Immune Evasion Mechanisms of Staphylococcus epidermidis Biofilm Infection

1Initially described in 1878, staphylococci are Gram-positive microorganisms that have been implicated in infections involving multiple systems of the human body, including the skin and soft tissue, the skeletal system, the respiratory system, the blood stream, and more recently, infections involving implanted medical devices. Staphylococci are further classified as being coagulase-positive, primarily identifying Staphylococcus aureus (S. aureus), or coagulase-negative (CoNS). Staphylococcus epidermidis (S. epidermidis) is the most important and best-studied member of the CoNS group.

2While conventionally regarded merely as an innocuous commensal of the human skin, and increasingly recognized for its beneficial role in skin immunity and within the skin microbiota, S. epidermidis has also emerged as an important human pathogen. This is mostly due to the increased number of implanted prosthetic materials and medical devices. Lacking most of the many aggressive virulence factors that S. aureus produces, S. epidermidis’ primary virulence mechanism is to form deeply seated microbial communities, known as biofilms, on implanted medical devices surfaces and native host tissues. For example, S. epidermidis is a leading cause of infections on central venous catheters, which occur at a frequency of ∼80,000 annually in the United States alone and may result in severe blood infections. Furthermore, together with S. aureus, it is the premier pathogen causing prosthetic joint infections. Finally, S. epidermidis causes 15–40% of prosthetic valve endocarditis cases, a less common but very serious infection. All these infections are easily recognized macro- or microscopically to proceed as biofilm-associated. Within biofilms, S. epidermidis are protected from the effects of antimicrobial therapy as well as the host immune system. Consequently, medical therapy in biofilm-associated infections can be exceptionally challenging, with attempts at infection eradiation often entailing complete removal of the infected foreign body as well as administration of prolonged courses of antimicrobial therapy, approaches that incur risks to patients and excess cost to the health care system.

3Herein, we attempt to outline our current understanding of the host- and pathogen-derived characteristics that favor effective immune evasion by S. epidermidis, such as those contributing toward the microorganism’s successful seeding, colonization of device surfaces, and securing survival in the form of indolent biofilm communities.

4Biofilm development has been modeled to occur in three stages: (1) attachment, (2) proliferation/formation of the matured biofilm, and (3) detachment/dispersal. During attachment, staphylococcal surface-attached proteins known as microbial surface components recognizing adhesive matrix molecules (MSCRAMMs) establish non-covalent interactions with device surfaces coated by host proteins and host tissues. After attachment, proliferation and maturation of the biofilm follows, with the production of an extracellular matrix consisting of the staphylococcal biofilm exopolysaccharide, polysaccharide intercellular adhesin (PIA), also called poly-N-acetylglucosamine (PNAG), teichoic acids, proteins, and extracellular DNA (eDNA). During this second stage of biofilm expansion, channels and mushroom-shaped structures form to facilitate nutrient delivery to deeper layers of the biofilm. The last stage of biofilm development is characterized by the detachment and subsequent dispersal/dissemination of biofilm clusters to distal sites, a process mostly due to the activity of the surfactant-like phenol-soluble modulin (PSM) peptides (;; Figure 1).

5The exopolysaccharide PIA/PNAG is an important and abundant component of the S. epidermidis biofilm matrix. Initially described in S. epidermidis clinical isolates, PIA is a linear homopolymer of β-1,6-linked N-acetylglucosamine monomers, which is positively charged due to partial deacetylation. These charged moieties facilitate interactions between components of the biofilm extracellular matrix and the staphylococcal cell wall. PIA is the product of the icaADBC operon, which encodes three membrane proteins (IcaA, IcaC, IcaD), and an additional protein (IcaB) that is exported and subsequently attached to the staphylococcal cell surface by non-covalent interactions. IcaA, an N-acetylglucosaminyltransferase, along with IcaD, catalyzes the conversion of UDP-N-acetylglucosamine to 10–20-mers of β-1,6-linked poly-N-acetylglucosamine. IcaC is needed to further extend the N-acetylglucosamine oligomers, and probably represents a PIA export system, while IcaB deacetylates the poly-N-acetylglucosamine molecule, enabling localization of the PIA molecule to the cell surface (Figure 2).

