Effect of Poly (Lactic Acid/ε-Caprolactone) Bilayer Membrane on Tooth Extraction Socket Wound Healing in a Rat Model

1. Introduction

Maintaining sufficient alveolar ridge height and width is critical for preserving natural teeth, enabling successful prosthetic rehabilitation, and ensuring safe implant placement [1]. However, the alveolar bone frequently undergoes significant resorption following periodontal disease progression, prolonged use of ill-fitting dentures, or extended periods after tooth extraction without intervention. This bone loss complicates prosthetic treatment and restricts the options for implant placement [2].

Once resorbed, the alveolar bone does not regenerate spontaneously. To address this, various approaches, including autologous bone grafts and the application of bone sub-stitutes, have been developed to reconstruct alveolar defects [3]. Nevertheless, during the bone regeneration process, the infiltration of soft tissue into the defect site can inhibit osteogenesis. This is largely because fibroblasts (FBs), derived from soft tissues, proliferate more rapidly than osteoblasts [4,5]. Consequently, preventing soft tissue invasion and preserving the regenerative space are critical prerequisites for successful bone aug-mentation [6,7].

Barrier membranes, a central component of the guided tissue regeneration (GTR) and guided bone regeneration (GBR) techniques, have been developed to selectively guide cellular repopulation and promote tissue-specific regeneration. GTR was first introduced in 1982 by Nyman et al. to facilitate periodontal regeneration by creating a space for periodontal ligament-derived cells to migrate, while GBR, introduced in 1988, specifically targets alveolar bone defects by encouraging osteoblast migration and simultaneously preventing soft tissue ingrowth [8,9,10,11].

Membranes used in GTR and GBR are broadly classified into non-resorbable and re-sorbable types [12,13]. Non-resorbable membranes, such as expanded polytetrafluoro-ethylene and titanium mesh, provide excellent space maintenance but require a second surgery for removal, which can cause patient discomfort and risks damaging newly formed tissues [14,15]. By contrast, resorbable membranes—primarily made of collagen or synthetic polymers like poly(lactic-co-glycolic acid) (PLGA)—biodegrade over time, eliminating the need for retrieval surgery and reducing patient morbidity [16,17,18,19].

Despite their advantages, conventional resorbable membranes face challenges, including uncontrolled degradation rates, insufficient mechanical strength, and potential immu-nogenicity associated with animal-derived materials. Therefore, the development of synthetic, biocompatible membranes with predictable resorption profiles has gained significant attention [20,21,22].

Recently, a novel bilayer resorbable membrane composed of poly(L-lactic ac-id-co-ε-caprolactone) (P(LA/CL); PBM) has been introduced. PBM features a dual-layer structure: a smooth, solid outer layer designed to prevent soft tissue invasion and maintain space, and a porous inner layer that promotes tissue integration and supports osteogenesis. Furthermore, its slow degradation profile and elasticity make it suitable for large defect sites requiring prolonged healing periods [23,24,25].

In this study, we aimed to evaluate the wound healing effects and biological performance of PBM compared to conventional PLGA membranes, focusing on soft tissue healing, bone regeneration, and cell–membrane interactions through comprehensive in vivo and in vitro analyses.

2. Materials and Methods

2.1. Materials

Two types of commercially available resorbable membranes were used in this study: a PLGA membrane (GC Membrane, GC Corporation, Tokyo, Japan) and a bilayer membrane composed of P(LA/CL) (PBM; Cytrans Elashield, GC Corporation). Both membranes were utilized in in vitro and in vivo experiments.

2.2. Cell Culture

Oral epithelial cells (OECs) and FBs were isolated from rat oral mucosa [26]. Cells were seeded onto either the surface or the reverse side of each membrane at a density of 2 × 10⁴ cells/membrane. For seeding, a custom-designed titanium apparatus was used (Figure 1). This device consisted of two parts: a cylindrical base (part a) supporting the membrane and a donut-shaped cover (part b), which is secured to the base by tightening side screws using a small hex wrench. This mechanism firmly locks the membrane in place, pre-venting displacement and ensuring that cells cannot pass underneath during culture. Cells were cultured under standard conditions (37°C, 5% CO₂) for 5 days prior to further analysis.

2.3. Scanning Electron Microscopy (SEM)

After 5 days of culture, cell morphology on the membranes was examined by SEM. Samples were fixed with 2.5% glutaraldehyde, dehydrated through a graded ethanol series, dried, sputter-coated with gold, and observed. In parallel, for quantification of adherent cells, membranes were treated with TrypLE Express (Thermo Fisher Scientific, MA, USA) to detach the cells, and the number of cells was counted using an automated cell counter (TC20, Bio-Rad, Tokyo, Japan) [27].

