PHARMACEUTICAL COMBINATION FOR THE TREATMENT OF MYELOPROLIFERATIVE NEOPLASMS

20230158032 · 2023-05-25

    Inventors

    Cpc classification

    International classification

    Abstract

    The present invention relates to the treatment of myeloproliferative neoplasms through targeted elimination of malignant clones and overcome of diseases persistence. The treatment is based on a combination of inhibitors of mRNA splicing and processing factors together with Jak inhibitors.

    Claims

    1. A method for the treatment of a myeloproliferative neoplasm, comprising administering to a patient a combination comprising: at least one inhibitor of a mRNA splicing and processing factor, and at least one JAK inhibitor.

    2. The method according to claim 1, wherein the at least one inhibitor of a mRNA splicing and processing factor is selected from the group comprising a Cpsf7 inhibitor, a Cstf2 inhibitor, a Hnrnpk inhibitor, a Hnrnpu inhibitor, a Pcbp1 inhibitor, a Polr2a inhibitor, a Rbm39 inhibitor, a Rbm8a inhibitor, a Sf3b1 inhibitor, a Snrnp200 inhibitor, a Srrm1 inhibitor, a Srsf2 inhibitor, a Srsf6 inhibitor, a Srsf9 inhibitor, a Srsf11 inhibitor, and a Ybx1 inhibitor.

    3. The method according to claim 1, wherein the at least one inhibitor of a mRNA splicing and processing factor is a Ybx1 inhibitor.

    4. The method according to claim 1, wherein the at least one inhibitor of a mRNA splicing and processing factor is selected from the group comprising a mRNA expression inhibitor, a protein expression inhibitor, a post-translational modification inhibitor, a protein arginine methyl transferase inhibitor, and a functional inhibitor.

    5. The method according to claim 1, wherein the myeloproliferative neoplasm is selected from the group comprising polycythemia vera, primary myelofibrosis, idiopathic myelofibrosis, essential thrombocythemia, chronic myeloid leukemia, acute myeloid leukemia, chronic eosinophilic leukemia, hypereosinophilic syndrome, chronic myelomonocytic leukemia, atypical chronic myelogenous leukemia, chronic neutrophilic leukemia systemic mastocytosis, juvenile myelomonocytic leukemia, and myeloma, post-polycythemia vera myelofibrosis or post-essential thrombocythemia myelofibrosis.

    6. The method according to claim 1, wherein the myeloproliferative neoplasm is caused by and/or associated with a JAK2V617F mutation.

    7. The method according to claim 1, wherein the myeloproliferative neoplasm is persistent to JAK-inhibitor.

    8. The method according to claim 1, wherein the JAK inhibitor is selected from the group consisting of ruxolitinib, pacritinib, NS-018, CEP-33779, NVP-BVB808, TG101209, fedratinib, momelotinib, baricitinib, AZD960, AZD1480, tofacitinib, gandotinib, XL019, NVP-BSK805, peficitinib, pyridone 6, filgotinib, itacitinib, decernotinib, janex1, and JAK3-IN-1.

    9. A combination comprising: at least one inhibitor of a mRNA splicing and processing factor, and at least one JAK inhibitor.

    10. A pharmaceutical composition comprising at least one inhibitor of a mRNA splicing and processing factor, and at least one JAK inhibitor, together with at least one pharmaceutically acceptable vehicle, excipient and/or diluent.

    11. The pharmaceutical composition according to claim 10, wherein the at least one inhibitor of a mRNA splicing and processing factor is selected from the group comprising a Cpsf7 inhibitor, a Cstf2 inhibitor, a Hnrnpk inhibitor, a Hnrnpu inhibitor, a Pcbp1 inhibitor, a Polr2a inhibitor, a Rbm39 inhibitor, a Rbm8a inhibitor, a Sf3b1 inhibitor, a Snrnp200 inhibitor, a Srrm1 inhibitor, a Srsf2 inhibitor, a Srsf6 inhibitor, a Srsf9 inhibitor, a Srsf11 inhibitor, and a Ybx1 inhibitor.

    12. The pharmaceutical composition according to claim 10, wherein the at least one inhibitor of a mRNA splicing and processing factor is a Ybx1 inhibitor.

    13. The pharmaceutical composition according to claim 10, wherein the at least one inhibitor of a mRNA splicing and processing factor is selected from the group comprising a mRNA expression inhibitor, a protein expression inhibitor, a post-translational modification inhibitor, a protein arginine methyl transferase inhibitor, and a functional inhibitor.

    14. The pharmaceutical composition according to claim 10, wherein the JAK inhibitor is selected from the group consisting of ruxolitinib, pacritinib, NS-018, CEP-33779, NVP-BVB808, TG101209, fedratinib, momelotinib, baricitinib, AZD960, AZD1480, tofacitinib, gandotinib, XL019, NVP-BSK805, peficitinib, pyridone 6, filgotinib, itacitinib, decernotinib, janex1, and JAK3-IN-1.

    15. (canceled)

    Description

    DESCRIPTION OF THE FIGURES

    [0278] FIG. 1 shows on the top, the interplay between Ybx1, Mknk1, and ERK in JAK-inhibitor persistent cells with or without Ybx-1 inactivation. The workflow on the bottom depicts the combination therapy according to the present invention.

    [0279] FIG. 2 shows schematic of the phosphoproteome workflow. a) Following sample collection, phosphopeptides were enriched using EasyPhos work flow (Humphrey, et al., 2015. High-throughput phosphoproteomics reveals in vivo insulin signaling dynamics. Nat Biotechnol 33), and analyzed in single-run LC-MS/MS. Data were analyzed in Maxquant and Perseus. b) Summary of identified and quantified class-I phosphosites (localization probability of >0.75) corresponding to number of proteins of this experiment.

    [0280] FIG. 3 shows the results of the functional phosphoproteomics screen identifying mRNA splicing and processing factors downstream of Jak2V617F. Unsupervised hierarchical clustering of significantly (t-test with permutation based FDR<0.01) regulated phosphosites in Jak2WT and Jak2V617F cells. The numbers between square brackets indicate the range of Z-score. [0-2]: phosphorylation increase; [−2-0] phosphorylation decrease The panel highlight shows the top-upregulated phosphoproteins in WT (cluster-1, up) and VF (cluster-2, bottom) cells and the individual phosphosites with the amino acid position of significantly regulated Jak2-Stat signaling pathway members. Heatmap on the right shows enrichment of known Jak2 targets in both WT and VF cells.

    [0281] FIG. 4 shows sub-network map of significantly enriched Gene ontology (GO) terms (p-value <0.01) of differentially phosphorylated proteins in Jak2V617F. The colours of the subnetwork nodes range from bright to transparency based on their p-value. Highlighted nodes indicate the core components of the sub-network based on the p-value.

    [0282] FIG. 5 shows ranking of significantly phosphorylated proteins in Jak2V617F. Phosphorylated proteins participating in mRNA splicing and processing are highlighted in black with top 15 proteins displayed.

    [0283] FIG. 6 shows western blot validation of shRNA library targets, Pcbp1 protein (on the top) and Ybx1 protein (on the bottom) in murine Jak2-V617F cells. The experiment was performed three times with similar results.

    [0284] FIG. 7 a), b), c) show shRNA validation of selected top 15 targets essential for Jak2V617F cell survival and growth in the presence and absence of Jak-i (Rux, 0.5 μM) measured by MTS assay. The graphs in c) show changes in cell viability depending on knockdown of respective differentially phosphorylated and splicing relevant genes (light grey dots)—with (right) or without (left) JAK-inhibitor treatment. shRNAs against Ybx1 (medium grey) and non-targeting control (dark grey) show dependency on Ybx1 only upon Jak-inhibitor treatment (right panel, dropout to the lower left quadrant of the right panel) but not per se (without JAK-I treatment) (left panel). Data obtained from 4 independent experiments each with 8 technical replicates.

    [0285] FIG. 8 show in vitro viability of murine BaF3 Jak2V617 cells following lentiviral infection with shRNAs targeting the top 15 mRNA processing factors (selected in FIG. 5) or with non-targeting control (shNT) in the presence and absence of Jak-i (Rux, 0.5 μM). Data are mean±SD from 4 independent experiments each with 8 technical replicates, as measured by MTS proliferation assay.

    [0286] FIG. 9 a) shows quantification of Ybx1 positivity in bone marrow sections of Jak2V617F-positive patients using the multiplied M-score. Ybx1 positivity was analysed by immunohistochemistry in primary bone marrow biopsies of 93 myeloproliferative neoplasm patient samples compared to 23 healthy donor controls. Healthy donors (HD; n=18), BCR-ABL-positive chronic myelogeneous leukemia (CML; n=17). Jak2V617F positive biopsies: essential thrombocythemia (ET; n=32), polycythemia vera (PV; n=23) or primary myelofibrosis (PMF; n=21). b) Immunoprecipitation of Jak2 receptor from murine Jak2V617F cells showing binding of Ybx1 to mutated Jak2-receptor. The experiment was performed three times with similar results.

    [0287] FIG. 10 shows cell growth curve of Jak2V617F cells following lentiviral infection with shRNAs targeting Ybx1 (Ybx1_sh1, Ybx1_sh2) or non-targeting control (shNT) and treatment with increasing doses of Jak-inhibitor (1 nM-10 μM ruxolitinib) measured by MTS assay. Data obtained from 4 independent experiments each with 8 technical replicates.

    [0288] FIG. 11 shows functional consequences of Ybx1 depletion in Jak2-mutated cell lines. a) Percentage of cell growth in Jak2WT and in Jak2V617F cells following lentiviral infection with shRNAs targeting Ybx1 or non-targeting control (shNT) measured by MTS assay. Data obtained from 4 independent experiments each with 8 technical replicates. b) Representative histogram and bar plot showing ROS levels measured in Jak2V617F cells following treatment with shRNAs targeting Ybx1 or non-targeting control (shNT). Data obtained from 4 independent experiments. c) Bar plot showing proliferative marker PCNA levels measured by flow cytometry in Jak2V617F cells following treatment with shRNAs targeting of Ybx1 or non-targeting control (shNT). Data obtained from 4 independent experiments. d) Cell cycle analysis after Ybx1 inactivation in Jak2V617F cells following treatment with shRNAs targeting Ybx1 or non-targeting control (shNT).

