Investigation and development of bilayer scaffolds: ZnONPs-loaded poly (vinyl alcohol) incorporated with Platostoma palustre aqueous extract and sponge collagen derived from fish skin

Abstract

This study introduces the design of two bilayer scaffolds that offer safety and show potential for biomedical applications, especially for wound healing in the skin or oral cavity. Each scaffold is composed of a collagen sponge layer derived from warm-water fish skin (seabass and tilapia) and a PVA layer incorporated with Hsiantsao extract and biosynthesized ZnONPs. Both extracted collagens were identified as type I, with their purity and triple-helical structure confirmed by electrophoresis, FT-IR, UV-Vis, and EDX analyses. TEM characterization revealed that the ZnONPs were small (7.95 ± 1.45 nm) and spherical. The bilayer scaffolds utilize the unique functions of each layer: the denser PVA layer, integrated with nanoparticles, acts as a barrier against dust and bacteria and releases bioactive compounds from the Hsiantsao extract, while the sponge collagen layer supports cell proliferation. Mechanically, the scaffolds showed high flexibility, with a tensile strength of about 4 MPa and an elongation at break of around 300%. They also absorbed fluids rapidly and maintained a slightly acidic pH (6.5–6.8). Additionally, the scaffolds exhibited excellent biocompatibility (cell viability > 115% after 48 h and a hemolytic percentage < 1.5%), strong antioxidant activity (69–70% DPPH and 80% ABTS scavenging), and antimicrobial properties against both Gram-positive and Gram-negative bacteria. These findings suggest that the Hsiantsao/ZnONPs-loaded PVA/Seabass and Hsiantsao/ZnONPs-loaded PVA/Tilapia scaffolds are promising candidates for medical treatments.

Graphical Abstract

Introduction

Collagen serves as a crucial structural protein and is often utilized in biomedical fields because of its bioactivity, biodegradability, and compatibility with biological systems [1, 2]. Collagen for medical and pharmaceutical use typically comes from two sources: recombinant human production or animals [3]. However, recombinant human collagen, produced through sophisticated biotechnology and genetic engineering [4, 5], is expensive and yields are low [6]. Meanwhile, animal-derived collagen carries risks of disease transmission and may not be acceptable due to religious reasons [1, 7]. Research has shown that collagen derived from aquatic sources has lower cytotoxicity and higher cell viability compare to bovine collagen [8]. Fins, scales, skin, and bones from freshwater and marine fish are important collagen sources. Utilizing these materials provides collagen and reduces fish waste from markets and processing plants, helping to mitigate environmental pollution [9]. The amino acid profile of collagen differs depending on the species and environmental conditions, leading to differences in its physical properties [10, 11]. Notably, hydroxyproline content strongly influences the mechanical and thermal stability of collagen by promoting intermolecular hydrogen bonding within the triple helix [12, 13]. Collagen from warm-water fish typically contains higher hydroxyproline levels compared to that from cold-water fish, resulting in greater thermostability and rigidity [14]. Asian seabass (Lates calcarifer) and Mozambique tilapia (Oreochromis mossambicus) are warm-water fish species widely cultivated for aquaculture, generating substantial non-edible by-products such as skin [15, 16]. This skin is considered a safe and promising source of protein for collagen extraction [17]. Porous collagen sponges are widely utilized in wound care due to their ability to promote fibroblast migration and encourage the synthesis of new collagen throughout the repair process [18]. A recent study showed that shad collagen sponge promotes rapid hemostasis, immune regulation, and wound healing by enhancing angiogenesis, cell proliferation, and collagen remodeling [3]. However, native collagen has limited antioxidant properties [7, 19] and is susceptible to microbial contamination [20]. The combination of herbal ingredients with natural polymers in wound dressings represents an innovative approach in wound care, as these additions enhance antioxidant activity and support wound contraction, vascularization, and epithelialization [21]. In a previous study, solid lipid nanoparticles containing marshmallow and clove oil were incorporated into collagen sponges to improve their ability to heal mouth ulcers. These collagen sponges act as mechanical supports that adhere securely to the ulcer site and minimize the need for frequent dressing changes, both of which are important for effective management of mouth ulcers [22].

Platostoma palustre (Mesona chinensis), an annual plant common in tropical and subtropical areas and usually uses as “Hsiantsao” herbal jelly in Taiwan [23]. It contains a wealth of bioactive substances, such as polysaccharides, phenolics, alkaloids, and flavonoids [23, 24], which endow it with diverse pharmacological effects including antioxidant, immunomodulatory, anti-inflammatory, anti-tumor, hypoglycemic effects, hepatoprotective, and regulation of gut microbiota [23, 25]. Thanks to its beneficial biological properties and safety, Platostoma palustre (Hsiantsao) is considered a promising candidate for enhancing wound healing therapies [24].

