Critical Role of SHP2 in the Initiation and Maintenance of Myeloproliferative Neoplasms Evoked By JAK2V617F

The JAK2V617F mutation has been found in a majority of patients with Philadelphia chromosome-negative myeloproliferative neoplasms (MPNs) including polycythemia vera (PV), essential thrombocythemia (ET) and primary myelofibrosis (PMF). Expression of JAK2V617F results in constitutive activation of the JAK2 tyrosine kinase and its downstream signaling pathways. JAK2 inhibitor therapy can reduce splenomegaly and constitutional symptoms but is not sufficient to cure or produce complete remission in patients with MPNs. Identification of critical signaling pathways/molecules required for the initiation and maintenance of MPNs will facilitate the development of more effective targeted therapies. To determine the role of JAK2V617F mutation in MPNs, we previously generated a conditional Jak2V617F knock-in mouse. We observed that expression of heterozygous Jak2V617F in mice hematopoietic compartments is sufficient to induce a PV-like MPN, whereas homozygous Jak2V617F expression accelerates the development of myelofibrosis. We have found that SHP2, a protein tyrosine phosphatase that positively regulates hematopoietic signaling, is constitutively phosphorylated in mouse and human hematopoietic cells expressing JAK2V617F. However, the contribution of SHP2 in the pathogenesis of MPNs induced by JAK2V617F remains elusive. Here, we sought to determine the role of SHP2 in JAK2V617F-evoked MPNs using conditional SHP2 knockout (SHP2 floxed) and Jak2V617F knock-in mice and MxCre line. Whereas expression of heterozygous Jak2V617F induced a PV-like MPN characterized by increase in red blood cells (RBC), white blood cells (WBC), neutrophils and platelets in the peripheral blood and splenomegaly, deletion of SHP2 normalized the blood parameters and spleen size in Jak2V617F knock-in mice. Flow cytometric analysis showed that deletion of SHP2 inhibited the expansion of erythroid, megakaryocytic, and granulocytic precursors in the bone marrow (BM) and spleens of Jak2V617F mice. Deletion of SHP2 also inhibited the expansion of hematopoietic stem cells and megakaryocyte-erythroid progenitors in the BM and spleens of Jak2V617F knock-in mice. Furthermore, deletion of SHP2 markedly inhibited the erythropoietin (Epo)-independent CFU-E colonies in the BM and spleens of Jak2V617F mice. In order to determine whether the effects of SHP2 deletion in Jak2V617F mice were cell autonomous, BM cells from uninduced control, MxCre;V617F/+ and MxCre;V617F/+;SHP2fl/fl mice were transplanted into lethally irradiated syngeneic recipient mice. Four weeks after transplantation, mice were injected with pI-pC to induce deletion of SHP2 and expression of Jak2V617F simultaneously. Transplanted animals receiving MxCre;V617F/+ BM developed a PV-like disease within 6 weeks after pI-pC induction. However, recipients of MxCre;V617F/+;SHP2fl/fl BM failed to induce the PV disease due to deletion of SHP2. Together, these data suggest that SHP2 is required for the initiation of PV-like MPN mediated by Jak2V617F. To determine if SHP2 is required for the maintenance of MPN evoked by Jak2V617F, uninduced wild type or MxCre;SHP2fl/fl mice BM cells were transduced with retroviruses expressing Jak2V617F and transplanted into lethally irradiated syngeneic recipient animals. Five weeks after transplantation, we assessed the peripheral blood counts to confirm the development of MPN disease. Transplanted animals receiving Jak2V617F-transduced wild type or MxCre;SHP2fl/fl mice BM were then divided into two groups- one group was injected with pI-pC (to induce deletion of SHP2 after establishment of the MPN disease); another group (control group) was injected with saline. We observed that deletion of SHP2 by pI-pC induction significantly inhibited the RBC, hematocrit, WBC and platelet counts in the peripheral blood and reduced the spleen size in transplanted animals expressing Jak2V617F. Moreover, deletion of SHP2 markedly inhibited the Epo-independent erythroid colonies in the BM and spleens of transplanted animals expressing Jak2V617F. Thus, SHP2 is not only required for the initiation of MPN but also required for the maintenance of MPN mediated by Jak2V617F. We have also observed that deletion of SHP2 blocks the development of myelofibrosis in Jak2V617F knock-in mice. In conclusion, our results suggest that SHP2 plays a critical role in the initiation and maintenance of MPNs evoked by Jak2V617F. Disclosures No relevant conflicts of interest to declare.

The capacity of mitogen-responsive T cells to home to their tissue of origin, analyzed by means of chromosome markers

Lance and Taub (1) showed that when radioactively labeled lymphocytes were injected into a syngeneic mouse and the lymph node cells of this animal transferred to a second syngeneic recipient, the proportion of radioactivity found in the lymph node relative to the amount present in the spleen of the secondary recipient had increased markedly. The interpretation of this result was that some lymphocytes have the capacity to home to their organ of origin. The purpose of the experiments described here was to test the homing copacity of T cells by a method that did not involve radioactive labeling. It has been shown elsewhere that some or all mouse T cells are stimulated to divide in culture by the mitogens phytohemagglutinin (PHA) and concanavalin A (Con A) (2). We therefore elected to inject karyotypically distinct lymphocytes into syngeneic recipients and to follow their subsequent distribution by culture of lymph node and spleen cells of the recipient with PHA or Con A. In this manner the homing capacities of spleen and lymph node T cells could be determined, and furthermore, the effects of labeling with chromium-51 ((51)Cr) could be assayed with respect to the persistence of mitogen responsiveness in the injected cells.