    [0289] FIG. 12 shows that inactivation of Ybx1 sensitizes Jak2V617F positive cells to Jak-inhibitor induced apoptosis. Percentage of apoptotic cells (AnnexinV.sup.+/Sytox.sup.+) of Jak2V617F positive cells following lentiviral mediated knockdown of Ybx1 (shYbx1-1 and shYbx1-2) compared to non-targeting control (scrambled shRNA, shSCR). a) murine BaF3-Jak2V617F cells, b) human Jak2V617F positive cell lines HEL and SET-2, c) primary murine lineage-negative bone marrow cells (Jak2.sup.+/+ and Jak2.sup.+/VF).

    [0290] FIG. 13 a) shows peripheral blood counts of recipient mice, week 8 and 16. White blood count (WBC), hemoglobin concentration (HGB) and platelet count (PLT) in Jak2V617F-Ybx1.sup.−/− animals compared to Jak2V617F-Ybx1.sup.+/+ controls. b) peripheral blood chimerism of lethally irradiated (12Gy) recipient mice. Percentage of Jak2V617F-Ybx1+1+ or Jak2V617F-Ybx1−/−(CD45.2) cells/CD45.1-competitors, weeks 4 to 16. Each data point represents an individual mouse biological replicate. Statistical analysis by Mann-Whitney-U test unless otherwise specified. c) FACS plots showing percentage of apoptotic cells in Jak2+/+ and Jak2V617F/+ murine BM cells following Jak-i treatment (Rux 0.5 μM) after lentiviral knockdown of Ybx1 (sh1 and sh2) compared to non-targeting control (shNT).

    [0291] FIG. 14 shows percentage of animals with loss (dark grey) or persistence (grey) of Jak2V617F clones in Ybx1.sup.+/+ WT (upper panel) or Ybx1.sup.−/− KO recipients, respectively.

    [0292] FIG. 15 a) shows peripheral blood chimerism of lethally irradiated (12Gy) recipient mice. FACS plots showing abundance of CD45.2 myeloid cells in Jak2V617F-Ybx1.sup.+/+ and Jak2V617F-Ybx1.sup.−/− recipient mice at week 20 after bone marrow transfer (BMT). b) histology of liver, spleen and lung of Jak2V617F-Ybx1.sup.+/+ and Jak2V617F-Ybx1.sup.−/− recipient mice at week 20 after BMT. Hematoxylin and eosin stain (H & E) at 10× magnification. Focal leukocyte infiltration (arrows) and haemorrhage (stars) of liver, spleen and lung, respectively.

    [0293] FIG. 16 a) shows a protocol for a JAK2-dependent xenograft model. b), c) Kaplan-Meier survival curves of irradiated (2Gy, single dose) recipient NSGS mice transplanted with Jak2V617F positive HEL cells treated with the indicated shRNAs (shSCR: scrambled negative control; shYBX1: targeting Ybx1).

    [0294] FIG. 17 shows functional consequences of Ybx1 deletion in mouse hematopoietic stem- and progenitor cells. a) design for assessment of steady state haematopoiesis. b) white blood count (WBC), c) Gr-1 positive cells (Gr1.sup.+), hemoglobin (HGB) and e), platelets (PLT) following genetic inactivation of Ybx1 (Ybx1.sup.−/− mice) compared to control mice (Ybx1.sup.+/+).

    [0295] FIG. 18 shows further results from the mice treated as shown in FIG. 6. a) FACS plots showing comparable percentage of LSK-cells and HSCs (SLAM.sup.+CD34.sup.−L.sup.−S.sup.+K.sup.+ cells) following genetic inactivation of Ybx1 in conditional knockout mice (compared to wildtype littermate controls). b) total stem and progenitor cell numbers per 1×10.sup.6 whole bone marrow cells at week 16 after genetic inactivation of Ybx1. c) FACS plot showing comparable abundance of mature myeloid and erythroid cells following genetic deletion of Ybx1, d) total numbers of mature blood cells of the myeloid (Gr1.sup.+), erythroid (TER119.sup.+), B-lymphoid (CD19) and T-lymphoid (CD3) lineage at week 16 after genetic inactivation of Ybx1. LSK: Lineage.sup.−Sca-1.sup.+Kit.sup.+ cells. e) colony numbers of Ybx1.sup.+/+versus Ybx1.sup.−/− murine stem/progenitor cells. Colonies were counted at day 8 after plating. Each sample was plated in duplicate, and three repeats of the experiment were performed. Data shown as mean±SD. f) spleen colony numbers counted on day 12 after injection of Ybx1.sup.+/+ or Ybx1.sup.−/− LSK-cells into lethally irradiated (12Gy) recipient mice (CFU-S12). CMP: common myeloid progenitor cells, GMP: granulocyte-monocyte progenitor cells, MEP: megakaryocyte-erythrocyte progenitor cells

    [0296] FIG. 19 a) shows dot plot showing the successful inhibition of respective targets of the corresponding kinase inhibitor used in this study (ANOVA test with Permutation based FDR<0.01). The size and colour of the dots are proportional to the phosphosite intensity, Z-scored (log 2 intensity). b) Dot plot showing changes in quantified Ybx1 phosphosites after various kinase inhibitor treatment (ANOVA test, permutation based FDR<0.01). The highlighted Ybx1 pS30 phosphosite is the only site highly significantly downregulated upon MEK/ERK inhibitor treatment compared to 10 controls. The size and colour of the dots are proportional to the phosphosite intensity, Z-scored (log 2 intensity).

    [0297] FIG. 20 shows a network representation of Ybx1 interactome based on annotation keywords. The keywords are highlighted in colours according to the protein function. n=4 biological replicates, t-test with Permutation based FDR<0.05.

    [0298] FIG. 21 a) shows scatter plot of Ybx1 interactome in Jak2V617F vs control. Ybx1 interactome is enriched for GO term mRNA splicing factors (green) and ribonucleoproteins (blue) assessed by Fisher's exact test. Fold enrichment of Ybx1 and Mapk1 in Jak2V617F cells compared to IgG control plotted against −log 10. b) Scatter plot of Ybx1 interactome in DMSO vs Jak-i (Rux, 0.5 μM treatment for 4 hours) treated Jak2V617F cells. Fold enrichment of Mapk1 in DMSO vs Jak-i plotted against −log 10 student t-test p-value. c) Immunoprecipitation of Ybx1 from murine Jak2V617F cells with and without Jak-i treatment (Rux, 0.5 μM for 4 hours) and analyzed for interaction with Mapk1 by western blot analysis using ERK1/2 antibody.

    [0299] FIG. 22 shows that Jak2-mutated clones are selectively vulnerable to inhibition of ERK signaling. a) Unsupervised hierarchical clustering of significantly down regulated phosphosites (n=2012 sites) in murine Jak2-mutated cells following inactivation of Ybx1 by 2 shRNAs compared to non-targeting control. Biological replicates were averaged (n=4), heatmap represents Z-scored log 2 transformed phosphosite intensity values of sample assessed by ANOVA test with Permutation based FDR<0.01. On the right, kinase-substrate motifs significantly down-regulated in Ybx1 targeted Jak2V617F cells are shown including Benjamin-Hochberg FDR value (−log 10). b) Network map for proteins assigned as ERK substrates (significantly downregulated phosphosites, ANOVA test with permutation-based FDR<0.01). c) Representative flow cytometry histograms showing pERK levels measured in Jak2 mutated human HEL cells (left) and primary patient cells (right) following RNAi (shNT versus shYbx1) and/or drug treatment (Ruxolitinib 500 nM, Trametinib 200 nM).

    [0300] FIG. 23 shows proteome analysis of murine Jak2V617F cells following inactivation of Ybx1 by 2 shRNAs compared to non-targeting control. Heat map representation of significantly enriched GO term biological processes is assessed by Fisher's exact test (P-value (−log.sub.10) shown).

    [0301] FIG. 24 a) shows quantification of Mcl-1 phosphosite pT144 (ANOVA test with Permutation based FDR<0.01) following genetic inactivation of Ybx1 with two shRNAs (shYbx1-1 and shYbx1-2) compared to shNT control in murine Jak2V617F cells. The y-axis is the log 2 intensity of the phosphopeptide. b) Scatter dot plot of Mcl-1 phosphosite pT144 after respective kinase inhibitor treatment (ANOVA test with Permutation based FDR<0.01). The y-axis is the z-scored, log 2 intensity of the phosphopeptide. c) Western blot analysis for relative protein abundance of Mcl1, Ybx1, Bim and Bcl-XL following genetic inactivation of Ybx1 with 4 different shRNA constructs (shYbx1 1-4), compared to non-targeting control (shNT). d) Measurement of apoptosis (AnnexinV/Sytox-positive cells) after genetic inactivation of Ybx1 and concomitant Jak-i treatment (Rux 100 nM, 500 nM), and ectopic overexpression of Mcl1. Results of 3 independent experiments. *: P=0.0286 as determined by Mann-Whitney test.

