Histopathology 2000, 37, 27±32
Calretinin expression in ameloblastomas M Altini, H Coleman, C Doglioni,1 G Favia2 & E Maiorano3
Division of Oral Pathology, Department of Anatomical Pathology, University of the Witwatersrand, Johannesburg, South Africa; 1Servizio di Anatomia Patologica, Ospedale Civile, Belluno; 2Clinica Odontostomatologica, UniversitaÁ degli Studi; 3 Instituto di Anatomia e Istologia Patologica, UniversitaÁ degli Studi, Bari, Italy Date of submission 13 July 1999 Accepted for publication 9 December 1999
Altini M, Coleman H, Doglioni C, Favia G & Maiorano E (2000) Histopathology 37, 27±32
Calretinin expression in ameloblastomas Aims: Calretinin is a 29-kDa calcium-binding protein which is expressed in the central and peripheral nervous systems as well as in many other normal and pathological human tissues. The objective of this study was to determine whether calretinin was expressed in the epithelium of ameloblastomas. Methods and results: Twenty-seven cases of unicystic ameloblastoma and 31 cases of solid and multicystic ameloblastoma were studied. Five-micron sections were immunohistochemically stained using either antiserum 7696 from Swant, or antiserum 18±0211 from Zymed with a standard ABC method. The results were identical with the two antisera. Positive staining was seen in 22 cases (81.5%) of unicystic ameloblastomas. This generally consisted of diffuse, intense nuclear and cytoplasmic staining of several cell layers of the more super®cial cells both in the characteristic and non-
descript areas of the cyst linings. In the solid and multicystic ameloblastomas, staining occurred in 29 (93.5%) of the cases. The staining was almost always restricted to the stellate reticulum-like epithelium, was generally intense and diffuse in distribution. Areas of squamous metaplasia stained particularly intensely as did the cells surrounding micro- and macro-cysts. In both groups of lesions, most of the cases that did not stain were intensely in¯amed. Conclusions: The biological signi®cance of calretinin expression in ameloblastomas is not known and its use as a distinctive, speci®c immunohistochemical marker for ameloblastic tissues remains to be con®rmed. However, the results of this study raise the possibility that calretinin may be an important diagnostic aid in the differential diagnosis of cystic and solid ameloblastic tumours.
Keywords: ameloblastoma, calretinin, odontogenic tumours, unicystic ameloblastoma
Introduction Calretinin is a calcium-binding protein of 29-kilodalton (29 kDa) and is a member of the large family of EF-hand proteins to which the S100 protein also belongs. EFhand proteins are characterized by a peculiar amino acid sequence that folds up into a helix-loop-helix which acts as the calcium-binding site; calretinin contains six such EF-hand stretches.1 Calretinin is widely expressed in central and peripheral neural tissues particularly in the retina and in neurones of sensory pathways.2,3 The expression of calretinin as well as of other Address for correspondence: Professor M Altini, Division of Oral Pathology, Private Bag 3, WITS, 2050, South Africa. e-mail:
[email protected] q 2000 Blackwell Science Limited.
calcium-binding proteins such as calmodulin, parvalbumin and calbindin has been of particular use to neuro-anatomists in identifying speci®c subsets of neurones.4 However, expression outside the nervous system has been less extensively characterized and most studies have been performed on non-human tissues. The exact biological function of calretinin remains unknown but possible roles as a calcium buffer4 and regulator of apoptosis have been postulated.5 Calcium-binding proteins such as calbindin6,7 and calmodulin8 have been documented in odontogenic epithelium during tooth development in the rat. In addition a recent, as yet unpublished, study in our department has shown calretinin immunoreactivity in the enamel organs during odontogenesis in the same
28
M Altini et al.
q 2000 Blackwell Science Ltd, Histopathology, 37, 27±32.
Calretinin in ameloblastomas
experimental animal. However, these calcium-binding proteins have not been demonstrated in neoplastic or cystic odontogenic tissues. The aim of this study was to investigate calretinin expression in ameloblastomas and its possible speci®city as an immunohistochemical marker of ameloblastic differentiation.
