Pergsmon

01452126(95)00056-9

Leukemia Research Vol. 19, No. 10, PD. 741-748, 1995. Copyright 0 1995 &via Science Ltd

Printed in Great Britain. All rights reserved 0145-2126/95 $9.50 + 0.00

POSSIBLE CO-EXISTENCE OF RAS ACTIVATION AND MONOSOMY 7 IN THE LEUKAEMIC TRANSFORMATION OF MYELODYSPLASTIC

SYNDROMES

Janet Stephenson, He Lizhen and Ghulam J. Mufti Department of Haematological Medicine, Kings College School of Medicine and Dentistry, Bessemer Road,

London SE5 9PJ, U.K.

(Received 10 February 1995. Accepted ‘7 April 1995)

Abstract-The frequency of RAS activation was studied in 48 patients with acute myeloid leukaemia (AML) or with myelodysplastic syndromes (MDSI, in order to address the question of whether patients possessing monosomy 7 or other alterations of chromosome 7 have a higher incidence of RAS activation than those lacking chromosome 7 abnormalities. Samples were screened for oncogenic point mutation by DNA amplification followed by oligonucleotide hybridization analysis at codons 12, 13 and 61 of N-RAS and codons 12 and 13 of K-RAS. Two additional samples were considered to have activated RAS due to additional karyotypic abnormalities t(5;12) or loss of both copies of chromosome 17 and, hence, the neurofibro- matosis (NFI) loci. The group of chronic myelomonocytic leukaemia (CMML) patients had activated RAS in 4111 cases and inclusion of two CMMLt patients (with monosomy 7) brings this incidence to 5113. No change in frequency of RAS activation was seen between groups containing de nova AML samples with or without chromosome 7 abnormalities (l/5 and 2/12, respectively). However, assessment of MDS samples in the process of, or subsequent to, leukaemic progression showed a difference between the two groups. The frequency of RAS activation in samples with monosomy 7 was 4/9 samples while none of the seven samples without chromosome 7 changes showed RAS activation. The co-existence of RAS activation and monosomy 7 in MDS indicates that these lesions can co-operate in the multistep process of leukemogenesis.

Key words: RAS activation, monosomy 7, AML, MDS.

Introduction

RAS mutations in codons 12, 13, and 61, which alter the intrinsic GTPase activity of the molecule and produce constitutively activated RAS have been detected in approximately 25% of myelodysplastic syndromes (MDS), myeloproliferative syndromes (MPS) and acute myeloid leukaemias (AML)(reviewed in [l]). In addi- tion to oncogenic point mutations, two other mechan- isms by which RAS may be deregulated have recently been implicated in the development of myeloid malignancies. A specific translocation t(5;12)(q33;p13) which is most common in chronic myelomonocptic leukaemia (CMML), has recently been identified as producing a hybrid TEL-PDGF-jl receptor [2], assumed to be oncogenic due to constitutive receptor activation and therefore increased RAS activation [3]. This event appears equivalent to constitutive activation of the c-

Correspondence to: Janet Stephenson Ph.D., 3 Ullswater Road, Barnes, London SW13 9PL, U.K. (Tel: 081 563 7038).

FMS receptor, also known to occur in MDS and leukaemia. The neurofibromatosis gene NFl encodes a GTPase activating protein responsible for inactivating RAS [4:], so children with defects in NFl have deregulated RAS activation: such children are predis- posed to MDS, MPS, and AML. However, in adult myeloid malignancies, inactivating mutations of NFl are uncommon: one out of 28 MDS samples of unspecified FAI3 classification [5], and one of eight CMML samples included in a study of 27 AML patients and 57 MDS patients [6]. The most common method of RAS activation in adults is by oncogenic point mutation of the N-RAS and, to a lesser extent, the K-RAS genes.

The significance of such mutations is uncertain: RAS mutation may be an early or late change in MDS [7,8], and de nova AML [9], and while some studies have indicated an association of mutant RAS with poor prognosis and increased risk of leukaemia in MDS [lo, 111, others have observed no such association [12], or have noted the loss of RAS mutation at leukaemic

741

142 .I. Stephenson et al.

