated. The titer was defined as the point at which the titration curve crossed an OD of 0.1.
The concentration of serum cytokines was determined using the Beadlyte mouse 10-plex cytokine detection system. The levels of interleukin-1α (IL-1α), IL-1β, IL-6, IL-10, IL-12(p70), IL-13, gamma interferon (IFN-γ), KC, MIP-1β, and tumor necrosis factor alpha (TNF-α; Millipore, St. Charles, MO) were determined according to the manufacturer's protocol, using a Luminex HT200 machine (Luminex Corporation). Positive controls provided by the manufacturer were run as quality controls with serum samples. Sera from sIgM−/− and control mice were obtained in the naïve state and 1 week and 3 weeks postinfection, stored at −80°C, and centrifuged at 4°C for 15 min at 13,000 × g (to separate lipids) prior to use at a dilution of 1:2. Peritoneal lavage fluid was obtained from mice at 1 week and 3 weeks postinfection, using PBS for lavaging, and cytokine levels for sera were determined.
The secretion of IL-12 and TNF-α by spleen cells and peripheral blood mononuclear cells (PBMCs) of infected mice was determined with enzyme-linked immunospot (ELISPOT) assay kits (R&D Biosystems, Minneapolis, MN), according to the manufacturer's protocols. Briefly, spleen cells and/or PBMCs obtained from infected mice 8 days postinfection were plated onto a 96-well polyvinylidene difluoride-backed microplate precoated with a polyclonal antibody specific for mouse IL-5, IL-12, IL-13, or TNF-α (R&D Biosystems) and incubated for 18 h in a 37°C incubator with CO2. Following incubation, the plates were washed, detected with a biotinylated polyclonal antibody specific for the cytokine, and incubated overnight at 2 to 8°C. Thereafter, streptavidin-AP was added, and the plates were developed with 5-bromo-4-chloro-3′ indolyl phosphate p-toluidine salt and Nitro Blue Tetrazolium chloride. ELISPOT-forming cells were counted using an automated ELISPOT reader system (Autoimmun Diagnostika, Strasburg, Germany). To account for variability in the yield of cells from spleens or PBMCs when the ELISPOT assays were repeated, the results were normalized based on the input of cells (2.5 × 105 cells/well for spleen cells; 105 cells/well for PBMCs).
The ability of CD5+ spleen cells to mediate the killing of C. neoformans was evaluated with a killing assay performed as described previously, with modifications for splenocytes (47, 78, 79). Briefly, splenocytes were isolated from naïve sIgM−/− and control mice, ~2 × 108 cells per strain were resuspended in PBS containing 0.5% BSA and 2 mM EDTA (Sigma), such that 90 μl contained 107 cells, and the cells were incubated with 10 μl of paramagnetic MACS microbeads labeled with CD5 (Ly-1) (Miltenyi Biotec, Auburn, CA) for 15 min at 4°C, with gentle inversion for continuous mixing. Following incubation, cells were washed, resuspended in 500 μl of the buffer, loaded onto an LS column (Miltenyi Biotec), and placed in the magnetic field of a MACS separator (Miltenyi Biotec) to select the CD5+ population. Positively selected CD5+ splenocytes were counted in a hemocytometer and pelleted at 1,200 rpm at 4°C for 10 min. The pellet was resuspended in tissue culture medium [RPMI 1640 (Mediatech) containing 10% fetal calf serum (Atlanta Biologicals, Lawrenceville, GA) and 1% minimal essential medium (MEM) nonessential amino acids (Mediatech)]. The purity of the isolated cells was confirmed by fluorescence-activated cell sorting (FACS). Approximately 106 CD5+ cells were added in a 200-μl volume to each well of a 48-well tissue culture plate (Becton Dickinson) for an overnight incubation at 37°C with 5% CO2. After overnight incubation, the cells were stimulated with 100 U/ml of IFN-γ (PeproTech, Rocky Hill, NJ) to induce differentiation of the CD5+ cells into macrophage-like cells, as described previously (40). On day 2 and day 3 after stimulation, separate wells of cells were incubated with C. neoformans 24067 at an effector/target (E:T) ratio of 1:5 for 4 h at 37°C with 5% CO2, as described previously (70, 77). After incubation, the supernatants were collected. Next, cells in the wells were lysed with 100 μl of sterile distilled water for 30 min at room temperature, followed by aspiration and hard ejection of the fluid with a pipette several times to achieve complete cellular disruption, and the lysate was collected. Finally, 100 μl of PBS was used to rinse each well, and the cell supernatant, lysate, and rinse were pooled and mixed by vortexing. Tenfold serial dilutions were made and plated on Sabouraud's dextrose agar (Becton Dickinson). Plates were incubated at 37°C for 48 h, and colonies were counted after 2 days. Percent killing was defined as follows: [(initial input of C. neoformans − number of CFU recovered)/initial input of C. neoformans] × 100.
