Srsf2P95H mutant reduces polycythemia phenotype in Jak2V617F knock-in mice

In order to investigate the consequences of concurrent SRSF2P95H and JAK2V617F mutations in MPN, we generated Cre-inducible Srsf2P95H/+Jak2V617F/+ knock-in mice by crossing conditional Jak2V617F knock-in mice [2] with conditional Srsf2P95H knock-in [16] and Mx1Cre transgenic [18] mice. The expression of Srsf2P95H/+ and Jak2V617F/+ mutants were induced in the hematopoietic compartments of these mice at 4 weeks after birth following intraperitoneal injection of polyinosine-polycytosine (pI-pC). We analyzed four groups of mice: control (WT or Mx1Cre only), Mx1Cre; Srsf2P95H/+ (hereafter Srsf2P95H/+), Mx1Cre; Jak2V617F/+ (hereafter Jak2V617F/+) and Mx1Cre; Srsf2P95H/+Jak2V617F/+ (hereafter Srsf2P95H/+Jak2V617F/+) mice. Mice were analyzed at 24 weeks after pI-pC induction (i.e., at 28 weeks after birth). Consistent with our previous report [2], mice expressing heterozygous Jak2V617F (Jak2V617F/+) showed a PV disease characterized by significant increase in white blood cell (WBC), neutrophil (NE), platelet (PLT), red blood cell (RBC), hemoglobin (Hb) and hematocrit (HCT) levels in their peripheral blood compared to control mice (Fig. 1A–F). Mice expressing heterozygous Srsf2P95H (Srsf2P95H/+) displayed decreased hemoglobin but increased mean corpuscular volume (MCV) relative to control mice (Fig. 1E, G), consistent with published report [16]. Srsf2P95H/+Jak2V617F/+ mice expressing both Srsf2P95H and Jak2V617F mutants exhibited significantly reduced WBC, neutrophil, platelet, RBC, hemoglobin and hematocrit parameters compared to Jak2V617F/+ mice (Fig. 1A–F). While Jak2V617F/+ mice displayed significantly reduced MCV, Srsf2P95H/+Jak2V617F/+ mice had higher MCV values compared to Jak2V617F/+ mice (Fig. 1G). Jak2V617F/+ mice exhibited marked splenomegaly, whereas Srsf2P95H/+Jak2V617F/+ mice had significantly reduced spleen size/weight compared to Jak2V617F/+ mice (Fig. 1H). H&E staining of the BM sections from WT (control) and Srsf2P95H/+ mutant mice showed normal BM cellularity (Fig. 1I). Jak2V617F/+ mice BM sections exhibited hypercellularity with significant increase in erythroid precursors and megakaryocyte clusters (Fig. 1I). Srsf2P95H/+Jak2V617F/+ mice BM sections exhibited normal BM cellularity and a reduction of erythroid precursors and megakaryocyte clusters compared to Jak2V617F/+ mice BM (Fig. 1I). Reticulin staining of the BM sections from Jak2V617F/+ mice showed mild to no bone marrow fibrosis at 24 weeks after induction (Fig. 1I). Srsf2P95H/+Jak2V617F/+ mice also did not exhibit bone marrow fibrosis at this time (Fig. 1I). A few Srsf2P95H/+Jak2V617F/+ mice were monitored for longer period and they were assessed for bone fibrosis at one year after induction. We also did not observe bone marrow fibrosis in Srsf2P95H/+Jak2V617F/+ mice at one year after induction (data not shown). Together, these results suggest that co-expression of Srsf2P95H reduces polycythemia but does not promote myelofibrosis in Jak2V617F/+ mice.

Fig. 1: SRSF2P95H mutant reduces polycythemia in Jak2V617F knock-in mice.