Table 1. RAS activation in samples grouped according to karyotype

Disorder Karyotype RAS

Strength of hybridization

signal

Group 1 974

1135 594 344

1023

Atypical CGL to AML MDS-AML

De nova AML-M2 De nova AMLM4

MDS-AML

1111 MDS-AML

894 De nova AML-M2 197 De ROVO AML-M4 908 De novo AML

1121 MDS-AML

1120 MDS-AML

1248” MDS-AML

1107 MDS-AML 841 CMMLt

953 CMMLt 1109 MDS-RA 858 MDS-RA

Group 2 522 De nova AML-MS

516 De nova AMLM2 648 De novo AMLMS 933 De novo AML 960 De novo AML-M4 969 De novo AML-Ml

1009 De nova AML-MS

1011 De ROVO AML-M3

1013 De novo AML

1025 De nova AMGM4 503 RAPBt

1113 De nova AML-MS

510 RAEBt

731 RAEB 883 RAEBt 897 RAEB

761 RAEBt 1005 MDS-AML

1015 MDS-AML

1193 MDS-AML +8 15~

de1 (7)(q31.2q32) 2Oc de1 (7)(q22q36) 15~

de1 (7)(qllq32) 13c 46XX 2c de1 (7)(q22q32) 14c 46XY lc

t(7;14)(q21;q32) llc t(7;14)(q21;q32) del(l)(p22p35) 8c 46xx lc

- 7 -4 t(4;7)(q12;q22) t(5:12)(q33;p13) 7c

-7 -4 t(4;7)(q12;q22) 3c 46XY 1Oc -714c

-7 de1(12)@llp12) 19~ -7 2oc

- 7 de1(5)(q13q23.2) - 17 - 17 der(l)@36) 4c 46XY 6c

inv(16)(p13q13) 13c inv(16)@13q13) -7, der(17)@13),

der(5), -3,-5,-9 2c -7 complex karyotype including loss

of 17p 15c -7 complex karyotype 8c 46XX lc

de1(7q)(q22q34) 12c 46XY 3c

monosomy 7 14c 46XY 5c monosomy 7 llc

dic(7;15)@ll;pll) 13c dic(7;15)@ll;pll) del(17) 6c

de1(16)(q22q24), +8 7c 46XY 8c

Random loss 4/15 cells de1(17)(q23.lq24.3) 1,lc 46XX 9c

Diploid 7 46XY 2Oc

del(18Xq22.2q23) 8c 46XY 12~

- 1 +8 derl 12~ 46XX 2c

t(15;17)(q22;qll) 13c 46XX 2c

de1(2)(p12.2p13) 15c

46XX 16c + 8 t20;21(qll;qll) 7c

Complex karyotype 17 c 46XX 6c

de1 12(p11.2~12) 4c 46XY 6c 46XX 15~ 46XY 15~

de1 12(p12.2~13) 8c 46XY 12c Diploid 7’

de1(2)(qllq13) 13c 46XY 17c

de1(2)(qllq13) llc 46XX 9c

RAS Assumed activated

[Presence of t(5;12)] Cysl2 +

RAS Assumed activated [Absence of NPl]

Asp13 t Ser12 + Ala13 +

Asp12 Asp13 Va113

+++tt ttt

t

Asp12 ttt Asp13 tt Asp13 tt

RAS activation and monosomy 7 743

Table 1 continued

Strength of hybridization

Disorder Karyotype RAS signal

Group 3 201 CMML 46XY Asp 12 ++++

Asp 13 + Val 13 +

445 CMML t(10;16)(p15;q13) 15c Asp 13 ++++ 453 CMML de1(12)(pllp12.2) 6c

46XY 19c 502 CMML del(ll)(p15.lp15.3) 4c 46XY llc 202 CMML 46XX 15~ 403 CMML 46XY 15~

1031 CMML 46XX 2Oc 903 CMML 46XX 2Oc 636 CMML 46XX 15~ Arg12 + 402 CMML Diploid 7’ 566 CMML 46XX 15~ Arg12 ++++

cys12 +

‘Diploid 7 indicates samples where karyotyping failed or was not possible. br these samples chromosome 7 status in the majority of cells was normal according to the signal produced when genomic DNA was blotted on to nylon, probed with chromosome 7 cD:A probes (D-actin 7pter-q22, MDR-1 7q21, ACHE 7q22) and control on chromosome 11 and analysed by densitometry.

