ALBERT

Description: Abstract 4522 Background: Conditioning regimens utilized to prepare patients for stem cell transplantation uniformly result in suppression of marrow function, leading to prolonged recovery of neutrophils and platelets. Neutropenia increases the risk of infection while thrombocytopenia and prolonged platelet engraftment time can result in increased risk of bleeding as well as morbidities associated with subsequent transfusions. The variations in duration of cytopenias are affected by multiple factors including regimen intensity and other clinical factors. Methods: In this retrospective study of consecutive patients undergoing stem cell transplantation at the University of Rochester between 2005 and 2009 for radiation—containing regimens, and between 2007 and 2009 for non-radiation containing regimens, we assessed the relationship between platelet engraftment times and source of stem cells, type of graft, and the effect of conditioning regimen intensity. Platelet engraftment was defined per standard CIBMTR definitions – platelet engraftment occurred at the first of 3 days of platelet count 〉20,000/μ L without platelet transfusions for 7 days, and/or the first day of platelet count 〉100,000/μ L. Guidelines for transfusion support stipulated a 10,000/μ L threshold for platelets in the absence of bleeding and a hemoglobin of 8.0 g/dL for red cells in the absence of other symptoms. The time to platelet engraftment and the number of platelet transfusions required prior to engraftment were analyzed in those patients with platelet nadir

Description: Abstract 2329 Age is the most important risk factor for myelodysplastic syndrome (MDS), a premalignant state that transforms into acute myelogenous leukemia in one third of cases. Indeed with normal aging, hematopoietic stem cell (HSC) regenerative potential diminishes and differentiation skews from lymphopoiesis toward myelopoiesis. The expansion in the HSC pool with aging provides sufficient but abnormal blood production, and animals experience a decline in immune function. Previous studies from our lab established that the DNA methyltransferase 3a (Dnmt3a) enables efficient differentiation by critically regulating epigenetic silencing of HSC genes (Challen et al. 2012) Interestingly, Dnmt3a expression is decreased in old HSCs, leading us to hypothesize that epigenetic changes in old HSCs may partially mimic the changes seen in Dnmt3a mutant HSCs. We propose that revealing the genome-wide DNA methylation and transcriptome signatures will lead to a greater understanding of HSC aging and MDS, which is characterized by frequent epigenetic abnormalities. In this study, we investigated genome-wide DNA methylation and transcripts by whole genome bisulfite sequencing (WGBS) and transcriptome sequencing (mRNA-seq)in young and old HSCs. For WGBS, we generated ∼600M raw reads resulting in ∼ 60 raw Gb of paired-end sequence data and aligned them to either strand of the reference genome (mm9), providing an average 40-fold sequencing depth. Globally, there was a 1.1% difference in the DNA methylation between young and old HSCs. Of these differences, 38% (172,609) of the CpG dinucleotides were hypo-methylated, and 62% (275,557) were hyper-methylated in old HSCs. To understand where the methylation changes predominantly occurred, the genome was subdivided into 77 features. Among these features, SINEs, especially Alu elements, exhibited the highest level of DNA methylation (90.94% in young HSCs, and 91.87% in old HSCs). CpG islands (CGIs) adjacent to the transcription start sites (TSS) exhibited the lowest level of DNA methylation (2.02% in young HSCs, and 2.11% in old HSCs). Interestingly strong hypo-methylation was observed in ribosomal RNA regions (68.04% in young HSCs, 59.04% in old HSCs), and hyper-methylation was observed in LINEL1 repetitive elements (88.62% in young HSCs, 90.12% in old HSCs). Moreover, the examination of differentially methylated promoters identified enrichment of developmentally important transcription factors such as Gata2, Runx1, Gfi1b, Erg, Tal1 Eto2, Cebpa and Pu.1. Additionally, we compare our ∼10,000 differentially methylation regions (DMRs, regions with clustered DNA methylation changes) with a chip-seq data set containing binding of 160 ChIP-seq analyses of hematopoietic transcription factors in different hematopoietic cells. We found significant overlaps between DMRs and transcription factor binding regions. We found DMRs which were hypermethylated showed association with differentiation-promoting Ets factors, in particular Pu.1 from a range of different