A White blood cell (WBC), B neutrophil (NE), C platelet (PLT), D red blood cell (RBC), E hemoglobin (Hb), F hematocrit (HCT) and G mean corpuscular volume (MCV) counts in the peripheral blood of control (n = 11), Srsf2P95H/+ (n = 5), Jak2V617F/+ (n = 11) and Srsf2P95H/+Jak2V617F/+ (n = 11) mice were assessed at 24 weeks after pI-pC induction. H Spleen size/weight in control (n = 11), Srsf2P95H/+ (n = 6), Jak2V617F/+ (n = 8) and Srsf2P95H/+Jak2V617F/+ (n = 11) mice. I Bone marrow histology. Representative images of the H&E and Reticulin staining of the BM sections from control, Srsf2P95H/+, Jak2V617F/+ and Srsf2P95H/+Jak2V617F/+ mice (n = 5–6 per group) at 24 weeks after pI-pC induction. Scale bar, 20 μm. Data are presented in bar graphs as mean ± SEM. (*P < 0.05; ***P < 0.001; ****P < 0.0001; ns not significant). Statistical significances were determined using one-way ANOVA with Tukey’s multiple comparison test.

Effects of concurrent Srsf2P95H and Jak2V617F mutations on hematopoietic stem/progenitors and precursor cells in mice

We next assessed the effects of concurrent expression of Srsf2P95H and Jak2V617F mutants on mice hematopoietic stem/progenitor cells (HSPC) by flow cytometry. The representative flow cytometry plots of HSPC analysis are depicted in Fig. 2A. Jak2V617F/+ mice exhibited significant increase in frequencies and total numbers of LSK (Lin-Sca-1+c-kit+), LT-HSC (long-term hematopoietic stem cells), ST-HSC (short-term hematopoietic stem cells) and MPP (multipotent progenitors) in their BM, while concurrent expression of Srsf2P95H and Jak2V617F significantly reduced the frequencies and total numbers of LSK, LT-HSC, ST-HSC and MPP populations in the BM of Srsf2P95H/+Jak2V617F/+ mice (Fig. 2B–E and Supplementary Fig. 1A–D). Co-expression of Srsf2P95H and Jak2V617F mutants also significantly reduced the frequencies and total numbers of LK (Lin-c-kit+; myeloid progenitors), CMP (common myeloid progenitors), GMP (granulocyte-macrophage progenitors) and MEP (megakaryocyte-erythroid progenitors) in the BM of Srsf2P95H/+Jak2V617F/+ mice compared to Jak2V617F/+ mice (Fig. 2F–I and Supplementary Fig. 1E–H). However, we did not observe significant changes of HSPC in the spleens of Srsf2P95H/+Jak2V617F/+ mice compared to Jak2V617F/+ mice (Supplementary Fig. 2A–H).

Fig. 2: Effects of concurrent SRSF2P95H and Jak2V617F mutations on hematopoietic stem/progenitors and precursor cells.