Sample 1148 was not amplified for N-RAS codon 61 screening.

progression [9,13], or emergence of a different mutant at relapse [14,15]. One study has indicated that RAS point mutation correlates with increased survival in de nova AML patients [16].

RAS activation may be regarded as a post-initiation event affecting cell proliferation and differentiation [17], with a variable influence on leukaemic progression dependent upon the other changes within the cell. The identification of additional lesions which increase leukaemic progression in the presence of RAS mutation is therefore desirable.

p53 is one such candidate since it co-operates with RAS in oncogenic transformation in vitro [18], and in some solid tumours, such as colon carcinoma [19]. However, p53 mutation is rare in myeloid disorders [20-221 and such lesions have not been associated with RAS mutation [23].

Other candidates are the putative tumour suppressor genes on chromosome 7, marked by the frequent occurrence of monosomy 7 and deletion of 7q(22-q34) in myeloid disorders. There is at present no clear picture regarding the significance of monosomy 7 or del(7q) in association with RAS activation. The fact that children with defects in the NFl gene are at risk of developing myeloid malignancies, including a significant proportion with chromosome 7 abnormalities [24,25], carries the implication of a synergistic effect between RAS activation and monosomy 7.

Data on the association of monosomy 7 with RAS point mutation is variable and is largely confined to paediatric leukaemias and JCML, in which RAS

activation by point mutation appears to be a rare event. Certain studies on paediatric disease show a low incidence of RAS mutation in the presence of mono- somy 7 or de1 7q in JCML (O/4) [26], childhood monosomy 7 (3/20) [27], or childhood AML (O/5) [28]. Additional work [29] showed RAS mutation to be present in two out of six children with MDS and monosomy 7, and a possible association with disease progression. This study also observed RAS mutation to be rare in childhood AML (2/35 cases) with no association to the three cases of monosomy 7. The rarity of RAS point mutation in childhood AML and JCML may indicate that RAS activation in paediatric disease occurs by mechanisms other than point muta- tion.

Adult myelodysplastic syndromes and leukaemia have a higher incidence of RAS point mutation than childhood disorders, and although many studies in which RAS point mutations have been detected have used samples of unknown karyotype, sporadic cases of RAS activation in the presence of chromosome 7 abnormal- ities are known. One example is the observation of RAS point mutation in the presence of del(7q) [30] in adult MDS. Two further cases are included in this study in which rnonosomy 7 is associated with karyotypic abnormalities capable of resulting in RAS activation. The first of these is a patient with AML who shows a subclone containing the 5;12 translocation evolving in cells with both monosomy 7 and the translocation t(4;7)(q12;q22). The second is a case of AML in which one copy of chromosome 7 and both copies of

744 J. Stephenson et al.

1 2 3 4 5 6 7 6 9

10 11 12 13 14 15

1 2 3 4 5 6 7 6 9

10 11 12 13 14 15

GIY Asp 12

A B C

cys 12 Arg 12 Vall3 Ser 12

Asp 13 Ala 13

Fig. 1. Oligonucleotide hybridization results. Slot la = normal control; Slot 6a = sample 894; Slot 10a = sample 1120; Slot 15a = sample 841; Slot lb = sample 201; Slot 2b = sample 445; Slot 5b = sample 566; Slot 6b = sample 636; Slot 12b = normal

control; Slot llc = sample 1013; Slot 13~ = sample 1025.

chromosome 17 have been lost, thus losing both alleles of NFl, and inferred RAS activation.

