Determination of wrinkles in individual humans

Abstract

The present invention relates to a method for determining the predicted wrinkle grade of human skin, comprising: providing human skin cells; determining the methylation level of at least one CpG dinucleotide or the expression level of at least one gene; and determining the predicted wrinkle grade of the skin cells by comparing the determined methylation level with empirically determined data representing a correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade of at least one human individual; or determining the predicted wrinkle grade of the skin cells by comparing the determined gene expression level with empirically determined data representing a correlation between the gene expression level and the visual wrinkle grade of at least one human individual.

Description

Technical Field

[0001] The present invention relates to determining the predicted wrinkle grade of human skin by analyzing the epigenetic pattern of a specific site of DNA or by analyzing the expression level of a specific gene. The methylation level or gene expression level at the CpG site of a skin sample is evaluated to accurately determine the predicted wrinkle grade of human skin. The predicted wrinkle grade is then correlated with the visual wrinkle grade using an average absolute error of less than 20 wrinkle grade units.

Background Art

[0002] Aging is the time-dependent general loss of function and health over an individual's lifespan. The accumulation of molecular damage is considered to be one of the main features of aging. For example, epigenetic changes in DNA are known to occur over time and are recognized as an indicator of the degree of the aging process in human individuals. In particular, the degree of methylation of CpG dinucleotides has been confirmed to be useful in methods for determining biological age and predicting chronological age. Such known methods are based on the analysis of various CpG dinucleotides in both skin cells and blood cells. A good correlation between an individual's chronological age and biological age has been achieved by methods for evaluating the degree of methylation of CpG dinucleotides. In general terms, biological age is related to an individual's overall health status. Methods for predicting biological age can be useful for addressing a number of questions in research related to developmental biology, cancer, and aging, as discussed, for example, in Horvath, Genome Biology 2013, 14:R115.

[0003] Biological age provides an indicator of an individual's overall health status and degree of aging, but a direct correlation between biological age and a specific phenotype, such as a specific skin aging phenotype, has not yet been established. Measuring and monitoring specific properties related to the skin phenotype at the molecular level can be useful in many applications, such as identifying active agents to minimize the visual signs of aging.

[0004] Therefore, it is desirable to provide a method for determining the characteristic phenotype of an individual's skin. [Overview of the Initiative] [Means for solving the problem]

[0005] The present invention is defined in the appended claims.

[0006] According to the first aspect, a method for determining the predicted wrinkle grade of human skin: a) To provide human skin cells; b) Determining the methylation level of at least one CpG dinucleotide in a specific region of at least one chromosome of the skin cell, or determining the expression level of at least one gene in the skin cell; and c) A method is provided which includes determining the predicted wrinkle grade of the skin cell by comparing the determined methylation level with empirically determined data representing a correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade of at least one human individual, or determining the predicted wrinkle grade of the skin cell by comparing the determined gene expression level with empirically determined data representing a correlation between the gene expression level and the visual wrinkle grade of at least one human individual.

[0007] According to the second aspect, a method for testing an activator comprising the method according to the first aspect, further comprising the following steps: d) bringing the skin cells from step a) into contact with the activator; e) Determining the predicted wrinkle grade of the skin cells in step d) according to the method of the first embodiment; A method is provided which includes f) comparing the predicted wrinkle grade determined in steps a) to c) with the predicted wrinkle grade determined in step e).

[0008] According to a third embodiment, a composition is provided comprising one or more activators identified by the method of the second embodiment.

[0009] According to a fourth aspect, the use of DNA methylation levels or gene expression levels is provided for determining the predicted wrinkle grade of human skin according to the method of the first aspect.

[0010] According to the fifth aspect, a computer-readable medium storing computer-executable instructions for causing a computer to perform a method for determining the predicted wrinkle grade of human skin: a) inputting at least one value of the determined methylation level of at least one CpG dinucleotide in a specific region of at least one chromosome of the skin cell, or inputting at least one value of the determined expression level of at least one gene of the skin cell; b) comparing the determined methylation level to stored data representing a correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade of at least one human individual, or comparing the gene expression level to stored data representing a correlation between the gene expression level and the visual wrinkle grade of at least one human individual; c) A computer-readable medium is provided that includes displaying the predicted wrinkle grade.

[0011] A sixth aspect provides a kit for determining the predicted wrinkle grade of human skin according to the method of the first aspect, comprising at least one oligonucleotide primer for amplifying and / or sequencing at least one CpG nucleotide of at least one nucleotide sequence, or at least one oligonucleotide primer for amplifying and / or sequencing a fraction of at least one exon of at least one nucleotide sequence.

[0012] Certain embodiments of the present invention may offer one or more of the following advantages: ● Desired ease of sample collection from the individual; ● Desirable ease of acquiring bioinformatics data; ● Desired cost-effective data collection; ● Desired accuracy for wrinkle grade prediction; ● Desired ease of identifying activators for controlling wrinkle grade; and ● Desired ease of predicting the effect of the surfactant on wrinkle grade over time.

[0013] Details, examples, and selections provided in connection with any particular one or more of the described aspects of the present invention apply equally to all aspects of the present invention. Any combination of embodiments, examples, and preferred configurations described herein, in all possible variations thereof, are incorporated into the present invention unless otherwise shown herein or unless explicitly contradicted in the context. [Brief explanation of the drawing]

