Gene expression profiling of Philadelphia chromosome (Ph)-negative CD34+ hematopoietic stem and progenitor cells of patients with Ph-positive CML in major molecular remission during therapy with imatinib

Article metrics


Imatinib is a potent inhibitor of the BCR-ABL tyrosine kinase. Treatment of chronic myelogenous leukemia (CML) in different stages of disease has shown its efficacy and superiority to conventional therapies. Beside inhibition of BCR-ABL, imatinib also has inhibitory effects on other tyrosine kinases such as wild-type ABL, c-KIT and platelet-derived growth factor receptors (PDGF-R) alpha and beta. ABL interacts with several proteins involved in DNA repair mechanisms such as p73, DNA-PK and RAD51.1,2 C-KIT plays an essential role in differentiation and proliferation of hematopoietic stem cells.3 PDGF-R is expressed by bone marrow stromal cells that are constituents of the hematopoietic microenvironment. Inhibition of ABL, c-KIT or PDGF-R might therefore result in functional disturbance of hematopoiesis or even in secondary karyotypic abnormalities observed in Philadelphia (Ph)-negative hematopoiesis during imatinib therapy.4,5

Therefore, we compared gene expression profiles of immunomagnetically isolated CD34+ bone marrow cells from six healthy volunteers with Ph-negative CD34+ cells of bone marrow from eight patients with Ph-positive CML in chronic phase, who reached major molecular remission during imatinib therapy. The aim was to identify imatinib-induced transcriptional changes in vivo, which may promote the occurrence of secondary chromosomal aberrations in Ph-negative hematopoiesis.

Patients (six males, two females; median age: 52 years, range: 25–68 years) included into this study had Ph-positive CML in chronic phase and were treated with imatinib (400 mg/day) as first-line therapy. The median time of imatinib therapy was 28 months (range: 11–39 months). Cytogenetics and quantitative real-time RT-PCR for BCR-ABL transcripts showed that there were no secondary chromosomal aberrations and that all patients were in complete cytogenetic and in major molecular remission, which was defined as 3 log reduction of the BCR-ABL/G6PDH ratio in comparison with pretreatment value (Table 1 ). CD34+ cells (median: 5 × 105; range 2 × 105–1.8 × 106) were isolated immunomagnetically from bone marrow mononuclear cells of eight patients and six healthy volunteers with a purity >98% using the MACS system (Miltenyi Biotec, Bergisch Gladbach, Germany) as previously described.6 As assessed by FISH analysis following immunomagnetic isolation using LSI BCR/ABL Dual Color, Dual Fusion Translocation Probes (Vysis, Bergisch-Gladbach, Germany) according to the manufacturer's instructions (threshold for false-positive colocalization: 5%), we found no evidence for the t(9;22) translocation in the CD34+ cells from imatinib-treated patients examined in this study. Total RNA (median: 550 ng; range: 130–990 ng) from isolated CD34+ cells was used to generate biotin-labelled cRNA (median: 7.5 μg; range: 2.6–12.3 μg) by means of Enzo BioArray HighYield RNA Transcript Labelling Kit (Affymetrix Ltd, UK). Labelled cRNA was hybridized to Affymetrix HG-Focus GeneChips covering 8793 genes, representing a broad spectrum of different functional groups according to the manufacturer's instructions. For quality control, normalization, and data analysis, we used the affy package of functions of statistical scripting language ‘R’ integrated into the Bioconductor project ( Using histograms of perfect match intensities, boxplots, 5′ to 3′ RNA degradation side-by-side plots, or scatter plots, we estimated the quality of probes and hybridizations. To normalize raw data, we used a method of variance stabilizing transformations (VSN).7 To compare the normalized data from CD34+ cells of imatinib-treated patients and healthy volunteers, we used the Significance Analysis of Microarrays (SAM) algorithm v1.21 (, which contains a sliding scale for false discovery rate (FDR) of significantly up- and downregulated genes.8 All data were permuted 800 cycles by using the two class, unpaired data mode of the algorithm.

Table 1 Characteristics and results of cytogenetic and molecular BCR-ABL analyses of each patient

Comparing gene expression profiles of Ph-negative CD34+ hematopoietic progenitor cells during imatinib therapy with CD34+ cells of healthy volunteers, no genes were found that differed in expression with sufficient significance (see Supplementary information on Leukemia's website). The complete data of our array experiments are available in the gene expression omnibus (GEO) database (; accession no.: GSE1418) according to the M.I.A.M.E. standards.

Our results show that in CD34+ hematopoietic progenitor cells imatinib at a dose of 400 mg/day has no measurable uniform influence on the expression of the 8793 genes assessed in this study. This was surprising since in previous studies imatinib had inhibitory effects on wild-type ABL, c-KIT and PDGF-R, which play a role in hematopoiesis. Wild-type ABL negatively regulates cell growth, seems to have proapoptotic activity and is involved in the cellular response to genotoxic stress.1,2 C-KIT plays a role in maintenance and self-renewal of hematopoietic stem cells, has mitogenic and antiapoptotic activity in progenitor cells and is involved in differentiation into mature blood cells.3 Gene expression analysis does not allow to measure inhibition of phosphorylation in signalling pathways. However, we would have expected some transcriptional changes of genes involved in functional end points of the ABL or c-KIT signal transduction cascades. Therefore, inhibition of those tyrosine kinases in Ph-negative CD34+ cells in vivo does not seem to be an important issue of imatinib treatment with a dose of 400 mg/day. This might be different when using higher doses of imatinib.

