Professional Documents
Culture Documents
Sanchez Et Al 2018 5 Reductase Isozymes and Aromatase Mrna Levels in Plucked Hair PDF
Sanchez Et Al 2018 5 Reductase Isozymes and Aromatase Mrna Levels in Plucked Hair PDF
Sanchez Et Al 2018 5 Reductase Isozymes and Aromatase Mrna Levels in Plucked Hair PDF
https://doi.org/10.1007/s00403-017-1798-0
ORIGINAL PAPER
Received: 21 March 2017 / Revised: 20 November 2017 / Accepted: 21 November 2017 / Published online: 28 November 2017
© Springer-Verlag GmbH Germany, part of Springer Nature 2017
Abstract
Female pattern hair loss (FPHL) is an important hair disorder, especially when young women are affected. However, phar-
macological treatments are not successful in all women. Androgens, especially dihydrotestosterone (DHT), may play a role
in FPHL, but many women with this disorder have normal serum androgen levels. It therefore appears that hair follicle levels
of DHT depend on in situ testosterone (T) metabolism. Because T can be converted to DHT or estradiol (E2) by 5α-reductase
(5α-R) and aromatase, respectively, these enzymes would determine DHT and E2 concentrations and their ratio. We propose
and apply a low-invasive, sensitive and precise method for the absolute quantification of mRNA levels of aromatase and
5α-R isozymes (type 1, type 2 and type 3) in plucked hair from young women with FPHL. Normoandrogenic women with
FPHL and controls were studied. Plucked hair samples were obtained by trichogram from vertex scalp and mRNA levels
quantified by real-time RT-PCR. We revealed for the first time the presence of 5α-R3 mRNA in human hair. Interestingly,
one, two, or even three 5α-R isozymes were increased in some women with FPHL but not in others, which may explain the
lack of response to 5α-R inhibitors in some FPHL cases. Aromatase mRNA levels were significantly lower in women with
FPHL than in controls. It may therefore produce a reduction in oestrogen levels and an increase in the androgen/oestrogen
ratio in hair. The proposed low-invasive technique offers a molecular aetiologic diagnosis of FPHL for the selection of more
appropriate pharmacological treatments with early predicted effectiveness.
Keywords Female pattern hair loss · Aromatase · 5α-R isozymes · mRNA levels · Trichogram
13
Vol.:(0123456789)
78 Archives of Dermatological Research (2018) 310:77–83
the hair to enter telogen phase and the dermal papilla to progesterone, FSH, LH, prolactin, androstenedione, DHEA-
become senescent [18, 50]. DHT may also interfere with S, and estradiol within the normal range, and none showed
the follicular cycle by interacting with the Wnt signalling signs of clinical virilization or had undergone antiandrogenic
pathway [38]. or contraceptive hormonal therapy in the previous 6 months.
DHT is synthesized from testosterone (T) by the enzyme All individuals signed their written consent before inclusion
steroid 5α-reductase (5α-R). There are two well-character- in the study.
