Identification of genes potentially involved in disease ... - Nature

Leukemia (2005) 19, 998–1004 & 2005 Nature Publishing Group All rights reserved 0887-6924/05 $30.00 www.nature.com/leu

Identification of genes potentially involved in disease transformation of CML JJWM Janssen1, SM Klaver1, Q Waisfisz1, G Pasterkamp2, DPV de Kleijn2, GJ Schuurhuis1 and GJ Ossenkoppele1 1

Department of Hematology, VU University Medical Center, Amsterdam, The Netherlands; and 2Department of Experimental Cardiology, University Medical Center Utrecht, Utrecht, The Netherlands

In patients with chronic myeloid leukemia (CML) who do not reach a (near) complete cytogenetic response, the disease progresses over several years from an indolent, chronic phase into a rapidly fatal blast crisis. Events that are responsible for this transformation process are largely unknown. To identify changes in gene expression that occurred during the course of the disease, we performed cDNA subtraction on sequentially stored peripheral blood mononuclear cell pellets, collected throughout the course of disease of a single CML patient. In total, 32 differentially expressed sequences were identified, of which 27 corresponded to known genes. On quantitative PCR, eight of these genes, YWHAZ, GAS2, IL8, IL6, PBEF1, CCL4, SAT and MMRN, showed comparable differential expression in additional CML patient samples. This set of genes can be considered as a starting point for further research on causes of disease transformation in CML and may lead to new targets in the treatment of resistant CML. Leukemia (2005) 19, 998–1004. doi:10.1038/sj.leu.2403735 Published online 7 April 2005 Keywords: CML; transformation; subtractive hybridisation

Introduction Chronic myeloid leukemia (CML) is a malignant disorder of the hemopoietic stem cell that is characterized cytogenetically by the presence of the Philadelphia chromosome, which represents fusion of the BCR-gene derived from chromosome 22 with the ABL-gene from chromosome 9. The protein product of this fusion, BCR-ABL, has leukemogenic properties and is pivotal in the development of CML. Typically, after several years, the chronic phase of the disease transforms into a rapidly fatal blast crisis. In patients that obtain a (near-) complete cytogenetic response on interferon and/or imatinib, development of blast crisis is delayed and may even be prevented. Although the response percentage to imatinib is high, especially in patients with high Sokal or Euro-risk scores, primary and secondary resistance to the drug is seen. Moreover, a small number of patients present with disease in accelerated or even blastic phase, in which imatinib is clearly less effective.1 The molecular mechanisms that underlie BCR-ABL-mediated transformation are poorly understood; however, several events have been described to coincide with development of blast crisis. Mutations or deletions of P53 and its related gene TP73L (in around 25% of myeloid blast crises), doubling of the Philadelphia-chromosome, trisomy 8, upregulation of C-MYC and RB expression have been found.2–6 Changes in proliferation, differentiation, apoptosis sensitivity, DNA repair function, as well as loss of imprinting and genomic instability have also been observed upon disease transformation.7 In a small minority of Correspondence: Dr JJWM Janssen, Internist-hematologist, Department of Hematology, VU University medical center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands; Fax: þ 31 20 444 2601; E-mail: [email protected] Received 18 August 2004; accepted 31 January 2005; Published online 7 April 2005

cases of advanced stage CML, N- and K-RAS mutations have been reported.8 The question which of these changes occurred as a cause or as a result of the transformation itself has not been solved yet. To identify events responsible for disease progression, we performed PCR subtraction on samples from a CML patient of whom peripheral blood mononuclear cells had been collected throughout the course of the disease. A total of 32 differentially expressed sequences were identified, of which eight potentially relevant genes showed similar differential expression in an additional set of CML patients.

Materials and methods

Patients Suppression subtractive hybridization was performed on peripheral blood mononuclear cell pellets isolated at diagnosis, 2 years after diagnosis and at blast crisis of a single CML patient. This was a man of 51 years who was diagnosed with CML in 1995. Sokal score at diagnosis was 1.22, Euro score 2065. Cytogenetics revealed the standard translocation t(9;22)(q34;q11), RT-PCR showed a b3a2 fusion. He had no HLA-identical siblings and was given intensive treatment with induction chemotherapy as part of a clinical research protocol in August 1995.9 Next, interferon alpha 3–6 106 U s.c. q.d. was started, but because of lack of any response and intolerance to this drug, he switched to hydroxyurea after 1.5 years. At 3.6 years after diagnosis, the disease transformed into a blast crisis and he died of progressive bone marrow failure shortly thereafter. Blast cells at blast crisis were Sudan black positive and were positive on immunophenotyping for CD34, CD13, CD117 and CD45 (dim), but negative for CD41, CD5, CD2, CD19 and CD20, compatible with myeloid blast crisis. Characteristics of the index patient at the time points when samples for suppression subtractive hybridization were taken are shown in Table 1. Peripheral blood mononuclear cell pellets of an additional group of CML patients were used to validate the differential expression of genes as found in the index patient. None of these patients ever reached any cytogenetic response. Treatment consisted of interferon-alpha and/or hydroxyurea. Samples were collected throughout the course of the disease. Late chronic phase was defined as a time point at least 2 years after diagnosis and not fulfilling common criteria of blast crisis. Of several patients, sequential samples from diagnosis throughout blast crisis were available; in others, a more limited part of the disease course was represented, for example, from diagnosis to late chronic phase, or from late chronic phase to blast crisis.

