Document p4Ema8dbkojBNJm4pqKDoqLE

ICANCER RESEARCH 58. 2I76-2ISI. May 15. 1998] Increased Translocations and Aneusomy in Chromosomes 8 and 21 Among Workers Exposed to Benzene1 Martyn T. Smith,2 Luoping Zhang, Yunxia Wang, Richard B. Hayes, Guilan Li, Joseph Wiemels, Mustafa Dosemeci, Nina Titenko-Holland, Liqiang Xi, Prema Kolachana, Songnian Yin, and Nathaniel Rothman Division of Environmental Health Sciences, School of Public Health, University of California. Berkeley. California 94720-7360 M.T. S., L Z, Y. W.. J. W.. N. T-H., P. K.]: Division of Cancer Epidemiiihigv unti Genetics. National Cancer Institute. Bethesdii. Maryland 20X92 R.B. H.. M. D., N. K.:and Institute of Occupational Medicine. Chinese Academy of Preventive Medicine. Beijing 100050. China C.L. L. X.. S. Y.I ABSTRACT Chromosome aberrations in peripheral blood lymphocytes have been used for many years to monitor human populations exposed to potential carcinogens. Recent reports have confirmed the validity of this approach by demonstrating that elevated levels of chromosome aberrations in lym phocytes are associated with subsequent increased cancer risk, especially for increased mortality from hematological malignancies including acute myeloid leukemia (AMD. We postulated that this approach could be improved in two ways: (a) by detecting oncogenic disease-specific aber rations; and (/>) by using chromosome painting so that many more metaphases could be analyzed. Numerical and structural aberrations in chro mosomes Kand 21 are commonly observed in AMI . In the present study, we painted chromosomes 8 and 21 in lymphocyte metaphases from 43 healthy workers exposed to benzene, an established cause of AML, and from 44 matched controls. To examine dose-response relationships the workers were divided into two groups at the median exposure level, a lower-exposed group (31 ppm; n = 21), and a higher-exposed group (>31 ppm; n = 22). Benzene exposure was associated with significant increases in hyperdiploidy of chromosomes 8 (1.2, 1.5. and 2.4 per 100 metaphases; /' < O.O(H)l) and 21 (0.9, 1.1, and 1.9 per I(Mlmetaphases; /' < O.(HM)l).Translocations between chromosomes 8 and 21 were in creased up to 15-fold in highly exposed workers (0.01, 0.04, and 0.16 per 100 metaphases; P < 0.0001). In one highly exposed individual, these translocations were reciprocal and were detectable by reverse transcriptase-PCR. These data indicate a potential role for t(8;21) in benzeneinduced leukemogenesis and are consistent with the hypothesis that de tection of specific chromosome aberrations may be a powerful approach to identify populations at increased risk of leukemia. INTRODUCTION dromes (8-10). A variety of specific chromosome aberrations, includ ing trisomy of chromosomes 8 and 21 and translocations between these two chromosomes, are commonly found in AML and myelo dysplastic syndromes (11-13). Translocation (8;21)(q22;q22) is the most common translocation observed in AML. It occurs in 40% of patients with the AML-M2 subtype and disrupts both the AMLOE (CBFA2) gene on chromosome 21 and the ETO (MTG8) gene on chromosome 8 (13, 14). The fusion transcript AML1-ETO has been isolated (15) and can be detected by RNA-based RT-PCR (16). Chromosome translocations can also be detected by chromosome painting using FISH (17). FISH has a number of advantages over PCR including the fact that not only structural but also numerical chromo somal aberrations can be visualized and precisely quantitated. On the other hand. PCR holds the potential advantage of greater sensitivity. FISH and RT-PCR. therefore, complement one another. We propose that specific chromosomal aberrations that are known to be oncogenic may provide a better marker of future leukemia risk than conventional chromosomal aberrations. As an initial test of this hypothesis, we have used FISH and RT-PCR to determine the pres ence of AML-specific chromosomal aberrations in the peripheral blood of workers who are at increased risk of AML because they are exposed to high levels of benzene and in the peripheral blood of matched controls. We demonstrate that AML-specific aberrations are found at increased levels in the blood cells of workers exposed to benzene and propose that the detection of disease-specific chromo somal aberrations by chromosome painting and RT-PCR may be a powerful approach to identifying populations at increased risk of leukemia. The measurement of chromosome aberrations in peripheral blood lymphocytes has been used for many years to monitor human popu lations exposed to carcinogens or potential carcinogens (1. 2). For example, there are numerous studies in the literature that show that human exposure to the leukemogen benzene increases the level of chromosome aberrations in lymphocytes detected by conventional methods (3-5). Recent reports have confirmed the validity of this approach by demonstrating that elevated levels of chromosome aber rations in peripheral blood lymphocytes are associated with subse quent increased cancer risk (6, 7), especially for increased mortality from hematological malignancies including AML1 (7). Exposure to benzene in the workplace has been associated with increased risk of AML, aplastic anemia, and myelodysplastic syn- Received 11/14/97; accepted 3/17/98. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore he hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by funds from the National Cancer Institute: National Institute of Envi ronmental Health Sciences Grants ROI ES06721. P42 ES04705. and P30 ES01896; and a grant from the California Environmental Protection Agency (to M. T. S.). J. W. was supported by a Howard Hughes Predoctoral Fellowship. 