UV-Crosslinked Collagen and Gelatin Sheets as Cell Scaffold Materials: Nanoscale Surface Properties and Cell Proliferation Performance

Abstract

Collagen and gelatin are promising cell scaffold materials, but their structural instability under physiological conditions necessitates crosslinking treatment. This study evaluated UV crosslinking (254 nm, 0–180 min) as a non-toxic alternative to chemical crosslinking for collagen sheets (CS) and gelatin sheets (GS). Physicochemical properties were characterized by gel fraction analysis, atomic force microscopy (AFM) in PBS, and Fourier transform infrared spectroscopy (FT-IR), while NIH-3T3 fibroblast proliferation was evaluated by CCK-8 assay. UV crosslinking dramatically improved GS water resistance (gel fraction increased from 22% to 80% at 60 min) while maintaining smooth nanoscale surfaces (R_q: 1–4 nm), whereas CS exhibited inherent high stability (90% gel fraction without UV treatment, reaching 95–98% after irradiation). Both materials achieved maximum elastic modulus at 60 min (CS: 2.0 MPa; GS: 1.5 MPa). UV irradiation significantly enhanced cell proliferation on both substrates compared to untreated controls (p < 0.05). CS showed consistently high proliferation across all UV-treated conditions (day 3 absorbance: ~2.5–2.7), while GS exhibited progressive increases reaching a maximum at 180 min (absorbance: ~2.9). The continued GS enhancement despite slightly decreased elastic modulus suggests that chemical factors, possibly related to RGD motif accessibility, contribute beyond mechanical optimization. UV crosslinking effectively establishes structural stability essential for cell scaffold function, with both materials representing effective, biocompatible scaffolds for tissue engineering applications.

1. Introduction

In the fields of regenerative medicine and tissue engineering, the development of cell scaffold materials is critically important for achieving tissue and organ repair and regeneration [1,2]. Ideal cell scaffold materials must satisfy diverse requirements, including cell adhesiveness, biocompatibility, appropriate mechanical strength, and controllable biodegradability. As materials possessing these properties, naturally derived biopolymers, particularly collagen and its thermally denatured product gelatin, have been extensively studied as cell scaffold materials due to their excellent biocompatibility and cell affinity [3,4].

Collagen is a major component of the extracellular matrix in mammals and is a fibrous protein with a triple helix structure [5,6]. Within its molecular structure, various cell recognition sites exist, including the RGD (Arg-Gly-Asp) sequence that promotes cell adhesion, providing excellent biological functions that support cell adhesion, proliferation, and differentiation [7,8]. In contrast, gelatin is a natural polymer obtained by thermal denaturation of collagen, in which the triple helix structure of collagen dissociates into a random coil conformation [4,9]. Gelatin has advantages over collagen in terms of higher solubility and superior processability. Furthermore, it has been reported that thermal denaturation disrupts the higher-order structure of collagen, exposing cell adhesion sites, such as RGD sequences, that were previously buried internally, thereby exhibiting higher cell affinity [10,11].

However, the major challenge in the practical application of collagen and gelatin as cell scaffold materials is their insufficient structural stability in aqueous solutions. Particularly under physiological conditions (37 °C, pH 7.4), these materials readily dissolve or swell, making it difficult to maintain their structure over the period required for cell culture. To resolve this issue, it is essential to introduce crosslinking structures between molecules to enhance the water resistance and mechanical strength of the materials [12].

Various crosslinking methods for collagen and gelatin have been developed, including chemical crosslinking, physical crosslinking, and enzymatic crosslinking [13,14]. Chemical crosslinking agents such as glutaraldehyde, carbodiimide, and genipin are widely used [15,16,17]. These crosslinking agents can efficiently form intermolecular crosslinks and substantially improve the mechanical strength of materials. However, many chemical crosslinking agents possess cytotoxicity [18,19], and if unreacted crosslinking agents remain, they may adversely affect cell viability and function. Additionally, the requirement for crosslinking agent removal processes complicates the manufacturing process and raises cost concerns.

Physical crosslinking methods include dehydrothermal treatment (DHT) and UV irradiation [20,21]. UV irradiation crosslinking has the advantages of requiring no crosslinking agents and enabling non-contact processing. UV irradiation excites aromatic amino acid residues such as tyrosine and tryptophan within protein molecules, inducing intermolecular covalent bond formation [22,23]. Although dityrosine bond formation is considered a primary crosslinking mechanism, it should be acknowledged that native collagen and its derivative gelatin contain very low levels of tryptophan, which may limit the contribution of tryptophan-specific crosslinks [24]. Furthermore, endogenous photosensitizers such as riboflavin, if present in the preparation, may enhance or modify the photochemical reaction pathways [25]. While UV crosslinking is considered superior from a biocompatibility perspective compared to chemical crosslinking agents, the precise molecular mechanisms require further investigation using techniques such as mass spectrometry.

However, previous studies on UV crosslinking methods have primarily focused on macroscopic property evaluation (gel fraction, tensile testing, etc.), with insufficient investigation of nanoscale surface morphology and local elastic modulus changes. Cells adhere to scaffold material surfaces and are known to sense and respond to surface properties (roughness, stiffness, chemical composition, etc.) [26,27]. Therefore, quantitative evaluation of surface properties at the nanoscale is essential in cell scaffold material design. In particular, in situ observation in liquid using AFM can evaluate surface morphology and mechanical properties under conditions close to actual cell culture environments, providing important information for understanding cell–material interactions [28,29].

