Role of the cytoplasmic pattern recognition receptor Nod2 in Coxiella burnetii infection

Abstract

Q fever is caused by Coxiella burnetii, an obligate intracellular bacterium, which survives in an acidic vacuole and escapes killing by preventing phagosome maturation [1]. In acute Q fever, granuloma formation is dependent on migration of leucocytes through the vascular endothelium. During chronic Q fever, granulomas are lacking and monocytes do not migrate across the endothelium [1], suggesting that host resistance against C. burnetii requires innate immune components. Toll-like receptors (TLRs) are pattern recognition receptors important in C. burnetii recognition. Although TLR4 ⁄) and control mice have a similar pattern of tissue infection and clearance, granuloma formation and cytokine response are controlled, in part, by TLR4 [1]. TLR2 is also involved in granuloma formation but seems dispensable for bacterial clearance [2]. The obligate intracellular lifestyle of C. burnetii drew our attention to elucidate the contribution of another family of proteins implicated in microbial detection, the cytoplasmic nucleotide-binding oligomerization domain (Nod) proteins, and more specifically Nod2, in innate immunity to C. burnetii. Nod2 contains leucine-rich repeats similar to those found in TLRs and is thought to be responsible for recognition of peptidoglycans from Gram positive and Gram negative bacteria through direct or indirect interaction with muramyl dipeptide. Nod2 signalling contributes to inflammatory response via activation of the transcription factor NF-jB, which is necessary for clearing pathogens. The intracellular localisation of Nod2 and C. burnetii makes Nod2 a potential candidate for C. burnetii recognition through leaking into the cytoplasm of muramyl dipeptide-containing cell wall fragments. Bone marrow-derived macrophages (BMDM) from control and Nod2 ⁄) mice [3] were incubated with C. burnetii for 4 h and cultured for 12 days. Infection was measured by determining the number of C. burnetii DNA copies as previously described [2]. Uptake and survival of C. burnetii in BMDM did not depend on Nod2 as DNA copy number was similar in control and Nod2 ⁄) BMDM (data not shown). Next, we tested whether the production of cytokines ⁄ chemokines by C. burnetii-stimulated BMDM required Nod2. We assessed the expression of genes encoding pro-inflammatory (TNF) and immunoregulatory cytokines (IL-10). BMDM were stimulated for 8 h with C. burnetii and gene expression was monitored by quantitative RTPCR [2]. In Nod2 ⁄) BMDM, although TNF transcripts were induced, they were significantly reduced compared with control macrophages (Fig. 1a). This result agrees with other studies showing that TNF and IL-12p40 levels are reduced in the absence of Nod2 after stimulation with Mycobacterium tuberculosis [4]. Note that IL10 was not affected by the absence of Nod2 (Fig. 1a). IL-10 is known to interfere with macrophage activation by stimulating the replication of C. burnetii through decreasing TNF production and blocking phagosomal maturation [1]. IL-10 is also associated with the chronic evolution of Q fever [1]. The absence of IL-10 modulation observed in Nod2 ⁄) BMDM is consistent with these data, as we did not observe any increase in bacterial replication. We also measured the expression of RANTES and MCP-1, two chemokines modulated in response to C. burnetii [5]. RANTES expression was fourfold increased in Nod2 ⁄) BMDM whereas MCP-1 expression was decreased by 40% (Fig. 1a). It is likely that RANTES and MCP-1 play a role in Q fever pathophysiology, as their production is Corresponding author and reprint requests: B. Desnues, CNRS UMR 6236, Faculte de Medecine de la Timone, 27 Bd J. Moulin, 13385 Marseille Cedex 05, France. E-mail: benoit.desnues@univmed.fr