Hsiantsao has also been used as a reducing and stabilizing agent in the bio-synthesis of Zinc oxide nanoparticles (ZnONPs) from zinc salt solutions, taking advantage of its antioxidants [26, 27]. Biological nanoparticle synthesis is safer and more environmentally friendly than conventional methods, which often involve toxic chemicals, health risks, and high costs [27]. ZnONPs green synthesized using Hsiantsao aqueous extract have shown antibacterial effectiveness against Gram-positive and Gram-negative bacteria [26].

To utilize the unique properties of each component and enhance the overall efficiency of the scaffold for potential wound healing in various areas, such as skin and intraoral sites, we designed a bilayer structure with each layer serving a specific function. The sponge layer was fabricated from type I collagen extracted from the skin of warm-water fish species, specifically Lates calcarifer (seabass) and Oreochromis mossambicus (tilapia). This layer acts as a structural scaffold, absorbing exuded fluids and supporting cell proliferation. The other layer combines three ingredients with complementary effects: mechanical strength provided by the synthetic polymer polyvinyl alcohol (PVA), antioxidant activity from Hsiantsao aqueous extract, and antibacterial activity from green-synthesized zinc oxide nanoparticles. Based on this approach, we successfully fabricated two bilayer scaffolds: Hsiantsao/ZnONPs-loaded PVA/Seabass (PHZnO/S) and Hsiantsao/ZnONPs-loaded PVA/Tilapia (PHZnO/T).

Materials and methods

Materials

Fish skin from seabass and tilapia were obtained from the Kaohsiung market (Taiwan). Biosynthesized ZnONPs and the number 2 sample of Hsiantsao powder were supplied by our laboratory, with their properties referenced from previous research [24, 26]. Sodium chloride and tromethamine hydrochloride were purchased from Cyrusbioscience (Taiwan). Butanol-1-ol was sourced from Duksan Pure Chemicals (Korea). EDTA and trichloroacetic acid were acquired from Thermo Fisher Scientific (USA); β-mercaptoethanol from Bio Basic (Canada). American bacteriological agar was purchased from Condalab (Spain), and nutrient broth from HiMedia (India). The Cell counting kit-8 (CCK-8) reagent was purchased from Energenesis Biomedical (Taiwan). PVA (Mw 89,000–98,000), 99 + % hydrolyzed, along with other chemicals not mentioned above, were supplied from Sigma-Aldrich (USA).

Collagen extraction from fish skin

Acid-soluble collagen (ASC) from seabass and tilapia was extracted following Li’s method [28] with minor modifications. Fish skin was cleaned and soaked 6 h in 0.1 M NaOH (1:10 w/v), with the NaOH solution replaced every 2 h. Afterward, the seabass and tilapia skins were thoroughly washed with cold water until the water reached a neutral pH, treated 12 h with 10% butanol-1-ol (1:10 w/v) with the butanol-1-ol solvent replaced every 3 h, and rinsed again with cold water. The residues were extracted 3 days in 0.5 M acetic acid (1:10 w/v), filtered through a filter bag, and precipitated by adding 2.6 M NaCl and 0.5 M tris (hydroxymethyl) aminomethane at pH 7.5. The precipitate was centrifuged at 20,000 × g for 1 h at 4 °C using a centrifuge (5425 R, Eppendorf, Germany), re-dissolved in 0.5 M acetic acid. The solution was dialyzed using a membrane with a molecular cutoff of 14 kDa, against 10 volumes of 0.1 M acetic acid for half a day, followed by dialysis against deionized (DI) water for 48 h. Finally, the dialysates were lyophilized using a freeze dryer (Uniss, Taiwan). To avoid collagen degradation, all extraction processes are kept below 4 °C.

Characteristics of acid-soluble collagen

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)

Electrophoresis was conducted based on the Laemmli (1970) method [29] with slight modifications. 0.5% ASC solutions were purified 20 min using 10% trichloroacetic acid at 4 °C and then centrifuged 15 min at 20,000 × g. The precipitate was then mixed with a sample buffer (1 M tromethamine hydrochloride, 10% sodium lauryl sulphate, glycerol, 0.5 M EDTA, bromophenol blue, and β-mercaptoethanol). The mixture was boiled 7 min at 95 °C and immediately loaded onto a polyacrylamide gel (5% stacking gel and 8% separating gel). Electrophoresis was performed using an electrophoresis power supply (MP-310, Major Science, USA) at a constant voltage of 120 V. Afterward, the gel was stained for 2 h using 0.1% (w/v) Coomassie Brilliant Blue in 40% (v/v) methanol and 10% (v/v) acetic acid. After that, the gel was de-stained using 30% (v/v) methanol and 10% (v/v) acetic acid. A pre-stained protein marker (9-245 kDa) from Genomics was used as a ladder.

Ultraviolet-visible spectroscopy

The UV-Vis absorption spectra of fish skin collagen were recorded with a UV-Vis spectrophotometer (V-730, Jasco, Japan). Freeze-dried collagen was dissolved in 0.1 M acetic acid to produce 0.1% collagen solutions, followed by filtration to eliminate any particulates. The solution was transferred to a quartz cuvette and then examined within a wavelength range of 200 to 340 nm.