6PIA does not appear to be absolutely required for S. epidermidis biofilm formation, as S. epidermidis isolates from biofilm-associated catheter and prosthetic joint infections were found to be negative for the ica genes. In S. epidermidis ica strains, intercellular adhesion has been shown to be mediated through proteinaceous components, of which two of the best-defined are the accumulation-associated protein (Aap) and the extracellular matrix binding protein (Embp). However, PIA-dependent biofilms appear to be more structured and robust than those dependent on proteins.

7Accumulation-associated protein is a 240-kDa surface-bound protein consisting of two domains, designated as A and B. Proteolytic cleavage of domain A induces a conformation change in the protein, enabling domain B to mediate polymerization of Aap into fibrils, leading to aggregation and biofilm formation in the PIA-negative S. epidermidis clinical strain 5179.

8Extracellular matrix binding protein is a 1-MDa protein surface protein that has been implicated as an intercellular adhesin that mediates S. epidermidis biofilm formation. Embp is composed of 59 found-in-various-architecture (FIVAR) domains and 38 G-related albumin-binding (GA) domains. In vitro, the FIVAR domains were observed to bind to fibronectin, thus suggesting a role in the initial attachment phase of biofilm development. Furthermore, like the FIVAR domains described in the S. aureus protein Fmtb, which interacts with N-acetylglucosamine, the FIVAR domains of Embp in S. epidermidis are hypothesized to also bind to PIA. However, it appears that FIVAR domains alone are insufficient for biofilm aggregation.

9Phenol soluble modulins (PSMs) were first described in 1999 in S. epidermidis. Later it was found that S. epidermidis produces six PSM peptides, PSMα, PSMβ1, PSMβ2, PSMδ, PSM𝜀, and δ-toxin. Next to δ-toxin, β-PSMs are the primary PSMs produced in S. epidermidis. Besides these PSMs, which are all core genome-encoded, there is one PSM, PSM-mec, which is encoded within some types of the methicillin resistance-conferring mobile genetic element, SCCmec. S. epidermidis β-PSMs have been shown to be key effector molecules in biofilm structuring and dissemination, but investigation in S. aureus indicates all PSMs have similar capacities. The general mechanism by which PSMs contribute to biofilm structuring and dispersal is believed to be the disruption of non-covalent (hydrophobic, electrostatic) interactions between biofilm matrix macromolecules. The PSM structuring effect occurs independently of the mode of biofilm formation (PIA-dependent or -independent).

10In immunocompetent hosts, the primary innate immune response against planktonic staphylococcal infections involve the complement system as well as effector cells. The complement system’s primary role is to recruit effector molecules that label staphylococci and mark them for destruction by effector cells. In addition, the activated complement system produces a cell-killing membrane attack complex. Complement fixation occurs through the classical, alternative, and lectin pathways. While the alternative and lectin pathways are part of the innate immune system, activation of the classical pathway requires binding of an antibody to an antigen on the staphylococcal surface. Activation of these pathways leads to the generation of C3a, a pro-inflammatory chemoattractant, which recruits phagocytes to the infection site. While the alternative pathway seems to play a smaller role in host defense against S. epidermidis infections, the classical and lectin pathways are believed to be necessary for rapid killing of planktonic S. epidermidis by effector cells.

11Neutrophils (or, polymorphonuclear leukocytes, PMNs) have been identified as the primary effector cells in the innate immune defense against staphylococcal infection. Activation of neutrophils might involve: (1) recognition of pathogen-associated molecular patterns (PAMPs) on bacterial surfaces by the cellular recognition receptors (Toll-like receptors), (2) deposition of opsonins on the bacterial surface, or (3) the direct action of complement system. Leukocyte responses involve direct contact of the staphylococci with the host immune cells, and are channeled via pattern recognition receptor-dependent pathways. While the role of these pathways in staphylococcal infection have primarily been established in S. aureus, there is evidence that the cellular recognition receptor, Toll-like receptor 2 (TLR2), plays a significant role in S. epidermidis bloodstream infection. However, how exactly pathogen recognition receptors impact S. epidermidis blood infection remains unknown.

12Once activated, neutrophil-mediated killing involves reactive oxygen species as well as non-oxygen-dependent processes involving antimicrobial peptides (AMPs), such as defensins and cathelicidins, and antimicrobial proteins, such as lysozyme. Again, much of this is general knowledge obtained in other bacteria, but a significant role of AMP resistance mechanisms, such as proteolysis by the SepA protease and PIA-mediated resistance on the cell surface (see below), as well as sensing of the presence of AMPs by the S. epidermidis ApsRSX system, have been demonstrated to contribute to S. epidermidis survival in neutrophils and biofilm infection.