2 To whom requests for reprints should be addressed, at Division of Environmental Health Sciences. School ol Public Health. University of California. Berkeley. CA 947207360. 1The abbreviations used are: AML. acute myeloid leukemia: FISH, fluorescence in situ hybridi/ation: RT-PCR. reverse Iranscriptase-PCR: TWA. time-weighted average. MATERIALS AND METHODS Subject Enrollment and Exposure Assessment. The enrollment of sub jects and identification of factories for this study in Shanghai. China has been previously described in detail by Rothman et al. (18, 19). Forty-four exposed workers and 44 matched controls were enrolled in this study. Informed consent was obtained from each subject, and a questionnaire was administered to dclermine age, gender, current and lifelong tobacco use, current alcohol consumption, medical history, and work history. Exclusion criteria for all subjects were history of cancer, therapeutic radiation, chemotherapy, or current pregnancy. Three factories that used benzene were selected so that the study population would have a wide range of exposures to benzene. Two workplaces in the same geographic area that did not use benzene or other chemicals associated with bone marrow toxicity were selected as control factories. One factory manufactured sewing machines, whereas the second workplace was an administrative facility. Individual exposure was monitored by passive dosimetry badges, which were worn by each worker for a full workshift on 5 separate days during the 1-2-week period before phlebotomy. Badges were analyzed by gas chromatography with flame ionization detection. An 8-h TWA exposure to benzene was calculated as the geometric mean of the five air measurements. The median TWA was 31 ppm (range, 1-328 ppm). Current exposures to benzene were confirmed by the analysis of phenol and other metabolites in the urine (19). These exposures were reflective of workplace exposures for at least 1 year before the study, as determined from a review of factory work practice records (20). Cumulative exposure to benzene while working in these factories 2176 DETECTION OF t(H:2l) IN BENZENK-KXPOSKD WORKERS employed was also assessed as described previously (19. 21). All workers had been employed for at least 8 months, and their job practices were extremely stable. Remedial action was taken subsequently at the two workplaces with the highest benzene exposures. This included the substitution of toluene for benzene, the enclosure of reaction vessels, and an improvement in ventilation. Blood Sample Collection and Culture. Peripheral blood was obtained by phlebotomy: cultured in RPMI 1640 supplemented with 15% fetal bovine serum. 1% penicillin-streptomycin, 1% i.-glutamine (Life Technologies. Inc., Grand Island. NY), and 1% phytohemagglutinin-P (Pharmacia. Piscataway. NJ) at 37Cin a 5% CO2 moist atmosphere: and harvested 72 h after culture leukemia virus reverse transcriptase (200 units; Promega, Madison. WI) was added to a total volume of 20 /ail. The mixture was incubated at room temperature for IO min and then incubated at 37Cfor 60 min. Negative controls included samples with all components hut the reverse transcriptase or RNA. The positive controls used were the Kasumi-1 cell line (22) kindly provided by Dr. Nanao Kamada (Research Institute of Nuclear Medicine and Biology. Hiroshima. Japan) and RNA isolated from two AML patients with t(8;21). The cDNA mixture was PCR amplified with the primers AML1/ETO5A (GACCTCAGGTTTGTCGGTCGAAG) and AML1/ETO3A (CCATTGCT- initiation. Colcemid (0.1 tg/ml)was added 4 h before harvesting to oblain the metaphase spreads. After hypotonie treatment (0.075 M KC1) for 30 min at 37C,the cells were fixed three times with fresh Carney's solution (methanol: glacial acetic acid, 3:1). The fixed cells were then dropped onto prelabeled glass slides, air dried, stored at 2"0Cunder nitrogen, and shipped on dry ice GAAGCCATTGGGTGG) using a hot start protocol. Five /il of cDNA, 25 pmol of each primer. 5% DMSO. 50 mM KG. 10 mM Tris-HCI (pH 8.3), 2.5 pmol of each deoxynucleotide triphosphate. and 2.5 units of Taq polymerase (Perkin-Elmer Corp.. Foster City, CA) were amplified for 39 cycles of PCR (94Cfor 1 min. 55Cfor 1 min, and 72Cfor I min, followed by 72Cfor to the United States. Blood from one person could not be successfully cultured, and results are reported here for only 87 subjects. Blood smears were also prepared and used for the PCR-based analyses. FISH Analysis of Metaphase Spreads. Dual-color painting of whole 20 min). A second nested reaction using I /Iof the first reaction was performed using the 5'B and 3'B primers described previously (16). The second PCR reaction was electrophoresed through 2% agarose, blotted to nylon membranes (Genescreen; DuPont). and probed by a l:!P-labeled oligo- chromosomes 8 (green) and 21 (red) was performed on metaphase spreads prepared from cultured lymphocytes. The biotinylated human painting probe specific for chromosome 8 and the digoxigenin-labeled painting probe for chromosome 21 were purchased from Oncor, Inc. (Gaithersburg. MD). After the denaturation of cellular DNA in 70% formamide in 2x SSC at 72Cfor 2 nucleotidc probe encompassing the junction of the fusion transcript. Quality of the RNA was assessed by the ability to PCR-amplify Y-actinand hypoxanthine phosphoribosyltransferase genes from the cDNA. PCR products from the external reaction (5A and 3A primers) were directly sequenced using the 5'B and 3'B primers (PCR Product Sequencing Kit; Amersham). Both strands were min, the slides were quickly dipped in ice-cold 70, 85, and 100% ethanol for 2 min each to dehydrate them and then air dried. After prewarming for 5 min at 37C,each of the painting probes was denatured separately for 10 min at 70Cand then preannealed for 2.5 h at 37C.The differently