Expression of CD44 variant exons by normal oral epithelia

Oral Oneologv, Vol. 33, No. 4, pp. 26&262, 1997 0 1997 Elsevier Science Ltd. All rights reserved Printed in &ear Brirain 1368-8375197 $17.00+0.00

PII:

Expression

of CD44

SO964-1955(97)00081-4

Variant

Exons

D. T. Oliveira’ ‘Department 2Department

by Normal

Oral

Epithelia

and E. W. Ode112

of Pathology, School of Dentistry of Bauru, University of Sao Paulo, Brazil; of Oral Medicine and Pathology, UMDS Guy’s Hospital, London SE1 9RT,

and

U.K.

Altered expression of CD44H and CD44 splice variants is associated with metastasis in several malignancies but its analysis requires knowledge of the normal pattern of expression which is tissue specific. There are considerable regional variations in epithelial differentiation within the mouth, including differences in cell surface glycoproteins. The aim of this study was to determine whether there are regional variations in the expression of CD44 variants in oral epithelia. Frozen and formalinfixed paraffin-embedded sections of lip vermilion border, buccal mucosa, dorsum and ventrum of tongue, floor of mouth, gingiva and hard palate were stained immunohistochemically for CD44H and the product of variant exons v3, ~415, v6 and v9. Paraffin sections were subjected to microwave antigen retrieval. All epithelia stained strongly for all variants in basal, suprabasal and prickle cells and cornified and surface layers and the basal surface of basal cells were negative. Patterns of staining were identical in frozen and formalin-fixed tissue. Despite the structural heterogeneity within oral epithelium, no regional variation was detected. 0 1997 Elsevier Science Ltd Key words: Oral Oncology,

CD44, Vol.

CD44 33, No.

variants,

keratinocyte,

4, pp. 260-262,

squamous

INTRODUCTION

accepted

14 Nov.

oral

mucosa

1997

Human CD44 is a transmembrane hyaluronan-binding glycoprotein involved in many processes, including lymphocyte activation and homing, cell migration and tumour metastasis [l-3]. A family of isoforms expressed by different tissues are produced from the CD44 gene by alternative splicing of 9 variant exons (exons v2-~10) and 9 constant exons (exons sl-s9 or ~10) [4,5]. Keratinocytes express the haemopoietic CD44 isoform (CD44H) and larger isoforms containing many variant exons [6]. These are thought to be important in cell+ell adhesion and their expression reduces with terminal differentiation [7]. CD44 expression in malignancy is markedly deranged [8] and abnormalities correlate with both tumour progression [9] and metastasis [ 1 O-l 21. Overexpression, downregulation, exons abnormal for the tissue of origin and defective mRNA splicing may occur. Although some carcinomas express additional exons, others show loss of most or some exons from their normal repertoire [6,9, 131. Squamous carcinomas, including those from the upper aerodigestive tract, show selective reduced expression of exons [6,9,14]. To understand patterns of CD44 variant expression in oral carcinoma it is necessary to investigate the possibility of regional variation in expression. This is especially important in the oral mucosa where anatomically close sites have disCorrespondence to E. W. Odell. Received 27 Sep. 1996; provisionally manuscript received 21 Nov. 1996.

epithelium,

1996;

revised

tinctly different epithelial structure and susceptibility to malignancy [15]. Regional variation has been reported in many phenotypic characteristics including expression of keratinocyte surface glycoproteins [ 161, of which CD44 is one. Therefore, the aims of this study were to determine whether regional differences in the expression of CD44 variant exons v3, v4/5, v6 and v9 can be demonstrated and also to determine whether formalin-fixed tissue subjected to microwave antigen retrieval may be used rather than fresh frozen material. MATERIALS

AND

METHODS

Clinically normal mucosa and mucosa from the remote periphery of surgical resections for benign non-epithelial lesions were used. Normal histological appearances for site were confirmed by routine histology. Thirty-three samples of formalin-fured (10% buffered formalin) parathn-embedded normal human oral mucosa from defined sites were obtained from archives. Sites included were lip vermilion border (x=2), buccal mucosa (n=6), dorsum of tongue (n=6), ventral tongue and the floor of mouth (n=6), gingiva (x=7) and hard palate (x=6). Frozen tissue specimens (total 14) were quenched in isopentane, frozen and stored in liquid nitrogen. Sites included were lip vermilion border (n=l), buccal mucosa (n=5), dorsum of tongue (n=4), ventral tongue and the floor of mouth (n=2), gingiva (n=l) and hard palate (x=1) and in 8 cases matched frozen and formalin-fixed samples were used.

Variant

CD44 in Normal

Oral Mucosa

Fig. 1. Formal&tied buccal (upper) and hard palate (lower) mucosa stained immunocytochemically v3, ~415, v6 and v9 after antigen retrieval. Identical patterns of reactivity were found in hozen variant exons in the basal and prickle cell layers (x95).

