And Microsatellite Mutations in Radiation-Induced Acute Myeloid Leukaemia in the CBA/H Mouse J Fennelly, E Wright and M Plumb

Home , Mutation rate

Mini- and microsatellite mutations in radiation-induced acute myeloid leukaemia in the CBA/H mouse J Fennelly, E Wright and M Plumb

Radiation and Genetic Stability Unit, Harwell, Oxon OX11 0RD, UK

Radiation-induced acute myeloid leukaemia (AML) in the CBA/H sporadic cancers suggesting it may play a general role in mouse is a clonal disorder and therefore amenable to the analy- tumour evolution.14–16 sis of genetic instability during radiation leukaemogenesis. The genotype of a single minisatellite and 20 microsatellite loci was X-irradiation (3 Gy) of CBA/H mouse results in the induction compared in tail and leukaemic spleen DNA prepared from the of acute myeloid leukaemia (AML) with an incidence of 25% same mouse. Somatic mutation at the Ms6-hm minisatellite and a mean latency of 18 months suggestive of multi-stage locus was nearly seven times higher (27%, 4/15) than the spon- leukaemogenesis.17,18 As this haemopoietic neoplasia is taneous germline mutation rate (4%). Only 1/15 AMLs exhibited clonal in origin,19 mutations which occurred early in radiation microsatellite mutations, but 5/20 loci were mutated in the same leukaemogenesis can be analysed. We have therefore AML, indicating that it was deﬁcient in mismatch repair. Thus, whereas somatic minisatellite mutations, which are associated screened radiation-induced CBA/H AMLs for minisatellite and with complex intra-allelic gene conversion events, occur at a microsatellite mutations. Whereas somatic mutations at a sin- very high rate in the radiation-induced AMLs, microsatellite gle minisatellite locus occur at a high frequency (27%), only instability, which has been associated with the acquisition of one AML was detected which had a mismatch repair defect. the replication error repair (RER؉) phenotype, is infrequent but detectable. Keywords: minisatellite instability; microsatellite instability; radiation-induced AML Material and methods