    [0302] FIG. 25 shows pharmacologic inhibition of murine Jak2V617F cells using a) Jak-inhibitor (Ruxolitinib, Rux) alone or in combination with a Mknk1 inhibitor (2 μM). b) Jak-i (Ruxolitinib, Rux) alone or in combination with the ERK-inhibitor trametinib (100 nM, 200 nM). Bar plot represents n=4 individual experiments with statistical analysis by Mann-Whitney-U test. c) Jak2-i (Pacritinib) alone or in combination with the ERK-inhibitor trametinib (100 nM, 200 nM). Bar plot represents n=4 individual experiments with statistical analysis by Mann-Whitney-U test. d) Induction of apoptosis after pharmacologic inhibition of FACS sorted CD34.sup.+ human bone marrow cells from patients with Jak2VF mutated neoplasms (n=6) and CD34.sup.+ healthy bone marrow cells (n=6) using Jak1/2 inhibitor Ruxolitinib (Rux, 500 nM), Jak2 inhibitor Pacritinib (Pac, 200 nM) or combinations with MEK/ERK inhibitor Trametinib (Tram, 200 nM) as indicated. Aligned dot plot shows proportion of dead cells, p-values shown in the plot (Paired t-test). e) Representative Western blot showing downregulation of Mknk1 protein abundance in human HEL cells following overnight combination treatment with Ruxolitinb and Trametinib. Experiment was performed three times. f, g) Xenograft model using NSGS mice transplanted with Jak2VF positive HEL cells treated with Ruxolitinb alone (n=12) or in combination with Trametinib (n=12). f) Percentage of hCD45 positive cells in the bone marrow. Paired t-test. g) Percentage of disease penetrance in NSGS animals. hCD45 positivity (>1%, grey) or negativity (<1%, green) following RUX or RUX/Tram treatment, respectively. h) Peripheral blood and i) Bone marrow analysis of human cell chimerism in the treated mice in weeks 4 and 20. i) Molecular quantification of JAK2V617F allelic burden in sorted human cells derived from the bone marrow of recipient animals as determined by pyrosequencing. l) Pie chart of JAK2V617F allelic burden per individual mouse that were either RUX (grey) or RUX/Tram (green) treated; m) Pie chart of JAK2V617F allelic burden per individual mouse that were either RUX (top) or RUX/Tram (bottom) treated.

    [0303] FIG. 26 shows that targeting Ybx1 in Jak2VF cells promotes retained introns. a) gene-ontology enrichment analysis of differentially expressed genes after inactivation of Ybx1 in murine Jak2VF cells. Bar plot shows significantly deregulated pathways with p-values on the x-axis. b) gene-set-enrichment analysis (GSEA) depicting negative enrichment of Erk-signaling related genes following inactivation of Ybx1. c) schematic representation of alternative splicing events types (left panel). Percentage and type of splicing events (1943 differential splicing events with p-value <0.05) differentially regulated following inactivation of Ybx1. Retained intron (RI, colored yellow) being significantly upregulated splicing event (70% increase) following inactivation of Ybx1; d) Experimental design of RNA sequencing and data analysis performed in this study; e) Bars represent number of retained intron events significantly upregulated in Ybx1 depleted cells with p-value 0.05 (1064 RI events) then filtered for p-value 0.01 and delta-PSI 0.1 are down to 472 highly significant RI events. f) Network map displaying enrichment of gene sets in the 472 highly significant RI events. Each node represents significantly enriched gene sets. Clusters of functionally related gene sets are circled and labels are highlighted, e.g, pathways involved in the regulation of MAPK- and apoptosis-signaling.

    [0304] FIG. 27 shows in a) retained intron read density profile between indicated exon-intron locations of Araf, Braf and Mnkn1 genes in control and Ybx1 targeted. b) transcript (top) and protein (bottom) expression levels of Araf, Braf and Mnkn1 in control and inYbx1 targeted murine Jak2VF cells. The transcript levels in TPM were determined by RNA-seq and protein levels in Log 2-fold change were determined by proteome data (n=4, p-values using one tailed student t-test). c) Western blot validation of Mknk1 protein abundance following inactivation of Ybx1 in murine BaF3 Jak2VF and human HEL cells, experiment was performed three times with similar results.

    [0305] FIG. 28 shows in a) network representation of Ybx1 interacting spliceosomal proteins in Jak2VF cells. The size and color of the node indicates the abundance of the corresponding proteins (Z-scored protein intensity is used as mentioned in the figure) and the edges are connected by STRING database interactions. b) List of significant Ybx1 interacting spliceosomal proteins presented according to their spliceosome complex. c) Spliceosome proteins interacting with Ybx1 participate in spliceosome assembly reaction in a stepwise manner to excise intronic sequences from immature mRNA to form a mature mRNA. BPS

    [0306] FIG. 29 shows generation of Ybx1 phospho-mimetic and phospho-null mutants. a) Western blot showing the expression of Ybx1 phosphomutants in BaF3 Jak2VF cells as indicated. b) Bar plot shows quantification of nuclear Ybx1 expression in Ybx1 phosphomutants expressing BaF3 Jak2VF cells, analysed by confocal microscopy. Data from 3 independent imaging experiments (p-values using student t-test). c) Cell growth curve of Ybx1 phosphomutants expressing BaF3 Jak2VF cells following treatment with increase doses of Jak-inhibitor (1 nM-10 μM RUX) measured by MTS assay. Data obtained from 4 independent experiments each with 8 technical replicates. d) Results of FACS analysis of apoptosis induction in Ybx1 phosphomutants expressing BaF3 Jak2VF cells following Jak-i treatment (RUX 0.5 μM) compared to untreated Ybx1wt expressing BaF3 Jak2VF cells.

    [0307] FIG. 30 shows that MEK-inhibition prevents Ybx1 nuclear localization in Jak2-mutated cells. a) Bar plot shows quantification of nuclear Ybx1 expression, analysed by confocal microscopy, in BaF3 Jak2VF cells after treatment with Ruxolitinib (0.5 μM), MEKi (2 μM), Trametinib (100 nM and 200 nM), Ruxolitinib in combination with Trametinib or DMSO control for 2 hours. Data from 3-4 independent imaging experiments (p-values using student t-test). b) Bubble plot showing the regulation of human Ybx1 pS32 and pS36 phosphorylation (representative of the murine Ybx1 pS30 and pS34) in HEL cells following treatment with Ruxolitinib (0.5 μM), MEKi (10 μM), Trametinib (500 nM and 2 μM) for 4 hours in vitro. Phosphorylation status of Mapk and Jak2 is shown as successful inhibition of respective targets of the corresponding kinase inhibitors. Following biological quadruplicates sample collection (n=4 per group), phosphopeptides were enriched using EasyPhos workflow and analyzed in single-run LC-MS/MS. Each data point is the averaged median of biological quadruplicate. Size and color of the bubbles are proportional to the Z-scored (log 2 phosphosite intensity) and significance was tested using multiple sample test. c) Bar plot shows quantification of nuclear Ybx1 expression, analyzed by confocal microscopy, in HEL cells following treatment with respective inhibitors as shown for 2 hours. Data from 3 independent imaging experiments (p-values using student t-test); d) Immunoprecipitation of Ybx1 from Ybx1 phosphomutants expressing Ba/F3 Jak2VF cells and analyzed for interaction with Mapk1 by western blot using ERK1/2 antibody. Representative images from n=3 biological independent experiments.

    [0308] FIG. 31 shows Western blot showing the regulation of Mknk1 protein abundance in Ybx1 phosphomutants expressing Ba/F3 Jak2VF cells as indicated. Representative images from n=3 biological independent experiments.

    [0309] FIG. 32 a) Unsupervised hierarchical clustering of significantly down regulated phosphosites (n=2390 sites) in human HEL Jak2-mutated cells following inactivation of Ybx1 by 2 independent shRNAs compared to non-targeting control. Biological replicates were averaged (n=4), heatmap represents Z-scored log 2 transformed phosphosite intensity values of sample assessed by ANOVA test with Permutation based FDR<0.01. b) Kinase-substrate motifs significantly down-regulated in Ybx1 targeted HEL cells are shown including Benjamin-Hochberg FDR value (−log 10). c) ERK substrate motifs significantly downregulated and shared between Ybx1 targeted mouse and human Jak2VF cells. d) Western Blot analysis of total protein abundance and phosphorylation of Jak2-downstream targets upon Jak-i treatment and/or genetic inactivation of Ybx1 by RNAi. Levels of GAPDH were used as loading control for each individual blot. Blots are representative of at least 3 independently performed experiments with similar results. e-f) Bar plots show the mean fluorescence intensity of pERK levels measured e) in human HEL and f) patient Jak2 mutated cells following genetic inactivation by RNAi with or without drug treatment as indicated. Representative FACS plots shown in FIG. 22c. Data shown as mean±SD and p-value determined by two-tailed student t-test. g) Western blot validation of Mknk1 targeting shRNAs in murine BaF3 Jak2VF cells. h) Representative western blot analysis of pERK upon genetic inactivation of Mknk1 in BaF3 Jak2VF cells. The experiment was performed four times with comparable results. i) Growth curve of Jak2VF cells following lentiviral infection with shRNAs targeting Mknk1 or non-targeting control (shNT) and treatment with increase doses of Jak-inhibitor (1 nM-10 μM RUX) measured by MTS assay. Data obtained from 4 independent experiments each with 8 technical replicates. j) Percentage of apoptotic Jak2VF cells following lentiviral knockdown of Mknk1 (sh2 and sh3 with or without Ruxolitinib 0.5 μM) compared to non-targeting control (shSCR). Data obtained from 4 independent experiments and p-value determined by two-tailed student t-test.

    [0310] FIG. 33 a) Measurement of apoptosis (AnnexinV/Sytox-positive cells) after genetic inactivation of Ybx1 and concomitant Jak-i treatment (RUX 100 nM, 500 nM). Rescue by ectopic overexpression of Mcl-1. Results of 3 independent experiments. P=0.0286 as determined by Mann-Whitney test. b) Heatmap shows unsupervised hierarchical clustering of significantly regulated (t-test with permutation-based FDR<0.01) phosphosites with (n=24) and without (n=24) Jak-i treatment in Jak2 mutated primary patient samples. Phosphoproteome analysis of Jak2 mutated primary patient samples (total n=48) samples following in vitro (n=18, Jak-i treatment for 2 hours) or in vivo (n=6, 2 hours post dosing of Rux samples were collected) exposure to Ruxolitinib. c) Network map of significantly enriched GO terms (p-value <0.01) of dephosphorylated proteins upon Jak-i treatment. Proteins involved in mRNA splicing are enriched and listed. d) Box plot shows no significant changes in the Mapk1 and Mapk3 phosphorylation in control vs Jak-i treated patient samples. e) Box plot shows significant changes in Ikbkb, Stat3 and Stat5 phosphorylation in control vs Jak-i treated patient samples. p-values as determined by Mann-Whitney test.

