Tumor necrosis factor-alpha (TNFα) drives pathologic platelet hyperreactivity through signaling on hematopoietic stem/progenitor cells and expansion of canonical megakaryocytic progenitors. Free

Tumor necrosis factor-alpha (TNFα) has multiple roles for signaling inflammation and its two receptors, TNFαR1 and TNFαR2 exhibit alternative/competing signaling outcomes via continued cytokine production or cell death. The role of TNFɑ signaling on platelet function has been described in both age, systemic disease and myeloproliferative neoplasms leading to prothrombotic states. Notably, our group has previously shown that TNFɑ is the primary cytokine responsible for age-dependent platelet hyperreactivity. In this study we sought to better characterize the specific signaling effects of TNFɑ on megakaryocyte and megakaryocyte progenitors (MkPs). We generated a conditional mouse model to knockout megakaryocytic TNFαR1/2 (TNFαR1^flox/flox/2^flox/floxusing both Pf4-cre and Gp1b-cre.) Mice were treated with TNFα for 21 days and unexpectedly were not protected from developing hyperreactive platelets. To better understand why TNFαR1/2 deletion from megakaryocytes did not have an expected effect we performed single cell RNA sequencing on megakaryocyte-specific TNFαR1/2 knockout and wildtype (TNFαR1^flox/flox/2^flox/floxwith no Cre-recombinase) CD61-enriched bone marrow mononuclear cells after treatment with TNFα. After identification of hematopoietic stem cell and megakaryocyte clusters, pseudotime analysis demonstrated that transcripts for both TNFαR1/2 are not detected at the level of mature megakaryocytes. Further, upon query of the BloodSpot database of normal murine hematopoiesis, the lowest TNFαR1/2 expression is seen among megakaryocyte-erythroid progenitors and MkPs. Additionally, low/absent TNFαR1/2 transcripts are also seen in MkPs from human normal bone marrow samples, by single cell RNA-sequencing. We further confirmed this transcriptional finding in normal hematopoiesis by performing multicolor flowcytometry on isolated mouse bone marrow mononuclear cells and using biotinylated TNFα observed that the lack of receptor transcripts correlates to loss of the ability by MkPs (Lineage-/cKit+/Sca-1-/CD34-/CD41+) and megakaryocytes (Lineage-/CD41+/CD42d+) to bind TNFα, while it is retained in higher hematopoietic progenitors. We then hypothesized that if TNFα was not directly acting on megakaryocytes it was affecting hematopoietic stem and progenitor cells (HSPCs) upstream the traditional hematopoietic hierarchy to drive the observed hyperreactive phenotype. To test this hypothesis, we performed adoptive transplantation of either wildtype or germline TNFαR1/2 knockout (TNFαR1/2-KO) cKit-selected HSPCs into lethally irradiated mice. After allowing for engraftment, mice were administered TNFα for 21 days. Washed platelets were collected and following stimulation with 0.1 U/mL thrombin the platelets from TNFαR1/2-KO showed reduced platelet activation at multiple time points after thrombin activation.

To understand how chronic TNFα exposure alters hematopoiesis to support hyperreactive platelet formation we analyzed the HSPCs from mice treated with TNFα compared to vehicle in 8-12 week C57/Bl6 mice. Indeed, after chronic TNFα administration we observe expansion of both long-term and short-term HSCs as well as skewing of myeloid progenitors by immunophenotyping. Unexpectedly, TNFα decreased the frequency of mature megakaryocytes with a resultant expansion of MkPs. Importantly, there was no observed differences in platelet numbers of these mice. We performed immunophenotyping of MkPs to determine if the increase in MkPs was a result of direct/non-canonical derived stem cell ontogeny as recently, there has been emerging evidence of additional age-dependent changes to stem cell fitness, and their effects on megakaryopoiesis. Several groups have described atypical or non-canonical MK progenitors, including recently defined immunophenotypic markers, characterized by loss of SLAMF2 (CD48) and aberrant surface expression of JAM-A (CD321) that can be used to identify these progenitors. Utilizing a novel flow cytometry panel that identifies aberrant CD321 expression in MkPs we saw only expansion of canonical MkPs derived through stepwise differentiation.

Together these data show that chronic TNFα alters hematopoiesis favoring expansion of immunophenotypic stem cells with myeloid bias and megakaryocyte/platelet dysfunction. To our knowledge, this is the first study to demonstrate the effects of chronic TNFα on biasing in vivo hematopoiesis and to map TNFαR1/2's functional absence on megakaryocytes and their canonical progenitors.