Reduction of wrinkles: From a computational hypothesis to a clinical, instrumental, and biological proof

Abstract

Background

Facial wrinkles are clear markers of the aging process, being chronological, photo‐induced, or reflecting repetitive facial expressions. The aim of this study is to provide new insights into the biophysical and biological mechanisms involved in the formation, prevention, or elimination of the expression wrinkles.

Materials and methods

We use a computational model to get a better understanding of the wrinkle mechanical behavior and evolution after skin softening and suggesting a possible antiaging mechanism. Then, we provide a clinical demonstration of the anti‐wrinkle effect of a long‐term application of a 20% glycerol in a moisturizer formula (GBM) versus its vehicle on crow's feet. Skin hydration, elasticity, and wrinkles visibility were evaluated by a combination of clinical and instrumental in vivo data, inverse finite element analysis, and proteomic data.

Results

The computational model shows a predominantly compressive stress beneath the wrinkle and its significant decrease by the softening of stratum corneum. The associated clinical study confirmed a significant increase of skin hydration and elasticity as well as a decrease of wrinkle visibility after 2 and 4 months as application for both formulas; this effect being stronger for GBM. A softening effect on stratum corneum and dermis was also observed for the GBM. Furthermore, proteomic data revealed an effect of upregulation of four proteins associated with desquamation, cell‐glycan extracellular interactions, and protein glycation/oxidation, functions related to the tissue mechanics and adhesion.

Conclusions

We provide an in vivo demonstration of the anti‐ageing benefit of glycerol at high dose (20%) reflected by a cumulative skin surface softening effect. The use of high moisturizing potent formulations should bring additional performance to other conventional moisturizing formulations.

1. INTRODUCTION

Wrinkles emerge and become more pronounced with age as a result of a very complex process affecting the nature of the different skin layers. Expression wrinkles in particular appear in locations that undergo cumulative mechanical stresses induced into the skin by facial expressions over the years. Emphasized by intrinsic and external aggressions of the stratum corneum, all these factors are supposed to induce the biomechanical and structural changes in the deep skin layers responsible for the wrinkle's formation in the middle age.

In order to provide a physical interpretation of the expression wrinkles formation, several theoretical or computational models have been proposed in the literature. ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ They suggest that the expression wrinkles can be interpreted as skin folds induced by the deeper facial tissues’ contraction, namely, muscle contraction for dynamic temporary wrinkles (during facial gestures) or deep tissue volume loss for permanent wrinkles. ⁹ This is based on a well‐known mechanical instability phenomenon called buckling that, on multilayered systems under lateral compression, can be triggered when the strain exceeds a given value that depends on the layers’ properties. Following this hypothesis, the wrinkle dimensions depend upon the relative skin layers’ morphology and stiffness, and their occurrence and change with aging depend on the evolution of such properties.

Skin aging has been mostly addressed in relation to deeper skin layers, namely, papillary and reticular dermis, whereas much less attention has been devoted to most superficial layers, such as the stratum corneum. However, despite its fineness (about 10 μm on the face, i.e., 1% of the dermis thickness), it plays an important role on the mechanical properties of the skin, due to its much higher stiffness (ranging from MPa to few hundreds of MPa according to its water content ¹⁰ , ¹¹ ) compared to the living epidermis (tenths of kPa ¹² ) and dermis (tenths of kPa ¹³ to MPa ¹⁴ ) ones. Although it is very challenging to evaluate in vivo its mechanical properties, Hara et al. ¹⁵ employed a computational–experimental inverse method approach to demonstrate a stiffening of this layer with aging. As for its contribution to expression wrinkles, Potter et al. ² employed a computational approach to suggest that the depth of the dynamic temporary wrinkles could be decreased by decreasing the stratum corneum stiffness, which can be accomplished by hydrating the skin. This was experimentally confirmed by Hara et al., ¹⁶ who demonstrated that the dynamic temporary wrinkles induced by the face expressions were reduced after skin hydration and hypothesized that moisturizers could be an effective way to prevent the formation of permanent wrinkles.