[0014] The present invention will be further explained with reference to the following figures. [Figure 1] A graph showing the correlation between predicted wrinkle grade and visual wrinkle grade based on a test set using DNA methylation; the regression line (or fitting line) on this graph is the best estimate of the linear relationship between this predicted wrinkle grade and visual wrinkle grade; and Pearson correlation (R), mean absolute error (MAE), and significance of the correlation (p) are provided to offer additional statistical indicators. [Figure 2] A graph showing the correlation between predicted wrinkle grade and visual wrinkle grade based on an independent dataset; using DNA methylation; the regression line (or fitting line) on this graph is the best estimate of the linear relationship between this predicted wrinkle grade and visual wrinkle grade; providing Pearson correlation (R), mean absolute error (MAE), and significance of the correlation (p) (shown to provide additional statistical indicators). [Figure 3]Graph showing the correlation between predicted wrinkle grade and visual wrinkle grade based on a test set using a DNA methylation method based on low methylation regions (LMR); the regression line (or fitting line) on this graph is the best estimate of the linear relationship between this predicted wrinkle grade and visual wrinkle grade; Pearson correlation (R), mean absolute error (MAE), and significance of the correlation (p) (shown to provide additional statistical metrics) are provided. [Figure 4] Graph showing the correlation between predicted wrinkle grade and visual wrinkle grade based on an independent data set; using a DNA methylation method based on low methylation regions (LMR); the regression line (or fitting line) on this graph is the best estimate of the linear relationship between this predicted wrinkle grade and visual wrinkle grade; Pearson correlation (R), mean absolute error (MAE), and significance of the correlation (p) (shown to provide additional statistical metrics) are provided. [Figure 5] Graph showing the correlation between predicted wrinkle grade and visual wrinkle grade based on a test set using a gene expression method; the regression line (or fitting line) on this graph is the best estimate of the linear relationship between this predicted wrinkle grade and visual wrinkle grade; Pearson correlation (R), mean absolute error (MAE), and significance of the correlation (p) (shown to provide additional statistical metrics) are provided. [Figure 6] Graph showing the correlation between predicted wrinkle grade and visual wrinkle grade based on an independent data set; using a gene expression method; the regression line (or fitting line) on this graph is the best estimate of the linear relationship between this predicted wrinkle grade and visual wrinkle grade; Pearson correlation (R), mean absolute error (MAE), and significance of the correlation (p) (shown to provide additional statistical metrics) are provided.

[0015] It should be understood that the following description and reference to the drawings relate to exemplary embodiments of the present invention and do not limit the scope of the claims.

Best Mode for Carrying Out the Invention

[0016] This invention is based on the finding that the wrinkle grade of human skin can be predicted with good accuracy using the methylation level or gene expression level of selected CpG nucleotides.

[0017] Wrinkles are lines, folds, or sagging in the skin, such as in the dermal layers, and are a natural phenomenon associated with the aging process. Wrinkles are most prominent on skin exposed to the sun. Such areas include, but are not limited to, the face, neck, hands, and forearms (antebrachium), where the skin becomes more sensitive to photodamage, which can lead to the aforementioned wrinkles. The degree of appearance of wrinkles is also called the wrinkle grade. Wrinkle grade is typically assessed by visual analysis, which is referred to herein as the “visual wrinkle grade.” There are various methods that can be implemented to determine the visual wrinkle grade. In this disclosure, the visual wrinkle grade is determined using a method involving a visual analysis of an individual’s face by a trained individual, also known herein as the “assessor.” In this method, the assessor is required to evaluate the visual appearance of the individual’s wrinkles and assign the individual a wrinkle grade unit. The wrinkle grade units assigned to an individual are based on the overall appearance of the facial skin, specifically the degree of crow's feet. The wrinkle grade units associated with an individual can range from 0 to 100 and follow a predefined and / or known rating scale, as seen, for example, in "Atlas du Vieillisment Cutane-Population Europeenne" ("Skin Aging Atlas-European Population"); Pr. Roland BAZIN et Al.; 2007; ISBN: 978-2-914738-94-1. In some examples, to assess the wrinkle grade, the assessor is provided with 10 photographs of individuals with varying degrees of wrinkles, from no wrinkles (wrinkle grade unit = 0) to high degree of wrinkles (wrinkle grade unit = 100). Despite the guidance provided to the assessor by using this method, the assessment of wrinkle grade units using this method is subjective. Therefore, multiple judges are asked to provide an evaluation of a single individual, and the visual wrinkle grade is the average of the evaluations provided by the judges.In this method, the evaluation of at least 10 judges, for example, at least 12 judges, at least 15 judges, at least 18 judges, at least 20 judges, at least 25 judges, at least 30 judges is applied.

[0018] As used herein, the term "wrinkle prediction grade" refers to the wrinkle grade of an individual and is based on the methylation level of DNA (i.e., epigenetic level) and / or gene expression level. The methylation level of DNA (i.e., epigenetic level) and / or gene expression level is determined in a skin sample. The wrinkle prediction grade disclosed herein is determined using the method of the first aspect of the present invention. In particular, the methylation level of a specific CpG dinucleotide or the expression level of a specific gene is evaluated as part of the present invention.

[0019] DNA methylation levels can be determined using DNA methylation level detection methods. Examples of DNA methylation level detection methods, but not limited to, include: (a) base-resolution sequencing approaches, e.g., whole-genome bisulfite sequencing (WGBS, also known as Methyl-Seq), enzyme Methyl-Seq (EM-Seq), anchor-based bisulfite sequencing (ABBS), post-bisulfite adapter tagging (PBAT), reduced-expression bisulfite sequencing (RRBS), or shotgun bisulfite sequencing (BS-Seq); (b) base-resolution microarray approaches, such as oligonucleotide-based (e.g., Reinders et al. "Genome-wide, high-resolution DNA methylation profiling using bisulfite-mediated cytosine conversion"; Genome Res. 2008 Mar; 18(3): 469-76) or bead-tipped (e.g., Illumina) (c) non-basic-resolution sequencing approaches, e.g., methylated DNA immunoprecipitation sequencing (MeDIP-Seq), methylated CpG island recovery assay (MIRA), methylation-sensitive restriction enzyme sequencing (MRE-Seq), or MethylCap-Seq (also known as MBDCap-Seq), (d) non-basic-resolution microarray approaches such as MeDIP-Chip or Comprehensive High-throughput Arrays for Relative Methylation (CHARM), and (e) targeted approaches, e.g., targeted bisulfite sequencing, targeted enzyme methyl-seq, ligation-mediated PCR, methylation-specific PCR (MS-PCR), methylation-sensitive high-resolution melting (MS-HRM), or conjugated bisulfite restriction analysis (COBRA).

[0020] DNA methylation levels may be expressed as either beta or m values. The beta value is defined as the number of detected methylated (M) molecules (i.e., nucleotides) divided by the total number of methylated (M) and unmethylated (U) molecules. In other words, M is equal to the number of detected methylated molecules or the signal intensity of the methylation probe; U is equal to the number of detected unmethylated molecules or the signal intensity of the unmethylation probe (see Equation 1).

number

number

[0021] In addition to the DNA methylation levels determined by the detection methods disclosed above, in some embodiments, the methods according to the present invention are performed using specific regions known as low-methylation regions (LMRs). LMRs are characterized by relatively low levels of DNA methylation ranging from 10 to 50% and often coincide with gene regulatory regions such as enhancers and promoters. LMRs can be determined by a number of known methods, including, for example, the application of MethylSeekR.