Materials and methods Twenty-seven unicystic ameloblastomas and 31 solid and multicystic ameloblastomas were selected from the archives of the Division of Oral Pathology of the University of the Witwatersrand. The diagnosis in each case was con®rmed based upon adequate clinical, radiological and histological parameters. Five-micron sections were cut from the paraf®n wax embedded blocks and immunohistochemically stained with calretinin antiserum using a standard avidin±biotinylated peroxidase complex (ABC) method. In 10 of the cases from each group, the sections were pretreated with 0.01% pronase in phosphate-buffered saline for 15 min at room temperature and antiserum 7696 (Swant, Bellinzona, Switzerland) was used at a 1 : 8000 dilution with overnight incubation at room temperature. The remainder of the cases were microwaved before being stained with antiserum 18±0211 (Zymed Laboratories, San Francisco, CA) at a dilution of 1 : 25 and incubated for 60 min at room temperature. Human brain was used as positive and negative controls throughout the study. Immunoreactivity was evaluated with regard to the number and distribution of positive cells and the intensity, pattern and localization of staining, using conventional light microscopy.
29
thin and regular with no rete ridge formation and consisted both of areas which were nondescript in nature and areas showing typical ameloblastic features such as palisaded basal columnar cells and stellate reticulum-like epithelium. Staining was positive in 22 (81.5%) cases. In 18 this consisted of clumped, diffuse, intense cytoplasmic and nuclear staining of several layers of the more super®cial and luminal epithelial cells as well as of the stellate reticulum-like epithelium producing a band-like appearance (Figure 1). Only focally positive epithelial cells were present in the epithelial linings in four cases (Figure 1e). The luminal layer of cells did not stain in some cases (Figure 1c±f). Single basal and parabasal cells stained positively in only two cases. The immunoreactivity was not necessarily continuous throughout the entire length of the epithelium. There was no staining of the lining epithelium in ®ve cases; four of these were intensely in¯amed and the lining epithelium had undergone squamous metaplasia with proliferation and arcading. Only one case which did not stain was lined by typical ameloblastic epithelium. In general, the areas of nondescript lining stained more intensely and diffusely than the characteristic ameloblastic epithelium (Figure 1). In the type 2 lesions (plexiform unicystic ameloblastoma), the intraluminal nodules of proliferating epithelium showed focal staining of cells while in the type 3 lesions, the mural follicles were variable, some showing no staining while in others the stellate reticulum-like epithelium stained intensely. In all cases, numerous individual cells in the connective tissue wall of the cyst were strongly positive. These were interpreted as being mast cells (Figure 1a,b).
Results Both antisera used in this investigation yielded identical results in the control and study sections. In the light of this, the ®ndings will be reported together. UNICYSTIC AMELOBLASTOMAS
The 27 cases of unicystic ameloblastoma included examples of all three clinicopathological variants; type 1, cyst only; type 2, cyst with intraluminal epithelial proliferation; type 3, cyst with transmural tumour growth.9 In the majority of cases, the cyst linings were
S O L I D A N D M U LT I C Y S T I C A M E L O B L A S T O M A S
The sample consisted of 31 cases comprising approximately equal numbers of follicular, plexiform and mixed histological variants. Of these, 29 (93.5%) showed intense positive staining of the ameloblastic epithelium while two cases were negative. Staining was diffusely distributed in 26 cases while three cases stained focally. In all of the cases, staining was restricted to the stellate reticulum-like epithelium while in only one case did basal cells also show focal positivity (Figures 2±4). Areas of squamous metaplasia within the stellate
Figure 1. Calretinin immunostaining in unicystic ameloblastomas. a and b, Non-descript epithelial lining showing diffuse, `band-like' staining of cells in the super®cial cell layers. Individual stromal cells interpreted as mast cells also stained positively (arrow) (calretinin ´ 400). c, Diffuse staining of stellate reticulum-like cells (calretinin ´ 256). d, e and f, Focal and diffuse staining of the super®cial cell layers in characteristic ameloblastic epithelium (calretinin ´ 256, ´ 160 and ´ 400, respectively). q 2000 Blackwell Science Ltd, Histopathology, 37, 27±32.
30
M Altini et al.
