B. Steffensen*, D. E. Lopatin, Blood group substances as R. G. Caffesse and C. T. Hanks Departments of Periodontics, Oral Biology,

andUniversity Oral Pathology, of Michigan, School Ann ofArbor, . Michigan, differentiation markers in human USA dento-g i ng ival e pi t heI i um

Steffensen, B.. Lopatin, D. E., Caffesse, R. G.,Hanks, C. 7: Bloodgroup substances as differentiation markers in human dento-gingival . Journal of Perio- donral Research 1987: 22: 451-455.

The level of cellular differentiation of human oral, sulcular, and junctional epithe- lium was compared by immunohistochemical analysis of membrane-associated blood group-specific carbohydrates. Identification of the blood group A-specific carbohydrate and its two immediate precursor substances, type 2 chain H and N-acetyllactosamine. was accomplished by an indirect immunofluorescence tech- nique. Murine monoclonal antibodies reacting specifically with the antigenic determinants of the blood group substances were used as markers. The blood group A substance, indicating the highest level of cellular differentiation, was demonstrated on the cells in the upper layers of the oral epithelium. In the , the A substance was present on a few cells only, while type 2 chain H was observed frequently. This indicates an intermediate differentiation level of sulcular epithelium. The type 2 chain H precursor, N-acetyllactosamine, the indicator of the lowest level of cell differentiation among the tested substances, was the only blood group substance detected on the junctional epithelial cells and on the basal cells of the sulcular and oral epithelium. Based upon previous studies of cell renewal and differentiation in oral epithelium, the present results indicate that the variations in distribution of the different blood group substances correspond with the regional rates of cell division and the levels of cellular differentiation. The findings also suggest that the cells in the differentiate to a level similar to that of basal cells in the oral epithelium. Accepted for publication April 23, 1987

reestablishment of structure and func- Introduction data are available on the level of cellular tion of the junctional epithelium is not differentiation of the different epithelial The epithelium of the human dento-gin- yet fully understood (13). components of the dento-gingival area gival region has been divided into oral, Blood group-specific carbohydrate in humans (21). Results from studies in sulcular, and junctional epithelium (1, substances have been detected on epi- mice indicate, however, that the 2). The differences in structural and thelial cell membranes as well as on junctional epithelium is less differentiat- functional characteristics of these ep- erythrocytes (14). It has been demon- ed than sulcular or oral epithelium (8, ithelia, such as tissue permeability (3), strated, biochemically, that the speci- 22). variation in ultrastructure (4), rate of ficity of the blood group substances is The aim of the present investigation cellular division (5), and degree of kerat- determined by variations in the struc- was to compare the level of cellular dif- inization (6). have been attributed to ture of the terminal sugar residues on ferentiation between human oral, sulcu- embryogenic development, tissue inter- the carbohydrate chains (1 5, 16). Analy- lar, and junctional epithelium by immu- action and the special environment in sis of epithelial blood group substances nohistochemical identification of three the dento-gingival area (7, 8). The has been shown to be useful in studies well-defined epithelial blood group sub- junctional epithelium retains specific of cell differentiation of normal oral epi- stances. qualities which are clearly distinct from thelium (17, 18) and in conditions of the sulcular and oral gingival epithe- premalignancy, malignancy, and wound lium. However, after surgical removal, healing (19, 20). Current data indicate Material and methods a new junctional epithelium can regen- that the synthesis and elongation of the Tlssue samples erate from the oral epithelium (2, 9-12). blood group-specific carbohydrates are Twenty-one biopsies of marginal gin- The nature of the processes leading to correlated to the differentiation and ma- giva were obtained from 14 patients turity of epithelial cells (17, 18, 20). during periodontal surgery (1 6 speci- *Present address: Department of Peri- Thus, with increasing differentiation, mens) or in combination with ex- odontics, University of Texas Health Science the epithelial cells express more de- traction (5 specimens). Blood typing Center. San Antonio, Texas. U.S.A. veloped carbohydrates. Very few similar showed that 11 patients had blood type 452 Stej’ensen et al.

