Selective Induction of Leukemia-Associated Fusion Genes by High-Dose Ionizing Radiation1

[CANCER RESEARCH 58. 421-425. February I. 1

Michael W. N. Deininger, Shikha Bose, Joanna Gora-Tybor, Xiu-Hua Yan, John M. Goldman, and Junia V. Melo2

Leukaemia Research Fumi Centre for Adult Leukaemia. Department of Haemah>li>/;\: Royal Postgraduale Medical School, Ducane Road, London W12 ONN, United Kingdom

ABSTRACT event involves the acquisition of the genetic abnormality whose "success" in the production of a leukemic phenotype will depend on There is strong clinical and epidemiolÃ³gica! evidence that ionizing its capacity to impart to the target cell a proliferative and/or survival radiation can cause leukemia by inducing DNA damage. This crucial advantage over its normal neighbors. In molecular terms, the gener initiation event is believed to be the result of random DNA breakage and misrepair, whereas the subsequent steps, promotion and progression, ation of a potentially successful reciprocal chromosomal translocation must rely on mechanisms of selective pressure to provide the expanding requires that: (a) at least two independent DNA DSBs occur, one in leukemic population with its proliferative/renewal advantage. To investi each chromosome partner; (b) the two breaks occur simultaneously, gate the susceptibility of human cells to external agents at the genetic i.e., within the same cell cycle, so that the two ends of one broken recombination stage of leukemogenesis, we subjected two hematopoietic chromosome are available to interact and be ligated (misrepaired) to cell lines, KG1 and III.6(1, to high doses of y-irradiation. The irradiation the respective complementary broken ends of the other chromosome; induced the formation of fusion genes characteristic of leukemia in both and (c) the recombination observes the polarity of the DNA molecule. cell lines, but at a much higher frequency in M.I than in 111,6Â».In KG1 Various physical and chemical mutagens are known to induce DNA cells, the AML1-ETO hybrid gene [associated with the t(8;21) transloca DSBs that may ultimately be misrepaired in the form of reciprocal tion of acute myeloid leukemia] occurred significantly more often than the BCR-ABL [associated with t(9;22) chronic myeloid leukemia) or the chromosomal translocations. IR is the most extensively studied and DEK-CAN [associated with t(6;9) acute myeloid leukemia] fusion genes. has been shown to produce DNA DSBs directly, in contrast to UV These findings support the notion that ionizing radiation can directly radiation and most chemical mutagens that induce DSBs only sec generate leukemia-specific fusion genes but emphasize the differing sus ondarily in the course of DNA replication. ceptibility of different cell populations and the differing frequency with There is no doubt that IR can cause leukemia. Most of the evidence which the various fusion genes are formed. The selectivity observed at the for this association in humans is provided by studies of people who primary level of gene fusion formation may explain at least in part the received irradiation as a consequence of medical (diagnostic X-ray differential risk for development of some but not other forms of leukemia and radiotherapy), occupational (radiology and nuclear industry) or after high-dose radiation exposure. accidental (nuclear explosion) exposure (3). However, little is known of the molecular mechanism(s) by which IR generates leukemia- INTRODUCTION specific fusion genes or of the reasons why some forms of leukemia are associated with exposure to IR, whereas others are not. It is Leukemia is the commonest neoplastic disorder of the hematopoi generally believed that in the multistep process of leukemogenesis, the etic system. Like other human malignancies, leukemia arises as a initiation step represented by DNA damage and misrepair is a random consequence of an acquired genetic change in a cell capable of clonal phenomenon, and that selectivity in the establishment of a leukemic expansion. The commonest and most extensively characterized ge phenotype operates only at the subsequent stages of promotion and netic abnormalities that cause leukemia are the chromosomal translo progression. To test this concept, we have established an experimental cations that give rise to fusion genes encoding oncogenic proteins. The classical example of this type of abnormality is found in CML,3 system in which the formation of different types of fusion genes in hematopoietic cells can be assessed independently of the biological in which a t(9;22)(q34;ql 1) chromosomal translocation ( 1) generates consequence of the individual translocation, thereby eliminating the a BCR-ABL hybrid gene, which is transcribed into a chimeric mRNA and translated into a p210BCR"ABLfusion protein with elevated tyro- proliferative advantage component of the mechanism that gives rise to the actual leukemic process in vivo. This system is similar to that used sine kinase activity and transforming abilities (2). Several other types by Ito et al. (4) in their demonstration of radiation-induced BCR-ABL of chromosomal translocations and their derived fusion genes have translocations. We show here that the generation of leukemia-associ now been identified in acute leukemia. Some types seem to be lineage ated fusion genes by high-dose IR in different cell lines varies in specific, such as the t(l;19) PBX1-E2A gene and the t(l;ll) MLL- frequency according to the specific type of gene fusion, suggesting AF1P gene of acute lymphoblastic leukemia, the t(8;21) AML1-ETO that selectivity at the basic genetic recombination level may also gene, and the t(6;9) DEK-CAN gene of AML. or even restricted to a underlie the preferential induction of some forms of leukemia by IR. defined morphological subtype, such as the t(15;17) PML-RAR-a gene of M3-AML. MATERIALS AND METHODS It is widely accepted that leukemogenesis is a multistep process, consisting of initiation, promotion, and progression. The initiation Cell Cultures. A single batch from each of the HL60 (5) and KG1 (6) cell lines was used for all of the experiments after thorough karyotypic analyses to confirm the absence of t(9;22), 1(8:21), and t(6:9) translocations. The lines Received 10/7/97; accepted 11/25/97. were cultured in RPMI 1640 supplemented wilh 109f- FCS. y-irradiation of The cosls of publication of this article were defrayed in part by the payment of page replicale l()7-cell aliquots was carried out from a ' "Cs source operating al 1.85 charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely lo indicate this fact. Gy/min 24 h after subculturing. In the initial experiments, culture aliquots of ' Supported in pan by grants from the Leukaemia Research Fund (United Kingdom) the cell lines were irradiated in parallel with fast neutrons al upprnximalely and the lit. Mildred Scheel-Stiliung fur Krebsforschung (Germany). 2 To v.hom requests for reprints should be addressed, at LRF Centre for Adult 0.75 Gy/min. produced by a Van de Graat accelerator (7). After irradiation, the Leukaemia. Department of Haematology. Royal Postgraduate Medical School. Hammer cells were reincubated at 37Â°C,5% CO2, and 95% humidity for 24 or 48 h smith Hospital. Ducane Road, London W12 ONN. United Kingdom. Phone: 44-181-383- before being harvested. Cell viability in cultures before and alter irradiation 2167; Fax: 44-181-742-9335; E-mail: jmelo^rpms.ac.uk. 3 The abbreviations used are: CML. chronic myeloid leukemia; AML, acute myelo- was assessed by Irypan blue exclusion. blastic leukemia; DSB, double-strand break; IR. ioni/.ing radiation; RT-PCR, reverse RT-PCR Assays. RNA extraction and random hexamer-primed cDNA Iranscription-PCR: nt. nucleotide. synthesis from each cell aliquot were done as described previously (8. 9). 421

