Clinical Cancer Research Meeting Calendar Frontiers in Basic Cancer Research
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Clinical Cancer Research 13, 1726, March 15, 2007. Published Online First February 26, 2007;
doi: 10.1158/1078-0432.CCR-06-2569
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
1078-0432.CCR-06-2569v1
13/6/1726    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Starkova, J.
Right arrow Articles by Trka, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Starkova, J.
Right arrow Articles by Trka, J.

Human Cancer Biology

The Identification of (ETV6)/RUNX1-Regulated Genes in Lymphopoiesis Using Histone Deacetylase Inhibitors in ETV6/RUNX1-Positive Lymphoid Leukemic Cells

Julia Starkova1, Jozef Madzo1, Gunnar Cario2, Tomas Kalina1, Anthony Ford3, Marketa Zaliova1, Ondrej Hrusak1 and Jan Trka1

Authors' Affiliations: 1 Childhood Leukaemia Investigation Prague, Department of Paediatric Haematology/Oncology, 2nd Medical School, Charles University Prague, Prague, Czech Republic; 2 Department of Paediatrics, University Hospital Schleswig-Holstein, Kiel, Germany; and 3 Leukaemia Research Fund Centre, Institute of Cancer Research, Chester Beatty Laboratories, London, United Kingdom

Requests for reprints: Jan Trka, Department of Paediatric Haematology/Oncology, 2nd Medical School, Charles University, V uvalu 84, Prague 15006, Czech Republic. Phone: 420-224-436580; Fax: 420-224-436521; E-mail: jan.trka{at}lfmotol.cuni.cz.


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 References
 
Purpose: Chimeric transcription factor ETV6/RUNX1 (TEL/AML1) is believed to cause pathologic block in lymphoid cell development via interaction with corepressor complex and histone deacetylase. We wanted to show the regulatory effect of ETV6/RUNX1 and its reversibility by histone deacetylase inhibitors (HDACi), as well as to identify potential ETV6/RUNX1-regulated genes.

Experimental Design: We used luciferase assay to show the interaction of ETV6/RUNX1 protein, ETV6/RUNX1-regulated gene, and HDACi. To identify ETV6/RUNX1-regulated genes, we used expression profiling and HDACi in lymphoid cells. Next, using the flow cytometry and quantitative reverse transcription-PCR, we measured differentiation changes in gene and protein expression after HDACi treatment.

Results: Luciferase assay showed repression of granzyme B expression by ETV6/RUNX1 protein and the reversibility of this effect by HDACi. Proving this regulatory role of ETV6/RUNX1, we identified, using complex statistical analysis, 25 genes that are potentially regulated by ETV6/RUNX1 protein. In four selected genes with known role in the cell cycle regulation (JunD, ACK1, PDGFRB, and TCF4), we confirmed expression changes after HDACi by quantitative analysis. After HDACi treatment, ETV6/RUNX1-positive cells showed immunophenotype changes resembling differentiation process compared with other leukemic cells (BCR/ABL, ETV6/PDGFRB positive). Moreover, ETV6/RUNX1-positive leukemic cells accumulated in G1-G0 phase after HDACi whereas other B-lineage leukemic cell lines showed rather unspecific changes including induction of apoptosis and decreased proliferation.

Conclusions: Presented data support the hypothesis that HDACi affect ETV6/RUNX1-positive cells via direct interaction with ETV6/RUNX1 protein and that treatment with HDACi may release aberrant transcription activity caused by ETV6/RUNX1 chimeric transcription factor.

Neoplastic transformation is characterized by inadequate proliferation and changes in the mechanisms of differentiation and apoptosis (1). Aberrant transcription factors, frequently present at the diagnosis of leukemias, are probably responsible for one of the steps in leukemogenic transformation in these particular cases. The expression of chimeric transcription factors results in a block of differentiation and apoptosis by interfering with the function of their wild-type counterparts. The role for one of these chimeric transcription factors, ETV6/RUNX1 (TEL/AML1), in leukemogenesis is still not completely understood despite the frequency of t(12;21)-positive acute lymphoblastic leukemia (ALL; up to 25%; refs. 2–4).

In normal hematopoietic cells, the erythroblast transformation–specific (ETS) family transcription factor ETV6 (TEL) is composed of two major functional domains: an NH2-terminal pointed (PNT) domain and a COOH-terminal ETS domain (5–7). RUNX1 [AML1, core binding factor (CBF)-{alpha}2], a member of the RUNX protein family (8), represents an {alpha}-subunit of CBF and displays high homology with a segment of the Drosophila gene runt. RUNX1 is composed of some characteristic domains. The Runt domain is responsible for the binding with DNA and protein-protein interaction, and the transactivation domain is located at the COOH-terminal site (9, 10). The ETV6/RUNX1 fusion protein contains the NH2-terminal part of the ETV6 protein with the dimerization PNT domain and almost the complete RUNX1 protein with all functional domains. Deficiency in RUNX1 expression leads to an early block in hematopoietic differentiation. Therefore, RUNX1 knockout animals completely lack mature hematopoiesis and die embryonically (11, 12). RUNX1 seems to have a dual role in promoting cell cycle progression and differentiation, presumably depending on the presence of different factors that interact with it in different stages of development (13). RUNX proteins are able to either increase or actively inhibit the transcriptional activity of target genes, most likely depending on the specific cell type as well as the particular target gene (14). ETV6 function is essential for the establishment of hematopoiesis of all lineages in the bone marrow (15). ETV6–/– mice are embryonic lethal because of a yolk sac angiogenic defect.