Mouse Similar procedures for tissue processing Carbohydrate monoclonal have previously been shown to enable Antigen structure antibodies clear detection of epithelial blood group GalNAcal substances (1 7, 18). 1-3 \ “AH16” A 3 or Type 1 or 2 determinant Gal191 --4GlcNAcl --R Antibodies and conjugates chain f2 L-FucuI The epithelial blood group substances were identified by an indirect immuno- “BE2” fluorescent staining technique. Murine H L-Fucal--2Gal/II-4GlcNAcl --R Type 2 chain- monoclonal antibodies reacting with determinant specific carbohydrates specific for the complete blood group A substance (AH16) (24) “ 1B2” and the immediate precursor sub- N-acctyllactosamine Gal/?I-dGlcNAc--R Type 2 chain- stances, type 2 chain H (BE2) (25) or specific N-acetyllactosamine (1 B2) (25) were ap- Gal =Galactose. plied to the sections as the primary anti- GI = Glucose. bodies. The chemical composition of the FUC= FUCOSC. carbohydrate determinants for these AC= Acetyl. blood group substances is shown in Fig. GIcNAc = N-acet yl-D-glucosamine. 1. (The murine monoclonal antibodies Fig. 1. The chemical composition of the carbohydrate chains specific for the blood group were obtained as a gift from Dr. S.I. substances A, type 2 chain H, and N-acetyllactosamine. Hakomori, Fred Hutchinson Cancer Research Center, Seattle, WA). The sec- 0 and 3 patients had blood type A. Bi- and 5 specimens were fixed in Bouin’s ondary antibodies were goat anti-mouse opsies from patients with these blood fixative (23). Following fixation, the immunoglobulins (IgA + IgG + IgM, types were selected because previous specimens, which included tooth sub- heavy and light chain-specific) conju- studies have demonstrated the blood stance, were placed in a buffered 0.5 gated with fluorescein isothiocyanate group substances type 2 chain H and mol/l EDTA solution (pH 7.1) for decal- (FITC) for immunofluorescence visual- N-acetyllactosamine in epithelium from cification of the hard tissues. All speci- ization (Cappel Scientific Div., Mal- individuals with blood type 0 and, in mens were embedded in parafin and vern, PA). addition, the blood group A substance cut at 4-5 microns-thick sections. The in persons with blood type A (20). sections were deparafinized with xylene The patients were appraised of the and hydrated in graded ethanol and Staining procedure study and informed consents were ob- phosphate buffered saline at pH 7.4 The tissue sections were incubated in tained consistent with the policies of (PBS) before the staining procedure. a humidity chamber with the primary The University of Michigan and the NIH.

Tissue processing Sixteen specimens were fixed in a 4% neutral buffered formaldehyde solution,

lium

Fig. 2. Schematic drawing of a gingival tissue section illustrating the oral, sulcular, and junctional epithelium. The circles represent the areas in the subepithelial connective tis- Fig. 3. Oral epithelium from a patient having blood type A ( x 250). A: Immunofluorescence sue where the inflammatory cell infiltrate was staining of the blood group A substance using murine monoclonal antibodies. B: Hematoxylin- evaluated. eosin staining of the section shown in A. Blood group substances in gingiva 453

nous cells was determined in each area. After examination and photography in the IF microscope, the sections were re- stained with hematoxylin-eosin (HE), permitting comparison of the IF stain- ing reaction with the normal histology.

Degree of inflammation The degree of inflammation was deter- mined from the HE-stained sections. In each section, the total number of in- flammatory cells was counted (field size = 1.3 mm2 at 400 x ) in the connec- tive tissue located immediately beneath a) the oral epithelium, b) the sulcular epithelium, corresponding to the region where keratinization disappeared, and c) the junctional epithelium at the mid- point between the most coronal colla- gen fiber attachment and the most api- cal part of the sulcular epithelium. These cell counts were related to the IF staining patterns (Fig. 2).

Control reactions Control reactions included a) staining with FITC-conjugated goat anti-mouse antibodies alone, b) substitution of PBS in place of the murine monoclonal anti- bodies, and c) application of the hybri- doma cell culture media in place of the cultured hybridoma supernatant fluid containing antibody.

