(CANCER RESEARCH 53. 1883-1888. April 15. ]993|

Alterations of the p53 1\imor Suppressor Gene and Its Association with Activationof the c-K-ras-2 Protooncogene in Premalignant and Malignant Lesions of theHuman Uterine Endometrium1

Takayuki I immoto,' Masami Fujita, Masaki Inoue, Jerry M. Rice, Ryuichi Nakajima, Osanni Tanizawa, and

Taisei NomuraDepartments of Radiation Bii>log\ IT. E.. T. N.I and Obstetrics and Gynecologv ¡T.E., M. F., M. l.. R. N.. O. T.¡,Osaka University Faculty of Medicine, 2-2, Yamadaoka, Suita,

Osaka 565. Japan, and Laboratory of Comparative Carcinogenesis, National Cancer Institute, Frederick Cancer Research and Development Center, Frederick, Maryland21702-1201 ¡J.M. R.I

ABSTRACT

We previously reported (T. Enomoto et al.. Cancer Res., 50: 6139-6145,1990; T. Enomoto et al., Cancer Res., 5/: 5308-5314, 1991) a significantfrequency of activating point mutations in codini 12 of the c-K-ras-2

protooncogene in endometrial adenocarcinoma and its premalignant precursor lesions (series 1 and 2). To reveal the role of the pS3 tumor suppressor gene in the development of endometrial adenocarcinoma and tostudy the association of p53 alterations with K-ras activation, an addi

tional 28 endometrial adenocarcinomas and an additional 11 premalignantatypical uterine hyperplasias (series 3), as well as 12 cases of endometrialadenocarcinoma (10 having K- or N-ras activation) and 2 cases of atypical

hyperplasia from series 1 and 2, were screened for the presence of p53alterations. Alleile loss, recognized at the polymorphic site in codon 72 ofthep53 gene, was detected in 6 of 19 (32%) informative cases of endometrial adenocarcinoma and 1 of 4 (25%) informative cases of endometrialatypical hyperplasia by restriction fragment length polymorphism analysis and by single-strand conformation polymorphism analysis of poly-merase chain reaction (PCR)-amplified DNA fragments. Mutations in thehighly conserved regions of the p53 gene were detected by single-strandconformation polymorphism analysis of PCR-amplified DNA fragments.

Mutations were found in 9 of 40 (23%) endometrial adenocarcinomas and1 of 13 (8%) atypical hyperplasias that were studied. Mutations in p53were significantly more frequently found in clinical grade 3 (Gj) cancers(6 of 14,43%) than in G,-G2 cancers (3 of 26,12%) (P = 0.033). Mutations

were subsequently confirmed by direct sequencing. Single missense basesubstitutions were detected in 6 cases of endometrial carcinoma and in onecase of atypical hyperplasia. Deletions of a single base and of 2 bases wereeach detected in single cases of endometrial carcinoma, and a single baseinsertion was found in a third case. Point mutations in K-ras were alsoidentified in tumors of series 3 by direct sequencing of PCR-amplified

DNA fragments of exons 1 and 2. Point mutations in codons 12 and 13 inK-ras were detected by direct sequencing of PCR-amplified DNA in 7 of 28

adenocarcinomas in series 3, but none were found in exon 2 (codons59-63). The spectrum of point mutations in p53 in endometrial adenocarcinomas was almost identical to what we found in K-ras in series 1 and 2

and in series 3, suggesting the possible role of a mutagen that might beresponsible for mutations in both K-ras and p53. However, there was nocorrelation between the presence of p53 gene mutations and K-ras activation in these tumors. It appears that inactivation oip53, as well as K-ras

activation, plays a significant role in the development of endometrialadenocarcinoma. In contrast to K-ras activation, which commonly occurs

as an early event, inactivation of p53 usually occurs as a later event inendometrial carcinogenesis, independently of K-ras activation.

1960s to early 1970s (2-4). In Japan, they are the second most com

mon malignancy of the female urogenital tract, and their incidence isincreasing (5). However, the pathogenesis of endometrial carcinoma isnot yet fully understood.

We previously reported that activation of c-K-ras-2 (and occasionally N-ras) by point mutations plays a significant role in the patho

genesis of endometrial adenocarcinoma of the human uterus (6, 7).Point mutations in codon 12 or 13 of K-ras were found in 34% of

endometrial adenocarcinomas and less frequently in atypical hyperplasias (13%), which are considered to be premalignant, while noactivated K-ras genes were observed in adenomatous or cystic hyper

plasias, which are considered to be benign. Therefore, activation ofK-ras can occur as an early and perhaps initiating event in at least a

fraction of endometrial adenocarcinomas.While ras gene mutation is observed in some endometrial adeno

carcinomas, more than 60% of these tumors do not contain any detectable ras gene mutations. This indicates that there are alternativegenetic alterations which contribute to carcinogenesis of the humanendometrium. However, little is known about other genetic changesthat contribute to endometrial carcinogenesis. In colorectal carcinomas, deletions of chromosomes 5q, 17p, and 18q are frequently foundin addition to ras gene mutations. These observations suggest thatmultiple additive genetic changes are involved in the development ofcolorectal carcinomas (8) and, by inference, possibly in other carcinomas as well that, like the majority of colonie carcinomas, evolve byprogression from premalignant precursor lesions.

