A play in four acts: Staphylococcus aureus abscess formation

Abstract

Staphylococcus aureus is an important human pathogen that causes skin and soft tissue abscesses. Abscess formation is not unique to staphylococcal infection and purulent discharge has been widely considered a physiological feature of healing and tissue repair. Here we present a different view, whereby S. aureus deploys specific virulence factors to promote abscess lesions that are distinctive for this pathogen. In support of this model, only live S. aureus are able to form abscesses, requiring genes that act at one or more of four discrete stages during the development of these infectious lesions. Protein A and coagulases are distinctive virulence attributes for S. aureus, and humoral immune responses specific for these polypeptides provide protection against abscess formation in animal models of staphylococcal disease.

Staphylococci and Staphylococcus aureus

Sir Alexander Ogston observed ‘micrococci’ in pus from surgical wound infections and named the bacterium Staphylococcus. Using eggs to isolate pure cultures for the inoculation of rabbits, he demonstrated that staphylococci cause the development of abscesses in infected animals. During the Ninth Surgical Congress in Berlin (1880), Ogston reported his results, firmly establishing staphylococci as the etiological agent of suppurative abscesses. To date, 36 species and 18 subspecies of the genus Staphylococcus have been distinguished^{1, 2}. Ogston’s microbe, Staphylococcus aureus, remains the most prominent member among this group, as it is the most frequent cause of human skin or soft tissue abscesses³. In addition to abscesses, S. aureus strains can cause pneumonia, toxic-shock syndrome, exfoliative skin disease, and enteritis³. Several unique biological traits are associated with S. aureus, including the ability to bind to immunoglobulin, agglutinate and coagulate blood or plasma (Figure 1)³.

Figure 1.

Staphylococcus aureus characteristics and distinguishing features. (a) The immunoglobulin binding properties of staphylococcal protein A were revealed by confocal laser scanning immunofluoresence microscopy as previously described⁸⁰ . Alexafluor-647 conjugated immunoglobulin (red) binds to protein A (left), Bodipy-conjugated vancomycin (green) stains staphylococcal peptidoglycan (middle), and a merger of the two data sets reveals the deposition of immunoglobulin on the surface of staphylococci (right). (b) Shape and clustering of S. aureus cells imaged by scanning electron microscopy (SEM; image courtesy of Matt Frankel, University of Chicago). (c) S. aureus coagulates blood, a feature conferred by two secreted coagulases: staphylocoagulase (Coa) and von Willebrand factor binding protein (vWbp)¹⁶. Lepirudin anti-coagulated mouse blood was incubated with 10⁵ cells of wild-type S. aureus Newman or isogenic mutants lacking coa, vWbp, or coa/vWbp and examined over time for the formation of a blood clot. Blood containing staphylococci lacking both coagulase genes do not form a clot following a 16 hour observation period.

S. aureus and several other staphylococcal species colonize the human skin, nails, and nares and disseminate among recipient host populations via physical contact and aerosols³. The massive consumption of antibiotics over the past 50 years has led to the selection of drug-resistant strains, designated MRSA for methicillin-resistant S. aureus⁴. Although no longer in clinical use in the United States, methicillin was initially used for the treatment of penicillin-resistant strains; penicillinase is unable to cleave methicillin. MRSA isolates are now widespread⁵. The basis for their resistance is a penicillin-binding protein that cannot be inhibited by methicillin or oxacillin. As a result, vancomycin has been used as the last-resort antibiotic for MRSA infections. However, S. aureus isolates that have vancomycin-resistant (VRSA) or vancomycin-intermediate (VISA) phenotypes have emerged by generating altered substrates for cell wall synthesis and no longer associate with vancomycin or related glycopeptide drugs^6–8. MRSA are a frequent cause of hospital-acquired infection in the United States⁹. Over the past decade, community-acquired infections (CA-MRSA) increased in abundance and severity⁹. One such clone has swept over the United States and currently colonizes a large segment of the population¹⁰. An FDA approved vaccine that can prevent S. aureus infections caused by MSSA (methicillin-sensitive S. aureus), MRSA, CA-MRSA, VISA or VRSA strains is not yet available.

