An Emerging Role of B Cell Immunity in Susceptibility to Pneumocystis Pneumonia

Although clinical Pneumocystis pneumonia (PCP), has been associated with diseases affecting CD4⁺ T cell immunity (1), data over the last two decades point to a critical role of B cell immunity in this disease (2, 3). In this issue of the Journal, Hu and colleagues (pp. 322–331) performed a lung microarray study in a murine model of PCP—rendered susceptible to infection through corticosteroid treatment—an important risk factor in many HIV-negative cases of PCP (4). Importantly, they chose a 14-day time point after Pneumocystis inoculation, as this time point coincided with the induction of adaptive immune responses that are just initializing clearance, yet fungal burdens are quite similar between control and steroid-treated animals. Thus, differences in gene expression largely reflect defects in host immunity and are not skewed by differences in fungal burdens between the groups.

The authors conducted an unbiased analysis of the dataset, and observed a defect in many B cell–related genes, including Ig genes and Ms4a1 (CD20). This signature was confirmed by real-time PCR in lung tissue, which showed defects in a number of B cell genes, including Cd19, Cd22, Cxcr5, and Tnfrsf13c, the latter of which encodes the receptor for B cell activating factor. Aspects of this signature were also confirmed in the blood of steroid-exposed patients with clinical PCP. This is consistent with recent data showing that anti-CD20 has been associated with clinical PCP, and that anti-CD20 also renders otherwise immunocompetent mice susceptible to Pneumocystis murina infection (2).

Data in the murine model have shown that B cells play several key roles in immunity against PCP. First, they function as critical antigen-presenting cells in the lung, as absence of class II major histocompatibility complex on B cells results in defective CD4⁺ T cell immunity and susceptibility to PCP (3, 5). A similar defect in T cell priming occurs in mice treated with CD20 (2) or in mice lacking secreted IgM (6). Thus, in primary infection, B cells appear to be critical to elicit T cell immunity to this infection. Consistent with this notion is that some defects in Ig production have rarely been associated with PCP, including x-linked agammaglobulinemia (7) and common variable immunodeficiency (8). However, despite these reports, PCP is rare in pure B cell disorders, suggesting that there must exist B cell–independent mechanisms of CD4⁺ T cell priming. Indeed, adoptive transfer of CD4⁺ T cells into B cell–deficient Rag1^−/− mice can mediate clearance of the fungus (unpublished observations), suggesting that CD4⁺ T cells have intrinsic antifungal activity.

In addition to antigen presentation, class-switched Igs can prevent secondary infection, even in the absence of CD4⁺ T cells (9). Passive transfer of immune serum has been shown to prevent Pneumocystis infection as well (10, 11). Moreover, vaccination to elicit specific immune responses to surface antigens of Pneumocystis has been shown to be protective (12, 13). Taken together, these studies show that B cell immunity is critical—both in terms of activating CD4⁺ T cells, but also as an effector mechanism that can likely clear the organism through opsonic phagocytosis (14), as well as activation of complement (15). Some of this antibody response is directed against major surface glycoproteins of the organism (16), and now that we have draft genomes for the mouse, rat, and human organisms, this family of major surface glycoproteins is the most divergent between the fungal species. This may explain the host specificity of these organisms (rat Pneumocystis does not infect mice, and vice versa), similar to what has been elegantly reproved with other host-specific commensals, such as segmented filamentous bacteria (17). This raises the issue of whether infections in humans are mediated by quasispecies, such that, although the host may have had prior humoral immunity to PCP, as has been documented in young children (18), this humoral response may not be protective. A failure of these antibodies to cross-react with an infecting species may not be protective in an immunosuppressed host. These types of questions hopefully can be answered with more in-depth sequencing of human samples, as well assaying antibody responses to more than just major surface glycoprotein antigens. In conclusion, the current article shows that corticosteroids induce a major defect in B cells that is strongly correlated with PCP. These data need to be replicated, but this kind of biomarker may prove to be useful to more precisely identify and define individuals at risk for developing PCP in patients that require steroid-containing medical regimens.