Toll-like receptor 2 downregulation and cytokine dysregulation predict mortality in patients with Staphylococcus aureus bacteremia

Staphylococcus aureus is an opportunistic pathogen causing various pathological conditions, ranging from asymptomatic colonization to virulent invasive infections including bacteremia [1]. The clinical presentation of S. aureus bacteremia (SAB) is heterogeneous; while many patients experience only uncomplicated bacteremia, others exhibit metastatic infections in various organ systems that warrant prolonged antibiotic therapy and often lead to a poor prognosis due to serious sequelae [2]. Many risk factors affecting prognosis in SAB have been identified; host factors include old age [3, 4], presence of comorbidities [5], source of infection and severity of infection at time of presentation [6], and underlying immune status [7, 8], while pathogen-specific factors include methicillin- resistance [9], exotoxins produced by S. aureus, and specific S. aureus strains [10, 11]. Other than virulence factors and antibiotic susceptibility of the infecting S. aureus, underlying host conditions, and host immune responses contribute to case-by-case variation in host-pathogen relationships, thus resulting in clinical heterogeneity [12‐14]. Nevertheless, complex mechanisms underlying the host responses remain unclear.

Upon bloodstream invasion, S. aureus is initially recognized and cleared from circulation by the innate immune system, which is initiated through a sequence of events resulting in the production and secretion of various inflammatory cytokines and chemokines, phagocyte activation, and the initiation of adaptive immunity [14]. Toll-like receptors (TLRs), which are pattern recognition receptors (PRRs) on phagocytic cell membranes, play a pivotal role in recognizing pathogen-associated molecular patterns on bacterial cell walls and activating intracellular signaling pathways [15]. In particular, TLR2 has been reported to be involved in the recognition of staphylococcal peptidoglycan, lipoteichoic acid, and lipoproteins [14, 16, 17]. Upon activation, TLR2 triggers inflammatory signaling pathways, resulting in increased expression and secretion of various antimicrobial peptides, cytokines, and chemokines, which recruit immune cells to the site of infection and trigger the adaptive immune response [17‐19]. Notably, S. aureus produces staphylococcal superantigen-like protein 3 (SSL3) and TIR-containing protein (TirS) during infections, which interfere with TLR2 signaling and dampen appropriate TLR2 activation [20‐23].

Although TLR2-mediated immune responses during staphylococcal infections have been extensively studied, most evidence has been obtained from in vitro studies or in vivo murine models, with varying outcomes [17, 21]. While immune responses primarily triggered through TLR2 signaling are pro-inflammatory, immunoregulatory responses mediated by TLR2 have also been reported [16, 24]. Some human studies on TLR2 expression in sepsis patients have reported that TLR2 is upregulated in sepsis patients and down-regulated in severe sepsis and septic shock patients, leading to death [25‐27]. However, studies on SAB in humans are scarce [27, 28].

Therefore, the aims of this study were to assess TLR2 expression and associated host cytokine responses during the course of SAB, and to examine how they differ on the basis of clinical outcomes of SAB at different stages of the infection.

Host immune responses during infection are one of the factors contributing to the clinical outcomes of SAB, and previous studies have reported that the marked heterogeneity in the clinical spectrum of SAB may result from differential inflammatory responses to S. aureus [12, 13, 34, 38]. TLR2 is reportedly the key sensor of S. aureus infections and the primary trigger for innate immune responses [17]. However, available literature contains diverse descriptions of TLR2 function in the immune response to S. aureus infections [16, 17, 24], and evidence from human studies are limited. TLR2 expression and associated cytokine levels were analyzed during the course of SAB to investigate their associations with SAB severity and clinical outcomes. The present data show that TLR2 is expressed during early stage SAB, and TLR2 downregulation may be associated with mortality in SAB. Furthermore, the levels of pro- and anti-inflammatory cytokines were significantly higher in SAB non-survivors than in survivors, suggesting the dysregulation of inflammatory responses.