As the DNA-binding domain of RUNX1 is retained in the ETV6/RUNX1 fusion, it is expected that this fusion protein binds to DNA through this domain and functions in a dominant-negative fashion. Several theories propose a mechanism of repressional activity of ETV6/RUNX1. Siu et al. describe heterodimerization of ETV6/RUNX1 and ETV6 that prevents normal ETV6 activity. Another study suggests that the dominant-negative effect of ETV6/RUNX1 is caused by the binding with RUNX1 coactivators (e.g., p300) and sequestering them into a complex localized in the cytoplasm (16, 17). Another theory proceeds from a previously described association of the ETV6 part of the chimeric protein with the nuclear corepressors mSin3A, N-CoR, and histone deacetylase (HDAC)-3, which led to the assumption that ETV6/RUNX1 works as a transcriptional repressor by changing chromatin pattern (7, 18, 19). Published results confirmed a direct negative effect of the ETV6/RUNX1 protein by using reporter constructs driven by regulatory regions derived from hematopoiesis-specific genes including the lymphoid-specific T-cell receptor-ß enhancer and interleukin-3 promoter (10, 20). However, no other target genes of ETV6/RUNX1 expressed in hematopoietic cells have been identified thus far. The proposed mechanism of action via chromatin remodeling using HDACs has not been proved.

HDACs are characterized by their capability to cleave acetyl groups from the lysine residues localized at the ends of histones. The role of HDACs in hematological diseases was described for the first time in a study on acute promyelocytic leukemia. Retinoic acid receptor-{alpha} receptor is able to bind the corepressor complex including HDACs, and this function is extended by binding with its fusion partner promyelocytic leukemia in t(15;17). Another fusion partner of retinoic acid receptor-{alpha}, promyelocytic leukemia zinc finger protein in t(11;17), also binds the corepressors and inhibits transcription of the target genes, consequently blocking myeloid differentiation. The aberrant repression in both in vivo and in vitro models of acute promyelocytic leukemia was counteracted only by HDAC inhibitors (HDACi) in combination with retinoic acid (21). A number of HDACi have been characterized that inhibit the deacetylation of histones, which is associated with the reactivation of gene expression, leading to differentiation and thus abolishing tumor growth (1, 22, 23).

We attempted to identify genes regulated by ETV6/RUNX1 in lymphoblasts using expression profiling and studied whether their expression can vary depending on the acetylation/deacetylation of histones. We compared different genetically characterized subgroups (ETV6/RUNX1, BCR/ABL, and MLL/AF4) most frequently present in children with ALL to identify specific genes of ETV6/RUNX1 genotype. It was proposed that chromatin remodeling is responsible for the leukemogenic role of ETV6/RUNX1 fusion protein and that this effect could be reversible by the application of HDACi (24). The present work aimed to verify this assumption by using HDACi trichostatin A (TSA) and valproic acid (VPA). We tested the hypothesis that aberrant chromatin remodeling affects expression of genes originally transactivated by RUNX1 (25–27). We tested whether HDACi can release repression activity of ETV6/RUNX1 and induce differentiation in lymphoid leukemic cells by comparison with other B-lymphoid leukemic cells with different mechanisms of leukemogenesis (BCR/ABL, ETV6/PDGFRB).


   Materials and Methods
Top
 Abstract
 Materials and Methods
Results
 Discussion
 References
 
Cell cultivation. REH, a B-cell precursor leukemia cell line with translocation (12;21), kindly provided by R. Pieters (Erasmus Medical Center Rotterdam), Nalm-6 B-precursor leukemic cell line with translocation (5;12), Nalm-24 B-precursor leukemic cell line with translocation (9;22), and NC-NC, normal lymphoblastoid cells immortalized by EBV transformation, were used for experiments. Cell cultures except for REH were purchased from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (German Collection of Microorganisms and Cell Cultures; Braunschweig, Germany). Cells were cultured at the Roswell Park Memorial Institute (RPMI) 1640 with 2 mmol/L L-glutamine with 10% fetal calf serum and 10 mL/L antibiotic solution (100 units/mL penicillin, 100 µg/mL streptomycin). All suspension cultures were maintained at 37°C in 5% CO2. Cells were collected by centrifugation for 10 min at 240 x g and resuspended at a density of 5 x 105/mL in fresh medium 24 h before all experiments. Cell lines were treated in six-well plates for 24 and 48 h with either VPA at concentrations of 0.5 and 1.0 mmol/L or TSA at concentrations of 120 and 240 nmol/L. The doses of VPA were chosen according to the optimum serum concentrations in neurologic patients (0.35-0.7 mmol/L).

Isolation of mRNA and cDNA conversion. Total RNA was extracted from a standardized amount of mononuclear cells isolated from cell lines using a modified method described by Chomczynski and Sacchi (28) and from patient samples as previously described (29). The total extraction volume of RNA was adjusted to the number of processed cells and converted into cDNA using Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc., Carlsbad, TX) according to the manufacturer's instructions.

Flow cytometry. Immunophenotype measurement of treated cells was done with a combination of fluorescent monoclonal antibodies. Particular antibodies were chosen to cover specific stages of studied B-cell development. The selected immunostainings were measured by multicolored combinations: surface antigens—CD10 FITC/CD20 phycoerythrin/CD71 A1633/4',6-diamidino-2-phenylindole; intracellular antigens—iTdT FITC/iIgM phycoerythrin/20 PC5 and iRAG-1 A1633. For measurement, we used a BD FACSAria Cell Sorting System (Becton Dickinson, San Jose, CA) and data acquisition analyses were done by FlowJo (TreeStar, Ashland, OR) and by CELLQuest (Becton Dickinson) software applications. CycleTEST PLUS DNA Reagent Kit (Becton Dickinson) was used for DNA analysis. Proliferation activity was assessed as a percentage of cells in S and G2-M phases of cell cycle. All individual cultivations and cell measurements were done in independent triplicates. Statistics was calculated by ANOVA test.

Plasmids. Myc-tagged ETV6/RUNX1 in pcDNA3.1 vector (pcDNA3.1Myc-ETV6/RUNX1) for transient ETV6/RUNX1 model and pGZMB-luc for luciferase assay were kindly provided by A. Ford (Institute of Cancer Research, LRF, London, United Kingdom). Control vector pcDNA3.1-empty was prepared by EcoRI restriction and pGL3-basic and pRL-CMV were purchased from Promega (Madison, WI).