labeled probes of chromosomes 8 and 21 were mixed well, applied to prewarmed slides (at 37C) on a slide warmer, and hybridized overnight with target DNA at 37Cin a humidified chamber. The slides were washed the next day in 0.5 X SSC at 72Cfor 5 min after three brief washes in 0.1 M phosphate buffer (pH 8.0) at room temperature. The hybridization signals were detected in a mixed solution of 10 fig/ml FITC- conjugated avidin (Vector Laboratories, Burlingame, CA) in preblock solution (4X SSC, 0.1% Triton X-IOO, 0.02% sodium azide, and 5% Carnation nonfat dry milk) and 10 fig/ml antidigoxigenin antibody conjugated with rhodamine (Boehringer Mannheim. Indianapolis. IN) in phosphate buffer with 4% BSA for 15 min at 37C.After three washes in phosphate buffer for 2 min each with sequenced, and the results were verified to the published deletion junction sequence (15). Statistical Analysis. Study subjects were divided into three groups: () controls; (ft) workers exposed to S31 ppm of benzene; and workers exposed to >31 ppm of benzene as an 8-h TWA. We have previously shown that these exposure categories are highly correlated with urinary benzene metabolite levels and various measures of hematotoxicity ( 18). Summary data for aneu- ploidy of chromosomes 8 and 21 within these exposure groups are presented as the mean of the number of aberrations per 100 metaphases scored for each subject. A square root transformation was used to normalize each outcome, and a test for trend was performed by linear regression, controlling for age and gender, the original matching variables. There was no evidence of confounding by current or previous tobacco use. alcohol use. or body mass index, and these factors were excluded from final models. Because structural chromosomal aberrations occurred at 1-2 orders of magnitude less frequently than aneu- intermittent agitation at room temperature, the nuclei of the cells were counterstained with a blue fluorescent dye, 4'.6-diamidino-2-phenylindole (0.1 /xg/ml: Sigma. St. Louis, MO), prepared in a mounting medium (Vector Laboratories). The hybridization signals were viewed under a fluorescence microscope equipped with epifluorescent illumination and a X 100 oil immer sion lens. A triple-bandpass filter for 4',6-diamidino-2-phenylindole/FITC/ Texas Red (excitation at 405. 490, and 570 nm: emission at 460. 525. and 635 nm) was used. The stained slides were randomized and coded before scoring. For effi ciency, all scoratale metaphase spreads on two spots of each slide were analyzed. Metaphase cells were considered scorable if they met the following criteria: (a) the cells appeared to be intact; (h) the chromosomes were con densed and well spread: (c) the centromeres of all chromosomes were readily visible: and (d) staining was sufficiently bright to enable the detection of genetic rearrangements on chromosomes 8 and 21 and exchanges between targeted and nontargeted chromosomes. A total of 42,082 metaphase spreads from the 87 subjects (average, 483 cells/subject) were scored for the presence of numerical and structural aberrations in chromosomes 8 and 21. This is approximately 10 times more than the number routinely scored by conven tional cytogenetics (50 cells/subject). Molecular Analysis of Blood Smears for the Presence of AML1/ETO Chimeric cDNA Derived from t(8;21). One-half of the contents of coded blood smear slides previously stored at room temperature were scraped into microcentrifuge tubes, and total RNA was isolated with TRIzol reagent (Life Technologies, Inc.). Approximately 0.5-1.2 tgof RNA were isolated by this method, as determined by A260nm. One-third of the RNA was added to 50 ng of random primers, heated to 95Cfor 5 min, quick-cooled on ice, and then added to a mixture containing 50 mM Tris-HCI (pH 8.3), 75 mM KC1, 3 mM MgCU 50 mM DTP, 25 pmol each of four deoxynucleotide triphosphates, 20 units of RNasin. and 50 ng of random hexanucleotide. Moloney murine ploidy. the data were pooled for all subjects within each exposure category. Summary data are presented as the total number of events detected over the total number of metaphases scored within each group, and a test for trend was performed using exact methods. RESULTS Demographics and Exposure Levels. Exposed workers and con trols had similar demographic characteristics with mean SD ages of 35.3 7.8 and 35.4 7.3 years, respectively. Forty-eight percent of each group were women. None of the women in either group smoked, but 21 of 23 men in both the exposed and control groups were smokers. The exposed males smoked 11.0 7.8 cigarettes/day, and nonexposed males smoked 13.5 13.7 cigarettes/day. Both groups had a low mean alcohol intake of approximately 1.5 drinks/week. The workers exposed to benzene had worked in their respective factories for a mean of 6.3 4.4 years, with a range of 0.7-16 years. The median current benzene air exposure level was 31 ppm. The exposed workers were divided into two groups at the median exposure level, a lower-exposed group (s31 ppm) and a higher-exposed group (>31 ppm), to evaluate the dose-response relationships. Cumulative expo sure was assessed in terms of ppm-years. Fifteen exposed subjects (34%) had 1-100 ppm-years of exposure, 16 (36%) had 100-500 ppm-years of exposure, and 13 (30%) had >500 ppm-years of expo sure. Aneusomy of Chromosomes 8 and 21. Numerical chromosome aberrations (aneuploidy/aneusomy) can be divided into two catego ries: (a) hyperdiploidy (a gain in the chromosome number); or (b) 2177 DETECTION OF t(8;21) IN BENZENE-EXPOSED WORKERS hypodiploidy (a decrease in the chromosome number). Benzene ex posure was associated with a highly significant increasing trend in hyperdiploidy of both chromosomes 8 and 21 in the lymphocytes