This study has analysed the occurrence of RAS point

already known to have activation.

a high incidence of RAS

mutation in association with karyotypic abnormalities of chromosome 7 in adult de nova AML, in AML with a preceding preleukaemic phase, and MDS in transforma- tion to AML. Patients with CMML were regarded as a separate group since this subtype of MDS has char-

Materials and Methods

DNA was extracted from patient bone marrow, except for four cases of CMML when peripheral blood samples were used, using proteinase K digestion and phenol extraction according to standard methodologies. Two samples of normal

acteristics of a myeloproliferative syndrome and is peripheral blood were included as negative controls.

RAS activation and monosomy 7

Table 2. RAS activation in samples subdivided according to disease

745

RAS mutated

RAS activated No RAS activation

Total samples Frequency of RAS activation (%)

MDS-AML plus abnormal 7

1 CMMLt 1 MIX-AML 2 MDS-AML

1 CMMLt 4 MDS-AML

9

45

MDS-AML lacking De nova AML plus De nova AML lacking abnormal 7 abnormal 7 abnormal 7

0 1 2

0 0 0 2RAEB 4 9 4 RAEBt

3 MDS-AML 9 5 11

0 20 18

PCR gene amplification The following oligonucleotides were used for PCR: Nla SGACTGAGTACAAACTGGTGG Nlb SGG6CCTCACCrffATGGTG

Product length = 118 bp encompassing codons 12 and 13 of N-RAS.

N2a SGGTGAAACCITGTITGTTGGA N2b SATACACGAGGAAGCC’G

Product length = 103 bp encompassing codon 61 of N-RAS.

Kla SGGCCTGCTGAAAATGACI’GA Klb SGTCCTGCACCAGTAATATGC

Product length = 162 bp encompassing codons 12 and 13 of K-RAS.

One microgram of genomic DNA was used in PCR amplification [31], using the following conditions: 94°C 30 s, 55°C 30 s, 72°C 1 min for 30 cycles.

Oligonucleotide hybridization assay Twenty-five microlitres of each PCR reaction was applied to

Hybond N membrane (Amersham), according to the manufac- turer’s instructions, using a Schlier and Schuell slot blot apparatus. After removal from the slot blot manifold, the DNA was denatured, neutralized and fixed by W illumination according to standard procedures.

HPLC purified oligonucleotides for use as probes were obtained from Clontech. Probe sequences were as described elsewhere [32].

The probes were end-labelled with 32P-ATP according to the Clontech protocol, but purification of each probe was carried out using Nensorb-20 nucleic acid purification cartridges (Du Pont) according to the manufacturer’s instruc- tions. Pre-hybridization, hybridization and washing procedures were carried out by standard methodology (Clontech), and the washing temperature (61°C) to discriminate complete-match and single-mismatch probe binding was assessed empirically. The probes for point mutations which are most commonly used in MDS and AML were immediately used individually (Asp- N12, Val-N12, Arg-N12, Ser-N12, Cys-N12, Asp-N13, Arg- N13 Ala-N13, Asp-K12, Asp-K13, Val-K13), the others were used in pools of two to four probes per filter.

Chromosome 7 probe hybridization Genomic DNA (250 ng, 500 ng and 1 kg amounts) were

applied to Hybond N filters in a slot-blot manifold as described. Prehybridization, hybridization and non-stringent washing of filters was carried out at 65°C in cylinders using a rotary oven, with the exception of ACHE for which 70°C was used for washing. Prehybridization was carried out in

5 xSSPE, 5; xDenhardt’s solution and 0.5% SDS after addition of freshly denatured sonicated sheared salmon sperm DNA to a final concentration of 20 @ml. Hybridization was carried out overnight. The final stringent wash (0.1 x SSPE, 0.1 x SDS) was carried out in trays on a shaking platform. Plasmid preparation was carried out using Qiagen columns (Diagen) according to the manufacturer’s instructions. Insert-specific restriction fragments were generated, and purified by electrophoresis through 0.8% low melting temperature agarose (BRL). Probes were g;nerated without further purification by radiolabelling with P dCTP using a multiprime kit (Amersham). The following probes were used: ACHE: 2.0 kb Sac1 cDNA fragment [33]; MDR-I: 0.8 kb PvuII genomic fragment [34]; /& actin: 1.1 kb EcoRI cDNA fragment (ATCC 65128). Gamma globin control: this probe was generated by PCR using primers 5’aactgttgctttataggatttt and 5’ aggagcttattgataactcagac to pro- duce a 655 base pair fragment spanning the promotor and first exon of gamma globin G. The fragment was separated from PCR primers by two rounds of ethanol precipitation prior to labelling by random priming. Autoradiographs were analysed using WP gel scanning software to determine the intensity of the signal.

