Recurrent infections and immune evasion strategies of Staphylococcus aureus

Introduction

The Gram-positive pathogen Staphylococcus aureus causes a wide swath of human diseases including skin and soft tissue infections (SSTI) and invasive diseases that lead to bacteremia, sepsis, endocarditis or pneumonia [1]. S. aureus colonizes the skin and nares of 20-30% of the human population [2]. Due to the frequent use of antibiotics, S. aureus strains have evolved resistance against the most abundantly used therapeutics [3]. These drug-resistant strains are historically referred to as methicillin-resistant S. aureus (MRSA) [4]. Infections with methicillin-sensitive S. aureus (MSSA) or MRSA originate both in the community and in hospitals [5,6]. The therapy of severe MRSA infections is complicated by the fact that these strains are susceptible to only few antimicrobials – vancomycin, daptomycin or linezolid [5]. Because of the severity of invasive disease, MRSA infections are associated with a poor outcome even when appropriate antibiotic therapies have been implemented [7,8]. A key feature of staphylococcal SSTI is its recurrence, which occurs in approximately 30% of all cases. These clinical observations as well as experiments with animals that had been repeatedly challenged with S. aureus suggest that infections with this pathogen do not generate protective immune responses [9,10]. The current epidemic of community- and hospital-acquired MRSA infections in developed and developing countries is testimony for the successful spread and immune evasive attributes of this pathogen [11]. Here we review what is known about the immune evasive strategies of S. aureus.

Staphylococcal immune evasion strategies - an overview

Upon entry into subepidermal tissues or blood, S. aureus encounters the cellular and proteinaceous elements of host innate immune defenses. S. aureus is uniquely programmed to compromise the effectiveness of both components by secreting proteins that inhibit complement deposition or activation as well as the chemotaxis of polymorphonuclear leukocytes (neutrophils) [12-17]. Other secreted polypeptides display lytic activities towards neutrophils, the primary line of defense against S. aureus infections (reviewed in reference [16]). Several of these secreted products are encoded by two gene clusters designated Immune Evasion Cluster 1 and 2 (IEC) that appear to have been acquired through integration of prophages [18]. IEC encoded products include superantigen-like proteins that act on human cells or human complement components, not on those of other mammals, suggesting a pathogenic strategy of host specific adaptation [19,20]. Consequently, the contribution of IEC1 and IEC2 to disease pathogenesis has not yet been explored in animal models for S. aureus infection. IEC also encodes staphylokinase, a bacterial plasminogen activator that mediates cleavage of opsonins by activating plasmin [4], and provides resistance towards antimicrobial peptides such as α-defensins [5]. Details on the physiological and molecular attributes of IEC encoded immune evasion strategies and the variation of these genes between S. aureus strains can be found in several excellent reviews and are not described in detail here [17,20,21]. In addition to host-specific virulence factors, staphylococci also deploy universal modulators of inflammation by increasing the abundance of the host signaling molecule adenosine (Figure 1A). This immune evasion strategy was revealed through the discovery of AdsA, a virulence factor that is encoded in the core genome of the chromosome (see below).

Staphylococcal infections in humans result in a transient increase in anti-staphylococcal antibody levels [22-24]. Nevertheless, protective immunity is not observed and recurrent infections occur frequently [1]. Clearly, immunization with defined antigens can elicit protection in animal models of S. aureus disease, suggesting that during infection antibody levels may simply be too low to be protective [25,26]. Two classes of immunomodulatory factors are secreted by S. aureus and compromise the induction of both humoral and cell-mediated immunity: enterotoxins that function as T cell superantigens [27] and the B cell superantigen staphylococcal protein A (SpA) [28]. Injection of either of these factors in purified form into animals is associated with the depletion of T or B cell populations, respectively. The T cell superantigens bind MHC class II determinants on the surface of antigen-presenting cells even in the absence of antigen peptide [29]. This binding triggers functional association with T cell receptors (TCR) on the surface of T helper cells [30,31]. Enterotoxins are commonly expressed by clinical isolates of S. aureus and encoded on mobile genetic elements [32]. Each type of enterotoxin recognizes a specific subset of variable Vβ chains of T-cell receptors via its superantigen domain [33]. While it is clear that enterotoxins such as TSST-1 cause an emetic response when ingested, the contribution of superantigen activity during SSTI or other infections is still poorly understood. In rabbits, where purified T cell superantigens display extraordinary toxicity, at least some of the staphylococcal enterotoxins have been shown to contribute to the pathogenesis of lethal pneumonia and that preexisting immunity can prevent the lethal outcome of these infections [34]. Similarly, humans develop neutralizing antibodies against a T cell superantigen, however cross protection against other staphylococcal T cell superantigens has not been observed [35]. The considerable variability and functional redundancy of staphylococcal enterotoxins and toxic shock syndrome toxins has prevented the development of a superantigen vaccine capable of neutralizing at least several of these T cell superantigens.

