Document G6m59JbnKYoVxqgMNKO4eON7r

Toxicology and Applied Pharmacology 206 (2005) 237 245 www.elsevier.com/locate/ytaap Molecular biomarkers for the study of childhood leukemia Martyn T. Smitha,*, Cliona M. McHalea, Joseph L. Wiemelsb, Luoping Zhanga, John K. Wienckeb, Shichun Zhengb, Laura Gunna, Christine F. Skibolaa, Xiaomei Maa, Patricia A. Bufflera aDivision of Environmental Health Sciences, School of Public Health, University of California, 140 Warren Hall, Berkeley, CA 94720-7360, USA bDepartment of Epidemiology and Biostatistics, University of California Medical School, San Francisco, CA 94143-0560, USA Received 29 July 2004; accepted 22 November 2004 Available online 13 April 2005 Abstract Various specific chromosome rearrangements, including t(8;21), t(15;17), and inv(16), are found in acute myeloid leukemia (AML) and in childhood acute lymphocytic leukemia (ALL), t(12;21) and t(1;19) are common. We sequenced the translocation breakpoints of 56 patients with childhood ALL or AML harboring t(12;21), t(8;21), t(15;17), inv(16), and t(1;19), and demonstrated, with the notable exception of t(1;19), that these rearrangements are commonly detected in the neonatal blood spots (Guthrie cards) of the cases. These findings show that most childhood leukemias begin before birth and that maternal and perinatal exposures such as chemical and infectious agents are likely to be critical. Indeed, we have reported that exposure to indoor pesticides during pregnancy and the first year of life raises leukemia risk, but that later exposures do not. We have also examined aberrant gene methylation in different cytogenetic subgroups and have found striking differences between them, suggesting that epigenetic events are also important in the development of some forms of childhood leukemia. Further, at least two studies now show that the inactivating NAD(P)H:quinone acceptor oxidoreductase (NQO1) C609T polymorphism is positively associated with leukemias arising in the first 1 2 years of life and polymorphisms in the 5,10-methylenetetrahydrofolate reductase (MTHFR) gene have been associated with adult and childhood ALL. Thus, low folate intake and compounds that are detoxified by NQO1 may be important in elevating leukemia risk in children. Finally, we are exploring the use of proteomics to subclassify leukemia, because cytogenetic analysis is costly and time-consuming. Several proteins have been identified that may serve as useful biomarkers for rapidly identifying different forms of childhood leukemia. D 2005 Elsevier Inc. All rights reserved. Keywords: Chromosome translocations; Neonatal blood spots; Cytogenetics; Proteomics; Leukemia; Children; Biomarkers; Molecular epidemiology; Pesticides Introduction to childhood leukemia Leukemia, the number one disease killer of children less than 14 years of age in the United States, represents around 31% of all cancer cases among children in this age group. The incidence of leukemia is 10-fold lower in children than adults, with approximately 3250 children and adolescents younger than 20 years of age diagnosed each year in the US, of which 2400 are acute lymphoblastic leukemia (ALL). * Corresponding author. Fax: +1 510 642 0427. E-mail address: martynts@berkeley.edu (M.T. Smith). 0041-008X/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.taap.2004.11.026 Leukemia is a heterogeneous disease, characterized by the dysregulated proliferation of blood precursor cells of myeloid or lymphoid origin. It can be classified as acute (low level of differentiation) or chronic (high level of differentiation) and can be further classified by cytogenetic subtype. For example, t(12;21) which generates the TEL-AML1 fusion gene occurs in 25% of patients with common ALL (cALL), a subtype of ALL with peak incidence between ages 2 and 5 (Romana et al., 1995). Translocation t(1;19) and high hyperdiploidy (>50 chromosomes) are also common in childhood ALL (Harrison, 2001), while various specific chromosome rearrangements, including t(8;21), t(15;17), and inv(16), are found in acute myeloid leukemia (AML) (Hall, 2001). 238 M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 Overview of biomarker types Besides the various cellular and molecular characteristics discussed above, leukemias show further heterogeneity in response to therapy and subsequent risk of relapse. There is a clear need to develop techniques to delineate biomarkers of susceptibility and optimized therapy on an individual or subtype basis. This review outlines recent advances in molecular biomarkers of childhood leukemia. The term ``genotype'' will be discussed in two contexts. In the constitutive context, heritable germline susceptibility genes including rare mutations in high penetrance genes such as Breast Cancer 1 (BRCA1), retinoblastoma (Rb), Ataxia telangiectasia mutated (ATM), and/or common polymorphisms in low risk genes such as MTHFR, NQO1, glutathione S-transferase (GST) modulate an individual's risk of developing cancer. The tumor genotype describes mutations (at the gene or chromosomal level) that characterize the tumor, e.g., t(12;21), monosomy 7, or Ras mutation, and may evolve with disease progression. Epigenetic changes at the level of gene-specific