Topological Analysis of p21 Expression in Esophageal Squamous Dysplasia€¦ · Yasuhiro Shirakawa,1 Yoshio Naomoto, Masashi Kimura, Ryuichi Kawashima, Tomoki Yamatsuji, Takahiko

Topological Analysis of p21WAF1/CIP1 Expression in EsophagealSquamous Dysplasia

Yasuhiro Shirakawa,1 Yoshio Naomoto,Masashi Kimura, Ryuichi Kawashima,Tomoki Yamatsuji, Takahiko Tamaki,Madoka Hamada, Minoru Haisa, andNoriaki TanakaFirst Department of Surgery, Okayama University Medical School,700-8558 Okayama Japan

ABSTRACTIn the normal stratified squamous epithelium of the

esophagus, only the third to the fifth layers of cells expressthe cyclin-dependent kinase inhibitor p21WAF1/CIP1 (p21).Using immunohistochemical staining, we examined the top-ological distribution of cells expressing p21, p53, Ki67, andcytokeratin 10 (CK10), a differentiation marker of esopha-geal squamous cell carcinoma (SCC), in 25 superficial SCCsand 72 dysplastic lesions of the esophagus. Image analysis ofp21, p53, and Ki67 expression was also performed in 48dysplastic lesions. In superficial SCCs, although Ki67- andp53-expressing cells were mainly distributed in the deeplayers of tumors despite tumor differentiation, the distribu-tion of p21 correlated with tumor differentiation. In dysplas-tic lesions, p53- and Ki67-coexpressing cells tended to locatein the same layers and expand in the lower layers of epithe-lium with the progression of dysplasia. p21-expressing cellsshifted to the upper layers of the epithelium with the pro-gression of dysplasia. However, this change was heterogene-ous; in some lesions, p21-expressing cells were confined tothe superficial layers of atypical cells (confined type),whereas in others, p21-overexpressing cells were scatteredamong atypical cells (scattered type). CK10 expression wasobserved in 25% of dysplastic lesions, and the frequency ofCK10 expression was significantly higher in the scatteredthan in the confined type. Our results suggest that esopha-geal squamous dysplasia represents the earliest pathologicalprocess in esophageal squamous carcinogenesis. Our resultsalso suggest that differentiation of esophageal SCC is deter-mined at the stage of dysplasia, and that p21 plays a criticalrole in the differentiation process.

INTRODUCTIONEsophageal SCC2 has one of the worst prognoses among

digestive carcinomas. At present, early detection is the best andonly chance to improve the prognosis of esophageal SCC.Recent advances in diagnostic techniques have allowed a betterdetection of superficial esophageal SCCs, and EMR currentlyoffers absolute curative therapy for some types of SCC super-ficial lesions. With regard to superficial SCCs, the importanceof detection of dysplastic lesions that could be considered pre-cancerous lesions has also been emphasized in recent years.Furthermore, various biomarkers of esophageal squamous dys-plasia have been investigated, such as DNA content (1), Ki67(1), p53 (2), proliferating cell nuclear antigen (3), and argyro-philic nucleolar organizer region (4). Results of these studiessuggest that the nature of squamous dysplasia is analogous tothat of intraepithelial SCC. In fact, some pathologists especiallythose in Japan, argue that this lesion is a low-grade carcinomarather than a separate entity of esophageal squamous dysplasia.However, whether esophageal squamous dysplasia is a precan-cerous lesion remains to date a controversial issue, similar to thecase of colonic adenoma in the adenoma-carcinoma sequence.

The cyclin-dependent kinase inhibitor p21WAF1/CIP1 wasinitially identified as a downstream effector of the cell cyclearrest function of p53.p21WAF1/CIP1gene expression is directlyup-regulated by wild-type but not mutant p53 at the transcrip-tional level (5, 6). However,p21WAF1/CIP1gene is also regulatedby p53-independent factors, including growth-promoting factorsand differentiation-associated transcription factors (7–16). Onthe other hand, p21WAF1/CIP1 has been shown to be associatedwith senescence (17) and terminal differentiation (18–20).There is a striking compartmentalization of p21WAF1/CIP1-ex-pressing cells throughout the normal gastrointestinal tract epi-thelia. In the normal epithelium of the esophagus, only the thirdto fifth layers of cells express p21WAF1/CIP1, whereas most cellsin the second layer express Ki67, a well-known proliferatingmarker expressed during both the G1 and S phases of prolifer-ation but not in quiescent cells (21). In recent years, expressionof p21WAF1/CIP1 has been investigated in esophageal SCC aswell as in tumors of other tissues such as the larynx, stomach,and head and neck (22–28). Several studies have shown agood correlation between the quantity and distribution ofp21WAF1/CIP1 expression with proliferation and differentiationof SCC (23–25, 28). However, there have been only a fewreports on the expression of p21WAF1/CIP1 in squamous dyspla-sia. Furthermore, the objective topological analysis of variousbiomarkers has never been reported in these lesions.

