BRITISH JOURNAL OF CANCER Final: 30th October 2006
Title: Germline mutation rate in the hypervariable minisatellite CEB1 in the parents of children with leukaemia
Authors: BG Davies1, JM Birch2, TOB Eden3, M Reeves3, YE Dubrova4 and GM Taylor1
Addresses: 1
Cancer Immunogenetics Laboratory, 2CRUK Paediatric and Familial Cancer Study Group,
3
Academic Unit of Paediatric Oncology, Division of Human Development, University of
Manchester, Manchester, UK and 4Department of Genetics, University of Leicester, Leicester, UK.
Correspondence: G M Taylor; Email:
[email protected]
Keywords: Childhood leukaemia; Minisatellite; Germline; Mutation.
Word counts: Summary: 233 Text: 3,909
1
Summary Gardner and colleagues advanced the hypothesis that the Seascale leukaemia cluster may have been caused by germ cell mutagenesis due parental preconceptional irradiation (PPI) of fathers working at the Sellafield nuclear installation. Recent evidence has shown that PPI can lead to an increased germline mutation rate in certain minisatellite loci, allowing previously undetectable levels of germline sensitivity to IR to be measured. In this study, we have investigated the hypothesis that childhood leukaemia may be associated with parental germ cell sensitivity detected by an increased parental germline minisatellite mutation rate. To test this we compared the germline mutation rate of the hypervariable minisatellite locus, CEB1 in family trios (both parents and their child) of children with leukaemia (n = 109) compared with normal families (n = 64). We found no significant difference in mean mutation rate of parents of children with leukaemia and control children (0.0414 vs 0.0887; P > 0.05). The majority of germline mutations (95%) were paternal, and unaffected by the presence of leukaemia cells in the sample. We found no significant differences in mean minisatellite allele size (72 vs 68 repeats; p > 0.05), or mutational spectrum (gains or losses of repeats) in the parents of case and control children. Although preliminary, our results suggest that childhood leukaemia is unlikely to be associated with increased germline minisatellite instability resulting from exposure of parental germ cells to mutagens such as IR.
2
Introduction Acute leukaemia accounts for about 30% of childhood malignancies in the UK, with an average incidence rate of 4/100,000/year in the socio-economically developed countries (Little, 1999; Stiller, 2004). Although it is thought to be initiated by unrepaired chromosome breaks following exposure of the fetus or infant to environmental carcinogens (Greaves, 1997; 1999), epidemiological studies have failed to identify specific carcinogens with any certainty. A notable exception involves exposure of the unborn child to X-rays in utero (Stewart et al, 1956; Stewart and Kneale, 1968; Stewart and Kneale, 1970; Bithell and Stewart, 1975; Doll and Wakeford, 1997; Wakeford and Little, 2003).
The discovery of an unusual excess of childhood leukaemia cases in the village of Seascale in the vicinity of the Sellafield nuclear site in Cumbria, UK, raised concerns that this might have been caused by fetal or neonatal exposure to IR (reviewed in Beral, Roman and Bobrow, 1993). To identify the cause more precisely, a case-control study of the Seascale “cluster” was carried out by Gardner and colleagues (Gardner et al, 1990). They interpreted their findings as suggesting an effect of IR on fathers that may be leukaemogenic for their offspring. They proposed that this might occur by IR-induced de novo mutations in sperm during the preconceptional period, a mechanism later known as the preconceptional irradiation (PPI) hypothesis. The PPI hypothesis suggested a new mechanism of leukaemogenesis in humans, and was highly controversial, since there was little evidence from other sources to suggest that childhood leukaemia could be caused by germline mutation. Moreover, the fathers’ radiation exposure was considered to be insufficient to cause the mutation rate required for the number of cases observed (Evans, 1990; Doll, Evans and Darby, 1994; Wakeford et al, 1994). Additionally, there was no evidence for a germline effect among the children of Japanese atomic bomb survivors (Kodaira et al, 1995, 2004; Izumi et al, 2003). Since radiation damage is randomly distributed in the genome, levels of PPI causing childhood leukaemia should also cause an
3
increase in other single gene disorders (Doll et al, 1994). Parker et al (2000) reported an increased risk of stillbirth in men from Sellafield exposed to PPI, but this was not confirmed by Doyle et al (2000) in a more wide-ranging study of nuclear industry workers.