Figure 2. Follicular ameloblastoma showing calretinin immunostaining of the central parts of the stellate reticulum-like cells (calretinin ´ 160).
Figure 3. Granular cell ameloblastoma showing focal calretinin immunostaining of the stellate reticulum-like cells. Several positively stained stromal cells can clearly be seen (calretinin ´ 160).
reticulum-like epithelium were particularly prominently stained but the staining was by no means restricted only to squamous cells (Figure 2). Similarly, where there was macro- or micro-cyst formation, the cells lining the cysts stained prominently (Figure 4). The intense staining of the macro-cysts occurred even when the epithelium was nondescript and lacked ameloblastic features. In four cases of granular cell ameloblastoma focal positivity of the granular cells was also present (Figure 3). Of the two negative cases, one was in¯amed showing considerable metaplasia while the other consisted of typical ameloblastic epithelium. In the negative cases, single intensely stained nonepithelial cells were scattered throughout the tumour epithelium. In addition, all cases showed numerous scattered, intensely positive cells in the stroma. These were all interpreted as being mast cells (Figures 3 and 4). The possibility that those occurring in the epithelium might be Langerhans cells was also considered but they were nondendritic.
Figure 4. Ameloblastic follicle with microcyst formation showing calretinin immunostaining of the stellate reticulum-like cells. A strongly positive stromal cell is also present (arrow) (calretinin ´ 400). q 2000 Blackwell Science Ltd, Histopathology, 37, 27±32.
Calretinin in ameloblastomas
Discussion Although calretinin expression in the nervous system has been extensively used by neuro-anatomists to study and identify speci®c subsets of neurones, recently it has emerged as an immunohistochemical marker with a high discrimination potential in speci®c areas of pathology. For example, calretinin represents by far the most speci®c and sensitive marker of both benign and malignant mesothelial cells.4,10 In addition, calretinin is now also established as a marker of neuronal differentiation in central nervous system tumours.11 Calretinin may also have a role to play in the differential diagnosis of cutaneous in¯ammatory in®ltrates since it is consistently expressed in mast cells and in cutaneous mastocytomas.4 Its expression has also been investigated in many normal human tissues and other human neoplasms but its potential role as a speci®c immunohistochemical marker of these tissues has yet to be elucidated.10 Only a partial correlation between staining of normal cells and their neoplastic counterparts has been observed.10 In view of the widespread distribution of calretinin in many normal and neoplastic human tissues and because of the demonstration of calcium-binding proteins in odontogenic epithelium during odontogenesis,6±8 it occurred to us that calretinin might also be expressed in cystic and neoplastic odontogenic tissues and could have some application in the differential diagnosis of odontogenic tumours and cysts. A series of investigations have been undertaken to determine calretinin immunoreactivity in odontogenic tissues and the results of the ®rst of these is reported in this paper. Calretinin expression has been demonstrated in the neural elements of the tooth pulp, periodontal ligament and in viscerosensory nerve ®bres in oral and pharyngeal tissues in the rat12±14 but has not been reported in odontogenic epithelium. There are a number of papers reporting the immunohistochemical expression of intermediate ®laments,15 growth factors,16 basement membrane components,17 cell cycle regulating factors18±21 and apoptotic proteins22,23 in ameloblastomas but calcium-binding proteins have not been investigated in these tumours. The results of our study have demonstrated the frequent expression of calretinin in the epithelium of ameloblastomas. In both the unicystic and solid and multicystic variants calretinin expression occurred almost exclusively in the cells of the stellate reticulumlike epithelium while in the former it also occurred in several of the more super®cial cell layers of the nondescript epithelium which frequently lines large q 2000 Blackwell Science Ltd, Histopathology, 37, 27±32.