Fourier transform infrared spectroscopy (FT-IR)

FT-IR spectra were obtained with an FT-IR spectrometer (Jasco, Japan) across the range of 4000 to 400 cm^-1 to identify the characteristic functional groups in collagen. To ensure accuracy, each sample was prepared and measured three times.

Elemental analysis

The elemental composition of freeze-dried collagen extracted from fish skin was analyzed using Energy Dispersive X-ray (EDX; Oxford Instruments, Ultim Max 100) spectroscopy at an accelerating voltage of 5 kV and an acquisition time of 60 s, coupled with Scanning Electron Microscopy (SEM; Zeiss Gemini 450, Germany). EDX analysis was performed on three randomly selected areas per sample to ensure the representativeness of the results.

Characteristics of ZnONPs

The morphology and size of the ZnONPs were examined using Transmission Electron Microscopy (TEM). The ZnONP solution was sonicated to ensure dispersion, and then placed 10 μL of the solution onto a carbon-copper coated grid and dried in an oven at 105 °C for one day. To further analyze the elemental composition of the ZnONPs, EDX analysis was conducted.

Fabrication of bilayer scaffolds

The 1% (w/v) solutions of freeze-dried collagen in 0.1 M acetic acid were prepared and stored at 4 °C until use. Separately, a 12% PVA solution was prepared, and ZnONPs solution (final concentration of 0.1 mg/mL) was added, followed by gentle stirring. After the mixture cooled, Hsiantsao aqueous extract was added to reach a concentration of 0.5 mg/mL. Then, the resulting mixture was evenly spread in petri dishes and left at 25 °C for 24 h to form hydrogel films. These films were pre-frozen 1 h at -20 °C, after which the 1% collagen solution was applied as a second layer at a 1:1 ratio. The bilayer scaffolds were frozen 1 day at -20 °C, thoroughly rinsed with DI water, and freeze-dried to obtain the final bilayer scaffolds (Fig. 1).

Fig. 1.

Schematic representation of the bilayer scaffold fabrication procedure

Morphological characterization

The microstructure of the scaffolds was coated with gold for 1 min and then examined using SEM at an accelerating voltage of 3 kV. Additionally, EDX analysis was conducted to assess the distribution of ZnONPs.

Mechanical properties of bilayer scaffolds

Tensile strength

The structural integrity of the scaffold was assessed using a tensile testing machine (EZ test, Shimadzu, Japan). The scaffolds were pre-immersed in DI water for 10 s and mounted with a gauge length of 30 mm and stretched at 20 mm/min. Each bilayer scaffold specimen was tested four times. The tensile strength (MPa) and percent elongation were calculated using standard formulas (1-2):

$$\mathrm{Tensile\; strength}={\mathrm{F}}_{\max }/{\mathrm{A}}_{\mathrm{o}}$$ 1

$$\%\mathrm{Elongation}=\left[\right({\mathrm{L}}_{\mathrm{f}}-{\mathrm{L}}_{\mathrm{o}})/{\mathrm{L}}_{\mathrm{o}}]\mathrm{x}100\%$$ 2

Where F_max is represents the maximum load (N), A_o is the cross-sectional area (mm²), L_f is length at break, and L_o is original gage length.

Water contact angle

The water contact angle measurement was also employed to assess the wettability and hydrophilicity of the scaffolds. Specifically, a 5 μL drop of DI water was applied to each side of the bilayer scaffolds, and the contact angle between the water droplet and the scaffold surface was observed over 60 s and measured using ImageJ software. This assessment was performed four times.

Swelling ability

A 0.2 g dried sample was soaked in 40 mL of DI water and incubated at 37 °C. The change in weight was measured at time intervals, with each test performed four times. The swelling degree was determined using an Eq. (3).

$$\%\mathrm{Swelling\; degree}=\left[\right({\mathrm{W}}_{\mathrm{t}}-{\mathrm{W}}_{\mathrm{d}})/{\mathrm{W}}_{\mathrm{d}}]\mathrm{x}100\%$$ 3

Where W_t is the wet scaffold’s mass and W_d is the dried scaffold’s mass.

pH test

A 0.2 g sample was soaked in 40 mL of DI water and incubated at 37 °C. The pH values were measured at designated time intervals using a digital pH meter (Laqua, Horiba, Japan), with each experiment performed four times.

Antioxidant capacity

To evaluate antioxidant activity, both DPPH and ABTS radical scavenging assays were performed. For each method, a 0.2 g sample was soaked in 20 mL of DI water at 37 °C for 30 min, followed by filtration through Whatman paper. For the DPPH assay, 1 mL of the filtrate was combined with 1 mL of DPPH solution (0.01 mg/mL in absolute ethanol) and incubated in the dark for 30 min. For the ABTS assay, 1 mL of the filtrate was mixed with 1 mL of ABTS radical solution, which had been prepared by reacting 2.45 mM potassium persulfate with ABTS and incubating in the dark for 16 h, and then left in the dark for 30 min. Vitamin C (5 μL/mL) was used as a positive control in both assays. Absorbance was measured at 517 nm for DPPH and 734 nm for ABTS. All experiments were conducted in triplicate. The percentage of radical inhibition was calculated using the following Eq. (4):

$$\%\mathrm{Inhibition}=\left[\right({\mathrm{A}}_{\mathrm{c}}-{\mathrm{A}}_{\mathrm{s}})/{\mathrm{A}}_{\mathrm{c}}]\mathrm{x}100\%$$ 4

Where A_c and A_s are the absorbance values of the negative control and the sample, respectively.

