acute myeloid leukemia
Gomes e Silva, F.P.
Citation
Gomes e Silva, F. P. (2009, March 3). Genome wide characterization of minimally differentiated acute myeloid leukemia. Retrieved from https://hdl.handle.net/1887/13569
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/13569
Note: To cite this publication please use the final published version (if applicable).
2
Fernando PG Silva, Bruno Morolli, Clelia T Storlazzi, Luisa Anelli, Hans Wessels, Vladimir Bezrookove, Hanneke C Kluin-Nelemans & Micheline Giphart-Gassler
Oncogene 2003;22:538-547
tumor suppressor gene
Abstract
Based on our previous results indicating the presence of a tumor suppressor gene (TSG), chromosome 21 was analyzed for loss of heterozygosity (LOH) in eighteen patients with acute myeloid leukemia (17, AML-M0; one, AML-M1). Allelotyping at polymorphic loci was performed on purified material, allowing unequivocal detection of allelic loss and homozygous deletions. Six AML-M0 patients shared a common region of LOH harboring a single gene: RUNX1 (AML1), the most frequent site of translocations in acute leukemia and a well-known fusion oncogene. Fluorescence in situ hybridization allowed the identification of deletions with breakpoints within RUNX1 in two patients as the cause of LOH. In the four others the LOH pattern and the presence of two karyotypically normal chromosomes 21 were in line with mitotic recombination. Further molecular and cytogenetic analyses showed that this caused homozygosity of primary RUNX1 mutations: two point mutations, a partial deletion and, most significantly, a complete deletion of RUNX1. These findings identify RUNX1 as a classical TSG: both alleles are mutated or absent in cancer cells from 4 of the 17 AML-M0 patients examined. In contrast to AML-M0, the AML-M1 patient was trisomic for chromosome 21 and has two mutated and one normal RUNX1 allele, suggesting that the order of mutagenic events leading to leukemia may influence the predominant tumor type.
Keywords: RUNX1; tumor suppressor gene; AML; leukemia
Introduction
RUNX1 encodes the core binding protein (CBP) α subunit of a heterodimeric transcription factor complex involved in the regulation of hematopoiesis and contains a Runt DNA- binding domain.1 This 128 amino acid domain is an evolutionarily conserved protein motif that is responsible both for DNA binding and heterodimerization with the β subunit.2;3 The β subunit, encoded by the CBFB gene, does not directly bind DNA but enhances the ability of the α subunit to do so.4;5
A number of hematopoietic specific genes are activated by RUNX1 such as interleukin-3, granulocyte-macrophage colony-stimulating factor and myeloperoxidase.6-8 RUNX1 per se is not sufficient for transcriptional activation. It has been hypothesized that RUNX1 acts as an organizing factor that coordinates the assembly of transcriptional activation complexes.9-13 Supporting this view are reports of the interactions between RUNX1 and other transcription factors, such as Ets or C/EBPα,11;14-17 and between RUNX1 and co-activator proteins such as p300/CREB-binding protein and YAP.18;19
RUNX1 is the most frequent site of translocations in acute leukemia such as t(8;21)(q22;q22) associated with AML-M2 and t(12;21)(p13;q22) associated with pediatric ALL.20 The fusion oncoproteins resulting from these translocations AML1-ETO and TEL-AML1, respectively, were suggested to act as transcriptional repressors in a dominant-negative manner, by recruiting co-repressors, and blocking the transcription of the wild type allele.21 These chimeric proteins retain their RUNT domain and might also suppress transactivation by dimerizing with CBFB in a more efficient manner than wild type RUNX1.22
Mutations in RUNX1 have been observed in familial platelet disorder (FPD), a disease with predisposition to AML.22-24 In addition, mutations, some of which were biallelic, have been observed in sporadic leukemias.25;26 Furthermore RUNX3, a gene of the same family of RUNX1, was found to be a tumor suppressor gene in gastric carcinogenesis.27 This suggests that RUNX1 might also function as a tumor suppressor gene (TSG). In recent years the number of TSG alterations associated with hematological malignancies has grown.28 Genetically, recessive TSGs are inactivated by two sequential mutational events.29 the second often resulting in loss of heterozygosity (LOH). LOH analysis, although widely used to map regions of chromosomes harboring candidate TSGs, has rarely identified these genes and the usefulness of this technique has been questioned.30 The major pattern of LOH in primary and transformed lymphocytes is in agreement with mitotic recombination (MR) as the LOH- generating mechanism.31-34 MR reduces a chromosome region to homozygosity without changing chromosome ploidy and can be efficiently identified by genotyping polymorphic markers for telomeric regions.35 Using this approach in a previous study, we identified a candidate TSG on chromosome 21 in a patient diagnosed with acute minimally differentiated leukemia (AML-M0).35 In the present study, we identify RUNX1 as the specific target for LOH in 6 of 17 AML-M0 patients. Further molecular and cytogenetic analyses have identified two consecutive mutational hits completely deleting the RUNX1 gene in one patient and duplicating RUNX1 mutations in three other patients. Our data strongly suggest that, in addition to its previously reported role in dominantly acting fusion oncogenes, RUNX1 may also act as a classical tumor suppressor gene.