Mouse survival data were evaluated statistically by comparing Kaplan-Meier survival curves with a log rank (Mantel-Cox) test (Prism version 5.2 for Windows; GraphPad Software, CA). Student's t test or the Mann-Whitney U test was used for between-group comparisons depending upon the normality of data distributions and sample size, as appropriate. For all statistical evaluations, a two-tailed P value of ≤0.05 was considered significant, and P values of 0.05 to 0.1 were considered a trend toward significance; all P values were reported as numerical values.
The inoculum of C. neoformans is an important determinant of virulence in mouse models of cryptococcosis (69, 70). In this study, we used two inocula delivered i.p. to compare the survival of sIgM−/− and control mice. With a higher inoculum (1.1 × 107 CFU of C. neoformans 24067), the median survival of sIgM−/− mice (15 days) was longer than that of control mice (6 days) (Fig. (Fig.1A),1A), but this difference was not statistically significant (log rank test; P = 0.29). With a lower inoculum (3 × 105 to 5 × 105 CFU), sIgM−/− mice survived significantly longer than control mice (log rank test; P < 0.036) (Fig. (Fig.1B1B).
Previous studies have demonstrated the utility of measuring serum GXM levels to assess fungal clearance (19, 22, 77). Therefore, we measured the serum GXM levels of infected mice at 24 h and several weeks postinfection. Compared to the levels of control mice, sIgM−/− mice had lower serum GXM levels at all of the times evaluated; these differences were statistically significant at 24 h and 1 week postinfection (Mann-Whitney U test; P < 0.03) (Fig. (Fig.22).
We determined the levels of the total splenic, B- and T-lymphocyte, and B- and T-cell subsets in both mouse strains in the naïve state and at 1 week and 2 weeks postinfection. There was no difference in the numbers of total splenic lymphocytes, B cells, or T cells between the mouse strains (Fig. (Fig.33 and and4).4). Control mice had a higher level of total B cells at 2 weeks postinfection than at 1 week postinfection (P = 0.022; Student's t test). There were significant differences in B-1 B-cell expression between the mouse strains. In the naïve state, sIgM−/− mice had significantly higher levels of B-1a B cells (CD19+ CD5+) than control mice (P = 0.02; Student's t test) (Fig. (Fig.3C),3C), and at 1 week postinfection, sIgM−/− mice had a higher level of B-1b cells (CD19+ CD5−) than control mice (P = 0.020; Student's t test). Naïve control mice had a higher level of B-1b B cells than control mice at 1 week postinfection, as did control mice at 2 weeks postinfection compared to their counterparts at 1 week postinfection (P = 0.011 and 0.008, respectively; Student's t test) (Fig. (Fig.3D).3D). There were no significant differences in the number of splenic CD4 T cells, CD8 T cells, or mononuclear cells between the mouse strains (Fig. (Fig.44).
