The Inflammatory Cytokine IL17 Tunes and Amplifies the Erythropoietic Response to Erythropoietin In Vivo

Erythropoiesis is regulated by a well-established negative feedback loop that maintains tissue oxygen tension through the oxygen-sensitive transcription of erythropoietin (Epo). Epo is the principal and essential regulator of erythropoiesis in both the basal state and the response to stress. It is unclear, however, whether and how the feedback control of erythropoiesis is tuned by additional factors that signal environmental stresses, such as inflammation, which impact erythropoiesis. To date, only a handful of factors other than Epo are known to increase erythropoietic rate, and options in the clinic are limited. Using single-cell RNA-sequencing (scRNA-seq), we recently found that the receptor IL17RA is broadly expressed in early hematopoiesis, including in early erythropoiesis, and that adding its ligand, the pro-inflammatory cytokine IL17A, amplifies the Epo response of adult bone-marrow CFU-e in colony-formation assays (Tusi et al., Nature 2018). We also found that IL17A synergizes with Epo in activating Stat3 and Stat5 in freshly explanted bone-marrow erythroid progenitors and precursors, respectively. Here we investigate whether IL17A regulates erythropoiesis in vivo and how it interacts with the Epo-controlled feedback regulation of this process.

We first established which other IL17 family members regulate erythropoiesis. We found that the homodimeric ligands IL17A and IL17F, and the heterodimeric IL17A-IL17F, and no other IL17 ligands, increase CFU-e colony formation in vitro. This ligand specificity identifies the IL17RA/IL17RC heterodimeric receptor as the likely signaling receptor on erythroid progenitors. We next examined the erythropoietic effect of IL17A in vivo by administering 8- to 10-week-old mice with either vehicle, IL17A (200 ng/g twice daily), Epo (0.25 U/g once daily), or with both Epo and IL17A, for 72 hours. Analysis by cytometry and scRNA-Seq revealed that while there was little erythropoietic response to administration of IL17A alone, IL17A synergizes strongly with Epo, in both the spleen and bone marrow, and significantly increases reticulocytes in peripheral blood. The response in the spleen is especially striking: joint Epo/IL17A treatment expands splenic ProE by 270-fold over vehicle-treated controls, as compared with no significant increase and 40-fold increase upon treatment with IL17A or Epo alone, respectively.

We investigated whether IL17A plays a role in the erythroid stress response to hypoxia. Mice placed in a hypoxic (10% oxygen) chamber for 24 hours had elevated plasma IL-17A levels compared with matched normoxic mice (5.49 ± 2.09 pg/mL vs 1.19 ± 0.52 pg/mL, p=0.03). Further, 24-hour pre-treatment with IL17A accelerated the erythropoietic response of mice to hypoxia. Mechanistically, Epo and IL17A may combine to accelerate erythroid cell cycling. Using a mouse transgenic for a live-cell reporter of cell cycle length, we found that 24-hour hypoxia and IL17A resulted in shorter cycles throughout the erythroid differentiation trajectory, particularly in early progenitors (BFU-e, CFU-e).

Transcriptionally, the most notable differentially expressed genes in response to the joint IL17A and Epo treatment are genes that are already upregulated by stimulation of EpoR signaling, including Podxl, Cda, Cpd and Gdf3; the addition of IL17A results in their further increase. We also examined this effect by labeling erythroid progenitors and precursors with antibody to Podxl. We found that Podxl is not expressed in basal bone marrow but is rapidly induced on the cell surface of late CFU-e and ProE cells following Epo injection; this induction is further amplified in mice that are injected with both Epo and IL17A.

These findings suggest that IL17A, induced in response to stresses such as pulmonary infection, does not affect erythropoiesis by itself. However, its induction provides a crucial pre-emptive increased sensitivity of erythroid progenitors to Epo, thus accelerating the erythropoietic response should Epo also rise as disease progresses. Taken together, these findings reveal a mechanism for an integrated and coherent feedback control of erythropoiesis that allows it to respond efficiently to a variety of stresses. Similar control circuits may apply more broadly in the regulation of hematopoiesis.