Results
Six patients shared a small common region of LOH harboring a single gene: RUNX1 The number of TSGs identified in sporadic cancer by LOH analysis is limited, mainly due to tumor heterogeneity or impurity of the tumor sample.30 To obtain pure tumor and control material, we sorted leukemic blasts and control T-cells by flow cytometry from archive material of 18 patients, 17 diagnosed with acute minimally differentiated myeloid leukemia (AML-M0) and 1 with AML-M1 (Table 1).
Table 1. Classification and karyotype of patients
Patient no Diagnose Karyotyping COBRA
1 AML M0 n.d.
2 # AML M0 47,XX,t(4;12)(q12;p13), -21, +2 mar 47,XX,t(4;12),-21,+ mar(21) +mar(21)
3 x AML M0 46,XX
4 x AML M0 46,XY, t(2;3)(p23;q27)
5 x# AML M0
45,XY,-3,add(5)(q11),-6, der(7) t(6;7)(p?;p22), del(7)(q32),der(17) t(6;17)(?;p11.2),der(21) t(3;21) (?;q22),ins(21;6)(q22;?),+r(6)
45,XY,-3,der(5)t(3;5),-6, der(7)t(6;7), der(17) t(7;17),der(21)t(6;21), + mar
6 x AML M0 47,XY 7 AML M0 n.d.
8 x AML M0 46,XX
9 x AML M0 52,XX, t(1;4)(p13;p12),+6,+8, t(10;12) (q11;p11),+18,+19,+20,+21
10 x AML M0 45,XY, t(3;9)(q28;q21),del(7)(p15) 11 x AML M0 48,XY,+13,+14
12 # AML M1 48,XY,+9,+21
13 # AML M0 n.d. 46,XX
14 # AML M0
46,XY,der(1;7)(q10;p10),der(1;16) (q10;p10),
-6,+9
15 # AML M0 46,XY 46,XY,der(18)t(18;?)
16 x AML M0 46,XY
18 # AML M0 45,XY,-7, del(20)(q1?2) 45,XY,-7 19 xy AML M0 46,XY
n.d. : not done; x : karyotyping at diagnose; # : karyotyping of cultured cryo-preserved cells; y : equals patient 9 in Giphart-Gassler et al., 1998,35 previously classified as AUL . Patient 17 was excluded due to misclassification.
DNA isolated from these purified cells was analyzed with polymorphic markers derived from the q arm of chromosome 21 (Figure 1a). In the blasts of patients 2, 5, 7, 11, 14, 18 and 19 the signal of only a single allele was retained at one to several loci. In 4 patients (7, 11,
18 and 19) LOH covered a major part of the q arm to the telomere. In the patients from which a karyotype was made, no cytogenetic abnormalities in chromosome 21 were detected (Table 1), indicating that LOH occurred by MR. Evidently, little information about the exact location of a TSG is provided by the extended LOH patterns resulting from MR. Fortunately, a small region of LOH was present in patient 14. LOH observed in this patient at D21S211 warranted further investigation, since this marker appeared homozygously deleted in blasts from patient 18 and because of its close location to RUNX1.
Figure 1. Loss of heterozygosity analyses. A) LOH was accessed using chromosome 21 q arm polymorphic markers, in particular markers at the telomeric region to detect MR. The markers used and their genomic positions are listed on the left side. At markers where LOH is indicated the loss of one allele in the tumor cells was
>95%. SROs (small regions of overlap) are detected at D21S211 and D21S1446. The LOH at marker D21S1446 in patient 5 was considered less relevant, since we could not exclude non-specific LOH because of both the telomeric position of this marker and the many chromosomal abnormalities present in patient 5 (Table 1). B) Based on the LOH in patient 14 and the absence in patient 18 of D21S211, more detailed LOH studies were conducted for the region from 31,5 to 34,4Mb, which harbors D21S211 and RUNX1. The markers used are listed on the left alongside with the genomic organization of RUNX1 (exons 1 to 8). Gaps A, B and C represent approximately 780 Kb, 600 Kb and 800 Kb, respectively.
0 -LOH;
-non informative;
-heterozygous, no LOH;
-allelic imbalance (AI);
-deletion and
* -not done.
The 3-Mb region containing RUNX1 was analyzed with a further set of 8 markers (Figure 1b). Indeed, RUNX1 was the target for LOH: patient 14 showed LOH from markers 57-76 to D21S211, in patient 2 LOH started within or close to RUNX1 and in patient 18 LOH was interrupted by a region of complete absence of PCR products from D21S65 to D21S211.
Additionally, marker 47-237 could repeatedly not be amplified in DNA from blasts of patient 19, suggesting an intragenic homozygous deletion. The pattern of LOH as observed in patients 7, 11, 18 and 19 was consistent with LOH by MR (Figure 1a,b). In the remaining patients no LOH was observed in the RUNX1 region. PCR product quantification led to the detection of an imbalance of allelic products (AI) in patient 2, with a 1:2 ratio of multiple markers flanking the region of LOH (Figure 1a). The karyotype of patient 2 showed monosomy of chromosome 21 and two marker chromosomes. Combined binary ratio labeling (COBRA) analysis detected these as being derived from chromosome 21 (Table 1), explaining the LOH and AI patterns observed. AI in a 1:2 ratio was also detected in patient 12, throughout the entire q arm, in line with trisomy 21. The presence of 3 copies of RUNX1 was also confirmed by fluorescence in situ hybridization (FISH) experiments (Table 2 and below).
