Histomorphometric evaluation of the epithelial compartment in reconstructed tissues
At day 3 of development, the epithelium was comprised of 1–4 cell layers (37.73 ± 10.87 μm thickness, mean, and standard deviation) without any sign of differentiation in all OTs. At day 5, the epithelial component displayed 3–5 cell layers (49.79 ± 11.50 μm) and very little differentiation, as only one superficial cell layer had a more flatten morphology and could be considered as suprabasal. For both day 3 and day 5 OTs, there was no noticeable difference between OTs constructed using GF or PLF, as assessed by histological analysis of HE stained tissues (Fig. 1).
Figure 1. Hematoxilin–eosin stained tissues. (A) Organotypic models constructed with gingival epithelial cells grown on top of collagen matrices populated with GF (gingival fibroblasts) or PLF (periodontal ligament fibroblasts) harvested at different stages of growth. (B) Native tissues used as controls. Scale bar: 200 μm.
Download figure to PowerPoint
At day 7, the OT models had an epithelial component of 9–13 layers (130.93 ± 29.28 μm), increasing to 11–16 layers at day 9 (190.83 ± 16.78 μm). In all 7- and 9-day OT models, the differentiation was visible and the presence of the spinous layers could be noticed. No significant difference in the thickness could be observed between OTs constructed with GF or PLF at days 7 and 9 either. Variations from batch to batch could be observed regarding size of the cells, total number of layers or degree of differentiation. However, some of the tissues constructed with GF showed a well-differentiated superficial cell layer compared with those constructed with PLF at day 9, but such differences were not found consistently in all batches.
Distribution and quantification of proliferating cells (Ki-67)
In vivo, the cell proliferation rate is an important feature that differentiates JE, having the highest cell turnover among all types of periodontal and oral tissues. Therefore, we have quantified the Ki-67-expressing cells in the reconstructed OT models for the purpose of comparison. The distribution and number of Ki-67-positive cells varied in culture with the time allowed for development and differentiation. Day 3 and day 5 OT cultures showed proliferating cells distributed in all cell layers, similarly with the situation encountered in JE in vivo. In these OTs, we found significantly higher proliferation rate (P < 0.001), in cultures both with GF and with PLF, compared with those harvested at later time points (day 7 and 9) (Fig. 2A). The type of fibroblasts was not found to influence significantly (P = 0.38) the proliferation rate within the basal and parabasal layers of organotypic models. Interestingly, significant more Ki-67-positive cells were detected at day 5 in suprabasal layers (P = 0.018, paired t-test) in tissues with PLF (2.34 ± 0.39, mean and standard error) than in those with GF (1.28 ± 0.43). At day 7, fewer Ki-67-expressing nuclei were counted in suprabasal layers, (1.03 ± 0.51 in GF OTs and 1.13 ± 0.59 in those with PLF, respectively; Fig. 2B). At day 9, no Ki-67-expressing cells were present in suprabasal layers.
Figure 2. Quantification and distribution of proliferating cells (Ki-67 positive) in organotypic models. (A) Percentage of Ki-67-positive cells in basal and parabasal layers; (B) Mean (n = 4 independent experiments) of Ki-67-positive cells per 334 μm length in basal (light gray bars) and suprabasal (dark gray bars) layers. Statistical significance: ***P < 0.001; *P < 0.05. GF, gingival fibroblasts; PLF, periodontal ligament fibroblasts.
Download figure to PowerPoint
Expression pattern of differentiation markers in reconstructed and control tissues
A summary of expression of the differentiation markers, both in reconstructed tissues and in controls, is presented in Table 2. The CK 13, 16, and 19 were consistently found expressed in the epithelium of all OT cultures, constructed either with GF or with PLF. In OTs, the marker for stratified squamous non-keratinized epithelia CK 13 was present throughout all epithelial layers at day 3 but more variable at day 5, becoming stronger in upper layers. At days 7 and 9, the expression of CK 13 became polarized, as all cells of suprabasal layers stained strongly for this marker, while cells in basal layers were positive in a variable proportion from a few scattered cells to about 50% positive. Slight variations were observed from batch to batch or even within the same tissue model. The CK 13 staining in JE control tissue was positive in all layers, in a proportion slightly varying between samples, from 85% to 100%. The SE control tissues expressed CK 13 only in suprabasal layers, and the staining was gradually lost toward GE, where this marker was nearly absent.
Table 2. Results of immunostaining for a panel of different markers in control tissues and organotypic models. The presence (+) or absence (−) of immunopositive cells or their presence in only some of the samples (+/−) is marked in the table without indicating the intensity of staining or any quantification
|Marker||In organotypicsa||Control JE||Control SE||Control GE|
At day 3, the OTs expressed CK 16 in the cells of the upper layers and sparsely in the basal cell layer, where only a few positive cells were found, matching with the expression of this marker determined in control tissues. In JE, CK 16 was found positive only in half of the samples where was expressed by approximately 30% of cells. OT tissues constructed with both GF with PLF from day 5 to day 9 stained positive for CK 16 only in the suprabasal layers. SE control tissues displayed CK 16 immunopositive cells in the spinous layer and sparsely in basal cells and the staining faded or disappeared toward JE. In GE control tissues, we have found CK 16-positive cells in parabasal and suprabasal cells, much stronger stained in the spinous layer.