    [0311] FIG. 34 shows that combination of mRNA splicing and processing factor inhibitors and Jak inhibitors synergistically and significantly impairs growth and proliferation of murine JAK2V617F cells. a)-b): Dose dependent inhibition of growth and proliferation of Jak2V617F cells upon treatment with a splicing factor inhibitor in the presence and absence of Jak inhibitor Ruxolitinib (RUX, at 0.5 μM). Growth and proliferation was measured by MTS assay. Data shown are representative of 2 independent experiments with 8 technical replicates, mean±SD. The used splicing factor inhibitors are: a) GSK3326595, a Prmt5 inhibitor; b) Indisulam, a Rbm39 inhibitor; c) Herboxidiene; d) Pladienolide B.

    EXAMPLES

    Methods

    Cell Culture

    [0312] Murine Ba/F3 cells stably expressing Jak2WT and Jak2V617F, human SET-2 and HEL 92.1.7 cells (purchased from DSMZ, Braunschweig, Germany) were cultured in RPMI 1640 medium (Life Technologies, Carlsbad, Calif., USA) supplemented with 10% FBS (Life Technologies) in a humid atmosphere of 5% CO2 at 37° C. Cell lines were tested and maintained mycoplasma free throughout the study. Primary murine cells were cultured in StemSpan SFEM medium (Stemcell Technologies, Vancouver, Canada) supplemented with cytokines (100 ng/ml SCF, 10 ng/ml TPO, 6 ng/ml IL-3 and 10 ng/ml IL-6; Pepro Tech, Rocky Hill, N.J., USA). The functional inhibitors were GSK3326595 (Cat #58664, Selleckchem), Indisulam (Cat # SML1125, Sigma Aldrich), Herboxidiene (Cat #Cay25136, Biomol, Cayman chemicals), and Pladienolide B (Cat #6070, Tocris).

    Primary Patient Samples

    [0313] Patient samples and healthy donor controls were obtained after informed consent and according to the Helsinki declaration from the Tumor Banks in Jena and Magdeburg, approved by the respective local ethics committee (Ethics Committe, University Hospital Jena #4753/04-16 and Ethics Committee, Medical Faculty, OvGU Magdeburg #115/08).

    Focused Lentiviral shRNA Library Screen

    [0314] In brief, the Mission TRC lentiviral pLKO.1 shRNA vectors (Sigma Aldrich) targeting the top 15 hits of mRNA processing and splicing factors enriched in Jak2V617F were selected) and lentiviruses were individually produced per shRNA by co-transfecting with 3rd generation packaging plasmids pMDL, pRSV and pVSVG in HEK293T cells seeded in a 10 cm culture dishes. In total were used 74 shRNAs, 4-5 different shRNAs per each target (SEQ ID NO. 1-144), and 4 non targeting controls (Sigma Aldrich, catalog number: SHC016, SHC016V). Sequences for Cpsf7 were bought from Sigma Aldrich, catalog number SHCLNG-NM_172302, SEQ ID NOs 1-10. The viruses were collected at 48 hours and 72 hours after transfection, syringe filtered through a 0.45 μm syringe filter, concentrated using 30K Amicon Ultra-15 centrifugal filters (Merck) and the target murine BaF3 Jak2V617 cells were infected with polybrene (8 mg/ml, Sigma). For screening purpose 200,000 murine BaF3 Jak2V617 cells in 2 ml RPMI media with 10% heat-inactivated FBS (Invitrogen) were seeded in each well of a 6 well plates. Concentrated virus was added per well (with polybrene 8 mg/ml), centrifuged at 500×g for 1 hour at 35° C., subsequently 48 hrs after transduction cells were 1 μg/ml puromycin selected for 2 days. On day 3, cells were washed, viable cells were counted and plated in 96 well plates (8 technical replicates per sample condition per experiment. In addition, each experiment was performed in technical duplicates for growth assay with and without Jak2 inhibitor Ruxolitinib (0.5 μM). The plates were incubated at 37° C. and 5% CO2 for 72 hours and subjected to Cell Titre 96 aqueous one solution (Promega) according to the manufacture's protocol. Further viable cells were counted after 72 hours with the countess automated cell counter (ThermoFischer Scientific) using trypan blue. Western blot was carried out to assess knock down efficiency on protein level for Pcbp1 and Ybx1. For data analysis, the 8 technical duplicates were averaged and the values were normalized against non-target controls included in each plates. The analysis of the normalized data between biological replicates showed correlation coefficient between r=0.963 to 0.988 indicating high reproducibility of the procedure. Targets were considered potential candidates if 2 or more shRNA responded only in the Jak2 inhibitor treated group but not in the untreated controls of all the biological replicate experiments performed. Using these criteria, we choose our candidates for further characterization.

    Inactivation of Murine and Human Ybx1

    [0315] In order to inactivate Ybx1, 4 shRNAs targeting mouse Ybx-1 (SEQ ID NO. 137-144) and 4 shRNAs targeting human Ybx-1 (SEQ ID NO. 159-162, Sigma-Aldrich, St. Louis, Mo., USA) were validated and the two selected shRNAs (mouse Ybx1: SEQ ID NO: 141, 143; human Ybx1: SEQ ID NO: 159, 160) were used thereafter using respective non-targeting controls. In brief, Ba/F3, SET-2, and HEL cells were lentiviral transduced with lentiviral particles by centrifuging the cells at 872×g for 1.5 hours at 33° C. The cells were cultured for 2 days, puromycin selected for 48 hours and seeded (5×10.sup.6 cells) followed by inhibitor treatment or addition of diluent control as indicated below. Cells were harvested and Ybx-1 knock-down was checked by qPCR and western blotting.

    Jak-i Dose Dependent Cell Growth and Viability in Ybx1-Inactivated Cells.

    [0316] To analyse Jak inhibitor dose dependent cell growth and viability, murine BaF3 Jak2V617F cells treated with a non-targeting shNT control or shYbx1 (two different shRNA targeting Ybx1, sh1 and sh2) were counted with the Countess™ automated cell counter (Thermo Fischer Scientific) using trypan blue in 96 well plates after 2 days of 1 μg/ml puromycin selection. 3×10.sup.4 viable cells per well (8 technical replicates per sample condition) in 96 well plate were seeded in RPMI medium with 10% heat-inactivated FBS and exposed to different concentrations of the Jak2 inhibitor Ruxolitinib ranging between 1 nM-10 μM. The plates were incubated at 37° C. and 5%002 for 72 hours and subjected to Cell Titer 96 Aqueous One Solution (Promega) according to the manufacture's protocol. Viable cells were counted after 72 hours by trypan blue. Determination of IC.sub.50 inhibitory concentration of the Jak-i was calculated using GraphPad Prism.

    Drug Combination Treatments

    [0317] Viable BaF3 Jak2V617F cells were seeded at a density of 30,000 cells per well in RPMI medium with 10% heat inactivated FBS and exposed to Ruxolitinib (0.5 μM) alone or in combination with different concentration of the Mknk1 inhibitor CGP57380 (Sigma), MEK/ERK inhibitor Trametinib (Novartis), PI3K inhibitor LY294002, and p38 MAP kinase inhibitor SB203580 (Merck) and incubated for 72 hours. Cell Titer 96 Aqueous One Solution was added to the plates according to the manufacture's protocol and measurements were performed after 4 hours. The plates were read at 490 nm in Tecan Infinite M200 and the responses were analyzed using GraphPad Prism.

    Apoptosis Assays

    [0318] Cells stably infected with either non-targeting or Ybx1 specific shRNA were seeded in six-well plates and selected for 24 hours with puromycin. Primary murine lineage-depleted cells or FACS-sorted human CD34.sup.+ cells were incubated in 48 well plates. Inhibitor treatment was performed at concentrations as indicated for 48 hours unless otherwise stated. Apoptosis was measured by flow cytometry on a BD FACS Canto™ cytometer using Annexin V in combination with SYTOX™ Blue or SYTOX™ Green as dead cell stains.

    Proliferation Assay with PCNA

    [0319] After puromycin selection of murine BaF3 Jak2V617F treated with non-targeting shNT or shYbx1, cells were washed twice with ice cold 1×PBS, fixed in 70% ethanol and permeabilized with 0.1% Tween-20. The cells were stained with PCNA-Alexa fluro488 conjugate (Biozol) on ice for 20 min and measured by flow cytometry (FIG. 11).

    MTS Assay

    [0320] 3×10.sup.4 viable cells per well (8 technical replicates per sample condition) in 96 well plate were seeded in RPMI medium with 10% heat-inactivated FBS and exposed to different concentrations of the Jak inhibitor, mRNA splicing and processing factor inhibitor or MEK inhibitors. The plates were incubated at 37° C. and 5% CO2 for 72 hours and subjected to Cell Titer 96 Aqueous One Solution (Promega) according to the manufacture's protocol and measurements were performed after 4 hours. The plates were read at 490 nm in Tecan Infinite M200 and the responses were analyzed using Graph Pad Prism.

    Cell Cycle Analysis

    [0321] For cell cycle measurements, 2×10.sup.6 murine BaF3 Jak2V617F cells stably expressing shNT or shYbx1 were washed in ice cold 1×PBS twice, fixed in ice cold 70% ethanol for 30 minutes on ice and stored at 4° C. After collection of biological replicates, samples were Ribonuclease A treated and stained with Propidium Iodide (PI). The PI stained cells were measured using BD Canto flow cytometer and data analyzed in FlowJo.

    ROS Measurements Using Carboxy-H.SUB.2.DFFDA

    [0322] In brief, 1×10.sup.6 murine BaF3 Jak2V617F cells stably expressing shNT or shYbx1 were washed twice with 1×PBS and resuspended in 20 μM carboxy-H2DFFDA for 30 mins in dark at room temperature. Thereafter, the cells were washed thrice in 1×PBS and measured using BD FACSCanto™ cytometer. Data were analyzed in FlowJo.