Subsequently Dubrova and colleagues demonstrated that PPI could lead to an increased mutation rate in a type of tandem repeat DNA locus (TRDL) known in humans as minisatellites. Due to the high spontaneous germline mutation rate of some minisatellites (>1,000 x higher than most protein-coding loci), a PPI effect could be detected in a substantially smaller population than required to detect mutations in protein-coding loci (Dubrova et al, 1993; 1997). Studies of two different populations, one exposed to radiation (~0.5 Gy) from radionuclide contaminated land following the Chernobyl accident (Dubrova et al, 1996, 1997, 2002), the other due to radiation (>1 Sv) from the fallout of nuclear weapons tests (Dubrova et al, 2002), revealed an approximate doubling in germline minisatellite mutation rate due to PPI. These and other data suggest that minisatellite mutation rates may be useful biomarkers of germline genetic effects caused by environmental mutagens such as radiation (Yauk, 2004: Bouffler et al, 2006). Since minisatellites do not appear to be the direct target of mutagen-induced DNA damage (Dubrova et al, 1996, 1997), they may be acting as a marker for some other effect, such as genome sensitivity or instability (Niwa, 2003).
Based on the PPI hypothesis for the aetiology of the Sellafield leukaemia cluster (Gardner et al, 1990), and the increased germline minisatellite mutation due to PPI, we surmised that a proportion of sporadic childhood leukaemias might be associated with an increased germline minisatellite mutation rate due to undetected exposures of parental germ cells to mutagens. To address this, we compared germline mutation rates in the hypervariable human minisatellite, CEB1 (Vergnaud and Denoeud, 2000) in the families of children with leukaemia (n = 109) and control families (n = 64), using a sensitive PCR-based germline mutation assay.
4
Materials and Methods Study Group The case family series consisted of a cross-sectional series of 110 children with acute leukaemia treated at Royal Manchester Children's Hospital, UK, and both biological parents (n = 218 one family had monozygotic twins with leukaemia). Following parental consent, blood samples (10 20 ml) were collected between 1990 and 1995. The control family series consisted of buccal scrape samples obtained from 64 unaffected children also from the North West of England, and blood samples from both control parents (n = 128). All sample collection and analyses were carried out with the approval of the Local Research Ethics Committee.
Genomic DNA Extraction Genomic DNA (gDNA) was extracted from case and control blood samples using a Nucleon BACC2 genomic DNA extraction kit (Amersham Biosciences), and from buccal samples using the MasterAmp™ Buccal Swab DNA Extraction Kit (Epicentre Technologies). DNA was quantified by fluorimetry on a TBS-380 Mini-Fluorometer (Turner BioSystems, Inc. Sunnyvale, CA., USA) using a PicoGreen® (Molecular Probes) based assay, and working DNA dilutions of 50 ng/µl were prepared.
CEB1 Primer Sequences and PCR Amplification Conditions CEB1 alleles were amplified by PCR using the 3' CEB1 primer P14 (5'-ggatcctctcctgtgcctttcct-3') as described by Buard et al (1998). The 5' primer MAR1 (5'-gaattttcagtgagagtcggcc-3') was designed for this study to avoid single nucleotide polymorphisms (SNPs) flanking the CEB1 minisatellite, using the CEB1 sequence (accession no. AF048727) and SNP information kindly provided by Dr Jerome Buard (personal communication). PCRs were performed on an MJ Research DNA Engine (Waltham, MA., USA) in reaction mixtures prepared on ice in 0.2 ml thin-
5
walled tubes consisting of 1 µl of gDNA (50 ng) and 0.5 µl (2.5 units) of Platinum Taq DNA polymerase High Fidelity (Invitrogen) added to 48.5 µl of Mg2+-titrated PCR master-mix (see below). Amplification conditions were: 94° C for 1 min. 30 s, 30 cycles at 94° C for 15s, 60° C for 30 s, 70° C for 10 min., and a final extension step of 70° C for 15 min.