In this work, we start from a computational model of a permanent wrinkle ¹⁷ to suggest a mechanism of action of a long‐lasting softening of the stratum corneum as a potential anti‐wrinkle treatment. Then, we demonstrate its clinical efficacy by evaluating the evolution of the wrinkle visibility and other skin biophysical and biological markers due to a long‐term application of a strong moisturizer. Therefore, our integrated approach allows us to gain a deeper biophysical and biological comprehension of the mechanisms involved in the permanent wrinkle evolution.

2. MATERIALS AND METHODS

2.1. Computational wrinkle model

As illustrated in Figure 1, we developed a 3D computational mechanical model of one crow's feet wrinkle. The skin was described by a 5‐layer model, including stratum corneum, living epidermis, papillary dermis, reticular dermis, and hypodermis. Each skin layer is mechanically modeled as a nearly incompressible hyperelastic neo‐Hookean material, whose thickness and mechanical properties listed in Table 1 are obtained either by the literature or by internal measurements. The wrinkle morphology was extracted by profilometry data. The skin deformation due to a facial expression is modeled as a uniform mechanical compression by 15% (according to in vivo data ¹⁶ ) along a direction normal to the wrinkle fold.

FIGURE 1.

Computational model of a crow's feet wrinkle. On the left, skin surface topography whose data were obtained by a profilometer. On the right, 3D view of the 5‐layer wrinkle model, where the skin deformation due to a facial expression is modeled as a uniform mechanical compression by 15% along a direction normal to the wrinkle fold

TABLE 1.

Computational model parameters employed for each skin layer

Layer	Parameter	Value
Stratum corneum	Thickness	10 μm
	Young's modulus	5 MPa
Living epidermis	Thickness	70 μm
	Young's modulus	5 kPa
Papillary dermis	Thickness	200 μm
	Young's modulus	30 kPa
Reticular dermis	Thickness	1.2 mm
	Young's modulus	100 kPa
Hypodermis	Thickness	1.5 mm
	Young's modulus	20 kPa
All	Poisson ratio	0.49

Note: To better compare with the literature, we expressed the 𝐶₁₀ neo‐Hookean parameter in terms of the corresponding (secant) Young's modulus in the small deformation limit 𝐸 ≈ 6𝐶₁₀.

2.2. Clinical study

To assess the long‐term efficacy of a high moisturizing formula on the skin properties and the wrinkles visibility, we conducted a monocentric double‐blinded randomized study. A total of 31 women aged from 55 to 75 were selected and included from the general population with light to moderate crow's feet wrinkles (score from 2 to 3.6 according to the aging atlas ¹⁸ ) and prone to dry skin (hydration values from 37 to 55 arbitrary units (a.u.) measured on cheek with Corneometer CM825 Courage & Khazaka). Previous studies demonstrated that formulas containing high concentration of glycerol allowed a strong ¹⁹ and long‐lasting ²⁰ moisturizing effect. Therefore, a 20% glycerol in oil/water (O/W) emulsion was applied twice daily on split face versus an O/W emulsion placebo for 4 months. Both formulas were applied by the 31 women on randomized sides of her face, and each woman was her own control. Clinical assessments by a trained dermatologist and instrumental measurements were done at baseline before any products application (D0), 1 week after products application (D7), after 2 months (D56) and 4 months (D112), with a 12 h washout before each evaluation. The clinical assessment was performed on crow's feet wrinkles with an Evalux table to ensure standardized conditions of lightening.

Moreover, the skin hydration was evaluated on crow's feet using the Corneometer on subjects on lying position.

Finally, the skin mechanical properties on crow's feet were assessed by a Dermal Torque Meter (DTM, model DTM310 Courage & Khazaka—Dia‐Stron) with SOND01 and TORQ01, and a Cutometer (Courage + Khazaka electronic GmbH, Cologne, Germany). For the DTM, which is rather sensitive to the skin superficial layers, the torque applied to the skin was 100 cN cm. For the Cutometer, rather sensitive to the skin deeper layers, the suction probe with a hole of 2 mm was chosen, a Mode 1 depression of 450 mbar is applied during 2 s, followed by a relaxation period of 2 s, this cycle being restarted five times.

