Histone Deacetylase Inhibitor Valproic Acid Enhances the ...Histone Deacetylase Inhibitor Valproic Acid Enhances the Cytokine-Induced Expansion of Human Hematopoietic Stem Cells Lidia

Histone Deacetylase Inhibitor Valproic Acid Enhances

the Cytokine-Induced Expansion of Human

Hematopoietic Stem Cells

Lidia De Felice,1Caterina Tatarelli,

2Maria Grazia Mascolo,

1Chiara Gregorj,

1Francesca Agostini,

1

Roberto Fiorini,2Vania Gelmetti,

2,3Simona Pascale,

2,3Fabrizio Padula,

2Maria Teresa Petrucci,

1

William Arcese,4and Clara Nervi

2,3

Departments of 1Cellular Biotechnology and Hematology and 2Histology and Medical Embryology, University of Rome ‘‘La Sapienza’’;3San Raffaele Biomedical Science Park of Rome; and 4Department of Biopathology, University of Rome ‘‘Tor Vergata,’’ Rome, Italy

Abstract

Ex vivo amplification of human hematopoietic stem cells(HSC) without loss of their self-renewing potential representsan important target for transplantation, gene and cellulartherapies. Valproic acid is a safe and widely used neurologicagent that acts as a potent inhibitor of histone deacetylaseactivities. Here, we show that valproic acid addition to liquidcultures of human CD34++ cells isolated from cord blood,mobilized peripheral blood, and bone marrow stronglyenhances the ex vivo expansion potential of different cytokinecocktails as shown by morphologic, cytochemical, immuno-phenotypical, clonogenic, and gene expression analyses.Notably, valproic acid highly preserves the CD34 positivityafter 1 week (range, 40-89%) or 3 weeks (range, 21-52%)amplification cultures with two (Flt3L ++ thrombopoietin) orfour cytokines (Flt3L ++ thrombopoietin ++ stem cell factor ++interleukin 3). Moreover, valproic acid treatment increaseshistone H4 acetylation levels at specific regulatory sites onHOXB4 , a transcription factor gene with a key role in theregulation of HSC self-renewal and AC133 , a recognizedmarker gene for stem cell populations. Overall, our resultsrelate the changes induced by valproic acid on chromatinaccessibility with the enhancement of the cytokine effect onthe maintenance and expansion of a primitive hematopoieticstem cell population. These findings underscore the potenti-ality of novel epigenetic approaches to modify HSC fatein vitro . (Cancer Res 2005; 65(4): 1505-13)

Introduction

Hematopoietic stem cell (HSC) ex vivo expansion is an attractivestrategy to increase the number of self-renewing HSCs finalized totransplantation, gene, and stem cell therapies.Following the widespread use of the umbilical cord blood (CB)

as HSC source alternative to bone marrow (BM) in allogeneictransplantation, the cell dose infused has been markedly evidencedas pivotal item on the engraftment rate and clinical outcome (1–4).Several culture conditions and different cytokine combinations

inducing a significant expansion of HSC with engraftment potentialin animal models have been proposed (5, 6). There is a generalagreement on the key role played by Flt-3 ligand (Flt3L) and

thrombopoietin in the regulation of the early stages of hemato-poiesis (7–12). The combination of early acting cytokines like stemcell factor (SCF) and/or lineage specific growth factors such asinterleukin 3 (IL-3), IL-6, and IL-11 allows a wider HSCsamplification (9, 12, 14). However, the maintenance of primitiveHSCs with long-term engraftment potential is still controversial(5, 6, 13–16).Recent studies, have provided new insight on the identification

of intracellular factors that regulate self-renewing. Proteins that areactive during embryonic development, such as the members of theWnt, bone morphogenetic protein, Sonic hedgehog, and Notchfamilies induce self-renewal in adult HSCs (17–19). A functionalrole for the enhancement of the hematopoietic cell growth hasbeen shown for the transcriptional factor HOXB4 and thetranscriptional repressor Bmi-1, opening new therapeutic perspec-tives for the ex vivo amplification of HSCs (20–25).A mechanism by which upstream signaling can converge on

common targets to regulate gene expression might be provided bythe dynamic changes in the local and global chromatin organiza-tion induced by histone deacetylases (HDAC) and histoneacetyltransferases (26, 27). Deacetylation of the N-terminal tailsof nucleosomal core histone H3 and H4 by HDACs leads to achromatin conformation resulting in gene silencing, whereashistone acetylation by histone acetyltransferases is associated withthe activation of genomic regions. Recently, highly specificinhibitors of HDAC, acting as strong modulators of cellularproliferation and differentiation, have become available (28–30).Interestingly, the global chromatin remodeling activity of theseinhibitors affects only few (4-10%) selected genes (29). Thisspecificity, along with the low in vivo toxicity of these compounds,raises the possibility of their clinical use (28–31).Valproic acid, a strong HDAC inhibitor, has been used for >20

years in the treatment of several neurologic disorders (32). Theeffectiveness of valproic acid and other HDAC inhibitors inregulating the growth, differentiation, and apoptosis of tumorcells, including acute myeloid leukemias, has been documentedin vitro and in vivo (28–34). However, nothing is known yet aboutthe effect of valproic acid on proliferation, differentiation, andsurvival of normal human HSCs.

Materials and Methods

Cell Sources and Selection of CD34+ CellsHuman BM and mobilized peripheral blood (MPB) samples were

obtained, after informed consent from allogeneic transplant donors. The

cord blood units not eligible for the Italian CB Bank, were processed as

fresh samples or cryopreserved. Cryopreserved CB units were thawed

Requests for reprints: Clara Nervi, San Raffaele Biomedical Science Park of Romeand University of Rome ‘‘La Sapienza,’’ Via di Castel Romano 100, 00128 Rome, Italy.Phone: 39-06-80319049/52; Fax: 39-06-80319054; E-mail: [email protected] andLidia De Felice, E-mail: [email protected].

I2005 American Association for Cancer Research.

www.aacrjournals.org 1505 Cancer Res 2005; 65: (4). February 15, 2005

Research Article

Research. on November 3, 2020. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

according to Rubinstein et al. (35). Low-density mononuclear cells wereseparated by Ficoll-Hypaque density gradient centrifugation and incubated

overnight to remove the monocytic cells by adherence on plastic flasks.

CD34+ cells were positively selected by immunomagnetic separation using

the mini or midi MACS CD34 progenitor cell isolation kit (Miltenyi Biotec,Auburn, CA) according to the manufacturer’s instructions.

Ex vivo Expansion of CD34+ CellsCD34+ cells (2 � 104 cell/mL), were cultured in Iscove’s modified

Dulbecco’s medium (Life Technologies, Gaithersburg, MD) supplemented

with 10% human pooled AB serum. Recombinant human (rhu) Flt3L

(50 ng/mL; R&D Systems, Abingdon, United Kingdom), rhu-thrombopoie-

tin (50 ng/mL), rhu-SCF (100 ng/mL), and rhu-IL-3 (20 ng/mL; PeproTech,

London, United Kingdom) were used in the following combination: Flt3L +

thrombopoietin (FT) or Flt3L + thrombopoietin + SCF + IL-3 (FTS3).

