ALBERT

Description: LK, SS Equal contributions VG, JTP Equal credit as senior authors Prolyl hydroxylase 2 (encoded by EGLN1), the principle negative regulator of HIF-1 and HIF-2 (encoded by EPAS1), is an iron dependent enzyme (Kaelin WG et al. Mol Cell, 2008); thus iron deficiency can augment hypoxic responses (Zhang X et al Blood Cells Mol Dis, 2014). Because EPAS1 mRNA has a 5' iron regulatory element, iron deficiency may inhibit the translation of HIF-2α and downregulate its target genes (Sanchez M et al Nat Struct Mol Biol, 2007). The Partnership for Anemia: Clinical and Translational Trials in the Elderly conducted a trial to determine the efficacy of IV iron in patients with borderline iron deficiency (Price E et al. Blood Cells Mol Dis, 2014). We measured expression levels of HIF1A, EPAS1, EGLN1 and HIF target genes in platelets and granulocytes. Nineteen patients were treated with 200 mg IV iron sucrose weekly for 5 weeks. Blood was obtained at screening, one week after the second dose of IV iron and 8 weeks after completion of iron therapy. Granulocytes and platelets were isolated and their RNA reverse transcribed. Transcripts were quantified by real-time qT-PCR, and normalized to HPRT gene. The threshold cycle for the expression of each gene was calculated and compared to 5 healthy controls. We found a positive correlation between granulocyte and platelet mRNA of HIF1A, EPAS1 and EGLN1, and these relationships achieved a one-sided significance level of 0.1 or stronger for most data points for HIF1A and EPAS1 (Table 1). There was also a positive correlation for granulocyte and platelet mRNA of 5 of 7 hypoxia response genes (P of 0.1 or stronger in many comparisons; the strongest being for VEGF and PDK1). Table 1. Spearman correlation rho between granulocyte and platelet mRNAs (N = 17). Gene Screen TV2 FU2 Regulators of the hypoxic response HIF1A .52** .21 .40* EPAS1 .40* .54** .67** EGLN1 .28 .30 .09 Responders to the hypoxic response GLUT1 .44** .00 .07 HK1 .25 .20 .36* PDK1 .24 .55** .41* VEGF .68** .33* .62** FOXO3 -.10 .22 .45** BNIP3 .03 -.19 -.08 BNIP3L .15 -.05 .20 *one-sided P

Description: First and second authors deserve equal credit In each female somatic cell, most genes on either the maternal or paternal X-chromosomes are randomly inactivated at early embryogenesis. This remains constant in the progeny of these cells throughout life. The XIST gene is a principal regulator of X-chromosome inactivation expressed in inactive female X-chromosomes. It encodes a long noncoding RNA which leads to recruitment of histone H3 lysine 27 trimethylation (H3K27me3)-methylase, and eventual inactivation of transcripts of most X-chromosome genes on inactive X-chromosomes. The clonality of human tumors is central to our understanding of malignant and premalignant diseases such as polycythemia vera (PV) and essential thrombocythemia (ET). Clonal populations can be detected by expression of X-chromosome genes in females. In a normal female embryo, there are eight cells at the time of random embryonic X-chromosome inactivation; thus X-chromosome allelic ratios vary from female to female. This ratio is stable in time, identical in all blood cell lineages, and the same for any interrogated X-linked gene in females heterozygous for other X-chromosome inactivated genes (Prchal, J Exp Med, 1996). Based on polymorphisms for 5 X-chromosome genes (MPP1, FHL1, IDS, BTK, and G6PD), we developed a quantitative transcriptional clonality assay and found that allelic usage ratios are the same for each marker and in each myeloid cell line (Swierczek, Blood 2009). In two PV (P1, P3) and one ET (P2) females, we observed expression of both alleles of either the IDS or G6PD genes in platelets and granulocytes, whereas other X-chromosome genes for which these women were heterozygous were expressed only a single allele. We then analyzed their X- chromosome transcripts in clonogenic assays of early erythroid progenitors (BFU-E). Each of these females had some BFU-E which expressed both X-chromosome genes. Reactivation of inactivated X-chromosomes was reported in some human breast cancers (Richardson, Cancer Cell, 2006) and was associated with overexpression of a small subset of X-chromosomal oncogenes and/or tumor suppressor genes (Thakur, Mole Cancer Res, 2007). Further, conditional Xist deletion inhematopoietic stem cells of female mice resulted in global X-chromosome reactivation and complete penetrance of a highly aggressive PV/ET-like syndrome that progressed to fatal lymphoma, acute leukemia or myelodysplastic syndrome (Yildirim, Cell,2013). Our