Antibacterial

The antibacterial activity of the bilayer scaffolds was assessed using the agar disk diffusion method against Escherichia coli (ATCC 8739, Gram-negative) and Staphylococcus aureus subsp. aureus (ATCC 9144, Gram-positive), both obtained from the Bioresource Collection and Research Centre, Taiwan [30]. A 100 μL aliquot of bacterial suspension (10⁶ CFU/mL) was evenly spread onto nutrient agar plates, after which the bilayer scaffolds were placed on the agar surface and then incubated 1 day at 37 °C. Penicillin-streptomycin (0.1 mg/mL) was used as the positive control. The experiment was performed in triplicate.

Cell viability

Cytotoxicity of freeze-dried bilayer scaffold samples was assessed using mouse embryonic fibroblasts (NIH/3T3 cells). The NIH/3T3 cells were cultured in DMEM containing 10% fetal bovine serum and incubated at 37 °C in a 5% CO₂ atmosphere. Once confluence was reached, the cells were detached and seeded (1 × 10⁴ cells/well) onto scaffolds that had been sterilized by UV irradiation for 2 h [31]. At 24 and 48 h, the scaffolds were rinsed twice with PBS and then incubated 2.5 h in medium containing 10% CCK-8 reagent. Afterwards, 100 μL of the resulting mixture was transferred to a 96-well plate, and absorbance at 450 nm was recorded with a microplate reader (BIO-RAD, model 680). Control samples (cells without scaffold samples) and blank samples (medium with 10% CCK-8, no cells) were also tested. The experiment was performed in triplicate. Cell viability was determined using the following formula (5):

$$\%\mathrm{Cell\; viability}=\left[\right({\mathrm{A}}_{\mathrm{s}}-{\mathrm{A}}_{\mathrm{b}})/({\mathrm{A}}_{\mathrm{c}}-{\mathrm{A}}_{\mathrm{b}}\left)\right]\mathrm{x}100\%$$ 5

Where A_b, A_c, and A_s denote the absorbance readings for the blank, control, and sample, respectively.

Hemolysis study

Blood compatibility of the bilayer scaffolds was assessed using a hemolysis assay. A scaffold with a diameter of 1 cm was incubated with a mixture of 0.2 mL fresh anticoagulated blood and 5 mL of 0.9% NaCl at 37 °C for 1 h. After incubation, the mixture was centrifuged at 1000 rpm for 10 min. Then, 100 μL of the supernatant was transferred to a 96-well plate, and absorbance at 450 nm was measured. The negative control consisted of fresh anticoagulated blood and 0.9% NaCl without scaffolds, while the positive control contained fresh anticoagulated blood and 1% Triton X-100 in 0.9% NaCl without scaffolds. The test was performed in triplicate. The percentage of hemolysis was calculated as described below (6):

$$\%\mathrm{Hemolysis}=\left[\right({\mathrm{A}}_{\mathrm{s}}-{\mathrm{A}}_{\mathrm{n}})/({\mathrm{A}}_{\mathrm{p}}-{\mathrm{A}}_{\mathrm{n}}\left)\right]\mathrm{x}100\%$$ 6

Where A_s, A_p, and A_n represent the absorbance values of the samples, positive and the negative controls, respectively.

Statistical analysis

All calculations were carried out using SPSS Statistics 25 and OriginPro 2025. Data are presented as means ± SE from various sample lots. Statistical analysis was performed using one-way ANOVA, with a p-value < 0.05 regarded as statistically significant.

Results

Characteristics of collagen

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis

Figure 2a displays the electrophoretic band patterns of seabass and tilapia skin collagens. Collagen extracted exhibited two distinct α chains (α1, α2) as the main components, along with a higher molecular weight β chain. The band pattern indicated that both were type I, consistent with collagen isolated from the skin of silver pomfret (Pampus argenteus), crimson snapper (Lutjanus erythropterus) [7], and channel catfish skin (Ictalurus punctatus) [32].

Fig. 2.

Characteristics of ASC. SDS-PAGE (a), UV absorption spectra (b) and Fourier transform infrared spectrum (c)

Ultraviolet-visible spectroscopy

Pure type I collagen shows a maximum absorption peak between 210–240 nm, corresponding to -C = O, -COOH, and -CONH₂ groups in its polypeptide chains [33, 34]. A small peak at 280 nm, related to tyrosine and tryptophane, may appear if residual non-collagenous proteins are present [34, 35]. In this study, UV spectra were recorded from 200 to 340 nm for all fish samples, with the maximum absorption observed at 225 nm (Fig. 2b), consistent with previous findings for fish collagen [34, 35] and the absence of a 280 nm peak confirms the removal of non-collagenous proteins and indicates high collagen purity.

