Download Insulin gene polymorphism and premature male pattern baldness in

Clinical Science (1999) 96, 659ñ662 (Printed in Great Britain)
Insulin gene polymorphism and premature
male pattern baldness in the general population
Justine A. ELLIS, Margaret STEBBING and Stephen B. HARRAP
Department of Physiology, University of Melbourne, Parkville, Victoria 3052, Australia
A
B
S
T
R
A
C
T
Insulin is found in hair follicles and may play a role in the regulation of androgen metabolism and
the hair growth cycle, which are relevant to the loss of scalp hair known as male pattern baldness.
An excess of dihydrotestosterone on balding scalp indicates that the condition is androgen
dependent. Premature male pattern baldness may be the male phenotype of familial polycystic
ovary syndrome, a condition characterized by high levels of androgens and insulin that has been
linked to insulin gene polymorphism. Therefore, we studied possible associations between
relevant insulin gene polymorphisms and premature male pattern baldness in the general
community. We examined the distribution of three dimorphic restriction fragment length
polymorphisms : HphI, PstI and FokI in cases consisting of 56 men aged 18–30 years with significant
baldness, and in 107 control men aged 50 years or more with no indication of baldness. No
significant differences between cases and controls in allele, genotype or haplotype frequencies
were identified. We conclude that, in the general population, the insulin gene is not associated
with premature male pattern baldness.
INTRODUCTION
Male pattern baldness (MPB) is defined as the genetically
predisposed loss of scalp hair which follows a defined
pattern and is androgen dependent [1,2]. The mode of
inheritance of MPB is often cited as autosomal dominant
[3], but a polygenic aetiology appears more likely [4,5].
The genetic defect(s) underlying this condition are
unknown, although clues are provided by the excess
production of dihydrotestosterone on balding scalp [6].
Dihydrotestosterone is formed from the reduction of
testosterone by the action of the enzyme 5α-reductase.
The genes encoding the two forms of 5α-reductase were
for some time suspected of involvement in MPB, but we
have recently shown that common variants of these genes
do not appear to be associated with this disorder [5].
Polycystic ovary syndrome (PCOS) is a familial
endocrine disorder affecting as many as 10 % of females
of reproductive age. It is characterized by hyper-
androgenism and hyperinsulinaemia. Peripheral insulin
resistance appears to be the main cause of the hyperinsulinaemia, and PCOS patients have a 7-fold increased
risk of Type II (non-insulin-dependent) diabetes [7]. In
males from families with PCOS, an increase in the
prevalence of premature MPB (significant balding before
30 years of age) in an autosomal dominant pattern of
inheritance [8,9] has led investigators to suggest that
premature MPB is the male phenotype of the disease
[9,10]. The genetic disorder(s) underlying PCOS may
therefore provide clues for MPB. Recently, linkage and
association of familial PCOS with the insulin gene (INS)
regulatory variable number of tandem repeats (VNTR)
was reported [7].
Apart from their association in PCOS [11–13], androgen and insulin pathways appear to be closely linked
in both women and men [14]. There is a strong association
between circulating levels of insulin and sex hormonebinding globulin, and possibly between insulin and
Key words : androgen, androgenetic alopecia, hair, polycystic ovary syndrome.
Abbreviations : MPB, male pattern baldness ; PCOS, polycystic ovary syndrome ; PCR, polymerase chain reaction ; RFLP,
restriction fragment length polymorphism ; VNTR, variable number of tandem repeats.
Correspondence : Professor Stephen B. Harrap.
# 1999 The Biochemical Society and the Medical Research Society
659
660
J. A. Ellis and others
testosterone in normal, obese and diabetic men [14–20].
Furthermore, insulin found in hair follicles appears to
play a role in the control of the growth cycle of the hair
[21].
In view of its linkage with PCOS and potentially
important effects on androgen metabolism generally, the
INS gene is a good candidate for MPB. Importantly, the
INS polymorphism linked to PCOS is thought to have
functional effects on INS gene expression [22]. Therefore,
we performed a population-based case-control association study of the INS gene in premature MPB using
three dimorphic restriction fragment length polymorphisms (RFLPs), one of which is in strong allelic
association with the VNTR [7].
