Effect of a Topical Collagen Tripeptide on Antiaging and Inhibition of Glycation of the Skin: A Pilot Study

The clinical study included 22 Asian women with noticeable periorbital and glabellar wrinkles. Participants' mean age was 47.1 years (range, 30–54 years). No participant dropped out because of noncompliance. In the final analysis after 4 weeks, data from 22 participants were collected and analyzed.

All participants applied the Ever Collagen Corrector Collagen Tripeptide Ampoule (CTP ampoule; NEWTREE Co., Ltd., Seoul, Korea), the commercialized cosmeceutical product used in the clinical study, twice daily onto their facial skin for 4 weeks. The total ingredient included in the test ampoule is listed in Table S3. After 4 weeks, the mean periorbital skin roughness (Ra) measured by PRIMOS was significantly reduced from 20.77 ± 3.51 μm to 19.24 ± 3.52 μm (Figure 1A, p < 0.001). The mean glabellar Ra was also significantly reduced from 21.40 ± 3.09 μm to 20.21 ± 2.83 μm (Figure 1B, p < 0.001). In addition, the maximum periorbital wrinkle roughness (Rmax) significantly decreased from 191.11 ± 32.58 μm to 182.01 ± 33.63 μm (Figure 1C, p < 0.001). The maximum glabellar Rmax also significantly decreased from 169.92 ± 21.57 μm to 159.76 ± 21.58 μm (Figure 1D, p < 0.001).

Meanwhile, clinical improvement of wrinkles was also visualized by a three-dimensional (3D) skin analysis camera system (Antera 3D^®, Miravex, Dublin, Ireland). The mean depressed depth of periorbital wrinkles at baseline was 0.06 ± 0.01 mm, which significantly improved to 0.05 ± 0.01 mm at week 4 (Figure 2A,B, p < 0.001).

Within 4 weeks, the skin density measured by Ultrascan UC22 (Courage + Khazaka electronic GmbH, Köhn, Germany) showed a significant increase from 55.66 ± 7.61 to 59.67 ± 7.84 (Figure 3A,B, p < 0.001). Furthermore, the skin surface elasticity (total recovery/total elongation; gross elasticity (R2)) increased from 0.81 ± 0.03 to 0.83 ± 0.03 (Figure 3C, * p < 0.001). The maximum collagen strength (R4) after application of the topical CTP also significantly increased from 68.02 ± 5.48 to 70.24 ± 5.14 (Figure 3D, p < 0.001). Meanwhile, AGE accumulation in the skin was 2.26 ± 0.32 AU at baseline, and it was significantly reduced to 2.16 ± 0.29 AU at week 4 (Figure 3E, p < 0.001). The summary of the overall clinical studies data outcomes is listed in Table S1.

Overall, the average percentage of subjects who responded to "very satisfied" and "satisfied" to the 4 week use of topical CTP product and the overall improvement in facial skin problems caused by aging at the end of the study was 95.5% ("very satisfied": four subjects, "satisfied": 17 subjects). One subject rated "slightly satisfied" with respect to the test product and the overall improvement of the facial skin. No adverse events were observed during the 4 week study period, and none of the participants dropped out of the study because of adverse events, suggesting that the test product formulation was safe to use.

As the cell viability of 1000 μg/mL of the hydrolyzed fish skin extract (EverCTP^TM; NEWTREE Co., Ltd., Seoul, Korea) was 96.88%, the investigators concluded that EverCTP^TM had no cell cytotoxicity with treatment up to 1000 μg/mL (Figure S1).

To investigate the effect of EverCTP^TM on cellular aging, we performed a quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) to measure collagen 1, matrix metallopeptidase (MMP)1, MMP3, and MMP9 messenger (m)RNA expression levels. UVR exposure was used in this experiment for the generation of extrinsic cellular aging on HDFs. Transforming growth factor-beta (TGF-β; Peprotech, Rocky Hill, NJ, USA) and retinoic acid (RA; Sigma Aldrich, Saint Louis, MO, USA) were used as positive controls for collagen 1 and MMPs, respectively. When compared to the control group, cells treated with 500 μg/mL and 1000 μg/mL of EverCTP^TM showed significantly higher expressions of collagen 1 (Figure 4A, p < 0.05 and p < 0.005, respectively). The expression levels of MMP-1, -3, and -9 reduced significantly after the treatment with 250 μg/mL, 500 μg/mL, and 1000 μg/mL of EverCTP^TM compared to the UVR exposure group. (Figure 4B–D; p < 0.05, p < 0.01, and p < 0.005, respectively).

