Pii: s0923-1811(01)00129-3

Journal of Dermatological Science 27 (2001) 147 – 151 Potential anti-androgenic activity of roxithromycin in skin Shigeki Inui a,*,1, Takeshi Nakajima a,1, Yoko Fukuzato a, Naohiro Fujimoto b, Chawnshang Chang b, Kunihiko Yoshikawa a, Satoshi Itami a a Department of Dermatology, Course of Molecular Medicine, Graduate School of Medicine, Osaka Uni6ersity, Osaka, 2-2, C5, Yamada-oka, Suita-shi, Osaka 565-0871, Japan b George H. Whipple Laboratory for Cancer Research, Departments of Pathology, Urology, and Radiation Oncology, Uni6ersity of Rochester Medical Center, Rochester, 601 Elmwood A6e., Rochester, NY 14642, USA Received 12 April 2001; received in revised form 23 May 2001; accepted 23 May 2001 Abstract
Since acne formation is a multistep process accelerated by androgens, we examined whether a new anti-acne antibiotic roxithromycin (RXM) may act as anti-androgen using transient transfection assays in human skinfibroblasts. The result showed no significant effect of 0.5, 1 and 5 mg/ml RXM on 10−9 M R1881-induced androgenreceptor (AR) transcriptional activity. While the cotransfection of exogenous ARA55, a novel AR coactivator,increased AR transactivation up to 2.59-fold, this increase was attenuated by 5 mg/ml RXM to 64.7%. Semiquantita-tive RT-PCR results showed that 0.1 mM H O treatment increased ARA55 mRNA expression level, indicating that reactive oxygen species increase the expression of ARA55 in skin. These results suggest that RXM may serve asanti-androgen only in the hypersensitive state to androgen, but not in the physiological state, through modulatingend-organ hypersensitive condition to androgen possibly involving the pathway from reactive oxygen species toARA55. 2001 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Acne; Anti-androgen; ARA55 1. Introduction
been used as anti-acne drugs. These antibioticshave been revealed to have not only anti-bacterial Acne formation is a multistep process, involv- but also anti-inflammatory properties [3]. How- ing increased sebum production by androgens ever, so far it has not been examined whether [1,2], colonization of Propionibacterium acnes, and anti-acne antibiotics may act as anti-androgen.
resultant inflammation. The antibiotics against P.
Recently we have cloned out a novel androgen acnes, such as tetracyclines and macrolides, have receptor (AR) co-activator ARA55 from humanprostate cDNA library using a yeast two hybridassay system [4]. ARA55 enhances AR transacti- * Corresponding author. Tel.: + 81-6-6879-3031; fax: + 81- vation by around five-fold in the prostate cancer cell line, DU145 cells, indicating that ARA55 may E-mail address: inui@derma.med.osaka-u.ac.jp (S. Inui).
1 The first two authors contributed equally to this work.
play important roles in the progression of prostate 0923-1811/01/$ - see front matter 2001 Elsevier Science Ireland Ltd. All rights reserved.
PII: S 0 9 2 3 - 1 8 1 1 ( 0 1 ) 0 0 1 2 9 - 3 S. Inui et al. / Journal of Dermatological Science 27 (2001) 147 – 151 cancer by the modulation of AR activity [4]. We ethanol as a mock in some experiments. At 48 h examine here whether ARA55 can exert its func- after transfection, the cells were harvested for tion in human dermal fibroblasts and then overex- luciferase assays. Luciferase activities were mea- press ARA55 in dermal fibroblasts to recapitulate sured by the luminometer using the Dual-Luci- hypersensitive state to androgens in the skin.
feraseTM reporter assay system (Promega). The Moreover, we evaluate the potential anti-andro- results were summarized from three independent genic activity of roxithromycin (RXM), a new sets of transfection and presented as mean 9SE.
macrolide antibiotic, which have been reported to In the preliminary study, we confirmed that the be effective in the treatment of acne [5].
cotransfection of ARA55 has no significant effecton the basal reporter activity without the ligands.