methylation in leukemia will be discussed and preliminary data on the emerging field of proteomic biomarker discovery in leukemia will be outlined. The Northern California Childhood Leukemia Study Much of the research discussed in this review was based on samples from the Northern California Childhood Leukemia Study (NCCLS), a population-based case-control study begun at the University of California, Berkeley, in 1995 with the aim of investigating the relationship between environmental exposures and genetic risk factors for childhood leukemia (Fig. 1). To date, biological specimens (from peripheral blood, bone marrow and/or buccal cells), from over 800 cases occurring at 9 hospitals covering 35 counties, and a similar number of controls, have been processed and stored in our laboratories. Conventional cytogenetics and fluorescent in situ hybridization (FISH) analyses of NCCLS ALL cases show similar genotype proportions as those reported in the literature, with hyperdiploidy (29%) and t(12;21) (16%) being the most prevalent subtypes. Gene and protein expression profiling One approach to stratifying pediatric leukemia and discovering biomarkers is the use of gene expression profiling by microarray analysis to identify subtype discriminating genes. This typically involves statistical determination of patterns of gene expression against a background of global gene expression (30,000 genes). A recent study demonstrated the ability of such an approach to diagnose pediatric ALL subtypes including TEL-AML1 and hyperdiploidy, by virtue of unique gene expression profiles in less than 100 genes, with an overall accuracy of 97% (Ross et al., 2003). While this approach is very useful for gaining insight into the underlying biology of leukemia subtypes and developing targeted therapies, global gene expression analysis by microarray is not amenable to the clinical setting. Drawbacks include the need for RNA isolation from an enriched leukemic cell population to minimize background, inherent data complexity, and expense. RNA is also an inherently unstable molecule, making sample processing difficult in the typical hospital setting and gene expression may not reveal the true phenotype which is based on protein. For these reasons, we and other groups have begun exploring the use of proteomics to subclassify leukemia. Our approach involves the use of the ProteinChip\ system from Ciphergen, which combines affinity chromatography and mass spectrometry, allowing the rapid capture and identification of proteins based on their biochemical properties. Using this system, we initially analyzed crude cell lysates from 3 childhood leukemia cell lines: 697 which harbors t(1;19); REH which has t(12;21); and MV4;11 with t(4;11) (Gunn et al., 2004). Analysis of proteins retained on the array surface revealed 22 peaks using the Biomarker Wizard feature of the ProteinChip\ software. Average peak intensities were compared among the 3 cell lines. 697 and REH, which are both acute lymphoblastic cell lines, had few differences between them. However, one protein of around 4 kDa was expressed in 697 but not in REH (Fig. 2, Peak A). This may be a marker of ALL cells harboring t(1;19). When the spectra of 697 and REH cells were compared with that of MV4;11, which is a biphenotypic myelomonocytic cell line, and the Kasumi AML cell line, several major differences were noted. One protein of 8.3 kDa was expressed at much higher levels in the ALL cell lines than in MV4;11 or Kasumi cells (Fig. 2, Peak B). This protein is currently being identified and may be a marker of ALL. Yet another protein of around 4 kDa was expressed at a >50-fold higher level in MV4;11 cells (Fig. 2, Peak C). This protein may serve as a useful biomarker for rapidly identifying mixed lineage leukemias harboring translocations involving the MLL gene at 11q23. We are currently attempting to identify the differentially expressed proteins and analyzing the cell lysates on additional chip surfaces. Based on these findings, we also propose to analyze the protein spectra of leukemic cells from bone marrow of childhood leukemia cases to further test the idea that different types of leukemia may be differentiated using ProteinChip\ technology. Chromosome translocations as biomarkers: backtracking childhood leukemia to birth As discussed above, specific chromosome translocations characterize certain subtypes of childhood leukemia and themselves serve as markers of diagnosis, disease progres- M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 239 Fig. 1. Northern California Childhood Leukemia Study (NCCLS). NCCLS is a population-based case-control study begun at the University of California, Berkeley, in 1995 with the aim of investigating the relationship between environmental exposures and genetic risk factors for childhood leukemia. To date, biological specimens (from peripheral blood, bone marrow and/or buccal cells), from over 800 cases occurring at 9 hospitals covering 35 counties and a similar number of controls, have been processed and stored in our laboratories. sion, and relapse. Indeed, detection of translocations by quantitative PCR is widely used to determine minimal residual disease and early pre-clinical