Xiao-he Chen

CYP2A6, MAOA, DBH, DRD4, and 5HT2A genotypes, smoking behaviour and cotinine levels in 1518 UK adolescents

Objectives Smoking is a major cause of death and often initiates in adolescence. Mutations in CYP2A6 slow metabolism of nicotine to cotinine. Haploinsufficiency in adults is associated with lower cigarette consumption, lower cotinine level and higher quit rates. Other genes are also implicated in smoking behaviour. We explored smoking behaviour and cotinine levels in relation to genotypes in adolescents. Methods 1518 subjects from the Ten Towns Heart Health Study were genotyped for CYP2A6 alleles *1A, *1B, *2, *4, *5, *9 and *12 to classify predicted nicotine metabolism rate. DBH(rs77905), MAOA(rs1801291+VNTR), DRD4(VNTR) and 5HT2A(rs6313) were also studied. Smoking status was established by questionnaire and salivary cotinine measurement at 13–15 and 18 years. Results No significant associations were identified for DBH, MAOA, DRD4 and 5HT2A markers, with smoking status or cotinine level at either age. At age 18, haploinsufficiency (HI) for CYP2A6 was associated with a higher odds of being a current smoker compared with the *1B carriers (WT1B) (OR=2.23 (1.16, 4.27) for current versus ex); *1A homozygotes (WT1A) were also at slightly higher risk (OR=1.44 (1.01, 2.06)). Partial haploinsufficiency (PHI) was not associated with being a current smoker. There were no significant associations at age 13–15. PHI and HI were associated with higher cotinine levels amongst smokers at both 13–15 and at 18 years compared with WT1B and WT1A groups. Conclusions CYP2A6 haploinsufficiency increases likelihood of continuing smoking in teenagers. We hypothesize an explanatory ‘occupancy’ model to explain why haploinsufficiency results in faster progression to nicotine dependence, but lower subsequent consumption.

Replication of IGF2-INS-TH * 5 haplotype effect on obesity in older men and study of related phenotypes

Interindividual variation of the IGF2-INS-TH region influences risk of a variety of diseases and complex traits. Previous studies identified a haplotype (designated IGF2-INS-TH*5 and tagged by allele A of IGF2 ApaI, allele 9 of TH01 and class I alleles of INS VNTR) associated with low body mass index (BMI) in a cohort of UK men. We aimed here both to study whether previous findings relating *5 with weight are replicated in a different cohort of men (East Hertfordshire) characterised in more phenotypic detail and to test the effect of this haplotype on related subphenotypes. The PHASE program was used to identify *5 and not*5 haplotypes. A total of 490 haplotypes were derived from 131 men and 114 women, the frequency of *5 being around 9%. Specific tests of *5 haplotype (vs not*5 haplotypes) conducted included Students t-test and multiple regression analyses. We observed replication of weight effect for the *5 haplotype in men: significant associations with lower BMI (−1.81 kg/m2, P=0.009), lower waist circumference (−6.3 cm, P=0.001) and lower waist–hip ratio (−5%, P<0.001). This haplotype also marks nearly two-fold lower 120 min insulin (P=0.004) as well as low baseline insulin (−11.02 pmol/l, P=0.043) and low 30 min insulin (−64.44 pmol/l, P=0.072) in a glucose tolerance test. No association between *5 and these traits was found in women. Our results, taken together with other data on IGFII levels and TH activity, point to the importance of *5 as an integrated polygenic haplotype relevant to obesity and insulin response to glucose in men.

Higher resolution microplate array diagonal gel electrophoresis: application to a multiallelic minisatellite.

The 5′ polymorphic region of the insulin (INS, MIM# 176730) gene contains a variable tandem repetition of 14–15 bp (a variable number of tandem repeats (VNTR) locus). After PCR amplification, we achieved precise sizing of class I alleles (range 641 to 843 bp) on 96‐well open‐face polyacrylamide microplate array diagonal gel electrophoresis (MADGE) gels, obtaining resolution of the 2% mobility difference which represents one tandem repeat. PCR products were run double‐stranded, but no additional bands were generated except in the case of differences of three, two, and one repeat between alleles; none compromised allele identification, and in the latter case the heteroduplex was a useful confirmation signal. No end labelling of primers was required, as the sensitive Vistra Green™ intercalating dye for double strands was used for visualization of bands from diluted samples. Duracryl™, a high mechanical‐strength polyacrylamide derivative, proved to have good resolution properties for electrophoresis. A co‐run ladder ensured precise binning without inter‐lane variability. Simultaneous electrophoresis of gels in a thermostatically controlled tank allowed up to 1,000 samples to be run in 90 min. Gels were analyzed using a FluorImager® 595 fluorescent scanning system, and alleles identified using a combination of Phoretix™ software for band migration measurement and Microsoft® Excel® to compute allele sizes. Unlike other systems for minisatellite allele sizing, throughput was not limited (in time or cost) by electrophoresis. Hum Mutat 15:565–576, 2000.

Microplate array diagonal gel electrophoresis for mutation research in DNA banks

Molecular genetic epidemiology, association and linkage studies in populations, human or other species, is now yielding powerful new insights into disease and susceptibility genes. Inter alia, the subject requires laboratory analytical methodologies focused on achieving high throughput. Here we review one suite of methodology suitable for such laboratories. Microplate array diagonal gel electrophoresis (MADGE) was invented for molecular genetic epidemiological studies. It combines direct compatibility with microplates, convenient polyacrylamide gel electrophoresis and economy of time and reagents, at minimal capital cost, and enables one user to run up to several thousand gel lanes per day for direct assay of single‐base variations. Melt‐MADGE combines temporal thermal ramp apparatus to achieve similar throughput for de novo mutation scanning.