Results

Forty-eight patients were assessed for point mutation at codons 12, 13 and 61 of N-RAS, and at codons 12, and 13 of K-RAS by PCR amplification and oligonu- cleotide hybridization. Nine patients showed the pre- sence of N-RAS mutation (19%), with five of these being CMML or CMML in transformation to AML.

The results are shown in Table 1, in which cases are classified into three groups: those with monosomy 7, or chromosome 7 abnormalities, those lacking chromoso- ma1 abnormalities of 7, and cases of CMML which are known to be associated with a high incidence of RAS point mutation.

No mutations of the K-RAS gene at codons 12 and 13 were detected in this group of patients, although a separate study included patients with K-RAS mutation (unpublished results). The mutations which are most common in myeloid malignancy, the G to T transver- sions, replacing glycine with aspartate at codon 12 or 13 are both represented in our patients. The hybridization results are shown in Fig. 1.

146 J. Stephenson et al.

Five of the positive samples showed hybridization with more than one specific oligonucleotide. Sample 1120 gives weak signals with probes Aspl3, Ser12, and AIa13. Sample 841 shows a strong signal with probe Aspl2, a slightly weaker signal with Asp13 and a much weaker signal with probe Va113. Sample 1013 gives a strong signal with Asp12 and a weaker signal with Asp 13. Sample 566 gives a strong signal with Arg12 and a much weaker signal with CyslZ Sample 201 gives a strong signal with Asp12 and weak signals with Asp13 and Va113.

Discussion

Three groups of patients were analysed for the occurrence of R4S point mutations. The first of these were patients with myeloid disorders and monosomy 7 or other abnormalities involving chromosome 7. This group consisted of five cases of de nova AML, seven cases of AML after progression from MDS, two cases of CMML in progression to AML, one case of atypical (Philadelphia chromosome negative) CGL progressing to AML, and two cases of MDS-RA. The second group without chromosome 7 abnormalities consisted of 11 cases of de nova AML, three cases of AML subsequent to a myelodysplastic phase, four cases of RAEBt, and two cases of RAEB. The third group consisted of 11 cases of CMML.

The overall frequency of RAS point mutations in these samples was 9/48 (19%). The results confirm a high incidence of RAS point mutations in CMML patients since four of the 11 CMML samples in group 3 showed point mutations (36%). When the two CMMLt/ monosomy 7 samples are included as part of this group, the additional positive sample brings the incidence to 38%. In one case (636) a weak signal was obtained, but the remaining samples gave extremely strong signals, indicating that the mutations are to be found in the majority of cells, and are likely to have arisen as an early event. The positive signals detected in non-CMML samples were weak to moderate, indicating that the mutations are present in a subclone of cells. In the case of sample 1120, additional chromosomal changes, including monosomy 7, and three RAS point mutations, are apparent as late changes accompanying progression from MDS to acute leukaemia. The observation of signals with more than one probe has been seen in other studies [9, lo], and when investigated has been due to acquisition of differing point mutations by separate cells (rather than coexistence of two mutations within the same cell, or the same allele) [9]. In this study the observation of two different changes at the same codon is obviously not due to two changes within the same allele since the probes used would not fully match the mutated sequence. The relatively large number of

samples in this study with multiple RAS mutations may be due to observation of RAS as a late event occurring in samples that already possess karyotypic abnormalities. It is also possible that multiple probe binding is an artefact-while TMAC is used to eliminate differences in efficiency of probe binding due to base composition it may not do so completely, allowing additional probes to bind faintly to a mutated target sequence.

