Vox Sanguinis (2016) 110, 79–89 © 2015 International Society of Blood Transfusion DOI: 10.1111/vox.12303
ORIGINAL PAPER
Epigenetic and molecular signatures of cord blood CD34+ cells treated with histone deacetylase inhibitors D. Gajzer,1* J. Ross,1* L. Winder,1 S. Navada,1 W. Zhang,2 L. Silverman1 & P. Chaurasia1 1
Division of Hematology/Medical Oncology, Department of Medicine, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA 2 Department of Medicine Bioinformatics Core, Icahn School of Medicine at Mount Sinai, New York, NY, USA
Background and Objectives Epigenetic modifications tightly regulate the gene expression and cellular function of haematopoietic stem cells. Histone deacetylase inhibitors (HDACIs) alter the gene expression profile of cord blood (CB) CD34+ cells by controlling the genes involved in chromatin modification, thereby influencing the self-renewal, maintenance and expansion of haematopoietic stem and progenitor cells (HSPCs). Materials and Methods The class I and II HDACIs, valproic acid and scriptaid, were utilized to expand CB-CD34+ cells ex vivo. The gene profiling was performed on HSPC using Illumina microarray, GeneGO MetaCoreTM and Ingenuity pathway analyses. The molecular analyses were performed using Q-PCR and Western blotting. Results Each HDACI treatment of CB-CD34+ cells created unique epigenetic and molecular signatures that governed chromatin modification required for cellular and functional behaviour of stem cells. GeneGO MetaCoreTM and Ingenuity pathway analyses established the molecular understanding of epigenetically regulated HSPCs in the presence of scriptaid and VPA that revealed different network(s) of potential regulators during erythropoiesis. VPA induced transcriptional activation of the glucocorticoid receptor (GCR) and an increase in the intracellular signalling of signal transducers and activators of transcription (STAT) required during stress erythropoiesis. Canonical Wnt signalling and many epigenetically regulated chromatin remodellers were significantly influenced so as to establish maintenance and regulation of HSPC.
Received: 3 February 2015, revised 1 April 2015, accepted 27 April 2015, published online 17 June 2015
Conclusion Treatment with Individual HDACIs has demonstrated significantly unique epigenetic and molecular signatures of CB-HSPC. This study identifies potential key regulators of HSPC and gives insights into the clinically important processes of HSPC expansion and haematopoietic lineage development for transplantation purposes. Key words: cord blood CD34+ cells, epigenetic regulation and gene expression, haematopoietic stem and progenitor cells, histone deacetylase Inhibitors.
Introduction Haematopoiesis is a hierarchically co-ordinated continuum that arises from a pool of self-renewing pluripotent haeCorrespondence: Pratima Chaurasia, Division of Hematology/Medical Oncology, Department of Medicine, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY 10029, USA E-mail: [email protected] *These authors contributed equally to this work.
matopoietic stem cells. During the differentiation process, HSC loses self-renewal ability and acquires specialized progenitor cell types with ability to undergo terminal differentiation [1–6]. Recent reports suggest that epigenetic alterations play an important role in haematopoietic stem and progenitor cell (HSPC) division, expansion and lineage restriction [3, 4]. HSCs possess unique epigenetic signatures that are inherited by HSPC subpopulations and that allow their differentiation into multilineage mature blood cells through highly co-ordinated gene activation and
79
80 D. Gajzer et al.
silencing [7]. It is well documented that HDACIs affect the epigenetic code in stem cells and that such induced modifications are maintained through cell divisions in vitro as well as in vivo [1–8]. HDACIs selectively alter 2–10% of gene expression depending on their concentrations, culture period and cell type [8–11]. Recently, we demonstrated that the treatment of CB-HSC with HDACI in the presence or absence of serum differently affected the expansion of HSPC ex vivo [3, 4]. The class I HDACs (1, 2, 3 and 8) are ubiquitously expressed nuclear enzymes and are involved in cell growth and proliferation, whereas class II HDACs (4, 5, 6, 7, 9 and 10) shuttle between nucleus and cytoplasm and their expression and functions are tissue specific [12]. The screening study performed with a number of HDACIs including valproic acid (VPA) and scriptaid (Scr) demonstrated significantly greater expansion of CB-HSPC [4]. To identify epigenetic and molecular gene signatures associated with the expansion and maintenance of primary CD34+ cells (PC) and HSPC ex vivo in the presence or absence of VPA and Scr, CD34+ were subjected in parallel to microarray, Ingenuity, iReport and GeneGo MetaCoreTM functional analyses. Treatment with HDACIs led to epigenetic signatures, consisting of a number of genes that affect chromatin modification of HSPC distinct from epigenetic signatures observed in primary CD34+ cells (PC). Ingenuity analysis revealed that HDACIs generated an increased number of genes that modulate most key cellular and molecular function. In addition, iReport Wheel filtered most significantly high-ranked canonical pathways that are involved in cellular growth and proliferation in VPA-expanded CD34+ cells. The distinct epigenetic and molecular signatures created by individual HDACIs, scriptaid and VPA, were capable of altering the stem cell fate and differentiation programme of expanded HSPC. Alterations in expression of genes encoding transcriptional factors (TFs) are difficult to identify as these changes are small and their activity is mostly modulated at a post-transcriptional level. A comparison of the transcriptional network profiles generated by individual HDACI, using GeneGO MetaCoreTM analyses, identified a core set of molecular components that delineated a role for the chromatin remodellers in the maintenance and regulation of HSPC. We also observed that each individual HDACI generated a unique expression signature, such as functional transcriptional regulation revealed upregulation of glucocorticoid receptor (GCR)/ STAT network (required during erythropoiesis) in the presence of VPA, but not scriptaid. These analyses revealed that exposure to different HACIs leads to varying epigenetic alterations of potential regulators during various stages of HSPC self-renewal, maintenance and development.
Methods Isolation of CB-CD34+ cells, ex vivo culture and microarray analysis CB collections were purchased from the Placental Blood Program at the New York Blood Center, New York. CB mononuclear cells (MNCs) were isolated and CD34+ cells were purified by immunomagnetic selection (Miltenyi Biotec, San Diego, CA, USA) as previously described [3]. Primary CB-CD34+ cells (PC) with purity (>90%) were initially primed in IMDM media supplemented with 30% FBS plus cytokines cocktail (SCF (100 ng/ml), FLT3 (100 ng/ml), TPO (100 ng/ml) and IL3 (50 ng/ml) for 16– 18 h. 2 9 105 CB-CD34+ cells were incubated with freshly prepared culture media supplemented with cytokines cocktail plus HDACIs including scriptaid (SCR 1 lM; Cayman Chemical; Ann Arbor, MI, USA) or valproic acid (VPA 1 mM; Sigma-Aldrich, St. Louis, MO, USA) along with a control for 7 days as described in Fig. 1a and Table 1. After 7 days, CD34+ cells were enriched using MACS columns (Miltenyi Biotech) and purity between 95 and 99% was considered for experiments listed in Fig. 1a. All the experiments including microarray, Q-PCR and Western blot were performed from enriched CD34+ cells following various culture conditions. CB-CD34+ cells were processed individually and equal number of PC and enriched CD34+ cells from 3–4 samples were pooled after expansion to avoid sample variations for microarray and Western blot analysis. Total RNA was extracted from PC and enriched CD34+ cells using Trizol–RNeasy columns (Qiagen, Valencia, CA, USA) [3, 4]. Illumina human 6v2 arrays (43 000 genes) hybridization and scanning were carried out as described at http://medicine.yale.edu/keck/ycga/ microarrays/index.aspx. Gene expression levels were compared between PC and control/SCR/VPA groups by analysis of variance (ANOVA) utilizing the statistical package Partek Genomic Suite (Partek Incorporated, Saint Louis, MO, USA) with a cut-off of P < 0005 and ≥20fold change. The gene expression data have been deposited in the Gene Expression Omnibus (GEO) databank (http://www.ncbi.nlm.nih.gov/geo) under the Accession Number GSE5980 or at http://www.ncbi.nlm.nih.gov/ geo/query/acc.cgi?acc=GSE59803.
Q-PCR Total RNA was reverse transcribed using cDNA kit (Clontech, Mountain View, CA, USA) as per manufacturer’s protocol. A total of 13 genes, AURAK, ALDH6A1, BTG2, CCND1, CD36, CD96, COL4AS, AXUD1, HES6, MMP9, NCOA7, PRKCB2, RCOR2, and housekeeping genes © 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 81
(GAPDH) were chosen for Q-PCR analysis to confirm gene expression profile generated by microarray analysis. SYBR Green Q-PCR was performed using PCR primers designed for Illumina platform as per manufacturer’s instructions (SA Biosciences, Qiagen Inc, Valencia, CA, USA) and described previously [3, 4].
