2025 Volume 48 Issue 10 Pages 1611-1620
The extracellular matrix (ECM) is a noncellular component in all tissues and organs. Collagen, a component of the ECM, is the most abundant protein in the body. The amount of collagen decreases with aging, leading to wrinkles and skin sagging. Accelerated aging caused by UV radiation is known as photoaging. Skin fibroblasts exposed to UVB have reduced total collagen content due to accelerated production of collagen-degrading enzymes, matrix metalloproteinases (MMPs), and suppression of collagen production. However, the effects of UVB on the ECM surrounding the cell and its interaction with the cell have not been well reported. This study sought to elucidate the effects of UVB-induced changes in the extracellular environment on cells. UVB-irradiated collagen exhibited irradiation-dependent degradation and accelerated carbonylation, resulting in qualitative changes. UVB-induced changes in collagen led to reduced cell-collagen interactions, such as adhesion, proliferation, and contraction. Moreover, UVB-irradiated collagen was more susceptible to collagen degradation by MMP-1. UVB-induced changes in collagen were analyzed using mass spectrometry (LC-MS/MS). Although alternation of post-translational modification was not detected in the cell-bound regions of collagen, multiple sites of carbonylative modification were detected on proline (Pro), arginine (Arg), and lysine (Lys). The numerous carbonylative modifications to Pro, Arg, and Lys in collagen may cause changes in the overall structure of collagen and affect cellular functions that are regulated by the interaction with collagen. Overall, our findings highlight a photoaging mechanism that focuses on the effects of ECM changes on cells.
The extracellular matrix (ECM) not only provides a physical scaffold for cells but also regulates various functions, such as cell adhesion, differentiation, migration, cell–cell interactions, and intracellular signaling.1) Collagen, a major component of the ECM, is the most abundant protein in mammals, accounting for approximately 30% of all proteins in the human body.1) Vertebrates have 28 types of collagen superfamilies. These collagens are either fibrous or nonfibrous and form a triple-helical structure composed of 3 α-chains.2,3) Fibrous type I collagen is a major component of skin, bone, tendons, and other tissues.3,4) In the skin ECM, collagen constitutes 70–80% of the dry weight, with type I collagen accounting for 80–90%.5) Collagen fibers are oriented in various directions and provide mechanical tension to the skin. Aging and UV light exposure damage collagen fibers and contribute to loss of skin elasticity and wrinkle formation.6) Normal collagen metabolism is maintained by the balance between collagen production by dermal fibroblasts and collagen degradation by matrix metalloproteinases (MMPs) secreted by cells. In aging skin, increased production of reactive oxygen species (ROS) and activation of the mitogen-activated protein kinase (MAPK) pathway and nuclear factor-κB (NF-κB) not only suppress collagen production by dermal fibroblasts but also increase MMP-1 production, resulting in an imbalance in collagen metabolism and reduction in collagen content.7) Long-term sunlight exposure accelerates collagen loss. Photoaging is the formation of deep wrinkles and sagging skin, and is accelerated by sun exposure. ROS generated by UVB induces MMP-1 production via the c-Jun pathway and reduces procollagen synthesis by inhibiting the transforming growth factor-β signaling pathway.5,8) Disruption of the balance between collagen production and degradation is considered a major cause of photoaging.
Sunlight is a composite light of various wavelengths. Of these, UV rays are most closely related to photoaging. UV light is divided into 3 bands according to wavelength: UVA (320–400 nm), UVB (290–320 nm), and UVC (100–290 nm). As most UVC is absorbed by the ozone layer, only UVA and UVB reach the surface of the Earth. UVB has a particularly strong influence on skin damage. In fact, excessive exposure to UVB is believed to play a central role in photoaging, leading to sunburns, hyperpigmentation, erythema, skin thickening, and deep wrinkles. As a result, sunscreens that block UVB are widely available. Sunscreens are not only used to prevent photoaging but also to help prevent skin cancer.9)
Numerous studies have investigated the mechanisms underlying photoaging caused by UV radiation, particularly its direct effects on cells. UVB not only directly damages nucleic acids, proteins, and lipids within cells but also causes indirect damage by generating ROS, which oxidize these molecules. The direct absorption of UVB by DNA leads to covalent bond formation and the production of cyclobutane pyrimidine dimers. ROS can also produce damaged nucleotides, such as 8-hydroxy-2′-deoxyguanosine, which further harm DNA.10) UVB directly causes cross-linking, fragmentation, and carbonylation of proteins.