    DNA Damage Analysis Using yH2AX pS139

    [0323] 1×10.sup.6 murine BaF3 Jak2V617F cells stably expressing shNT or shYbx1 cells were seeded on Poly-L-Lysin coated dishes for 2-4 hours, washed in PBS, fixed with 4% Paraformaldehyde, blocked in blocking buffer (0.2% Triton-X, 1% BSA and 5% Normal rabbit serum) and incubated over night with rabbit yH2AX pS139 antibody (#CST 2577) overnight. After overnight incubation, samples were washed and incubated with secondary antibody (anti-rabbit-Alexa 568). DAPI was used for nuclear staining (NucBlue, ThermoScientific). Positive control samples were prepared by exposing the BaF3 Jak2VF cells stably expressing shNT to 20 mins UV under the cell culture hood. Imaging was performed using Zeiss LSM 780 microscope and processed in Zen Black software tool.

    Label Free Phospho-Proteome Sample Preparation

    [0324] Samples were collected as quadruplicate biological replicates for each condition, lysed in Guanidinium chloride (Gmdcl) buffer (6M Gdmcl, 100 mM Tris pH8.5, 10 mM TCEP and 40 mM CAA), heated for 5 mins at 95° C. and cooled on ice for 15 min. Lysed samples were then sonicated (Branson probe sonifier output 3-4, 50% duty cycle, 10×30 sec) and heated again. Proteins were precipitated with acetone, and quantified by bicinchoninic acid BCA assay. A protein sample of 2 mg was digested with LysC and Trypsin overnight at room temperature and phosphopeptides enriched by TiO.sub.2 beads. The enriched peptides were desalted, washed and eluted on StageTips with 2 layers of SDB-RPS material with elution buffer (80% Acetonitrile and 5% NH.sub.4OH). The eluted peptides were vacuum centrifuged until dryness and reconstituted in 2% ACN/0.1% TFA. All the samples were stored in −20° C. until measurement.

    Phosphoproteome of Primary Patient Samples

    [0325] In brief, peripheral blood samples from patients with Jak2 mutated myeloproliferative neoplasms were collected, granulocytes isolated, and treated with DMSO or Ruxolitinb 0.5 μM for 2 hours (either in vitro or in vivo). Cells were lysed and processed in 4% SDC buffer (4% SDC, 100 mM Tris pH8.5, 10 mM TCEP and 40 mM CAA), heated for 5 mins at 95° C. and cooled on ice for 15 min. Lysed samples were sonicated, heated again for 5 mins and BCA quantified. Approximately 350 μg of proteins were digested with LysC and Trypsin overnight at room temperature and phosphopeptides were enriched by TiO.sub.2 beads as described elsewhere.

    Deep-Proteome Quantification.

    [0326] Proteome samples of phosphoproteome analysis were collected after TiO.sub.2 enrichment. In brief, cells were lysed in 1% SDC buffer (1% SDC, 100 mM TrispH8.0, 40 mM CAA and 10 mM TCEP), heated for 5 mins at 95° C., cooled on ice for 15 mins and sonicated (Branson probe sonifier output 3-4, 50% duty cycle, 10×30 sec). 25 μg were digested with LysC and Trypsin overnight and peptides were eluted on Stage Tips with 3 layers of SDB-RPS material with elution buffer. The eluted peptides were vacuum centrifuged until dryness and reconstituted in 2% acetonitril (ACN)/0.1% trifluoro acetic acid (TFA). All the samples were stored in −20° C. until measurement.

    Drug-Perturbed Phosphoproteome Profiling

    [0327] In order to profile the kinase inhibitor action on Jak2V617F cells, murine Jak2V617F BaF3 were treated with 0.5 μM Jak2 inhibitor Ruxolitinib (Selleckchem, S1378) for 2 hours and 10 μM MEK inhibitor PD0325901 (Sigma), 10 μM p38 inhibitor SB203580 (Merck), 20 μM JNK inhibitor SP600125 (Sigma), 50 μM PI3K inhibitor LY294002 (Merck), 10 μM AKT inhibitor MK2206 (Enzo Life) and 100 nM mTOR inhibitor Torin-1 (Millipore) for 1 hour. The cells were lysed in Gmdcl buffer and processed as mentioned in the phospho-proteome sample preparation protocol.

    Liquid Chromatography (LC)-MS/MS Measurement

    [0328] For the LC-MS/MS analysis, Q-Exactive mass spectrometer with a nanospray ion source connected online to an Easy-nLC 1000 HPLC system was used. Peptides were separated on an in-house prepared 50 cm C18 columns (75 μM inner diameter with 1.9 μM C18 ReproSil particle, Dr. Maisch GmbH) in a 140 minute gradient between 5%-65% in buffer B (0.5% formic acid, 80% acetonitrile). The column temperature was maintained at 50° C. using column oven (in-house made). Peptides were analyzed with a full scan (300-1600 m/z, R=60,000 at 200 m/z) at a target of 3e6 ions, followed by high energy collisional disassociation-based fragmentation (HCD) of top10 most abundant isotope patterns with a charge MS/MS scan, detected in the orbitrap detector (R=15,000 at 200 m/z). Dynamic exclusion of sequenced peptides was set to 40 s and apex trigger (4 to 7 s) were on. All data were acquired using X-caliber software (Thermoscientific).

    Data Processing with Maxquant

    [0329] Mass spectrometric raw files were processed using the Andromeda search engine integrated into Maxquant 15 environment (1.5.5.2 version). The MS/MS spectra were matched against the mouse (UniProt FASTA 2015_08) database with an FDR<0.01 at the level of proteins, peptides and modifications. The search included fixed modification for carbamidomethyl and in the variable modifications table phosphoSTY was added additionally for the phosphorylated peptide search to the default settings. Peptides with at least seven amino acids were considered for identification. Maximum two missed cleavages were allowed for protease digestion. Match between run was enabled with the matching window of 1 min to transfer peptide identification to across runs based on normalized retention time and high mass accuracy.

    Phosphoproteome Data Analysis

    [0330] Perseus16 software (1.5.2.11 version) environment was used for all Maxquant output table analysis. For phosphoproteome analysis, sample for class-I phosphosites (localization probability >0.75) and required a minimum of 3 or 4 valid values in each of the biological quadruplicates Statistical analysis of was performed on the logarithmized intensities values. Significance was assessed by Student's t-test using permutation-based FDR, to identify the significantly regulated phosphosites. In group comparisons two sample t-test or for multiple samples comparison ANOVA test was performed with permutation-based FDR cut-off 0.01 or 0.05. The significantly regulated phosphosites were filtered, Z-scored and represented as either unsupervised hierarchical clustered heat maps or profile plot. Annotations were extracted from UniprotKB, Gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome. Kinase-substrates relationships were extracted from phosphosite plus database (phosphosite.org). Fischer exact test was performed to discover motifs and annotations that are significantly regulated in the sample groups. For phosphosite occupancy calculation, the proteome and phosphoproteome of corresponding samples were matched in maxquant to estimate occupancy, occupancy ratio and occupancy error scale using the extracted signal difference of modified peptide, unmodified peptide and the corresponding protein ratios (described earlier, 8). Phosphoprotein network architecture were obtained using String database and further networks and sub-networks were analyzed and visualized in Cytoscape.

    Ybx1 Interactome Preparation

    [0331] For Ybx1 affinity purification, cells were lysed in 150 mM NaCl, 50 mM Tris (pH7.5), 5% glycerol, 1% IGPAL-CA-630 (Sigma), protease inhibitors (EDTA free, Roche), 1% Benzonase, and 1 mM MgCl2 for 30 min on ice. 1 mg of total lysate was incubated with validated Ybx1 antibody (Abcam #76149) overnight and 30p1 of rec-protein A sepharose 4B conjugates (Invitrogen) for 2 hours. Non-specific binders were removed by three washes with wash buffer 1 (150 mM NaCl, 50 mM Tris (pH7.5), 5% glycerol, 0.05% IGPAL-CA-630) and three washes with wash buffer 11 (150 mM NaCl, 50 mM Tris (pH7.5)).

    [0332] The bound proteins were on-beads digested with Trypsin and LysC overnight. The peptides were desalted on C18 Stage tips and analyzed by mass spectrometry.

    RNA Sequencing

    [0333] Transcriptome profiling of BaF3 Jak2VF cells stably expressing shNT or shYbx1 cells was performed using a strand-specific RNA sequencing protocol described previously. In brief, total RNA was isolated using 2×10.sup.6 cells using NucleoSpin RNA Kit according to the Manufacturer's protocol (Macherey Nagel). RNA library for sequencing was prepared using NEBNext Poly(A) mRNA Magnetic Isolation Module. The quality was analyzed on a Bioanalyzer (Agilent 2100 Bioanalyzer) high sensitivity DNA assay. Samples were sequenced on Illumina Nexseq500 and multiplexed reads were demultiplexed on the basis of their barcodes. Sequencing reads were filtered, trimmed and then mapped to the Ensembl gene annotation and the mouse genome assembly GRCm38 using STAR aligner with ENCODE settings in two-pass mode considering splice junctions across all samples in the second mapping step. Gene counts were quantified using feature Counts 28 and differential expression calculated with the limma-voom pipeline. Gene and transcript expression levels were quantified using RSEM. Event level differential splicing was calculated with the EventPointer package in R.

    Ybx1 Conditional Knock Out Genotyping and Gene Expression by qPCR

    [0334] Genotyping of tails and fetal livers was performed using the following primers: Ybx1_cond_for (GCCTAAGGATAGTGAAGTTTCTGG SEQ ID NO 145), Ybx1_con_rev (CCTAGCACACCTTAATCTACAGCC, SEQ ID NO 146), Cre_for (CGTATAGCCGAAATTGCCAG, SEQ ID NO 147), Cre_rev (CAAAACAGGTAGTTATTCGG, SEQ ID NO 148). Genotyping PCR was performed using the Dream Taq Green PCR Master Mix (2×) (Thermo Fisher Scientific, Waltham, Mass., USA) following the manufacturers' protocol.