Quality control of the cDNA samples was assessed by a one-step PCR RESULTS amplification of the normal ABL message from 1 fil of cDNA. The remaining 79 /J.1of cDNA from each cell aliquot were then divided into 4 X KKl-^il Experimental Design. For establishment and optimization of the replicate PCR tests for the first-step amplification of a given fusion transcript, test system, several pilot experiments were carried out with variable followed by a second-step (nested) amplification of 1 p.1 of the first-step doses of low and high linear energy transfer irradiation produced by products. Amplification of BCR-ABL and of AML1-ETO fusion transcripts y-rays and neutron particles, respectively. Other variables in these were done as described previously (10-12), with minor modifications of the experiments included the number of cells/irradiated aliquot (from thermocycling conditions for the AMLI-ETO PCRs to increase the stringency 106-10S) and the time interval between irradiation and analysis. In the and specificity of the amplification: the annealing temperature was increased basic protocol, cells from exponentially growing myeloid cell lines from 62Â°Cto 66Â°Cand from 60Â°Cto 64Â°Cfor the first- and second-step PCRs. were exposed in liquid suspension to y-rays or to neutron particles, respectively: and the number of cycles in the first step was increased from 40 returned to standard culture for 24-48 h, and harvested for RT-PCR to 45. Positive controls for these PCRs were diluted cDNA preparations from the BV173 (for BCR-ABL) and the KASUMI-1 (for AMLI-ETO) cell lines, analysis for specific fusion gene transcripts. The cell viability in these cultures 24 and 48 h after y-irradiation was, on average, 55-79% and previously standardi/.ed in our single-test diagnostic protocols for reproducible detection of one leukemia cell in IO5 nonhematopoietic cells (murine fibro- did not differ significantly between the two cell lines studied (see blasts). PCR primers for amplification of DEK-CAN and DEK-ABL hybrid below) at each time point and dose (Table 1). The initial studies were transcripts were as follows: ECDE* (5'-tgaagaaTtCCAATGTTAAGAAAG- aimed at investigating the formation of the BCR-ABL hybrid gene, CAGAT) and HNCA~ (5'-gcataagcttTCCAGGTCACTGTCAA) for first-step the product of the t(9;22)(q34;ql 1) translocation of CML, in two DEK-CAN; HIDE* (5'-GATAGCAGCACCACCAAGAAGAATCA) and myeloid cell lines, HL60 and KG1, that do not contain a t(9;22). N2CA (5'-ACAGGCTCCACAGGGAAGTCTGTT) for second-step DEK- Preliminary results indicated that BCR-ABL transcripts could be CAN; ECDE* and HIAJ (5'-tattAAGCTTCCATTGATCCCGCT) for first- induced on rare occasions by both y-irradiation (at doses from 50-300 step DEK-ABL; and NÃŒDE*and Abl3' (5'-GGTACCAGGAGTGTTTCTC- Gy) and neutron particles (doses from 10-100 Gy) in both cell lines, CAGACTG) for second-step DEK-ABL. For positive controls for DEK-CAN and that the hybrid BCR-ABL mRNA molecules varied in size and and DEK-ABL amplifications, two artificial DNA constructs were engineered by cloning of individual DEK and CAN and DEK and ABL PCR-amplified sequence composition (Fig. 1Â«).Because the overall viability of neutron-irradiated cultures was poorer than that of y-irradiated cells sequences flanked by appropriate restriction sites into the polylinker of pEMBLIS. The resulting pEMBL-DECA and pEMBL-DEAB recombinant (data not shown) and no clear-cut advantage of using neutron irradi plasmids contain colinear arrangements of DEK-CAN or DEK-ABL sequences ation was observed in these initial experiments, we decided to con that include the expected exon junction points plus additional sequences from centrate on y-irradiation for the subsequent comprehensive analysis of (he downstream exon in DEK and the upstream exon in CAN or ABL. frequency. Likewise, in view of the apparent rarity of gene fusion respectively (maps and sequences available on request). After PCR with the induction at each individual dose of IR, the experimental protocol was second-step primers, these artificial constructs yield bands that are respectively redesigned and standardized as described below. 140 and 177 bp larger than those from the expected natural fusion transcripts. Induction of Different Fusion Genes. The pilot experiments es The composition of the PCRs for both DEK-CAN and DEK-ABL was: 1X tablished that the generation of transcriptionally functional BCR-ABL PCR buffer (Life Technologies, Inc.), 1.5 mM MgCU 0.25 Â¡JLMeachde- oxynucleotide triphosphate. 0.25 JAMeach 5' and 3' primer, and 0.025 unit//il leukemia fusion genes could be detected in irradiated cells under no selective pressure and provided the basis for a more detailed analysis Taq polymerase. Thermocycling parameters were as described previously (13). Each batch of PCR tests encompassed 40 tubes of test cDNAs from 10 cell of the frequency of the phenomenon and the possibility of inducing aliquots. 1 tube with the positive control cDNA and 4 negative controls derived other types of fusion genes. We therefore selected two other chromo from the RNA extraction. cDNA synthesis, and first-step and second-step PCR somal translocations, t(8;21)(q22;q22) and t(6:9)(p23:q34), which blanks. Mixes of reagents for cDNA synthesis and PCR amplification of each give origin, respectively, to the AMLI-ETO (14) and the DEK-CAN type of fusion transcript were prepared in bulk and stored as single-use aliquots (15) fusion genes both associated with AML and investigated whether with dedicated pipettes and plugged tips in a PCR-free laboratory by personnel these gene recombinations could also be induced by IR in HL60 and not involved with handling cells and/or actual PCR procedures. Aliquots from KG1 cells, which do not contain either chromosomal translocation. all mixes were extensively tested on mock 45-tube PCR assays containing no Furthermore, we asked whether IR at the high doses used in these DNA template (H,O blanks) before and at various intervals during the period experiments could also lead to the formation of irrelevant or nonsense of testing of actual cellular material and were always found to be contamina hybrid genes not known to be associated with human leukemia, such tion free. Tests of control nonirradiated cells were all done before those of as a DEK-ABL fusion gene. To obtain a valid estimate of the fre irradiated samples. The six stages of cell preparation. RNA extraction. cDNA synthesis, first-step PCR. second-step PCR. and electrophoresis of PCR prod quency at which each type of hybrid gene was formed, 4 sets of 40 identical 107-cell aliquots from each cell line were exposed each to 0, ucts were each carried out in six separate laboratory rooms in dedicated laminar air flow cabinets. No cellular material from patients with t(8;2I) or 50, and 100 Gy of y-rays for separate RT-PCR analyses of BCR-ABL, t(6;9) translocations had ever been processed in any of these rooms before or AMLI-ETO, DEK-CAN, and DEK-ABL fusion transcripts after 24-h during this study. postirradiation culture. The sensitivity of the RT-PCR assays was