A Representative plots of flow cytometric analysis of control, Srsf2P95H/+, Jak2V617F/+ and Srsf2P95H/+Jak2V617F/+ mice are shown. Percentages of B LSK (Lin−Sca-1+c-kit+), C LT-HSC (Lin-Sca-1+c-kit+CD34−CD135−), D ST-HSC (Lin−Sca-1+c-kit+CD34+CD135−), E MPP (Lin-Sca-1+c-kit+CD34+CD135+), F LK (Lin-Sca-1-c-kit+), G CMP (Lin-Sca-1-c- kit+CD34+CD16/32Low), H GMP (Lin-Sca-1-c-kit+CD34+CD16/32High) and I MEP (Lin−Sca-1−c-kit+CD34−CD16/32−) in the BM of control (n = 11), Srsf2P95H/+ (n = 6), Jak2V617F/+ (n = 10) and Srsf2P95H/+Jak2V617F/+ (n = 11) mice are shown in bar graphs as mean ± SEM. J Flow cytometric analysis of erythroid precursors using surface marker Ter119 and CD71 in the BM of control (n = 11), Srsf2P95H/+ (n = 6), Jak2V617F/+ (n = 10) and Srsf2P95H/+Jak2V617F/+ (n = 11) mice are shown in bar graphs as mean ± SEM. K Percentages of CD41+ megakaryocytic precursors in the BM of control (n = 11), Srsf2P95H/+ (n = 5), Jak2V617F/+ (n = 9) and Srsf2P95H/+Jak2V617F/+ (n = 10) mice are shown in bar graphs as mean ± SEM. L, M Hematopoietic progenitor colony assays. In total, 2 × 104 BM cells from control (n = 9), Srsf2P95H/+ (n = 4), Jak2V617F/+ (n = 9) and Srsf2P95H/+Jak2V617F/+ (n = 10) mice were plated in methylcellulose medium supplemented with cytokines. CFU-GM (L) and BFU-E (M) colonies were scored 7 days after plating. N Erythropoietin-independent CFU-E colony formation assay. In total, 1 × 105 spleen cells from control (n = 6), Srsf2P95H/+ (n = 6), Jak2V617F/+ (n = 10) and Srsf2P95H/+Jak2V617F/+ (n = 8) mice were plated in methylcellulose medium without any cytokines. CFU-E colonies were scored after 2 days. O CFU-Mk colonies derived from the BM of control (n = 4), Srsf2P95H/+ (n = 4), Jak2V617F/+ (n = 4) and Srsf2P95H/+Jak2V617F/+ (n = 4) mice. P Serum TGF-β1 levels in control (n = 9), Srsf2P95H/+ (n = 9), Jak2V617F/+ (n = 9) and Srsf2P95H/+Jak2V617F/+ (n = 9) mice were assessed by ELISA. (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant). Statistical significances were determined using one-way ANOVA with Tukey’s multiple comparison test.

While Jak2V617F/+ mice BM exhibited a significant increase of erythroid precursors (Ter119+/CD71+) compared to control mice, Srsf2P95H/+Jak2V617F/+ mice BM showed significantly reduced erythroid precursors (Ter119+/CD71+) compared to Jak2V617F/+ mice (Fig. 2J). The Srsf2P95H/+Jak2V617F/+ mice also had reduced percentage of megakaryocytic (CD41+) cells in their BM compared to Jak2V617F/+ mice (Fig. 2K). Hematopoietic progenitor colony assays showed significantly increased number of CFU-GM and BFU-E colonies derived from the BM of Jak2V617F/+ mice compared to control animals, while the number of CFU-GM and BFU-E colonies derived from the BM of Srsf2P95H/+Jak2V617F/+ mice were significantly lower compared to Jak2V617F/+ mice (Fig. 2L, M). Spleens from Jak2V617F/+ mice exhibited large numbers of Epo-independent CFU-E colonies (Fig. 2N), a hallmark feature of PV [19], whereas spleens from Srsf2P95H/+Jak2V617F/+ mice had significantly reduced Epo-independent CFU-E colonies compared to Jak2V617F/+ mice (Fig. 2N). The number of CFU-Mk colonies derived from the BM of Jak2V617/+ mice was significantly higher compared to control animals (Fig. 2O). Srsf2P95H/+Jak2V617F/+ mice BM exhibited significantly reduced number of CFU-Mk colonies compared to Jak2V617F/+ mice (Fig. 2O). Aberrant expression of transforming growth factor beta 1 (TGF-β1) has been linked to MF [1, 20]. So, we assessed the TGF-β1 levels by ELISA. Whereas Jak2V617F/+ mice exhibited elevated levels of serum TGF-β1, Srsf2P95H/+Jak2V617F/+ mice showed significantly reduced serum TGF-β1 levels compared to Jak2V617F/+ mice (Fig. 2P).