For this study, no REB approval was necessary as an ethical committee submission was not required for this noninvasive proof‐of‐concept study.

2.3. Inverse finite element analysis

To quantify and localize the mechanical effect of the long‐term treatment within the skin, we performed an inverse finite element analysis. ¹⁵ , ²¹ It combines DTM and the Cutometer results for the Ue parameter with a multilayered finite element computational model of the skin. This model assumes planar skin surface and interfaces, employs the same layers’ thicknesses than the computational wrinkle model (whose values are reported on Table 1) without distinguishing between papillary and reticular dermis for an easier retrieve of the unknown mechanical properties, and considers the skin layers as homogeneous isotropic nearly incompressible neo‐Hookean materials. Specifically, a cost function was built by adding the squares of the differences between the experimental points (at each pressure for the Cutometer and couple for the DTM) and the corresponding simulated values. This function describes the overall discrepancy between numerical simulations and experimental measurements. Then the four Young moduli of the finite element model (strictly speaking, the C ₁₀ parameter) were allowed to vary and, for each new set of these parameters, a new value of the cost function was obtained. This procedure was automatized by Matlab (MathWorks) by using a GlobalSearch routine, which allowed us to calculate the four Young moduli that minimize the cost function for each volunteer, treatment, and time point of the clinical study. The method is less accurate in retrieving the epidermis and hypodermis stiffnesses, because the experimental methods considered here are less sensitive to this layers. In fact, concerning the living epidermis, its mechanical contribution is masked by those of the stiffer stratum corneum and dermis, while concerning the hypodermis, its contribution is masked by the stiffer dermis, as long as the Cutometer 2 mm probe is used. ²²

2.4. Multiple reaction monitoring proteomics

Multiple reaction monitoring (MRM) is a targeted proteomic method that was used to determine the relative quantification of 54 abundant proteins of the stratum corneum sampled via D‐squames. MRM is increasingly employed for relative quantification of biomarkers in samples scarce in protein content. The proteins selected are involved in nine important biological functions of the skin which are 1‐defense against microorganisms, 2‐inflammation, 3‐desquamation, 4‐hydration, 5‐barrier function, 6‐ceramide production, 7‐differentiation, 8‐proliferation, and 9‐glycolysis, which enable us to detect some function impacted by the treatment.

The MRM method used in this paper is probably the most reliable and reproducible proteomic quantification method as a synthetic standard peptide is added for each of the targeted peptides, and the signal is normalized to this synthetic peptide. In our studies, we always add the same synthetic peptides so it increases the intra and inter assay reproducibility. ²³ , ²⁴

The analysis consists in four steps: (a) extraction/solubilization of proteins from stratum corneum, (b) digestion of proteins with trypsin, (c) MRM data acquisition, and (d) statistical analysis.

2.4.1. Protein extraction from D‐squames samples

D‐squames were placed in 2 ml Eppendorf tubes with 1 inox bead and 500 μl of DOC buffer (0.5% sodium deoxycholate, 50 mM ammonium bicarbonate, 50 mM DTT, 1 μM pepstatin, and a cocktail of protease inhibitor (Roche EDTA free cocktail)). Extraction was performed on a Retsch 400 Mixer‐Mill: 2 cycles of 2 min at 30 Hz. The supernatant was filtered on 0.45 μm nylon membranes by centrifugation to remove particulates (Millipore). Proteins were precipitated using five volumes of ice cold acetone during one night at −20°C. Proteins pellets were recovered after centrifugation at 16 000 g for 20 min at 4°C and resuspended in 50 μl of 50 mM ammonium bicarbonate/1% sodium deoxycholate. ²⁵

2.4.2. Digestion of proteins with trypsin

The proteins were then denatured at 95°C for 5 min, reduced using 0.2 mM DTT, and alkylated with 0.8 mM iodoacetamide. Finally, 0.5 μg of trypsin was added, and the samples were incubated at 37°C for 16 h for complete trypsin digestion of proteins. Digestion was stopped by an acidification of the buffer to pH2. Peptides were lyophilized and then suspended in 0.1% formic acid prior to peptide dosage on NanoDrop 2000 at 205 nm (Thermo Fisher Scientific). Samples were then dried and suspended in 0.1% formic acid at a concentration of 0.5 μg/μl.

