5α-Reductase Type 2 Is Constitutively Expressed in the Dermal Papilla and Connective Tissue Sheath of the Hair Follicle in Vivo But Not during Culture in Vitro1 (2024)

Search for other works by this author on:

Article history

ANDROGENS EXERT THEIR biological effects on their target cells as testosterone or dihydrotestosterone (DHT). The latter, the more potent androgen, is generated from testosterone by the enzyme steroid 5α-reductase (5αR). DHT binds to the androgen receptor (AR) with higher affinity than testosterone, and the resultant complex is more stable than the testosterone-AR complex. Thus, DHT plays a major role in the effects of androgens (1, 2).

There is much evidence to suggest that male-pattern baldness develops under the influence of androgens. Hypogonadal men or those castrated before puberty do not become bald, but such men begin to lose their hair when treated chronically with testosterone (3, 4). Moreover, XY individuals with testicular feminization, who have nonfunctional AR but normal male levels of circulating testosterone, have a female phenotype and do not develop male-pattern baldness (5, 6). Many target organs of androgens in men, including the prostate, are more responsive to DHT than to testosterone.

There are two isoforms of human 5αR, namely, type 1 (5αR1) and type 2 (5αR2). 5αR1, which is encoded by the SRD5A1 gene and is composed of 259 amino acids, has an optimum pH of 6–9, whereas 5αR2, encoded by the SRD5A2 gene and composed of 254 amino acids, has an optimum pH of 5.5. 5αR1 has been detected in various androgen-independent organs, such as the liver and brain (1). By contrast, 5αR2 is found predominantly in androgen-dependent organs, such as the epididymis and prostate (1). It is generally believed that 5αR1 and 5αR2 play a catabolic and an anabolic role, respectively, in the metabolism and action of androgens (1). 5αR2, rather than 5αR1, is thought to be important in the development of male-pattern baldness: male pseudohermaphrodites with 5αR2 deficiency do not develop male-pattern baldness (7–9). Furthermore, clinical studies have shown that a selective inhibitor of 5αR2, finasteride, promotes hair growth in men with male-pattern baldness in parallel with decreased levels of DHT in scalp skin (10, 11).

A hair follicle consists of epithelial components, which include the inner and outer root sheath (ORS) and hair shaft, and mesenchymal components, which include the dermal papilla (DP) and connective tissue sheath that form a three-dimensional tube. In the bulbar portion, proliferating cells of the primitive epithelial matrix envelope the DP. The DP and the upper portion of the follicle remain throughout the hair cycle, whereas the lower follicle undergoes a cyclic pattern of growth (anagen), regression (catagen), and resting (telogen). The interaction between cells of the DP and those of the matrix is considered to play an important role in this cycle. Moreover, in the morphogenesis of embryonic hair, the first inductive signal is thought to be of mesenchymal origin, instructing ectoderm to form a hair germ. In hair follicles of adult vibrissae, dermal papillae have the capacity to form new hair follicles (12).

In male-pattern baldness, the hair shaft gradually becomes thinner and the anagen phase becomes shorter during repeated hair cycles. DP cells have recently been shown to produce androgen-dependent biological factors that modulate the proliferation of follicular epithelial cells (13, 14).

The various findings described above suggest that the action of androgen in hair follicles is mediated via the DP. However, neither the localization of 5αR isozymes nor that of AR in human hair follicles has yet been clarified, in particular, in male-pattern baldness. To determine the localization of 5αR isozymes and AR in hair follicles, we examined the levels of the corresponding mRNAs in various portions of hair follicles that had been freshly dissected from balding and nonbalding scalp by semiquantitative RT-PCR. We also studied the expression of AR, 5αR1, and 5αR2 mRNAs and the activity of the isozymes in cultured mesenchymal cells to determine whether the various genes are expressed in culture.

Materials and Methods

Dissection of hair follicles into five portions

Hair follicles at the anagen phase were isolated from balding (five samples) and nonbalding (seven samples) scalp skin of 12 individuals who were undergoing plastic surgery for male-pattern baldness or resection of benign tumors after provision of written informed consent approved by the ethical committee of Oita Medical University (Table 1). Each hair follicle was dissected into five portions: the DP; the lower connective tissue sheath (LCTS); the matrix portion surrounding the DP that contained matrix cells (MX); the upper epithelial portion, which included the hair shaft and inner and ORS; and the upper connective tissue sheath, under a light microscope as described previously (16). In brief, each hair follicle was cut horizontally immediately above the tip of the DP. Then the LCTS of the bulbar portion was inverted and the loosely surrounding matrix portion was dissected from the DP. The DP was cut off from the inverted LCTS at the base of the papilla. The remaining upper epithelial and mesenchymal portions were loosely attached to each other, and they were easily separated mechanically (Fig. 1). Contamination of each portion by the others was confirmed histologically to be negligible, as demonstrated previously (16).