The incidence of RAS point mutations within group 1 as a whole is slightly increased compared to group 2: two patients in group 2 are positive (2/20), while three patients within group 1 showed RAS point mutations, one of whom was in progression from CMML (3/18). Inclusion of the two samples in which RAS is assumed to be activated brings this frequency to 5/18 (27%).

Group 1 patients can be further subdivided by karyotypic abnormality, since the abnormalities of chromosome 7 included in this study are heterogenous with regard to the molecular events likely to be responsible for leukaemia. There are three regions of chromosome 7 implicated: a region spanning 7p22-pter [24], and two regions on the long arm 7q: q32-q34 and 7q22-q31 [35]. The region at 7q21-22 is also observed to be interrupted by translocations in a subgroup of samples, as demonstrated by two samples in this study. Deletions and translocations of the long arm are more common in adult myelodysplastic syndromes and AML, and of the short arm in childhood leukaemia. There is some evidence that in familial monosomy 7, the predisposing locus is not on 7q [36]. Leukaemic patients with monosomy 7 may have changes in the retained chromosome at one or more of these three sites. It is possible that heterogeneity within group 1 may mask an association of RAS activation with a particular mole- cular lesion on chromosome 7, for instance on the short arm rather than within either critical deleted region on the long arm. The results here show no change in incidence of RAS activation when samples with loss or alteration of 7p are considered, compared to those with loss or alteration of 7q.

However, assessing patients who have progressed, or are in progression, from MDS or CMML to acute leukaemia within groups 1 and 2, produces a more convincing association (Table 2): of the nine patients in group 1, two have point mutations and two have RAS activated by other mechanisms, while of the seven MDS patients undergoing or subsequent to leukaemic trans- formation, none show RAS activation, nor do the two MDS patients at a earlier stage of leukaemic progres- sion. Fishers exact test (two-tailed) gave P = 0.08235 at a 5% level of significance, and therefore indicates that firm conclusions regarding an association between these two lesions and leukaemic progression will require a larger body of samples.

RAS activation and monosomy 7 747

The frequency of RAS activation when de nova AML patients are considered is similar in both groups: of the six de nova AML samples with chromosome 7 abnormalities, only one has a subclone of cells in which RAS is activated, and two of the 11 samples in group 2 show RAS activation.

The co-existence of RAS activation and monosomy 7 in MDS indicates that these lesions can co-operate in the multistep process of leukemogenesis. Characterization of the mechanism by which loss of gene function on chromosome 7 co-operates with RAS activation will await identification of the individual gene(s) responsi- ble.

References

1. Stephenson J., Mufti G. J. & Yoshida Y. (1993) Myelodysplastic syndromes: from morphology to molecu- lar biology. Part II. The molecular genetics of myelodysplasia. ht. J. Hematol. 57, 99.

2. Golub T. R., Barker G. F., Lovett M. & Gilliland D. G. (1994) Fusion of PDGF receptor beta to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t&12) chromosomal translocation. Cell 77, 307.

3. Sawyers C. L. & Denny C. T. (1994) Chronic myelomo- nocytic leukemia: Tel-a-kinase what Ets all about. Cell 77, 171.

4. Martin G. A., Viskochil D., Bollag G., McCabe P. C., Crosier W. J., Haubruck H., Conroy L., Clark R., O’Connell P., Cawthon R. M., et al. (1990) The GAP- related domain of the neurofibromatosis type 1 gene product interacts with ras ~21. Cell 63, 843.

5. Li Y., Bollag G., Clark R., Stevens J., Conroy L., Fults D., Ward K., Friedman E., Samowitz W., Robertson M., et al. (1992) Somatic mutations in the neurofibromatosis 1 gene in human tumors. Cell 69, 275.

6. Ludwig L., Janssen J. W. G., Schultz A. S. & Bartram C. R. (1993) Mutations within the FLR exon of NFl are rare in myelodysplastic syndromes and acute myelocytic leukaemia. Leukemia 7, 1058.