2. Materials and methods
2.4. Re6erse transcription-polymerase chainreaction (RT-PCR) Cultured human dermal fibroblasts at the sub- Human dermal fibroblasts were isolated from confluency were treated with 0.1 mM H O for 2, skin specimen obtained from plastic surgery oper- 4, 8, 12, and 24 h. Then the cells were harvested ation and then maintained in Dulbecco’s modified and total RNA was isolated by acid – guanidinium Eagle’s medium supplemented with 10% fetal thiocyanate – phenol – chloroform method. The re- covery and purity of the RNA was calculatedfrom the optical densities at 260 and 280 nm. One microgram of total RNA of each sample wasreverse-transcribed using 2.5 mmol/l random hex- 837.06) was supplied by Roussel-Uclaf (Paris, France). Its stock solution was prepared in deoxynucleotide triphosphate (dNTP), 1.0 U/ml Rnase inhibitor (Takara, Tokyo, Japan), and 2.5U/ml M-MLV Reverse Transcriptase (Gibco- 2.3. Transfection and reporter gene assays BRL). The mixture was incubated at 25 °C for 10min, followed by 42 °C for 45 min, 99 °C for 5 Human dermal fibroblasts were grown in Dul- min, then 4 °C for 5 min. The resultant DNA was becco’s modified Eagle’s medium supplemented amplified using a thermal cycler (Astec, Fukuoka, with 10% fetal bovine serum. At 50 – 70% conflu- Japan) in a final volume of 25 ml containing 2 ency in a 24-well plate, the cells were transfected using lipofectamine plusTM (Gibco-BRL) accord- HCl, pH 8.3, 0.025 U/ml recombinant Taq DNA ing to the manufacturer’s instruction. For luci- polymerase (Takara), and 0.5 mmol/l of each ferase assays, we transfected 0.3 mg of the reporter primer. The following sense and antisense primers plasmids, mouse mammary tumor virus luciferase (MMTV-Luc) [6]. In some samples, the expression plasmids for ARA55 and AR, pSG5-ARA55 [4] and pSG5-AR [6], were co-transfected to the cells.
primers amplifies 633 bp of the C-terminal region The pRL-CMV vector, the Renilla luciferase con- of ARA55 [4]. After an initial denaturation at 95 trol reporter vector driven by the CMV immedi- oC for 9 min, 27 cycles of amplification (denatura- ate-early enhancer/promoter, was co-transfected tion at 94 °C for 1 min, annealing at 63 °C for 1 as an internal control. At 24 h after transfection, min, and extension at 74 °C for 1 min) were we put fresh medium with methyltrienolone followed by a terminal extension at 74 °C for 1 (R1881, a synthetic androgen) or ethanol as a min. These ARA55 primers are the same as we mock. We also treated the cells with RXM or have used previously and amplify 633 bp of the S. Inui et al. / Journal of Dermatological Science 27 (2001) 147 – 151 C-terminal region of ARA55 [4]. PCR amplifica- 3.3. RXM effect on the R1881-induced tion was performed for 22 cycles (1 min at 94 °C, transcriptional acti6ity enhanced by ARA55 in 45 s at 60 °C, and 45 s at 72 °C) for glyceralde- hyde-3-phosphate dehydrogenase (G3PDH) as aninternal control to demonstrate comparable RNA amounts and quality among samples. The prod- anti-androgenic function in dermal fibroblasts without exogenous ARA55 transfection (Fig. 1), RXM may modulate the AR transactivation en-hanced by ARA55 overexpression. Thus, we 3. Results
transfected 0.3 mg of pSG5-AR and 0.6 mg ofpSG5-ARA55 and examined the effect of 1 and 5 3.1. Effect of RXM on the R1881-induced mg/ml RXM on R1881-induced transactivation in transcriptional acti6ity in dermal fibroblasts dermal fibroblasts. The co-transfection of ARA55increased MMTV-luciferase activity by around To examine whether RXM can suppress andro- 2.5-fold (Fig. 3, lane 3). Although 1 mg/ml RXM gen activity in terms of transcriptional regulation, did not show any significant effect on this en- we performed luciferase assays using dermal hanced induction (Fig. 3, lane 4), RXM at the fibroblasts transiently transfected with 0.3 mg of concentration of 5 mg/ml suppressed the AR pSG5-AR. We tested the concentrations from 0.5 transactivation to 64.7% induction (Fig. 3, lane to 5.0 mg/ml, because 5.0 mg/ml of RXM is consis- 5). These results suggest that RXM has an anti- tent with the maximal concentration in the serum androgenic activity under the overexpression of after taking 150 mg of RXM per os [7]. The result showed no significant effect of 0.5, 1 and 5 mg/mlRXM on 10−9 M R1881-induced AR transcrip-tional activity (Fig. 1). From this result, RXMdoes not seem to have anti-androgenic effect inskin dermal fibroblast.