Alteration of the p53 tumor suppressor gene by base substitution,deletion or insertion mutations, or allelic loss or rearrangements isobserved in a wide variety of human cancers and is currently the mostcommonly found alteration associated with human cancers (9, 10).Alteration of both p53 alíeles,one through deletion and the otherthrough a base substitution mutation, occurs frequently in colorectal(11), lung (12), breast (13), brain (10), and other human cancers (14).We therefore attempted to detect various possible alterations in p53 inmalignant and premalignant lesions of human uterine endometrium.We detected loss of heterozygosity by restriction fragment lengthpolymorphism analysis and SSCP3 analysis of PCR-amplified DNA

fragments (15) and screened for mutations by PCR-SSCP (16). We

established the identity of these mutations by genomic DNA sequencing.

INTRODUCTION

Endometrial carcinomas are common, representing 13% of all malignancies in women in the United States ( 1), where the incidenceappears now to have leveled off after an abrupt increase during the late

Received 9/8/92; accepted 2/10/93.The costs oí"publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1Supported in part by a grant from the Osaka Cancer Society.2 To whom requests for reprints should be addressed, at Department of Radiation

Biology. Osaka University Medical School. 2-2. Yamadaoka. Suita, Osaka 565. Japan.

MATERIALS AND METHODS

Tissues. Samples used in this study were from patients who had beenadmitted to the Department of Obstetrics and Gynecology at the Osaka University Hospital in Osaka, Japan. No initial chemotherapy or hormonal therapywas performed prior to tumor excision. Surgically removed tissues were sampled for histopathological diagnosis, and remainders were frozen for extractionof DNA. In some cases, especially those in which normal stromal cells orinfiltrative cells comprised more than 30% of cells in histological sections of

' The abbreviations used are: SSCP. single-strand conformation polymorphism; PCR.

polymerase chain reaction; HPV. human papillomavirus.

1883

Research. on February 14, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

p53 AND K-raj IN HUMAN UTERINE LESIONS

Table 1 Alterations of p53 and K-ras in endometriai adenocarcinomas of the human uterus (series 3)

K-ras

Codon 72polymorphismsCase

GradeInformative4647

4849505152535455565758NoNT

NoNoYesYesYesNoYesYesNoYesYes59

2Yes602No612Yes622Yes632No643No653Yes663No673No683Yes693No703NT713No723Yes733 NoLOH"NoYesNoNoYesYesNoNoNoNoNoNoYesLOH*YesYesYesYesNoYesYesExonWT

WTWTWTWT6WTWTWT7WTWTWTWTWTWTWTWT5WTWTWTWT7655WTMutationCodon204

GAG-»TAG233

CAC-»CACC149

TCC->ACC248

CGG-*CAG193CAT-»GAT1X1CGC-»TGC175

CGC-»TGCMutation

Codon12

GGT->TGT

WTWTWTWTWTI2GGT->TGT1

2GGT-»GATWT13

GGC-K3ACWTWTWTWTI2GGT->GTTWT12

GGT-»GCTWTWTWT13

GGC-»GACWTWTWTWTWTWTWT

LOH, loss of heterozygosity; NT, not tested; WT. wild type; A, deletion.'' LOH in this column was identified by the loss of bands that represented the normal alíelewhen PCR-SSCP analysis was performed to detect mutation.

a lesion, DNA was extracted from selected areas of formalin-fixed paraffin-

embedded tissue sections. The clinical stage of each carcinoma was establishedaccording to the International Federation of Gynecology and Obstetrics stagingsystem (17). Peripheral blood samples from the patients were also collected forextraction of DNA prior to tumor removal. The current series included 28 newcases of endometriai adenocarcinomas and 11 new cases of atypical hyperpla-

sia (series 3). Ten cases of endometriai adenocarcinomas which were shown tocontain K- or N-ras gene mutations in series 1 and 2 (6, 7), two cases from

series I and 2 in which no ras gene mutations could be detected, and 2 atypicalhyperplasias in series 2 (7) from which DNA from WBC was available werealso analyzed for the presence of p53 alterations.

Detection of Point Mutations in the K-ras Gene by Direct Sequencing.

All 28 cases of endometriai carcinoma and all 11 cases of atypical endometriaihyperplasia in series 3 were analyzed for the presence of point mutations inK-ras. PCR was performed to generate amplified fragments of exon 1 or exon2 segments of the K-ras gene, which were subsequently sequenced by the

dideoxy method as previously reported (7, 18).Detection of Loss of Heterozygosity in the p53 Gene. The second posi

tion of codon 72 of p53 is known to be naturally polymorphic; CGC (Arg) aswell as CCC (Pro) is frequently observed (19). This polymorphism can bedetected by the restriction enzyme BsiVl in cases where its restriction site(CGCG) is present, i.e.. those in which codon 72 is CGC (15). PCR wasperformed to generate an amplified DNA fragment of 247 base pairs surrounding codon 72 of p53. Primers used were 5'-GATGCTGTCCGCGGAC-GATATT-3' (upstream) and 5'-CGTGCAAGTCACAGACTTGGC-3' (down

stream) (15). One base in this upstream primer was altered to create an artificialinvariant BstVl site as an internal control for ß.cfUIdigestion. DNA extractedfrom the tumors and WBC of the patients was used as a PCR template.Conditions for PCR amplification and ß.vrUIdigestion were published previously (20). The 247-base pair DNA fragments were cut into two fragments by

B.ÃŽ/UIdigestion (236 and 11 base pairs) when no restriction site was present incodon 72, whereas three fragments (160, 76. and 11 base pairs) resulted whenthe restriction site was present in that codon.