Table I Cell viability in culture* harvested at 24 or 4K h after exposure tu y-irratiiatum This is expressed as the median percentage of viable cells as determined by trypan blue exclusion, the range in brackets, and (he number of experiments. Differences between viabilities of HL60 and KG1 cells al each lime point and dose are noi statistically significant (Mann-Whilney test).

OGy 50 Gy l(>OGy hHL6Ãœ 24 h93% 95% 193-97*1 [91-96%] [50-93%] [51-74%] [50-95%] [21-84%] (n = 6) (n = 5) (n = 3) (n = 3) (n = 5) (n = 5) KOI 96% 94% 70% 79% 79% 65% [95-97%] [92-96%] [68-84%] [63-85%| [72-87%] [65-69%] (n = 6)48 (n = 5)24h75% (n = 3)48h63% (n = 5)24h60% (n = 6)48h59% (n = 4}

422

1 2 3 5678 type of multiple replicate PCR assay is approximately 2.5-log more sensitive than our routine single-test nested RT-PCR protocol for molecular diagnosis and follow-up of minimal residual leukemia (10). allowing for the detection of fusion transcripts in at least 1 of IO7 cells. Extensive precautions were taken to avoid PCR contamination. The results showed a remarkable variation in the frequency of induction of each type of hybrid gene in the two cell lines, as summarized in Table 2. Thus, the mean frequency of BCR-ABL

123456 7 8 9 10 11 12 13 translocation events in irradiated KG1 cells was increased to only a minor degree as compared to that found in control nonirradiuted cells. The proportion of irradiated samples expressing DEK-CAN fusion genes did not differ from that seen in unirradiated cells. In contrast, the induction of AML1-ETO fusions in KG1 cells, but not in HL60 cells, was significantly higher in cultures exposed to y-irradiation, with nearly two-thirds of the cell aliquots irradiated at either 50 or KM) Gy exhibiting AML 1-ETO fusion transcripts, as compared to less than 8% of the nonexposed samples (P < 0.0001 for the x2 test). In irradiated KG1 cells, even the formation of a transcriptionally active DEK-ABL hybrid gene, a nonfunctional form of gene rearrangement, could be detected. The HL60 cell line showed an overall poor re sponse to IR. with no significant difference between exposed and nonexposed cells in the proportions of samples expressing the leuke mia fusion genes studied (Table 2). Structure of the Fusion Transcripts. Sequence analysis of the Fig. I. Representative second-step RT-PCR producÃsof the various types of fusion gene IR-induced BCR-ABL gene transcripts (Fig. la) showed that approx transcripts induced by IR in KG I and HL60 cells. The junction between sequences from the imately half of the samples produced mRNA molecules with the two partner genes is represented by u double-headed arrow (Â«->).In some transcripts, this junction is bridged by intervening sequences, most frequently of unknown origin. a, BCR- classical b3a2 or b2a2 junctions, as found in the great majority of ABL transcripts (observed in y- and/or neutron-irradiated cells!. Lane I. BCR exon patients with p210BCR ABLCML (2). In the other half, transcripts with b2Â«->ABLexon a2; Lane 2. BCR exon b3Â«-Â»ABLexona2: Lane 3. BCR exon b4Â«->ABLexon a2; Lane 4, BCR exon b5Â«-ABL exon a2; Lane 5. BCR exon el7Â«-Â»ABLexon a2; Lane