In the present study, using immunohistochemical stainingReceived 7/8/99; revised 11/1/99; accepted 11/1/99.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisementin accordance with 18 U.S.C. Section 1734 solely toindicate this fact.1 To whom requests for reprints should be addressed, at First Depart-ment of Surgery, Okayama University, 2-5-1 Shikata-cho Okayama700-8558, Japan. Phone: 81-86-235-7257; Fax: 81-86-221-8775.

2 The abbreviations used are: SCC, squamous cell carcinoma; CK10,cytokeratin 10; DAB, diaminobenzidine; EMR, endoscopic mucosalresection; TI, topological index.

541Vol. 6, 541–550, February 2000 Clinical Cancer Research

Research. on May 18, 2021. © 2000 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

and computer image analysis, we investigated the topologicaldistribution of p21WAF1/CIP1-, Ki67-, and p53-expressing cellsin esophageal superficial SCC and dysplasia and determined thecorrelation between the two types of lesions in the course ofesophageal squamous carcinogenesis. Furthermore, we per-formed immunohistochemical staining for CK10 in superficialesophageal SCC and dysplasia. CK10, an epidermis-relatedcytokeratin, is a marker of keratinizing squamous epithelium(29) that is not expressed in the normal esophageal epithelium.In esophageal SCC, the expression of CK10 correlates with thedegree of keratinization, a feature of tumor differentiation (30).Our results showed CK10 expression not only in superficialSCC but also in dysplasia. We also investigated the correlationbetween p21WAF1/CIP1 expression and the tendency for differ-entiation assessed by CK10 expression in dysplasia.

MATERIALS AND METHODSPatients and Tissue Sampling. We examined 25 super-

ficial SCCs, including 15 mucosal carcinomas and 10 submu-cosal carcinomas, obtained from previously untreated 23 pa-tients. Patients included 21 men and 2 women with a mean ageof 68.3 years (range, 44–73 years). We also examined 72dysplastic lesions, including 20 mild, 22 moderate, and 30severe dysplastic lesions, obtained from 53 patients. Patientswere 47 men and 6 women with a mean age of 61.9 years(range, 38–76 years). Tissue samples were not obtained bybiopsy because it has been our experience that such samples areoften too small for image analysis. Instead, tissue samples were

surgically or endoscopically resected at the First Department ofSurgery, Okayama University Medical School, Okayama, Ja-pan, during a period extending from 1992 to 1997.

The histological diagnosis of superficial SCC and squa-mous dysplasia was based on the WHO International Histolog-ical Classification of Tumors, 1990 (31). Superficial SCC rep-resented mucosal and submucosal carcinomas, and the grade oftumor differentiation was assessed by the degree of keratiniza-tion. On the other hand, in mild dysplasia, atypical cells (whichcontain enlarged and hyperchromatic nuclei) were limited to thebasal one-third of the thickness of the epithelium. In moderatedysplasia, the zone of atypical cells extends to two-thirds or lessof the thickness of the epithelium, whereas in severe dysplasia,this zone extends to two-thirds or more. In the present study,intraepithelial SCCs were included in the severe dysplasia groupbecause it was sometimes difficult to differentiate betweenintraepithelial SCC and severe dysplasia. Two experienced pa-thologists who were blinded to the clinical data and the resultsof other diagnostic tests determined the grade of dysplasia in allof the lesions.

Immunohistochemistry. p21WAF1/CIP1, p53, Ki67, andCK10 staining was performed using formalin-fixed, paraffin-embedded serial sections. Tissue sections (3mm thick) weremounted on silanized slides (DAKO Japan Co., Tokyo, Japan),were deparaffinized in xylene for 20 min, and were rehydratedin graded ethanol solutions. Endogenous peroxidase wasblocked by incubating the sections in 3.0% H2O2 in methanolfor 15 min. Antigen retrieval on paraffin sections was performed

Fig. 1 Using NIH Image program, we calculated the relative values of the vertical position of immunoreactive cells in the epithelial layer. Todemonstrate the topological distribution, the mean6 SD of relative values was defined as the TI.