Luciferase assay. The human carcinoma cell line HeLa was seeded at 1.5 x 105 per well and transfected with pcDNA3.1Myc-ETV6/RUNX1 or pcDNA3.1-empty (1.6 µg) by Lipofectamine (2 µL) in 600 µL of serum-free medium 24 h before transfection with pGZMB-luc. DNA fragment of granzyme B (GZMB) promoter was subcloned into the SmaI site of pGL3-basic luciferase reporter gene construct to form pGZMB-luc. Cells were transfected by pGL3-basic to normalize the luciferase activity (pGZMB-luc/pGL3-basic). pRL-CMV was transfected to each sample for normalizing transfection efficiency.

To test the effect of HDACi on ETV6/RUNX1, cotransfected HeLa cells were treated with VPA or TSA and luciferase activity was monitored. Subsequently, differences between untreated and treated cells were calculated. Luciferase activity was measured by a luminometer Microplate TLX2 using Dual-Luciferase Reporter Assay System (Promega) according to manufacturer's instructions. Luciferase activity of untreated cells was arbitrarily set to 100%. All experiments were done in triplicate. Data were expressed as relative luciferase activity (RLU).

Quantitative reverse transcription-PCR. Real-time quantitative reverse transcription-PCR (RT-PCR) was done in the LightCycler rapid thermal cycler system (Roche Diagnostic GmbH, Basel, Switzerland) and Q-Cycler (Bio-Rad, Hercules, CA) according to manufacturers' instructions. Oligonucleotide hydrolyzation probes were used in systems for quantification of genes selected from expression profiling, SYBR Green DNA-binding dye for RAG-1 and TdT genes, and oligonucleotide hybridization probes for ß2-microglobulin (housekeeping gene; ref. 4).

Gene expression measurements and analyses. Spotted cDNA microarrays were used that contain more than 43,000 features representing ~30,000 genes (Stanford Functional Genomics Facility, Stanford, CA). The reference RNA used for all the arrays was Universal Human Reference RNA (Stratagene Europe, Amsterdam, the Netherlands). We labeled each of the sample RNA (RNA of untreated REH cells, TSA-treated REH cells, VPA-treated REH cells) with a red fluorescent dye (Cy5-dUTP, Amersham Pharmacia Biotech Europe, Freiburg, Germany) and the reference RNA with a green fluorescent dye (Cy3-dUTP, Amersham Pharmacia Biotech Europe) and comparatively hybridized sample RNA and reference RNA to an array. Each experiment was done in triplicates.

The fluorescence intensities of Cy5 and Cy3 were measured using a GenePix 4000 scanner (Axon Instruments, Foster City, CA) and analyzed using GenePix Pro 4.1 software (Axon Instruments). Subsequent analysis and normalization of data was done as previously described (29, 30).

Statistical analysis. To analyze the data of REH samples treated with HDACi in the supervised analysis, we used significance analysis of microarrays (31). A list of significantly differentially expressed genes was obtained by carrying out a 1,000-fold permutation test and considering a false discovery rate of 5% and a fold change of >2. Analysis of ALL patient expression data was done following our previously described method (32).


   Results
Top
 Abstract
 Materials and Methods
 Results
Discussion
 References
 
Functional proof of ETV6/RUNX1 regulation of granzyme B
The transcription of granzyme B is known to be directly regulated by the RUNX1 protein (33, 34). Therefore, the granzyme B promoter region was cloned into the luciferase construct to determine whether ETV6/RUNX1 represses granzyme B expression on the basis of an existing RUNX1-binding site. Luciferase activity was measured in HeLa cells transfected with pcDNA3.1 Myc-ETV6/RUNX1 and compared with HeLa cells transfected with pcDNA3.1 empty vector. The mouse DNA fragment of granzyme B (bp –324 to –43) has two RUNX1-binding sites (–278 to –273 and –219 to –214) and is identical to the human gene (whole inserted sequence of granzyme B has 84% homology with human gene fragment). Cells were transfected with pGZMB-luc or pGL3-basic to normalize the luciferase activity. A decrease to 33% in ETV6/RUNX1 cells versus control cells with empty vector (100%) indicated that granzyme B was down-regulated by ETV6/RUNX1. Twenty-four hours of cultivation with HDACi released this expression block and increased luciferase activity in ETV6/RUNX1 cells by >180%. Values were calculated as relative luciferase units (RLU) of ETV6/RUNX1 cells divided by RLU of control cells (Fig. 1 ).


Figure 1
View larger version (10K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. Granzyme B expression regulation by ETV6/RUNX1 chimeric transcription factor. Normalized luciferase activities of GZMB-luc detected in HeLa cells with empty vector (assigned as 100%) and cells transiently transfected with ETV6/RUNX1. The ratio of RLU in HeLa GZMB-luc/ETV6/RUNX1-transfected cells to RLU of HeLa control cells (GZMB-luc/empty vector-transfected) was assigned to a value of 100%. Both control and ETV6/RUNX1 cells were then treated with HDACi (VPA 1 mmol/L, TSA 240 nmol/L). Increase of measured luciferase activity due to released repression activity of ETV6/RUNX1 protein was recalculated as a percentage.

 
Screening strategy for identification of downstream genes
Step I. Recalculation of expression profiling data from ALL patients was done by analyzing the data published by Fine et al. (29, 30), which are available through the Stanford Microarray Database4 as previously described (32). Differences between genotypic subgroups were calculated using data from patients with B-cell precursor ALL. For graphical comparison of the predictive value of genes correlating with a genotype assessed by different methods, the probe set–specific optimal cutoff value was determined using statistical software R.5 For each probe set, the frequencies of cases above all possible cutoff values were computed. The level that led to the largest distance from the diagonal of equality was used as the optimum cutoff value for each probe set (32).

This approach was applied to ETV6/RUNX1-positive versus ETV6/RUNX1-negative, genotypically defined patients (BCR/ABL and MLL/F4 positive) to select genes specific for ETV6/RUNX1-positive phenotype. Expecting the repression effect of ETV6/RUNX1 fusion protein, we searched for genes down-regulated in the ETV6/RUNX1-positive group. The scale of cutoff values for each probe set was from 0 to 100. This meant that genes selected for further analysis had to have a value of difference between txx (cutoff value of a single gene in ETV6/RUNX1-positive patients) and tyy (cutoff value of a single gene in ETV6/RUNX1-negative patients) ≤ –33.3. This approach allowed us to identify 2,539 down-regulated genes specific for the ETV6/RUNX1 phenotype in ascending order (order of importance).