Mg2+ titration PCR amplification of CEB1 alleles was found to be highly sensitive to Mg2+ concentration. A PCR master-mix (19.4 ml) was prepared with the following final reagent concentrations: 0.5 µM each HPLC-purified CEB1 primer (Eurogentec), 0.7 mM MgSO4 (50 mM MgSO4 with Platinum Taq DNA Polymerase), 0.2 mM dNTPs (PCR Nucleotide Mix, Amersham Biosciences) 5% v/v DMSO (Sigma), 0.2 mg/ml bovine serum albumin (BSA, (ultra pure, non-acetylated; Ambion), and 1 x High Fidelity PCR Buffer (50x High Fidelity PCR Buffer with Platinum Taq DNA Polymerase HF). The master-mix was kept on ice at 4° C while a test PCR on a known gDNA sample was performed. Lack of bands or faint bands indicated a requirement for more Mg2+, and this was added in 10 or 20 µl aliquots resulting in an increase of 0.025-0.05 mM Mg2+. This was repeated until robust bands (see below) were obtained, and the titrated master-mix was stored at -20° C.
Electrophoresis CEB1 alleles amplified in 20 µl PCR reaction mixtures from family trios (father, mother and child) were loaded onto a 40 cm long 0.7% (w/v) agarose gel (SeaKem LE Agarose, Cambrex BioScience). Gels were run at 2 Volts/cm in 1xTBE buffer (Crystal Buffers, Severn Biotech) for approximately 23 h, until the 400 bp marker from the DNA ladder (1 Kbp Plus DNA Ladder, Invitrogen) was about to run off the gel. Few alleles 0.2). Analysis of mutation rates in relation to leukaemia subtype (Table 3) was limited by the small number of patients in each group, and the results must be viewed with caution. The highest mutation rate was observed in T ALL (0.1250), but this is based on only 5 cases. We compared the ages of case fathers whose children had a germline mutation with fathers of children who did not, but found no significant difference (p = 0.28). Similarly we found no significant difference in the age of control fathers of children with or without germline mutations (p = 0.11). Among the germline mutations detected in the 8 case children, 4 were found in gDNA from remission blood and 4 in diagnostic samples (see Discussion).
10
Comparison of CEB1 allele sizes in case and control parents (Figure 3) revealed no significant difference in size distribution between the two groups (P > 0.05). Overall, parental alleles ranged in size from 215 bp (5 repeats) to 9,131 bp (228 repeats) with a mean case parent allele size of 2,896 bp (72 repeats), and a mean control parent allele size of 2,711 bp (68 repeats).
Germline mutation spectrum We compared the type of mutational event (gain or loss of repeats), and the size of the repeat for the case and control germline mutant alleles. In children with leukaemia, we observed 5 (62%) mutant alleles with an increase and 3 with a decreased number of repeats, whilst in the controls, 9 (81%) mutants were increased, and 2 were decreased in size. In both groups the size of the mutational events were similar, most mutations (68%) involving the gain or loss of 1 - 5 repeat units (Figure 4). Overall, the largest mutation detected was a deletion of 12 repeat units (progenitor allele size: 3,454 bp) in a control child. In the case series the largest mutation was a deletion of 9 repeats (progenitor allele size: 2,634 kbp) in a remission sample from a child with BCP ALL.