Results Blood group A substance The blood group A substance was found on single cells or groups of cells in the upper spinous cell layers in the oral epi- thelium (Fig. 3). Only single scattered cells in the most superficial cell layers Fig. 4. Gingival epithelium from a patient having blood type 0 ( x 250). Immunofluorescent of the sulcular epithelium showed pres- staining of the blood group substance typc 2 chain H using murine monoclonal antibodies in ence of the A substance. No substance (A) oral epithelium and (C) sulcular epithelium. B and C: Hematoxylin-eosin staining of the A was detected in the junctional epithe- sections shown in A and C. lium.

npe 2 chain H substance murine monoclonal antibodies at a di- In the ortho-keratinized oral epithe- lution of 1 :5 for 24 h at 4°C. After three Evaiuatlon Of reaction lium, the type 2 chain H substance was rinses with PBS, each for 5 min, a 1:lO Evaluation of the stained sections was demonstrated on cells in all spinous cell dilution of the secondary goat anti- performed with a Leitz epifluorescence layers (Fig. 4A) and on parabasal cells mouse antibodies was added to the sec- microscope which was equipped with a over the connective tissue papillae. This tions and they were incubated for 1 h at filter system for FITC immuno-fluore- substance was also present in the para- room temperature. Following repeated scence (IF) and an HB0/2 200 W mer- or non-keratinized sulcular epithelium PBS rinses and slight air-drying, the sec- cury superpressure lamp as light source. but was, in this region, limited to the tions were mounted in a glycerol/PBS- The staining reaction was evaluated parabasal and deeper spinous cells (Fig. based mounting medium containing qualitatively in the areas of oral, sulcu- 4C). A clear borderline of the staining para-phenylenediamine to reduce fa- lar, and junctional epithelium (Fig. 2); reaction for the type 2 chain H sub- ding (26). the reaction of basal, parabasal and spi- stance, corresponding to the morpho- 454 Steffensen et al.

hydrates (17, 20, 24, 25). Previously, Dabelsteen and coworkers (17,20) have used similar methods on oral epithelium and have demonstrated that the incom- plete carbohydrates are present on the least mature cells, and that the complete carbohydrates are found on more ma- ture and more highly differentiated cells. In agreement with their findings, the results from this investigation showed that the most complete blood group substance, the A substance, was located in the upper layers of the oral epithelium and the precursor sub- stances, type 2 chain H and N-acetyllac- tosamine, were present in the deeper cell layers. Thus, the location of the differ- ent carbohydrates can be interpreted as a reflection of the maturity and differen- tiation of the epithelial cells. The distribution of the blood group substances in the different regions of the dento-gingival epithelium seemed to be related to the rate of cell proliferation. Thus, the least developed carbohydrate, N-acetyllactosamine, was widespread in the junctional epithelium which has a Fig. 5. Junctional epithelium from a patient having blood type 0 ( x 250). A: Immunofluore- very high rate of cell division (5, 9, 27, scence staining of blood group substance N-acetyllactosamine using rnurine monoclonal 28). The rates of cell division in the antibodies. B: Hematoxylin-eosin staining of the section shown in A. regions where the type 2 chain H and the A antigens were detected (i.e. the sulcular and oral epithelium), are known to be lower than in the junction- al epithelium (5, 9, 27, 28). In other logical border between the sulcular and tory infiltrate in the subepithelial con- words, the results clearly indicated a re- junctional epithelial cells, could be ob- nective tissue and the location of the lationship between the level of cellular served. In the junctional epithelium, blood group substances was observed in differentiation and the regional rates of only single separate cells carried the the examined sections. cell division. type 2 chain H substance. In studies of epithelium from other parts of the oral cavity, the location of Axatlon blood group substances has been related N-acetyllacto8amine The staining reactions following forma- to the type and degree of keratinization N-acetyllactosamine was observed on lin fmation were clear and distinct. Bou- (18). This is in agreement with the find- cell in the whole width of the junctional in’s fmative caused a reaction of higher ings in the present study. The type 2 epithelium, especially in the more apical intensity but also with more back- chain H substance was found on cells areas (Fig. 5). Like the staining reaction ground stain. The distribution of the at a higher cell layer in ortho-keratin- for the antibodies to type 2 chain H blood group substances was similar with ized oral epithelium compared to the substance, the demarcation of the stain- either of the two fixatives and the choice non- or para-keratinized sulcular epi- ing reaction for N-acetyllactosamine of fixative was judged to have no influ- thelium (Fig. 4 A-D). Previous obser- corresponded to the morphological de- ence on the interpretation of the results. vations have demonstrated an associ- lineation between junctional and sulcu- ation between the degree of keratiniz- lar epithelium. The N-acetyllactosamine ation and the rate of cellular division Dlscussion was also detected on the basal cells of of the sulcular epithelium (6, 29). The the oral and sulcular epithelium, and on In the present study, it has been possible present findings may, however, also in- the parabasal cells of the oral epithelial to identify three different blood group dicate that the type of keratinization ridges. substances in the human dento-gingival is associated not only with the rate of epithelium, and to locate these to cer- cellular division but also with the level tain areas and epithelial layers. The of cellular differentiation as represented Degree of inflammation method of identification was based by blood group substances. The degree of inflammation in the speci- upon an immunohistochemical reac- It has been reported that an inflam- mens ranged from a total of 32 to 385 tion between cell membrane-associated matory infiltrate in the subepithelial inflammatory cells per field (1.3 nun2). carbohydrates and murine monoclonal connective tissue influences the mitotic No relationship between the inflamma- antibodies specific for each of the carbo- rate of the overlying epithelium (30). Blood group substances in gingiva 455