    Histology Staining and Immunohistochemistry

    [0335] Formalin-fixed and paraffin-embedded bone marrow biopsies with proven myeloproliferative neoplasia or primary samples without histopathologic abnormalities were retrieved from the archival files of the Institute of Pathology, Otto-von-Guericke University Medical Center, Magdeburg, Germany. All MPNs were diagnosed and classified according to the World Health Organization (WHO) 2008 classification in synopsis with clinical data and presentation. The study comprises of biopsies derived from 76 MPN patients (PV (n=23), ET (n=32) and MF (n=21)) compared to healthy donor controls (n=18) or BCRABL positive CML (n=17). Immunohistochemistry was performed using a monoclonal Rabbit-anti-human Ybx1 antibody (Abcam; ab76149) in a dilution of 1:100.

    Hematopoietic Progenitor Cell Assays

    [0336] Colony formation assay: For investigation of colony formation in methylcellulose, LSK (Lin−Sca1+KIT+) cells were sorted from bone marrow of the respective donor mice as previously described. 1×10.sup.3 cells were seeded in MethoCult M3434 (Stem Cell Technologies), respectively. Colony numbers were counted on day 10 after plating using standard methods. Spleen colony formation assays (CFU-S12): bone marrow cells were collected from donor mice and 1×10.sup.2 LSK cells were FACS sorted and injected via tail vein into lethally irradiated (12Gy TBI) C57BL/6 recipient mice. At day 12 post-injection, spleens from recipient mice were harvested and stained with Bouin's fixative solution (Sigma-Aldrich), and colonies were counted using standard methods.

    Experimental Animals

    [0337] All mice were housed under pathogen-free conditions in the accredited Animal Research facility at the Animal Research Facility of the Otto-von-Guericke University—Medical Faculty, Magdeburg. All experiments were approved by the Landesverwaltungsamt Saxony-Anhalt, Halle, Germany. Conventional Ybx1 knockout mice have been generated as previously described 23. Mice harboring a ‘floxed’ (flanked with loxP sites) allele of Ybx1 have been generated at Taconic-Artemis in a pure C57BL/6 background.

    Transplantation Assays and In Vivo Treatment

    [0338] For competitive repopulation assays 2×10.sup.6 BM cells (for Jak2WT) or 5×10.sup.4 sorted LSK cells (for Jak2V617F) of 6-8 week old Ybx1.sup.−/− or Ybx1.sup.+/+ (CD45.2) littermates and 2×10.sup.6 (CD45.1/2) competitor cells (derived from intercrossing CD45.1 animals with CD45.2 animals purchased from Charles River) were transplanted via lateral tail vein injection into lethally irradiated (12Gy, single dose) 6-8 week old Ly45.1 mice (Jackson Laboratories, Bar Harbor, Me.). For serial transplantation experiments whole BM of primary recipient mice was harvested and 2×10.sup.6 whole BMC were injected into lethally irradiated secondary recipients. Ruxolitinib was purchased at Selleckchem (Selleckchem, S1378) and formulated for administration by oral gavage as previously described. Mice received the Jak1/2 inhibitor ruxolitinib at a dose of 90 mg/kg or vehicle control by oral gavage BID. For xenografting of Jak2-mutated human cells, HEL cells were either infected with lentiviral particles for transduction of the respective shRNAs (shNT or shYbx1) or incubated for 24 hours with inhibitors as indicated. 1×10.sup.6 viable cells were injected in each irradiated (2Gy) recipient NSGS mouse via lateral tail vain injection. Engraftment and expansion of human cells was monitored weekly by the presence of hCD45-positive cells in the peripheral blood. For patient derived xenograft experiments (PDX) we used an improved model for human HSC transplantation and analysis, that has been developed by her group from immune-deficient mouse strains containing Kit mutations (NSGW41). These mice can be engrafted without prior conditioning and therefore maintain an intact niche and microenvironment. Primary bone marrow samples were acquired during routine biopsies and cells were isolated by Ficoll gradient centrifugation followed by depletion of CD3 positive cells. 1.8-2×10.sup.6 stem- and progenitor cells (HSPCs) were engrafted (pairwise) per animal. Mice were followed for 4 weeks and peripheral blood chimerism of human CD45 positive cells was measured by flow cytometry. Between weeks 4 and 20 all animals were treated for 5 days every 4 weeks with either Ruxolitinib (90 mg/kg BID per gavage) or the combination of Ruxolitinib with the MEK/ERK-inhibitor Trametinib (1 mg/kg QD per gavage).

    Quantification of JAK2VF Mutant Cells by Pyrosequencing

    [0339] In order to assess for the relative abundance of JAK2-mutated cells within the patient derived xenograft (PDX) model, we sorted human CD45 positive cells from the bone marrow at week 20 and performed pyrosequencing for the JAK2V671F mutation. DNA isolation and whole-genome amplification were carried out on FACS-sorted hCD45 positive cells using the REPLI-g Single Cell Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Amplicons were generated using AmpliTaq Gold DNA Polymerase (Thermo Fisher Scientific, Waltham, Mass., USA; biotinylated forward primer: GAAGCAGCAAGTATGATGAGCA (SEQ ID NO 149); reverse primer: TGCTCTGAGAAAGGCATTAGAA (SEQ ID NO 150)) according to standard protocols. Samples were then analyzed by pyrosequencing (PyroMark Q96 ID, Qiagen, Hilden, Germany; sequencing primer: TCTCGTCTCCACAGA (SEQ ID NO 151)) to assess for the mutational status of the JAK2V617F variant of the individual subpopulations.

    Rescue of shRNA-Inactivated Endogenous Ybx1 with Exogenous Enforced Expression of Mcl-1

    [0340] Ba/F3 cells expressing EpoR (MSCV-EpoR-Neo) and Jak2V617F-GFP (MSCV-Jak2V617F-GFP) were infected with retrovirus expressing empty vector (MSCV-Puro) or Mcl-1 (MSCV-Mcl-1-Puro). Knockdown of Ybx1 was performed as indicated.

    Immunoprecipitation and Immunoblotting

    [0341] Ba/F3 EpoR and wild-type Jak2 or Jak2V617F, respectively, were washed twice with PBS and starved for 4 h in serum reduced (0.5%) medium at a density of 1×10.sup.6/ml. For immunoprecipitation, the TrueBlot Anti-Rabbit Ig IP Beads Kit (Rockland Immunochemicals, Gilbertsville, Pa., USA) was used following the manufacturers instruction. The following antibodies were purchased from Cell Signaling (Danvers, Mass., USA) and used at a 1:1000 dilution: p-Akt (9271), Akt (9272), p-p44/42 MAPK (9106), p44/42 MAPK (9102), p-cRaf (9427), cRaf (9422) and p-Ybx1 (Ser102) (2900). GAPDH antibody (H86504M, 1:5000) was purchased from Meridian Life Sciences (Memphis, Tenn., USA), p-Stat5 antibody (05-495, 1:1000) was purchased from Millipore (Darmstadt, Germany) and Stat5 (sc-1081, 1:100) antibody was purchased from Santa Cruz Biotechnologies (Dallas, Tex., USA). Mcl-1 antibody (600-401-394, 1:1000) was delivered by Rockland (Limerick, Pa., USA) and Ybx1 antibody (ab76149, 1:1000) was delivered by Abcam (Cambridge, UK).

    Flow Cytometry

    [0342] For immunophenotype analysis, peripheral blood cells, bone marrow or spleen cells were resuspended in PBS/1% FBS after erythrocyte lysis (PharmLyse™, BD Pharmingen). Unless otherwise stated, the following antibodies were used: Sorting and analysis of LSK-cells or Sca-1+ cells were performed as previously described 25,26. Biotinylated antibodies against Gr-1 (RB6-8C5), B220 (RA3-6B2), CD19 (6D5), CD3 (145-2C11), CD4 (GK1.5), CD8 (53-6.7), TER119 and IL7Ra (A7R34) (all Biolegend, San Diego, Calif., USA) were used for lineage staining. An APC-Cy7- or BV421-labeled streptavidin-antibody (Biolegend) was used for secondary staining together with an APC-anti-KIT (clone 2B8) and a FITC- or PE-anti-Sca-1 antibody (clone E13-161.7). Cells were analyzed using a FACSCantoll™ (Becton-Dickinson) cytometer. Analysis was performed using FlowJo™ software (Treestar, Ashland, Oreg.). Fix&Perm Kit (Life Technologies) was used for intracellular staining according to the manufacturer's protocol.

    Immunofluorescence Analysis (Jak2-Ybx1 Co-Localization)

    [0343] 2×10.sup.4 cells were washed and seeded onto adhesion slides (Marienfeld). Attached cells were fixed in PBS/2% PFA/0.01% glutaraldehyde for 15 min on ice followed by permeabilization for 10 min with PBS/0.02% Triton X100 at room temperature. After blocking for 30 min with PBS/1% BSA/0.1% Tween 20, Ybx-1 (Abcam ab76149, 1:50) or pYbx-1 labeled with Alexa488 (Ser102) (BIOSSUSA bs-3477R-A488, 1:50) was incubated in blocking solution 1 h. The samples were washed 5 times 5 min with PBS followed by incubation of Alexa Fluor 488 donkey anti-rabbit antibody (life technologies A21206, 1:200) for 1 h. After additional washing, Jak2 labeled with Cy3 (BIOSSUSA bs-0908R-Cy3, 1:50) was incubated in blocking solution for 1 h. In the following, the samples were washed, incubated with DAPI for 10 min and mounted using ProTags® Mount Fluor (quartett, 401603095). Samples were analyzed using confocal microscope Leipa SP8.

    Generation of Ybx1 Phosphorylation Mutants

    [0344] Phosphorylation mutants mimicking hyperphosphorylation or de-phosphorylation of Ybx1 were generated by site-directed mutagenesis at amino acid residues that were (i) highly conserved and (ii) differentially phosphorylated in the absence or presence of mutated JAK2 kinase. These aspects applied to the murine serine residues S30, S34, S172 and S174. In detail the following mutants were generated by site directed mutagenesis using a retroviral MSCV-IRES-GFP backbone (Addgene, plasmid #20672): (1) MIG-mYbx1-S30A/S34A (SEQ ID NO 152); (2) MIG-mYbx1-530A (SEQ ID NO 153); (3) MIG-mYbx1-534A (SEQ ID NO 154); (4) MIG-mYbx1-S30D/S34D (SEQ ID NO 155); (5) MIG-mYbx1-530D (SEQ ID NO 156); (6) MIG-mYbx1-534D (SEQ ID NO 157) and (7) MIGmYbx1-S172A/S174A (SEQ ID NO 158). Constructs were expressed in murine Jak2-mutated (Ba/F3-JAK2VF) cell lines. In brief, cells were infected by co-localization of virus supernatant (containing the respective constructs as indicated above) with Ba/F3-Jak2-V617F(VF) cells on retronectin-coated plates. Infection has been repeated after 24 hours and GFP-positive cells were sorted to ensure expression of the mutants in a homogeneous population.