Levels of total IgG, IgG1, IgG2a, IgG2b, and IgG3 were determined in sera from naïve and C. neoformans-infected mice at 1 week and 3 weeks postinfection. There was no significant difference in total IgG between sIgM−/− and control mice in the naïve state or after infection (Fig. (Fig.5A).5A). Among the IgG subtypes, control mice had significantly more total IgG1 than sIgM−/− mice in the naïve state and at 1 week postinfection (Student's t test; P = 0.014 and P = 0.041, respectively), whereas sIgM−/− mice had significantly more total IgG2a at all the times evaluated (Student's t test; P of <0.0001 for the naïve state, P of 0.0009 for 1 week postinfection, P of 0.0013 for 3 weeks postinfection) (Fig. (Fig.5C).5C). There were no significant differences in the levels of total IgG2b between the mouse strains; however, sIgM−/− mice had significantly more total IgG3 at 1 week and 3 weeks postinfection (Student's t test; P = 0.037 and P = 0.018, respectively) (Fig. (Fig.5E).5E). Using the ratio of total IgG2a to IgG1 as a surrogate for a Th1-like response, as reported by others (18, 21, 25, 44, 71, 74), sIgM−/− mice had significantly higher ratios in the naïve state and at each postinfection time evaluated than control mice (Fig. (Fig.5F5F).
In the naïve state, the levels of GXM-specific IgG2a and IgG3 were significantly higher in sIgM−/− mice, whereas the level of GXM-specific IgG2b was higher in control mice (Fig. (Fig.6).6). Among the times evaluated, sIgM−/− mice had significantly more GXM-specific IgG2a at 3 weeks postinfection (Student's t test; P = 0.02) and more GXM-specific IgG3 at each postinfection time (Student's t test; P of 0.0048 for the naïve state, P of 0.0095 for 1 week postinfection, P of 0.0348 for 3 weeks postinfection) (Fig. (Fig.6E).6E). In contrast, control mice had more GXM-specific IgG2b at all the times evaluated, but this effect was statistically significant only in the naïve state (Student's t test; P = 0.035) (Fig. (Fig.6D).6D). Based on specific IgG titers, sIgM−/− mice had a higher ratio of GXM-specific IgG2a/IgG1 on days 7 and 21 postinfection, with the following median ratios. At day 7 postinfection (P = 0.078), sIgM−/− mice and control mice had ratios of 18 (n = 8) and 1 (n = 8), respectively; at day 21 postinfection (P < 0.002), sIgM−/− mice and control mice had ratios of 6 (n = 8) and 0.6 (n = 8), respectively (Student's t test) (Fig. (Fig.6F6F).
Levels of IgE in sera from mice in the naïve state and at 3 weeks postinfection were determined, using IgE levels as an indicator of Th2-like immunity (see Fig. S1 in the supplemental material). In the naïve state, both sIgM−/− and control mice had similarly low levels of serum IgE, with mean optical densities (ODs) of 0.047 and 0.057, respectively. At 3 weeks postinfection, both mouse strains exhibited a significant increase, with mean OD values of 0.1 and 0.13 for the sIgM−/− and control mice, respectively.
The importance of Th-1 cytokines and chemotactic/inflammatory chemokines in host resistance to C. neoformans is well documented (32, 33, 38, 39, 65). We determined serum cytokine/chemokine levels in each mouse strain in the naïve state and at 1 week and 3 weeks postinfection. In the naïve state, sIgM−/− mice had significantly larger amounts of IL-1β, IL-6, IFN-γ, TNF-α, MIP-1β, IL-10, and IL-13 than control mice (Fig. (Fig.7).7). Following infection, control mice had a significantly higher level of IFN-γ at 1 week postinfection, and sIgM−/− mice had significantly higher levels of IL-10 at 1 week postinfection (1) and IL-6 at 3 weeks postinfection (3). In addition to serum inflammatory mediators, we also measured the cytokine/chemokine response in the peritoneal cavities of both strains of mice at 1 week and 3 weeks postinfection (see Table S1 in the supplemental material). sIgM−/− mice had significantly more IL-1α than controls at 1 week postinfection and had a significantly higher level of IL-6 at 3 weeks postinfection (3) than at 1 week postinfection.
We also determined the level of cytokine secretion by analyzing splenocytes and PBMCs isolated from C. neoformans-infected mice at 1 week postinfection. Spleen cells and PBMCs from both mouse strains produced IL-5, IL-12, IL-13, and TNF-α (Fig. (Fig.8).8). Spleen cells from sIgM−/− mice produced significantly more TNF-α than those from control mice (Student's t test; P = 0.042).
