The NF-κB transcription factor c-Rel controls host defense against Citrobacter rodentium.

The NF-B transcription factor c-Rel controls host defense against Citrobacter rodentium Maik Luu1, Rossana Romero1, Jasmin Bazant1, Elfadil Abass2, Sabrina Hartmann1, Hanna Leister1, Florence Fischer1, Rouzbeh Mahdavi1, Carlos Plaza-Sirvent3, Ingo Schmitz3,4, Ulrich Steinhoff1*, Alexander Visekruna1* 1Institute for Medical Microbiology and Hygiene, Philipps-University Marburg, Marburg, Germany. 2Department of Clinical Laboratory Science, College of Applied Medical Sciences, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia. 3Institute of Molecular and Clinical Immunology, Otto-von-Guericke University Magdeburg, Magdeburg, Germany. 4Systems-Oriented Immunology and Inflammation Research Group, Dept. of Experimental Immunology, Helmholtz Centre for Infection Research, Braunschweig, Germany Correspondence: Dr. Alexander Visekruna Institute for Medical Microbiology and Hygiene, Philipps-University, Marburg, Biomedical Research Center (BMFZ) Hans Meerwein Straße 2, 35032 Marburg, Germany [email protected] Tel: +49 (0)6421 / 2864359 Fax: +49 (0)6421 / 5866420 Authorship note: *AV and *US contributed equally to this work Keywords: Citrobacter rodentium, c-Rel, CD4+ T cells, B cells, NF-B The family of NF-B transcription factors comprises five closely related subunits that are involved in multiple aspects of adaptive and innate immune responses. Bacterial infections and chronic inflammation activate predominantly the canonical NF-B pathway consisting of RelA/p50 and c-Rel/p50 dimers [1]. The activation of NF-B leads to rapid proteasomal degradation of IB, a main inhibitory protein involved in the regulation of the canonical NF-B pathway, resulting in translocation of RelAand c-Rel-containing dimers into the nucleus and induction of target gene expression [2]. The potential role of NF-B in the protective immune responses against C. rodentium has not been investigated in the detail. The functional analysis of mice deficient for the p50 subunit of NF-B, which can differentially regulate immune responses by building either homodimers or heterodimers with various NF-B proteins, revealed that this protein is crucial for the eradication of C. rodentium infection [3]. Further, mice lacking the atypical IB protein IBNS showed impaired Th17 cell responses upon infection with C. rodentium [4]. The important role of c-Rel in regulating the function of immune cells prompted us to investigate if this transcription factor is crucial for orchestrating protective immune defense against C. rodentium. To examine the consequences of c-Rel deficiency for the course of C. rodentium infection, rel-/- and WT mice were orally infected with this pathogen and the bacterial numbers (colony forming units, CFU) were determined in the faeces. While WT mice were capable of eliminating the infection within approximiately 20 days, rel-/- animals were not able to clear the pathogen. Of note, the vast majority of c-Rel-deficient mice died between days 105 and 120 post infection (Fig. 1, A and B). At day 120 of infection, histological data demonstrated significant crypt hyperplasia and chronic infiltration of inflammatory cells into the colonic lamina propria of mice lacking cRel, while the WT intestine did not exhibit any immunopathology due to the clearance of infection within first three weeks (Fig. 1C). When we tested for the localization of C. rodentium on day 90 post infection, we found that, in contrast to the non-detectable levels of the pathogen in WT mice, this bacterium was detected in all examined extra-intestinal organs of rel-/- animals such as pancreas, liver, kidney and spleen, indicating for translocation and systemic spreading of C. rodentium (Fig. 1D). The high bacterial load outside of the intestine illustrates that c-Rel is an essential factor, which restricts the infection to the intestinal tissue. Innate immune cells such as dendritic cells (DCs), macrophages, neutrophils and innate lymphoid cells (ILCs) play a central role in the early phase of inflammatory responses during infection with C. rodentium. Notably, Rag1-/- rel-/- mice infected with C. rodentium displayed a prolonged survival as compared to Rag1deficient animals (Supporting Information Fig. 1A), suggesting that innate immune responses may compensate for partially defective function of intestinal DCs and ILCs in mice lacking c-Rel that was previously described [5]. T and B lymphocytes play an essential role during the elimination of C. rodentium. Rag1-/- mice lacking mature T and B cells are not able to eradicate this pathogen [6]. The NF-B signaling pathway is known to regulate the activity of lymphocytes, however the contribution of individual NF-B subunits to protective immune responses against pathogenic bacteria is only partially understood. To define the key effector mechanism involved in the c-Rel-mediated defense against C. rodentium infection, we next examined CD4+ T cells isolated from the colonic lamina propria at day 12 post infection. We found that the frequency of IL-17A+CD4+ and IFN+CD4+ T cells was significantly increased in the colon of WT mice as compared to rel-/- animals during the course of infection. Particularly, the double-positive