Table 2. FISH with various BAC and PAC probes
Patient no 2 12 14 18
BAC/PAC Exons of
RUNX1 *
Position
on map* # Number of signals per metaphase#
pZ21A cen 3
bC067E3 12,6 Mb 3
bC215A8 13,9 Mb 1
RP11-164H11 1
RP11-8P19 32,2 Mb 2
RP11-79A12 32,4 Mb 0
RP1-125H6 32,5 Mb 0
RP1-140K16 6 - 8 32,7 Mb 1 3 2 0
RP1-122C7 2 - 6 3 1
RP11-353B2 2 - 3 3
RP11-17O20 1 33,0 Mb 3 3 1 0
RP1-220P20 1 0
RP1-169K17 2 0
RP1-27A22 33,9 Mb 2 0
RP1-24J14 34,2 Mb 2 0
RP11-12N9 34,4 Mb 2
RP11-433E24 44,9 Mb 3 3 2 2
* data are from www.ncbi.nlm.nih.gov/genome/guide/human and BLAST searches. # see also Figure 2.
Figure 2. Localization and delimitation of deletions by FISH analysis. A) The open rectangles represent BAC/PAC probes and their relative positions to RUNX1. The thin lines below the map represent the extension of the heterozygous deletions in patients 2 and 14. The thick line represents the homozygous deletion in patient 18. The dotted lines represent the putative borders of the deletions.
The proximal deletion of patient 2 is not shown in this picture. In the upper part of the map, polymorphic markers used for LOH analyses (Figure 1b) are represented. Gap A represents a physical distance of approximately 200 Kb. B) Pseudocolored images of FISH co-hybridizations using probes delimiting proximal (panel a,c,e) and distal (panel b,d,f) breakpoint regions of deletions in patients 2 (a,b), 14 (c,d) and 18 (e,f) respectively (see also Table 2). On the upper part of each panel are the metaphase spreads. On the lower part of each panel an enlargement of chromosome 21 and derivatives. On the bottom the name of the probe written in the same color as the signal.
A
B
Homozygous and hemizygous deletions of RUNX1 were detected by fluorescence in situ hybridization (FISH)
The LOH breakpoints and homozygous deletions strongly indicate that RUNX1 functions as a tumor suppressor. To confirm and precisely map the deletions, metaphase chromosome spreads derived from patients 2, 14 and 18 were analyzed with a set of BAC/PAC probes by FISH (Table 2 and Figure 2). In patient 2, three hybridization signals were observed using probes for the chromosome 21 centromeric region and probes distal from RP1-122C7, identifying the chromosome 21 and both marker chromosomes. The remaining probes gave single signals, indicating the existence of a deletion of approximately 21-Mb starting within RUNX1 and ending at bC067E3 (Table 2, Figure 2b, panels a,b). FISH combined with LOH analysis (Figure 1b) indicated that several exons (6 to 8) of RUNX1 were deleted. A hemizygous deletion of approximately 1 Mb was detected in patient 14 (Table 2, Figure 2b panels c,d). The deletion started in RP1-140K16 between markers 56-85 and 57-76, included D21S211 and ended within RP1-169K17, indicating a deletion that includes the first 4 to 6 exons of RUNX1. In patient 18 the homozygous deletion, including the entire RUNX1 gene, was confirmed and stretched over 2 Mb from RP1-24J14 to RP11-79A12. Both regions flanking the deletion were present in two copies (Figure 2b, panels e,f) in line with MR as the mechanism of LOH.
Figure 3. Mutation analyses of RUNX1.
A) DNA sequence chromatograms of the RUNX1 mutations in patients 7, 11 and 12. Mutated DNA sequenced from blasts of the patients (top) is compared with normal DNA sequences of control T-cells of the same patients (bottom). Arrows indicate the mutations. In patient 12 the solid, black peak is the wild-type nucleotide G. The ratio between the wild-type nucleotide G and the mutant nucleotide A is approximately 1:2. B) In patient 7 the mutation G>T results in a Trp>Cys change at position 79.
The G>A substitution in patient 11 changed the conserved GT splice donor site of intron 4 to AT. Finally, in patient 12, the G>A mutation results in an Asp>Asn change in position 171. Amino acid position is given according to Miyoshi et al., 1995.36
Loss-of-function point mutations were detected in RUNX1 binding domain in three patients
To further evaluate the possibility that RUNX1 is a TSG, we sequenced both tumor and control cells DNA derived from all patients at exons 3 to 5 encoding the DNA binding domain.36 The DNA binding domain is the region where most mutations have been reported.22;23;25;26 Of the five patients showing LOH or AI (excluding patient 18 and 19 with deletions), we detected mutations in three (patients 7, 11 and 12 (Figure 3)). No mutations were found in patients without LOH. The G to T mutation in patient 7 resulted in a Trp>Cys change at position 79.