Collagen has a slower turnover than most proteins in the body, and a half-life of approximately 15 years.11,12) Thus, collagen is one of the proteins most susceptible to aging and environmental stresses. The accumulation of post-translationally modified proteins, such as oxidation, carbonylation, carbamylation, glycation, and racemization, has been found in aged skin.13–17) Thus, skin aging is believed to progress not only due to the quantitative decrease in collagen and direct damage to cells from UV radiation but also qualitative changes in the ECM resulting from post-translational modifications. UVB-absorbing amino acids and UV chromophores (Cys, His, Phe, Trp, and Tyr) are associated with the sensitivity of the ECM components to UV irradiation.18,19) Despite previous investigations on the degradation of collagen by UVB,20) the effects of UV-induced degradation or modification of the ECM on cellular function remain to be thoroughly evaluated. In this study, we aimed to elucidate how UVB-induced oxidative modifications of type I collagen affect fibroblast behavior. To isolate the effects of UVB-specific damage, we employed a simplified model in which purified collagen scaffolds were oxidized by UVB irradiation alone, without the involvement of other ECM components or systemic factors present in photoaged skin. This approach allowed us to specifically evaluate the cellular responses to UVB-modified collagen, independent of confounding variables.
GL15SE (Sankyo Denki, Tokyo, Japan) was used as the broadband UVB source. The UVB irradiation intensity was measured using a UVX-Radiometer connected to a Radiometer Sensor (UVX-31; Analytik Jena GmbH, Jena, Germany). Bovine type I collagen (acid-extracted from dermis; Nippi, Tokyo, Japan) diluted to 1 mg/mL with distilled water was packed into a quartz cell and irradiated with UVB (0.8 mW/cm2) in the cold room set to 4°C.
Carbonylation AssayUVB-irradiated collagen (180 μg) was mixed with 10 mM 2,4-dinitrophenylhydrazine (DNPH) in 2 M HCl (100 μL) and incubated at 37°C for 1 h in the dark. After DNPH labeling, 3 times the volume of cold acetone was added to the solution, which was then allowed to stand at −30°C for 3 h. The solution was centrifuged at 14000 × g for 10 min, and the supernatant was removed. The precipitate was washed 3 times with 300 μL of cold acetone and redissolved in 100 μL of 6 M urea/50 mM Tris–HCl (pH 7.5). The carbonylation level was evaluated by measuring the absorbance at 375 nm using a microplate reader (2300 EnSpire; PerkinElmer Inc., Shelton, CT, U.S.A.). Protein carbonylation levels for each sample were calculated using the absorption coefficient (Ε = 22000 M−1 cm−1). The protein carbonyl content was expressed as nmol/mg protein.21)
Circular Dichroism (CD) SpectroscopyNormal and UVB-irradiated collagen were diluted to 0.1 mg/mL with purified water. Circular dichroism (CD) measurements were performed using a J-1500 spectropolarimeter (JASCO, Tokyo, Japan) with a quartz cuvette of 0.1 cm path length, scanning from 180 to 260 nm. The measurement parameters were as follows: bandwidth, 1 nm; scan speed, 100 nm/min; and each spectrum was averaged over 10 accumulations. Thermal denaturation curves were recorded by monitoring ellipticity at 220 nm while increasing the temperature from 20 to 60°C at a rate of 4°C/min.
