Blocking Neuronal Signaling to Immune Cells Treats Streptococcal Invasive Infection

Abstract

The nervous system, the immune system, and microbial pathogens interact closely at barrier tissues. Here, we find that a bacterial pathogen, Streptococcus pyogenes, hijacks pain and neuronal regulation of the immune response to promote bacterial survival. Necrotizing fasciitis is a life-threatening soft tissue infection in which "pain is out of proportion" to early physical manifestations. We find that S. pyogenes, the leading cause of necrotizing fasciitis, secretes streptolysin S (SLS) to directly activate nociceptor neurons and produce pain during infection. Nociceptors, in turn, release the neuropeptide calcitonin gene-related peptide (CGRP) into infected tissues, which inhibits the recruitment of neutrophils and opsonophagocytic killing of S. pyogenes. Botulinum neurotoxin A and CGRP antagonism block neuron-mediated suppression of host defense, thereby preventing and treating S. pyogenes necrotizing infection. We conclude that targeting the peripheral nervous system and blocking neuro-immune communication is a promising strategy to treat highly invasive bacterial infections. VIDEO ABSTRACT.

Keywords: CGRP; botulinum neurotoxin; infection; neuroimmune; neuroimmunology; neutrophil; nociceptor; pain; streptococcus pyogenes; streptolysin.

Figure 1. S. pyogenes induces pain associated behaviors and directly activates sensory neurons

(A) Representative images of mouse hind paws at different time points after subcutaneous injection of S. pyogenes M1 or M3 strains (5×10⁷ cfu). (B) Histopathology of skin and soft tissue biopsies 72 h after injection of vehicle, M1 or M3 (5×10⁷ cfu). Scale bars, 100 µm. (C) Spontaneous pain reflexes (lifting/licking of hind paw) over 1 h after injection of different inoculums of M1 (n=4/group) or M3 (n=9–11/group). (D) Mechanical sensitivity after injection of vehicle, S. pyogenes M1 (n=8–9/group) or M3 (n=7–10/group). (E) Heat sensitivity after injection of vehicle, S. pyogenes M1 (n=8–9/group) or M3 (n=7–9/group). (F) Representative Fura-2 ratiometric fields (Left) and calcium traces (Center) of DRG neurons at baseline and after stimulation in vitro with live S. pyogenes M3 (5×10⁹ cfu/mL), capsaicin (1 µm), and KCl (40 mM). Scale bars, 50 µm. Proportions (Right) of capsaicin non-responsive (Cap−) and capsaicin responsive (Cap+) neurons that responded to M3 (n=3–4 fields/condition). Statistical analysis: (C) One-way ANOVA, Tukey post-tests. (D–E) Two-way ANOVA, Bonferroni post-tests. (F) Two-way ANOVA, Bonferroni post-tests. (C,F) *p<0.05 **p<0.01 ***p<0.001 ****p<0.0001. (D,E) veh vs 5×10⁷ cfu: ***p<0.001 ****p<0.0001, veh vs 5×10⁶ cfu: †p<0.05 ††p<0.01 †††p<0.001 ††††p<0.0001, veh vs 5×10⁵ cfu: ^§§p<0.01 ^§§§§p<0.0001. ns=not significant. Mean±SEM. See Figure S1 for related data.

Figure 2. S. pyogenes induces neuronal activation and CGRP release through SLS

(A) Representative Fura-2 ratiometric fields (left) and calcium traces (center) of DRG neurons responding to filtered supernatant from S. pyogenes M1 (5×10⁹ cfu/mL), capsaicin (1 µm), and KCl (40 mM). Proportions (Right) of capsaicin non-responsive (Cap−) and capsaicin responsive (Cap+) neurons that responded to M1 supernatant (n=3–4 fields/condition). (B–D) Representative Fura-2 ratiometric fields (B) and calcium traces (C) of DRG neurons stimulated with filtered supernatant from S. pyogenes M1 (wt) or isogenic mutants lacking SLS (ΔsagA), both SLO and SLS (ΔsloΔsagA), or double mutant bacteria in which sagA expression was restored (ΔsloΔsagA+pDL:sagA). (D) Proportions of responding DRG neurons to bacterial supernatant from S. pyogenes M1 (wt) or isogenic mutant strains (n=3 fields/condition). (E) DRG neurons stimulated for 30 min with supernatant from S. pyogenes M1 (wt), isogenic mutants, or medium, analyzed for in vitro release of CGRP (n=5 samples/group). Statistical analysis: (A) Two-way ANOVA, Bonferroni post-tests. (D,E) One-way ANOVA, Tukey post-tests. *p<0.05 ***p<0.001 ****p<0.0001. ns=not significant. Scale bars, 50 µm. Mean±SEM. See Figure S2 for related data.