(a)
GeneGo MetaCoreTM analysis The identification of significantly regulated transcriptional network, gene hubs (protein–protein), stem cell maintenance and regulation was performed using GeneGo MetaBaseTM software (Thomson Reuters, New York, NY, USA). The algorithm begins with a set of differentially expressed genes or proteins derived from PC and CD34+ cells enriched from control, SCR and VPA cultures. These genes or proteins are distinguished by the standard statistical analysis of data generated by Illumina platform at P < 0005- and ≥20-fold change. Using GeneGo’s MetaBaseTM, we mapped sets of genes or proteins onto a global database of protein–protein interactions. (b)
Ingenuity pathways analysis (IPA)
Fig. 1 (a) Effects of HDACIs on CB-HSPC expansion. A schematic representation of the ex vivo culture strategies of the day 0 primary cord blood (CB) CD34+ cells (PC). Freshly isolated PC were primed in a serumcontaining media plus cytokines cocktail for 16–18 h. Cells were then further incubated with freshly prepared culture media supplemented with cytokines cocktail plus HDACI (scriptaid or VPA) along with a control (untreated-without HDACIs) for additional 7 days. The expanded cells were enriched for CD34+ cells and were used for further analysis. (b) Percentage subpopulation of CD34+ cells after treatment in the presence or absence of HDACI. Cells were stained with APC-conjugated CD34 monoclonal antibody and analysed by flow cytometry (n = 4).
Table 1 Terminology referred to the cell populations studied. Condition
Terminology
PC Control
Primary cord blood CD34+ cells (day 0) Cultures supplemented with cytokines cocktail for 7 days (without HDACI) Cultures supplemented with cytokines cocktail and 1 mM VPA for 7 days Cultures supplemented with cytokines cocktail and 1 lM SCR for 7 days
VPA SCR
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
The Ingenuity pathway analysis tool (Ingenuity Systems) was utilized to identify cellular functional networks underlying HDACI modifications of CD34+ cells. The gene networks were algorithmically generated depending upon their functional connectivity from the data obtained using Illumina platform and Fischer’s exact test was applied to calculate a P-value (P = 005) so that each biological function assigned to that network or data set is due to chance alone. Molecular function based on genes involved in various biological processes and canonical pathways was analysed using Ingenuity iReport Wheel for red cell production.
Western blot The cell lysates were prepared from enriched CD34+ from control, scriptaid-/VPA-treated cultures and cell lines (HeLA and MCF7). A total of 15–20 lg protein was fractionated and immunoblotted with glucocorticoid antibody (Santa Cruz Biotechnology Inc., Dallas, TX, USA), STAT3 and b-actin antibody (Cell Signaling Technology, Danvers, MA, USA) as described previously [4].
Statistical analyses Results were expressed as either the mean – SD or SE of varying numbers of individual experiments.
82 D. Gajzer et al.
Study approval Informed consent or subject approval was not required for this study, as low-volume unidentifiable CB units for research purposes were purchased from the New York Blood Center.
Results Treatment of CB-CD34+ cells with individual HDACIs identifies differentially expressed genes A schematic representation of primary CB-CD34+ cells (day 0) expansion strategies utilized for ex vivo expansion in the presence and absence of HDACIs (scriptaid and VPA) and the terminology used to refer primary CB-CD34+ cells (PC), control (untreated-without HDACIs) and HDACIs-treated cells (Scr and VPA) are provided in Fig. 1a. The addition of individual HDACI (scriptaid or VPA) led to an increase in the expansion of CD34+ cells as compared to control cultures (Fig. 1b). However, VPA and scriptaid (P = 002) generated a significantly greater subpopulation of CD34+ cells (587 – 54%, P < 00001; 415 – 127%, P = 002) as compared to control (241 – 47%) (ANOVA P = 00001). We performed Illumina BeadChip analysis to explore the gene expression profile underlying the expansion of CD34+ cells in the presence of HDACI. The heatmap revealed a distinct expression profile that distinguished CD34+ cells from the control cultures and cultures containing scriptaid or VPA (Fig. 2a) while comparing the gene expression profile of control, scriptaid- or VPAtreated CD34+ cells to that of PC. A total of 559 differentially expressed genes (DEGs) were identified with a
(a)
≥–20-fold expression difference that was significant (P < 0005) using ANOVA. Commonly regulated genes between CD34+ cells from control cultures and cultures containing scriptaid and VPA are indicated in the Venn diagram (Fig. 2b). A total of 100 genes were commonly expressed in CD34+ cells exposed to control, scriptaid and VPA as compared to PC. A total of 209 genes were similarly regulated CD34+ cells treated with scriptaid and VPA. Of these 209 DEGs, 108 were upregulated (within a range of 20- to 110-fold change), and 101 genes were downregulated (-220- to -20-fold change). A similar degree of overlap (109 genes) between scriptaid and VPA suggests that a core pattern of epigenetically altered gene expression could be associated with a molecular signature of HSPC genes. This gene signature distinguishes scriptaidand VPA-enriched CD34+ cells and CD34+ cells from control cultures as well as PC. A complete list of differentially regulated genes obtained from microarray analysis is presented in the supplemental table (Table S1). Next, we performed Q-PCR analysis to cross-validate the microarray gene expression profile for a number of genes. As can be seen in Fig. S1, BTG2, AXUD1 and MMP9 were uniformly downregulated in control, VPA- and scriptaid-treated CD34+ cells. However, CD36, CD96, COL4AS, HES9, NCOA7 and cell cycle and cell division genes including AURAK, CCND1, PRKCB2 and RCOR2 were upregulated in both VPA- and scriptaid-treated CD34+ cells (≥17-fold change).
HDACIs influence expression of genes involved in chromatin modification Accumulating evidence supports that chromatin modification plays an important role in the epigenetic alterations
(b)
Fig. 2 HDACIs alter the gene profiles. (a) Gene expression profiles of primary CB-CD34+ cells (PC), enriched CD34+ cells from control and HDACI-containing cultures from two independent experiments. PC and enriched CD34+ cells were pooled after 7 days of cultures (3–4 CB samples) for each individual experiment. Heatmap was generated by ANOVA model of Partek Genomic Suite Program with a cut-off of false discovery rate of P < 005. Genes expressing near background are shown in grey, upregulated genes in red and downregulated genes in blue. (b) Venn diagram shows differentially expressed genes (DEGs) among control, SCR and VPA (P < 0005)).
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 83
that govern overall gene expression. A list of 15 genes involved in chromatin remodelling is shown in Fig. 3a. RERE expression was upregulated in PC and CD34+ cells reisolated from scriptaid and control cultures. Although RERE is critical in cell survival, over expression of RERE leads to caspase 3-mediated cell death [13]. Interestingly, expression levels of PAD14 in scriptaid- and VPA-treated cells were similar to PC but were upregulated in control CD34+ cells. PAD14 may be involved in granulocyte and macrophage development leading to inflammation and immune response [14]. A set of four genes including CHD7, ASF1A, LEF1 and MYB were unanimously upregulated in HDACI-treated CD34+ cells. Recently, Hsu et al. demonstrated that CHD7 is evolutionarily conserved and is a negative epigenetic regulator of HSCs/HPC by interacting with Runx1 and c-MYB during haematopoiesis [15]. ASF1A transcription factor plays an important role in encoding histone chaperone proteins and the Wnt signalling pathway. LEF1 associates with histone deacetylase (HDAC1) at the
(a)
Fig. 3 HDACIs influence expression of genes that affect chromatin modification. (a) Cluster diagram depicting the gene expression patterns associated with chromatin modification in PC and CD34+ cells enriched from control, scriptaid and VPA cultures. (b) Validation of microarray data by Q-PCR. ASF1A, CHD7, LEF1 and PADI4 were differentially expressed as compared to primary CB-CD34+ cells (PC). The fold change in expression relative to PC was calculated by SYBR Green Q-PCR. GAPDH was used as an internal housekeeping gene. Measurements were obtained in triplicate and a negative control (lacking the cDNA template) was included in each experiment (n = 3).
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
(b)
upstream region of the c-myc gene that leads to an increase in multipotent progenitors in an autonomous fashion in PADI4-deficient mice [15]. As can be seen in Fig. 3a, HDCAI-treated HSPC are associated with downregulation of PADI4 and upregulation of LEF1 as compared to control, which could be associated with generation of a large number of progenitor cells ex vivo. In addition, SMYD3, BNP3 and ING2 are upregulated in CD34+ cells enriched from the cultures in the presence or absence of HDACIs. However, epigenetic mark SMYD3 has been shown to associate with JAB1 in order to regulate transcription of the INK4a gene and haematopoietic cells [16]. BNP plays an important role in ES cell self-renewal and pluripotency [17]. The recruitment of the ING2associated HDAC1 component by WDR5 on b-globin gene promoter induces H3K4me3 and consequently silences expression of b-globin [18]. We also validated the expression levels of several chromatin-modifying genes including CHD7, ASF1A, LEF1 and PADI4 by Q-PCR (Fig. 3b).
84 D. Gajzer et al.
Functional pathway analysis of HDACIs-treated CB-CD34+ cells demonstrates significant alterations in cellular and signalling pathways CB-HSC treated with an individual HDACI resulted in the generation of far greater numbers of HSPC. Therefore, we next decided to explore the genes and pathways involved in the expansion of HSPC, either in control culture or HDACI-treated cultures. As shown in Fig. 4a, top biological cellular functions include cell cycle, death and sur-
vival, cell signalling, cell-to-cell signalling and interaction, cellular assembly and organization development, maintenance and movement. In the presence of scriptaid or VPA, there was a significant increase in the number of genes related to epigenetic alterations and different cellular functions, as listed in Table S2. The epigenetic reprogramming of these stem cells plays a critical role in the ex vivo expansion and generation of effector blood cells through a series of downstream progenitors with progressively limited potential. We next analysed
(a)
(b)
Fig. 4 Ingenuity pathway analysis of the cellular function involved in the enriched CD34+ cells treated in the presence or absence of HDACIs including VPA and scriptaid. (a) A data analysis capability tool within the Ingenuity pathway analysis was used to determine the altered gene expression associated with functional and cellular properties of HSPC expanded in the presence or absence of HDACI. (b) Wheel chapter that highlights a total of 77 transcription regulators that are differentially expressed in the VPAexpanded cord blood CD34+ cells and sorted by molecular function. The genes are grouped by the number of associated processes and colour coded for expression fold change (iReport Ingenuity). The relative expression levels of some of these transcriptional regulators (red outline of circles) involved in the canonical pathway are also shown. The statistical significance of these pathways was calculated with the right-tailed Fisher’s exact test.