[0022] In some embodiments, the method according to the first embodiment may include determining the methylation levels of at least two CpG dinucleotides, or at least three CpG dinucleotides, or at least four CpG dinucleotides, or at least five CpG dinucleotides, or at least seven CpG dinucleotides, or at least ten CpG dinucleotides, or at least twelve CpGs, or at least fourteen CpG dinucleotides, or at least sixteen CpG dinucleotides, or at least nineteen CpG dinucleotides, or at least thirty CpG dinucleotides, or at least fifty CpG dinucleotides, or at least one hundred CpG dinucleotides. In some embodiments, the method according to the first embodiment may include determining the methylation levels of two CpG dinucleotides, or three CpG dinucleotides, or four CpG dinucleotides, or five CpG dinucleotides, or seven CpG dinucleotides, or ten CpG dinucleotides, or twelve CpG dinucleotides, or fourteen CpG dinucleotides, or sixteen CpG dinucleotides, or nineteen CpG dinucleotides, or thirty CpG dinucleotides, or fifty CpG dinucleotides, or one hundred CpG dinucleotides.In some embodiments, the method according to the first aspect may include determining the methylation levels of 2 to 19 CpG dinucleotides, or 5 to 50 CpG dinucleotides, or 10 to 100 CpG dinucleotides.

[0023] Gene expression levels can be determined using gene expression level detection methods. Examples of gene expression level detection methods, but not limited to: (a) sequencing-based high-throughput assays, e.g., ribonucleic acid sequencing (RNA-Seq), single-cell RNA sequencing (scRNA-Seq), spatial transcriptomics (e.g., Visium spatial gene expression and FISSEQ), sequential gene expression analysis (SAGE), or variations of sequential gene expression analysis (SAGE) (e.g., LongSAGE, RL-SAGE, and SuperSAGE); (b) hybridization-based high-to-medium-throughput assays, e.g., microarrays (e.g., in-situ synthesized oligonucleotide-based, cDNA-based, bead-based, and tiling-based microarray platforms), Luminex xMAP Technology, NanoString nCounter technology, multiplexed error-robust fluorescence in-situ hybridization (MERFISH), digital spatial transcriptomics (e.g., GeoMx DSP), or RNA Scope. Examples include (c) polymerase chain reaction (PCR)-based medium-to-low throughput assays, e.g., digital PCR or quantitative and reverse transcription PCR (qRT-PCR and RT-PCR), e.g., TaqMan assays and low-density arrays (LDA), and (d) hybridization-based medium-to-low throughput assays, e.g., QuantiGene RNA assays, fluorescence in situ hybridization (FISH), Northern blotting, dot blotting, or slot blotting.

[0024] Gene expression levels may be expressed based on counts (e.g., raw counts, raw counts normalized against housekeeping genes and / or internal standards, as well as normalized counts such as TPM, RPKM, and FPKM), fluorescence signal intensity (e.g., of hybridization-based assays), or changes (e.g., log2FC or FC).

[0025] In addition to the gene expression levels determined by the detection methods disclosed above, further training of models is also possible with a reduced number of genes, for example, 17,653, 11,769, 5,885, 1,177, 100, 10, 2, or 1 gene. Such gene selection may be performed by obtaining a model trained with all genes and performing in silico knockout on each gene to calculate the overall effect size of the gene on wrinkle grade prediction.

[0026] In some embodiments, the method according to the first embodiment may include determining the expression levels of at least two genes, or at least three genes, or at least five genes, or at least ten genes, or at least fifty genes, or at least 100 genes, or at least 1,000 genes, or at least 2,000 genes, or at least 5,000 genes, or at least 10,000 genes, or at least 15,000 genes, or at least 20,000 genes, or at least 25,000 genes, or at least 30,000 genes, or at least 35,000 genes. In some embodiments, the method according to the first aspect may include determining the expression levels of two genes, or three genes, or five genes, or ten genes, or fifty genes, or 1,000 genes, or 1,000 genes, or 2,000 genes, or 5,000 genes, or 10,000 genes, or 15,000 genes, or 20,000 genes, or 25,000 genes, or 30,000 genes, or 35,000 genes, or 38,892 genes.In some embodiments, the method according to the first embodiment may include determining the expression levels of 2 genes to 38,892 genes, or 5 genes to 35,000 genes, or 10 genes to 30,000 genes, or 50 genes to 25,000 genes, or 100 genes to 20,000 genes, or 500 genes to 15,000 genes, or 1,000 genes to 10,000 genes, or 2,000 genes to 8,000 genes.

[0027] In some embodiments, the gene may be selected from one or more of the following:

[0028] The empirically determined data representing the correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade according to the method of the present invention are obtained as follows: Reference data for a subject is obtained by 1) measuring the DNA methylation level of at least one CpG dinucleotide in a specific region of at least one chromosome of one subject, and 2) determining the visual wrinkle grade of the subject. The methylation level of at least one CpG dinucleotide is determined from skin cells collected from a part of the subject's human body, such as the arm, face, neck, décolleté and / or hand. Reference data is collected from at least 100 subjects to establish the empirically determined data. For example, reference data is collected for at least 110 subjects, or at least 130 subjects, or at least 150 subjects, or at least 200 subjects, or at least 250 subjects, or at least 300 subjects, or at least 350 subjects, or at least 400 subjects, or at least 450 subjects, or at least 500 subjects. The subjects are: a) aged 1 to 120 years; b) of any sex, including male, female, and mixed sex; b) of any race, including Caucasian, African, Asian, or mixed race; and d) of any phototype, including phototype 1, phototype 2, phototype 3, phototype 4, phototype 5, phototype 6, or mixed phototype. Empirically determined data are obtained by correlating measured DNA methylation levels with corresponding visual wrinkle grades from at least 100 subjects. The correlation can be expressed mathematically, for example, through linear regression and / or machine learning approaches such as support vector machines, generalized linear models, or artificial neural networks. This correlation represents the empirically determined data.

