Chemokine-cleaving Streptococcus pyogenes protease SpyCEP is necessary and sufficient for bacterial dissemination within soft tissues and the respiratory tract

Abstract

SpyCEP is a Streptococcus pyogenes protease that cleaves CXCL8/IL-8 and its activity is associated with human invasive disease severity. We investigated the role of SpyCEP in S. pyogenes necrotizing fasciitis and respiratory tract infection in mice using isogenic strains differing only in SpyCEP expression. SpyCEP cleaved human CXCL1, 2, 6 and 8 plus murine CXCL1 and 2 at a structurally conserved site. Mice were infected in thigh muscle with a strain of S. pyogenes that expresses a high level of SpyCEP, or with an isogenic non-SpyCEP expressing strain. SpyCEP expression by S. pyogenes hindered bacterial clearance from muscle, and enhanced bacterial spread, associated with cleavage of murine chemoattractant CXCL1. Mice were then infected with Lactococcus lactis strains that differed only in SpyCEP expression. In contrast to the parent L. lactis strain (lacks SpyCEP), which was avirulent when administered intramuscularly, infection with a strain that expressed SpyCEP heterologously led to dramatic systemic illness within 24 h, failure to clear bacteria from muscle and marked dissemination to other organs. In the upper airways, SpyCEP expression was required for survival of L. lactis but not S. pyogenes. However, dissemination of S. pyogenes to the lung was SpyCEP-dependent and was associated with evidence of chemokine cleavage. Taken together, the studies provide clear evidence that SpyCEP is necessary and sufficient for systemic bacterial dissemination from a soft tissue focus in this model and also underlies dissemination in the respiratory tract.

Fig. 1

SpyCEP cleaves human CXCL1, 2, 6 and 8 and murine CXCL1/2 after the first four amino acids of the C-terminal α-helix. A. Colloidal Coomassie blue stained SDS-PAGE gels showing human chemokines cleaved by SpyCEP. Chemokines were co-incubated either alone (left lane); with supernatant containing SpyCEP, H292 (+SpyCEP, middle lane); or supernatant without SpyCEP, H575 (-SpyCEP, right lane). B. Mass spectroscopy analysis of uncleaved and cleaved chemokines using electrospray ionization generated a series of multiply charged ions (indicated as m/z; mass-to-charge ratio) from which the average molecular mass (m) of each was deduced. C. Site of SpyCEP-mediated cleavage of various chemokines determined in this study and previously for CXCL8/IL-8 and CXCL2/MIP-2 (Edwards et al., 2005): arrowhead. The position of the α-helix in this region is indicated by a grey shaded box.

Fig. 2

SpyCEP influences bacterial clearance in soft tissue and systemic spread of S. pyogenes. A. SpyCEP production by S. pyogenes parent strain H292 and mutant cepA strain H575. Western blots were performed on culture supernatants (left panel) and cell wall preparations (right panel). Wild-type SpyCEP was detected as an approximate 160 kDa band in both the supernatant and the cell wall preparation from strain H292. Strain H575 produced an inactive truncated N-terminal SpyCEP fragment of approximately 40 kDa in the culture supernatant, but not the cell wall due to lack of a cell wall anchor motif. B. Expression of SpyCEP by strain H292 enhanced S. pyogenes survival at the site of infection following intramuscular infection of CD1 female mice compared with strain H575 (P = 0.0019; inoculum 2 × 10⁸ cfu). C. Expression of SpyCEP also enhanced spread of S. pyogenes to the regional lymph node (P = 0.0001). n = 10 mice per group in each experiment, representative of two experiments. ND, not detected; cfu, colony-forming units; horizontal line, median.

Fig. 3

Levels of KC (murine CXCL1) in mouse serum (left axis) and mouse whole thigh homogenate (right axis) 24 h after infection with S. pyogenes. KC was measured by ELISA. Infection with parent strain H292 reduced chemokine levels compared with those mice infected with mutant strain H575 in both serum and thigh muscle. n = 10 mice per group.

Fig. 4

Heterologous expression of SpyCEP in L. lactis is sufficient to resist bacterial clearance and promote dissemination. A. Heterologous expression of SpyCEP by L. lactis strain H487 and parental control L. lactis strain H486. Western blots for SpyCEP were performed on culture supernatants (left panel) and cell wall preparations (right panel). H487 produced SpyCEP in both supernatant and cell wall. Lower molecular weight bands represent autocatalytic cleavage of SpyCEP. B. Bacterial quantification in muscle tissue demonstrating heavy growth in mice infected with SpyCEP-expressing L. lactis strain H487 compared with control strain H486 (P = 0.0001). C. Bacterial dissemination of L. lactis to regional lymph node. Quantification of bacteria in each inguinal lymph node shows greater bacterial spread to node in mice infected with H487 compared with H486 (P = 0.0015). D. Bacterial quantification in liver. SpyCEP was also essential for systemic spread of L. lactis to liver with greater bacterial burden in liver of mice infected with H487 compared with H486 (P = 0.012). E. Bacterial quantification in spleen with greater bacterial burden in spleen of mice infected with H487 compared with H486 (P = 0.0003). n = 10 per group representative of two experiments. ND, not detected; cfu, colony-forming units; horizontal line, median.

Fig. 5

Tissue sections from thigh muscle of L. lactis-infected mice obtained at 3 and 6 h after onset of infection suggest that SpyCEP is sufficient to reproduce some features of invasive necrotizing infection. Gram stained muscle sections from L. lactis-infected mice at 3 and 6 h after infection (left and central panels). Upper panels, mice infected with control L. lactis H486 (empty plasmid); Lower panels, mice infected with L. lactis strain expressing SpyCEP, H487. Right hand panels, haematoxylin and eosin-stained tissues at 6 h. L, lactococci; N, neutrophils; nN, necrotic neutrophil. Representative of three mice at each time point, magnification ×200.

Fig. 6

SpyCEP enhances S. pyogenes dissemination to lung from the upper respiratory tract. S. pyogenes studies are shown in (A), (B) and (C) (10 mice per group). Following intranasal infection, S. pyogenes survival in the nasal tissues (A) and NALT (B) was unaffected by disruption of SpyCEP. However, S. pyogenes dissemination into the lower respiratory tract (C) was enhanced by SpyCEP expression (P = 0.02). Lactococcal studies are shown in (D), (E) and (F) (8 mice per group). SpyCEP expression by L. lactis was sufficient to allow survival in the nasal tissues (P = 0.001) (D) and also enhanced survival in NALT (P = 0.046) (E). SpyCEP did not significantly affect survival of L. lactis in the lower respiratory tract (F). ND, not detected; cfu, colony-forming units; horizontal line, median. Data from one experiment for each strain.

Fig. 7

Levels of KC (murine CXCL1) in mouse lung homogenate (left side axis) and mouse serum (right side axis) 24 h after infection with S. pyogenes. Infection with parent strain H292 reduced chemokine levels compared with those mice infected with mutant strain H575 in both lung and serum. n = 10 mice per group.