METHODS
Sampling
The details of recruitment and the characteristics of
participants have been reported previously [5]. Approximately 3000 Caucasian individuals were recruited as part
of the Victorian Family Heart Study. This is a general
population survey of coronary risk factors and consists
of approximately 800 families, comprising a mother,
father and at least one natural offspring. Baldness survey
questionnaires were sent to all males (approximately one
half of the sample number) asking them to detail their
degree of baldness, if any, in terms of the Hamilton
baldness scale [2,23]. The survey described all categories
of the scale, and provided diagrams to assist in the selfassessment. Validation of this self-reported questionnaire
approach has been reported previously [5].
A total of 1073 men returned the completed baldness
questionnaires. This group contained 529 sons aged 18 to
30 years and 544 fathers aged 50 years and above.
Cosmetically significant baldness of type III or greater
[23] was reported by 58 (11.0 %) sons aged 18 to 30 years.
These individuals were defined as cases and DNA was
available for analysis in 56 of these subjects. A total of 114
fathers aged 50 and above (21.0 %) reported no baldness
(Hamilton baldness score of I). Two individuals were
identified as fathers of young balding men used in the
case group, and were therefore excluded from the study.
DNA was available from 107 of the remaining individuals
who comprised the control group for our genetic
analyses.
Analysis of RFLPs of the INS gene
DNA was extracted from leucocytes using standard
phenol\chloroform techniques. Approximately 50 ng of
DNA from each individual was used in polymerase chain
reactions (PCRs). The HphI RFLP, located 573 bp distal
to the regulatory VNTR and k23 bp from the start site
of the INS gene, was detected by amplification of DNA
# 1999 The Biochemical Society and the Medical Research Society
using primers ins04 (5h-TCCAGGACAGGCTGCATCAG-3h) and ins05 (5h-AGCAATGGGCGGTTGGCTCA-3h) [24]. DNA was added to a mix containing
0.5 µM of each primer, 1iPCR buffer (Perkin–Elmer
Applied Biosystems, Norwalk, CT, U.S.A.), 250 µM
dNTP (Perkin–Elmer Applied Biosystems), 1.5 mM
MgCl (Perkin-Elmer Applied Biosystems) and 1 unit
#
Amplitaq Gold DNA polymerase (Perkin–Elmer Applied Biosystems), to give a total reaction volume of
20 µl. The thermal conditions required for the reaction
were 95 mC for 10 min (for activation of the Amplitaq
Gold enzyme), followed by 35 cycles of 95 mC for 1 min,
64 mC for 1 min and 72 mC for 1 min, followed by a final
extension time of 72 mC for 10 min. Products were then
digested by the addition of 5 units of HphI restriction
endonuclease (New England Biolabs, Beverly, MA,
U.S.A.) in the presence of 1ibuffer ‘ 4 ’ (New England
Biolabs) at 37 mC for 1 h. Digested products were then
electrophoresed through a 2 % agarose gel (Type II-A,
Sigma, St. Louis, MO, U.S.A.) containing ethidium
bromide. DNA was visualized by placing the gel on an
UV transilluminator.
The PstI RFLP located j1127 bp from the INS start
site was detected in a similar manner. Primers ins01 (5hGAAGGAGGTGGGACATGT-3h) and ins02 (5hGCTGGTTCAAGGGCTTTA-3h) were used to amplify
DNA [24]. Again, approximately 50 ng of DNA was
added to a reaction mixture as described above. Thermal
conditions required for the reaction were 95 mC for
10 min, followed by 35 cycles of 95 mC for 1 min, 55 mC
for 1 min and 72 mC for 1 min, followed by a final
extension of 72 mC for 10 min. PCR products were
digested with 5 units of PstI (Boehringer Mannheim,
Indianapolis, IN, U.S.A.) in the presence of 1ibuffer
‘ H ’ (Boehringer Mannheim) at 37 mC for 1 h. Similarly,
the FokI RFLP located j1428 bp from the INS start site
was detected by amplification of DNA with primers
ins13 (5h-TAAAGCCCTTGAACCAGC-3h) and DS02
(5h-CAGCCCAGCCTCCTCCCTCCACA-3h) [24].
