The role of GRP94, an endoplasmic reticulum (ER) stress protein with both pro- and anti-inflammatory functions, has not been investigated in macrophages during ER stress, whereas ER stress has been reported in many diseases involving macrophages. In this work, we studied GRP94 in M1/LPS + IFNγ and M2/IL-4 primary macrophages derived from human monocytes (isolated from buffy coats), in basal and ER stress conditions induced by thapsigargin (Tg), an inducer of ER calcium depletion and tunicamycin (Tm), an inhibitor of N-glycosylation. We found that GRP94 was expressed on the membrane of M2 but not M1 macrophages. In M2, Tg, but not Tm, while decreased GRP94 content in the membrane, it induced its secretion. This correlated with the induction of a pro-inflammatory profile, which was dependent on the UPR IRE1α arm activation and on a functional GRP94. As we previously reported that GRP94 associated with complement C3 at the extracellular level, we analyzed C3 and confirmed GRP94-C3 interaction in our experimental model. Further, Tg increased this interaction and, in these conditions, C3b and cathepsin L were detected in the extracellular medium where GRP94 co-immunoprecipitated with C3 and C3b. Finally, we showed that the C3b inactivated fragment, iC3b, only present on non-stressed M2, depended on functional GRP94, making both GRP94 and iC3b potential markers of M2 cells. In conclusion, our results show that GRP94 is co-secreted with C3 under ER stress conditions which may facilitate its cleavage by cathepsin L, thus contributing to the pro-inflammatory profile observed in stressed M2 macrophages.

Background We previously showed that supernatants of Lactobacillus biofilms induced an anti-inflammatory response by affecting the secretion of macrophage-derived cytokines, which was abrogated upon immunodepletion of the stress protein GroEL. Methods We purified GroEL from L. reuteri and analysed its anti-inflammatory properties in vitro in human macrophages isolated from buffy coats, ex vivo in explants from human biopsies and in vivo in a mouse model of DSS induced intestinal inflammation. As a control, we used GroEL purified (LPS-free) from E. coli. ResultsWe found that L. reuteri GroEL (but not E. coli GroEL) inhibited pro-inflammatory M1-like macrophages markers, and favored M2-like markers. Consequently, L. reuteri GroEL inhibited pro-inflammatory cytokines (TNFa, IL-1b, IFNc) while favouring an anti-inflammatory secretome. In colon tissues from human biopsies, L. reuteri GroEL was also able to decrease markers of inflammation and apoptosis (caspase 3) induced by LPS. In mice, we found that rectal administration of L. reuteri GroEL (but not E. coli GroEL) inhibited all signs of haemorrhagic colitis induced by DSS including intestinal mucosa degradation, rectal bleeding and weight loss. It also decreased intestinal production of inflammatory cytokines (such as IFNc) while increasing anti-inflammatory IL-10 and IL-13. These effects were suppressed when animals were immunodepleted in macrophages. From a mechanistic point of view, the effect of L. reuteri GroEL seemed to involve TLR4, since it was lost in TRL4 -/mice, and the activation of a non-canonical TLR4 pathway. Conclusions L. reuteri GroEL, by affecting macrophage inflammatory features, deserves to be explored as an alternative to probiotics.