Discussion This study was designed to address the hypothesis that childhood leukaemia is associated with an increased rate of germline mutation in the hypervariable minisatellite, CEB1. The results suggest, with some qualifications, that there is no such association. The hypothesis was based on the on the notion that the exposure of parental germ cells to mutagenic agents such as IR, can lead to an increased risk of childhood leukaemia - the PPI hypothesis (Gardner et al, 1990), with minisatellite mutation acting directly or indirectly as a biomarker of germline leukaemogenic damage. As far as we are aware no previous analysis of this type has been reported in childhood leukaemia. We could find no evidence for quantitative (germline mutation rate) or
11
qualitative (mutation spectrum) differences in CEB1 minisatellite mutations between case and control families.
Unlike other studies of germline minisatellite mutation rate in relation to parental IR exposure (reviewed by Bouffler et al, 2006), no radiation dosimetry was available for our leukaemia case series. The catchment area for our sporadic cases covers an area of North West England south of Cumbria, where levels of occupational or environmental radiation exposure are far below that reached at Sellafield. In view of the high rate of spontaneous and PPI-induced CEB1 mutation, we believe that this provides the most sensitive current measure of germ cell mutagenesis, and given the absence of any increase in association with childhood leukaemia provides reassuring, albeit preliminary, evidence that childhood leukaemia arises post-zygotically.
Germline minisatellite mutation analysis has generally been carried out using Southern blotting or probing for PCR-amplified gDNA (Dubrova et al, 1996; Stead and Jeffreys, 2000). However, this can be expensive on gDNA and leads to difficulties in testing children where the amount of sample is a limiting factor. We therefore selected a PCR-based assay that directly visualised CEB1 alleles on long-read agarose gels using a UV transilluminator. We found direct analysis of PCR products to be simple, and bands clearly resolvable, allowing fine resolution and sizing of mutations. However, most minisatellites are difficult to amplify by PCR due to their high GCcontent, large allele sizes, and large differences in relative allele size, and this limits the number of minisatellites that can be analysed by our method. We found that the most critical aspect of CEB1 amplification was the Mg2+ concentration. Deviation from the optimum concentration by as little as 0.1 mM resulted in poor, or no amplification. We overcame this by titrating Mg2+ using a large volume of PCR master-mix. This gave reproducible amplification of CEB1 alleles to visible levels, and should be useful for the analysis of other PCR-able minisatellite loci, as we have found for the minisatellite B6.7 (Tamaki et al, 1999).
12
The majority of all germline mutations detected in both series of subjects in the present study (18/19) were paternal in origin, a result previously reported for CEB1 (Vergnaud et al, 1991). In addition to comparing germline mutation rates in case and control families, we examined the size change and polarity of mutational events (increase or decrease in allele size), but we found no differences between the two series. In both groups, expansions in allele size were more prevalent than deletions (with an overall ratio of 3:1 in favour of expansion), and a large proportion of mutations involved the gain or loss of only 1 - 3 repeat units (75% cases, 64% controls). Small mutations such as these are reminiscent of the stepwise model for microsatellite expansion (Nag, 2003) though the evidence suggests that virtually all CEB1 germline mutations are initiated by recombinogenic events occurring during gametogenesis (Buard et al, 1998; Buard et al, 2000), and not as a result of DNA replication slippage.
In view of the tendency of some single gene disorders to increase in frequency with paternal age, we compared paternal ages at the birth of case and control children with germline mutations. We could find no evidence that fathers transmitting germline mutations were significantly older than those who did not. Mean allele sizes differed slightly between the groups (72 and 68 repeats in cases and controls respectively), but no significant overall difference between the distributions was found (Kolmogorov-Smirnov two-independent samples test), suggesting that differences in allele size distribution is probably not a factor in the risk of leukaemia. Analysis of CEB1 mutation rate in relation to leukaemia subtype was limited by the small number of cases in the study, but our preliminary result suggesting that the mutation rate may be higher in fathers of patients with T ALL deserves further study. As far as we are aware there are no data on the sub-types of leukaemia in the Seascale leukaemia cluster that might suggest a bias toward a specific subtype.