relapse (Moppett et al., 2003). This is possible because the signature fusion gene sequence arising from a specific translocation is consistent across patients at the level of RNA, allowing the use of a generic assay for each translocation type. However, as the chromosome breaks occur for the most part randomly in very large (up to 160 Kb) introns, the fusion gene sequences at the genomic DNA level are clonotypic or patient-specific. This characteristic has been exploited in the process of backtracking where certain translocations in childhood 240 M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 Fig. 2. Surface-enhanced laser desorption ionization time of flight (SELDI-TOF) mass spectrometry proteomics of childhood leukemia cell lines. The ProteinChip\ system from Ciphergen, which combines affinity chromatography and mass spectrometry allowing the rapid capture and identification of proteins based on their biochemical properties, was used to determine proteins differentially expressed among leukemic cell lines. 22 peaks from proteins retained on the IMAC-Cu array surface were identified using the Biomarker Wizard feature of the ProteinChip\ software. Resulting peaks for an 8.3-kDa protein which was expressed >10-fold more highly in 697 and REH, two acute lymphoblastic cell lines than in MV4;11, which is a biphenotypic myelomonocytic cell line, are shown. This peak is currently being identified. leukemia were shown to arise in utero, preceding the disease by as long as 14 years. The process, illustrated in Fig. 3, involves sequencing genomic DNA from patients at diagnosis, developing clonotypic, highly-sensitive, nested PCR assays, and testing for the presence of the patientspecific fusion gene sequences on Guthrie cards, blood spots taken at birth for routine screening for metabolic disorders. The first direct evidence for an in utero origin of specific translocations was for t(4;11) in children aged 5 months to 2 years (Gale et al., 1997). Following up on this pioneering work in Greaves' laboratory, we sequenced the translocation breakpoints of 56 patients with childhood ALL or AML harboring t(12;21), t(8;21), t(15;17), inv(16), and t(1;19), and demonstrated, with the notable exception of t(1;19), that these rearrangements are commonly detected in the neonatal blood spots (Guthrie cards) of the cases (see Table 1, McHale et al., 2003a, 2003b; Wiemels et al., 1999a, 2002b). Indirect support for a prenatal origin of several ALL cases lacking specific chromosomal translocations was provided by tracking clonotypic immunoglobulin heavy chain (IgH) and T cell receptor (TCR) rearrangements including cALL (Yagi et al., 2000), T-ALL (Fasching et al., 2000), and hyperdiploid ALL (Taub et al., 2002). Hyperdiploid ALL was also recently demonstrated to have an in utero origin by the identification of CD34(+)/CD19(+) B-lineage progenitor cells with triploid chromosomes in the stored cord blood of an individual who subsequently developed hyperdiploid ALL (Maia et al., 2004b). While t(12;21), characteristic of common acute lymphoblastic leukemia (c-ALL), has a peak incidence in children 2 5 years of age (Romana et al., 1995), other translocations of demonstrated prenatal origin such as t(8;21), t(15;17), and inv(16) characteristic of different AML subtypes occur more frequently in young adults (age 20 40) (Moorman et al., 2001). Indeed, in our M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 241 2003; McHale and Smith, 2004). Overall, these findings show that most childhood leukemias begin before birth and that maternal and perinatal exposures such as chemical and infectious agents are likely to be critical. Indeed, we have reported that exposure to indoor pesticides during pregnancy raises leukemia risk (Ma et al., 2002a). Chromosome translocations in normal individuals Translocations t(8;21) and t(12;21) have been shown to occur in the normal population (before birth) at a frequency that is 100-fold greater than the risk of developing the corresponding leukemia (Mori et al., 2002). Other translocations including t(9;22) and t(14;18) (Janz et al., 2003) have been reported in the literature and we have observed t(6;11), t(15;17) (unpublished findings), and del(1p) (Curry and Smith, 2003) in small numbers of normal individuals. The predictive value of these combined observations is unknown. The low concordance rate (10%) observed in cALL in twins (Greaves and Wiemels, 2003), and data from animal models (Bernardin et al., 2002; Castilla et al., 1999), imply that the presence of translocations alone is not sufficient to cause leukemia. Translocations may arise in cells with an environment non-permissive for neoplastic progression because of inappropriate lineage or differentiation stage, or lack of complementary mutations at the spontaneous or germline genotype level. Fig. 3. Backtracking TEL-AML1 rearrangement in newborn blood spots. Genomic DNA from each leukemic patient is amplified by long-distance PCR and the unique patient gene rearrangement is sequenced. A nested PCR assay is designed for the breakpoint junction and used to determine if the sequence is present Guthrie Card (heel-prick at birth) DNA. studies, protracted postnatal latencies