2.2.1 Modification of the WAW Formula

The original WAW formula consists of C Herba, D. polystachya, S. chinensis, E. ulmoides, A. bidentata, C. chinensis, P. cocos, A. orientalis, H Rubrum, R. glutinosa, C. officinalis, and M. officinalis. This prescription was tailored to meet the needs of external use, such as color and texture considerations. The names of the medicinal substances included in the formula were standardized according to the 2015 Pharmacopeia of the People's Republic of China and the nomenclature conventions for modern Chinese medicine studies. Any drug components not included in the pharmacopeia could not be entered and systematically analyzed and thus were excluded. The formulas were constructed according to the interactions between medicinal substances and the laws of compounding, as well as factors such as the nature and attributes of the medicinal substances and the requirements of modern skincare products.

2.2.2 In Vitro Experiments to Explore the Effects of the Tailored WAW Formula on Skin

Each medicinal substance was extracted according to the design of the experimental programme, changing as necessary the factors and parameters of the extraction process conditions and using ultrasound-assisted extraction and then reduced-pressure extraction and filtration to generate 10 g of each substance. The eight medicinal substances were evaluated for collagen content, NO content, tyrosinase activity, and melanin content to determine whether these substances had any effect on the skin and to perform a preliminary exploration to a provide research direction for subsequent experiments.

2.2.3 Network Pharmacological Analysis

2.2.4 Optimisation of the Extracted Individual Medicinal Substances in the Tailored WAW Formula

2.2.5 Determination of Optimal Formulas and Characterization of Components by HPLC

Based on the principles of traditional Chinese medicine prescriptions, the results of in vitro efficacy experiments of single herbs, and the network pharmacology results, the combinations and ratios of the single herbs in the WAW formula were adjusted, resulting in three different formulation ratios: M1, M2, and M3 (Table 8). The three formulas were subsequently subjected to anti-inflammatory evaluation (NO inhibition assay and whitening evaluation), tyrosinase activity measurement, and moisturizing evaluation (filaggrin, AQP3, and HAS2 gene expression level measurements as described in section 2.2.6.4) to identify the most effective formulas to undergo subsequent identification of the components by HPLC and cellular assay validation.

After the compounding ratio was determined, the quality markers of the formulas were studied using HPLC. The characteristic components of the single herbs were selected to optimize the chromatographic conditions. The qualitative analyses were carried out by retention time and ultraviolet spectrograms under suitable chromatographic conditions.

High-performance liquid chromatography (HPLC) parameters: The chromatographic column used was Poroshell HPH-C18 (3 × 100 mm, 2.7 μm); the mobile phase was water:acetonitrile (99:1); the flow rate was 0.8 mL/min; the detection wavelengths for morroniside, loganin, schisandrin A, pinoresinol diglucoside, and β-ecdysterone were 240 nm, 238 nm, 215 nm, 200 nm, and 248 nm, respectively; the column temperature was 30°C; the detector was a diode array detector with an injection volume of 20 μL; and the gradient elution was 0~25 min, 95%–0% A, 5%–100% B; 25~30 min, 100% B; 30~35 min, 0%–95% A, 100%–5% B; and 35~37 min, 95% A, 5% B (A: acetonitrile, B: water).

2.2.6 Validation of the Efficacy of MGO-Induced Cellular Senescence Modeling

3.1.1 Effects of the WAW Formula After Addition or Subtraction of Substances

Experts in traditional Chinese medicine pharmacology, theory, and modern research, as well as in the development of botanical ingredients for cosmetics, have analyzed and explored the components of the classic herbal formula “Wan An Wan” (WAW). In accordance with the principles of herb interaction, compatibility, flavor, and meridian tropism and considering the requirements of modern skincare products, the tailored WAW formula has been reformulated to include D. polystachya, C. chinensis, C. officinalis, S. chinensis, E. ulmoides, A. bidentata, A. orientalis, and P. Cocos.