[0029] The empirically determined data representing the correlation between gene expression levels and visual wrinkle grades using the method of the present invention are obtained as follows: Reference data for a subject is obtained by 1) measuring the gene expression level of at least one gene in one subject, and 2) determining the visual wrinkle grade of the subject. The gene expression level is determined from skin cells collected from a part of the subject's human body, such as the arm, face, neck, décolleté, and / or hand. Reference data is collected from at least 100 subjects to establish empirically determined data. For example, reference data is collected for at least 110 subjects, or at least 130 subjects, or at least 150 subjects, or at least 200 subjects, or at least 250 subjects, or at least 300 subjects, or at least 350 subjects, or at least 400 subjects, or at least 450 subjects, or at least 500 subjects. The subjects are: a) aged 1 to 120 years; b) of any sex, including male, female, and mixed sex; b) of any race, including Caucasian, African, Asian, or mixed race; and d) of any phototype, including phototype 1, phototype 2, phototype 3, phototype 4, phototype 5, phototype 6, or mixed phototype. Empirically determined data are obtained by correlating gene expression levels with corresponding visual wrinkle grades from at least 100 subjects. The correlation can be expressed mathematically via linear regression and / or machine learning approaches, such as support vector machines, generalized linear models, or artificial neural networks. This correlation represents the empirically determined data.

[0030] Visual wrinkle ratings and / or predicted wrinkle ratings may be influenced by many parameters, including genetic background, age, UV exposure, smoking, exposure to pollution, and / or repeated facial expressions.

[0031] Visual wrinkle grades and / or predicted wrinkle grades can be influenced by several factors, such as protecting the skin from sunlight, moisturizing, avoiding smoking, and maintaining a healthy diet. Wrinkle grades can also be affected by the use of surfactants.

[0032] In certain embodiments, the skin cells used according to the method of the present invention are collected from a human individual. In certain embodiments, the human skin cells used according to the method of the present invention are obtained by collecting the required whole skin sample (i.e., all skin layers) from the individual. In certain embodiments, the human skin cells used according to the method of the present invention are obtained by collecting skin samples from the arm, face, neck, décolleté and / or hand. In certain embodiments, the human skin cells used according to the method of the present invention are obtained by culturing skin cells using an in vitro method.

[0033] Samples from individuals may be collected using aspiration blistering, punch biopsy, shave biopsy, microbiopsy, tape stripping, or during any surgical procedure such as reconstructive surgery, lifting, or transplantation.

[0034] Skin samples may be taken from the epidermis and / or dermis. Skin samples may be taken from the entire skin, i.e., from both the epidermis and dermis.

[0035] Human skin cells may be cultured from small samples of skin cells taken from an individual. The collected human skin cells are grown in vitro in a container such as a Petri dish, in a culture medium or substrate that supplies essential nutrients.

[0036] The human skin cells used may be a mixture of harvested cells and cultured cells.

[0037] In certain embodiments, the specific region of a chromosome containing the CpG dinucleotide according to the present invention may be a coding region or a non-coding region. In certain embodiments, the CpG dinucleotide may be present in both coding and / or non-coding regions. The CpG dinucleotide may be found in a single specific region or in different specific regions.

[0038] In certain embodiments, the method of the first embodiment further includes the step of estimating a predicted wrinkle grade for a human individual. The correlation between the predicted wrinkle grade and the visual wrinkle grade is high using the method according to the present invention, and the difference between the predicted wrinkle grade and the visual wrinkle grade can be expressed using the mean absolute error (MAE). The mean absolute error (MAE) for wrinkle grade units determined by the method according to this specification is about 20 or less, about 18 or less, about 16 or less, about 14 or less, about 12 or less, about 11 or less, about 10 or less, about 9 or less, about 8.8 or less, about 8 or less, about 6 or less, about 4 or less, and 2 or less.

[0039] The difference between the predicted wrinkle grade and the visual wrinkle grade may be determined for a single individual. The deviation of a single data point from the best-fitting line superimposed on data points from all individuals can vary in units of approximately 0 to 20 wrinkle grades. For example, the deviation of a single data point from the best-fitting line is in units of approximately 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, and 20 wrinkle grades.

[0040] In certain embodiments, the predicted wrinkle grade of an unknown individual is determined as described above. From this predicted wrinkle grade, the individual's visual wrinkle grade can be estimated by applying the expected MAE to the CpG data points used, or by applying the expected MAE to the gene expression levels used.

[0041] Visual wrinkle grades and / or predicted wrinkle grades can also be reduced by applying activators to skin cells in the form of pharmaceuticals and / or cosmetics. The method of the present invention also allows for reliable prediction of medium-term and / or long-term reductions in wrinkle grades upon application of activators. This prediction may be made at an early stage and without the need to monitor the progression of skin phenotypes over a long period, which can be costly and time-consuming. While we do not wish to be bound by theory, it is conceivable that a method for predicting the progression of wrinkle grades over time using the method of the present invention may be due to the relationships discovered by the inventors between specific biological changes determined by this method and associated wrinkle grades. Such a method is not known in the art and has not been attempted. Activators can be applied to skin cells in vitro or in vivo. Using an in vitro method, activators are added to the culture medium of skin cells. Using an in vivo method, skin cells of a human organism can be applied to the activator using topical, subcutaneous, or intradermal administration.

[0042] As used herein, “activator” is any agent having therapeutic and / or cosmetic effects on an individual. Therapeutic effects are the treatment and / or prevention of disease. Cosmetic effects are improvements in appearance, such as treating and / or preventing signs of molecular aging. Cosmetic effects may also be reductions in wrinkle grade, including visual wrinkle grade and / or predicted wrinkle grade. In certain embodiments, the activator is a cosmetic agent. When applied to an individual, a cosmetic agent can enhance attractiveness, alter appearance, beautify, and / or cleanse. This cosmetic agent can also prevent and / or treat signs of phenotypic aging of human skin, such as wrinkle formation, pale complexion, reduced wound healing ability, loss of elasticity, and sagging.