542 Topological Analysis of p21WAF1/CIP1 in Esophagus



by heating at 121°C for 15 min in 10 mM citrate buffer solution(pH 6.0) in an autoclave. After the blocking of nonspecificreactivity with rabbit serum for 10 min at room temperature(Histofine SAB PO kit; Nichirei, Tokyo), sections were incu-bated overnight at 4°C with the primary antibody. Identificationof the distribution of the primary antibody was achieved bysubsequent application of a biotinylated antiprimary antibody(Histofine SAB PO kit) and streptavidin-peroxidase (HistofineSAB PO kit). Immunostaining was developed using DAB/hy-drogen peroxide solution (Histofine DAB substrate kit; Nichirei,Tokyo, Japan), and sections were counterstained with Mayer’shematoxylin. As a negative control, some sections were sub-jected to normal serum blocking and omission of the primaryantibody.

In each lesion, the assessment of p21WAF1/CIP1, p53, andKi67 was based on the nuclear staining pattern, whereas theCK10 assessment was based on the cytoplasmic staining pattern.Expression of p21WAF1/CIP1and Ki67 was considered positive ifit was equal or stronger than its expression in normal epithelium.Expression of p53 and CK10 was considered positive whenintense immunoreactivity was present in.10% of the atypicalcells examined. Furthermore, the percentage of immunoreactivecells in the lesions was semiquantitatively evaluated by countingat least 500 cells in the most representative area.

Monoclonal Antibodies. Clone EA10 for p21WAF1/CIP1

was obtained from Oncogene Science (Cambridge, MA); clonesDO-7 for p53 and DE-K10 for CK10 were from DAKO(Glostrup, Denmark); and MIB-1 for Ki67 was purchased from

Immunotech (Westblock, ME). EA10, DO-7, and DE-K 10 werediluted at 1:100, and MIB-1 was ready for use.

Computer Image Analysis. In this study, analysis ofsquamous cell carcinogenesis was based on the stereologicaland conformational concept. To investigate the distribution ofp21WAF1/CIP1, p53, and Ki67 expressions in squamous dyspla-sia, digital image analysis was performed using NIH Imageanalysis program.3 For objective estimation of the topologicaldistribution, we used the TI, calculated as follows: (a) usingNIH image analysis software, immunoreactive cells weremarked manually, and, at the same time, the coordinates (xi, yi)of each marked cell were plotted individually; (b) the top(luminal) and bottom (basal) side of the epithelium were markedmanually [the coordinates of the top marker were (x0, y0), andthose of the bottom marker were (xt, yt)]; and (c) r i 5 (yi 2yt)/(y0 2 yt), where r represented the relative value of thevertical position of immunoreactive cells in the epithelium. Themean and SD ofr values of all of the immunoreactive cells wasdefined as the TI, and this index reflected the topology of theimmunoreactive area in the epithelium (Fig. 1).

Statistical Analysis. Differences between groups wereexamined for statistical significance using the unpairedt test(Student, Welch) and thex2 test. AP less than 0.05 denoted thepresence of a statistically significant difference.

3 http://rsb.inf.nih.gov/nih.image/.

Table 1 Summary of p21WAF1/CIP1, CK10, p53 expression in superficial SCC

Lesion no. Age Sex DifferentiationDepth ofinvasion Therapy p21 expression CK10 expression p53 expression

1 74 M Well ma EMR Scatteredb Positive Positive2 49 M Well m Operation Scattered Positive Positive3 68 M Well m Operation Scattered Positive Positive4 60 F Well m Operation Scattered Positive Positive5 62 M Well sm Operation Scattered Positive Positive6 54 M Well sm Operation Scattered Negative Positive7 68 M Well sm Operation Scattered Negative Positive8 69 M Well sm Operation Scattered Positive Positive

9 52 M Moderately m EMR Scattered Positive Positive10 51 M Moderately m Operation Negative Negative Positive11 74 M Moderately m Operation Scattered Positive Positive12 63 M Moderately sm Operation Negative Negative Positive13 63 M Moderately sm Operation Negative Negative Positive14 52 M Moderately sm Operation Scattered Negative Negative15 66 M Moderately sm Operation Scattered Positive Positive

16 68 F Poorly m Operation Negative Negative Positive17 49 M Poorly m EMR Negative Negative Positive18 60 M Poorly m Operation Topc Negative Positive19 74 M Poorly m Operation Negative Negative Negative20 62 M Poorly m EMR Top Negative Positive21 48 M Poorly m Operation Top Negative Positive22 71 M Poorly m Operation Top Negative Positive23 62 M Poorly sm Operation Negative Negative Positive24 59 M Poorly sm Operation Negative Negative Positive25 69 M Poorly sm Operation Negative Negative Positive

a m, mucosal carcinoma; sm, submucosal carcinoma.b Scattered, expression scattered in the tumor.c Top, expression limited to the top layer of the tumor.

