Boldo extract benefits facial skin

“Acne” ‘is a general term which describes all cutaneous conditions characterised by disorders of the pilosebaceous follicles. In anatomical terms, common polymorphous juvenile acne begins with the formation of a comedo, followed or not by an inflammatory phase. As the acne progresses, it is possible to observe four major physiological disorders, namely sebaceous hyper-secretion, an anomaly of the keratinisation of the infundibulo-ostial part of the pilo-sebaceous follicle, the appearance of opened or closed comedones, papules, pustules and a proliferation of bacteria leading to the follicle inflammation.1,2

The bacterial flora in normal skin is constantly colonising the epidermis forming a cutaneous bacterial ecosystem and establishes interactions between the commensal cutaneous flora and the contaminating bacterial pathogenic agents. The number of saprophyte germs on the skin varies from one person to another and from one area of the body to another. The composition of the normal flora is fairly constant. In acneic skin, we know that comedones coming from skin keratinisation troubles can be infected by bacteria. Moreover, bacteria release chemical factors particularly attractive to polynuclear neutrophils (PNs) which will then converge to the site of the comedo.

These polynuclear neutrophils will degranulate in order to enhance the inflammatory phenomenon. The proteolytic enzymes of the PNs are involved at the same level as the bacterial enzymes in the fragmentation of the comedo wall.3 In inflammatory phenomena, the cytokine interleukin-1a play a major role in both the hypercornification and the regulation of immune factors.4 Moreover, Propionibacterium acnes-induced IL-8 secretion plays a key role in the initiation of inflammatory events in acne.5 The chemokine IL-8 has mitogenic activity on keratinocytes and is a potent chemoattractant for neutrophils.

Chemokines are involved in the control of dendritic cell (DC) trafficking which is critical for the immune response. Especially, MIP-3a/CCL20 may have a cell-recruitment role in the skin epidermal layer through the CCR6 receptor expressed on Langerhans cells.6

Furthermore, ß-defensins produced by epidermal keratinocytes on proinflammatory stimulations were capable of attracting immature dendritic cells and memory T cells through CCR6.7 Normal skin controls the microorganism intrusion by production of peptide antibiotics such as defensins.8 A large number of antimicrobial peptides exists with different origins and functions. In general these peptides are small molecules (10-50 amino acids) with an amphipathic structure: separating cationic (positively charged) and hydrophobic regions. These antimicrobial peptides are synthesised in the epithelial tissues of higher organisms. In general, the specificity of antimicrobial peptides lies with the specific characteristics of cell membranes.

Moreover, eukaryotic cell membranes have only neutral phospholipids (phosphatidylcholine and sphingomyelin), whereas bacterial membranes possess negatively charged phospholipids (phosphatidylglycerol and cardiolipin). Antimicrobial peptides will therefore specifically target negatively charged bacterial membranes and hence be selectively toxic against bacteria. Different mechanisms of action have been identified for these peptides, such as Interaction with lipids in bacterial membranes or interaction with LPS (lipopolysaccharides).

In order to induce membrane permeability and cell death, the peptide cationic character allows antimicrobial peptides to interact with negatively charged lipids of bacterial membranes. LPS are major components of the external membranes of Gram- bacteria and are released during lysed pathogenic bacterial infections. Thereby, these endotoxins activate the inflammatory defence reactions. LPS possess bivalent ion binding sites to which cationic peptides have a very strong affinity. When reacting with cationic peptides, LPS release ions which destabilise the membrane structure. This mode of action is being closely studied for medical applications, since it would solve problems of endotoxicity due to uncontrolled release of LPS found with existing antibiotics treatments.

In humans, the role of antimicrobial peptides was first studied in the literature in two major diseases of the skin: atopic dermatitis and psoriasis. The presence of pro-inflammatory factors resulting from the skin’s pathological state triggers the increased expression of beta-defensins.9 Beta-defensins are a particular type of antimicrobial peptides acting in the innate immune defense of epithelial tissues, including skin and mucosa. Up to now, 34 types of human beta-defensins have been registered in databases, 28 of which have been identified by bio-informatics and only four have been found and studied in relation to the skin: human beta-defensins 1, 2, 3 and 4 (hBD1, hBD2, hBD3 and hBD4). While hBD1 is constitutively synthesised in skin epidermis and acts against Gram- bacteria, hBD2 is inducibly synthesised in the skin by the action of pro-inflammatory factors. IL-1a, IL-1ß, TNFa and LPS induce hBD2 synthesis which acts on Gram- bacteria but not on Gram+ bacteria. hBD3 is inducibly synthesised by the action of proinflammatory factors such as TNFa, IFNg and LPS. Because hBD3 is active on Gram- and Gram+ bacteria hBD3 has a wider bactericidal action than hBD2.