Valproic acid (1 mmol/L, Sigma-Aldrich, Milan, Italy) was added 30

minutes before the addition of cytokines. The cultures were incubated at

37jC and 5% CO2 for 7 days. In long-term expansion, the cultures were fed

every week with an equal volume of fresh medium supplemented with

valproic acid and cytokines. On days 7, 14, and 21 the nucleated cell

number, viability, clonogenic potential, immunophenotype, and gene

expression of amplified cells were determined. Cell expansion was

expressed as fold increase, calculated by dividing the output absolute

number of cells and progenitors after amplification cultures by the

respective input cell number on day 0.

Clonogenic AssayCells were suspended in 1 mL methylcellulose medium supplemented

with SCF, GM-CSF, IL-3, and EPO (Methocult GFH4434, Stem Cell

Technologies, Vancouver, British Columbia, Canada), and plated in

duplicate in 35-mm culture dishes. Each plate was scored for granulocyte

macrophage colony-forming unit, burst forming unit-erythroid, and

granulocyte, erythroid, macrophage, megacaryocyte colony-forming unit

growth after 10 and 14 days incubation.Replating Tests. The ability of primary granulocyte macrophage colony-

forming unit colony-forming unit, burst forming unit-erythroid, and

granulocyte, erythroid, macrophage, megacaryocyte colony-forming unit

to give rise to secondary colonies was evaluated by picking up individualcolonies from the Petri dishes. Each colony was dispersed to single cell

suspension in 200 AL of methylcellulose medium (Methocult) and seeded

into separate wells. The percentage of positive wells and the number of

secondary colonies derived from each primary colony was established after10 and 14 days incubation.

Flow CytometryCell immunophenotype was assessed by two-color fluorescence analysis

using FITC-conjugated mouse anti-human CD34, CD38, CD90, CD45, CD15

(Becton Dickinson, Mountain View, CA), and phycoerythrin-conjugated

mouse anti-human CD34, CD33, CD13 (BD), CD133/1 (AC 133)-

phycoerythrin (Miltenyi Biotec), CD117-phycoerythrin (Beckman Coulter,

Fullerton, CA). Isotype controls were mouse immunoglobulin G conjugat-

ed to FITC or phycoerythrin. A minimum of 50,000 events were collected

for each sample by a FACScan flow cytometer (Becton Dickinson) using

CellFit software (Becton Dickinson) for data acquisition and analysis. For

cell cycle analysis and apoptosis, cells were resuspended in 50% FCS, fixed

in 70% ethanol for 24 hours, incubated with 50 Ag/mL propidium iodide

(Sigma-Aldrich) and 50 units/mL DNase-free RNase A (Sigma-Aldrich) and

analyzed after 3 hours (10,000 events) using a Epics XL Cytometer

(Beckman Coulter).

Cell Differentiation Assays by Morphologic andCytochemical AnalysisMyelograms were evaluated on Wright-Giemsa stained cytospins by light

microscopy by counting at least 200 cells per experimental condition. Cells

were classified as immature, granulocytes and monocytes based on

chromatin and cytoplasm characteristics. Cytospins were evaluated for

positive staining of (i) MPO, an enzyme restricted to the primary granules of

granulocytes and monocytes; (ii) a naphthyl acetate esterase positive in

cells of the monocytic lineage and virtually absent in granulocytes; (iii) a

naphthyl acetate esterase inhibition by sodium fluoride, occurring in

monocytes, as per manufacturer’s instructions (Sigma-Aldrich).

Immunofluorescence and Immunoblot analyses were done using anti-

bodies recognizing histone H3 or the acetylated forms of histone H3 and H4

(Upstate) or HOXB4 (Santa Cruz Biotechnology, Santa Cruz, CA) asdescribed (33).

RNA Preparation and Real-time PCRRNA was extracted at days 0, 7, 14, and 21 as described (33).

Quantitative reverse transcription-PCR was done in ABI PRISM 7000

(Applied Biosystems, Foster City, CA) using Taqman oligonucleotides(Assay on Demand, Applied Biosystems) for glyceraldehyde phosphate

dehydrogenase (GAPDH) , CD34 antigen (CD34) , c-kit receptor (CD117) ,

AC133 (CD133) , kinase domain receptor (KDR) , GATA binding protein 1

(GATA1) , colony-stimulating factor 2 receptor (GM-CSFr) , and colony-stimulating factor 3 receptor (G-CSFr) , glycoprotein Ib (Gp1ab) . Delta-delta

Ct values were normalized with those obtained from the amplification of

GAPDH. For the SYBR green dye detection method, primers were designed

using the Primer Express software (Applied Biosystems): GAPDH( forward) 5V-ATCAGCAATGCCTCCTGCAC-3V, GAPDH (reverse) 5V-TGGCATGGACTGTGGTCATG-3V, monocyte-specific esterase 1 (MSE,

forward) 5V-GAACCACAGAGATGCTGGAGC-3V, and MSE (reverse) 5V-TCCCCGTGGTCTCCTATCAC-3V. Reactions were done in triplicates.

Chromatin Immunoprecipitation. Cells (2 � 106) were treated with

valproic acid and/or FTS3 alone or in combination for 2 hours. Cross-

linking of histones to DNA was obtained by the addition to the medium offormaldehyde at 1% final concentration for 10 minutes at 37jC to cross-

link the histones to DNA. Chromatin was sonicated and immunopreci-

pitated overnight with an anti-acetylated histone H4 antibody (Upstate,

Lake Placid, NY) as recommended by the manufacturer. DNA wasanalyzed by PCR using primers for HOXB4 P1 252-bp promoter region

encompassing both HxRE-1 and E-box regions (36, 37): HOXB4-P1

( forward, �208 to �184 bp) 5V-GAAATAAACCTCTTTGGCTGGAGTGG-3V; HOXB4-P1 (reverse, +21 to +44 bp) 5V-GGGTGCACATAGTTTGAGTT-GATC-3V; HOXB4-P5 upstream 236-bp promoter region used as control:

HOXB4-P5 ( forward, �748 to �726 bp) 5V-GCGAAGTCTCCCCGAAT-TAGTG-3V; HOXB4-P5 (reverse, �512 to �488 bp) 5V-GTCTCTATGGG-GAGTTAGGTTACT-3V, AC133-1A , 191-bp promoter 1/exon 1A region used

by CD34+ progenitors (38): AC133-EX1A ( forward) 5V-CTACAGGAAATG-

Table 1. Immunophenotypic features and proliferativepotential of HSC from different cell sources

CB MPB BM

CD34+ (%) 73.4 F 9.0 90.4 F 3.7 95.9 F 1.4

CD34+90+ (%) 28.7 F 4.4 44.4 F 6.2* 13.3 F 2.5*

CD34+38� (%) 17.4 F 2.9 16.1 F 8.1 5.7 F 1.5

GM-CFU (n . coloniesper 500 CD34+)

224.6 F 98.7 63.9 F 13.4 12.8 F 0.7

BFU-E (n . colonies

per 500 CD34+)

37.7 F 9.7 86.7 F 22.6 13.7 F 5.2

CFU-GEMM (n. coloniesper 500 CD34+)

31.5 F 6.2c

12.3 F 3.5b

3.0 F 2.7c

Total CFU (n . colonies

per 500 CD34+)

292.5 F 112.4 163.0 F 38.8 29.4 F 3.2

NOTE: Mean F SE.