The G to A missense mutation in patient 12 resulted in an Asp>Asn change at position 171. In patient 11, the mutation changed the conserved GT splice donor site of intron 4 to AT. None of the three mutations were found in the control T-cells, which shown they where somatically acquired. In two patients (7 and 11), only the mutated allele was present, whereas in patient 12 the wild-type allele and the mutated allele were present in a ratio of 1:2 in accordance with the karyotype and LOH results.
Discussion
We have shown that RUNX1 is a frequent target for LOH, occurring in 35% (6/17) of patients with the undifferentiated form of AML. Seven out of eighteen patients, 39%, harbored hetero- or homozygous RUNX1 mutations. For AML-M0 the percentage is 1,7-fold higher than reported previously using single-stranded conformation polymorphism (SSCP) analyses.26 Of the seven RUNX1 mutations found, six have not been described before. In 24% of the AML-M0 cases studied, RUNX1 became inactivated by two sequential “hits” in accordance with Knudsons “two hit” hypothesis.29 By detecting the homozygous mutations in patients 7, 11, 18 and 19 and revealing the mechanisms of biallelic inactivation, we show that RUNX1 is genetically a TSG whose absence or biallelic mutation can lead to AML-M0. In patients 18 and 19 the first hit was a deletion, whereas in patients 7 and 11, the first hit was a point mutation. The extent of loss of heterozygosity combined with the presence of two apparently normal chromosomes 21 indicates that in at least three cases the first hit was duplicated by MR, resulting in a total loss of the wild-type allele (Figure 4). In patients 2 and 14, only one allele of RUNX1 was found mutated, as a result of deletions starting within RUNX1.
Although no point mutations were found in exons 3 to 5 of the other allele, the existence of mutations in the remaining RUNX1 sequence is still a possibility.
It has been shown that mutations truncating the DNA binding domain will result in loss-of- function of the protein.22 In patient 11, a G>A substitution changed the conserved GT splice donor site of intron 4 to AT. Exon skipping or use of a cryptic splice site as a result of this mutation will severely affect or truncate the RUNT domain. Missense mutations were also reported as causing loss-of-function of RUNX1.37 Position 79, changed in patient 7 from Trp>Cys, is part of a loop in the DNA binding domain responsible for the binding to the DNA major groove. Known mutations in the neighboring Arg80, Cys81 and Lys83 residues result in loss of binding function.37;38 Recently an identical mutation has been described in an AML- M0 patient.39 Residue 171, changed in patient 12 from Asp>Asn, interacts by hydrogen bonds with the DNA major groove and with Arg174 in order to fix the orientation of its guanidinium
group for interaction with DNA.38 A similar biallelic Asp>Gly change at position 171 found in an AML-M0 patient resulted in loss of DNA binding.26;37;38 An hereditary mono-allelic mutation in the same position, Asp>Tyr, resulted in FDP which subsequently evolved to AML in two members of a family with this mutation.24
This is the first study where a total homozygous deletion of RUNX1 is reported, linking total loss of RUNX1 to AML-M0. In RUNX1-/- mice a complete block in fetal liver hematopoiesis was observed underscoring the importance of RUNX1 in normal definitive hematopoiesis.40 Thus, AML-M0 cells may well correspond to differentiation-blocked precursor cells that underwent clonal expansion. As with other tumor suppressor genes, RUNX1 per se will probably not be the sole responsible gene for the development of leukemia and its progression must be accompanied by somatic mutations in other target genes leading to a growth advantage.41 However, RUNX1 is probably a very important factor in defining the type of malignancy. To date, in RUNX1 biallelic mutations and homozygous deletions are found only in AML-M0 patients.25;26 Furthermore, various translocations involving RUNX1, which result in dominant-negative oncoproteins that repress the transcription of RUNX1 target genes, are always associated with a defined leukemic phenotype.20 Amplification and over-expression of RUNX1 was found related to childhood acute lymphoblastic leukemia.42 The importance of the type of the RUNX1 mutation in defining the phenotype is well illustrated by a case of essential thrombocythemia reported by Preudhomme et al., 2000.26 Strikingly, this patient, who presented two mutated alleles and one wild type allele (trisomy), progressed to AML- M0 by loss of the wild-type allele (disomy).
Heterozygous germline mutations in RUNX1 result in haploinsufficiency in FPD patients with predisposition to AML. Some of the mutations found were common to sporadic AML, including AML-M0.41 In light of our results and the high percentage of MR-type of LOH, one would expect a high frequency of AML-M0 among the leukemias arising in FPD patients.
