J Clin Pathol 1992;45:1007-101 1

p53 protein in odontogenic cysts: Increasedexpression in some odontogenic keratocysts

G R Ogden, D M Chisholm, R A Kiddie, D P Lane

AbstractAims: To assess p53 protein expression ina range of odontogenic cysts arising in themouth, including those of developmentaland inflammatory origin.Methods: p53 protein was identified usingthe polyclonal antibody CM-1, togetherwith a standard immunoperoxidase tech-nique. A total of 36 cystic lesions wereexamined, all of which were histologicallybenign.Results: Expression of p53 protein wasidentified within the lining of five of 12odontogenic keratocysts but was notdetected in the other cystic lesions in theseries.Conclusions: This is believed to be thefirst report that identifies increasedexpression of p53 protein in benign cysticepithelium. The increased expression ofp53 protein in the nucleus is usuallyassociated with malignant disease. Thesefindings are relevant to the managementof odontogenic keratocysts which have atendency to recur, and also to Gorlin Goltzsyndrome in which keratocysts and multi-ple basal cell carcinomas are features.

(7 Clin Pathol 1992;45:1007-1010)

genic cysts: odontogenic keratocysts, dentig-erous cysts, and radicular cysts. Histologically,the odontogenic keratocyst is characterised byan epithelial lining five to eight cells thick, witha prominent basal cell layer of columnar orcuboidal cells (fig 1). The keratinisation of theepithelium is predominantly parakeratotic,although sometimes orthokeratosis is seen.The cyst wall is usually thin unless inflamma-tion is present. The surrounding connectivetissue may contain islands of epithelium orseparate small cysts.`4-'This cyst is of partic-ular interest because of the reported high rateof mitotic activity within its lining,'7 its tend-ency to recurrence,'4 18 and the associationwith Gorlin Goltz syndrome,'9 in which multi-ple naevoid basal cell carcinomas may develop.It is important to note also that some odonto-genic keratocyst linings show epithelial dyspla-sia and it is generally accepted that keratinisingcysts have a greater tendency to malignanttransformation than others. 5 Dentigerouscysts are found near the crown of an uneruptedtooth. Histologically, the cyst wall is composedof a thin layer of connective tissue lined bystratified squamous epithelium or by epithe-lium that is only two to three cells thick. Theepithelium may include mucus producing cellsand sometimes ciliated cells. In some cases thecyst lining is keratinised.'6 Odontogenic kera-

Department of DentalSurgery,DentalHospital and School,Dundee UniversityG R OgdenD M ChisholmR A KiddieCancer ResearchCampaignLaboratories,BiochemistryDepartment, DundeeUniversityD P LaneCorrespondence to:Dr G R Ogden, Departmentof Dental Surgery, DentalHospital and School, ParkPlace, Dundee, ScotlandDDI 4HNAccepted for publication6 May 1992

p53 protein was first reported by Lane andCrawford in 1979.' It is a product of the p53gene which is now classified as a tumoursuppressor gene.21 The gene seems to be afrequent target for mutation,4 being seen as acommon step in the pathogenesis of mosthuman cancers.5 In normal cells p53 proteinhas a very short half life,6 such that it cannot bedetected immunohistochemically.7 Mutationof the p53 gene often results in increasedstability of the gene product (p53 protein).8Thus the mutant form can be detected usingimmunohistochemical techniques; numerousstudies report that it has been identified inmalignant cells but not in normal cells.9'-2We examined a range of oral mucosal lesions

and only identified p53 expression in patientswith oral cancer. 3 However, cystic lesionsaffecting the oral cavity have not been pre-viously examined for the presence of p53. Theepithelial cells which comprise the lining ofodontogenic cysts are derived from the prima-tive oral epithelium of the oral mucosa andcontribute to tooth formation. During andafter this process such epithelia are a commonsource of cystic change within the jaw bones.This study examines three types of odonto-

Figure I Classic histopathological features displayed bythe odontogenic keratocyst obtained from case 4 (table 2)and illustrated in fig 2 and 3.