Breaches in local defense, for example skin cuts or hair follicle trauma, provide opportunities for staphylococcal invasion, which typically manifests as abscess formation: the accumulation of pus accompanied by severe inflammation of surrounding tissues³. Staphylococci in pus disseminate onto skin surfaces or enter circulating lymph and blood, which leads to the formation of abscesses at new sites³. S. aureus is a frequent cause of sepsis, often the consequence of invasive disease originating from one or more abscesses. About one-fifth of staphylococcal skin and soft tissue infections that improve under antibiotic therapy flare up again at a later time¹¹. This phenomenon, as well as clinical data on S. aureus infected individuals, suggests that some humans do not acquire protective immunity following infection³. Here we review what is known about S. aureus abscess formation and the molecular mechanisms that enable the pathogen to form these lesions and suppress the immune responses of its host.

Abscess formation in stages

Staphylococci that disseminate through the bloodstream enter peripheral tissues and seed infectious lesions, initially inducing inflammatory responses that attract neutrophils, macrophages, and other phagocytes³. The responsive invasion of immune cells to the site of infection is accompanied by liquefaction necrosis and the release of pus with replicating staphylococci into circulating body fluids^12–14. Defensive host responses to staphylococcal invasion include fibrin deposits to shield healthy tissues from the disseminating microbe. Upon abscess drainage or surgical excision of infected lesions, lost organ tissues can be replaced with fibrotic scars. However, these events cannot be considered specific host responses to S. aureus invasion; they are induced even with chemical trauma, physical insult or injection of sterile biological materials into organ tissues¹⁵.

We suggest that abscess formation and staphylococcal entry into purulent material is a pathogen driven process that usurps the default responses of its infected host to enhance microbial replication and dissemination. In support of this hypothesis, several pathogen-specific genes are required for the ability of S. aureus to establish abscess lesions (vide infra)^{14, 16}. Following entry into organ tissues, staphylococci attract immune cells and replicate in their presence to eventually replace physiological epithelia with distinctive lesions¹⁴. Within four days of infection, the pathogen has organized itself as a staphylococcal abscess community (SAC) at the center of the lesion, shielded from host immune cells by a surrounding pseudocapsule comprised of fibrin deposits¹⁶. These lesions grow through the arrival of immune cells and the destruction of healthy tissues associated with neutrophil necrosis. Staphylococci disseminate from these lesions with the draining pus into circulating body fluids.

Using C57BL/6 and BALB/c mice in an intravenous challenge model, infected animals produced abscess lesions in all tissues examined similar to those observed in humans¹⁴. We liken the pathogenesis of staphylococcal disease to a play that is staged in four acts (Figure 2). Following intravenous injection, S. aureus survives in the bloodstream as an extracellular pathogen or when phagocytosed by neutrophils (Stage I). Upon distribution of staphylococci to peripheral tissues, neutrophils as well as other immune cells are recruited to the site of infection (Stage II). S. aureus reprograms infected lesions to replicate as a SAC at their center, separated from surrounding immune cells by a pseudocapsule with fibrin deposits (Stage III). Movement of lesions to organ capsules and their subsequent rupture releases S. aureus with purulent exsudate to other sites, where the pathogen repeats this infectious program (Stage IV). Without antimicrobial therapy, staphylococci cannot be cleared by host immune responses from abscess lesions and infected animals typically succumb to disease within 30–60 days¹⁴.

Figure 2.

Working model for staphylococcal abscess formation and persistence in host tissues. Stage I: following intravenous inoculation, S. aureus survives in the bloodstream and disseminates via the vasculature to peripheral organ tissues. Stage II: staphylococci in renal tissue attract a massive infiltrate of polymorphonuclear leukocytes and other immune cells. Stage III: abscesses mature showing a central accumulation of the pathogen (SAC) surrounded by a pseudocapsule of fibrin deposits (pink rim), and zones of necrotic and healthy polymorphonuclear neutrophils (PMNs; purple and light blue cells, respectively), and finally a rim of eosinophilic material (orange rim). Stage IV: abscesses mature and rupture on the organ surface to initiate new rounds of infections. Genes required for specific stages of staphylococcal abscess development are in red above the corresponding stage of infection. Figure adapted with the authors’ permission from an article published by Cheng and colleagues¹⁴.