2.4.3. Liquid chromatography‐tandem mass spectrometry analysis

The MRM analysis was performed on 56 proteins for which specific peptides were selected based on their ionization properties, sequence uniqueness, and retention time.

For each selected peptide to be analyzed by MRM, an equivalent peptide was synthetized using heavy isotopes containing either a heavy lysine or heavy arginine at the N‐Ter of the peptide. These standard peptides were prepared in 0.1% acid formic (concentration 0.2–1.8 μg/μl) and added to 4 μl of sample prior to injection in the mass spectrometer. ²⁶

2.4.4. Statistical analysis

To analyze each parameter individually with the group variable (vehicle vs. 20% glycerol), a student test (t test) for paired data was performed. A significant difference between the two groups was concluded when the p‐value (risk of wrongly concluding that there is a difference) was less than 5%.

A univariate analysis was carried out, where the effect of the 20% glycerol formula was evaluated after 56 days of application and compared to the effect of the vehicle through hemi face protocol. As a consequence, the analysis was performed using paired statistical t test and calculating the corresponding p value.

Then, multivariate analysis was performed to increase our understanding of the mechanisms underlying the increase of the skin hydration by the glycerol formula. Such analysis employed partial least squares‐discriminant analysis (PLS‐DA) and was undertaken to study protein and clinical/instrumental measurements together. Partial least squares regression (PLS regression) is a statistical method that constructs a linear regression model by projecting the predicted variables and the observable variables to a new space. The new space is also called the space of the latent variables, that is, why PLS is also denominated as projection to latent structure. The latent classes will be constructed by considering the correlation between the variables. PLS‐DA is usually applied to explain the difference between groups of people and was here employed to explain the differences between the vehicle group and the 20% glycerol group. The most common visualization of the results is the biplot.

Here, the predicted variable in the PLS‐DA corresponds to the treatment variable Y (20% glycerol group or vehicle group). In the model, we have two groups of explicative variables: variables that represent the proteins with the prefix P_ and those measured by an instrument (Cutometer, DTM, or Corneometer) with the prefix I_. The complete list of variables is given in Table 2. In our example of PLS‐DA, we highlight the most important variables, among proteins variables and instrument variables described precisely above, which discriminate the best the treatment group 20% glycerol treatment and vehicle treatment. The parameters that best define the 20% glycerol treatment are as follows: Corneo cheeks, Corneo crow's feet, DTM_UR, galectin 7, and kallikrein 5. It means that subjects with 20% glycerol treatment are higher values for these parameters than subjects with vehicle treatment.

TABLE 2.

List of explicative variables used in the partial least squares‐discriminant analysis (PLS‐DA)

Variable name	Variable description
P_CASP14	Caspase 14 protein
P_AZGP1	Zinc‐alpha‐2‐glycoprotein
P_SOD1	Superoxide dismutase protein
P_PLACK	Junction plakoglobin protein
P_PRDX2	Peroxirédoxine‐2 protein
P_ENO1	Enolase 1 protein
P_TGM3	Protein‐glutamine gamma‐glutamyltransferase E
P_PARK7	PARK7 protein
P_KLK7	Kallikrein 7 protein
P_KLK5	Kallikrein 5 protein
P_LGALS7	Galectin 7 protein
I_CUTO_R0	Cutometer parameter R0
I_CUTO_R3	Cutometer parameter R3
I_CUTO_R4	Cutometer parameter R4
I_CUTO_R8	Cutometer parameter R8
I_DTM_UR	DTM parameter UR
I_CORNEO_CROWSFEET	Corneometer measurement on crow's feet
I_CORNEO_CHEEK	Corneometer measurement on the cheek