1 007

on August 24, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.45.11.1007 on 1 Novem

ber 1992. Dow

nloaded from

Ogden, Chisholm, Kiddie, Lane

Table I Expression ofp53 in cysts of the oral cavity

Cyst type Examined Positive Negative

Radicular cyst 12 0 12Dentigerous cyst 12 0 12Odontogenic keratocyst 12 5 7

tocyst and dentigerous cyst are both devel-opmental cysts, but radicular cysts are

inflammatory in origin. They arise as sequelaeof dental caries from the effects of inflamma-tion on epithelial residues situated in the apicalportion of the periodontal ligament. They are

the most common cysts found within the jaw,usually associated with a non-vital tooth. His-tologically, most radicular cysts are lined by a

non-keratinising stratified squamous epithe-lium. Other features include mucus met-aplasia, hyaline bodies, cholesterol clefts withan associated foreign body giant cell reactionand an inflammatory cell infiltrate.'6

MethodsThis study was based on the examinations of36 odontogenic cysts, all ofwhich were routinesurgical specimens fixed in 10% neutral buf-fered formalin. The cysts comprised 12 radicu-lar cysts (three mandibular, nine maxillary), 12dentigerous cysts (six each from the mandibleand maxilla), and 12 odontogenic keratocysts(10 mandibular, two maxillary). For eachspecimen, sections were chosen randomly fromparaffin wax embedded blocks which had beenstained with haematoxylin and eosin p53expression before. Four 5 ,um sections were cutfrom each block and dried overnight at 37GC.The specimens were dewaxed and then washedin TRIS-buffered saline (TBS) and incubatedovernight at 4°C, with the polyclonal antibodyCM-1 (diluted 1 in 1000 of 10% fetal calfserum in phosphate buffered saline (PBS)).CM-1 is a rabbit polyclonal antibody raisedagainst the whole p53 protein.20 A biotinylatedantirabbit immunoglobulin was applied for onehour before applying the avidin:biotin complex(Vectastain ABC Kit, Vector Labs, England)for one hour at room temperature. The speci-mens were then incubated with a solution of

diaminobenzidine and hydrogen peroxide inPBS for 10 minutes. A brown precipitate seen

in the nucleus confirmed the presence of thep53 protein. The tissue specimens were notcounterstained.Two negative controls were used in each

case. For one section a solution of CM-1previously blocked with p53 protein was

applied, whilst normal goat serum was appliedto the second section.

Further staining of the odontogenic kerato-cysts was carried out using the monoclonalantibody PC 10, which recognises the pro-liferating cell nuclear antigen (PCNA). Thiswas used at a dilution of 1 in 250. A bio-tinylated antimouse immunoglobulin was

applied instead of the biotinylated antirabbitimmunoglobulin used with CM-1.To obtain an indication of the rate of cell

division, the number of suprabasal mitoses wascounted in a randomly selected area of 1500epithelial cells in each specimen of odonto-genic keratocyst, as described by Browne.4

ResultsThe results for p53 expression in a variety oforal cysts are given in tables 1 and 2. None ofthe inflammatory radicular cysts nor devel-opmental dentigerous cysts expressed p53. Ineach case the histopathological features were

typical and unremarkable. The only cyst typein which p53 was detectable was the odonto-genic keratocyst (fig 2) and a positive resultwas obtained in five of 12. Staining with thePCNA antibody (PC10) indicated that the p53positive cells were actively dividing, becausesimilar regions were positive for both anti-bodies (fig 3). CM-1 positivity was identified inmost of the basal cells; stainin,g with PC 10 wasmuch more intense, being present in all basalcells and most parabasal cells. All odontogenickeratocysts were positive for PCNA. Two of thefive cases positive for p53 were recurrentlesions compared with five of the seven p53negative keratocysts. One female patient hadfeatures associated with the Gorlin Goltzsyndrome (table 2), namely hypertelorism,spina bifida, and odontogenic keratocysts.However, to date, no basal cell neoplasms have

Table 2 Characteristics of the odontogenic keratocysts assessed for p53 expression

PatientCyst Multiple Recurrent p53No Age Sex Site cysts cysts Histopathological features expression

1 24 F L maxilla (molar) + - Moderate focal inflammation +mild epithelial hyperplasia

2 14 F L mandible (molar) - - Mild focal inflammation +3 16 F L maxilla (premolar) - Mild epithelial hyperplasia +4 75 M L mandible (ramus) + + Mild diffuse inflammation +5 48 F L mandible (body) - Mild focal inflammation +6 21 F L mandible (molar) + - Moderate focal inflammation

Mild epithelial hyperplasiaSatellite cysts (T3rpes I and 2)

7 45 F L mandible (ramus) + - Satellite cyst (Type 1)8 59 F R mandible (ramus) - Mild focal inflammation

Cholesterol, hyaline9 16 M L mandible (ramus) Mild focal inflammation

Satellite cyst (Type 1), hyaline10 35 M R mandible (ramus) + - Mild diffuse inflammation

Cholesterol, hyaline11 75 M L mandible (angle) + + Mild focal inflammation12 63 F L mandible (premolar) + + Mild diffuse inflammation

Satellite cysts (Types 1 and 3)

1008

on August 24, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.45.11.1007 on 1 Novem

ber 1992. Dow

nloaded from

p53 protein in oral cysts

Figure 2 Identification ofp53 positive cells usingCM-I in most basal cellsin the lining of a recurrentodontogenic keratocyst(case 4).