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 85
red cell-related, functionally relevant, epigenetically altered gene expression by Web-based tool iReport. In the Ingenuity Knowledge Base, as compared to control, 45 and 77 DEGs with a 15-fold change in expression (cutoff P < 005) for scriptaid and VPA, respectively (Table S3). In addition, CD34+ cell treatment with VPA also identified a greater number of pathways (n = 225) as compared to scriptaid (n = 27) by filtering through keyword ‘red cell production’ (data not shown). The cellular growth, proliferation and development of HSPC in the presence of VPA led to the most significantly regulated pathways including B cell-activating factor signalling (P = 00005) GM-CSF signalling, erythropoietin signalling, angiopoietin signalling and PEDF signalling (P = 003) (Fig. 4b) and most regulated effector molecules (Table 2). The degree of alterations in gene expression is related to several possible mode(s) of action and emphasizes the alterations in cellular functions and signalling pathways that basically play critical roles during haematopoiesis.
Transcriptional regulation analysis indicates upregulation of Wnt/tcf/b-catenin pathway and erythroid differentiation genes networks The relative importance of the low levels of TF expression or post-translational modifications in HSPC can be obscured if evaluated entirely with microarray analyses. To overcome this problem, we utilized the transcriptional regulation network mapping algorithm within MetaCore to identify the TF which regulate all the DEGs (P < 005) generated from microarray data. This analysis identified 20 significant TF networks. A list of few important commonly expressed TFs and target genes (CREB1, c-Myc, SP1, ESR1 (nuclear), GCRa, androgen receptor, STAT3, E2F1, STAT1) in HSPC that were altered by scriptaid and Table 2 The molecular function representing the most regulated effector molecules in VPA-enriched HSPC generated by ‘iReport’ Wheel after filtering through key word ‘red cell production’ Gene Transcription factors (TFs)
Transmembrane receptors Cytokine Other proteins
NFKBIA NFKB1 STAT1 STAT4 LEF1 ID2 CSF2RA TNFSF IGFBP4 TUBB2B CCNA1
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Fold change -1994 -1504 153 189 270 234 -154 -1862 1897 549 219
VPA treatment are provided in Table S4. These data suggest that fates of HSCs are regulated by a fine tuning of a number of TFs and other proteins. To determine a molecular signature gene expression characteristic of HSPC, we next performed GeneGo analysis on the gene lists obtained using Illumina platform. The most significant individual molecular signature was associated with many proteins, and these proteins were sorted based on the number of interactions with global interactome (GeneGo Global Network) and within the individual protein lists. The interaction by protein function scored and ranked relative connectivity between genes (hubs) network (TFs, receptors, ligands, kinases, phosphatases, proteases, enzymes and others) by VPA and scriptaid. These highly connected genes should be further investigated as they play crucial roles in regulating haematopoiesis. We have previously demonstrated that in the presence VPA plus serum, the fate of CB-CD34+ cells skews towards erythroid lineage [3, 6]. Interestingly, a number of erythroid progenitor and erythropoiesis related genes were identified that were connected to the molecular network (Table S5). VPA significantly regulated a variety of genes including AHR, EKLF1, LCKBP1 (HS1), miR-186-5p AHR, b-myb, EZH2, EKLF1, LSD1, DNMT1 and related proteins. These genes are involved in defining the phenotype and maintenance of erythroid progenitor cells continuum rather than a hierarchy. Next, we built a network for process including stem cell maintenance (GO:0035019) and regulation of stem cell maintenance (GO:2000036) using GeneGo analysis. A list of gene network(s) provided in Table S6 demonstrates a significant upregulation of canonical Wnt pathway in HSPC reisolated after treatment with either VPA or scriptaid (Fig. S2a and b). The stem cell maintenance was more significantly regulated by canonical Wnt/Tcf/b-catenin pathway in VPA-treated HSPC (total network pathways = 1308) as compared to scriptaid (total network pathways = 938). A number of histone-modifying enzymes, chromatin remodellers (PAX2, Cullin4a, CDX2, SFRP1, LTX, BRG1, PAF1, LIN-28, YAP1, TRAP, RTF1 and TLX), were also connected to the networks involved in stem cell maintenance and regulation. Many other chromatin remodellers showed unique connectivity in scriptaid (PROX1, Dicer, RIF1, MOZ and VPS72 (SWC2))- and VPA (ARS2, CTR9, TRG20, SHC and CRSP8)-treated HSPC. In addition, many stem cell and progenitor self-renewal genes including Wnt, DLL1, SOX2, OCT3/4 HES1, c-kit, ZFP36L2 and GATA2 were distinguished as common molecular network (s) regulators. These data clearly suggest that different combinations of chromatin modifications appear to provide dynamic epigenetic alterations to the functional networking of genes required for HSPC maintenance and regulation.
86 D. Gajzer et al.
Role of glucocorticoid receptor (GCR) in erythropoiesis It has been well documented that GCR is a key regulator of continuous proliferation of erythroid progenitor self-renewal [19]. We next analysed the GCRa (Table
S4) transcriptional regulation network. As can be seen from Fig. 5a (i, ii and iii) that GCRa transcription network was upregulated by HMG-1 and HMG-2 in HSPC either enriched from control cultures or cultures containing scriptaid/VPA. Elevated levels of HMGB-1 and highly related HMG-2 have been shown to interact
(a) (i)
(ii)
(iii)
(b)
Fig. 5 Effect of HDACIs on biological network of transcriptional regulation of GCR. (a) GeneGO MetaCoreTM analysis was performed using direct interactions of the significantly regulated 161 genes in control, 240 genes in scriptaid 494 genes in VPA at FDR ≤ 005 (Fig. 2b). Networks represent (i) control, (ii) scriptaid and (iii) VPA and red broken line encircles GCR in the centre of the each network. The red circle and blue circle represent over- and underexpression, respectively; activation (green solid line), inhibition (red solid line) and unspecified (grey solid line). A list of legend symbols is provided at https://portal.genego.com/legends/MetaCoreQuickReferenceGuide.pdf (b) Western blot analysis of GCR. Total cell lysates were prepared from the CD34+ cells enriched from the control cultures and cultures containing HDACIs. The cell lysates were analysed using WB analysis for GCR receptor (a- and bsubunit) and STAT3. HeLA and MCF7 cell lines were utilized as positive controls for GCR and STAT3, respectively. b-actin was utilized as a loading control. One of the three representative experiments is shown.
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 87
with DNA-binding proteins including HOX and OCT, steroid hormone receptors including oestrogen, androgen, and glucocorticoid receptors, RAG1 and 2, Rel and p53 [20]. VPA generated a more complex network(s) of GCRa as compared to control and scriptaid. To further evaluate the significance of the GCRa and STAT3 expression (Table S4) levels, these particular TFs were selected for Western blot analysis. The elevated protein expression of GCRa and STAT3 in the CD34+ cells treated with VPA was consistent with the predicted network analysis (Fig. 5b). However, an increased expression of STAT3 and GCR in VPA-treated CD34+ cells could be attributed to increase level of recruitment of STAT3 to GCR [21]. The recruitment of STAT3 to GCR either by protein–protein interactions or their corecruitment to the adjacent sites led to synergistic transcription. By contrast, recruitment of GCR to DNA-bound STAT3 was linked with transrepression [22].