[0043] In certain embodiments, the predicted wrinkle grade is determined before and after contacting skin cells with the activator. For example, the predicted wrinkle grade is determined, skin cells are contacted with the activator, and then the predicted wrinkle grade is determined again. The predicted wrinkle grade is compared before and after contacting the skin cells with the activator. A decrease in the predicted wrinkle grade after treatment indicates the therapeutic and / or cosmetic effect of the activator. The time between contacting the skin cells with the activator and determining the predicted wrinkle grade may vary. The steps of contacting the skin cells with the activator and determining the predicted wrinkle grade may be repeated to obtain information on the effect of the activator on skin cells over time. This method may have the advantage of providing a rapid and cost-effective method for identifying active cosmetic and / or pharmaceutical compounds and / or extracts, particularly for treating phenotypic characteristics of skin, such as wrinkle grade. In certain embodiments, the activators identified by the method according to the present invention are formulated to produce a dermatological composition. The dermatological composition may be a cosmetic composition and / or a pharmaceutical composition. Dermatological compositions can be used to reduce predicted and / or visual wrinkle levels.

[0044] This disclosure may be described by one or more of the following paragraphs: A. A method for determining the predicted wrinkle grade of human skin: a) To provide human skin cells; b) Determining the methylation level of at least one CpG dinucleotide in a specific region of at least one chromosome of the skin cell, or determining the expression level of at least one gene in the skin cell, and c) A method comprising determining the predicted wrinkle grade of the skin cells by comparing the determined methylation level with empirically determined data representing a correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade of at least one human individual, or determining the predicted wrinkle grade of the skin cells by comparing the determined gene expression level with empirically determined data representing a correlation between the gene expression level and the visual wrinkle grade of at least one human individual. B. The method according to paragraph A, wherein the skin cells are collected from a human individual. C. The method according to paragraph A, wherein the skin cells are cultured in vitro. The method according to any one of the preceding paragraphs, wherein the methylation level in step b) is determined using a DNA methylation level detection method. E. The method according to paragraph D, wherein the methylation level determined by the DNA methylation level detection method is expressed as a beta value or an m value. F. A method according to paragraph D or E, wherein the method for detecting the DNA methylation level is selected from a) a base-resolution sequencing approach, b) a base-resolution microarray approach, c) a non-base-resolution sequencing approach, d) a non-base-resolution microarray approach, or e) a targeted approach. A method according to paragraph F, wherein the base-resolution sequencing approach is selected from whole-genome bisulfite sequencing (WGBS), enzyme methyl-Seq (EM-Seq), anchor-based bisulfite sequencing (ABBS), post-bisulfite adapter tagging (PBAT), reduced-expression bisulfite sequencing (RRBS), or shotgun bisulfite sequencing (BS-Seq). H. The method according to paragraph F, wherein the base-resolution microarray approach is based on oligonucleotides or bead chips. I. The method according to paragraph F, wherein the non-basic-resolution sequencing approach is selected from methylated DNA immunoprecipitation sequencing (MeDIP-Seq), methylated CpG island recovery assay (MIRA), methylation-sensitive restriction enzyme sequencing (MRE-Seq), or MethylCap-Seq (MBDCap-Seq). J. The method according to paragraph F, wherein the non-base resolution microarray approach is selected from MeDIP-Chip or Comprehensive High-throughput Arrays for Relative Methylation (CHARM). K. The method according to paragraph F, wherein the targeting approach is selected from targeted bisulfite sequencing, targeted enzyme methyl-seq, ligation-mediated PCR, methylation-specific PCR (MS-PCR), methylation-sensitive high-resolution melting (MS-HRM), and conjugated bisulfite restriction analysis (COBRA). The method according to any one of paragraphs A to C, wherein the expression level in step b) is determined using a gene expression level detection method. M. The method according to any one of paragraph L, wherein the gene expression level detection method is selected from a) a sequencing-based high-throughput assay, b) a hybridization-based high-to-medium-throughput assay, c) a polymerase chain reaction (PCR)-based medium-to-low-throughput assay, or d) a hybridization-based medium-to-low-throughput assay. N. The method according to paragraph M, wherein the sequencing-based high-throughput assay is selected from ribonucleic acid sequencing (RNA-Seq), single-cell RNA sequencing (scRNA-Seq), spatial transcriptomics, sequential gene expression analysis (SAGE), and variations of sequential gene expression analysis (SAGE). O. The method according to paragraph N, wherein the spatial transcriptomics is selected from Visium spatial gene expression and FISSEQ. P. The method according to paragraph N, wherein the variation of sequential gene expression analysis (SAGE) is selected from LongSAGE, RL-SAGE, and SuperSAGE. Q. The method according to paragraph M, wherein the hybridization-based high-to-medium throughput assay is selected from microarrays, Luminex xMAP technology, NanoString nCounter technology, multiplexed error-robust fluorescence in situ hybridization (MERFISH), digital spatial transcriptomics, and RNAscope ISH. R. The method according to paragraph Q, wherein the microarray is selected from in-situ synthesis oligonucleotide-based, cDNA-based, bead-based, and tiling-based microarray platforms. S. The method according to paragraph Q, wherein the digital space transcriptomics is a GeoMx DSP. T. The method according to paragraph M, wherein the polymerase chain reaction (PCR)-based medium-to-low throughput assay is selected from digital PCR, quantitative PCR, reverse transcription PCR, and low-density array (LDA). U. The method according to paragraph T, wherein the reverse transcription PCR comprises qRT-PCR and RT-PCR. V. The method according to paragraph T or U, wherein the reverse transcription PCR is a TaqMan assay. W. The method according to paragraph M, wherein the hybridization-based medium-to-low throughput assay is selected from QuantiGene RNA assay, fluorescence in situ hybridization (FISH), Northern blotting, dot-blotting, and slot-blotting. X. The method according to paragraphs A, B, or D-W, further comprising the step of estimating the visual wrinkle grade of the human individual. Y. The method according to paragraph X, wherein the predicted wrinkle grade is estimated to be within 20 wrinkle grade units of the visual wrinkle grade. Z. The method according to any one of the preceding paragraphs, further comprising the step of bringing the skin cells into contact with an activator. AA. The method according to paragraph Z, wherein the predicted wrinkle grade is determined before and after skin cells come into contact with the activator. A method for testing an activator comprising the method described in any one of paragraphs A to W, further comprising the following steps: d) bringing the skin cells of step a) into contact with the activator; e) Determining the predicted wrinkle grade of the skin cells in step d) according to the method described in any one of paragraphs A to W; f) A method comprising comparing the predicted wrinkle grade determined in steps a) to c) with the predicted wrinkle grade determined in step e). The method according to paragraph BB, wherein step CC) is performed either in vivo or in vitro. DD. The method according to paragraph BB or paragraph CC, wherein the activator is a cosmetic agent and / or a therapeutic agent. EE. The method according to any one of paragraphs BB-DD, for identifying an activator that prevents and / or treats signs of phenotypic aging of human skin. A skin composition comprising one or more activators as specified by the method described in any one of paragraphs BB to EE. GG. Use of paragraph FF composition to reduce predicted wrinkle grade. HH. Use of DNA methylation levels or gene expression levels to determine the predicted wrinkle grade of human skin according to the method described in any one of paragraphs A-W. II. A computer-readable medium storing computer-executable instructions for causing a computer to execute a method for determining the predicted wrinkle grade of human skin: a) inputting at least one value of the determined methylation level of at least one CpG dinucleotide in a specific region of at least one chromosome of the skin cell, or inputting at least one value of the determined expression level of at least one gene of the skin cell; b) Comparing the determined methylation level with stored data representing a correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade of at least one human individual, or comparing the gene expression level with stored data representing a correlation between the gene expression level and the visual wrinkle grade of at least one human individual; c) A computer-readable medium that includes displaying the predicted wrinkle grade. JJ. The stored data is a computer-readable medium as described in paragraph II, including at least one linear regression equation. A kit for determining the predicted wrinkle grade of human skin according to the method described in any one of paragraphs A to W of KK, comprising at least one oligonucleotide primer for amplifying and / or sequencing at least one CpG nucleotide of at least one nucleotide sequence, or at least one oligonucleotide primer for amplifying and / or sequencing a fraction of at least one exon of at least one nucleotide sequence.