In UV irradiation crosslinking, parameters such as irradiation wavelength, irradiation time, and irradiation distance significantly influence the properties of the resulting materials [30]. Among these, the UV wavelength is a particularly critical factor. Short-wavelength UV (200–280 nm) carries high photon energy and can efficiently induce protein crosslinking reactions. In this study, we employed 254 nm UV irradiation, a wavelength widely used in germicidal lamps. This wavelength falls within the absorption range of aromatic amino acid residues, especially tyrosine and tryptophan, and is therefore expected to facilitate efficient crosslinking [31]. It should be noted, however, that 254 nm UV irradiation does not exclusively induce crosslinking; competing photochemical events—including oxidation of tyrosine, phenylalanine, and cysteine residues, as well as backbone cleavage and partial denaturation—may occur simultaneously [32]. Disentangling specific crosslinking pathways from these competing modifications requires analytical techniques beyond FT-IR, such as mass spectrometry.

Furthermore, excessively prolonged UV irradiation can promote not only crosslinking but also main-chain scission and protein degradation. Sionkowska et al. reported that extended UV exposure damages the molecular structure of collagen, leading to reductions in denaturation temperature and mechanical properties [33]. These considerations highlight the importance of optimizing irradiation conditions to achieve an appropriate balance between crosslink formation and structural preservation of the material.

As factors determining cell scaffold material performance, physical properties such as surface roughness and elastic modulus have traditionally been emphasized. However, by comparing materials with different molecular structures but the same origin, such as collagen and gelatin, we can answer the fundamental question of whether chemical properties (exposure of cell adhesion sites) or physical properties dominate cell responses. Furthermore, to ensure the generalizability of our findings, we employed NIH-3T3 cells, a well-established mouse fibroblast cell line widely utilized in biomaterial research, for cell proliferation evaluation. The purpose of this study was to systematically evaluate the effects of introducing crosslinking structures into gelatin sheets and collagen sheets using short-wavelength UV irradiation (254 nm) on their physicochemical properties and cell affinity. Specifically, by varying UV irradiation time from 0 to 180 min, we systematically evaluated: (1) crosslinking degree by gel fraction analysis through dissolution testing; (2) microscale surface morphology by scanning electron microscopy (SEM) in the dried state; (3) nanoscale surface topography and local elastic modulus by AFM in physiological buffer; (4) chemical structural changes using FT-IR; (5) cell proliferation capacity using NIH-3T3 mouse fibroblasts by CCK-8 assay; and (6) cell adhesion and spreading morphology by bright-field microscopy. This multi-scale characterization approach enabled comprehensive evaluation of the structure–property–function relationships in UV-crosslinked protein scaffolds.

This study is expected to contribute to the rational design of biomaterials for regenerative medicine by elucidating the mechanisms of crosslinking structure formation by UV crosslinking from multiple perspectives and clarifying the relative importance of chemical and physical factors in cell proliferation. Additionally, by comparing two types of materials with different structures—collagen and gelatin—new insights into the correlation between molecular structure and cell affinity are anticipated.

2. Results

2.1. Crosslinking Structure Formation and Water Resistance by UV Irradiation

To evaluate the degree of crosslinking structure formation introduced by UV irradiation, gel fraction measurements by dissolution testing were performed (Figure 1). The gel fraction of gelatin sheets (GS) was approximately 22% in the UV-unirradiated (0 min) state, confirming that the majority dissolved upon PBS immersion. This indicates that uncrosslinked gelatin molecules retain water solubility and cannot maintain their structure under physiological conditions. With increasing UV irradiation time, the gel fraction increased markedly, reaching approximately 74% at 30 min irradiation and approximately 80% at 60 min irradiation. Even when irradiation time was extended to 180 min, the gel fraction reached a near-equilibrium state at approximately 82%, suggesting that crosslinking reaction progression saturates after 60 min.

In contrast, collagen sheets (CS) exhibited a high gel fraction of approximately 90%, even in the UV-unirradiated state. This reflects that collagen possesses a certain degree of insolubility even without crosslinking treatment due to intermolecular interactions derived from its triple helix structure [5]. After UV irradiation, the gel fraction of CS was maintained in the range of approximately 95–98% at all irradiation times, with smaller changes due to irradiation time compared to GS. This result suggests that CS originally possesses high structural stability, and the additional effect of UV crosslinking is relatively small.

2.2. Preservation of Surface Microstructure in the Dried State

The results of surface morphology observation in the dried state by SEM are shown in Figure 2. For both GS and CS samples, no significant changes in surface morphology were observed before and after UV irradiation. All samples exhibited relatively smooth surfaces, and no clear fibrous structures or porous structures were observed. This result indicates that 254 nm short-wavelength UV irradiation primarily induces covalent bond formation between protein molecules but does not significantly affect surface morphology or structural orientation at the micrometer scale.

The surface morphology observed by SEM is considered to depend mainly on drying conditions during sample preparation. In this study, the adoption of slow drying conditions at 10 °C and 40% relative humidity is inferred to have resulted in relatively uniform orientation of protein molecules and formation of smooth surfaces. Introduction of crosslinking structures by UV irradiation involves bond formation at the molecular level and does not reach the point of altering the macroscopic surface structure already formed. This finding has important implications for cell scaffold material design, in that UV crosslinking can improve water resistance while maintaining surface topography.

2.3. Quantitative Evaluation of Nanoscale Surface Morphology and Roughness in Liquid

To evaluate surface properties in the environment where cells actually contact—that is, in aqueous solution—AFM measurements were performed in a state of being immersed in PBS (Figure 3). The surface of GS was extremely smooth, with surface roughness (R_q) maintained in the range of 1–4 nm under all UV irradiation conditions. From AFM height images, uniform surfaces at the nanoscale were confirmed, and no clear morphological changes due to UV irradiation were observed. These results demonstrate that UV crosslinking primarily induces chemical changes (formation of intermolecular covalent bonds) without producing topographical changes detectable by AFM. Previous studies have shown that UV-induced crosslinking and photochemical modifications can alter mechanical, chemical, or surface-energy properties while maintaining surface topography at the nanoscale [34,35], and our findings are consistent with this understanding.