Isolation of peripheral blood mononuclear cells Peripheral blood was collected by venipuncture into lithium heparin-coated glass containers. Peripheral blood mononuclear cells were isolated using density gradient centrifugation with

Genes involved in disease transformation of CML JJWM Janssen et al

999 Table 1

and directly sequenced thereafter on the ABI Prism 310 Genetic Analyzer (Applied Biosystems-Perkin Elmer, Foster City, CA, USA).

Clinical characteristics of index patient Year 1995

1997

1999

Diagnosis Late chronic phase Blast crisis Total leukocytes ( 109/l) Thrombocytes ( 109/l) Basophils (%) Blasts (%) Promyelocytes (%) Myelocytes (%) Metamyelocytes (%) Stabs (%) Segments (%) Monocytes (%) Ph+ (%)

176 368 1 2 5 22 10 9 43 2 100

7.5 275 6 2 2 8 6 10 46 14 100

57.9 2 9 85 0 0 1 0 1 2 ND

ND: not done.

Ficoll-Paque (1.077 g/ml). Cells were pelleted, snap-frozen in liquid nitrogen and then stored at 801C until use.

Suppression subtractive hybridization and library screening Total RNA was extracted using the Qiagen RNeasy kit (Qiagen, Westburg BV, The Netherlands) according to the manufacturer’s protocol. cDNA synthesis, subtraction PCR, library screening and sequence analysis were performed as described previously.10 Briefly, cDNA was prepared by using the SMART PCR cDNA synthesis kit (Clontech, Beckton Dickinson Biosciences BV, Woerden, The Netherlands) and samples were mixed for the PCR-Select cDNA subtraction kit according to the manufacturer’s protocol. Subtracted cDNA products were subsequently ligated into pGEM-T-Easy cloning vectors (Promega Benelux BV Leiden, The Netherlands) and transformed into supercompetent XL1Blue cells (Stratagene Europe, Amsterdam, The Netherlands). Plasmid DNA was spotted in an identical manner onto four separate nylon membranes (Hybond-N, Amersham Pharmacia) for every subtraction procedure. This was performed with libraries derived from the subtraction of (I) chronic phase minus diagnosis and reverse, (II) late chronic phase minus blast crisis and reverse and (III) blast crisis minus late chronic phase and reverse. After baking the membranes at 801C for 2 h, each blot was hybridized with a different radioactively labeled cDNA probe: for the subtraction procedure of chronic phase minus diagnosis, blot 1 was hybridized with unsubtracted cDNA of the late chronic phase sample, blot 2 with unsubtracted cDNA of the diagnosis sample, blot 3 with subtracted DNA of chronic phase minus diagnosis and blot 4 with subtracted cDNA of diagnosis minus chronic phase. Likewise, hybridizations were performed for all subtraction procedures. Only clones that showed a clear difference between the two subtracted probes and no or a low signal for the unsubtracted probes were selected for sequence analysis.

P53 sequence analysis For P53 sequence analysis, DNA from mononuclear cell pellets was isolated with the QIAamp DNA Blood Mini Kit (Qiagen) according to the manufacturer’s protocol. P53 exons 5–9 were amplified using PCR (primer sequences: see Table 2a)

Quantitative PCR (Q-PCR) PCR amplification was performed with a LightCycler real-time PCR machine (Roche Diagnostics, Almere, The Netherlands). Reaction volumes were 20 ml, consisting of 2 ml cDNA, 2 ml of LightCycler Fast Start DNA SYBR Green Mastermix (Roche) and 0.5 mM reverse and forward primers each. MgCl2 was added to a final concentration of 4 mM in all reactions except for GAPDH amplification to which 1.5 mM was added. In all PCR reactions, conditions were an initial denaturation step at 951C for 10 min, followed by 45 cycles, each for 15 s at 951C, 10 s at 591C and 10 s at 721C. For quantification, comparisons were made with standard curves of PCR products cloned into PGEM-T-easy vectors (Promega). GAPDH signals were used for normalization of expression for highly expressed genes, Beta-glucuronidase (GUS) signals for normalization of genes with a relative low expression. For primer design the GeneFisher program, accessible at http://bibiserv.techfak.uni-bielefeld.de/genefisher/ was used.11 Primers were purchased from Isogen (Maarssen, The Netherlands). Primer sequences are shown in Table 2b.