4 0

presented in this case. Of the two cystsexamined from this patient with Gorlin Goltzsyndrome, the first, removed from the man-dibular molar region, proved p53 positive. Theother cyst, removed three years later from themolar region of the maxilla, was p53 neg-ative.Two cysts from the same elderly male

patient, removed at one operation also behaveddifferently. The first, removed from the angle ofthe mandible, was p53 negative; the second,removed from the ramus of the mandible, onthe same side as the first, was p53 positive.Three of the 12 cysts were recurrent, and ofthese one was p53 positive. A history ofmultiple cysts of the jaws was given by two offive patients with cysts showing p53 expres-sion. This history was noted in five of sevenpatients with cysts that did not express p53.By definition, all odontogenic keratocysts

reported in this paper had typical histopatho-logical features. A striking feature was theremarkable consistency of the histopatholog-ical appearance by the epithelium. All cystcapsules showed mild to moderate non-specificchronic inflammatory cell infiltration. Of theseven odontogenic keratocysts in which p53expression was absent, six capsules or liningscontained either cholesterol clefts, hyalinebodies, or satellite cysts. The latter feature

Figure 3 Identification ofPCNA using PCIO in thebasal and parabasal cells ofthe p53 positive recurrentodontogenic keratocystillustrated in fig 2.

comprises three types14 and each, in turn,represents a proliferative, mature, or degen-erative phase in the life cycle of the cyst.Interestingly, the five linings which were p53

* rW positive lacked the presence of cholesterol,hyaline, and satellite cysts in the sectionsexamined. However, a larger series will be

ek.8 * necessary before a firm conclusion with regardto the clinical relevance of this finding can bedrawn. The number of suprabasal mitoticfigures per 1500 cells varied from 0 to 4 but nodifference could be made between the groupsp53 positive and p53 negative linings.Thus the two groups of odontogenic kerato-

cysts could not be distinguished in terms eitherof patient medical or dental history, includingage and sex, nor recurrence or mitotic activityof the cyst lining.

DiscussionMany papers have recently reported the asso-ciation ofp53 protein with a variety of differentmalignant tumours.3 However, it is impor-tant to emphasise that the CM1 antibodyrecognises both wild and mutant forms of p53.Thus identification ofp53 in the lining ofsomeodontogenic keratocysts should not necessarilyimply an association with malignant disease. Itis known that the cellular environment canaffect the stability of p53.6 For example, theenzymatic pathways for p53 degradation maybe inactivated.2' Certain viruses, such as ade-novirus, can also stabilise p53 in non-malig-nant tissue.22 Furthermore, because CM1recognises the whole p53 protein it is morelikely to detect its presence than monoclonalantibodies that are directed against specificepitopes of the p53 protein. CM1 seems toproduce a much stronger "signal" than someof the earlier p53 monoclonal antibodies.2'Hence it seems to be more sensitive. Certainlythe staining seen in odontogenic keratocysts ismuch weaker than that seen in oral cancers.'3Various factors peculiar to the keratocyst mayhelp to explain the occurrence of p53 pos-itivity. Firstly, the cyst has a tendency to recur,suggesting increased epithelial activity.'4Although the rate of recurrence may depend

23more on the method oftreatment, recurrenceup to 60% has been reported.'5 This may inpart be due to its high mitotic rate. Main foundan average of eight mitoses for every 1 cm oflining examined (range 0-22), which wasmuch higher than the other cysts he exam-ined"7 (although Browne"4 calculated a rate of3-9 in his study of 130 primary and ninerecurrent keratocysts). The position of thesemitoses may also be influential. The number ofmitoses situated in the basal region of oralepithelium from various sites has been esti-mated to be between 30-50%.2426 However,only 10% were found in the basal region ofodontogenic keratocysts as reported byBrowne"' and the present study supports thisview. If PCNA positivity can be related to cellproliferation,27 then our results would supportthe finding of increased nuclear activity.The other factor of possible importance is

the association between odontogenic kerato-

1 009

Ix%`V..... %.-.