[0045] In certain embodiments, the method according to the present invention may have one or more of the following effects: ● An efficient method for predicting the degree of wrinkles; ● A reliable correlation between visual wrinkle grade and predicted wrinkle grade; ● Reducing the time required to determine the wrinkle grade; ● An efficient method for determining the effect of pharmaceuticals or cosmetic agents on the degree of wrinkles; ● An efficient method for predicting the effect of surfactants on wrinkle grade; ● An effective method for predicting the decline in wrinkle grade over time as a result of applying an activator, for example, over a long period of time such as more than one year; ● An efficient method for predicting the long-term effects of activators on an individual, providing personalized predictions.

[0046] The present invention may include any combination of the features and / or limitations referred to herein, except for any mutually exclusive combinations of features. The foregoing description applies to specific embodiments of the invention for the purpose of illustrating the invention. However, it will be apparent to those skilled in the art that many modifications and variations are possible to the embodiments described herein. Such modifications and variations shall be within the scope of the invention as defined in the appended claims. [Examples]

[0047] Example 1: Aspirated blisters with a diameter of 7 mm were collected from the palmar forearms of 461 female participants aged 29 to 84 years in the Population-Based Study of Health in Pomerania (SHIP-TREND-1) (Volzke et al., “Cohort Profile Update: The Study of Health in Pomerania (SHIP)”; Int J Epidemiol, 51, e372-e383, 2022). Aspirated blisters were obtained by applying a negative pressure of 180 mbar for 30 minutes, followed by 320 mbar until blister formation. The epithelium (roof) of the aspirated blisters was prepared with surgical scissors and forceps, rapidly frozen in liquid nitrogen, and stored at -80°C until further use.

[0048] DNA derived from the epithelium of aspirated blister tissue from each of the 461 participants was isolated using the QIAmp DNA Investigator kit (Qiagen) according to the manufacturer's instructions. The DNA methylation patterns of the 461 female epidermal samples were determined using the Infinium® MmethylationEPIC BeadChip array (Illumina).

[0049] Methylation data analysis was performed using the R Bioconductor packages minfi (Aryee et al., “Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays”; Bioinformatics, 30, 1363-9, 2014) and limma (Ritchie et al., “limma powers differential expression analyses for RNA-sequencing and microarray studies”; Nucleic Acids Res, 43, e47, 2015). Specifically, raw .idat files were read and preprocessed. Methylated loci (probes) were filtered based on several criteria: high detection p-value (threshold P > 0.01, parameter provided by minfi), location on sex chromosomes, ability to self-hybridize, and potential SNP contamination. Sequence normalization was performed using the preplaceQuanile function available in minfi. Quality control was performed after every preprocessing step.

[0050] In addition to DNA methylation levels, portrait photographs were taken of each of the 461 research participants. Before the photograph was taken, each volunteer had to close their eyes, cover their hair, and remove all makeup and accessories. The images were captured using the following standardized setup: a flash diffuser was employed to ensure uniform illumination, non-polarization, and color control. Based on these photographs, a panel of experts consisting of more than 30 specialists assigned a wrinkle grade to each research participant on a scale of 0 to 100. Each expert was trained with a panel set of 100 images, provided with image cards of exemplary crow's feet (the area around the eyes) on a 10-point wrinkle grade scale as orientation. To avoid bias in the volunteers' judgments, the photographs were presented to each expert randomly, ensuring that the 30 experts did not judge the volunteers in the same order. Finally, the average of these expert judgments became the individual visual wrinkle grade for each volunteer.

[0051] Next, the sample group was divided into two: epidermal samples from 378 female participants were used for training, and 83 epidermal samples from female participants were used for testing. Dividing the sample group into two subsets—approximately 80% allocated for training and approximately 20% for validation—was thought to ensure that the target variable was equally distributed between the training and validation sets. The wrinkle grade predictor was trained using the cv.glmnet function (alpha=0, lambda=1708.46) from the glmnet R package, with the beta value of the training dataset applied (Friedman et al., "Regularization Paths for Generalized Linear Models via Coordinate Descent"; Journal of Statistical Software, 33, 1-22; 2010), in 10x cross-validation mode with visual wrinkle grade as the outcome variable, i.e., the target variable. This method establishes empirically determined data representing the correlation between the methylation level of the above-mentioned CpG nucleotides and the visual wrinkle grade of at least one human individual, and is referred to herein as a “trained wrinkle grade predictor.” The trained wrinkle grade predictor was validated using a test set and independent publicly available datasets (Holzscheck et al., “Multi-omics network analysis reveals distinct stages in the human aging progression in epidermal tissue”; AGING, 12; 2020). Predicting each volunteer's wrinkle grade from the test set of empirically determined data compared to the corresponding visual wrinkle grade showed a very high correlation (R=0.86 and P=1.03 × 10⁻⁶). -21 A mean absolute prediction error of 8.81 wrinkle grade units was obtained (see Figure 1). The same applies to the validation of the trained wrinkle predictor using an independent dataset, which yielded a mean absolute prediction error of 9.66 wrinkle grade units (see Figure 2).

