Group A Streptococcus interactions with the host across time and space

Abstract

Group A Streptococcus has a fantastically wide tissue tropism in humans, manifesting as different diseases depending on the strain’s virulence factor repertoire and the tissue involved. Activation of immune cells and proinflammatory signaling has historically been considered an exclusively host-protective response that a pathogen would seek to avoid. However, recent advances in human and animal models suggest that in some tissues group A Streptococcus will activate and manipulate specific proinflammatory pathways to promote growth, nutrient acquisition, persistence, recurrent infection, competition with other microbial species, dissemination, and transmission. This review discusses molecular interactions between the host and pathogen to summarize how infection varies across tissue and stages of inflammation. A need for inflammation for GAS survival during common, mild infections may drive selection for mechanisms that cause pathological and excess inflammation severe diseases like toxic shock syndrome, necrotizing fasciitis, and rheumatic heart disease.

Introduction

The obligate human pathogen Streptococcus pyogenes (Group A Streptococcus; GAS) is one of the top infectious causes of human mortality and is responsible for more than 500,000 annual deaths worldwide (reviewed in [1]). Disease disproportionally occurs in low-resource settings and there is no vaccine [2]. Most infections, more than one billion annually, are superficial and relatively mild, such as impetigo and pharyngitis. However, these infections are the trigger for immune sequelae like rheumatic heart disease [3]. Annually, millions of these mild infections develop into severe disease, including bacteremia, cellulitis, puerperal sepsis, Streptococcal toxic shock syndrome, and necrotizing fasciitis.

Since few infections turn severe, it can be inferred that the physical barrier formed by the skin and mucosa are critical determinants in preventing invasion. GAS often lives at these barrier sites with little-to-no disease, which can present a challenge for appropriately scaling the immune response. Conventional pattern recognition receptors are adept at recognizing microbe-associated molecular patterns, including those of GAS [4]. However, these can be too non-specific to discriminate a threat from microbiota that must be tolerated. Detection of virulence factors needed for advanced disease could provide a clue to the immune system on how much threat a specific species or strain poses. We discuss recent mechanistic insights and new experimental models that inform the fundamental principles of how this host-pathogen interaction varies between body sites (space) and the escalation of immune response (time).

Host inflammatory response to upper respiratory tract infection

Acute pharyngitis and tonsillitis are common childhood ailments occurring when the nasopharyngeal mucosa, tonsils, and adenoids inflame in response to infection. Bacterial growth at this site requires 1) binding to cells and resisting removal by mucus, 2) competition with the microbiota for nutrients, and 3) resistance to killing by immune factors. Most of our knowledge about the specific immunology of this site primarily comes from experimental murine infections, where intranasally-inoculated GAS adhere to the mucosa and exhibits tropism for lymphoid tissue. Despite anatomical differences between mice and humans, there is inflammation and pathology resembling human disease, and this allows for studies on the mechanistic contributions of specific virulence factors or immune effectors using knockout bacteria or mice. A new acute pharyngitis challenge using healthy volunteers is providing insights into the commonalities of human immunity to that observed in the mouse model [5,6]. Over three days, individuals who developed signs of pharyngitis had increased saliva IL-1β, IL-6, and IL-18, and systemic increases of these cytokines and IFNγ [5,6]. This strong acute-phase proinflammatory cytokine profile is similar to the experimental mouse pharyngitis model [7–10]. The elevations of IL-1β and IFNγ stand out since prior studies in the mouse pharyngitis model suggest that the inflammatory programs they control potentially have significant mechanistic contributions to disease progression [7–9].

IL-1 receptor knockout mice and mice administered the IL-1 receptor antagonist drug Anakinra had broad decreases in other inflammatory cytokines, suggesting IL-1 has a central role in the regulation of inflammation at this site [7]. Surprisingly, however, reducing this inflammation also reduced the bacterial burden, showing there are circumstances where inflammation may be beneficial for GAS [7]. GAS growth is antagonized by microbiota species [11], and IL-1 signaling dependent on the virulence factor SpeB drove neutrophil recruitment that allowed GAS to overcome this interference [7]. Capsule and the secreted DNase Sda1 have been recently shown be required to resist neutrophil killing in this model [10,12], meaning their requirement is at least in part to resist the inflammation induced by SpeB. However, these factors have additional virulence contributions. For example, capsule can promote cell adherence and experimental pharyngitis by binding the hyaluronic acid receptor CD44 on airway cells [13]. In addition to subverting immune pathways, Sda1 may contribute to microbiota interactions, since DNases secreted by other Streptococci have been recently shown to degrade DNA-containing biofilms made by other upper respiratory tract species for an advantage in niche competition [14].