3.2. Effect of co-transfection of ARA55 with ARon R1881-induced transacti6ation in dermalfibroblasts It has been demonstrated that ARA55 potenti- ates AR transcriptional activity by around five-fold in DU145 cells [4]. To ascertain whetherARA55 can enhance AR transcriptional activityin dermal fibroblast, we examined the effect ofexogenous ARA55 overexpression on R1881-in-duced AR transcriptional activity. pSG5-AR (0.3mg) was transiently transfected into all samples.
Fig. 1. RXM effect on the R1881-induced transcriptional The co-transfection of 0.3 mg or 0.6 mg pSG5- activity in dermal fibroblasts without transfection of ARA55.
ARA55 showed 2.35- or 2.59-fold increase in We transfected 0.3 mg of pSG5-AR and 0.3 mg of MMTV-Luc R1881-induced AR transcriptional activity (Fig.
into the cell cultured at 50 – 70% confluency in a 24-well plate 2). Although, compared with the ARA55 effect in using lipofectamine plusTM (Gibco-BRL) and treated with 0.5 prostate cancer cells [4], its effect in dermal (lane 3), 1 (lane 4) or 5 mg/ml (lane 5) of RXM and 10−9 MR1881 (lanes 2-5) or ethanol as a mock (lane 1). The relative fibroblasts is relatively weak, it was confirmed reporter gene activities were compared with the luciferase that ARA55 can potentiate AR transcriptional activity in the presence of R1881 and the absence of RXM S. Inui et al. / Journal of Dermatological Science 27 (2001) 147 – 151 Fig. 2. Exogenous ARA55 effect on the R1881-induced tran- Fig. 3. RXM effect on the R1881-induced transcriptional scriptional activity in dermal fibroblasts. We transfected 0.3 mg activity in dermal fibroblasts with transfection of ARA55. We of pSG5-AR, 0.3 mg of MMTV-Luc (lanes 1–4), and 0.3 (lane transfected 0.3 mg of pSG5-AR, 0.3 mg of MMTV-Luc (lanes 3) or 0.6 mg (lane 4) of pSG5-ARA55 into the cells cultured at 1 – 5), or 0.6 mg of pSG5-ARA55 (lanes 3–5) into the cells 50 – 70% confluency in a 24-well plate using lipofectamine cultured at 50 – 70% confluency in a 24-well plate using lipofec- plusTM (Gibco-BRL), and treated with 10−9 M R1881 (lanes tamine plusTM (Gibco-BRL), and treated with 10−9 M 2 – 4) or ethanol as a mock (lane 1). The relative reporter gene R1881 (lanes 2 – 5) or ethanol as a mock (lane 1). The relative activities were compared with the luciferase activity in the reporter gene activities were compared with the luciferase presence of R1881 and the absence of ARA55 (lane 2).
activity in the presence of R1881 and ARA55 and the absenceof RXM (lane 3).
3.4. H O increases ARA55 mRNA expression in testosterone and acne severity in women and pro- posed a model in which variation in skin sensitiv- Reactive oxygen species (ROS), generated from ity to androgen and the level of androgen equally infiltrating neutrophils around the lesions of acne, contribute to the pathogenesis of acne [9]. Very has been known to play important roles in the recently it has been reported no positive correla- acne formation mainly by progressing inflamma-tion [3,8]. We assumed that ROS may modulateandrogen sensitivity in the skin, possibly throughupregulating the expression of ARA55. To testthis possibility, we treated cultured dermal fibrob-lasts with 0.1 mM H O and examined the change of ARA55 mRNA expression level by semiquanti-tative RT-PCR. As shown in Fig. 4, ARA55mRNA was gradually increased by the H O treatment, suggesting that ROS increase the ex-pression of ARA55 and consequently modulateandrogen sensitivity in the skin.