t>. atypical junctions such as b5a2. b5alb, b4a2, b2a3, and el7a2 fusions BCR exon b4Â«->l4l bp of unknown sequenceÂ«-Â»ABLexona2; lane 7, BCR exon b5Â«->ABL or with intervening sequences between BCR and ABL were detected. exon Ib + exon a2; Lane X. BCR exon b5Â«-Â»l54bp of ABL intron la + exon a2. h. These intervening sequences comprised 62- or 154-bp fragments of AML1-ETO transcripts. Lane I, AML exon 5<-Â»ETOexon 2; Lane 2, AML exon 6<->ETO exon 2: Lane 3, AML exon 6Â«-*ETO exon Ib + exon 2; Lane 4 AML exon 5Â«-Â»ETOexonIh ABL intron la (corresponding to nt numbers 47310-47371 and + la + exon 2; Lane 5, AML exon 5Â«->l06bp of unknown sequence*-Â»ETOexon 2; Lane rt. 47218-47371. respectively, of the HSABLGR3 locus. GenBank ac AML exon 5 + exon 7bÂ«-Â»ETOexon2; Lane 7, AML exon 6Â«-Â»ETOexonIb + 68 bp of ETO intron + ETC) exon 2; Lane H. AML exon 5Â«->68bp of ETO intron + ETO exon 2; Lane 9, cession number U07563) inserted between BCR exon b5 and ABL AML exon 5Â«-Â»ETOexon la + exon 2; Lane II), AML exon 5 + exon 7bÂ«->59-bphuman exon a2 in one and three samples, respectively. One additional sample rRNA repeating unit sequenceÂ«-Â»ETOexonIb + exon 2; Lane II, AML exon 5Â«-Â»228hpof showed transcripts containing an insertion of 141 bp of unknown unknown sequenceÂ«-Â»ETOexon2 (larger fragment) and AML exon 5*-Â»158bp of unknown sequence*->ETO exon Ib + exon 2 (shorter fragmenu; Urne 12. AML exon 5*-*ETO exon origin between BCR exon b4 and ABL exon a2. Similar forms of 2 + 65-bp duplication of exon 2 (larger fragment) and AML exon 5Â«-Â»ETOexon2 (shorter atypical BCR-ABL transcripts in X-irradiated HL60 cells were re fragment); Lane 13, AML exon 5<-Â»ETOexon Ib + exon 2 (larger fragment) and AML exon 5Â«ETO exon 2. i, DEK-CAN transcripts. Lane I, pEMBL-DECA construct; Lane 2, DEK ported by others (4). (nt I080)Â«Can (nl 2531); Lane 3, DEK (nt 1()37)<-Â»CAN (nt 2070); Lane 4, DEK (nl The types of AML 1-ETO fusion transcripts induced by IR also 1003(Â«-Â»CAN(nt2250); Urne 5, DEK (nt 905-1003) + (nl 772-8l8)Â«-Â»CAN (nl 2753). The varied extensively in sequence composition (Fig. \h). One-third of the unavailability of data on the intron/exon boundaries of DEK and CAN precludes the precise assignment of donor and acceptor exons al the junction in the DEK-CAN transcripts, nt translocation products had a junction between AML1 exon 5 and ETO numbers are according to the GenBank sequence database under accession numbers X64229 exon 2 (A5-E2), which is characteristic of the t(8;21) leukemia- (DEK) and X64228 (CAN), ci. DEK-ABL transcripts. Lane 1, DEK (nt I080)<->ABLexon a2; Lane 2, pEMBL-DEAB construct. associated fusion gene (11); in approximately 40% of samples, AML1-ETO transcripts with an in-frame A6-E2 junction were found, optimized individually for each fusion gene by using diluted cDNAs and in the remainder, the hybrid mRNA molecules contained fusions from cells known to express the given gene or from artificially between AML1 and ETO alternatively spliced exons. Coexpression of constructed DNA fusions in a plasmici vector. We estimate that this two or more types of transcript was detected in several aliquots from