Figure 3. SLS is necessary for pain during S. pyogenes infection

(A,B) Spontaneous lifting/licking pain over 1 h after injection of vehicle, S. pyogenes M1 or M3 (5×10⁸ cfu) wt, ΔsagA, Δslo, or ΔsloΔsagA strains (M1, n=8/group; M3, n=12/group). (C,D) Mechanical (n=10/group) and heat (n=9–10/group) sensitivity after injection of S. pyogenes M3 wt or isogenic mutants (5×10⁷ cfu). (E) Spontaneous pain over 1 h in mice injected with S. pyogenes M1 (5×10⁸ cfu) and treated with anti-SLS or control IgG (n=4–5/group). (F) Spontaneous pain over 1 h after injection of S. pyogenes M1 (5×10⁸ cfu) wt or isogenic mutants complemented with plasmid encoding sagA (pDL:sagA) or empty plasmid (pDL278) (n=8/group). Statistical analysis: (A,B,E,F) One-way ANOVA, Tukey post-tests, **p<0.01 ***p<0.001 ****p<0.0001; (C,D) Two-way ANOVA, Bonferroni post-tests, ΔsagA vs wt *p<0.05 ***p<0.001 ****p<0.0001, ΔsloΔsagA vs wt ††††p<0.0001. ns=not significant. nd=not detected. Mean±SEM. See Figure S3 for related data.

Figure 4. TRPV1 neurons that mediate pain inhibit host defenses against S. pyogenes infection

(A) Heat sensitivity measured in Trpv1-Cre/Dta mice and control littermates after S. pyogenes M1 injection (5×10⁷ cfu, n=7–8/group). (B) Mechanical sensitivity of ipsilateral (ipsi) and contralateral (contra) hind paws after S. pyogenes M1 injection (5×10⁷ cfu) in Trpv1-Cre/Dta and control littermates (n=5/group). (C) Heat sensitivity in RTX and vehicle-treated mice after S. pyogenes M1 injection (5×10⁷ cfu, n=5/group). (D) Mechanical sensitivity of ipsilateral and contralateral hind paws of RTX and vehicle-treated mice after S. pyogenes M1 injection (5×10⁷ cfu, n=5/group). (E–H) S. pyogenes M1 (5×10⁶ cfu) was injected subcutaneously into the flank of mice: (E) Representative pictures of flank lesions, and (F) Quantification of dermonecrosis and weight loss at different time points after S. pyogenes M1 injection in Trpv1-Cre/Dta or control littermates (n=14–16/group). (G) Representative pictures of flank lesions, and (H) Quantification of dermonecrosis and weight loss after S. pyogenes M1 injection in RTX or vehicle-treated mice (n=15/group). Statistical analysis: (A–H) two-way ANOVA, Bonferroni post-tests. *p<0.05 **p<0.01 ***p<0.001 ****p<0.0001. (B) Trpv1-Cre/Dta (ipsi) vs Control (ipsi). (D) RTX (ipsi) vs Veh (ipsi). Mean±SEM. See Figure S4 for related data.

Figure 5. Nociceptors suppress recruitment of neutrophils that mediate host protection against S. pyogenes infection