The absence of IFNγ in some mouse models [7] compared to the human challenge [5] suggests the involvement of superantigens. These toxins bridge antigen-presenting cells MHC-II receptor to the T cell receptor, leading to antigen non-specific activation of human but not mouse T cells and a cytokine storm that includes IFNγ [9]. While IFNγ is typically important for controlling intracellular pathogens, transgenic mice carrying the human Vβ chain of MHC-II, superantigens promote nasopharyngeal inflammation and the replication of GAS [8]. T cell ablation broadly decreases the production of proinflammatory cytokines in this model, arguing that they, and superantigens, are key arbiters of inflammation [9]. T cells ablation also decreased GAS numbers, suggesting as with IL-1 and neutrophils, a robust inflammatory response involving T cells is also required for GAS infection of the nasopharynx. Since IL-1 also acts on T cells, such as to drive the expansion of pathological GM-CSF-expressing CD4 T cells in acute rheumatic fever [15], these inflammatory pathways are not necessarily divergent.

Host inflammatory response to skin infection

The naturally declining incidence of diseases in the skin’s descending layers suggests that GAS’s sufficiency to bypass the cornified and underlying epidermal layers of the skin is limited and that infections of the deeper tissue are promoted by dermal injury. As in pharyngeal infection, necrotizing soft tissue infections of humans and mice feature highly elevated levels of the proinflammatory cytokines IL-1β, IL-6, IL-18, and TNFα [16,17], with SpeB directly leading to increased activation of IL-1β and IL-18 [18,19]. Elevated risk of invasive infection is associated with NSAIDs or biologics targeting IL-1β or IL-6, suggesting rapid induction of inflammation is important for controlling infection once this barrier is breached [18,20,21]. Notably, for IL-1β, this immune requirement is the opposite of what is seen in the pharyngitis model previously discussed, where it instead may drive disease. Activities have recently been separated between the IL-1 cytokines; IL-1β is important for inducing emergency granulopoiesis that restricts GAS dissemination, while IL-1α has a non-redundant role that does not enhance immunity at the infection site, but promotes tolerance and prolongs mouse survival [22].

Other inflammatory pathways may have similar roles throughout the body. Superantigens increase the lethality of experimental GAS soft-tissue infection [23]. The mechanism for this is unknown, but increased lethality is likely not a factor selected in the evolution of GAS. One clue comes from the other major superantigen-producing human pathogen, Staphylococcus aureus, which uses superantigens to promote bacterial survival in a bacteremia model [24]. This was recently shown to result from T cells producing increased IFNγ, which, instead of promoting the killing of phagocyted bacteria as typical, was pathological and impaired antimicrobial activity [24]. Given the conservation of superantigens between species, GAS skin and soft tissue infections may involve similar mechanisms.

Cell death is also an important contributor to inflammation and the anti-GAS immune response. Pyroptosis was originally described by Cookson and Brennan [25] as “pyro,” relating to fire, and “ptosis”, falling, in contrast to the immunologically silent “apoptosis.” It is worth noting that early literature may refer to any death program as apoptosis, which may not proceed by the mechanisms we currently define as apoptosis. Pyroptosis is initiated by gasdermin-family cell death effectors regulated by proteolysis, such as by the inflammasome-associated caspases. Gasdermin cleavage leads to lysis, which releases damage-associated molecular patterns that promote strong innate and adaptive immune responses (reviewed in [26]) and deprives pathogens of a protected intracellular niche, which can contribute to antibiotic failure and recurrent infection [27]. GAS has been widely observed in vitro and in vivo to efficiently invade epithelial cells [28], and evade autophagic killing by degrading essential regulatory ubiquitin-LC3 adaptor proteins with SpeB [30]. Two recent papers show that skin cells defend against this fate by committing to pyroptosis through a unique mechanism of gasdermin A (GSDMA) cleavage by SpeB [31,32]. Since no endogenous protease has been found to activate GSDMA, this gasdermin may act as a pathogen sensor to detect foreign proteases, following the model of effector-triggered immunity [33]. Since SpeB is essential for intracellular survival, and can only activate GSDMA after GAS have already directed their internalization and escape into the cytosol, GSDMA may thereby sense the pathogenicity through the confluence of these activities [31]. Indeed, the immune response to GAS is impaired in GSDMA-knockout mice, and while GSDMA is not essential in the defense against SpeB-mutant bacteria, the essential virulence activities of SpeB renders these strains attenuated [31,32].

Microbial responses to inflammation

While excess inflammation underlies complications and post-infectious sequelae, this may not be evolutionarily beneficial for GAS, which likely instead gains specific benefits to help it avoid clearance from the host or promote its transmission. A recent work suggests that inflammation is advantageous during its competition with resident species [7]. Prior antibiotic administration has long been known to be associated with infection in humans and to increase the susceptibility of mice in the nasopharyngeal infection model, suggesting antibiotic-sensitive members of the resident microbiota ordinarily interfere with GAS growth [34]. After antibiotic disruption, neutrophils and IL-1 signaling are no longer required for GAS growth and shedding, arguing a mechanistic role for this immune response in overcoming competing microbiota species [7]. Whether the growth advantage that GAS gains from targeting T cells with superantigens occurs in part from effects on the microbiota is not yet known. However, high superantigen expression is directly responsible for scarlet fever, which is highly infectious, supporting a role in transmission [8,35]. Furthermore, superantigen-activated T cells can kill B cells, reducing antibody production [36,37]. While this mechanism may not be responsible for the rapid attenuation of superantigen-mutant GAS in the murine acute pharyngitis model, over the arc of natural disease the impaired responses may allow for recurrent infections [36].

