In this study, we report that introduction of TEL-Syk into fetal liver hematopoietic cells leads to a rapidly progressive myelodysplasia with dramatic splenic and bone marrow fibrosis. Expression of TEL-Syk in progenitor cells induced a rapid (within 30 days following fetal liver cell transfer) expansion in the number of myeloid cells (neutrophils, monocytes, and eosinophils) in the peripheral blood as well as significant splenic and hepatic extramedullary hematopoiesis. However, with time (by 60 days following fetal liver cell transfer) TEL-Syk chimeric mice developed dramatic bone marrow, splenic and hepatic fibrosis that correlated with the appearance of anemia and thrombocytopenia, which likely contributed to the mortality seen in these animals. The expression of TEL-Syk in progenitors also lead to significant hematopoietic cell apoptosis, as shown by increased cleaved caspase-3 staining, which in combination with the fibrosis was likely the cause of the bone marrow and splenic hypocellularity in older (60 days) chimeras. TEL-Syk mice showed elevated inflammatory cytokines in serum with increases in MMPs, IGFBPs and other angiogenic-related factors. We further demonstrated that fetal liver hematopoietic cells expressing TEL-Syk manifest elevated levels of STAT5 phosphorylation in both resting and cytokine stimulated cells, which was partially resistant to JAK2 inhibition.
Expression of TEL-Syk in fetal liver progenitor cells induces colony formation and proliferation at very low cytokine levels (which otherwise do not support normal progenitor proliferation) due to hyperactivation of cytokine signaling pathways such as JAK2/STAT5. Besides being hyperresponsive to proliferation-inducing cytokines, we found that expression of TEL-Syk leads to overproduction of a number of proinflammatory cytokines. It is likely that cytokine overproduction establishes a paracrine feedback loop that contributes to myeloid cell proliferation and dysplasia in TEL-Syk expressing cells. In other words, both overproduction and hypersensitivity to growth-promoting cytokines could contribute to the MDS caused by TEL-Syk expression in progenitors. The cytokine hypersensitivity also caused skewing of myeloid cell development in in vitro
assays with increased numbers of abnormal appearing CFU-M colonies arising from Syk-deficient fetal liver cells. Surprisingly, at least in in vitro
liquid culture assays, we did not observe a significant difference in the proliferation rate of TEL-Syk expressing progenitors compared to vector transduced cells (data not shown). Hence, the increased cell numbers in the TEL-Syk CFU assays must be due to increased cell survival in vitro
. Since we observed just the opposite in vivo
(ie increased apoptosis in spleens of TEL-Syk chimeric mice), the complex developmental affects of TEL-Syk expression in progenitors is only partially reflected in standard methylcellulose CFU assays. Perturbation of hematopoietic progenitor populations has also been demonstrated in a mouse model of BCR-ABL-induced myelodysplasia [21
]. Expression of BCR-ABL in hematopoietic stem cells leads to a significant increase in splenic-derived myeloid progenitor populations, which contributes to myeloid cell expansion. In this model, overproduction of the proinflammatory cytokine IL-6 is crucial to maintain the myeloid cell expansion.
While expression of TEL-Syk in fetal liver hematopoietic cells induced rapid myeloproliferation with myleodysplasia, we did not observe outgrowth of blast-like cell types in these mice. Moreover, we were unsuccessful in adoptively transferring the myeloproliferative disease to secondary recipient mice using either irradiated or non-irradiated hosts (data not shown). Therefore it is unlikely that the disease process that we observed represents a myeloid cell malignancy, as is seen in mice receiving BCR-ABL transformed progenitor cells [22
]. It is likely that the high level of apoptosis induced in myeloid cells by expression of TEL-Syk prevents establishment of myeloproliferation in secondary recipient mice.
The fact that TEL-Syk expression in fetal liver hematopoietic cells leads to a myeloproliferative disease rather than lymphoid leukemia demonstrates a key difference between our data and experiments conducted by Wossning et al [16
]. In that work, the authors introduced TEL-Syk into differentiated pre-B cells, rather than a mixed population of hematopoietic cells, leading to CD19+
lymphoid leukemia. The variability in disease phenotype is likely to be context dependent such that TEL-Syk introduced into a mixed population of hematopoietic progenitors yields a myeloid disease, while TEL-Syk introduced into a lymphoid precursor yields a lymphoid leukemia. This effect has also been demonstrated in BCR-ABL+
], in which paracrine factors maintain lineage status, but the genetic lesion drives proliferation by deregulated signaling.
