Procurement of oral mucosa samples

The protocol for obtaining human oral mucosa samples had internal review board approval. Patients who had been subjected to 3rd molar removal were given sufficient information regarding this study, and all individuals (total 16: 6 males and 10 females, with a mean age of 23.6 years) signed an informed consent form.

Primary oral keratinocyte and fibroblast cultures

Keratinized oral mucosa was harvested from the area adjacent to the site of tooth extraction without causing any morbidity. The mucosal sample was transported in a 15-ml conical tube containing 5 ml of a basic keratinocyte culture medium, EpiLife (Life Technologies, Carlsbad, CA, USA). Primary oral keratinocyte cultures were established, and cells were serially passaged as described previously. The majority of cells used in this study were 4th and 5th passage, with the range of passages used being 3rd to 7th.

Oral fibroblast cultures were established from the underlying connective tissue after the epithelial layer was scraped off. Primary fibroblasts were isolated from those mucosa samples by an explant culture technique. The connective tissue was minced into 2-mm ² explants. These were placed in a 60-mm Petri dish (Corning, New York, NY, USA) and incubated in a moist atmosphere of 5% CO ₂ at 37 °C in a Dulbecco’s modified Eagle’s medium (DMEM) culture medium (Wako Chemical, Osaka, Japan) supplemented with 10% foetal bovine serum (FBS) (Nichirei, Tokyo, Japan), gentamicin (5.0 μg/ml; Life Technologies), and amphotericin B (0.375 μg/ml; Life Technologies). Cells were fed every other day. Once the cell outgrowth was sufficient, cells were detached with a 0.025% trypsin/ethylenediaminetetraacetic acid (EDTA) solution (Life Technologies), neutralized with a 0.0125% defined trypsin-inhibitor solution (DTI) (Life Technologies), then re-plated in another culture vessel (0.8 × 10 ⁴ cells/cm ² ). The 3rd–5th passaged cells were used in this study.

Fabrication of a 3D, full-thickness living oral mucosa equivalent (LOME) ( n = 10)

The LOME, consisting of a stratified squamous epithelium (oral keratinocytes) grown on human cadaver dermis in which oral fibroblasts were repopulated, was reconstructed as reported previously elsewhere. Briefly, an equal number of oral keratinocytes and oral fibroblasts (1.5 × 10 ⁵ ) was resuspended in EpiLife supplemented with EpiLife defined growth supplement and 1.2 mM Ca ²⁺ . The cell suspension was seeded onto a papillary surface of human cadaver dermis, AlloDerm (LifeCell, Branchburg, NJ, USA), which had been cut into circular pieces that fit in a 48-well microplate (1 cm ² ), and pre-soaked with human type IV collagen (5 μg/cm ² ; Sigma-Aldrich, St. Louis, MO, USA) overnight at 4 °C. After the LOME had been cultured in a submerged condition for 5 days with the medium changed every 24 h, it was raised to an air–liquid interface and cultured for another 3 days to allow oral keratinocytes to differentiate and stratify with the same culture medium. The medium was changed every 3 days.

Ex vivo wound healing assay

After culturing the LOME for 8 days, the middle of the epithelial surface of the LOME was wounded using a 2-mm-wide linear excision tool. The 2-mm-wide strip composed of an epithelial layer and the underlying AlloDerm was excised using sterile scissors. This resulted in a linear, partial thickness mucosa defect. ZA (Zometa) was purchased from Novartis Pharmaceuticals Corp. (East Hanover, NJ, USA), dissolved in sterile ddH ₂ O and used at a final concentration of 10 μM. To examine the effects of ZA on subsequent epithelial regeneration and re-epithelialization over the wounded area on the LOME, 10 μM of ZA was added to the culture medium and the LOME cultured for an additional 11 days. As a control, no ZA was added.