Protagonists of the plot

Stage I: survival of staphylococci in blood

Following intravenous inoculation of mice with a sublethal dose of bacteria (~10⁷ colony forming units, CFU), 99.9% of the original inoculum disappears from the vasculature within 6 h (Figure 3a). Dissemination to organ tissue occurs within 3 h followed by rapid decline (3–12 h) and measurable replication after 24 h (Figure 3a)¹⁴. Unless staphylococci survive or replicate in blood and circulating lymph fluids, this pathogen could not disseminate to new sites within an infected host. Thus, the first stage of the plot is to describe the actors in the play that lead to disease crisis and dramatic endings of S. aureus infections. To survive in blood, S. aureus must escape a variety of innate immune mechanisms, including antimicrobial peptides, complement, and phagocytic killing^{17, 18}. The principal defense against S. aureus is provided by neutrophilic polymorphonuclear leukocytes, which comprise 60–70% of human white blood cells¹⁹. Reactive oxygen species, along with the acidic vacuolar environment, prove toxic to many microbes, and help curtail bacterial spread. S. aureus produces the carotenoid staphyloxanthin^{20, 21}, a membrane-bound pigment that scavenges reactive oxygen species and protects S. aureus from neutrophil killing²². Further, staphylococci deploy a group of secreted peptides, termed phenol-soluble modulins (PSMs), to target human neutrophils for destruction by disrupting the integrity of their plasma membrane²³.

Figure 3.

Monitoring and visualizing Staphylococcus aureus abscess formation. (a)–(e) Visualization of lesions in kidneys following infection at Stages II, III and IV as previously described [14]. Paraffin embedded kidneys were thin-sectioned, stained with hematoxylineosin, mounted on glass slides and examined by light microscopy for the formation of staphylococcal abscesses. Micrographs on the left (a, c, e) show sections of entire kidneys. Micrographs on the right (b, d) represent magnifications of the area framed in the white box in (a) and (c). (b) Stage II: Site of inflammation with immune cell infiltrates (green arrowhead). (d) Stage III: staphylococci are clearly distinguishable as central nidus, staphylococcal abscess community (SAC), within the maturing abscess (yellow arrowheads). The SAC is surrounded by an amorphous, eosinophilic capsule (see Figure 4) followed by a zone of dead PMNs (polymorphonuclear neutrophils; white box), apparently healthy PMNs (red box), and necrotic PMNs (green box). The entire abscess region is separated from healthy kidney tissue by a second eosinophilic layer, which appears to increase in size over time. (e) Stage IV: Abscesses migrate to the periphery of organ tissue. The organization of SACs and immune cells during Stage III is abolished prior to rupture of the lesion onto the organ surface. (f) Representative distribution of bacteria in blood (dashed line) and kidney tissue (solid line) over time following intravenous infection of mice with 1 × 10⁷ colony forming units (CFU) of S. aureus. Bacterial loads are represented as CFU per g of blood or renal tissue over time following infection.

The complement cascade plays a central role in innate host defenses to S. aureus^24–26. By binding to microbial surfaces, C3 and C5 convertases deposit complement and mediate phagocytic killing of bacteria. The associated release of C3 and C5 fragments further serves as a chemoattractant for phagocytes. S. aureus subverts these mechanisms by secreting factors inhibitory for complement activation and neutrophil chemotaxis^{27, 28}. Although not part of the core S. aureus genome, these prophage-encoded factors are expressed by many virulent strains²⁹.