Statistical analysis Gene expression levels as expressed relative to GUS or GAPDH copy numbers were loge-transformed. Median values of the additional groups of patients at the three different disease stages were compared using the Mann–Whitney U-test for nonparametric variables. A P-value of less than 0.05 was considered to indicate a statistically significant difference.

Results and discussion Differential expression of the identified genes was validated in the index patient samples by Q-PCR. The results from subtractive hybridization and Q-PCR correlated for all tested genes, except for MEN1, which showed only weak differences on the library screening blots. Table 3 gives the extent of differential expression for all tested genes. Four genes showed less than two-fold difference, which was considered too low for reliable determination of differential expression. The differences Table 2a Exon Exon Exon Exon Exon Exon Exon Exon Exon Exon

5 5 6 6 7 7 8 8 9 9

Primer sequences p53

forward reverse forward reverse forward reverse forward reverse forward reverse

50 -GGAATTCTCTGTTCACTTGTGCCCTGACTTT-30 50 -GGAATTCGGGCAACCAGCCCTGTCGTC-30 50 -GGTTGCCCAGGGTCCCCAG-30 50 -CGGAGGGCCACTGACAACCA-30 50 -CTTGCCACAGGTCTCCCCAA-30 50 -AGGGGTCAGCGGCAAGCAGA-30 50 -ACCTGATTTCCTTACTGCCTCTTGC-30 50 -GGCATAACTGCACCCTTGGTCTCCT-30 50 -AGGAGACCAAGGGTGCAGTTATGCC-30 50 -ACCAGGAGCCATTGTCTTTGAGGC-30

Sequence primers Exon 5 50 -TCTCTCCAGCCCCAGCT-30 Exon 6 50 -CAACCACCCTTAACCC-30 Exon 7 50 -GCTGGGGCACAGCAGGC-30 Exon 8 50 -CTTCTTGTCCTGCTTGC-30 Exon 9 50 -TTAGACTGGAAACTTTC-30

Leukemia


1000 Table 2b APEX1 F APEX1 R CCL4 F CCL4 R CD96 F CD96 R CHI3L2 F CHI3L2 R DC2 F DC2 R DNAJB1 F DNAJB1 R EMR1 F EMR1 R FPR 1 F FPR 1 R GAPDH F GAPDH R GAS2 F GAS2 R GUS F GUS R HSPA8 F HSPA8 R IL6 F IL6 R IL8 F IL8 R MEN1 F MEN1 R MMRN F MMRN R NACSINF NACSINR NXF1 F NXF1 R IVNS1ABP F IVNS1ABP R PBEF1 F PBEF1 R SAT F SAT R YWHAZ F YWHAZ R

Primer sequences Q-PCR 50 -TGGCAAACCTGCCACACTCAAG-30 50 -CCAGGCAGCTCCTGAAGTTCAG-30 50 -CAGCGCTCTCAGCACCAATGG-30 50 -GATCAGCACAGACTTGCTTGCTTC-30 50 -CCACTCCTCAACCCAGCAATTCC-30 50 -GGCTGTGCTGGTAAAGACATTGTC-30 50 -TCTTGACTGCGGGCGTATC-30 50 -CCCAAGACCCATGGAAGTCA-30 50 -CTGGTGGTGGTGTCTTACTTCC-30 50 -GAAGCTGGATGCAAGTCCTTCC-30 50 -TACACATTCCATGGAGACCCTCATG-30 50 -GGCCAAAGTTCACGTTGGTGAAG-30 50 -GCATCTTTTTGGAAACCCTCAG-30 50 -AGTCCAACGTCACATTCTCTTC-30 50 -CCTGCTGTATCTGCCTGGCTA-30 50 -GAATCCAGCCACCCAGATCA-30 50 -AGGTCATCCCTGAGCTGAACGG-30 50 -CGCCTGCTTCACCACCTTCTTG-30 50 -CTGCCGAATGCTGCAGATC-30 50 -TGGCAGAGACCACCAAGTAG-30 50 -GAAAATATGTGGTTGGAGAGCTCATT-30 50 -CCGAGTGAAGATCCCCTTTTTA-30 50 -ATTGCAGAAGCCTACCTTGG-30 50 -CGTTTCTTTCTGCTCCAACC-30 50 -CCAGAGCTGTGCAGATGAGT-30 50 -TCAGCAGGCTGGCATTTGT-30 50 -GGCTCTCTTGGCAGCCTTCC-30 50 -GCGCAGTGTGGTCCACTCTC-30 50 -AGAAGGTCTCCGATGTCATATGG-30 50 -ACCGGAGCTGTCCAATTTGG-30 50 -CAGATTTAAGGCGTTGGAAGC-30 50 -GTAGCATTCAGTTCTGACACAC-30 50 -CCACCAGTCCTCTCACCAAA-30 50 -TTGGCTTTCGCCCCAAAAC-30 50 -AGAGGCGGTTCTGGTATTCG-30 50 -GCACAGTAACATGAATGCGATC-30 50 -ACCGTTGTAATGCAGGAGTGTG-30 50 -TCACAGACTGCAGACTGGTGTC-30 50 -CAAGGACCCAGTTGCTGATC-30 50 -CTCCTTTTCCTTCCTCCAGTG-30 50 -CACAGCATTGTTGGTTTTGC-30 50 -GGATGGTTCATTCCATTCTGC-30 50 -GAAAGCCTGCTCTCTTGCAAAGAC-30 50 -CCTCCTTCTCCTGCTTCAGCTTC-30

