222 JOURNAL OF COSMETIC SCIENCE SYNERGIES IN COMBATING ACNE PRONE SKIN: OLEANOLIC ACID AND NORDIHYDROGUAIARETIC ACID INHIBIT 5-ALPHA REDUCTASE• KERATINOCYTE PROLIFERATION AND INFLAMMATORY RESPONSE: IN VITRO AND IN VIVO STUDIES* Claire Mas-Chamberlin, Philippe Mondon, Ph.D., Francois Lamy, Ph.D. and Karl Lintner, Ph.D. Sederma, Le Perray en Yvelines, France * This paper is being presented by Dr. Rob Comber on behalf of Sederma. Key words: oily sId• P. aches, 5a-reductase. prostaglandin Introduction Mild (i.e. non-pathological) ache and oily skin constitute a multifaceted aesthetic problem with mainly four interrelated causes and symptoms: hyperseborrhea, hyperkeratinisation, cutaneous bacterial proliferation and local inflammation. A successful approach to treat ache prone skin should take into account all four factors in a concerted manner by combining active ingredients that tackle all biochemical and physiological aspects. Hyperseborrhea is most often observed as a consequence of increased concentration of dihydrotestosterone, asteroid hormone generated by the reduction of testosterone by 5-at reductase and responsible for the stimulation of sebum production and secretion. The same hormone testosterone also leads to increased proliferation of keratinocytes (hyperkeratosls) which then causes the skin pores of the pilo-sebacious canal to clog and to the formation of microcomedos. These sites become ideal grounds for bacterial colonization. Chemicals produced and released by the microorganisms finally cause an inflammatory response that is typical of the papules, pustules and macrolesions ofacneic skin. Materials and Methods O!eanolic acid is obtained in pure form by extraction (95%) from the leaves of olive trees. NDGA is obtained either by synthesis or by extraction (95%) from Larrea divaricata. 5-c•-reductase type I assay follows the protocol of Zu-Yue (detection of NADPH Ill) and is confirmed by HPLC quantification of testosterone. Keratinocyte proliferation (neocultured human skin cells) is measured by BrdU incorporation and ELISA assay. Microbial inactivation is visualized on agar gel inoculated with 10 4 or 10 5 cfu/ml and incubated in presence or absence of the products. P. ovale is cultured in Sabouraud gel with olive oil. In vivo tests are carried out on various panels of young volunteers with informed consent. Sebum secretion is assayed with the Sebutape© method. Results and Discussion: Hyperseborrhea: Large-scale screening of plant derived pure molecules has yielded oleanolic acid as a powerful candidate for the inhibition of 5-ct-reductase type I activity. Figure I shows the concentration dependent inhibition of this enzyme by as low as 3 and 9 ppm (I and II) of oleanolic acid The synergy with NDGA (4 ppm and 12 ppm) is also 0 ' '••1 f evident from this graph, as NDGA alone has no effect on 5-a- -,0 reductase. (OA). ' .3o I Hyperkeratosis: -40 I'ø^ +NO0^ I To check the rapid proliferation of keratinocytes, a growth- -so moderating molecule is needed that is, however, devoid of cytotoxicity. NDGA, by acting on the protein processing in the Golgi apparatus, and on the intrace!lular pool of ATP, has a reversin!e (and thus non-toxic) effect on keratinocyte multiplication as is shown in figure 2. Again this effect is powerful and concentration dependent. lOO 80 80 40 20 o , Control 4ppm 12ppm 13relMive growht rMeJ Fig 2: kera•3,tc 8mwth rate mdu•xt ia presence eofNDGA Inflammation: NDGA is well known for a second, intrinsically unrelated biological activity, that is as a non- steroidal cyclooxygenase inhibitor. Effects on prostaglandin synthesis (PGE2) inhibition can be seen at concentrations as low as 3 ppm, once more in concentration dependent manner (fig. 3). This in vitro observed anti-inflammatory activity is of the same order of magnitude as acetyl salicylic acid, the active ingredient of aspirin.

2002 ANNUAL SCIENTIFIC MEETING 223 100 . 80. 40. 20. 0 Control 3ppm •,pprn l•ppm • Fi8.3 :Pro•and• PGE2 synlhes•r'a•in Ine•ne.• of NDGA orA½. Salicylic n•l (3ppm) Bacterial colonization: 5 strains of bacteria commonly found in ncne lesions and ache prone skin (l¾opionefiacterium aches, Coynebacrteriutn tninutissitnutn, $taphylococcus epidermidis, $taphylococcus aureus and Acetinobacter calcoaceticus) and a yeast (Pityrosporum ovale, also known as Malassezia furfur) were investigated. The strains were inoculated in agar gelose in presence or absence of oleanolic acid (15ppm) and NDGA (20ppm) dissolved in butyleneglycol, and a combination of the two at these concentrations levels formulated in a carbomer gel containing 1% of butyleneglycol, glycerin and caprylylglycol. Whereas selective bacterial inhibition by OA or NDGA or the excipient gel could be observed, a total bacterial and yeast growth inhibition was only obtained in the final synergistic mixture. In vivo studies: Clinical studies were undertaken to confirm and substantiate the observations from the in vitro data. At first, the connection between 5-•x-reductase inhibition with sebum production, was established in a clinical test on 9 volunteers with oily skin. The Sebutape© method was used in this double blind vehicle controlled study to measure sebum secretion rates before and after a 6 week treatment with a mixture of Oleanolic acid (9 ppm) and NDGA (12) ppm formulated in a light emulsion. Table I shows the decrease in the number of active sebum glands and the concomitant decrease in quantity of sebum produced (-44%). These values are signifil antly different from the placebo effects. 'Table ! TO T42 Active glands ] Collected Sebum (mm2) Active glands I Collected Sebum (mm2) Mean + s.d. 181+ 23 6.98+1.49 125 + 17 3.+1.25 % variation - -31 -44 The lesional aspects of acne prone skin (comedos, papules, pustules, microcysts, so called retentional and inflammatory lesions) were evaluated in a second clinical study on 20 volunteers of 18-35 years of age, :- .- Table II DO D58 DecreaSe % Microcysts 14.9+2.4 I 0.9+ 1.9 -27% Comedos 7.9+1.8 5.2+1.1 -35% Papules 8: pustules 5.8+1.0 3.9+_0.7 -32% Total lesions 28.6+_3.3 19.9+2.5 -30% Significance P0.05 P0.05 presenting with symptoms of acne type skin. A two month treatment with a Carbomer /butylen glycol/polyol based gel containing 9 ppm of OA and 12 ppm of NDGA gave the following results (dermatological assessment: counting of lesions before and after the trial period (table II). This result compares favorably with retinoic acid treatment [2] the panelists observed the improvement, too, and noted particularly the comfortable skin feel all throughout the day. Conclusion: Four causes and interrelated symptoms of ache prone skin led us to develop a blend of active molecules with which it is possible to act upon all four as we have shown in the specific in vitro and the related in vivo tests. The reduction in sebum production and in cell proliferation, coupled with a light antibacterial effect and anti- inflammatory activity leads to a visible improvement of the skin, without leaving scars or provoking irritation and side effects (fig. 4). References [1] Zu-YUE and Zeg-Hong, Meth. Find Exp. Clin. Pharmacol. 20, 283,1998 [2] Cuncliffe N. et al. Dermatology196, 9-15, 1998