[0052] Example 2: Next, Example 1 was repeated by reducing the determined DNA methylation patterns of 378 female epidermal samples to the DNA patterns of specific low-methylated regions (LMRs) used as input for model training, via the Infinium® MmethylationEPIC BeadChip array (Illumina).

[0053] The LMRs used were determined by combining whole-genome bisulfite sequencing (WGBS) methylation data derived from the epidermis of young and elderly subjects (Raddatz, G. et al., "Aging is associated with highly defined epigenetic changes in the human epidermis"; EPIGENETICS CHROMATIN; 6(1), 36; 2013), resulting in an average 14-fold coverage of strand-specific CpGs. MethylSeekR (Burger, L. et al., "Identification of active regulatory region from DNA methylation data", NUCLEIC ACIDS RES; 41(16), e155; 2013) was then applied to this dataset using default parameters to identify skin-specific LMRs, and those overlapping with at least one probe from the EPIC methylation array were ultimately selected for further analysis. This resulted in a set of 40,140 LMRs. For training the wrinkle grade predictor, the mean beta value of each LMR was calculated, taking into account the CpG located within its specific region, and used as data input according to the same method as described in Example 1, with the parameter alpha set to 0 and lambda set to 125.62.

[0054] Predicting each volunteer's wrinkle grade from an empirically determined data set, compared to the corresponding visual wrinkle grade, showed a very high correlation (R=0.86 and P<2.2×10⁻⁶). -16A mean absolute prediction error of 8.73 wrinkle grade units was obtained (see Figure 3). The mean absolute prediction error of 13.95 wrinkle grade units was obtained for the independent dataset (R=0.89 and P<2.2x10). -16 The same applies to the validation of pre-trained wrinkle predictors using (see Figure 4).

[0055] Example 3: Aspirated vesicular epithelium from 437 female participants aged 29–84 years in the Pomerania Cohort Health Study (SHIP-TREND-1) (Volzke et al., "Cohort Profile Update: The Study of Health in Pomerania (SHIP)"; Int J Epidemiol, 51, e372–e383, 2022) was suspended in RNA extraction lysis buffer and homogenized using an MM 301 bead mill (Retsch). RNA was extracted using the RNeasy Fibrous Tissue Mini kit (Qiagen) according to the manufacturer's instructions.

[0056] Transcriptome libraries were prepared using the TruSeq Library Preparation Kit (Illumina) according to the manufacturer's instructions. Sequencing was performed on an Illumina HiSeq system in single-ended mode with a read length of 50 bp, up to a final sequencing throughput of 100 million reads per sample. The generated raw reads were processed as follows: (i) Quality control using FastQC 0.11.7 (Andews, Ss-andews / FastQC. GitHub https: / / github.com / s-andrews / FastQC), (ii) Trimmomatic 0.36 (Bolger, AM, Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics) (iii) The read sequences were trimmed according to 30,2114-2120 (2014)., with the following parameters for adapter trimming: use TruSeq3-SE adapter sequence, seed mismatch set to 2, palindromic clipping threshold set to 30, simple clipping threshold set to 10, parameters for sliding window set to window size 4, required quality set to 20, parameters for reads and trailing set to 20 as the minimum quality required to maintain the base, and the minimum required read length to retain after trimming set to 40 bp, and the reads were mapped to the human transcriptome of the GRCh38.106 build by Salmon 0.8.1 (Patro, R., Duggal, G., Love, MI, Irizarry, RA & Kingsford, C. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods). 14,417-419 (2017).) The parameters "bootstrap" were set to 0, "flsd" to 10, and "flmean" to 180, and (iv) the reads were quantified as transcripts per million (TPM) reads.

[0057] The gene expression data was converted to log10 by applying the following formula.

number

number

[0058] Predictors for wrinkle severity based on gene expression data were obtained by training a fully connected neural network using Keras (Chollet, F. keras-team / keras. GitHub https: / / github.com / keras-team / keras.) in R (R Core Team. R: A Language and Environment for Statistical Computing. The R Foundation (2018)) with a TensorFlow backend (Abadi, M. et al. TensorFlow: A system for large-scale machine learning. In 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI 16) 265-283 (The Advanced Computing Systems Association, 2016)). The neural network had the following configuration: (i) input layer (number of nodes equal to the number of features), (ii) first hidden layer (number of nodes is the square root of the first layer), (iii) second hidden layer (number of nodes is the square root of the second layer), and (iv) output layer (one node representing the wrinkle grade / target variable). The weights of the two hidden layers were initialized using the He function (He, K., Zhang, X., Ren, S. & Sun, J. Delving Deep into Rectifiers: Surpassing HumanLevel Performance on ImageNet Classification. Preprint at arXiv:1502.01852[cs](2015).), and the "elu" (exponential linear unit) activation function (Clevert, D.-A., Unterthiner, T. & Hochreiter, S. Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs). Preprint at arXiv:1511.07289[cs](2015).) was used for all hidden layers.The loss function for training was defined as the mean squared error. To improve the model's generalization ability, weight decay was implemented and the regularization coefficient (L2) was set to 0.01. Furthermore, the model was trained using stochastic gradient descent with Adam (Kingma, DP & Ba, J. Adam: a method for stochastic optimization. Preprint at arXiv:1412.6980[cs](2017)) and a learning rate of 0.001 with a batch size of 15 samples for training up to 100 epochs. All parameters of the neural network are summarized in Table 1. [Table 1]

[0059] The trained wrinkle grade predictors were validated using the test set and independent publicly available datasets (Holzscheck et al., "Multi-omics network analysis reveals distinct stages in the human aging progression in epidermal tissue"; AGING, 12; 2020).