3. RESULTS

3.1. Simulation of the impact of the moisturizer on the wrinkle mechanical behavior

In Figure 2, we simulated the impact of the stratum corneum mechanical softening on the skin mechanical stress (third principal stress) during the wrinkle compression induced by the facial expression. The stratum corneum softening was modeled by decreasing its Young modulus (i.e., its C ₁₀ parameter) according to the value obtained in vivo by inverse analysis. The results demonstrate that a softer stratum corneum exerts a lower stress on the deeper layers close to the wrinkle bottom when it bends during the facial expression. This in turn translates into a lower compressive (i.e., negative) mechanical stress into the deeper layers, especially into the epidermis and the dermis. Our results, obtained for a crow's feet wrinkle, are in‐line with the simulated decrease of the stress field depth under a skin microrelief due to the stratum corneum softening obtained by Leyva‐Mendivil et al. ⁷

FIGURE 2.

Cross‐sectional view of the simulated impact of the mechanical softening of the stratum corneum on the skin mechanical stress (third principal stress) during the wrinkle compression induced by the facial expression. On the left, stress field before the stratum corneum mechanical softening, on the right stress field after softening. Although the computer model contains all skin layers, here only the stratum corneum (SC), the living epidermis (Epi), and the papillary dermis (PD) are shown. The negative values indicate that the stress is compressive.

We hypothesize that it is specifically the high repetitive compressive stress developed into the skin beneath and around the wrinkle during the face expressions that triggers the mechanobiological events responsible for the tissue remodeling and the dermis degradation resulting in the wrinkle formation and deepening. Therefore, our results provide a plausible anti‐wrinkle mechanisms induced by a long‐lasting hydration: by softening the stratum corneum, a long‐lasting moisturizer, applied on the skin surface, partially relieves the compressive stress generated during the facial expressions into the deep layers, thereby providing a protective effect on the epidermis and the dermis and a potential mechanism to prevent the wrinkle formation or to decrease their visibility.

3.2. Clinical study

The results of the clinical study reinforced such hypothesis.

3.2.1. Hydration

The results illustrated on Figure 3 show that the skin hydration measured on crow's feet increased with 20% glycerol emulsion, more than with placebo emulsion. This difference is very significant (p value <0.0001 with moderate effect size) from D7: +12.9 a.u. on crow's feet with 20% glycerol and no increase with placebo. This difference in skin hydration increase remains high and significant (p value <0.0001 with moderate effect size) after 2 and 4 months (about two times more important with 20% glycerol emulsion: +17.6 vs. +8.3 a.u. at M2, +17 vs. 6.7 a.u. at M4).

FIGURE 3.

Skin hydration (arbitrary units) measured by Corneometer on crow's feet

3.2.2. Crow's feet wrinkles visibility

Figure 4 demonstrates that both formulas significantly decreased the visibility of crow's feet wrinkles after 2 months of bi‐daily applications: −0.59 grades with 20% glycerol emulsion (p value <0.0001 with moderate effect size) and −0.53 with placebo emulsion (p value = 0.001 with moderate effect size); and after 4 months: −0.67 grades with 20% glycerol (p value <0.0001 with moderate effect size) and −0.52 with placebo (p value = 0.002 with moderate effect size). Nevertheless, there is no significant difference between both formulas in this wrinkles visibility reduction, which could suggest a nonlinear relation between moisturizing potent and reduction in wrinkles visibility.

FIGURE 4.

Crow's feet wrinkles score (scale from 0 to 6) assessed by a dermatologist

Figure 5 provides an example of the effect obtained after 4 months of bi‐daily application of 20% glycerol moisturizing emulsion.

FIGURE 5.

