Download Polymorphism of the Tumor Necrosis Factor Alpha Gene and Waist

Mol. Cells, Vol. 18, No. 3, pp. 340-345
Molecules
and
Cells
KSMCB 2004
Polymorphism of the Tumor Necrosis Factor Alpha Gene and
Waist-Hip Ratio in Obese Korean Women
Jae-Young Um1,3, Byung-Ku Kang2, Si-Hyeong Lee2, Jo-Young Shin2, Seung-Heon Hong3, and
Hyung-Min Kim1,*
1
Department of Pharmacology, College of Oriental Medicine, Kyung Hee University, Seoul 130-701, Korea;
College of Oriental Medicine, Wonkwang University, Iksan 570-749, Korea;
3
Department of Oriental Pharmacy, College of Pharmacy, Wonkwang University, Iksan 570-749, Korea.
2
(Received July 5, 2004; Accepted August 13, 2004)
A number of candidate genes have been in implicated in
the pathogenesis of obesity in humans. Tumor necrosis
factor-alpha (TNFα) is expressed primarily in adipocytes, and elevated levels of this cytokine have been
linked to obesity and insulin resistance. Recently, the A
allele of a polymorphism at position –308 in the promoter region of TNFα (G-308A) has been shown to increase transcription of the gene in adipocytes. We therefore designed this study to test whether obese and nonobese subjects differ in terms of TNFα genotype distribution. We also investigated whether the genotypes affect anthropometric parameters, such as body mass
index (BMI). The study included 153 obese healthy
women and 82 non-obese women. Total fat mass and
percent body fat were determined by dual-energy X-ray
absorptiometry. Genomic DNA was extracted and used
for NcoI restriction fragment length polymorphismbased genotyping of TNFα. No differences were observed in allele and genotype frequencies between obese
and non-obese women, and no association of TNFα
polymorphism with BMI was observed for genotype in
the obese women. In addition, age, percent body fat,
BMI, and cholesterol levels did not vary with TNFα
genotype. However, waist-to-hip ratio (WHR) was significantly lower in subjects with TNFα GA or AA genotypes (0.94 ± 0.07 vs. 0.92 ± 0.03, P < 0.005). These results indicate that polymorphism at position –308 of the
TNFα promoter is not a significant factor for BMI, but
affects WHR in obese healthy Korean women.
Keywords: Koreans; Obese Women; Polymorphism; Tumor Necrosis Factor-Alpha; Waist-Hip Ratio.
* To whom correspondence should be addressed.
Tel: 82-2-961-9448; Fax: 82-2-968-1085.
E-mail: hmkim@khu.ac.kr
Introduction
Obesity is increasing in prevalence and is associated with
several adverse health problems, including type 2 diabetes,
hyperlipidemia, and hypertension. The influence of obesity on the development of obesity-related disorders is
complex and probably involves interactions between genetic, nutritional, and metabolic factors (Comuzzie and
Allison, 1998; Perusse and Bouchard, 1999).
The cytokine, tumor necrosis factor (TNF)α, acting as
a modulator of gene expression in adipocytes, has been
implicated in the development of insulin resistance and
obesity (Hotamisligil and Spiegelman, 1994). Fat tissue is
a significant source of endogeneous TNFα production,
and the expression of this cytokine is elevated in human
obesity in both adipose (Hotamisligil et al., 1995; Kern et
al., 1995) and muscle tissues (Saghizadeh et al., 1996).
Increased expression of TNFα strongly correlates with the
level of hyperinsulinemia (Hotamisligil et al., 1993) and
glucose disposal rate during application of the euglycemic
clamp technique (Saghizadeh et al., 1996).
The gene for human TNFα is located on the short arm
of chromosome 6 (Nedwin et al., 1985), and a G → A
substitution has been identified at position –308 upstream
from the transcription initiation site in the promoter region of the gene (Wilson et al., 1992). In vitro experiments have demonstrated that this NcoI restriction fragment length polymorphism (RFLP) increases transcriptional activation of the TNFα gene (Wilson et al., 1997).