of exposed workers (P < 0.0001; Table 1). In the higher-exposure category, the rise in hyperdiploidy was 2-fold. The effect of benzene on both chromosomes was approximately equal and was mainly in the form of trisomy. In contrast to the significant correlation observed between current benzene exposure and hyperdiploidy, no correlation was found with cumulative exposure. As expected, hypodiploidy was considerably more prevalent than hyperdiploidy in the metaphase spreads. Exposure to benzene caused a small but significant increase in the hypodiploidy of chromosome 8 (P = 0.0007) but not that of chromosome 21 (P = 0.84; Table 1). Interestingly, the background level of chromosome 21 hypodiploidy was 2- to 3-fold higher than that of chromosome 8 (Table 1). Structural Aberrations in Chromosomes 8 and 21. Three types of structural aberrations in chromosomes 8 and 21 could be detected by chromosome painting; namely, breaks, deletions, and transloca tions. Breaks were observed as painted acentric fragments. Unpainted acentric fragments were excluded from the analysis. Deletions were detected by comparing the length of two chromosomes painted in the same color. Three types of translocations could be detected by paint ing: (a) translocations between chromosomes 8 and 21 [t(8;21)]; (b) translocations between chromosome 8 and another unidentified chro mosome [t(8;?)l; and (c) translocations between chromosome 21 and another unidentified chromosome [t(21;?)]. The frequency of these different types of structural aberrations is shown in Fig. 1 using pooled metaphase data. Benzene caused a highly significant increase in chromosome 8 breaks (P < 0.0001), but not in deletions (P = 0.48: Fig. 1). The level of t(8;21) was increased 15-fold in workers exposed to >31 ppm over the level found in control subjects. It was also significantly higher in the lower-exposed group (^31 ppm) than it was in the controls. The increase in other translocations, i.e., t(8;?) and t(21;?), was smallerai around 4-fold in the group exposed to > 31ppm and was not significantly higher in the group exposed to 31 ppm as compared to controls (Fig. 1). To further explore the distribution of aberrations within the differ ent exposure categories shown in Fig. 1, the data are presented in Table 2 as the proportion of subjects in each exposure group with at least one detectable event. A tendency toward an increased number of individuals with detectable structural aberrations in their blood in the exposed groups is clearly observable (Table 2). Note that 2% of the control subjects ( 1 individual) had detectable t(8;21), whereas 19% of the workers in the s31 ppm lower-exposed group (four individuals) and 27% of workers in the higher (>31 ppm)-exposed group (six individuals) had detectable t(8:21 ). Similar increases in the number of individuals with t(8;?), t(21;7), and chromosome 8 breaks were also observed. To determine whether individuals with the highest level of struc tural aberrations in their blood biased this analysis, we repeated the trend tests excluding the subject with the highest value in each category. Similar highly significant trends were found: t(8;21), P = 0.00013; t(8;?), P < 0.0001 ; t(21 ;?), P = 0.045; and chromosome t(8;21) 8 Break 8 Deletion Fig. 1. Structural chromosome aberrations in chromosomes 8 and 21 in lymphocytes from the blood of workers exposed to different levels of benzene and unexposed controls, as detected by chromosome painting. P values equal to trend test by exact methods. 8 breaks, P < 0.0001. This demonstrates that outliers or subjects with clonal abnormalities are not the cause of the elevated levels of structural aberrations seen in the benzene-exposed workers as a whole. Interestingly, the worker with the highest level of structural chromosome aberrations did not have the highest level of t(8;21). However, the worker with the highest level of t(8;21) (7 of 100 metaphases) was also the only individual to have detectable reciprocal translocations between chromosomes 8 and 21 (2 of 100 metaphases). In all other t(8;21 (-positive individuals (10 of 43 individuals), part of chromosome 21 was translocated to chromosome 8, or part of chro mosome 8 was translocated to chromosome 21, but no reciprocal exchanges were detectable. Molecular Analysis of t(8;21) by RT-PCR. To further character ize the t(8;21) translocations described above, we performed a mo lecular analysis using RT-PCR of RNA isolated from the blood of a subset (n = 21) of 11 highly exposed workers and 10 matched unexposed controls (Fig. 2). Five members of the exposed group and one member of the unexposed group had t(8;21) detectable by FISH. The groups were matched on age, gender, and smoking status. RNA was isolated from coded blood smear slides and reverse-transcribed, and AML1/ETO chimeric cDNA was PCR-amplified in a nested reaction. PCR products were blotted and probed with a junction- specific oligonucleotide and sequenced for confirmation. The exper iment was performed blind, so that lanes on the gel were identified after the fact. In one exposed individual who had reciprocal t(8;21) detectable by FISH, we were able to detect the AML1-ETO fusion transcript, demonstrating that the t(8;21) translocation detected by FISH in this worker's blood was in-frame (Fig. 2). Sequencing showed that it was identical to that found in the t(8;21)-positive Kasumi-1 cell line and two AML patients positive for t(8;21) by G-banding analysis. We did not detect fusion transcripts in other Table 1 Aneusomy of Chromosomen8 unii 21 in benzene-exposed workers and matched control* 8Benzene Chromosome 21Hypodiploidy14.2 (n)Conterxoplsosure (44) 31ppm (21) ,>3I ppm (22) Trend (P value)''Hypodiploidy5.8 3.2" 6.2 2.7 8.9 4.8 0.0007Hyperdiploidy1.2 0.8 1.5 0.8 2.4 1.1 -C0.0001Chromosome ' Mean SD events per 100 scored metaphase cells. ' P values obtained by linear regression on square root-transformed values, adjusted for age and gender. 