7. Yunis J. J., Boot A. J., Mayer M. G. & Bos J. L. (1989) Mechanisms of ras mutation in myelodysplastic syndrome. Oncogene 4, 609.

8. Lyons J., Janssen J. W., Bartram C., Layton M. & Mufti G. J. (1988) Mutation of Ki-ras and N-ras oncogenes in myelodysplastic syndromes. Blood 71, 1707.

9. Farr C. J., Saiki R. K., Erlich H. A., McCormick F. & Marshall C. J. (1988) Analysis of RAS gene mutations in acute myeloid leukemia by polymerase chain reaction and oligonucleotide probes. Proc. natn. Acad. Sci. USA 85, 1629.

10. Padua R. A., Carter G., Hughes D., Gow J., Farr C., Oscier D., McCormick F. & Jacobs A. (1988) RAS mutations in myelodysplasia detected by amplification, oligonucleotide hybridization, and transformation. Leukemia 2, 503.

11. Paquette R. L., Landaw E. M., Pierre R. V., Kahan J., Lubbert M., Lazcano O., Isaac G., McCormick F. & Koeffler H. P. (1993) N-ras mutations are associated with poor prognosis and increased risk of leukemia in myelodysplastic syndrome. Blood 82, 590.

12. Melani C., Haliassos A., Chomel J. C., Miglino M., Ferraris A. M., Gaetani G. F., Kaplan J. C. & Kitzis A.

(1990) Ras activation in myelodysplastic syndromes: clinical and molecular study of the chronic phase of the disease. Br. J. Haematol. 74, 408.

13. Nakagawa T., Saitoh S., Imoto S., Itoh M., Tsutsumi M., Hikiji K., Nakao Y. & Fujita T. (1991) Loss of multiple point mutations of RAS genes associated with acquisition of chromosomal abnormalities during dis&e progression in myelodysplastic syndrome. Br. J. Haematol. 77, 250.

14. Senn H. P., Jiricny J., Fopp M., Schmid L. & Moroni C. (1988) Relapse cell population differs from ,acute onset clone as shown by absence of the initially activated N-ras oncogene in a patient with acute myelomonocytic leukemia. Blood 72, 931.

15. Layton D. M., Mufii G. J., Lyons J., Janssen J. W. & Bartram C. R. (1988) Loss of ras oncogene mutation in a myelodysplastic syndrome after low-dose cytarabine therapy. N. Engl. J. Med. 318, 1468.

16. Neubauer A., Dodge R. K., Davey F. R., Silver R. T., Schiffer C. A., Mayer R. J., Ball E. D., Wurster-Hill D., Bloomfield C. D. & Liu E. (1994) Prognostic importance of mutations in the RAS proto-oncogenes in de novo acute myeloid leukaemia. Blood 83, 1603.

17. Maher J., Colonna F., Baker D., Luzzatto L. & Roberts I. (1994) Retroviral-mediated gene transfer of a mutant H-ras gene into normal human bone marrow alters myeloid cell proliferation and differentiation. Exp. Hematol. 22, 8.

18. Parada L. F., Land H., Weinberg R. A., Wolf D. & Rotter V. (1984) Cooperation between gene encoding ~53 tumour antigen and ras in cellular transformation. Nature 312,649.

19. Shaw P., Tardy S., Benito E., Obrador A. & Costa J. (1991) Occurrence of Ki-ras and ~53 mutations in primary colorectal tumors. Oncogene 6, 2121.

20. Jonveax P., Fenaux P., Quiquandon I., Pignon J., Lai J., Loucheux-Lefebvre M., Goosens M. & Berger R. (1992) Mutations in the ~53 gene in myelodysplastic syndromes. Oncogene 6, 2243.

21. Fenaux P., Jonveaux P., Quiquandon I., Lai J. L., Pignon J. M., Loucheux Lefebvre M. H., Bauters F., Berger R. & Kerckaert J. P. (1991) P53 gene mutations in acute myeloid leukemia with 17p monosomy. Blood 78, 1652.

22. Slingerland J. M., Minden M. D. & Benchimol S. (1991) Mutation of the ~53 gene in human acute myelogenous leukemia. Blood 77, 1500.