Abbreviation: BFU-E, burst forming unit-erythroid; GM-CFU, granulo-

cyte macrophage colony-forming unit; CFU-GEMM, granulocyte,

erythroid, macrophage, megacaryocyte colony-forming unit.*P = 0.030.cP = 0.041.bP = 0.034.

Cancer Research

Cancer Res 2005; 65: (4). February 15, 2005 1506 www.aacrjournals.org



GATGCTGTCCAG-3V; AC133-EX1A (reverse) 5V-GGATTCCTTAAACATACT-

CACCGC-3V; AC133-EXE , alternative 251-bp region on exon E used as

control: AC133-EXE ( forward) 5V-GCTCTGATGCTCTGATGAGAC-3V;

AC133-EXE (reverse) 5V-CTCTATGCCTGGGAACTCCTG-3V; GAPDH

to detect nonrelevant cellular DNA sequences: GAPDH ( forward) 5V-

ACAGTCCATGCCATCACTGCC-3V; GAPDH (reverse) 5V GCCTGCTTCAC-

CACCTTCTTG-3V.

StatisticsDescriptive statistics are reported as mean F SE unless otherwise

stated. Student’s t test was used to test the probability of significant

differences between samples. Differences were considered significant if

P V 0.05.

Results

Effect of Valproic Acid on Short-term Amplification Inducedby Four Cytokines. The effect of valproic acid on cytokine-

induced HSCs amplification was assessed in 7-day cultures

stimulated by the combination of four growth factors: Flt3L,

thrombopoietin, SCF, and IL-3 (FTS3). The baseline characteristicsof the CD34+ cells selected from CB (n = 3), MPB (n = 4), and BM

(n = 2) are shown in Table 1. CB and MPB were more enriched in

the early stem cell fraction than BM as documented by the higherpercentage of CD34+CD90+ and CD34+CD38� cells. The prolif-

erative potential of committed progenitors was higher in CB than

in the other cell sources with a frequency of one clonogenic cellout of 1.7 CB CD34+ cells, one of three MPB CD34+ cells, and 1 of

17 BM CD34+ cells.The acetylation status of histone H3 or H4 was evaluated in

CD34+ cells by immunofluorescence and immunoblotting analysisusing specific antibodies. Low levels of the acetylated form ofhistone H3 and H4 were detectable in untreated CD34+ cells from

all the cell sources. Sixteen hours treatment with either FT or FTS3induced a moderate increase of the acetylated forms of both thesehistones that was further enhanced in the presence of valproic acid(Fig. 1A and B).Because HDAC inhibitors have been reported to induce either

differentiation or apoptosis in cancer cells and in leukemicprogenitors (28–33), the effect of FTS3 + valproic acid (FTS3 +VPA) on differentiation and apoptosis of HSC was analyzed. Insharp contrast with the results obtained in tumor cells, themorphology of normal HSC treated for 7 days with FTS3 + VPAseemed significantly more immature than after treatmentwith FTS3 alone. Indeed, morphologic features of myeloiddifferentiation (chromatin condensation with nuclear segmenta-tion, decreased nucleus/cytoplasm ratio, decreased cytosolicbasophilia, and appearance of specific granules) were detectablein cytokine-treated HSC but not in FTS3 + VPA–treated cells(Fig. 1C).Fluorescence-activated cell sorting analysis of propidium iodide–

stained cells revealed a consistent increase of cells in G0-G1 phaseof the cell cycle after treatment with FTS3 + VPA, whereas in FTS3-treated cultures a higher percentage of cells was in S-G2M phase.Valproic acid in combination with FTS3 also reduced of 1.5- to 6-fold the frequency of apoptotic cells induced by 7 days of FTS3treatment (median, 7.1; range, 0-22.8%) in all the HSC sources (datanot shown).Accordingly, immunophenotypic analysis of 7-day FTS3 ampli-

fication cultures showed that the percentage of total CD34+ cellswas significantly decreased whereas in cultures treated with FTS3 +VPA the CD34+ cells were highly preserved either for CB (76.6%versus 11.2%, P < 0.001), MPB (88.9% versus 42.2%, P = 0.003), orBM (40% versus 19.6%, P = 0.007; Fig. 2A). More impressive was theeffect of valproic acid on the early stem cell subsets CD34+CD90+

Figure 1. Effect of cytokines eithercombined or without valproic acid (VPA) onthe accumulation of acetylated histone H3and H4 and differentiation of human CD34+HSC. A, whole cell lysates were preparedfrom CD34+ cells isolated from two MPBand two BM cases after treatment for 16hours with FT or FTS3 in the presenceor in the absence of VPA as described inthe method section. Immunoblot analysiswas done using the anti-acetylated histoneH3 and H4 antibodies. H3 and H4acetylation levels of freshly isolateduntreated cells. Expression of histoneH3 was used as a control for histoneacetylation levels and samples loading. B,immunofluorescence analysis of theacetylation status of histone H3 inCD34+ HSC from representative CB, MPB,and BM samples treated with FTS3 orFTS3 + VPA as indicated for 16 hours,immunostained with anti-acetylated histoneH3 antibody (FITC, green ), counterstainedwith Hoechst 33258 to visualize nuclei(blue ), and analyzed at 40� or 100�magnification. Untreated cells as controls.C, Wright-Giemsa staining of CD34+ HSCisolated from representative CB, MPB, andBM samples treated for 7 days with FTS3or FTS3 + VPA. Cell morphology followingthe isolation and before treatment (day 0).Magnification, 100�.

Ex vivo Expansion of CD34+ Cells by Valproic Acid




and CD34+CD38�. The percentage of CD34+CD90+ detected incultures with FTS3 + VPA or FTS3 alone were respectively 37.2%versus 0.26% (P < 0.001) for CB, 65.3% versus 1.9% (P < 0.001) forMPB and 25.6% versus 1.3% (P = 0.015) for BM. Similar results wereobtained for the CD34+CD38�cell fraction with higher values inFTS3 + VPA than in FTS3 cultures (Fig. 2A and B). Interestingly,in all the three HSC sources the percentages of the early cellfraction CD34+CD90+ after 7 days of amplification with FTS3 +VPA were higher than at the start of cultures.Cell expansion was evaluated in terms of fold increase by

dividing the total number of nucleated cells recovered at day 7 bythe input cell number. As shown in Fig. 3, a lower cell proliferationwas detected in FTS3 + VPA cultures than in cultures treated withcytokines alone (with an increase of nucleated cells of 25.1- versus55.2-fold for CB, 21.7- versus 30.6-fold for MPB and 2.2- versus 3.5-fold for BM). The average viability, measured by trypan blue dyeexclusion, in FTS3 and FTS3 + VPA cultures, was unchanged (95%,96%, and 97%, respectively). These results, along with the increasednumber of cells in G0/G1 phase and the reduced incidence ofapoptosis in FTS3 + VPA compared with FTS3-treated cells (datanot shown), suggest that valproic acid lengthened the duration ofthe cell cycle rather than induce cell death. Notably, despite thelower cell proliferation, the coexposure to FTS3 + VPA induced asignificantly higher amplification of CD34+, CD34+CD90+, andCD34+CD38� cells than FTS3 alone (Fig. 3).Consistent with a higher immaturity of the cells amplified with

