There is remarkably little pertinent information available about the correlation between oral epithelium and the possible role in the etiology of dentigerous cysts and ameloblastomas (benign or malignant). Monoclonal antibodies, the wide spectrum screening type was applied on paraffin section, using the peroxidase-antiperoxidase (PAP) procedure. Seven human embryonic cap, bell-stage teeth with the associated dental lamina and oral epithelium, eleven cases of ameloblastoma including three malignant cases, and seven cases of dentigerous cysts were immunocytochemically studied. Keratin profiles were similar in distribution in ameloblastoma (benign and malignant), the mature oral epithelium and dentigerous cysts. The reaction was strong in the malignant ameloblastoma while the dental organs, thin non-keratinized embryonic oral epithelium and dental lamina displayed negative reaction. Interestingly, certain individual cells revealed negative reaction at the stratified squamous adult oral epithelium. These clear negative cells were also present in ameloblastoma. Since the keratin reaction is related to the development and differentiation of tissues and was negative in the thin non-keratinized epithelium, dental lamina, dental organ but positive in adult epithelium and ameloblastoma, it seems that both dental organs and dental lamina are the principal origin of ameloblastoma and dentigerous cysts.

 
Recent progress in understanding the biology of keratin together with the development of monoclonal antibodies to individual keratin proteins provide the foundation for studying keratin expression in normal and pathological oral epithelia.
Keratins are a group of water insoluble proteins characterized by the occurrence of a zonal distribution of different molecular weights keratin that form 10 mm to non-filament in a wide variety of epithelial cells.1,2 Compared with other types of intermediate filament protein (vimentin, desmin, neurofilament and glial filament) keratin is very complicated in terms of sub-unit composition.3 The sub-unit composition of keratin filaments varies with cell type,4,5 period of embryonic development,6 stage of histologic differentiation,7,8 cellular growth environment,9 and disease state.10-14
Ameloblastoma are odontogenic tumors of epithelial origin. The resemblance of this tumor's epithelium to the normal enamel organ indicates that ameloblastoma arises from the dental epithelium, or at least very closely connected with it, but the precise point of origin is unknown.15,16 Spouge17 reported that this tumor arises from the dental lamina or its derivatives. Additionally, other authors suggested that ameloblastoma could also originate from either surface epithelium, or epithelium of odontogenic cysts, particularly the dentigerous cysts.18-24 The aims of this work are:

1.   To study the keratin profiles and the cyto-keratin expression in oral epithelium including:
       a.    Embryonic non-keratinized epithelium:
           i.)    thin up to 2-3 layers
           ii.)    thick, up to 10 layers
      b.   Dental lamina
      c.   Ameloblastic epithelium of ameloblastoma d.           Dentigerous cyst epithelium
      e.   Adult oral epithelium:
           i.)    non-keratinized
           ii.)    keratinized
2.   To examine   whether   immuno-histological localization of epithelial proteins could be used       to determine the origin of ameloblastoma and dentigerous cysts.

 
Eleven cases of ameloblastoma (including three malignant), 7 cases of dentigerous cysts, 7 human embryonic cap and bell stage teeth with the associated dental laminae, and adult oral epithelium were taken from the files of Oral Biology and Oral Pathology Departments, Faculty of Dentistry, University of Alexandria.
Five of the ameloblastomas were of the follicular type, 3 were plexiform type, whereas the other 3 were malignant. All tumors, cysts, and embryonic dental tissues were fixed in 10% formalin and embedded in paraffin. Sections, 4 microns thick, were cut for the immunohisto-chemical detection of keratin proteins with the use of wide spectrum screening monoclonal antibodies to correlate between the oral epithelium and the role in the etiology of dentigerous cysts and ameloblastoma (benign and malignant).

Immunohistochemical Methods

Deparaffinized sections were treated for 20 minutes with a methanol solution containing 0.3% H2O2 to permit inactivation of endogenous peroxidase. The sections were rinsed well and treated with normal rabbit serum for 30 minutes and plotted dry with filter paper. Subsequently, they were reacted with a monoclonal antibody to the wide spectrum keratin (poloclonal) for one hour, rinsed three times in phosphate-buffered saline (PBS), and reacted for 30 minutes with horse radish peroxidase (HRP). The sections were labelled with rabbit anti-mouse immunoglobulin (1:20 dilution, Dakopatts, Copenhagen, Denmark) and rinsed well. Finally, the sections were immersed for 5 minutes in 0.005M TRIS buffer solution (pH 7.6) with 0.05% 3.3 diaminobenzidine tetrahydrochlo-ride (DAB) or by using another chromogen, 3-amino-9-ethyl-carbazole (AEC) in dimethyl formide.

Trypsin Pretreatment

Deparaffinized sections were treated for 30 minutes at 37°C with 100 ml of phosphate-buffered saline (PBS), which contained 0.1 geach of trypsin and CaCI2 prior to immuno-histochemical staining.