Proliferation AssayPrimary human dermal fibroblasts (HDFs) derived from pooled foreskin donors were purchased from CELLnTEC (Bern, Switzerland) and cultured in DMEM supplemented with 10% (v/v) heat-inactivated fetal bovine serum at 37°C in a 5% CO2 atmosphere. Once confluent, the cells were detached with trypsin, and a cell suspension of 2.4 × 105 cells/mL was prepared using 5-fold concentrated (5×) DMEM. The cell suspension (1.2 mL) was mixed with collagen (2 mL), lactalbumin hydrolysate (0.6 mL), and distilled water (2.2 mL) on ice. The cell mixture was seeded at 250 μL/well in a 24-well plate and cultured for 2 h at 37°C with 5% CO2. Mixtures of normal and UVB collagen were prepared at percentages of 0, 25, 40, 50, and 60% and used for the embedding culture. After 24 h of incubation, 250 μL of CellTiter Blue (Promega, Madison, WI, U.S.A.) was added on top of the gel and incubated at 37°C for 2 h. The gel was washed 3 times with 500 μL of phosphate-buffered saline (PBS), and 500 μL of isopropanol was added to extract the CellTiter Blue. Finally, 150 μL of the extracted solution was transferred to a 96-well plate, and the fluorescence intensity of Ex/Em = 535/595 was measured using a microplate reader (Infinite F200; TECAN, Männedorf, Switzerland).
Contraction AssayHDFs were cultured in DMEM supplemented with 10% (v/v) heat-inactivated fetal bovine serum at 37°C in a 5% CO2 atmosphere. Once confluent, the cells were detached with trypsin, and a cell suspension of 2.4 × 105 cells/mL was prepared using 5× DMEM. The cell suspension (0.5 mL) was mixed with normal or UVB collagen (2 mL) on ice. The cell mixture was then seeded onto a 35-mm glass base dish and incubated at 37°C for 1 h. Upon confirmation of collagen gel solidification, DMEM (1 mL/dish) was added. The collagen gel was peeled from the bottom of the dish using a flame-sterilized microspatula. The area of the collagen gel was measured after culture at 37°C for 20 h.
Expression Analysis of ProMMP-1HDFs were cultured in DMEM supplemented with 10% (v/v) heat-inactivated fetal bovine serum at 37°C in a 5% CO2 atmosphere. Once confluent, the cells were detached with trypsin, seeded at 1.0 × 105 cells per well in 12-well plates, and incubated at 37°C for 19 h. Either normal collagen or UVB collagen was added to a final concentration of 100 μg/mL and incubated at 37°C for 24 h. IL-1α was added at a final concentration of 100 ng/mL as a positive control for MMP production. The culture supernatants were subjected to protein quantification, and 100 μg of protein was applied on a 12% polyacrylamide gel for sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). Separated proteins were transferred onto a 0.45-μm-pore polyvinylidene fluoride membrane (IPVH304FO; Millipore, Bedford, MA, U.S.A.). After a 30-min block in 5% (w/v) skim milk/PBS with Tween-20, the membranes were probed with primary and secondary antibodies diluted with 1% skim milk. Anti-proMMP-1 (1 : 2000, M4696, Lot SLCD5065; Sigma-Aldrich, St. Louis, MO, U.S.A.) served as the primary antibody, while horseradish peroxidase-conjugated anti-goat IgG (1 : 10000; GE Healthcare, Chicago, IL, U.S.A.) served as the secondary antibody. Immunoreactive bands were detected using ECL Prime Western blotting Detection Reagent (RPN2232; GE Healthcare). Blot images were acquired using an Image Analyzer LAS 3000 (GE Healthcare) and ImageJ software (National Institute of Health, Besthesda, MA, U.S.A.) was used for densitometric analysis.
Collagen Degradation by MMP-1Activated MMP-1 was prepared by incubating 13.2 ng of in-house purified human proMMP-1 with 1 mM 4-aminomethylmercury acetate for 4 h at 37°C.22) Normal or UVB collagen was mixed with activated MMP-1 in a 1 : 15 ratio and digested in 50 mM Tris–HCl (pH 7.4)/200 mM NaCl/5 mM CaCl2 at 37°C overnight. MMP-1-digested collagen was electrophoresed on a 7.5% polyacrylamide gel and silver-stained with EzStain Silver (ATTO, Tokyo, Japan). Images of the gels were acquired using a high-quality scanner. Finally, to determine the difference in MMP-1 specificity for normal and UVB collagen, the band intensity of non-digested collagen α(I) chains was measured using ImageJ.