ized 5α-R isozymes in the hair follicle, type 1 (SRD5A1) In all participants, human hair samples were obtained by
and type 2 (SRD5A2) [42]. The importance of 5α-R2 in trichogram from the vertex area of the scalp, plucking 30–40
androgenetic alopecia is demonstrated by the absence of hairs with tightly closing epilation forceps. To avoid differ-
this condition in males with congenital 5α-R2 deficiency ences in tractional (epilation) force between the inner and
[17]; however, little is known about the role of 5α-R1 in hair outer area of the sample, all samples were taken in identical
growth because a human 5α-R1 genetic deficiency syndrome conditions by the same dermatologist following Serrano-Fal-
has not been identified [24]. The more recently discovered con et al. [44]. The epilated hair roots were then placed on
5α-R type 3 isozyme is expressed throughout the dermis a glass slide and covered with coverslip, using light micros-
and epidermis [2], but to our best knowledge studies on this copy to examine hair root types. The anagen/telogen ratio
isozyme in the hair follicle has not been carried out. upon root examination was about 80/20. Only hair roots in
For many decades, androgens have dominated endo- anagen phase were selected for study. Examination of the
crine research into hair growth control. However, the role proximal end of the hair shaft helps to determine whether
of androgens in FPHL remains controversial. Around a hair is in anagen, telogen, or catagen phase. Anagen hair
22–34.8% of women with polycystic ovary syndrome pre- shafts are longer, have a uniform diameter, a rectangular
sent FPHL [33, 37]; however, FPHL can also occur in com- shape, and a slight distal angle. Anagen hair bulbs are seen
plete androgen insensitivity syndrome [6], pointing to other as darkly pigmented triangular or delta-shaped bulbs with
pathogenic factors besides an androgenetic aetiology [26]. an angle to the hair shaft and there is presence of inner root
Oestrogens may also play a role in scalp hair growth, sheath. The sample included hair longer than 2 cm, with
[30]. Thus, aromatase, the key enzyme required to convert some miniaturized hairs but no fuzz. The 10 anagen hairs
androgens to oestrogens, has been localized in scalp hair showing the most intact sheaths were selected for this study.
follicles [41]. Samples were immediately transferred to individual
Pharmacological treatments against FPHL include the Eppendorf tubes prefilled with 0.5 mL RNAlater (Ambion)
administration of 5α-R inhibitors, but their effectiveness to avoid RNA degradation and were stored at room tempera-
remains unclear [13]. Furthermore, the therapeutic response ture for 24 h followed by storage at − 80 °C.
varies among patients, probably due to the different aetiolo-
gies of the disease. Hence, an adequate approach to FPHL RNA isolation
requires an understanding of androgen and oestrogen metab-
olism in the hair follicle and of the involvement of the key Around 10 hair roots in anagen phase per participant were
enzymes 5α-R and aromatase, which may be responsible for homogenized in phenol–guanidine isothiocyanate (Trizol
its androgen/oestrogen ratio. Invitrogen). Total RNA was extracted with chloroform
The objective of this study was to propose and apply a and precipitated with isopropanol by 12,000g centrifuga-
low-invasive, sensitive and precise method for the abso- tion at 4 °C. The RNA pellet was washed twice with 75%
lute quantification of mRNA levels of aromatase and 5α-R ethanol and resuspended in diethylpyrocarbonate-treated
isozymes (type 1, type 2 and type 3) in plucked hair from water. RNA samples were then treated with Turbo DNase
young women with FPHL. (Ambion) to remove contamination with genomic DNA.
RNA yield was determined spectrophotometrically by A260/
A280 ratio using a NanoDrop ND-1000 spectrophotometer
Materials and methods (ThermoFisher). Isolated total RNA integrity was electro-
phoretically verified by ethidium bromide staining.
Subjects and sampling
Reverse transcription and quantitative real‑time
This study included 10 women (age range 20–30 years) PCR