13As compared to S. aureus, infections by S. epidermidis are characterized by proceeding with decidedly less inflammation and overall morbidity. This is also true for biofilm-associated infections. Specific mechanisms that S. epidermidis employs to limit the inflammatory response will be discussed in the following. On the other hand, there are specific interactions between S. epidermidis and the host that can be described as pro-inflammatory, which will also be discussed. These may be less important given the overall outcome.

14Early research has suggested that biofilm formation protects S. epidermidis from phagocytosis by effector cells. When compared to the response against planktonic infection, the innate immune response against S. epidermidis biofilm infection has been characterized as being less pronounced. However, vigorous induction of the complement system has been described. Yet, deposition of C3b and IgG on S. epidermidis biofilm surfaces was paradoxically diminished, when compared to the planktonic mode of growth, suggesting that subsequent activation and killing of S. epidermidis biofilm by effector cells might also be impaired.

15Diminished activation of leukocytes has been described in S. epidermidis biofilm infection and seems to involve multiple mechanisms. First, observation of a blunted response to pro-inflammatory compounds by macrophages after exposure to S. epidermidis biofilm suggests that interference of signaling by cellular recognition receptors might partially contribute to the observed quiescent immune response toward biofilms, when compared to the planktonic mode of growth. Additionally, when rabbit polyclonal PIA/PNAG antiserum was utilized as opsonin, fewer deaths of S. epidermidis derived from disrupted biofilms were detected, when compared to their isogenic planktonic form, implicating opsonin deposition as one aspect of innate immune response that may be modulated in S. epidermidis biofilm infection.

16Moreover, it has been hypothesized that sufficient contact of the bacteria with components of the host immune system is necessary for efficient activation of effector cells. In fact, deficient uptake of bacteria by macrophages and reduced generation of an NF-κB-mediated macrophage inflammatory response have been described in S. epidermidis biofilm, phenotypes probably linked to the immune evasion effect of PIA (see below) and biofilm formation per se. Additionally, selective modulation of the immune response seems to occur, as discriminatory activation of the weakly pro-inflammatory J774A.1 macrophage by S. epidermidis biofilms was reported. Once consumed by macrophages, biofilm-derived S. epidermidis appears to be able to survive more effectively within these effector cells, than their isogenic planktonic counterpart.

17The exopolysaccharide PIA represents a particularly important constituent of the microorganism’s immune evasion strategies. When compared to the isogenic 1457 ica M10 strain, S. epidermidis 1457 is killed by PMNs less efficiently, likely by reducing phagocytosis. Reduced uptake has been shown directly for macrophages. This is likely mostly due to PIA being a preeminent biofilm constituent and a molecule that leads to the formation of bacterial aggregates. However, there are also more specific interactions of PIA with the immune system that have been reported. Some of them are probably due to PIA forming a sort of positively charged “capsule” around S. epidermidis, which is a general mechanism to shield the bacteria from immune recognition. In accordance with that notion, differences in PMN-mediated killing have been attributed to decreased availability of opsonizable surfaces in ica(+) biofilm and differences in PIA-mediated opsonization mechanisms. Furthermore, PIA appears to prevent neutrophil attacks when cell clusters are disintegrated.

18Animal experiments as well as observations from human S. epidermidis biofilm infections have largely recapitulated an overall attenuated immune response to biofilm infection due to wildtype S. epidermidis strains when compared to those derived from the isogenic ica isolates. Levels of inflammatory cytokines (including TNF-α, IL-6, IFN-γ) and CD11b expression in human whole blood were higher in infections with S. epidermidis 1457 ica M10 than those with the isogenic wildtype strain. Moreover, skin and soft tissue inflammatory changes surrounding subcutaneous catheter insertion sites were found to be more severe in mice infected with wildtype S. epidermidis compared to the ica strain, which correlates with in vitro findings of differential activation of the complement pathways. Furthermore, levels of IL-6 in blood cultures of ica(+) S. epidermidis isolated from neonates were lower as compared to that of an ica control strain. Finally, there was a significant correlation between low levels of C-reactive protein (CRP) and in vitro biofilm-forming capacity in S. epidermidis isolates from neonate blood infection.