Discussion Epigenetic alterations in the chromatin remodelling complex govern chromatin structure and expression of genes involved in the self-renewal and maintenance of HSPC. We aimed to understand the relative effects of class I and class II HDACIs on gene profiles and to identify the commonalties and uniqueness in the epigenetic and molecular networking of these genes required for ex vivo expansion of HSPC. In this study, we identified a total of 559 DEGs with ≥–20-fold change that fits in a more stringent cut-off (P < 0005). The gene expression profile exhibited a far greater number of genes that were uniquely regulated in the presence of VPA as compared to scriptaid and control. While some of the individual genes were not significantly altered, the functional annotation(s) that group these individual gene(s) together led to a significant alteration in a particular pathway. Previously, we demonstrated that serum-free and serum-containing cultures in the presence of HDACI can govern differential HSC fate and fold expansion [3, 4]. The serum-containing cultures likely antagonize the gene silencing of HSPC and establish open chromatin configuration of HSPC for transcription of genes required for the stem cell differentiation programme. Therefore, the fold expansion of CD34+ cells in the presence of HDACI was lower in serum-containing cultures as compared to serum-free cultures [4]. The treatment of CB-HSPC with VPA in the serum-free media led to epigenetic reprogramming and a far greater expansion of true HSC that were capable of human cell engraftment in primary and secondary NOD/SCID mice [4]. © 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
A key finding of this study revealed that as compared to PC, scriptaid and VPA led to the identification of epigenetic signatures (CHD7, ASF1A, LEF1 and MYB) on the expanded HSPC. The upregulation of these genes suggests their effect on the target genes is likely required for maintenance and expansion of the functional HSPC. By contrast, activation of PADI4 in control HSPC can act as an epigenetic coactivator and regulate histone repressive marks H3R2me2a at Tal1/PADI4 target gene IL6ST. IL6ST transcription plays a role during lineage differentiation of HSPC [23]. The LRWD1 binds to histone methylation repressive marks (H3K9me3, H3K27me3 and H4K20me3) in a co-operative manner with methylated DNA in heterochromatin regions. In addition, BRPE1, BRPE3 and PHF15 play critical roles in histone H4-K12, H4-K16, H4-K5 and H4-K8 acetylation. The biological pathway analysis of DEGs revealed that cellular and functional processes of expanded HSPC were significantly modulated in a greater number of genes in scriptaid- and VPA-treated HSPC as compared to control. Interestingly, Wnt pathway was most significantly regulated in the stem cell maintenance and regulation network. Upregulation of Notch3 promotes cell growth and also co-operates with Wnt/tcf pathway in regulating cell function [24]. We have also identified the sets of significantly regulated gene hub(s) by selecting network(s) connectivity via protein–protein interaction in class I and class II HDACIs which led to expansion of HSPC. However, erythroid-specific protein–protein interaction network(s) were highly connected and did not reflect scale-free topologies. In addition, iReport generated a variety of haematopoiesis-related canonical pathways including JAK/STAT, aryl hydrocarbon receptor, GM-CSF and erythropoietin, which were over-represented in the presence of VPA. It has been reported that GCR-mediated regulation of erythroid progenitor proliferation is essential for stress erythropoiesis in vivo [22]. The role of STATs in haematopoietic cell signalling has been well documented and STAT3 expression is upregulated during erythroid maturation [25]. TF, as an intrinsic determinant, allows entry for uncovering the molecular path required for HSPC phenotype and to programme lineage-specific differentiation. Moreover, insights into the molecular mechanism for the differential function of protein–protein connectivity are critically lacking. These analyses have provided a molecular framework for developing and testing new hypotheses about how epigenetics can alter genes and biological protein–protein networks that could contribute to the physiological and functional properties of ex vivo expanded HSPC and their roles in generation of transfusion products in clinical settings.
88 D. Gajzer et al.
Acknowledgements DG and JR performed experiments; LW SN and SL discussed, critically reviewed and provided valuable suggestions for the manuscript; WZ contributed to microarray data analysis and wrote the manuscript; PC designed and performed research, analysed GeneGO MetaCoreTM data and wrote the manuscript. We are grateful to Dr. R Hoffman (Icahn School of Medicine at Mount Sinai) for
his support. We thank Dr. S Mane (Yale University) for providing microarray services and Ingenuity team for performing the iReport data analysis. This study was support by NYSTEM IDEA (Contract # C026431) to RH and PC.
Conflict of interests The authors declare no competing financial interests.
References 1 Araki H, Mahmud N, Milhem M, et al.: Expansion of human umbilical cord blood SCID-repopulating cells using chromatin-modifying agents. Exp Hematol 2006; 34:140–149 2 Bug G, Gul H, Schwarz K, et al.: Valproic acid stimulates proliferation and self-renewal of hematopoietic stem cells. Cancer Res 2005; 65:2537–2541 3 Chaurasia P, Berenzon D, Hoffman R: Chromatin-modifying agents promote the ex vivo production of functional human erythroid progenitor cells. Blood 2011; 117:4632–4641 4 Chaurasia P, Gajzer DC, Schaniel C, et al.: Epigenetic reprogramming induces the expansion of cord blood stem cells. J Clin Invest 2014; 124:2378–2395 5 De Felice L, Tatarelli C, Mascolo MG, et al.: Histone deacetylase inhibitor valproic acid enhances the cytokineinduced expansion of human hematopoietic stem cells. Cancer Res 2005; 65:1505–1513 6 Chaurasia P, Berenzon D, Hoffman R: Epigenetic regulation of normal hematopoietic development in L€ ubbert M, Jones PA (ed): Epigenetic Therapy of Cancer, Preclinical Models and Clinical Approaches, Heidelberg, Springer Verlag publication, 2014, pp 67–88 7 Feng S, Jacobsen SE, Reik W: Epigenetic reprogramming in plant and animal development. Science 2010; 330:622–627 8 Maes J, Maleszewska M, Guillemin C, et al.: Lymphoid-affiliated genes are associated with active histone modifications in human hematopoietic stem cells. Blood 2008; 112:2722–2729 9 Tollervey JR, Lunyak VV: Epigenetics: judge, jury and executioner of
10
11
12
13
14
15
16
17
stem cell fate. Epigenetics 2012; 7:823–840 Mahmud N, Petro B, Baluchamy S, et al.: Differential effects of epigenetic modifiers on the expansion and maintenance of human cord blood stem/ progenitor cells. Biol Blood Marrow Transplant 2014; 20:480–489 Cohen KS, Cheng S, Larson MG, et al.: Circulating CD34(+) progenitor cell frequency is associated with clinical and genetic factors. Blood 2013; 121: e50–e56 Delcuve GP, Khan DH, Davie JR: Roles of histone deacetylases in epigenetic regulation: emerging paradigms from studies with inhibitors. Clin Epigenetics, 2012; 4:5 Mohanan S, Cherrington BD, Horibata S, et al.: Potential role of peptidylarginine deiminase enzymes and protein citrullination in cancer pathogenesis. Biochem Res Int 2012; 2012:895343 Hsu J, Huang H-T, Lee C-T, et al.: Chromatin Remodeling Enzyme CHD7 Negatively Regulate Hematopoietic Stem Cell Function [abstract]. Blood (ASH Annual Meeting Abstracts) November 15, 2013; Blood: 122 (21). Nakashima K, Arai S, Suzuki A, et al.: PAD4 regulates proliferation of multipotent haematopoietic cells by controlling c-myc expression. Nat Commun 2013; 4:1836 Mori M, Yoneda-Kato N, Yoshida A, et al.: Stable form of JAB1 enhances proliferation and maintenance of hematopoietic progenitors. J Biol Chem 2008; 283:29011–29021 Abdelalim EM, Tooyama I: NPR-A regulates self-renewal and pluripotency of embryonic stem cells. Cell Death Dis 2011; 2:e127
18 Xu Z, He Y, Ju J, et al.: The role of WDR5 in silencing human fetal globin gene expression. Haematologica 2012; 97:1632–1640 19 Wessely O, Deiner EM, Beug H, et al.: The glucocorticoid receptor is a key regulator of the decision between selfrenewal and differentiation in erythroid progenitors. EMBO J 1997; 16:267–280 20 Dintilhac A, Bernues J: HMGB1 interacts with many apparently unrelated proteins by recognizing short amino acid sequences. J Biol Chem 2002; 277:7021–7028 21 Arambasic J, Poznanovi G, IvanovicMatic S, et al.: Association of the glucocorticoid receptor with STAT3, C/ EBPbeta, and the hormone-responsive element within the rat haptoglobin gene promoter during the acute phase response. IUBMB Life 2010; 62:227– 236 22 Langlais D, Couture C, Balsalobre A, et al.: The Stat3/GR interaction code: predictive value of direct/indirect DNA recruitment for transcription outcome. Mol Cell 2012; 47:38–49 23 Kolodziej S, Kuvardina ON, Oellerich T, et al.: PADI4 acts as a coactivator of Tal1 by counteracting repressive histone arginine methylation. Nat Commun 2014; 5:3995 24 Wang T, Holt CM, Xu C, et al.: Notch3 activation modulates cell growth behaviour and cross-talk to Wnt/TCF signalling pathway. Cell Signal 2007; 19:2458–2467 25 Cokic VP, Bhattacharya B, BeleslinCokic BB, et al.: JAK-STAT and AKT pathway-coupled genes in erythroid progenitor cells through ontogeny. J Transl Med 2012; 10:116
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 89
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1 Validation of Microarray data by Q-PCR. Figure S2 A list of genes network(s) provided in Table S6, demonstrates a significant upregulation of canonical Wnt pathway in the both HSPC reisolated after treatment with either VPA (a) or scripted (b). Table S1 Microarray gene lists by Venn diagram: Genes significantly differentially expressed in reisolated CD34+ cells treated in the absence(control) and presence of HDACIs (scriptaid and VPA) for 7 days and compared with Primary CD34+ cells (Day0) (PC). Table S2 Epigenetic alterations in the cellular function associated genes in reisolated CD34+ cells treatted in the presence or absence of HDACIs including VPA and scriptaid by Ingenuity Pathways Analysis. Table S3 A total of 77 Significantly differentially expressed transcripts generated from Illumina dataset by iREPORT with p-value cutoff 0.05 and fold change cutoff 1.5. Table S4 GeneGo MetacoreTM analyses generated network list of transcription factors and target genes in the reisolated CD34+ cells after treatment with VPA. Table S5 Significantly connected genes involved during the developmental process of of erythroid progenitor cells/erythropoiesis in the presence of scriptaid and VPA. Table S6 GeneGo MetacoreTM analyses generated network for processes stem cell maintenance and stem cell maintenance regulation in the reisolated CD34+ cells after treatment with VPA.