Only limited data are available but progression of FPD to AML-M0 has not been reported. In one of the studies, an FPD patient progressed from MDS to AML-M1. This progression was accompanied by a change in karyotype to trisomy 21.24 In the current study, patient 12 also shows trisomy and AML-M1. Five other non-AML-M0 leukemias with acquired trisomy and mutation in RUNX1 involving two of the three copies of the gene have been found.26 These data raise the possibility that starting from a heterozygous somatic or germ line mutation, AML may develop following the duplication of the chromosome carrying the mutated allele (Figure 4). Interestingly, in two other cases of FPD with leukemia, a wild type allele could still be found in the leukemic cells, but no information was provided about the karyotype.23 In transformed lymphoblastoid cells both MR and chromosome loss with concomitant duplication resulting in complete uniparental disomy seems to be the mechanisms contributing equally to LOH.31 This isodisomy may occur by two successive rounds of mitotic non- disjunction involving trisomy after the first non-disjunction event. The trisomy 21 observed in non AML-M0 patients might reflect this type of mechanism. We therefore speculate that the type of AML that will develop in FPD patients or individuals with precursor cells harboring a sporadic RUNX1 mutation will depend on the order of further mutations. In
AML-M0 inactivation of the second RUNX1 allele may occur before mutations resulting in a proliferative advantage. Cells in which growth is deregulated first may develop non-AML- M0 because of the increased chance of nondisjunction.
Figure 4. Inactivation of RUNX1 according to the Knudson two hit hypothesis 29, based on LOH, cytogenetic and sequencing data. Two chromosome 21 homologues including the normal RUNX1 ( ) gene are depicted at the top. The first “hit”
is a point mutation ( ) in patient 7, 11 and 12, an intragenic deletion in p.19 and a deletion covering the entire RUNX1 gene in p.18. Mitotic recombination appears to be the second “hit”, duplicating the first mutation in patient 11,18, 19 and most likely 7. In patient 14 and 2 only one
“hit”, inactivating RUNX1, was detected: a deletion in p.14 and a more complex chromosomal breakage and duplication in p.2. In patient 12 (AML-M1) duplication of the chromosome 21 with the mutated gene does not follow the Knudson hypothesis. One active RUNX1 allele is still present and one additional non-disjunction would be required for complete inactivation.
It is interesting to note that given the high number of extended LOH patterns including the telomere (22%) observed in this study, LOH analysis at telomeric regions appears an effective approach to identify chromosomes harboring critical TSG. More precise localization of the potential TSG usually requires identification of local deletions. Through use of purified populations of patient-derived cells, both hemi- and homozygous deletions could be detected in 4/18 patients in this study, indicating that LOH analysis may prove to be a valuable general method to localize TSG.
Materials and methods
Patients samples
Eighteen patients diagnosed at the Dept. of Hematology of the Leiden University Medical Center either with AML-M0 or AML-M1 (Table 1) were used in this study. Patients were selected based on their similarity with patient 19 described previously and on the availability of cryo-preserved pre-treatment cell samples.35 Classification was based on morphology, cytochemistry and immunophenotype, as determined by flow cytometry, following the French-American-British classification.43;44
Sorting by flow cytometry
Leukemic blasts and control T-cells were sorted by flow cytometry from cryo-preserved mononuclear cells isolated from bone marrow or peripheral blood at the time of diagnose. An average of 20×106 cells was labeled with CD3-FITC and CD13-, CD33- or CD34-PE (Becton
Dickinson) according to the blast-specific immunophenotype. Viable cells (viability being based upon the forward and side scatter profile) were analyzed and sorted with a Becton Dickinson FACS Vantage following standard procedures. Leukemic cells were sorted by gating the CD13+, CD33+ or CD34+ population (PE/FL2) and T-cells were obtained by gating the CD3+ population (FITC/FL1). Care was taken to avoid the inclusion of nuclei by using an appropriate threshold level.
Culturing
For diagnostic karyotyping, samples were cultured for 24, 48 or 72 hours. For further cytogenetic analysis, 3.3-10×106 cryo-preserved cells were grown for 72 hours in Amniogrow medium (Cytocell).
In some patients the fraction of CD3+ cells was too low (0,5-1%) to obtain sufficient control material. For these, T-cells were expanded in vitro from 1.6-5×106 cells as previously described.45 Culturing continued until the percentage of CD3+ cells was larger than 95%, as determined by flow cytometry.
LOH analysis
For DNA isolation >105 sorted cells were spun down for 10 minutes at 2000 rpm. The cell pellet was resuspended in DNA lysis buffer: 10 mM Tris-HCl (pH 8.3), 0.5% Tween-20, 1 mM EDTA, and 0,4 mg/ml proteinase K at a concentration of 100-1000 cells/μl. Cells that were sorted in lower yields were resuspended directly into lysis buffer at a minimal concentration of 50 cells/μl. Following incubation at 56oC overnight, the DNA lysates were incubated at 100oC for 10 min to inactivate proteinase K and stored at 4oC. An Invisorb TwinPrep DNA/
RNA Kit (Invitek) was used to isolate DNA from > 2×106 total cells or expanded T-cells.