Step II. Then we asked whether an ETV6/RUNX1-positive cell line REH could serve as a model in the next intended in vitro experiments. Therefore, we compared data from two independent expression profiling experiments, one on ALL patients and the other using REH samples. It was essential for our future strategy to choose only those genes from step one of our analyses that had similar relative expressions in both tested sets. Mathematically, genes selected in step two had to have differences ("closeness") between their expression in ETV6/RUNX1-positive patients and that in REH cells closer than the minimum difference in expression between ETV6/RUNX1-positive and ETV6/RUNX1-negative patients. The output of the second step of analysis resulted in 927 genes. This set of down-regulated genes characterized equally ETV6/RUNX1-positive patient samples and the REH cell line. For exact algorithm, see Fig. 2 .


Figure 2
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. Diagrammatic representation of four steps of expression profiling data analysis for identification of genes regulated by ETV6/RUNX1. Step I: Identification of genes discriminating subgroup of ETV6/RUNX1-positive patients. Selected range characterized only by down-regulated genes. txx, cutoff value of each probe set in ETV6/RUNX1-positive patients; tyy, cutoff value of each probe set in ETV6/RUNX1-negative patients. Step II: Selection of the genes on the basis of similarity (closeness) of REH (ETV6/RUNX-positive cell line) cells and ETV6/RUNX1-positive patients. Median pER, median of relative expression of screened gene in all ETV6/RUNX1-positive patients; median pBA, median of relative expression of screened gene in all BCR/ABL-positive patients; median pMA, median of relative expression of screened gene in all MLL/AF4-positive patients. Step III: Selection made on the basis of direction of change in the expression level of selected genes after VPA treatment. Step IV: Selection of genes with significantly changed expression (≤0.05) after VPA treatment.

 
Step III. In the next step of analysis, we selected genes that changed their expression after VPA treatment in samples of REH cells. Again, only genes with increased expression after VPA were chosen, thus corresponding with the expected underlying mechanism of ETV6/RUNX1 action. The difference between median gene expression in treated samples (with VPA) and median gene expression in untreated samples (median VPA – median C) had to be higher than zero. This third step limited our list to 759 genes.

Step IV. In the last step, we selected genes with a significantly changed expression pattern after VPA treatment (P ≤ 0.05) and these genes were listed in descending order. This fourth step separated only 72 genes. We present these top 25 genes in Table 1 .


View this table:
[in this window]
[in a new window]

  		Table 1. Selection of genes with significantly changed expression (≤0.05) after the treatment with VPA

 
Expression profiling versus quantitative RT-PCR
For further analysis, we chose genes with a known proliferative or cell cycle regulative function. Using quantitative RT-PCR, we confirmed these expression changes after the treatment with HDACi (VPA, TSA) in four genes, three of them down-regulated and one up-regulated in ETV6/RUNX1-positive leukemia cells. First, ACK1 (alias TNK2, activated p21cdc42Hs kinase) and PDGFRB (platelet derived growth factor receptor ß) were chosen from our list of 25 genes (see above). Moreover, differential expression in ETV6/RUNX1-positive and ETV6/RUNX1-negative patients was confirmed by comparison with other publicly available expression profiling data6 (35, 36). Gene JunD (junD proto-oncogene) was chosen from the more extensive list of the first 72 genes, and gene transcription factor 4 (TCF4) was selected as an example of genes up-regulated in ETV6/RUNX1-positive cells. Expression of JunD, ACK1, and PDGFRB in ETV6/RUNX1-positive patients as well as in the cell lines was low with a clear increase after HDACi treatment. TCF4 gene expression was high and the treatment with HDACi led to a decrease. We confirmed changes of expression levels by quantitative RT-PCR between treated and untreated cells: JunD, TSA P = 0.013, VPA P = 0.0008; ACK1, VPA P = 0.07, TSA P = 0.5; PDGFRB, TSA P < 0.0001, VPA P = 0.016; TCF4, TSA P < 0.0001, VPA P = 0.0002 (Fig. 3 ).


Figure 3
View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 3. Analysis of selected genes expression using quantitative RT-PCR. Quantification of four genes selected from microarray data was done by quantitative RT-PCR comparing untreated and treated (VPA, TSA) REH cells. Gene expression of target was normalized to ß2-microglobulin as a housekeeping gene. All samples were measured in triplicates and gene expression of each replicate was detected in duplicate. P value was calculated from the mean of ratio of selected gene to ß2-microglobulin for all replicates with absolute error.

 
Effect of HDACi on proliferation activity and apoptosis
We tested cell cycle progress and apoptosis rate after HDACi treatment in ETV6/RUNX1-positive cells and compared them with pre–B-leukemic cells with different leukemogenic mechanism (ETV6/PDGFRB) to look if there are any differences in drug effect. EBV-transformed mature lymphocytes were used as a control to test the toxicity of the compounds in nonleukemic B-lymphoid cells. An increase in the number of apoptotic cells was expressed as the ratio of apoptotic fraction in treated cells to apoptotic fraction in untreated cells. In the REH cell line, the apoptosis rate was 8% using 1 mmol/L VPA and 26% using 240 nmol/L TSA, in contrast to 3.7% and 1.7% in control lymphocytes (NC-NC cell line). Similarly, in Nalm-6, the apoptosis rate reached 17% with VPA and 41% with TSA. The Nalm-6 and REH cell lines reduced their proliferation activity (S + G2-M) by 19% (VPA), 28% (TSA) and 7% (VPA), 15% (TSA), respectively. These data suggest that treatment with TSA preferentially leads to apoptosis in both ETV6/RUNX1-positive and ETV6/RUNX1-negative leukemic cell lines. However, this effect was clearly seen after VPA treatment only in Nalm-6 cells, whereas in REH cells, we observed a relative increase in G1-G0. In contrast, in NC-NC cells the proportion of retained cell cycle phases was almost unchanged (Fig. 4 ).