Crow's feet wrinkles improvement after 4 months of bi‐daily application of 20% glycerol moisturizing emulsion

3.2.3. Skin mechanical properties

Figure 6 shows that there is a significant increase of skin elasticity (Ue) measured by DTM on crow's feet with 20% glycerol emulsion from 2 months of bi‐daily applications (+0.47 degrees, p value = 0.029 with weak effect size) to 4 months (+1.00 degree, p value <0.0001 with moderate effect size). There is also a significant increase of Ue with placebo emulsion at D112 (+0.53 degrees, p value = 0.006 with weak effect size) but twice lower than the glycerol one. These data suggest a mechanical softening of the skin superficial layers, to which the DTM is mostly sensitive.

FIGURE 6.

Skin elasticity (Ue in degrees) measured by a Dermal Torque Meter

The inverse final element analysis allowed us to calculate the Young modulus of the stratum corneum and the dermis, whose results are reported on Figures 7 and 8, respectively. In these figures, only the values at 0 and 2 months were reported because the Cutometer experimental data at 4 months were acquired with the wrong pressure (lower than the 450 mbar preconized for the inverse analysis routine). For the stratum corneum, we observed a time effect with the 20% glycerol emulsion which induces a mechanical softening at 2 months, in‐line with the DTM data in Figure 6. From this figure, we argue that the stratum corneum softening at 4 months is even stronger that at 2 months. For the dermis, we also observed a time effect with the 20% glycerol emulsion which significantly softened the stiffness of the dermis after 2 months of bi‐daily applications (−0.03 MPa, p value = 0.004 with weak effect size), whereas placebo had no effect.

FIGURE 7.

Young modulus of the stratum corneum (MPa) calculated by inverse finite element analysis

FIGURE 8.

Young modulus of the dermis (MPa) calculated by inverse finite element analysis

3.3. Proteomics

3.3.1. Univariate analysis

As reported in Table 3, the univariate analysis reveals two proteins significantly modulated by the treatment with 20% glycerol versus the vehicle: kallikrein 5 (×6.0) and galectin 7 (×2.7). Both proteins are implicated in differentiation mechanism, with kallikrein 5 being an important enzyme implicated in the desquamation process of the skin.

TABLE 3.

List of the proteins significantly impacted by the treatment with 20% glycerol, with their gene name, protein name, main biological function, ratio, CV (%), and p value

Gene name	Protein name	Main biological function	Mean ratio	CV (%)	p Value
KLK5	Kallikrein 5	Differentiation/desquamation	6.009141671	2.20003706	0.01883383
LGALS7	Galectin 7	Terminal differentiation	2.728148518	1.45929699	0.03699697
KLK7	Kallikrein 7	Differentiation/desquamation	5.75153250441152	3.04791591806175	0.19493416074303
PARK7	Protein/nucleic acid deglycase DJ 1	Antioxidant	1.99163275826689	1.61151647897117	0.0826937615924165

Note: In gray: p value <0.05.

None of the other proteins were significantly modulated, suggesting that the application of glycerol did not impact the following biological functions: defense against microorganisms, inflammation, hydration, barrier function, ceramide production, proliferation, and glycolysis.

3.3.2. Multivariate analysis

The multivariate analysis illustrated in Figure 9 reveals a cluster of clinical signs and protein biomarkers around the 20% glycerol group and summarizes 43% (first and second dimension) of the information given by the parameters (protein and clinical/instrumental measurements). No clinical signs and proteins were clustered with the vehicle.

FIGURE 9.

PLS‐DA (partial least squares‐discriminant analysis) combining the clinical results and protein modulation. Black triangles: clinical/instrumental measurements; green circles: proteins; gray circles: active 20% glycerol and vehicle

The parameters that best define the 20% glycerol treatment can be classified into two groups based on their distance from the 20% glycerol group:

1. Group of parameters with higher correlation to the 20% glycerol group: Corneo cheeks, corneo crow's feet, DTM_UR with galectin 7, and KLK5,

2. Group of parameter with a high correlation with the 20% glycerol group: addition of I_CUTO_R0, I_CUTO_R8, I_CUTO_R4, KLK7, and PARK7.

