Abstract
The aim of this study was to evaluate the suitability of tissue-engineered mucosa (TEM) as a model for studying the acute effects of ionizing radiation (IR) on the oral mucosa. TEM and native non-keratinizing oral mucosa (NNOM) were exposed to a single dose of 16.5 Gy and harvested at 1, 6, 24, 48, and 72 h post-irradiation. DNA damage induced by IR was determined using p53 binding protein 1 (53BP1), and DNA repair was determined using Rad51. Various components of the epithelial layer, basement membrane, and underlying connective tissue were analyzed using immunohistochemistry. The expression of cytokines interleukin-1β (IL-1β) and transforming growth factor beta 1 (TGF-β1) was analyzed using an enzyme-linked immunosorbent assay. The expression of DNA damage protein 53BP1 and repair protein Rad51 were increased post-irradiation. The expression of keratin 19, vimentin, collage type IV, desmoglein 3, and integrins α6 and β4 was altered post-irradiation. Proliferation significantly decreased at 24, 48, and 72 h post-irradiation in both NNOM and TEM. IR increased the secretion of IL-1β, whereas TGF-β1 secretion was not altered. All observed IR-induced alterations in TEM were also observed in NNOM. Based on the similar response of TEM and NNOM to IR we consider our TEM construct a suitable model to quantify the acute biological effects of IR.
The field of tissue engineering of mucosal equivalents has developed significantly during the last 10 years. We and others have reported the development of tissue-engineered mucosal constructs using autologous keratinocytes seeded onto biomaterials, whether or not repopulated with fibroblasts. Tissue-engineered mucosal constructs (TEM) have been successfully used to study the dynamics of wound healing and to test cytotoxicity and working mechanisms of new treatments. A major advantage of TEM over the use of native non-keratinizing oral mucosa (NNOM) is in the limited availability of NNOM tissue; many cells can be cultured and multiple TEM constructs of consistent quality can be engineered from a single biopsy, therefore many different parameters can be studied in a single experiment.It is well known that radiotherapy, essential during the treatment of oral cancer, not only kills tumour cells, but also damages the surrounding healthy tissue. This damage to healthy tissue manifests in a certain order. The first stage, characterized by erythema, is followed by the manifestation stage, sub-acute stage, chronic stage, and finally the late stage, each with their specific characteristics. Many studies have reported on the side effects of ionizing radiation (IR) on the oral mucosa. These side effects include erythema and dry or moist desquamation or ulceration, and they can be very severe, resulting in the interruption or termination of the radiation treatment.
Immediately after exposure to IR, several biological processes occur. Certain proteins such as p53 binding protein 1 (53BP1) and γH2Ax are elevated within 5 min after IR, both indicating the presence of DNA double-strand breaks, a well-documented result of IR. Additionally a number of cytokines, such as interleukin (IL)-1, IL-6, IL-8, tumour necrosis factor alpha (TNF-α), and transforming growth factor beta (TGF-β), are increased, thereby initiating the biological response to IR .
Keratins are the predominant proteins found in the epithelial layer, and as the expression of certain keratins is a clear indication of the maturation and stratification of the epithelium, we studied keratin expression in the epithelium in non-irradiated and irradiated TEM. As mentioned earlier, radiotherapy is known to cause severe tissue damage, hence we used markers for cell proliferation and apoptosis to evaluate the viability of non-irradiated and irradiated TEM. To study if radiotherapy affects the structural integrity of TEM we evaluated the expression of (hemi-)desmosomal proteins and the expression of major basement membrane component collagen type IV.
The aim of this study was to evaluate the suitability of TEM as a model for studying the acute effects of IR on the oral mucosa. To this end we evaluated epithelial proliferation and differentiation, expression of basement membrane components, cell–cell adhesion, attachment of the epithelium to the underlying connective tissue, and DNA damage and repair on TEM. Additionally, we evaluated the presence of pro-inflammatory cytokines in the conditioned culture media of irradiated and non-irradiated TEM. Finally, the results observed in TEM were compared to those observed in NNOM.