8.7 15.2 6.6 13.7 5.4 0.84Hyperdiploidy0.9 2178 0.7 1.1 0.6 1.9 0.9 <0.0001 DETECTION OF Il8;21) IN BENZENE-EXPOSEDWORKERS Table 2 Distribution of structural aberrations involving chromosomes K and 21 in ben-ene-exposed workers und mulched controls (Ben)zCeonneteroxlpsosure (44) (if s31 ppm (21) 4(19) 4(19) >31 ppm (22)t(8;21)1' 6(27)t(8;?)5(11) 11 (50)1(21;?)*4(9) " Translocation between chromosome 8 and another unidentified chromosome. ' NTruamnsbleorcaotfiosnubbjeetcwtseewnitchhsropmecoifsiocmsetru2c1tuarnadl cahnaontghee.r unidentified chromosome. rfPercentage of subjects with specific structural change. 1(5) 8(36)Breakage of chro8m1o0s(o2m3)e 7(33) 12(55)Deletion in chrom85o(s1o1m) e 0(0) 5(23) exposed workers or control subjects (Fig. 2). RNA was amplified from all subjects using Y-actinand hypoxanthine phosphoribosyl- transferase primers showing that failure to detect fusion transcripts was not due to a lack of RNA (data not shown). DISCUSSION Acute leukemia is often characterized by the presence of specific numerical and structural chromosome aberrations in the malignant clone. For example, malignant clones in AMLs commonly have numerical changes in chromosomes 5, 7, and 8 or contain transloca tions between chromosomes 8 and 21 (11-13, 23, 24). These chro mosomal changes confer a selective advantage to proliferate, resist differentiation, and/or survive apoptosis, and the malignant clone of myeloid origin becomes the dominant WBC type in the blood and marrow (25, 26). It is widely believed that cells of the lymphocytic lineage do not possess the same clonal abnormalities as those found in the myeloid lineage (27), but they may be present at far lower levels, undetectable by conventional methods. Because we prepared met- aphases from whole blood, it is possible that the specific chromosome aberrations we detected in the present study were in a small popula tion of myeloid cells. This is highly unlikely, however, because myeloid cells are terminally differentiated and will not divide in response to phytohemagglutinin or survive 72 h of culture. Because AML-specific changes will not confer a selective growth advantage on lymphoid cells (28), the presence of AML-specific chromosomal changes in lymphocytes in nondiseased individuals reflects the rate at which these events are being produced by genotoxic insults. Further more, because lymphocytes are much longer-lived than granulocytes, the production of aberrations in the former is cumulative over a period of months. The quantitation of AML-specific aberrations in peripheral blood lymphocytes may therefore provide a measure of damage and leukemia risk incurred by exposure to genotoxic agents such as benzene and radiation and to the environment in general. This idea is a subtle modification of the long-established approach of monitoring conventional chromosome aberrations in the peripheral blood lym phocytes of people exposed to potential carcinogens. The use of chromosome aberrations as biomarkers of early effect recently received substantial support from the findings of Hagmar et al. (6) and Bonassi et al. (7) that showed that high levels of chromo some aberrations in peripheral blood lymphocytes were associated with an increased risk of subsequently developing cancer, especially hematological malignancies (7). We reasoned that this approach could be improved in two ways: (a) by detecting oncogenic disease-specific aberrations; and (b) by using FISH, so that many more metaphases could be analyzed, thereby increasing the statistical power. As an initial test of this hypothesis, we have used FISH to study metaphase spreads prepared from the lymphocytes of workers exposed to ben zene. Chromosome painting by FISH allowed us to detect aberrations specific to AML, notably hyperdiploidy 8 and t(8;21). It also allowed us to study an average of 483 metaphases/individual, an order of magnitude more metaphases than is routinely scored by conventional cytogenetics. The results of this analysis showed that exposure to benzene was associated with increased levels of hyperdiploidy of chromosome 8, mainly in the form of trisomy 8, and translocations between chromosomes 8 and 21 in the lymphocytes of the exposed but otherwise healthy workers. Benzene exposure also increased the num ber of breaks in chromosome 8 and the number of translocations between chromosomes 8 and 21 and other chromosomes. Transloca tion (8;21) was increased as much as 15-fold in the highly exposed workers. Benzene exposure was associated not only with increased levels of specific aberrations in the groups as a whole but also with a dose- dependent increase in the number of individuals with detectable structural aberrations. This is illustrated by the fact that there was a dramatic increase in the percentage of the population with detectable translocations in the high-dose group. For example, t(8;21) was ob served in only 2% of control subjects but was detected in 27% of the high-dose group (Table 2). Likewise, the percentage of the population 185 Fig. 2. Southern blot to detect the presence of AML1/ETOchimeric cDNA derived from t(8;21) in the blood of workers exposed to high levels of benzene and unexposcd controls. Amplification and blotting of cDNA prepared from RNA isolated from the 11 benzene-exposed workers is shown in Lanes 3, 5, 10, 11, 14-18, 21, and 24. Five of these workers had t(8;21) detectable by FISH (Lanes 3, 5, 10, 11. and 24). A positive response is seen in Lane 24. showing that this worker had detectable AMLI/ETO fusion transcripts in his peripheral blood. RNA from the 10 matched unexposed controls was amplified and blotted in Lanes I. 2. 