Sections were stained with monoclonal antibody against CD44H (clone 2C5) and with exon-specific monoclonal antibodies (R&D Systems, Abingdon, U.K.) against products of variant exons v3 (3G5), v4/5 (302), v6 (2FlO) and v9 (lE8) raised by immunisation with human chimeric fusion protein CD44v3-lo-Fc. The strategy used to raise these antibodies and evidence of their specificity against native and formalin-fixed cells is described by Fox et al. [ 133. Immunohistochemistry was performed using a standard streptavidin-biotin-peroxidase complex technique (Dako, U.K.). Formalin fixed 4 pm sections were stained after microwave antigen retrieval by microwaving at 700 W for 10min in O.lM citrate buffer, pH6.0 [17]. Frozen sections 5 pm thick were fixed in acetone for 10 min at 4°C. In either case, endogenous peroxidase was blocked in methanol hydrogen peroxide and the slides incubated for 1 h at room temperature with the following primary antibodies: CD44H (diluted 1:2000), CD44v3 (diluted 1:500), CD44v4/5 (diluted l:lOO), CD44v6 (diluted 1:lOOO). The antibody to CD44v9 (R&D Systems) supplied as supernatant was diluted 1:2. Optimal dilutions were determined in a separate series of experiments. Antibody binding was visualised using 3.3’diaminobenzidine (Sigma, Poole, U.K.) and sections were counterstained with haematoxylin. Appropriate positive and negative controls were included in each run. RESULTS The patterns of reactivity were identical in both frozen tissue and formalin-lixed material subjected to microwave antigen retrieval but staining was consistently stronger in the formalin fixed material. All fixed specimens stained positively and strongly for CD44H and product of variant exons v3, v4/5, v6 and v9 after microwaving (Fig. 1) but not without antigen retrieval. Only very minor qualitative differences in staining

specimens;

for (left to right) CD44H, both sites express all

pattern were detected between keratinised or non-keratinised epithelium in either frozen or formalin-fixed mucosa. Variant exon staining was limited to the epithelium but stromal cells were positive for the more widely expressed CD44H. Expression of CD44H and all variant exons demonstrated a differentiation-related pattern, being strongest in the basal and parabasal layers. Cells in mitosis were stained. No staining for CD44H or variant exons was present along the basal surface of basal cells and staining weakened evenly and progressively towards the epithelial surface. Either no staining or weak staining was found in the upper stratum spinosum of buccal mucosa, dorsum of tongue filiform and interpapillary epithelium, floor of mouth and ventral tongue. In keratinised epithelia of the lip vermilion, hard palate and gingiva, variant CD44 exon staining remained strong and uniform to the junction with the stratum comeum. No CD44 staining was present in the stratum comeum, or in the surface squamous layers of non-keratinising epithelia. DISCUSSION Epidermal keratinocytes express relatively large amounts of many CD44 isoforms. The predominant keratinocyte mRNA contains all variant exons (except vl which is non-functional in human) and mRNA for at least five isoforms is spliced in normal and malignant keratinocyte cell lines; ~8-10, ~6-10, ~4-10, ~3-10 and ~2-10 [7,18,19]. Immunocytochemistry of skin keratinocytes confirms the expression of these exons individually and suggests that all are expressed at all stages of terminal differentiation [7]. Epithelia vary markedly in their CD44 variant exon repertoire and oral epithelia have not been extensively investigated. Expression of v4, v6 and v9 has been reported in oesophagus [61 and tonsilar epithelium [6,9] and v5, v6, v7, ~7-8 and vl0 in oral and pharyngeal epithelia using frozen tissue [ 141.

262

D. T. Oliveira

However, the diversity of epithelial structure in oral mucosa makes it an excellent model to study protein expression in relation to differentiation. Our results have confirmed expression of CD44H and variant exons v3, v4/5, v6 and v9 by oral epithelia at all oral sites tested, despite their widely differing structures and functions. These findings are in keeping with the accepted hypothsis that exon v9 is expressed by most epithelia in the ‘epithelial’ form of CD44 (exons ~1-5 ~8-10 ~6-8 ~10) and that exons v6, v4/5 and v3 are expressed in the ‘keratinocyte’ form (~1-5 ~3-10 ~6-8 ~10) restricted to stratified epithelia [5,6]. Expression by junctional epithelium, previously reported to show only weak expression of CD44H [20] has not been investigated. It is difficult to obtain uninflamed nonhyperplastic specimens from this site, an important consideration because IFNy and possibly other cytokines can modulate keratinocyte CD44 expression [6]. The strong expression by basal cells, the lack of expression on their basal surface, and the fading of staining with progressive terminal differentiation matches that described for various exons in skin [7], oesophagus [6] and that recently reported for exons v5, v6, v7, v7-v8 and v10 in frozen oral epithelia [14]. We have not found different staining intensity for individual exons in either frozen or formalin-fixed tissue as has been reported previously for oesophagus [6] and for exon v5 in oral epithelia [ 141. We have made no attempt to quantitate expression because differences in glycosylation and epitope masking prevent quantitative immunocytochemistry of CD44 isoforms even when antibodies have equivalent binding affinity [ 141. Immunocytochemistry using these antibodies is effective on formalin-fixed paraffin-embedded tissue after microwave antigen retrieval and the same patterns of reactivity are seen as in frozen tissue. This advance will simplify the study of oral carcinoma by enabling the use of archival tissue. In conclusion, expression of CD44 exons v3, v4/5, v6 and v9 has been demonstrated using both fresh frozen and paraffin-fixed oral epithelia and no regional variation has been detected. The uniformity of expression in normal tissue will simplify interpretation of results in the study of oral carcinoma.