Among these, D. polystachya and C. chinensis are the sovereign medicinal agents. D. polystachya, with its sweet flavor and neutral nature, moisturizes and nourishes the lungs and tonifies the kidneys to secure essence. C. chinensis, known since “Shennong's Classic of Materia Medica,” has a spicy-sweet flavor and neutral nature and is moisturizing and nourishing. From the perspective of traditional Chinese medicine, the nourishment and circulation of blood and essential maintenance are closely related to skin health and the promotion of natural, radiant skin.

The minister medicines among the components of WAW are C. officinalis and S. chinensis. C. officinalis tastes sour and slightly warm, with astringent properties to consolidate and stop leakage. S. chinensis tastes sour–sweet and is warm, with astringent properties to contract and solidify. Both are sour medicines with astringent and solidifying properties, which are beneficial for dry and flaking skin and maintaining facial stability. With abundant essence, qi, blood, and body fluids promoting flourishing and facial radiance, C. officinalis and S. chinensis can consolidate essence, thereby nourishing the skin.

The assistant medicines in WAW are E. ulmoides and A. bidentata. E. ulmoides tastes sweet and warm, tonifies yang, and nourishes essence. A. bidentata tastes bitter-sweet and is neutral, promoting blood circulation and alleviating stasis.

The courier medicines are P. cocos and A. orientalis, which promote fluid production and clear turbidity. P. cocos tastes sweet and bland, with a neutral nature, promoting diuresis and dampness dispersion, whereas A. orientalis tastes sweet and bland, with a cold nature, promoting diuresis, dampness dispersion, heat clearing, and turbidity transformation. Both P. cocos and A. orientalis belong to the category of sweet and bland medicines.

3.1.2 Results of In Vitro Efficacy Tests

The three different groupings of compound ratios were nontoxic to cells at concentrations of 0.1%–2% (Figure 1A), and the eight individual herbs were nontoxic to cells at concentrations of 0.5%–2% (Figure 1B–E); therefore, overall concentrations of 2%, 1%, and 0.5% were selected for the subsequent experiments.

To assess the potential skin health benefits of the eight herbs in the tailored WAW formula and to understand their effects on skin improvement from a modern perspective, we evaluated the eight herbs of WAW in terms of their ability to increase collagen synthesis in the skin, which in turn improves skin firmness and reduces fine lines; to effectively modulate tyrosinase activity to reduce hyperpigmentation or blotchiness; and to regulate nitric oxide levels to improve skin and reduce inflammation. The experimental results (Figure 1F–I) revealed that the samples, when applied at 2% concentrations, ranked in terms of collagen content as follows: D. polystachya>P. cocos>C. officinalis>A. bidentata; the other herbs did not promote collagen. They ranked in terms of inhibition of tyrosinase activity as follows: S. chinensis>A. bidentata>D. polystachya>C. officinalis>P. cocos>C. chinensis>E. ulmoides>A. orientalis. They ranked in terms of inhibition of NO production as follows: E. ulmoides>C. officinalis>P. cocos>D. polystachya>A. orientalis>C. officinalis>S. chinensis>C. chinensis. They ranked in terms of melanin inhibition as follows: S. chinensis>C. officinalis>E. ulmoides; other herbs did not have a melanin-inhibiting effect. When applied at a 1% concentration, C. officinalis and A. bidentata produced some collagen synthesis-promoting effects, whereas the other herbs did not significantly promote collagen synthesis. S. chinensis, A. orientalis, A. bidentata, and P. cocos had certain effects on inhibitory tyrosinase activity. D. polystachya and C. officinalis better inhibited NO production. Only C. officinalis and S. chinensis had some inhibitory effects on melanin synthesis. When applied at a concentration of 0.5%, the eight herbs had no obvious effect on promoting collagen synthesis or inhibiting melanin; only C. chinensis inhibited tyrosinase activity; D. polystachya and C. officinalis inhibited the production of NO better; and only C. officinalis had some inhibitory effects on melanin synthesis.