Recently, hBD4 has been detected in inflammatory areas of acneic skin where its over-expression was associated with the pro-inflammatory factor IL-8. hBD4 has a bactericidal action on Gram- bacteria and Gram+ bacteria. Because hBD3 has a wide bactericidal action we have examined the ability of Boldo extract to induce human hBD3 mRNA expression. Moreover, we have studied whether the Boldo extract-induced hBD3 is independent or not on proinflammatory cytokines and chemokines production. We have shown that Boldo extract stimulates significantly hBD3 mRNA expression and even reduces cytokine synthesis in keratinocyte monolayers cultures with high concentration of calcium.

We have determined, in vivo, the efficacy of the Boldo extract applied on one side of the face versus a placebo applied on the other side of the face, on mild to moderate acneic and non-acneic subjects. Our study has shown a significant improvement of skin compared to the placebo of various skin parameters such as evenness, smoothness, oiliness/shininess and pore size in both acneic and non-acneic subjects.

Materials and methods

Human keratinocytes isolation and culture
Normal human keratinocytes derived from neonatal foreskins were used in this study. Freshly isolated cells were cultured in serum-free medium K-SFM (Invitrogen, Cergy Pontoise, France) supplemented with 1% penicillin/streptomycin (Invitrogen), epidermal growth factor (EGF) and bovine pituitary extract (BPE) according to the manufacturer’s instructions.

Cells were grown and amplified at 37°C and 5% CO2 in air. In fourth passage, keratinocytes were seeded in 96-well plate in K-SFM at 1.7 ml calcium chloride concentration obtained by addition of a 100 mM solution of calcium chloride (CaCl2; Sigma, L’IsIe d’Abeau, France).

Cell treatment
Human keratinocyte was incubated for 24 hours at 37°C, in the absence or presence of Boldo extract, and tested at 5 concentrations from 0.01 to 5% inclusive.

Cell viability
Viability tests were performed on keratinocytes cultivated at confluence in 96-well plates for 24 h with 1% (v/v) of the tested ingredient. The culture medium was eliminated and replaced by 200 ml of a solution of 5 mM paranitrophenyl phosphate (Sigma) in 0.1 M acetate buffer at pH 5. After 2 h incubation at 37°C, in 5% CO2, the reaction was stopped by adding 20 ml of 1 N NaOH.

Absorption was read at 405 nm with a VICTOR2 (Perkin Elmer, Wellesley, MA, USA) spectrophotometer and reflected cellular viability. We consider that a product was cytotoxic when the cellular viability was lower than 75% of the non-treated control. Sodium dodecyl sulfate (SDS) at 0.05% was used as a positive control to validate the experimentation.

Total RNA extraction
Total keratinocyte RNA was extracted using the “SV 96 total RNA Isolation System” method according to the manufacturer’s protocol (Promega, Charbonnières, France). For each trial, 80 to 100 µl of total RNA was obtained and quantified at 260/280 nm with a spectrophotometer, then frozen at -80°C before use.

Real-time reverse transcriptasepolymerase chain reaction
Dose-dependent effects of Boldo extract on hBD3 mRNA expression were analysed by real-time RTPCR. hBD3 primers were designed as 5’-AGC CTA GCA CCT ATC AGG ATC-3’ (forward) and 5’-CTT CCC CAC CAT TTT CCC CCA-3’ (reverse). Amplification using these primers resulted in a 206-bp fragments. Real-time RT-PCR analyses were performed using the Quantitect TM SYBR green RT-PCR kit (Qiagen, Courtaboeuf, France) in a fluorescence thermocycler (Opticon; MJ Research, Biorad, France) according to manufacturers’ instructions.

