Thrombosis is a frequent, life-threatening complication of systemic infection associated with multiple organ damage. We have previously described a novel mechanism of inflammation-driven thrombosis induced by Salmonella Typhimurium infection of mice. Thrombosis in the liver develops 7 days after infection, persisting after the infection resolves, and is monocytic cell dependent. Unexpectedly, thrombosis was not prominent in the spleen at this time, despite carrying a similar bacterial burden as the liver. In this study, we show that thrombosis does occur in the spleen but with strikingly accelerated kinetics compared with the liver, being evident by 24 hours and resolving rapidly thereafter. The distinct kinetics of thrombosis and bacterial burden provides a test of the hypothesis that thrombi form in healthy vessels to trap or remove bacteria from the circulation, often termed immunothrombosis. Remarkably, despite bacteria being detected throughout infected spleens and livers in the early days of infection, immunohistological analysis of tissue sections show that thrombi contain very low numbers of bacteria. In contrast, bacteria are present throughout platelet aggregates induced by Salmonella in vitro. Therefore, we show that thrombosis develops with organ-specific kinetics and challenge the universality of immunothrombosis as a mechanism to capture bacteria in vivo.

Clearance of intracellular infections caused by Salmonella Typhimurium (STm) requires IFN-g and the Th1-associated transcription factor T-bet. Nevertheless, whereas IFN-g 2/2 mice succumb rapidly to STm infections, T-bet 2/2 mice do not. In this study, we assess the anatomy of immune responses and the relationship with bacterial localization in the spleens and livers of STminfected IFN-g 2/2 and T-bet 2/2 mice. In IFN-g 2/2 mice, there is deficient granuloma formation and inducible NO synthase (iNOS) induction, increased dissemination of bacteria throughout the organs, and rapid death. The provision of a source of IFN-g reverses this, coincident with subsequent granuloma formation and substantially extends survival when compared with mice deficient in all sources of IFN-g. T-bet 2/2 mice induce significant levels of IFN-g 2 after challenge. Moreover, T-bet 2/2 mice have augmented IL-17 and neutrophil numbers, and neutralizing IL-17 reduces the neutrophilia but does not affect numbers of bacteria detected. Surprisingly, T-bet 2/2 mice exhibit surprisingly wild-type-like immune cell organization postinfection, including extensive iNOS + granuloma formation. In wild-type mice, most bacteria are within iNOS + granulomas, but in T-bet 2/2 mice, most bacteria are outside these sites. Therefore, Th1 cells act to restrict bacteria within IFN-g-dependent iNOS + granulomas and prevent dissemination.

Krüppel-like factor 2 (KLF2) is a potent regulator of lymphocyte differentiation, activation and migration. However, its functional role in adaptive and humoral immunity remains elusive. Therefore, by using mice with a B cell-specific deletion of KLF2, we investigated plasma cell differentiation and antibody responses. We revealed that the deletion of KLF2 resulted in perturbed IgA plasma cell compartmentalization, characterized by the absence of IgA plasma cells in the bone marrow, their reductions in the spleen, the blood and the lamina propria of the colon and the small intestine, concomitant with their accumulation and retention in mesenteric lymph nodes and Peyer’s patches. Most intriguingly, secretory IgA in the intestinal lumen was almost absent, dimeric serum IgA was drastically reduced and antigen-specific IgA responses to soluble Salmonella flagellin were blunted in KLF2-deficient mice. Perturbance of IgA plasma cell localization was caused by deregulation of CCR9, Integrin chains αM, α4, β7, and sphingosine-1-phosphate receptors. Hence, KLF2 not only orchestrates the localization of IgA plasma cells by fine-tuning chemokine receptors and adhesion molecules but also controls IgA responses to Salmonella flagellin.