19Moreover, PIA confers protection against the action of host AMPs. The mechanism behind the protective effect of PIA against AMPs is thought to involve repulsion between the cationic PIA and the commonly cationic AMPs. However, a similar effect was reported in the same study for the anionic AMP dermcidin, suggesting that different mechanisms of PIA-mediated resistance to AMPs also exist.

20Polysaccharide intercellular adhesion has also been reported to act in a pro-inflammatory fashion. For example, it has been implicated in the activation of the complement system during infection, based on the comparison of wild-type 1457 and isogenic ica M10 strains. This phenotype was also achieved using a PIA preparation in the same study. These results await verification by a demonstration of a specific receptor-mediated mechanism.

21In contrast to PIA, which forms a meshwork of extracellular matrix that embeds adjacent S. epidermidis cells, Aap is covalently attached to the S. epidermidis surface through its LPXTG motif and extends radially to form tuffs of fibrils. Embp is also found on the S. epidermidis surface, forming a proteinaceous matrix. Like in a predominantly PIA-dependent biofilm, biofilms dependent on Aap or Embp protect S. epidermidis from J774A.1 macrophage phagocytosis.

22Complete genome analyses of S. epidermidis have shown that besides from their crucial roles in S. epidermidis biofilm structuring and dissemination, PSMs are the only gene products with cytolytic properties in S. epidermidis, with the PSMδ of the α-type PSM subclass discovered as the first highly potent cytolysin produced by S. epidermidis. PSMδ has been shown to be highly cytolytic against human neutrophils. Nevertheless, when culture filtrates of S. epidermidis were examined, low levels of neutrophil lysis were measured, and this is thought to be a consequence of selective production of PSMs by S. epidermidis, favoring production of the non-cytolytic β-type PSMs over the α-type PSMs, with the exception of the δ-toxin (PSMγ). This strategy results in a low inflammatory profile and likely contributes toward S. epidermidis’ successful immune evasion and subsequent colonization of device surfaces.

23The most frequent and serious complication of biofilm-associated infection on indwelling medical devices is bloodstream infection that arises from hematogenous seeding of dispersed biofilm clusters, which can develop into full-blown sepsis. As the most frequent cause of biofilm infections on such devices, S. epidermidis is also a leading cause of hospital-associated bloodstream infections, in particular in neonates. Believed for the longest time to be due to an over-reacting immune response to invariant bacterial surface structures such as lipoteichoic acids and lipopeptides, a recent study has implicated PSM-mec, which is highly expressed in many methicillin-resistant isolates of S. epidermidis. Like other PSMs, PSM-mec is pro-inflammatory due to activation of the formyl peptide receptor 2. Most likely, other PSMs have similar effects, which remain to be determined.

24Finally, it needs to be mentioned that there are reports on the presence of other toxins in S. epidermidis, such as a pathogenicity-island encoded enterotoxin, but this has to be considered a very rare exception. Reports claiming much more widespread presence of toxins in S. epidermidis have to be regarded as due to inappropriate species identification.

25In summary, by modulating multiple aspects of the host innate immune response using secreted exopolysaccharides, peptides, and other proteinaceous biofilm components, S. epidermidis is able to orchestrate an overall low inflammatory profile, escape killing by the host innate immune system, and persist on the surfaces of implanted devices (Figure 1). While often considered innocuous partly due to its low inflammatory infection profile, it is precisely the low immunogenicity that makes S. epidermidis difficult to diagnose at initial stages of infection. Often, S. epidermidis biofilm infections are diagnosed at the third stage of biofilm infection, when infection sequelae are more severe and difficult to manage and treat, suggesting a need for a more complete basic understanding of the mechanisms behind this relatively quiescent immune response, so that safer avenues of therapy as well as novel approaches to prevention might be pursued.

26Presently, our knowledge on the pro-inflammatory capacity of S. epidermidis in comparison to S. aureus during infection is poor. Future research should include an in-depth analysis of how only recently discovered pro-inflammatory molecules of S. epidermidis, such as the PSMs, contribute to inflammation and immune priming. This research will also benefit from a comparison with recent advances in the understanding of how S. epidermidis primes the skin immune system during the commensal state.

27All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication.

28The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.