543Clinical Cancer Research



RESULTSDistribution of p21WAF1/CIP1, p53, and Ki67 in Esoph-

ageal Superficial SCC. p53 was overexpressed in 23 (92%)of 25 esophageal superficial SCCs. In most lesions, the distri-bution of p53-expressing cells was similar to that of Ki67-expressing cells; they were mainly located in the deep layers ofthe tumor. This distribution was independent of the grade oftumor differentiation (Table 1; Figs. 2 and 3).

p21WAF1/CIP1 immunoreactivity was detected in 13

(59%) of 25 of esophageal superficial SCCs. Expression ofp21WAF1/CIP1 correlated with the grade of differentiation.Expression of p21WAF1/CIP1 was high in 8 (100%) of 8well-differentiated and 4 (57%) of 7 moderately differenti-ated but was faint in 4 (40%) of 10 poorly differentiatedsuperficial SCCs. The latter pattern of expression was alsolimited to the top layer of the tumor. In differentiated SCCs,p21WAF1/CIP1-expressing cells were mostly scattered in theupper and middle layers of the tumor and were considerably

Fig. 2 Expression of p21 (a),Ki67 (b), p53 (c), and CK10 (d)in well-differentiated esopha-geal superficial SCCs. Ki67-and p53-expressing cells are lo-cated mainly in the deep layersof the tumor, and the distribu-tion of p21WAF1/CIP1-expressingcells shows overexpression andmixing of these cells with Ki67expressing cells. The presenceof CK10 expression is notable.

Fig. 3 Expression of p21WAF1/CIP1

(a), Ki67 (b), p53 (c), and CK10(d) in poorly differentiated esoph-ageal superficial SCC. Similar towell-differentiated lesions, Ki67-and p53-expressing cells aremainly located in the deep layersof the tumor. However, faintp21WAF1/CIP1-expression wasfound only in the top layers of thetumor. There is a lack of CK10expression.




intermingled with those of Ki67-expressing cells. However,p21WAF1/CIP1and Ki67 expression was mutually exclusive atthe cellular level. In keratinized areas, peripheral cell layerswere positive for Ki67, whereas adjacent internal layers werepositive for p21WAF1/CIP1 (Table 1; Figs. 2 and 3).

Distribution of p21WAF1/CIP1, p53, and Ki67 in Esoph-ageal Squamous Dysplasia. The percentages of esophagealsquamous dysplastic lesions that were positive for p21WAF1/CIP1

and p53 were 92% (66 of 72) and 78% (56 of 72), respectively.We were able to use 48 (67%) of 72 lesions for image analysesof various biomarkers in the following studies. These consisted

of 12 mild, 16 moderate, and 20 severe dysplastic lesions(Table 2).

No p53-expressing cells were detected in eight dysplasias,and there was no significant difference between the TI of p53and that of Ki67 in 82% (33 of 40) of p53-positive lesions. Withthe progression of dysplasia, the distribution of Ki67- and p53-expressing cells expanded in the lower layers of the epitheliumin many lesions (Table 2; Fig. 4).

Overall, the distribution of p21WAF1/CIP1-expressing cellsshifted to the upper layers of the epithelium with the progressionof dysplasia. Furthermore, there were significant differences in

Table 2 Summary of p21WAF1/CIP1, Ki67, p53, and CK10 expression in dysplasia

Lesion no. Age/Sex Grade of dysplasia TI of p212 TI of Ki67p53 expression

(TI of p53 2 TI of Ki67) CK10 expression

1 40/M Mild nsa Positive (ns) Positive2 75/M Mild ns Positive (ns) Negative3 64/M Mild P , 0.05 Positive (ns) Negative4 72 /F Mild ns Positive (ns) Negative5 53/M Mild P , 0.05 Positive (ns) Negative6 74/M Mild P , 0.01 Positive (ns) Negative7 52/M Mild P , 0.01 Positive (ns) Negative8 67/M Mild P , 0.01 Positive (ns) Negative9 68/M Mild P , 0.01 Positive (ns) Negative

10 40/M Mild P , 0.01 Positive (ns) Negative11 69/M Mild P , 0.01 Positive (ns) Negative12 54/M Mild P , 0.01 Positive (ns) Negative