It was. therefore, anticipated that less 4. Schroeder HE, Munzel-Pedrazzoli S. 18. Vedtofte P, Dabelsteen E, Hakomori S-I, mature epithelial cells might be detected Morphometric analysis comparing et al. Regional variations of cell surface in areas with a high inflammatory cell junctional and oral epithelium of normal carbohydrates in human oral stratified count than in areas with a low degree human gingiv. Helv Odont Acta 1970; 14: epithelium. DiffPrentiation 1984; 25: 22 I. of inflammation. However, compari- 53. 19. Hakomori S-I. Glycosphingolipids in 5. Hill MW. Cell renewal in oral epithelia. cellular interaction: Differentiation and sons of the distribution of blood group In: Meyers J, Squier CA, Gerson SJ, eds. oncogenesis. Annu Rev Biochem 1981; substances between different tissue sec- The ,structureand function of oral mucusa. 50: 733. tions in the present study did not reveal Oxford: Pergamon Press, 1984: 53. 20. Dabelsteen E. Receptors on the keratin- such alterations of cellular maturation 6. Caffesse RG, Kornman KS, Nasjleti CE. ocyte surface. In: Meyers J, Squier CA, due to subepithelial connective tissue in- The effect of intensive antibacterial ther- Gerson SJ, eds. The Structure and Func- flammation. This may, in the oral epi- apy on the sulcular environment in mon- /ion of . Oxford: Pergamon thelium, be the result of an overall low keys. Part I. Inflammation, mitotic ac- Press, 1984: 83. degree of inflammation. As even the tivity and keratinization of the sulcular 21. Brandtzaeg P. Localization of blood- least inflamed sections of the sulcular epithelium. J Periodontol 1980; 51: 155. group substances A and B in alcohol- 7. Mackenzie IC, Binnie WH. Recent ad- fixed human gingivae by indirect immu- and, in particular, the junctional epithe- vances in oral mucosal research. J Oral nofluorescence technique. Acta Odont lium demonstrated blood group sub- Path 1983; 12: 389. Scand 1965; 23: 335. stances representative for cells with a 8. Mackenzie IC. The role of epithelial- 22. Mackenzie IC, Rittman BR, Vercelotti P. low level of differentiation, a further mesenchymal interactions in epithelial Phenotypic markers of murine junctional decline may not have been detectable. migration and differentiation. J Period epithelium. J Dent Res 1986; 65 (spec Detection and analysis of cell mem- Res 1984; 19: 656. iss): Abs. 68. brane-associated carbohydrates, i.e. 9. Engler WO, Ramfjord SP, Hinniker JJ. 23. Sheehan DC, Hrapchak BB. Theury and blood group substances, has already for Healing following simplc . Practice of Histutechnology. 2nd cd. St. A tritiated thymidine radiographic study. some time been used for investigations Louis: C. V. Mosby Co., 1980. 1. Epithelialization. J Perioduntol 1966; 24. Abe K, Levery SB, Hakomori S-I. The of cell behavior and differentiation in 37: 298. antibody specific to type 1 chain blood conditions of premalignancy and malig- 10 Listgarten MA. Electron microscopic group A determinant. J Immunol 1984; nancy in the human oral epithelium (1 9, study of the junction between surgically 132: 1951. 