of over 10 years were demonstrated. The biological significance of such protracted natural histories in childhood leukemia has been reviewed (Maia et al., 2004a). It remains to be determined if leukemias presenting in young adulthood with these translocations also have a prenatal origin. We and others have recently reviewed the origins of chromosomal translocations in childhood leukemia (Greaves and Wiemels, Using genetics to find environmental risk factors for childhood leukemia Aside from benzene, radiation, and chemotherapy (Pedersen-Bjergaard, 2002), there is currently little direct evidence linking environmental exposure with leukemia. Most likely, gene environment interactions modulate risk of leukemia development and progression. This is exemplified by Greaves' infectious hypothesis, which proposes that infectious exposures, subsequent to the in utero acquisition of a leukemia translocation, may result in highly dysregulated responses in susceptible individuals (those lacking important infection-driven modulation of the naive immune system in infancy). This could drive pre-leukemic Table 1 Chromosome abnormalities arising in utero in childhood leukemia Abnormality Reference t(4;11) t(12;21) t(8;21) t(15;17) inv(16) Hyperdiploidy t(1;19)a Gale et al., 1997 McHale et al., 2003b; Wiemels et al., 1999a Wiemels et al., 2002b McHale et al., 2003a McHale et al., 2003a Maia et al., 2003; Panzer-Grumayer et al., 2002 Wiemels et al., 2002a a Only rarely arises in utero. 242 M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 clones to fully disseminated leukemia (Greaves, 1997). In support of this theory are the findings that lifestyle factors, such as exposure to multiple infections such as in daycare settings (Ma et al., 2002b), and genetic factors such as variant human leukocyte antigen (HLA) alleles that influence the immune system (Taylor et al., 2002), may reduce risk. In light of the difficulties in assessing risks from specific exposures, genotype association studies in the case-control setting may be useful for the ascertainment of potential gene environment interactions. Single nucleotide polymorphisms (SNPs) are common DNA sequence variations among individuals, which may underline differences in response to exposures. Genotype association studies include genotyping of an individual candidate gene, e.g., NQO1, or a candidate pathway, e.g., folate metabolism in the members of a study population. DNA pools, compiled from the DNA of many individuals, reduce the genotyping effort and can be used to analyze candidate pathways, e.g., cell cycle, DNA repair, apoptosis, or for genome scanning, but are quite costly ($2 million) and require large amounts of sample material (50 Ag DNA per person). New methods involving multiplex analysis of hundreds or thousands of SNPs are becoming available from several companies and will soon be widely employed. NQO1 and leukemia risk NQO1 (formerly DT-diaphorase, diaphorase 4; EC 1.6.99.2) is a cytosolic flavoenzyme (homodimer of 273 residues) which catalyzes the obligatory two-electron reduction of quinones to hydroquinones, thus protecting against oxidative stress and toxicity of quinones and other electrophiles. NQO1 is highly inducible. With regard to the functional genomics of NQO1, the homozygous 609C Y T mutation (187Pro Y Ser) is associated with lack of activity and protein. NQO1 609CC has high activity, 609CT has low to intermediate activity, while 609TT has no activity. The homozygous mutation rates among Caucasian/AfricanAmerican, Hispanic, and Asian populations are 5%, 15%, and 20%, respectively. Table 2 summarizes studies examining the effect of the inactivating NQO1 C609T polymorphism on leukemia Table 2 NQO1 C609T and leukemia risk NQO1 study Benzene hematoxicity (Rothman et al., 1997) Therapy-related AML (Larson et al., 1999) Adult de novo AML (Smith et al., 2001) Childhood ALL (Krajinovic et al., 2002) Infant leukemia (Wiemels et al., 1999b) Infant leukemia (Smith et al., 2002) Infant t(4;11) (Smith et al., 2002) OR (95% CI) 2.4 (1.0 5.7) 1.4 1.6 1.5 (1.1 1.9) 1.7 (1.2 2.4) 2.5 (1.1 6.0) 2.5 (1.1 5.7) 7.7 (1.7 V) risk. A 2.4-fold increased risk of benzene-induced hematoxicity was shown in individuals lacking NQO1 activity in a case-control study of occupational benzene exposure in Shanghai, China (Rothman et al., 1997). Adult individuals with inactivating mutation were shown to be at increased risk of developing therapy-related acute myeloid leukemia (t-AML) (OR 1.4 1.6), probably through increased vulnerability to chemotherapeutic carcinogens (Larson et al., 1999). Similarly, an increased risk of de novo adult AML was demonstrated in a large casecontrol study (Smith et al., 2001). Three studies have shown that the inactivating NQO1 C609T polymorphism is positively associated with childhood leukemias, especially those arising in the first 1 2 years of life, particularly those with t(4;11) (Krajinovic et al., 2002; Smith et al., 2002; Wiemels et al., 2001). These data suggest that compounds, which are substrates for NQO1 or cause oxidative stress, are causative factors in producing leukemia. Such compounds include benzene, phenol, hydroquinone, quinones, azo, and nitro compounds (the