Phenotypes observed in Srsf2P95H/+Jak2V617F/+ mice are cell autonomous

To assess whether the phenotypes observed in Srsf2P95H/+Jak2V617F/+ mice were cell intrinsic, we transplanted BM cells from control, Srsf2P95H/+, Jak2V617F/+ and Srsf2P95H/+Jak2V617F/+ mice into lethally irradiated C57BL/6 wild type recipient mice as outlined in Fig. 3A. At 4 weeks after transplantation, mice were injected with pI-pC to induce expression of Srsf2P95H/+ and Jak2V617F/+ in hematopoietic compartments. Transplanted animals expressing Jak2V617F/+ exhibited elevated neutrophil (NE), red blood cell (RBC), hemoglobin (Hb) and hematocrit (HCT) levels but decreased MCV in the peripheral blood compared to control animals (Fig. 3B–F). Co-expression of Srsf2P95H/+ and Jak2V617F/+ mutants significantly reduced neutrophil, RBC, hemoglobin and hematocrit levels but increased MCV in the recipient animals compared to mice expressing Jak2V617F/+ (Fig. 3B–F). Transplanted animals expressing Jak2V617F/+ showed marked splenomegaly, whereas mice co-expressing Srsf2P95H/+ and Jak2V617F/+ mutants exhibited significantly reduced spleen weights compared to Jak2V617F/+ mice (Fig. 3G). Flow cytometry analyses showed decreased percentages of LSK, LT-HSC, ST-HSC and MPP in the BM of transplanted mice co-expressing Srsf2P95H/+ and Jak2V617F/+ compared to mice expressing Jak2V617F/+ (Fig. 3H–L). The percentages of LSK, LT-HSC and MPP were also significantly reduced in the spleens of transplanted mice co-expressing Srsf2P95H/+ and Jak2V617F/+ compared to mice expressing Jak2V617F/+ (Supplementary Fig. 3A–D). However, the percentages of LK, CMP, GMP and MEP were not significantly altered in the BM of transplanted mice co-expressing Srsf2P95H/+ and Jak2V617F/+ compared to mice expressing Jak2V617F/+ (Fig. 3M–P).

Fig. 3: Phenotypes observed in the SRSF2P95H/+Jak2V617F/+ mice are cell autonomous.

A Experimental design for cell autonomous bone marrow transplantation (BMT) assay. BM cells from control, Srsf2P95H/+, Jak2V617F/+ and Srsf2P95H/+Jak2V617F/+ mice at 8 weeks after birth (without pI-pC) were transplanted into lethally irradiated wild type C57BL/6 recipient mice (1 × 106 cells/recipient). At 4 weeks after BMT, pI-pC injections were given to induce the expression of Srsf2P95H and Jak2V617F mutants in the recipient animals. Recipient mice were analyzed at 36 weeks after pI-pC induction. Peripheral blood B neutrophil (NE), C red blood cell (RBC), D hemoglobin (Hb), E hematocrit (HCT) and F mean corpuscular volume (MCV) counts of control (n = 10), Srsf2P95H/+ (n = 7), Jak2V617F+ (n = 6) and Srsf2P95H/+Jak2V617F/+ (n = 8) mice are shown in bar graphs. G Spleen weights of control (n = 4), Srsf2P95H/+ (n = 4), Jak2V617F/+ (n = 5) and Srsf2P95H/+Jak2V617F/+ mice (n = 4). H Representative plots of flow cytometric analysis of control, Srsf2P95H/+, Jak2V617F/+ and Srsf2P95H/+Jak2V617F/+ BMT mice. Frequencies of I LSK, J LT-HSC, K ST-HSC, L MPP, M LK, N CMP, O GMP and P MEP in the BM of control (n = 8), Srsf2P95H/+ (n = 4), Jak2V617F/+ (n = 7) and Srsf2P95H/+Jak2V617F/+ mice (n = 5) are shown in bar graphs as mean ± SEM. (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant). Statistical significances were determined using one-way ANOVA with Tukey’s multiple comparison test.