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
ORIGINAL PAPER
Epigenetic and molecular signatures of cord blood CD34+ cells treated with histone deacetylase inhibitors D. Gajzer,1* J. Ross,1* L. Winder,1 S. Navada,1 W. Zhang,2 L. Silverman1 & P. Chaurasia1 1
Division of Hematology/Medical Oncology, Department of Medicine, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY, USA 2 Department of Medicine Bioinformatics Core, Icahn School of Medicine at Mount Sinai, New York, NY, USA
Background and Objectives Epigenetic modifications tightly regulate the gene expression and cellular function of haematopoietic stem cells. Histone deacetylase inhibitors (HDACIs) alter the gene expression profile of cord blood (CB) CD34+ cells by controlling the genes involved in chromatin modification, thereby influencing the self-renewal, maintenance and expansion of haematopoietic stem and progenitor cells (HSPCs). Materials and Methods The class I and II HDACIs, valproic acid and scriptaid, were utilized to expand CB-CD34+ cells ex vivo. The gene profiling was performed on HSPC using Illumina microarray, GeneGO MetaCoreTM and Ingenuity pathway analyses. The molecular analyses were performed using Q-PCR and Western blotting. Results Each HDACI treatment of CB-CD34+ cells created unique epigenetic and molecular signatures that governed chromatin modification required for cellular and functional behaviour of stem cells. GeneGO MetaCoreTM and Ingenuity pathway analyses established the molecular understanding of epigenetically regulated HSPCs in the presence of scriptaid and VPA that revealed different network(s) of potential regulators during erythropoiesis. VPA induced transcriptional activation of the glucocorticoid receptor (GCR) and an increase in the intracellular signalling of signal transducers and activators of transcription (STAT) required during stress erythropoiesis. Canonical Wnt signalling and many epigenetically regulated chromatin remodellers were significantly influenced so as to establish maintenance and regulation of HSPC.
Received: 3 February 2015, revised 1 April 2015, accepted 27 April 2015, published online 17 June 2015
Conclusion Treatment with Individual HDACIs has demonstrated significantly unique epigenetic and molecular signatures of CB-HSPC. This study identifies potential key regulators of HSPC and gives insights into the clinically important processes of HSPC expansion and haematopoietic lineage development for transplantation purposes. Key words: cord blood CD34+ cells, epigenetic regulation and gene expression, haematopoietic stem and progenitor cells, histone deacetylase Inhibitors.
Introduction Haematopoiesis is a hierarchically co-ordinated continuum that arises from a pool of self-renewing pluripotent haeCorrespondence: Pratima Chaurasia, Division of Hematology/Medical Oncology, Department of Medicine, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, NY 10029, USA E-mail: [email protected] *These authors contributed equally to this work.
matopoietic stem cells. During the differentiation process, HSC loses self-renewal ability and acquires specialized progenitor cell types with ability to undergo terminal differentiation [1–6]. Recent reports suggest that epigenetic alterations play an important role in haematopoietic stem and progenitor cell (HSPC) division, expansion and lineage restriction [3, 4]. HSCs possess unique epigenetic signatures that are inherited by HSPC subpopulations and that allow their differentiation into multilineage mature blood cells through highly co-ordinated gene activation and
79
80 D. Gajzer et al.
silencing [7]. It is well documented that HDACIs affect the epigenetic code in stem cells and that such induced modifications are maintained through cell divisions in vitro as well as in vivo [1–8]. HDACIs selectively alter 2–10% of gene expression depending on their concentrations, culture period and cell type [8–11]. Recently, we demonstrated that the treatment of CB-HSC with HDACI in the presence or absence of serum differently affected the expansion of HSPC ex vivo [3, 4]. The class I HDACs (1, 2, 3 and 8) are ubiquitously expressed nuclear enzymes and are involved in cell growth and proliferation, whereas class II HDACs (4, 5, 6, 7, 9 and 10) shuttle between nucleus and cytoplasm and their expression and functions are tissue specific [12]. The screening study performed with a number of HDACIs including valproic acid (VPA) and scriptaid (Scr) demonstrated significantly greater expansion of CB-HSPC [4]. To identify epigenetic and molecular gene signatures associated with the expansion and maintenance of primary CD34+ cells (PC) and HSPC ex vivo in the presence or absence of VPA and Scr, CD34+ were subjected in parallel to microarray, Ingenuity, iReport and GeneGo MetaCoreTM functional analyses. Treatment with HDACIs led to epigenetic signatures, consisting of a number of genes that affect chromatin modification of HSPC distinct from epigenetic signatures observed in primary CD34+ cells (PC). Ingenuity analysis revealed that HDACIs generated an increased number of genes that modulate most key cellular and molecular function. In addition, iReport Wheel filtered most significantly high-ranked canonical pathways that are involved in cellular growth and proliferation in VPA-expanded CD34+ cells. The distinct epigenetic and molecular signatures created by individual HDACIs, scriptaid and VPA, were capable of altering the stem cell fate and differentiation programme of expanded HSPC. Alterations in expression of genes encoding transcriptional factors (TFs) are difficult to identify as these changes are small and their activity is mostly modulated at a post-transcriptional level. A comparison of the transcriptional network profiles generated by individual HDACI, using GeneGO MetaCoreTM analyses, identified a core set of molecular components that delineated a role for the chromatin remodellers in the maintenance and regulation of HSPC. We also observed that each individual HDACI generated a unique expression signature, such as functional transcriptional regulation revealed upregulation of glucocorticoid receptor (GCR)/ STAT network (required during erythropoiesis) in the presence of VPA, but not scriptaid. These analyses revealed that exposure to different HACIs leads to varying epigenetic alterations of potential regulators during various stages of HSPC self-renewal, maintenance and development.
Methods Isolation of CB-CD34+ cells, ex vivo culture and microarray analysis CB collections were purchased from the Placental Blood Program at the New York Blood Center, New York. CB mononuclear cells (MNCs) were isolated and CD34+ cells were purified by immunomagnetic selection (Miltenyi Biotec, San Diego, CA, USA) as previously described [3]. Primary CB-CD34+ cells (PC) with purity (>90%) were initially primed in IMDM media supplemented with 30% FBS plus cytokines cocktail (SCF (100 ng/ml), FLT3 (100 ng/ml), TPO (100 ng/ml) and IL3 (50 ng/ml) for 16– 18 h. 2 9 105 CB-CD34+ cells were incubated with freshly prepared culture media supplemented with cytokines cocktail plus HDACIs including scriptaid (SCR 1 lM; Cayman Chemical; Ann Arbor, MI, USA) or valproic acid (VPA 1 mM; Sigma-Aldrich, St. Louis, MO, USA) along with a control for 7 days as described in Fig. 1a and Table 1. After 7 days, CD34+ cells were enriched using MACS columns (Miltenyi Biotech) and purity between 95 and 99% was considered for experiments listed in Fig. 1a. All the experiments including microarray, Q-PCR and Western blot were performed from enriched CD34+ cells following various culture conditions. CB-CD34+ cells were processed individually and equal number of PC and enriched CD34+ cells from 3–4 samples were pooled after expansion to avoid sample variations for microarray and Western blot analysis. Total RNA was extracted from PC and enriched CD34+ cells using Trizol–RNeasy columns (Qiagen, Valencia, CA, USA) [3, 4]. Illumina human 6v2 arrays (43 000 genes) hybridization and scanning were carried out as described at http://medicine.yale.edu/keck/ycga/ microarrays/index.aspx. Gene expression levels were compared between PC and control/SCR/VPA groups by analysis of variance (ANOVA) utilizing the statistical package Partek Genomic Suite (Partek Incorporated, Saint Louis, MO, USA) with a cut-off of P < 0005 and ≥20fold change. The gene expression data have been deposited in the Gene Expression Omnibus (GEO) databank (http://www.ncbi.nlm.nih.gov/geo) under the Accession Number GSE5980 or at http://www.ncbi.nlm.nih.gov/ geo/query/acc.cgi?acc=GSE59803.
Q-PCR Total RNA was reverse transcribed using cDNA kit (Clontech, Mountain View, CA, USA) as per manufacturer’s protocol. A total of 13 genes, AURAK, ALDH6A1, BTG2, CCND1, CD36, CD96, COL4AS, AXUD1, HES6, MMP9, NCOA7, PRKCB2, RCOR2, and housekeeping genes © 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 81
(GAPDH) were chosen for Q-PCR analysis to confirm gene expression profile generated by microarray analysis. SYBR Green Q-PCR was performed using PCR primers designed for Illumina platform as per manufacturer’s instructions (SA Biosciences, Qiagen Inc, Valencia, CA, USA) and described previously [3, 4].
(a)
GeneGo MetaCoreTM analysis The identification of significantly regulated transcriptional network, gene hubs (protein–protein), stem cell maintenance and regulation was performed using GeneGo MetaBaseTM software (Thomson Reuters, New York, NY, USA). The algorithm begins with a set of differentially expressed genes or proteins derived from PC and CD34+ cells enriched from control, SCR and VPA cultures. These genes or proteins are distinguished by the standard statistical analysis of data generated by Illumina platform at P < 0005- and ≥20-fold change. Using GeneGo’s MetaBaseTM, we mapped sets of genes or proteins onto a global database of protein–protein interactions. (b)
Ingenuity pathways analysis (IPA)
Fig. 1 (a) Effects of HDACIs on CB-HSPC expansion. A schematic representation of the ex vivo culture strategies of the day 0 primary cord blood (CB) CD34+ cells (PC). Freshly isolated PC were primed in a serumcontaining media plus cytokines cocktail for 16–18 h. Cells were then further incubated with freshly prepared culture media supplemented with cytokines cocktail plus HDACI (scriptaid or VPA) along with a control (untreated-without HDACIs) for additional 7 days. The expanded cells were enriched for CD34+ cells and were used for further analysis. (b) Percentage subpopulation of CD34+ cells after treatment in the presence or absence of HDACI. Cells were stained with APC-conjugated CD34 monoclonal antibody and analysed by flow cytometry (n = 4).