Loss of heterozygosity in the p53 gene was also detected by an alternativePCR-SSCP analysis in all cases. PCR amplification was performed to generate66-base pair DNA fragments surrounding codon 72. Primers used were 5'-CAGATGAAGCTCCCAGAA-3' (upstream) and 5'-GTGTAGGAGCT-GCTGGTG-3' (downstream) (15). The PCR reaction mixture (total 3 ul)

contained genomic DNA (0.1 ug), deoxynucleotide triphosphates (60 UM)."P-end-labeled primer (0. l UMeach), MgCK ( 1.5 mm). Tris (pH 8.3) (10 min).

KCl (50 raw), and Taq polymerase (O.I unit; Perkin Elmer Cetus Corp.. Nor-walk, CT). One cycle of PCR consisted of 95°Cfor 1 min, 57°Cfor 30 s. and

72°Cfor 30 s, and a total of 30 cycles of PCR amplification were performed.

SSCP analysis was conducted following the procedures previously published(20).

Detection of Point Mutation \np53 by PCR-SSCP Analysis. Exons 5. 6,

7. and 8 were amplified by PCR using primers, the sequences of which wereprovided by Dr. U. Schlegel4 and published previously (20). Conditions of

PCR amplification and SSCP analysis were the same as previously published

(20).DNA Direct Sequencing Analysis. PCR-amplified DNA fragments of p53

which showed bands with altered mobilities by SSCP were subsequentlydirectly sequenced by the dideoxy termination method as previously described(20).

Statistics. The significance of differences in the frequency with whichmutations occurred in different categories of lesions was estimated usingFisher's exact test (21).

RESULTS

Point Mutations in K-ras. Point mutational activation of K-ras

was detected by direct sequencing. Of 28 endometriai adenocarcinomas in series 3, 7 tumors (25%) contained demonstrable point mutations in exon 1 of K-ras, 5 cases in codon 12, and 2 cases in codon 13

(Table 1). As was previously the case for series 1 and 2, no mutationswere detected in exon 2 (codons 59-63) of K-ras. The incidence ofpoint mutations in K-ras detected by direct sequencing was 28% (8 of

29) in series 1 and 2 (7), which was consistent with the incidence inseries 3 but slightly underestimates the frequency identifiable by themore sensitive oligonucleotide hybridization method (38% for K-rasplus N-ras in series 1 plus 2).

Point mutations in K-ras in atypical hyperplasias were also exam

ined. Of 11 atypical hyperplasias in series 3, one case contained aGGT—»GTTtransversion in codon 12 and one case contained aGGC->GAC transition in codon 13 (Table 2).

Loss of Heterozygosity in p53. Loss of heterozygosity in the p53gene was initially surveyed by restriction fragment length polymorphism analysis of the PCR-amplified 247-base pair DNA fragment.

Loss of heterozygosity was further analyzed by SSCP analysis of a

4 U. Schlegel, personal communication. 1991.

1884

Research. on February 14, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

pS3 AND K-ras IN HUMAN UTERINE LESIONS

Table 2 Alterations of p53 and K-ras in atypical endometrial hyperplasias of the human uterus (series 2 and J>

Case

p53 K-ras

Codon 72 polymorphisms Mutation

Informative LOH" LOH' Exon Codon Mutation Codon

31327475767778798081828384NoNTNTNoYesNoNTNoYesNTYesYesNoSeries

2WTYes

7 248CGG->CAGSeries

3WTWTYes

WTWTWTWTNo

WTWTNo

WTNoWTWTI3GGC-K3ACWTWTWTWTWT12

GGT-»GTTWTWTWT13

GGC-»GACWTWT

' LOH. loss of heterozygosity; NT, not tested; WT, wild type.' LOH in this column was identified by the loss of bands that represented the normal alíelewhen PCR-SSCP analysis was performed to detect mutation.

PCR-amplified 66-base pair DNA fragment. The second step wasnecessary since the 247-base pair DNA fragments were not success

fully amplified in some cases in which DNA was extracted fromformalin-fixed paraffin-embedded sections. Of 37 cases of endome

trial adenocarcinomas from series 1, 2, and 3 in which DNA derivedfrom WBC was available, 19 cases (52%) showed heterozygosity incodon 72 of the p53 gene. Loss of heterozygosity was observed in 6of 19 (32%) informative cases of adenocarcinoma (Tables 1 and 3).There was no discrepancy between observations by restriction fragment length polymorphism analysis and by SSCP analysis in thosecases in which loss of heterozygosity could be detected by bothmethods. There was no significant association between loss of heterozygosity and any clinical parameters of the tumors such as stage,grade, presence of distant métastases,or depth of myometrial invasion.

Loss of heterozygosity in the p53 gene in atypical hyperplasias inseries 2 and 3 was also detected. Of 9 cases of atypical hyperplasia inwhich DNA derived from WBC was available, 4 cases (44%) showedheterozygosity in codon 72 of the p53 gene. Loss of heterozygositywas observed in one of these 4 informative cases (25%) (Table 2).