2.4. Membrane Permeability Test

The permeability of membranes was assessed using a modified titanium apparatus de-signed to hold the membranes in place (Figure 2a). Each membrane (PLGA and PBM surfaces, PBM reverse side, or filter paper as control) was mounted on the device, which was submerged in distilled water within a 12-well culture plate. Toluidine blue dye or fetal bovine serum (FBS) was applied onto the top of the membranes. After incubation, the degree of penetration was evaluated by determining the absorbance of the underlying water using a spectrophotometer.

2.5. Establishment of the Rat Membrane Model

Sixty male Wistar rats (6 weeks old, weight 130–160 g) were used for the in vivo model (Kyudo Co.) [28]. Under general anesthesia, the maxillary right first and second molars were extracted. After extraction, the septum between the extraction sockets was removed using a round bur (CA φ1.0, Dentech, Tokyo, Japan). Each extraction socket was covered with either a PLGA or a PBM membrane. The membranes were sutured to the sur-rounding mucosa using 4-0 silk thread (two stitches per site) (Figure 3a). Surgery was performed under systemic anesthesia (0.3 mg/kg medetomidine, 4.0 mg/kg midazolam, and 5.0 mg/kg butorphanol), and the animals were maintained in accordance with the guidelines approved by the Animal Care and Use Committee of Kyushu University (Approval Number: A25-240-0). Sutures were removed 3 days postoperatively.

2.6. Euthanasia and Sample Preparation

The experimental schedule is shown in Figure 3b, at 1-4 and 8 weeks post-surgery, animals were euthanized by anesthesia overdose. Phosphate-buffered saline perfusion followed by Zamboni’s fixative perfusion was performed. The maxillae were harvested for further analysis. For macroscopic evaluation, specimens were observed using a ste-reomicroscope. For histological analysis, the specimens were trimmed to preserve the area of interest, fixed in 4% paraformaldehyde for 48h, and subsequently decalcified in Kalkitox™ solution (Wako, Osaka, Japan) at 4°C for 24 hours. After decalcification, samples were embedded in Optimal Cutting Temperature compound (Sakura Finetek, Tokyo, Japan) and cryosectioned at a thickness of 10 μm using a cryostat .

2.7. Micro-Computed Tomography (Micro-CT) Analysis

For radiographic assessment, micro-CT scanning was performed (SkyScan 1076, Bruker, Kontich, Belgium) at 49 kV tube voltage and 201 μA current, with a pixel size of 18 μm. Three-dimensional image reconstruction and bone volume quantification within the extraction sockets were performed using CTAn software (Bruker microCT) [29].

2.8. Histological Analysis

Frozen sections were stained with hematoxylin and eosin and observed by light mi-croscopy to evaluate soft tissue healing and new bone formation beneath the membranes.

2.9. Statistical Analysis

Quantitative data were expressed as mean ± standard deviation. Statistical comparisons between groups were performed using one-way analysis of variance followed by Tukey’s post hoc test for multiple comparisons. A p-value of < 0.05 was considered to be statistically significant.

3. Results

3.1. Comparison of OEC and FB Adhesion on PBM Membrane Surfaces

Using a custom-designed titanium culture apparatus (Figure 1), OECs and FBs were cultured separately on the surface and reverse sides of the PBM membrane. After 5 days of culture, OECs displayed extensive proliferation on the surface side of the PBM membrane, whereas only minimal adhesion was observed on the reverse side (Figure 4a). Higher magnification SEM images confirmed numerous well-spread epithelial cells with extended lamellipodia on the surface side, whereas few cells were present on the reverse side (Figure 4b). No notable differences in FB adhesion were observed between the two sides of the PBM membrane.

3.2. Permeability Differences Between Membranes

Toluidine blue dye and FBS permeability tests were conducted to evaluate membrane barrier properties. The PBM surface and PLGA effectively blocked dye and serum penetration (Figure 2b and c). By contrast, the reverse side of the PBM membrane exhibited increased permeability over time, similar to that of the filter paper control.

3.3. Mucosal Wound Healing over Membranes

Gross examination showed progressive mucosal closure in both the PBM and PLGA groups (Figure 3c). By 3 weeks postoperatively, complete mucosal closure was observed in almost all cases in both groups. Quantitative analysis revealed no statistically significant difference in the rate of mucosal closure between the two groups (Figure 3d).