Fourier transform infrared spectroscopy (FT-IR)

FT-IR spectra of the collagen extracted from seabass and tilapia skin (Fig. 2c) showed characteristic amide A, B, I, II, and III peaks, confirming the presence of type I collagen. The amide A bands were detected at 3306 cm^-1 for seabass and 3314 cm^-1 for tilapia, which correspond to N-H stretching vibrations [28, 35]. Typically, amide A appears between 3400 and 3440 cm^-1, but shifts to lower frequencies, around 3300 cm^-1, when the N-H group forms hydrogen bonds with carbonyl groups [36]. The amide B bands, found at 2939 cm^-1 for seabass and 2935 cm^-1 for tilapia, are related to asymmetric stretching of CH₂ groups [28, 37]. Both seabass and tilapia showed amide I bands at 1641 cm^-1, which are attributed to C = O stretching along the polypeptide backbone and typically fall within 1600-1700 cm^-1 [36]. Amide II bands, normally observed between 1550 and 1600 cm^-1 [38], were present at 1542 cm^-1 in seabass and 1547 cm^-1 in tilapia. The shift to lower frequencies suggests increased hydrogen bonding, indicating well-preserved collagen activity and a stable triple-helical structure [39]. Amide III bands were recorded at 1239 cm^-1 for both species, and their position around 1240 cm^-1 signals the presence of a triple helix [40]. Both amide II and III bands correspond to N-H bending, C-N stretching, and C-H stretching vibrations, further confirming the structural integrity of the collagen [36, 39, 40]. The absorption ratio of the amide III to the 1454-1450 cm^-1 peak was approximately 1 (1.04 for seabass and 0.95 for tilapia), suggesting that the triple-helix structure of collagen remained intact throughout the extraction process [28]. Preserving the triple-helical structure is essential for type I collagen’s functionality and applications [40]. The extraction conditions in this study successfully maintained this structure, resulting in non-denatured collagen.

Elemental analysis

The EDX analysis (Table S1) revealed that heavy metals such as mercury and lead were absent in the ASC from both seabass and tilapia. In contrast, previous research on collagen derived from jellyfish (Cyanea nozakii) detected trace amounts of mercury and lead [41]. The EDX results showed that carbon content was higher in tilapia skin compared to seabass. Additionally, magnesium, phosphorus, and sulfur were completely undetected in the collagen samples from both seabass and tilapia. Furthermore, calcium and potassium were not present in seabass collagen, which is consistent with findings from earlier studies on silver pomfret (Pampus argenteus) collagen [7].

Characteristics of ZnONPs

As shown in Table S2, the elemental composition of the sample was analyzed and revealing distinct phases of ZnO. Zinc and oxygen displayed weight percentages of 72.02% and 22.67%, respectively, with an atomic percentage of 37.21% for zinc and 47.85% for oxygen, consistent with its stoichiometric formula. The TEM image (Fig. 3a) reveal that the ZnONPs are well-dispersed, spherical, and uniformly sized. Histogram analysis (Fig. 3b) indicates their sizes range from 5 to 13 nm, with an average diameter of 7.95 ± 1.45 nm.

Fig. 3.

Shape of Zinc oxide nanoparticles under Transmission Electron Microscopy (TEM) (a). Scale bar: 10 nm. The white dashed circles represent nanoparticles. The chart of size distribution of ZnONPs (b)

Morphological characterization of bilayer scaffolds

SEM images of the bilayer scaffolds showed that a distinct boundary between the two layers is clearly visible, and the layers are well integrated without delamination, indicating good structural stability (Fig. 4a, d). The PHZnO layers are dense, smooth, and continuous (Fig. 4b, e), while the collagen sponge layers (Fig. 4c, f) exhibit a highly porous, interconnected network with larger pores. These features highlight the complementary functions of the bilayer design (Fig. 4g): the dense PHZnO layer acts as a physical barrier to external contaminants and helps retain moisture for tissue regeneration, while the porous, sponge-like structure efficiently absorbs and retains exudate and blood. Additionally, EDX analysis confirmed the relatively uniform distribution of ZnO nanoparticles within the PHZnO layer (Fig. 4h).

Fig. 4.