13
A potential confounding problem in our study concerns the type of sample analysed from the leukaemia patients. We collected 68% of blood samples at diagnosis and 32% in remission. Since the diagnostic samples contained varying percentages of leukaemic blasts (1-96%), there was a possibility that CEB1 somatic mutations could have influenced the result. However, the CEB1 somatic mutation rate is known to be very low (Vergnaud et al, 1991; Buard et al, 1998; 2000), and no increase in somatic minisatellite mutation rate in leukaemic blasts has previously been reported. A comparison of the mutations in the diagnostic (4 mutations in 68 children; 5.9%) and remission samples (4 mutations in 32 children; 12.5%) revealed a lower frequency in the former which may account for the lower overall germline mutation rate in the cases than the controls (0.0414 vs 0.0887). In 7 case families, a paired diagnostic and a remission sample were available for analysis. We found no differences in the CEB1 alleles in the paired samples, even though 3 of the diagnostic samples had >73% blasts and 1 had 12% (% blast values were not available for 3 samples). This suggests that CEB1 somatic mutations make little if any contribution to the mutation rate that we detected in case families, and it is thus unlikely that leukaemic blasts obscure a higher constitutional mutation rate in normal leucocytes.
In summary, our preliminary results find no evidence of an increased rate of CEB1 germline minisatellite mutation in the parents of children with leukaemia. However, we have only studied one minisatellite locus, in contrast to normal practice in PPI studies of using up to 8 loci (Bouffler et al, 2006). Using CEB1 as a biomarker of germ cell sensitivity to mutagens including PPI, our data suggest that germline mutations affecting genetic stability are unlikely to be a major factor in the aetiology of childhood leukaemia. However, these results will need confirmation with more cases analysed, using more minisatellite loci, before a final conclusion can be reached, since our data do not rule out a role for pre-zygotic mutagenesis in a minority of cases. Furthermore, the recent report by Dickinson and Parker (2002) confirming the findings of Gardner et al (1990)
14
of a statistical association between father's radiation dose at Sellafield and the child's risk of developing leukaemia still requires a biological explanation.
Acknowledgements This project was carried out with funding provided by the Department of Health, U.K. (contract RRX81), to Dr GM Taylor. We very gratefully acknowledge receipt of specimens from the patients of the late Dr. RF Stevens and Dr AM Will during the course of this study. We thank Cancer Research (U.K.) for supporting JM Birch and TOB Eden, and the Medical Research Council and the Paediatric Oncology Fund Royal Manchester Children's Hospital for funding M Reeves. We also thank Ruth Barber (Department of Genetics, University of Leicester) for technical advice and Jerome Buard (Institut de Génétique Humaine, Montpellier, France) for kindly providing CEB1 sequence information.
REFERENCES Beral V, Roman E, Bobrow M. (1993) Childhood cancer and nuclear installations. BMJ Publishing Group, London, UK
Bithell JF, Stewart AM. (1975) Pre-natal irradiation and childhood malignancy: a review of British data from the Oxford survey. Br J Cancer 31: 271-287.
Bouffler SD, Bridges BA, Cooper DN, Dubrova Y, McMillan TJ, Thacker J, Wright EG, Waters R. (2006) Assessing radiation-associated mutation risk to the germline: repetitive DNA sequences as mutational targets and biomarkers. Radiation Research 165: 249-268.
15
Buard J, Bourdet A, Yardley J, Dubrova Y, Jeffreys AJ. (1998) Influences of array size and homogeneity on minisatellite mutation. EMBO J 17: 3495-502.
Buard J, Collick A, Brown J, Jeffreys AJ. (2000) Somatic versus germline mutation processes at minisatellite CEB1 (D2S90) in humans and transgenic mice. Genomics 65: 95-103.