FTS3 + VPA the number of the CD34+ cells giving rise to primarycolonies (plating efficiency) was reduced: 5% versus 56% (P = 0.017)for CB, 3.3% versus 7.3% (P = 0.012) for MPB, and 3.1% versus 8.8%for BM (Fig. 2A). Furthermore, the replating tests showed that ahigher number of secondary colonies were obtained fromindividual granulocyte macrophage colony-forming unit, burstforming unit-erythroid, and granulocyte, erythroid, macrophage,megacaryocyte colony-forming unit generated in amplificationcultures with FTS3 + VPA than with FTS3 alone (6.9 F 5.3 versus1.3 F 0.3). Therefore, despite the low absolute number (Fig. 3), the

hematopoietic progenitors preserved a higher replating potential invalproic acid cultures.Effect of Valproic Acid on Short-term Amplification Induced

by the Early Acting Cytokines FL and Thrombopoietin. Theeffect of valproic acid was also evaluated in expansion culturesstimulated only by early acting cytokines FL + thrombopoietin (FT)in three MPB, and one CB samples. The results have beenexpressed as mean of the four cases because no significantdifference was detected among the different cell sources.Although the CD34+ cells were more effectively preserved incultures stimulated by FT compared with FTS3, the addition ofvalproic acid further increased the percentage of either CD34+cells (79% versus 61.9%) or the early cell fraction CD34+CD90+(65.4% versus 8.6%, P < 0.001; Fig. 2C). The plating efficiency ofCD34+ cells amplified with FT + VPA was 50% of cells amplifiedwith FT alone. Notably, immunoblot analysis, showed thatvalproic acid was effective in increasing the histone acetylationlevels induced by 16 hours FT treatment (Fig. 1A).Effect of Valproic Acid on Hematopoietic Stem Cell

Expansion over 3-Week Cultures. The long-term effect of valproicacid on HSCs amplification was investigated in 3-week cultures withFTS3 in six cases (MPB = 4 and CB = 2). At day 21, morphologicanalysis showed an immature phenotype in <50% of MPB or CB cellsamplified with FTS3 alone, and in more than 75% of cells expandedwith FTS3 + VPA (Fig. 4A). Accordingly, more pronounced features ofmyelomonocytic differentiation were detected in FTS3-treated cellsthan in cultures supplemented with FTS3 + VPA. Cytochemicalstaining that allow a functional characterization of myeloid cells wasalso done. Myeloperoxidase (MPO) activity was detected in about35% of FTS3-treated cells, and in 4% to 8% of cells treated with FST3+ VPA. Likewise, staining for sodium fluoride–inhibited naphthylacetate esterase activity revealed a higher percentage of monocytesin FTS3-treated cells than in cells treated with FTS3 + VPA (Fig. 4B).Immunophenotypic analysis was done with a wide panel of

monoclonal antibodies to evaluate the early stem cell pool, theintermediate progenitor cell compartment, and the differentiated

Figure 2. VPA enhances the frequency ofearly HSCs induced by cytokines in 7 and21 day amplification cultures. Thefrequency of total CD34+ cells, their moreprimitive cell subsets and the clonogenicpotential expressed as plating efficiency(PE) of CD34+ cells, was evaluated before(5) and after amplification with cytokinesalone ( ) or cytokines + VPA (n). A, effectof valproic acid (VPA ) was evaluated ondifferent cell sources (CB, n = 3; MPB,n = 4; BM, n = 2), after 7 days amplificationwith the combination of four cytokines(FTS3). B, fluorescence-activated cellsorting plots of CD34+ cell fractions from arepresentative MPB sample before (day 0)and after 7 and 21 days amplification withFTS3 F VPA. C, effect of VPA oncytokine-induced amplification evaluatedafter 7 days amplification with only twoearly acting cytokines (FT) in three MPBand one CB samples. Columns, mean;bars, FSE. The difference between theeffect induced by cytokines alone orcytokines + VPA was evaluated byStudent’s t test. *, P < 0.05; **, P < 0.001.

Cancer Research




cells generated along the expansion. A significantly slower decreaseover time of the number of CD34+ cells was observed in culturessupplemented weekly with FTS3 + VPA (1st week, 84.6%; 2nd week,46.8%; 3rd week, 32.5%) compared with cultures stimulated withFTS3 alone (1st week, 32.3%; P = 0.005; 2nd week, 3.1%; P < 0.001;3rd week, 0.6% P = 0.002; Fig. 5). On the 3rd week in FTS3 + VPAcultures, the percentage of CD34+ cells (32.5 F 7.5%) wascomparable to the number of positive cells (32.2 F 11.0%) detectedat day 7 in amplification cultures with FTS3 alone.Notably, at the 3rd week of culture, a higher number of cells

amplified in presence of valproic acid maintained the character-istics of primitive stem cells as documented by the significantincrease of CD34+CD90+ cells (17.7% versus 0.14%, P = 0.002) andCD34+CD38� cells (3.4% versus 0.04%, P = 0.018). Finally, aconsistent fraction (10.2%) of the CD34+ cells generated inpresence of valproic acid coexpressed the AC133, which is an earlyhematopoietic cell marker rapidly down regulated during differen-tiation (Figs. 2B and 5). The c-kit receptor (c-kitR/CD117) wasdetected at a low level in valproic acid cultures compared with thelack of expression of this marker among cells amplified with FTS3alone (CD34+CD117+ 2.0% versus 0%).Despite the significantly lower number of total nucleated cells

recovered after amplification with FTS3 + VPA, the CD34+cells generated after 3 week expansion increased over the inputcell number 35.1-fold compared with 4.5-fold in presence of FTS3.Notably, the early subpopulations CD34+CD90+, CD34+CD38�, andCD34+CD133+ were amplified respectively 16.1-, 7.2-, and 16.2-foldby the 3rd week, whereas no amplification of these cell fractionswas detected among cells expanded with FTS3.The progenitor cells (CD34+CD33+) and early myeloid cells

(CD33+ and CD13+) were more represented in valproic acidamplification cultures although the differences were not statis-tically significant. Indeed, a significantly lower percentage ofdifferentiated cells (CD15+) was produced along the 3 weeks ofamplification in presence of valproic acid than after expansionwith only FTS3 (1st week, 10.5% versus 34%, P = 0.05; 2nd week,12.1% versus 48.4%, P = 0.014; 3rd week, 15.4% versus 57.9%,P < 0.001). By contrast, differentiated cells of erythroid (CD235a+)and megakaryocytic lineages (CD61+) were more frequent in thepresence of valproic acid than in cultures treated with FTS3alone (Fig. 5).In amplification cultures with FTS3 the plating efficiency of

CD34+ cells, after a drop on day 7, increased progressively and wascomparable to baseline level by the 3rd week. In agreement withthe higher immaturity of the cells, a progressive decrease of theplating efficiency was detectable in valproic acid cultures (Fig. 5).Effect of Valproic Acid on Genes Expressed at Different