Since sIgM−/− mice had more B-1a B cells in the naïve state and lower GXM levels than control mice at 24 h postinfection, we wondered if this CD5+ cellular subset had antifungal activity against C. neoformans in the naïve state. To address this question, we determined the ability of CD5+ splenocytes from each mouse strain to kill C. neoformans. Incubation of CD5+ splenocytes from both mouse strains resulted in the killing of >80% of an input inoculum of 1 × 107 C. neoformans CFU, with cells from the sIgM−/− mice manifesting more of a fungicidal effect than those from control mice at 3 days after stimulation with IFN-γ (Student's t test; P = 0.009) (Fig. (Fig.9).9). The mean number of C. neoformans organisms killed, based on CFU recovery/the number of input of C. neoformans CFU/the number of CD5+ cells plated were as follows: 8.6 × 106/1 × 107/2.4 × 106 for sIgM−/− mice, 5.4 × 106/7 × 106/1.4 × 106 for control mice with day 2 cells, 8.8 × 106/1 × 107/2.4 × 106 for sIgM−/− mice, and 5.8 × 106/7 × 106/1.4 × 106 for control mice with day 3 cells.
In this study, we used sIgM−/− mice (which lack secreted IgM) to determine the importance of naturally occurring IgM in resistance to systemic murine CD. We found that sIgM−/− mice survived significantly longer than control mice after infection with 3 × 105 to 5 × 105 CFU and had evidence of earlier fungal clearance (lower levels of serum GXM at 24 h and 1 week postinfection), though we note that the level of serum GXM may not reflect the fungal burden in tissue (19). Consistent with the original description of sIgM−/− mice (12), naïve sIgM−/− mice had higher levels of splenic B-1a (CD5+) B cells than control mice, suggesting that B-1a B cells could have a contributory role in the relative resistance of sIgM−/− mice to infection with a lower inoculum of C. neoformans. This was not the case for a higher inoculum, after which the survival of the mouse strains was comparable. B-1a (CD5+) B cells are a subset of B-1 B cells that are predominantly localized to the peritoneal cavity and spleen (30, 75). The human homolog of this cell type, memory IgM B cells (15), has been linked to the ability to respond to pneumococcal vaccine and pneumococcal resistance (41, 60) and, recently, to a lower risk for HIV-associated CD (68). To our knowledge, a role for B-1a B cells in murine CD has not been considered previously.
The expanded B-1a B cell compartment in sIgM−/− mice provides a unique opportunity to investigate the role of B-1a B cells in murine C. neoformans infection, with the caveat that sIgM−/− and most other mouse models of CD are primary infection models, whereas human disease occurs as a consequence of reactivation in the setting of chronic (latent) infection (reviewed in reference 19). We also note that naturally occurring IgM has been implicated in host defense against the relevant pathogen in other infection models in sIgM−/− mice (5, 7, 14, 20). In contrast, our model implicates sIgM-independent mechanisms, namely, a superior specific IgG response and/or fungicidal effects of CD5+ (B-1) B cells (26) in the resistance phenotype of the sIgM−/− mice to infection using a lower inoculum of C. neoformans. These proposed mechanisms are based on our data showing an IgG2a-dominated GXM response, a larger B-1a B cell compartment in naïve sIgM−/− mice, and the ability of CD5+ splenocytes from both mouse strains to kill C. neoformans. However, these mechanisms are likely to be interrelated, since the main mechanism by which specific antibody mediates protection against C. neoformans is by enhancing the antifungal activity of effector phagocytes (reviewed in reference 17).