Thermal conditions required were 95 mC for 10 min,
followed by 35 cycles of 95 mC for 1 min, 62 mC for 1 min
and 72 mC for 1 min, followed by a final extension of
72 mC for 10 min. PCR products were digested with 5
units of FokI (Boehringer Mannheim) in the presence of
1ibuffer ‘ M ’ (Boehringer Mannheim) at 37 mC for 1 h.
Digested products were electrophoresed and visualized
as above.
RESULTS
Lack of association of the INS RFLPs with
baldness
Table 1 summarizes the distribution of the resulting
alleles after digestion of PCR products by HphI, PstI and
Male pattern baldness and the insulin gene
Table 1 Distribution of INS RFLP alleles in young bald men (cases) and older non-bald men
(controls)
Values are n (%). A represents the uncut alleles and B represents the cut alleles.
Cases
Controls
Hph I RFLP alleles
Pst I RFLP alleles
Fok I RFLP alleles
A
B
A
B
A
B
34 (30.4)
63 (29.4)
78 (69.6)
151 (70.6)
84 (75.0)
156 (72.9)
28 (25.0)
58 (27.1)
83 (74.1)
155 (72.4)
29 (25.9)
59 (27.6)
Distribution of genotypes for each INS RFLP analysed in young bald men (cases) and older non-bald men (controls)
Values are n ( %). A represents the uncut alleles and B represents the cut alleles.
Table 2
Hph I RFLP genotypes
Cases
Controls
Pst I RFLP genotypes
Fok I RFLP genotypes
AA
AB
BB
AA
AB
BB
AA
AB
BB
8 (14.3)
7 (6.5)
18 (32.1)
49 (45.8)
30 (53.6)
51 (47.7)
35 (62.5)
56 (52.3)
14 (25.0)
44 (41.1)
7 (12.5)
7 (6.5)
34 (60.7)
56 (52.3)
15 (26.8)
43 (40.2)
7 (12.5)
8 (7.5)
FokI. No significant differences were found between
cases and controls for either allele of each RFLP (HphI :
χ# l 0.029, df l 1, P l 0.86 ; PstI : χ# l 0.168, df l 1,
P l 0.68 ; FokI : χ# l 0.099, df l 1, P l 0.75). The
possible individual genotypes for each RFLP (i.e. AA,
AB or BB) were also compared in cases and controls. The
distribution of genotypes is shown in Table 2. Again,
no significant differences in genotypes were detected
between groups (HphI : χ# l 4.33, df l 2, P l 0.11 ; PstI :
χ# l 4.89, df l 2, P l 0.087 ; FokI : χ# l 3.26, df l 2,
P l 0.20). Finally, we analysed haplotypes of the three
INS RFLPs in cases and controls. A total of 27 different
haplotypes resulting from combinations of all alleles of
the three RFLPs was possible. No significant differences
in haplotype distributions were demonstrated between
the groups (χ# l 10.26, df l 26, P l 0.997) (results not
shown).
DISCUSSION
We have shown that the INS gene does not appear to play
a role in the aetiology of premature MPB in the general
population. Our primary interest was in a polymorphism
of functional relevance to INS gene expression. The INS
VNTR, located within the promoter region of the
INS gene, is thought to regulate insulin expression [22]
and it is this region which has been linked to PCOS. In
our study we utilized the HphI RFLP which is located
573 bp 3h to the VNTR. Our marker has been shown to
be in very close allelic association with the INS VNTR
and also in strong association with PCOS [7]. Nevertheless, our analysis showed that the distribution of
alleles and genotypes relevant to the HphI RFLP was
indistinguishable statistically between subjects with
premature MPB and controls. Additionally, we analysed
two RFLPs 3h to the INS gene (FokI and PstI) [24] to
investigate the possibility of functional mutations elsewhere in the INS gene. The FokI and PstI RFLPs were in
strong allelic association but were not associated with
MPB in our population.