Protein biomarkers clustering with the active group (upper right part of Figure 9) are galectin 7, kallikrein 5, kallikrein 7, and PARK7, which were all found increased in the treatment with 20% glycerol compared to the vehicle, as demonstrated in Figure 10. Kallikrein 5 and 7 are the most abundant kallikrein found in stratum corneum and are implicated in the desquamation process.

FIGURE 10.

Box plot representation of the protein modulation with the vehicle and the treatment with glycerol 20% for the four proteins associated with the glycerol treatment based on partial least squares‐discriminant analysis (PLS‐DA)

4. DISCUSSION

4.1. Role of a cumulative softening effect induced by an intensive hydration toward an anti‐ageing biophysical effect

This long‐term clinical study confirms the effects of repeated hydration on the transformation of the skin hydration and mechanics. These effects are all the more important as the applications are repeated (measured effects more pronounced at D112 versus D56), which makes it possible to envisage biological modifications of the stratum corneum. As suggested by digital simulation, an improvement of crow's feet clinical scoring is demonstrated, which can be perceived by the consumer.

The results of this long‐term study suggest that a cumulative hydration provides anti‐ageing and perceivable effects to the consumer.

Specific effects of high concentration of glycerol are significant effects on long‐term hydration (D7, D56, and D112), a greater elasticity of the skin with the 20% glycerol formula and more specifically of the stratum corneum as measured by the DTM. These experimental results suggest that the skin transformation induced by glycerol is more specifically located in the superficial skin layers. Nevertheless, the two formulas have a moisturizing and softening effect, which induces a reduction of crow's feet scoring. In the future, it would be interesting to perform a new clinical study with a more neutral placebo (with very low moisturizing and softening effect), in order to more clearly demonstrate the impact of high concentration of humectants/plasticizers on crow's feet wrinkles.

Thanks to the inverse finite element analysis used in this clinical study, it was demonstrated that the 20% glycerol formula induced as follows:

1. A softening of the stratum corneum on the face, which corresponds to a rejuvenation effect of the skin. ²

2. A softening of the dermis (−40% in Young's modulus on the face), which also corresponds to a skin rejuvenation, because previous in vivo studies ²⁷ , ²⁸ , ²⁹ demonstrated a mechanical stiffening of the skin with aging measured by suction experiments.

4.2. Potential antiaging biological mechanisms involved in stratum corneum and epidermis: mechanobiological hypothesis

Mechanical forces are known to regulate homeostasis of the skin and play a role in different skin diseases, such as fibrosis, blistering diseases, or alopecia. Regeneration of lost tissue after injury is also determined by the local wound environment stiffness. Changing the mechanical forces and stress in the epidermal and dermal layers could modify the mechano‐transduction signals perceived by keratinocytes and fibroblasts. ³⁰ The biological effects measured in this clinical study could be linked to a decrease of compressive stress in the epidermis and to a new balance between the extracellular forces exerted on cells and cytoskeletal tensegrity. Additionally, a previous work ³¹ demonstrated that formula with high concentration of glycerol can have marked effects and reduce the SC drying stresses. Computational modeling of full thickness skin (human cheek) revealed that this reduction of SC drying stress results in a drastic change of strains in the epidermal layer and in the lower mechanoreceptor‐laden layers of skin (resulting in a decrease of tightness perception). This confirms the mechanobiological effect in the living epidermis induced by the surface stratum corneum softening.

From the 54 proteins of the stratum corneum analyzed after 56 days of treatment, 4 proteins are upregulated in the zone treated with 20% glycerol. The four proteins are kallikrein 5, kallikrein 7, galectin 7, and PARK7. The proteomics results illustrate the biological and physiological impacts that glycerol can have on the skin, beyond its direct biophysical effects, potentially explaining some long‐term effects.