4, 8, 12, 13, 19, 20, 22, and 23. One of these unexposed workers had t(8;21) detectable by FISH, but a positive response was not observed (Lane 12). RNA isolated from two t(8;2i)-positive leukemia patients and the Kasumi-1 cell line [which harbors the t(8;21) translocation1was included in Lanes 6, 9. and 25, respectively, as positive controls. RNA from one 1(8:21-negativleukemia patient was included as a negative control (Lune 7}.The analysis was performed blind, so that the lanes were identified after the fact. 2179 DETECTION OF t(8;21) IN BENZENE-EXPOSED WORKERS with t(8;?) rose from 11 to 50<7r (Table 2). Thus, in the high-dose group, one-fourth to one-half of the population had detectable trans- locations by the FISH methodology described here, whereas a much smaller proportion (approximately one-tenth) had detectable translo study the blood smears. RNA was isolated from the blood smears of a subset of 21 exposed workers and matched controls. We were only able to detect identifiable AMLI-ETO fusion transcripts in the RNA isolated from the blood smear of the worker with the highest level of cations of any kind in the unexposed controls. Interestingly, increased levels of specific chromosome aberrations were correlated only with current exposure, reflective of exposures over the previous 6-12 months, and not to cumulative exposure levels measured in ppm-years. This would tend to indicate that the majority of the aberrations detected had a half-life of a few months at most and thus would not be detectable many years after exposure. We are currently conducting a study using the same methodology of workers with past exposure, but no current exposure, to high levels of benzene to expand on this finding. However, the present results are quite different from those reported in this same population using the gly- cophorin A mutation assay, which measures mutations in erythroid progenitors in the bone marrow, in which cumulative exposure but not current exposure correlated with increased mutant frequency (21). The fact that different cell types are affected in the production of glyco- t(8:21 ), and the only subject with reciprocal translocations detected by FISH. This fusion transcript was identical in sequence to that found in the Kasumi-1 cell line. This demonstrates that the translocation de tected by FISH was in-frame and of etiological significance to leukemogenesis. RT-PCR therefore confirmed the FISH analysis, show ing the power of this combined approach. RT-PCR and PCR have previously been used to detect transloca tions (9;22) and (14; 18) in unexposed individuals of different ages and in smokers (38, 39). Both translocations were found to increase with age. and the t( 14; 18) translocation was increased in cigarette smokers. We are not aware of any previous studies showing detectable t(8;21) by RT-PCR in nondiseased, untreated individuals. The results ob tained here and those described previously by others clearly demon strate the potential of RT-PCR in monitoring specific aberrations in populations exposed to suspected or established leukemogens. Future phorin A mutations and specific chromosome aberrations may explain this difference. The results are also inconsistent with the prevailing theory that all translocations are persistent stable aberrations. How ever, the recent findings of Spruill el al. (29) suggest that some translocation frequencies decline with time and are not completely persistent (30). Chromosome painting has previously been used to study humans and cell lines exposed to ionizing radiation (31, 32). The effects of radiation are considered to be random, so that the number of breaks and translocations in a particular chromosome is proportional to the size ofthat chromosome. Indeed, it has been argued that the effects of radiation on the whole genome can be studied using only two or three chromosomes (31). If breaks and subsequent translocations were random, one would expect that translocations between chromosomes 8 and 21 would occur 60 times less frequently than translocations between chromosome 8 and all of the other chromosomes and 19 times less frequently than translocations between chromosome 21 and other chromosomes. These values were calculated using the formulas devised by Savage and Papworth (33) and Lucas el al. (34) using the genomic contents of human chromosomes described by Morton (35). In fact, a striking selectivity for stable translocations between chro mosomes 8 and 21 was observed in controls as well as in exposed workers. In controls, the ratio of t(8;?):t(8;21) was only 3, not 60; and the ratio of t(21;?):t(8;21) was only 2, not 19 (Fig. 1). Thus, in unexposed controls, chromosome 8 was 10-20 times more likely to exchange with chromosome 21 than with other chromosomes. In the workers exposed to lower levels of benzene (31 ppm), the ratio of t(8;?):t(8;21) fell to 1 (0.042:0.042), and for t(21;?):t(8;21), it was even lower (0.4; 0.017:0.042). Similar ratios (1.125 and 0.5, respec tively) were observed in the highly exposed workers (>31 ppm). Thus, in the benzene-exposed workers, chromosome 8 was 38-60 times more likely to exchange with chromosome 21 than with other chromosomes, suggesting that some type of site-specific recombina studies are planned to collect samples from workers exposed to benzene and from the general population to examine the level of different translocations in their peripheral blood by RT-PCR. The mechanism by which benzene exposure leads to increased levels of translocations between chromosomes 8 and 21 is not clear at present. One possible mechanism may be related to