Table 2 Detection of fusion genes in cell lines exposed lo y-irratiialion Results are expressed as the number of posilive samples, i.e., the number of 107-cell aliquols (of 40 aliquois tested for each fusion gene) with specific I'CR producÃs,and as the number of samples with multiple fusions, i.e., positive samples with more than one type of fusion transcripts, as shown by PCR products with different si/es and junction sequences. NT, not tested. -Fusion No. of positive samples" (n 40)HL6050 samplesfusions''100 with multiple

geneBCR-ABL Gy6 Gy4 Gy2 Gy2 Gy0 Gy2 Gy0 Gy()

AML1-ETO 3 25 25 0 5 ? 0 6NT 19 0 1NT 0 DEK-CAN 20KG150 NTNT100 23OGy4 4 NTNT100 0 0 0 0 0 DEK-ABL(Kiy1 0= 0OGy0 0No.KG150 NTof 0OGy0 0HL6050 NT100 0 " Statistically significant difference between irradiated (at either 50 or KM) dy) and nonirradialed (0 Gy> samples only for AMLI-F-TO in KG1 samples (P < (UHH)I for the %~ test). ''Statistically significant difference between 50 and 100 Ciy irradiation for AMLI-ETO in K(il samples (/> < 0.004 for the \2 lest).