IL-17A+IFN-+CD4+ T cell population, which is required for the clearance of C. rodentium, was significantly diminished in mice lacking the transcription factor c-Rel, suggesting that a defective T cell response contributes to the observed phenotype (Fig. 2, A-D). It is known that mice lacking c-Rel have a substantially impaired IL-2 production and that low secretion of IL-12 and IL-23 may also contribute to defective expression of cytokines by T cells during infection with C. rodentium [7, 8]. When we supplied IL-2 exogenously into rel-/- T cell cultures, we observed similar proliferation and IFN- production in Th1 cells compared to WT controls (Supporting Information Fig. 1, B and C). Of note, the nuclear translocation of RelA, a transcription factor related to c-Rel, was comparable between WT and rel-/- CD4+ T lymphocytes following T cell activation, suggesting for an intact RelA signaling in the absence of c-Rel (Supporting Information Fig. 1D). Our previous data revealed that the atypical IB protein IBNS was able to interact with c-Rel in the nucleus of T cells [9]. Since IκBNS does not contain a DNA-binding motif, this protein may require interaction with c-Rel to bind to target DNA sequences. Interestingly, both proteins, c-Rel and IκBNS appear to be required for the clearance of C. rodentium. It was previously shown that mice lacking c-Rel were resistant to collagen-induced arthritis and were not able to elicit a specific IgG antibody response to collagen type II (CII) [10]. Because c-Rel plays an important role for antibody generation, we next investigated C. rodentium-specific antibody production at day 12 post infection in serum and colon cultures. During infection with C. rodentium, rel-/- mice had only marginal C. rodentium-specific IgG antibody levels in serum and colon as compared to high amount of bacteria-specific antibodies in WT animals (Fig 2, E and F). Since clearance of C. rodentium infection is crucially dependent on humoral immune responses to bacterial proteins, we conclude that c-Rel-mediated generation of C. rodentium-specific antibodies is the key step in prevention of systemic spreading and elimination of this bacterium. Thus, c-Rel is required for protective antibacterial responses and for limitation of C. rodentium colonization to intestinal tissues. Acknowledgments: This study was supported by the Von Behring-Röntgen-Stiftung (Ulrich Steinhoff und Maik Luu), Manchot Stiftung (Rossana Romero), Loewe center novel drug targets against poverty related neglected tropical infectious disease (DRUID, Ulrich Steinhoff und Rouzbeh Mahdavi), Studienstiftung des deutschen Volkes (Maik Luu) and FAZIT-Stiftung (Hanna Leister and Alexander Visekruna). Conflict of interest: The authors declare no commercial or financial conflict of interest. References 1Zhang, Q. et al., Cell 2017. 168: 37-57. 2Grivennikov, S. I. et al., Cell 2010. 140: 883-899. 3Dennis, A. et al., Infect Immun 2008. 76: 4978-4988. 4Annemann, M. et al., J Immunol 2015. 194: 2888-2898. 5Visekruna, A. et al., Mucosal Immunol 2015. 8: 307-315. 6Vallance, B. A. et al., Infect Immun 2002. 70: 2070-2081. 7Liou, H. C. et al., Int Immunol 1999. 11: 361-371. 8Reinhard, K. et al., Eur J Immunol 2011. 41: 1388-1398. 9Schuster, M. et al., Immunity 2012. 37: 998-1008. 10 Campbell, I. K. et al., J Clin Invest 2000. 105: 1799-1806. Figure 1: Lack of c-Rel results in enhanced susceptibility to infection with C. rodentium. (A) WT and rel-/- mice were infected with 1x1010 CFU of C. rodentium and bacterial titers were measured at indicated time points (CFU per g stool). Data are shown for one out of two individual experiments, each with 6 mice per group. (B) Survival of WT and rel-/- mice was analyzed over the indicated time after infection with C. rodentium. Data are pooled from two individual experiments (n = 12 mice per group). (C) H&E-stained cryosections of colon of WT and rel-/- mice at day 120 post C. rodentium infection (left panel). Scale bar represents 100 µm. Right panel displays the histopathological scoring of colonic tissue from infected WT and rel-/- mice on day 120. Data are one representative of two individual experiments, each with 6 mice per group. Data are means ± SEM, ***P < 0.001 (Student´s t-test). (D) Bacterial titers (CFU) in indicated organs of WT and rel-/- mice were analysed on day 90 post C. rodentium infection. n.d. = not detectable. Data are one representative of two experiments, each with 6 mice per group. Figure 2: Impact of c-Rel on CD4+ T cell activity and antibody production during C. rodentium infection. (A-D) WT and rel-/- mice were infected with C. rodentium and colonic lamina propriaderived CD4+ T cells were analyzed by flow cytometry for intracellular cytokine production on day 12 post infection. The frequency of IL-17A and IFN- of gated CD4+ T cells is illustrated in representative dot plots and diagrams. Data represent one representative out of two individual experiments, each with 4 mice per group. (E and F) C. rodentium-specific IgG antibody titers were analyzed in serum and colon on day 12 post infection of WT and rel-/- and mice. Data are shown for one out of two individual experiments, each with 4-5 mice per group. Results are represented as the mean ± SEM, ***P < 0.001 (Student´s t-test).