Figure 4
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 4. Detection of changes in cell cycle progress. Cell cycle was monitored 48 h after HDACi treatment of REH, Nalm-6, and NC-NC cell lines. Cell cycle fractions (apoptosis, G1-G0, S + G2-M phases) are presented as a percentage of all cells.

 
Detection of differentiation antigen shift
Three leukemic cell lines with known fusion genes, REH (ETV6/RUNX1), Nalm-24 (BCR/ABL), and Nalm-6 (ETV6/PDGFRB), were selected for monitoring of HDACi-driven changes in immunophenotype. Initially, CD10 and CD20 levels were detected 24 and 48 h after HDACi (VPA, TSA) treatment. Due to different developmental stages of selected cell lines, mean fluorescence area intensity instead of genuine cell count was measured. This allowed a simple comparison of differentiation pattern shift. The pattern of expression for native REH cells was CD10++/CD20–. After cultivation with HDACi, the CD10 hyperexpression decreased (CD10+; VPA P = 0.0004, TSA P = 0.0002) and CD20 expression increased to low positivity (CD20dim/+; VPA P = 0.005, TSA P = 0.01). The expression pattern for native Nalm-24 cells was CD10+/CD20+. After treatment, expression of both studied antigens decreased (VPA P = 0.0008, TSA P < 0.0001; VPA P = 0.08, TSA P = 0.0004, respectively). The expression pattern for native Nalm-6 cells was CD10++/CD20–. Expression of CD10 was dysregulated in an up-and-down manner depending on the time of acquisition and dose and type of HDACi. Although its level oscillated, its expression tended to be slightly higher. Surface antigen CD20 maintained its initial negativity under treatment with HDACi (Fig. 5 ).


Figure 5
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 5. Expression of differentiation markers monitored in three types of leukemias. Typical B-lineage differentiation markers (CD10, CD20, and TdT) were detected in three lymphoid leukemia cell lines, REH (ETV6/RUNX1 positive), Nalm-24 (BCR/ABL positive), and Nalm-6 (ETV6/PDGFRB positive), by flow cytometry. Changes of expression of individual markers after the treatment with VPA (1.0 mmol/L) are given as the fluorescence mean normalized to control (untreated cells). *, P ≤ 0.05; **, P ≤ 0.01 (ANOVA test).

 
Whereas Nalm-24 and Nalm-6 cell lines did not display evident differentiation drift, the REH cell line expressed CD10 and CD20 in a pattern similar to more mature normal blasts (37). We used another differentiation marker of B-cell development, TdT. After VPA administration, the expression of TdT in the REH cell line significantly decreased (VPA P = 0.0001), whereas in Nalm-24 cells, the TdT level decrease was not significant (VPA P = 0.24). RAG-1 expression in REH cell line paralleled that of TdT (VPA P = 0.006). Lower doses (0.5 mmol/L VPA, 120 nmol/L TSA) of VPA and TSA were less effective; however, the same trends were observed. After 48 h, the effect of HDACi was less pronounced.

The expression decrease of RAG-1 and TdT antigen was confirmed at the mRNA level by quantitative RT-PCR. The transcription levels of RAG-1 and TdT were normalized to the expression of the housekeeping gene ß2-microglobulin. Normalized expression data of treated and untreated samples were compared.

The mRNA level of RAG-1 gene rapidly decreased after 24 h from the initial level 3.30 ± 0.321 to 1.004 ± 0.053 after VPA (P = 0.0002) and 0.197 ± 0.046 after TSA (P < 0.0001) administration. Similarly, the mRNA level of TdT significantly decreased within 24 h after TSA or VPA treatment from initial RNA level 1.767 ± 0.178 to 0.888 ± 0.110 (P = 0.01) and 0.072 ± 0.020 (P = 0.0001), respectively (Fig. 6 ).


Figure 6
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 6. Comparison of changes in gene expression and protein expression levels of TdT and RAG-1 after HDACi treatment. The expression of differentiation markers was measured using flow cytometry 24 and 48 h after the treatment with HDACi in REH cells. A, dot blot images show effect of TSA (120 and 240 nmol/L) and VPA (0.5 and 1.0 mmol/L) on the expression of TdT and RAG-1 48 h after the treatment. Data are presented as a percentage of cells of specific quadrants. B, normalized gene expression level by quantitative RT-PCR 24 h after incubation of REH cells with HDACi. *, P ≤ 0.05; **, P ≤ 0.01 (Mann-Whitney test).

 

   Discussion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
References
 
The functional consequences of fusion gene formation are either constitutive kinase activation (e.g., BCR/ABL) or altered transcriptional regulation. Published studies suggest that RUNX1 fusion proteins may act as negative regulators of genes that, in normal hematopoiesis, are transactivated by the RUNX1 transcription factor (18, 38–40). RUNX1 is known to regulate promoters of various genes involved in myeloid differentiation, such as macrophage colony-stimulating factor receptor, granulocyte-macrophage colony-stimulating factor, interleukin-3, neutrophil elastase, myeloperoxidase, and others (38, 41). RUNX1 regulates hematopoietic myeloid cell differentiation and transcriptional activation but its role in lymphoid development is not yet fully understood (42). Only a few RUNX1 target genes or regions have been described in lymphoid cells (i.e., B-lymphoid kinase, complement receptor 1, enhancers of T-cell receptor ß, and immunoglobin {alpha}; refs. 10, 17, 43).

We hypothesized that the normal role of RUNX1 in B-cell precursors is compromised in a dominant-negative fashion by the ETV6/RUNX1 fusion. ETV6/RUNX1 recruits a corepressor complex including HDAC, which likely induces chromatin remodeling and consequently blocks the transcription of genes normally transactivated by RUNX1. To prove the direct involvement of HDACs in ETV6/RUNX1 function, we used HDACi to overcome this presumptive block. Despite expected broad effects of HDACi treatment on cellular transcriptional regulation, these compounds act very selectively to alter the transcription of fewer than 2% of expressed genes (44).