Morphological characteristics of the scaffolds. SEM images of the cross-sectional layer at a magnification of 160× (a, d), the PVA/Hsiantsao/ZnONP layer at 4.6 K× (b, e), and the collagen layer at 200× (c, f). Image of a freeze-dried bilayer scaffold (g). EDX image of the distribution of Zn (red dots) (h)

Mechanical properties of bilayer scaffolds

Tensile strength

The toughness of wound dressings is crucial for retaining the moisture necessary for soft tissue healing and maintaining the integrity of the dressing during use. Since wounds differ in etiology and anatomical location, each requires specific mechanical properties to facilitate optimal healing. For skin applications, wound dressings should possess a tensile strength of 2.5–16 MPa and an elongation at break greater than 70% [42]. Intraoral wound healing also demands biomaterials with suitable mechanical strength to protect wounds from the mechanical stresses generated by daily oral activities such as chewing, swallowing, speaking, and sneezing [43]. Therefore, it is crucial to perform tensile tests on the scaffolds [44]. Pure collagen membrane derived from fish exhibit low tensile strength and are mechanically fragile, often being crushed or damaged during tensile testing procedures. This inherent weakness limits their practical application as wound dressings unless further crosslinking or combining with other polymers is employed to enhance their mechanical properties. Therefore, we created the scaffolds with the hydrogel layer contain PVA as a major factor for increasing the mechanical properties of the scaffold. The results of our study (Table 1) show that freeze-dried PHZnO/S and PHZnO/T scaffolds showed high flexibility and tensile strength upon contact with liquid, with values of 4.45 ± 0.45 MPa for scaffolds using ASC from seabass and 4.22 ± 1.30 MPa for those from tilapia. There was no significant difference in tensile strength, and similarly, no significant difference in elongation at break between the two groups.

Table 1.

Mechanical properties of the PHZnO/S and PHZnO/T scaffolds

Sample	Tensile strength (MPa)	Elongation at break (%)
PHZnO/S	4.45 ± 0.45	297.33 ± 10.27
PHZnO/T	4.22 ± 1.30	304.26 ± 72.26

Data are shown as mean ± SE, n = 4.

Water contact angle

Contact angle measurements (Fig. 5a, b) demonstrated that surface wettability is strongly influenced by the structural configuration of the scaffolds, which were designed with two distinct physical characteristics. The inner layer, composed of sponge collagen, is highly suited for direct contact with the wound bed due to its rapid fluid absorption capability, absorbing water within 25 s. This property enables efficient absorption of wound exudate and reduces the risk of bacterial proliferation in stagnant fluid. In contrast, the outer PVA hydrogel layer, characterized by a denser pore structure, is less hydrophilic and helps regulate moisture evaporation. This prevents excessive drying and maintains an optimal environment for tissue regeneration. Upon initial contact with the hydrogel layer samples, the water exhibited a contact angle of 99.5° ± 2.2°. Over time, the contact angle gradually decreased as the droplets spread, reaching approximately 73.5° ± 4.1° after one minute.

Fig. 5.

Contact angle images of water on scaffold surfaces within 60 s after droplet deposition (a). Contact angle values over time for different sample surfaces (b). Swelling ratio (c). Data are shown as mean ± SE, n = 4, *p < 0.05

Swelling ability

A suitable scaffold should have strong swelling capacity to quickly absorb exudate from the wound surface [45]. Superior swelling characteristics help concentrate coagulation factors, promote hemostasis, and enhance the uptake of tissue fluids and metabolic waste [46]. This maintains a moist environment, which supports wound healing and skin regeneration [18]. The swelling behavior of PHZnO/S and PHZnO/T bilayer scaffolds was evaluated over a 24 h period. As shown in Fig. 5c, both PHZnO/S and PHZnO/T scaffolds rapidly increased in swelling upon immersion in water, remaining quite stable thereafter and the equilibrium swelling ratios reach approximate 500%, indicating strong swelling performance. No significant difference in swelling capacity was observed between PHZnO/S and PHZnO/T scaffolds, indicating that the fish source used for collagen extraction does not influence swelling behavior.

pH test

The pH value plays a critical role in the fabrication of scaffolds for biomedical applications, especially in wound healing, as it directly affects their performance at the wound site. Healthy skin has a mildly acidic pH, which supports its barrier function and defends against bacterial invasion. In contrast, an alkaline environment promotes bacterial growth, increasing the risk of infection [47]. Using wound dressings may also change the pH at the wound location [48]. When treating oral diseases, it is important to select materials with a pH level close to that of saliva (5.6–7.9) in order to prevent irritation or damage to oral tissues [49, 50]. Therefore, it is essential to carefully consider the pH of scaffolds when designing them. Moreover, recent studies suggest that pH fluctuations can be utilized to trigger the release and activation of bioactive compounds, potentially enhancing the effectiveness of drug delivery systems [51]. In this study, we monitored the pH changes of the scaffolds over 24 h. The bilayer scaffolds from acid-soluble collagen extracted from both fish species (seabass and tilapia) (Fig. 6) maintained a relatively stable, slightly acidic pH, starting at approximately 6.5 and gradually increasing to around 6.8 after 24 h. Overall, the bilayers exhibited a pH profile with potential for wound healing and intraoral wound care, as well as for inhibiting bacterial growth.

Fig. 6.