Histological and immunohistochemical examinations

ZA-treated as well as untreated LOMEs that had been cultured for a total of 19 days, as well as native oral mucosa tissue not used for cell culture, were fixed with 4% paraformaldehyde (PFA) in 0.1 M PBS and embedded in paraffin. Paraffin sections of 4.5 μm in thickness were stained with haematoxylin and eosin for histological examination. A part of the tissue sample fixed with PFA was incubated with 30% sucrose overnight, embedded in OCT compound (Sakura Co. Ltd., Tokyo, Japan) and stored at −80 °C. For immunohistochemistry, after quenching the endogenous peroxidase activity with 0.3% hydrogen peroxide in absolute methanol, heat-induced epitope retrieval (HIER) was performed in 10 mM citric acid buffer, pH 6.0, for 20 min. After blocking with 5% normal goat serum for 1 h at room temperature (RT), sections were incubated with one of the following antibodies in a humidified chamber overnight at 4 °C. For integrin αvβ6, sections 6 μm in thickness were made on a cryostat and immunostained as mentioned above without HIER. Primary antibodies and dilutions used in this study included: laminin (1:80; PROGEN Biotechnik GmbH, Heidelberg, Germany), Ki-67 (1:100; Dako Japan, Tokyo, Japan), TGF-β1 (1:100; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), TGF-β receptor I (1:50; Abcam, Cambridge, UK), TGF-β receptor II (1:100; Biorbyt, Cambridge, UK), p-Smad2 (1:2000; Millipore, Burlingame, CA, USA), p-Smad3 (1:50; Abcam), and integrin αvβ6 (1:100; Millipore). As a control, the primary antibody for the respective specimen was omitted. The slides were stained with Envision+ System horseradish peroxidase (HRP)-labelled anti-mouse or anti-rabbit polymers (Dako Japan) for 90 min and visualized with DAB (Liquid DAB+ Substrate System; Dako Japan) within 2 min at RT. Sections were counter-stained with methylene blue.

Quantifications of epithelial thickness, Ki-67 labelling index, and re-epithelialization

The slides were all photographed and examined using an Axioplan 2 imaging microscope (Carl Zeiss, Göttingen, Germany) at different magnifications through an AxioCam HRc digital camera (Carl Zeiss) connected to a computer. Using the AxioVision software, the following three parameters were evaluated: (1) to determine the thickness of the viable epithelial layer in the unwounded area, the linear distance perpendicular to AlloDerm between the basal surface of basal cells and the keratinized layer was randomly measured at 10 points per one section and the mean of 10 values was assigned to the thickness of the sample. (2) To analyze the Ki-67 labelling index (LI) in an unwounded area, the total number of basal cells of the individual LOMEs and the Ki-67-immunolabelled basal keratinocytes were counted separately at a magnification of 200× by a single trained examiner. The LI of Ki-67 in the unwounded area was calculated by dividing the number of immunopositive cells by the total number of basal keratinocytes, and multiplying by 100. (3) To determine the length of re-epithelialization, the laminin-positive sections ( n = 10) were analyzed by two examiners in a blinded fashion. After the wound edge, where the continuous laminin immunoreaction seen on AlloDerm was disrupted, and the tip of the newly formed epithelium were determined, the linear length of the two points was measured. The mean value was assigned to the length of the sample.

Enzyme-linked immunosorbent assay (ELISA)

For quantification of TGF-β1 in the conditioned media during manufacturing of the LOME, ELISA reagents were used in accordance with the manufacturer’s specifications (Quantikine ELISA; R&D systems, Minneapolis, MN, USA).

Immunocytochemistry

Oral keratinocytes were seeded in chamber slides (Iwaki, Co. Ltd., Tokyo, Japan), cultured until 70% confluent, and incubated with or without 10 μM of ZA for 72 h before they were fixed with 4% paraformaldehyde in PBS for 20 min. They were then permeabilized in PBS with 0.3% Triton X-100 for 10 min, rinsed several times with PBS, and incubated with primary antibodies overnight at 4 °C. Alexa 594-conjugated secondary antibodies against mouse or rabbit immunoglobulins (Life Technologies) were applied for 60 min at RT. After removing the chamber, glass coverslips were mounted onto slides in a mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) (Vectashield; Vector Laboratories Inc., Burlingame, CA, USA), and cells were observed by epifluorescence with an Axioplan Zeiss microscope coupled to a Coolsnap Fx Camera (Roper Scientific, Evry, France).

Statistical analysis

Data are all presented as the mean ± standard deviation (SD). For all pairwise comparisons, a paired Student’s t -test was performed. A difference of P < 0.05 was considered significant.

