Inflammatory monocyte recruitment is regulated by interleukin-23 during systemic bacterial infection

Abstract

Listeria monocytogenes is a gram-positive intracellular pathogen that causes meningitis and septicemia in immunocompromised individuals and spontaneous abortion in pregnant women. The innate immune response against L. monocytogenes is primarily mediated by neutrophils and monocytes. Interleukin-23 (IL-23) is an important proinflammatory cytokine well known for its role in neutrophil recruitment in various infectious and autoimmune diseases. We have previously shown that IL-23 is required for host resistance against L. monocytogenes and for neutrophil recruitment to the liver, but not the spleen, during infection. Despite efficient neutrophil recruitment to the spleen, IL-23p19 knockout (KO) mice have an increased bacterial burden in this organ, suggesting that IL-23 may regulate the recruitment/function of another cell type to the spleen. In this study, we show that specific depletion of neutrophils abrogated the differences in bacterial burdens in the livers but not the spleens of C57BL/6 (B6) and IL-23p19 KO mice. Interestingly, L. monocytogenes-infected IL-23p19 KO mice had fewer monocytes in the spleen than B6 mice, as well as a reduction in the monocyte-recruiting chemokines CCL2 and CCL7. Additionally, the overall concentrations of tumor necrosis factor alpha (TNF-α) and nitric oxide (NO(•)), as well as the percentages and total numbers of monocytes producing TNF-α and NO(•), were reduced in IL-23p19 KO mice compared to levels in B6 mice, leading to increased bacterial burdens in the spleens of L. monocytogenes-infected IL-23p19 KO mice. Collectively, our data establish that IL-23 is required for the optimal recruitment of TNF-α- and NO(•)-producing inflammatory monocytes, thus revealing a novel mechanism by which this proinflammatory cytokine provides protection against bacterial infection.

Fig 1

Depletion of neutrophils increases the susceptibility of IL-23p19 KO mice to L. monocytogenes infection. B6 and IL-23p19 KO mice were infected with ∼10⁴ L. monocytogenes CFU. These mice were treated with anti-Ly6G or isotype antibody at 1 day prior to infection and at 3 days p.i. The susceptibility of the mice was monitored for 14 days p.i. These data are combined from two independent experiments (n = 9 to 10/group). A log rank analysis was performed to determine statistical differences between groups. An asterisk (*) indicates a significant difference from all the other groups.

Fig 2

Depletion of neutrophils eliminates the differences in bacterial burden in the livers, but not the spleens, of B6 and IL-23p19 KO mice during L. monocytogenes infection. B6 and IL-23p19 KO mice were infected with ∼3,000 L. monocytogenes CFU. The mice were treated with anti-Ly6G or isotype antibody (Ab) at 1 day prior to infection and at 3 days p.i. At day 5 p.i., spleens and livers were harvested, and the bacterial CFU were enumerated in the spleen and the liver. A two-way ANOVA was performed to determine statistical differences between groups (*, significantly different from B6 mice; #, significantly different from isotype-treated controls). All data are expressed as means ± standard errors of the means (n = 5/group). These data are representative of two independent experiments.

Fig 3

IL-23 is required for the optimal recruitment of monocytes during L. monocytogenes infection. B6 and IL-23p19 KO mice were infected with ∼10⁴ L. monocytogenes CFU. The monocyte percentages in the blood, spleen, and liver were determined at day 0 (A), day 1 (C), and day 3 p.i. (E). The monocyte numbers in the spleen were examined at day 0 (B), day 1 (D), and day 3 p.i. (F). Monocytes were identified as cells expressing high levels of Ly6C and intermediate levels of CD11b (Ly6C^hi CD11b^int) in splenocytes from day 1 L. monocytogenes-infected mice (G). A t test was performed to determine statistical differences between groups (*, significantly different from B6 mice). All data are expressed as means ± standard errors of the means (n = 5/group). These data are representative of two independent experiments.

Fig 4

IL-23 is required for the optimal production of CCL2 and CCL7 during L. monocytogenes infection. B6 and IL-23p19 KO mice were infected with ∼10⁴ L. monocytogenes CFU for 1 day. ELISAs were performed to determine the concentrations of CCL2 and CCL7 in the serum (A), CCL2 in spleen homogenates (B), and CCL2 in supernatants from overnight cultures of splenocytes (C) and bone marrow cells (D) that were stimulated with HKLM or left unstimulated. A two-way ANOVA (A) and t tests (B, C, and D) were used to determine statistical differences between groups (*, significantly different from B6 mice). All data are expressed as means ± standard errors of the means (n = 5/group). These data are representative of two independent experiments.

Fig 5

IL-23 is required for the optimal production of TNF-α and NO^• in the spleen during L. monocytogenes infection. B6 and IL-23p19 KO mice were infected with ∼10⁴ L. monocytogenes CFU for 1 or 3 days (D1 or D3, respectively). The overall production of TNF-α and NO^• was determined by performing ELISAs on supernatants from overnight cultures of splenocytes stimulated with HKLM (A and B) and B6 splenocytes from day 3 p.i. stimulated with rIL-23 in the presence or absence of HKLM (C and D). A two-way ANOVA (A and B) and one-way ANOVA (C and D) were performed to determine statistical differences between groups (*, significantly different from B6 mice; **, significantly different from unstimulated and rIL-23-stimulated mice; #, significantly different from rIL-23 stimulation. All data are expressed as means ± standard errors of the means (n = 5/group). These data are representative of two independent experiments.

Fig 6

Optimal recruitment of TNF-α- and iNOS-producing inflammatory monocytes to the spleen is dependent on IL-23 during L. monocytogenes infection. B6 and IL-23p19 KO mice were infected with ∼10⁴ L. monocytogenes CFU for 1 or 3 days. Flow cytometric analysis was performed on isolated splenocytes to determine the expression of TNF-α or iNOS by monocytes (Ly6G⁻ Ly6C^hi). The percentages (A) and numbers (C) of TNF-α-producing monocytes in the spleen at days 1 and 3 p.i. were determined from dot plots (B). The percentages (D) and numbers (F) of iNOS-expressing monocytes in the spleen at day 3 p.i. were obtained after the corresponding isotype control values determined from dot plots (E) were subtracted. A two-way ANOVA (A and C) and t tests (D and F) were performed to determine statistical differences between groups (*, significantly different from B6 mice). All data are expressed as means ± standard errors of the means (n = 5/group). These data are representative of two independent experiments.