Mutations in p53 in Endometrial Adenocarcinomas. A total of40 adenocarcinomas, including 28 in series 3 and 12 from series 1 and2, were screened for p53 mutations in exons 5 through 8, where themajority of the point mutations are reported in human neoplasms, byPCR-SSCP analysis of each exon. Altered mobilities of PCR products,

which suggest the existence of mutation, were observed in 9 of 40endometrial adenocarcinomas (23%), 3 in exon 5 (cases 64, 71 and

72), 2 in exon 6 (cases 51 and 70), 3 in exon 7 (cases 15, 55, and 69),and one in exon 8 (case 6) (Tables 1 and 3). One or 2 bands withmobility shifts which were clearly distinguishable from the 2 bandscorresponding to the normal alíelewere observed in these 9 cases,whereas only the normal alíelebands were seen in the remaining 31cases. In 8 of 9 cases which showed altered mobilities, signals of 2bands corresponding to the normal alíelewere significantly reduced ornot detectable. We concluded that the normal alíeleof p53 was lost insuch cases.

Mutations in p53 were defined by direct sequencing. Missensepoint mutations were found in 5 cases, and a nonsense mutation wasfound in one case (Fig. 1). A G:C-*T:A transversion resulted inGlu—>newstop codon (GAG—>TAGin codon 204) in case 51, and aG:C->C:G transversion (CAT-»GAT in codon 193) resulted inHis—»Aspsubstitution in case 70. G:C—>A:Ttransitions were found infour cases: CGC—»CAGin codon 248 in case 69, resulting inArg—>Glnsubstitution; CGC—>TGCin codon 181 in case 71 and incodon 175 in case 72, resulting in Arg—>Cys substitution; andCGG—>TGGin codon 248 in case 15, resulting in Arg—>Trpsubsti

tution. A single base pair deletion in codon 149 in exon 5(TCC—>ACC)resulting in a new stop codon (TGA) in codon 169 was

observed in case 64 (Fig. 1), and a 2-base pair deletion in codon 293in exon 8 (GGG—>AAG) resulting in a new stop codon (TAA) in

codon 304 was observed in case 6. A single base pair insertion incodon 233 (CAC—>CACC) in exon 7 resulting in a new stop codon

(TAA) in codon 304 was detected in case 55.

Table 3 Alterations of p53 and K-ras in endornetrial adenocarcinomas of the human uterus (series I and 2)

p53

Codon 72 polymorphisms Mutation

K-ras

Case"Gn316171820252

:is:26:23:i:6ïde

InformativeNTNoNoYesYesNo!

No!Yes1Yes1Yes1

YesNoLOH''NoNoYesNoYesNoLOH

cExonWTWTWTWTWTWTWTYes

7WTWTWTYes

8Codon

MutationCodon12

GGT-+GTT12GGT-»GAT(N-ra.r)I2GGT-»TGT&''GAT12GGT-»<'GAT12

GGT-»''GTTWT1

2GGT-»GAT248CGG->TGG12GGT-»GAT12GGT-»GAT12

GGT-»GAT12GGT->GAT&JGCT293

GGG->AAG WT

" Case 1 is identical to DNA no. I in series I. Case 2 is identical to DNA no. 33 in series 1. Case 3 is identical to DNA no. 23 in series 1, and Case 6 is identical to DNA no. 12

in series 1.* LOH. loss of heterozygosily; NT. not tested; WT, wild type.' LOH in this column was identified by the loss of bands that represented the normal alíelewhen PCR-SSCP analysis was performed to detect mutation.(/The mutant alíeleindicated was very dilute, comprising only 1/8-1/64 of all K-ras gene copies present, and therefore could be detected only by dot-blot analysis.

1885

Research. on February 14, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

pSJ AND K-ras IN HUMAN UTERINE LESIONS

B

G A T CT-A]

205G-,

A 204G-l-TG

-1T 203G-1T-iG

202CJT~|

201

T-1-~_P25'••»

C 1250-

•¿� G —¿�iG249—

A-lG

-,248B?>4A

|247A—¿�'

r\ —¿�i

Fig. l. Demonstration of point mutations in p53 by direct sequencing of PCR-amplifiedgenomic DNA fragments. A. a CAT—»GATtransversion in codon 193 in exon 6 in case 70;B, a GAG~*TAG transversion in codon 204 in exon 6 in case 51; C, a CGG—*CAG

transition in codon 248 in exon 7 in case 69. In case 64 (£)),due to a 1-base pair deletionin codon 149 (TCC—*ACC) in exon 5. signals from both wild-type p53 derived from the

normal stromal cells and mutated p53 derived from the tumor cells are seen.

A mutation in the p53 gene in one premalignant lesion of uterineendometrium was also identified. The existence of a mutation in exon7 was suggested in one of 13 atypical hyperplasias studied by PCR-

SSCP analysis (Fig. 2). Point mutation was subsequently confirmedby direct sequencing in this single case from series 2, a CGG—>CAGtransition in codon 248 resulting in Arg—»Ginsubstitution (Table 3),

for an overall incidence of one in 13 (8%).