Procurement of oral mucosa samples

The protocol for obtaining human oral mucosa samples had internal review board approval. Patients who had been subjected to 3rd molar removal were given sufficient information regarding this study, and all individuals (total 16: 6 males and 10 females, with a mean age of 23.6 years) signed an informed consent form.

Primary oral keratinocyte and fibroblast cultures

Keratinized oral mucosa was harvested from the area adjacent to the site of tooth extraction without causing any morbidity. The mucosal sample was transported in a 15-ml conical tube containing 5 ml of a basic keratinocyte culture medium, EpiLife (Life Technologies, Carlsbad, CA, USA). Primary oral keratinocyte cultures were established, and cells were serially passaged as described previously. The majority of cells used in this study were 4th and 5th passage, with the range of passages used being 3rd to 7th.

Oral fibroblast cultures were established from the underlying connective tissue after the epithelial layer was scraped off. Primary fibroblasts were isolated from those mucosa samples by an explant culture technique. The connective tissue was minced into 2-mm ² explants. These were placed in a 60-mm Petri dish (Corning, New York, NY, USA) and incubated in a moist atmosphere of 5% CO ₂ at 37 °C in a Dulbecco’s modified Eagle’s medium (DMEM) culture medium (Wako Chemical, Osaka, Japan) supplemented with 10% foetal bovine serum (FBS) (Nichirei, Tokyo, Japan), gentamicin (5.0 μg/ml; Life Technologies), and amphotericin B (0.375 μg/ml; Life Technologies). Cells were fed every other day. Once the cell outgrowth was sufficient, cells were detached with a 0.025% trypsin/ethylenediaminetetraacetic acid (EDTA) solution (Life Technologies), neutralized with a 0.0125% defined trypsin-inhibitor solution (DTI) (Life Technologies), then re-plated in another culture vessel (0.8 × 10 ⁴ cells/cm ² ). The 3rd–5th passaged cells were used in this study.

Fabrication of a 3D, full-thickness living oral mucosa equivalent (LOME) ( n = 10)

The LOME, consisting of a stratified squamous epithelium (oral keratinocytes) grown on human cadaver dermis in which oral fibroblasts were repopulated, was reconstructed as reported previously elsewhere. Briefly, an equal number of oral keratinocytes and oral fibroblasts (1.5 × 10 ⁵ ) was resuspended in EpiLife supplemented with EpiLife defined growth supplement and 1.2 mM Ca ²⁺ . The cell suspension was seeded onto a papillary surface of human cadaver dermis, AlloDerm (LifeCell, Branchburg, NJ, USA), which had been cut into circular pieces that fit in a 48-well microplate (1 cm ² ), and pre-soaked with human type IV collagen (5 μg/cm ² ; Sigma-Aldrich, St. Louis, MO, USA) overnight at 4 °C. After the LOME had been cultured in a submerged condition for 5 days with the medium changed every 24 h, it was raised to an air–liquid interface and cultured for another 3 days to allow oral keratinocytes to differentiate and stratify with the same culture medium. The medium was changed every 3 days.

Ex vivo wound healing assay

After culturing the LOME for 8 days, the middle of the epithelial surface of the LOME was wounded using a 2-mm-wide linear excision tool. The 2-mm-wide strip composed of an epithelial layer and the underlying AlloDerm was excised using sterile scissors. This resulted in a linear, partial thickness mucosa defect. ZA (Zometa) was purchased from Novartis Pharmaceuticals Corp. (East Hanover, NJ, USA), dissolved in sterile ddH ₂ O and used at a final concentration of 10 μM. To examine the effects of ZA on subsequent epithelial regeneration and re-epithelialization over the wounded area on the LOME, 10 μM of ZA was added to the culture medium and the LOME cultured for an additional 11 days. As a control, no ZA was added.