Stages of Hematopoiesis. The effect of FTS3 and FTS3 + VPAon the expression of genes implicated in the early developmentalstages of HSC (CD34 , KDR , c-kit R , and AC133) or during lineagecommitment and differentiation (GATA1 , MPO , GM-CSFr, G-CSFr,MSE , and Gp1ab) was investigated in four cases (MPB = 2 andCB = 2) by quantitative reverse transcription-PCR. At days 7, 14,and 21, either FTS3 or FTS3 + VPA modulated the expression ofmost genes when compared with freshly isolated MPB or CBCD34+ cells (day 0), although with differences in potency andkinetics between the two cell sources (Fig. 6A and B). Inagreement with the results of immunophenotypic analysis, CD34expression decreased during the long-term cultures in both FTS3and FTS3 + VPA–treated cells. However, CD34 mRNA was about3- to 7-fold more abundant in valproic acid-treated cells than in

cultures treated with cytokines alone either in MPB or CBexpansion cultures (Fig. 6 and data not shown). Remarkably, theexpression of KDR , AC133 and c-kitR were induced by FTS3 +VPA with levels higher than those measured in freshly isolatedCD34+ cells on days 14 and 21. GATA1 , which is present eitherin HSCs or in erythroid and megacaryocytic progenitors was alsoincreased by FTS3 + VPA at 14 to 21 days compared with FTS3cultures or freshly isolated MPB or CB CD34+ cells (data notshown), whereas GM-CSFr, a common myeloid progenitor-affiliated gene expressed during the early and late stages ofgranulocytic and monocytic differentiation was expressedat comparable levels in valproic acid and control cultures(Fig. 6A and B).Expression of genes characterizing late stages of hematopoietic

differentiation, such as G-CSFr, glycoprotein 1ab (Gp1ab), MPO ,and MSE was significantly higher in FTS3-treated cells than in cellstreated with FTS3 + VPA along 3 weeks amplification (Fig. 6B).We next assessed the effect of valproic acid on the expression

of genes involved in HSC self-renewal such as Wnt3A , b catenin ,Notch-1 , and HOXB4 (18, 19, 39). The expression of b catenin andNotch-1 remained unchanged and Wnt3A was undetectable(Fig. 6C and data not shown). In contrast, the homeogeneHOXB4 was significantly induced in FTS3 + VPA cultures at both

Figure 3. Effect of valproic acid (VPA ) on the amplification of early hematopoieticstem cell subpopulations and committed progenitors in 7 day cultures withFTS3. The expansion of total nucleated cells (NC ), CD34+, CD34+CD90+,CD34+CD38�, cells and committed progenitors (CFU-GM , BFU-E , andCFU-GEMM) after 7 day cultures with FTS3 alone or in association with VPA hasbeen evaluated as fold increase over the input (day 0) values in CB (n = 3), MPB(n = 4), and BM (n = 2) samples. Columns, mean; bars, FSE. The statisticaldifferences between the mean fold increase detected in the two amplificationconditions were evaluated by Student’s t test. *, P < 0.05.





RNA and protein levels (Fig. 6C and D), with a strong increaseat day 21 compared with a minimal and transient increase inFTS3-treated cells.Valproic Acid Increases Histone Acetylation Levels on

Chromatin at AC133 and HOXB4 Regulatory Sites. To addresswhether chromatin remodeling events are involved in the effect of

valproic acid on the ex vivo amplification of CD34+ cells by

cytokines, we investigated by ChiP analysis on chromatin prepared

from CD34+ cells isolated from MPB (two cases), the changes in theacetylation status of histone H4 at regions surrounding the exon1A

of AC133 whose expression seems restricted to CD34+ cells and the

HOXB4 promoter, which contain the promoter response element 1(HxRE-1) and the E-box regulatory regions (36–38). Upon 2 hours

treatment, FTS3 changed the acetylation levels on histone H4 at

these regulatory sites on both AC133 and HOXB4 that becamehyperacetylated after the addition of valproic acid either in absence

or presence of FTS3 (Fig. 6E). The specificity of the chromatin

changes at these sites was indicated by the absence of modulationin the acetylation status of histone H4 in the same samples when

distal sequences were amplified by PCR using specific primers.

Thus, the functional effects of valproic acid on AC133 and HOXB4

gene expression positively correlated with the acetylation status ofhistone H4 at regulatory sites present on their promoters.To relate the changes on chromatin accessibility by valproic acid

with the enhancement of the cytokine effect on the maintenance ofa primitive HSC population, we analyzed the effect of a singleexposure to valproic acid on 3 weeks amplification of CD34+ cellsfrom 3 MPB and 1 CB. The immunophenotype of cells exposed to asingle dose of valproic acid and cultured 3 weeks in its absence wassimilar to that of control cells never exposed to valproic acid andcultured with FTS3 alone (Fig. 5). However, by the 3rd week ofculture, the fraction of CD34+CD90+ cells was significantly higherin cells exposed to valproic acid than in control cultures (0.48 F0.13% versus 0.14 F 0.06, P = 0.034; Fig. 5).

Discussion

A requisite for successful in vitro expansion of human HSC forapplication in stem cell transplantation settings and gene therapy

would be to efficiently induce stem cell proliferation withouta concomitant loss of the self-renewing potential. Ex vivoamplification protocols have shown that several cytokine combi-nations may induce a rapid cell cycling of HSCs resulting in asignificant expansion of a partially differentiated HSC compartmentthat can be useful to shorten the period of post-transplant aplasia(40). Nevertheless, the expansion or maintenance of the early HSCswith long-term repopulating ability is still controversial (40).In this study, we have investigated the effect of valproic acid,

an inhibitor of HDAC, in amplification cultures stimulated bycombinations of either four (FTS3) or two early acting cytokines(FT). Our results provide evidence that valproic acid significantlyenhances the effect of cytokines on the amplification of HSC andselectively increases the early HSC compartment in 1- to 3-weekexpansion cultures. This effect of valproic acid has beendocumented in all the HSC sources and is correlated with thehyperacetylation of the nucleosomal histones H3 and H4.In all the HSC sources, the addition of valproic acid to cultures

stimulated by either FTS3 or FT decreased the overall number ofex vivo generated cells. However, the percentage of the cellsretaining the expression of CD34 antigen after 7-day cultures wassignificantly higher (1.5- to 3-fold) than with FT and FTS3 alone.Previous studies on CB amplification reported percentages of

CD34+ cells consistently <20% with a 10-fold expansion of CBCD34+ (41, 42). In agreement with these results, our data showed a9.4-fold increase of CB CD34+ cells after 7 days of expansion withFTS3. Notably, the addiction of valproic acid significantly improvedthe amplification of CD34+ cells with an increase of about 25-foldin samples from either fresh or thawed CB. Furthermore, theexpression of CD34+ cells was kept significantly higher in presenceof valproic acid till the 3rd week of liquid culture, whereas previousamplification studies showed a marked decrease with time of thepercentage of CD34+ cells, independently from the cytokinecombination used (43).Surprisingly, further characterization of the CD34+ cells showed

that the CD34+ subpopulations lacking the CD38 antigen orcoexpressing the antigens CD90 and AC133 were significantly morepreserved in amplification cultures with FTS3 and valproic acidthan with growth factors alone either after 1- or 3-week cultures inall the HSC sources. Several studies have documented that these