Although controversial, the majority of the data support a
direct role for this biallelic polymorphism in the elevation
of TNFα levels observed in homozygotes for the −308 A
allele (Abraham and Kroeger, 1999). Recent studies of the
relationship between this polymorphism and insulin resistance have given inconsistent results (Deng et al., 1996;
Fernandez-Real et al., 1997; Hamann et al., 1995; Wal-
Jae-Young Um et al.
ston et al., 1999). Fernandez-Real et al. (1997) showed
that the polymorphism influenced insulin sensitivity via
an increase in body fat in a group of non-diabetic normotensive Spanish subjects. A recent study showed no
difference between type 2 diabetes mellitus patients and
healthy control subjects in the frequencies of alleles at –
308 (Hamann et al., 1995) and Walston et al. reported
that TNFα polymorphism at –308 site did not relate to
any traits of obesity and insulin resistance in a group of
non-diabetic subjects (Walston et al., 1999).
Because the etiology of human obesity is complex and
multifactorial, we excluded obese subjects with metabolic
disorders including diabetes and dyslipidemia, in order to
investigate only the relationship between TNF genotype
and obesity. We designed this study to test whether obese
and non-obese subjects differ in TNFα genotype distribution, and whether the genotypes affect the anthropometric
parameters of obese healthy women.
Materials and Methods
Subjects Subjects were healthy obese (body mass index, BMI ≥
25 kg/m2, range 25−39.1) women between the ages of 15−40
years. All subjects were nonsmokers and had no evidence of
cancer, liver, renal, hematological disease or metabolic disorder
other than obesity. A total of 153 women met all study criteria
and were enrolled. All subjects were normotensive, and subjects
with dyslipidemia (total cholesterol > 250 mg/dl, HDLcholesterol < 35 mg/dl) and/or diabetes were excluded. All
methods and procedures for the study were approved by the
institutional review board of our hospital. Each participant provided written informed consent.
Phenotypic measurements
Anthropometry Height (in cm) and weight (in kg) were
measured to calculate BMI as weight (kg)/height (m) squared.
Waist circumference (measured at the narrowest point
superior to the hip) was divided by the circumference of
the hip (measured at its greatest gluteal protuberance) to
obtain the waist-to-hip ratio (WHR).
Dual-energy X-ray absorptiometry Fat mass was determined by dual-energy X-ray absorptiometry (DEXA).
Genotyping Genomic DNA was extracted from whole
blood by an inorganic procedure (Miller et al., 1988). The
single base pair polymorphism at position –308 in the
promoter region of the TNFα gene was examined by the
NcoI (Takara, Japan) RFLP method (Cabrera et al., 1995).
The primers used were 5′-AGGCAATAGGTTTTGAGGGCCAT-3′ and 5′-TCCTCCCTGCTCCGA-TTCCG-3′.
341
Statistical analysis The mean levels of all numerical values were tested by Student’s t-test or ANOVA. Comparisons of genotypes and allele frequencies of the TNFα
genotypes between groups were carried out using Pearson’s chi-squared test. Odds ratios were calculated with
95% confidence intervals, and all statistical analyses were
performed with SPSS v10.00 (SPSS Inc.) statistical
analysis software. A P-value of less than 0.05 was considered statistically significant.
Results
Clinical characteristics of obese women The clinical
characteristics of the obese women and controls are presented in Table 1. 27.5% of the obese women were classified as BMI 25−26 (n = 42), 35.9% as BMI 27−29 (n =
55), and 36.6% as BMI 30−40 (n = 56). As expected,
weight, fat mass, percent body fat (PBF), and WHR differed between the three BMI groups (Table 1).
Frequencies of alleles and genotypes Genotype frequencies in all groups were in accordance with the HardyWeinberg equilibrium. The distribution of TNFα genotype in the obese women (BMI ≥ 25) was as follows; GG,
129 (84.3%); GA, 22 (14.4%); and AA, 2 (1.3%). This
was not significantly different from the distribution in the
non-obese women (BMI < 25): GG, 68 (82.9%); GA, 12
(14.6%); and AA, 2 (2.4%). The allele frequencies in the
obese women were: G, 280 (91.5%); and A, 26 (8.5%).
This also was not significantly different from the distribution in non-obese women: G, 1480 (90.2%); and A, 16
(9.8%) (Table 2). In addition, TNFα –308 G > A did not
show any association with obesity by logistic regression
analysis controlling for confounding factors.
Characteristics of obese women according to TNFα
genotype Table 3 describes the relationships between the
TNFα genotype and anthropometric parameters in obese
women. Interestingly, WHR was significantly lower in
women with GA + AA genotypes than in those with the
GG genotype (0.92 ± 0.03 vs. 0.94 ± 0.07, P = 0.004).