Non-recombining chromosome Y haplogroups and centromeric HindIII RFLP in relation to blood pressure in 2,743 middle-aged Caucasian men from the UK

Evidence from rodents and association analyses in humans suggest the presence on chromosome Y of one or more genes affecting blood pressure (BP). The HindIII centromeric alphoid polymorphism has been reported to be associated with BP in three independent human populations, although other studies reported null associations with this trait. Our objective was to test for association between BP and genetic variation of the Y chromosome. To this end, 2,743 unrelated Caucasian men recruited from nine UK practices were analysed for five SNPs (including the HindIII site) and two microsatellites spanning the non-recombining region of the Y chromosome. Systolic and diastolic BP were analysed both as quantitative traits and as categorical variables. Differences between locations were tested. Haplotypic and linkage disequilibrium (LD) analyses were also performed. Overall, no significant association was found between any of the loci analysed and BP, although post hoc analyses suggest a possible relation of specific Y haplogroups to BP. The HindIII polymorphism marks major structural differences in the Y centromere which could infuence mitotic loss during ageing, or other somatic events. However, this study does not support a causal effect on BP, although association of one or more Y haplogroups cannot be excluded.

Evidence of Admixture from Haplotyping in an Epidemiological Study of UK Caucasian Males: Implications for Association Analyses

Objective: Cohort and case-control genetic association studies offer the greatest power to detect small genotypic influences on disease phenotypes, relative to family-based designs. However, genetic subdivisions could confound studies involving unrelated individuals, but the topic has been little investigated. We examined geographical and interallelic association of SNP and microsatellite haplotypes of the Y chromosome, of regions of chromosome 11, and of autosomal SNP genotypes relevant to cardiovascular risk traits in a UK-wide epidemiological survey. Results: We show evidence (p = 0.00001) of the Danelaw history of the UK, marked by a two-fold excess of a Viking Y haplotype in central England. We also found evidence for a (different) single-centre geographical over-representation of one haplotype, both for APOC3-A4-A5 and for IGF2. The basis of this remains obscure but neither reflect genotyping error nor correlate with the phenotypic associations by centre of these markers. A panel of SNPs relevant to cardiovascular risks traits showed neither association with geographical location nor with Y haplotypes. Conclusion: Combinations of Y haplotyping, autosomal haplotyping, and genome-wide SNP typing, taken together with phenotypic2 associations, should improve epidemiological recognition and interpretation of possible confounding by genetic subdivision.

Paucimorphic Alleles versus Polymorphic Alleles and Rare Mutations in Disease Causation: Theory, Observation and Detection

Definitions of polymorphism in a gene include occurrence of a rarer allele of at least 1% frequency; or occurrence of the commonest allele at less than 95% frequency. Many alleles of single nucleotide polymorphisms (SNPs) in genes occur at much higher frequency (up to 50%). Many common diseases have a substantial genetic component. The prevailing hypothesis for the molecular basis of common diseases is that it involves the combinatoric action of common polymorphic alleles of minor effect (common disease / common variant, CD / CV hypothesis). The ready development of genome-wide databases of high frequency SNPs is enabling the testing of this hypothesis. A contrasting approach has been the study of very highly selected cases and families by linkage and mutation detection techniques to identify rare mutations of large effect on a gene, often private to a single family (rare disease / rare variant, RD / RV hypothesis. These approaches have formed the mainstay of disease gene discovery, the latter having been feasible for a decade, the former just now becoming feasible. However, an intermediate possibility exists. Sequence changes at an intermediate frequency (herewith, “paucimorphisms”, arbitrarily 0.0005<q<0.05) may exist and may have a moderate effect. A number of different loci may predispose to the same disease, although only one paucimorphic allele of one particular gene will be found in any one individual. Exploring the “paucimorphisms hypothesis” will require mutation detection applied both at the level of large numbers of relatively unselected cases and at the population level. In this review we consider the foundations of this hypothesis, relevant available technologies and possible future approaches to systematically explore this hypothesis.

Microplate-array diagonal-gel electrophoresis (MADGE) systems for high-throughput electrophoresis

▼Microplate-array diagonal-gel electrophoresis (MADGE) was invented by Day in 1994 (Ref. 1, 2) to enable compatibility between microplate-based liquid phase reactions and polyacrylamide or agarose-gel electrophoretic analysis. In MADGE, a microplate-compatible array of wells is turned on a diagonal relative to the line of electrophoresis so that samples can run on a long track between the wells (Fig. 1). This means that the 96-well format can be coupled to high sample resolution. Microplates occupy an analytical ‘middle ground’ in throughput between test tubes and microarrays, and MADGE can now meet the associated electrophoretic requirements at this level. The natural scalability of MADGE enables high throughput in large laboratories while its simplicity facilitates distributability between smaller laboratories. The set-up and running costs of MADGE compare favourably with other techniques (Ref. 3) and most electrophoresis-based analysis should be feasible using MADGE. The invention of MADGE was driven by the higher-throughput requirements in human singlenucleotide polymorphism (SNP) genotyping, and similar requirements for de novo mutation scanning have further spawned the invention of melt-MADGE. This set of core protocols reflects the origins of MADGE but the scope of MADGE extends in principle to any higher throughput genomic analysis. In a short series of four articles for Technical Tips Online, we describe the developed components of MADGE. This first article reviews the use of MADGE for analysing simple patterns of polymorphism such as those generated in