Figure 4. Valproic acid (VPA ) addition maintains a highpercentage of morphologically and functionally immature CD34+HSC in amplification cultures with multiple cytokines. Morphologicand functional analyses were done in CD34+ cells from MPB(n = 2) and CB (n = 2) before (day 0) and after 7, 14, and 21 daysof treatment with FTS3 or FTS3 + VPA. A, % immature cells(Immature, 5), myelomonocytic cells of the granulocytic (Gr, )and monocytic (Mo, n) lineage. B, functional characterization ofmyelomonocytic cells was done by cytochemical staining of MPBand CB cells. Cells were evaluated for positive staining of MPO,a naphthyl acetate esterase (ANAE ) with or without sodiumfluoride (NaF ) as described in Materials and Methods. At day 0,all the cells were negative for these cytochemical staining.Means FSD of two separate experiments.

Cancer Research




markers identify cell subsets in early stages of hematopoieticdevelopment. The AC133 has been shown to be expressed on stem/progenitor cell fractions including CD34+CD38� cells that have along-term repopulating potential as assessed in fetal sheep andNOD/SCID transplantation models (44, 45). Interestingly, we foundthat in MPB samples, valproic acid addition greatly hyperacetylatethe chromatin surrounding the exon 1A, a region generating theAC133 mRNA isoform specifically expressed in CD34+ HSC (38). Inour study, the higher immaturity of CD34+ cells generated incultures supplemented by FTS3 + VPA was also confirmed by the 2-to 4-fold increased expression of genes associated to an earlyhematopoietic stem cell compartment (KDR , CD117 , and HOXB4),compared with the levels detected in cultures with FTS3 alone.Recently, several studies have shown differences in long-term

engraftment potential of resting and cycling hematopoietic stemcells providing evidences that phenotypic changes associated withcell cycle phases are reversible (46, 47). Growth factors exposureinduces quiescent, early HSC to proceed through cell cycle givingrise either to cells retaining HSC properties, or to cells committed

toward differentiation. In amplification cultures supplementedwith valproic acid, a lower number of cells were in the active phaseof cell cycle (S-G2-M) and a significantly higher percentage of cellsstill was in G1 phase on the 3rd week of expansion compared withthe higher proliferative rate induced by growth factors alone (datanot shown). The lower rate of proliferation induced by valproic acidcorrelates with the immaturity of the cells generated in cultures.Indeed, the lower plating efficiency of committed hematopoieticprogenitors in primary cultures and the higher number ofsecondary colonies originated from individual clones suggests thatvalproic acid treatment induces one or more self-renewal divisions

Figure 5. Immunophenotypic profile of cells amplified over 3 weeks with FTS3alone or in association with valproic acid (VPA ). The characterization of theamplified cells from MPB (n = 4) and CB (n =2) was done weekly on cellsgenerated in vitro along 3 weeks expansion cultures supplemented with FTS3alone (�) and after coexposure to FTS3 and VPA either added to cultures everyweek (��) or only once at the start of cultures (��). Early, intermediate, andlate hematopoietic cells were characterized by a wide panel of MoAbs. Points,mean; bars, FSE. The statistical differences between the mean values of cellsubpopulations detected in FTS3 and FTS3 + VPA amplification cultures wereevaluated by Student’s t test. *, P < 0.05; **, P < 0.001. Fold amplification (FI )of the different cell subsets over the input cell number.

Figure 6. Valproic acid (VPA) modifies the gene expression profiles inducedby cytokines in CD34+ HSC cells. Time-dependent effect of FTS3 (�) andFTS3 + VPA (��) on the expression of (A) genes of HSC earlydevelopmental stages (CD34 , KDR , AC133 , CD117 , and GATA-1), (B) genesof different stages during late hematopoietic differentiation (MPO , GM-CSF ,G-CSF , MSE , and Gp1ab), and (C ) genes potentially involved in HSCself-renewal (HOXB4 , hcatenin , and Notch-1 ). Total RNA was extracted fromCD34+ MPB cells before (day 0) or after 7, 14, and 21 days treatment withFTS3 or FTS3 + VPA. Relative gene expression in treated cells compared withfreshly isolated CD34+ cells (day 0) was determined by quantitative reversetranscriptase-PCR. Points, means of two separate experiments; bars, FSD. D,immunoblot analysis of HOXB4 expression in CD34+ MPB cells treated ornot with FTS3 in the presence or absence of VPA done as described inMaterials and Methods. Histone H3 was probed as a protein loading control.Representative case of two that gave similar results. E, effect of cytokines and/or VPA on histone H4 hyperacetylation at regulatory chromatin sites ofHOXb4 and AC133 genes. CD34+ MPB cells were treated or not (control)with VPA, FTS3, and FTS3 + VPA for 2 hours. ChiP assay was done with ananti-acetyl-histone H4 antibody (H4-Ac ) or no antibody (no Ab) and analyzedby PCR using primers corresponding to the regulatory sites on the HOXB4(P1) and AC133 (Ex1A) genes that are active in hematopoietic cells or to distalregions on HOXB4 (P5) and AC133 (ExE) as described in Materials andMethods. The amplification of GAPDH was used to detect nonrelevant cellularDNA sequences present in the samples. A sample representing 0.02% of totalinput chromatin (input ) was included in the PCR analysis. Representative caseof two that gave similar results.





of the colony-forming cells before the enter a pathway of terminaldifferentiation. Although these results suggest that valproic acidaffects the amplification of HSC by promoting self-renewal, thedifferentiation capacity toward all the hematopoietic lineages ismaintained as indicated by the detection of mature cells expressingspecific markers for all the hematopoietic lineages (CD13, CD15,CD235a, and CD61).While this article was in preparation, another group (48)

reported that human bone marrow CD34+ cells amplified in atwo-step culture system with cytokine cocktails specific to promoteHSC cycling (1st step) or differentiation (2nd step) and sequentiallyexposed to 5-aza-2Vdeoxycytidine alone or in combination withTrichostatin A, increased hematopoietic progenitor cells with aprimitive phenotype (CD34+CD90+) that were able to engraft NOD/SCID mice (48). 5-Aza-2Vdeoxycytidine is a chromatin remodelingagent that function as an inhibitor of DNA methyltransferaseactivities. Trichostatin A is an HDAC inhibitor structurally andfunctionally different from valproic acid (29, 32, 49).The effectiveness of different agents altering chromatin structure