Buard J, Brenner C, Jeffreys AJ. (2002) Evolutionary fate of an unstable human minisatellite deduced from sperm-mutation spectra of individual alleles. Am J Hum Genet 70: 1038-1043.
Dickinson HO, Parker L. (2002) Leukaemia and non-Hodgkin's lymphoma in children of male Sellafield radiation workers. Int J Cancer 99: 437-444.
Doll R, Evans HJ, Darby SC. Paternal exposure not to blame. Nature 367: 678-680.
Doll R, Wakeford R. (1997) Risk of childhood cancer from fetal irradiation. Br J Radiol 70: 130139.
Doyle P, Maconochie N, Roman E, Davies G, Smith PG, Beral V. (2000) Fetal death and congenital malformation in babies born to nuclear industry employees: report from the nuclear industry family study. Lancet 356: 1293-1299.
Dubrova YE, Jeffreys AJ, Malashenko AM. (1993) Mouse minisatellite mutations induced by ionizing radiation. Nature Genetics 5: 92-94.
16
Dubrova YE, Nesterov VN, Krouchinsky NG, Ostapenko VA, Vergnaud G, Giraudeau F, Buard J, Jeffreys AJ. (1997) Further evidence for elevated human minisatellite mutation rate in Belarus eight years after the Chernobyl accident. Mutation Research 381: 267-278.
Dubrova YE, Nesterov VN, Krouchinsky NG, Ostapenko VA, Neumann R, Neil DL, Jeffreys AJ. (1996) Human minisatellite mutation rate after the Chernobyl accident. Nature 380: 683-686.
Dubrova YE, Grant G, Chumak AA, Stezhka VA, Karakasian AN. (2002) Elevated minisatellite mutation rate in the post-Chernobyl families from Ukraine. Am J Hum Genet 71: 801-809.
Dubrova YE, Bersimbaev RI, Djansugurova LB, Tankimanova MK, Mamyrbaeva ZZ, Mustonen R, Lindholm C, Hulten M, Salomaa S. (2002) Nuclear weapons tests and human germline mutation rate. Science 295: 1037.
Evans HJ. (1990) Ionising radiations from nuclear establishments and childhood leukaemias-an enigma. BioEssays 12: 541-549.
Gardner MJ, Snee MP, Hall AJ, Powell CA, Downes S, Terrell JD. (1990) Results of case-control study of leukaemia and lymphoma among young people near Sellafield nuclear plant in west Cumbria. BMJ 300: 423-429.
Greaves MF. (1997) Aetiology of acute leukaemia. Lancet 349: 344-349.
Greaves M. (1999) Molecular genetics, natural history and the demise of childhood leukaemia. Eur J Cancer 35: 1941-1953.
17
Izumi S, Koyama K, Soda M, Suyama A. (2003) Cancer incidence in children and young adults did not increase relative to parental exposure to atomic bombs. Br J Cancer 89: 1709-1713.
Kasai K, Nakamura Y, White R. (1990) Amplification of a variable number of tandem repeats (VNTR) locus (pMCT118) by the polymerase chain reaction (PCR) and its application to forensic science. J Forensic Sci 35: 1196-1200.
Kodaira M, Satoh C, Hiyama K, Toyama K. (1995) Lack of effects of atomic bomb radiation on genetic instability of tandem-repetitive elements in human germ cells. Am J Hum Genet 57: 1275-1283.
Kodaira M, Izumi S, Takahashi N, Nakamura N. (2004) No evidence of radiation effect on mutation rates at hypervariable minisatellite loci in the germ cells of atomic bomb survivors. Radiat Res 162: 350-356.
Little J. (1999) Epidemiology of childhood cancer. International Agency for Research on Cancer, Lyon, France.
Nag DK. (2003) Trinucleotide repeat expansions: timing is everything. Trends Mol Med 9: 455457.