13 54/M Moderate ns Positive (ns) Negative14 48/M Moderate P , 0.01 Positive (P, 0.01) Negative15 53/M Moderate ns Positive (ns) Negative16 50/M Moderate ns Positive (ns) Positive17 71/M Moderate P , 0.01 Positive (P, 0.01) Negative18 68/M Moderate ns Positive (ns) Negative19 71 /F Moderate ns Positive (ns) Positive20 71/M Moderate P , 0.01 Positive (P, 0.01) Negative21 67/M Moderate P , 0.01 Positive (ns) Negative22 76 /F Moderate ns Negative Positive23 63/M Moderate ns Positive (ns) Negative24 53/M Moderate P , 0.01 Positive (P, 0.01) Negative25 65/M Moderate ns Negative Positive26 38/M Moderate P , 0.01 Positive (P, 0.05) Negative27 64/M Moderate ns Negative Negative28 54/M Moderate P , 0.01 Positive (ns) Negative

29 40/M Severe ns Negative Positive30 65/M Severe ns Positive (ns) Negative31 55/M Severe ns Negative Negative32 52/M Severe ns Negative Negative33 60 /F Severe ns Positive (ns) Positive34 71 /F Severe ns Positive (ns) Positive35 74/M Severe ns Positive (ns) Positive36 59/M Severe ns Positive (ns) Negative37 65/M Severe ns Negative Positive38 65/M Severe P , 0.01 Positive (ns) Negative39 60/M Severe ns Negative Positive40 71/M Severe P , 0.01 Positive (ns) Negative41 67/M Severe P , 0.01 Positive (ns) Negative42 74/M Severe P , 0.01 Positive (ns) Negative43 68/M Severe P , 0.01 Positive (ns) Negative44 56/M Severe P , 0.01 Positive (ns) Negative45 66/M Severe P , 0.01 Positive (ns) Negative46 72 /F Severe P , 0.01 Positive (ns) Positive47 62/M Severe P , 0.01 Positive (ns) Negative48 72 /F Severe P , 0.01 Positive (ns) Negative

a ns, not significant.




the meanr-values of p21WAF1/CIP1-expressing cells among eachgrade (P, 0.01). However, this change was heterogeneous. Insome lesions, p21WAF1/CIP1-expressing cells were found in afew lines among the superficial layers of atypical cells, whereasin others, cells overexpressing p21WAF1/CIP1 were scatteredamong atypical cells. In the latter group, p21WAF1/CIP1-express-ing cells were mixed with those of Ki67-expressing cells. Therewere 22 dysplastic lesions (46%) in which no significant differ-ence was detected in TI between p21WAF1/CIP1-expressing cellsand Ki67-expressing cells. However, similar to superficial SCC,

p21WAF1/CIP1 and Ki67 expression was mutually exclusive tothe cellular level in dysplastic lesions (Table 2; Figs. 5 and 6).

CK10 Expression in Esophageal Superficial SCC andDysplasia. In superficial SCC, CK10 expression correlated withdifferentiation. CK10 immunoreactivity was detected in 6 (75%) of8 well-differentiated, 3 (43%) of 7 moderately differentiated, and 0(0%) of 10 poorly differentiated superficial SCCs (Table 1). Indysplastic lesions, CK10 expression was detected in 12 (25%) of48 lesions [1 (8%) of 12 mild, 4 (25%) of 16 moderate, and 7 (35%)of 20 severe dysplastic lesions (Table 2; Fig. 7)].

Fig. 4 TI of Ki67- and p53-express-ing cells in esophageal squamousdysplasia. No p53-expressing cellswere detected in 8 dysplastic lesions,and there was no significant differ-ence between the TI of p53-express-ing cells and the TI of Ki67-express-ing cells in 33 dysplastic lesions.With the progression of dysplasia,the topological distribution of Ki67-and p53-expressing cells overlappedwith each other and expanded in thelower layer of the epithelium in manylesions.

Fig. 5 TI of p21WAF1/CIP1-express-ing cells in esophageal squamousdysplasia. Overall, the topologicaldistribution of p21WAF1/CIP1-ex-pressing cells shifted significantly tothe upper layers of the epithelium(P , 0.01).




