Divergent roles of IL-23 and IL-12 in host defense against Klebsiella pneumoniae

Abstract

Interleukin (IL)-23 is a heterodimeric cytokine that shares the identical p40 subunit as IL-12 but exhibits a unique p19 subunit similar to IL-12 p35. IL-12/23 p40, interferon gamma (IFN-gamma), and IL-17 are critical for host defense against Klebsiella pneumoniae. In vitro, K. pneumoniae-pulsed dendritic cell culture supernatants elicit T cell IL-17 production in a IL-23-dependent manner. However, the importance of IL-23 during in vivo pulmonary challenge is unknown. We show that IL-12/23 p40-deficient mice are exquisitely sensitive to intrapulmonary K. pneumoniae inoculation and that IL-23 p19-/-, IL-17R-/-, and IL-12 p35-/- mice also show increased susceptibility to infection. p40-/- mice fail to generate pulmonary IFN-gamma, IL-17, or IL-17F responses to infection, whereas p35-/- mice show normal IL-17 and IL-17F induction but reduced IFN-gamma. Lung IL-17 and IL-17F production in p19-/- mice was dramatically reduced, and this strain showed substantial mortality from a sublethal dose of bacteria (10(3) CFU), despite normal IFN-gamma induction. Administration of IL-17 restored bacterial control in p19-/- mice and to a lesser degree in p40-/- mice, suggesting an additional host defense requirement for IFN-gamma in this strain. Together, these data demonstrate independent requirements for IL-12 and IL-23 in pulmonary host defense against K. pneumoniae, the former of which is required for IFN-gamma expression and the latter of which is required for IL-17 production.

Figure 1.

Animals with targeted gene deletion in the IL-12, IL-17, and IL-23 signaling pathways demonstrate increased mortality during intrapulmonary K. pneumoniae infection. (A) WT C57BL/6, IL-23 p19^−/−, IL-12 p35^−/−, IL-12 p40^−/−, or IL-17R^−/− mice were challenged with 10⁴ CFU intratracheal K. pneumoniae and survival was recorded every 12 h (n = 16–20 per group). IL-12 p40 knockout mice showed the greatest susceptibility to infection (*P < 0.01 compared with C57BL/6 [log rank test]). Survival differences between IL-23 p19^−/−, IL-17R^−/−, and IL-12 p35^−/− mice were not statistically significant. (B) WT and p19^−/− mice were also challenged with a lower dose (10³ CFU) of bacteria, demonstrating significant mortality in p19^−/− mice to a sublethal dose of K. pnuemoniae (*P < 0.01 compared with WT mice; n = 10 per group).

Figure 2.

Cytokine mRNA expression following pulmonary K. pneumoniae infection. Animals were administered 10⁴ CFU bacteria and killed at specified time points. BAL cell pellet and lung homogenate mRNA were assayed via real-time RT-PCR. (A) Increases in BAL cell IL-23 p19, IL-12/23 p40, and IL-12 p35 mRNA expression during infection (n = 4–5 per group). (B) Whole lung tissue IL-23 p19, IL-17, and IL-17F mRNA expression following infection (n = 4–5 per group). Data are normalized for 18s ribosomal RNA content and plotted as fold change over baseline (time 0) expression. *Earliest significant (P < 0.05) increase in expression compared with time zero transcripts. Error bars represent mean ± SD.

Figure 3.

AM IL-23 expression is required for induction of splenocyte IL-17 expression in response to K. pneumoniae. (A) AMs from naive WT, p35^−/−, p40^−/−, and p19^−/− mice were recovered via BAL and exposed in vitro to K. pneumoniae. After 24 h, supernatants were harvested, centrifuged, and placed onto adherent cell-depleted WT splenocytes for 24 h to assay IL-17 induction (n = 5 per group; *P < 0.05 compared with media control). (B) IL-23 p19 mRNA expression in unexposed mDCs (mDC-control) or pDCs (pDC-control) or mDCs, AMs, and pDCs following 2-h in vitro exposure to K. pneumoniae. Data are expressed as fold increase in p19 expression compared with mDC- or pDC-control (n = 4–5 per group; *P < 0.05 compared with mDC-control). Error bars represent mean ± SD.

Figure 4.

IL-23 expression is required for lung IL-17 and IL-17F expression, whereas IL-12 is necessary for IFN-γ induction in response to K. pneumoniae infection. WT, p35^−/−, p40^−/−, and p19^−/− mice were infected with 10⁴ CFU K. pneumoniae and killed 24 h after infection. (A, B) Whole lung homogenate IL-17 (A) and IL-17F (B) mRNA expression as measured via real-time RT-PCR. (C) Whole lung homogenate IL-17 protein expression in repeat experiments. (D) Lung homogenate IFN-γ content in response to infection, indicating IL-23 is not sufficient to induce IFN-γ in the absence of IL-12. (E) Lung homogenate IFN-γ mRNA 24 h after K. pneumoniae infection confirms equivalent IFN-γ expression in WT and IL-23 p19^−/− mice (n = 6 per group; *P < 0.05 compared with WT). Error bars represent mean ± SD.

Figure 5.

IL-23 p19^−/− mice have reduced cytokine and chemokine responses to K. pneumoniae infection. Whole lung cytokine and chemokine content were measured in lung homogenates 24 h after 10⁴ CFU K. pneumoniae delivery. Values are normalized to homogenate protein concentration (n = 6 per group; *P < 0.05 compared with WT). Error bars represent mean ± SD.

Figure 6.

Defects in pulmonary host defense in IL-23 p19^−/− mice are rescued by IL-17 treatment. WT C57BL/6 or IL-23 p19^−/− mice were challenged with 10⁴ K. pneumoniae followed by intratracheal administration of 1.5 μg recombinant murine IL-17 (or phosphate-buffered saline vehicle) 12 h later. Animals were then killed 24 h after rmIL-17 (or vehicle) delivery. (A) IL-17 improves bacterial clearance in IL-23 p19^−/− mice. (B) IL-17 restores pulmonary concentrations of G-CSF, KC, and LIX in IL-23 p19^−/− mice. (C) IL-17 partially improves bacterial clearance in IL-12 p40^−/− mice. *P < 0.05 compared with WT vehicle-treated control. **P < 0.05 compared with vehicle-treated group of same genotype (n = 4–6 per group). Error bars represent mean ± SD.