These results show that

1. In addition to its roles described as a booster of proteases activities involved in desquamation by water supply of dry skins, glycerol promotes the expression of the two key desquamation associated proteases‐kallikreins: kallikrein 5 and 7. ³² Corneodesmosomes (proteins), which are the targets of these two proteases, are elements of the tissue mechanical properties through their roles in cell junctions. Computational cellular models ³³ suggest the important role of these junctions in the mechanical stiffness of the stratum corneum. The long‐lasting softening of stratum corneum, as measured in this clinical study, would confirm this hypothesis. ²

2. Glycerol may have an important role in maintaining epidermal homeostasis through its potential to induce the expression of galectin 7, a major epidermal lectin that has been shown to be involved in the preservation of epidermal regeneration of skin subjected to environmental stress like UVs ³⁴ or protecting the barrier from inflammatory stress like in AD. ³⁵ This glycan‐binding protein may directly influence cell/extracellular matrix interactions and thus play a role in tissue mechanical properties.

3. Besides, these roles of induction of epidermal proteins linked to skin renewal and regeneration capacities through the upregulation of the biomarkers KLK5, KLK7, and LGALS7, and glycerol has an impact on the expression of another major biomarker: PARK7 recently described as a key player in skin antioxidant defenses ³⁶ and also described as an enzymatic protector against glycation disorders associated with skin aging ³⁷ and lack of epidermal plasticity. ³⁸

4.3. Numerical simulation and mechanobiological hypothesis

The use of numerical simulation of wrinkles or skin made it possible to explain the levers of actions at the origin of this extra performance:

1. On the short term, a mechanical softening of the stratum corneum leads to a reduced skin wrinkling, ² emphasized during the lateral compressions exerted by the muscles during facial gestures (immediate effect).

2. On the long term, the reduction of compressive stresses at the level of the epidermis, the papillary and reticular dermis under the wrinkle, would allow a regeneration of the extracellular matrix by mechanobiological effects.

In fact, it is now accepted that the application of compressive forces leads to resorption of skin tissue, whereas the application of tensile forces stimulates the cellular proliferation. ³⁹ The cumulative compressive stress induced by the repetitive skin folding during facial expressions would lead to a tissue degradation and resorption, making such locations preferential folding zones where wrinkles can permanently set and worsen over the time. The compressive stress relief induced by the long‐lasting softening of the stratum corneum would not only delay such degradation (potential wrinkle prevention) but also would encourage the tissue regeneration (potential wrinkle repair). This would contribute to the long‐lasting reduction of crow's feet wrinkles and to the long‐lasting improvement of global skin elasticity.

Proteomics results suggest that mechanobiological effects are induced by the decrease of the mechanical compressive stress in the epidermidis, which could lead to a regenerative effect in the epidermal living layers and in the stratum corneum. In vitro mechanobiological tests could be performed to validate this hypothesis.

5. CONCLUSION

The combination of instrumental in vivo data, inverse finite element analysis, and proteomic data gives new insights into the biophysical and biological mechanisms involved after a cumulative effect of a 20% glycerol based moisturizer over 4 months. Several skin parameters have been improved, with a time effect as shown in the following:

1. A lasting increase in hydration versus baseline and vehicle.

2. A prevailing skin elasticity increase (DTM and Cutometer) and stratum corneum softening versus baseline and vehicle.

3. A softening of the dermis, associated with potential antiaging effect.

4. An improvement in the clinical score of crow's feet wrinkles, although no extra performance with increasing moisturizing potent has been demonstrated.

Stratum corneum and epidermal relaxation could be achieved through the observed induction of proteins related to modifications of the extracellular/ultrastructural proteins associated with the desquamation process, the cell‐glycan extracellular interactions, and the glycation/oxidation linked proteins properties alterations (respectively KLK5–KLK7, LGALS7, and PARK7).

Our results provide an in vivo demonstration that a cumulative skin surface softening treatment might have an anti‐ageing impact, both on crow's feet wrinkles and skin elasticity. The use of high moisturizing potent formulations should bring additional performance to simple moisturizing formulations.

CONFLICT OF INTEREST

This work was supported by L'Oréal Research & Innovation.

Santoprete R, Hourblin V, Foucher A, et al. Reduction of wrinkles: From a computational hypothesis to a clinical, instrumental, and biological proof. Skin Res Technol. 2023;29:1–11. 10.1111/srt.13267

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.