diagnosed with grade II alopecia according to the Ludwig
classification, considered as FPHL, presenting more than First-strand cDNA was synthesized from 1 µg of total RNA
20% of miniaturized hairs, and 7 healthy women of similar following Castro et al. [4]. Absolute quantification of mRNA
age with no hair loss or thinning, who served as controls. of 5α-R1, 5α-R2, 5α-R3 and aromatase in the plucked hairs
All participants had serum testosterone, SHBG, 17-OH was performed by real-time RT-PCR using the Techne
13
Archives of Dermatological Research (2018) 310:77–83 79
Quantica™ Real-time PCR system with SYBR Green PCR Table 1 Primers for PCR amplification of each gene studied and sizes
Master Mix (Promega). The final optimized concentra- of amplified fragments
tion used in the reactions was 0.4 µM for both the forward Gene Sequence (5′–3′) Size frag-
and reverse 5α-R1, 5α-R2, 5α-R3 and aromatase primers. ment (refer-
Twenty-two microliters of the prepared PCR master mix ence)
was added to each well of the M icroAmp® Optical 96-well SRD5A1
reaction plate (Thermo Scientific). Three microliters of each Sense AGCCATTGTGCAGTGTATGC 163
reverse-transcribed RNA sample was added to the appropri- Antisense AGCCTCCCCTTGGTATTTTG
ate well to make a final reaction volume of 25 µl. SRD5A2
Absolute quantification of mRNA is achieved by extrap- Sense TGAATACCCTGATGGGTGG 154
olation of fluorescence signals from test samples against Antisense CAAGCCACCTTGTGGAATC
standard curves representing the initial copy numbers for SRD5A3
a defined fluorescence signal; an external in vitro synthesis Sense TCCTTCTTTGCCCAAACATC 212
cRNA standard was used for their construction. Synthesis Antisense CTGATGCTCTCCCTTTACGC
of cRNAs was performed by in vitro transcription with T7 CYP19A1
RNA polymerase followed by reverse transcription and Sense TATTAGGGCCCTGTGTCTGC 193
real-time PCR, using the method described by Fronhoffs Antisense TGGGTTGGGACTTTTCCTCC
et al. [10]. The standard curve was generated by performing
five independent serial dilutions of the cRNA standard and
assaying each dilution in duplicate. To maximize the accu-
racy, dilutions were made over the range of copy numbers
that included the amount of target mRNA expected in the
experimental RNA samples. Samples and cRNA standards
were reverse-transcribed and amplified in a parallel proce-
dure using identical primer pairs and hybridization probes.
This process allows the parallel amplification of cRNA
standards with a sequence identical to the target gene itself.
In comparison to relative quantification, this method offers
the advantage of giving an absolute copy number for a spe-
cific target. The amount of mRNA was expressed as number
of mRNA copies per micrograms of total RNA.
The PCR profile was as follows: denaturation at 94 °C
for 30 s; annealing at 55 °C for the SRD5A1 gene, 54 °C
for the SRD5A2 gene, 60 °C for the SRD5A3 gene, 60 °C
for the CYP19A1 gene for 30 s; and extension at 72 °C for Fig. 1 mRNA levels of 5α-reductase type 1 (5α-R1) in anagen hairs
30 s. The number of cycles was 40 in all cases. At the end plucked from young women with female pattern hair loss (FPHL)
(n = 10) and controls (n = 7). Data are expressed as means ± SD
of the amplification phase, a melting curve analysis was car-
ried out on the products formed in order to confirm that a
single PCR product was detected by the SYBR Green dye. non-normal by the Kolmogorov–Smirnov test. GraphPad
All reactions were run in triplicate and no cDNA was added Prism 5.0 software (San Diego, CA) was used for the statis-
to the negative reactions. tical analysis. P < 0.05 was considered significant.
Primers for 5α-R1 (SRD5A1 mRNA, Genbank accession
no. NM_001047.3), 5α-R2 (SRD5A2 mRNA, GenBank
accession no. NM_000348.3), 5α-R3 (SRD5A3 mRNA Results
GenBank accession no. NM_024592.4) and aromatase
(CYP19A1 mRNA GenBank accession no. NM_000103.3) In plucked hair samples from the global series (young
were designed using Primer 3 software. The primer women with FPHL and controls), the highest mRNA levels
sequences (5′–3′) are given in Table 1. were for 5α-R1 followed by 5α-R3 and then by 5α-R2.