DISCUSSION

The combined results of our previous studies and the present investigation are summarized in Table 4. We previously reported that 11of 29 endometrial adenocarcinomas of series 1 and 2 (38%) containedraígene mutations detectable by dot-blot hybridization; 10 were inK-ras (34%) and one in N-ras (4%) (6). We subsequently performeddirect sequencing, which is less sensitive than dot-blot hybridization,

to confirm ras mutations. We could not detect mutated alíelesby thismethod when they comprised 12% or less of all K-ras gene copies;only 8 of 29 tumors (28%) revealed mutated K-ras sequences by this

method (6). In the present study, we found that 7 of 28 tumors (25%)in series 3 contained K-ras mutations detectable by direct sequencing,

which was consistent with the incidence in series 1 and 2. When the29 endometrial adenocarcinomas of our series 1 and 2 were added, atleast 17 of 57 cases (30%) contained activating point mutations in theK-ras gene, and 15 of 57 (26%) contained one or more mutated K-rasalíelescomprising more than 12% of K-ras gene copies. Recently

Mizuuchi et al. (22) confirmed the existence of activated K-ras in

endometrial adenocarcinoma in Japanese women, although they reported that only 6 of 49 tumors in their series (12%) contained mutations. These authors extracted DNA from frozen tissue sampleswhich were estimated to consist of more than 80% tumor cells, butthey did not microdissect the sections to obtain tissue enriched intumor cells. In our experience, in even the most carefully preparedsamples, it was rare to obtain tissue in which tumor cells constitutedmore than 80% of total cells, since most endometrial cancer tissuescontain abundant stromal cells and infiltrative cells. If microdissectionfrom paraffin-embedded sections was not performed, wild-type K-ras

alíelesfrom normal cells usually diluted the mutated alíelebelow thelevel of detection, and significant underestimation of the true incidence would result (18).

All of the activating mutations we found in K-ras occurred in either

codon 12 (15 tumors) or codon 13 (2 tumors); no mutations werefound in the exon 2, codon 59-63 region. If three double mutations wedetected by dot-blot hybridization in series 1 and 2 are counted separately, we found 20 point mutations in G:C pairs: G:C—>A:Ttransitions in 10 cases; G:C—>T:Atransversions in 7 cases; and G:C—>C:G

transversions in 3 cases. No mutations were observed in A:T pairs.We also studied K-ras activation in 11 cases of atypical hyperplasia

in series 3 by direct sequencing. A codon 12 GGT—>GTTtransversionwas detected in one case and a codon 13 GGC—>GACtransition in a

second case. When 16 atypical hyperplasias in series 1 and 2 wereadded, point mutations in K-ras were detected in a total of 4 of 27

atypical hyperplasias (15%). The incidence of point mutations inK-ras in G, cancers (10 of 33; 30%) was higher than that in atypicalhyperplasias, but not to a statistically significant extent (P = 0.13),

and there was only a limited correlation between histological gradeand the presence of point mutations in K-ras in endometrial adenocarcinomas. This suggests that activation of K-ras more commonly

occurs in atypical hyperplasias that progress to cancer, rather thanduring progression of G, cancer to more highly malignant grades.K-ras gene activation apparently does occur as a later event in some

cases, from the previous observations that some carcinomas (but, sofar, no atypical hyperplasias) contain two different mutations differingin relative abundance. Also, in one case, a K-ras codon 12 mutation

was predominantly found in a region of a carcinoma which showed amore aggressive histological pattern (7). More such cases are neededto assess the significance of these observations.

Both K-ras activation and incidence of allelic loss and mutations in

p53 were much more frequent in endometrial adenocarcinoma of the

Fig. 2. Demonstration of point mutations in p53in atypical hyperplasia of uterine endometrium. Amutation in exon 7 in case 32 (A ) was detected byPCR-SSCP analysis IB) and subsequently identified as a CGG—>CAGtransition by direct dideoxy

sequencing (C).

Ba b e d e G A T C

251

250

249

248

247

246

1886

Research. on February 14, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

p5.< AND K-raj IN HUMAN UTERINE LESIONS

Table 4 Summary of p53 and K-ras gene alterations in primary human uterinelesions (series I. 2. and 3)

p53

Lesion LOH MutationActivated

K-ras genes"

CorpusPremalignant

Atypical hyperplasia 1/4(25%) 1/13 (8%) 4/27(15%)

MalignantEndometrial adenocarcinoma 6/19(32%) 9/40(23%) 17/57(31%)

G, 3/9(33%) 2/19(11%) 10/33(30%)''

G2 1/4(25%) 1/7(14%) 3/8(38%)Gj 2/6(33%) 6/14(43%)'" 4/16(25%)

CervixMalignantCervical carcinoma'' 0/10 (0*)' 2/36 (6%/ 1/23(4%)«

" Data included from series 1 (6) and series 2 (7).* Not significantly higher than in atypical hyperplasia (P = 0.13).' Significantly higher than GÃŒcancers (P = 0.042) or than G|-Gz cancers combined

(P = 0.033).d Data from Refs. 7 and 20.*"Lower than in endometrial adenocarcinoma but of borderline significance

(P = 0.057). Data from Ref. 20.^Significantly lower than in endometrial adenocarcinoma (P = 0.035). Data from Ref.

20.* Significantly lower than in endometrial adenocarcinoma (P = 0.0099). Data from

Ref. 7.

uterine corpus than in squamous carcinomas of the uterine cervix(7, 20) (Table 4). In cervical carcinoma, K-ras activation was found inonly one of 23 tumors (4%, P = 0.0099) (7), loss of heterozygosity inp53 was found in none of the informative cases (0%, P = 0.057), andmutations in p53 were found in only 2 of 36 tumors (6%, P = 0.035).