Expression of Srsf2P95H mutant reduces the competitiveness of Jak2V617F HSPC

To evaluate the effects of concurrent expression of Srsf2P95H and Jak2V617F mutants on HSPC function, we performed competitive repopulation assays (outlined in Fig. 4A). We generated Mx1Cre; Jak2V617F/+ GFP+ and Mx1Cre; Srsf2P95H/+Jak2V617F/+ GFP+ mice. Equal numbers of BM cells from these donor mice (5 × 105) were mixed with WT (non-GFP) mice BM cells (5 × 105) at a ratio of 1:1 and then transplanted into lethally irradiated WT C57BL/6 mice. At 4 weeks after BMT, the recipient animals were injected with pI-pC to induce Srsf2P95H/+ and Jak2V617F/+ expression. The percentages of donor-derived mutant (GFP + ) cells were determined in the peripheral blood leukocytes of the chimeric mice by flow cytometry every 4 weeks and the mice were analyzed at 12 weeks after pI-pC induction (i.e., 16 weeks after BMT). We observed significantly higher percentages of GFP+ granulocyte (Gr-1+), erythroid (Ter119+), megakaryocyte (CD41+), B-lymphocyte (B220+) and T-lymphocyte (TCRβ+) cells in the peripheral blood of chimeric mice receiving Jak2V617F/+ BM compared with chimeric mice receiving Srsf2P95H/+Jak2V617F/+ BM (Fig. 4B–F). We also observed significantly reduced percentages of GFP+ Gr-1+, Ter119+, CD41+, B220+ and TCRβ+ cells in the BM of chimeric recipient animals receiving Srsf2P95H/+Jak2V617F/+ BM compared with Jak2V617F/+ BM (Fig. 4G–K). Similarly, we observed significantly reduced percentages of GFP+ Gr-1+, Ter119+, CD41+, B220+ and TCRβ+ cells in the spleens of chimeric animals receiving Srsf2P95H/+Jak2V617F/+ BM compared with Jak2V617F/+ BM (Supplementary Fig. 4A–E). Whereas the majority (70–80%) of LSK and LK cells in the BM and spleens of chimeric animals receiving Jak2V617F/+ BM were GFP+ at 12 weeks after pI-pC induction (16 weeks after transplantation), the percentages of GFP + LSK and LK cells in the BM and spleens were significantly lower in chimeric animals receiving Srsf2P95H/+Jak2V617F/+ BM compared with Jak2V617F/+ BM (Fig. 4L, M and Supplementary Fig. 4F, G). These data suggest that co-expression of Srsf2P95H mutant reduces the hematopoietic progenitor function and diminishes the clonal advantage of Jak2V617F mutant HSPC.

Fig. 4: Srsf2P95H mutant reduces the competitiveness of Jak2V617F HSPC.

A A scheme on competitive BM transplantation assay. BM cells (5 × 105) from Jak2V617F/+GFP+ or Srsf2P95H/+Jak2V617F/+GFP+ mice without pI-pC injection were mixed with non-GFP WT BM (5 × 105) cells at a 1:1 ratio and transplanted into lethally irradiated non-GFP WT C57BL/6 recipient mice. pI-pC injections were given to the recipients at 4 weeks after BMT to induce Srsf2P95H and Jak2V617F expression. The recipient mice were analyzed at 12 weeks after pI-pC injections. Percentages of donor derived (GFP + ) B Gr-1+, C Ter119+, D CD41+, E B220+ and F TCRβ+ cells in the peripheral blood of recipients at 4, 8 and 12 weeks after pI-pC injections are shown in bar graphs as mean ± SEM. Percentages of GFP+ G Gr-1+, H Ter119+, I CD41+, J B220+, K TCRβ+, L LSK and M LK cells in the BM of recipient mice are shown in bar graphs as mean ± SEM (Jak2V617F/+GFP: WT = 1:1, n = 6–8; Srsf2P95H/+Jak2V617F/+GFP: WT = 1:1, n = 5). (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001). Statistical significances were determined using two-tailed unpaired t test.