in expression between time points for the remaining genes varied from 2- to 106-fold (linear values). Although possibly interesting, expressed sequence tags and genes that were only predicted were not validated by quantitative PCR. Some of the identified differentially expressed genes presumably represent differences in maturation stages between undifferentiated blasts at blast crisis and more mature myeloid progenitors at diagnosis and late chronic phase. Their identification demonstrates the power of the suppression subtractive hybridization technique; however, these genes are probably irrelevant to the disease transformation process. These genes include: MPO, LYZ, encoding the cytoplasmic enzymes myeloperoxidase and lysozyme, respectively; PRG1, encoding a proteoglycan that is stored in more mature myeloid cells;12 FPR1, a neutrophil G-protein-coupled receptor;13 S100A8, encoding calgranulin A that is abundantly expressed in neutrophil cytoplasm.14 The upregulation of LARS (leucyl tRNA synthetase)15 was interpreted as reflecting increased transcriptional activity in blast crisis cells. The differential expression of the other genes was verified by Q-PCR in an additional group of CML patients. At least six (range 6–47) individual patient samples were analyzed for every Leukemia

Table 3 List of differentially expressed genes in index patient and level of differential expression Name of gene

Accession number

Fold difference (by Q-PCR)a

Diagnosis4blast crisis PRG1 NM_002727 IVNS1ABP NM_006469 MPO NM_000250

ND o2 (o0.7) ND

Diagnosisoblast crisis EMR1 NM_001974 CD96 NM_198196 GAS2 NM_005256 CHI3L2 NM_004000

o2 3 12 11

(o0.7) (1.1) (2.5) (2.4)

Diagnosis4late chronic phase DNAJB1 NM_006145 HSPA8 NM_006597 NXF1 NM_006362 DC2 NM_021227 APEX1 NM_080648 RPS6 NM_001010 EEF1A1 NM_001402 NM_152546 FLJ25286b

35 5 2 o2 4 o2 3

(3.6) (1.6) (0.7) (o0.7) (1.4) (o0.7) (1.1) ND

7 106 3 7 42 9

(1.9) (4.7) (1.1) (1.9) (3.7) (2.2)

Diagnosisolate chronic phase SAT NM_002970 CCL4 NM_002984 YWHAZ NM_145690 PBEF1 NM_182790 IL6 NM_000600 IL8 NM_000584 Late chronic phase4blast crisis IL8 NM_000584 S100A8 NM_002964 FPR1 NM_002029 LYZ NM_000239 MEN1 AF001893 Late chronic phaseoblast crisis MMRN NM_007351 NACSIN NM_015252 LARS NM_020117 NM_152546 FLJ25286b d NM_004772 C5orf13 NM_016304 C15orf15d NM_018243 FLJ10849b

2 (0.7) ND ND ND c

6 (1.8) 24 (3.2) ND ND ND ND ND

a

Between brackets: loge-transformed difference. Genes hypothetically encoding proteins. c Reverse to subtraction result. d Orf, open reading frame; ND, not done; NS, nonsignificant; D, diagnosis; LCP, late chronic phase; BC, blast crisis. b

time point of the disease for each gene. For 14 genes, subtraction results could not be confirmed in this additional group of patients. However, YWHAZ, GAS2, IL8, IL6, PBEF1, CCL4, SAT and MMRN showed a similar pattern of expression in the additional patient samples as was found in the index patient. Figure 1a–h shows the median relative expression levels of these genes in this set of patient samples as observed in samples taken at diagnosis, late chronic phase or blast crisis. These consistently differentially expressed genes will be discussed in more detail below.

Apoptosis related genes Delayed apoptosis is a feature of early phase CML,16 but even more so of advanced disease stages in which upregulation of


1001

a

e

YWHAZ

1.2 0.8

0.0

n=19

n=14

D

LCP

n=11

NS

expression (% of GUS)