blood cell types. In contrast, hypomethylated DMRs showed associations with HSC-associated transcription factors such as Scl and Gata2. Further examination of the differentially methylated gene bodies, intragenic and intergenic DMRs identified some previously noted targets for epigenetic silencing or alteration in AML and also novel transcripts including long non-coding RNAs (lincRNA) and upstream regulatory elements (URE). We found significant correlation between RNA-seq expression and DMRs within +1kb upstream of TSS. RNA-sequencing provided complementary and distinct information about HSC aging. We identified differentially expressed genes, novel RNA transcripts, differential promoter, coding sequence, and splice variant usage with age. Gene set enrichment analysis of up- and down- regulated genes, revealed ribosomal protein and RNA metabolism as critical contributors to HSC aging. In conclusion, our study marks a milestone in the mouse HSC epigenome, reporting the first complete methylome and transcriptome of pure HSC using whole-genome bisulfite sequencing and RNA-seq. These provide novel information about the magnitude and specificity of age-related epigenetic changes in a well-defined HSC population. Understanding the roles of DNA methylation and transcription in normal HSC function will allow for greater therapeutic exploitation of HSCs in the clinic. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 4440 Background: The optimal strategy to mobilize hematopoietic stem cells into peripheral blood for collection has not been defined, and some patients do not successfully mobilize. Failure to harvest the desired number of CD34+ cells results in emotional disappointment for the patient and ineffective utilization of resources. In this study we analyzed various factors influencing CD34+ yields during stem cell collection in patients being considered for autologous transplantation whose peripheral blood CD34+ analysis triggered stem cell collection (usually 〉10/μ L), and we determined thresholds for successful collection in a single apheresis session. Method: Retrospective chart review of 244 consecutive patients who underwent stem cell collection at the University of Rochester between 2005 and 2008 was conducted. Cells were collected via Cobe semi-automated protocols. For each of the patients, diagnosis, age, gender, type of infusion, marrow status, number of prior chemotherapy treatments, mobilization regimen, CD34+ count/μ l on first day, CD34+/kg after first apheresis, total time of apheresis and volume processed (liters) were recorded. The various mobilization regimens utilized were 1)GCSF, 2)CHEMO (salvage therapy with DHAP, ESHAP, RICE) with GCSF, 3)cyclophosphamide with GCSF, 4)AMD3100 (Plerixafor) with GCSF, and 5) GCSF with GMCSF. The marrow status was defined as no involvement, mild involvement (10-20%), and moderate involvement (〉20%). Another factor considered was the number of prior chemotherapy regimens, and the total number of apheresis cycles required to reach targeted yields was recorded. All statistical analyses were conducted using SAS 9.2. All tests were two-sided with p-values ≤ 0.05 considered significant. Result: Analysis revealed a positive linear relationship between the log of initial CD34+ counts and the log transformed number of CD34+ cells/kg on the first day of apheresis (R2= 0.57, p 〈 0.001). Multivariate analysis suggested that both CD34+ count per μ L blood prior to first apheresis (p 〈 0.001) and time on the machine (p = 0.08) were positively associated with CD34+ end yield after adjusting for age and gender. A similar analysis on all the mobilization regimens revealed statistical significance for a higher initial CD34 cell count/μ l predicting a higher apheresis yield (p 〈 0.001). Also, for each of the major diagnoses (AL, HL, MM. NHL), there was a positive relationship between peripheral CD34+ cells/μ l prior to apheresis and CD34+ cells/kg after first apheresis yield (p 〈 0.001) Another objective of this study was to see if there is a threshold number of CD34+ cells/μ l blood that would predict for lymphoma patients (combined HL, NHL) that at least 4 × 106 CD34+ cells/kg would be reached on the first day of apheresis, and a similar threshold for the multiple myeloma patients (MM) to reach 6 × 106 CD34+ cells/kg on first day. Receiver operating characteristics (ROC) curves were used to determine an optimal initial CD34+ count cutoff and the odds ratio of achieving such a threshold between the two groups was assessed using logistic regressions. Based on the cutoff of 48.3 cells/μ L, lymphoma patients were classified into two groups depending on their initial CD34+ count greater or less than this cutoff. The odds of reaching at least 4 × 106 CD34+ cells/kg on the first day of apheresis when the initial CD34+ count was greater than 48.3 cells/μ L were 24.4 times than those with the initial CD34+ count less than the cutoff (p