Histological and immunohistochemical examinations

ZA-treated as well as untreated LOMEs that had been cultured for a total of 19 days, as well as native oral mucosa tissue not used for cell culture, were fixed with 4% paraformaldehyde (PFA) in 0.1 M PBS and embedded in paraffin. Paraffin sections of 4.5 μm in thickness were stained with haematoxylin and eosin for histological examination. A part of the tissue sample fixed with PFA was incubated with 30% sucrose overnight, embedded in OCT compound (Sakura Co. Ltd., Tokyo, Japan) and stored at −80 °C. For immunohistochemistry, after quenching the endogenous peroxidase activity with 0.3% hydrogen peroxide in absolute methanol, heat-induced epitope retrieval (HIER) was performed in 10 mM citric acid buffer, pH 6.0, for 20 min. After blocking with 5% normal goat serum for 1 h at room temperature (RT), sections were incubated with one of the following antibodies in a humidified chamber overnight at 4 °C. For integrin αvβ6, sections 6 μm in thickness were made on a cryostat and immunostained as mentioned above without HIER. Primary antibodies and dilutions used in this study included: laminin (1:80; PROGEN Biotechnik GmbH, Heidelberg, Germany), Ki-67 (1:100; Dako Japan, Tokyo, Japan), TGF-β1 (1:100; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), TGF-β receptor I (1:50; Abcam, Cambridge, UK), TGF-β receptor II (1:100; Biorbyt, Cambridge, UK), p-Smad2 (1:2000; Millipore, Burlingame, CA, USA), p-Smad3 (1:50; Abcam), and integrin αvβ6 (1:100; Millipore). As a control, the primary antibody for the respective specimen was omitted. The slides were stained with Envision+ System horseradish peroxidase (HRP)-labelled anti-mouse or anti-rabbit polymers (Dako Japan) for 90 min and visualized with DAB (Liquid DAB+ Substrate System; Dako Japan) within 2 min at RT. Sections were counter-stained with methylene blue.

Quantifications of epithelial thickness, Ki-67 labelling index, and re-epithelialization

The slides were all photographed and examined using an Axioplan 2 imaging microscope (Carl Zeiss, Göttingen, Germany) at different magnifications through an AxioCam HRc digital camera (Carl Zeiss) connected to a computer. Using the AxioVision software, the following three parameters were evaluated: (1) to determine the thickness of the viable epithelial layer in the unwounded area, the linear distance perpendicular to AlloDerm between the basal surface of basal cells and the keratinized layer was randomly measured at 10 points per one section and the mean of 10 values was assigned to the thickness of the sample. (2) To analyze the Ki-67 labelling index (LI) in an unwounded area, the total number of basal cells of the individual LOMEs and the Ki-67-immunolabelled basal keratinocytes were counted separately at a magnification of 200× by a single trained examiner. The LI of Ki-67 in the unwounded area was calculated by dividing the number of immunopositive cells by the total number of basal keratinocytes, and multiplying by 100. (3) To determine the length of re-epithelialization, the laminin-positive sections ( n = 10) were analyzed by two examiners in a blinded fashion. After the wound edge, where the continuous laminin immunoreaction seen on AlloDerm was disrupted, and the tip of the newly formed epithelium were determined, the linear length of the two points was measured. The mean value was assigned to the length of the sample.

Enzyme-linked immunosorbent assay (ELISA)

For quantification of TGF-β1 in the conditioned media during manufacturing of the LOME, ELISA reagents were used in accordance with the manufacturer’s specifications (Quantikine ELISA; R&D systems, Minneapolis, MN, USA).

Immunocytochemistry

Oral keratinocytes were seeded in chamber slides (Iwaki, Co. Ltd., Tokyo, Japan), cultured until 70% confluent, and incubated with or without 10 μM of ZA for 72 h before they were fixed with 4% paraformaldehyde in PBS for 20 min. They were then permeabilized in PBS with 0.3% Triton X-100 for 10 min, rinsed several times with PBS, and incubated with primary antibodies overnight at 4 °C. Alexa 594-conjugated secondary antibodies against mouse or rabbit immunoglobulins (Life Technologies) were applied for 60 min at RT. After removing the chamber, glass coverslips were mounted onto slides in a mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) (Vectashield; Vector Laboratories Inc., Burlingame, CA, USA), and cells were observed by epifluorescence with an Axioplan Zeiss microscope coupled to a Coolsnap Fx Camera (Roper Scientific, Evry, France).

Statistical analysis

Data are all presented as the mean ± standard deviation (SD). For all pairwise comparisons, a paired Student’s t -test was performed. A difference of P < 0.05 was considered significant.