Moreover, we found no evidence that first-line imatinib treatment results in uniform up- or downregulation of genes contributing to oncogenesis or DNA repair, for example, genes encoding interaction partners of wild-type ABL such as p73, DNA-PK, ATM or RAD51.2 The lack of transcriptional changes supports the view that secondary chromosomal aberrations in Ph-negative hematopoiesis observed in some patients during imatinib therapy1,4,5 are not induced by the tyrosine kinase inhibitor. Since secondary aberrations such as trisomy 8, monosomy 7 or loss of the Y chromosome mainly occurred during imatinib second-line therapy, they might be the result of previous treatment with other drugs such as cytarabin, which might have induced genetic instability in Ph-negative progenitor cells.

Trying to explain our results, one might postulate that inhibition of one or two receptors in normal CD34+ cells can easily be compensated by alternative signalling pathways. Alternatively, the differences in gene expression may have been too subtle to be detected by our method, or they may have affected genes not represented on the chip. A further explanation is that exposure to a specific drug does not result in uniform changes of the gene expression profile and heterogenous changes cannot be detected with our relatively small sample number. This view is supported by the results from hierarchical cluster analyses including different numbers of genes with heterogenous expression, which were selected according to defined criteria (Figure 1). We found several genes that were heterogenously expressed in our samples indicating that a heterogenous alteration of gene expression by imatinib could be possible. However, we could not identify groups that were associated with imatinib treatment supporting the results from the SAM algorithm that there were no uniform imatinib-induced changes in gene expression.

Figure 1

Hierarchical cluster analyses. Hierarchical cluster analysis was performed by the ‘hclust’ function of the statistical scripting language ‘R’ integrated into the Bioconductor project ( using a complete linkage clustering algorithm with Manhattan distance function. Selection of genes included into each analysis was based on three criteria: (a) genes with expression values above the 5% quantile in two or more samples and a difference of at least 1.96 standard deviations (95% confidence intervals) from the mean value in at least one sample (4771 genes), (b) genes with expression values above the 5% quantile in two or more samples and a difference of at least 1.96 standard deviations (95% confidence intervals) from the mean value in at least two samples (271 genes), (c) genes with expression values above the 5% quantile in two or more samples and an at least 1.5-fold difference from the mean value in at least two samples (527 genes). Red fields indicate higher values than mean, blue fields indicate lower values than mean. The dendrogram visualizes the degree of similarity of the different samples by the branch length.

In conclusion, after reaching major molecular remission during first-line treatment with 400 mg imatinib per day, no uniform influence of the tyrosine kinase inhibitor on gene expression patterns in Ph-negative CD34+ stem and progenitor cells was observed in vivo. Therefore, based on our gene expression data we found no evidence for a major functional disturbance of Ph-negative hematopoiesis using imatinib as first-line therapy in CML.


  1. 1

    Bumm T, Muller C, Al-Ali HK, Krohn K, Shepherd P, Schmidt E et al. Emergence of clonal cytogenetic abnormalities in Ph− cells in some CML patients in cytogenetic remission to imatinib but restoration of polyclonal hematopoiesis in the majority. Blood 2003; 101: 1941–1949.

  2. 2

    Yuan ZM, Shioya H, Ishiko T, Sun X, Gu J, Huang YY et al. p73 is regulated by tyrosine kinase c-Abl in the apoptotic response to DNA damage. Nature 1999; 399: 814–817.

  3. 3

    Lyman SD, Jacobsen SE . C-kit ligand and Flt3 ligand: stem/progenitor cell factors with overlapping yet distinct activities. Blood 1998; 91: 1101–1134.

  4. 4

    Andersen MK, Pedersen-Bjergaard J, Kjeldsen L, Dufva IH, Brondum-Nielsen K . Clonal Ph-negative hematopoiesis in CML after therapy with imatinib mesylate is frequently characterized by trisomy 8. Leukemia 2002; 16: 1390–1393.

  5. 5

    Terre C, Eclache V, Rousselot P, Imbert M, Charrin C, Gervais C et al. France Intergroupe pour la Leucemie Myeloide Chronique. Report of 34 patients with clonal chromosomal abnormalities in Philadelphia-negative cells during imatinib treatment of Philadelphia-positive chronic myeloid leukemia. Leukemia 2004; 18: 1340–1346.

  6. 6

    Steidl U, Kronenwett R, Rohr U-P, Fenk R, Kliszewski S, Maercker C et al. Gene expression profiling identifies significant differences between the molecular phenotypes of bone marrow-derived and circulating human CD34+ hematopoietic stem cells. Blood 2002; 99: 2037–2044.

  7. 7

    Huber W, von Heydebreck A, Sultmann H, Poustka A, Vingron M . Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 2002; 18: 96–104.

  8. 8

    Goss Tusher V, Tibshirani R, Chu G . Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci 2001; 98: 5116–5121.

Download references


This work was supported by the Leukämie Liga e.V., Düsseldorf.

Author information

Correspondence to R Kronenwett.

Additional information

Supplementary Information

Supplementary Information accompanies the paper on the Leukemia website (

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Neumann, F., Teutsch, N., Kliszewski, S. et al. Gene expression profiling of Philadelphia chromosome (Ph)-negative CD34+ hematopoietic stem and progenitor cells of patients with Ph-positive CML in major molecular remission during therapy with imatinib. Leukemia 19, 458–460 (2005) doi:10.1038/sj.leu.2403615

Download citation

Further reading