3.2.1 Drug-Active Ingredient-Target Network

In the search box of the TCMSP homepage, search for the eight medicinal herbs identified [7], as shown in Table 3. The number of compounds for each herb was as follows: D. polystachya (71 compounds), C. chinensis (29 compounds), C. officinalis (226 compounds), S. chinensis (130 compounds), E. ulmoides (147 compounds), A. bidentata (176 compounds), P. cocos (34 compounds), and A. orientalis (46 compounds).

Using Cytoscape 3.9.0 software, the drug-active ingredient-target network of the tailored WAW formula was constructed (Figure 2A), with blue rectangles representing the 412 action targets, hexagons representing the active ingredients of the 8 single medicinal substances, pale yellow hexagons representing the common active ingredients of the herbs, and each edge representing the interaction between the drug, the active ingredient, and the target. This regulatory network had a total of 825 nodes, 9805 edges, and an average degree value of 15.084. Core node screening was performed based on the network topological features such as the node degree value, and the information table of the 10 active ingredients with the highest degree values in the specific drug-active ingredient-target-of-action network is shown in Table 4.

The main active ingredients of the tailored WAW formula, as determined by network pharmacological research methods, are cyclic enol ether terpenoids, alkaloids, lignans, polysaccharides, flavonoids, etc. These findings provide suggestions and references for further research.

3.2.2 KEGG Pathway and GO Enrichment Analysis

We obtained 23 747 and 15 094 potential target genes from the GeneCards and NCBI Gene databases, respectively, with the search terms “Skin/Skin disease/dermatology” and “Homo sapiens”, respectively, and 27 387 skin-related target genes were obtained after duplicates. By intersecting the active ingredient regulatory targets of the tailored WAW formula with skin-related target genes, 412 skin-related active ingredient targets of the tailored WAW formula were obtained (Figure 2B).

The 412 skin-related targets of the tailored WAW formula active ingredient action targets were subjected to PPI analysis via the STRING platform, and the results were imported into Cytoscape 3.9.0 for network topology analysis. In this analysis, nodes denote the target genes, degree values denote the number of lines connected to the nodes to assess the importance of each node in the network, and each edge denotes the interaction between the target relationships. The top 20 targets are shown in Figure 2C.

To gain insight into the functions of these protein targets, we performed functional enrichment analysis using the GO and KEGG databases. After the targets of all eight drugs were combined, 429 genes were subjected to KEGG and GO signaling pathway enrichment. A total of 8 KEGG pathways (p < 0.05) and 13 GO entries (p < 0.05) associated with skin function were screened, and the main biological processes associated with skin aging are shown in Tables 5 and 6. The results revealed that the MAPK signaling pathway, the PI3K–Akt signaling pathway, and the collagen catabolic process were related to skin aging, with the related genes being TGFB, MMPs, MAPKs, etc.

In addition, we analyzed the biological process (BP), molecular function (MF), and cellular component (CC) terms related to the enriched GO pathways. Among the BP, MF, and CC categories, the top 20 terms with significantly enriched rankings are displayed visually (Figure 2D–F). The main BP terms were positive regulation of response to external stimulus, response to extracellular stimulus, protein heterodimerisation activity, and positive regulation of cell adhesion; the main MF terms were protein kinase activity, protein tyrosine kinase activity, and protein serine/threonine/tyrosine kinase activity; and the main CC terms were focal adhesion, cell–cell junction, perinuclear region of cytoplasm, and extracellular matrix.

3.2.3 Orthogonal Experimental Design and Results

As described in section 2.2.4, based on the results of the above single-factor experiments, an orthogonal 2-factor 2-level experiment, namely, the liquid-ingredient ratio (A) and ultrasonication time (B), was designed, and a total of 4 test points were analyzed; the results are shown in Table 7. Based on the orthogonal experiments, the optimal extraction process of the eight medicinal substances from the tailored WAW formula was obtained, as shown in the following table:

3.2.4 Adjustment of the Compounding Ratio of the Tailored WAW Formula

The results of network pharmacology revealed that among the top 10 compounds ranked according to degree, the single herb hyssop had many effective compounds, so the proportion of hyssop was greater than that of the original formula.