423

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1998 American Association for Cancer Research. INDUCTION OÃ l.irKKMIA I-TSION C,l:NI-;s BY IR both cell lines (Fig. le); this might have resulted from multiple and KG 1 cells by being mutant for p53 ( 19, 20), although a back translocation events and/or from alternative splicing from a single ground level of fusion gene formation has also been observed in AMLI-ETO fusion gene. In KG1 cells, this phenomenon was signif several other hematopoietic cell lines with normal p53 genes studied icantly more frequent after 100-Gy irradiation than after 50-Gy irra in our laboratory.4 The detection of such evidence of constitutional diation (76 and 24% of the positive samples, respectively; P < 0.004 genomic instability in cultured cells should probably not be surprising for the x1 test; Table 2). and is not without precedents, because similar infrequent events seem The DEK-CAN fusion transcripts observed in irradiated KG 1 cells to occur even in the absence of the stress conditions of normal cell had exactly the same configuration as those derived from the AML- culture, as shown by the observations in vivo of low levels of BCR- associated t(6;9) translocation (16), i.e., a junction between DEK nt ABL (21) and BCL2-IgH (22) fusion gene transcripts in blood cells 1080 and CAN nt 2531 (GenBank accession numbers X64229 and from normal individuals. X64228. respectively). In control nonirradiated HL60 and KG1 cells. (b) The second conclusion is that high-dose IR is able to induce per DEK-CAN transcripts with either this classical junction or with fu se a selective nonrandom increase in the background level of genetic sions between other DEK and CAN exons were also found at low recombination, with preferential formation of some genes as com frequency (Fig. lc). The nonsense DEK-ABL hybrid genes generated pared to others (e.g., AMLI-ETO versus BCR-ABL, DEK-CAN, and in irradiated KG1 cells were transcribed into fusion mRNA molecules DEK-ABL). It should be emphasized that in addition to their common with a perfect splicing of DEK nt 1080 to ABL exon 2 (Fig. \J). oncogenic characteristics, the three leukemia hybrid genes as well as The messages encoded by most of the atypical transcripts from the leukemia-unrelated DEK-ABL fusion chosen for this analysis BCR-ABL. AMLI-ETO, and DEK-CAN hybrid genes induced by IR have very similar genomic structures, because the genes from each were out of frame because of a wrong combination of exons at the partner pair are positioned in the same orientation in relation to the junction and/or due to the insertion of unknown, possibly intronic respective chromosome centromere, and all are fused in a head-to-tail sequences. The presence of premature termination codons in these configuration. One might therefore have expected that in the absence transcripts as well as in the DEK-ABL fusion results in the coding of of biological selective pressure, all of the four types of hybrid genes truncated, nonfunctional fusion proteins and is predicted to cause a would be generated at similar frequencies if the phenomenon of reduction in the respective mRNA levels (17, 18). It is therefore radiation-induced DNA double-strand breakage/rejoining were totally interesting to note that although the PCR assay used in this study is random. This was not. in fact, the case. The mechanism underlying the not strictly quantitative, these abnormal mRNA molecules were de target selectivity of this radiation-induced recombination bias is not tected at frequencies similar to those of in-frame transcripts not only known, but it could be related to a sequence-dependent variation in the after IR exposure, but also in nonirradiated cells. predominant type of DNA damage/repair process on an affected gene region (23); a requirement for compatibility between unusual DNA DISCUSSION conformations, such as Z DNA helix structures, at the breakpoint regions of the two partner genes (24, 25); the size of the intronic target We have shown that high-dose IR of hematopoietic cell lines can regions for breakpoints; or to the physical distance between the two induce the formation of a variety of fusion genes with corresponding chromosomes/genes in interphase nuclei (26). The higher susceptibil mRNA transcripts that are in some cases identical to those found in ity of certain chromosomal regions (fragile sites) to damage and naturally occurring human leukemias. We also found other analogous recombination could also be due to differences in chromatin organi fusion genes not known to be associated with human leukemia. The zation (27-29), which is an integral part of the DNA tertiary structure relative frequencies with which the four fusion genes studied were and modulates the accessibility of regulatory information for effective detected differed considerably, and their respective frequencies were transcription (30). Whatever the mechanism, our experiments provide also different in the two cell lines. It could be argued that the HL60 evidence that the reason why certain forms of chromosomal translo and KG I cell lines, being already of leukemic origin, could be more cations and gene fusions are more successful than others is not prone to undergo spontaneous or induced genetic recombination. restricted to their capacity to encode functional proteins but is also Insofar as general or random mutation is concerned, this is probably dependent on intrinsic differences in the likelihood of their being a correct assumption, and the known karyotypic instability of some generated by the initiating event. malignant cell lines in long-term culture is evidence that clonal Among atomic bomb survivors (31), leukemias accounted for the evolution and succession is a common phenomenon in selected largest proportion of the excess cancer cases attributable to the irra growth conditions. It seems unlikely, however, that in the conditions diation, demonstrating that leukemia is more easily induced by irra set up in our experimental strategy, the leukemic nature of the target diation than most other cancers. Furthermore, the magnitude and the cell would be a major contributor to the specificity of an induced temporal patterns of the radiation-induced risk differed significantly chromosomal translocation or gene fusion. Thus, although hemato between the different types of leukemia, with a remarkably high risk poietic cell lines reflect normal hematopoiesis occurring in vivo to for AML, CML, and acute lymphoblastic leukemia and no increased only a limited degree, it is possible that high-dose irradiation would incidence of chronic lymphocytic leukemia, adult T-cell leukemia, or induce comparable changes in normal human cells. multiple myeloma (31). The different time courses of the various The results of this study lead to two important conclusions: leukemias could be explained if the kinetics of proliferation in the ((/) At least in vitro, there must be a low background level of transformed cell populations were governed differently by the various transcriptionally productive hybrid gene formation in hematopoietic fusion gene products associated with each of the leukemias and/or cells, as demonstrated by the detection of different types of specific possibly by differing requirements for additional, cumulative muta fusion transcripts in two different cell lines in the absence of exposure tions before expression of the full leukemic phenotype. However, the to irradiation. The possibility that this observation might have resulted preferential occurrence of some but not other types of leukemia could from a technical artifact was thoroughly investigated and effectively excluded, because the strategy adopted in the RT-PCR assays in be related not only to the higher capacity of some oncogenic proteins cluded comprehensive positive and negative controls. The inherent 4 S. Bose, M. W. N. Deininger. J. Gora-Tybor. J. M. Goldman, and J. V. Melo. The genomic instability in cell lines may be exacerbated by the oxidative presence of BCR-ABL fusion genes in leukocytes of normal individuals, manuscript in stress of normal culture conditions. This may be compounded in HL60 preparation. 424

