CXCL4’s “Gliful” subversion of BM in MPN

In this issue of Blood, Gleitz et al report on a critical role of CXLC4 in inducing bone marrow (BM) fibrosis and inflammation, 2 hallmark features of primary myelofibrosis (PMF).¹ 

Classic BCR-ABL1^– myeloproliferative neoplasms (MPNs) include PMF, polycythemia vera, and essential thrombocythemia, all of which are characterized by disease-initiating somatic mutations in JAK2, CALR, or MPL genes.²  The mutations are mutually exclusive and converge to activate MPL-JAK-STAT signaling.³  However, despite the common apical molecular events, MPNs display variation in BM pathology, are associated with different risks of disease progression, and require management strategies targeting a unique set of morbidities associated with each subtype. Even as the mechanisms underlying the biological differences among different MPNs remain an area of active investigation, it has become clear that the mutant hematopoietic stem cells (HSCs) and their progeny exert cell extrinsic effects resulting in inflammation and BM fibrosis, which in turn, further contribute to disease progression.³  This is evidenced by the tight association of BM fibrosis with increased cytogenetic aberrations, greater number of disease-specific mutations, and mutations associated with adverse outcome.⁴ 

It is now well established that the target cell that acquires the MPN-initiating mutation is an undifferentiated HSC with multilineage potential. By contrast, the identity of cells responsible for BM fibrosis has remained elusive until recently when LepR⁺ and Gli1⁺ mesenchymal stromal cells were identified as the progenitors of fibrosis that cause myofibroblasts in the BM.^5,6  Once the identity of the cells contributing to BM fibrosis was uncovered, the next obvious question was unraveling the pathways involved in the cross-talk between the malignant hematopoietic cells and the fibrosis-causing stromal cells.

In their article, Gleitz et al leveraged in vitro coculture experiments as well as several models of fibrosis-associated MPNs to uncover the interaction between hematopoietic cells carrying the MPN-defining mutations and Gli1⁺ myofibroblasts. It was previously shown that megakaryocytes contribute to BM fibrosis in a multitude of ways,⁷  and now the authors show that mutant HSCs overexpress multiple CXC chemokines, particularly CXCL4, and can reprogram Gli1⁺ stromal cells with as little as 72 hours of coculture by inducing upregulation of matrisome-associated genes. In primary BM biopsies from patients with MPN, CXCL4 overexpression seemed to precede reticulin fibrosis. Furthermore, in JAK2^V617F and MPL^W515L mouse models of MPN, CXCL4 knockdown improved anemia, leukocytosis, thrombocytosis, megakaryocytic atypia, and spleen size. In addition, CXCL4 knockdown prevented increased expression of profibrotic transforming growth factor β (TGF-β) in megakaryocytes and of inflammatory pathways such as nuclear factor kappa B (NF-κB), tumor necrosis factor α (TNF-α), and TNF-related apoptosis-inducing ligand (TRAIL) in stromal cells in a thrombopoietin-induced model of BM fibrosis. However, even though there was reduced invasion of Gli1⁺ stromal cells into the marrow space, there was only modest reduction in BM fibrosis. More importantly, overexpression of CXCL4 alone was not sufficient to cause BM fibrosis, implicating additional pathways in the process.

CXCL4 was first implicated in the pathogenesis of PMF almost 4 decades ago.⁸  Gleitz and colleagues advance the field by uncovering the pathways downstream of CXCL4 and providing evidence that places CXCL4 upstream of many known inducers of inflammation and BM fibrosis. Moreover, the authors show that CXCL4 can positively amplify JAK/STAT signaling, a hallmark of all MPNs. CXCL4 therefore emerges as an attractive molecule for developing targeted therapy. However, even more fascinating is the observation that HSCs carrying MPN-associated mutations globally alter the gene expression profile of stromal cells in as little as 72 hours. Despite the fact that the data are derived from an ex vivo experiment, this observation raises several important questions. Is this reprograming irreversible or does it become imprinted and independent of the mutant HSCs? If indeed the reprogramming of fibrosis-inducing mesenchymal stromal cells can be induced early by mutant HSCs, this is likely to happen long before clinical presentation of the disease. Finally, if confirmed, this study has profound implications on how we manage our expectations for response to even the most targeted and fibrosis-reversing therapy in PMF. Surely, a reprogrammed stromal cell is unlikely to reverse its phenotype through targeting of 1 specific pathway. Although this hypothesis is nihilistic, it might explain only partial reversal of BM fibrosis seen in previous studies despite significant reduction in inflammatory cytokines.^9,10