1. Laurent, T. C. and Fraser, J. R., Hyaluronan. FASEB Journal, 1992,6,2397-2404. 2. Jutila, M. A., Function and regulation of leukocyte homing receptors. Journal of Leukocyte Biology, 1994, 55, 133-140. 3. Isacke, C. M., The role of the cytoplasmic domain in regulation CD44 function. Journal of Cell Science, 1994, 107, 2353-2359. 4. Underhill, C., CD44: the hyaluronan receptor. Journal of Cell Science, 1992, 103, 293-298. 5. Gunthert, U., CD44: a multitude of isofomis with diverse functions. Current Topics in Microbiology &3 Immunology, 1993, 184, 47-63.

and

E. W. Ode11

6. Mackay, C. R., Terpe, H. J., Stauder, R., Marston, W. L., Stark, H. and Gunthert, U., Expression and modulation of CD44 variant isofonns in humans. Journal of Cell Biology, 1994, 124, 7182. 7. Hudson, D. L., Sleeman, J. and Watt, F. M., CD44 is the major peanut lectin-binding glycoprotein of human epidermal keratinocytes and plays a role in intercellular adhesion. Journal of Cell Science, 1995, 108, 1959-1970. 8. Matsumura, Y., Sugiyama, M., Matsumura, S., et al., Unusual retention of introns in CD44 gene transcipts in bladder cancer provides new diagnostic and clinical oncological opportunities. Journal of Pathology, 1995, 177, 1 l-20. 9. Salmi, M., Gron-Virta, K., Sointu, I’., Grenman, R., Kalimo, H. and Jalkanen, S., Regulated expression of exon v6 containing isoforms of CD44 in man: downregulation during malignant transformation to tumors of squamocellular origin. Journal of Cell Biology, 1993, 122, 431-442. 10. Koopman, G., Heider, K. H., Horst, E., et al., Activated human lymphocytes and aggressive non-Hodgkin’s lymphomas express a homologue of the rate metastasis-associated variant of CD44. Journal of Experimental Medicine, 1993, 177, 897-904. 11. Kaufmann, M., Heider, K. H., Sit-m, H. I’., Minckwitz, G. von, Ponta, H. and Herrlich, I’., CD44 variant exon epitopes in primary breast cancer and length of survival. Lancer, 1995, 345, 615-619. 12. Tanabe, K. K. and Saya, H., The CD44 adhesion molecule and metastasis. Critical Reviews in Oncogenesis, 1994, 5, 201-212. 13. Fox, S. B., Fawcett, J., Jackson, D. G., et al., Normal human tissues, in addition to some tumors, express multiple different CD44 isoforms. Cancer Research, 1994, 54, 4539-4546. 14. Herold-Mende, C., Seiter, S., Bomo, A., et al., Expression of CD44 splice variants in squamous epithelia and squamous carcinomas of the head and neck. Journal of Pathology, 1996, 179, 66-73. 15. Morgan, I’. R., Leigh, I. M., Purkis, I-‘. E., Gardner, I. D., Muigen, G. N. I’. van and Lane, E. B., Site variation in keratin expression in human oral epithelia - an immunocytochemical study of individual keratins. Epithelia, 1987, 1, 31-43. 16. Dabelsteen, E., Cell adhesion molecules in oral mucosa. Current Opinion in Dent&y, 1991, 1, 802-808. 17. Cattoretti, G., Becker, M. H., Key, G., et al., Monoclonal antibodies against recombinant parts of the Ki-67 antigen (MIB 1 and MIB 3) detect proliferating cells in microwave-processed formalin-fured paraffin sections. Journal of Pathology, 1992, 168, 357-363. 18. Brown, T. A., Bouchard, T., St. John, T., Wayner, E. and Carter, W. G., Human keratinocytes express a new CD44 core protein (CD44E) as a heparan-sulfate intrinsic membrane proteoglycan with additional exons. Journal of Cell Biology, 199 1, 113, 207-221. 19. Kugelman, L. C., Ganguly, S., Haggerty, J. G., Weissman, S. M. and Milstone, L. M., The core protein of epican, a heparan sulfate proteoglycan on keratinocytes, is an alternative form of CD44. Journal of Investigative DermatoloD, 1992, 99, 886-89 1. 20. Hakkinen, L., Oksala, O., Salo, T., Rahemtulla, F. and Latjava, H., Immunohistochemical localization of proteoglycans in human periodontium. Journal of Histochemistry & Cytochemisty, 1993,41, 1689-1699. Acknowledgement-The support of The Conselho Pesquiso e Desenvolvimento Cientifico e Tecnologico 201317195-B is gratefully acknowledged.

National Process0

de nr.