Srsf2P95H mutant-induced overexpression of S100A9 contributes to decreased erythropoiesis in Jak2V617F/+ mice

A previous study has shown increased expression of S100a8 and S100a9 mRNA in hematopoietic progenitors of Srsf2P95H/+ mice [21]. Furthermore, increased expression of S100A8 and S100A9 has been linked to erythroid differentiation defects and MDS pathogenesis [22, 23]. Since we observed decreased erythrocytosis in Srsf2P95H/+Jak2V617F/+ mice, we assessed the expression of S100A8 and S100A9 in MEP (megakaryocyte-erythroid progenitors) by RT-qPCR. We found significantly increased expression of S100a8 and S100a9 mRNA in Srsf2P95H/+Jak2V617F/+ mice MEP compared with Jak2V617F/+ mice MEP (Fig. 5A). We also observed significantly increased expression of S100a9 mRNA in Srsf2P95H/+ mice MEP compared with WT mice MEP (Fig. 5A). Immunoblot analyses also revealed increased expression of S100A8 and S100A9 proteins in the BM of Srsf2P95H/+ and Srsf2P95H/+Jak2V617F/+ mice compared with WT or Jak2V617F/+ mice BM (Fig. 5B). We further performed functional validation by retroviral overexpression of S100A8 and S100A9 into Jak2V617F/+ mice BM and progenitor colony assays. We observed significantly reduced CFU-GM and BFU-E colony formation by overexpression of S100A8 and S100A9 in the BM of Jak2V617F/+ mice (Supplementary Fig. 5A, B).

Fig. 5: S100A9 overexpression induced by Srsf2P95H contributes to decreased erythrocytosis in Jak2V617F/+ mice.

A mRNA expression of S100a8 and S100a9 was determined in sorted MEP from the BM of control, Srsf2P95H/+, Jak2V617F/+ and Srsf2P95H/+Jak2V617F/+ mice by RT-qPCR. Relative expression levels were normalized to housekeeping gene Hprt (n = 3). B Immunoblots showing increased expression of S100A8 and S100A9 proteins in Srsf2P95H/+Jak2V617F/+ BM compared with WT (control) and Jak2V617F/+ BM. β-actin was used as a loading control. C A scheme on the experimental design is depicted. Jak2V617F/+ BM cells were transduced with retroviruses expressing vector or S100A9 and transplanted into lethally irradiated C57BL/6 mice. The recipient mice were analyzed at 24 weeks after BMT. D Immunoblot analysis of S100A9 expression in the Jak2V617F/+ BM transplanted animals expressing vector or S100A9. Erk2 was used as a loading control. E Peripheral blood RBC, Hb, HCT, WBC, NE and PLT counts of transplanted mice receiving Jak2V617F/+ BM expressing vector (n = 5) or S100A9 (n = 5) were measured at 24 weeks after BMT. F Representative plots of flow cytometric analysis of erythroid precursors in the BM and spleens of recipient mice expressing vector or S100A9 using surface markers CD71 and Ter119. G Percentages of erythroid precursor cells at different stages of differentiation (stage I–IV, from immature to mature) are shown in bar graphs as mean ± SEM. (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns not significant). Significance was determined in (A) using one-way ANOVA with Tukey’s multiple comparison test. Statistical significances in (E, G) were determined using two-tailed unpaired t test.

To assess the in vivo effects of S100A9 overexpression in Jak2V617F/+ mice, we performed bone marrow transplantation assays following retroviral expression of empty vector or S100A9 into Jak2V617F/+ BM cells (outlined in Fig. 5C). Immunoblot analysis confirmed increased S100A9 protein levels in the BM of Jak2V617F/+ mice expressing S100A9 (Fig. 5D). Transplanted animals receiving Jak2V617F/+ BM overexpressing S100A9 exhibited significantly reduced RBC, hemoglobin and hematocrit counts compared to recipients of Jak2V617F/+ BM expressing vector at 24 weeks after transplantation (Fig. 5E). However, WBC, neutrophil and platelet counts were not significantly altered by S100A9 overexpression (Fig. 5E). Flow cytometric analysis showed that recipients of Jak2V617F/+ BM expressing S100A9 had reduced erythroid precursors (stage II, CD71highTer119high) in their BM and spleens compared to recipient animals expressing vector (Fig. 5F, G). However, we did not observe bone marrow fibrosis in transplanted animals receiving S100A9 transduced Jak2V617F/+ mice BM (data not shown). Together, these results suggest that Srsf2P95H mutant induces overexpression of S100A9 (and S100A8) and contributes to reduced erythropoiesis in Jak2V617F/+ mice.