Description: Abstract 1700 To better understand molecular bases of MDS pathogenesis, we performed a genome-wide CHIP-Seq analysis of H3K4me3, a histone mark associated with gene activation. In MDS CD34+ cells (N=4), 36 genes showed higher levels of promoter-H3K4me3 compared to controls. 10 of 11 randomly selected genes from these 36 showed increased mRNA expression in MDS CD34+ cells (N=100), supporting the positive correlation between expression and increased promoter H3K4me3. Pathway Analysis indicated that a majority of these genes are involved in innate immunity signal and NF-kB activation. This was validated by increased phospho-p65 in MDS bone marrow CD34+ specimens (N=15). Knock-down of 4 of these genes (C5AR1, FPR2, PTAFR and TYROBP) in OCI-AML3 cells resulted in reduction of p-p65. The observation of innate immunity signal activation and epigenetic deregulation led us screen key innate immunity activators, the Toll-like Receptor (TLR) family genes, and histone methylation regulators in MDS CD34+ cells. Among 8 TLRs and 24 histone methylation regulators, TLR1, 2 and 6, and Jmjc-domain histone demethylase JMJD3 were found to be significantly overexpressed in MDS CD34+ cells compared to control counterparts. This is of biological significance because TLR1 and 6 form functional hetero-dimmers with TLR2. Also, JMJD3 expression can be activated by TLR-NFkB in macrophages. To study TLR2 activation in HSC, we treated CD34+ cells with TLR2 agonists. This led to increased expression of JMJD3, supporting the biological function of TLR2 signal in HSC. We observed that JMJD3 knockdown in OCI-AML3 cells led to reduced expression of innate immunity genes (N=7), accompanied with changes of promoter H3K27me3 and H3K4me3, suggesting that JMJD3 forms a positive feedback to perpetuate innate immunity pathway. To further study the role of TLR-JMJD3 in MDS, we performed capture deep sequencing in 40 MDS bone marrow mononuclear cells (TLR1, 2, 4 and 6, JMJD3, UTX, UTY and JMJD1A). Seven different rare SNP in the coding regions of JMJD3, UTY, JMJD1A, and TLR2 were identified in 5 patients. Among them, one SNP of TLR2 causes a missense mutation, changing a conserved hydrophobic Phe217 to a hydrophilic Ser. We then analyzed the presence of TLR2 F217S in 150 MDS samples by Sanger sequencing. TLR2 F217S was observed in 17 (11%) patients. To evaluate the somatic nature of this alteration, CD3+ cells from 15 corresponding patients were sequenced and only two (13%) CD3+ cell samples carried TLR2 F217S. In the available 9 CD34+ cDNA samples, TLR2 F217S was observed in 8 (90%). We then expressed wild-type or F217S TLR2 in 293T cells, a cell line without endogenous TLR2 expression. Reporter assays indicated that in the absence of TLR2 agonist wildtype and F217S mutant TLR2 led to similar levels of NF-kB activation, whereas F217S led to an increased NF-kB activation compared to wildtype at the presence of TLR2 agonists. F217S also led to increased activation (phospho- and polyubiquitin-) of IRAK1, a key signal mediator for TLR signaling. These results suggest that TLR2 F217S led to more robust innate immunity signal activation when stimulated by agonists. We further studied the impact of TLR2 activation on hematopoietic differentiation. Colony formation of CD34+ cells indicated that TLR2 agonists led to decreased number of erythroid colonies, which was confirmed by flowcytometry that demonstrated TLR2 agonist treatments could reduce the number of CD71 high/HLADR low featured erythroid precursors. To examine the effect of targeting TLR2-JMJD3 in primary MDS cells, we transduced MDS CD34+ cells with shRNAs. In 4 CD34+ cases isolated from lower risk MDS, 3 transduced with JMJD3 shRNA and 4 transduced with TLR2 shRNA had increased ratio of erythroid colonies. In average, JMJD3 and TLR2 shRNA transduction led to a 50% increase in erythroid colonies. This was accompanied by increased expression of genes positively involved in differentiation of erythroid lineage (GYPA, GATA1 and EPOR). Finally, CD34+ cell mRNA expression levels of four genes in this study (NCF2, AQP9, MEFV and TLR1) were associated with overall survival of patients. Taken together, these studies highlight the implication of the deregulation of TLR2-JMJD3 mediated innate immunity signals in MDS pathogenesis and suggest that intervention of this pathway may have therapeutic potential in MDS. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 1188 Chromatin regulators, including both Trithorax-group (trxG) and Polycomb-group (PcG) families, have been reported to maintain hematopoietic stem cells (HSC) self-renewal and differentiation. However, their primary targets in HSC remain to be fully identified. Moreover, those histone modifiers have been recently found to control longevity in C.elegans, which leads us to further investigate their roles in HSC aging. HSCs increase in number, decrease in regeneration capacity, and exhibit a myeloid biased differentiation with age. In this study, we profiled global mRNA (RNA-seq) and chromatin changes (ChIP-seq) in highly purified young (4 month) and old (24 month) murine bone marrow-derived HSCs (SP-KSL-CD150+). One key challenge in this study is that mouse HSCs represent less than 0.01% of all bone marrow cells. Thus, we first developed and optimized a method for successful application of ChIP-seq to a limited number of cells (

Description: Background: Prior studies using microarray platforms have shown alterations of gene expression profiles (GEPs) in MDS CD34+ marrow cells related to clinical outcomes (Sridhar et al, Blood 2009, Pellagatti et al, JCO 2013). Given the increased sensitivity and accuracy of high-throughput RNA sequencing (RNA-Seq) (Mortazavi et al, Nat Meth 2008, Soon et al, Mol Syst Bio 2012) for detecting and quantifying mRNA transcripts, we applied this methodology for evaluating differential gene expression between MDS and normal CD34+ marrow cells. Methods:RNA was isolated from magnetic bead affinity-enriched CD34+ (〉90%) marrow aspirate cells (Miltenyi Biotec, Auburn, CA) and amplified using the Smarter Kit (Clontech, Mt View, CA). The amplified product (ds DNA) was fragmented to a size distribution of ~200-300bp using the E220 Focused Ultrasonicator (Covaris Inc, Woburn, MA). End repair, adapter ligation and PCR amplification were performed using the NEBNext Ultra RNA library prep kit for Illumina (New England Biolabs, Ipswich, MA). The indexed cDNA libraries were sequenced (paired end, 100bp) on an Illumina HiSeq2000 platform with median read counts of 69 million. The sequences were aligned to Human Reference sequence hg19 using DNAnexus mapper with gene detection focused on known annotated genes. The differential expression was analyzed using edgeR. DAVID and Ingenuity IPA programs were used for pathway analyses. Gene Set Enrichment Analysis (GSEA) was used to identify biologic processes in our dataset present across phenotypes. Results: Correlations of RNA-Seq data from unamplified to amplified transcripts demonstrated high fidelity of transcripts obtained (Pearson and Spearman R2 = 0.80). After filtering samples for adequate read counts, 12,323 genes were evaluated. Differential expression analysis yielded 719 differentially expressed genes (DEGs) in MDS (n=30) vs normal (n=21) with FDR 50% of the patients. Hierarchical cluster analysis using these DEGs confirmed clear separation of MDS patients from normals, with 2 differential expression clusters—one region overexpressed and one underexpressed. A distinctive trend toward clustering of the patients was seen which related to their IPSS categories and marrow blast %. In functional pathway analysis of the 2 distinctive gene clusters which distinguished MDS from normal, the underexpressed MDS DEGs demonstrated enrichment of inflammatory cytokines, oxidative stress and interleukin signaling pathways, plus mitochondrial calcium transport; whereas the MDS overexpressed DEG cluster showed enrichment of adherens junction/cytokeletal remodeling, cell cycle control of chromosome replication and DNA damage response pathways. Using GSEA analysis, significantly increased numbers of genes in MDS vs normal, common to those in gene sets present within curated public databases, were involved with TP53 targets and mTOR signaling pathways. Conclusions: Our study demonstrated that RNA-Seq methodology, a high-throughput and more comprehensive technique than most gene expression microarrays, was capable of showing significant and distinctive differences in gene expression between MDS and normal marrow CD34+ cells. Specific clustering of the DEGs was demonstrated to distinguish patient subsets associated with their major clinical features. Further, the stringently identified DEGs shown to be engaged in functional pathways and biologic processes highly relevant for MDS were extant within the patients’ CD34+ cells. These transcriptomic data provide information complementary to exomic mutational findings contributing to improved understanding of biologic mechanisms underlying MDS. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 848 DNA methylation is an epigenetic modification in vertebrate genomes critical for regulation of gene expression. DNA methylation is catalyzed by a family of DNA methyltransferase enzymes, Dnmt1, Dnmt3a, and Dnmt3b. Dnmt1 is primarily a maintenance methyltransferase, targeting hemimethylated DNA to reestablish methylation marks after DNA replication. Dnmt3a and Dnmt3b are de novo methyltransferases that are essential for normal embryonic development. In humans, somatic mutations in DNTM3A have been identified in ∼20% of human acute myeloid leukemia (AML) and ∼10% of myelodysplastic syndrome (MDS) patients, but the mechanisms through which these mutations contribute to pathogenesis is not well understood. Congenital mutations in DNMT3B can cause ICF (immunodeficiency, centromeric instability, and facial anomalies) syndrome. These patients exhibit chromosomal instability due to heterochromatin decondensation and demethylation of satellite DNA. Our group has recently reported that Dnmt3a is essential for HSC differentiation (Challen Nature Genetics, 2011). Conditional knockout of Dnmt3a (Dnmt3a-KO) resulted in HSCs that could not sustain peripheral blood generation after serial transplantation, but phenotypically defined HSCs accumulated in the bone marrow. Dnmt3b is also highly expressed in HSCs, but its contribution to gene regulation in hematopoiesis is unclear. Here, we examine the role of Dnmt3b, alone and in combination with Dnmt3a KO, in the regulation of hematopoiesis. We performed conditional ablation of Dnmt3b, as well as Dnmt3a and Dnmt3b simultaneously using the Mx1-cre system. Unlike the Dnmt3a-KO HSCs, loss of Dnmt3b had a minimal impact on blood production. Even after several rounds of transplantation, 3b-KO HSCs performed similarly to WT controls. However, the Dnmt3ab-dKO (double knock-out) peripheral blood contribution was quickly and severely diminished, accompanied by a dramatic accumulation of Dnmt3ab-dKO HSCs in the bone marrow (Figure 1). The dKO phenotype paralleled that of the 3a-KO HSC, but was more extreme. To examine the impact of loss of Dnmt3a and -3b on DNA methylation in HSCs, we performed Whole Genome Bisulfite Sequencing (WGBS) on Dnmt3a-KO, Dnmt3ab- dKO and control HSCs. As we previously found with more limited DNA methylation analysis, loss of Dnmt3a led to both increases and decreases of DNA methylation at distinct genomic regions (Challen, Nature Genetics, 2011). However, loss of both Dnmt3a and -3b primarily resulted in loss of DNA methylation that was much more extensive than that seen in the 3a-KO. In addition, RNAseq of the mutant HSCs revealed increased expression of repetitive elements, inappropriate splicing, and truncation of 3ÕUTRs. To gain insight into the accumulation of Dnmt3ab-dKO HSCs in the bone marrow, we performed a time course analysis of the proliferation and apoptosis status of the HSCs. Every four weeks after transplantation of HSCs, we sacrificed a cohort of 3 control and 3 dKO mice, counted donor derived HSCs in the bone marrow, and analyzed their Ki67 and Annexin V expression. Up to 12 weeks post-transplant, no significant differences are seen in the expression of Ki67 or Annexin V. These data show that while Dnmt3b alone has minimal impact on DNA methylation in HSCs, Dnmt3a and -3b act synergistically to effect gene expression changes that permit HSC differentiation. In the absence of both of these de novo DNA methyltransferases, there is an immediate and extreme shift toward self-renewal of dKO HSCs. The Ki67 and Annexin V expression patterns suggest that a lack of de novo DNA methylation does not affect the proliferation or apoptosis of HSCs, but instead that the accumulation of HSCs and lack of peripheral blood contribution is primarily due to an imbalance between self-renewal and differentiation. By understanding the mechanisms through which Dnmt3a and -3b exert these effects, we should identify genes that are critical for normal hematopoietic differentiation. These genes may serve as targets for therapeutic intervention in malignancies caused by defective DNA methyltransferases. Figure 1: HSC composition of the bone marrow after secondary transplantation of control (left) and double Dnmt3a/3b KO (right) HSCs. After control HSC transplantation, HSCs comprise ∼0.01% of whole bone marrow. After transplantation of dKO HSCs, phenotypically-defined HSCs (KLS CD34–Flk2–) comprise ∼0.48% of bone marrow. Figure 1:. HSC composition of the bone marrow after secondary transplantation of control (left) and double Dnmt3a/3b KO (right) HSCs. After control HSC transplantation, HSCs comprise ∼0.01% of whole bone marrow. After transplantation of dKO HSCs, phenotypically-defined HSCs (KLS CD34–Flk2–) comprise ∼0.48% of bone marrow. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 291 Although cytogenetic abnormalities are common in MDS, search for genetic alterations has been less informative with few prevalent abnormalities thus far known. To identify genes aberrantly activated in MDS, we developed a novel approach based on chromatin immuno-precipitation combined with massive parallel sequencing (CHIP-Seq) using the Solexa 1G sequencing technology. To our knowledge this is the first example of the use of this technology in primary human samples. For CHIP analysis we used an antibody against H3K4me3 (histone-H3-lysine 4-trimethylation). H3K4me3 is a chromatin mark of gene activation that localizes to active gene promoter regions. CHIP-Seq was performed in CD34+, CD34 neg cells and whole bone marrow (WBM) from 6 patients with MDS and 4 normal controls. In total 30 samples were sequenced. Patients samples were obtained at the time of initial referral at MDACC and were sorted immediately using standard separation procedures. When compared to normal controls for each cellular compartment, we identified 36, 156 and 32 potential active gene promoters associated with H3K4me3 in CD34+, CD34 neg cells and WBM respectively. Of importance, gene promoter regions identified did not overlap among the different cellular compartments analyzed (differences were observed comparing normal vs MDS but also among different MDS compartments), indicating that chromatin structure and gene expression profiles are aberrant and distinct in non-CD34+ cells that may also contribute to the pathobiology of MDS. Here we focus on H3K4me3-associated gene promoters in CD34+ cells. To confirm the results obtained with the CHIP-seq approach, we studied the expression levels of the top 9 CHIP-Seq identified genes in an independent cohort of in CD34+ cells obtained from 54 MDS at the time of initial diagnosis. Patient characteristics have been previously reported (Leukemia, in press): 11 (20%) low risk, 20 (37%) int-1, 15 (27%) int-2 and 8 (14%) high risk by IPSS. We confirmed gene expression up-regulation of 7 (C5AR1, FPR1, FPR2, AQ9, FYB, FCAR, IL8RA) of 9 genes detected by CHIP-Seq. Using Ingenuity Pathway Analysis of the 36 genes identified in CD34+ cells revealed NF-κB as central activated knot in CD34+ cells. This was confirmed by phospho-p65 immuno-staining in primary cells. Furthermore up-regulation of all 10 NF-κB activation associated genes was confirmed in MDS CD34+ cells by Q-RT-PCR. Transfection of OCI-AML3 cells with a siRNAs cocktail targeting 4 of the CD34+ NF-κB activation genes dramatically repressed NF-κB activation as well as expression and promoter NF-κB association of JMJD3 gene, a known NF-κB transcriptional target. JMJD3 encodes a Jmjc-domain K27me3 demethylase, which positively regulates H3K4me3. We further characterized expression levels of 17 known histone demethylases known in human in 35 patients with MDS and identified JMJD3 as the only histone demethylase overexpressed in MDS CD34+ cells. siRNA targeting JMJD3 reduced expression and promoter H3K4me3 levels of several CHIP-Seq detected MDS- CD34+-NF-κB activation genes. Finally expression profile of JMJD3 and the panel CD34+-NF-κB activation genes in the 54 patients with MDS indicated that expression levels were consistently overexpressed in patients with higher-risk (high and int-2) disease compared to patients with lower (low and int-1) risk disease. In view of the known antiapoptotic and proliferative role of the NF-κB pathway, this data indicates that expression of upstream and downstream modulators of NF-κB signaling, regulated at the chromatin level by JMJD3, have a role in MDS progression and could serve as therapeutic targets. Through this novel in vivo CHIP-Seq analysis, we demonstrated that a positive regulatory loop exists in MDS CD34+ cells. This loop contains JMJD3 promoted gene activation through positive regulation of H3K4me3, which leads to NF-κB signaling activation, and then further promotion of JMJD3 expression and activation of the whole signaling cascade. Our study also demonstrates that in vivo CHIP-Seq can be used to discover disease specific targets. Disclosures: No relevant conflicts of interest to declare.