31
parts of this cystic neoplasm. Interestingly the surface or luminal cell layers did not always stain. Positivity of the basal cells was hardly ever seen, occurring only in one of the solid tumours and in two of the unicystic variants where single cells expressing calretinin were observed. The reasons for this peculiar distribution of calretinin in ameloblastic epithelium remains unknown. It is not yet clear whether any correlation exists between calretinin expression in normal odontogenic tissues and their neoplastic counterparts. The expression of calretinin was almost always very intense and diffuse being band-like in areas but was not always equally distributed throughout the section with some areas inexplicably being completely devoid of staining. This did not appear to be due to a technical artefact as some sections showed intense staining of one follicle but absolutely no staining of follicles in the immediate vicinity. Calretinin expression was particularly intense in the epithelium lining the micro- and macro-cysts and also in areas where squamous metaplasia of the stellate reticulum-like epithelium had occurred. This propensity for expression by squamous cells concurs with the documentation by Doglioni et al.10 of calretinin expression in keratinizing epithelial cells of the thymus and in the super®cial layer of the pilar infundibulum. Surprisingly, however, only one of the squamous carcinomas in their series stained positively. In the unicystic ameloblastomas, epithelium which completely lacked ameloblastic features frequently expressed calretinin while typical ameloblastic epithelium often showed little or no immunostaining. The impression was gained that the better the differentiation of the epithelium the less expression of calretinin occurred but this was not a completely consistent ®nding. This may indicate that calretinin expression in some cells varies according to their metabolic activity and may be lost when this activity changes. Our results have also shown that ameloblastic epithelium which was associated with an in¯ammatory in®ltrate lost calretinin expressivity although this was not always the case. In areas of epithelial hyperplasia without signi®cant in¯ammation, calretinin expressivity appeared not to be affected. We have previously referred to the fact that strongly stained nonepithelial cells were observed within the tumour epithelium of two solid and multicystic ameloblastomas and that similar appearing cells were observed within the ®brous stroma of all the solid and multicystic and unicystic lesions. We interpreted these as being mast cells and this interpretation is entirely consistent with previous documentation that mast cells do occur within the epithelium and connective tissue of
32
M Altini et al.
odontogenic tumours and cysts24 and that they stain positively with calretinin antiserum.4,10 We also considered the possibility that these cells within the epithelium might be Langerhans cells, which are also known to occur within odontogenic cyst and tumour epithelium,25 but they were clearly nondendritic and this seems unlikely. The biological signi®cance of calretinin expression in ameloblastomas is not known and its use as a speci®c immunohistochemical marker for ameloblastic tissue remains to be con®rmed. However, a concurrent study comparing calretinin expression in unicystic ameloblastomas with other odontogenic cysts has shown that calretinin is preferentially expressed by ameloblastic tissues and does not occur in the other cysts.26 This raises the possibility that calretinin may be an important diagnostic aid in the differential diagnosis of cystic odontogenic lesions and ameloblastic tumours.
11. 12. 13. 14. 15.
16.
17.
References 1. Rogers J, Khan M, Ellis J. Calretinin and other CaBPs in the nervous system. Adv. Exp. Med. Biol. 1990; 269; 195±203. 2. Andressen C, Blumcke I, Celio MR. Calcium-binding proteins: selective markers of nerve cells. Cell Tissue Res. 1993; 271; 181± 208. 3. Schwaller B, Buchwald P, Blumcke I, Celio MR, Hunziker W. Characterisation of a polyclonal antiserum against puri®ed human recombinant calcium binding protein calretinin. Cell Calcium. 1993; 14; 639±648. 4. Dei Tos AP, Doglioni C. Calretinin: a novel tool for diagnostic immunohistochemistry. Adv. Anat. Pathol. 1998; 5; 61±66. 5. Gandler JC, Gotzos V, Fellay B, Schwaller B. Inhibition of the proliferative cycle and apoptotic events in WiDr cells after downregulation of the calcium-binding protein calretinin using antisense oligodeoxynucleotides. Exp. Cell. Res. 1996; 225; 399±410. 6. Taylor AN. Tooth formation and the 28,000-Dalton vitamin Ddependent calcium-binding protein. J. Histochem. Ctyochem. 1984; 32; 159±164. 7. Elms TN, Taylor AN. Calbindin-D28K localisation in rat molars during odontogenesis. J. Dent. Res. 1987; 66; 1431±1434. 8. Goldberg M, Escaig F, Feinberg J, Weinman S. Ultrastructural localization of calmodulin in rat incisor ameloblasts and odontoblasts during early stages of development. Adv. Dent. Res. 1987; 1; 227±235. 9. Ackermann GL, Altini M, Shear M. The unicystic ameloblastoma: a clinicopathological study of 57 cases. J. Oral. Pathol. Med. 1988; 17; 541±546. 10. Doglioni C, Dei Tos AP, Laurino L et al. Calretinin: a novel
18. 19. 20.