Radiation-induced AMLs Introduction Male CBA/H mice, 8–12-week-old, were 3 Gy X-irradiated Radiation-induced genetic instability may contribute to neo- and leukaemias diagnosed as described.17,18,20 The animal plastic transformation. The ability to detect the effects of ionis- studies here described were carried out under guidance issued ing radiation on the mutation rate of a single genetic locus by the MRC in ‘Responsibility in the use of animals for medi- depends on the background spontaneous mutation rate, the cal research’ (July 1993) and Home Office project licence No. magnitude of the effect of the radiation on the spontaneous PPL 30/00689. DNA was prepared from the leukaemic spleen mutation rate, and the ability to score a statistically significant and tail from the same mouse.20 number of clonal descendants of an irradiated cell. Hypervariable tandem repeat minisatellite loci have spontaneous germline and somatic mutation rates of up to 10%,1–5 and therefore potentially enable the detection of Southern blot analyses induced mutation rates in small population samples. Thus, there is evidence that paternal irradiation increases the germ- DNA (10 ␮g) was digested with HinfI and resolved by 1% aga- line mutation rates of minisatellite loci in a dose-dependent rose gel electrophoresis, alkali denatured and transferred to fashion in their clonal offspring.6–9 Somatic mutations at spe- Genescreen (DuPont, Boston, MA, USA). The Ms6-hm probe cific human minisatellite loci have been associated with type was a (GGGCA)Ͼ6 synthetic oligonucleotide repeat labelled I diabetes mellitus (INS) and multiple forms of cancer by random priming (GIBCO BRL, Paisley, UK; oligolabelling (HRAS1).5 The induction of somatic minisatellite mutations in kit) in the presence of 50 ␮Ci of ␣-32P-dATP (Amersham, UK; rodent model systems has been reported in chemically 3000 Ci/mmol), and hybridised to Southern blots as induced tumours in the mouse10–12 and in X-irradiated described.2 Blots were washed to a stringency of 0.1 × SSC at C3H/10T1/2 cells where the mutation rate approached 100% 65°C. Molecular weight markers were the 1 kbp ladder from if the cells were passaged in vivo.13 Sequence analysis of a GIBCO/BRL. number of minisatellite loci and their germline mutant variants revealed the polar gain or loss of repeat units at one end of the tandem repeat array, and thus implicated complex inter- Polymerase chain reaction of microsatellite loci and intra-allelic gene conversion events.3,4 Microsatellite mutations are associated with mismatch Genomic DNA (10–50 ng) was PCR amplified in a final vol- repair defects and have been described as the replication error + ume of 15 ␮l of a solution containing 200 ␮M of dGTP, dTTP repair (RER ) mutator phenotype in cancer.14,15 Although first and dCTP, 20 ␮M dATP, 1 ␮Ci of ␣-32P-dATP (3000 Ci/mmol; described in hereditary nonpolyposis colorectal cancer Amersham), 60 mM Tris-HCl (pH 8.0), 15 mM ammonium sul- (HNPCC), microsatellite instability has also been detected in phate, 15 mM MgCl and 1 unit of Taq polymerase a number of other cancer prone families and in a variety of 2 (Boehringer Mannheim, UK). After a 1.5 min pre-incubation at 95°C, DNA was amplified for 30 cycles: 95°C for 45 s; 55°C for 45 s; 72°C for 45 s. PCR products were resolved by denat- Correspondence: M Plumb uring 6% polyacrylamide, 50% urea gel electrophoresis and Received 16 December 1996; accepted 27 February 1997 autoradiography. Mini- and microsatellite mutations in radiation-induced CBA/H AML J Fennelly et al 808 Results microsatellite germline mutation rate in normal wild-type CBA/H mice is less than 0.003% (Figure 2, and data not Ms6-hm germline and somatic mutation rates shown). Comparison of 15 control and leukaemic DNA samples at 19 dinucleotide microsatellite repeat loci and one The murine hypervariable Ms6-hm minisatellite locus maps trinucleotide repeat locus, revealed five microsatellite near the brown coat color gene (b) on chromosome 4.2 The mutations in one AML DNA (case 8, Figure 2, and data not spontaneous germline mutation rate of Ms6-hm was estimated shown), and we conclude that it had acquired the RER+ to be 0.024–0.032 per gamete in BXD RI, BXH RI and phenotype during multi-stage radiation leukaemogenesis. All (C57BL/6J X DBA/2J)F1 and F2 crosses.2 To obtain an indi- five microsatellite mutations appeared to be caused by the cation of the instability of Ms6-hm in the CBA/H colony, its deletion of a single dinucleotide repeat unit (Figure 2, and spontaneous germline mutation rate in the CBA/H colony was data not shown), but, given the small numbers analysed, it is first estimated. unclear whether this apparent bias towards the delection of The germline mutation rate (␮) of a neutral allele in an single repeat units in the AMLs is significant. inbred strain is related to the heterozygosity (H) of the locus by the following equation: H = 4N ␮/(1 + 4N ␮) e e Discussion