Protein A has been shown to bind to the V_H3 region of IgM molecules [36] exposed on the surface of B lymphocytes [37] (Figure 1B). V_H3 type B-cell receptors (BCR) are present in about 30% of human B cell populations. When injected into mice, purified SpA promotes the proliferation and apoptosis of B cells in both the spleen and the bone marrow [37] (Figure 1B). In mice, B-1 and marginal zone B cells as well as innate B cells, which secrete natural antibodies, are most susceptible to SpA treatment in comparison with B-2 cells or follicular B cells that generate antigen-specific antibodies [28,38]. Once depleted, B cells cannot be replenished readily in SpA-treated animals [37]. SpA interacts with several host proteins and is a key anti-opsonic factor (see below). In agreement with this model, mutants that no longer produce protein A display reduced virulence in multiple animal models for staphylococcal diseases [39-41]. Most relevant to the immune-modulatory attributes of this molecule is the observation that mice infected with spa mutants, but not wild-type S. aureus, elicit protective immunity against secondary infection with wild-type strains in a manner that correlates with the production of anti-staphylococcal antibodies [42]. Like humans, mice with a history of staphylococcal infection have not acquired protective immunity and mount only weak humoral responses to S. aureus infection [42]. These observations support a model whereby S. aureus modulates B cell responses via its SpA surface antigen, a process that may be amenable to therapeutic interference.

Controlling inflammation - adenosine synthase A (AdsA)

Adenosine synthase A (AdsA) is an immune evasion factor that was initially identified in a genetic screen probing for the relative contributions of cell wall anchored surface proteins towards staphylococcal survival in blood [43]. Both wild-type and adsA variants are phagocytosed by neutrophils when inoculated into fresh blood, however wild-type S. aureus survives within neutrophils whereas adsA mutants are killed [43]. Similar to SpA, AdsA is synthesized as a precursor with an N-terminal signal peptide and a C-terminal sorting signal. Following initiation into the secretory pathway, the C-terminal sorting signal is cleaved by sortase and covalently linked to the cell wall envelope of staphylococci [44] (Figure 2). AdsA harbors two 5’-nucleotidase signature sequences and purified recombinant AdsA catalyzes the hydrolysis of phosphate esterified at carbon 5’ of ribose or deoxyribose within purine nucleotides with adenosine monophosphate (AMP) being the preferred substrate [43]. The resulting conversion of adenosine monophosphate (AMP) to adenosine had been observed earlier in 5’-nucleotidases that are present in the cytoplasm of many cells [45]. Following inoculation of staphylococci into blood, AdsA surface protein is responsible for a dramatic increase in the overall abundance of extracellular adenosine, an activity that is absolutely dependent on its 5’-nucleotidase sequences [43].

Adenosine represents the perhaps most potent immunosuppressive signaling molecule. It is formed by cells following hypoxic stress, exposure to reactive oxygen species (ROS) or cell lysis associated with severe cell damage [46,47] (Figure 1A). Extracellular adenosine interacts with one or more of its adenosine receptors - A1, A2A, A2B and A3 - all members of the family of G protein coupled receptors [47,48]. Adenosine receptor interaction triggers anti-inflammatory signaling cascades that cause the inhibition of platelet aggregation [49], of neutrophil superoxide bursts or degranulation [50], of T cell adherence to fibroblasts and interleukin 1α (IL-1) production as well as the increased production of IL-10, an anti-inflammatory cytokine [51]. Thus, adenosine is thought to be the key signal controlling excessive inflammation within mammalian organisms. Staphylococci evolved adsA to provide the adenosine signaling molecule during infection, thereby escaping both innate and adaptive defense mechanisms otherwise essential for the clearance of bacterial pathogens by the host.

The finding that AdsA is necessary for staphylococcal survival within neutrophils provides a rational appreciation for earlier observations, demonstrating that S. aureus escapes the bactericidal attributes of leukocytes [9,52]. Gresham and colleagues demonstrated that the injection of neutrophils harboring S. aureus into the blood stream of naïve mice allowed the pathogen to disseminate within the host and to develop staphylococcal disease [53]. Taken together, these observations suggest a model whereby staphylococci exploit their intracellular location in neutrophils to disseminate from the bloodstream into peripheral tissues. Presumably this journey is completed by triggering neutrophil diapedesis through endothelial barriers [54]. Entry into the neutrophil vehicle comes at a price, as staphylococci have to tame all bactericidal responses of these cells including reactive oxygen species (ROS), hydrolytic enzymes and peptide defensins [54]. There are, however, considerable gains. The intracellular location of staphylococci may provide protection from complement, antibodies and extracellular host defensive peptides as well as the many other cells of the host’s immune system, most notably phagocytes (macrophages and dendritic cells) that can kill bacteria and present antigens to elicit adaptive immune responses [55].

