Priming innate immune responses to infection by cyclooxygenase inhibition kills antibiotic-susceptible and -resistant bacteria

Abstract

Inhibition of cyclooxygenase (COX)-derived prostaglandins (PGs) by nonsteroidal anti-inflammatory drugs (NSAIDs) mediates leukocyte killing of bacteria. However, the relative contribution of COX1 versus COX2 to this process, as well as the mechanisms controlling it in mouse and humans, are unknown. Indeed, the potential of NSAIDs to facilitate leukocyte killing of drug-resistant bacteria warrants investigation. Therefore, we carried out a series of experiments in mice and humans, finding that COX1 is the predominant isoform active in PG synthesis during infection and that its prophylactic or therapeutic inhibition primes leukocytes to kill bacteria by increasing phagocytic uptake and reactive oxygen intermediate-mediated killing in a cyclic adenosine monophosphate (cAMP)-dependent manner. Moreover, NSAIDs enhance bacterial killing in humans, exerting an additive effect when used in combination with antibiotics. Finally, NSAIDs, through the inhibition of COX prime the innate immune system to mediate bacterial clearance of penicillin-resistant Streptococcus pneumoniae serotype 19A, a well-recognized vaccine escape serotype of particular concern given its increasing prevalence and multi-antibiotic resistance. Therefore, these data underline the importance of lipid mediators in host responses to infection and the potential of inhibitors of PG signaling pathways as adjunctive therapies, particularly in the con-text of antibiotic resistance.

Figure 1

Characterisation of infectious peritonitis. Injecting GBS i.p. to W/T mice triggers an immediate (A) leukocytes infiltrate comprising (B) GR1-positive PMNs and F4/80-positive macrophages as well as (C) CD3-positive T and CD19-positive B cells. Total peritoneal leukocytes were prepared for (D) COX 1 and COX 2 protein expression while (E) PGE₂, (F) prostacyclin, measured as its stable metabolites 6-keto PGF1α and 2,3-dinor 6-keto PGF_1α, (G) TNFα and (H) IL-10 were determined in the cell-free inflammatory exudates. Values are expressed as the mean ± SEM of 5-12 mice/group.

Figure 2

COX 1 is the predominant isoforms functional during infectious peritonitis. Selective COX inhibitors (SC-560 for COX 1 and NS-398 for COX 2) and non-selective COX inhibitors (aspirin and indomethacin) were dosed orally to W/T animals 1h before GBS injection. Cell-free peritoneal exudate levels of (A) PGE₂ and (B) PGD₂ were measured 3h after GBS with similar results obtained for (C) PGE₂ and (D) PGD₂ using COX 1 knockout mice. In addition, COX inhibitors were dosed to mice 1h before i.p. zymosan (non-infectious stimulus) with cell-free peritoneal exudate levels of (E) PGE₂ and (F) PGD₂ measured 3h later. NS-398 was used at dosing levels that do not inhibit (G) COX 1-derived plasma TxB₂ measured 3h after i.p. GBS injection. Data were analysed by one-way ANOVA and Dunnett’s multiple comparison test or by unpaired Student T-test. Values are expressed as the mean ± SEM of between 5-12 mice/group. * P < 0.05; ** P < 0.01 and *** P <0.001.

Figure 3

COX inhibition enhances bacteria killing in rodents. Mice were given dual (indomethacin/aspirin), COX 1 (SC-560) or COX 2 (NS-398) selective inhibitors orally either (A) 1h before or (B) therapeutically 1h after i.p. inoculation of GBS. To exclude dampened bacterial translocation as a potential explanation for reduced plasma GBS titres following COX inhibition, peritoneal exudate levels of (C) GBS were measured 3h after inoculation. GBS was also injected i.p. to (D) COX 1 knockout mice with plasma taken 3h later for overnight cfu culture number determination as well as (E) incubated in culture media only with NSAIDs in the absence of leukocytes. Data were analysed by one-way ANOVA and Dunnett’s multiple comparison test or by unpaired Student T-test. Values are expressed as the mean ± SEM of between 5-12 mice/group. * P < 0.05 and ** P < 0.01.