Relationship between Distribution of p21WAF1/CIP1-ex-pressing Cells and Proliferation and Differentiation ofEsophageal Squamous Dysplasia.The relationship amongheterogeneous p21WAF1/CIP1 expression, differentiation, andproliferation was investigated in severe dysplastic lesions (n520). As stated earlier, severe dysplastic lesions were classified inthe present study into two types: confined type and scattered

type. In the confined type (n5 10), p21WAF1/CIP1-expressingcells were confined to the upper one-half of layers ofthe epithelium, whereas in the scattered type (n5 10),p21WAF1/CIP1-expressing cells were found beyond the upperone-half of layers. To investigate the tendency of these twodifferent types of dysplasias for differentiation, serial sectionswere immunostained for CK10. There was a significant differ-

Fig. 6 p21WAF1/CIP1 expression and TI values in normal esophageal epithelium (a), moderate dysplasia (b,d) and severe dysplasia (c,e). Overall,p21WAF1/CIP1-expressing cells shifted to the upper layers of the epithelium. However, p21WAF1/CIP1expression in dysplasia was heterogeneous. In thepresent study, severe dysplastic lesions were classified as confined type (c) and scattered type (e) by estimating the topological distribution ofp21WAF1/CIP1-expressing cells.

Fig. 7 CK10 expression in se-vere dysplastic esophageal le-sions of scattered type.




ence in CK10 expression between scattered type and confinedtype (P, 0.05). CK10-expressing cells were detected in 60% (6of 10) of scattered-type lesions but in only 1 (10%) of 10 lesionsin the confined-type lesions (Table 3; Fig. 7). Furthermore, thetopological distribution of CK10 seemed to overlap with that ofp21WAF1/CIP1 in dysplastic lesions (Figs. 6eand 7). To investi-gate the proliferative activity, we determined the percentage ofKi67-expressing cells in these lesions. There was no significantdifference in the percentage of Ki67 expressing cells betweenconfined and scattered lesions (Fig. 8).

DISCUSSIONIn the present study, we focused on the topological distri-

bution of p21WAF1/CIP1-, Ki67-, and p53-expressing cells inesophageal squamous dysplasia and superficial SCC to charac-terize esophageal squamous carcinogenesis, proliferation, anddifferentiation. In normal squamous epithelium of the esophagusand other tissues, p21WAF1/CIP1-expressing cells are locatedabove the proliferative compartment (which express Ki67) inparabasal layers (21), and cells with only faint expression of p53are detected on rare occasions (25). In the present study, thedistribution of Ki67 in many dysplastic lesions overlapped withthat of p53-expressing cells but expanded in the lower layers ofepithelium with the progression of dysplasia. On the other hand,Ki67- and p53-coexpressing cells were mainly located in thedeep layers of many superficial SCCs. Overall, in dysplasticlesions, the distribution of p21WAF1/CIP1-expressing cells shiftedto the upper layers of the epithelium with the progression of thegrade of dysplasia. In superficial SCCs, p21WAF1/CIP1-express-ing cells were mainly located in the upper layer of the tumor. Onthe basis of these findings, it seems that esophageal SCCs arisein the parabasal layer of the epithelium, and dysplasia is theearliest pathological process that grows upward toward theluminal surface. Our results also showed thatp53gene mutationcorrelated with esophageal squamous carcinogenesis and tumorprogression in the early stages in many lesions.

Heterogeneous expression of p21WAF1/CIP1 was observednot only in SCC but also in dysplasia. In esophageal superficialSCCs, the pattern and topological distribution of p21WAF1/CIP1

expression correlated with the grade of differentiation. In recentyears, a similar correlation between p21WAF1/CIP1 expressionand tumor differentiation has been reported in esophageal SCCs(23, 25) as well as in head and neck SCCs (24, 28), cutaneousSCCs, non-small cell lung carcinomas (32), and colon adeno-carcinomas (33). However, to our knowledge, there are noreports on the heterogeneous expression of p21WAF1/CIP1

in squamous dysplasia. Yanget al. (25) reported thatp21WAF1/CIP1-expressing cells were found surrounding the top

of the hyperplastic region. In other studies, p21WAF1/CIP1-ex-pressing cells were detected throughout the whole thickness ofthe epithelium in severe dysplasia (24, 28). Our results showedthat in some dysplastic lesions (confined type), p21WAF1/CIP1-expressing cells were present in a few lines in the superficiallayers of atypical cells similar to p21WAF1/CIP1-positive, poorlydifferentiated-type superficial SCCs. In dysplastic lesions of thescattered type, p21WAF1/CIP1-overexpressing cells were scat-tered among atypical cells similar to differentiated-type super-ficial SCCs. Furthermore, CK10, a marker of differentiation ofSCC, was detected in dysplastic lesions, and the pattern ofp21WAF1/CIP1 expression correlated with CK10 expression in amanner similar to that observed in SCCs. Taken together, thesefindings suggest that the tendency for differentiation of SCC isinitiated at the early stages of carcinogenesis, and p21WAF1/CIP1