LOH was assessed using polymorphic microsatellite markers (see web sites: http://www.gdb.
org/, http://www.ncbi.nlm.nih.gov and http://hgp.gsc.riken.gov.jp). Polymorphic markers within RUNX1 are described elsewhere.23 Polymerase chain reaction (PCR) conditions and LOH analysis was performed as described previously.32
Karyotyping
Metaphase preparations from the cultured cells were obtained according to standard cytogenetic techniques. Cytogenetic analysis was performed on GTG-banded chromosomes and the karyotype was assigned according to the International System for Human Cytogenetic Nomenclature (ISCN) criteria.46
COBRA
The 24-colour COBRA-FISH experiments were performed as previously described.47 Image analysis was performed using software implemented on a Power Macintosh 8100 designed to classify chromosomes on the basis of COBRA labelling.
Fluorescence in situ hybridization analyses
Chromosome preparations were co-hybridized in situ with Cy3 (NEN)- or FluorX (Amersham)-labeled probes prepared by nick translation.48 Briefly, 600 ng of labeled probe was used per FISH hybridization, which was performed at 37 °C in 2xSSC, 50% (v/v)
formamide, 10% (w/v) dextran sulphate, 5 μg COT1 DNA (BRL), and 3 μg sonicated salmon sperm DNA, in a volume of 10 μl. Post-hybridization washing was at 42 °C in 2xSSC- 50% formamide (x3) followed by three washes in 0.1xSSC at 60 °C. Chromosomes were identified by simultaneous DAPI (blue) staining which produces a Q-banding pattern. Digital images were captured using a Leica DMRXA epifluorescence microscope equipped with a cooled CCD camera (Princeton Instruments). Cy3, FluorX and DAPI fluorescence signals were recorded separately as gray scale images, using appropriate filters. Pseudocoloring was performed using Adobe Photoshop software.
As FISH probes the P. De Jong RP1 PACs and RP11 BACs used were chosen by querying their sequences at the RIKEN Center site (http://hgp.gsc.riken.go.jp/chr21/) and NCBI database (http://www.ncbi.nlm.nih.gov/). The precise position of the end sequences of the probes was established by BLASTN search against the complete sequence of chromosome 21. BACs bC215A8 and bC067E3 were obtained from CEPH, as reported in the Web site (http://www.biologia.uniba.it/rmc/2-YAC-BAC/YAC-BAC-Chromosome/BAC-21.html).
pZ21A is an alphoid probe specific for chromosomes 13 and 21 (http://www.biologia.uniba.
it/rmc/5-alfoidi/alf-13-21.html).
DNA sequencing
We amplified exons 3, 4 and 5 of RUNX1 from genomic DNA in two independent PCRs. The sequencing primers for exons 3, 4 and 5 of RUNX1 are as follows: 3F external: ATC CCA AGC TAG GAA GAC CGA C, 3R external: AGC TGC TTG CTG AAG ATC C, 3F internal:
GCC TGT CCT CCC ACC ACC CTC TC, 3R internal: TGT TTG CAG GGT CCT AAC TCA ATC, 4F external: GAC CGA GTT TCT AGG GAT TC, 4R external: CAT TGC TAT TCC TCT GCA ACC, 4F internal: GTG GGT TTG TTG CCA TGA AAC G, 4R internal: CAT CCC TGA TGT CTG CAT TTG TCC, 5F external: GGG GGA AAG GTT GAA CCC AA, 5R external: TTG TGC TGA AGG GCT GGA CA, 5F internal: CCA CCA ACC TCA TTC TGT TT and 5R internal: AGA CAT GGT CCC TGA GTA TA (where F stands for forward primer and R for reverse primer). For genomic amplification we used the following PCR conditions: denaturation at 95 oC for 5 min; 34 subsequent amplification cycles performed at 95 oC for 30 s, at 58 oC for 1 to 3 min and at 72 oC for 1 min: and a final step at 72 oC for 10 min. PCR products were purified with QIAquick PCR Purification Kit (Quiagen) and sequenced by capillary electrophoresis on an ABI Prism 3700 DNA Analyser (PE-Applied Biosystems), using the Big Dye terminator cycle sequencing kit (PE-Applied Biosystems).
Both forward and reverse strands were analyzed by BLAST (http://www.ncbi.nlm.nih.gov/
BLAST/). Mutations were confirmed by repeating PCR and sequencing reactions.
Acknowledgements
We thank H. Vrieling, P. Devilee and L.G. Fradkin for discussions and critical review of this manuscript; R.A.Th.W. Gouw and K.H.G. Kroeze-Jansema (Department of Human Genetics) for technical assistance, R. van der Linden and M. van der Keur (Department of Hematology) for cell sorting and G. de Groot-Swings (Department of Hematology) for her help with the collection of cells. This work was supported by the Inter University Institute for Radio Pathology and Radiation Protection (IRS) and by the grants from the Calouste
Gulbenkian Foundation and the Foundation for Science and Technology (Portugal) to F.P.G.
Silva.
Reference list
Tanaka T, Tanaka K, Ogawa S et al. An acute myeloid leukemia gene, AML1, regulates 1.
transcriptional activation and hemopoietic myeloid cell differentiation antagonistically by two alternative spliced forms. Leukemia 1997;11 Suppl 3:299-302.
Meyers S, Downing JR, Hiebert SW. Identification of AML-1 and the (8;21) translocation 2.
protein (AML-1/ETO) as sequence-specific DNA-binding proteins: the runt homology domain is required for DNA binding and protein-protein interactions. Mol.Cell Biol. 1993;13:6336-6345.