main source of these compounds is diet) (Smith, 1996, 1999). Others, potentially metabolized by NQO1, include dietary flavonoids, which are topoisomerase II inhibitors and have been linked with infant leukemia (Ross, 1998). Oxidative stress may also be generated by inflammation and low antioxidant intake, and the effects of low NQO1 activity may be indicative of their role in the progression of leukemia. Folate metabolism and leukemia risk Folate is critical in DNA synthesis and repair and in methylation processes. It is particularly important in cells with high turnover rates and thus having the greatest requirement for DNA synthesis. Deficiencies in folate contribute to DNA damage and impair DNA synthesis. Folate or one-carbon metabolism is channeled into two major pathways: purine and thymidine synthesis for DNA formation and repair, and synthesis of methionine and Sadenosylmethionine for protein and polyamine synthesis and methylation reactions. Several functional polymorphisms have been identified in the folate pathway. The folate-metabolizing gene 5,10methylenetetrahydrofolate reductase (MTHFR) shunts methyl groups between the DNA synthesis branch and the methylation branch. The C677T low function variant shunts one-carbon groups towards thymidine and purine synthesis, potentially protecting against DNA strand breaks and translocation formation. An association between polymorphic variants at positions C677T and A1298C in MTHFR and a decreased risk of adult ALL was reported by our group (Skibola et al., 1999, 2002). To confirm the role of the folate pathway in leukemia susceptibility, polymorphisms were examined in 3 additional genes involved in folate metabolism (Skibola M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 243 et al., 2002). Thymidylate synthase (TS) plays a critical role in maintaining a balanced supply of deoxynucleotides required for DNA synthesis. TS is polymorphic at a unique tandem repeat sequence site in the 5V untranslated region, containing either 2 (2R) or 3 (3R) 28-bp repeats (Kaneda et al., 1987). Heterozygotes (2R3R) exhibited a 2.8-fold reduction in ALL risk while the TS 3R3R genotype conferred an even greater level of protection (Skibola et al., 2002). Cytosolic serine hydroxymethyltransferase (SHMT1) catalyzes 2 reactions in the folate pathway and may play an important role in maintaining one-carbon unit homeostasis (Girgis et al., 1998; Stover and Schirch, 1991). A C1420T polymorphism, resulting in reduced plasma and red blood cell folate thereby mimicking folate deficiency, exists in the SHMT1 gene (Heil et al., 2001). Individuals heterozygotic for SHMT1 C1420T exhibited a 2.1-fold decrease in ALL risk, whereas homozygotes had a further reduction (3.3-fold) in risk (Skibola et al., 2002). Overall, this research revealed an association between alterations in the folate metabolic pathway affecting thymidylate synthesis and ALL that may underscore the widespread importance of compromised DNA fidelity and insufficient folate pools in the pathogenesis of adult ALL. Low intake of folic acid and folate cofactors may exacerbate ALL risk in individuals with ``high-risk'' genotypes. A combination of unfavorable genotypes, diet, and vitamin B status may conceivably be the key factor in susceptibility to adult ALL that may also hold true for certain subtypes of childhood ALL. Two studies have addressed the role of MTHFR polymorphisms in modulating childhood ALL risk (Krajinovic et al., 2004; Wiemels et al., 2001). Wiemels et al. examined the role of MTHFR polymorphisms in molecularly defined subtypes of childhood ALL and showed a significant association between C677T and MLL translocations as well as hyperdiploidy (Wiemels et al., 2001). The association between MTHFR and childhood ALL was confirmed in a case-control study (270 ALL patients and 300 controls) and furthermore showed a potential gene environment link between the protective effect of the MTHFR variant and folate intake. Indeed, maternal folate supplementation in pregnancy was recently shown to protect against childhood ALL in a case-control study (83 cases and 166 controls) of children in Australia (Thompson et al., 2001). Folate and fragile histidine locus (FHIT) gene methylation Besides a crucial role in DNA synthesis, folate is involved in the formation of S-adenosylmethionine, a universal methyl donor critical for DNA and protein methylation throughout the body. Aberrant DNA methylation, including global hypomethylation and gene-specific hypermethylation at cytosine-guanine (CpG) islands in gene promoters, may result in dysregulated gene expression. We examined aberrant gene methylation in different cytogenetic subgroups of our NCCLS population and found striking differences between them, suggesting that epigenetic events are also important in the development of some forms of childhood leukemia. The FHIT gene encodes a tumor suppressor gene with proapoptotic function whose aberrant expression was previously reported in leukemia (Peters et al., 1999; Yang et al., 1999). FHIT is hypermethylated in a promoter CpG island in several cancer types (Tanaka et al., 1998; ZochbauerMuller et al., 2001). Our study demonstrated aberrant methylation of FHIT as