Briefly, the reaction mix was composed of SYBR Green 1X buffer, the two specific primers used at 0.5µM, 0.5µl of the enzyme mix, 50ng total RNA, the whole adjusted to 50 µl using RNAse/ DNAse free water. The Q-RT-PCR reaction progresses according to a first reverse transcription stage at 50°C for 30 minutes, followed by thermal denaturation of the reverse transcriptase at 95°C for 15 minutes, before a second PCR amplification stage consisting of up to 50 PCR cycles (95°C for 15s, 60°C for 30s and 72°C for 30s). SYBR green I fluorescence was detected at the end of each cycle to monitor the amount of amplified product formed during that cycle. At the end of each run, melting curve profiles were produced (cooling the sample to 50°C and then heating slowly at 0.1°C s-1 up to 95°C with continuous measurement of fluorescence) to confirm amplification of specific transcripts. Cycleto- cycle fluorescence emission readings were monitored and analysed using Opticon Software (MJ Research, Biorad). It allows the determination of the cycle number at which each sample reached a threshold corresponding to the beginning of the signal. This threshold cycle number correlates inversely to the log of the initial template concentration. Relative hBD3 transcript levels were corrected by normalisation based on the 3-actin transcript levels. 3-Actin primers were designed as 5’-GTG CCC CCC CCC AGG CAC CA-3’ (forward) and 5’- CTC CTT AAT CTC ACG CAC CAT TTC-3’ (reverse) and were predicted to amplify a 540-bp fragment.

MIP-3a/CCL20, IL-8 and IL-1a levels in culture supernatant
The well culture supernatants were tested with ELISA kits to measure levels of MIP-3a, IL-8 and IL-1a. Levels were compared with RNA concentrations measured in each well. This allowed the absence of pro-inflammatory effects to be with Boldo capable of stimulating hBD3 without inducing the concomitant stimulation of pro-inflammatory cytokines. MIP-3a, IL-1a and IL-8 levels were measured by enzyme-linked immunosorbent assay (ELISA; R & D systems, Abingdon, UK) using 96-well microtiter plates in accordance with the manufacturer’s instructions. Supernatants were diluted in order to measure the three cytokines levels in the same supernatant. The levels of MIP-3a, IL-1a and IL-8 were calculated using standard curves obtained with human recombinant (hr) MIP-3a (from 0.47 to 500 pg mL-1), human recombinant (hr) IL-1a (from 1 to 250 pg mL-1) or human recombinant (hr)IL-8 (from 10 to 2000 pg mL-1). The results were expressed as the mean ± SD in pg per RNA.

Statistics

The results are given in the form of averages +/- standard deviation (SD). Statistical significance of the differences observed was evaluated by analysis of variance (ONE WAY ANOVA), followed by the Bonferroni test (* p<0.05; ** p<0.01 and *** p<0.001).

In vivo evaluation of the efficacy of Boldo extract on acneic and non-acneic subjects
The study was performed at the National Hospital of Calcuta in India. Two oil/water formulations have been tested, containing either 4% of Boldo extract or no active compound (placebo), on 50 subjects.

Twenty five subjects clinically diagnosed as suffering from mild to moderate acne of the facial skin were in one group and 25 non-acneic subjects were in the other one. Subjects (either male or female and from 13 to 25 years old) were recruited to participate to this clinical study, following specific inclusion and non-inclusion criteria.

The randomised, double blind comparative against placebo study lasted 10 weeks (2 weeks washout + 8 weeks treatment). The subjects were required to use a commercial face wash during the course of the study. Boldo extract and the placebo were used twice a day, during 8 weeks, in half face, by the subjects from both the acneic group and the non-acneic group, at home, under normal conditions of use, instead of their usual product.

Global acne evaluation
The investigator graded the overall severity of acne on each half face of the acneic subjects by using a scale from 0 (normal, clear skin with no evidence of acne vulgaris) to 4 (inflammatory lesions are more apparent: many comedones and papules/pustules).

Evaluation of overall skin condition
Overall skin clarity, evenness, smoothness, oiliness, pore size and condition/appearance of the skin of all subjects was evaluated using a 10-point (0-9) scale, where 0 indicates perfect condition and 9 indicates severe condition. Overall skin condition/appearance was an overall assessment of the skin, taking into account all the pertinent skin parameters clarity, smoothness and oiliness.