Materials and methods
Cell culture
Biopsies of approximately 2 cm 2 of buccal tissue were taken from four healthy individuals (three males and one female, all Caucasian, mean age 62.75 ± 6.39 years) upon informed consent. Single cell suspensions of keratinocytes or fibroblasts were obtained as described previously. Briefly, keratinocytes were isolated from the epithelial sheet by incubation in trypsin–ethylenediaminetetraacetic acid (EDTA), and the single cell suspension was seeded onto lethally irradiated 3T3 fibroblast feeder layers, in accordance with the protocol of Rheinwald and Green. Fibroblasts were isolated by mincing the dermis using scalpels. This was followed by incubation in collagenase/dispase (1.5 mg/ml/2.5 mg/ml) solution to obtain a single cell suspension. The cells used in this study were within passage 3–6.
Preparation of de-epidermized dermis
Human cadaver skin, cryopreserved in 10% glycerol, and tested negative for cytomegalovirus, human immunodeficiency virus, and hepatitis B virus, was obtained from the Euro Skin Bank (Beverwijk, the Netherlands). The epidermis was removed by gently shaking the skin in phosphate-buffered saline (PBS) supplemented with 200 IU/ml penicillin, 200 μg/ml streptomycin, and 5 μg/ml amphotericin B. The skin was kept in the PBS solution for 3 weeks, and PBS was changed 3 times a week. The de-epidermized dermis (DED) was trimmed into pieces of approximately 1 cm 2 .
Tissue engineered mucosa (TEM)
TEM was created as described previously. Briefly, per construct, 1 × 10 5 fibroblasts were spun into the lamina propria of the DED and 1 × 10 keratinocytes were seeded into a steel ring (diameter 10 mm) placed onto the papillary side of the DED. This was kept under submerged conditions for 24 h. Next, the construct was raised to the air/liquid interface and cultured for 14 days before harvesting. Each test was done in triplicate and three independent replicates of the experiment were performed.
Radiation protocol
TEM and NNOM were gamma-irradiated with 16.5 Gy, a dose based on preliminary studies in our laboratory (unpublished data), using a 137-Cs source. The control group was not irradiated. TEM constructs and NNOM were cultured at 37 °C, 10% CO 2 and harvested at 1, 6, 24, 48, or 72 h post-irradiation by snap freezing with liquid nitrogen for cryosectioning.
Histology and collection of TEM-conditioned culture media
Culture medium was collected prior to harvesting of TEM, centrifuged at 400 × g for 5 min at 4 °C, and stored at −80 °C until further analysis. Cryosections (6 μm) were stained with haematoxylin–eosin (Klinipath, Duiven, the Netherlands) and overall morphology was assessed using a light microscope (Olympus). The thickness of the viable epithelium was determined from two consecutive images, and the average thickness (μm) was measured using Hamamatsu software (Hamamatsu Photonics, Japan) by averaging 12 measurements per image.
Immunohistochemistry
For immunohistochemical analysis, sections were fixed for 10 min with acetone. After incubation with primary antibodies for keratins K10 (1:200; Euro-Diagnostica), K13 (1:200; Euro-Diagnostica), and K19 (1:500; Novus Biologicals), collagen type IV (1:500; Euro-Diagnostica), Ki-67 (1:200; DAKO), cleaved caspase 3 (1:1500; Abcam), vimentin (1:200; Euro-Diagnostica), desmoglein 3 (1:500; Novus Biologicals), integrin α6 (1:1000; Novus Biologicals) or integrin β4 (1:2000; Novus Biologicals), sections were incubated with diaminobenzidine (DAB) substrate or New Fuchsin substrate. All sections were counterstained using Mayer’s haematoxylin. To visualize DNA damage and repair, sections were fixed with 4% formaldehyde in PBS for 15 min. After incubation with primary antibody for either 53BP1 or Rad51, sections were stained with secondary antibody Alexa Fluor 594 (1/1000, goat anti-rabbit IgG, Molecular Probes, Leiden, the Netherlands). Finally the sections were mounted using DAPI/DAPCOA/VectaShield to visualize cell nuclei. Positive controls were biopsies of NNOM, and for the negative controls, PBS replaced the primary antibody.