an important mechanism for inactivation of this tumor suppressor gene in hematological malignancies (Zheng et al., 2004). FHIT methylation was more common in pediatric ALL (27.4%) than AML (5.9%) and was strongly associated with the hyperdiploid subtype (55.4%) and inversely associated with chromosomal translocations. Need for a childhood leukemia consortium based on the Interlymph Consortium model The studies to date on genetic risk for childhood leukemia have all been quite small and even our study of over 800 cases and the British study of over 1000 cases are too small to effectively study subtypes or gene environment interactions. A consortium needs to be built that will allow thousands of cases to be studied along with similar numbers of controls. A suitable model could be the Interlymph Consortium, which is an international collaboration of 17 non-Hodgkin lymphoma (NHL) studies in 12 countries with over 14,000 cases and extensive environmental data. The scale of this study overcomes limitations of smaller molecular cancer epidemiology studies and enhances power for analyses of gene environment interactions. Further, it allows the study of NHL subtypes based on molecular and genetic criteria. A similar approach should be taken for childhood leukemia allowing us to determine if specific constitutive genotypes are related to certain subtypes of the disease. Conclusions In summary, it has been shown that most childhood leukemias harboring translocations begin before birth. These translocations are present at birth at 100 times the leukemia rate, showing that additional changes are also critical and that screening for translocations at birth is not going to be a suitably predictive biomarker for the disease. It does, however, tell us that the prenatal period is critical for the initiation of childhood leukemia and that exposures during pregnancy should be further examined by epidemiologists. Indeed, in our own study, preliminary findings show that indoor pesticide exposure during 244 M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 pregnancy raises risk of childhood leukemia. The exact nature of the pesticides responsible or other chemical components in the commercial mixtures used are being investigated further. With regard to genetic susceptibility, studies to date have shown that the NQO1 C609T polymorphism raises risk of leukemia especially in infants. Polymorphisms in the folate metabolizing pathway may also be involved, showing that folate and perhaps other dietary components are important risk factors. International collaboration is needed to produce large studies so that these associations can be definitively examined in the case-control setting. Acknowledgments This work was supported by National Institute of Environmental Health Sciences Grant P42ES04705 and ROI ES0098137. References Bernardin, F., Yang, Y., Cleaves, R., Zahurak, M., Cheng, L., Civin, C.I., Friedman, A.D., 2002. TEL-AML1, expressed from t(12;21) in human acute lymphocytic leukemia, induces acute leukemia in mice. Cancer Res., 62, 3904 3908. Castilla, L.H., Garrett, L., Adya, N., Orlic, D., Dutra, A., Anderson, S., Owens, J., Eckhaus, M., Bodine, D., Liu, P.P., 1999. The fusion gene Cbfb-MYH11 blocks myeloid differentiation and predisposes mice to acute myelomonocytic leukaemia. Nat. Genet., 23, 144 146. Curry, J.D., Smith, M.T., 2003. Measurement of SIL-TAL1 fusion gene transcripts associated with human T-cell lymphocytic leukemia by realtime reverse transcriptase-PCR. Leuk. Res., 27, 575 582. Fasching, K., Panzer, S., Haas, O.A., Marschalek, R., Gadner, H., PanzerGrumayer, E.R., 2000. Presence of clone-specific antigen receptor gene rearrangements at birth indicates an utero origin of diverse types of early childhood acute lymphoblastic leukemia. Blood, 95, 2722 2724. Gale, K.B., Ford, A.M., Repp, R., Borkhardt, A., Keller, C., Eden, O.B., Greaves, M.F., 1997. Backtracking leukemia to birth: identification of clonotypic gene fusion sequences in neonatal blood spots. Proc. Natl. Acad. Sci. U.S.A., 94, 13950 13954. Girgis, S., Nasrallah, I.M., Suh, J.R., Oppenheim, E., Zanetti, K.A., Mastri, M.G., Stover, P.J., 1998. Molecular cloning, characterization and alternative splicing of the human cytoplasmic serine hydroxymethyltransferase gene. Gene, 210, 315 324. Greaves, M.F., 1997. Aetiology of acute leukaemia. Lancet, 349, 344 349. Greaves, M.F., Wiemels, J., 2003. Origins of chromosome translocations in childhood leukaemia. Nat. Rev. Cancer, 3, 639 649. Gunn, L., Shiao, R.T., Hegedus, C.M., Skibola, C.F., Smith, M.T., 2004. Proteomic analysis of childhood leukemia cell lines. Proceedings of the American Association for Cancer Res., 95th Annual Meeting 45, Abstract #3542. Hall, G.W., 2001. Childhood myeloid leukemias. Best Pract. Res. Clin. Haematol., 14, 573 591. Harrison, C.J., 2001. Acute lymphoblastic leukaemia. Best Pract. Res. Clin. Haematol., 14, 593 607. Heil, S.G., Van der Put, N.M., Waas, E.T., den Heijer, M., Trijbels, F.J., Blom, H.J., 2001. Is mutated serine hydroxymethyltransferase (SHMT) involved in the etiology of neural tube defects? Mol. Genet. Metab., 73, 164 172. Janz, S., Potter, M., Rabkin, C.S., 2003. Lymphoma- and leukemia- associated