where Ne is the effective population size which, in an inbred colony derived from repeated brother–sister matings, has been The somatic mutation rate at a single chromosome 4 mini- estimated to be 2.6.2,21 Heterozygosity, as deﬁned by the satellite locus in AMLs which arose in vivo after 3 Gy whole hybridization of at least two HinfI genomic restriction frag- body X-irradiation of CBA/H mice is almost seven-fold higher ments to the Ms6-hm probe in Southern blot analyses (0.27) than the spontaneous germline mutation rate (0.04). (Figure 1, lanes 8 and 12, and data not shown), was detected The mutations were detected by Southern blot analysis of gen- in 25/84 control un-irradiated CBA/H mice (H = 0.3) giving omic DNA suggesting that the majority of the leukaemic cells ␮ = 0.04, a value which is consistent with the germline contained the mutation, and, as the leukaemias are clonal in mutation rates estimated directly and using heterozygosity origin, that the mutations occurred early in radiation leukae- analyses in other inbred mouse strains.2 mogenesis prior to, or shortly after the clonal expansion of the Fifteen CBA/H radiation-induced AMLs were analysed for leukaemic cell. In contrast, microsatellite instability and the + Ms6-hm mutations by Southern blot analysis of DNA prepared RER phenotype is infrequently detected in the same AMLs. from tail (control) and leukaemic spleen from individual mice. The spontaneous somatic mutation rate of the Ms6-hm As shown in Figure 1, gain or loss of genomic restriction frag- locus (0.028) is very similar to its germline mutation rate ments was observed in the AMLs when compared to the con- (0.025) in C57BL/6, DBA/2 or their F1 offspring, and although trols (cases 1, 3 and 6). In total, somatic mutations at the Ms6- a much higher (0.20) somatic mutation rate was detected at hm locus were detected in 4/15 AMLs giving a somatic the Hm-2 minisatellite locus, the mutations occurred early in mutation frequency of 0.27, almost seven-fold higher than the development as the mosaicism was indistinguishable in sev- spontaneous germline mutation rate.

Microsatellite mutations in radiation-induced AMLs

Analysis of 20 microsatellite loci in 15 CBA/H mice revealed complete homozygosity at the loci, indicating that the

Figure 1 Somatic mutations at the Ms6-hm locus. Genomic DNA Figure 2 Microsatellite instability in an AML. DNA prepared from (10 ␮g) prepared from tail (T, lanes 2, 4, 6, 8, 10 and 12) or leukaemic tail (T) or leukaemic spleen from three mice (cases 7, 8 and 9) was spleen (L, lanes 1, 3, 5, 7, 9 and 11) from the same mouse in six PCR ampliﬁed using primers that detect the following chromosome-