This indicates a significant difference in oncogenic pathway betweenendometrial carcinoma and cervical carcinoma. Cervical carcinoma isknown to be strongly associated with specific strains of HPVs (23),and E6 and E7 proteins of certain oncogenic HPVs (HPV-16 and -18)

form stable complexes with p53 and pRB, respectively (24). Loss ofnormal tumor suppressor functions of these two genes may play keyroles in the etiology of cervical carcinomas positive for HPV infection. In fact, we detected HPV DNA by Southern blot analysis andfound that 19 of 36 cervical carcinomas (53%) were HPV-positive

(20), whereas only 2 of 27 endometrial carcinomas in series 3 (7%)were HPV-positive (P = 0.00012).5 No p53 mutations were detected

in any of 19 HPV-positive cervical carcinomas. In these tumors, p53may be inactivated by complexing with E6 protein. Therefore, inac-

tivation of p53 by allelic loss and/or mutations may not be obligatory.Since no such viruses that can inactivate p53 have been identified inendometrial carcinoma, inactivation of p53 by allelic loss and/or mutations may play a significant role in these tumors. However, p53mutations were detected in only 2 of 10 cervical carcinomas whichcontained low copy numbers of HPV DNA that was detectable only byPCR (1 OOpy/10-10* cells). Moreover, none of 9 cervical carcinomas

without detectable HPV contained p53 mutations. Therefore, inactivation of the p53 gene by allelic loss or by mutations is infrequent incervical carcinoma, irrespective of the presence or absence of HPVinfection. In cervical carcinomas, especially those without HPV infection or with infection but at low HPV copy number, there shouldexist an alternative oncogenic pathway which is independent of p53inactivation.

Association between the presence of K-ras activation and the pres

ence of p53 mutation in endometrial carcinoma was evaluated. Incidence of p53 mutations in the tumors with ras activation (2 of 17,11.8%) was lower (but not significantly so) than in the tumors withoutras activation (7 of 23, 30%; P = 0.16). This indicates that ras and

p53 gene mutations occur independently in endometrial adenocarcinoma, which is consistent with the findings in non-small-cell lung

5 M. Fujita and T. Enomoto, unpublished data.

cancer (25). In non-small-cell lung cancer, however, p53 mutations

which coexist with ras mutations tend to cluster in exon 8 (26). We didnot observe any such tendency in the only two endometrial adenocar-

cinomas with both raíand p53 mutations, both of which had p53

mutations in exon 7.Correlations between the presence of p53 mutations and histolog-

ical grade of endometrial carcinomas were attempted. Mutations weresignificantly more frequently found in G3 cancers (6 of 14, 43%) thanin G, cancers (2 of 19, 11%; P = 0.042) or in G, and G2 cancerscombined (3 of 26, 12%; P = 0.033). No correlations were established

between the presence of p53 mutations and other clinical parameters,such as clinical stage, presence of lymph node or distant métastases,or depth of myometrial invasion.

Association between allelic loss and the presence of mutations inp53 was evaluated. The frequency of allelic loss, based on the codon72 polymorphism (6 of 19, 32%), was slightly higher (but not significantly so) than the frequency of mutations in p53 (9 of 40, 23%; P =

0.33). Of 9 tumors with mutations in p53, 8 were also analyzed forallelic loss in p53 detectable by the codon 72 polymorphism (Tables1 and 3). Four of these 8 cases were informative, and allelic loss wasobserved in all 4 cases. Moreover, in 4 of 4 carcinomas with mutationsin p53 in which allelic loss could not be evaluated by codon 72polymorphism, loss of a normal alíelewas suggested by the absenceof two bands which represent the wild-type sequences in PCR-SSCP

analysis. Of 31 tumors that did not contain mutations in highly conserved regions of the p53 gene, 15 cases were informative (48%) andone showed allelic loss in p53 by the codon 72 polymorphism criterion (case 57). It is possible that mutations exist in exons that were notevaluated in this case. However, studies that included outlying exonssuggest mutations outside exons 5-8 are rare (14).

Although mutations in p53 were more frequently detected in G3cancers than in G,_2 cancers, and mutations in p53 may thus occur asa later event in tumor progression, it is important to note that alterations in p53 were detected in one case of atypical hyperplasia. Thisshows that p53 alteration can also occur in earlier stages of tumordevelopment. Atypical hyperplasia is considered to have the potentialto develop into carcinoma. Although the risk of invasive cancer forwomen with atypical hyperplasia is considered to be 5-12% (26), little

is known about prognostic indicators for individual cases of atypicalhyperplasia. Evaluation of oncogene activation and tumor suppressorgene alteration, including c-K-ras-2 and p53, may be a useful indi

cator of prognosis for atypical endometrial hyperplasia.Of 7 point mutations in p53 which were detected in endometrial

carcinoma or in atypical hyperplasia, 5 were G:C—>A:Ttransitions,one was a G:C—>T:Atransversion, and one was a G:C—»C:Gtrans-

version. Like the mutations in K-ras, none were observed at any A:Tbase pairs. All G:C—>A:Ttransversions were at CpG sites, and 3 of

them were in codon 248, which was the most common site for mutation in our combined series of endometrial lesions. When 10 mis-