Based on the ability of B-1a B cells to differentiate into phagocytes (4, 40) and exhibit direct fungicidal activity against C. neoformans (26) and the ability of the CD5 ectodomain to bind fungal cells, including C. neoformans (26), we determined the ability of CD5+ splenocytes to kill C. neoformans in vitro. Upon stimulation with IFN-γ ex vivo, CD5+ splenocytes from both mouse strains were able to kill C. neoformans. While this hypothesis requires validation in vivo, it is logical to posit that the larger CD5+ B-1 B cell compartment of naïve sIgM−/− mice could mount a larger antifungal response. We also found that splenocytes from both mouse strains secreted IL-12, a potent inducer of IFN-γ (59). Although we did not determine secretion of IFN-γ, PBMCs and splenocytes from sIgM−/− mice produced significantly more TNF-α than those from control mice. TNF-α, which is produced by B-1 B cells (43, 73), induces T-cell production of IFN-γ (43), and IFN-γ induces B-1a cells to differentiate into phagocytes (40). One caveat to our findings is that we used CD5+ splenocytes to assess fungicidal activity. CD5 can be expressed by mature T cells (10), and CD5+ T cells were more abundant in sIgM−/− mice than in IgM sufficient mice (12). Although sIgM−/− and control mice had similar levels of splenic T cells in our study, T cells could have enhanced the antifungal activity of B-cell-derived phagocytes in sIgM−/− mice. This question, and whether CD5+ splenocytes have fungicidal activity in vivo, are under evaluation in our laboratory but were beyond the scope of this study.
Specific IgG has been shown to enhance the ability of effector cells to phagocytose C. neoformans (reviewed in reference 58), and GXM-specific IgG2a is superior to other IgG subtypes in promoting phagocytosis in vitro (47, 58) and mediating protection against C. neoformans in vivo (8, 46, 47). Our data show that naïve sIgM−/− mice had higher levels of total IgG2a and that C. neoformans-infected sIgM−/− mice had higher levels of GXM-specific IgG2a than control mice. An elevated level of IgG2a is a characteristic of sIgM−/− mice that has been proposed to compensate for the lack of IgM (12). Hence, GXM-specific IgG2a could be compensatory in sIgM−/− mice. We recognize that the association between higher levels of GXM-specific IgG2a and the survival advantage of sIgM−/− mice does not establish causality; however, it is consistent with a large body of data that has demonstrated the superiority of IgG2a against C. neoformans. In contrast, naïve control mice had higher levels of GXM-specific IgG2b. GXM-specific IgG2b conferred the shortest survival time and/or was detrimental in murine CD (47, 57, 70) and was the least effective IgG subtype in promoting phagocytosis of C. neoformans (58, 70). Hence, the antifungal activity of effector phagocytes in control mice might have been hampered by the relatively high level of specific IgG2b and low level of IgG2a in these mice. sIgM−/− mice also had a higher level of GXM-specific IgG3. Given that B-1a B cells produce IgG3 (12), this could also be due to the expanded B-1a B-cell population in these mice. While IgG3 was not protective against CD in A/J and C57BL/6 mice, it did protect C57BL/6 × 129Sv mice (54). Hence, the IgG3 response of control (C57BL/6 × 129Sv) mice in this study might reflect a genetic factor, although the specificity of naturally occurring IgG3 might not be the same as the specific IgG3 (45) that was used in the foregoing studies.
Our data show that serum levels of TNF-α, IFN-γ, IL-6, IL-1β, and MIP-1β were significantly higher among naive sIgM−/− mice than among control mice, with the most pronounced differences being in the levels of IL-1β and IL-6. While the differences between mouse strains for other mediators were more modest, they were statistically significant. The importance of IFN-γ in driving the development of Th-1 immunity to C. neoformans has been well documented (37, 39), and IL-6 and TNF-α have each been associated with fungal clearance (2, 11, 37, 61, 62, 76). To our knowledge, IFN-γ levels in naïve sIgM−/− mice have not been examined previously. However, it was hypothesized that the elevated levels of IgG2a and IgG3 in these mice could be due to increased levels of IFN-γ (12), a switch factor for these IgG subtypes (64). Given that IFN-γ can also inhibit B-1 B-cell proliferation (12), the lower level of B-1b B cells in control mice than in sIgM−/− mice at 1 week postinfection might stem from the higher level of IFN-γ in control mice at this time. The role of IL-6 in phagocytosis of C. neoformans (8, 11, 62) and that of beta chemokines in antifungal activity and promotion of adaptive immunity (27, 34, 62, 72) are well documented. Hence, the inflammatory milieu of naïve sIgM−/− mice, featuring factors with the ability to induce fungal clearance (IL-6 and beta chemokines) and promote Th-1 immunity (IFN-γ and TNF-α), and differentiation of B-1a B cells to phagocytes (IFN-γ) (40), might provide a “ready-made” environment for generating an antifungal response that improves survival. Caveats to our results are that the levels of these factors that are required to mediate biological activity against C. neoformans are unknown and that another methodology, such as expression profiling, might reveal a more dynamic picture of the inflammatory profiles of the mice.