CK 19 was expressed in all OT reconstructed tissues, in all cell layers, stronger at 3 and 5 days of culture and slightly fading at days 7 and 9, especially in OTs constructed with GF (Fig. 4). The percentage of CK 19-positive cells decreased gradually with the time allowed for maturation of the tissue cultures. In GF tissues, from 70.3 ± 2.83% (mean and standard error) CK19-positive cells determined at day 3 and 67.91 ± 2.03% at day 5, the percentage decreased further to 53.17 ± 2.65% at day 7 and 31.24 ± 4.97% at day 9. The length of time the in vitro reconstructed tissues were allowed to develop in culture before harvesting and analyzing was found statistically significant by ANOVA test for repeated measurements (P = 0.003) as having the main effect for expression of this marker in OT culture. The OTs constructed with PLF expressed CK 19 in a higher percentage than those with GF, as 76.95 ± 0.87% of the cells at day 3 stained positively and 72.81 ± 2.72% at day 5, respectively. The proportion of CK 19-expressing cells decreased further, reaching 56.29 ± 2.04% at day 7 and 35.45 ± 5.44% at day 9 in PLF organotypics. Although the expression of CK 19 was maintained higher in PLF than in GF organotypics throughout all development stages, when these groups were compared by use of paired t-test, statistical significance was found only at day 5 (P = 0.047). In vivo, CK 19 was expressed intensely in all samples in all layers of JE, while in control SE the expression changed from intensely positive cells toward apical margin, until almost disappearing toward the coronal region. In GE, there were very few CK 19-positive cells, scattered in the basal layers in rete ridges (Fig. 4).
The marker for simple epithelia CK 8 was weakly positive in few of our OT cultures constructed with both GF and PLF, only at day 3 and not in cultures at day 5 or older. The control tissues did not stain positive for CK 8 either, but few single cells in the basal layers of rete ridges in GE were observed, as for CK 19. The positive control (prostate tissue) showed intense staining for CK 8.
In OT cultures grown to day 3 and 5, the marker for stratified keratinized epithelia CK 10 was absent. In samples harvested at day 7, CK10-positive cells were found sparsely in spinous layers, in approximately 20% of the samples, in OT tissues constructed with both GF and PLF. The expression of CK 10 OTs was much stronger at day 9, in spinous and superficial layers, in cultures with both types of fibroblasts. This pattern was observed also in control tissues: absence of CK 10 in JE, very rare positive cells in SE (more toward gingival margin), and strong expression in spinous and superficial layers of GE.
Transglutaminase stabilizes the epithelium by cross-linking constituent proteins; thus, it has been used as a marker for epidermal cell maturation. It was expressed in all layers of day 3 cultures, constructed with either GF or PLF, while in days 5, 7, and 9 it was present mainly in the upper layers and seemed stronger in OT with GF. In control JE, transglutaminase was found positive in the upper part of the tissue. In SE, the staining faded toward apical margin, but it was stronger in spinous layers toward gingival margin. Transglutaminase was expressed in spinous layers of GE control tissues.
The other marker of late keratinisation, filaggrin, was not found positive in any of OT cultures. This marker lacked as well from JE and SE control samples, but was found positive in upper layers of GE control tissues.
Collagen IV, characteristic for basal membranes and external basal lamina in JE, was not detected in OTs of day 3. At day 5, collagen IV was found sparsely deposited in OTs constructed with both GF and PLF (Fig. 5). The deposition of collagen IV was still weak and irregular at day 7, but was observed as a continuous line in OTs grown till day 9. Laminin-1, another protein characteristic for basal membranes, was expressed by the epithelial cells nearby epithelium-collagen matrix interface in all OTs, but its deposition extracellularly was not detected till day 5, when it was sparce and discontinuous in OTs with both GF and PLF (Fig. 5). The deposition of laminin-1 became more obvious at day 7 and further more developed at day 9, when it was visible as a continuous line in OTs constructed with both type of fibroblasts.
Figure 5. Immunohistochemistry showing similarities in protein expression found between a representative organotypic culture, PLF – day 5, and control JE. Scale bar: 200 μm.
Download figure to PowerPoint
Concluding, based on the analyses of tissue morphology and differentiation, correlated with the expression pattern of different markers, and the decline of proliferating cells in the basal layers, the OT cultures could be grouped in ‘early’ (day 3 and 5) and ‘late’ (day 7 and 9) culture models. The ‘early’ OT cultures had low number of epithelial layers, none, or very little differentiation (one suprabasal layer at day 5) and high number of proliferating cells distributed in all cell layers. The specific markers (FDC-SP and ODAM), the CKs (CK 19, 13, 16, and 8), and transglutaminase normally expressed by native JE [5, 7, 15, 17] were also determined in our ‘early’ OT models (especially in those constructed with PLF) as having a distribution pattern similar with the one encountered in vivo. The markers CK 10 and filaggrin normally lacking from JE [7, 13, 16, 24] were also lacking from day 3 and 5 OT models. Deposition of basement membrane proteins was detected only from day 5 OTs.