21. 22. 23.
24. 25.
26.
immunocytochemical marker for mesothelioma. Am. J. Surg. Pathol. 1996; 20; 1037±1046. Doglioni C, Dei Tos AP, Furlanetto A et al. Calretinin: a novel immunocytochemical marker of neuronal differentiation in tumours of the brain. Mod. Pathol. 1996; 9; 140A (Abstract). Ichikawa H, Deguchi T, Mitani S et al. Neural parvalbumin and calretinin in the tooth pulp. Brain Res. 1994; 647; 124±130. Ichikawa H, Jacobowitz DM, Sugimoto T. Coexpression of calretinin and parvalbumin in Ruf®ni-like endings in the rat incisor periodontal ligament. Brain Res. 1997; 770; 294±297. Ichikawa H, Jacobowitz DM, Sugimoto T. Calretinin-immunoreactivity in the oro-facial and pharyngeal regions of the rat. Neurosci. Lett. 1992; 146; 155±158. Heikinheimo K, Hormia M, Stenman G, Virtanen I, Happonen RP. Patterns of expression of intermediate ®laments in ameloblastoma and human fetal tooth germ. J. Oral. Pathol. Med. 1989; 18; 264±273. Myoken Y, Myoken Y, Okamoto T, Sato JD, Takada K. Immunohistochemical localisation of ®broblast growth factor-1 (FGF-1) and FGF-2 in cultured human ameloblastoma epithelial cells and ameloblastoma tissues. J. Oral. Pathol. Med. 1995; 24; 387±392. Vigneswaran N, Whitaker SB, Budnick SD, Waldron CA. Expression patterns of epithelial differentiation antigens and lectin-binding sites in ameloblastomas. A comparison with basal cell carcinoma. Human Pathol. 1993; 24; 49±57. Kim J, Yook JI. Immunohistochemical study on proliferating cell nuclear antigen expression in ameloblastomas. Oral. Oncol. Eur. J. Cancer. 1994; 30B; 126±131. Li T-J, Brown R, Matthews J. Expression of proliferating cell nuclear antigen (PCNA) and Ki67 in unicystic ameloblastoma. Histopathology 1995; 26; 219±228. Funaoka K, Arisue M, Kobayashi I et al. Immunohistochemical detection of proliferating cell nuclear antigen (PCNA) in 23 cases of ameloblastoma. Oral. Oncol. Eur. J. Cancer. 1996; 32B; 328± 332. Slootweg PJ. p53 protein and Ki-67 reactivity in epithelial odontogenic lesions. An immunohistochemical study. J. Oral. Pathol. Med. 1995; 24; 393±397. Mitsuyasu T, Harada H, Higuchi Y et al. Immunohistochemical demonstration of bcl-2 protein in ameloblastoma. J. Oral. Pathol. Med. 1997; 26; 345±348. Kumamoto H. Detection of apoptosis-related factors and apoptotic cells in ameloblastomas: analysis by immunohistochemistry and an in situ DNA nick end-labelling method. J. Oral. Pathol. Med. 1997; 26; 419±425. Smith G, Smith AJ, Basu MK. Mast cells in human odontogenic cysts. J. Oral. Pathol. Med. 1989; 18; 274±278. Murase N, Tatemoto Y, Iwai Y, Okada Y, Mori M. Langerhans cells in odontogenic tumours and cysts as detected by S-100 protein immunohistochemistry. Basic Appl. Histochem. 1990; 34; 135± 141. Altini M, Maiorano E, Coleman HG, Doglioni C. Calretinin expression in ameloblastomas and odontogenic cysts. J. Dent. Res. 1998; 77; 1278 (Abstract).
q 2000 Blackwell Science Ltd, Histopathology, 37, 27±32.