3.0 2.0 1.0

BC

GAS2 p = 0.025 2.0

p = 0.008

1.6 1.2 0.8 n=36

n=26

n=46

f

6.0 5.0

0.0 D

LCP

n=9

n=6

n=13

D

LCP

BC

CCL4 p = 0.006

3.0 2.0 1.0

n=11

n=9

n=10

D

LCP

BC

g


2.0 1.0 n=47

n=38

n=26

D

LCP

BC

0.0

IL6 p = 0.021

NS

5.0 4.0 3.0 2.0 1.0 0.0

n=11

n=7

n=9

D

LCP

BC

h

MMRN p = 0.02

p = 0.029

3.0 2.0 1.0


p < 0.001

3.0

4.0

SAT p = 0.044

p = 0.009

n=17

n=10

n=11

D

LCP

BC

0.0 -1.0


expression (% of GAPDH)

p < 0.0001

4.0

IL8

p = 0.027

4.0

0.0

BC

c

d

p = 0.01

0.0

b

0.4


p = 0.01 1.6

p = 0.005

4.0



2.0

0.4

PBEF1

5.0

p = 0.045

5.0

p = 0.02

4.0 3.0 2.0 1.0 n=11 0.0

D

n=9

n=9

LCP

BC

Figure 1 (a–h) Median relative expression levels of consistently differentially expressed genes as observed in samples taken at diagnosis, late chronic phase or blast crisis. Expression levels have been loge-transformed and are relative to the expression of GUS (Beta-glucuronidase) or GAPDH (Glyceraldehyde-3-phosphate dehydrogenase). Abbreviations: YWHAZ ¼ tyrosine 3-monooxygenase/tryptophan 5-monooxygenase; GAS2 ¼ growth arrest specific-2; IL8 ¼ interleukin-8; IL6 ¼ interleukin-6; PBEF1 ¼ Pre-B-cell colony-enhancing factor precursor; CCL4 ¼ small inducible cytokine A4; SAT ¼ spermidine/spermine-N-acetyltransferase; MMRN ¼ multimerin. NS ¼ not significant. D ¼ diagnosis; LCP ¼ late chronic phase; BC ¼ blast crisis.

BCR-ABL is a common event.17.YWHAZ (tyrosine 3-monooxygenase/tryptophan 5-monooxygenase;14-3-3 zeta protein) is involved in the regulation of apoptosis. Subtractive hybridization revealed higher expression in late chronic phase than at diagnosis. Q-PCR results in the group of CML patients at three disease stages show that YWHAZ expression peaks at late

chronic phase (Figure 1a). YWHAZ has a central role in the regulation of apoptosis, cell growth and cycling by modulating the functional activities of several key signaling proteins.18 For instance, the proapoptotic protein BAX can bind to the YWHAZ protein, which leads to increased levels of homodimerized BCL2 and hence to decreased apoptosis. A YWHAZ family Leukemia


1002

GAS2 (% of GUS)

2.8

GAS2 in blast crisis

2.4 2.0 1.6 1.2 0.8 0.4 0.0 MYELOID

LYMPHOID/BIPHENOTYPIC

Figure 2 Median expression levels of GAS2 in myeloid vs lymphoid or biphenotypic blast crisis. Expression levels have been loge-transformed and are relative to the expression of GUS.

member, BAP-1, shows 80% homology with YWHAZ and is essential for the transforming properties of BCR-ABL.19 Although this association has not been described for the YWHAZ protein, it may similarly be involved in the BCR-ABL-induced apoptosis resistance. The protein product of the GAS2 (growth arrest specific-2) gene, whose expression has not previously been reported in hematopoietic cells, is a caspase-3 substrate. It plays a role in the regulation of microfilament and cell shape changes during apoptosis.20 Overexpression of the protein stabilizes P53 and thereby increases cell susceptibility to apoptosis following DNA-damaging agents.21 Therefore, the upregulation of GAS2 in blast crisis compared to diagnosis seems to be contradictory to its known physiological role. Q-PCR in a large group of patients shows a remarkable difference in expression between diagnosis and blast crisis, as was found by subtractive hybridization, but even more between late chronic phase and blast crisis (see Figure 1b). Separating blast crises into myeloid vs lymphoid or biphenotypic reveals higher, although nonsignificant GAS2 expression in myeloid crises (Figure 2). As P53 mutations mainly occur in myeloid blast crisis, we hypothesized that high GAS2 expression might represent P53 abnormalities. Therefore, exons 5–9 of the P53 gene (the hot spot region for mutations) were sequenced in the blast crisis sample of the index patient, but no mutations were found (results not shown).

Cytokines Some genes encoding proliferation-inducing cytokines and interleukins were also identified as being higher expressed in late chronic phase than at both diagnosis and blast crisis, such as IL8. Consistent results were found with Q-PCR in other CML patients (see Figure 1c). IL8 has mitogenic, motogenic and angiogenic properties.22 A high level of expression has been related to disease progression in CLL23 and it was strongly linked to refractory disease in T-ALL.24 If high expression in late chronic phase CML has relevance for CML transformation as well is unknown but warrants further study. A second cytokine gene, IL6, was much higher expressed in late chronic phase compared to both diagnosis and blast crisis in the index patient, as well as in the additional group of patients (see Figure 1d). In agreement with this, it has previously been shown that IL6 levels were higher in cultures of CML monocytes of treated CML than those obtained at diagnosis and at blast crisis.25 At blast crisis, levels were very low, although the difference with late chronic phase was not significant. ConLeukemia