Table 1 Terminology referred to the cell populations studied. Condition
Terminology
PC Control
Primary cord blood CD34+ cells (day 0) Cultures supplemented with cytokines cocktail for 7 days (without HDACI) Cultures supplemented with cytokines cocktail and 1 mM VPA for 7 days Cultures supplemented with cytokines cocktail and 1 lM SCR for 7 days
VPA SCR
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
The Ingenuity pathway analysis tool (Ingenuity Systems) was utilized to identify cellular functional networks underlying HDACI modifications of CD34+ cells. The gene networks were algorithmically generated depending upon their functional connectivity from the data obtained using Illumina platform and Fischer’s exact test was applied to calculate a P-value (P = 005) so that each biological function assigned to that network or data set is due to chance alone. Molecular function based on genes involved in various biological processes and canonical pathways was analysed using Ingenuity iReport Wheel for red cell production.
Western blot The cell lysates were prepared from enriched CD34+ from control, scriptaid-/VPA-treated cultures and cell lines (HeLA and MCF7). A total of 15–20 lg protein was fractionated and immunoblotted with glucocorticoid antibody (Santa Cruz Biotechnology Inc., Dallas, TX, USA), STAT3 and b-actin antibody (Cell Signaling Technology, Danvers, MA, USA) as described previously [4].
Statistical analyses Results were expressed as either the mean – SD or SE of varying numbers of individual experiments.
82 D. Gajzer et al.
Study approval Informed consent or subject approval was not required for this study, as low-volume unidentifiable CB units for research purposes were purchased from the New York Blood Center.
Results Treatment of CB-CD34+ cells with individual HDACIs identifies differentially expressed genes A schematic representation of primary CB-CD34+ cells (day 0) expansion strategies utilized for ex vivo expansion in the presence and absence of HDACIs (scriptaid and VPA) and the terminology used to refer primary CB-CD34+ cells (PC), control (untreated-without HDACIs) and HDACIs-treated cells (Scr and VPA) are provided in Fig. 1a. The addition of individual HDACI (scriptaid or VPA) led to an increase in the expansion of CD34+ cells as compared to control cultures (Fig. 1b). However, VPA and scriptaid (P = 002) generated a significantly greater subpopulation of CD34+ cells (587 – 54%, P < 00001; 415 – 127%, P = 002) as compared to control (241 – 47%) (ANOVA P = 00001). We performed Illumina BeadChip analysis to explore the gene expression profile underlying the expansion of CD34+ cells in the presence of HDACI. The heatmap revealed a distinct expression profile that distinguished CD34+ cells from the control cultures and cultures containing scriptaid or VPA (Fig. 2a) while comparing the gene expression profile of control, scriptaid- or VPAtreated CD34+ cells to that of PC. A total of 559 differentially expressed genes (DEGs) were identified with a
(a)
≥–20-fold expression difference that was significant (P < 0005) using ANOVA. Commonly regulated genes between CD34+ cells from control cultures and cultures containing scriptaid and VPA are indicated in the Venn diagram (Fig. 2b). A total of 100 genes were commonly expressed in CD34+ cells exposed to control, scriptaid and VPA as compared to PC. A total of 209 genes were similarly regulated CD34+ cells treated with scriptaid and VPA. Of these 209 DEGs, 108 were upregulated (within a range of 20- to 110-fold change), and 101 genes were downregulated (-220- to -20-fold change). A similar degree of overlap (109 genes) between scriptaid and VPA suggests that a core pattern of epigenetically altered gene expression could be associated with a molecular signature of HSPC genes. This gene signature distinguishes scriptaidand VPA-enriched CD34+ cells and CD34+ cells from control cultures as well as PC. A complete list of differentially regulated genes obtained from microarray analysis is presented in the supplemental table (Table S1). Next, we performed Q-PCR analysis to cross-validate the microarray gene expression profile for a number of genes. As can be seen in Fig. S1, BTG2, AXUD1 and MMP9 were uniformly downregulated in control, VPA- and scriptaid-treated CD34+ cells. However, CD36, CD96, COL4AS, HES9, NCOA7 and cell cycle and cell division genes including AURAK, CCND1, PRKCB2 and RCOR2 were upregulated in both VPA- and scriptaid-treated CD34+ cells (≥17-fold change).
HDACIs influence expression of genes involved in chromatin modification Accumulating evidence supports that chromatin modification plays an important role in the epigenetic alterations
(b)
Fig. 2 HDACIs alter the gene profiles. (a) Gene expression profiles of primary CB-CD34+ cells (PC), enriched CD34+ cells from control and HDACI-containing cultures from two independent experiments. PC and enriched CD34+ cells were pooled after 7 days of cultures (3–4 CB samples) for each individual experiment. Heatmap was generated by ANOVA model of Partek Genomic Suite Program with a cut-off of false discovery rate of P < 005. Genes expressing near background are shown in grey, upregulated genes in red and downregulated genes in blue. (b) Venn diagram shows differentially expressed genes (DEGs) among control, SCR and VPA (P < 0005)).
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 83
that govern overall gene expression. A list of 15 genes involved in chromatin remodelling is shown in Fig. 3a. RERE expression was upregulated in PC and CD34+ cells reisolated from scriptaid and control cultures. Although RERE is critical in cell survival, over expression of RERE leads to caspase 3-mediated cell death [13]. Interestingly, expression levels of PAD14 in scriptaid- and VPA-treated cells were similar to PC but were upregulated in control CD34+ cells. PAD14 may be involved in granulocyte and macrophage development leading to inflammation and immune response [14]. A set of four genes including CHD7, ASF1A, LEF1 and MYB were unanimously upregulated in HDACI-treated CD34+ cells. Recently, Hsu et al. demonstrated that CHD7 is evolutionarily conserved and is a negative epigenetic regulator of HSCs/HPC by interacting with Runx1 and c-MYB during haematopoiesis [15]. ASF1A transcription factor plays an important role in encoding histone chaperone proteins and the Wnt signalling pathway. LEF1 associates with histone deacetylase (HDAC1) at the
(a)
Fig. 3 HDACIs influence expression of genes that affect chromatin modification. (a) Cluster diagram depicting the gene expression patterns associated with chromatin modification in PC and CD34+ cells enriched from control, scriptaid and VPA cultures. (b) Validation of microarray data by Q-PCR. ASF1A, CHD7, LEF1 and PADI4 were differentially expressed as compared to primary CB-CD34+ cells (PC). The fold change in expression relative to PC was calculated by SYBR Green Q-PCR. GAPDH was used as an internal housekeeping gene. Measurements were obtained in triplicate and a negative control (lacking the cDNA template) was included in each experiment (n = 3).
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
(b)
upstream region of the c-myc gene that leads to an increase in multipotent progenitors in an autonomous fashion in PADI4-deficient mice [15]. As can be seen in Fig. 3a, HDCAI-treated HSPC are associated with downregulation of PADI4 and upregulation of LEF1 as compared to control, which could be associated with generation of a large number of progenitor cells ex vivo. In addition, SMYD3, BNP3 and ING2 are upregulated in CD34+ cells enriched from the cultures in the presence or absence of HDACIs. However, epigenetic mark SMYD3 has been shown to associate with JAB1 in order to regulate transcription of the INK4a gene and haematopoietic cells [16]. BNP plays an important role in ES cell self-renewal and pluripotency [17]. The recruitment of the ING2associated HDAC1 component by WDR5 on b-globin gene promoter induces H3K4me3 and consequently silences expression of b-globin [18]. We also validated the expression levels of several chromatin-modifying genes including CHD7, ASF1A, LEF1 and PADI4 by Q-PCR (Fig. 3b).
84 D. Gajzer et al.
Functional pathway analysis of HDACIs-treated CB-CD34+ cells demonstrates significant alterations in cellular and signalling pathways CB-HSC treated with an individual HDACI resulted in the generation of far greater numbers of HSPC. Therefore, we next decided to explore the genes and pathways involved in the expansion of HSPC, either in control culture or HDACI-treated cultures. As shown in Fig. 4a, top biological cellular functions include cell cycle, death and sur-
vival, cell signalling, cell-to-cell signalling and interaction, cellular assembly and organization development, maintenance and movement. In the presence of scriptaid or VPA, there was a significant increase in the number of genes related to epigenetic alterations and different cellular functions, as listed in Table S2. The epigenetic reprogramming of these stem cells plays a critical role in the ex vivo expansion and generation of effector blood cells through a series of downstream progenitors with progressively limited potential. We next analysed
(a)
(b)
Fig. 4 Ingenuity pathway analysis of the cellular function involved in the enriched CD34+ cells treated in the presence or absence of HDACIs including VPA and scriptaid. (a) A data analysis capability tool within the Ingenuity pathway analysis was used to determine the altered gene expression associated with functional and cellular properties of HSPC expanded in the presence or absence of HDACI. (b) Wheel chapter that highlights a total of 77 transcription regulators that are differentially expressed in the VPAexpanded cord blood CD34+ cells and sorted by molecular function. The genes are grouped by the number of associated processes and colour coded for expression fold change (iReport Ingenuity). The relative expression levels of some of these transcriptional regulators (red outline of circles) involved in the canonical pathway are also shown. The statistical significance of these pathways was calculated with the right-tailed Fisher’s exact test.