Cell culture

Individual dermal papillae were isolated from hair follicles of the anagen stage under a light microscope and put into explant culture (17, 18). Fibroblasts derived from the connective tissue sheath were also cultured in a similar manner. Cells were grown in DMEM supplemented with 10% FBS, 20 mml-glutamine, 50 U/mL penicillin, and 50 μg/mL streptomycin in a humidified atmosphere of 5% CO₂ in air at 37 C.

Semiquantitative analysis of the expression of mRNAs for the 5αR isozymes and AR by RT-PCR

The individual five portions dissected from about 20 hair follicles per skin sample were pooled to yield five samples per individual for analysis. Total RNA was extracted from each pooled sample and from cultured cells with ISOGEN (Nippon Gene, Tokyo, Japan) according to the manufacturer’s instructions. The RNA was dissolved in double-distilled water, which had been treated with diethyl pyrocarbonate, and stored at −70 C. cDNA was synthesized as follows. One microgram of total RNA and 0.5 μg of oligo primers (Promega Corp., Madison, WI) were heated to 70 C for 10 min in 17 μL of diethyl pyrocarbonate-treated double-distilled water. After chilling on ice for 2 min, the mixture was supplemented with the reaction buffer to give a final volume of 40 μL. The final reaction mixture included 400 U Moloney murine leukemia virus reverse transcriptase (Life Technologies, Rockville, MD), 10–40 U RNase inhibitor (WAKO, Osaka, Japan), 0.5 mm dNTPs (TAKARA, Kyoto, Japan), 50 mm Tris-HCl (pH 8.3), 75 mm KCl, 3 mm MgCl₂, and 10 mm dithiothreitol. The final mixture was incubated at 42 C for 1 h and then heated at 70 C for 10 min to inactivate the enzyme. The cDNA was amplified by PCR, with 30 cycles, in a thermal cycler (ASTEC, f*ckuoka, Japan). The 25 μL aliquots of the reaction mixture for PCR contained 0.65 U of Taq DNA polymerase (TAKARA), 200μ m dNTPs, 0.5 μm sense and antisense primers, 50 mm KCl, 10 mm Tris-HCl (pH 8.3), 1 mm MgCl₂, and 0.5 μL of cDNA as template. The nucleotide sequences of primers used for amplification were as follows: For AR, the sense and antisense primers were 5′-CTCTCTCAAGAGTTTGGATGGCT-3′ and 5′-CACTTGCACAGAGATGATCTCTGC-3′, respectively; for 5αR1, they were 5′-CAATGGCGCTTCTCTATGG-3′ and 5′-TACACACAGCACCTGACACG-3′; for 5αR2, they were 5′-TGAGGTTACATGCTGCTTGC-3′ and 5′-TCCAATTACAAGCGTTCGG-3′; and for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), they were 5′-CCCATCACCATCTTCCAG-3′ and 5′-CCTGCTTCACCACCTTCT-3′. Each cycle of PCR was programmed for denaturation (94 C; 1 min), annealing (55 C; 1 min) and extension (72 C; 1 min), with the exception that the first denaturation step was allowed to proceed for 5 min, and the final extension step for 3 min. Twenty microliters of each reaction mixture after PCR were combined with 100 μL of 0.4 N NaOH, neutralized with an equal volume of cold 2 m ammonium acetate, and transferred to a nylon membrane (Hybond N+; Amersham Pharmacia Biotech, Tokyo, Japan) through a slot window of a blotting apparatus (Bio-Rad Laboratories, Inc., Hercules, CA) according to the manufacturer’s instructions. After UV-cross-linking at 312 nm, the amplified fragments on the membrane were allowed to hybridize with oligonucleotide probes specific for AR or 5αR1 or 5αR2 cDNAs, which had been conjugated with digoxigenin with a DIG oligonucleotide tailing kit (Roche Diagnostics, Mannheim, Germany), overnight at 54 C. The hybridized probes were detected on x-ray film with a DIG luminescence detection kit (Roche Diagnostics) according to the manufacturer’s instruction. The products of PCR specific for GAPDH were subjected to electrophoresis on a 1.5% agarose gel that contained ethidium bromide and photographed under UV light at 254 nm. The bands on x-ray films and agarose gels were recorded as digital images and intensities were determined with NIH image computer software (version 1.57). The intensity of each band specific for AR, 5αR1, and 5αR2 was normalized by reference to the band specific for GAPDH. Student’s t test was used for statistical analysis.