in enhancing the amplification of primitive HSC subsets is inagreement with the hypothesis that reversible phenotypic shifts inhematopoietic cells depend on chromatin remodeling and theconsequent expression of genes in response to environment-derived stimuli (50). Chromatin itself may serve as a genomicintegrator of various extracellular and intracellular pathways.Among the intracellular factors that induce self-renewal of HSC,ectopic expression of the transcription factor HOXB4 increases thenumber of transplantable HSCs retaining the ability to repopulatein vivo myeloid and lymphoid lineages (24, 25). Here we show thatin MPB CD34+ cells the acetylation status of histones on specificDNA binding sites within the HOXB4 promoter is increased by 2hours of exposure to valproic acid. Notably, the addition of FTS3further enhances the valproic acid effect on the chromatinacetylation status at this site. Increased expression levels of HOXB4gene and protein were also detectable after 1 or 3 weeks of culturein FTS3 + VPA–treated cells, compared with FTS3 alone. Thus, itseems that valproic acid by inducing chromatin modifications atcritical DNA-binding sites leads to transcriptional derepression ofspecific target genes and enhances the effect of early actingcytokines. Accordingly, a single exposure to valproic acid at thestart of the amplification culture maintained a significantly higher

fraction of CD34+CD90+ cells over 3 weeks compared with culturesnever exposed to valproic acid.A significant role of histone deacetylases in hematopoiesis is

further highlighted by findings showing that an abnormalregulation of HDAC activities is a leukemogenic event underlyingthe block of differentiation in acute myeloid leukemias. Regardlessof their primary genetic lesion, acute myeloid leukemias can bealso induced to differentiate by treatment with HDAC inhibitorsvalproic acid and Trichostatin A in vitro and in vivo (28–33).These evidences suggest a scenario where valproic acid can act

as epigenetic regulator of both stem cell maintenance anddifferentiation of committed progenitors into different lineages(28–33). The chromatin accessibility at specific DNA-binding sitesmight represent the key event for the activity of cytokines ortranscription factors either involved in the maintenance of earlyHSC population or in lineage commitment. Some of these factorsmight be present in the same cellular context and dictate cell fatechoice by targeting enzymes with chromatin remodeling activitysuch as HDACs, histone acetyltransferases, or methylases atspecific gene loci. Valproic acid can therefore initiate a series ofevents leading to the maintenance of the undifferentiated state inearly HSC, or generating a chromatin code coupled to specificdifferentiation decision in HPC and acute myeloid leukemia blasts.In conclusion, this study present in vitro evidence for a new and

relevant effect of valproic acid on human hematopoietic stem cellsand highlight the potentiality of novel epigenetic approaches forthe ex vivo amplification of HSC aimed to transplantation, geneand stem cell therapies. However, whether valproic acid enhancesthe plasticity, self-renewal, and/or engraftment potential of long-term repopulating stem cells in vivo are relevant questions thatneed further investigation in animal models.

Acknowledgments

Received 8/24/2004; revised 11/19/2004; accepted 12/10/2004.Grant support: Italian Ministry of Health, National Program on Stem Cells,

European Community grants QLK3-CT-2002-01918 and QLG1-2001-01935, ItalianMinistry of University and Research (Cofinanziamento and Fondo per gli investimentidella ricerca di base), and Associazione Italiana per la Ricerca sul Cancro.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Drs. Paolo Bianco, Carmelo Carlo Stella, Giulio Cossu, and DariaBrambilla for critical reading of the article and helpful advice.

References1. Gluckman E, Rocha V, Boyer-Chammard A, et al.Outcome of cord-blood transplantation from relatedand unrelated donors. Eurocord Transplant Group andthe European Blood and Marrow TransplantationGroup. N Engl J Med 1997;337:373–81.

2. Sanz GF, Saavedra S, Planelles D, et al. Standardized,unrelated donor cord blood transplantation in adultswith hematologic malignancies. Blood 2001;98:2332–8.

3. Laughlin MJ, Barker J, Bambach B, et al. Hematopoi-etic engraftment and survival in adult recipients ofumbilical-cord blood from unrelated donors. N Engl JMed 2001;344:1815–22.

4. Wagner JE, Barker JN, DeFor TE, et al. Transplantationof unrelated donor umbilical cord blood in 102 patientswith malignant and nonmalignant diseases: influence ofCD34 cell dose and HLA disparity on treatment-relatedmortality and survival. Blood 2002;100:1611–8.

5. Conneally E, Cashman J, Petzer A, Eaves C. Expansionin vitro of transplantable human cord blood stem cellsdemonstrated using a quantitative assay of their

lympho-myeloid repopulating activity in nonobesediabetic-scid/scid mice. Proc Natl Acad Sci U S A 1997;94:9836–41.

6. Piacibello W, Sanavio F, Severino A, et al. Engraftmentin nonobese diabetic severe combined immunodefi-cient mice of human CD34(+) cord blood cells afterex vivo expansion: evidence for the amplification andself-renewal of repopulating stem cells. Blood 1999;93:3736–49.

7. Shah AJ, Smogorzewska EM, Hannum C, Crooks GM.Flt3 ligand induces proliferation of quiescent humanbone marrow CD34+. Blood 1996;87:3563–70.

8. Piacibello W, Sanavio F, Garetto L, et al. Extensiveamplification and self-renewal of human primitivehematopoietic stem cells from cord blood. Blood1997;89:2644–53.

9. Petzer AL, Zandstra PW, Piret JM, Eaves CJ. Differen-tial cytokine effects on primitive (CD34+ CD38�)human hematopoietic cells: novel responses to Flt3-ligand and thrombopoietin. J Exp Med 1996;183:2551–8.

10. Yonemura Y, Ku H, Lyman SD, Ogawa M. In vitroexpansion of hematopoietic progenitors and mainte-

nance of stem cells: comparison between FLT3/FLK-2ligand and KIT ligand. Blood 1997;89:1915–21.

11. Dao MA, Hannum CH, Kohn DB, Nolta, JA. FLT3ligand preserves the ability of human CD34+ progen-itors to sustain long-term hematopoiesis in immune-deficient mice after ex vivo retroviral-mediated trans-duction. Blood 1997;89:446–56.

12. De Felice L, Di Pucchio T, Mascolo MG, et al. Flt3Linduces the ex-vivo amplification of umbilical cordblood committed progenitors and early stem cells inshort-term cultures. Br J Haematol 1999;106:133–41.

13. Kobari L, Pflumio F, Giarratana M, et al. In vitro andin vivo evidence for the long-term multilineage(myeloid, B, NK, and T) reconstitution capacity ofex vivo expanded human CD34(+) cord blood cells. ExpHematol 2000;28:1470–80.

14. Bryder D, Jacobsen SE. Interleukin-3 supportsexpansion of long-term multilineage repopulatingactivity after multiple stem cell divisions in vitro . Blood2000;96:1748–55.

15. Traycoff CM, Cornetta K, Yoder MC, Davidson A,Srour EF. Ex vivo expansion of murine hematopoietic

Cancer Research




progenitor cells generates classes of expanded cellspossessing different levels of bone marrow repopulatingpotential. Exp Hematol 1996;24:299–306.

16. Guenechea G, Segovia JC, Albella B, et al. Delayedengraftment of nonobese diabetic/severe combinedimmunodeficient mice transplanted with ex vivo -expanded human CD34(+) cord blood cells. Blood1999;93:1097–105.

17. Kondo M, Wagers AJ, Manz MG, et al. Biology ofhematopoietic stem cells and progenitors: implicationsfor clinical application. Annu Rev Immunol 2003;21:759–806.