In addition to secreted factors, staphylococci rely on cell wall-anchored surface proteins for their survival in blood³⁰. Clumping factor (ClfA), a sortase-anchored product whose N-terminal domain is displayed on the bacterial surface, binds fibrinogen³¹. This biochemical attribute triggers staphylococcal agglutination in the presence of fibrinogen or fibrin, i.e. in lymph fluids or blood³¹. How agglutination ensures staphylococcal survival is not entirely clear. On one hand, the extremely large size of agglutinated staphylococci can be thought to interfere with phagocytosis^{32, 33}. On the other, ClfA enables staphylococci to adhere to fibrin deposits on the vascular endothelium, providing a niche for pathogen replication at the periphery of circulating fluids. The latter mechanism is certainly essential for the pathogenesis of endocarditis³⁴. Adenosine synthase A (AdsA), another sortase-anchored product, harbors a 5’-nucleotidase signature sequence. AdsA cleaves ADP or AMP to form adenosine. When inoculated into blood, S. aureus synthesizes large amounts of adenosine from AMP using AdsA³⁵. Adenosine is a key regulator of innate and acquired immune responses and is generated at sites of inflammation, hypoxia, organ injury, and traumatic shock³⁶. Host inflammatory responses, e.g. in response to bacterial infection, are regulated to prevent excessive tissue damage during infection; production of adenosine is the key signal to end host inflammatory response³⁶. Thus, staphylococci evade immune responses by signaling to the host that its inflammatory activities must be canceled.

As the pathogen has endowed itself with such a multitude of defense mechanisms, it comes as no surprise that S. aureus, although quickly taken up by polymorphonuclear neutrophils (PMNs) or monocytes, survives within these leukocytes^{19, 37}. Gresham and colleagues demonstrated that PMNs harboring S. aureus can indeed seed infectious lesions when transferred into naïve mice, suggesting that staphylococci use PMNs as vehicles to enter the tissues of an infected host³⁸.

Stage II: preparing a site for the infectious lesion

Although staphylococci arrive within 1–3 hours in peripheral organs where abscess lesions are being established (Figure 3a), a histopathological correlate for this activity is not detectable until about 48 hours after intravenous challenge (Figure 3)¹⁴. At that time, sufficiently large numbers of immune cells (predominantly neutrophils) have accumulated in a small area to replace physiological epithelia with microscopically discernable lesions (Figure 3b and 3c)¹⁴. During Stage II of abscess maturation, early lesion sites contain a low abundance of staphylococci. It seems unlikely that the dramatic recruitment of immune cells can be explained by the massive replication of bacteria – micro- or macro-colonies of staphylococci are not observed early during infection. Molecular genetic studies on lipoprotein maturation revealed a fascinating phenotype for the S. aureus lgt mutant (lipoprotein diacyl-glyceride transferase), which is unable to link diacyl-glycerol to the cysteine-thiol of lipoprotein precursors³⁹. The lgt mutant multiplied in host tissues with minimal recruitment of immune cells and did not form typical abscess lesions⁴⁰. Thus, in addition to producing formylated peptides that are perceived by the immune system, S. aureus could actively release lipoproteins for recognition via Toll-like receptor 2 (TLR2)⁴⁰. The ensuing pro-inflammatory signals could then recruit large numbers of immune cells to the site of infection, thereby contributing to the destruction of organ tissues.

Staphylococcal replication in organ tissues must meet the pathogen’s nutritional requirement for iron⁴¹. A key pathway for the scavenging of iron at this stage involves the hemoglobin hemophore IsdB, as well as two other sortase anchored products, IsdA and IsdC, which pass the heme-moiety across the cell wall envelope for subsequent transport across the staphylococcal membrane⁴². Depending on the site of infection, too much import of heme is not a good thing. The associated production of oxygen radicals proves toxic for staphylococci and immune cells. Mutations in S. aureus hrtA, at the center of a pathway that perceives heme and transports superfluous heme out of bacteria, cause heme-toxicity in mutants but also interfere with neutrophil recruitment, resulting in liver-specific hypervirulence⁴³. HrtA and HrtB form an ABC-type transport system that is essential for alleviating hemin toxicity. In the absence of hrtA, hemin toxicity triggers overexpression of the orphan permease HrtB. This triggers membrane stress, thereby augmenting the secretion of a subset of immunomodulatory proteins that are directed to inhibit neutrophil migration to the site of infection⁴⁴. Although this is not yet known in detail, we surmise that staphylococci require several unique biosynthetic pathways during this developmental program, in particular because the pathogen must carry out its metabolism in an environment that restricts the availability of zinc⁴⁵. SdrD, another sortase-anchored surface protein, is also required at Stage II¹⁴. Its specific activity during host infection is not yet known.