TUMOR INDUCTION BY IMMUNOLOGICALLY ACTIVATED MURINE LEUKEMIA VIRUS

A graft-vs.-host reaction (GVHR) was induced in young male CAFI and CB6F1 mice by the administration of BALB/cJ spleen cells. A proportion of such mice subsequently developed lymphoreticular rumors. Cell-free extracts (CFEs) prepared from the reticular tissues of CAF1 mice killed at intervals after the induction of the GVHR were tested for their capacity to produce the same tumors in a litter of syngeneic mice inoculated at birth. 12 of 29 (41.4%) such extracts were positive, causing lymphoreticular tumors in one or more littermate recipients. The positive CFEs came from donors killed at all stages of the GVHR, from tumor-bearing mice as well as from non-tumor-bearing mice. However, whereas less than 30% of CFEs from mice killed within 12 mo of GVHR induction were oncogenic, the incidence of oncogenic extracts from mice killed 12–15 mo after GVHR induction rose to 75%. None of the CFEs prepared from nine normal uninjected male CAF1 mice killed between the ages of 8 and 18 mo transmitted tumors to recipients. CFEs prepared from CAF1 mice with the GVHR were tested for infectious murine leukemia virus (MuLV) using the XC assay and also for complement-fixing (CF) group-specific MuLV antigen. Substantial titers of B-tropic MuLV and CF antigen were detected in at least half the extracts from mice killed 11–14 mo after GVHR induction. During the first few months of GVHR induction, MuLV titers were low and CF antigen was absent. Neither infectious MuLV nor CF antigen were detected in CFEs prepared from normal control mice. Serially passed CFEs originating from a CB6F1 GVHR-induced RCN caused similar tumors in successive generations of syngeneic recipient mice. These lymphoreticular tumors were shown to contain infectious MuLV, CF MuLV antigen, and C-type particles. These data together provide evidence that MuLV is activated during the GVHR and that it is responsible for the eventual development of lymphoreticular tumors.

CELLULAR BASIS OF THE GENETIC CONTROL OF IMMUNE RESPONSES TO SYNTHETIC POLYPEPTIDES

DBA/1 mice are high responders to the (Phe, G) determinant of the synthetic polypeptide (Phe, G)-Pro--L, whereas SJL mice respond well to the Pro--L region of this macromolecule (6). In order to determine whether the phenomenon described above is related to the number of antigen-sensitive units detected for both specificities, and whether responses to these determinants can be transferred independently, graded and limiting inocula of spleen cells from SJL, DBA/1, and F1 donors were injected into X-irradiated, syngeneic, recipient mice with (Phe, G)-Pro--L. By this approach, one antigen-sensitive unit specific for (Phe, G) was detected in 1.7 x 106 and 8.5 x 106 spleen cells from immunized and nonimmunized DBA/1 donors, respectively. In contrast, one (Phe, G) relevant precursor was detected in 20 x 106 SJL spleen cells, irrespective of whether the donors had been immunized. On the other hand, for the Pro--L specificity, one limiting splenic precursor was found in 1.3 x 106 and in 3.4 x 106 cells for immunized and nonimmunized SJL donors, respectively; whereas one response unit was estimated for this determinant in 9.4 x 106 and in 38 x 106 spleen cells from immunized and nonimmunized DBA/1 mice. The findings reported here indicate that the phenotypic expression of the genetic control(s) for immune responsiveness to different immunopotent regions of (Phe, G)-Pro--L is directly correlated with the number of immunocompetent response units detected in two inbred mouse strains. In the spleens of immunized F1 donors, similar frequencies of one limiting precursor in 3.0 x 106 and in 2.8 x 106 cells were detected for (Phe, G) and Pro--L, respectively. The results of a chi-square test for independence of (Phe, G) and Pro--L responses in F1 animals is compatible with the hypothesis that the transferred spleen cells limiting the response to (Phe, G)-Pro--L are restricted to generate antibodies specific for only one of the two determinants of this macromolecule.

THE DISTRIBUTION OF LARGE DIVIDING LYMPH NODE CELLS IN SYNGENEIC RECIPIENT RATS AFTER INTRAVENOUS INJECTION

The distribution of large dividing lymph node or thoracic duct lymph cells, labeled in vitro with 3H-thymidine, was studied in syngeneic recipient rats after intravenous injection. In most experiments the donor rats had been immunized with Bacillus pertussis 4 days earlier, but in some instances cells from nonimmunized donors were used. In smears, the labeled donor cells had the appearance of large lymphocytes or large pyroninophilic cells. By electronmicroscopy, the majority of labeled donor cells were seen to have only scanty endoplasmic reticulum. It was found that the labeled cells rapidly "homed" to lymphoid tissue and recirculated in the recipient, in a fashion resembling that of small lymphocytes. However, the distribution of labeled cells was found to depend upon the source of the donor cells. Cells from mesenteric lymph nodes or thoracic duct lymph showed a marked preferential accumulation in lymphoid tissue within or adjacent to the intestine, whereas cells from peripheral nodes accumulated preferentially in peripheral lymph nodes. Cells from any of these sources showed an equal tendency to accumulate in the white pulp of the spleen. Suspensions of small lymphocytes, labeled in vitro with 3H-uridine, did not display a similar tendency to localize preferentially in lymphoid tissue in certain regions. It was also found that large dividing lymph node cells from donors immunized with an antigen (2,4-dinitrophenyl-bovine gamma globulin (DNP-BGG) or B. pertussis) showed a greater tendency to accumulate in a recipient lymph node containing that antigen than in the contralateral node. It was not determined whether the selective accumulation of large dividing lymphoid cells from different sources in lymphoid tissue of different regions in recipients was due to an antigen recognition mechansim or was the result of two different populations of cells with different "homing" mechanisms.