 
Negative keratin reaction was found in epithelium of all the human embryonic cap and bell stages of enamel organs and dental laminae. The embryonic thin 2-4 layers of cell, on-keratinized oral epithelium, were also keratin negative [Fig, 1] The thick 10-15 rows of embryonic cells of non-keratinized oral epithelium revealed very weak staining for keratin in the basal and parabasal layers. The spinous cells developed keratin protein expressions and stained moderately orange red color when using AEC chromogen [Figs. 2,3].
The parakeratinized adult oral epithelium expressed keratin in a sequential order of reaction from the weak stain in the basal and parabasal cell layers to a moderate stain in the superficial spinous cell layer. The parakeratinized layer revealed strong antigen reaction.
The same distribution of keratin proteins were expressed   in   the   ortho-keratinized    oral   adult
epithelium (weak staining reaction of the basal and parabasal cell layers and moderate staining in the spinous layers). The superficial spinous cell layers and the granular cell layers exhibited strong reaction assuming a deep brown color when using DAB chromogen and deep red color with AEC [Fig. 4].
Immunohistochemistry study of the seven dentigerous cysts demonstrated keratin reactivity in all layers of the epithelial lining (a regular layer of para-keratinized stratified epithelium). The antigen was localized exclusively in the cytoplasm of the cells. The spinous cell layers and the superficial parakeratinized layers were more strongly labelled [Fig.5].
The 11 cases of ameloblastoma were immuno-cytochemically reacted with the antikeratin antiserum. The tumor epithelium in all cases of ameloblastoma showed strong red or brown stain when reacted with antibodies against keratin. The pre-ameloblast, like peripheral cells, and the stellate reticulum, like central cells, in the follicles were keratin positive. However, cells undergoing degeneration appeared as vacuolated clear cells. The piexiform ameloblastoma revealed anastomosing or continuous inter-connected strands of epithelium that were keratin positive. Also, in the case of acanthomatous ameloblastoma, the tumor epithelium were strongly positive in keratin reaction [Fig. 6].
In malignant ameloblastoma, two of the cases were undifferentiated (malignant transformation) in certain areas. The malignant criteria were hyperchromatism, pleomorphism, and increase in mitotic figures. The third tumor was a poorly differentiated type (ameloblastic carcinoma) that appeared similar to a squamous cell carcinoma. The tumor exhibited keratin pearl formation, individual cell keratinization, hyperchromatism, pleomorphism and increased mitotic figures. All these three cases revealed strong keratin antigen expression that was variable in intensity at different locations. Clear cells were detected in ameloblastoma, which do not express any keratin protein reaction [Fig. 7].

 
It has been shown that several types of epithelial cells have the potential to differentiate into odontogenic epithelium when exposed to proper inductive signals.25 Thesleff reported that oral mucosa was readily differentiated into ameloblasts when experimentally combined with dental mesenchyme.26 However, there is little pertinent information available about the pathway and life span of the dental lamina in the permanent molar region.27-29 It was stated that the morphology of dental lamina epithelium shows a striking resemblance to the epithelial lining of a keratocysts.14,21 Also, the palisading of the basal cells, the hyperchromatic nuclei along with the slight parakeratosis, the apical polarization of the nuclei, were common features in the epithelial lining of cysts in which ameloblastoma arise.30
In the present study, wide spectrum screening monoclonal keratin antibodies were used to correlate between various types of epithelium and their possible role in the etiology of dentigerous cysts and ameloblastoma (benign and malignant). The reactivity was identified in all epithelia except in the primitive oral epithelium (thin non-keratinized embryonic stratified squamous epithelium), the dental laminae and the enamel organs (cap and bell stage).
The findings of negative reaction on thin non-keratinized embryonic epithelium, dental laminae, enamel organs strong in adult epithelium, ameloblastoma and cysts suggest that the keratin expression for epithelial cells depends on the stage of development and maturation. It also indicate that these epithelia could be the primary tissue of origin of the ameloblastoma and dentigerous cysts.
In accordance with Steolinga,24 dental lamina and developing tooth epithelium are unlikely to be the principal origin of ameloblastoma and dentigerous cysts. Keratin profiles observed in normal squamous epithelia were similar to dentigerous cysts, ameloblastoma, and malignant ameloblastoma. Accordingly, very few ameloblastoma are likely to arise from dentigerous cysts, which is in agreement with the findings of Kuusela et al.11
Using the monoclonal antibodies, it was shown that the keratin pattern of the epidermis varies during normal differentiation.31,32 Also, in oral mucosa the turn-over of non-keratinized epithelia is generally faster than that of keratinized mucosa.27,29,32
One of the rather striking immunohistochemical results of this study was the presence of clear negative cells to keratin proteins in ameloblastoma. It seems likely that keratin genes expression have similar profiles in oral epithelium and ameloblastoma and is closely linked to the differentiation of epithelium, hence it could be used as epithelial marker.
Accordingly, it was possible to provide localization and distribution of the monoclonal keratin antibodies profile in the malignant ameloblastoma. On the basis of this finding, we suggest that such strong reaction might be due to various levels of differentiation, or it could be considered as a marker of epithelial maturation.
Antibodies, against other epithelial proteins, may be used in the future to solve the problem of whether there are specific cell types in the enamel organ or the oral mucosa, which are progenitors of ameloblastoma. It would be necessary to combine cell kinetic data with immunocytochemical results
in exploring a possible functional relationship.
Nevertheless, the keratins, which is tightly linked to differentiation, constitute important biological markers. The proteins are stable, relatively resistant to degradation, show great fidelity of expression, and are very antigenic. Therefore, it is probable that the monospecific antikeratin antibodies will prove to have fruitful applications, not only in oral biology but also in many areas of oral diagnosis.
Further immunocytochemical studies against other epithelial proteins, utilizing electron microscopy, may throw light on the origin of ameloblastoma and dentigerous cysts. Such a finding may establish a basic foundation in accurate diagnosis and treatment.