Fig. 4. H O effect on ARA55 mRNA expression in dermal fibroblasts. Semi-quantitative RT-PCR was performed using 4. Discussion
the same amount of cDNA reverse-transcribed from totalRNA isolated from cultured human dermal fibroblasts treatedwith 0.1 mM H O for 0, 2, 4, 8, 12, and 24 h. The upper The previous study showed the striking vari- panel shows the 633 bp products of ARA55 mRNA, and the ability in the relationship between plasma free lower shows G3PDH mRNA as the internal controls.
S. Inui et al. / Journal of Dermatological Science 27 (2001) 147 – 151 tion between the grade of acne severity and the this manuscript and Dr Motoko Takahashi at clinical or laboratory markers of androgenicity Department of Biochemistry, Graduate School of tested [10]. From these observations, the end-or- Medicine, Osaka University, for her useful discus- gan sensitivity may be a very important factor in the pathogenesis of acne. In the present study, werecapitulated the hypersensitive state to androgenby overexpression of exogenous ARA55 in vitro.
References
Since RXM showed anti-androgenic activity onlyunder overexpression of ARA55, we suggest that [1] Henze C, Hinney B, Wuttke W. Incidence of increased RXM serves as anti-androgen only in the state of androgen levels in patients suffering from acne. Derma-tology 1998;196:53 – 4.
androgen hypersensitivity, but not in the physio- [2] Akamatsu H, Zouboulis CC, Orfanos CE. Control of logical state. We showed here that H O treat- human sebocyte proliferation in vitro by testosterone and 5-alpha-dihydrotestosterone is dependent on the localiza- suggesting that ROS may modulate androgen sen- tion of the sebaceous glands. J Invest Dermatol sitivity through ARA55 in the skin. Therefore, [3] Miyachi Y, Yoshioka A, Imamura S, Niwa Y. Effect of RXM may exert the potential anti-androgenic antibiotics on the generation of reactive oxygen species. J activity in the hypersensitive state to androgen such as acne, which has been reported to involve [4] Fujimoto N, Yeh S, Kang HY, Inui S, Chang HC, ROS production by neutrophils in the inflamma- Mizokami A, et al. Cloning and characterization of an- drogen receptor coactivator, ARA55, in human prostate.
J Biol Chem 1999;274:8316 – 21.
So far, the concentration of H O in the tissue [5] Akamatsu H, Nishijima S, Akamatsu M, Kurokawa I, of acne in vivo is unknown. Previously it was Asada Y. Clinical evaluation of roxithromycin in patients reported that H O at higher concentrations than with acne. J Int Med Res 1996;24:109 – 14.
0.2 mM triggered apoptosis of dermal fibroblasts, [6] Hsiao PW, Chang C. Isolation and characterization of but not at 0.05 and 0.1 mM [11]. Therefore, we ARA160 as the first androgen receptor N-terminal-associ-ated coactivator in human prostate cells. J Biol Chem determined for the biological relevance that the concentration of H O should be 0.1 mM in this [7] Shiba K, Saito A, Shimada J, Kaji M, Okuda S, Hori S, study. Since we used skin dermal fibroblasts from Miyahara T, Matsumoto F, Sakurai I, Hojo T, Ueda Y.
normal human, but not sebocytes from acne pa- Basic and clinical studies on RU28965. Chemotherapy tients, it is questionable whether our experimental [8] Akamatsu H, Horio T. The possible role of reactive system can reflect the pathological condition of oxygen species generated by neutrophils in mediating acne acne inflammation. Further studies using cultured inflammation. Dermatology 1998;196:82 – 5.
human sebocytes are necessary to conclude our [9] Reingold SB, Rosenfield RL. The relationship of mild hypothesis. However, we report here for the first hirsutism or acne in women to androgens. Arch Dermatol time that anti-acne antibiotics may have anti-an- [10] Cibula D, Hill M, Vohradnikova O, Kuzel D, Fanta M, Zivny J. The role of androgens in determining acneseverity in adult women. Br J Dermatol 2000;143:399 –404.
[11] Bladier C, Wolvetang EJ, Hutchinson P, de Haan JB, Acknowledgements
Kola I. Response of a primary human fibroblast cell lineto H O : senescence-like growth arrest or apoptosis? Cell We thank Mrs. Karen Wolf for her helping for

Source: http://dc150.4shared.com/download/N6VbRpp6/CV160-JDS-01-Inui.pdf