Niwa O. (2003) Induced genomic instability in irradiated germ cells and in the offspring; reconciling discrepancies among the human and animal studies. Oncogene 22: 7078-7086.
Parker L, Pearce MS, Dickinson HO, Aitkin M, Craft AW. (1999) Stillbirths among offspring of male radiation workers at Sellafield nuclear reprocessing plant. Lancet 354: 1407-1414.
18
Stead JDH, Jeffreys AJ. (2000) Allele diversity and germline mutation at the insulin minisatellite. Hum Mol Genet 9: 713-723.
Stewart AM, Webb J, Giles D, Hewitt D. (1956) Malignant disease in childhood and diagnostic irradiation in utero. Lancet ii: 447.
Stewart AM, Kneale GW. (1968) Changes in the cancer risk associated with obstetric radiography. Lancet i: 104-107.
Stewart AM, Kneale GW. (1970) Age-distribution of cancers caused by obstetric X-rays and their relevance to cancer latent periods. Lancet ii: 4-8.
Stiller CA (2004). Epidemiology and genetics of childhood cancer. Oncogene 23: 6429-6444.
Tamaki K, May CA, Dubrova YE, Jeffreys AJ. (1999) Extremely complex repeat shuffling during germline mutation at human minisatellite B6.7. Hum Mol Genet 8: 879-888.
Vergnaud G, Mariat D, Apiou F, Aurias A, Lathrop M, Lauthier V. (1991) The use of synthetic tandem repeats to isolate new VNTR loci: cloning of a human hypermutable sequence. Genomics 11: 135-144.
Vergnaud G, Denoeud F. (2000) Minisatellites: mutability and genome architecture. Genome Research 10: 899-907.
19
Wakeford R, Tawn EJ, McElvenny DM, Binks K, Scott LE, Parker L. (1994) The Seascale childhood leukaemia cluster cases – the mutation rates implied by paternal preconceptional radiation doses. J Radiol Prot 14: 17-24.
Wakeford R. (1995) The risk of childhood cancer from intrauterine and preconceptional exposure to ionising radiation. Environmental Health Perspectives 103: 1018-1025.
Wakeford R, Little MP. (2003) Risk coefficients for childhood cancer after intrauterine irradiation: a review. Int J Radiat Biol 79: 293-309.
Yauk CL (2004). Advances in the application of germline tandem repeat instability for in situ monitoring. Mutation Research 566: 169-182.
20
Table 1 Details of leukaemia cases series in this study Number of samples obtained at: Leukaemia
Total
Diagnosis
Remission
B-cell precursor ALL (BCP ALL)
48
24
72
Progenitor B-cell ALL (Pro-B ALL)
6
2
8
T-cell ALL (T ALL)
5
4
9
Acute myeloid leukaemia (AML)
7
2
9
Others*
2
0
2
Total
68
32
100
*One case of chronic myeloid leukaemia (CML) and one case of myelodysplastic syndrome (MDS).
21
Table 2 Germline CEB1 allele mutation rates in leukaemia cases and controls Case:control Comparison
Cases
Controls
P* ratio
Number of informative families
100
64
8
10
(parental origin: P, paternal; M, maternal)
(8P:0M)
(10P:1M)
Mean parental allele mutation rate
0.0414
0.0887
(193)
(121)
0.0825
0.1613
(97)
(62)
0
0.0161
(96)
(62)
Number of children with germline mutations
1.56:1
0.8:1
(No. parental alleles detected) Mean paternal allele mutation rate (No. paternal alleles detected) Mean maternal allele mutation rate (No. maternal alleles detected)
*Probability using Fisher's exact test of independence (two-tailed).
22
0.46:1
> 0.05
0.51:1
0.20
/
/
Table 3 Leukaemia subtype in relation to paternal allele mutation rate Paternal allele No. mutant Subtype
No. paternal alleles
mutation rate paternal alleles (%)
BCP ALL
70
7
10.0
Pro-B ALL
8
0
0
T ALL
8
1
12.5
AML
9
0
0
Other*
2
0
0
Total
97
8
8.0
Controls
62
10
16.1
23
B.