expression seems to be involved in differentiation.Overexpression of p21WAF1/CIP1 is associated with sup-

pression of tumor growth in experimental models (18). How-ever, p21WAF1/CIP1expression might not overcome the progres-sion of esophageal squamous dysplasia and SCC. In dysplasia,the heterogeneity of p21WAF1/CIP1 expression did not correlatewith the proliferative activity assessed by the percentage ofKi67-expressing cells, and similar phenomenon has been de-scribed in SCC of different tissues using several proliferationmarkers (28, 34).

Several studies have demonstrated p53-independent ex-pression of p21WAF1/CIP1 in different normal and malignanttissues (21, 24, 34–36). However, in the present study, it wasnot clear whether p21WAF1/CIP1expression in squamous dyspla-

Fig. 8 Relationship between type of p21WAF1/CIP1 expression and per-centage of Ki67-expressing cells in severe dysplastic lesions. There wasno significant difference in the percentage of Ki67-expressing cellsbetween confined-type lesions and scattered-type lesions.

Table 3 Relationship between type of p21WAF1/CIP1 expression andCK10 expression in severe dysplastic lesionsa

p21 expression

CK10 expression

Positive Negative

Confined type (n5 10) 1 (10.0%) 9 (90.0%)Scattered type (n5 10) 6 (60.0%) 4 (40.0%)a P , 0.05 with statistic significant.




sia was independent ofp53 gene status. Although p53-positivesevere dysplastic lesions showed both confined-type and scat-tered-type p21WAF1/CIP1 expression, all of the p53 negativesevere dysplastic lesions showed scattered-type p21WAF1/CIP1

expression. We, therefore, speculate that the pattern ofp21WAF1/CIP1 expression in esophageal squamous dysplasiatends to depend onp53 gene status and function. However, itcould also be debated that p53 overexpression is not alwayssynonymous withp53 gene mutation. Overestimation ofp53gene mutation by immunohistochemistry could sometimes berelated to certain mechanisms including inactivation of theenzymatic pathway responsible for p53 degradation (37) orinactivation of wild-type p53 through binding with anotherprotein that prolongs its half-life (38). False-negative casesrepresent about 10% of all of thep53 gene mutations andpreferentially occur in certain types ofp53 gene mutation (39,40). Additional molecular biological studies are, thus, necessaryto clarify the mechanism of p21WAF1/CIP1 expression in differ-entiation of esophageal dysplasia and to address whetherp21WAF1/CIP1 expression alone could explain differentiation ofesophageal squamous dysplasia and esophageal SCC.

In conclusion, we suggest that esophageal squamous dys-plasia is the earliest pathological process in esophageal squa-mous carcinogenesis. The degree of differentiation of esopha-geal SCC may be determined at the stage of dysplasia and iscorrelated with p21WAF1/CIP1expression. It is, therefore, essen-tial to follow up these early lesions at appropriate intervals andto resect at an appropriate stage. Furthermore, p21WAF1/CIP1

expression may be a useful indicator for treating esophagealsquamous dysplasia.

ACKNOWLEDGMENTSWe thank Tae Yamanishi for the technical support and Satomi

Segawa for the preparation of the manuscript.

REFERENCES1. Itakura, Y., Sasano, H., Mori, S., and Nagura, H. DNA ploidy inhuman esophageal squamous dysplasias and squamous cell carcinomasas determined by image analysis. Mod. Pathol.,7: 867–873, 1994.2. Wang, L. D., Hong, J. Y., Qiu, S. L., Gao, H., and Yang, C. S.Accumulation of p53 protein in human esophageal precancerous lesions:a possible early biomarker for carcinogenesis. Cancer Res.,53: 1783–1787, 1993.3. Wang, L. D., Shi, S. T., Zhou, Q., Goldstein, S., Hong, J. Y., Shao,P., Qiu, S. L., and Yang, C. S. Changes in p53 and cyclin D1 proteinlevels and cell proliferation in different stages of human esophageal andgastric-cardia carcinogenesis. Int. J. Cancer,59: 514–519, 1994.4. Wang, D. Y. Image analysis of AgNORs in esophageal ballooncytosmears. Chung Hua Chung Liu Tsa Chih,15: 30–33, 1993.5. Harper, J. W., Adami, G. R., Wei, N., Keyomarsi, K., and Elledge,S. J. The p21 Cdk-interacting protein Cip1 is a potent inhibitor of G1