sense point mutations reported from other groups are included (27, 28,29), 13 G:C->A:T transitions, 3 G:C-»T:A transversions, and oneG:C—»C:Gtransversion have been reported in endometrial lesions. Of

the 13 transitions, 10 were at CpG sites, and 5 of 10 were in codon248. G:C—>A:Ttransitions in CpG sites are thought to occur in part

through deamination of methylated cytosine to thymidine (30). Certain toxic agents may catalyze this deamination reaction, such as nitricoxide (NO), which is a cigarette smoke constituent, air pollutant, andendogenous bioregulatory agent, resulting in G:C—»A:Ttransition(31). G:C—>A:Ttransitions in CpG sites therefore are not necessarily"spontaneous" in origin. The fact that the spectrum of missense point

mutations in p53 is similar to that in K-ras in endometrial lesions may

suggest a role of exogenous or endogenous chemical agent(s) that maycause mutations in both genes.

1887

Research. on February 14, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

pS3 AND K-ras IN HUMAN UTERINE LESIONS

In the present study, we found that 13 of 28 endometrial adenocar-cinomas of series 3 contained either K-ras and/or p53 mutations.However, 15 of these 28 tumors contained neither K-ra.v mutations

detectable by direct sequencing nor p53 mutations in the highly conserved regions. This may indicate that there should exist other onco-

genes or tumor suppressor genes that are involved in endometrialcarcinogenesis. Further study will be necessary to identify such genes.

ACKNOWLEDGMENTS

The authors wish to express their appreciation to Dr. G. Buzard and Dr. C.Weghorst for helpful comments and to K. Enomoto and K. Breeze for preparing the typescript.

REFERENCES

1. Young, C. R., Fuks. Z.. Knapp. R. C., and Disaia, P. J. Cancer in ovary. In: V. T.DeVita, Jr.. S. Hellman. and S. A. Rosenberg (eds.l. Cancer: Principles and Practiceof Oncology, pp. 1083-1117. Philadelphia: J. B. Lippincott Co.. 1985.

2. Jones. H. W. Endometrial hyperplasia. In: H. W. Jones, A. C. Wentz, and L. S. Burnett(eds.), Novak's Textbook of Gynecology. Ed. II, pp. 716-727. Baltimore: Williams

& Wilkins. 1988..V Jones. H. W. Endometrial carcinoma. In: H. W. Jones, A. C. Wentz. and L. S. Burnett

(eds.), Novak's Textbook of Gynecology, Ed. 11, pp. 728-760. Baltimore: Williams

& Wilkins, 1988.4. Gompel. C.. and Silverberg, S. G. (eds.). The corpus uteri. In: Pathology in Gyne

cology and Obstetrics, Ed. 3, pp. 149-277. Philadelphia: J. B. Lippincott Co., 1985.5. Masubuchi, K. Increasing incidence of endometrial carcinoma in Japan. Gynecol.

Oncol., 3: 335-346. 1975.6. Enomoto, T., Inoue. M, Perantoni, A. O.. Terakawa. N., Tanizawa, O., and Rice, J. M.

K-ros activation in neoplasms of the human female reproductive tract. Cancer. Res.,50: 6139-6145, 1990.

7. Enomoto, T.. Inoue, M.. Perantoni. A. O.. Buzard, G. S., Miki, H., Tanizawa, O., andRice, J. M. K-rcw activation in premalignant and malignant epithelial lesions of thehuman uterus. Cancer Res., SI: 5308-5314. 1991.

8. Fearon, E. R.. and Vogelstein. B. A genetic model for colorectal tumorigenesis. Cell,61: 759-767, 1990.

9. Levine, A. J.. Momand. J.. and Finlay. C. A. The p53 tumor suppressor gene. Nature(Lond.), 351: 453-156. 1991.

10. Nigro. J. M.. Baker, S. J., Preisinger, A. C.. Jessup. J. M.. Hosteller, R., Cleary, K.,Bigner, S. H.. Davidson. N., Baylin, S.. Devilee, P.. Glover. T., Collins. F. S., Weston,A.. Modali. R.. Harris. C. C., and Vogelstein, B. Mutations in the p5) gene occur indiverse human tumour types. Nature (Lond.), 342: 705-708, 1989.

11. Baker, S. J., Fearon. E. R., Nigro, J. M., Hamilton. S. R., Preisinger, A. C.. Jessup, J.M., van Tuinen. P., Ledbeiter. D. H., Barker, D. F.. Nakamura, Y.. White, R., andVogelslein. B. Chromosome 17 deletions and p53 gene mutations in colorectal carcinomas. Science (Washington DC). 244: 217-221, 1989.

12. Takahashi, T.. Ñau, M. M., Chiba, !.. Birrer. M. J.. Rosenberg, R. K., Vinocour. M..Levitt, M., Pass, H., Gazdar. A. F., and Minna, J. D. p53: A frequent target for geneticabnormalities in lung cancer. Science (Washington DC). 246; 491^494. 1989.

13. Bartek, J., Iggo, R., Gannon. J.. and Lane, D. P. Genetic and immunochemical analysisof mutant p53 in human breast cancer cell lines. Oncogene, 5: 893-899, 1990.

14. Hollstein, M.. Sidransky. D.. Vogelstein. B., and Harris, C. C. p53 mutations in humancancers. Science (Washington DC). 253: 49-53, 1991.