Kagoshima H, Akamatsu Y, Ito Y, Shigesada K. Functional dissection of the alpha and beta 3.
subunits of transcription factor PEBP2 and the redox susceptibility of its DNA binding activity.
J.Biol.Chem. 1996;271:33074-33082.
Ogawa E, Inuzuka M, Maruyama M et al. Molecular cloning and characterization of PEBP2 4.
beta, the heterodimeric partner of a novel Drosophila runt-related DNA binding protein PEBP2 alpha. Virology 1993;194:314-331.
Wang S, Wang Q, Crute BE et al. Cloning and characterization of subunits of the T-cell receptor 5.
and murine leukemia virus enhancer core-binding factor. Mol.Cell Biol. 1993;13:3324-3339.
Uchida H, Zhang J, Nimer SD. AML1A and AML1B can transactivate the human IL-3 promoter.
6.
J.Immunol. 1997;158:2251-2258.
Cockerill PN, Osborne CS, Bert AG, Grotto RJ. Regulation of GM-CSF gene transcription by 7.
core-binding factor. Cell Growth Differ. 1996;7:917-922.
Friedman AD, Britos-Bray M, Suzow J. The murine myeloperoxidase gene contains a bipartite 8.
distal enhancer, including a novel region regulated by PEBP2/CBF. Leuk.Res. 1996;20:809-815.
Redondo JM, Pfohl JL, Hernandez-Munain C et al. Indistinguishable nuclear factor binding to 9.
functional core sites of the T-cell receptor delta and murine leukemia virus enhancers. Mol.Cell Biol. 1992;12:4817-4823.
Wang SW, Speck NA. Purification of core-binding factor, a protein that binds the conserved core 10.
site in murine leukemia virus enhancers. Mol.Cell Biol. 1992;12:89-102.
Giese K, Kingsley C, Kirshner JR, Grosschedl R. Assembly and function of a TCR alpha 11.
enhancer complex is dependent on LEF-1-induced DNA bending and multiple protein-protein interactions. Genes Dev. 1995;9:995-1008.
Meyers S, Hiebert SW. Indirect and direct disruption of transcriptional regulation in cancer: E2F 12.
and AML-1. Crit Rev.Eukaryot.Gene Expr. 1995;5:365-383.
Lenny N, Westendorf JJ, Hiebert SW. Transcriptional regulation during myelopoiesis. Mol.Biol.
13.
Rep. 1997;24:157-168.
Petrovick MS, Hiebert SW, Friedman AD et al. Multiple functional domains of AML1: PU.1 and 14.
C/EBPalpha synergize with different regions of AML1. Mol.Cell Biol. 1998;18:3915-3925.
Mao S, Frank RC, Zhang J, Miyazaki Y, Nimer SD. Functional and physical interactions between 15.
AML1 proteins and an ETS protein, MEF: implications for the pathogenesis of t(8;21)-positive leukemias. Mol.Cell Biol. 1999;19:3635-3644.
Zhang DE, Hetherington CJ, Meyers S et al. CCAAT enhancer-binding protein (C/EBP) and 16.
AML1 (CBF alpha2) synergistically activate the macrophage colony-stimulating factor receptor promoter. Mol.Cell Biol. 1996;16:1231-1240.
Westendorf JJ, Yamamoto CM, Lenny N et al. The t(8;21) fusion product, AML-1-ETO, 17.
associates with C/EBP-alpha, inhibits C/EBP-alpha-dependent transcription, and blocks granulocytic differentiation. Mol.Cell Biol. 1998;18:322-333.
Sudol M. Yes-associated protein (YAP65) is a proline-rich phosphoprotein that binds to the SH3 18.
domain of the Yes proto-oncogene product. Oncogene 1994;9:2145-2152.
Kitabayashi I, Yokoyama A, Shimizu K, Ohki M. Interaction and functional cooperation of 19.
the leukemia-associated factors AML1 and p300 in myeloid cell differentiation. EMBO J.
1998;17:2994-3004.
Look AT. Oncogenic transcription factors in the human acute leukemias. Science 1997;278:1059- 20.
1064.
Hiebert SW, Lutterbach B, Amann J. Role of co-repressors in transcriptional repression mediated 21.
by the t(8;21), t(16;21), t(12;21), and inv(16) fusion proteins. Curr.Opin.Hematol. 2001;8:197- 200.
Michaud J, Wu F, Osato M et al. In vitro analyses of known and novel RUNX1/AML1 mutations 22.
in dominant familial platelet disorder with predisposition to acute myelogenous leukemia:
implications for mechanisms of pathogenesis. Blood 2002;99:1364-1372.
Song WJ, Sullivan MG, Legare RD et al. Haploinsufficiency of CBFA2 causes familial 23.
thrombocytopenia with propensity to develop acute myelogenous leukaemia. Nat.Genet.
1999;23:166-175.
Buijs A, Poddighe P, van Wijk R et al. A novel CBFA2 single-nucleotide mutation in familial 24.
platelet disorder with propensity to develop myeloid malignancies. Blood 2001;98:2856-2858.