(A) Histopathology of flank biopsies from vehicle or RTX-treated mice 3 days after injection of S. pyogenes M1 (5×10⁶ cfu). Scale bars, 50 µm. (B) Bacterial load recovery (log₁₀ cfu) from flank lesions and spleens in RTX or vehicle-treated mice after S. pyogenes M1 injection (5×10⁶ cfu, n=4/group). (C–E) Flow cytometry of leukocyte recruitment in necrotizing lesions 1 day after S. pyogenes M1 injection (5×10⁶ cfu): (C) Representative FACS plots showing neutrophils (CD11b⁺Ly6G⁺ gates) in lesion samples. (D–E) Quantification of immune cell populations by flow cytometry in flank biopsies from infected Trpv1-Cre/Dta mice or control littermates (n=4/group), or from uninfected mice, infected vehicle-treated mice, or infected RTX-treated mice (n=4–5/group). (F–H) Measurement of CGRP release ex vivo from flank skin punch biopsies. (G) CGRP release from uninfected skin (0 h), 7 h, or 24 h after S. pyogenes M1 injection (5×10⁶ cfu) (n=3/group). (H) CGRP release from uninfected skin or 7 h after S. pyogenes M1 (5×10⁶ cfu) injection of Trpv1-Cre/Dta mice or control littermates, or Vehicle or RTX-treated mice (n=3/group). Statistical analysis: (B,D,E,H) Two-way ANOVA, Bonferroni post-tests. (G) One-way ANOVA, Tukey post-tests. *p<0.05 **p<0.01 ***p<0.001 ****p<0.0001. ns=not significant. nd=none detected. Mean±SEM. See Figure S5 for related data.

Figure 6. Local vs. intrathecal BoNT/A injection dissociates pain perception from peripheral neuro-immune suppression

(A–D) Subcutaneous administration of BoNT/A (25 pg/100 µL) or vehicle 6 days prior to S. pyogenes M1 injection in flank skin (5×10⁶ cfu). (B) Representative images of lesions (day 8), (C) Dermonecrosis size measurements, and (D) Weight loss over time after injection of S. pyogenes (n=5–10/group). (E–H) Intrathecal administration of BoNT/A or vehicle 1 day prior to S. pyogenes M1 injection in flank skin (5×10⁶ cfu). (F) Representative images of lesions (day 8), (G) Dermonecrosis size measurements, and (H) Weight loss over time after injection of S. pyogenes (n=6/group). (I) DRG neurons exposed to BoNT/A (25 pg/200 µL) or medium for 24 h were stimulated with S. pyogenes supernatant (5×10⁹ cfu/mL) for 30 min, and CGRP was measured in neuronal supernatant (n=5/group). (J) CGRP release from skin punch biopsies of mice treated intrathecally or locally with BoNT/A, 7 h after S. pyogenes M1 (5×10⁶ cfu) injection (n=3/group). Statistical analysis: (C,D,G,H) Two-way ANOVA, Bonferroni post-tests. (I,J) One-way ANOVA, Tukey post-tests. *p<0.05 **p<0.01 ***p<0.001 ****p<0.0001. ns=not significant. Mean±SEM. See Figure S6 for related data.

Figure 7. BoNT/A and CGRP antagonism block neural modulation of immunity to treat bacterial invasion

(A) DRG neurons were pretreated with BoNT/A for 24 h, or with CGRP antagonists (CGRP_8–37 or BIBN4096) immediately before co-incubation with mouse neutrophils and S. pyogenes M1 for 1 h. Bacterial survival was measured as the multiplication factor of surviving colonies/starting inoculum (n=3–4 replicates/group). (B) Mouse neutrophils were incubated with S. pyogenes M1 in presence of CGRP or vehicle for 1 h, and bacterial survival measured (n=4/group). (C) Human whole blood was incubated with S. pyogenes M1 in presence of CGRP or vehicle for 3 h, and bacterial survival measured (n=3/group). (D) Representative images of lesions at day 8 (left) and dermonecrosis size (right) of mice treated 2 h after S. pyogenes M1 injection (5×10⁶ cfu) with vehicle, BoNT/A, or BIBN4096 (n=6–7/group). (E–G) Mice were treated subcutaneously with BoNT/A or vehicle at day 2 and day 9 following flank injection of S. pyogenes M1 (5×10⁶ cfu). Representative images show lesions before and after treatment (E). Dermonecrotic lesions (F) and abscess sizes (G) were measured over time (n=10/group). Blue dots show injection sites at day 2 and day 9. Arrows show BoNT/A treatments. Statistical analysis: (A–C) One-way ANOVA, Tukey post-tests. (D–G) Two-way ANOVA, Bonferroni post-tests. (A–C,F–G) *p<0.05 **p<0.01 ***p<0.01 ****p<0.0001. (D) BIBN4096 vs veh: *p<0.05 **p<0.01 ***p<0.001 ****p<0.0001, BoNT/A vs veh: †p<0.05 ††p<0.01 †††p<0.001 ††††p<0.0001. ns=not significant. Mean±SEM. See Figure S7 for related data.