The extremely high rate of apoptosis we observed in TEL-Syk expressing mice is likely a major contributor to the bone marrow and splenic hypocellularity that developed in these animals. Increased hematopoietic cell apoptosis is a clinical characteristic of myelofibrosis associated with myeloproliferation in patients [23
]. An increased rate of apoptosis could limit the ability of fetal liver hematopoietic cells expressing TEL-Syk to develop secondary genetic changes which would allow establishment of more long lived myeloproliferation or even leukemia. Indeed, chronic myeloproliferative diseases such as CML are associated with reduced rates of apoptosis often through inhibition of stress responses [5
Analysis of serum cytokines demonstrated an elevation of proinflammatory cytokines such as MCP-1, IL-13, MIP-1α, IL-6, IP-10, MIG, and TCA in mice receiving TEL-Syk expressing fetal liver hematopoietic cells. Though we found no direct evidence of inflammatory induced tissue damage (such as pneumonitis or glomerulonephritis as is seen in other mouse models of inflammatory disease [25
]), in combination with the anemia and thrombocytopenia, the proinflammatory nature of the MDS in the TEL-Syk chimeras may contribute to their poor survival. Elevated circulating levels of proinflammatory cytokines have been observed in a number of MPNs in humans. Patients with primary myelofibrosis, with or without the presence of JAK2V617F
, develop a proinflammatory cytokine signature that contains IL-6, MCP-1, MIG, MIP-1α, TNF-α and IP-10 [26
]. The pathologic role of these cytokines is undetermined but their increase correlates with disease prognosis, in particular elevation of IL-6, IL-2R, IL-1RA, MIP-1α, MIG, IL-8, IL-12, IP-10 correlate with shortened primary myelofibrosis survival. Furthermore, numerous proinflammatory cytokines were found in the plasma of PV patients, in which IL-1β, -4, -5, -7, -10, -17, EGF, IFNα, TNF-α, GM-CSF, MIP-1α, MIP-1β, and MCP-1 correlated with reduced survival [27
]. Finally, patients with MDS also have elevated cytokines, increases in IP-10, IL-7, and IL-6 being poor prognostic factors for survival [24
]. MCP-1, MIG, G-CSF, TNF-α, IL-13, -8, -15, IFN-γ, and HGF were also significantly increased. The authors of these studies suggest that the initial myeloproliferation and the presence of inflammatory circulating cytokines are due to an abnormal bone marrow microenvironment propagated by pathogenic myeloid cells.
Induction of myeloproliferation with myelofibrosis has also been linked to expression of other tyrosine kinases fused to TEL. In particular, expression of a TEL-Lyn fusion protein, originally isolated from a patient with idiopathic myelofibrosis [29
], in mouse fetal liver hematopoietic progenitors also leads to an aggressive MPD with excessive bone marrow fibrosis, culminating in lethality by 60-90 days following cell transfer [30
]. As we observed with TEL-Syk, the kinase inactive version of TEL-Lyn fails to induce hematopoietic progenitor proliferation, indicating the requirement for kinase activity in both models. Similarly, activation of STAT5 is observed in progenitors transduced with TEL-Lyn. TEL fusion proteins with other tyrosine kinases, such as ABL1, ABL2, JAK2, NTRK3, FFGR3, and PDGFRB, have also been associated with various hematologic malignancies [6
], though direct comparisons between these fusion proteins in mouse models is not complete. Recent evidence demonstrated that BCR-ABL circumvents JAK2 and drives STAT5 signaling independently of cytokines in the context of CML [31
]. Hantschel et al. observed that BCR-ABL induces STAT5 phosphorylation and CML in JAK2-deficient hematopoietic stem cells. Administration of various JAK2 inhibitors failed to block phosphorylation of STAT5 in Ba/F3 cells transduced with BCR-ABL. These results are consistent with our in vitro
data demonstrating that TEL-Syk drives phosphorylation of STAT5 and colony formation in fetal liver cells even in the presence of JAK inhibitors. Mechanistically, it is evident that these fusions (BCR-ABL, TEL-Lyn, and TEL-Syk) uncouple the JAK2-STAT5 pathway to drive disease progression.
Overall, these results demonstrate that expression of the TEL-Syk fusion protein in fetal liver hematopoietic cells leads to rapid myeloproliferation and fatal myelofibrosis in a mouse model. The extent and the aggressive nature of this disease is unusual and will serve as a useful model for studying deregulated signaling in the context of myeloproliferative neoplasms with myelofibrosis. Future work with this model may allow identification of the soluble factor(s) derived from the proliferating myeloid cells that contribute to the extensive stromal fibrosis. Though many investigators have implicated myeloid cell-derived factors in myelofibrosis none have been defined.