New insight into the virulence and inflammatory response of Staphylococcus aureus strains isolated from diabetic foot ulcers

Abstract

Staphylococcus aureus strains isolated from diabetic foot ulcers (DFUs) have less virulence, but still cause severe infections. Furthermore, hypovirulent S. aureus strains appear to be localized in the deep tissues of diabetic foot osteomyelitis, indicating that the unique environment within DFUs affects the pathogenicity of S. aureus. In this study, the cell-free culture medium (CFCM) of S. aureus strains isolated from DFUs exhibited higher cytotoxicity to human erythrocytes than those isolated from non-diabetic patients with sepsis or wounds. Among these S. aureus strains isolated from DFUs, β-toxin negative strains have less virulence than β-toxin positive strains, but induced a higher expression of inflammatory cytokines. Our study and previous studies have shown that the synergistic effect of phenol-soluble modulin α and β-toxin contributes to the higher hemolytic activity of β-toxin positive strains. However, lysis of human erythrocytes by the CFCM of β-toxin negative strains was greatly inhibited by an autolysin inhibitor, sodium polyanethole sulfonate (SPS). A high level of glucose greatly reduced the hemolytic activity of S. aureus, but promoted the expression of interleukin-6 (IL-6) in human neutrophils. However, 5 mM glucose or glucose-6-phosphate (G6P) increased the hemolytic activity of SA118 (a β-toxin negative strain) isolated from DFUs. Additionally, patients with DFUs with growth of S. aureus had lower level of serum IL-6 than those with other bacteria, and the CFCM of S. aureus strains significantly reduced lipopolysaccharide-induced IL-6 expression in human neutrophils. Therefore, the virulence and inflammatory response of S. aureus strains isolated from DFUs are determined by the levels of glucose and its metabolites, which may explain why it is the predominant bacteria isolated from DFUs.

Keywords: Staphylococcus aureus; diabetic foot ulcers; inflammatory response; virulence; β-toxin.

Figure 1

S. aureus strain isolated from DFUs had higher hemolytic activity to human erythrocytes. Lysis of human erythrocytes was measured by quantitative hemolysis assays with the CFCM of S. aureus strains isolated from DFUs (n=32), non-diabetic sepsis (n=22) and wounds (n=32). ^** p<0.01, ^*** p<0.001.

Figure 2

β-toxin positive strains isolated from DFUs had higher virulence in vitro. (A) Hemolytic phenotype of representative S. aureus strains isolated from DFUs (SA19 and SA118 strains) was cultured in 5% sheep blood agar plates. (B) Synergistic hemolysis between a strain of S. agalactiae ATCC13813 (vertical streak of growth) and five representative S. aureus strains (three clinical strains, a positive control ATCC25923 and a negative control ATCC29213). The cytotoxicity of β-toxin positive and negative strains isolated from DFUs to human erythrocytes (C) or neutrophils (D) was analyzed. ^** p<0.01, ^*** p<0.001.

Figure 3

β-toxin positive strains isolated from DFUs had higher virulence in vivo. Mice were injected subcutaneously with ~10⁷ CFUs per mouse of β-toxin positive and representative β-toxin negative strains isolated from DFUs (n=8), and control mice received only sterile PBS. (A) Representative abscess results were shown on day 2 after infection. (B) Representative histological results (H&E stain) were shown on day 2 after infection. (C) The abscess areas were measured on day 2 after infection. (D) The bacterial load in the infected skin was measured 2 days after infection. (E) The expression of IL-1β in the infected tissues was measured by real-time RT-PCR. ^*** p<0.001.

Figure 4

The hemolytic activity of β-toxin positive and negative strains was affected by different cytotoxic factors. The hemolytic activity of SA19 and SA118 strains was analyzed by increasing concentration of FBS (A) or HDL (B). (C) The inhibition of hemolytic activity by FBS or HDL was analyzed in β-toxin positive strains isolated from DFUs. (D) Lysis of human erythrocytes by CFCM of RJ-2, _△PSMα or _△agr cultures was analyzed at increasing dilutions. (E) The turbidity of skim milk mixed with SA19 and SA118 supernatants was analyzed by pre-incubated with EDTA or SPS. (F) The hemolytic activity of β-toxin negative CFCM was analyzed by pre-incubated with EDTA (n=24). The hemolytic activity of SA19 and SA118 supernatants was analyzed by pre-incubated with increasing concentration of PMSF (G) or SPS (H). (I) The inhibition ratio of hemolytic activity by SPS was analyzed in β-toxin positive and negative strains isolated from DFUs. ^*** p<0.001.

Figure 5

The effect of glucose on the hemolytic activity of β-toxin negative strains isolated from DFUs. Lysis of human erythrocytes by the CFCM of SA19 (A) and SA118 (B) strains when grown in LB medium with increasing concentration of glucose. Lysis of human erythrocytes was analyzed by the CFCM of β-toxin negative strains (C) or β-toxin positive strains (D) from DFUs when grown in the LB medium with 5 mM glucose. (E) Lysis of human erythrocytes was analyzed by the CFCM of S. aureus strains isolated from non-diabetic wounds grown in LB medium with 5 mM glucose (n=16). (F) Lysis of human erythrocytes was analyzed by the CFCM of SA118 strain grown in LB medium with increasing concentration of G6P. ^* p<0.05, ^** p<0.01, ^*** p<0.001.

Figure 6

The inflammatory response of S. aureus strains isolated from DFUs in vitro. (A) The hemolytic phenotype of S. aureus strains isolated from bone marrow of a patient with diabetic foot osteomyelitis in 5% sheep blood agar plates. (B) Lysis of human erythrocytes was analyzed by the CFCM of hemolytic and non-hemolytic colonies. (C) The secretion of IL-6 in neutrophils was analyzed by stimulated with the CFCM of hemolytic and non-hemolytic colonies. (D) The secretion of IL-6 was analyzed in neutrophils stimulated with the CFCM of SA118 strain when grown in LB medium with increasing concentration of glucose. Human neutrophils were stimulated with CFCM of β-toxin positive or negative strains in presence of FBS for 15 h, and the levels of IL-6 (E), IL-1β (F) and LDH activity (G) in the culture supernatants were measured. (H) The number of CFUs was detected at 0 and 6 h to calculate the rates of survival for β-toxin positive and negative isolates exposed to whole human blood (n=8), and the level of IL-6 in the plasma was measured (I). ^** p<0.01, ^*** p<0.001.

Figure 7

The inflammatory response was impaired by S. aureus strains isolated from DFUs. (A) The level of serum IL-6 was retrospectively analyzed in patients with DFUs without bacterial growth (n=13) as well as in those with bacterial growth of S. aureus (n=11), other Gram-positive bacteria (n=10, included 3 Enterococcus faecalis, 3 Streptococcus dysgalactiae, 2 S. agalactiae and 2 Streptococcus anginosus), Gram-negative bacteria (n=21, included 8 Proteus mirabilis, 4 Enterobacter aerogenes, 4 Morganella morganii, 3 Escherichia coli and 2 Pseudomonas aeruginosa), or both Gram-negative bacteria (included 3 E coli, 2 E aerogenes and 1 P. aeruginosa) and S. aureus (n=7). (B) The release of IL-6 was analyzed in neutrophils induced by LPS with the CFCM of SA19, SA118 or S. epidermidis RP62A. ^** p<0.01, ^*** p<0.001.