Previous murine studies have reported that TLR2-deficient mice are more susceptible to S. aureus infections with a higher bacterial burden and resulted in a higher mortality rate [15]. TLR2 deficiency is suggested to impair phagocytosis, attenuate pro-inflammatory cytokine production, and lead to high mortality [15, 39]. Similarly, in a study on sepsis patients, TLR2 was upregulated in sepsis; however, mortality was associated with TLR2 downregulation [25]. Our results, although not statistically significant, are consistent with these findings, suggesting that TLR2-activity triggered upon bacteremia onset is important for early bacterial clearance and more favorable clinical outcomes of SAB. The lack of statistical significance could be explained by the complex interaction between TLR2 and S. aureus in vivo. First, TLR2-activity seems to be influenced by multiple bacterial factors including proliferative activity, capsule formation, protein synthesis, and cell-wall active factors. Hilmi et al. reported marked variability in TLR activity toward S. aureus isolates, displaying low to absent TLR2-activity in 64% (68/106) of the isolates tested [40]. Furthermore, Hanzelmann et al. reported strain-specific TLR2-activity among S. aureus isolates, showing that strong TLR2 stimulation depends on an active AGR system and is associated with high level phenol-soluble modulin (PSM) production [41]. Second, the route and site of infection seem to influence TLR2-mediated immune responses, as reported in murine studies using TLR2-deficient mice [21]. Although protective roles of TLR2 signaling have been suggested in models of intravenous infection, localized responses to infections have displayed variable outcomes [15, 21]. Third, antimicrobial therapy is also likely to affect the expression of PRRs, including TLR2, by various mechanisms. Testro et al. reported that TLR4 expression was upregulated by antibiotic prophylaxis targeting Gram-negative bacteria in patients with liver cirrhosis [42], and Moore et al. showed that beta-lactams enhanced TLR2 activation by inducing structural changes on the surface of pneumococci, whereas vancomycin did not [43]. Using an in vitro sepsis model, Bode et al. examined the immunomodulatory effects of quinolones, tetracyclines, and macrolides and showed that TLR expression on monocytes and peripheral blood mononuclear cells was differently modulated according to antibiotic treatment [44]. Fourth, host responses to S. aureus infections are complex and other PRRs, including NOD-like receptors and C-type lectin receptors, are also involved in bacterial recognition and the induction of inflammatory reactions in response to S. aureus infections [22]. The complexity of real-life circumstances dampens the significance of TLR2-activity under controlled in vitro conditions.

Both pro- and anti-inflammatory responses occur during S. aureus-induced TLR2 activation [15, 16], along with dysregulated cytokine responses, characterized by elevated IL-10 levels and IL-10/TNF ratios, which are considered responsible for the variations in the clinical outcomes of SAB [12, 13, 38]. Previous studies involving sepsis patients have reported that elevated IL-6 levels are associated with sepsis severity and death [45], and McNicholas et al. reported that IL-6 levels are significantly increased in patients with complicated SAB [12]. Elevated IL-10 levels and a high IL-10/TNF-α ratio have been associated with mortality in severe sepsis patients [46], and similar findings have been previously reported among SAB patients [38]. Consistent with the literature, in this study, patients who died from SAB had significantly higher IL-6 and IL-10 levels during early-stage bacteremia, and patients who died within 7 days of SAB onset further displayed a significantly elevated IL-10/TNF-α ratio and TLR2 downregulation. These results suggest that dysregulated immune responses contribute to SAB-related mortality.

This is the first study to investigate the association among TLR2 expression, cytokine responses, and clinical outcomes in SAB patients. We measured TLR2 expression and cytokine levels throughout the course of SAB (range D2–48) to examine their association with SAB severity and clinical outcomes. Although a correlation between cytokine levels and TLR2 expression could not be determined, our results indicate TLR2 downregulation and significant immune dysregulation in SAB patients with early mortality during early-stage SAB. Furthermore, TNF-α, IL-6, and IL-10 levels were elevated in persistent SAB patients on D10–13.

This study has some limitations, particularly as this study was performed under actual clinical conditions. First, this study was part of a prospective cohort study, not an interventional study; hence, clinical samples could not be obtained at the same time points post-bacteremia, i.e., the timing and total numbers of post-bacteremia blood samples from each patient was different. Therefore, for statistical analysis, we arbitrarily assigned sampling time points as ≤D5, D6–9, D10–13, D14–19 and ≥ D20, thus potentially resulting in bias. However, as in most patients, since the first sample was collected at D3–5, we believe that the measurements during early-stage SAB are largely comparable. Second, since we only included patients who consented to participate, clinically unstable patients were less likely to be enrolled, and the number of patients with worse outcomes was relatively small. Although SAB mortality was lower than that reported in previous studies [47], the clinical characteristics of the enrolled patients and the excluded patients did not significantly differ. Third, we did not analyze bacterial factors including the AGR status, toxigenicity, or inoculum doses, which could have influenced the immune responses and clinical outcomes.

This study indicated that TLR2 downregulation, and IL-6 and IL-10 elevation during early SAB may be associated with mortality from SAB. Furthermore, dysregulated inflammatory responses characterized by elevated IL-10/TNF-α ratios were associated with mortality within 7 d of SAB onset. Therefore, TLR2 expression levels and associated cytokine reactions during the early-stage may be considered potential prognostic factors in SAB.