Irritation and skin condition assessment
The investigator assessed safety by grading erythema, oedema, dryness, and scaling as per a scale from 0 (none) to 3 (severe).

The appearance frequency of these parameters was counted in acneic and non-acneic subjects.

Self assessment after 2, 4 and 8 weeks of product application
After 2, 4 and 8 weeks of product application, subjects were asked to fill out a self-assessment questionnaire rating the appearance and feel of each half of their facial skin. The scoring was based on a 1 to 10 scale (1 = very unfavourable, 10 = favourable). Week 8 (final self assessment): subjects also completed a final self-assessment questionnaire for each half of their face. Subjects were asked to compare the appearance and feel of their skin 8 weeks after product application as compared to the time prior to the study start. The scoring was based on a 1 to 5 scale (1 = disagree completely, 3 = neutral, 5 = agree completely).

Statistics
Non-parametric analysis was conducted. The Wilcoxon test was used to determine differences between products as well as difference between timepoints compared to Day 0 (significativity: p<0.05).

Results

Activation of beta-defensin 3 synthesis
To analyse the dose-dependent effect of Boldo extract on hBD3 mRNA levels, Boldo extract was further applied on foreskin keratinocyte in hypercalcic medium during 16 h. Effect on mRNA defensin expression by increasing concentrations of Boldo extract was thereafter analysed by realtime RT-PCR. A first evaluation of Boldo extract cytotoxicity profile (n = 6) allowed us to determine the 2% maximal concentration giving a cellular viability of nearly 75% (Data not shown).

Results shown in Figure 1 confirmed the strong stimulatory dose-effect of Boldo extract on hBD3 mRNA expression. At 1.25% Boldo extract concentration the increase in hBD3 mRNA level was 2-fold higher compared with non-treated control whereas at 2% concentration Boldo extract induced a 10-fold stimulation (p< 0.001).

Inhibition of inflammatory cytokines production
Changes in pro-inflammatory IL-1a, chemokine IL-8 and chemokine MIP-3· proteins were evaluated in cells supernatants by ELISA assay. Levels were compared with RNA concentrations measured in each well at all concentrations. Results showed (Fig. 2) that cytokine IL-1a synthesis and IL-8 chemokine synthesis were respectively decreased by 30 and 40% with 2% Boldo extract treatment. On the other hand, chemokine MIP-3a presented a high inhibition (60%) from 1% Boldo extract treatment (p<0.001) and was inhibited by 80% at 2% Boldo extract treatment (p<0.001).

In vivo evaluation of the efficacy of Boldo extract on acneic and non-acneic subjects

Global acne status assessment
Global acne status assessment was graded by the investigator regarding the overall severity of acne on each half face of the acneic subjects by using a scale from 0 (normal, clear skin with no evidence of acne vulgaris) to 4 (Inflammatory lesions are more apparent: many comedones and papules/pustules). The results have shown that the effect of Boldo extract concerning the global acne evaluation (Fig. 3) was significantly better than the effect of the placebo at the week 4 and week 8 time points. Percentages in graphs give the improvement after week 4 and 8 with regard to the clinical score obtained at day 0. At week 8, the global acne status was evaluated by 20% improvement. Boldo extract also has shown significant effect comparatively to the baseline (day 0) at week 4 and week 8. As far as the placebo is concerned, it has shown significant effect in comparison to baseline only at week 8.

Regarding details of the global assessment, the effect of Boldo extract concerning the papule and pustules count was significantly better than the effect of the placebo at the week 4 and week 8 time points (Fig. 4). Percentages in graphs give the improvement after week 2, week 4 and 8 with regard to the clinical score obtained at day 0. Concerning the decreasing of papule number, Boldo extract at 4% has shown at 4 and 8 weeks a decrease by respectively 33 and 40% of the papule number. In the same way, pustule number has shown very important decrease by 45% at week 4 against 74% at week 8 after Boldo extract treatment. The effect of Boldo extract concerning the closed comedones count was significantly better than the effect of the placebo at the week 4 and week 8 time points. As opened comedones have shown a decrease by 32% and 38% at 4 and 8 weeks of treatment, closed comedones have presented a decrease number by 33 and 43% respectively at week 4 and 8.

Evaluation of overall skin condition