Osato M, Asou N, Abdalla E et al. Biallelic and heterozygous point mutations in the runt domain 25.
of the AML1/PEBP2alphaB gene associated with myeloblastic leukemias. Blood 1999;93:1817- 1824.
Preudhomme C, Warot-Loze D, Roumier C et al. High incidence of biallelic point mutations 26.
in the Runt domain of the AML1/PEBP2 alpha B gene in Mo acute myeloid leukemia and in myeloid malignancies with acquired trisomy 21. Blood 2000;96:2862-2869.
Lund AH, van LM. RUNX: a trilogy of cancer genes. Cancer Cell 2002;1:213-215.
27.
Krug U, Ganser A, Koeffler HP. Tumor suppressor genes in normal and malignant 28.
hematopoiesis. Oncogene 2002;21:3475-3495.
Knudson AG. Antioncogenes and human cancer. Proc.Natl.Acad.Sci.U.S.A 1993;90:10914- 29.
10921.
Devilee P, Cleton-Jansen AM, Cornelisse CJ. Ever since Knudson. Trends Genet. 2001;17:569- 30.
573.
de Nooij-Van Dalen AG, van Buuren-van Seggelen VH, Lohman PH, Giphart-Gassler M.
31.
Chromosome loss with concomitant duplication and recombination both contribute most to loss of heterozygosity in vitro. Genes Chromosomes.Cancer 1998;21:30-38.
de Nooij-Van Dalen AG, Morolli B, van der Keur M. et al. Intrinsic genetic instability of normal 32.
human lymphocytes and its implication for loss of heterozygosity. Genes Chromosomes.Cancer 2001;30:323-335.
Holt D, Dreimanis M, Pfeiffer M et al. Interindividual variation in mitotic recombination.
33.
Am.J.Hum.Genet. 1999;65:1423-1427.
Vrieling H. Mitotic maneuvers in the light. Nat.Genet. 2001;28:101-102.
34.
Giphart-Gassler M, Nooij-Van Dalen A, Buuren-Van Seggelen V et al. A search for novel tumour 35.
suppressor genes for adult acute leukaemia by allelotyping at sub-telomeric chromosomal regions. Mutat.Res. 1998;400:409-419.
Miyoshi H, Ohira M, Shimizu K et al. Alternative splicing and genomic structure of the AML1 36.
gene involved in acute myeloid leukemia. Nucleic Acids Res. 1995;23:2762-2769.
Nagata T, Werner MH. Functional mutagenesis of AML1/RUNX1 and PEBP2 beta/CBF beta 37.
define distinct, non-overlapping sites for DNA recognition and heterodimerization by the Runt domain. J.Mol.Biol. 2001;308:191-203.
Tahirov TH, Inoue-Bungo T, Morii H et al. Structural analyses of DNA recognition by the 38.
AML1/Runx-1 Runt domain and its allosteric control by CBFbeta. Cell 2001;104:755-767.
Langabeer SE, Gale RE, Rollinson SJ, Morgan GJ, Linch DC. Mutations of the AML1 gene in 39.
acute myeloid leukemia of FAB types M0 and M7. Genes Chromosomes.Cancer 2002;34:24-32.
Wang Q, Stacy T, Binder M et al. Disruption of the Cbfa2 gene causes necrosis and 40.
hemorrhaging in the central nervous system and blocks definitive hematopoiesis. Proc.Natl.
Acad.Sci.U.S.A 1996;93:3444-3449.
Speck NA, Gilliland DG. Core-binding factors in haematopoiesis and leukaemia. Nat.Rev.
41.
Cancer 2002;2:502-513.
Mikhail FM, Serry KA, Hatem N et al. AML1 gene over-expression in childhood acute 42.
lymphoblastic leukemia. Leukemia 2002;16:658-668.
Bennett JM, Catovsky D, Daniel MT et al. Proposed revised criteria for the classification of acute 43.
myeloid leukemia. A report of the French-American-British Cooperative Group. Ann.Intern.Med.
1985;103:620-625.
Bennett JM, Catovsky D, Daniel MT et al. Proposal for the recognition of minimally 44.
differentiated acute myeloid leukaemia (AML-MO). Br.J.Haematol. 1991;78:325-329.
van Dam FJ, Natarajan AT, Tates AD. Use of a T-lymphocyte clonal assay for determining HPRT 45.
mutant frequencies in individual rats. Mutat.Res. 1992;271:231-242.
ISCN (1995). An International System for Human Cytogenetic Nomenclature. Basel: S. Karger;
46.
1995.
Tanke HJ, Wiegant J, van Gijlswijk RP et al. New strategy for multi-colour fluorescence in situ 47.
hybridisation: COBRA: COmbined Binary RAtio labelling. Eur.J.Hum.Genet. 1999;7:2-11.
Lichter P, Ledbetter SA, Ledbetter DH, Ward DC. Fluorescence in situ hybridization with Alu 48.
and L1 polymerase chain reaction probes for rapid characterization of human chromosomes in hybrid cell lines. Proc.Natl.Acad.Sci.U.S.A 1990;87:6634-6638.