To prove the direct effect of HDACi on ETV6/RUNX1 in cells in vitro, we chose a previously identified target gene of RUNX1, granzyme B. To our knowledge, this is the first time that a direct effect of the interaction between ETV6/RUNX1 and any potential target genes of RUNX1 has been shown. For this purpose, we have established a transient ETV6/RUNX1 expressing model to show the down-regulation of granzyme B in the presence of the ETV6/RUNX1 protein compared with ETV6/RUNX1-negative controls. Next, we showed that HDACi treatment relieves the repression observed in the presence of the ETV6/RUNX1 fusion protein. This experiment confirms the hypothesis that the ETV6/RUNX1 fusion protein can act as a transcriptional repressor and its activity is reversible by HDACi. Therefore, HDACi can be used as a tool for the identification of genes potentially regulated by ETV6/RUNX1.

Childhood ALL subgroups known thus far are characterized by the presence of specific genetic aberrations, mainly chromosomal translocations leading to the fusion genes formation. They are part of different pathways, thus contributing to leukemogenesis by different mechanisms. To study neoplastic transformation in the presence of ETV6/RUNX1 and to identify potential ETV6/RUNX1 target genes, we compared expression data of ALL patients of three genotypically defined groups: ETV6/RUNX1, BCR/ABL, and MLL/AF4. Identification of specific genes of the ETV6/RUNX1-positive phenotype was based on the analysis of particular cutoffs; a selection of genes specifically overexpressed or underexpressed in the cells of ETV6/RUNX1-positive patients was made. This group of genes was compared with the genes whose expression patterns changed significantly after treatment with HDACi. From the resulting pool of genes that fulfilled all of these criteria of analysis, we selected four genes (JunD, ACK1, PDGFRB, and TCF4) to compare expression data from expression profiling and data from quantitative RT-PCR. These four genes were selected from the extensive list on the basis of their previously described fundamental role in cell proliferation and cell cycle regulation (45–48). Our analysis pinpointed top 25 genes that either are directly regulated by the ETV6/RUNX1 fusion protein or are in ETV6/RUNX1-regulated pathways. Most likely, they represent target genes of the RUNX1 transcription factor involved in B-lineage lymphopoiesis. We assume that these genes (albeit not exclusively) are likely to be involved in ETV6/RUNX1-driven leukemogenesis.

We also investigated the influence of HDACi on cell cycle progress and differentiation in ETV6/RUNX1-positive cells. As the morphologic shift was hardly to be expected in lymphoblasts, we concentrated on flow cytometry analysis. These immunophenotype changes differed between ETV6/RUNX1-positive and ETV6/RUNX1-negative cells. ETV6/RUNX1-positive cells showed indications of differentiation (decrease of CD10, TdT, and RAG-1 expression and up-regulation of CD20) compared with cell lines derived from leukemias with aberrant tyrosine kinase fusion genes (BCR/ABL, ETV6/PDGFRß), which underwent unspecific immunophenotype changes without a differentiation effect.

We are aware of the effect of HDACi on apoptotic pathways in malignant cells that can complicate analysis of apoptosis in our model cell lines. Indeed, a proapoptotic effect was observed in both ETV6/RUNX1-positive and ETV6/RUNX1-negative leukemic cells. In contrast to leukemic cells, the administration of HDACi did not change the progress of the cell cycle in normal mature lymphocytes. HDACi administration decreased the number of proliferating cells in the leukemic cell lines. TSA had a more pronounced effect with a significant shift towards apoptosis in both ETV6/RUNX1-positive and ETV6/RUNX1-negative cells, whereas VPA acted more selectively, arresting ETV6/RUNX1-positive cells in the G1-G0 phase of the cell cycle instead of driving them to apoptosis. This result is in agreement with the previously published data suggesting that VPA preferentially acts on corepressor-associated HDACs and inhibits class I HDACs more efficiently than class II enzymes. Previous studies also showed lower toxicity of VPA to normal hematopoietic precursors (25).

HDACi are currently among the most promising compounds in drug development for cancer therapy, and first-generation HDACi are currently being tested in phase I/II clinical trials (49). ALL patients carrying the ETV6/RUNX1 fusion have a good prognosis compared with other subgroups of childhood ALL, but relapses still occur in this group. And although generally the prognosis after relapse tends to be good, second relapses also occur. As in all subgroups with overall superior prognosis, the main challenge is to identify the sporadic high-risk patients. Minimal residual disease quantitation represents an available tool for the identification of slow responders to the initial phase of treatment as well as patients with adverse dynamics of the leukemic clone. As recently showed, minimal residual disease assessment can even help to test the efficacy of individual drugs in the multidrug induction therapy in ETV6/RUNX1-positive patients (50).

We presume that HDACi can potentially be used in the treatment of patients with detectable residual disease or with molecular relapses in the relapse treatment and after the transplantation in this genotypically defined subgroup of patients. The doses of VPA we used in our study correspond to the serum concentrations of this agent in neurologic patients. The use of this agent for human medicine predisposes VPA for testing in high-risk ETV6/RUNX1-positive leukemia patients.


   Footnotes
 
Grant support: Czech Ministry of Health grant no. 8316; Grant Agency of the Czech Republic grant no. 301/D189; Grant Agency of Charles University, no. 56/2005 and 75/2004; and Czech Ministry of Education, grant MSM0021620813.

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

Note: J. Starkova and J. Madzo contributed equally to this work and should be considered co-first authors.

4 http://genome-www.stanford.edu/microarray Back

5 http://www.r-project.org Back

6 Available at: http://www.stjuderesearch.org/ALL1 and http://www.stjuderesearch.org/ALL3. Back

Received 10/23/06; revised 12/11/06; accepted 12/13/06.