In contrast, CS showed markedly different surface morphology changes. In the UV-unirradiated (0 min) state, R_q showed a high value of approximately 47 nm, and swollen fiber bundle-like structures were observed in AFM images. This is considered to result from the triple helix structure of collagen swelling in liquid, with water molecules penetrating into spaces between molecular chains, forming surfaces with large irregularities. At 30 min and 60 min UV irradiation, R_q decreased to approximately 26 nm and approximately 21 nm, respectively, and surface smoothing progressed. In AFM images, it was confirmed that swollen fiber bundles contracted and changed to finer network structures. This smoothing can be interpreted as a result of suppression of swelling due to the restriction of collagen molecular chain mobility by intermolecular crosslinking formation through UV irradiation.

However, at 180 min UV irradiation, R_q rose again to approximately 35 nm, and surface non-uniformity increased. From AFM images, localized aggregated structures were observed, suggesting that excessive crosslinking reactions induced non-uniform aggregation of molecular chains. According to reports by Sionkowska et al., prolonged UV irradiation can induce scission of collagen polymer main chains and damage molecular structure [30]. The results of this study also suggest that excessive irradiation of 180 min promotes crosslinking reactions while simultaneously causing partial degradation of main chains and formation of non-uniform crosslinking points, resulting in increased surface irregularity.

2.4. Quantitative Evaluation of Local Elastic Modulus in Liquid

AFM force curve measurements revealed the distribution of local elastic modulus, as shown in Figure 4. For CS, the elastic modulus increased markedly with UV irradiation time, rising from approximately 0.5 MPa (500 kPa) at 0 min to 1.2 MPa (1200 kPa) at 30 min and reaching 2.0 MPa (2000 kPa) at 60 min. This increase in elastic modulus directly demonstrates that intermolecular crosslinks formed by UV irradiation strengthen the collagen molecular network and enhance resistance to deformation. The response to indentation deformation measured by force curves reflects the crosslink density at the nanoscale, confirming that sufficient crosslink network formation was achieved by 60 min irradiation [36].

However, at 180 min irradiation, the elastic modulus decreased to approximately 1.5 MPa (1500 kPa), falling below the peak value observed at 60 min. This decrease in elastic modulus, similar to the re-increase in surface roughness, is attributed to polymer backbone degradation caused by excessive UV irradiation [37]. While crosslink points are formed, concurrent progression of molecular backbone scission compromises the continuity of the network structure, resulting in deterioration of mechanical properties. These results suggest that optimization of irradiation time is critically important in UV crosslinking, requiring a balance between promotion of crosslinking reactions and avoidance of structural damage.

GS exhibited a similar trend, although the absolute elastic modulus values were lower than those of CS, increasing from approximately 0.3 MPa (300 kPa) at 0 min to 1.5 MPa (1500 kPa) at 60 min, and maintaining approximately 1.4 MPa (1400 kPa) at 180 min. The lower elastic modulus of GS compared to CS is attributed to gelatin’s random coil structure, which lacks the intermolecular interactions arising from collagen’s ordered triple helix structure. Nevertheless, the approximately 5-fold increase in elastic modulus upon UV irradiation indicates efficient crosslink formation between gelatin molecules.

2.5. Analysis of Chemical Structure Changes

From spectra obtained by FT-IR measurements, chemical structure changes in proteins by UV irradiation were evaluated (Figure 5). For both GS and CS samples, absorption bands characteristic of proteins—Amide I (approximately 1650 cm⁻¹), Amide II (approximately 1550 cm⁻¹), and Amide III (approximately 1240 cm⁻¹)—were clearly observed [38,39]. These peaks are derived from vibrations of peptide bonds and confirm that the samples are composed of proteins.

A particularly noteworthy change is the wavenumber shift in the Amide II peak (Figure 5, bottom panel). The Amide II peak of CS showed a tendency to shift to lower wavenumbers with increasing UV irradiation time. The Amide II band is primarily derived from coupled modes of N–H in-plane bending vibration and C–N stretching vibration [40], and it is sensitive to protein secondary structure and intermolecular hydrogen bonding state [41,42]. The shift to lower wavenumbers suggests formation or strengthening of intermolecular hydrogen bonds or changes in secondary structure. Crosslinking structures formed by UV irradiation are considered to be mainly dityrosine bonds between tyrosine residues and oxidative crosslinking of tryptophan residues [22,23], but with the formation of these covalent bonds, the orientation of surrounding peptide chains and reorganization of the hydrogen bonding network may occur.

In contrast, no significant changes due to UV irradiation were observed in the position and shape of the Amide I peak. The Amide I band is primarily derived from the C=O stretching vibration and reflects the content of secondary structures (α-helix, β-sheet, and random coil). The results of this study suggest that UV crosslinking does not greatly change the secondary structure of proteins themselves but primarily brings about crosslinking point formation between molecules and reorganization of the hydrogen bonding network.

2.6. Effect of UV Irradiation on Cell Proliferation Capability

The results of cell proliferation capability evaluated by CCK-8 assay are shown in Figure 6. For both CS and GS, UV irradiation significantly enhanced cell proliferation compared to untreated controls at all time points measured.

For GS (Figure 6, right panel), cell proliferation on UV-unirradiated (0 min) samples was markedly suppressed, with absorbance on day 1 of culture at approximately 0.2 and remaining at an extremely low value of approximately 1.2 on day 3. This is attributed to the rapid dissolution and swelling of uncrosslinked gelatin sheets in PBS, preventing them from providing a stable scaffold structure to serve as a foundation for cell adhesion. In bright-field observations described later, cells on unirradiated samples were observed in an aggregated state, confirming that the function as a scaffold material was not sufficiently exhibited.