The remaining variables were similar in the GG and GA +
AA genotypes as well as among the three genotypes. We
also investigated the relationships between TNFα genotype and anthropometric parameters in non-obese women.
However, we did not find any differences between GG
and GA + AA genotypes in non-obese women (data not
shown).
TNFα polymorphism and BMI We failed to detect any
statistically significant association between TNFα polymorphism and BMI (Table 4). We note that the frequency
of the A allele was lower in the groups with BMI 27−40
kg/m2 than in the group with BMI 25−26 kg/m2, although
342
TNF-α Polymorphism and WHR in Obese Women
Table 1. Characteristics of obese women according to BMI (n = 153).
Non-obese
Controls
82
025.6 ± 7.5
062.2 ± 19.5
161.3 ± 5.2
163.7 ± 25.4
077.3 ± 38.8
019.5 ± 2.1
033.3 ± 2.6
00.85 ± 0.04
112.9 ± 9.5
072.9 ± 9.5
N (%)
Age (year)
Weight (kg)
Height (cm)
Total cholesterol (mg/dl)
Triglyceride (mg/dl)
Fat mass (kg)
PBF (%)
WHR
SBP (mm Hg)
DBP (mm Hg)
~26
42(27.5)
028.0 ± 9.8
064.4 ± 4.9
158.7 ± 6.5
182.9 ± 45.3
116.7 ± 112.0
023.2 ± 2.6
035.9 ± 2.9
00.89 ± 0.03
110.8 ± 11.6
069.1 ± 7.9
BMI (kg/m2)
27~29
55(35.9)
031.5 ± 10.0
070.8 ± 6.0
158.2 ± 6.0
174.6 ± 34.7
122.2 ± 79.5
026.1 ± 3.2
036.1 ± 6.2
00.93 ± 0.04
116.7 ± 17.2
071.3 ± 11.3
30~
56(36.6)
030.2 ± 11.3
085.7 ± 14.6
163.5 ± 7.0
196.2 ± 41.5
163.2 ± 140.6
034.4 ± 8.4
139.6 ± 6.0
10.98 ± 0.07
121.7 ± 11.5
077.4 ± 7.5
P*
0.258
< 0.001
< 0.001
0.027
0.086
< 0.001
0.001
< 0.001
0.083
0.023
Values are means ± SD; BMI, body mass index; PBF, percentage body fat; WHR, ratio of waist-to-hip circumstance; SBP, systolic blood pressure; DBP, diastolic blood pressure.
* By one-way ANOVA (among BMI groups).
Table 2. Genotype and allele frequencies of TNFα in non-obese and obese women.
Non-obese, n (%)
n = 82
Genotype
GG
GA
AA
Allele
G
A
Obese, n (%)
n = 153
P*
068 (82.9)
012 (14.6)
002 (02.4)
129 (84.3)
022 (14.4)
002 (01.3)
x 2 = 0.416,
P = 0.812
148 (90.2)
016 (09.8)
280 (91.5)
026 (08.5)
x 2 = 0.208,
P = 0.649
OR
(95% CI)
0.859 (0.447-1.652)
OR, odds ratio; CI, confidence interval.
* By x 2 test (two-sided).
the difference was not statistically significant. In obese
women with BMI ≥ 30 kg/m2, carriers of the GA + AA
genotypes tended to exhibit a lower frequency of the A
allele than did women with BMI < 30 kg/m2 (14.3% vs.
16.5%).
Discussion
The purpose of the present study was to determine whether
the G/A polymorphism of the TNFα gene is associated
with obesity and anthropometric parameters in women.
Obesity is a complex metabolic disorder with a strong
genetic component, and there are many candidate genes
for obesity and its related phenotypes. Most of these genes
are candidates because mutations in them cause rare genetic syndromes affecting adipocyte differentiation (Bouchard et al., 1998). Recent interest has focused on the role
of the TNFα gene in insulin-resistance and obesity. We
therefore examined the association between the polymorphism of TNFα and obesity without metabolic disease.
However, we did not find significant differences between
obese women and non-obese women in the frequencies of
TNFα polymorphism.