23. Buhler-Leclerc M., Gratwohl A. & Senn H.-P. (1994) Occurrence of point mutations in ~53 gene is not increased in patients with acute myeloid leukaemia carrying an activating N-ras mutation. Br. J. Haematol. 84, 443.

24. Kaneko Y., Maseki N., Sakurai M., Shibuya A., Shinohara T., Fujimoto T., Kanno H. & Nishikawa A. (1989) Chromosome pattern in juvenile chronic myelogenous leukemia, myelodysplastic syndrome, and acute leukemia associated with neurofibromatosis. Leukemia 3, 36.

25. Shannon K. M., O’Connell P., Martin G. A., Paderanga D., Olson K., Dinndorf P. & McCormick F. (1994) Loss of the normal NFl allele from the bone marrow of children with type 1 neurofibromatosis and malignant myeloid disorders. N. Engl. J. Med. 330, 597.

26. Miyauchi J., Asada M., Sasaki M., Tsunematsu Y., Kojima S. & Mizutani S. (1994) Mutations of the N-ras gene in juvenile chronic myelogenous leukemia. Blood 83, 2248.

27. Neubauer A., Greenberg P., Negrin R., Ginzton N. & Liu E. (1994) Mutations in the ras proto-oncogenes in patients with myelodysplastic syndromes. Leukemia 8, 638.

28. Farr C., Gill R., Katz F., Gibbons B. & Marshall C. J. (1991) Analysis of RAS gene mutations in childhood myeloid leukaemia. Br. J. Haematol. 77, 323.

748 J. Stephenson er al.

29. Lubbert M., Mirro J. J., Kitchingman G., McCormick F., Mertelsmann R., Herrmann F. & Koeffler H. P. (1992) Prevalence of N-ras mutations in children with myelodys- plastic syndromes and acute myeloid leukemia. Oncogene 7, 263.

30. Mansoor A. M., Bharadwaj T. P., Sethuraman S., Chandy M., Pushpa V., Kamada N. & Murthy P. B. (1993) Analysis of karyotype, SCE, and point mutation of RAS oncogene in Indian MDS patients. Cancer Genet. Cytogenet. 65, 12.

31. Saiki R. K., Gelfand D. H., Stoffel S., Scharf S. J., Higuchi R., Horn G. T., Mullis K. B. & Erlich H. A. (1988) Primer- directed enzymatic amplification of DNA with a thermo- stable DNA polymerase. Science 239, 487.

32. Padua R. A., Carter G., Hughes D., Gow J., Farr C., Oscier D., McCormick F. & Jacobs A. (1988) RAS mutations in myelodysplasia detected by amplification, oligonucleotide hybridization, and transformation. Leukemia 2, 503.

33. Soreq H., Ben Aziz R., Prody C. A., Seidman S., Gnatt A.,

Neville L., Lieman Hurwitz J., Lev Lehman E., Ginzberg D., Lipidot Lifson Y., et al. (1990) Molecular cloning and construction of the coding region for human acetylcholi- nesterase reveals a G + C-rich attenuating structure. Proc. natn. Acad. Sci. USA 87, 9488.

34. Roninson I. B., Chin J. E., Choi K. G., Gros P., Housman D. E., Fojo A., Shen D. W., Gottesman M. M. & Pastan I. (1986) Isolation of human mdr DNA sequences amplified in multidrug-resistant KB carcinoma cells. Proc. natn. Acad. Sci. USA 83,4538.

35. Ix Beau M. M. & Rowley J. D. (1986) Chromosomal abnormalities in leukemia and lymphoma: clinical and biological significance. Adv. Hum. Genet. 15, 1.

36. Shannon K. M., Turban A. G., Chang S. S., Bowcock A. M., Rogers P. C., Carroll W. L., Cowan M. J., Glader B. E., Eaves C. J., Eaves A. C., et al. (1989) Familial bone marrow monosomy 7. Evidence that the predisposing locus is not on the long arm of chromosome 7. J. Clin. Invest. 84, 984.