Thermolysin Digestion of CollagenNormal and UVB collagen (10 μg) were diluted to 0.1 mg/mL using 50 mM Tris–HCl (pH 8.0)/0.5 mM CaCl2. Each collagen was heat-denatured at 80°C for 30 min, reduced with dithiothreitol (10 μg) at 37°C for 1 h, and alkylated with iodoacetamide (25 μg) at 37°C for 1 h in the dark. The proteins were digested with thermolysin (Promega), derived from Bacillus thermoproteolyticus Rokko, at 75°C for 3 h.
Trypsin Digestion of CollagenNormal and UVB collagen (10 μg) were diluted to 0.1 mg/mL using 50 mM Tris–HCl (pH 8.6). Each collagen was heat-denatured at 80°C for 30 min, reduced with dithiothreitol (10 μg) at 37°C for 1 h, and alkylated with iodoacetamide (25 μg) at 37°C for 1 h in the dark. The protein was digested with Sequencing Grade Modified Trypsin (Promega), derived from porcine pancreas, overnight at 37°C.
Identification of the Carbonylation Site in UVB CollagenThe thermolytic and tryptic digests were analyzed using a Q-Exactive mass spectrometer (Thermo Fisher Scientific, Waltham, MA, U.S.A.) connected to a nanoflow LC system (Easy nLC 1200; Thermo Fisher Scientific). The sample solutions were loaded onto a tipping column (C18 column, 0.15 mm id ×100 mm; 3 μm particle size; Nikkyo Technos, Tokyo, Japan) with a linear gradient of 5–60% acetonitrile in the presence of 0.1% formic acid at a flow rate of 200 nL/min. The spray voltage was 1.5 kV and the capillary temperature was 250°C. Full mass spectrometry (MS) scans were obtained using an Orbitrap analyzer (Thermo Fisher Scientific) at 70000 resolution and m/z 400 (m/z 150–2000). Higher energy collisional dissociation MS/MS scans of the top 10 most abundant ions were collected using an Orbitrap analyzer at a 17500 resolution, normalized collision energy of 27, and exclusion time of 15 s.
The raw data file was processed using Mascot Distiller software (ver. 2.6; Matrix Science, Chicago, IL, U.S.A.) and converted into a Mascot generic format (mgf) file. The mgf file was then leveraged to perform peptide searches using the sequence data of bovine type I collagen on the Mascot search engine (ver. 2.6.2; Matrix Science). The following search conditions were employed: enzyme, thermolysin or trypsin; maximum number of missed cleavages, 6 for thermolysin and 2 for trypsin; fixed modification, carbamidomethyl (C); variable modifications, proline (Pro) - > hydroxyproline, lysine (Lys) - > hydroxylysine, ammonia loss (N-term C [carbamidome-thyl]), arginine (Arg) - > GluSA, carbonyl (KR), Lys - > allysine, oxidation (M), Pro - > pyro-Glu; MS1 tolerance of 5 ppm, and MS2 tolerance of 0.05 Da. The resulting files (.dat) were exported from Mascot in CSV format and processed using Microsoft Excel. Peptides with a false discovery rate of 1% or less were selected.