B-1a B cells produce IL-10 (49), most likely explaining the markedly elevated level of this cytokine in naïve sIgM−/− mice in our study. IL-10 has been increasingly recognized to have a regulatory role (1, 31, 36), including modulating the murine inflammatory response to C. neoformans (29). B-1a B-cell-derived IL-10 has been linked to B-1a B-cell localization in the peritoneum (6). Such localization might explain in part why sIgM−/− mice exhibited a resistant phenotype in our intraperitoneal infection model but were more susceptible to pulmonary infection with other pathogens (7, 14) and why B-1a B cells were detrimental in pulmonary Paracoccidioides brasiliensis infection (53). The level of IL-13 was also markedly elevated in naïve sIgM−/− mice. Although IL-13 had a detrimental effect in pulmonary models of CD (28, 48, 66), its role in systemic CD is unknown. Interestingly, consistent with evidence that it can also have a beneficial regulatory function (52), pulmonary clearance of Staphylococcus aureus in mice required IL-13 (50). Given that phagocytosis enhances the secretion of inflammatory mediators (56, 70) and that IgG2a binds activating phagocyte Fcγ receptors (FcγRs) and complement, IL-10 and IL-13 might enhance host defense in our model by regulating the inflammatory response. While this hypothesis requires validation, it is consistent with observations from other models of infection and inflammation (50, 52). Increased levels of IL-10 and IL-13 in naïve sIgM−/− mice, and the induction of IgE in both mouse strains, are suggestive of a Th2-skewed response. On the other hand, based on specific IgG2a/IgG1 ratios, which have been used by many groups as a surrogate marker for Th-1 (higher ratio)/Th-2 (lower ratio) balance (18, 21, 25, 44, 71, 74), sIgM−/− mice manifested a more pronounced Th-1 response than control mice at 1 and 3 weeks postinfection. Nonetheless, our data revealed relatively small differences in the cytokine profiles of sIgM−/− and control mice and, as reported for antibody immunity to CD (8), demonstrated that both Th-1 and Th-2 cytokines were associated with the resistance phenotype of sIgM−/− mice.
Our lower-inoculum model (105 CFU) suggests that the survival advantage of sIgM−/− mice after systemic C. neoformans infection stemmed from the ability of naïve sIgM−/− mice to enhance phagocytosis and limit fungal proliferation. A similar trend that was not statistically significant was observed in the higher-inoculum model (107 CFU). Given that the inoculum influences the nature of the murine response to C. neoformans (69, 70), our results suggest that the host factors which govern fungal immunity might differ as a function of the inoculum and that the inoculum, which can be controlled before the development of a mature, adaptive response is limited. Based on this scenario, memory IgM B cells, the human homolog of B-1a B cells (15) which have been linked to resistance to HIV-associated CD (68), might only be able to control the proliferation of a certain fungal burden, beyond which their antifungal effect would be overwhelmed. We note that the novelty of our discovery that sIgM−/− mice are more resistant to systemic C. neoformans infection in an inoculum-dependent manner pertains to an intraperitoneal infection model. Our findings might not pertain to a pulmonary model. Although this question is now under investigation, our findings hold promise for advancing our understanding of resistance to HIV-associated CD, which presents predominantly as a systemic disseminated disease.
This research was supported by the National Institutes of Health grants R01 AI 35370 and 45459 to L.P. and the Molecular Pathogenesis of Infectious Diseases Training Grant T32 AI 007506-10 to K.S.S.
None of the authors have a conflict of interest relevant to this article.
Editor: J. L. Flynn
Published ahead of print on 9 November 2009.
†Supplemental material for this article may be found at http://iai.asm.org/.