Our interest in the INS gene arose because of reported
genetic linkage with PCOS. PCOS is reportedly one of
the most common female endocrine disorders [12], and
may be expressed in males as premature MPB [9,10]. The
negative findings in this study might be explained if, in
our population, PCOS is not common, premature MPB
is not a male phenotype of PCOS or the INS gene is not
linked with PCOS.
The prevalence of PCOS in the Australian population
has not been reported but the generally accepted figure is
5–10 % of women of reproductive age [12]. In comparison, the overall prevalence of premature MPB in our
participants was 11 %. In males, the phenotypic expression of PCOS related to hair growth includes
premature baldness and excessive hairiness. In firstdegree male relatives of female cases of PCOS, 19.7 %
had early baldness and\or excessive hairiness, compared
with 6.5 % of first-degree male relatives of controls (P
0.01) [9]. However, when premature balding is separated
from excessive hairiness, the percentage of first-degree
male relatives of PCOS patients was only 11.1 %. It is
possible that the earlier appearance of MPB in firstdegree relatives is secondary to high androgen levels per
se rather than sharing a common pathophysiology related
to the INS gene.
Further independent confirmation of linkage between
INS VNTR and PCOS has not yet been reported.
# 1999 The Biochemical Society and the Medical Research Society
661
662
J. A. Ellis and others
Interestingly, the LOD score achieved by Waterworth
et al. [7] was borderline (maximum LOD score l 3.25)
[25] and was observed in only 9 of 17 families.
It should be noted that there are several limitations to
this type of association study, such as the possibility of
population stratification between the case and control
groups, which we have discussed previously [5]. It is
unlikely, however, that such limitations are masking an
association of MPB to INS, since population stratification
is normally associated with a type I, or false-positive,
error. In addition, we have surveyed three separate
RFLPs of the INS gene and all produce a similar result.
We conclude that the gene encoding insulin does not
appear to be associated with premature MPB in the
general population. This does not necessarily indicate
that MPB and PCOS are not in some way associated.
However, given the link between the androgen and
insulin pathways, and the role of insulin in hair growth
cycle regulation, other genes involved in the insulin
pathway may be worthy of investigation.
ACKNOWLEDGMENTS
We thank Dr John Hopper (Australian NHMRC Twin
Registry), Dr Graham Giles (Collaborative Cohort
Study, Health 2000), the general practitioners and research nurses for their contributions to subject recruitment and Angela Lamantia for her technical assistance
and DNA extraction. This work was supported in part
by the Victorian Health Promotion Foundation.
REFERENCES
1 Hamilton, J. B. (1942) Male hormone stimulation is a
prerequisite and an incitant in common baldness. Am. J.
Anat. 71, 451–480
2 Hamilton, J. B. (1951) Patterned loss of hair in man : types
and incidence. Ann. N.Y. Acad. Sci. 53, 708–728
3 Osborn, D. (1916) Inheritance of baldness. J. Hered. 7,
347–355
4 Kuster, W. and Happle, R. (1984) The inheritance of
baldness : two B or not two B ? J. Am. Acad. Dermatol. 11,
921–926
5 Ellis, J. A., Stebbing, M. and Harrap, S. B. (1998) Genetic
analysis of male pattern baldness and the 5α-reductase
genes. J. Invest. Dermatol. 110, 849–853
6 Dallob, A. L., Sadick, N. S., Unger, W. et al. (1994) The
effect of finasteride, a 5α-reductase inhibitor, on scalp skin
testosterone and dihydrotestosterone concentrations in
patients with male pattern baldness. J. Clin. Endocrinol.
Metab. 79, 703–706
7 Waterworth, D. M., Bennett, S. T., Gharani, N. et al.
(1997) Linkage and association of insulin gene VNTR
regulatory polymorphism with polycystic ovary
syndrome. Lancet 349, 986–990
8 Ferriman, D. and Purdie, A. W. (1979) The inheritance of
polycystic ovarian disease and a possible relationship to
premature balding. Clin. Endocrinol. 11, 291–300
9 Lunde, O., Magnus, P., Sandvik, L. and Hoglo, S. (1989)
Familial clustering in the polycystic ovarian syndrome.
