Identification of a new plant extract for androgenic alopecia treatment using a non-radioactive human hair dermal papilla cell-based assay

The list of the Thai plants and plant parts used in this study is shown in Table 1. All plant materials were purchased from of Chao Krom Poe Dispensary in Bangkok, Thailand. Voucher specimens of the plants were deposited in the herbarium collection of Thailand’s Forestry Department with the following specimen numbers (SNs): Avicennia marina, SN145917; Micromelum minutum, SN055802; Diospyros mollis, SN142793; Alternanthera sessilis, SN097887; Salacia verrucosa, SN099212; Crotalaria retusa, SN043656; Senna timoriensis, SN094415; Dalbergia parviflora, SN200944; Afgekia sericea, SN124647; Bacopa monnieri, SN176524; Tarenna hoaensis, SN176524; Pterygota alata, SN059871; Scoparia dulcis, SN163093; and Senna garrettiana. SN008620. For preparation of crude extracts, the plant part used for each species was ground into powder and extracted through maceration using 100 % methanol at room temperature for two days. The methanolic extracts were then evaporated to dryness at 45 °C using a rotary evaporator (Buchi, Switzerland) and kept at -20 °C until used.

All of the organic solvents used were analytical grade and purchased from RCI Labscan (Bangkok, Thailand). Ultrapure grade dimethyl sulfoxide (DMSO) was purchased from Ameresco® (Framingham, USA). Testosterone (T) and 5α-dihydrotestosterone (5α-DHT) were purchased from Sigma-Aldrich (St. Louis, USA). Dutasteride was purchased from BDG Synthesis (Wellington, New Zealand). Agarose-LE was purchased from Affymetrix (Santa Clara, USA). Mesenchymal stem cell medium and its supplements were purchased from Sciencell Research Laboratories (Carlsbad, USA). Foetal bovine serum, 100X antibiotic-antimycotic solution, 10X PrestoBlue®, RPMI medium, 50X Tris-acetate-EDTA (TAE) buffer, 0.25 % trypsin-EDTA and Platinum® Taq polymerase were purchased from Invitrogen (Grand Island, USA). A GeneRuler 1-kb DNA ladder was purchased from Thermo Fisher Scientific (Pittsburgh, USA). RNeasy® mini kits were purchased from Qiagen (Valencia, USA). DNase I enzyme, first-strand cDNA synthesis kit, dATP, dTTP, dCTP and dGTP were purchased from Fermentas (Walthan, USA).

HHDPCs, obtained from Sciencell Research Laboratories (Carlsbad, USA), were grown in mesenchymal stem cell medium containing 5 % foetal bovine serum (FBS), mesenchymal stem cell medium supplement and 1X antibiotic-antimycotic solution at 37 °C in 5 % CO₂. The cells between passages 2 to 6 were used in this study.

Reverse-transcriptase polymerase chain reaction (RT-PCR) was used to identify the type of 5α-reductase enzymes (i.e., 5α-R1 and/or 5α-R2) and AR, expressed in passages 2, 4, 5 and 6 of HHDPCs. Total RNA was extracted from HHDPCs according to the manufacturer’s instructions for the RNeasy® mini kit. The RNA was subsequently treated with DNase I to remove any genomic DNA contamination. Complementary DNA (cDNA) was synthesised from the treated RNA using the first-strand cDNA synthesis kit. The cDNA obtained from this reaction was used as a DNA template for PCR reactions. The PCR reaction comprised of 1X PCR buffer, 2 mM MgCl₂, 0.4 mM dNTP mix, 0.4 μM each of forward and reverse primer and 2.5 units of Platinum® Taq polymerase. The forward and reverse primers for the two isoforms of 5α-R (5α-R1 and 5α-R2) and AR, shown in Table 2, were designed based on the protein region of the full-length sequence obtained from the NCBI GenBank using Clone manager (Scientific & Educational Software, USA) and made to order at 1^st Base Laboratories (Selangor, Malaysia). The PCR-thermal profile started with an initial denaturation at 94 °C for 3 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing for 30 s at 52 °C, and extension at 72 °C for 2 min, followed by a final extension at 72 °C for 10 min. The PCR products were analysed using 1 % agarose gel electrophoresis.

For the cytotoxicity test, HHDPCs were seeded at a cell density of 1x10⁵ cells/ml onto 96-well plates (100 μl of 10,000 cells/well). After 24 h, the cells were separately treated with 100 μl of 5, 10, 20, and 40 μg/ml of each extract to obtain the final concentrations of 2.5, 5, 10 and 20 μg/ml, respectively, and 1 % DMSO (control). Cell viability was measured at 24 h after treatment using 1X PrestoBlue® reagent in RPMI medium. In the presence of viable cells, PrestoBlue® changes from a non-fluorescent blue colour to a fluorescent purple-pink colour, detected using Multimode Detector DTX 880 (Beckman Coulter®, Indianapolis, USA), a bottom-read fluorospectrophotometer with excitation/emission wavelengths of 535 and 615 nm, respectively. The highest final non-toxic concentration of each extract was then used for further studies.