    Statistical Analysis

    [0345] For survival analysis, Kaplan-Meier curves were plotted using GraphPad Prism™ version 6.0 h (GraphPad Software, SanDiego, Calif.). Differences between survival distributions were analyzed using the logrank test. Statistical analyses were performed using Student t test (normal distribution) or Mann-Whitney U test (when normal distribution was not given). P less than 0.05 was considered statistically significant (p<0.05 indicated as *, p<0.01 indicated as **, and p<0.001 indicated as ***).

    Example 1. Phosphoproteomic Analysis of JAK2V617F Mutants

    [0346] To identify downstream effectors of the mutant Jak2 kinase that drive the evolution of persistent clones, we performed in-depth mass spectrometry (MS)-based phosphoproteomics in murine hematopoietic cells expressing erythropoietin-receptor and either Jak2-wildtype (Jak2WT) or mutated Jak2-V617F (Jak2V617F) kinase (FIG. 2a). Cells were exposed to either erythropoietin (EPO) alone or combined with a Jak-inhibitor (Jak-i; Ruxolitinib, Rux) to investigate EPO-independent and ruxolitinib-responsive signaling in Jak2V617F compared to Jak2WT. In total, we quantified 21,764 distinct class-I phosphorylated sites on 4135 different proteins (FIG. 2b). Of 6517 phosphosites significantly regulated in Jak2V617F and Jak2WT at a False Discovery Rate of 5%, 5191 were distinctly regulated in Jak2V617F cells on 1758 proteins, including bona fide Jak2 targets such as STAT5, STAT3, and PIM (FIG. 3). Kinase motifs of glycogen synthase kinase-3 (GSK3), ERK, and cyclin-dependent kinases (CDKs) were also enriched. Notably, the most highly enriched cellular processes in Jak2V617F cells were those of mRNA splicing and processing (FIG. 4).

    Example 2. Inactivation of mRNA Splicing and Processing Factors Sensitizes Jak2V617F Mutant Cells to Treatment with Jak Inhibitors

    [0347] To assess the potential functional role of protein members (n=47) of the mRNA splicing and processing pathway, we chose 15 members that were significantly phosphorylated in Jak2V617F mutant on the basis of their statistical significance, i.e. p-value (FIG. 5), and subjected them to an RNA interfering (RNAi) based screening. Each candidate gene was targeted with 4-5 shRNAs, so that 70 shRNAs targeting 15 candidates and 4 non-targeting controls were used (western blotting, FIG. 6). shRNA-treated cells were analysed for knock-down efficiencies, and for growth with and without the Jak-inhibitor (Jak-i). Targeting Hnrnpk, Polr2a Srsf2, Srsf9, and Snrnp200 resulted in dropouts (FIG. 8), with at least 3 of 4 shRNAs displaying significant growth impairment. Interestingly, some of the targeted proteins identified in this screen appeared dispensable for proliferation and survival of Jak2V617F cells but their inactivation sensitized persistent cells to Jak-i treatment (FIG. 7a).

    [0348] The most prominent candidate (4 shRNAs out of 4 targeted) that sensitized Jak2V617F cells to Jak-i treatment was the pleiotropic Y-box binding protein 1 (Ybx1) (FIG. 7b, 8), a member of core spliceosomal proteins and regulates mRNA splicing in various cellular contexts.

    [0349] Ybx1 was highly expressed in 76 primary bone marrow (BM) biopsies of patients diagnosed with Jak2-mutated MPN (FIG. 9a) compared to normal bone marrow (n=18) or Jak2-negative CML (n=17). Moreover, Ybx1 colocalized with and bound to Jak2 in Jak2V617F-positive cells (FIG. 9b).

    Example 3. Genetic Inactivation of Ybx1 In Vitro

    [0350] Genetic inactivation of Ybx1 in BaF3 Jak2V617F cells, resulted in a dose-dependent reduction of in vitro proliferation after exposure to Jak-i Ruxolitinib (IC50 reduction: 1000 nM to 275 nM; FIG. 10). Surprisingly, Ybx1-depletion alone did not affect viability, proliferation, cell cycle activity, ROS production and DNA damage in Jak2WT and Jak2V617F cells (FIGS. 11a-11d).

    [0351] Reduction of Jak2V617F cell growth exposed to Jak-i in combination with Ybx1 inactivation could be attributed to induction of apoptosis. These findings were confirmed in Jak2V617F mutated murine (FIG. 12a) and human (FIG. 12b) cell lines, as well as primary lineage depleted Jak2V617F (Jak2.sup.+/VF) murine bone marrow cells (FIG. 12c).

    Example 4. Genetic Inactivation of Ybx1 In Vivo

    [0352] To assess Ybx1 as a potential therapeutic target in Jak2-mutated neoplasms, it was investigated if Ybx1 genetic inactivation would lead to a reduction of Jak-i persistent clones in vivo.

    [0353] A conditional knockout mouse model was generated with Exon 3 of Ybx1 flanked with loxP sites crossed with conditional Jak2V617F knock-in mice harboring an inducible Mx1-Cre recombinase. Bone marrows from Ybx1F/F Jak2V617F Mx+ and Ybx1.sup.+/+ or Ybx1.sup.+/− Jak2V617F Mx.sup.+ littermate controls (CD45.2) was transplanted in a competitive manner along with 45.1 competitor cells. Following engraftment of transplanted cells, recipient animals received plpC injections to activate Mx1-Cre and genetically delete Ybx1 with concomitant Jak-i medication by gavage.

    [0354] Recipients of Jak2V617F Ybx1.sup.+/+ bone marrow showed hyperleukocytosis, thrombocytosis, and onset of symptomatic myeloproliferation (splenomegaly). In contrast, Ybx1-deficient Jak2V617F clones did not lead to symptomatic disease within 16 weeks after transplantation (FIG. 13a).

    [0355] Peripheral blood (PB) chimerism revealed an increasing percentage of PB CD45.2/Jak2V617F positive cells, while genetic inactivation of Ybx1 resulted in suppression or loss of the Jak2-mutated clone (FIG. 13b and FIG. 13c). When followed for a total of 20 weeks, 5 of 9 recipients lost the Jak2V617F clone (<1% CD45.2 cells in PB and bone marrow, BM) while all controls notably increased PB chimerism (FIG. 14). 4 of 9 recipients of Ybx1F/F Jak2V617F Mx.sup.+ bone marrow showed counter-selection of clones with incomplete excision of Ybx1, indicating selective pressure. Consistently, development of myeloid hyperplasia in the BM (FIG. 15a) and organ infiltration (FIG. 15b) was blunted in recipients of Ybx1-depleted cells.

    [0356] In a xenograft model of Jak2V617F mutated human cells, shRNA mediated inactivation of Ybx1 followed by saline injection did not result in significant delay of disease progression (FIG. 16a, 16b). However, concurrent Jak-i treatment with Ybx1 inactivation resulted in significantly prolonged survival and loss of disease penetrance in 4/12 (30%) recipients (FIG. 16c). Notably, Jak-i alone delayed disease progression only to a minor extent (Not treated (NT)+diluent: Overall Survival (OS)=27 vs 30 days; NT+Jak-i: OS=27 vs 72 days).

    [0357] These data suggest that the evolution of persistent Jak2-clones under pharmacologic Jak inhibition can be disrupted in human and murine cells by eliminating Ybx1.

    Example 5. Safety of Ybx1 Inactivation

    [0358] To be clinically relevant, a treatment based on Ybx1 inactivation must have a suitable therapeutic index between hematopoietic stem- and progenitor cells (HSPCs) and their malignant counterpart. Fortunately, genetic inactivation of Ybx1 did not perturb steady-state haematopoiesis (FIG. 17a-e) and Ybx1-deficient HSPCs (Lineage-Sca-1.sup.+Kit.sup.+ cells; LSK) did not reveal any disadvantage in colony forming potential or lineage commitment compared to Ybx1+/+ controls (FIG. 18a-f). Thus, inactivation of Ybx1 does not impair HSPC function in vitro and in vivo, demonstrating to be clinically relevant.

    Example 6 Analysis of Ybx1-MAPK Interaction

    [0359] To further gain insight into the regulation of Ybx1 by mutant Jak2, the phosphoprofile of Ybx1 was investigated in Jak2V617F cells. Following treatment with EPO and Jak-i, Ybx1 was specifically phosphorylated at several phosphosites in Jak2V617F compared to Jak2WT cells (phosphosites: pS2, pS3, pS27, pS30, pS34, pS172, and pS174). To further characterize post-translational modulators of Ybx1, phosphoproteome analysis was performed following pharmacologic short-term inhibition of several bona fide Jak2 downstream effectors AKT, JNK, MEK, mTOR, p38, PI3K, and Jak (FIG. 19a). This identified significant changes in Ybx1 phosphorylation, and specifically the MEK/ERKi (PD0325901) responsive pS30 and pS34 residues (FIG. 19b).