The procollagen type I peptide (PIP1) concentration increased significantly when participants were treated with 250 μg/mL and 500 μg/mL of EverCTP^TM compared to the control (Figure 5A, p < 0.01 and p < 0.005, respectively). Treatment with 1000 μg/mL of EverCTP^TM also increased the PIP1 concentration (885.65 ± 89.69) compared to the control (824.65 ± 8.12), but this was not statistically significant. On the other hand, the MMP1 protein secretion after ultraviolet (UV) exposure showed a significant reduction when treated with 250 μg/mL, 500 μg/mL, and 1000 μg/mL of EverCTP^TM (Figure 5B, p < 0.005).

To measure the elastase activation that increases due to UV light and reactive oxygen species (ROS) during the aging process, an elastase inhibition assay was performed. Accordingly, 1 mM phosphoramidon (PR; Sigma Aldrich) was used as the positive control. Elastase inhibition activity was significantly increased after treatment with EverCTP^TM dose-dependently. Compared to the negative control, 250 μg/mL, 500 μg/mL, and 1000 μg/mL of EverCTP^TM showed 42.12%, 42.74%, and 48.47% increases in elastase inhibition activity, respectively (Figure 6, p < 0.005).

After treatment with 10 μM hydrogen peroxide (H₂O₂), the increased reactive oxygen species (ROS) level was significantly reduced when cotreated with 1000 μg/mL of EverCTP^TM (Figure 7, p < 0.05, p < 0.01, p < 0.005). Meanwhile, the effect of EverCTP^TM on the glycation process in human dermal fibroblast (HDF) cells was investigated by enzyme-linked immunosorbent assay (ELISA) after UV exposure. Treatment with 1000 μg/mL of EverCTP^TM significantly decreased the productions of pentosidine (from 211.682 ± 0.425 to 156.975 ± 0.135), methylglyoxal (from 5.586 ± 0.014 to 5.307 ± 0.001), and AGEs (from 16.688 ± 0.235 to 10.762 ± 0.424), while significantly increasing the production of glyoxalase I compared to the UV group (from 0.008 ± 0.001 to 0.777 ± 0.003; Figure 8, p < 0.005).

Additional investigations of the CTP ampoule were performed to assess its effect on the productions of glycated collagen and AGEs. The fluorescence intensity of glycated collagen production of the control group was 0.999 ± 0.058, while that of the CTP ampoule control group without collagen component (CTP control) was 0.09 ± 0.027 (Figure 9A, p < 0.005). Compared to the control group, the CTP ampoule group (with collagen) showed a significant decrease in glycated collagen production to 0.517 ± 0.114 (Figure 9A, p < 0.005). Moreover, the fluorescence intensity of AGE production of the control group was 2.476 ± 0.005, while that of CTP ampoule group was significantly decreased to 2.296 ± 0.002 (Figure 9B, p < 0.005). Lastly, we measured the degree of denatured collagen production induced by heat treatment to observe the effect of CTP ampoule cotreatment. The fluorescence intensity of denatured collagen production after heat treatment was 13.361 ± 2.532, and the degree of intensity significantly decreased to 8.676 ± 0.623 after treatment with the CTP ampoule (Figure 9C, p < 0.05).

The efficacy of collagen 1 absorption after topical application of gelatin (nonhydrolyzed collagen) and EverCTP^TM to the skin was evaluated in reconstructed human micro-tissue (KeraSkin-FT, Bio Solution Co., Ltd., Seoul, Korea). The fluorescence intensity of collagen 1 absorbed into the skin after 1000 μg/mL of EverCTP^TM treatment was 2.316 ± 0.444, which was increased compared to that after 1000 μg/mL of gelatin treatment (2.051 ± 0.550) and the negative control (0.266 ± 0.111; Figure 10A,B, p < 0.005). Despite an increase in the EverCTP^TM treatment group, the difference in florescent intensities compared to the gelatin group did not reach a statistical significance (p = 0.5520).