topoisomerase II inhibition, because t(8:21) is a clonal abnormality associated with AML caused by topoisomerase-inhibiting drugs (28), and benzene metabolites are topoisomerase II poisons (40). We are currently per forming further analysis of the metaphase spreads collected from these populations to measure changes at chromosome loci known to contain topoisomerase recognition sites, including 1Iq23, and at sites on the genome also related to leukemia development such as 5q31 and 7q22. In summary, we have shown that benzene exposure is associated with markedly elevated levels of t(8;21) and of hyperdiploidy 8 and 21, in the circulating lymphocytes of otherwise healthy workers exposed to benzene compared to unexposed controls. This suggests a role for these aberrations in benzene-induced leukemia and that their detection in peripheral blood cells by chromosome painting may be useful biomarkers of increased risk for hematological malignancy for benzene and other potential leukemogens. Prospective studies are being considered to further test this hypothesis. ACKNOWLEDGMENTS We are indebted to the study subjects and the staff of the Chinese Academy of Preventive Medicine, Beijing and the Shanghai Hygiene and Anti-Epidemic Center. Shanghai. China for assistance and participation in this study. We gratefully acknowledge the technical assistance of Pravina Venkatesh and Weihong Guo. and we thank Prof. Rainer K. Sachs (University of California. Berkeley. CA) and Dr. Roben E. Tarone (National Cancer Institute) for expert statistical advice. tion or selection process must be occurring. However, the ratio of 38-60-fold in benzene-exposed workers was not significantly greater REFERENCES than the ratio of 10-20-fold found in controls. The increased frequency of t(8;21) in the workers highly exposed to benzene detected by FISH encouraged us to attempt to detect AMLOEETO fusion transcripts in the blood of these workers by RT-PCR. AMLOEand ETO are the two genes that are fused in the translocation (8;21). Only limited material was available for this analysis in the form of 4-year-old blood smears. However, we have recently been able to detect gene expression by RT-PCR in stored Guthrie card blood spots: therefore, we applied this same technology (36, 37) to 1. Galloway. S. M.. Berry, P. K., Nichols. W. W., Wolman, S. R., Soper, K. A., Stolley, P. D.. and Archer. P. Chromosome aberrations in individuals occupationally exposed to ethylene oxide, and in a large control population. Mutt.Res., 170: 55-74. 1986. 2. Littlefield. L. G.. and Joiner. E. E. Analysis of chromosome aberrations in lympho cytes of long-term heavy smokers. Mutt.Res.. 770: 145-150. 1986. 3. Tough. 1. M.. and Court Brown. W. M. Chromosome aberrations and exposure to ambient benzene. Lancet. 73X7: 684-685, 1965. 4. Fomi, A. Chromosome changes and ben/ene exposure. A review. Rev. Environ. Health.. J: 5-17, 1979. 5. Major, J.. Jakab. M.. Kiss. G., and Tompa, A. Chromosome aberration, sister- chromatid exchange, proliferative rale index, and serum thiocyanale concentration in smokers exposed to low-dose benzene. Environ. Mol. Mutagen., 23: 137-142. 1994. 2180 DETECTION OF 1(8:21) [N BENZENE-EXPOSED WORKERS 6. Hagmar, L., Brogger, A., Hansteen, I. L.. Heim, S., Hogstedt. B.. Knudsen, L., Lambert, B., Linnainmaa. K., Mitelman, F.. Nordenson. 1., Reuterwall. C.. Salomaa. S., Skerfving, S., and Sorsa, M. Cancer risk in humans predicted by increased levels of chromosomal aberrations in lymphocytes: Nordic Study Group on the Health Risk of Chromosome Damage. Cancer Res.. 54: 2919-2922. 1994. 7. Bonassi, S.. Abbondandolo. A.. Camurri, L... Dal PrL. .. De Ferrari, M., Degrassi, F.. Forni. A.. Lamberti. L.. Lando. C., Padovani. P.. Sbrana. I.. Vecchio. D.. and Puntoni. R. Are chromosome aberrations in circulating lymphocytes predictive of future cancer onset in humans? Preliminary results of an Italian cohort study. Cancer Genet. Cytogenet.. 79: 133-135, 1995. 8. Aksoy. M. Benzene Carcinogenicily, pp. 113-151. Boca Raton, FL: CRC Press, Inc., 1988. 9. Hayes, R. B.. Yin, S. N., Dosemeci. M., Li. G. L.. Wacholder, S., Travis. L. B., Li. C. Y., Rothman, N., Hoover, R. N., and Linet, M. S. Benzene and the dose-related incidence of hmatologie neoplasms in China. Chinese Academy of Preventive Medicine-National Cancer Institute Ben/ene Study Group. J. Nati. Cancer Inst.. 89: 1065-1071, 1997. 10. Rinsky. R. A.. Smith. A. B.. Homung. R. F.. Filloon. T. G., Young, R. J.. Okun, A. H.. and Landrigan, P. J. Benzene and leukemia. An epidemiologie risk assessment. N. Engl. J. Med.. 316: 1044-1050, 1987. 11. Pedersen-Bjergaard. J., and Rowley. J. D. The balanced and the unbalanced chromo some aberrations of acute myeloid leukemia may develop in different ways and may contribute differently to malignant transformation. Blood. H3: 2780-2786. 1994. 12. Sandberg, A. A. The Chromosomes in Human Cancer and Leukemia, 2nd d.p.p. 223-312. New York: Elsevier. 1990. 13. Kagan. J. Molecular biology of chromosomal aberrations in leukemia/lymphoma Hematol. Pathol.. 7: 159-201, 1993. 14. Nucifora, G., and Rowley. J. D. AML1 and the 8:21 and 3:21 translocations in acute and chronic myeloid leukemia. Blood. S6. 1-14. 1995. 15. Erickson. P.. Gao, J.. Chang, K. S., Look. T.. Whisenant. E.. Raimondi. S.. Lasher. R., Trujillo, J., Rowley. J.. and Drabkin. H. Identification of breakpoints in 1(8:21 ) acute myelogenous leukemia and isolation of a fusion transcript. AMLl/ETO. with simi larity to Drosophila segmentation gene. runt. Blood. HO: 1825-1831. 1992. 16. Downing. J. R.. Head, D. R., Curcio-Brint, A. M.. Hulshof, M. G.. Motroni, T. A.. Raimondi. S. C.. Carroll. A. J., Drabkin. H. A., Willman. C. Theil. K. S., Civin, C. I., and Erickson, P. An AMLl/ETO fusion transcript is consistently detected by RNA-based polymerase chain reaction in acute myelogenous leukemia containing the (8;2I)(q22;q22) translocation. Blood. 