Various post-translational modifications have been detected in photoaged skin, in addition to cross-linking and fragmentation of proteins. In this study, bovine collagen solution (1 mg/mL) was irradiated with UVB at 0, 25, 50, 75, and 100 J/cm2, and the collagen bands were analyzed using SDS-PAGE. Previously, the lifetime cumulative UV dose for Japanese individuals was calculated using 1 minimal erythema dose (MED) as 20 mJ/cm2. Based on the findings, individuals received 200 MED of UVB exposure per year.23) Over 20 years, this exposure is reported to be approximately 4000 MED, which is equivalent to 80 J/cm2. UVB irradiation was found to induce collagen fragmentation, with a dose-dependent increase in molecular bands below 100 kDa and a decrease in the band intensity of whole collagen (Fig. 1A). A carbonylation assay using DNPH also revealed dose-dependent increases in the amount of carbonylated proteins (Fig. 1B). DNPH is the most popular method used in carbonylated protein evaluation. DNPH reacts with carbonyl groups to form stable 2,4-dinitrophenylhydrazones, which can be detected and quantified spectrophotometrically as they exhibit a maximum absorption wavelength between 365 and 375 nm.24,25) Collagen irradiated with 75 J/cm2 UVB exhibited sufficient whole-collagen bands and significant carbonylation. Therefore, collagen irradiated with 75 J/cm2 of UVB could be defined as UVB collagen and was used in the subsequent experiments. To assess whether UVB irradiation affected the secondary structure of collagen, CD spectra were recorded (Fig. 2). The spectra of UVB collagen were nearly identical to those of normal collagen, with only a slight decrease in ellipticity observed at approximately 190 nm, suggesting that UVB irradiation did not cause substantial changes to the overall secondary structure, including the characteristic triple-helix conformation. Furthermore, to evaluate the thermal stability of collagen, temperature-dependent CD measurements were performed at 220 nm while increasing the temperature from 20 to 60°C at a rate of 4°C/min. In normal collagen, a decrease in ellipticity began at approximately 42°C, indicating the onset of denaturation. In contrast, UVB-irradiated collagen showed a similar decrease starting at approximately 38°C. These results suggest that UVB irradiation slightly reduced the thermal stability of collagen.
UVB-irradiated collagen was analyzed using SDS-PAGE. Lane 1: protein marker; lanes 2–5: 0, 25, 50, 75, and 100 J/cm2 UVB-irradiated collagen, respectively (A). Collagen solution (1 mg/mL) was irradiated at 25–100 J/cm2. The UVB-irradiated collagen solutions were assayed for carbonylation levels using DNPH. Data are indicated as mean ± S.D. of 3 samples. *p < 0.05 and ***p < 0.01, significantly different from the normal collagen solution (B).
CD spectra of normal and UVB-irradiated collagen measured at 25°C. The 2 spectra showed similar profiles, with a slight decrease in ellipticity at approximately 190 nm in the UVB-treated sample (A). Thermal denaturation profiles were monitored at 220 nm from 20 to 60°C (4°C/min). Denaturation began at approximately 42°C for normal collagen and approximately 38°C for UVB-irradiated collagen (B).
To determine the effect of UVB collagen on HDFs, cells were seeded on collagen gels prepared using normal or UVB collagen. Cells seeded on normal collagen gels adhered as expected; however, cells seeded on gels made from UVB collagen remained rounded and accumulated at the center of the well (Fig. 3A). Normally, HDFs adhering to the collagen matrix spread over time and the cells adopt an elongated morphology.26) However, HDFs seeded on UVB collagen gels did not show any morphological changes. Such a finding suggests reduced cell adhesion on UVB collagen. Therefore, we performed an embedded culture to analyze the interaction between cells and collagen. In the contraction assay, HDFs embedded in normal collagen significantly contracted the collagen gel, whereas those embedded in UVB collagen displayed minimal contraction (Fig. 3B). The collagen contraction assay is a visual method for evaluating the mechanical interaction between ECM and cells.27) The ECM provides mechanical support and adhesion sites for cells. Therefore, reduced contraction of UVB collagen indicates weakened cell adhesion and mechanical strength. As the proportion of UVB collagen increased, the collagen gels became more fragile. A mixture of normal and UVB collagen was therefore used for the embedded culture. Based on proliferative activity assessment, cell proliferation decreased when the percentage of UVB collagen reached 60% (Fig. 3C). These results suggest that a higher percentage of UVB collagen reduces the scaffold function.
HDFs were seeded onto normal or UVB collagen gels and incubated at 37°C for 24 h. The cell adhesion ability of HDFs was weakened on UVB collagen gels (A). HDFs were embedded in normal or UVB collagen gels, and the gel contraction activities were compared. The contraction of fibroblasts embedded in normal (left) and UVB (right) collagen gels was found to differ. The contracted collagen gel area was measured; the data are presented as mean ± S.D. of 3 samples. ***p < 0.01, significantly different from cells embedded in normal collagen gel (B). HDFs were embedded in collagen gels, and their cell proliferative activity was compared. Normal and UVB collagen solutions were mixed at ratios of 1 : 3 (UVB collagen 25%), 1 : 1.25 (40%), 1 : 1 (50%), and 1.25 : 1 (60%). Cell proliferation was evaluated using CellTiter-Blue. Data are expressed as mean ± S.D. of 3 samples. *p < 0.05, significantly different from cells embedded in 0% UVB collagen gel (C).