Additionally, the results of the additional quantitative evaluation on the rates of collagen absorption via high-performance liquid chromatography (HPLC) compared between the gelatin and EverCTP^TM treatments are shown in Figure 10C. The rates of collagen absorption in the gelatin treatment group and EverCTP^TM treatment group were 0.425 ± 0.100 and 12.221 ± 0.684, respectively, thus showing a significantly improved quantitative absorption rate of collagen in the EverCTP^TM group (p < 0.005).

This prospective, single-arm clinical trial was approved by the Institutional Review Board (IRB) of the Global Medical Research Center (IRB number: GIRB-21624-DX), and informed consent was obtained from each participant. The investigation was performed in full compliance with the principles of Good Clinical Practices and the Declaration of Helsinki.

All participants were instructed to apply the test product, the CTP ampoule, to their entire face twice daily for 4 weeks to evaluate its clinical efficacy and safety. To assess its safety, all participants were asked to report any adverse events they experienced while using the topical product.

Twenty-two healthy women aged between 30 and 50 years with noticeable periorbital and glabellar wrinkles were considered eligible for inclusion in the study. The main exclusion criteria were as follows: major internal disorders, major or active skin diseases, current pregnancy or lactation, history of allergy or hypersensitivity reactions, history of undergoing dermatologic procedures (e.g., lasers, filler, botulinum toxin injection) in the past 3 months, use of topical medications with any skin reactions (e.g., glucocorticoids, retinoids, and topical immunomodulators) in the past 3 months, use of systemic medications, such as glucocorticoids and immunomodulators within the last 1 month, and application of skin care products or topical medications with similar functions to topical collagen peptides within the last 3 months before participating in the study.

The CTP ampoule containing a hydrolyzed fish skin extract (EverCTP^TM) was used in the clinical study. EverCTP^TM is a form of collagen hydrolysate obtained from the skin of Pangasius hypophthalmus that contains 4% Gly–Pro–Hyp with the tripeptide content exceeding 25%; it was used as the test ingredient in our laboratory study. The preliminary human skin ex vivo penetration study with analysis of Franz diffusion cells showed 6.74% permeability of the Gly–Pro–Hyp tripeptide.

All participants were followed up at baseline and week 4 after the initiation of the treatment. The clinical efficacy of EverCTP^TM was investigated by assessing biophysical parameters determined by an optical 3D measurement system (PRIMOS CR, SnT Lab, Seoul, Korea), Cutometer Dual MPA580 (Courage Khazaka electronic GmbH, Köln, Germany), and Ultrascan UC22 (Courage + Khazaka electronic GmbH). The wrinkle volume and skin roughness of the periorbital area (crow's feet) and glabellar area were measured by the phase-shift rapid in-vivo measurement of skin (PRIMOS) at baseline and week 4. Among the skin surface descriptors of PRIMOS, Ra and Rmax were chosen to measure the 3D wrinkle changes during the study period. Antera 3D was used to perform a quantitative assessment of the depressions caused by the wrinkles.

Changes in skin elasticity were measured using the Cutometer MPA580. Among the various parameters of the Cutometer, we chose R2 and R4 to represent the efficacy of the test product in improvement of skin elasticity. The Ultrascan UC22 was used to measure changes in the density of the skin.

To noninvasively assess AGE accumulation in the skin, we used skin autofluorescence (skin AF) measured by the AGE Reader (DiagnOptics Technologies B.V., Groningen, The Netherlands). Skin AF, expressed as arbitrary units (AU), was measured in triplicate at the skin site on participants' inner forearm, which is a body part easily accessible and rarely exposed to the sun, as previously described [22].

All subjects underwent a physical examination to assess safety outcomes at every visit. Participants were also asked to report side-effects during the study period. Additionally, the participants were questioned to score their rates of satisfaction on the use of topical CTP product and the overall improvement in facial skin problems caused by aging using the following scale: 1 = unsatisfied, 2 = no change, 3 = slightly satisfied, 4 = satisfied, and 5 = very satisfied.

HDF cells purchased from ATCC^® (Manassas, VA, USA) were cultured in Dulbecco's modified Eagle's medium (Lonza, Walkersville, MD, USA) supplemented with 1% penicillin–streptomycin (Gibco, Grand Island, NY, USA) and 10% fetal bovine serum (Gibco). HDF cells were incubated in a humidified atmosphere containing 5% carbon dioxide (CO₂) at 37 °C. TGF-β, RA, and PR were used as appropriate positive controls. UV exposure and the treatment with H₂O₂ were performed for the generation of extrinsic cellular aging and the induction of ROS stress on HDFs, respectively.