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 85
red cell-related, functionally relevant, epigenetically altered gene expression by Web-based tool iReport. In the Ingenuity Knowledge Base, as compared to control, 45 and 77 DEGs with a 15-fold change in expression (cutoff P < 005) for scriptaid and VPA, respectively (Table S3). In addition, CD34+ cell treatment with VPA also identified a greater number of pathways (n = 225) as compared to scriptaid (n = 27) by filtering through keyword ‘red cell production’ (data not shown). The cellular growth, proliferation and development of HSPC in the presence of VPA led to the most significantly regulated pathways including B cell-activating factor signalling (P = 00005) GM-CSF signalling, erythropoietin signalling, angiopoietin signalling and PEDF signalling (P = 003) (Fig. 4b) and most regulated effector molecules (Table 2). The degree of alterations in gene expression is related to several possible mode(s) of action and emphasizes the alterations in cellular functions and signalling pathways that basically play critical roles during haematopoiesis.
Transcriptional regulation analysis indicates upregulation of Wnt/tcf/b-catenin pathway and erythroid differentiation genes networks The relative importance of the low levels of TF expression or post-translational modifications in HSPC can be obscured if evaluated entirely with microarray analyses. To overcome this problem, we utilized the transcriptional regulation network mapping algorithm within MetaCore to identify the TF which regulate all the DEGs (P < 005) generated from microarray data. This analysis identified 20 significant TF networks. A list of few important commonly expressed TFs and target genes (CREB1, c-Myc, SP1, ESR1 (nuclear), GCRa, androgen receptor, STAT3, E2F1, STAT1) in HSPC that were altered by scriptaid and Table 2 The molecular function representing the most regulated effector molecules in VPA-enriched HSPC generated by ‘iReport’ Wheel after filtering through key word ‘red cell production’ Gene Transcription factors (TFs)
Transmembrane receptors Cytokine Other proteins
NFKBIA NFKB1 STAT1 STAT4 LEF1 ID2 CSF2RA TNFSF IGFBP4 TUBB2B CCNA1
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Fold change -1994 -1504 153 189 270 234 -154 -1862 1897 549 219
VPA treatment are provided in Table S4. These data suggest that fates of HSCs are regulated by a fine tuning of a number of TFs and other proteins. To determine a molecular signature gene expression characteristic of HSPC, we next performed GeneGo analysis on the gene lists obtained using Illumina platform. The most significant individual molecular signature was associated with many proteins, and these proteins were sorted based on the number of interactions with global interactome (GeneGo Global Network) and within the individual protein lists. The interaction by protein function scored and ranked relative connectivity between genes (hubs) network (TFs, receptors, ligands, kinases, phosphatases, proteases, enzymes and others) by VPA and scriptaid. These highly connected genes should be further investigated as they play crucial roles in regulating haematopoiesis. We have previously demonstrated that in the presence VPA plus serum, the fate of CB-CD34+ cells skews towards erythroid lineage [3, 6]. Interestingly, a number of erythroid progenitor and erythropoiesis related genes were identified that were connected to the molecular network (Table S5). VPA significantly regulated a variety of genes including AHR, EKLF1, LCKBP1 (HS1), miR-186-5p AHR, b-myb, EZH2, EKLF1, LSD1, DNMT1 and related proteins. These genes are involved in defining the phenotype and maintenance of erythroid progenitor cells continuum rather than a hierarchy. Next, we built a network for process including stem cell maintenance (GO:0035019) and regulation of stem cell maintenance (GO:2000036) using GeneGo analysis. A list of gene network(s) provided in Table S6 demonstrates a significant upregulation of canonical Wnt pathway in HSPC reisolated after treatment with either VPA or scriptaid (Fig. S2a and b). The stem cell maintenance was more significantly regulated by canonical Wnt/Tcf/b-catenin pathway in VPA-treated HSPC (total network pathways = 1308) as compared to scriptaid (total network pathways = 938). A number of histone-modifying enzymes, chromatin remodellers (PAX2, Cullin4a, CDX2, SFRP1, LTX, BRG1, PAF1, LIN-28, YAP1, TRAP, RTF1 and TLX), were also connected to the networks involved in stem cell maintenance and regulation. Many other chromatin remodellers showed unique connectivity in scriptaid (PROX1, Dicer, RIF1, MOZ and VPS72 (SWC2))- and VPA (ARS2, CTR9, TRG20, SHC and CRSP8)-treated HSPC. In addition, many stem cell and progenitor self-renewal genes including Wnt, DLL1, SOX2, OCT3/4 HES1, c-kit, ZFP36L2 and GATA2 were distinguished as common molecular network (s) regulators. These data clearly suggest that different combinations of chromatin modifications appear to provide dynamic epigenetic alterations to the functional networking of genes required for HSPC maintenance and regulation.
86 D. Gajzer et al.
Role of glucocorticoid receptor (GCR) in erythropoiesis It has been well documented that GCR is a key regulator of continuous proliferation of erythroid progenitor self-renewal [19]. We next analysed the GCRa (Table
S4) transcriptional regulation network. As can be seen from Fig. 5a (i, ii and iii) that GCRa transcription network was upregulated by HMG-1 and HMG-2 in HSPC either enriched from control cultures or cultures containing scriptaid/VPA. Elevated levels of HMGB-1 and highly related HMG-2 have been shown to interact
(a) (i)
(ii)
(iii)
(b)
Fig. 5 Effect of HDACIs on biological network of transcriptional regulation of GCR. (a) GeneGO MetaCoreTM analysis was performed using direct interactions of the significantly regulated 161 genes in control, 240 genes in scriptaid 494 genes in VPA at FDR ≤ 005 (Fig. 2b). Networks represent (i) control, (ii) scriptaid and (iii) VPA and red broken line encircles GCR in the centre of the each network. The red circle and blue circle represent over- and underexpression, respectively; activation (green solid line), inhibition (red solid line) and unspecified (grey solid line). A list of legend symbols is provided at https://portal.genego.com/legends/MetaCoreQuickReferenceGuide.pdf (b) Western blot analysis of GCR. Total cell lysates were prepared from the CD34+ cells enriched from the control cultures and cultures containing HDACIs. The cell lysates were analysed using WB analysis for GCR receptor (a- and bsubunit) and STAT3. HeLA and MCF7 cell lines were utilized as positive controls for GCR and STAT3, respectively. b-actin was utilized as a loading control. One of the three representative experiments is shown.
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 87
with DNA-binding proteins including HOX and OCT, steroid hormone receptors including oestrogen, androgen, and glucocorticoid receptors, RAG1 and 2, Rel and p53 [20]. VPA generated a more complex network(s) of GCRa as compared to control and scriptaid. To further evaluate the significance of the GCRa and STAT3 expression (Table S4) levels, these particular TFs were selected for Western blot analysis. The elevated protein expression of GCRa and STAT3 in the CD34+ cells treated with VPA was consistent with the predicted network analysis (Fig. 5b). However, an increased expression of STAT3 and GCR in VPA-treated CD34+ cells could be attributed to increase level of recruitment of STAT3 to GCR [21]. The recruitment of STAT3 to GCR either by protein–protein interactions or their corecruitment to the adjacent sites led to synergistic transcription. By contrast, recruitment of GCR to DNA-bound STAT3 was linked with transrepression [22].