stitutively active IL6 induces continuous JAK/STAT activation in AML.26 This, in turn, induces cell cycle transition through upregulation of cyclins D2, D3, A and cdc25A, and the concomitant downregulation of p21 and p27.27 It may be hypothesized that constant IL6 stimulation in chronic phase CML induces increased proliferation, which could predispose to secondary abnormalities that can culminate in blast crisis. Otherwise, low IL6 levels at diagnosis and at blast crisis may be considered to reflect inadequate monocyte function.25 PBEF1 (Pre-B-cell colony-enhancing factor precursor) is, like IL6 and IL8, highest expressed in late chronic phase (Figure 1e). Previous reports showed the gene to be expressed in bone marrow stromal cells and in activated lymphocytes. PBEF1 stimulates the growth of pre-B-cell colonies but no effect was seen on myeloid or erythroid colonies.28 Using a comparable suppression subtractive hybridization as ours, it was found to be upregulated in colonic carcinoma samples as compared to normal mucosa.29 What role it may play in CML cells remains speculative, but if PBEF1 also increases proliferation of myeloid progenitors, this might, as already stated for IL6, facilitate the occurrence of secondary genetic events. Remarkably, also the CC chemokine gene CCL4 (‘small inducible cytokine A4’; SCYA4; MIP-1 beta) showed highest expression at late chronic phase, while expression at both diagnosis and blast crisis were clearly lower (see Figure 1f). CCL4 is an antagonist of the proliferation-inhibiting effect of MIP-1 alpha30 and induces cytotoxic T- and NK-cell proliferation and cytolytic capacity.31

Other genes Higher expression of the growth- and cell cycle-associated gene SAT (spermidine/spermine N-acetyltransferase) in late chronic phase compared to diagnosis was found at subtractive hybridization. Q-PCR in additional patient samples reveals that expression at diagnosis is lowest compared to the other two disease stages (see Figure 1g). SAT encodes a ratelimiting enzyme in the catabolism of polyamines, substances involved in many cellular processes, including chromatin condensation, maintenance of DNA structure, RNA processing, translation and protein activation. Upregulation of SAT leads to apoptosis in melanoma cells.32 Its role in CML remains speculative. The gene for MMRN (multimerin), encoding a specific factor V/Va-binding protein, was upregulated at blast crisis compared to late chronic phase according to the subtractive hybridization and this was confirmed by Q-PCR in the additional patient samples (Figure 1h). The expression levels at diagnosis were comparable with those at late chronic phase. Multimerin is a constituent of platelet alpha granules33 and thus elevated expression of this gene in blast cells may reflect differentiation into the megakaryocytic lineage.34 However, immunophenotyping showed negativity for the megakaryocytic marker CD41 on the blasts of all five patients where it was tested. Previous approaches to elucidate mechanisms of disease transformation with similar techniques as described by us found downregulation of, among others, PIASy, a potential inhibitor of the STAT proteins with apoptosis-promoting and growth suppressive properties35 and upregulation of PRAME36 in blast crisis progenitor cells derived from the bone marrow. In another, probably less appropriate experiment, overexpression of ribosomal proteins, of a c-myc-related protein and of a cellular adhesion molecule was found in K562 cells, serving as a model of CML blast crisis, compared to primary CML spleen cells. This


1003 comparison may have revealed many aspecific differences in gene expression that are unrelated to disease progression, but are due to secondary clonal aberrations that may occur during long-term culture of the cell line or to specific individual properties of the cells that constitute the K562 cell line. None of these studies showed overlapping results with each other or with our findings; further data on the role of these genes in CML is lacking. We are aware that the gene expression profiling as performed by us potentially harbours several pitfalls. Firstly, we used peripheral blood samples from only one patient for the subtraction procedure. This may have led to patient-specific changes in gene expression. However, a significant number of highly interesting genes showed a comparable expression pattern in a larger set of patients. Secondly, differences in the cellular composition of mononuclear cell fractions between the separate phases existed and may have caused irrelevant differential expression results. At diagnosis of CML, more myelocytes and fewer monocytes compared to late chronic phase are present and these were part of the mononuclear cell fractions studied by us. This is also true for the subtraction analyses comparing late chronic phase with blast crisis or diagnosis and blast crisis since the blast crisis samples typically have high blast counts. However, differential expression of blasts at blast crisis vs more mature myeloid cells at diagnosis or late chronic phase may actually reveal genes that are in fact responsible for the maturation block; therefore, we feel that our approach is at least complementary to others where progenitor cells are used to perform gene expression comparisons. Thirdly, differences in expression levels may also result from treatment. For most of the identified genes, no literature data exist on the potential effect of hydroxyurea or interferon alpha on their expression. Only for IL8 it has been described that treatment with interferon alpha of CML cells and with hydroxyurea for sickle cell anemia decreases its expression,37,38 which is contradictory to what we found by subtractive hybridization and Q-PCR. As patients cannot be left untreated when a diagnosis of CML is made, drug-induced changes are an inherent problem of investigations into the causes of disease transformation using patient samples. Moreover, genetic events that are responsible for disease transformation occur while patients are on medication. We used peripheral blood samples for the subtraction procedure. Although the use of bone marrow samples may also appear appropriate in these experiments, a drawback of bone marrow aspirates is the unquantifiable contamination with peripheral blood cells. These may have different properties compared to bone marrow-derived cells, and as immature cells are only present in the peripheral blood in large numbers at diagnosis and at blast crisis, comparisons between bone marrow cell populations taken at different stages of the disease may reveal differential expression reflecting the nature of the cells rather than specific changes reflecting stage of disease. Secondly, many CML patients have inaspirable bone marrow due to fibrosis, leading to selection of patients if only bone marrow cells would be used. We conclude that several genes involved in apoptosis regulation and cell proliferation, processes that are thought to play a role in the transformation process, as well as genes encoding cytokines, are differentially expressed in subsequent disease stages. This interesting set of genes can be considered as a starting point for further research on causes of disease transformation in CML and may lead to new targets in the treatment of resistant CML.