To determine if EverCTP^TM was cytotoxic, a cell counting kit-8 (CCK-8; Dojindo, Mashiki, Japan) assay was used according to the manufacturer's instructions to evaluate HDF cells (ATCC^®). In brief, HDF cells (5 × 10⁴ cells/well) were seeded onto a 96-well plate. When cell confluence was >80%, the HDF cells were treated with EverCTP^TM and incubated for 24 h. The ELISA microplate reader (VersaMax; Molecular Devices, San Jose, CA, USA) was used to measure the absorbance at 450 nm.

Total RNA was extracted from the HDF cells treated with EverCTP^TM after no irradiation or 10 mJ/cm² of ultraviolet-B (UVB) irradiation using a UVB lamp (BLX26, BIO-LINK^®-Crosslinker, FRA). Using the TRIzol reagent (Invitrogen, Waltham, MA, USA) according to the manufacturer's instructions, total RNA extraction was performed. For quantification and reverse transcription of total RNA, the NanoDrop 2000 spectrophotometer (ThermoFisher Scientific, Carlsbad, CA, USA) and RNA to complementary DNA EcoDry Premix Kit (Takara Sake, Berkley, CA, USA) were used. qRT-qPCR was conducted using the SYBR Green Master MIX (4309155, Promega Co., Madison, WI, USA), specific primer pairs, and the QuantStudio 3 Real-Time Polymerase Chain Reaction System (Applied Biosystems, Waltham, MA, USA). Primer sequences used in the present study are shown in Table S2. Relative mRNA expression levels were normalized to GAPDH and calculated using the 2^−ΔΔCt method.

PIP and MMP1 protein concentrations were detected using ELISA kits (Takara, Kusatsu, JPN; Amersham Life Science, Amersham, UK). MMP1 production was induced by 10 mJ/cm² of UVB irradiation. The Human Pentosidine ELISA kit (MyBioSource, San Diego, CA, USA), Methylglyoxal ELISA kit (Abcam, Cambridge, UK), Glyoxalase I assay kit (Abcam), and AGE assay Kit (Abcam) were used to measure pentosidine, methylglyoxal, gyloxalase I, and overall AGE production. All kits were used according to the manufacturers' instructions. The absorbance detection was performed using an ELISA microplate reader (VersaMax) at 450 nm.

To measure elastase inhibition activity, an elastase solution extracted by HDF cells was prepared by freezing/thawing the cells more than three times. After the solution was centrifuged (3000 rpm) at 4 °C for 20 min, the supernatant was harvested and used as the elastase solution. The elastase solution was placed in 100 μg of protein per well on a 96-well plate, and we added 0.2 M Tris–HCl buffer (pH 8.0) to make the total volume 88 μL. We added 2 μL of N-succinyl-tri-alanyl-p-nitroanilide, a substrate of elastase, and 10 μL of EverCTP^TM of three concentrations or 10 μL of 1 mM PR (Sigma Aldrich, Saint Louis, MO, USA) to each well containing the elastase solution and cultured the plate at 37 °C for 90 min. We added 10 μL of 0.2 M Tris–HCl buffer as a control group. Finally, using an ELISA microplate reader (VersaMax), the absorbance was measured at 405 nm at 37 °C. The elastase inhibition activity was expressed as a decrease (%) in absorbance of the test group with and without the sample solution. Elastase inhibition activity (%) = (A − B)/A × 100, where A is the absorbance of control, and B is the absorbance of sample.