A.
0.8 0.9 1.0 1.1 0.8 0.9 1.0 1.1 mM Mg2+ 5 kb
2 kb
1 kb
Figure 1. Effect of Mg2+ concentration on amplification of CEB1 alleles in two gDNA samples (A,B) by PCR . No alleles were resolved at 0.8 or 1.1 mM Mg2+, whereas alleles were optimally detected at 0.9 mM.
24
A. F
C
B. M
F
C
M 4 kb
3 kb
2 kb
1 kb
Figure 2. Detection of germline CEB1 mutations in two family trios (A, B) following PCR and gel electrophoresis. Mutant alleles are indicated by white arrows. Family A has a paternal mutation corresponding to a 7 repeat expansion. Family B has a paternal mutation corresponding to a 1 repeat deletion. To ensure accurate band comparison between adjacent lanes samples were loaded in the order father (F), child (C), mother (M).
25
Case parents
1.6
Control parents
Frequency
1.2
0.8
0.4
0 25 0.5
1 1.5
50 2 2.5
75
100
3 3.5
125
4 4.5 5 5.5
150
175
6 6.5 7 7.5
200
225
8 8.5 9 9.5
Allele size (repeats / Kbp)
Figure 3. Size distribution of minisatellite alleles in the parents of case and control children (number of alleles: case, n = 358; control, n = 235). DNA fragments were grouped into 0.5 kb intervals from 0.5 to 9.5 Kbp, with the number of repeat units shown, where 1 consensus repeat = 40 bp; ref. 24). Differences between case and control parent allele distributions were not significant in the Kolmogorov-Smirnov two-sample test (P > 0.05).
26
Gain
Loss
Case children
Frequency
0.6
Control children
0.4
0.2
0 6-10
1-5
1-5
6-10
11-15
Number of repeat units gained or lost
Figure 4. Frequency distribution of germline CEB1 mutations in case and control children by size and polarity, with mutations grouped into repeat unit changes of up to 5 units. In leukaemia case children all gains are 1-5 repeats, and losses 1-10 repeats, whereas in control children gains are 1-10 repeats, and losses 1-15 repeats.
27
Tables and Figures Table 1 Details of leukaemia cases series in this study Table 2 Germline CEB1 allele mutation rates in leukaemia cases and controls Table 3 Leukaemia subtype in relation to paternal allele mutation rate
Figure 1. Effect of Mg2+ concentration on amplification of CEB1 alleles in two gDNA samples (A,B) by PCR . No alleles were resolved at 0.8 or 1.1 mM Mg2+, whereas alleles were optimally detected at 0.9 mM.
Figure 2. Detection of germline CEB1 mutations in two family trios (A, B) following PCR and gel electrophoresis. Mutant alleles are indicated by white arrows. Family A has a paternal mutation corresponding to a 7 repeat expansion. Family B has a paternal mutation corresponding to a 1 repeat deletion. To ensure accurate band comparison between adjacent lanes samples were loaded in the order father (F), child (C), mother (M).
Figure 3. Size distribution of minisatellite alleles in the parents of case and control children (number of alleles: case, n = 358; control, n = 235). DNA fragments were grouped into 0.5 kb intervals from 0.5 to 9.5 Kbp, with the number of repeat units shown, where 1 consensus repeat = 40 bp; ref. 24). Differences between case and control parent allele distributions were not significant in the Kolmogorov-Smirnov two-sample test (P > 0.05).
Figure 4. Frequency distribution of germline CEB1 mutations in case and control children by size and polarity, with mutations grouped into repeat unit changes of up to 5 units. In leukaemia case children all gains are 1-5 repeats, and losses 1-10 repeats, whereas in control children gains are 1-10 repeats, and losses 1-15 repeats.
28