Mean mRNA levels of 5α-R1 (Fig. 1), 5α-R2 (Fig. 2)
Statistical analysis and 5α-R3 (Fig. 3) isozymes were higher in the women with
FPHL than in the controls, but statistical significance was
The non-parametric Mann–Whitney U test was used for not reached. There was a wide variability of results for these
comparisons because the data distribution was found to be isozymes among the women with FPHL, as can be observed
13
80 Archives of Dermatological Research (2018) 310:77–83
Fig. 2 mRNA levels of 5α-reductase type 2 (5α-R2) in anagen hairs Fig. 4 mRNA levels of aromatase in anagen hairs plucked from
plucked from young women with female pattern hair loss (FPHL) young women with female pattern hair loss (FPHL) (n = 10) and con-
(n = 10) and controls (n = 7). Data are expressed as means ± SD trols (n = 7). Data are expressed as means ± SD. *P < 0.05 vs. Control
13
Archives of Dermatological Research (2018) 310:77–83 81
women with FPHL can be suppressed by a dual 5α-R1 FPHL has been associated with high androgen levels
and 5α-R2 inhibitor but not by a selective 5α-R2 inhibi- [48] and/or elevated androgen receptor sensitivity [40],
tor [11], suggesting a role for 5α-R1 in FPHL. In line with explaining the administration of androgen receptor inhibi-
our results, immunohistochemical studies have described a tors in these cases. However, FPHL has also been observed
greater expression of 5α-R1 than of 5α-R2 in the outer root in women with complete androgen insensitivity syndromes
sheath of hair follicles in both sexes [42]. Furthermore, a [6], suggesting that another mechanism may contribute to
study using relative quantification of mRNA levels in hair its aetiopathogenesis. In this study, aromatase mRNA lev-
from women found high mRNA levels of 5α-R1 but did not els were significantly lower in hair follicles from women
detect 5α-R2 [31]. On the contrary, other authors have found with FPHL than in controls. In line with our results, aro-
higher expression of 5α-R2 and relatively lower of 5α-R1 matase levels in women were higher at the frontal hairline
in DP from AGA [15, 19]. In the present study, although than in other areas of the scalp, in agreement with the
5α-R2 mRNA levels were lower than those for 5α-R1, it relative preservation of the frontal hairline in FPHL [27].
should be borne in mind that the affinity of 5α-R2 for T is The decrease in aromatase may lead to a decrease in oes-
much higher than that of 5α-R1 [36]. 5α-R2 may therefore trogens in situ, unfortunately not measured in this study. The
play a major role in FPHL, as reported in male androgenetic role of oestrogens in FPHL is controversial. Thus, ex vivo
alopecia [17]. In this paper we have quantified mRNA but studies in an animal model reported that oestrogens may
unfortunately, due to the low starting material obtained in inhibit hair growth [30]. However, other studies may point to
plucking hair, enzymatic activity have not been measured, the contrary suggesting that oestrogens may prevent FPHL
this being one of the limitations of the study. by modulating cytokines and growth and transcription fac-
In the present series, mean 5α-R1, 5α-R2 and 5α-R3 tors [30]. Oestradiol increases the secretion of both IGF-1
mRNA levels appeared to be higher in the normoandrogenic and VEGF [21, 35]. IGF1 is a potent stimulator of hair folli-
women with FPHL than in the controls. However, the differ- cle growth and prevents the follicle from entering a catagen-
ences did not reach statistical significance, which is likely like state [35], while VEGF promotes perifollicular vascu-
attributable to the wide variability in levels observed among larization and regulates the hair follicle blood supply [52].
the women with FPHL. We highlight the very high mRNA Conversely, oestrogen inhibits the release of TNF-alpha,
levels of one, two, or even three 5α-R isozymes in some IL-6 and TGF-beta [25, 29], whose overexpression leads to
of the women with FPHL, while none of these levels were hair loss [45, 47]. In this sense, it has been found that preg-
increased in others (Table 2). Probably, FPHL is a complex nant women (period with known elevated estrogens levels)
trait in which sex steroids are important in some individuals present a higher number of hairs in anagen phase [28], while
but not in others. Therefore, the pharmacological approach women with lower oestrogen levels during menopause, post-
to this disorder should take account of this disparity among partum, treatment with aromatase inhibitor or selective oes-
the women, which would explain why treatment with 5α-R trogen receptor modulators are more likely to develop FPHL
inhibitors is not always successful [43]. Furthermore, these [1]. However, in spite of these observations, Kanty et al.
drugs are associated with undesirable adverse effects [12], [23] have reported that the aromatase inhibitor tamoxifen
making their inappropriate administration a highly relevant did not affect hair growth parameters. On the other hand,
issue. many physiological changes during menopause pregnancy
and postpartum should be also kept in mind. Thus, it is very
difficult to strictly dissociate the effects of estrogens in the
hair growth from other factors in these situations.