Intracellular ROS accumulation assessment was performed using the 2′7′-dichlorofluorescin diacetate (DCFDA) cellular ROS Detection Assay Kit (Abcam). Briefly, the HDF cells (5 × 10⁴ cells/dish) were seeded onto a glass bottom dish, and, when the cell confluence was >80%, we added 200 μL of 25 μM DCFDA to each dish and incubated the cells at 37 °C for 45 min. To induce ROS stress as the positive control, the treatment with 10 μM H₂O₂ was performed. For the test group, 10 μM H₂O₂ was added and incubated for 24 h prior to DCFDA. After diffusion into the cells, DCFDA was deacetylated and subsequently oxidized by ROS to form dichlorofluorescein, which is highly fluorescent, and it was assessed by confocal microscopy. After discarding the DCFDA solution, each dish was washed using phosphate-buffered saline (PBS). Fluorescence intensity was imaged and measured using a confocal microscope (LSM 700, Carl Zeiss, Jena, Germany), and compared between groups using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Glycated collagen production was estimated using the Collagen Glycation Assay Kit (COSMO BIO, Tokyo, Japan), which was used according to the manufacturer's instructions. In brief, 50 μL of a collagen solution was placed onto a 96-well black plate and incubated at 37 °C for 18 h. After it reacted for 18 h, the sample dilution buffer was added to the buffer group. In order to check the glycation reaction of the test product itself, the topical CTP ampoule without collagen was added to the CTP control group (without collagen). In the control group, fructose was added, while both fructose and the topical CTP ampoule (with collagen) were added to the CTP ampoule group. Using the microplate reader (VersaMax) at 370 nm and 440 nm, the absorbance was immediately detected after 4 weeks of reaction at 37 °C. Detected absorbance was used to compare the glycated collagen production between the groups.

In the assessment of AGE production, a mixture of bovine serum albumin (BSA) (5 mg/mL; Sigma-Aldrich, St. Louis, MO, USA), glucose (25 mM), and PBS or the CTP ampoule was placed in a 1:1:1 ratio and reacted at 37 °C for 2 weeks. PBS only was added to the buffer group, while BSA and glucose were added to the control group. The CTP ampoule was additionally treated along with BSA and glucose in CTP ampoule group. In order to verify the glycation reaction of the CTP ampoule itself, a control for the CTP ampoule excluding BSA was tested as the CTP control group.

The HDF cells (6 × 10⁴ cells/well) were seeded on four-well chamber slides and incubated with the CTP ampoule for 24 h when cell confluence reached > 80%. The slides were washed, treated with heated PBS (95 °C) for 1 min, and fixed with 4% paraformaldehyde. To induce denatured collagen production by HDFs as the positive control, heat stimulation was performed. After fixation, the slides were incubated with a denatured collagen detection reagent (Funakoshi, Tokyo, Japan) and incubated with FITC-conjugated streptavidin. Finally, the slides were fixed using the VECTASHIELD^® antifade mounting medium with DAPI (Vector Laboratories Inc., Burlingame, CA, USA) and evaluated using a fluorescence microscope (M2, Carl Zeiss). The fluorescence intensity was calculated using obtained fluorescence images and ImageJ software (National Institutes of Health).

The reconstructed human micro skin tissue model KeraSkin-FT was cultured in the KeraSkin-FT growth medium (Bio Solution Co., Ltd.) and incubated for 24 h in a humidified atmosphere containing 5% CO₂ at 37 °C after covering with 10 μL of 1000 μg/mL of EverCTP^TM or gelatin (non-hydrolyzed collagen) extracted from fish skin. Following fixation with 4% paraformaldehyde, the samples were made into 6 µm thick frozen sections. The frozen sections were incubated with primary anti-fish collagen 1 (Abcam) and a goat anti-rabbit immunoglobulin G H&L secondary antibody (Abcam). Then, the frozen sections were fixed using the VECTASHIELD^® mounting medium with DAPI and evaluated using a fluorescence microscope. The fluorescence intensity was calculated using obtained fluorescence images and ImageJ software (National Institutes of Health).

Additionally, after incubation, the medium containing collagen absorbed through the reconstructed human micro skin tissue model was analyzed to quantify hydroxyproline using HPLC. The increased absorption rates were calculated as previously described.

All data were expressed as the mean ± standard deviation. Datasets were assessed for normality using the Kolmogorov–Smirnov test. Clinical parameters at baseline and week 4 were compared between groups using the Wilcoxon signed-rank test or paired samples t-test according to the results from the normality test. The data were statistically analyzed using SPSS statistics version 25.0 software (IBM Corp., Armonk, NY, USA). Mean differences were considered significant when p < 0.05, p < 0.01, and p < 0.005. All laboratory experiments in the present study were performed at least three times (n ≥ 3); for statistical data analyses, the obtained data were used.