Quantification of radiation damage and repair
DNA damage was quantified by counting the number of cells containing DNA double-strand breaks in both the epithelial layer and the connective tissue in 12 randomly chosen microscopic views (100× magnification). The index was established as the ratio of the positive cells to all the cells in the basal layer or connective tissue (×100%). Quantification of repair, using Rad51, was done in a similar manner. Only sections from TEM harvested at 1, 6, and 24 h were quantified, as the characteristic kinetics of both 53BP1 and Rad51 take place within 24 h post-irradiation.
Proliferation and apoptosis index
To determine the proliferation index, the basal layer of the epithelium was analyzed. Images were taken from 12 randomly chosen microscopic views using a 100× magnification. The proliferation index was established as the ratio of the Ki-67-positive cells to all cells of the basal layer (×100%). Apoptotic cells were detected using an antibody against cleaved caspase 3. The apoptotic index was established as the ratio of the caspase 3-positive cells to all the cells in the basal layer or connective tissue (×100%).
ELISA assay on conditioned media
Concentrations of IL-1β, TGF-β, and tissue inhibitor of matrix metalloproteinase types 1 and 2 (TIMP-1 and TIMP-2) in the conditioned culture media were measured using DuoSet sandwich ELISA kits in accordance with the manufacturer’s instructions (R&D Systems). Results are expressed as pg or ng/cm 2 tissue, with each sample consisting of 4 ml supernatant derived from 1 cm 2 tissue.
Zymography for matrix metalloproteinases MMP-2 and MMP-9
Gelatinolytic proteinases in conditioned culture media were assessed using zymography. Samples of 10 μl conditioned medium were 1:1 diluted with sample buffer (0.1 M Tris–HCl, 4% sodium dodecyl sulfate (SDS), 20% glycerol, 0.005% bromophenol blue, 10 mM EDTA) and loaded onto a 10% polyacrylamide gel containing 2% gelatin. After electrophoresis, gels were washed with 2.5% (v/v) Triton X-100 to remove the SDS. After overnight incubation at 37 °C in incubation buffer consisting of 50 mM Tris–HCl, 1% Triton X-100, and 5 mM CaCl 2 , gels were stained with Coomassie brilliant blue. MMP-2 and MMP-9 appeared as unstained bands. Gels were scanned using a Kodak Image Station 440CF (Kodak) and the relative intensity of each band was quantified.
Native oral non-keratinizing mucosa (NNOM)
To determine whether the observations in TEM regarding proliferation and expression patterns of components of the epithelium, basement membrane, and connective tissue post-irradiation are similar to those in native oral mucosa, we exposed biopsies of NNOM to 16.5 Gy and assessed the above-mentioned components at 1, 6, 24, 48, and 72 h post-irradiation using immunohistochemistry.
Statistical analysis
All data are expressed as the mean ± standard error of the mean. Tests of normality were done using the Shapiro–Wilk W -test using SPSS software (IBM Nederland, the Netherlands). Statistical analyses were performed using the Student’s t -test with SPSS software. Comparisons between group means were made using the Tukey–Kramer test with GraphPad Instat software (GraphPad Software, USA). P -values of ≤0.05 were considered significant.
Statistical analysis
All data are expressed as the mean ± standard error of the mean. Tests of normality were done using the Shapiro–Wilk W -test using SPSS software (IBM Nederland, the Netherlands). Statistical analyses were performed using the Student’s t -test with SPSS software. Comparisons between group means were made using the Tukey–Kramer test with GraphPad Instat software (GraphPad Software, USA). P -values of ≤0.05 were considered significant.
Results
Confirmation of radiation damage
To confirm that a single dose of 16.5 Gy did result in DNA double-strand breaks, the number of cells containing 53BP1 foci was analyzed. As shown in Fig. 1 A , the number of 53BP1-positive cells in irradiated TEM was increased significantly by 1.7-fold at 1 h post-irradiation. After 6 and 24 h, a 1.3-fold and 1.6-fold increase, respectively, was observed when compared with non-irradiated TEM. Additionally, DNA repair was quantified, and as shown in Fig. 1 B, a 1.3-fold and 1.7-fold increase in irradiated TEM was observed at 1 and 6 h post-irradiation, respectively, as compared to non-irradiated TEM. After 24 h, the number of Rad51-positive cells in irradiated TEM returned to baseline.