independent AMLs (cases 1–6) was digested with HinfI, Southern blot- specific microsatellite loci: (A) D5Mit145, (CA)34 repeat; (B) ted and probed with the Ms6-hm probe. Novel genomic restriction D7Mit291, (TG)13(GA)23 repeats; (C) D2Mit52, (CA)14 repeat; and (D) fragments which are present in AMLs but not the control DNA from D2Mit436, (CA)20 repeat. The radiolabelled products were resolved the same mouse (cases 1, 3 and 6) are shown with arrows. Molecular by denaturing polyacrylamide gel electrophoresis and autoradio- weight markers (kbp) are shown on the left of the panel. graphy. Mini- and microsatellite mutations in radiation-induced CBA/H AML J Fennelly et al 809 eral adult tissues.1,2 There is the formal possibility that the References Ms6-hm mutations in the AMLs represent the clonal expansion of a bone marrow cell which had undergone a mutation 1 Gibbs M, Collick A, Kelly RG, Jeffreys AJ. A tetranucleotide repeat earlier in development which was unrelated to radiation leu- mouse minisatellite displaying substantial somatic instability dur- kaemogenesis. However, the clonal expansion of one cell ing early preimplantation development. Genomics 1993; 17: carrying a mutant genotype would result in the loss of a gen- 121–128. omic restriction fragment when compared to DNA from a 2 Kelly R, Bulfield G, Collick A, Gibbs M, Jeffreys AJ. Characteriz- complex tissue in the mosaic adult, unless the somatic ation of a highly unstable mouse minisatellite locus: evidence for somatic mutation during early development. Genomics 1989; 5: mutation occurred late in development. The Ms6-hm 844–856. mutations in the radiation-induced AMLs are defined by the 3 Buard J, Vergnaud G. Complex recombination events at the hyper- gain of Ms6-hm-related genomic restriction fragments mutable minisatellite CEB1 (D2S90). EMBO J 1994; 13: 3203– (Figure 1, lanes 1, 5 and 11) which were undetectable in the 3210. putative mosaic adult tissue (tail), and therefore either 4 Jeffreys AJ, Tamaki K, MacLeod A, Monckton DG, Neil DL, occurred late in development or during radiation leukaemog- Armour JAL. Complex gene conversion events in germline mutation at human minisatellites. Nat Genet 1994; 6: 136–145. enesis. 5 Krontiris TG. Minisatellites and human disease. Science 1995; The effects of ionising radiation on minisatellite germ- 269: 1682–1683. line mutation rates in the mouse and human is well docu- 6 Sadamoto S, Suzuki S, Kamiya K, Kominami R, Dohi K, Niwa O. mented,6–9,13 and high levels of somatic minisatellite mutations Radiation induction of germline mutation at a hypervariable mini- (50–100%) have also been detected in C3H/10T1/2 cells satellite locus. Int J Radiat Biol 1994; 65: 549–557. which were transformed by X-rays and passaged in vivo,13 in 7 Fan YJ, Wang Z, Sadamoto S, Ninomiya Y, Kotomura N, Kamiya 12 10 K, Dohi K, Kominami R, Niwa O. Dose-response of a radiation chemically induced mouse sarcomas and liver tumours, induction of a germline mutation at a hypervariable mouse minisa- and in chemically mutagenised Chinese hamster lung cells,11 tellite locus. Int J Radiat Biol 1995; 68: 177–183. data which support our conclusion that the somatic mutations 8 Dubrova YE, Nesterov VN, Krouchinsky NG, Ostapenko VA, Neu- in the AMLs arose during radiation leukaemogenesis. Given mann R, Neil DL, Jeffreys AJ. Human minisatellite mutation rate the high incidence of mutations at a single genetic locus in after the Chernobyl accident. Nature 1996; 380: 683–686. the AMLs, it is unlikely that the mutations were caused directly 9 Dubrova YE, Jeffreys AJ, Malashenko AM. Mouse minisatellite mutations induced by ionizing radiation. Nat Genet 1993; 5: by radiation-induced DNA damage, and implies that the ionis- 92–94. ing radiation had a more general indirect effect on an 10 Ledwith BJ, Joslyn DJ, Troilo P, Leander KR, Clair JH, Manam S, inherently error prone genomic DNA processing mechanism. Prahalada S, van Zwieten MJ, Nichols WW. Induction of mini- Minisatellite germline mutations are polar and have been satellite DNA rearrangements by genotoxic carcinogens in mouse associated with complex inter- and intra-allelic gene conver- liver tumours. Carcinogenesis 1995; 16: 1167–1172. sion events.3,4 11 Kikuno T, Honma M, Ogura S, Mizusawa H, Hayashi M, Sofuni T. DNA fingerprint analysis in chemically mutagenized Chinese In contrast, microsatellite instability which is associated + hamster lung cells. Mutation Res 1995; 338: 87–93. with the RER mutator phenotype caused by mismatch repair 12 Niwa O, Kamiya K, Furihata C, Nitta Y, Wang Z, Fan Y-J, Nino- defects14,15 is an infrequent event in radiation-induced AMLs miya Y, Kotomura N, Numoto M, Kominami R. Association of in the mouse. Microsatellite instability in spontaneous human minisatellite instability with c-myc amplification and K-ras haemopoietic malignancies has been extensively studied, with mutation in methylcholanthrene-induced mouse sarcomas. Can- very few cases detected in AML, NHL or ALL.22,23 However, cer Res 1995; 55: 5670–5676. 