Cell proliferation capability was dramatically improved in samples with crosslinking structures introduced by UV irradiation. For samples with 30 min of irradiation, absorbance on day 3 of culture reached approximately 2.4, showing approximately 2.0-fold increase compared to unirradiated samples. For samples with 60 min of irradiation, absorbance was approximately 2.7, and for samples with 180 min of irradiation, it was approximately 2.9, showing the highest cell proliferation capability. Statistical analysis confirmed that all UV-irradiated conditions (30, 60, and 180 min) showed significant enhancement compared to the 0 min control on days 2 and 3 (p < 0.05). This result clearly demonstrates that UV crosslinking improved the structural stability of scaffold materials and provided a foundation for cells to adhere and proliferate.

Interestingly, cell proliferation on GS showed continuously increasing values with extended UV irradiation, reaching a maximum at 180 min (absorbance: ~2.9). This trend differs from the elastic modulus pattern, which peaked at 60 min and slightly decreased at 180 min, suggesting that chemical factors (such as RGD motif accessibility) may provide additional proliferation enhancement beyond mechanical optimization.

The relationship between substrate stiffness (elastic modulus) and cell proliferation has been extensively reported in the literature. For fibroblasts, the optimal stiffness varies depending on substrate chemistry, dimensionality (2D vs. 3D), and cell source, but many studies report optimal proliferation in the range of approximately 1–15 kPa, particularly near 7–15 kPa [43,44,45]. Proliferation is generally suppressed on very soft substrates (below 1 kPa), reaches an optimum at intermediate stiffness, and then either decreases or plateaus on stiffer substrates.

The local elastic modulus measured by AFM in this study (GS 60 min: approximately 1.5 MPa = 1500 kPa; CS 60 min: approximately 2.0 MPa = 2000 kPa) is more than 100-fold higher than these values from the literature. However, several factors must be considered when interpreting these measurements. First, there is a fundamental difference in measurement scale: AFM nanoindentation measures local mechanical response at the scale of tens of nanometers, whereas cells sense substrates at the scale of several to tens of micrometers through their focal adhesions [43,44,45]. Second, the scaffold sheets in this study are thin films several micrometers thick that are supported by nylon membranes. Cells may sense the effective stiffness of the underlying support through the scaffold material, resulting in a perceived elastic modulus lower than the AFM-measured surface values. This architectural consideration is particularly relevant for thin-film scaffolds where substrate effects can dominate cell-sensed mechanics.

The favorable cell proliferation observed may be substantially influenced by the abundance of chemical cell adhesion sites (RGD sequences) rather than by substrate stiffness optimization. For GS in particular, the exposure of RGD sequences accompanying gelatinization may have exerted a powerful proliferation-promoting effect that supersedes the influence of physical stiffness. Indeed, as discussed below, the fact that GS consistently exhibited superior cell proliferation capacity compared to CS at 180 min—despite similar or even slightly lower elastic modulus values—is difficult to explain solely by physical factors, suggesting that the chemical factor of RGD sequence exposure is dominant [10,11]. Therefore, we conclude that the optimality of the 60 min irradiation condition in this study results not from the absolute elastic modulus value itself, but from the optimal balance of multiple factors: (1) securing structural stability through crosslinking; (2) avoiding structural damage from excessive irradiation; and (3) effective presentation of cell adhesion sites.

2.7. Observation of Cell Adhesion and Spreading Morphology

The results of cell morphology observation by bright-field microscopy are shown in Figure 7. This observation morphologically supports the differences in cell proliferation capability evaluated by the CCK-8 assay. On GS unirradiated (0 min) samples, cells could not adhere to the substrate surface and were observed as spherical clusters aggregated with each other. This aggregated morphology directly indicates that the scaffold material lost structural stability due to dissolution and swelling and could not provide a foundation for cell adhesion.

On GS samples with crosslinking structures introduced by UV irradiation, cell morphology changed dramatically. At all irradiation conditions of 30, 60, and 180 min, cells adhered firmly to the sample surface and exhibited spindle-shaped or polygonal morphology with greatly extended cytoplasm. This spreading morphology indicates that cells recognized the scaffold material, formed focal adhesions, and developed cytoskeletons [44,45]. Particularly on 180 min irradiation samples, cells showed the most favorable spreading, with high cell density and cell–cell contacts observed, confirming morphology suggesting active proliferation activity. Cell spreading area is an important indicator reflecting cell survival, proliferation, and differentiation state, and spread cells generally show active metabolic activity and proliferation capability [46].

For CS samples, cell adhesion and spreading were also promoted by UV irradiation. Unlike the severe aggregation observed on GS unirradiated samples, CS unirradiated samples showed partial cell adhesion to the substrate surface, reflecting the inherent structural stability of collagen. After UV irradiation, cell spreading was further promoted, and cell density increased substantially. On UV-irradiated CS samples (60 and 180 min), cells exhibited well-spread morphology comparable to that observed on UV-irradiated GS samples, which is consistent with the similar CCK-8 absorbance values observed for both materials after UV treatment. This observation result is consistent with CCK-8 assay results and demonstrates that UV crosslinking effectively enhances cell adhesion and proliferation on both CS and GS substrates.

Collectively, these comprehensive characterization results demonstrate that UV crosslinking effectively modifies both the physicochemical properties (gel fraction, surface morphology, elastic modulus, and chemical structure) and biological performance (cell proliferation and adhesion) of collagen and gelatin sheets. The results reveal distinct material-specific responses, with gelatin showing continued enhancement at extended irradiation times, while collagen maintains consistent performance across UV-treated conditions.