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 

  1. Marks P, Rifkind RA, Richon VM, Breslow R, Miller T, Kelly WK. Histone deacetylases and cancer: causes and therapies. Nat Rev Cancer 2001;1:194–202.[CrossRef][Medline]
  2. Hrusak O, Trka J, Zuna J, Houskova J, Bartunkova J, Stary J; The Pediatric Hematology Working Group in the Czech Republic. Aberrant expression of KOR-SA3544 antigen in childhood acute lymphoblastic leukemia predicts TEL-AML1 negativity. Leukemia 1998;12:1064–70.[CrossRef][Medline]
  3. Zuna J, Hrusak O, Kalinova M, Muzikova K, Stary J, Trka J; Czech Paediatric Haematology Working Group. TEL/AML1 positivity in childhood ALL: average or better prognosis? Leukemia 1999;13:22–4.[CrossRef][Medline]
  4. Krejci O, Starkova J, Otova B, et al. Up-regulation of asparagine synthetase fails to avert cell cycle arrest induced by L-asparaginase in TEL/AML1-positive leukaemic cells. Leukemia 2004;18:434–41.[CrossRef][Medline]
  5. Golub TR, McLean T, Stegmaier K, Carroll M, Tomasson M, Gilliland DG. The TEL gene and human leukemia. Biochim Biophys Acta 1996;8:M7–10.
  6. Kim CA, Phillips ML, Kim W, et al. Polymerization of the SAM domain of TEL in leukemogenesis and transcriptional repression. EMBO J 2001;20:4173–82.[CrossRef][Medline]
  7. Lopez RG, Carron C, Oury C, Gardellin P, Bernard O, Ghysdael J. TEL is a sequence-specific transcriptional repressor. J Biol Chem 1999;274:30132–8.[Abstract/Free Full Text]
  8. Blyth K, Cameron ER, Neil JC. The RUNX genes: gain or loss of function in cancer. Nat Rev Cancer 2005;5:376–87.[CrossRef][Medline]
  9. Friedman AD. Leukemogenesis by CBF oncoproteins. Leukemia 1999;13:1932–42.[CrossRef][Medline]
  10. Hiebert SW, Sun W, Davis JN, et al. The t(12;21) translocation converts AML-1B from an activator to a repressor of transcription. Mol Cell Biol 1996;16:1349–55.[Abstract]
  11. Okuda T, van Deursen J, Hiebert SW, Grosveld G, Downing JR. AML1, the target of multiple chromosomal translocations in human leukemia, is essential for normal fetal liver hematopoiesis. Cell 1996;84:321–30.[CrossRef][Medline]
  12. Wang Q, Stacy T, Binder M, Marin-Padilla M, Sharpe AH, Speck NA. Disruption of the Cbfa2 gene causes necrosis and hemorrhaging in the central nervous system and blocks definitive hematopoiesis. Proc Natl Acad Sci U S A 1996;93:3444–9.[Abstract/Free Full Text]
  13. Peterson LF, Boyapati A, Ranganathan V, et al. The hematopoietic transcription factor AML1 (RUNX1) is negatively regulated by the cell cycle protein cyclin D3. Mol Cell Biol 2005;25:10205–19.[Abstract/Free Full Text]
  14. Otto F, Lubbert M, Stock M. Upstream and downstream targets of RUNX proteins. J Cell Biochem 2003;89:9–18.[CrossRef][Medline]
  15. Wang LC, Swat W, Fujiwara Y, et al. The TEL/ETV6 gene is required specifically for hematopoiesis in the bone marrow. Genes Dev 1998;12:2392–402.[Abstract/Free Full Text]
  16. Song H, Kim JH, Rho JK, Park SY, Kim CG, Choe SY. Functional characterization of TEL/AML1 fusion protein in the regulation of human CR1 gene promoter. Mol Cells 1999;9:560–3.[Medline]
  17. Rho JK, Kim JH, Yu J, Choe SY. Correlation between cellular localization of TEL/AML1 fusion protein and repression of AML1-mediated transactivation of CR1 gene. Biochem Biophys Res Commun 2002;297:91–5.[CrossRef][Medline]
  18. Guidez F, Petrie K, Ford AM, et al. Recruitment of the nuclear receptor corepressor N-CoR by the TEL moiety of the childhood leukemia-associated TEL-AML1 oncoprotein. Blood 2000;96:2557–61.[Abstract/Free Full Text]
  19. Heibert SW, Lutterbach B, Durst K, et al. Mechanisms of transcriptional repression by the t(8;21)-, t(12;21)-, and inv(16)-encoded fusion proteins. Cancer Chemother Pharmacol 2001;48:S31–4.
  20. Uchida H, Downing JR, Miyazaki Y, Frank R, Zhang J, Nimer SD. Three distinct domains in TEL-AML1 are required for transcriptional repression of the IL-3 promoter. Oncogene 1999;18:1015–22.[CrossRef][Medline]
  21. Melnick A, Licht JD. Deconstructing a disease: RAR{alpha}, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia. Blood 1999;93:3167–215.[Free Full Text]
  22. Meinke PT, Liberator P. Histone deacetylase: a target for antiproliferative and antiprotozoal agents. Curr Med Chem 2001;8:211–35.[Medline]
  23. Kutko MC, Glick RD, Butler LM, et al. Histone deacetylase inhibitors induce growth suppression and cell death in human rhabdomyosarcoma in vitro. Clin Cancer Res 2003;9:5749–55.[Abstract/Free Full Text]
  24. Hiebert SW, Lutterbach B, Amann J. Role of co-repressors in transcriptional repression mediated by the t(8;21), t(16;21), t(12;21), and inv(16) fusion proteins. Curr Opin Hematol 2001;8:197–200.[CrossRef][Medline]
  25. Gottlicher M, Minucci S, Zhu P, et al. Valproic acid defines a novel class of HDAC inhibitors inducing differentiation of transformed cells. EMBO J 2001;20:6969–78.[CrossRef][Medline]
  26. Phiel CJ, Zhang F, Huang EY, Guenther MG, Lazar MA, Klein PS. Histone deacetylase is a direct target of valproic acid, a potent anticonvulsant, mood stabilizer, and teratogen. J Biol Chem 2001;276:36734–41.[Abstract/Free Full Text]
  27. Blaheta RA, Cinatl J, Jr. Anti-tumor mechanisms of valproate: a novel role for an old drug. Med Res Rev 2002;22:492–511.[CrossRef][Medline]
  28. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156–9.[Medline]