Table 2 FPHL patients with increased (⇑) 5α-R isozymes or Probably, the key issue for FPHL is the 5α-R/aro-
decreased (⇓) aromatase matase ratio in hair follicle. Thus, elevated 5α-R levels
FPHL 5α-R1 5α-R2 5α-R3 Aromatase and decreased aromatase levels may divert T to DHT
in situ rather oestrogen and vice versa. In this context, it
1 ⇑ ⇑ ⇓
is important to consider the regulation of aromatase and
2 ⇑ ⇑ ⇑ ⇓
5α-R isozymes. Thus, it has been demonstrated that oes-
3 ⇓
trogen induces aromatase activity in intact human anagen
4 ⇑
hair follicles [16], and may also lead to a direct inhibition
5 ⇑ ⇓
of 5α-R [30]. On the other hand, DHT itself may induce
6 ⇑ ⇑ ⇓
the expression of 5α-R isozymes by a feed-forward mecha-
7
nism [46]. The fact that both DHT and E2 may activate the
8 ⇑ ⇑ ⇓
key enzymes implicated in their own biosynthesis compli-
9
cates even further our understanding of the mechanisms
10 ⇑ ⇑ ⇑ ⇓
involved.
13
82 Archives of Dermatological Research (2018) 310:77–83
13
Archives of Dermatological Research (2018) 310:77–83 83
24. Kaufman KD (2002) Androgens and alopecia. Mol Cell Endo- Increased androgen receptor messenger RNA in frontal-parietal
crinol 198:89–95 hair follicles of women with androgenetic alopecia. Genet Mol
25. Kim MS, Chae HJ, Shin TY, Kim HM, Kim HR (2001) Estrogen Res 12:1834–1840
regulates cytokine release in human mast cells. Immunopharmacol 41. Sawaya ME, Penneys NS (1992) Immunohistochemical distribu-
Immunotoxicol 23:495–504 tion of aromatase and 3B-hydroxysteroid dehydrogenase in human
26. Langan EA, Paus R (2010) Female pattern hair loss: beyond an hair follicle and sebaceous gland. J Cutan Pathol 19:309–314
androgenic aetiology? Br J Dermatol 163:1140–1141 42. Sawaya ME, Price VH (1997) Different levels of 5alpha-reductase
27. Levy LL, Emer JJ (2013) Female pattern alopecia: current per- type I and II, aromatase, and androgen receptor in hair follicles
spectives. Int J Womens Health 5:541–556 of women and men with androgenetic alopecia. J Invest Dermatol
28. Lynfield YL (1960) Effect of pregnancy on the human hair cycle. 109:296–300
J Invest Dermatol 35:323–327 43. Seale LR, Eglini AN, McMichael AJ (2016) Side effects related to
29. Matsuda T, Yamamoto T, Muraguchi A, Saatcioglu F (2001) 5 α-reductase inhibitor treatment of hair loss in women: a review.