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1998 American Association for Cancer Research. INDUCTION OF LEUKEMIA FUSION GENES BY IR to confer a survival advantage on the target cell but also, as our 11. Kozu, T., Miyoshi. H.. Shimizu. K.. Maseki. N.. Kaneko. Y., Asou, H.. Kamada. N., experimental data suggest, to the major differences in the probability and Ohki, M. Junctions of the AMLI/MTG8(ETO) fusion are constant in t(8;2l) acute myeloid leukemia detected by reverse transcription polymerase chain reaction. that a specific leukemia-associated gene fusion is induced by irradi Blood. Â«2:1270-1276. 1993. ation. Finally, the clear difference between the KG1 and HL60 cell 12. Maruyama. F., Stass, S. A., Estey. E. H., Cork, A., Hirano. M.. Ino, T.. Freireich. E. J., Yang, P., and Chang, K. S. Detection of AMLI/ETO fusion transcript as a tool for lines in their susceptibility to acquire some types of fusion gene, diagnosing t(8;21) positive acute myelogenous leukemia. Leukemia (Baltimore), 8: notably AML1-ETO, makes it unlikely that the phenomenon is de 40-45, 1994. pendent on or exclusive to the preexisting malignant genotype of these 13. Diamond, J.. Goldman, J. M., and Melo, J. V. BCR-ABL. ABL-BCR, BCR. and ABL genes are all expressed in individual granulocyte-macrophage colony-forming unit cells. Instead, the difference could be explained if, in addition to local colonies derived from blood of patients with chronic myeloid leukemia. Blood. 85: sequence/structural requirements for generating chromosomal trans- 2171-2175. 1995. locations, intrinsic or inherited factors in the cell lines determined in 14. Miyoshi. H.. Kozu, T.. Shimizu, K.. Enomoto. K.. Maseki. N.. Kaneko. Y.. Kamada. N., and Ohki, M. The t(8;2l) translocation in acute myeloid leukemia results in part their susceptibility to generating new fusion genes. In other production of an AML1-MTG8 fusion transcript. EMBO J.. 12: 2715-2721. 1993. words, the different susceptibilities of the different cell lines could 15. Soekarman. D., von Lindern, M., Daenen. S., de Jong. B., Fonatsch, C., Heinze. B.. Bartram. C., Hagemeijer. A., and Grosveld. G. The translocalion (6;9) (p23;q34) parallel differences in susceptibility to leukemia in different persons shows consistent rearrangement of two genes and defines a myeloproliferative dis exposed to the same given dose of irradiation, differences presumably order with specific clinical features. Blood. 79: 2990-2997. 1992. due to inherited factors (32) or other acquired somatic mutations. The 16. von Lindern, M.. Fomerod. M., van Baal, S., Jaegle. M., de Wit. T.. Buijs, A., and Grosveld. G. The translocalion (6;9). associated with a specific subtype of acute myeloid availability of the experimental model described in this study should leukemia, results in the fusion of two genes, dek and can, and the expression of a chimeric, provide a sensitive tool for investigating these and other aspects of the leukemia-specific dek-can mRNA. Mol. Cell. Biol.. 12: 1687-1697. 1992. 17. Maqual, L. E. When cells stop making sense: effects of nonsense codons on RNA effects of IR on hematopoietic cells. metabolism in venebrate cells. RNA. /: 453-465, 1995. 18. Carter. M. S., Li, S., and Wilkinson. M. F. A splicing-dependent regulatory mecha nism thai detects translation signals. EMBO J., 15: 5965-5975. 1996. ACKNOWLEDGMENTS 19. Kuerbilz, S. J., Plunkett. B. S., Walsh, W. V.. and Kastan. M. B. Wild-type p53 is a cell cycle checkpoint determinant following irradiation. Proc. Nati. Acad. Sci. USA. We are grateful to Dr. Kevin Prise and Prof. Barry Michael of the Gray 89: 7491-7495, 1992. Laboratory, Mount Vernon Hospital (London, United Kingdom) for expert 20. Calabresse, C., Venturini. L.. Ronco, G., Villa. P., Degos, L., Belpomme, D., and advice and permission for use of the Van de Graaf accelerator. Chomienne, C. Selective induction of apoptosis in myeloid leukemic cell lines by monoacelone glucose-3 butyrate. Biochem. Biophys. Res. Commun.. 201: 266-283. 1994. REFERENCES 21. Biernaux, C.. Loos, M.. Sels, A.. Huez, G., and Stryckmans. P. Detection of major bcr-abl gene expression al a very low level in blood cells of some healthy individuals. 1. Rowley, J. D. A new consistent chromosomal abnormality in chronic myelogenous Blood, 86: 3118-3122, 1995. leukaemia identified by quinacrine fluorescence and Giemsa staining. Nature (Lond.). 