chromosomal translocations in healthy individuals. Genes Chromosomes Cancer, 36, 211 223. Kaneda, S., Takeishi, K., Ayusawa, D., Shimizu, K., Seno, T., Altman, S., 1987. Role in translation of a triple tandemly repeated sequence in the 5V-untranslated region of human thymidylate synthase mRNA. Nucleic Acids Res., 15, 1259 1270. Krajinovic, M., Sinnett, H., Richer, C., Labuda, D., Sinnett, D., 2002. Role of NQO1, MPO and CYP2E1 genetic polymorphisms in the susceptibility to childhood acute lymphoblastic leukemia. Int. J. Cancer, 97, 230 236. Krajinovic, M., Lamothe, S., Labuda, D., Lemieux-Blanchard, E., Theoret, Y., Moghrabi, A., Sinnett, D., 2004. Role of MTHFR genetic polymorphisms in the susceptibility to childhood acute lymphoblastic leukemia. Blood, 103, 252 257. Larson, R.A., Wang, Y., Banerjee, M., Wiemels, J., Hartford, C., Le Beau, M.M., Smith, M.T., 1999. Prevalence of the inactivating 609C Y T polymorphism in the NAD(P)H:quinone oxidoreductase (NQO1) gene in patients with primary and therapy-related myeloid leukemia. Blood, 94, 803 807. Ma, X., Buffler, P.A., Gunier, R.B., Dahl, G., Smith, M.T., Reinier, K., Reynolds, P., 2002a. Critical windows of exposure to household pesticides and risk of childhood leukemia. Environ. Health Perspect., 110, 955 960. Ma, X., Buffler, P.A., Selvin, S., Matthay, K.K., Wiencke, J.K., Wiemels, J.L., Reynolds, P., 2002b. Daycare attendance and risk of childhood acute lymphoblastic leukaemia. Br. J. Cancer, 86, 1419 1424. Maia, A.T., van der Velden, V.H., Harrison, C.J., Szczepanski, T., Williams, M.D., Griffiths, M.J., van Dongen, J.J., Greaves, M.F., 2003. Prenatal origin of hyperdiploid acute lymphoblastic leukemia in identical twins. Leukemia, 17, 2202 2206. Maia, A.T., Koechling, J., Corbett, R., Metzler, M., Wiemels, J.L., Greaves, M., 2004a. Protracted postnatal natural histories in childhood leukemia. Genes Chromosomes Cancer, 39, 335 340. Maia, A.T., Tussiwand, R., Cazzaniga, G., Rebulla, P., Colman, S., Biondi, A., Greaves, M., 2004b. Identification of preleukemic precursors of hyperdiploid acute lymphoblastic leukemia in cord blood. Genes Chromosomes Cancer, 40, 38 43. McHale, C.M., Smith, M.T., 2004. Prenatal origin of chromosomal translocations in acute childhood leukemia: implications and future directions. Am. J. Hematol., 75, 254 257. McHale, C.M., Wiemels, J.L., Zhang, L., Ma, X., Buffler, P.A., Feusner, J., Matthay, K., Dahl, G., Smith, M.T., 2003a. Prenatal origin of childhood acute myeloid leukemias harboring chromosomal rearrangements t(15;17) and inv(16). Blood, 101, 4640 4641. McHale, C.M., Wiemels, J.L., Zhang, L., Ma, X., Buffler, P.A., Guo, W., Loh, M.L., Smith, M.T., 2003b. Prenatal origin of TEL-AML1-positive acute lymphoblastic leukemia in children born in California. Genes Chromosomes Cancer, 37, 36 43. Moorman, A.V., Roman, E., Willett, E.V., Dovey, G.J., Cartwright, R.A., Morgan, G.J., 2001. Karyotype and age in acute myeloid leukemia. Are they linked? Cancer Genet. Cytogenet., 126, 155 161. Moppett, J., Burke, G.A., Steward, C.G., Oakhill, A., Goulden, N.J., 2003. The clinical relevance of detection of minimal residual disease in childhood acute lymphoblastic leukaemia. J. Clin. Pathol., 56, 249 253. Mori, H., Colman, S.M., Xiao, Z., Ford, A.M., Healy, L.E., Donaldson, C., Hows, J.M., Navarette, C., Greaves, M., 2002. Chromosome translocations and covert leukemic clones are generated during normal fetal development. Proc. Natl. Acad. Sci. U.S.A., 99, 8242 8247. Panzer-Grumayer, E.R., Fasching, K., Panzer, S., Hettinger, K., Schmitt, K., Stockler-Ipsiroglu, S., Haas, O.A., 2002. Nondisjunction of chromosomes leading to hyperdiploid childhood B-cell precursor acute lymphoblastic leukemia is an early event during leukemogenesis. Blood, 100, 347 349. Pedersen-Bjergaard, J., 2002. Chemicals and leukemia. In: Henderson, E.S., Lister, T.A., Greaves, M.F. (Eds.), Leukemia. Saunders, Philadelphia, pp. 171 199. M.T. Smith et al. / Toxicology and Applied Pharmacology 206 (2005) 237 245 245 Peters, U.R., Hasse, U., Oppliger, E., Tschan, M., Ong, S.T., Rassool, F.V., Borisch, B., Tobler, A., Fey, M.F., 1999. Aberrant FHIT mRNA transcripts are present in malignant and normal haematopoiesis, but absence of FHIT protein is restricted to leukaemia. Oncogene, 18, 79 85. Romana, S.P., Poirel, H., Leconiat, M., Flexor, M.A., Mauchauffe, M., Jonveaux, P., Macintyre, E.A., Berger, R., Bernard, O.A., 1995. High frequency of t(12;21) in childhood B-lineage acute lymphoblastic leukemia. Blood, 86, 4263 4269. Ross, J.A., 1998. Maternal diet and infant leukemia: a role for DNA topoisomerase II inhibitors? Int. J. Cancer, Suppl., 11, 26 28. Ross, M.E., Zhou, X., Song, G., Shurtleff, S.A., Girtman, K., Williams, W.K., Liu, H.C., Mahfouz, R., Raimondi, S.C., Lenny, N., Patel, A., Downing, J.R., 2003. Classification of pediatric acute lymphoblastic leukemia by gene expression profiling. Blood, 102, 2951 2959. Rothman, N., Smith, M.T., Hayes, R.B., Traver, R.D., Hoener, B., Campleman, S., Li, G.L., Dosemeci, M., Linet, M., Zhang, L., Xi, L., Wacholder, S., Lu, W., Meyer, K.B., Titenko-Holland, N., Stewart, J.T., Yin, S., Ross, D., 1997. Benzene poisoning, a risk factor for hematological malignancy, is