[0060] Predicting each volunteer's wrinkle grade from an empirically determined test set of data, compared to the corresponding visual wrinkle grade, showed a very high correlation (R=0.752 and P<2.23×10⁻⁶). -16 A mean absolute prediction error of 9.49 wrinkle grade units was obtained (see Figure 5). Similarly, validation using an independent dataset showed a very high correlation (R=0.858 and P=P<2.2×10) between predicted wrinkle grade and visual wrinkle grade. -16 A Pearson correlation was observed, and the mean absolute error was 12.83 wrinkle grade units (see Figure 6).

[0061] Example 4: Next, Example 3 was repeated by reducing the determined gene expression levels of 437 female epidermal samples to the gene expression levels of various gene subsets used as input for model training via RNA-Seq.

[0062] The selection of target genes(s) was carried out as follows: Based on the model initially trained from Example 3, in silico knockout was performed on each gene to calculate the overall “effect size” of the gene on the predicted wrinkle grade. More specifically, the predicted wrinkle grade values ​​of the test set were compared to the predicted wrinkle grade values ​​based on the altered expression levels of the test set, where the expression of the target gene was set to 0, and this set is referred to herein as the “modified test” set. As used herein, the “effect size” of the gene is the difference between the predicted values ​​of the “normal” test set, i.e., the test set from Example 3, and the predicted values ​​of the “modified” test set, i.e., the test set from Example 3 with the expression level of the target gene set to 0. The effect size of the gene was then averaged among the subjects in this test set. Next, the genes are sorted in descending order by absolute mean effect size, and the top 75% of genes (i.e., 17,653 genes), 50% of genes (i.e., 11,769 genes), 25% of genes (i.e., 5,885 genes), and 5% of genes (i.e., 1,177 genes) are used to train the reduced-number-of-gene models. In some cases, the top 100 genes, 10 genes, and / or 2 genes are selected to train the models.

[0063] For training the wrinkle grade predictor, the parameters were modified as listed in Table 2, and gene expression levels, expressed as log10-converted TPM values ​​according to the method described in Example 3, were used as data input. [Table 2]

[0064] Predicting each volunteer's wrinkle grade from an empirically determined data set, compared to the corresponding visual wrinkle grade, showed a high and significant correlation (R>0.34 and P<2.7×10⁻⁶). -4 A mean absolute prediction error of 13.78 wrinkle grade units was obtained (see Figure 3). (Pearson correlation) [Table 3]

Claims

1. A method for determining the predicted wrinkle grade of human skin, which is: a) To provide human skin cells; b) Determining the methylation level of at least one CpG dinucleotide in a specific region of at least one chromosome of the skin cell, or determining the expression level of at least one gene in the skin cell; and c) A method comprising determining the predicted wrinkle grade of the skin cells by comparing the determined methylation level with empirically determined data representing a correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade of at least one human individual, or determining the predicted wrinkle grade of the skin cells by comparing the determined gene expression level with empirically determined data representing a correlation between the gene expression level and the visual wrinkle grade of at least one human individual.

2. The method according to claim 1, wherein the skin cells are collected from a human individual.

3. The method according to claim 1, wherein the skin cells are cultured in vitro.

4. The method according to any one of claims 1 to 3, wherein the methylation level in step b) is determined using a DNA methylation level detection method.

5. The method according to claim 4, wherein the method for detecting the DNA methylation level is selected from a) a base-resolution sequencing approach, b) a base-resolution microarray approach, c) a non-base-resolution sequencing approach, d) a non-base-resolution microarray approach, or e) a targeted approach.

6. The method according to any one of claims 1 to 3, wherein the expression level in step b) is determined using a gene expression level detection method.

7. The method according to claim 6, wherein the gene expression level detection method is selected from a) a sequencing-based high-throughput assay, b) a hybridization-based high-to-medium-throughput assay, c) a polymerase chain reaction (PCR)-based medium-to-low-throughput assay, or d) a hybridization-based medium-to-low-throughput assay.

8. The method according to claim 1, 2, or 4-7, further comprising the step of estimating the visual wrinkle grade of the individual human, wherein the predicted wrinkle grade is estimated to be within 20 wrinkle grade units of the visual wrinkle grade.

9. The method according to any one of claims 1 to 8, further comprising the step of bringing the skin cells into contact with an activator, wherein the predicted wrinkle grade is determined before or after the skin cells are brought into contact with the activator (optional).

10. A method for testing an activator comprising the method according to any one of claims 1 to 7, further comprising the following steps: d) bringing the skin cells of step a) into contact with the activator; e) Determining the predicted wrinkle grade of the skin cells in step d) according to the method described in any one of claims 1 to 7; f) A method comprising comparing the predicted wrinkle grade determined in steps a) to c) with the predicted wrinkle grade determined in step e).

11. The method according to claim 10, wherein step d) is performed either in vivo or in vitro, and / or the activator is a cosmetic agent and / or a therapeutic agent.

12. The method according to claim 10 or claim 11, for identifying an activator that prevents and / or treats signs of phenotypic aging of human skin.

13. A composition comprising one or more activators specified by the method described in any one of claims 10 to 12.

14. Use of DNA methylation levels or gene expression levels to determine the predicted wrinkle grade of human skin according to the method of any one of claims 1 to 7.

15. A computer-readable medium storing computer-executable instructions for causing a computer to perform a method for determining the predicted wrinkle grade of human skin: a) inputting at least one value of the determined methylation level of at least one CpG dinucleotide in a specific region of at least one chromosome of the skin cell, or inputting at least one value of the determined expression level of at least one gene of the skin cell; b) Comparing the determined methylation level with stored data representing a correlation between the methylation level of the CpG nucleotide and the visual wrinkle grade of at least one human individual, or comparing the gene expression level with stored data representing a correlation between the gene expression level and the visual wrinkle grade of at least one human individual; c) A computer-readable medium that includes displaying the predicted wrinkle grade.

16. A kit for determining the predicted wrinkle grade of human skin according to the method of any one of claims 1 to 7, comprising at least one oligonucleotide primer for amplifying and / or sequencing at least one CpG nucleotide of at least one nucleotide sequence, or at least one oligonucleotide primer for amplifying and / or sequencing a fraction of at least one exon of at least one nucleotide sequence.

Images