3. Discussion

3.1. Enhancement of Cell Scaffold Performance by UV Crosslinking

The primary finding of this study is that UV crosslinking significantly enhances the cell scaffold performance of both collagen and gelatin sheets. These proliferation results, detailed in Section 2.6, directly reflect the improved structural integrity established by crosslinking. This enhancement directly correlates with the establishment of structural stability, as confirmed by gel fraction measurements showing improved water resistance (GS: 22% to 80% at 60 min UV).

These results are consistent with the general understanding that scaffold structural integrity is a prerequisite for effective cell culture [12,31]. Charulatha and Rajaram [31] demonstrated that crosslinking degree significantly influences collagen membrane performance, and our findings extend this concept to UV-crosslinked systems with quantitative nanoscale characterization.

The elastic modulus values achieved represent successful crosslinking, though the relationship between local AFM measurements and cell-sensed substrate stiffness requires careful interpretation (see Section 3.3). This mechanical optimization likely contributes to the favorable cell responses observed. Importantly, both materials demonstrated comparable cell proliferation after UV treatment, suggesting that once appropriate structural stability and mechanical properties are achieved, both collagen and gelatin provide suitable environments for fibroblast culture.

3.2. Differential Responses of Collagen and Gelatin to Extended UV Irradiation

An intriguing observation was the divergent behavior of GS and CS at extended irradiation times. GS exhibited progressive increases in cell proliferation with irradiation time, reaching maximum values at 180 min (absorbance ~2.9), while CS showed consistently high proliferation across all UV-treated conditions (30–180 min; absorbance ~2.5–2.7).

This difference is attributed to distinct molecular structures. In collagen, the triple helix structure sterically constrains RGD sequence accessibility [5,7], whereas gelatin’s denatured structure exposes these motifs. Taubenberger et al. [11] demonstrated that collagen denaturation enhances RGD exposure and cell adhesion. The continued proliferation enhancement of GS at 180 min—despite a slight decrease in elastic modulus (from ~1.5 MPa at 60 min to ~1.4 MPa at 180 min)—strongly suggests that chemical factors, particularly RGD motif accessibility, contribute beyond mechanical optimization for gelatin-based materials.

For CS, the consistent performance across UV-treated conditions (absorbance ~2.5–2.7) indicates a broader operational window advantageous for practical applications. The inherent stability of the collagen triple helix structure provides a stable foundation that is effectively enhanced by crosslinking.

It should be noted that 254 nm UV irradiation may induce photochemical side effects in proteins, including oxidative modifications of aromatic amino acids, formation of carbonyl groups, and main-chain scission [30]. While the gel fraction and AFM data suggest that such degradation becomes significant only at extended irradiation times (180 min), the potential accumulation of oxidative modifications even at shorter irradiation times cannot be excluded. From a biological safety perspective, any remaining reactive species generated during UV crosslinking are expected to dissipate during the subsequent PBS washing and overnight preincubation steps performed prior to cell seeding. The consistent cell proliferation enhancement observed at all UV-irradiated conditions supports the absence of significant cytotoxic effects under the conditions tested.

3.3. Relationship Between Nanoscale Surface Properties and Cell Response

AFM measurements in PBS revealed distinct surface morphologies: GS maintained extremely smooth surfaces (R_q: 1–4 nm) regardless of UV irradiation time, while CS showed irradiation-dependent roughness changes (47 nm at 0 min to 21 nm at 60 min, increasing to 35 nm at 180 min).

The smoothing of CS surfaces at 60 min UV correlates with optimal elastic modulus values, suggesting that appropriate crosslinking restricts molecular chain swelling and creates more uniform surfaces. The roughness increase at 180 min, accompanied by decreased elastic modulus, is consistent with previous reports of polymer chain degradation under excessive UV exposure [30].

Interestingly, despite the markedly different surface topographies of GS and CS, both materials supported comparable cell proliferation after UV treatment. This observation suggests that within the nanoscale roughness ranges examined (GS: 1–4 nm; CS: 21–47 nm), surface topography is not the dominant factor determining NIH-3T3 proliferation. Rather, scaffold stability and appropriate mechanical properties appear more influential, with potential additional contributions from chemical adhesion signals.

These findings contrast somewhat with studies emphasizing nanoscale topography effects on cell behavior [24,43], but are consistent with the concept that optimal substrate conditions represent a multifactorial balance rather than single-parameter optimization [25].

To contextualize the mechanical properties achieved in this study, comparison with previous reports on UV-crosslinked collagen-based materials is instructive. Tirella et al. (as cited in Sarvari et al. [46]) demonstrated that the elastic modulus of collagen hydrogels could be tuned between approximately 0.9 and 3.6 kPa by varying collagen concentration and employing riboflavin-mediated UV irradiation combined with freeze-drying. The elastic modulus values obtained in the present study—measured by AFM nanoindentation in PBS (CS: 0.5–2.0 MPa; GS: 0.3–1.5 MPa)—are substantially higher than these previously reported bulk hydrogel values. This difference is attributable to the distinct material format: the sheet materials in the present study are dry-cast thin films supported on nylon membranes, rather than hydrated bulk hydrogels, resulting in a more densely packed protein matrix. Furthermore, as discussed in this section, AFM-measured local nanoscale elastic modulus values inherently overestimate the cell-perceived effective stiffness due to the difference in measurement scale between nanoindentation and cell–substrate interaction through focal adhesions. These comparisons underscore that the mechanical characterization modality and material format must be carefully considered when comparing across studies and highlight the complementary nature of nanoscale AFM measurements and bulk mechanical testing for comprehensive scaffold characterization.