Assay of 5α-reductase activity

Cells were seeded at 1 × 10⁵ cells/well in a 24-well tissue culture dish and were cultured to subconfluence in DMEM supplemented with 10% FBS for 2 days. Then after cells had been washed twice with DMEM, they were incubated with DMEM that contained 20 nm 1α,2α-³H[ N]-testosterone (SA 51 Ci/mmol, NEN Life Science Products, Boston, MA) as substrate for 6 h in a humidified atmosphere of 5% CO₂ in air at 37 C, in the presence or absence of MK386, an inhibitor of 5αR1, or MK906, an inhibitor of 5αR2, which were kindly provided by Merck & Co., Inc. (Rahway, NJ). The reaction was stopped by addition of four volumes of chloroform-methanol (2/1, vol/vol) that contained 10 μg each of the following nonradioactive carrier steroids that were obtained from Sigma (St. Louis, MO): androstenedione, testosterone, DHT, androstanedione, androstane-3α,17β-diol, androstane-3β,17β-diol and androsterone. The extracted steroids were analyzed by thin-layer chromatography, and the purity of DHT was confirmed by high-performance liquid chromatography. DHT was actually the only 5α-reduced steroid generated under our conditions.

Results

Expression of genes for AR, 5αR1, and 5αR2 in various portions of hair follicles

We compared the levels of expression of GAPDH mRNA semiquantitatively among the five portions of hair follicles from each individual by RT-PCR using serially diluted aliquots of each sample of cDNA as the template. The diluted samples of cDNA that generated similar amounts of products of PCR specific for GAPDH in the case of each of the five portions were used for amplification by PCR of the genes for AR, 5αR1, and 5αR2. The levels of expression of AR, 5αR1, and 5αR2 mRNAs were assessed in 12 individuals. AR and 5αR1 mRNAs were expressed in all five portions from almost all individuals at similar respective levels (Figs. 2 and 3³). By contrast, 5αR2 mRNA was expressed in the mesenchymal portions, such as the DP and the upper and lower connective tissue sheaths, but hardly any was expressed in epithelial portions (Fig. 4). We also compared the levels of expression of AR, 5αR1, and 5αR2 mRNA obtained from mesenchymal portions between balding (four samples) and nonbalding (six samples) scalp. There were no statistically significant differences when cDNA concentrations were adjusted to appropriate levels for comparisons (P > 0.4; data not shown).

Expression of genes for AR, 5αR1, and 5αR2 in cultured DP cells and connective tissue sheath cells

To investigate whether the expression in vivo of AR, 5αR1, and 5αR2 by the mesenchymal cells continued in culture, we performed RT-PCR with total RNA extracted from cultured DP cells and connective tissue sheath cells after five or six passages. The AR and 5αR1 mRNAs were easily amplified (Fig. 5), but no obvious products of PCR were detected in the case of 5αR2 (data not shown in the figure). Again, levels of expression did not differ between cells that originated from balding scalp and those that originated from nonbalding scalp.

Effects of inhibitors of types 1 and 2 isozymes on the 5α-reductase activity of cultured DP cells and connective tissue sheath cells

To confirm that 5αR1 was also functionally predominant in cultured DP cells and connective tissue sheath cells, we examined the effects of inhibitors specific for each isozyme on 5αR activity in both kinds of cells obtained from three balding and three nonbalding individuals. MK386, an inhibitor of 5αR1, decreased the 5αR activity in a dose-dependent manner, and finasteride, an inhibitor of 5αR2, had hardly any effect (Fig. 6).