Stage III: crisis and disease

The molecular and cellular events that promote the formation of a staphylococcal abscess can be imagined when studying the histopathological features of maturing lesions. Within four or five days following infection, a large mass of the invading pathogen (SAC), replicates at the center of the lesion (Figure 3d and 3e). The SAC is separated from surrounding immune cells by an eosinophilic pseudocapsule that is composed of fibrin deposits¹⁴. This pseudocapsule is best seen by electron microscopy (Figure 4). In the periphery, staphylococcal abscesses are characterized by layers of immune cells: a ring of largely necrotic neutrophils in the immediate vicinity of SACs, followed by a layer of healthy appearing immune cells and eventually a layer of necrotic immune cells, whose border with healthy tissues is demarcated by a layer of fibrin or extracellular matrix deposits (Figure 3d and 3e)¹⁴.

Figure 4.

SACs during Stage III. At the center of an abscess lesion, S. aureus cells are found as tightly associated mass covered with a granular, electron-dense substance. Bacteria are enclosed by an amorphous pseudocapsule (white arrowheads) that separates the SAC from neighboring immune cells. Image was captured from the kidney of an animal infected with S. aureus Newman. Organ tissue was thin sectioned, fixed, dehydrated and sputter-coated with platinum/palladium and analyzed by scanning electron microscopy.

What are the molecular events that enable this astonishing metamorphosis? S. aureus secrete coagulase (Coa), which associates with and non-enzymatically activates prothrombin^{46, 47}. The staphylocoagulase-prothrombin complex cleaves fibrinogen to generate fibrin peptides, promoting clot formation^{48, 49}. One hundred years after the discovery of Coa, a second coagulase was identified⁵⁰. von Willebrand factor-binding protein (vWbp) was initially discovered to interact with its namesake host protein, a key factor in the recruitment of platelets to blood clots⁵¹, however vWbp also activates prothrombin to promote fibrin coagulation⁵². Although mutants lacking either coa or vWbp continue to form abscesses, variants lacking both genes were unable to form these lesions and coagulate blood (Figure 1c)¹⁶. Further, prothrombin, fibrin and coagulases localize to the eosinophilic pseudocapsule of staphylococcal abscesses¹⁶. A simple explanation for these observations is that staphylococci usurp the clotting cascade via secreted coagulases to form fibrin deposits in organ tissues. The deposits function as a barrier for immune cells preventing their penetration into SACs, where the pathogen replicates without interference while manipulating inflammatory responses, the fate of immune cells in the periphery and the progression of the entire infectious lesion from a safe distance. Additional evidence for the key contribution of coagulases to virulence and abscess formation was obtained through the discovery that S. aureus strains adapted to ruminants or horses require additional vWbp homologs to coagulate blood or plasma of these species and establish staphylococcal disease⁵³.

Mutations in two other genes affect the maturation of staphylococcal abscesses. spa encodes staphylococcal protein A (SpA, Figure 1a), a sortase-anchored product that is released from the bacterial envelope in large amounts. SpA impedes phagocytosis by binding the Fcγ component of immunoglobulin^{54, 55}, activates platelet aggregation via its binding to von Willebrand factor⁵⁶, and functions as a B-cell superantigen by crosslinking the Fab region of V_H3 bearing immunoglobulin M (IgM)^57–59. Also, through its activation of tumor necrosis factor receptor 1 (TNFR1), SpA initiates staphylococcal pneumonia⁶⁰. It is unknown which of the many biochemical attributes of SpA contributes to abscess formation, however, spa mutants appear unable to form abscesses¹⁴. Emp is an envelope associated protein of staphylococci, whose virulence attributes have been associated with the ability to promote adherence to endothelia⁶¹ and biofilm formation⁶². Numerous studies have attempted to correlate staphylococcal virulence with biofilm formation; the evidence is equivocal^{63, 64}. SACs are reminiscent of biofilm communities formed in vitro, however several genes that contribute to biofilm formation, e.g. those specifying capsular polysaccharide or poly N-acetylglucosamine synthesis^{65, 66}, are dispensable for abscess formation¹⁴. Immunofluorescence microscopy revealed Emp on the surface of staphylococci at the center of abscess lesions¹⁴, suggesting that this protein contributes to abscess formation by promoting growth as biofilm-like SACs.