associated with the NQO1 609C Y T mutation and rapid fractional excretion of chlorzoxazone. Cancer Res., 57, 2839 2842. Skibola, C.F., Smith, M.T., Kane, E., Roman, E., Rollinson, S., Cartwright, R.A., Morgan, G., 1999. Polymorphisms in the methylenetetrahydrofolate reductase gene are associated with susceptibility to acute leukemia in adults. Proc. Natl. Acad. Sci. U.S.A., 96, 12810 12815. Skibola, C.F., Smith, M.T., Hubbard, A., Shane, B., Roberts, A.C., Law, G.R., Rollinson, S., Roman, E., Cartwright, R.A., Morgan, G.J., 2002. Polymorphisms in the thymidylate synthase and serine hydroxymethyltransferase genes and risk of adult acute lymphocytic leukemia. Blood, 99, 3786 3791. Smith, M.T., 1996. The mechanism of benzene-induced leukemia: a hypothesis and speculations on the causes of leukemia. Environ. Health Perspect., 104 (Suppl. 6), 1219 1225. Smith, M.T., 1999. Benzene, NQO1, and genetic susceptibility to cancer. Proc. Natl. Acad. Sci. U.S.A., 96, 7624 7646. Smith, M.T., Wang, Y., Kane, E., Rollinson, S., Wiemels, J.L., Roman, E., Roddam, P., Cartwright, R., Morgan, G., 2001. Low NAD(P)H:quinone oxidoreductase 1 activity is associated with increased risk of acute leukemia in adults. Blood, 97, 1422 1426. Smith, M.T., Wang, Y., Skibola, C.F., Slater, D.J., Lo Nigro, L., Nowell, P.C., Lange, B.J., Felix, C.A., 2002. Low NAD(P)H:quinone oxidoreductase activity is associated with increased risk of leukemia with MLL translocations in infants and children. Blood, 100, 4590 4593. Stover, P., Schirch, V., 1991. 5-Formyltetrahydrofolate polyglutamates are slow tight binding inhibitors of serine hydroxymethyltransferase. J. Biol. Chem., 266, 1543 1550. Tanaka, H., Shimada, Y., Harada, H., Shinoda, M., Hatooka, S., Imamura, M., Ishizaki, K., 1998. Methylation of the 5VCpG island of the FHIT gene is closely associated with transcriptional inactivation in esophageal squamous cell carcinomas. Cancer Res., 58, 3429 3934. Taub, J.W., Konrad, M.A., Ge, Y., Naber, J.M., Scott, J.S., Matherly, L.H., Ravindranath, Y., 2002. High frequency of leukemic clones in newborn screening blood samples of children with B-precursor acute lymphoblastic leukemia. Blood, 99, 2992 2996. Taylor, G.M., Dearden, S., Ravetto, P., Ayres, M., Watson, P., Hussain, A., Greaves, M., Alexander, F., Eden, O.B., 2002. Genetic susceptibility to childhood common acute lymphoblastic leukaemia is associated with polymorphic peptide-binding pocket profiles in HLA-DPB1*0201. Hum. Mol. Genet., 11, 1585 1597. Thompson, J.R., Gerald, P.F., Willoughby, M.L., Armstrong, B.K., 2001. Maternal folate supplementation in pregnancy and protection against acute lymphoblastic leukaemia in childhood: a case-control study. Lancet, 358, 1935 1940. Wiemels, J.L., Cazzaniga, G., Daniotti, M., Eden, O.B., Addison, G.M., Masera, G., Saha, V., Biondi, A., Greaves, M.F., 1999a. Prenatal origin of acute lymphoblastic leukemia in children. Lancet, 354, 1499 1503. Wiemels, J.L., Pagnamenta, A., Taylor, G.M., Eden, O.B., Alexander, F.E., Greaves, M.F., 1999b. A lack of a functional NAD(P)H:quinone oxidoreductase allele is selectively associated with pediatric leukemias that have MLL fusions. United Kingdom Childhood Cancer Study Investigators. Cancer Res., 59, 4095 4099. Wiemels, J.L., Smith, R.N., Taylor, G.M., Eden, O.B., Alexander, F.E., Greaves, M.F., 2001. Methylenetetrahydrofolate reductase (MTHFR) polymorphisms and risk of molecularly defined subtypes of childhood acute leukemia. Proc. Natl. Acad. Sci. U.S.A., 98, 4004 4009. Wiemels, J.L., Leonard, B.C., Wang, Y., Segal, M.R., Hunger, S.P., Smith, M.T., Crouse, V., Ma, X., Buffler, P.A., Pine, S.R., 2002a. Site-specific translocation and evidence of postnatal origin of the t(1;19) E2A-PBX1 fusion in childhood acute lymphoblastic leukemia. Proc. Natl. Acad. Sci. U.S.A., 99, 15101 15106. Wiemels, J.L., Xiao, Z., Buffler, P.A., Maia, A.T., Ma, X., Dicks, B.M., Smith, M.T., Zhang, L., Feusner, J., Wiencke, J., Pritchard-Jones, K., Kempski, H., Greaves, M., 2002b. In utero origin of t(8;21) AML1ETO translocations in childhood acute myeloid leukemia. Blood 99, 3801 3805. Yagi, T., Hibi, S., Tabata, Y., Kuriyama, K., Teramura, T., Hashida, T., Shimizu, Y., Takimoto, T., Todo, S., Sawada, T., Imashuku, S., 2000. Detection of clonotypic IGH and TCR rearrangements in the neonatal blood spots of infants and children with B-cell precursor acute lymphoblastic leukemia. Blood, 96, 264 268. Yang, H.W., Piao, H.Y., Taki, T., Chen, T., Hashizume, K., Ohnishi, H., Bessho, F., Yanagisawa, M., Matsuo, Y., Hayashi, Y., 1999. Pattern of FHIT gene expression in normal and leukaemic cells. Int. J. Cancer, 81, 897 901. Zheng, S., Ma, X., Zhang, L., Gunn, L., Smith, M.T., Wiemels, J.L., Leung, K., Buffler, P.A., Wiencke, J.K., 2004. Hypermethylation of the 5VCpG island of the FHIT gene is associated with hyperdiploid and translocation-negative subtypes of pediatric leukemia. Cancer Res., 64, 2000 2006. Zochbauer-Muller, S., Fong, K.M., Maitra, A., Lam, S., Geradts, J., Ashfaq, R., Virmani, A.K., Milchgrub, S., Gazdar, A.F., Minna, J.D., 2001. 5V CpG island methylation of the FHIT gene is correlated with loss of gene expression in lung and breast cancer. Cancer Res., 61, 3581 3583.