A number of polymorphisms have been identified in
TNFα including the TNFα –308G > A promoter polymorphism (Shin et al., 2004). However, our study focused
on the latter, which is particularly interesting because of
its known involvement in differential transcriptional activation (Wilson et al., 1997), elevated plasma TNFα levels,
and higher levels of TNF on stimulation in vivo and ex
vivo (Louis et al., 1998). Nevertheless, there should be
further studies of other polymorphisms of TNF genes
since all the single-nucleotide polymorphisms (SNP) in
TNF genes are in very strong linkage disequilibrium.
Interestingly, we found that WHR was lower in women
with GA + AA genotypes than in those with the GG genotype. This finding differs from those reported by Rosmond et al. (2001), who found no association between
TNFα and obesity (BMI) and body fat distribution (WHR
and abdominal sagittal diameter). Contrary to their findings, some studies have suggested an association of TNFα
Jae-Young Um et al.
343
Table 3. Characteristics of obese women according to TNFα genotype (n = 153).
Genotypes
GG
129 (84.3)
030.1 ± 11.3
074.8 ± 13.7
160.3 ± 6.5
186.2 ± 43.2
135.6 ± 115.3
028.7 ± 7.9
029.2 ± 4.1
037.5 ± 5.9
00.94 ± 0.07
119.0 ± 14.6
074.1 ± 9.7
N (%)
Age (year)
Weight (kg)
Height (cm)
Total cholesterol (mg/dl)
Triglyceride (mg/dl)
Fat mass (kg)
BMI (kg/m2)
PBF (%)
WHR
SBP (mm Hg)
DBP (mm Hg)
GA + AA
24 (15.7)
029.4 ± 08.20
072.4 ± 11.50
159.6 ± 9.10
178.9 ± 26.70
125.0 ± 84.22
026.6 ± 4.60
028.3 ± 2.20
036.8 ± 3.60
00.92 ± 0.03
112.2 ± 6.70
072.2 ± 6.7
P*
0.522
0.422
0.654
0.477
0.694
0.212
0.265
0.605
0.004
0.182
0.577
Values are means ± SD; BMI, body mass index; PBF, percentage body fat; WHR, ratio of waist-to-hip circumstance; SBP, systolic blood pressure; DBP, diastolic blood pressure.
* By t-test (GG genotype vs. the other genotypes).
Table 4. Frequencies of TNFα genotypes according to BMI in obese women (n = 153).
BMI (kg/m2)
Genotypes
P*
25−26, n (%)
27−29, n (%)
30−40, n (%)
GG
34 (81.0)
47 (85.5)
48 (85.7)
GA + AA
08 (19.0)
08 (14.5)
08 (14.3)
x2 = 0.496,
P = 0.780
< 30, n (%)
≥ 30, n (%)
81 (83.5)
48 (85.7)
16 (16.5)
08 (14.3)
x2 = 0.131,
P = 0.717
OR
(95% CI)
0.84 (0.34-2.12)
BMI, body mass index; OR, odds ratio; CI, confidence interval.
* By x 2 test (two-sided).
polymorphism with obesity in several European populations (Fernandez-Real et al., 1997; Herrmann et al., 1998).
However, the design of those studies and the ethnic populations examined were quite different from the present
study, and the results are hardly comparable. Moreover,
obesity in women (BMI ≥ 25) was not the main focus of
those studies.
It is well known that body fat distribution generally differs between men and women. Men often have upperbody obesity, whereas peripheral obesity is the most common form among women. In fact, several of the polymorphisms have a different impact on body fat in men and
women. Many studies suggest that Trp64Arg polymorphism in the β3-adrenoceptor is only associated with obesity in women (Arner and Hoffstedt, 1999). Gln27Glu
polymorphism in the β2-adrenoceptor has been linked to
obesity in women but not in men in a Swedish population
(Hellstrom et al., 1999). These data indicate that men and
women have different obesity genes. Therefore, further
studies are needed to investigate the association of TNFα
gene polymorphism with male obesity.