Gynecol. Obstet. Invest. 28, 23–30
10 Carey, A. H., Chan, K. L., Short, F., White, D.,
Williamson, R. and Franks, S. (1993) Evidence for a single
gene effect causing polycystic ovaries and male pattern
baldness. Clin. Endocrinol. 38, 653–658
11 Dunaif, A., Segal, K. R., Shelly, D. R., Green, G.,
Dobrjansky, A. and Licholai, T. (1992) Evidence for
distinctive and intrinsic defects in insulin action in
polycystic ovary syndrome. Diabetes 41, 1257–1266
12 Dunaif, A. (1995) Hyperandrogenic anovulation (PCOS) :
a unique disorder of insulin action associated with an
increased risk of non-insulin-dependent diabetes mellitus.
Am. J. Med. 98 (Suppl. 1A), 33S–39S
13 Norman, R. J., Masters, S. and Hague, W. (1996)
Hyperinsulinemia is common in family members of
women with polycystic ovary syndrome. Fertil. Steril. 66,
942–947
14 Pasquali, R., Casimirri, F., De Iasio, R. et al. (1995) Insulin
regulates testosterone and sex hormone-binding globulin
concentrations in adult normal weight and obese men.
J. Clin. Endocrinol. Metab. 80, 654–658
15 Birkeland, K. I., Hanssen, K. F., Torjesen, P. A. and
Vaaler, S. (1993) Level of sex hormone-binding globulin is
positively correlated with insulin sensitivity in men with
type 2 diabetes. J. Clin. Endocrinol. Metab. 76, 275–278
16 Ebeling, P., Stenman, U.-H., Seppala, M. and Koivisto,
V. A. (1995) Acute hyperinsulinemia, androgen
homeostasis and insulin sensitivity in healthy man.
J. Endocrinol. 146, 63–69
17 Ibanez, L., Potau, N., Georgopoulos, N., Prat, N.,
Gussinye, M. and Carrascosa, A. (1995) Growth hormone,
insulin-like growth factor-I axis, and insulin secretion in
hyperandrogenic adolescents. Fertil. Steril. 64, 1113–1119
18 Haffner, S. M., Shaten, J., Stern, M. P., Smith, G. D. and
Kuller, L. (1996) Low levels of sex hormone-binding
globulin and testosterone predict the development of noninsulin-dependent diabetes mellitus in men. Am. J.
Epidemiol. 143, 889–897
19 Katsuki, A., Sumida, Y., Murashima, S. et al. (1996) Acute
and chronic regulation of serum sex hormone-binding
globulin levels by plasma insulin concentrations in male
non-insulin-dependent diabetes mellitus patients. J. Clin.
Endocrinol. Metab. 81, 2515–2519
20 Pasquali, R., Macor, C., Vicennati, V. et al. (1997) Effects
of acute hyperinsulinemia on testosterone serum
concentrations in adult obese and normal-weight men.
Metabolism 46, 526–529
21 Philpott, M. P., Sanders, D., Westgate, G. E. and Kealey,
T. (1994) Human hair growth in vitro : a model for the
study of hair follicle biology. J. Dermatol. Sci. 7 (Suppl.),
S55–S72
22 Kennedy, G. C., German, M. S. and Rutter, W. J. (1995)
The minisatellite in the diabetes susceptibility locus
IDDM2 regulates insulin transcription. Nat. Genet. 9,
293–298
23 Norwood, O. T. (1975) Male pattern baldness :
classification and incidence. South Med. J. 68, 1359–1365
24 Julier, C., Hyer, R. N., Davies, J. et al. (1991) InsulinIGF2 region on chromosome 11p encodes a gene
implicated in HLA-DR4-dependent diabetes susceptibility.
Nature (London) 354, 155–159
25 Lander, E. and Kruglyak, L. (1995) Genetic dissection of
complex traits : guidelines for interpreting and reporting
linkage results. Nat. Genet. 11, 241–247
Received 4 January 1999/12 March 1999; accepted 19 March 1999
# 1999 The Biochemical Society and the Medical Research Society