3.4. Bone Formation Beneath the Membranes

Histological analysis at 4 weeks demonstrated differences in the spatial relationship between the membrane and new bone (Figure 5a). In the PBM group, newly formed bone appeared to extend along the membrane. In the PLGA group, a concave defect area was observed beneath the membrane. Micro-CT analysis confirmed these observations (Figure 5b, 5c), bone thickness was significantly higher in the PBM group than in the PLGA. As shown in Figure 6a, newly formed bone in the PBM group was observed in direct contact with the underside of the membrane. By contrast, the PLGA membrane was surrounded predominantly by connective tissue, with no direct bone contact.

4. Discussion

Cell adhesion experiments were performed only on the PBM membrane using a custom titanium apparatus. OECs displayed robust adhesion and spreading with extended lamellipodia on the surface side of PBM, whereas minimal attachment was observed on the reverse side. This indicated that the PBM’s surface side offers a favorable environment for epithelial cell adhesion, likely due to surface topography or hydrophilic properties. Rapid epithelial coverage is critical to prevent postoperative contamination and may support improved early-stage soft tissue healing [30,31]. The lack of adhesion on the reverse side also reflects the directional nature of PBM, emphasizing the importance of correct membrane orientation during surgical placement.

In the permeability assay using toluidine blue and FBS, both, the PBM and PLGA surface sides effectively prevented dye permeation. However, the reverse side of PBM exhibited high permeability comparable to filter paper. This finding underscores a functional limitation of the bilayer PBM design—although the solid surface layer provides strong barrier function, the porous reverse layer may allow undesirable molecular or cellular infiltration if placed incorrectly [32,33]. PLGA, by contrast, is a non-layered uniform membrane, and although permeability testing was limited to one side, it displayed consistent barrier performance. These differences highlight the need for precise membrane orientation when using bilayer membranes like PBM in GBR procedures.

Gross examination showed no statistically significant difference in the mucosal healing rate between the PBM and PLGA groups, with both achieving near-complete closure by week three. These results suggested that both materials exhibit adequate biocompatibility to support soft tissue closure. However, considering PBM’s favorable epithelial cell adhesion in vitro, it may offer advantages in more challenging healing environments or in mucosa-compromised patients [34].

Histological and micro-CT evaluation revealed that the PBM group exhibited a larger amount of new bone formation within the extraction sockets compared to the PLGA group. In the PBM group, bone tissue filled the defect space more thoroughly, and a clear boundary was maintained between the membrane and the regenerating bone [35]. This suggested that PBM functions effectively as a space maintainer, promoting osteogenesis while preventing collapse or early soft tissue invasion [36,37,38].

Notably, as shown in Figure 6, new bone was observed in direct contact with the underside of the PBM membrane, suggesting that the membrane surface provided a favorable scaffold for osteogenesis. By contrast, in the PLGA group, the membrane was frequently observed in close contact with connective tissue, and bone regeneration appeared less robust. This may reflect the PLGA membrane’s tendency to integrate with soft tissue earlier, leading to partial loss of regenerative space during the healing process [39]. Taken together, these findings suggested that PBM’s bilayer structure—combining a solid barrier side with a tissue-friendly porous layer—may create an optimal environment for bone healing, provided that its orientation is correct [40].

The study results suggested that PBM offers several clinical advantages:

• Enhanced epithelial compatibility for early soft tissue sealing
• Effective space maintenance for bone regeneration
• Synthetic and fully resorbable, free from animal-derived components

However, the performance of PBM is strongly orientation-dependent. Incorrect placement could result in reverse-side exposure to soft tissues, leading to increased permeability and reduced barrier function. By contrast, PLGA membranes, though simpler and less supportive of epithelial attachment, offer a consistent barrier function regardless of placement direction, and may be preferable in cases where membrane orientation is difficult to maintain.

This study has several limitations. First, it used a small animal model with a limited bone defect size and rapid healing capacity, which may not replicate the challenges of human clinical GBR procedures. Second, only short-term healing (up to 4 weeks) was assessed. Long-term membrane degradation, integration, and bone quality outcomes remain unclear.

Future studies should:

• Evaluate PBM performance in larger animal models with more complex defects
• Investigate long-term space maintenance and bone maturation
• Consider design improvements to enhance the barrier function of the reverse side

5. Conclusions

The synthetic bilayer PBM membrane exhibited favorable characteristics for maintaining bone-soft tissue separation and promoting wound healing. These results suggested that PBM may serve as an effective alternative to conventional resorbable membranes for GBR procedures. Further investigations are required to elucidate its long-term clinical applicability.