The pH value over 24 h. Data are shown as mean ± SE, n = 4

Antioxidant capacity

The antioxidant activity of PHZnO/S and PHZnO/T bilayer scaffolds was assessed using DPPH and ABTS assays (Fig. 7). The results showed that when evaluated using the DPPH assay, both PHZnO/S and PHZnO/T scaffolds exhibited significant antioxidant activity, with scavenging rates of approximately 69% and 70%, respectively (Fig. 7a, b). Additionally, these scaffolds achieved over 80% ABTS radical scavenging (Fig. 7c, d). This pronounced antioxidant effect allows the scaffold to efficiently neutralize free radicals, thereby offering enhanced cellular protection within wound environments. The strong antioxidant performance of the scaffold can be attributed to the bioactive constituents present in Hsiantsao, such as flavonoids, phenols, alkaloids, tannins, and saponins [24]. These compounds are known to against oxidative stress, which is a key factor in the development of chronic wounds. Previous research by Luong et al. [24] further supports these findings, demonstrating that the aqueous extract of Hsiantsao - whether used alone or incorporated into a bilayer scaffold - contains high levels of total flavonoids and phenols and exhibits potent free-radical scavenging activity as measured by DPPH and ABTS assays.

Fig. 7.

Comparison of antioxidant activities of PHZnO/S and PHZnO/T scaffolds and vitamin C (5 µg/mL) by DPPH (%) (a, b) and ABTS (%) (c, d) assays. Data are shown as mean ± SE, n = 3, *p < 0.05, ****p < 0.0001

Antibacterial activity

For antibacterial evaluation, Gram-negative E. coli and Gram-positive S. aureus were chosen as representative microorganisms, as they are among the most common pathogenic bacteria associated with wound infections [52]. Figure 8 and Table 2 illustrate the antibacterial effectiveness of bilayer scaffolds included of PHZnO/S and PHZnO/T, which demonstrated inhibitory activity against both E. coli and S. aureus after 24 h of incubation. There was no significant difference between the inhibition zones of both scaffolds. This suggests that the antibacterial activity comes from the hydrogel layer, specifically the ZnONPs component.

Fig. 8.

Antibacterial activity of the bilayer scaffolds against E. coli (a) and S. aureus (b). Scale bar: 1 cm

Table 2.

Antibacterial of PHZnO/S and PHZnO/T scaffolds against E. coli and S. aureus

Sample	Zone of inhibition (mm)
	E. coli	S. aureus
Penicillin-Streptomycin (0.1 mg/mL)	11.00 ± 0.29	31.83 ± 0.60
Pure PVA	0.00	0.00
PHZnO/S	1.00 ± 0.35	1.80 ± 0.30
PHZnO/T	1.03 ± 0.29	1.67 ± 0.27

Data are shown as mean ± SE, n = 3.

Nanoparticles (NPs) produced through green synthesis often demonstrate superior antibacterial properties compared to those made by physical or chemical methods [53, 54]. This is attributed to the presence of several pharmacologically active biomolecules coating their surfaces, which enable multiple ligand-based conjugation of nanoparticles with receptors on bacterial membranes [53]. These biomolecules - including flavones, aldehydes, ketones, amides, polysaccharides, organic acids, and quinones - are recognized for their substantial therapeutic effects against a wide range of human pathogens [55]. NP-mediated toxicity is influenced by several physiological properties, including shape, size, concentration, and surface defects [53]. There is a direct relationship between NP size and cellular interactions, with smaller NPs interacting more efficiently with cell membranes and exhibiting higher toxicity [56]. Smaller ZnONPs can more effectively penetrate bacterial membranes due to their larger surface area, leading to membrane disruption and cell death [53, 57]. Additionally, the dissolution of ZnONPs into Zn²⁺ ions is size-dependent, and some studies suggest that this ion release contributes to the observed toxicity of ZnONPs [57]. The shape of NPs also plays a significant role in their interaction with bacterial membranes. For example, spherical ZnONPs release Zn²⁺ ions more effectively than rod-shaped ones, further influencing their antibacterial performance [58].

Cell viability

To preserve wound moisture, hygiene, and facilitate drainage at the wound site, conventional dressings in hospitals are typically replaced every 12 to 48 h [59]. Therefore, this experiment was designed to evaluate whether the scaffolds induce any cytotoxic effects on cells within 48 h. In addition to exhibiting in vitro biocompatibility, the bilayer scaffolds containing a sponge layer must also maintain their structural integrity to support cell attachment and proliferation. To ensure that the sterilization process did not affect scaffold structure, various methods were considered. UV irradiation was selected because it does not significantly affect the integrity of collagen scaffolds, including morphology, porosity, degradation properties, or cellular attachment and proliferation [60]. In contrast, other methods such as treatment with 70% ethanol induce protein denaturation and dehydration, altering scaffold structure and cellular interactions, while gamma irradiation disrupts the microstructure by altering collagen chemistry and breaking hydrogen bonds [61]. To assess the cytotoxicity of the PHZnO/S and PHZnO/T scaffolds, an in vitro assay was performed using NIH/3T3 cells and the CCK-8 reagent to evaluate cell viability. NIH/3T3 cells exhibited robust growth on the scaffolds after 48 h of incubation (Fig. 9). The results showed that the scaffolds are non-toxic, with cell viability exceeding 100% for both scaffolds fabricated from collagen derived from two fish species (116.9 ± 7.4% for PHZnO/S and 115.2 ± 2.4% for PHZnO/T, respectively).