previous, unbiased, whole exome sequencing of 31 JAK2V617F positive PV patients identified 87 somatic and germline mutations, with most patients having two or more somatic mutations (Wang, Leukemia, 2014). The majority of mutated genes were epigenetic modifiers; i.e. ASXL1, DNMT3A, TET2, SF3B1 and PDE4C. We interrogated the effects of epigenetic modifiers on methylation of DNA of CD34 cells and granulocytes by whole genome methylome analysis and demonstrated inconsistent, scattered hypomethylation changes in autosomes of both males and females. However, there were large regions of the hypomethylated X-chromosome genome in many but not all PV and ET females. We examined transcript levels of the hypomethylated genes by QT-RTPCR and found augmentation of transcripts of many but not all hypomethylated genes in granulocytes and platelets in these PV females. However, XIST transcripts were not decreased. The ChIP analysis of clonal granulocytes revealed decreased level of H3K27me3 at hypomethylated regions of biallelically expressed G6PD and IDS genes. We then performed unbiased transcriptome analysis by RNA seq of 24 PV and ET males and females (Figure). We observed more overexpressed genes (75%) on X chromosome than autosomes (66%).The X-cromosome encoded miR-221 (reported dysregulated in ET; Navarro, Blood Cancer Journal, 2016) and ARHGAP6 were overexpressed in PV and ET but more in females than males (Figure). The miR-221 targets SOCS1 and SOCS3 genes, negative regulators of the JAK2/STAT pathway, and the ARHGAP6 gene is a component of PDGF signaling pathway. It remains to be determined which hypomethylated X-chromosome genomic regions have functional activity in hematopoiesis in some PV females but miR-221 and ARHGAP6 are potential candidates. These data may contribute to a better understanding of gender differences of the PV phenotype and PV morbidity and mortality. One such an example is higher prevalence of Budd-Chiari syndrome in females. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2826 Polycythemia vera (PV) is a clonal blood disorder arising from a multipotent hematopoietic stem cell (HSC) associated in ∼95% of patients with the acquired somatic mutation, JAK2 V617F. Although important for the PV phenotype, we and others demonstrated that the JAK2 V617F mutation is not the initial and causative somatic event in PV pathogenesis. One of the major challenges of studying the molecular events in PV is to isolate and expand the disease-initiating HSC clones in vitro. To overcome this hurdle, we have utilized the recently developed induced pluripotent stem cell (iPSC) technology to generate disease-specific iPSC lines that preserve the genetic identities of patient HSC clones. We previously demonstrated that interferon (IFN) a is the only therapy that converts PV hematopoiesis from clonal to polyclonal (Liu, Blood 2003). A female patient with typical PV and a high allele burden (99%) of JAK2 V617F, and ∼1% of wild-type JAK2 was treated with peg-IFNa. JAK2 allele burden decreased to ∼65%, yet the majority of her myeloid cells remained clonal. Using her blood and bone marrow progenitors as well as blood samples from other PV patients, we generated dozens of iPSC clones by retroviral or episomal vectors with several distinct JAK2 genotypes (see Table below). We examined the erythroid differentiation of 6 representative PV-iPSC lines and normal control iPSCs. The hematopoietic progenitor cells (HPCs) derived from JAK2 V617F iPSCs had enhanced erythropoiesis compared to wild-type JAK2 iPSC cells. Additionally, some EPO-independent BFU-Es also formed from homozygous JAK2 V617F iPSCs, the hallmark of PV erythropoiesis. Using a quantitative X-chromosome transcriptional assay (Swierczek, Blood 2008), we examined the clonality of the iPSC clones (with and without the JAK2 mutation) derived from this single female patient and showed the same single X-chromosome usage in all clones as in her native PV granulocytes and platelets. These data indicate that epigenetic X-chromosome silencing is not reverted in the process of generating iPSC clones. Whether these two JAK2 V617F-negative iPSC lines originated from the same PV clone or from dormant normal HSC cells cannot be yet discerned but their EPO sensitivity is currently under analysis. Analyses of whole exome sequencing of these iPS clones as well as their germ-line control are currently underway. Additionally, whole genome and epigenome analyses and high density expression array of these iPSC clones would further characterize the clonal evolution of PV. These data underscore the heterogeneity of somatic mutations within single PV patient. These studies will lead to a better understanding of the genetic lesions in PV-initiating clones. (Note: The last two authors are both considered senior authors for this work)Table.Human iPSC lines from a female PV patient and healthy donorsRepresentative iPS clone (# of clones characterized)DonorParental cell typeJAK2 WT alleleJAK2 V617F alleleKaryotypePVB1.4 (3)PVPB MNC0246, XXPVB1.1 (4)PB MNC1247, XX, +der(1;9)(q10;p10)PVB1.11 (2)PB MNC2046, XXPVM1.1 (2)BM MSC2046, XXBC1 (〉5)Healthy donorBM CD34+2046, XYPB: peripheral blood; BM: bone marrow; MSC: mesenchymal stem cell. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 5167 The myeloproliferative neoplasms (MPNs) polycythemia vera (PV), essential thrombocytosis (ET), and primary myelofibrosis (PMF) are caused by an unregulated clonal proliferation of cells derived from a pluripotent stem cell. The exon 14 JAK2 V617F somatic mutation and cMPL mutations in codon 515 of exon 10 have provided excellent diagnostic tools to distinguish MPNs from other diseases with similar clinical features. JAK2 V617F mutations are present in ∼95% of PV patients and ∼50% of ET or PMF patients. Other somatic mutations have also been found in JAK2 exon 12 in patients with PV. Recently, a new somatic mutation of JAK2 in exon 14, L611V, was reported in 3 out of 168 PV patients in cis with JAK2 V617F (Cleyrat Leukemia 2010). Two cMPL mutations, cMPL W515L and cMPL W515K, have been identified in MPNs. The cMPL W515L mutation is reported to be present in ET and PMF patients, while cMPL W515K in some PMF patients. We designed and validated a new rapid and sensitive allele-specific quantitative PCR assay (AS-qPCR) for determination of the JAK2 L611V mutation. In our design, we included a locked nucleic acid and a mismatch in the allelic specific primers in order to enhance allelic discrimination in the PCR reaction. The assay was sensitive to mutant frequencies of 〉2% in the background of wild-type alleles. Allelic discrimination of the assay is 14 cycles for wild type and 11 cycles for mutant alleles. We studied 584 patients referred with clinical diagnoses of MPN belonging to two different groups: 1) 390 patients, who, based on clinical and laboratory evaluations, had diagnoses of PV (n=74), ET (n=60), MF (n=30), other unspecified MPNs (n=189), and other non-PV patients with erythrocytosis (n=37); 2) DNA from patients referred to ARUP Laboratories with otherwise unspecified clinical diagnoses of MPN who were found to be 196 JAK2 V617F-positive but appropriate clinical and other testing was not available for classification. The first group was tested for the JAK2 V617F, JAK2 L611V, cMPL W515L and W515K mutations. Out of 390 patients, 37% (148/390) were JAK2 V617F positive (over 〉90% of PV patients), 9 (ET or PMF) patients carried the cMPL W515L mutation, and 1 PMF patient had 93% allelic burden for the cMPL W515K mutation. While cMPL W515L mutation was not previously reported in PV, one PV patient was positive for both the cMPL W515L and JAK2 V617F mutations. Of the 10 cMPL W515L mutations, two MF patients carried JAK2 V617F mutations concurrently. No JAK2 L611V mutation was detected in this cohort. The additional 196 JAK2 V617F-positive samples from the patients with unspecified MPNs were also screened for the JAK2 L611V mutation and none were positive. We conclude that the rarity of the JAK2 L611V mutation does not justify its routine screening. We suggest that those patients who fulfill the WHO criteria for PV should be further evaluated by more productive laboratory studies such as screening for JAK2 exon 12 mutations, clonality studies in females and if negative assays for germline mutations such as EPOR and VHL gene mutations. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 4968 Elucidation of the molecular aberrations underlying the development of myeloproliferative neoplasms (MPNs) has progressed rapidly during the previous years. In addition to the JAK2V617F mutation, which is found in over 90% of patients with polycythemia vera (PV) and in around 50% of patients with essential thrombocythemia (ET) and primary myelofibrosis (PMF), several novel markers have recently been described. These include mutations in the Ten-Eleven-Translocation-2 (TET-2) gene, as well as overexpression of the hematopoietic transcription factor nuclear factor erythroid-2 (NF-E2). While the individual frequency of these abnormalities is well known, so far, studies that evaluate the relationship between these markers and their