HHDPCs were seeded at a cell density of 1x10⁵ cells/ml onto 96-well plates (100 μl of 10,000 cells/well). After 24 h, the cells were separately treated (in a total volume of 200 μl) with the final concentrations of 0.1 mM T and 0.5 % DMSO (internal control), 0.1 mM T and 0.25-20 μg/ml (non-toxic concentrations of the methanolic plant extracts), and of 0.5 % DMSO as negative control. After 48 h, the culture medium of each treatment was collected in Eppendorf tubes, and the attached treated cells were tested for cell viability using the 1X Prestoblue® reagent in RPMI medium. The remained substrate T and the product 5α-DHT formed from the activity of cellular 5α-R were extracted from the culture medium using an equal volume of ethyl acetate. The ethyl acetate layer was dried and reconstituted using 20 μl of methanol and spotted onto a TLC Silica gel 60 F₂₅₄ aluminium plate (Merck, Darmstadt, Germany). The TLC plate was developed using toluene:acetone (8:2) as the mobile phase [9]. The developed TLC plate was dipped briefly in a solution of 42.5 % phosphoric acid and heated at 120 °C for 20 min for the visual detection of 5α-DHT at 366 nm using a TLC Reprostar Imager (Camag, Switzerland), and the amount 5α-DHT formed was quantitated by scanning the product band of 5α-DHT at 366 nm using TLC Densitometer Scanner 3 (Camag, Switzerland) to obtain its value of peak area which was then converted to the amount of 5α-DHT using a calibration curve. The calibration curve of 5α-DHT was constructed by plotting peak areas and various amounts of 5α-DHT which showed linearity from 50 to 500 ng with the coefficient of determination (r²) value of 0.9874. The extract of AM displaying the highest potential in inhibiting the 5α-R enzyme was studied further to determine its effects on androgen-sensitive genes.

The effect of androgens on hair growth is exerted through the interactions with 5α-R and AR in HHDPCs. Therefore, the presence of both the enzyme and the receptor is important for this study, and expression in HHDPCs was evaluated. The RT-PCR analysis revealed that the genes of the Type 1 enzyme, 5α-R1, and AR were both expressed in passages 2, 4, 5 and 6 of HHDPCs, whereas the Type 2 gene, 5α-R2, was not expressed in any of the passages (Fig. 1). β-actin, used as an internal control, was constantly expressed in all passages.

To obtain a suitable starting concentration for screening methanolic plant extracts for 5α-R inhibitory activity, the cytotoxicity of each plant extract on HHDPCs at various concentrations was subsequently tested using Prestoblue® cell viability reagent. The results revealed that the extracts, at their final concentrations, exhibited different toxic levels on the cells (Fig. 2). Above the cell viability of 85 %, the extracts were considered to be non-toxic. The plant extracts of DM and SG were the most toxic, exhibiting toxicity above 2.5 μg/ml, followed by the ST, ASC and DP extracts, which displayed toxicity above 5 μg/ml (Fig. 2a). The other extracts with moderate toxicity above 10 μg/ml were SD, OM, AM and BC (Fig. 2b), whereas the remaining extracts, including TF, CA, AS, LH, MF and BM, displayed no toxicity even at 20 μg/ml (Fig. 2c). Based on these results, the highest final non-toxic concentration of each extract was used for the subsequent inhibitory activity analyses.

As previously mentioned, AGA is induced through the over-production of 5α-DHT, and the enzyme responsible for the conversion of T to 5α-DHT within HHDPCs is 5α-R. Thus, one potential method for reducing the effect of the androgens on hair growth is to inhibit this enzymatic reaction. HHDPCs expressed only 5α-R1, and therefore, the assay system was tested using a well-known 5α-R1 inhibitor, dutasteride, as a positive control. To obtain the starting concentration for testing the inhibitory activity, first, the cytotoxicity of dutasteride for HHDPCs was determined. It was observed that dutasteride exhibited toxicity (i.e., cell viability less than 85 %) above the final concentration of 10 μg/ml (Fig. 3a). At this concentration, the cell-based assay displayed complete 5α-R1 inhibitory activity (Fig. 3b). Lowering the concentration of dutasteride to 0.01 and 0.001 μg/ml, the assay displayed the 5α-R1 inhibition of 90.5 and 16.5 %, respectively, with an IC₅₀ value of 0.005 μg/ml or 9.7 nM (Fig. 3b). The assay system was then used to screen the methanolic plant extracts for 5α-R1 inhibitory activity. Using the highest final non-toxic concentration (Table 3) in the incubation mixture, each individual extract was evaluated for this activity. As shown in Figs. 4a-4c, each lane represents the inhibitory potential of each extract at its highest final non-toxic concentration. The lane “Cell + T” was the internal control for the conversion of T to 5α-DHT, and the lane “Cell-T” was the negative control. Among all extracts screened, the plant extracts of BM, DM, OM, MC, BC and BA displayed only 10 % inhibition, whereas the extracts of KG and MM displayed approximately 20 % inhibition. The highest inhibitory activity was observed from the crude extract of AM at a final concentration of 10 μg/ml through the reduction in 5α-DHT formation by more than 50 % (Table 3). The IC₅₀ value of AM was determined to be 9.21 ± 0.38 μg/ml.