We selected TNFα, a candidate gene for obesity in
Pima Indian, mainly because of the reported overexpression of TNFα mRNA in the adipose tissue of rodents,
which are genetic models for insulin resistance and associated obesity (Hotamisligil et al., 1993). Levels of TNFα
mRNA and protein in the adipose tissue of obese human
subjects were found to be 2.5 times higher than in lean
controls and highly correlated with fasting insulin concentration (Hotamisligil et al., 1995). In addition, it was reported that substitution of adenine (A) for guanine (G) at
position –308 in the promoter region of TNFα enhanced
the transcription of this cytokine in cultured cells (Wilson
et al., 1997). Indeed, some of the previous studies suggest
that homozygotes for the –308 variant are more obese
than groups with other genotypes (Fernandez-Real et al.,
1997; Hoffstedt et al., 2000; Norman et al., 1995). We
therefore expected to find higher WHR in GA + AA subjects. Instead we observed lower WHR in women with
GA + AA genotypes. Nonetheless, these results tend to
confirm that several cytokines (e.g., TNFα, IL-1α, IL-6,
and IFN-γ) suppress food intake in animals and humans in
344
TNF-α Polymorphism and WHR in Obese Women
the same manner as leptin (Langstein and Norton, 1991).
It may be premature to view the polymorphism at position
−308 of the TNFα gene as having a key role in the pathogenesis of obesity since it is probably related to differences in TNFα synthesis, secretion, and activity. However,
we believe that the increased basal expression of TNFα as
a consequence of the polymorphism could be a protective
factor against increases WHR. Further studies are needed
to test this idea.
In summary, we observed no differences in allele and
genotype frequencies between obese women and nonobese women, and no association of TNFα polymorphism
with BMI for genotype in obese women. However, WHR
was significantly lower in subjects with TNFα GA and
AA genotype. These results suggest that TNFα promoter
polymorphism at position –308 does not have a significant influence on BMI, but affects WHR in obese healthy
Korean women.
Acknowledgment This research was supported by Wonkwang
University in 2003.
References
Abraham, L. J. and Kroeger, K. M. (1999) Impact of the -308
TNF promoter polymorphism on the transcriptional regulation of the TNF gene: relevance to disease. J. Leukoc. Biol.
66, 562−566.
Arner, P. and Hoffstedt, J. (1999) Adrenoceptor genes in human
obesity. J. Intern. Med. 245, 667−672.
Bouchard, C., Perusse, L., Leblac, C., Tremblay, A., and Theriault, G. (1998) Inheritance of the amount and distribution of
human body fat. Int. J. Obesity 12, 205−215.
Cabrera, M., Shaw, M. A., Sharples, C., Williams, H., Castes,
M., Convit, J., and Blackwell, J. M. (1995) Polymorphism in
tumor necrosis factor genes associated with mucocutaneous
leishmaniasis. J. Exp. Med. 182, 1259−1264.
Comuzzie, A. G. and Allison, D. B. (1998) The search for human obesity genes. Science 280, 1374−1377.
Deng, G. Y., Maclaren, N. K., Huang, H. S., Zhang, L. P., and
She, J. X. (1996) No primary association between the 308
polymorphism in the tumor necrosis factor alpha promoter
region and insulin-dependent diabetes mellitus. Hum. Immunol. 45, 137−142.
Fernandez-Real, J. M., Gutierrez, C., Ricart, W., Casamitjana,
R., Fernandez-Castaner, M., Vendrell, J., Richart, C., and
Soler, J. (1997) The TNF alpha gene NcoI polymorphism influences the relationship among insulin resistance, percent
body fat, and increased serum leptin levels. Diabetes 46,
1468−1472.
Hamann, A., Mantzoros, C., Vidal-Puig, A., and Flier, J. S.
(1995) Genetic variability in the TNF-alpha promoter is not
associated with type II diabetes mellitus (NIDDM). Biochem.
Biophys. Res. Commun. 211, 833−839.
Hellstrom, L., Large, V., Reynisdottir, S., Wahrenberg, H., and
Arner, P. (1999) The different effects of a Gln27Glu beta 2adrenoceptor gene polymorphism on obesity in males and in
females. J. Intern. Med. 245, 253−259.
Herrmann, S. M., Ricard, S., Nicaud, V., Mallet, C., Arveiler,
D., Evans, A., Ruidavets, J. B., Luc, G., Bara, L., Parra, H. J.,
Poirier, O., and Cambien, F. (1998) Polymorphisms of the
tumour necrosis factor-alpha gene, coronary heart disease
and obesity. Eur. J. Clin. Invest. 28, 59−66.
Hoffstedt, J., Eriksson, P., Hellstrom, L., Rossner, S., Ryden, M.,
and Arner, P. (2000) Excessive fat accumulation is associated with the TNF alpha-308 G/A promoter polymorphism in
women but not in men. Diabetologia 43, 117−210.