Discussion

In an attempt to clarify the role of androgen in the development of male-pattern baldness, it is necessary to identify the part of the hair follicle that is a target for androgen because the follicle has a complex structure. The localization of isozymes of 5αR in the hair follicle is of particular interest because the isozymes are associated with the anabolism or catabolism of androgen in various tissues. At present, the physiological role of 5αR isozymes in the hair follicle is poorly understood.

Our present study revealed that AR mRNA was expressed in almost all the dissected portions of follicles, suggesting that the expression of AR alone does not necessarily imply responsiveness to androgen. In one earlier study, immunohistochemical staining of AR revealed that AR was restricted to DP cells (19), but another study found that almost all components of hair follicles could be positively stained for AR (20). These conflicting results are probably due to differences in the specificities of the antibodies used. Levels of AR in cultured DP cells from balding scalp have been reported to be slightly higher than those from nonbalding scalp (21). In addition, AR mRNA was found to be expressed at a higher level in cultured DP cells from androgen-responsive hair follicles, such as beard and axillary hairs, than in those from follicles of occipital hairs (22). However, we failed to detect any quantitative differences in terms of the level of expression of AR mRNA in DP cells between balding and nonbalding scalp.

We found that 5αR1 mRNA was also expressed in almost all portions of the hair follicle, a result that might reflect previous reports that 5αR1 is the main isozyme in nongenital skin (23–25). Our finding of the exclusive expression of 5αR1 and little 5αR2 in epithelial portions is compatible with previous reports on the expression of transcripts (26) and the characteristics of the 5αR1 activity (27) in plucked (containing epithelial portions only) and freshly isolated hair follicles (28). Immunohistochemical studies have also shown that 5αR1 is present in most portions of hair follicles (29, 30). It remains unknown whether 5αR1 is associated with androgen action in tissues other than sebaceous glands (31) and apocrine glands (32), both of which are androgen target tissues and contain the 5αR1 isozyme only.

The present study demonstrated that 5αR2 mRNA was expressed exclusively in mesenchymal portions of follicles. This observation might indicate that 5αR2 not only in the DP but also in the connective tissue sheath plays a pivotal role in the action of androgen on hair growth as it does in the external genitalia and prostate (9). Indeed, in a coculture system, DP cells from the beard, which have an isozyme with characteristics specific to 5αR2 (33), stimulate the proliferation of ORS cells in the presence of testosterone (34), whereas DP cells from balding scalp of the stumptailed macaque inhibit the proliferation of cocultured epithelial cells under similar conditions (13). Recently it was shown that 5αR activity in the hair follicle is concentrated in the DP, although the specific type of isozyme was not reported (35). Other authors have stated that 5αR2 is the major isozyme in the freshly isolated dermal papillae of occipital hair follicles (36). The results of immunohistochemical localization of 5αR2 in hair follicles in previous studies are entirely different from our present results: 5αR2 was localized in very narrow regions of the epithelial portion in two studies (37, 38) and at lower intensity in the ORS in another study (30) but not in dermal papillae.

The ability of mesenchymal cells to express 5αR2 mRNA was lost in culture, whereas the ability of expressing AR and 5αR1 mRNAs was retained. The effects of the type-specific inhibitors of 5αR supported these findings. After many passages, cultured DP cells fail to induce the formation of hair follicles (39). Similarly, cultured fibroblasts from cases of benign prostatic hyperplasia no longer express either type of 5αR isozyme after several passages (40). To retain patterns of mRNA expression by mesenchymal cells in culture, some factors produced by epithelial cells might be necessary (41). Thus, mesenchymal cells, cultured alone, are not necessarily suitable for studies of the mechanisms of action of androgen in the hair follicle.

We did not find any differences in terms of the expression of AR, 5αR1 and 5αR2 mRNAs, or in terms of 5αR activity between hair follicles from balding and nonbalding scalp. By contrast, in a previous study, hair follicles from balding scalp had higher levels of AR, 5αR1, and 5αR2, both in terms of mRNA expression and in terms of enzymatic activity, than those from nonbalding scalp (30). The reasons for these discrepancies remain unknown. The role of 5αR at a cellular level in the development of male-pattern baldness remains unknown although clinical findings clearly indicate that 5αR2 is a prerequisite for this condition (9, 10).

¹

This work was supported in part by Taisho Pharmaceutical Co., Ltd. (Tokyo, Japan).

Copyright © 2001 by The Endocrine Society

Issue Section:

Article

Email alerts

Citing articles via

More from Oxford Academic