20). The results from this study indicate denuded root surfaces and regenerated 25. Young WW, Portoukalian J, Hakomori that the method might be a valuable periodontal tissues. J Period Res 1972; S-I. Two monoclonal anticarbohydrate tool for understanding early healing and 7: 68. antibodies directed to glycosphingolipids cell differentiation in the dento-gingival II. Listgarten MA. Ultrastructure of the with a lacto-N-glycosyl Type I1 chain. J area. dento-gingival junction after gingivecto- Biol Chem 1981; 256 10967. my. J Period Res 1972; I: I5 I, 26. Johnson GD, Davidson RS, McNamee 12. Braga AM, Squier CA. Ultrastructure of KC, et al. Fading of immunofluorcscen- Acknowledgements the regenerating junctional epithelium of ce during microscopy: A study of the the monkey. J Periodontol 1980; 51: 386. phenomenon and its remedy. J Immunol The authors gratefully acknowledge Dr. 13. Mackenzie IC, Hill MW. Maintenance h4ethod.s 1982; 55: 231. Leif Hellden for critical review of the of regionally specific patterns of cell pro- 27. McHugh WD, Zander WA. Cell division manuscript. This investigation was sup- liferation and differentiation in trans- in the of developing and ported by USPHS grant DE-0273 I from planted skin and oral mucosa. Cell Tis- erupted teeth. Den/ Pract 1965; 15: 451. the National Institute of Dental Re- sue Res 1981; 219: 597. 28. Skougard M. Turnover of the gingival search, Bethesda, and the LeGro Fund 14. Szulman AE. The histological distri- epithelium in marmosets. Acta farh bution of blood group substances A and Scand 1965; 23: 623. from the School of Dentistry, Ann Ar- B in man. J Exp Med 1960; 111: 785. 29. Caffesse RG, Karring T, Nasjleti CE. bor, Michigan. 15 Watkins WM. The glycosyltransferase Keratinizing potential of sulcular epithe- products of the A, B, H, and Le genes lium. J Periodontol 1977; 48: 140. and their relationship to the structure of References 30. Hopps RM, Johnson NW. Relationship the blood group antigens. In: 5th Inter- between histological degree of inflam- Engler WO, Ramfjord SP. Hinniker JJ. national Convocation on Immunology. mation and epithelial proliferation in Development of epithelial attachment Buffalo, N. Y., 1977: 267. Macaque gingiva. J Periodont Res 1974; and in Rhesus monkeys. 16 Hakomori S-I. Blood group A, B, H and 9: 273. J Periodontul 1965; 36: 45. Ii antigens of human erythrocytes: Schroeder HE, Listgarten MA. Fine Chemistry, polymorphism, and their de- Address: structure of the developing epithelial at- velopmental change. Semin Hematol tachment of human teeth, monograph. 1981; 18: 39. Dr. Bjorn Steffensen, Basel: S. Karger, 1977. 17 Dabelsteen E, Vedtofte P, Hakomori S- Department of Periodontics, University of Osborn JW, Ten Cate AR. The dento- I, et al. Carbohydrate chains specific for Texas Health Science Center, gingival junction. In: Advanced Dental blood group antigens in differentiation 7703 Floyd Curl Drive Hi.ytolugy. 4th ed. Bristol: Wright PSG of human oral epithelium. J Invest Der- Sun Antonio, Texas 78284-7894. 1983: 142. matol 1982; 79: 3. U.S.A.