Figure 4

COX inhibition alters phagocytosis, cytokine synthesis and bacterial killing mechanisms in a cAMP-dependent manner. GBS was injected i.p. to (A) W/T mice dosed orally 1h earlier with indomethacin (dual COX inhibitor), SC-560 (COX 1 inhibitor) or NS-398 (COX 2 inhibitor) as well as to (B) COX 1 knockout mice. 3h after GBS injection (A-B) cAMP was measured in cell-free exudates while (C) phagocytosis and (D-E) NADPH oxidase activity was determined in total leukocytes. Cell-free exudate levels of (F-G) TNFα and IL-10 were also determined in NSAID-treated W/Ts and (H-I) COX 1 knockouts. This COX-inhibited differential change in W/T mice (J) bacterial killing, (K) NADPH oxidase activity and (L-M) cytokine synthesis was reversed by 15(S)-15 methyl PGE₂ (EP agonist) or db-cAMP given 5mins before GBS and therefore 55mins after NSAIDs. Data is represented and analysed by an unpaired Student’s T-test or ANOVA followed by either Dunnett’s or Bonferroni multiple comparison tests. Values are expressed as the mean ± SEM of 5-12 mice/group. * P < 0.05 and ** P < 0.01. *** P < 0.001.

Figure 4

COX inhibition alters phagocytosis, cytokine synthesis and bacterial killing mechanisms in a cAMP-dependent manner. GBS was injected i.p. to (A) W/T mice dosed orally 1h earlier with indomethacin (dual COX inhibitor), SC-560 (COX 1 inhibitor) or NS-398 (COX 2 inhibitor) as well as to (B) COX 1 knockout mice. 3h after GBS injection (A-B) cAMP was measured in cell-free exudates while (C) phagocytosis and (D-E) NADPH oxidase activity was determined in total leukocytes. Cell-free exudate levels of (F-G) TNFα and IL-10 were also determined in NSAID-treated W/Ts and (H-I) COX 1 knockouts. This COX-inhibited differential change in W/T mice (J) bacterial killing, (K) NADPH oxidase activity and (L-M) cytokine synthesis was reversed by 15(S)-15 methyl PGE₂ (EP agonist) or db-cAMP given 5mins before GBS and therefore 55mins after NSAIDs. Data is represented and analysed by an unpaired Student’s T-test or ANOVA followed by either Dunnett’s or Bonferroni multiple comparison tests. Values are expressed as the mean ± SEM of 5-12 mice/group. * P < 0.05 and ** P < 0.01. *** P < 0.001.

Figure 5

Prophylactic but not therapeutic COX inhibition triggers pro-inflammatory cytokine synthesis. W/T mice were injected i.p. with GBS 1h after oral administration of NS-398 (COX 2 inhibitor), SC-560 (COX 1 inhibitor) or SC-560 plus infliximab with (A) complete reversal of COX 1-mediated cytokine storm (TNFα synthesis, Figure 4F) afforded by infliximab. SC-560 and NS-398 were dosed orally 1h after GBS injection causing little effects on (B-C) inflammatory cytokine levels in cell-free inflammatory exudates 3h later. Survival analysis was completed on 8 mice/group and the log-rank test was used to compare each group against one another. ** values represent P < 0.01 for controls versus Sc-560/infliximab. For cytokines, values are expressed as the mean ± SEM of 5-7 mice/group and analysed by one-way ANOVA and Dunnett’s multiple comparison test. *values represent P < 0.05. Survival experiments had 8 mice/group.

Figure 6

COX inhibition primes human whole blood to kill bacteria. Peripheral blood of 8 healthy male volunteers between the ages of 25-50 was taken before and 1h after ingestion of 500mg naproxen and incubated ex vivo with opsonised GBS. After 60min aliquots were taken to determine (A i) numbers of viable bacteria determined by overnight incubation on agar plates or (A ii) leukocyte NADPH oxidase activity. Blood from volunteers taken naproxen was further incubated with (B i-iii) penicillin (0.075μg/ml) to determine NSAID inhibitory/additive effects on bacterial killing with antibiotics. (C) Blood was also treated 400μM naproxen for 30min and stimulated with FAM-SE labelled S. pneumoniae (ST193) 5min to determine bacterial phagocytosis and to show that (D) NSAIDs do not interfere with opsinisation of bacteria. Data were analysed by paired Student’s T-test. * values represent P < 0.05 and ** values represent P < 0.01.

Figure 7

Inhibiting PG signalling primes human whole blood to kill antibiotic-resistant bacteria. Healthy human peripheral blood was pre-incubated with/without penicillin and a pan EP/DP PG receptor antagonist targeted against EP₁, EP₂, EP₃-III, and DP₁ receptors as well as against individual PG receptors including EP4, DP1 and IP. Whole blood was then stimulated with either (A) serum-opsonised penicillin-susceptible (ST193) or (B) intermediate penicillin-resistant (ST199) S. pneumonia for 1h. After which time, numbers of surviving S. pneumoniae (cfu) were quantified 24h later on agar plates. Data were analysed by one-way ANOVA and Bonferroni’s multiple comparison test. * values represent P < 0.05, ** values P < 0.01 and *** values P < 0.001.