Hotamisligil, G. S. and Spiegelman, B. M. (1994) Tumor necrosis factor alpha: a key component of the obesity-diabetes link.
Diabetes 43, 1271−1278.
Hotamisligil, G. S., Shargill, N. S., and Spiegelman, B. M. (1993)
Adipose expression of tumor necrosis factor-alpha: direct
role in obesity-linked insulin resistance. Science 259, 87−91.
Hotamisligil, G. S., Arner, P., Caro, J. F., Atkinson, R. L., and
Spiegelman, B. M. (1995) Increased adipose tissue expression of tumor necrosis factor-alpha in human obesity and insulin resistance. J. Clin. Invest. 95, 2409−2415.
Kern, P. A., Saghizadeh, M., Ong, J. M., Bosch, R. J., Deem, R.,
and Simsolo, R. B. (1995) The expression of tumor necrosis
factor in human adipose tissue. Regulation by obesity, weight
loss, and relationship to lipoprotein lipase. J. Clin. Invest. 95,
2111−2119.
Langstein, H. N. and Norton, J. A. (1991) Mechanisms of cancer cachexia. Hematol. Oncol. Clin. North Am. 5, 103−123.
Louis, E., Franchimont, D., Piron, A., Gevaert, Y., SchaafLafontaine, N., Roland, S., Mahieu, P., Malaise, M., De
Groote, D., Louis, R., and Belaiche, J. (1998) Tumour necrosis factor (TNF) gene polymorphism influences TNF-alpha
production in lipopolysaccharide (LPS)-stimulated whole
blood cell culture in healthy humans. Clin. Exp. Immunol.
113, 401−406.
Miller, S. A., Dykes, D. D., and Polesky, H. F. (1988) A simple
salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. 16, 1215.
Nedwin, G. E., Naylor, S. L., Sakaguchi, A. Y., Smith, D.,
Jarrett-Nedwin, J., Pennica, D., Goeddel, D. V., and Gray, P.
W. (1985) Human lymphotoxin and tumor necrosis factor
genes: structure, homology and chromosomal localization.
Nucleic Acids Res. 13, 6361−6373.
Norman, R. A., Bogardus, C., and Ravussin, E. (1995) Linkage
between obesity and a marker near the tumor necrosis factoralpha locus in Pima Indians. J. Clin. Invest. 96, 158−162.
Perusse, L. and Bouchard, C. (1999) Genotype-environment
interaction in human obesity. Nutr. Rev. 57, S31-S37; discussion S37−S38.
Rosmond, R., Chagnon, M., Bouchard, C., and Bjorntorp, P.
(2001) G-308A polymorphism of the tumor necrosis factor
alpha gene promoter and salivary cortisol secretion. J. Clin.
Endocrinol. Metab. 86, 2178−2180.
Saghizadeh, M., Ong, J. M., Garvey, W. T., Henry, R. R., and
Kern, P. A. (1996) The expression of TNF alpha by human
muscle. Relationship to insulin resistance. J. Clin. Invest. 97,
1111−1116.
Shin, H. D., Park, B. L., Kim, L. H., Jung, J. H., Wang, H. J.,
Kim, Y. J., Park, H. S., Hong, S. J., Choi, B. W., Kim, D. J.,
Jae-Young Um et al.
and Park, C. S. (2004) Association of tumor necrosis factor
polymorphisms with asthma and serum total IgE. Hum. Mol.
Genet. 13, 397−403.
Walston, J., Seibert, M., Yen, C. J., Cheskin, L. J., and Andersen, R. E. (1999) Tumor necrosis factor α -238 and -308
polymorphisms do not associate with traits related to obesity
and insulin resistance. Diabetes 48, 2096−2098.
Wilson, A. G., di Giovine, F. S., Blakemore, A. I., and Duff, G.
345
W. (1992) Single base polymorphism in the human tumour
necrosis factor α (TNF α) gene detectable by NcoI restriction
of PCR product. Hum. Mol. Genet. 1, 353.
Wilson, A. G., Symons, J. A., McDowell, T. L., McDevitt, H.
O., and Duff, G. W. (1997) Effects of a polymorphism in the
human tumor necrosis factor α promoter on transcriptional
activation. Proc. Natl. Acad. Sci. USA 94, 3195−3199.