Inflammatory states seen in infections and chronic disorders are often characterized by a condition called anemia of Inflammation (AI). Using a mouse model of AI generated by a single injection of the heat killed pathogen Brucella abortus (BA), we have previously shown that mice lacking IL6 (IL6 -KO mice) exhibited protected erythropoiesis in the bone marrow (BM) and a faster recovery from anemia compared to controls. To study the mechanism of IL6 mediated improved erythropoiesis under AI, we investigated erythroid recovery in WT and IL6 -KO mice injected with BA. 72 hours following BA administration, both genotypes showed impaired BM erythropoiesis associated by a surge in inflammatory cytokines such as IFNγ and TNFα and a concurrent increase in mitochondrial ROS (reactive oxygen species), specifically superoxide (SO) in erythroid progenitors. Cytokines levels were normalized after 24 hours. However, during the second phase of erythroid recovery (10-14 days following BA treatment), mice showed a second surge of inflammatory cytokines. During this phase, IL6 -KO mice showed significantly improved BM erythropoiesis and normalization of ROS levels. We analyzed SO levels by Mitosox red, generic ROS production using the chloromethyl derivative of 2′,7′-dichlorofluorescin diacetate and total protein oxidation (by looking at total protein carbonylation) by Western blot. These results were in sharp contrast to WT animals which continued to show upregulated ROS, total protein oxidation and slackened erythroid recovery.

Given the extensive association of iron overloading conditions and oxidative stress, we investigated the improved erythroid recovery within IL6 -KO mice in the context of increased iron absorption. We asked if and how BM erythropoiesis would change in IL6 -KO mice lacking Hepcidin (Hamp), a key iron regulatory hormone whose absence induces iron overload. In double knock out mice lacking both IL6 and Hamp (DKO), we observed an initial improved erythroid recovery in the BM, as indicated by increased number of total BM erythroid progenitors and reticulocytes, and improved RBC lifespan. However, all of these factors regressed after 10 days, with a simultaneous increase of cytokine production. The deterioration of erythropoiesis in the BM was accompanied by a concurrent increase in ROS. Of note, compared to non-treated mice, erythroid progenitors in the spleen of DKO-BA did not show increased ROS levels, and this was associated with an increased number of splenic erythroid progenitors and RBC. These results point to a potential role in the microenvironment in influencing erythropoiesis under inflammation. Furthermore, these results suggest that while lack of IL6 protects BM erythropoiesis, iron overload could antagonize it. In fact, iron overloaded IL6 -KO mice (with iron dextran) did show an impairment in BM erythroid recovery and upregulation of ROS, similarly to the DKO. Finally, we hypothesized that while there is a distinct positive correlation between upregulated ROS and weakened BM erythropoiesis, this comes into effect only in the context of inflammation. In other words, we speculated that some level of ROS upregulation is otherwise a physiological byproduct of erythropoietic stress associated with the onset of anemia. In line with this theory, we found that ROS levels were indeed upregulated in BM erythroid progenitors of phlebotomized animals. Our data suggests a model of erythroid recovery under AI wherein inflammation exacerbates ROS production triggered by stress erythropoiesis, and that this phenomenon exquisitely impairs BM erythropoiesis. This negative effect on BM erythropoiesis is mitigated by the absence of IL6, but the lack of IL6 mediated protective function is reversible under conditions of iron overload (Figure 1).

Our results, in combination with the upregulation of TNFa, suggest the involvement of non-apoptotic ROS associated programmed cell death such as necroptosis or ferroptosis, which are under investigation. We also have preliminary data pointing to mechanism of increased erythroid iron uptake that could potentially contribute to oxidative stress under AI. With these additional studies, we hope to further characterize the mechanisms surrounding improved BM erythropoiesis in the absence of ROS and IL6. Our studies could potentially characterize cellular responses that can be targeted in conditions of inflammation for therapeutic benefits.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.