3.4. Implications for Scaffold Material Design and Future Directions

This study provides several insights for cell scaffold material design. First, UV crosslinking at 254 nm effectively enhances the performance of both collagen and gelatin sheets, establishing this as a practical, non-toxic crosslinking method. The optimal irradiation time (60 min) provides balanced physicochemical properties suitable for general applications, while extended irradiation (180 min) may offer advantages for gelatin-based applications requiring maximum cell proliferation.

Second, the choice between collagen and gelatin should consider application-specific requirements. Collagen offers inherent stability and consistent performance across a range of processing conditions. Gelatin provides potentially higher maximum proliferation capability (absorbance ~2.9 at 180 min) and may offer advantages in applications where enhanced cell adhesion is critical, such as cell expansion or early-stage tissue formation.

Future studies should address several limitations of the current work. First, the use of a single cell type (NIH-3T3) limits generalizability to other cell types with different adhesion characteristics. Long-term culture studies and functional assessments—including protein expression and matrix production—would provide a more comprehensive evaluation of scaffold performance. Furthermore, direct quantification of RGD accessibility through biochemical methods would strengthen the mechanistic interpretation of the differential responses observed.

From a mechanistic perspective, direct confirmation of specific crosslink chemistry—including dityrosine bond quantification and identification of tryptophan-derived crosslink products—represents an important future direction. Mass spectrometry-based peptide mapping would provide the definitive molecular-level evidence that FT-IR alone cannot supply [47]. The potential influence of endogenous photosensitizers and the relative contributions of oxidative amino acid modifications versus backbone crosslinking also warrant systematic investigation.

Beyond the current system, investigation of UV crosslinking effects on other naturally derived materials (e.g., fibrin, silk fibroin) and optimization of irradiation parameters for specific tissue engineering applications represent promising directions for future research. Future studies should also incorporate quantitative morphometric analysis—such as automated measurement of cell spreading area, aspect ratio, and cell density—to complement the qualitative observations reported here.

4. Materials and Methods

4.1. Materials

The collagen used in this study was Type I collagen acidic solution (I-AC, 5 mg/mL; Koken Co., Ltd., Tokyo, Japan). Gelatin was prepared by heat-treating the above 0.5 wt% atelocollagen solution at 60 °C for 30 min. Complete denaturation by heat treatment was confirmed by circular dichroism measurements described below. Phosphate-buffered saline (PBS, pH 7.4), culture media, and cell culture reagents used in cell culture experiments were all purchased from Fujifilm Wako Pure Chemical Corporation, Osaka, Japan. Nylon membrane filters (Cytiva Amersham Hybond-N Membranes, pore size 0.2 μm, processed into ring shapes; Cytiva, Marlborough, MA, USA) were used as cell scaffold supports.

4.2. Sample Preparation

Cell scaffold sheets were prepared by the following procedure (Figure 8). First, ring-shaped circular nylon membrane filters were placed in 55 mm diameter polystyrene culture dishes (Iwaki, Tokyo, Japan). Three mL of 0.5 wt% atelocollagen solution was uniformly spread on the culture dishes. For gelatin sheet preparation, the atelocollagen solution was heat-treated at 60 °C for 30 min, and then, similarly, 3 mL was spread. Culture dishes with spread solutions were left standing in an incubator at 37 °C with 5% CO₂ for 2 h to promote protein molecule aggregation and initial gelation. Subsequently, culture dishes were transferred to a constant temperature and humidity chamber (temperature 10 °C, relative humidity 40%) and slowly dried over 48 h. These drying conditions were set to avoid rapid structural changes in proteins and form uniform sheet structures. After drying, the thin film-like collagen sheets (hereinafter CS) and gelatin sheets (hereinafter GS) formed on membranes were obtained.

4.3. UV Irradiation

Crosslinking treatment by short-wavelength UV irradiation was performed on the prepared CS and GS. As a UV light source, low-pressure mercury lamps with a main wavelength of 254 nm (output 15 W, 5 lamps; the BIO-LINK BLX, Vilber Lourmat, Marne La Vallee, France) were used. Sheet samples with membranes were placed horizontally in an irradiation chamber, and the distance between the sample surface and lamps was fixed at 20 cm. Irradiation was performed at room temperature in an atmospheric environment, and irradiation times were four conditions: 0 min (unirradiated control), 30 min, 60 min, and 180 min. The UV irradiance at the sample surface was measured using a UV radiometer (SP-82UV, Mather Tool Co., Ltd., Nagano, Japan) and was 1.82 mW/cm². Accordingly, the UV energy doses delivered at each irradiation time were: 0 mJ/cm² (0 min), 3276 mJ/cm² (30 min), 6552 mJ/cm² (60 min), and 19,656 mJ/cm² (180 min). During irradiation, to minimize temperature rise in samples, a ventilation fan was installed in the chamber to maintain sample temperature at 30 °C or below. After irradiation, samples were stored in light-shielding containers in desiccators and preserved at room temperature until measurements.

4.4. Gel Fraction Measurements

To evaluate the degree of crosslinking structure formation introduced by UV irradiation, gel fraction measurements were performed by dissolution testing. CS and GS samples prepared under each irradiation condition were placed in 35 mm culture dishes (with samples), and initial weight (W₀) in the dried state was measured with a precision electronic balance (resolution 0.01 mg; (ATX324R; Shimadzu Corporation, Kyoto, Japan)). Next, 2 mL of PBS solution (37 °C) was added to the 35 mm dishes containing samples, and they were incubated at 37 °C for 24 h under static conditions. After incubation, the solution was removed with an aspirator, and rinse washing with 5 mL of distilled water for 5 min was repeated three times. After incubation, the solution was removed with an aspirator, and rinse washing with 5 mL of distilled water for 5 min was repeated three times. After washing, samples were dried in a constant temperature and humidity chamber (10 °C, 40% RH) for 72 h, and weight after drying (W₁) was measured. Gel fraction (%) was calculated by the following equation:

Gel Fraction (%) = (W₁/W₀) × 100

Three specimens (n = 3) were measured for each condition, and mean values and standard deviations were calculated.