15. Meltzer, S. J.. Yin. J.. Huang. Y., McDaniel, T. K.. Newkirk, C, Iseri, O., Vogelstein,B., and Resau, J. H. Reduction to homozygosity involving p53 in esophagea! cancersdemonstrated by the polymerase chain reaction. Proc. Nati. Acad. Sci. USA. 88:4976^980. 1991.

16. Orila. M.. Suzuki. Y, Sekiya, T.. and Hayashi. K. Rapid and sensitive detection ofpoint mutations and DNA polymorphisms using the polymerase chain reaction. Ge-nomics, 5: 874-879, 1989.

17. Poulsen. H. E., Taylor, C. W.. and Sobin. L. H. Histological typing of female genitaltract tumors. In: International Histological Classification of Tumors, no. 13. pp.15-18. Geneva: WHO. 1975.

18. Enomoto. T., Weghorst. C. M., Inoue. M.. Tanizawa. O., and Rice. J. M. K-ra.sactivation occurs frequently in mucinous adenocarcinomas and rarely in other common epithelial tumors of the human ovary. Am. J. Palhol., 139: 777-785, 1991.

19. Buchman. V. L.. Chumakov. P. M, Ninkina. N. N.. Samarina, O. P.. and Georgiev. G.P. A variation in the structure of the protein-coding region of the human p53 gene.Gene, 70: 245-252, 1988.

20. Fujila, M.. Inoue. M.. Tanizawa, O.. Iwamoio. S., and Enomolo, T. Aheraiions of Ihep53 gene in human primary cervical carcinoma with and without human papilloma-virus infection. Cancer Res., 52: 5323-5328, 1992.

21. Gad. S., and Weil, C. S. Stalistics and Experimental Design for Toxicologisis, Ed. 2.pp. 54-55. Caldwell. NJ: Telford Press. 1988.

22. Mizuuchi, H.. Nasim, S.. Kudo, R., Silverberg. S. G., Greenhouse, S., and Garre«,C.T. Clinical implicaiions of K-ras mulalions in malignanl epiihelial lumors of iheendomelrium. Cancer Res., 52: 2777-2781, 1992.

23. zur Hausen, H. Papillomaviruses in human cancer. Appi. Palhol., 5: 19-24, 1987.24. Howley, P. M. Role of human papillomavirus in human cancer. Cancer Res.. 51

(Suppl.).- 50l9s-5022s, 1991.25. Milsudomi. T. Steinberg, S. M., Nau, M. M., Carbone, D.. D'Amico. D., Bodner, S..

Oie, H. K.. Linnoila. R. !.. Mulshine. J. L.. Minna, J. D.. and Gazdar. A. F. p53 genemutations in non-small-cell lung cancer cell lines and iheir correlaiion wilh thepresence of ran mutations and clinical features. Oncogene. 7: 171-180, 1992.

26. Disaia. P. J., and Creasman, W. T. Endometrial hyperplasia. In: P. J. Disaia and W. T.Creasman (eds.). Clinical Gynecologic Oncology. Ed. 3. pp. 133-160. Si. Louis: C. V.Mosby Company. 1989.

27. Okamolo, A.. Sameshima. Y., Yamada. Y, Teshima, S.. Terashima, Y, Terada, M.. andYokola, J. Allelic loss on chromosome 17p and p53 mutations in human endometrialcarcinoma of ihe uterus. Cancer Res., 51: 5632-5636, 1991.

28. Risinger, J. I.. Georgette, A. D., Ignar-Trowbridge. D., McLachlan, J. A., Tsao, M-S.,

Senterman. M.. and Boyd. J. p53 gene mutations in human endometrial carcinoma.Mol. Carcinog.. 5: 250-253, 1992.

29. Kohler. M. F.. Berchuck. A.. Davidoff. A. M., Humphrey. P. A.. Dodge, R. K.,Iglehan. J. D.. Soper, J. T., Clarke-Pearson. D. L.. Bast, R. C., Jr.. and Marks. J. R.Overexpression and mutaiion of p53 in endometrial carcinoma. Cancer Res., 52:1622-1627, 1992.

30. Sved, J.. and Bird. A. The expected equilibrium of the CpG dinucleotide in vertebralegenomes under a mulaiion model. Proc. Nail. Acad. Sci. USA, 87: 4692^696. 1990.

31. Wink, D. A., Kasprzak, K. S.. Maragos. C. M.. Elespuru. R. K.. Misra. M.. Dunams.T. M., Cebula, T. A., Koch, W H.. Andrews. A. W.. Allen. J. S.. and Keefer. L. K.DNA deaminating abilily and genotoxicily of nitric oxide and its progenilors. Science(Washingion DC). 254: 1001-1003, 1991.

1888

Research. on February 14, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

1993;53:1883-1888. Cancer Res   Takayuki Enomoto, Masami Fujita, Masaki Inoue, et al.   EndometriumPremalignant and Malignant Lesions of the Human Uterine

-2 Protooncogene inrasAssociation with Activation of the c-K- Tumor Suppressor Gene and Itsp53Alterations of the

  Updated version

  http://cancerres.aacrjournals.org/content/53/8/1883

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/53/8/1883To request permission to re-use all or part of this article, use this link

Research. on February 14, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from