Fig. 9.

Cell viability of NIH/3T3 cells seeded on the bilayer scaffolds. Data are shown as mean ± SE, n = 3, p < 0.05

Hemolysis study

Because the scaffolds are designed for direct wound contact, in vitro blood compatibility is the foremost prerequisite for their use in medical applications, as it reflects how much the material causes red blood cell rupture [24]. As illustrated in Fig. 10, both PHZnO/S and PHZnO/T scaffolds exhibited very low hemolysis rates ( <1.5%), which comply with the requirements ( < 2%) of the International Organization for Standardization 10993-4 for non-hemolytic materials [62]. This demonstrates that the scaffolds possess excellent biocompatibility, making them suitable for wound treatment and supportive of the hemostatic phase of the healing process.

Fig. 10.

Hemolysis ratio of the PHZnO/S and PHZnO/T scaffolds. Data are shown as mean ± SE, n = 3, p < 0.05

Discussion

Modern wound dressings are designed to protect against contamination and regulate exudate to foster a moist healing environment. Selecting the appropriate dressing based on wound size, location, and exudate level is critical, as physico-chemical properties vary widely among commercial products. For instance, fiber-based dressings (e.g., Kaltostat^®, Aquacel Ag^®) are preferred for high-exudate burns due to superior absorbency, whereas flexible film dressings (e.g., DuoDERM^® ET) suit mobile anatomical sites. Foam dressings (e.g., Mepilex^®) balance moderate absorbency with mechanical strength to protect pressure-sensitive areas. While multi-layered dressings generally offer superior mechanical integrity, they often fail to match the high swelling capacity of mono-layered alternatives [63]. This highlights a critical need for advanced dressings that simultaneously offer robust mechanical strength and high absorption while ensuring optimal cytocompatibility. In this study, our bilayer scaffolds demonstrated moderate mechanical strength within the recommended range alongside strong swelling performance, attributed to their highly porous structure.

Regarding infection control, silver nanoparticles are frequently incorporated into commercial dressings (e.g., Mepilex^® Ag, Acticoat™). However, in vitro studies indicate that many silver-containing dressings can decrease cell viability and inhibit proliferation [64, 65]. Exceptions exist, such as PROMOGRAN PRISMA^®, a bovine collagen/ORC matrix that achieves a critical therapeutic balance by delivering silver levels sufficient to eradicate wound pathogens while maintaining cytocompatibility with cells, particularly dermal fibroblasts, thereby promoting optimal healing [66]. However, careful consideration of the collagen source is essential to ensure widespread applicability and avoid religious, cultural, or safety constraints. Our study overcomes this limitation by utilizing aquatic collagen extracted from fish waste. This sustainable approach not only mitigates risks but also valorizes waste products, offering a cost-effective solution that reduces the financial burden on patients. Furthermore, instead of silver, we incorporated green-synthesized ZnONPs, which demonstrated antibacterial activity against E. coli and S. aureus within 24 h while maintaining excellent cytocompatibility with NIH/3T3 fibroblasts over 48 h. Additionally, both scaffolds showed excellent hemocompatibility, creating a safe interface that maintains the viability of red blood cells.

The synergistic combination of antibacterial and antioxidant properties is a key feature of advanced commercial wound healing products. A prime example is L-Mesitran^® Soft, which utilizes medical-grade honey as its primary antibacterial agent. This formulation is further fortified with Vitamins C and E, which function as potent antioxidants to protect against oxidative stress. Together, these components fulfill a synergistic activity that significantly enhances the pro-healing effects of the product [67]. Aligning with this dual-action strategy, our scaffolds incorporate Hsiantsao extract, a natural ingredient exhibiting significant antioxidant potential. The inclusion of Hsiantsao imparts intrinsic antioxidant activity to the scaffold, and as demonstrated in previous studies, the release of its bioactive compounds from the matrix can further improve therapeutic outcomes.

Overall, this study primarily assessed their effects at the in vitro level. Comprehensive impacts, such as in vivo performance and underlying molecular mechanisms, remain to be investigated. Further analyses are necessary to provide stronger evidence for the efficacy of PHZnO/S and PHZnO/T scaffolds in healthcare applications.

Conclusion

This study developed and showcased the potential of PHZnO/S and PHZnO/T bilayer scaffolds for biomedical applications. By utilizing eco-friendly collagen derived from byproducts, together with Hsiantsao aqueous extract and green-synthesized ZnO nanoparticles, the fabricated scaffolds exhibited notable biocompatibility. Furthermore, the beneficial properties of each ingredient were harnessed to create multifunctional scaffolds: the ZnO nanoparticles imparted antimicrobial properties, whereas Hsiantsao extract offered significant antioxidant benefits to support treatment. The synergistic combination of these components enabled the scaffolds to serve as promising candidates for biomedical applications, including skin wound care and intraoral care.

Compliance with ethical standards

Conflict of interest

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1007/s10856-026-07015-4.