4.5. Scanning Electron Microscopy (SEM)

To observe the fine structure of sample surfaces in the dried state, SEM observations were performed. Samples of each irradiation condition were cut into 3 mm × 3 mm pieces and fixed to SEM sample stages using carbon tape. Platinum–palladium alloy was deposited on sample surfaces to a thickness of approximately 5 nm using a sputter coater (ion sputtering device; (JFC-1600; JEOL Ltd., Tokyo, Japan)). Deposition conditions were a vacuum degree of 2 Pa, a current of 20 mA, and a deposition time of 60 s. Observations were performed using a scanning electron microscope (JSM-6510; JEOL Ltd., Tokyo, Japan) under conditions of an acceleration voltage of 10.0 kV, a working distance of 8 mm, and a magnification of ×3000. For each sample, images were acquired from at least three different fields of view to confirm surface morphology uniformity.

4.6. Atomic Force Microscopy (AFM)

To evaluate the nanoscale morphology and local elastic modulus of sample surfaces in liquid, AFM measurements were performed. For measurements, an AFM device (Nanowizard 4; JPK Instruments, Berlin, Germany) was used with a cantilever for liquid measurements (HQ-13-Au; Oxford Instruments Asylum Research, Santa Barbara, CA, USA; rectangular shape, length: 485 μm, width: 33 μm, tip height: 14 μm, tip radius: <10 nm, and spring constant: 0.2 N/m). Samples were fixed to the AFM stage in 35 mm dishes (with samples) as they were. As a measurement solution, PBS (pH 7.4) was added in an amount of approximately 3 mL, and measurements were performed with sample surfaces completely immersed in solution. After waiting approximately 10 min for laser stabilization (to minimize thermal drift and ensure stable baseline measurements), measurements were started.

Surface morphology observation and local elastic modulus measurements were performed by QI mode (Quantitative Imaging mode). In QI mode, force curve measurements are performed for each pixel, and height and stiffness are simultaneously measured and imaged. Scanning range was 5 μm × 5 μm, and measurement conditions were set point: 500 pN, pixels: 50 × 50 (total 2500 points), and speed: 5 μm/s. From the obtained height images, root mean square roughness (R_q) was calculated as a surface roughness parameter. For each sample, measurements were performed at three different positions near the center of the dishes (n = 3), and mean values and standard deviations were obtained.

For local elastic modulus calculation, Hertz’s contact theory model [48] was applied to obtain force curves. In this study, as premises for data analysis, sample deformation was assumed to be elastic, the cantilever tip shape was assumed to be spherical (radius of curvature < 10 nm), and the sample Poisson’s ratio was assumed to be 0.5 (incompressible). For each irradiation condition, at least three samples were measured, and statistical analysis was performed.

4.7. Fourier Transform Infrared Spectroscopy (FT-IR)

To evaluate chemical structure changes by UV irradiation, FT-IR measurements were performed. For measurements, an FT-IR spectrophotometer (VERTEX70; Bruker Optics, Ettlingen, Germany) was used, and measurements were performed by the attenuated total reflectance (ATR) method using a built-in diamond ATR crystal (Bruker Optics Inc., Germany). Samples were cut into 7 mm × 7 mm pieces and brought into close contact with an ATR crystal. Measurement conditions were a wavenumber range of 4000–400 cm⁻¹, a resolution of 2 cm⁻¹, and a number of scans of 64. Twelve different locations were measured for each sample (n = 12), and the obtained spectra were subjected to baseline correction after correcting for the effects of water vapor and carbon dioxide in the atmosphere.

Focusing on absorption bands characteristic of proteins—Amide I (1700–1600 cm⁻¹), Amide II (1600–1500 cm⁻¹), and Amide III (1300–1200 cm⁻¹)—changes in peak positions and peak intensities by UV irradiation were analyzed. Particularly, the Amide II peak is known to reflect protein secondary structure and intermolecular hydrogen bonding state [41,42], and in this study, wavenumber shifts in this peak were analyzed in detail. Peak positions were determined using second derivative spectra, and changes under each irradiation condition were quantitatively evaluated.

4.8. Circular Dichroism (CD) Spectroscopy

To confirm that heat treatment implemented during gelatin solution preparation completely dissociated the triple helix structure of collagen, circular dichroism (CD) measurements were performed. For measurements, a circular dichroism spectrometer (J-820; JASCO Corporation, Tokyo, Japan) was used. As measurement samples, a 0.5 wt% atelocollagen solution diluted to approximately 0.03 mg/mL with 20 mM acetic acid was prepared, and two types of samples were measured: before heat treatment (Native) and after heat treatment at 60 °C for 30 min.

For measurements, a cylindrical quartz cell with an optical path length of 10 mm was used, and approximately 3 mL of sample solution was filled in the cell. Measurements were performed under conditions of a measurement wavelength range of 190–260 nm, a scanning speed of 100 nm/min, a data acquisition interval of 0.1 nm, a resolution of 1 nm, and a number of scans of 8. Measurement temperature was set at 25 °C, and the temperature was maintained constant by a Peltier-type temperature control device connected to the cell holder. The obtained CD spectra were background-subtracted for the solvent. The degree of structural denaturation was evaluated by the presence or absence of a positive peak around 220 nm that is characteristic of collagen triple helix structure.