The results of in vitro efficacy experiments of single herbs revealed that at the 2% treatment concentration, compared with the control group, D. polystachya, C. officinalis, P. Cocos, and A. bidentata had greater effects on promoting collagen synthesis, of which D. polystachya had the greatest effect (158.28%), and all of them inhibited tyrosinase activity, NO production, and melanin production. Among them, S. chinensis had the greatest ability to inhibit tyrosinase activity and melanogenesis (47.01% and 75.84%, respectively). C. officinalis and A. bidentata still promoted collagen at the 1% treatment concentration. Therefore, the proportion of herbs was greater than that of the original formula. E. ulmoides and A. orientalis did not clearly promote collagen production or inhibit tyrosinase activity, and as adjuvants and enablers, the samples were also darker in color, and the dosages were poorer than those of the other herbs.

Thus, by synthesizing the principles of TCM formulation with the results of network pharmacology and in vitro efficacy experiments, the combinations and compounding ratios of the single herbs of the tailored WAW formula were rearranged, as shown in Table 8.

Screening of the three ratios and identification of the chemical compositionIn summary, compounding should be carried out according to the above three different ratios, and parameters such as tyrosinase activity, NO production, and moisturizing genes should be measured afterwards. The results of the experiments (Figure 3A–C) revealed that M1, M2, and M3 inhibited tyrosinase activity and NO production at 2%, 1%, and 0.1% treatment concentrations, respectively. The ability of M3 at a 1% concentration to increase filaggrin mRNA expression was 1.823 times greater than that of the control group, and HAS2 mRNA expression was 1.871 times greater than that of the control group. Therefore, M3, which has the best overall effect on skin improvement, was chosen as the formula for the subsequent verification of the cellular-level aging effect.

After the extraction process of the formula was optimized with reference to the results of network pharmacology and the optimization of the extraction process of the single herbs, the formula quality was investigated using HPLC to evaluate the characteristic components of the single herbs by retention time and ultraviolet spectrograms. The key evaluated components included morroniside, loganin, schisandrin A, pinoresinol diglucoside, and β-ecdysterone (Figure 4).

3.3.1 Modeling Results

In accordance with previous laboratory experience, a cell aging model stimulated with MGO and high AGE expression was established in human skin fibroblasts. The results revealed that 4 mM MGO stimulated fibroblasts for 48 h, and the cell survival rate was 79.25% (Figure 5A). The highest expression level of CML was 53.04 pg/μg, which significantly differed from that of the normal group (18.10 pg/μg) (Figure 5B). Therefore, 4 mM MGO stimulation of fibroblasts for 48 h was selected as the model condition.

3.3.2 Results of Detecting Cellular Dysfunction Behavior

The toxicity of M3 to fibroblasts at each concentration was examined using the CCK-8 method, and the results are shown in Figure 1A. M3 was nontoxic to cell concentrations ranging from 2% to 0.5%. Therefore, 2%, 1%, and 0.5% were selected for subsequent experiments.

The cell proliferation ability associated with damage to cellular aging was assessed using the CCK8 assay. As shown in Figure 5C, M3 effectively restored the MGO-induced proliferation of damaged cells at concentrations of 2%, 1%, and 0.5%.

A fluorescence assay revealed that the cell adhesion capacity was associated with cellular senescence injury and that M3 ameliorated MGO-induced senescence injury-related cell adhesion capacity (Figure 5D,F). Compared with 2% and 1% M3, 0.5% M3 had a weaker ability to restore the adhesion of MGO-injured fibroblasts.

