Targeting the JAK2V617F Hematopoietic Stem Cell through Beta1 Integrin

The major barrier to progress in treatment of primary myelofibrosis (PMF) is the lack of effective therapies targeting the malignant JAK2 ^V617F hematopoietic stem cell (HSC), the source of disease progression and relapse. Expression of JAK2 ^V617F induces abnormal proliferation and differentiation of HSC, affecting all blood lineages but most prominently the erythro-megakaryocytic lineage, leading to clonal megakaryocyte proliferation, erythrocytosis and thrombocytosis.

In our previously published work ( Matsuura et al, Blood 2020) we identified elevated expression of the α5 subunit of the α5β1 integrin receptor as a driver of JAK2 ^V617F megakaryocyte proliferation. Expression of β1 integrin subunit was similar in wild-type (WT) and JAK2 ^V617F megakaryocytes, but activation of β1 integrin, detected by the conformation-specific 9EG7 antibody, was elevated in JAK2 ^V617F megakaryocytes. Following up on those findings, we detected elevated β1 integrin activation in long-term hematopoietic stem cells (HSC ^LT, LSKCD48-CD150+) from JAK2 ^V617F transgenic mice ( Xing et al, Blood 2008) when compared to WT controls.

To determine whether inhibition of β1 integrin in vivo reduces the number of HSC ^LT in JAK2 ^V617F transgenic mice, we examined HSC ^LT 24 hours after intravenous administration of 2mg/Kg of anti-β1 integrin inhibitory antibody HMβ1-1. This regimen led to a sharp reduction in number and percentage of HSC ^LT in JAK2 ^V617F transgenic mice, to 45% of control antibody-treated animals. The effect on WT HSC ^LT was negligible. HSC ^ST (LSKCD48-CD150-) and MPP2 (LSKCD48+CD150+) were also reduced, but the reduction was not statistically significant. MPP3/4 (LSKCD48+CD150-) and LSK were not affected. Megakaryocyte progenitors (LKCD41+CD150+) were significantly reduced. Inhibition of β1 integrin also sharply reduced peripheral blood platelet numbers in WT animals. The reduction in JAK2 ^V617F transgenic mice was not statistically significant. A longer observation time of 7 days demonstrated that while WT animals were able to mount a robust regenerative response with slight but significant increases in HSC ^LT and normalization of platelet numbers, in JAK2 ^V617F transgenic mice the number of HSC ^LT remained depressed, as did downstream stem/progenitor populations at various significance levels.

To gain insight into the mechanisms that lead to reduction in HSC ^LT, we examined apoptosis, cell cycle and HSC mobilization, all processes known to be affected by β1 integrin. Although basal expression of the apoptotic marker annexin V was decreased in JAK2 ^V617F HSC ^LT compared to WT, no differences were detected upon administration of HMβ1-1 in both genotypes at 4 hours post-administration. Similarly, peripheral blood colony assays revealed a basally higher level of stem/progenitor mobilization in JAK2 ^V617F, which did not significantly increase upon HMβ1-1 administration. In contrast, the cell cycle profile, examined at 4 hours post-administration to account for the earliest effects, demonstrated a decrease in G0 phase cells and increase in S/G2/M cells indicating cell cycle induction and exit from quiescence in both WT and JAK2 ^V617F HSC ^LT. The results indicate that inhibition of β1 integrin disrupts the adhesion-mediated cell cycle suppression in both WT and JAK2 ^V617F HSC ^LT leading to exit from quiescence. However, the downstream consequences of cell division in HSC ^LT differ significantly, with an apparent loss of regenerative potential in JAK2 ^V617F HSC ^LT, at least at the observed timepoints. Longer observation times are required to determine whether continuous β1 integrin inhibition can lead to JAK2 ^V617F HSC ^LT exhaustion. More sensitive assays to monitor mobilization are warranted to quantify possible minute amounts of mobilized HSC ^LT.

Our preclinical studies indicate that β1 integrin is an attractive target to specifically reduce the number of JAK2 ^V617F HSC ^LTin vivo. The role of β1 integrin in HSC ^LT has not been thoroughly studied due to lack of phenotype in β1 integrin-deficient HSC ( Brakebush et al, Immunity 2002). Our studies using the anti-β1 integrin antibody in WT HSC ^LT confirm previous findings. However, β1 integrin seems to be critical for maintenance of JAK2 ^V617F HSC ^LT, offering a possible therapeutic window. Understanding how β1 integrin supports quiescence in HSC ^LT will bring us a step closer to understanding the mechanisms that drive malignancy and fitness in JAK2 ^V617F HSC ^LT.