-")A ` "ki-k"I,.A,.k ii A'.'

on August 24, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.45.11.1007 on 1 Novem

ber 1992. Dow

nloaded from

Ogden, Chisholm, Kiddie, Lane

cysts and Gorlin Goltz syndrome, an auto-somal dominant genetic disorder with highpenetrance.28 The syndrome has variableexpression, and includes odontogenic kerato-cysts, basal cell naevi/carcinomas, bifurcationof the ribs, rib or vertebral fusion, spina bifida,intracranial calcification, hypertelorism, men-

tal retardation, cleft lip and palate and palmarand plantor pitting.'8 1929 The lack of con-sistency of these physical signs in the syn-

drome18 19 29 would tend to support theopinion of Browne'4 that the occurrence of jawcysts represents the minimal expression of thesyndrome. However, this remark may havebeen based on his finding of no difference inthe histology of single, multiple, or GorlinGoltz syndrome associated keratocysts. Sincethen differences have been found in the histol-ogy of odontogenic keratocysts that presentedas single cysts compared with those in theGorlin Goltz syndrome,30 principally an

increase in mitotic activity in the lining of cysts

from the Gorlin Goltz syndrome.p53 expression may therefore be a marker

for the syrndrome particularly as p53 expres-

sion has previously been associated with malig-nancy, and the syndrome is associated withdevelopment of basal cell carcinomas. How-ever, not all cysts removed from our one case ofGorlin Goltz syndrome were p53 positive. Onepossibility is that p53 positivity may be indica-tive of potential malignant change within thecyst lining. Malignant transformation in jawcysts, although rare, is thought to occur more

frequently in keratinising cysts'6 30 31 and hasbeen reported in odontogenic keratocysts32 33

A more extensive study of keratocysts fromGorlin Goltz syndrome and, in particular,those cases associated with basal cell neo-

plasms is now required: particularly as othertumours may be associated with the GorlinGoltz syndrome.34 It has been suggested thatthere might be a hereditory embryologicalmaldevelopment in these patients which pre-

disposes them to further cysts and recurrences.

It may be important, then, that of the odonto-genic cysts examined in this study, only odon-togenic keratocysts expressed p53 positivecells, a cyst whose characteristics have beencompared with that of an aggressivetumour.

We thank Mr G Carmichael for photographic assistance, Mrs VWilson for technical assistance, and Mrs E Mitchell for typingthe manuscript.

1 Lane DP, Crawford LV. T-antigen is bound to host protein inSV40-transformed cells. Nature 1979;278:261-3.

2 Lane DP, Benchimol S. p53: oncogene or anti-oncogene?Genes Develop 1990;4:1-8.

3 Levine AJ, Momand J, Finlay CA. The p53 tumour

suppressor gene. Nature 1991;351:453-6.4 Hollstein M, Sidransky D, Vogelstein B, Harris C. p53

mutations in human cancer. Science 1991;253:49-53.5 Lane DP. Mutation of the p53 gene and accumulation of the

p53 protein: Common steps found in the majority ofhuman cancers. General Motors Cancer Research Foun-dation. In: Fortner JC, Rhodes JE, eds. Accomplishments incancer research 1990. Philadelphia: JP Lippincott, 1991:22-65.

6 Reihsaus E, Kohler M, Kraiss S, Oren M, Montenarh M.Regulations of the level of the oncoprotein p53 in non-transformed and transformed cells. Oncogene 1990;5:137-45.

7 Gannon JV, Greaves R, Iggo R, Lane DP. Activatingmutations in p53 produce a common conformationaleffect. A monoclonal antibody specific for the mutantform. EMBO J 1990;9:1595-602.

8 Finlay CA, Hinds PW, Levine AJ. The p53 proto-oncogenecan act as a suppressor of transformation. Cell 1989;57:1083-93.

9 Bartek J, Bartkova J,Vojtesek B, et al. Patterns of expressionof the p53 tumour suppressor in human breast tissues andtumours in situ and in vitro. Int J Cancer 1990;46:839-44.

10 Chiba I,TakahashiT, Nau MM, et al. Mutations in the p53gene are frequent in primary resected non-small cell lungcancer. Oncogene 1990;5:1603-10.

11 Iggo R, Gatter K, Bartek J, Lane DP, Harris AL. Increasedexpression of mutant forms of p53 oncogene in primarylung cancer. Lancet 1990;335:675-9.