Stage IV: persistence and an end to the drama

Staphylococcal abscesses mature over weeks and, following rupture and release into the peritoneal cavity, lead to new infectious lesions (Stage IV). Variants lacking eap (extracellular adhesion protein) or a non-canonical protein secretion pathway designated Ess (ESAT-6 like secretion system) are defective in persistence and are, therefore, assigned to Stage IV^{14, 67}. Ess secretion is reminiscent of the ESX-1 dependent secretion of Mycobacterium tuberculosis⁶⁸. ESAT-6 and CFP10, products of the ESX-1 locus, have been shown to have immune evasive effects, including TLR2 downregulation and macrophage phagosome lysis. Deletion of the ESX-1 locus results in an attenuated strain (BCG) that forms irregular granulomas and is ultimately unable to persist within human hosts^{69, 70}. Therefore, it appears that this secretion system secretes immunomodulatory factors that are required for persistence. S. aureus has only one ESX-1-like locus and mutations in the ess cluster survive stages I-III of infection but fail to persist at stage IV⁶⁷, suggesting that this pathway is also involved in prolonged staphylococcal survival within the abscess⁷¹.

All S. aureus isolates appear to secrete Eap, although there is significant strain variation⁷². Structural studies showed that Eap and the C-terminal half of TSST1 (toxic shock syndrome toxin 1) as well as SEB (staphylococcal enterotoxin B) display homology, and Eap was thus shown to have superantigenic activity and to induce pro-inflammatory (IL6, TNFα) cytokine production by monocytes⁷³. However, Eap has also been shown to exert numerous anti-inflammatory effects by inhibiting neutrophil adherence to the endothelium⁷⁴, leukocyte recruitment in a murine bacterial peritonitis model⁷⁵, T cell activation and recruitment to the dermis⁷⁶.

Distinctive or universal: lessons learned from S. aureus abscess formation

Abscesses, defined as inflammatory lesions releasing purulent material, are standard responses for many different biological, chemical or physical insults to host tissues. Using a mouse model, we describe the concept that S. aureus infection may modify this default program of tissue repair by building complex lesions that ensure the pathogen’s persistence in a hostile environment, its dissemination throughout many different organ systems and its spread to new hosts. In support of our model we note that the injection of heat-killed staphylococci does not trigger abscess formation in mice¹⁴. This is unlike other biological materials, for example purified capsular polysaccharide of Bacteroides fragilis, that alone can activate host immune systems and promote abscess formation in a murine peritoneal challenge model^{77, 78}. Further, some, but certainly not all, genes that contribute to the pathogenesis of different S. aureus diseases are also required for abscess formation¹⁴. Of those, we note that coagulases and SpA are unique to S. aureus isolates. A reciprocal argument is that the absence of protein A and coagulase genes from S. epidermidis correlates with the inability of this microbe to cause abscesses of the kind that S. aureus generates. The obvious conclusion for these findings is that S. aureus abscesses could indeed be distinctive. Whatever could perturb their formation may also benefit the outcome of infections. The benefit of such research could be significant, for example through the development of vaccines that prevent S. aureus skin and soft tissue infections. In accord with this, antibodies raised against either protein A or staphylococcal coagulases can each prevent abscess formation in the aforementioned mouse model^{16, 79}. Future work will need to determine whether these genes are key virulence factors and vaccine antigens in other animal models of S. aureus infection, for example rats, guinea pigs, rabbits or non-human primates. If so, SpA and coagulases could represent important candidates for the development of a human vaccine. Recently, Kim et al. reported immunization of animals with a non-toxigenic variant of SpA (SpA_KKAA) that cannot bind Fcγ and Fab⁷⁹. The resulting immune responses promoted opsonophagocytic clearance and protected animals against abscess formation⁷⁹. Further, animals immunized with Spa_KKAA were able to mount antibody responses against numerous S. aureus antigens upon challenge⁷⁹. Typically, infection with S. aureus generates a modest immune response that does not afford protection against subsequent infections⁷⁹. Thus, protective immunity against S. aureus infection could be achieved via the neutralization of protein A with specific antibodies.