A scratch assay was used to investigate the ability of M3 to improve the migration ability of MGO-induced senescent cells at 24 h and 48 h, and the effect became more significant with increasing time (Figure 5E,G). At the 24 h modeling time range, 2% and 1% concentrations of M3 restored the migration ability of MGO-damaged cells, and both achieved more than 50% recovery, which significantly differed from that of the MGO model group (3.51%). At 48 h, the recovery effect of 2% and 1% M3 reached over 95%, whereas the effect of 0.5% M3 on restoring the migration ability of MGO-damaged cells reached 80% compared with that of the MGO model group (4.86%).

3.3.3 Results of the ECM Protein Component Content Assay

The ECM is a complex noncellular network mainly composed of fibroblasts, proteoglycans, hyaluronic acid, adhesion glycoproteins (fibronectin and laminin) and fibronectin (collagen), as well as growth factors and cytokines [8]. Among these proteins, laminin regulates cell adhesion, growth, and differentiation [9], and laminin-5 supports cell adhesion and migration more efficiently than other laminin proteins do, and increased expression of laminin-5 rejuvenates aging skin [10]. The multimodular ECM protein tenascin C (TNC) is not an obligatory structural component of the ECM but binds to structural proteins in the ECM and to cell surface receptors [11, 12]. The binding of TNC to these receptors activates downstream pathways and affects functions such as cell proliferation, adhesion, and migration [13, 14]. ECM glycation is manifested by increased skin stiffness and decreased elasticity, as well as the induction of fibroblast senescence and apoptosis [15]. AGEs (CML) regulate the expression of ECM proteins, altering the expression of enzymes responsible for their degradation and inhibiting pathway-synthesized proteins [16]. M3 at 2%, 1%, and 0.5% significantly restored the ability of MGO-injured fibroblasts to synthesize COL1, FN1, TNC, and LM5, all of which were significantly different from those of the MGO model group (Figure 6A–D). The ability of 0.5% M3 to restore the ability of MGO-injured fibroblasts to synthesize TNC and LM5 was weaker than that of 2% and 1% M3.

3.3.4 Results of the ECM Protein Component Production and Degradation Pathway Assays

The accumulation of functional abnormalities in aging dermal fibroblasts leads to increased degradation and decreased synthesis of the ECM [17, 18]. A decrease in the number of fibroblasts increases alterations and the degradation of the ECM, which manifests as progressed dermal thinning, increased wrinkling, and a loss of elasticity [19]. This degradation process is mediated mainly by the matrix metalloproteinase family (metalloproteinases, MMPs). ECM synthesis and degradation in dermal fibroblasts are regulated mainly by the TGF-β signaling pathway, which is able to (i) promote the expression of ECM genes and (ii) downregulate MMPs to promote ECM protein synthesis [20, 21]. The results revealed that 2% M3 significantly increased the expression of the TGF-β1 mRNA in MGO-damaged fibroblasts, whereas the other concentrations did not restore TGF-β1 mRNA expression. M3 at a concentration of 2% inhibited MMP-2 mRNA expression in MGO-damaged fibroblasts, whereas 1% M3 minimally inhibited MMP-2 and MMP-9 mRNA expression, and 0.5% M3 did not inhibit the ability of MGO-damaged fibroblasts (Figure 6E–G).

In summary, M3 was better at restoring the proliferation ability and migration ability of damaged cells and restoring the adhesion ability of damaged cells. Moreover, its ability to restore the expression of the ECM protein components type I collagen/COL1 and fibronectin/FN1 was greater, and its ability to restore the expression of the tendon proteins C/TNC and laminin 5/LM 5 was weaker. M3 promoted ECM protein production mainly by upregulating TGF-β. M3 also inhibited the degradation of ECM protein components by downregulating MMP-2 and MMP-9, thereby restoring the functions of migration, adhesion, and proliferation of damaged cells; slowing cellular aging; and exerting firming and anti-wrinkle effects on the skin (Figure 7).