Discussion Epigenetic alterations in the chromatin remodelling complex govern chromatin structure and expression of genes involved in the self-renewal and maintenance of HSPC. We aimed to understand the relative effects of class I and class II HDACIs on gene profiles and to identify the commonalties and uniqueness in the epigenetic and molecular networking of these genes required for ex vivo expansion of HSPC. In this study, we identified a total of 559 DEGs with ≥–20-fold change that fits in a more stringent cut-off (P < 0005). The gene expression profile exhibited a far greater number of genes that were uniquely regulated in the presence of VPA as compared to scriptaid and control. While some of the individual genes were not significantly altered, the functional annotation(s) that group these individual gene(s) together led to a significant alteration in a particular pathway. Previously, we demonstrated that serum-free and serum-containing cultures in the presence of HDACI can govern differential HSC fate and fold expansion [3, 4]. The serum-containing cultures likely antagonize the gene silencing of HSPC and establish open chromatin configuration of HSPC for transcription of genes required for the stem cell differentiation programme. Therefore, the fold expansion of CD34+ cells in the presence of HDACI was lower in serum-containing cultures as compared to serum-free cultures [4]. The treatment of CB-HSPC with VPA in the serum-free media led to epigenetic reprogramming and a far greater expansion of true HSC that were capable of human cell engraftment in primary and secondary NOD/SCID mice [4]. © 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
A key finding of this study revealed that as compared to PC, scriptaid and VPA led to the identification of epigenetic signatures (CHD7, ASF1A, LEF1 and MYB) on the expanded HSPC. The upregulation of these genes suggests their effect on the target genes is likely required for maintenance and expansion of the functional HSPC. By contrast, activation of PADI4 in control HSPC can act as an epigenetic coactivator and regulate histone repressive marks H3R2me2a at Tal1/PADI4 target gene IL6ST. IL6ST transcription plays a role during lineage differentiation of HSPC [23]. The LRWD1 binds to histone methylation repressive marks (H3K9me3, H3K27me3 and H4K20me3) in a co-operative manner with methylated DNA in heterochromatin regions. In addition, BRPE1, BRPE3 and PHF15 play critical roles in histone H4-K12, H4-K16, H4-K5 and H4-K8 acetylation. The biological pathway analysis of DEGs revealed that cellular and functional processes of expanded HSPC were significantly modulated in a greater number of genes in scriptaid- and VPA-treated HSPC as compared to control. Interestingly, Wnt pathway was most significantly regulated in the stem cell maintenance and regulation network. Upregulation of Notch3 promotes cell growth and also co-operates with Wnt/tcf pathway in regulating cell function [24]. We have also identified the sets of significantly regulated gene hub(s) by selecting network(s) connectivity via protein–protein interaction in class I and class II HDACIs which led to expansion of HSPC. However, erythroid-specific protein–protein interaction network(s) were highly connected and did not reflect scale-free topologies. In addition, iReport generated a variety of haematopoiesis-related canonical pathways including JAK/STAT, aryl hydrocarbon receptor, GM-CSF and erythropoietin, which were over-represented in the presence of VPA. It has been reported that GCR-mediated regulation of erythroid progenitor proliferation is essential for stress erythropoiesis in vivo [22]. The role of STATs in haematopoietic cell signalling has been well documented and STAT3 expression is upregulated during erythroid maturation [25]. TF, as an intrinsic determinant, allows entry for uncovering the molecular path required for HSPC phenotype and to programme lineage-specific differentiation. Moreover, insights into the molecular mechanism for the differential function of protein–protein connectivity are critically lacking. These analyses have provided a molecular framework for developing and testing new hypotheses about how epigenetics can alter genes and biological protein–protein networks that could contribute to the physiological and functional properties of ex vivo expanded HSPC and their roles in generation of transfusion products in clinical settings.
88 D. Gajzer et al.
Acknowledgements DG and JR performed experiments; LW SN and SL discussed, critically reviewed and provided valuable suggestions for the manuscript; WZ contributed to microarray data analysis and wrote the manuscript; PC designed and performed research, analysed GeneGO MetaCoreTM data and wrote the manuscript. We are grateful to Dr. R Hoffman (Icahn School of Medicine at Mount Sinai) for
his support. We thank Dr. S Mane (Yale University) for providing microarray services and Ingenuity team for performing the iReport data analysis. This study was support by NYSTEM IDEA (Contract # C026431) to RH and PC.
Conflict of interests The authors declare no competing financial interests.
References 1 Araki H, Mahmud N, Milhem M, et al.: Expansion of human umbilical cord blood SCID-repopulating cells using chromatin-modifying agents. Exp Hematol 2006; 34:140–149 2 Bug G, Gul H, Schwarz K, et al.: Valproic acid stimulates proliferation and self-renewal of hematopoietic stem cells. Cancer Res 2005; 65:2537–2541 3 Chaurasia P, Berenzon D, Hoffman R: Chromatin-modifying agents promote the ex vivo production of functional human erythroid progenitor cells. Blood 2011; 117:4632–4641 4 Chaurasia P, Gajzer DC, Schaniel C, et al.: Epigenetic reprogramming induces the expansion of cord blood stem cells. J Clin Invest 2014; 124:2378–2395 5 De Felice L, Tatarelli C, Mascolo MG, et al.: Histone deacetylase inhibitor valproic acid enhances the cytokineinduced expansion of human hematopoietic stem cells. Cancer Res 2005; 65:1505–1513 6 Chaurasia P, Berenzon D, Hoffman R: Epigenetic regulation of normal hematopoietic development in L€ ubbert M, Jones PA (ed): Epigenetic Therapy of Cancer, Preclinical Models and Clinical Approaches, Heidelberg, Springer Verlag publication, 2014, pp 67–88 7 Feng S, Jacobsen SE, Reik W: Epigenetic reprogramming in plant and animal development. Science 2010; 330:622–627 8 Maes J, Maleszewska M, Guillemin C, et al.: Lymphoid-affiliated genes are associated with active histone modifications in human hematopoietic stem cells. Blood 2008; 112:2722–2729 9 Tollervey JR, Lunyak VV: Epigenetics: judge, jury and executioner of
10
11
12
13
14
15
16
17
stem cell fate. Epigenetics 2012; 7:823–840 Mahmud N, Petro B, Baluchamy S, et al.: Differential effects of epigenetic modifiers on the expansion and maintenance of human cord blood stem/ progenitor cells. Biol Blood Marrow Transplant 2014; 20:480–489 Cohen KS, Cheng S, Larson MG, et al.: Circulating CD34(+) progenitor cell frequency is associated with clinical and genetic factors. Blood 2013; 121: e50–e56 Delcuve GP, Khan DH, Davie JR: Roles of histone deacetylases in epigenetic regulation: emerging paradigms from studies with inhibitors. Clin Epigenetics, 2012; 4:5 Mohanan S, Cherrington BD, Horibata S, et al.: Potential role of peptidylarginine deiminase enzymes and protein citrullination in cancer pathogenesis. Biochem Res Int 2012; 2012:895343 Hsu J, Huang H-T, Lee C-T, et al.: Chromatin Remodeling Enzyme CHD7 Negatively Regulate Hematopoietic Stem Cell Function [abstract]. Blood (ASH Annual Meeting Abstracts) November 15, 2013; Blood: 122 (21). Nakashima K, Arai S, Suzuki A, et al.: PAD4 regulates proliferation of multipotent haematopoietic cells by controlling c-myc expression. Nat Commun 2013; 4:1836 Mori M, Yoneda-Kato N, Yoshida A, et al.: Stable form of JAB1 enhances proliferation and maintenance of hematopoietic progenitors. J Biol Chem 2008; 283:29011–29021 Abdelalim EM, Tooyama I: NPR-A regulates self-renewal and pluripotency of embryonic stem cells. Cell Death Dis 2011; 2:e127
18 Xu Z, He Y, Ju J, et al.: The role of WDR5 in silencing human fetal globin gene expression. Haematologica 2012; 97:1632–1640 19 Wessely O, Deiner EM, Beug H, et al.: The glucocorticoid receptor is a key regulator of the decision between selfrenewal and differentiation in erythroid progenitors. EMBO J 1997; 16:267–280 20 Dintilhac A, Bernues J: HMGB1 interacts with many apparently unrelated proteins by recognizing short amino acid sequences. J Biol Chem 2002; 277:7021–7028 21 Arambasic J, Poznanovi G, IvanovicMatic S, et al.: Association of the glucocorticoid receptor with STAT3, C/ EBPbeta, and the hormone-responsive element within the rat haptoglobin gene promoter during the acute phase response. IUBMB Life 2010; 62:227– 236 22 Langlais D, Couture C, Balsalobre A, et al.: The Stat3/GR interaction code: predictive value of direct/indirect DNA recruitment for transcription outcome. Mol Cell 2012; 47:38–49 23 Kolodziej S, Kuvardina ON, Oellerich T, et al.: PADI4 acts as a coactivator of Tal1 by counteracting repressive histone arginine methylation. Nat Commun 2014; 5:3995 24 Wang T, Holt CM, Xu C, et al.: Notch3 activation modulates cell growth behaviour and cross-talk to Wnt/TCF signalling pathway. Cell Signal 2007; 19:2458–2467 25 Cokic VP, Bhattacharya B, BeleslinCokic BB, et al.: JAK-STAT and AKT pathway-coupled genes in erythroid progenitor cells through ontogeny. J Transl Med 2012; 10:116
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
Histone deacetylase inhibitors and cord blood 89
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1 Validation of Microarray data by Q-PCR. Figure S2 A list of genes network(s) provided in Table S6, demonstrates a significant upregulation of canonical Wnt pathway in the both HSPC reisolated after treatment with either VPA (a) or scripted (b). Table S1 Microarray gene lists by Venn diagram: Genes significantly differentially expressed in reisolated CD34+ cells treated in the absence(control) and presence of HDACIs (scriptaid and VPA) for 7 days and compared with Primary CD34+ cells (Day0) (PC). Table S2 Epigenetic alterations in the cellular function associated genes in reisolated CD34+ cells treatted in the presence or absence of HDACIs including VPA and scriptaid by Ingenuity Pathways Analysis. Table S3 A total of 77 Significantly differentially expressed transcripts generated from Illumina dataset by iREPORT with p-value cutoff 0.05 and fold change cutoff 1.5. Table S4 GeneGo MetacoreTM analyses generated network list of transcription factors and target genes in the reisolated CD34+ cells after treatment with VPA. Table S5 Significantly connected genes involved during the developmental process of of erythroid progenitor cells/erythropoiesis in the presence of scriptaid and VPA. Table S6 GeneGo MetacoreTM analyses generated network for processes stem cell maintenance and stem cell maintenance regulation in the reisolated CD34+ cells after treatment with VPA.
© 2015 International Society of Blood Transfusion Vox Sanguinis (2016) 110, 79–89