References 1 Druker BJ, Sawyers CL, Kantarjian H, Resta DJ, Reese SF, Ford JM et al. Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med 2001; 344: 1038–1042. 2 Bi S, Lanza F, Goldman JM. The involvement of ‘tumor suppressor’ p53 in normal and chronic myelogenous leukemia hemopoiesis. Cancer Res 1994; 54: 582–586. 3 Yamaguchi H, Inokuchi K, Sakuma Y, Dan K. Mutation of the p51/ p63 gene is associated with blastic crisis in chronic myelogenous leukemia. Leukemia 2001; 15: 1729–1734. 4 Griesshammer M, Heinze B, Bangerter M, Heimpel H, Fliedner TM. Karyotype abnormalities and their clinical significance in blast crisis of chronic myeloid leukemia. J Mol Med 1997; 75: 836–838. 5 Furukawa Y, DeCaprio JA, Belvin M, Griffin JD. Heterogeneous expression of the product of the retinoblastoma susceptibility gene in primary human leukemia cells. Oncogene 1991; 6: 1343–1346. 6 Handa H, Hegde UP, Kotelnikov VM, Mundle SD, Dong LM, Burke P et al. Bcl-2 and c-myc expression, cell cycle kinetics and apoptosis during the progression of chronic myelogenous leukemia from diagnosis to blastic phase. Leuk Res 1997; 21: 479–489. 7 Shet AS, Jahagirdar BN, Verfaillie CM. Chronic myelogenous leukemia: mechanisms underlying disease progression. Leukemia 2002; 16: 1402–1411. 8 LeMaistre A, Lee MS, Talpaz M, Kantarjian HM, Freireich EJ, Deisseroth AB et al. Ras oncogene mutations are rare late stage events in chronic myelogenous leukemia. Blood 1989; 73: 889–891. 9 Janssen JJWM, van Rijn RS, van der Holt B, Schuurhuis GJ, Vellenga E, Verhoef GE et al. Mobilisation of haemopoietic progenitors in CML: a second course of intensive chemotherapy does not improve Ph-negativity in stem cell harvests. Bone Marrow Transplant 2000; 25: 1147–1155. 10 de Kleijn DP, Smeets MB, Kemmeren PP, Lim SK, Van Middelaar BJ, Velema E et al. Acute-phase protein haptoglobin is a cell migration factor involved in arterial restructuring. FASEB J 2002; 16: 1123–1125. 11 Giegerich R, Meyer F, Schleiermacher C. GeneFisher – software support for the detection of postulated genes. Proc Int Conf Intell Syst Mol Biol 1996; 4: 68–77. 12 Stevens RL, Avraham S, Gartner MC, Bruns GA, Austen KF, Weis JH. Isolation and characterization of a cDNA that encodes the peptide core of the secretory granule proteoglycan of human promyelocytic leukemia HL-60 cells. J Biol Chem 1988; 263: 7287–7291. 13 Le Y, Murphy PM, Wang JM. Formyl-peptide receptors revisited. Trends Immunol 2002; 23: 541–548. 14 Passey RJ, Xu K, Hume DA, Geczy CL. S100A8: emerging functions and regulation. J Leukoc Biol 1999; 66: 549–556. 15 Bullard JM, Cai YC, Spremulli LL. Expression and characterization of the human mitochondrial leucyl-tRNA synthetase. Biochim Biophys Acta 2000; 1490: 245–258. 16 Gordon MY, Dazzi F, Marley SB, Lewis JL, Nguyen D, Grand FH et al. Cell biology of CML cells. Leukemia 1999; 13 (Suppl. 1): S65–S71. 17 Johansson B, Fioretos T, Mitelman F. Cytogenetic and molecular genetic evolution of chronic myeloid leukemia. Acta Haematol 2002; 107: 76–94. 18 Hermeking H. The 14-3-3 cancer connection. Nat Rev Cancer 2003; 3: 931–943. 19 Reuther GW, Fu H, Cripe LD, Collier RJ, Pendergast AM. Association of the protein kinases c-Bcr and Bcr-Abl with proteins of the 14-3-3 family. Science 1994; 266: 129–133. 20 Sgorbissa A, Benetti R, Marzinotto S, Schneider C, Brancolini C. Caspase-3 and caspase-7 but not caspase-6 cleave Gas2 in vitro: implications for microfilament reorganization during apoptosis. J Cell Sci 1999; 112 (Part 23): 4475–4482. 21 Benetti R, Del Sal G, Monte M, Paroni G, Brancolini C, Schneider C. The death substrate Gas2 binds m-calpain and increases susceptibility to p53-dependent apoptosis. EMBO J 2001; 20: 2702–2714. 22 Xie K. Interleukin-8 and human cancer biology. Cytokine Growth Factor Rev 2001; 12: 375–391. Leukemia


