NADPH oxidase regulates efficacy of vaccination in aspergillosis

Abstract

Invasive aspergillosis is often a consequence of immune suppression, and accumulating evidence points to a role for adaptive immunity. Hence, it may be possible to manipulate the adaptive immune system to enhance protective immunity in at-risk individuals. In this issue of the JCI, De Luca and colleagues describe the ontogeny of adaptive immune responses to murine aspergillosis infection in relation to vaccination. Their thought-provoking findings reveal the complexities of vaccine-induced immunity and could be used to improve vaccine efficacy.

Aspergillus spp constitute an omnipresent mold, found in air, water, and soil. Conidia — spores that are released from hyphae — are inhaled daily by humans. The major variable in human exposure is the number of particles, and this value may be low indoors and strikingly high in areas that contain decaying matter. The species of Aspergillus that cause invasive infection are limited, but these share an important trait: the ability to grow at 37°C, human body temperature. Among those that cause invasive disease in humans, A. fumigatus is the most prevalent followed, by A. flavus (1, 2).

Despite constant contact with conidia, the immunocompetent host rarely develops invasive disease, largely because immune defenses in lungs are highly effective. Infection is initiated when host resistance is dampened, and spores convert into hyphae that disseminate through the vascular system. Several risk factors predispose individuals to infection with Aspergillus. Chief among them are the use of glucocorticoids to treat inflammatory disorders, severe neutropenia induced by cytotoxic agents, or, less frequently, the use of TNF-α antagonists. The risk of disease acquisition escalates in parallel with the length of glucocorticoid administration or the duration of neutropenia (1). Qualitative defects in neutrophil function also increase the probability of acquiring disease. One of the more common disorders associated with dysfunctional phagocytes is chronic granulomatous disease (CGD), an inherited disease in which the fundamental genetic defect is in the assembly of the NADPH oxidase (3). Since the leukocytes from these patients fail to mount a vigorous respiratory burst, their phagocytes are impaired in their ability to combat numerous pathogens including bacteria and Aspergillus spp. Despite many advances in both early diagnosis and antifungal agents, the attributed mortality for all patients still ranges from 40% to 70% (2).

Cellular immunity to Aspergillus

Although innate immunity appears to be the principal host resistance mechanism, there is mounting evidence that cellular immunity also is crucial. Exposure to A. fumigatus spores induces the differentiation of T cells into Th1, Th2, and Th17 cells. The former are pivotal for generation of protective immunity to invasive aspergillosis, whereas Th2 cells are central to the pathogenesis of allergic bronchopulmonary aspergillosis, and they blunt Th1-mediated immunity in invasive infection. Th17 cells promote immunity, but apparently not with the same vigor as Th1 (4–7). Transfer of Th1-biased CD4⁺ cells ameliorates the severity of invasive infection, suggesting that they are the principal mediators of resistance (8, 9). The differentiation of CD4⁺ T cells into one of the aforementioned families is shaped by inflammatory monocytes and Dectin-1 (4, 10). Less well appreciated is the influence of CD8⁺ T cells on the host response. Antigen-reactive CD8⁺ T cells can be elicited by Aspergillus antigens, and protection relies on recognition of fungal RNA by TLR3 (11, 12). That T cells exert a strong influence on the host response has major implications not only for the immunopathogenesis but also for the design of prophylactic or therapeutic vaccines for this pathogen.

In this issue of the JCI, De Luca et al. (13) provide vital new information regarding the contribution of CD4⁺ and CD8⁺ T cells in host defenses against this devastating fungal pathogen. In this creative work, they use the p47^phox–/– model of CGD, in which the mice have defective NADPH oxidase, to decipher the antigen-presenting pathways in DCs that prompt CD4⁺ and CD8⁺ T cell activation. The authors report that antigen presentation by DCs that lack NADPH oxidase fails to prompt activation of CD8⁺ T cells.

Divergence of antigen presentation pathways in vaccine-induced immunity to Aspergillus

Unlike in other studies that have examined the T cell response to Aspergillus, De Luca and colleagues employed a model in which animals are immunized either with live conidia or with one of several recombinant antigens. They found that in wild-type mice, CD8⁺ T cells are essential for protection upon vaccination with conidia, whereas immunization with soluble Aspergillus proteins admixed with CpG confers protection in a CD4⁺ T cell–dependent manner (13). The finding that CD8⁺ T cells are the mediators of protection following immunization with conidia is somewhat surprising in light of previous work reporting that exposure to these fungal elements in primary infection induces a pronounced CD4⁺ T cell differentiation (4).