4.9. Cell Culture

For cell culture experiments, the mouse fibroblast cell line NIH-3T3 (derived from a Swiss mouse embryo) (purchased from the Japanese Collection of Research Bioresources Cell Bank, Ibaraki, Japan) was used. Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Fujifilm Wako Pure Chemical Corporation, Osaka, Japan) supplemented with 10% fetal bovine serum (FBS; (Sigma-Aldrich, St. Louis, MO, USA)) and 1% penicillin–streptomycin solution (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan). Culture conditions were 37 °C, 5% CO₂, and a saturated humidity environment, with medium changes every 2–3 days. When cells reached approximately 70–80% confluence, subculture was performed using a trypsin–EDTA solution prepared in our laboratory. Briefly, 0.2 g of trypsin (Nacalai Tesque, Inc., Kyoto, Japan) and 0.05 g of EDTA (Disodium Dihydrogen Ethylenediaminetetraacetate Dihydrate; Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) were dissolved in 200 mL of PBS (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan). Cells of passages 10–20 were used for experiments.

To evaluate functionality as cell scaffolds, 48-well multiwell plates for suspension cells (Iwaki, Tokyo, Japan) with CS and GS samples prepared under each UV irradiation condition were used. Samples were placed in each well (n = 5). Samples were surface sterilized by immersion in 70% ethanol for 5 min, followed by three washes with sterile PBS. Furthermore, by preincubating overnight with culture medium, samples were sufficiently swollen, and serum protein adsorption to surfaces was promoted.

4.10. Cell Proliferation Assay

For evaluation of cell proliferation capability, the Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Kumamoto, Japan) assay was used [49]. This assay is a highly sensitive cell viability and proliferation capability measurement method using intracellular dehydrogenase activity as an indicator. NIH-3T3 cells were seeded at a density of 6 × 10⁴ cells/well in 48-well plates, with each sample placed. Cell viability measurements using CCK-8 reagent were performed at 1, 2, and 3 days after cell seeding. At each measurement time point, medium was removed, and fresh medium (270 μL) and CCK-8 reagent (30 μL) were added. After incubating at 37 °C with 5% CO₂ for 2 h, 100 μL of medium was transferred from each well to a 96-well plate, and absorbance was measured using a microplate reader (wavelength 450 nm; (Thermo Fisher Scientific, Waltham, MA, USA)). Five wells (n = 5) were measured for each condition, and mean values and standard deviations were calculated.

4.11. Cell Morphology Observation

To observe cell adhesion and spreading morphology, bright-field microscopy observations were performed. On day 2 after cell seeding, the medium was removed from each well and washed twice with PBS to remove non-adherent cells. Subsequently, cell morphology was observed immediately using a phase-contrast microscope (BZ-X1000; Keyence Corporation, Osaka, Japan). A 10× objective lens was used, and images were acquired from at least five fields of view for each sample to ensure representative sampling. From the obtained images, the cell adhesion state (isolated dispersion or aggregation), cell morphology (spread or spherical), and cell density were qualitatively evaluated. No quantitative morphometric analysis was performed; observations served to support the CCK-8 proliferation data. Morphological assessment in this study was qualitative in nature. No quantitative morphometric analysis, such as automated measurement of cell spreading area or cell density, was performed. This represents a limitation of the current study, and future work should incorporate quantitative image analysis to provide a more rigorous evaluation of cell–material interactions.

4.12. Statistical Analysis

All quantitative data are expressed as mean ± standard deviation. For comparison between groups, one-way analysis of variance (ANOVA) was performed, and when significant differences were observed, post hoc testing was performed by Tukey’s multiple comparison test. Statistical significance was set at p < 0.05.

5. Conclusions

This study demonstrates that UV crosslinking (254 nm) effectively enhances the cell scaffold performance of both collagen sheets (CS) and gelatin sheets (GS). The key findings are summarized as follows.

First, UV crosslinking establishes the structural stability essential for cell scaffold function. Gel fraction measures confirmed that 60 min UV irradiation increased GS water resistance from 22% to 80%, while CS maintained high stability (>95%) across all conditions. This structural stabilization is the fundamental prerequisite enabling cell adhesion and proliferation.

Second, UV crosslinking optimizes the mechanical properties of both materials. AFM measurements revealed maximum elastic modulus at 60 min irradiation (CS: 2.0 MPa; GS: 1.5 MPa), values within the optimal range for fibroblast proliferation. Extended irradiation (180 min) caused slight decreases in elastic modulus, which is consistent with reported UV-induced chain degradation.

Third, both CS and GS demonstrated significantly enhanced NIH-3T3 cell proliferation following UV irradiation compared to untreated controls (p < 0.05). On day 3 of culture, UV-treated samples showed absorbance values of 2.5–2.9 compared to 1.2–1.7 for untreated materials. GS exhibited progressively increasing proliferation with irradiation time, reaching a maximum at 180 min (absorbance: ~2.9), while CS showed consistently high proliferation across all UV-treated conditions (30–180 min, absorbance: ~2.5–2.7).

Fourth, the observation that GS achieves maximum cell proliferation at 180 min despite a slight decrease in elastic modulus from 60 min (1.5 MPa) to 180 min (1.4 MPa) strongly suggests that chemical factors—specifically RGD motif exposure through thermal denaturation and continued structural reorganization during extended UV irradiation—provide additional enhancement beyond structural and mechanical optimization.

In conclusion, UV crosslinking represents an effective, non-toxic method for preparing high-performance cell scaffold materials from both collagen and gelatin. The choice between materials and optimization of irradiation conditions should be guided by specific application requirements, with 60 min irradiation recommended as a baseline condition that provides balanced properties for general tissue engineering applications.