12 Field JK, Spandidos DA, Malliri A, et al. Elevated p53expression correlates with a history of heavy smoking insquamous cell carcinoma of the head and neck. Br JCancer 1991;64:573-7.

13 Ogden GR, Kiddie RA, Lunny DP, Lane DP. Assessment ofp53 protein expression in normal, benign and malignantoral mucosa. J Pathol 1992;166:389-94.

14 Browne RM. The odontogenic keratocyst histologicalfeatures and their correlation with clinical behaviour. BrDentJ7 197 1;131:249-59.

15 Toller PA. Origin and growth of cysts of the jaws. Ann R CollSurg Engl 1967;40:306-36.

16 Pindborg JJ, Kramer IRH. Histological typing of odonto-genic tumours, jaw cysts and allied lesions. Internationalhistological classification of tumours No 5. Geneva: WorldHealth Organisation, 1971:39-42.

17 Main DMG. The enlargement of epithelial jaw cysts.Odontologisk Revy 1970;21:29-49.

18 Pritchard U, Delfino JJ, Ivey DM, Sclaroff A, Giglio JA.Variable expressivity of the multiple nevoid basal cellcarcinoma syndrome. J Oral Maxillofac Surg 1982;40:261-9.

19 Gorlin RJ, Pindborg JJ. Multiple basal cell nevi odontogeniccysts and skeletal abnormalities. In: Syndromes of the headand neck. New York: McGraw-Hill Book Co, 1964:400-9.

20 Midgley CA, Fisher CJ, Bartek J, Vojtesek B, Lane DP,Barnes DM. Analysis of p53 expression in humantumours: an antibody raised against human p53expressed in E-coli. J Cell Sci 1992;101: 183-9.

21 BennettWP, Hollstein MC, He A, et al. Archival analysis ofp53 genetic and protein alterations in Chinese esophagealcancer. Oncogene 1991;6:1779-84.

22 van den Heuvel SJL, van Laar T, Kast WM, Melief CJM,Zantema A, van der Eb AJ. Association between thecellular p53 and the adenovirus 5 EIB-55kd proteinsreduces the oncogenicity of Ad-transformed cells. EMBOJ 1990;9:2621-9.

23 Bramley PA. Treatment of cysts of the jaws. Proc R Soc Med1971;64;547-50.

24 Meyer J, Marwah AS, Weinmann JP. Mitotic rate of gingivalepithelium in two age groups. J Invest Dermatol1956;27:237-47.

25 Gargiulo AW, Wentz FM, Orban B. Mitotic activity ofhuman oral epithelium exposed to 30% hydrogen per-oxide. Oral Surg 1961;14:474-92.

26 Hayes RL, Silberkweit M, Soni NN, Simpson TH. Patternof mitotic activity and cell densities in normal gingivalepithelium of children. J Dent Res 1964;43:217-23.

27 Hall PA, Levison DA, Woods AL, et al. Proliferating cellnuclear antigen (PCNA) immunolocalization in paraffinsections: an index of cell proliferation with evidence ofderegulated expression in some neoplasms. J Pathol1990;162:285-94.

28 Rogerson KC. Gorlin's Syndrome: An update on diagnosisand management. Oral MaxiUofac Surg Clin North Am1991;3:155-64.

29 Gundlach KKH, Kiehn M. Multiple basal cell carcinomasand keratocysts-the Gorlin and Goltz syndrome. JMaxiUofac Surg 1979;7:299-307.

30 Woolgar JA, Rippin JW, Browne RM. A comparativehistological study of odontogenic keratocysts in basal cellnaevus syndrome and control patients. Jf Oral Pathol1987;16:75-80.

31 Shear M. Primordial cyst (keratocyst). In: Cysts of the oralregions. 2nd Ed. Bristol: PSG Wright. 1983:29.

32 High AS, Quirke P, Hume WI. DNA-ploidy studies in akeratocyst undergoing subsequent malignant transforma-tion. J Oral Patwol 1987;16:135-8.

33 Moos KF, Rennie JS. Squamous cell carcinoma arising in amandibular keratocyst in a patient with Gorlin's syn-drome. Br J Oral Maxillofac Surg 1987;25:280-4.

34 Partridge M, Towers JF. The primordial cyst (odontogenickeratocyst): its tumour-like characteristics and behaviour.BrI Oral MaxiU-ofac Surg 1987;25:271-9.

1010

on August 24, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.45.11.1007 on 1 Novem

ber 1992. Dow

nloaded from