13 Paquette B, Little JB. In vivo enhancement of genomic instability reports of microsatellite instability in chronic myeloid in minisatellite sequences of mouse C3H/10T1/2 cells. Cancer Res 24 leukaemia range from 52.6% in Japan (10/19 of cases )to 1994; 54: 3173–3178. undetectable in Britain (0/39 of cases25). As microsatellite 14 Eshleman JR, Markowitz SD. Mismatch repair defects in human instability has also been detected in Japanese patients with carcinogenesis. Hum Mol Genet 1996; 5: 1489–1494. myelodysplastic syndrome (3/19 cases26), there may be a 15 Loeb LA. Microsatellite instability: marker of a mutator phenotype genetic component within the human population which in cancer. Cancer Res 1994; 54: 5059–5063. 16 de Wind N, Dekker M, Berns A, Radman M, te Riele H. Inacti- determines the incidence of microsatellite instability in vation of the mouse Msh2 gene results in mismatch repair haematological malignancies. deficiency, methylation tolerance, hyperrecombination and pre- Genetic lesions in radiation-induced AMLs in the CBA/H disposition to cancer. Cell 1995; 82: 321–330. mouse, as indicators of genetic instability which either arose 17 Major IR, Mole RH. Myeloid leukaemia in X-ray irradiated CBA as the direct consequence of the exposure to ionising radi- mice. Nature 1978; 272: 455–456. ation, or during multistage leukaemogenesis, including the 18 Mole RH, Papworth DG, Corp MJ. The dose-response for X-ray 27 induction of myeloid leukaemia in male CBA/H mice. Br J Cancer delection of chromosome 2 sequences, the secondary loss 1983; 47: 285–291. of the Y chromosome,20 somatic minisatellite mutations and 19 Bessho M, Hirashima K. Single cell origin of radiation-induced the infrequent but detectable acquisition of the RER+ pheno- myeloid leukemia in mice with cellular mosaicism. Exp Hematol type. It would be of interest to re-examine therapy-related leu- 1987; 15: 589. kaemias in man for similar indicators of genetic instability in 20 Fennelly J, Crabtree G, MacDonald D, Lorimore S, Laval S, Proffitt radiation leukaemogenesis. J, Boyd Y, Wright E, Plumb M. Complex Y chromosome aber- rations are a recurrent secondary event in radiation-induced murine acute myeloid leukaemia. Leukemia 1995; 9: 506–512. 21 Kimura M, Crow JF. The number of alleles that can be maintained in a finite population. Genetics 1964; 49: 725–738. Acknowledgements 22 Robeldo M, Martinez B, Arranz E, Trujillo MJ, Ageitos AG, Rivas C, Benitez J. Genetic instability of microsatellites in hematological We thank Dr Y Dubrova and Prof A Jeffreys for the Ms6-hm neoplasms. Leukemia 1995; 9: 960–964. 23 Sill H, Goldman JM, Cross NCP. Rarity of microsatellite alterations probe. This work was supported by the Leukaemia Research in acute myeloid leukaemia. Br J Cancer 1996; 74: 255–257. Fund and the Medical Research Council. 24 Wada C, Shionoya S, Fujino Y, Tokuhiro H, Akahoshi T, Uchida T, Ohtani H. Genomic instability of microsatellite repeats and its Mini- and microsatellite mutations in radiation-induced CBA/H AML J Fennelly et al 810 association with the evolution of chronic myelogenous leukemia. bility is an early genetic event in myelodysplastic syndrome but Blood 1994; 83: 3449–3456. is infrequent and not associated with TGF-␤ receptor type II gene 25 Silly H, Chase A, Mills KI, Apfelbeck U, Sormann S, Goldman JM, mutation. Leukemia 1996; 10: 1696–1699. Cross NCP. No evidence for microsatellite instability or consistent 27 Alexander BJ, Rasko JEJ, Morahan G, Cook WD. Gene deletion loss of heterozygosity at selected loci in chronic myeloid leu- explains both in vivo and in vitro generated chromosome 2 aber- kaemia blast crisis. Leukemia 1994; 8: 1923–1928. rations associated with murine myeloid leukaemia. Leukemia 26 Kaneko H, Horiike S, Taniwaki M, Misawa S. Microsatellite insta- 1995; 9: 2009–2015.