Recent studies proposed that ROS in neutrophils exert not only a canonical bactericidal activity but also activate hydrolytic enzymes such as myeloperoxidase, cathepsin G, elastase, proteinase 3 as well as other hydrolases that degrade bacterial invaders [56]. It is not yet known whether staphylococcal adenosine production in extracellular fluids or within neutrophils can disrupt the formation or maturation of phagolysosomes, interfere with the generation or secretion of bactericidal molecules or induce cellular death mechanisms such as apoptosis or necrosis. Whatever the detailed mechanisms may be, the ability of S. aureus to avoid bactericidal activities of professional phagocytes has profound consequences on the development of humoral immunity. Adenosine decreases the expression of MHC-II on macrophages and dendritic cells and it decreases macrophage production of IL-12, a pivotal stimulus for Th1-type immune responses [48,57,58]. Skewing adaptive immune responses during infection may also precipitate inadequate T cell effector mechanisms, thereby further disabling adaptive immune responses (Figure 1A).

Immunoglobulin binding molecules of S. aureus

Therapeutic and preventive strategies to tame staphylococcal immune evasion

As is outlined above, staphylococci deploy a wide spectrum of strategies to avoid innate immune attacks such as complement deposition and killing by neutrophils. The complexity and functional redundancy of factors engaged in innate immune evasion renders the development of therapeutic or preventive approaches arduous. Similarly, variability of an immune-modulatory trait among S. aureus strains is an exclusion criterion for vaccine development. In contrast to most genes in the IEC clusters, SpA, Sbi and AdsA are universal traits of all S. aureus strains and may therefore offer opportunities for the development of vaccines or therapeutics [42].

AdsA is a homolog of CD73, the mammalian adenosine synthase (5’nucleotidase) [43]. Although this has not yet been tested, AdsA immunization may elicit antibodies that interfere with CD73 function, thereby promoting inflammatory responses in vaccinees. To establish a rational basis for AdsA vaccines or therapeutics, future work would need to study the specificity of immune responses to this molecule and the ability of specific antibodies to neutralize the bacterial enzyme without affecting the host’s ability to synthesize adenosine.

Wild-type SpA cannot be used as a vaccine antigen because of its B cell superantigen activity [10]. Each SpA IgBD domain adopts a triple α-helical bundle structure [79] and the atomic interactions with Fab V_H3 [66] or Fcγ [80] have been elucidated. In a recent study, glutamine 9 and 10 as well as aspartate 36 and 37 were identified as critical for the association of protein A with immunoglobulins [10]. When four substitutions - Q⁹K, Q¹⁰K, D³⁶A, and D³⁷A - were introduced in each of the five IgBDs domains, the new variant, designated SpA_KKAA, had lost B cell superantigen and Fcγ binding activities [10]. Immunization of rabbits or mice with SpA_KKAA, but not with wild-type SpA, generated high titer antibodies to this antigen. Further, ^SpAKKAA derived antibodies blocked the association of protein A with immunoglobulin as well as its B cell superantigen activity. Non-toxigenic SpA_KKAA elicited protective immunity against S. aureus challenge [10]. The resulting protein A specific antibodies facilitated opsonophagocytic killing of staphylococci in naïve mouse blood [10]. Further, following challenge with S. aureus, animals immunized with SpA_KKAA developed humoral immune responses to many different antigens, demonstrating that neutralization of SpA B cell-superantigen activity is paramount to the establishment of immunity [10].

Future prospects

Rapid spread of antibiotic resistance traits among S. aureus isolates as well as the acquisition of enhanced virulence attributes have precipitated a public health crisis that can no longer be addressed with the development of new antibiotics alone [81]. Several recent efforts have been directed towards the development of staphylococcal vaccines [82-86]. Although several staphylococcal antigens have shown promise in preclinical trials, those subjected to protective efficacy evaluation in clinical trials have not yet met the required end points for vaccine licensure [81,87,88]. The S. aureus vaccine field appears to recognize the need for multiple vaccine antigens to neutralize the key virulence strategies of this pathogen [25,26]. Nevertheless, the possibility of boosting the host’s ability to respond to S. aureus infection with protective immune response has not yet been explored. The recent observation that S. aureus spa mutants are unable to prevent animals from mounting a protective immune response correlates well with the know B cell-toxigenic activity of SpA [28,42]. This knowledge may be exploited for the development of new therapies by generating protein A variants that cannot escape host recognition and that elicit both opsonophagocytic and neutralizing antibodies [10].