Conidia are engulfed into phagosomes, and no evidence exists that they escape into the cytosol, indicating that cross-presentation by DCs is the mechanism whereby MHC class I gains access to peptides from conidia (Figure 1 and refs. 14–16). Although cross-presentation is a property of both DCs and macrophages, De Luca et al. report that lung macrophages lack the capability of cross-presentation; however, this observation was not expanded upon. The efficacy of vaccination required the presence of IFN-γ and to a lesser extent IL-17; these results mimic those reported for primary infection (5). In striking contrast, De Luca and colleagues found that p47^phox–/– mice develop immunity only when immunized with the soluble antigens, indicating that processing of conidia in the cross-presentation pathway is impaired when DCs lack NADPH oxidase. This finding is not a consequence of an intrinsic dysfunction of CD8⁺ cells in these mice, which have been reported to have an elevation in the number of these cells (17).

Figure 1. Cross-presentation pathway by antigen-presenting cells.

DCs or macrophages engulf antigens into phagosomes. The phagosomes may fuse with endoplasmic reticulum, and antigens from the particle are directed to the proteasome where they are digested; cognate peptides with the appropriate linear amino sequence then bind to MHC I. This complex moves to the surface of an antigen-presenting cell and is subsequently engaged by a CD8⁺ T cell. An alternative pathway is that antigens “leak” or escape from the phagosome and enter the cytosol in the absence of fusion of phagosome/endoplasmic reticulum; the antigens in the cytosol are digested by the proteasome, bind to MHC I, and traffic to the cell surface.

Pattern recognition receptors dictate the antigen presentation pathways

The shaping of the host response whether by direct invasion of a pathogen or by a vaccine requires the engagement of pattern recognition receptors (PRRs). These molecules are influential in determining the differentiation of T cells. Recognition of conidia by CD8⁺ T cells required TLR3 and TIR domain–containing adapter-inducing interferon–β (TRIF). On the other hand, MyD88 and TLR6 were necessary for the action of one of the soluble antigens, and other soluble antigens had distinct TLR requirements for driving CD4⁺ T cell activation (13). This work highlights the complexity of antigen recognition at the level of PRRs.

The intracellular highway for conidia and soluble antigens

T cells recognize small peptide epitopes from antigens in the context of the cognate MHC. Proper trafficking of particulate or soluble antigens is a key step in the optimal generation of the epitopes that drive T cell activation. De Luca et al. demonstrated that protective soluble antigens engaged the mannose receptor and occupied organelles that contained Rab GTPases (13). The importance of the mannose receptor for soluble antigens is not surprising, given that the authors employed glycosylated antigens. Indeed, removal of the carbohydrate moieties abolished their protective activity, emphasizing the utility of glycosylation on immunogens.

De Luca et al. describe an orderly spatial and temporal sequence as the antigens navigated through the DC (Figure 2). Conidia trafficked to the late endosomes and subsequently to Rab 14⁺ compartments, whereas soluble antigen never occupied these compartments. Autophagy and endosomal alkalinization were fundamental for directing conidia to the Rab14-containing organelles and for cross-presentation. In the absence of phox47, conidia do not reach Rab14⁺ compartments, and cells displayed diminished autophagy. Restoration of autophagy enabled DCs to cross-present conidial antigens to CD8⁺ T cells (13). These results raise the possibility that antigens from conidia that have been digested by the intracellular machinery of DCs are captured within autophagosomes and processed for class I MHC export (15, 18).

Figure 2. CD4⁺ and CD8⁺ T cell recognition of Aspergillus antigens.

(A) Soluble antigens such as Pep1, when admixed with soluble antigen, enter the MHC class II pathway whether NADPH oxidase is present or not and promote the development of protective CD4⁺ cells. (B) On the other hand, particulate immunogens such as conidia enter the MHC class I pathway but only activate CD8⁺ cells in the presence of NADPH oxidase. The lack of this oxidase, as in CGD, is associated with a failure to induce autophagy and endosomal alkalinization. This defect results in absent activation of CD8⁺ cells.

Clinical implications

De Luca et al. raise a number of issues that are germane to clinical medicine. Their findings clearly point to a defective CD8⁺ T cell response in mice that lack NADPH oxidase and suggest an underlying mechanism that explains the defect. One concern is the hierarchy of protective immunity conferred by T cells in aspergillosis. Are CD8⁺ T cells subservient to CD4⁺ T cells? The present article would suggest that this might not be the case, but more definitive proof will need to be accumulated. It is also unclear whether these findings will translate to human patients. One study published nearly 20 years ago indicated that memory CD4⁺ cells were defective in patients with CGD (19), but virtually nothing is known about human CD8⁺ cells from CGD patients and cross-presentation pathways. The third issue is that induction of autophagy enhanced cross-presentation and consequently activation of CD8⁺ T cells, suggesting perhaps a more general means to improve antigen recognition. The data presented in this article are congruent with prior reports that rapamycin, an mTOR inhibitor and inducer of autophagy, promotes CD8⁺ T cell differentiation and memory responses directly or enhances the efficacy of vaccination by acting on DCs (20, 21). This agent is principally used to prevent graft rejection. Utilizing it as a means to boost vaccine responses and improve CD8⁺ T cells responses may be problematic given its side effects, but new compounds that act similarly may help to prevent infections with organisms that remain difficult to cure such as Aspergillus spp.