To fully benefit from inflammation, it would be advantageous for GAS to detect it to alter its gene regulation to the changing environment. During the escalation of the host-pathogen interaction, GAS has numerous virulence factors that can have condition-specific benefits including capsule, M protein, SLO, SpeB, NAD glycohydrolase, SpyCEP, Sda1, and IdeS. Most of these are regulated directly or indirectly by the two-component system CovRS (or CsrRS), a sensor of the host defense peptide LL-37 [38]. LL-37 has both immune signaling and antimicrobial activity, but many GAS are fully resistant to direct killing by this peptide [39]. LL-37 levels are low and may not fully induce CovRS signaling in saliva and healthy skin, but are induced upon infection and highly expressed by infiltrating professional immune cells [40,41]. Detecting changes in LL-37 concentration can thus give GAS a readout of its inflammatory environment (Figure 1). Similarly, since some superantigens are induced by iron chelation [42], lactoferrin in mucus, transferrin in blood, calprotectin from neutrophils, and possibly iron species released from toxin-mediated hemolysis can all influence their regulation as infection progresses.

Figure 1:

Interplay between inflammation-induced cathelicidin/LL-37 and reactive oxygen species (ROS) and bacterial virulence. Concentrations of LL-37 range from undetectable in healthy skin, 50–300 nM in saliva, which can induce CovRS signaling. Many GAS are resistant to 10 μM or higher concentrations. Concentrations of the ROS hydrogen peroxide range from low μM levels in the phagosome concentrations with GAS resistant to 500 μM or higher concentrations. Inflammation increases LL-37 and ROS concentrations in phagocytes, phagosomes, and neutrophil extracellular traps, and when released, are proinflammatory themselves.

The changing tissue environment can also provide clues for the post-translational regulation of microbial virulence factors. One such mechanism relies upon changes in the redox potential studied in several recent manuscripts. Like many other enzymes with reactive cysteines, SpeB can be inactivated in a reducing environment [43]. Neutrophils can release reactive oxygen species and create an oxidative environment at the infection site with the potential to inactive SpeB [44] and the pore-forming toxin SLO [35], both major pro-inflammatory toxins. Glutathione is a primary antioxidant in humans and promotes the growth and reactive oxygen species resistance of GAS [45]. SLO can release glutathione from cellular stores [45], allowing its antioxidant activity to rescue SLO function [35], superantigen expression [35], and SpeB activity [44]. Thus, while the oxidative burst from immune cells can lead to a virulence switch by inactivating pro-inflammatory virulence factors, glutathione released by cell intoxication and death could counter this. Other redox sinks, such as the hyaluronic acid capsule, may also contribute to buffering these oxidation effects [17]. Additional recent evidence suggests local redox impacts which carbon utilization pathways are active and that this has an important role in disease pathogenesis [46].

Conclusions

GAS is highly heterogeneous with seemingly few “one-size-fits-all” virulence strategies active for all clones, all body sites, and all times of infection (Figure 2). Isolates can be clustered into groups by homology broadly reflecting common activities and association with infection of different body sites [47]. Most serotypes can colonize the nasopharynx of children [48], while fewer have a skin tropism [49]. These observations suggest that specific, variable, virulence factors are needed to overcome challenges specific to each niche, and that these are regulated by inflammatory cues in the environment. A better understanding of when and where bacterial factors are important is needed to identify the best targets for a broad-acting vaccine, and a better understanding of the complexities of inflammation during infection can lead to improved treatments to prevent rheumatic fever, Sydenham’s chorea, and other immune sequela.

Figure 2:

Host-pathogen interactions are divergent between body sites. GAS infects the skin and upper respiratory tract, secreting exotoxins that trigger a host inflammation. The GAS exotoxin SpeB activates IL-1ß and pyroptosis via the cell death effector Gasdermin A, both of which recruit neutrophils to the infection site. GAS superantigens activate T lymphocytes, resulting in IFNγ, IL-6, and TNFα production. While the molecular mechanisms of these interactions are similar in the upper respiratory and skin, the consequence of this inflammation is tissue-dependent.

Highlights.

• Group A Streptococcus infects many body sites, causing clinically-distinct diseases

• Immune responses can promote or inhibit pathogenesis, depending on tissue and timing

• Infection of a site depends on the specific virulence factor repertoire, which is variable

• Inflammation alters bacterial virulence factor expression and function in advanced disease

Data Availability

No data were used for the research described in the article.