18. Dick JE. Self-renewal writ in blood. Nature 2003;423:231–3.

19. Reya T. Regulation of hematopoietic stem cell self-renewal. Recent Prog Horm Res 2003;58:283–95.

20. Lessard J, Sauvageau G. Bmi-1 determines theproliferative capacity of normal and leukaemic stemcells. Nature 2003;423:255–60.

21. Park IK, Qian D, Kiel M, et al. Bmi-1 is required formaintenance of adult self-renewing haematopoieticstem cells. Nature 2003;423:302–5.

22. Amsellem S, Pflumio F, Bardinet D, et al. Ex vivoexpansion of human hematopoietic stem cells by directdelivery of the HOXB4 homeoprotein. Nat Med 2003;9:1423–7.

23. Krosl J, Austin P, Beslu N, Kroon E, Humphries RK,Sauvageau G. In vitro expansion of hematopoietic stemcells by recombinant TAT-HOXB4 protein. Nat Med2003;9:1428–32.

24. Sauvageau G, Thorsteinsdottir U, Eaves CJ, et al.Overexpression of HOXB4 in hematopoietic cells causesthe selective expansion of more primitive populationsin vitro and in vivo . Genes Dev 1995;9:1753–65.

25. Antonchuk J, Sauvageau G, Humphries RK. HOXB4-induced expansion of adult hematopoietic stem cellsex vivo . Cell 2002;109:39–45.

26. Jaenisch R, Bird A. Epigenetic regulation of geneexpression: how the genome integrates intrinsicand environmental signals. Nat Genet 2003;33Suppl:245–54.

27. Grewal SI, Moazed D. Heterochromatin andepigenetic control of gene expression. Science 2003;301:798–802.

28. Marks PA, Richon VM, Rifkind RA. Histone deacety-lase inhibitors: inducers of differentiation or apoptosisof transformed cells. J Natl Cancer Inst 2000;92:1210–6.

29. Johnstone RW, Licht JD. Histone deacetylase inhib-itors in cancer therapy: is transcription the primarytarget? Cancer Cell 2003;4:13–8.

30. Egger G, Liang G, Aparicio A, Jones PA. Epigenetics inhuman disease and prospects for epigenetic therapy.Nature 2004,429:457–63.

31. Minucci S, Nervi C, Lo Coco F, Pelicci PG. Histonedeacetylases: a common molecular target for differen-tiation treatment of acute myeloid leukemias? Onco-gene 2001;20:3110–5.

32. Gottlicher M, Minucci S, Zhu P, et al. Valproic aciddefines a novel class of HDAC inhibitors inducingdifferentiation of transformed cells. EMBO J 2001;20:6969–78.

33. Ferrara FF, Fazi F, Bianchini A, et al. Histonedeacetylase targeted treatment restores retinoic acidsignaling and differentiation in acute myeloid leukemia.Cancer Res 2001;61:2–7.

34. Kuendgen A, Strupp C, Aivado M, et al. Treatment ofMyelodysplastic Syndromes with valproic acid alone orin combination with all-trans retinoic acid. Blood 2004;104:1266–9.

35. Rubinstein P, Dobrila L, Rosenfield RE, et al.Processing and cryopreservation of placental/um-bilical cord blood for unrelated bone marrowreconstitution. Proc Natl Acad Sci U S A 1995;9210119–22.

36. Giannola DM, Shlomchik WD, Jegathesan M, et al.Hematopoietic expression of HOXB4 is regulated innormal and leukemic stem cells through transcriptionalactivation of the HOXB4 promoter by upstreamstimulating factor (USF)-1 and USF-2. J Exp Med 2000;192:1479–90.

37. Zhu J, Giannola DM, Zhang Y, Rivera AJ, EmersonSG. NF-Y cooperates with USF1/2 to induce thehematopoietic expression of HOXB4. Blood 2003;102:2420–7.

38. Shmelkov SV, Jun L, St Clair R, et al. Alternativepromoters regulate transcription of the gene thatencodes stem cell surface protein AC133. Blood2004;103:2055–61.

39. Reya T, Duncan AW, Ailles L, et al. A role for Wntsignalling in self-renewal of haematopoietic stem cells.Nature 2003;423:409–14.

40. Devine SM, Lazarus HM, Emerson SG. Clinicalapplication of hematopoietic progenitor cell expansion:current status and future prospect. Bone MarrowTransplant 2003;31:241–52.

41. Encabo A, Mateu E, Carbonell-Uberos F, Minana MD.CD34+. Transfusion 2003;43:383–9.

42. Zhai QL, Qiu LG, Li Q, et al. Short-term ex vivoexpansion sustains the homing-related properties ofumbilical cord blood hematopoietic stem and progen-itor cells. Haematologica 2004;89:265–73.

43. Tanavde VM, Malehorn MT, Lumkul R, et al. Humanstem-progenitor cells from neonatal cord blood havegreater hematopoietic expansion capacity than thosefrom mobilized adult blood. Exp Hematol 2002;30:816–23.

44. Yin AH, Miraglia S, Zanjani ED, et al. AC133, a novelmarker for human hematopoietic stem and progenitorcells. Blood 1997;90:5002–12.

45. de Wynter EA, Durig J, Cross MA, Heyworth CM,Testa NG. Differential response of CD34+ cells isolatedfrom cord blood and bone marrow to MIP-1 a and theexpression of MIP-1 a receptors on these immaturecells. Stem Cells 1998;16:349–56.

46. Habibian HK, Peters SO, Hsieh CC, et al. Thefluctuating phenotype of the lymphohematopoieticstem cell with cell cycle transit. J Exp Med 1998;188:393–8.

47. Lambert JF, Liu M, Colvin GA, et al. Marrow stemcells shift gene expression and engraftment pheno-type with cell cycle transit. J Exp Med 2003;197:1563–72.

48. Milhem M, Mahmud N, Lavelle D, et al. Modificationof hematopoietic stem cell fate by 5aza 2Vdeoxycytidineand trichostatin A. Blood 2004;103:4102–10.

49. Kramer OH, Zhu P, Ostendorff HP, et al. The histonedeacetylase inhibitor valproic acid selectively inducesproteasomal degradation of HDAC2. EMBO J2003;22:3411–20.

50. Quesenberry PJ, Colvin GA, Lambert JF. The chiar-oscuro stem cell: a unified stem cell theory. Blood2002;100:4266–71.





2005;65:1505-1513. Cancer Res Lidia De Felice, Caterina Tatarelli, Maria Grazia Mascolo, et al. CellsCytokine-Induced Expansion of Human Hematopoietic Stem Histone Deacetylase Inhibitor Valproic Acid Enhances the

Updated version

http://cancerres.aacrjournals.org/content/65/4/1505

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/65/4/1505.full#ref-list-1

This article cites 49 articles, 26 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/65/4/1505.full#related-urls

This article has been cited by 17 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/65/4/1505To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/content/65/4/1505.full#ref-list-1

http://cancerres.aacrjournals.org/content/65/4/1505.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Documents

Histone Deacetylase Inhibitor Valproic Acid Enhances the ...Histone Deacetylase Inhibitor Valproic Acid Enhances the Cytokine-Induced Expansion of Human Hematopoietic Stem Cells Lidia