    [0360] To corroborate the Ybx1-MAPK interaction, we performed Ybx1 affinity purification combined with quantitative interaction proteomics, which revealed 260 Jak2-VF specific interactors. Among these, ribonucleoproteins, mRNA splicing factors and ribosomal proteins were significantly enriched, which also explains the large number of interactors (p<0.05). Most notably, several bona fide members of mRNA splicing complexes were identified as Ybx-1 interactors in Jak-2 mutated cells (FIG. 20), that are known to interact during spliceosome formation and activation (44/134 core splicing factors that participate in several steps of dynamic spliceosome assembly, (FIG. 28a-c)). Notably, it was also discovered that Jak2 and Mapk1 interact (FIG. 20, 21a) indicating that Ybx1 may be recruited and regulated by the Jak2-Mapk1 axis. Furthermore, the interactome of Jak-i treated Jak2-VF cells confirmed that Ybx1-Mapk1 interaction is Jak-i independent (FIG. 21b, c). Collectively these data demonstrate that Ybx1 undergoes Jak2V617F-Mapk1 dependent phosphorylation and that Ybx1-Mapk1 interaction persists upon Jak-i treatment.

    Example 7 Ybx1 Phosphomutants

    [0361] In order to delineate the relevance of regulation of Ybx1 phosphorylation sites, we expressed Ybx1 phospho-mutants mimicking hyper- or hypo-phosphorylated states at conserved and relevant serine residues in Jak2-mutated cells (FIG. 29a).

    [0362] Expression of pS30A, pS34A and pS30A/34A double phosphomutants (but not pS30D, pS34D, pS30D/34D or pS170A/S172A mutants) resulted in reduction of nuclear Ybx1 translocation (FIG. 29b,c). Moreover, these modifications recapitulated the phenotype of genetic inactivation and sensitized Jak-mutated cells to Jak-i mediated cell death (FIG. 29d,e).

    [0363] Consistently, MEKi treatment alone or in combination with ruxolitinib significantly prevented Ybx1 nuclear import, an effect that was not detectable on ruxolitinib treatment alone, in murine and human Jak-mutated cells (FIG. 30a-c).

    [0364] These data suggest that MAPK signaling stabilizes nuclear Ybx1, and this notion was further supported by impaired binding of Ybx1 pS30A and pS34A-containing phosphomutants to Mapk1 (FIG. 30d).

    [0365] Collectively our data demonstrate that Ybx1 undergoes Jak2V617F-Mapk1 dependent phosphorylation and that Ybx1-Mapk1 interaction is crucial for Ybx1 nuclear translocation and persists despite Jak-i treatment.

    [0366] Moreover, these data show that inhibitors of Ybx1 phosphorylation could be used as therapeutic agent to sensitize Jak-mutated cells to Jak-i mediated cell death.

    Example 8 Transcription Regulation by YBX1

    [0367] In order to identify the transcriptional pathways controlled by Ybx1, RNA sequencing (RNAseq) analysis of murine and human Jak2-mutated cells was performed following inactivation of Ybx1 by RNA interference. Gene-ontology (GO) analysis of the differentially expressed coding genes revealed strong signatures of inflammation, chemotaxis and cytokine production but also of MAPK and ERK signaling and programmed cell death (FIG. 26a, b).

    [0368] Given the established role of Ybx1 in mRNA splicing, altered splicing events were analysed. Results showed an increase in intron retention (IR; 70% increase, 1064 events, p<0.05 & ΔPSI>0.1) compared to other alternative splicing events (FIG. 26c,d). Gene-ontology term analysis of intron retention highlighted that the involved genes (472 events, p<0.01 & ΔPSI>0.1) are enriched for RNA splicing, nonsense-mediated decay, apoptosis and MAPK signaling (FIG. 26 c,d,e). Depending on their localization, intron retention results in initiation of translation, nuclear degradation or mRNA stabilization.

    [0369] Thus, it was analysed whether ERK-signaling molecules that had increased intron retention such as Araf, Braf and Mknk1 were regulated in Jak-mutated cells (FIG. 27a). Comparison of global mRNA and protein abundance of ERK signaling proteins, showed a significant decrease of Braf (mRNA) and Mknk1 in Jak2-mutated cells (mRNA and protein) in Ybx1 targeted Jak-mutated cells (FIG. 27b). Western blot further confirms loss of Mknk1 expression in Ybx1 depleted Jak2V617F murine and human cells (FIG. 27c). Of note, Ybx1 ChIP-seq revealed no significant binding to Mknk1 in Jak2V617F cells (FIG. 31a,b). Consistently, expression of Ybx1 phosphomutants (S30A, S34A, and S30A/S34A) resulted in reduction or abrogation of Mknk1 (FIG. 31c), confirming that pS30/pS34 phosphorylation of Ybx1 in Jak2V617F cells is a critical requirement for efficient mRNA splicing and transcriptional regulation of Mknk1 transcripts and a mechanistic link to the Ybx1 dependent disruption of ERK-signaling.

    Example 9 Role of Ybx-1 in ERK-Signaling and Regulation of Apoptotic Factors Mcl1, Bim

    [0370] Phosphoproteome profiling of Ybx1-inactivated Jak2V617F cells revealed significant downregulation of phosphosites (downregulated phosphosites, 2012 in murine and 2390 in human, (FIG. 22a and FIG. 32a) with key phosphorylation changes in substrates enriched for motifs of ERK1/2, GSK3 and CDK shared between the two shRNAs (FIG. 22a and FIG. 32a). Of these, Mknk1 and Mcl-1 were identified as relevant ERK targets (FIG. 22b).

    [0371] To confirm these effects on cell signaling we investigated Jak2-dependent pathways in the presence or absence of pharmacologic Jak-i treatment. As observed on a global scale, inactivation of Ybx1 led to a considerable reduction of ERK phosphorylation in murine and human cells (FIG. 22c and FIG. 32d-f) while leaving STAT signaling largely unaffected. Concomitant pharmacologic Jak inhibition resulted in abrogation of ERK-signaling, while decreasing STAT3/5 signaling irrespective of Ybx1 inactivation. These findings indicate that Ybx1 is required for maintenance of ERK-signaling downstream of Jak2V617F.

    [0372] Likewise, genetic inactivation of Mknk1 by RNAi abrogated ERK signaling and sensitized the cells to Jak-i induced cell death (FIG. 32g-j), which links Ybx1 dependent Mknk1 expression to maintenance of ERK signaling in Jak2V617F cells.

    [0373] In the proteome of Ybx1 depleted Jak2V617F cells, affected cellular functions besides RNA splicing and processing included positive regulation of programmed cell death and apoptosis (FIG. 23). Supporting this, inactivation of Ybx1 in Jak2V617F cells by RNAi was mirrored by reduction of pro-apoptotic Mcl-1 phosphorylation, quantified on the basis of pT144 phosphorylation required for Mcl-1 stability (FIG. 24a,b). Likewise, Ybx1 loss resulted in concomitant decrease of Mcl-1 protein abundance and induction of Bim in a gene-dose dependent manner (FIG. 24c). Consistently, induction of apoptosis after Ybx1-kd and RUXOLITINIB treatment could be rescued by forced expression of Mcl1 (FIG. 24d).

    Example 10 Pharmacological Modulation of Mknk1- or ERK-Signaling in Combination with Jak2 Inhibitors

    [0374] Finally, pharmacological modulation of Mknk1- or ERK-signaling in combination with two different Jak-i resulted in induction of apoptosis in murine (FIG. 25a-c) and primary human VF-mutated CD34+BM cells but not in CD34+ healthy donor controls (FIG. 25d). This therapeutic combination of Jak2 and ERK inhibition dramatically downregulated Mknk1 protein levels in both murine and human cells indicating that Mknk1 is the primary target of this combination (FIG. 25e and FIG. 33a). Likewise, treatment of human Jak2V617F-mutated cell-lines with a combination of RUXOLITINIB and the ERK-inhibitor Trametinib (Tram) in a xenograft model reduced persistent cells (FIG. 25f) and disease penetrance (FIG. 25g) in vivo. These findings could be recapitulated using primary human Jak2V617F-mutated BM cells in a patient-derived xenograft (PDX) model of MPN. Here, in vivo treatment with a combination of RUXOLITINIB and Tram following engraftment of human cells resulted in significant decrease of hCD45 PB and BM chimerism (FIG. 25h,i) and abrogated the Jak2V617F clone in 2/5 recipients as confirmed by pyrosequencing of FACS-isolated hCD45 cells (FIG. 25l, m). To further consolidate ERK as a relevant target in Jak2 mutant cells, we performed in-depth phosphoproteome analysis following in vitro and in vivo exposure to Jak-i (ruxolitinib) in primary human Jak2-mutated cells. In primary cells, Jak-i treatment significantly altered 618 phosphosites including relevant mRNA splicing factors (FIG. 33b, c). Consistent with our previous findings, Jak-i treatment did not affect MAPK phosphorylation to a major extent (FIG. 33d), unlike NfKB and STAT signaling (FIG. 33e).

    Example 11. Inhibition of JAK2V617F Cell Growth and Proliferation by a Synergistic Combination of Splicing Factor Inhibitors and Jak Inhibitors

    [0375] Functional inhibitors of mRNA splicing and processing factors in combination with Jak inhibitors were able to synergistically inhibit JAK2V617F cell growth and proliferation. The tested functional inhibitors were GSK3326595, Indisulam, Herboxidiene and Pladienolide B.

    [0376] To this aim, murine Jak2 V617F cells were incubated with a functional inhibitor at increasing concentration in the presence and absence of Jak inhibitor Ruxolitinib (0.5 μM). After 72 hours, the incubation was stopped and growth measured by MTS assay.

    [0377] The results (FIG. 34A)-D)) showed that there was a synergistic effect between the functional inhibitor and Ruxolitinib on cell growth inhibition. Indeed the concentration of functional inhibitor (GSK3326595, Indisulam, Herboxidiene or Pladienolide B) needed to inhibit cell growth of 50% (IC.sub.50) was much lower in presence of Ruxolitinib than without. Indeed, the IC.sub.50 for GSK3326595 alone was 1000 nM, whereas it was 4 nM in the presence of Ruxolitinib. Indisulam alone was not able to reach a 50% inhibition of cell growth, whereas the IC.sub.50 in the presence of Ruxolitinib was 0.5 μM. The IC.sub.50 for Herboxidiene alone was approximatively 16 nM, whereas it was approximatively 8 nM in the presence of Ruxolitinib. The IC.sub.50 for Pladienolide B alone was approximatively 15 nM, whereas it was approximatively 7 nM in the presence of Ruxolitinib.