Cross-talk between transforming growth factor-beta and estrogen J Drugs Dermatol 15:414–419
receptor signaling through Smad3. J Biol Chem 276:42908–42914 44. Serrano-Falcón C, Fernández-Pugnaire MA, Serrano-Ortega S
30. Ohnemus U, Uenalan M, Inzunza J, Gustafsson JA, Paus R (2006) (2013) Hair and scalp evaluation: the trichogram. Actas Dermosi-
The hair follicle as an estrogen target and source. Endocr Rev filiogr 104:867–876
27:677–706 45. Stenn KS, Paus R (2001) Controls of hair follicle cycling. Physiol
31. Oliveira IO, Lhullier C, Brum IS, Spritzer PM (2003) The 5alpha- Rev 81:449–494
reductase type 1, but not type 2, gene is expressed in anagen hairs 46. Torres JM, Ruiz E, Ortega E (2003) Development of a quantitative
plucked from the vertex area of the scalp of hirsute women and RT-PCR method to study 5alpha-reductase mRNA isozymes in
normal individuals. Braz J Med Biol Res 36:1447–1454 rat prostate in different androgen status. Prostate 56:74–79
32. Orme S, Cullen DR, Messenger AG (1999) Diffuse female hair 47. Tsuji Y, Denda S, Soma T, Raftery L, Momoi T, Hibino T (2003)
loss: are androgens necessary? Br J Dermatol 141:521–523 A potential suppressor of TGF-beta delays catagen progression in
33. Ozdemir S, Ozdemir M, Görkemli H, Kiyici A, Bodur S (2010) hair follicles. J Investig Dermatol Symp Proc 8:65–68
Specific dermatologic features of the polycystic ovary syndrome 48 Vexiau P, Chaspoux C, Boudou P, Fiet J, Abramovici Y, Rueda
and its association with biochemical markers of the metabolic MJ, Hardy N, Reygagne P (2000) Role of androgens in female-
syndrome and hyperandrogenism. Acta Obstet Gynecol Scand pattern androgenetic alopecia, either alone or associated with
89:199–204 other symptoms of hyperandrogenism. Arch Dermatol Res
34. Pfaffl MW, Hageleit M (2001) Validities of mRNA quantification 292:598–604
using recombinant RNA and recombinant DNA external calibra- 49 Vujovic A, Del Marmol V (2014) The female pattern hair loss:
tion curves in real-time RT-PCR. Biotech Lett 23:275–282 review of etiopathogenesis and diagnosis. Biomed Res Int
35. Philpott MP, Sanders DA, Kealey T (1994) Effects of insulin and 2014:767628
insulin-like growth factors on cultured human hair follicles: IGF-I 50 Winiarska A, Mandt N, Kamp H, Hossini A, Seltmann H, Zou-
at physiologic concentrations is an important regulator of hair boulis CC, Blume-Peytavi U (2006) Effect of 5alpha-dihydrotes-
follicle growth in vitro. J Invest Dermatol 102:857–861 tosterone and testosterone on apoptosis in human dermal papilla
36. Poletti A, Celotti F, Motta M, Martini L (1996) Characterization cells. Skin Pharmacol Physiol 19:311–321
of rat 5alpha-reductases type 1 and type 2 expressed in Saccharo- 51 Yamana K, Labrie F, Luu-The V (2010) Human type 3
myces cerevisiae. Biochem J 314:1047–1052 5α-reductase is expressed in peripheral tissues at higher levels
37. Quinn M, Shinkai K, Pasch L, Kuzmich L, Cedars M, Huddleston than types 1 and 2 and its activity is potently inhibited by finas-
H (2014) Prevalence of androgenic alopecia in patients with poly- teride and dutasteride. Horm Mol Biol Clin Investig 2:293–299
cystic ovary syndrome and characterization of associated clinical 52 Yano K, Brown LF, Detmar M (2001) Control of hair growth
and biochemical features. Fertil Steril 101:1129–1134 and follicle size by VEGF-mediated angiogenesis. J Clin Invest
38. Ramos PM, Miot HA (2015) Female Pattern Hair Loss: a clinical 107:409–417
and pathophysiological review. An Bras Dermatol 90:529–543 53 Zhou CK, Levine PH, Cleary SD, Hoffman HJ, Graubard BI,
39. Rasmussen R (2001) Quantification on the LightCycler. In: Meuer Cook MB (2016) Male pattern baldness in relation to prostate
S, Wittwer C, Nakagawara K (eds) Rapid cycle real-time PCR, cancer-specific mortality: a prospective analysis in the NHANES
methods and applications. Springer, Heidelberg, pp 21–34 I epidemiologic follow-up study. Am J Epidemiol 183:210–217
40. Richeti F, Kochi C, Rocha MN, Sant’Anna Corrêa C, Lazzarini
R, Guazzelli RM, Mendonça RF, Melo MR, Longui CA (2013)
13