Leptin, a hormone secreted by adipocytes, plays a pivotal role in the control of body weight. Rodents with mutations in the leptin receptor gene develop morbid.
1997 Oxford University Press
Human Molecular Genetics, 1997, Vol. 6, No. 6
869–876
Leptin receptor gene variation and obesity: lack of association in a white British male population Takanari Gotoda1, Brian S. Manning1,3, Anthony P. Goldstone2, Helen Imrie1, Alison L. Evans1, A. Donny Strosberg3, Paul M. McKeigue4, James Scott1,2 and Timothy J. Aitman1,2,* 1Molecular
Medicine Group, MRC Clinical Sciences Centre and 2Department of Medicine, Royal Postgraduate Medical School, Hammersmith Hospital, Du Cane Road, London W12 0NN, UK, 3Institut Cochin de Génétique Moléculaire, Laboratoire d’Immunopharmacologie Moléculaire, Paris 75014, France and 4Department of Epidemiology and Population Sciences, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK Received January 30, 1997; Revised and Accepted February 27, 1997
Leptin, a hormone secreted by adipocytes, plays a pivotal role in the control of body weight. Rodents with mutations in the leptin receptor gene develop morbid obesity. It is possible, therefore, that leptin receptor gene mutations contribute to human obesity. To test this possibility, we determined the entire coding sequence of the human leptin receptor cDNA from peripheral blood lymphocytes of 22 morbidly obese patients with body-mass index (BMI) between 35.1 and 60.9 kg/m2. We identified five common DNA sequence variants distributed throughout the coding sequence at codons 109, 223, 343, 656 and 1019, one rare silent mutation at codon 986 and one novel alternatively spliced form of transcript. None of the five common variants, including the three that predict amino acid changes, are null mutations causing morbid obesity, because homozygotes for the variant sequences were also found in lean subjects. Furthermore, the frequency of each variant allele and the distribution of genotypes and haplotypes were similar in 190 obese (BMI >28 kg/m2) and 132 lean (BMI 28 kg/m2) and 132 lean (BMI 0.1) between the obese and lean groups. 2Genotype
872
Human Molecular Genetics, 1997, Vol. 6, No. 6 Table 2. Haplotype frequencies in obese and lean groups Haplotypesa
Obese group
Lean group
11111
35.8%
40.1%
22112
18.3%
17.7%
11122
16.2%
15.5%
12211
14.8%
15.3%
22111
8.1%
7.7%
11112
2.1%
0.5%
Others
4.7%
3.2%
aNumbers
on each haplotype from left to right correspond to alleles at codons 109, 223, 343, 656 and 1019 variants respectively [allele 1, wild-type allele (13); allele 2, variant allele]. Only haplotypes with a frequency >1% at least in one group are shown. No significant difference is observed (P >0.1) between the two groups.
Linkage disequilibrium and haplotype analysis All five common variants were in strong linkage disequilibrium in the white British male population. The gametic linkage disequilibrium coefficient D between variants at codons 109 and 223, 223 and 343, 343 and 656, and 656 and 1019 was 0.148 (χ2 = 209), 0.087(χ2 = 78.2), –0.023(χ2 = 8.54) and 0.095 (χ2 = 98.7) respectively (all equivalent to P 35 kg/m2, defined here as morbidly obese, (range = 35.1–60.9 kg/m2) and amplified by RT-PCR as described previously (25) with some modifications. RNA (1 µg) was reverse-transcribed in 25 µl of reaction mixture with 100 pmol of random hexamer primers. After incubation for 5 min at 95C, 1/10 of the products were amplified by PCR with each of six pairs of oligonucleotide primers (Table 4A), which resulted in the amplification of six cDNA fragments covering the entire coding region of human OB-R cDNA (Fig. 1). Amplification was performed during 35 cycles of PCR (94C 30 s, 50C 45 s, 72C 60 s) with 1.5 mM MgCl2 concentration, followed by an additional extension (72C 5 min). In the PCR amplification from lymphocytes frozen for a long period, an additional preceding PCR amplification was carried out with primers external to the above PCR products (Table 4B). PCR products (50 ng) were sequenced by cycle-sequencing with Taq polymerase FS dye-terminator sequencing kits (Perkin-Elmer) on a model 373A or 377 automated DNA sequencer (Applied Biosystems). Partial genomic analysis of the OB-R gene Several OB-R gene fragments could be amplified from leukocyte DNA by PCR using primers with exonic sequences. A 14 kb gene fragment comprising introns W, X and Y (Fig. 3) was amplified by a Gene Amp XL PCR kit (Perkin-Elmer) with a forward primer on exon W (5′-CCATTGAGAAGTACCAGTTC-3′) and a reverse primer on exon Z (5′-AGGACCACATGTCACTGATG-3′). DNA sequencing was carried out for the 14 kb fragments from a normal control subject and the obese patient with the abnormal cDNA fragment 5, focusing on the six exon–intron boundaries shown in Figure 3C. Genotyping Genotyping of the five common DNA sequence variants was carried out by PCR-RFLP analysis. Genomic DNA (100 ng in 50 µl reactions) was amplified by PCR with a pair of oligonucleotide primers specific for each sequence variant (Table 4C), and the PCR products were digested with an appropriate restriction endonuclease (see Fig. 4). Reverse primers for the codon 109, 656 and 1019 variants and a forward primer for the codon 343 variant have a sequence mismatch in their 3′ region in order to create an artificial restriction site. PCR amplification was performed as described above, except for a MgCl2 concentration of 2.0 mM for the codon 109 assay, 1.25 mM for the codon 343 assay and 1.0 mM for the codon 223 assay. Study population White British men aged 40–64 years with both parents born in Britain were selected from the 1988–1990 population-based epidemiological survey of factory workers and Family Practitioner lists in Southall, West London (23). Of 1262 consecutively sampled white males from the original survey, blood samples
875 Human Genetics, 1997, 6, No. NucleicMolecular Acids Research, 1994, Vol. Vol. 22, No. 1 6
875
Table 4. Sequences of primers for amplification of human leptin receptor cDNA and for genotyping of five common leptin receptor gene variants
1Nucleotide
positions on human leptin receptor cDNA (13).
were available for DNA extraction from 97% of the males. Because extremes in populations provide the most power to detect genetic effects on quantitative traits, and because we wished to examine differences in genotype frequencies between obese and lean individuals, white males of British descent with BMI >28 kg/m2, defined here as obese, (range = 28.0–51.2 kg/m2; n = 190) or