81: 2860-2865, 1993. 17. Gray. J. W.. and Pinkel. D. Molecular cytogenetics in human cancer diagnosis. Cancer (Phila.), 69: 1536-1542, 1992. 18. Rothman. N.. Li. G-L.. Dosemeci. M.. Bechtold. W.. Marti. G. E.. Wang. Y-Z.. Linet. M.. Xi, L-Q.. Lu. W., Smith. M. T.. Titenko-Holland. N., Zhang. L., Blot, W., Yin, S.. and Hayes. R. Hematotoxicity among Chinese workers heavily exposed to ben zene. Am. J. Ind. Med., 29: 236-246, 1996. 19. Rothman. N.. Smith, M. T., Hayes, R. B., Li. G. L.. Irons. R. D.. Dosemeci, M., Haas. R.. Stillman, W. S.. Linet. M.. Xi. L. Q.. Bechtold. W. E.. Wiemels, J.. Campleman. S.. Zhang. L.. Quintana. P. J., Titenko-Holland. N., Wang. Y. Z.. Lu. W.. Kolachana. P., Meyer. K. B.. and Yin. S. An epidemiologie study of early biologic effects of benzene in Chinese workers. Environ. Health Perspect.. 6 (Suppl. 1041; 1365-1370. 19%. 20. Dosemeci, M., Li, G. L., Hayes. R. B., Yin, S. N., Linet, M., Chow. W. H.. Wang. Y. Z.. Jiang. Z. L., Dai. T. R., Zhang, W. U.. Chao. X-J, Ye, P-Z.. Kou. Q-R., Fan. Y-H., Zhang. X-C. Lin, X-F., Meng, J-F., Zho, J-S., Wacholder, S., Kneller. R., and Blot, W. J. Cohort study among workers exposed to benzene in China. II. Exposure assessment. Am. J. Ind. Med.. 2ft: 401-411. 1994. 21. Rothman, N., Haas, R.. Hayes, R. B.. Li. G. L.. Wiemels. J.. Campleman, S., Quintana, P. J., Xi, L. J.. Dosemeci. M., Titenko-Holland. N.. Meyer. K. B.. Lu. W., Zhang. L.. Bechtold. W.. Wang. Y. Z.. Kolachana. P., Yin. S. N.. Blot, W. J., and Smith. M. T. Benzene induces gene-duplicating but not gene-inactivating mutations at the glycophorin A locus in exposed humans. Proc. Nati. Acad. Sci. USA. 92: 4069-4073. 1995. 22. Asou. H., Tashiro. S.. Hamamoto. K., Otsuji, A.. Kita, K.. and Kamada. N. Estab lishment of a human acute myeloid leukemia cell line (Kasumi-1) with 8:21 chro mosome translocation. Blood, 77: 2031-2036. 1991. 23. Le Beau. M. M.. Espinosa. R. D.. Neuman. W. L.. Stock. W.. Roulston. D.. Larson. R. A.. Keinanen. M., and Wcstbrook. C. A. Cytogenetic and molecular delineation of the smallest commonly deleted region of chromosome 5 in malignant myeloid diseases. Proc. Nati. Acad. Sci. USA. 90: 5484-5488. 1993. 24. Le Beau. M. M.. Espinosa. R. D.. Davis. E. M.. Eisenbart. J. D., Larson. R. A., and Green. E. D. Cytogenetic and molecular delineation of a region of chromosome 7 commonly deleted in malignant myeloid diseases. Blood, XX: 1930-1935. 1996. 25. Russell. N. H. Biology of acute leukaemia. Lancet. 349: 118-122. 1997. 26. Irons. R. D.. and Stillman. W. S. The process of leukemogenesis. Environ Health Perspecl.. 6 (Suppl. 104).- 1239-1246. 1996. 27. Knuutila. S.. Teerenhovi. L.. Larramendy. M. L., Elonen. E.. Franssila. K. ().. Nylund. S. J.. Timonen. T.. Heinonen. K.. Mahlamaki. E.. Winqvist. R.. and Ruutu. T. Cell lineage involvement of recurrent chromosomal abnormalities in hmatologieneo plasms. Genes Chromosomes Cancer. 10: 95-102. 1994. 28. Pedersen-Bjergaard. J.. Pedersen. M., Roulston. D.. and Philip. P. Different genetic pathways in leukemogenesis for patients presenting with therapy-related myclodysplasia and therapy-related acute myeloid leukemia. Blood. H6: 3542-3552. 1995. 29. Spruill. M. D.. Ramsey, M. J., Swiger. R. R.. Nath, J.. and Tucker. J. D. The persistence of aberrations in mice induced by y radiation as measured by chromosome painting. Mutt.Res.. 356: 135-145, 1996. 30. Matsumoto. K.. Ramsey. M. J., Nelson. D. O.. and Tucker. J. D. Persistence of radiation-induced translocations in human peripheral blood determined by chromo some painting. Radial. Res., in press. 1998. 31. Lucas. J. N.. Awa. A.. Straume. T.. Poggensee. M.. Kodama. Y.. Nakano. M., Ohiaki. K.. Weier. H. U.. Pinkel. D.. Gray. J.. and Liulefield. G. Rapid translocation frequency analysis in humans decades after exposure to ionizing radiation. Int. J. Radial. Bio]., 62: 53-63. 1992. 32. Matsuoka. A.. Tucker. J. D.. Huyashi, M.. Yama/aki. N.. and Sofuni. T. Chromosome painting analysis of X-ray-induced aberrations in human lymphocytes in v//m. Mutagenesis, 9: 151-155. 1994. 33. Savage. J. R., and Papworth. D. G. Frequency and distribution studies of asymmet rical trr.vH.vsymmetrical chromosome aberrations. Mutt.Res., 95: 7-18. 1982. 34. Lucas. J. N.. Tenjin. T.. Straume. T.. Pinkel. D.. Mixire. D. D.. Litt. M., and Gray. J. W. Rapid human chromosome aberration analysis using fluorescence in situ hybridi/alion [published erratum appears in Int. J. Radit.Biol., 56 (Suppl. 2): 201, 1989|. Int. J. Radial. Bio]., 56: 35-44, 1989. 35. Morton. N. E. Parameters of Ihe human genome. Proc. Nati. Acad. Sci. USA. AS: 7474-7476, 1991. 36. Melo. J. V., Yan. X. H.. Diamond. J.. Lin. F.. Cross, N. C.. and Goldman. J. M. Reverse transcription/polymerase chain reaction (RT/PCR) amplification of very small numbers of transcripts: the risk in misinterpreting negative results. Leukemia (Baltimore). IO: 12I7-122I. 1996. 37. Zhang. Y. H.. and McCabc. E. R. RNA analysis from newborn screening dried blood specimens. Hum. Genet., 9:311-314. 1992. 38. Liu. Y.. Hernandez. A. M.. Shibata. D.. and Cortopassi. G. A. BCL2 translocation frequency rises with age in humans. Proc. Nati. Acad. Sci. USA, 91: 8910-8914. 1994. 39. Biernaux. C.. Loos. M.. Sels, A., Huez. G., and Stryckmans. P. Detection of major bcr-ahl gene expression at a very low level in blood cells of some healthy individuals. Blood. 6:3118-3122. 1995. 40. Chen. H., and Eastmond, D. A. Topoisomerase inhibition by phenolic metabolites: a potential mechanism for benzene's clastogenic effects. Carcinogenesis (Lond). 16: 2301-2307. 1995. 2181