1004 23 Wierda WG, Johnson MM, Do KA, Manshouri T, Dey A, O’Brien S et al. Plasma interleukin 8 level predicts for survival in chronic lymphocytic leukaemia. Br J Haematol 2003; 120: 452–456. 24 Chiaretti S, Li X, Gentleman R, Vitale A, Vignetti M, Mandelli F et al. Gene expression profile of adult T-cell acute lymphocytic leukemia identifies distinct subsets of patients with different response to therapy and survival. Blood 2004; 103: 2771–2778. 25 Anand M, Chodda SK, Parikh PM, Nadkarni JS. Abnormal levels of proinflammatory cytokines in serum and monocyte cultures from patients with chronic myeloid leukemia in different stages, and their role in prognosis. Hematol Oncol 1998; 16: 143–154. 26 Schuringa JJ, Wierenga AT, Kruijer W, Vellenga E. Constitutive Stat3, Tyr705, and Ser727 phosphorylation in acute myeloid leukemia cells caused by the autocrine secretion of interleukin-6. Blood 2000; 95: 3765–3770. 27 Fukada T, Ohtani T, Yoshida Y, Shirogane T, Nishida K, Nakajima K et al. STAT3 orchestrates contradictory signals in cytokineinduced G1 to S cell-cycle transition. EMBO J 1998; 17: 6670–6677. 28 Samal B, Sun Y, Stearns G, Xie C, Suggs S, McNiece I. Cloning and characterization of the cDNA encoding a novel human pre-B-cell colony-enhancing factor. Mol Cell Biol 1994; 14: 1431–1437. 29 Hufton SE, Moerkerk PT, Brandwijk R, de Bruine AP, Arends JW, Hoogenboom HR. A profile of differentially expressed genes in primary colorectal cancer using suppression subtractive hybridization. FEBS Lett 1999; 463: 77–82. 30 Eaves CJ, Cashman JD, Wolpe SD, Eaves AC. Unresponsiveness of primitive chronic myeloid leukemia cells to macrophage inflammatory protein 1 alpha, an inhibitor of primitive normal

Leukemia

31

32 33 34

35

36 37

38

hematopoietic cells. Proc Natl Acad Sci USA 1993; 90: 12015–12019. Taub DD, Ortaldo JR, Turcovski-Corrales SM, Key ML, Longo DL, Murphy WJ. Beta chemokines costimulate lymphocyte cytolysis, proliferation, and lymphokine production. J Leukoc Biol 1996; 59: 81–89. Chen Y, Kramer DL, Jell J, Vujcic S, Porter CW. Small interfering RNA suppression of polyamine analog-induced spermidine/spermine n1-acetyltransferase. Mol Pharmacol 2003; 64: 1153–1159. Hayward CP, Kelton JG. Multimerin. Curr Opin Hematol 1995; 2: 339–344. Hayward CP, Bainton DF, Smith JW, Horsewood P, Stead RH, Podor TJ et al. Multimerin is found in the alpha-granules of resting platelets and is synthesized by a megakaryocytic cell line. J Clin Invest 1993; 91: 2630–2639. Ohmine K, Ota J, Ueda M, Ueno S, Yoshida K, Yamashita Y et al. Characterization of stage progression in chronic myeloid leukemia by DNA microarray with purified hematopoietic stem cells. Oncogene 2001; 20: 8249–8257. Watari K, Tojo A, Nagamura-Inoue T, Nagamura F, Takeshita A, Fukushima T et al. Identification of a melanoma antigen, PRAME, as a BCR/ABL-inducible gene. FEBS Lett 2000; 466: 367–371. Aman MJ, Rudolf G, Goldschmitt J, Aulitzky WE, Lam C, Huber C et al. Type-I interferons are potent inhibitors of interleukin-8 production in hematopoietic and bone marrow stromal cells. Blood 1993; 82: 2371–2378. Saleh AW, Hillen HF, Duits AJ. Levels of endothelial, neutrophil and platelet-specific factors in sickle cell anemia patients during hydroxyurea therapy. Acta Haematol 1999; 102: 31–37.