MMP-1 is a collagen-degrading enzyme secreted by HDFs.7) When normal or UVB collagen was added to HDFs, the protein expression of proMMP-1 was promoted in both types of collagen. However, proMMP-1 expression did not differ between normal and UVB collagen (Fig. 4A). Therefore, UVB collagen is suggested to be recognized by cells in the same manner as normal collagen. In contrast, decreased collagen content and the accumulation of post-translationally modified proteins have been reported in photoaged skin.28,29) Accordingly, we evaluated the degradation of normal and UVB collagen by activated MMP-1. As shown in Fig. 4B, the extent of degradation by activated MMP-1 was greater for UVB collagen than for normal collagen. These results suggest that carbonylation partially denatures the triple-helical structure of collagen, enhancing its accessibility to MMP-1.
ProMMP-1 production by HDFs upon the addition of normal or UVB collagen solutions. HDFs were treated with normal collagen, UVB collagen solution, or IL-1α (10 ng/mL) for 24 h. Culture supernatants were subjected to Western blotting. Relative proMMP-1 expression was semi-quantified using densitometry scanning; data are presented as mean ± S.D. of triplicate experiments. *p < 0.05 and ***p < 0.01, significantly different from untreated cells (control) (A). Normal or UVB collagen was degraded with active MMP-1 to evaluate substrate specificity. ProMMP-1 was activated with 1 mM APMA, and each collagen was digested overnight at 37°C. MMP-1-degraded collagen was analyzed using SDS-PAGE. Collagen α(I) chain band intensities were measured using ImageJ; data are expressed as mean ± S.D. of triplicate experiments. *p < 0.05 and ***p < 0.01, significantly different from normal collagen (B).
Carbonylation sites in normal and UVB collagen were identified by MS. According to previous studies, oxidative modifications are known to occur on Pro, Lys, and Arg residues, each associated with characteristic mass shifts.30) Therefore, each collagen sample was digested with trypsin, and the resulting peptides were analyzed by high-resolution LC-MS/MS to search for carbonylation at the 3 amino acids. For Pro, a +15.99 Da mass shift typically results from hydroxylation to hydroxyproline, but it may also indicate oxidative deamination forming glutamic semialdehyde. Because both modifications have the same mass, they cannot be distinguished even by MS/MS analysis. For Lys, carbonylation can result in the formation of aminoadipic semialdehyde, leading to a mass shift of +1.99 Da (loss of NH2, gain of = O); however, in another case, +14.01 Da shifts have been reported. For Arg, carbonylation may yield glutamic semialdehyde via side-chain cleavage, typically associated with a mass shift of +14.01 Da. In our data, +15.99 Da modifications were predominantly observed on Pro-containing peptides, while other oxidative mass shifts were detected on Lys and Arg residues, as shown in Table 1. As shown in Fig. 3, collagen carbonylation significantly reduced collagen–cell interactions. To investigate carbonylation in regions involved in cell adhesion, we digested collagen with thermolysin, which produces peptides suitable for analyzing sequences such as GFOGER and RGD. We first analyzed carbonylation near the GFOGER sequence, which has been reported to play a critical role in cell adhesion via integrin α2β1.31) However, no carbonylated sites were detected in this region. The RGD sequence is also involved in binding to ECM components, such as fibronectin and laminin. Therefore, we further analyzed peptides containing the RGD sequences present in collagen to examine possible carbonylation. However, no carbonylated sites were detected in these peptides. On the other hand, tracing the mass of the thermolytic peptide AGPQGPRGDKGETGEQGDRG (including the RGD sequence at positions 1092–1094) revealed 2 distinct peaks in both normal and UVB-treated collagen samples (Fig. 5A). MS/MS analysis confirmed that these peaks corresponded to peptides with identical sequences (Fig. 5B). This finding suggests that these peptides possess different physicochemical properties. MS/MS spectra showed no modifications from the C-terminus up to y10, and from y15 to the N-terminus, demonstrating that one of the amino acids within the PRGDK sequence is oxidatively modified. The difference in retention times may have resulted from variation in the site of oxidative modification. In addition, previous studies have reported that peptides containing isomeric forms of Asp or Glu exhibit the same MS/MS fragmentation patterns but elute at different times.32) Therefore, it is also possible that the 2 observed peaks represent isomeric forms of the same peptide. In any case, a modification has clearly occurred in the immediate vicinity of the RGD sequence, altering the physicochemical properties of the peptide. Such a modification is likely to contribute to the reduced interaction between collagen and cells observed in UVB-treated samples.