correlation with hematological parameters have not been conducted. The International Myeloproliferative Research Consortium (MPD-RC) has instituted a Tissue Bank (MPD-RC Trial 106) allowing the collection of MPN patient samples from member institutions in the US and Europe. Using this resource, we have investigated the relationship among the following molecular markers and hematological parameters in a series of 66 MPN patients: TET-2 mRNA expression, JAK2V617F allele burden, NF-E2 mRNA expression, granulocyte clonality (female patients) as well as hematocrit, hemoglobin concentration, platelet numbers and leukocyte counts. Our crossectional cohort included 37 PV patients, 14 ET, 4 PMF and one post PV MF patient. Sixtyeight percent of the patients were treated, medication including hydroxyurea, anagrelide, ASA, interferon and one patient treated with imatinib mesylate. We acknowledge that treatment may affect the molecular markers being investigated. In addition, the desired effect of treatment on hematological parameters may not be paralleled by a similar effect on molecular markers. Therefore, investigation of treated patients may not reveal biological relationships present in untreated diseases states. We thus tested the effect of cytoreductive treatment on the expression of molecular markers by comparing treated and untreated patients in our cohort. The JAK2V617F allele burden was significantly lower in treated patients than in untreated patients (p = 0.008). The same difference was noted when PV patients were analyzed alone (p = 0.006) In contrast, neither TET-2 nor NF-E2 mRNA expression were affected by treatment. In the 66 MPN patients evaluated, a significant inverse Spearman correlation of -0.25 was noted between NF-E2 mRNA expression and hemoglobin concentration (p = 0.05). This relationship was more pronounced when PV patients were analyzed alone (Spearman's r = -0.41, p = 0.01). In addition, a correlation of 0.51 between JAK2V617F allele burden and NF-E2 mRNA expression was noted (p = 0.0007). Occurrence of the recently discovered TET-2 gene mutations, which are present in around 15% of MPN patients, was measured indirectly by determining TET-2 mRNA expression. None of the other markers and parameters assessed was significantly correlated with the amount of TET-2 mRNA expressed. We report a previously unrecognized inverse relationship between NF-E2 mRNA expression and hemoglobin concentration. These data complement several recent findings in MPN patients. While NF-E2 is overexpressed in granulocytes of all three MPN subtypes, PV, ET and PMF, its overexpression in CD34+ hematopoietic stem cells has so far only been observed in PMF. We have recently shown that ex vivo overexpression of NF-E2 in CD34+ hematopoietic stem cells drastically reduces erythroid colony formation. It was previously noted that mean JAK2V617F allele burdens in PMF patients are higher than in PV or ET. Here, we confirm the previously noted positive correlation between JAK2V617F allele burden and NF-E2 mRNA expression. Our data therefore suggest that high levels of JAK2V617F in PMF patients directly or indirectly augment NF-E2 overexpression, which may contribute to the anemia of Primary Myelofibrosis. Disclosures No relevant conflicts of interest to declare.

Description: RUNX1 encodes a DNA-binding α subunit of the core-binding factor, a heterodimeric transcription factor. RUNX1 is a master regulatory gene in hematopoiesis and its disruption is one of the most common aberrations in acute leukemia. Inactivating or dominant-negative mutations in the RUNX1 gene have been also identified in pedigrees of familial platelet disorders with a variable propensity to develop acute myeloid leukemia (FPD/AML). We performed analysis of hematopoiesis from 2 FPD/AML pedigrees with 2 distinct RUNX1 germline mutations, that is, the R139X in a pedigree without AML and the R174Q mutation in a pedigree with AML. Both mutations induced a marked increase in the clonogenic potential of immature CD34+CD38− progenitors, with some self-renewal capacities observed only for R174Q mutation. This increased proliferation correlated with reduction in the expression of NR4A3, a gene previously implicated in leukemia development. We demonstrated that NR4A3 was a direct target of RUNX1 and that restoration of NR4A3 expression partially reduced the clonogenic potential of patient progenitors. We propose that the down-regulation of NR4A3 in RUNX1-mutated hematopoietic progenitors leads to an increase in the pool of cells susceptible to be hit by secondary leukemic genetic events.