22. Limpens, J.. Stad, R., Vos, C., de Vlaam, C., de Jong, D., van Ommen, G. J., 243: 290-29.1, 1973. Schuuring, E., and Kluin, P. M. Lymphoma-associated translocation t( 14;I8| in blood 2. Melo, J. V. The diversity of BCR-ABL fusion proteins and their relationship to B cells of normal individuals. Blood. 85: 2528-2536, 1995. leukemia phenotype. Blood, 88: 2375-2384, 1996. 23. Frankenberg Schwager, M.. Jha. B.. Bar. K., and Frankenberg. D. Molecular mech 3. Lord. B. I., and Hendry, J. H. Radiation toxicology: bone marrow and leukaemia. In: anism of potentially lethal damage repair. I. Enhanced fidelity of DNA double-strand i. H. Hendry and B. I. Lord (eds.). Radiation Toxicology: Bone Marrow and break rejoining under conditions allowing potentially lethal damage repair. Int. J. Leukaemia, pp. 1-21. London: Taylor & Francis, 1995. RadiÃ¢t.Biol., 67: 277-285, 1995. 4. Ito, T., Seyama, T.. Mizuno. T.. Hayashi, T., Iwamoto, K. S., Doni, K., Nakamura, N.. 24. Boehm, T., Mengle Gaw. L., Kees, U. R.. Spurr, N.. Lavenir, I., FÃ¶rster, A., and and Akiyama. M. Induction of BCR-ABL fusion genes by in vitro X-irradiation. Jpn. Rabbitts, T. H. Alternating purine-pyrimidine tracts may promote chromosomal J. Cancer Res., 84: 105-109, 1993. translocations seen in a variety of human lymphoid tumours. EMBO J., 8: 2621-2631, 5. Gallagher. R.. Collins, S.. Trujillo. J., McCredie. K.. Ahearn. M.. Tsai, S.. Metzgar. 1989. R.. Aulakh. G.. Ting. R.. Ruscelli. F.. and Gallo, R. Characterization of the contin 25. Arnheim. N. The possible role of Z DNA in chromosomal translocations. Cancer uous, differentiating myeloid cell line (HL-60) from a patient with acute promyelo- Cells (Cold Spring Harbor), 2: 121-122. 1990. cytic leukemia. Blood. 54: 713-733, 1979. 26. Kozubek. S., Lukaova. E.. Ryznar. L.. Kozubek. M., Likova. A.. Govorun, R. D.. 6. Koeffler, H. P., and Golde, D. W. Acute myelogenous leukemia: a human cell line Krasavin. E. A., and Horneck. G. Distribution of ABL and BCR genes in cell nuclei responsive to colony-stimulating activity. Science (Washington DC), 200; 1153- of normal and irradiated lymphocytes. Blood, 8V: 4537-4545. 1997. 1154, 1978. 27. Popescu, N. C. Chromosome fragility and instability in human cancer. Crit. Rev. 7. Prise, K. M.. Davies. S., and Michael, B. D. The relationship between radiation- Oncog.. 5: 121-140, 1994. induced DNA double-strand breaks and cell kill in hamster V79 fibroblasts irradiated 28. Hecht, F., and Hecht. B. K. Genetic activity and fragile sites. Cancer Genet. Cyto- with 250 kVp X-rays, 2.3 MeV neutrons or 238Pu a-particles. Int. J. Radial. Biol. genet., 55: 113, 1991. Relat. Stud. Phys. Chem. Med., 52: 893-902, 1987. 29. Ward, J. F. The yield of DNA double-strand breaks produced intracellularly by 8. Chomczynski, P., and Sacchi. N. Single-step method of RNA isolation by acid ionizing radiation: a review. Int. J. RadiÃ¢t.Biol., 57: 1141-1150. 1990. guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 762.- 156- 30. Beato, M. Chromatin structure and the regulation of gene expression: remodeling at 159, 1987. the MMTV promoter. J. Mol. Med.. 74: 711-724, 1996. 9. Cross, N. C, Melo, J. V., Feng. L., and Goldman. J. M. An optimized multiplex 31. Preston. D. L.. Kusumi. S., Tomonaga. M.. Izumi. S.. Ron. E.. Kuramoto. A.. polymerase chain reaction (PCR) for detection of BCR-ABL fusion mRNAs in Kamada, N., Dohy, H.. Matsuo, T.. Matsui. T., Nonaka. H.. Thompson. D. E.. Soda. hematological disorders. Leukemia (Baltimore), 8: 186-189, 1994. M.. and Mabuchi. K. Cancer incidence in atomic bomb survivors. Part III. Leukemia, 10. Cross, N. C, Feng, L., Chase, A., Bungey, J., Hughes. T. P.. and Goldman, J. M. lymphoma and multiple myeloma. 1950-1987. Radial. Res., 137: S68-S97, 1994. Competitive polymerase chain reaction to estimate the number of BCR-ABL tran 32. Taylor. G. M. Genetic effects of ionizing radiation with respect to leukaemia. In: J. H. scripts in chronic myeloid leukemia patients after bone marrow transplantation. Hendry and B. 1. Lord (eds.). Radiation Toxicology: Bone Marrow and Leukaemia, Blood. 82: 1929-1936, 1993. pp. 277-310. London: Taylor & Francis. 1995.

425

Michael W. N. Deininger, Shikha Bose, Joanna Gora-Tybor, et al.

Cancer Res 1998;58:421-425.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/3/421

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/58/3/421. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.