| Amino acid | Modification | Structure | Mass difference (Da) | Normal | UVB |
|---|---|---|---|---|---|
Proline |
Hydroxyproline |  |
+15.99 | 157 | 187 |
| Glutamic semialdehyde |  |
+15.99 | |||
| Pyro glutamic acid |  |
+13.98 | 1 | 15 | |
Lysine |
Aminoadipic semialdehyde |  |
−1.03 | 1 | 4 |
| Carbonyl |  |
+13.98 | 1 | 3 | |
Arginine |
Carbonyl |  |
+13.98 | 2 | 1 |
| Glutamic semialdehyde |  |
−43.05 | 1 | 8 |
The post-translational modifications were analyzed using Mascot Distiller software, and the number of amino acid residues that underwent modification was counted.
Collagen was digested with thermolysin and analyzed using LC-MS/MS. Peptides with the sequence, AGPQGPRGDKGETGEQGDRG, had different retention times in normal and UVB collagen.
Photoaging is an aesthetic phenomenon characterized by deep wrinkles resulting from chronic sun exposure.33,34) The formation of deep wrinkles due to photoaging is believed to be caused by a decrease in collagen content. UV radiation not only inhibits collagen production in skin fibroblasts via the activation of the MAPK pathway and NF-κB but also promotes MMP-1 production, leading to collagen degradation.7) Thus, photoaging may progress by altering the metabolic balance of collagen in skin fibroblasts. It is also known that various post-translational modifications, such as cross-linking, degradation, glycation, oxidation, and racemization, accumulate in the proteins of photoaged skin. Therefore, photoaging can be considered a complex phenomenon that not only affects cells but also various surrounding molecules.
In photoaging research, direct UVB effects on cellular DNA and signaling have been widely studied. However, it remains unclear how UVB-induced alterations to the ECM itself affect cell behavior. To address this, we employed a simplified in vitro model using purified collagen, excluding confounding cellular responses. This allowed us to examine the isolated effects of UVB-induced ECM modifications on fibroblast function, providing insight into indirect but critical pathways of photoaging.
First, by evaluating the effects of UVB collagen on cells, we found that UVB collagen decreased cell adhesion, proliferation, and contraction. UVB irradiation promoted collagen fragmentation and carbonylation. In particular, the contraction assay revealed significant differences in the interaction between HDFs and collagen, indicating that changes in collagen significantly reduce its function as a cell scaffold. According to previous studies, cell adhesion is reduced in collagen that is enzymatically degraded by 80%, indicating that collagen fragmentation is associated with cell adhesion.35) In contrast, the UVB collagen used in this study was partially degraded; however, 77% remained as protein based on the BCA assay (data not shown). Therefore, UVB collagen-induced cellular dysfunction may be primarily due to carbonylation of collagen. Although various ECM quality changes are known to occur in photoaged skin, the effects of ECM changes on cells have not been reported in detail. This study highlights the importance of the quality of the extracellular environment for cellular homeostasis.
We proceeded to evaluate the metabolism of UVB collagen. Notably, UVB collagen does not promote MMP-1 expression but is actively degraded by MMP-1. Following the addition of collagen to cultured cells, normal and UVB collagen enhanced MMP-1 production in HDFs. However, no significant difference in expression was found between the groups. In contrast, collagenolysis by activated MMP-1 was enhanced in UVB collagen compared to normal collagen. Therefore, the rate of collagen degradation increased with photoaging. In addition to reduced cell proliferation due to scaffold damage, the synergistic effect of accelerated collagen degradation by MMP-1 may contribute to an overall decrease in dermal collagen content and the formation of deep wrinkles.