  29. Cario G, Stanulla M, Fine BM, et al. Distinct gene expression profiles determine molecular treatment response in childhood acute lymphoblastic leukemia. Blood 2005;105:821–6.[Abstract/Free Full Text]
  30. Fine BM, Stanulla M, Schrappe M, et al. Gene expression patterns associated with recurrent chromosomal translocations in acute lymphoblastic leukemia. Blood 2004;103:1043–9.[Abstract/Free Full Text]
  31. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 2001;98:5116–21.[Abstract/Free Full Text]
  32. Vaskova M, Mejstrikova E, Kalina T, et al. Transfer of genomics information to flow cytometry: expression of CD27 and CD44 discriminates subtypes of acute lymphoblastic leukemia. Leukemia 2005;19:876–8.[CrossRef][Medline]
  33. Babichuk CK, Bleackley RC, Wargnier A, et al. Mutational analysis of the murine granzyme B gene promoter in primary T cells and a T cell clone Identification of human granzyme B promoter regulatory elements interacting with activated T-cell-specific proteins: implication of Ikaros and CBF binding sites in promoter activation. J Biol Chem 1997;272:18564–71.[Abstract/Free Full Text]
  34. Wargnier A, Legros-Maida S, Bosselut R, et al. Identification of human granzyme B promoter regulatory elements interacting with activated T-cell-specific proteins: implication of Ikaros and CBF binding sites in promoter activation. Proc Natl Acad Sci U S A 1995;92:6930–4.[Abstract/Free Full Text]
  35. Ross ME, Zhou X, Song G, et al. Classification of pediatric acute lymphoblastic leukemia by gene expression profiling. Blood 2003;102:2951–9.[Abstract/Free Full Text]
  36. Yeoh EJ, Ross ME, Shurtleff SA, et al. Classification, subtype discovery, and prediction of outcome in pediatric acute lymphoblastic leukemia by gene expression profiling. Cancer Cell 2002;1:133–43.[CrossRef][Medline]
  37. LeBien TW. Fates of human B-cell precursors. Blood 2000;96:9–23.[Abstract/Free Full Text]
  38. Lutterbach B, Westendorf JJ, Linggi B, Isaac S, Seto E, Hiebert SW. A mechanism of repression by acute myeloid leukemia-1, the target of multiple chromosomal translocations in acute leukemia. J Biol Chem 2000;275:651–6.[Abstract/Free Full Text]
  39. Fenrick R, Amann JM, Lutterbach B, et al. Both TEL and AML-1 contribute repression domains to the t(12;21) fusion protein. Mol Cell Biol 1999;19:6566–74.[Abstract/Free Full Text]
  40. Chakrabarti SR, Nucifora G. The leukemia-associated gene TEL encodes a transcription repressor which associates with SMRT and mSin3A. Biochem Biophys Res Commun 1999;264:871–7.[CrossRef][Medline]
  41. Tenen DG, Hromas R, Licht JD, Zhang DE. Transcription factors, normal myeloid development, and leukemia. Blood 1997;90:489–519.[Free Full Text]
  42. Tanaka T, Tanaka K, Ogawa S, et al. An acute myeloid leukemia gene, AML1, regulates hemopoietic myeloid cell differentiation and transcriptional activation antagonistically by two alternative spliced forms. EMBO J 1995;14:341–50.[Medline]
  43. Libermann TA, Pan Z, Akbarali Y, et al. AML1 (CBF{alpha}2) cooperates with B cell-specific activating protein (BSAP/PAX5) in activation of the B cell-specific BLK gene promoter. J Biol Chem 1999;274:24671–6.[Abstract/Free Full Text]
  44. Van Lint C, Emiliani S, Verdin E. The expression of a small fraction of cellular genes is changed in response to histone hyperacetylation. Gene Expr 1996;5:245–53.[Medline]
  45. Zhao DH, Hong JJ, Guo SY, et al. Aberrant expression and function of TCF4 in the proliferation of hepatocellular carcinoma cell line BEL-7402. Cell Res 2004;14:74–80.[CrossRef][Medline]
  46. Ahmed I, Calle Y, Sayed MA, et al. Cdc42-dependent nuclear translocation of non-receptor tyrosine kinase, ACK. Biochem Biophys Res Commun 2004;314:571–9.[CrossRef][Medline]
  47. Kobayashi S, Yamashita K, Takeoka T, et al. Calpain-mediated X-linked inhibitor of apoptosis degradation in neutrophil apoptosis and its impairment in chronic neutrophilic leukemia. J Biol Chem 2002;277:33968–77.[Abstract/Free Full Text]
  48. Bjorling-Poulsen M, Seitz G, Guerra B, Issinger OG. The pro-apoptotic FAS-associated factor 1 is specifically reduced in human gastric carcinomas. Int J Oncol 2003;23:1015–23.[Medline]
  49. Minucci S, Pelicci PG. Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat Rev Cancer 2006;6:38–51.[CrossRef][Medline]
  50. Kwok CS, Kham SK, Ariffin H, Lin HP, Quah TC, Yeoh AE. Minimal residual disease (MRD) measurement as a tool to compare the efficacy of chemotherapeutic drug regimens using Escherichia coli-asparaginase or Erwinia-asparaginase in childhood acute lymphoblastic leukemia (ALL). Pediatr Blood Cancer 2006;47:299–304.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 J. Leukoc. Biol.Home page
K. Schmidt, K. Seeger, C. Scheibenbogen, R. Bender, M. Abdulla, S. Sussmilch, A. Salama, and A. Moldenhauer
Histone deacetylase inhibition improves differentiation of dendritic cells from leukemic blasts of patients with TEL/AML1-positive acute lymphoblastic leukemia
J. Leukoc. Biol., March 1, 2009; 85(3): 563 - 573.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
1078-0432.CCR-06-2569v1
13/6/1726    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Starkova, J.
Right arrow Articles by Trka, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Starkova, J.
Right arrow Articles by Trka, J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online