To further investigate the qualitative changes in collagen, we identified the carbonylation site in collagen induced by UVB. Notably, the amount of carbonylation sites was found to increase at the Pro, Arg, and Lys residues. Metreveli et al.36) demonstrated that the exposure of collagen solutions to UV (100–400 nm) from a mercury-quartz lamp generated free radicals on Pro residues. When collagen is irradiated with UV, the aromatic amino acids in collagen (especially phenylalanine and tyrosine) are photoionized, releasing free electrons. The released free electrons combine with protons in the solution to produce hydrogen radicals. These hydrogen radicals interact with the pyrrolidine ring of Pro to extract the α-hydrogen from Pro. In the present study, carbonylation sites were particularly abundant in Pro residues, aligning with previous findings.
In addition to carbonylation, collagen in vivo undergoes numerous modifications, such as glycation, carbamylation, and calcification, with aging.13–17) Panwar et al.22) revealed that collagen modified by glycation or mineralization is resistant to cleavage by MMPs. In our study, carbonylated collagen showed increased degradation by activated MMP-1, demonstrating an opposing outcome. Advanced glycation end-products (AGEs) formation is a structural modification reaction involving sugar linkages, resulting in markedly larger molecules than carbonylation.24) Therefore, the presence of such large modification groups may limit the access of MMPs to collagen, thereby contributing to its reduced degradation.
UVB irradiation alters the thermal stability and degrades the triple-helical structure of collagen.37) However, our CD spectroscopy analysis showed that UVB-irradiated collagen maintains a secondary structure profile that is nearly identical to that of non-irradiated collagen, indicating that the overall triple-helical conformation remains intact.
Nonetheless, temperature-dependent CD analysis revealed that the thermal denaturation of UVB collagen begins at a lower temperature than that of normal collagen, suggesting a slight decrease in thermal stability. In addition, quantitative assays confirmed a significant increase in carbonylation levels following UVB exposure. These results indicate that, although the gross secondary structure is preserved, UVB irradiation induces widespread, microscopic chemical modifications at the amino acid level.
Notably, site-specific analysis of thermolysin-digested peptides suggested the occurrence of several modifications, including oxidative modifications, within or near the RGD sequence, a critical motif for cell adhesion. These modifications appear to alter the physicochemical properties of the peptide, as evidenced by distinct chromatographic behavior despite identical MS/MS spectra. While the exact nature of the modification could not be determined, oxidation and/or isomerization events near the RGD motif are likely responsible for disrupting collagen–cell interactions.
In conclusion, UVB-induced carbonylation does not result in large-scale structural collapse of collagen but instead induces multiple, localized chemical changes. These subtle yet widespread modifications alter collagen’s thermal stability and potentially impair its biological functions, such as cell adhesion and matrix remodeling.
These findings highlight the broader significance of UV effects beyond cellular damage.
Previous studies on photoaging mainly focused on the physiological effects of UV on cells; however, the present study highlights the importance of UV effects on the extracellular environment. In photoaging skin, a negative cycle of aging may be triggered by the reciprocal mechanisms of ECM degradation via cell-derived enzymes and cellular dysfunction caused by qualitative changes in the ECM due to UV light. This study focused on biophysical phenotypes, such as cell adhesion and contraction, as the manifestation of the effects of UVB collagen on HDFs. We showed that UVB exposure alters skin ECM and affects cellular functions. In the future, exploration of the physiological effects and molecular mechanisms of UVB collagen-cultured HDFs is expected to aid further understanding of the mechanisms of photoaging.
This work was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI, Grant Number JP20K19719.
Conceptualization: HS and TS; experiment execution: AA, TK, and MO; statistical analysis: AA, TK, and MO; methodology: HS; project administration: HS and TS; writing: AA and HS; writing—review and editing: HS, KM, and TS. All authors read and approved the final manuscript.
The authors declare no conflict of interest.
This article contains supplementary materials.