cyclin-dependent kinases. Cell,75: 805–816, 1993.6. el-Deiry, W. S., Tokino, T., Velculescu, V. E., Levy, D. B., Parsons,R., Trent, J. M., Lin, D., Mercer, W. E., Kinzler, K. W., and Vogelstein,B. WAF1, a potential mediator of p53 tumor suppression. Cell,75:817–825, 1993.7. Chin, Y. E., Kitagawa, M., Su, W. C., You, Z. H., Iwamoto, Y., andFu, X. Y. Cell growth arrest and induction of cyclin-dependent kinaseinhibitor p21 WAF1/CIP1 mediated by STAT1. Science (WashingtonDC), 272: 719–722, 1996.8. Datto, M. B., Li, Y., Panus, J. F., Howe, D. J., Xiong, Y., and Wang,X. F. Transforming growth factorb induces the cyclin-dependent kinase

inhibitor p21 through a p53-independent mechanism. Proc. Natl. Acad.Sci. USA,92: 5545–5549, 1995.

9. Liu, M., Iavarone, A., and Freedman, L. P. Transcriptional activationof the humanp21WAF1/CIP1gene by retinoic acid receptor. Correlationwith retinoid induction of U937 cell differentiation. J. Biol. Chem.,271:31723–31728, 1996.10. Liu, M., Lee, M. H., Cohen, M., Bommakanti, M., and Freedman,L. P. Transcriptional activation of the Cdk inhibitor p21 by vitamin D3

leads to the induced differentiation of the myelomonocytic cell lineU937. Genes Dev.,10: 142–153, 1996.11. Michieli, P., Chedid, M., Lin, D., Pierce, J. H., Mercer, W. E., andGivol, D. Induction of WAF1/CIP1 by a p53-independent pathway.Cancer Res.,54: 3391–3395, 1994.12. Parker, S. B., Eichele, G., Zhang, P., Rawls, A., Sands, A. T.,Bradley, A., Olson, E. N., Harper, J. W., and Elledge, S. J. p53-independent expression of p21Cip1 in muscle and other terminally dif-ferentiating cells. Science (Washington DC),267: 1024–1027, 1995.13. Somasundaram, K., Zhang, H., Zeng, Y. X., Houvras, Y., Peng, Y.,Zhang, H., Wu, G. S., Licht, J. D., Weber, B. L., and El-Deiry, W. S.Arrest of the cell cycle by the tumour-suppressor BRCA1 requires theCDK-inhibitor p21WAF1/CiP1. Nature (Lond.),389: 187–190, 1997.14. Tanaka, N., Ishihara, M., Lamphier, M. S., Nozawa, H., Mat-suyama, T., Mak, T. W., Aizawa, S., Tokino, T., Oren, M., and Tan-iguchi, T. Cooperation of the tumour suppressors IRF-1 and p53 inresponse to DNA damage. Nature (Lond.),382: 816–818, 1996.15. Timchenko, N. A., Wilde, M., Nakanishi, M., Smith, J. R., andDarlington, G. J. CCAAT/enhancer-binding proteina (C/EBPa) inhib-its cell proliferation through the p21WAF-1/CIP-1/SDI-1 protein. GenesDev., 10: 804–815, 1996.16. Zeng, Y. X., Somasundaram, K., and el-Deiry, W. S. AP2 inhibitscancer cell growth and activates p21WAF1/CIP1 expression. Nat. Genet.,15: 78–82, 1997.17. Noda, A., Ning, Y., Venable, S. F., Pereira-Smith, O. M., andSmith, J. R. Cloning of senescent cell-derived inhibitors of DNA syn-thesis using an expression screen. Exp. Cell Res.,211: 90–98, 1994.18. Halevy, O., Novitch, B. G., Spicer, D. B., Skapek, S. X., Rhee, J.,Hannon, G. J., Beach, D., and Lassar, A. B. Correlation of terminal cellcycle arrest of skeletal muscle with induction of p21 by MyoD. Science,267: 1018–1021, 1995.19. Jiang, H., Lin, J., Su, Z. Z., Herlyn, M., Kerbel, R. S., Weissman,B. E., Welch, D. R., and Fisher, P. B. The melanoma differentiation-associated genemda-6, which encodes the cyclin-dependent kinaseinhibitor p21, is differentially expressed during growth, differentiationand progression in human melanoma cells. Oncogene,10: 1855–1864,1995.20. Steinman, R. A., Hoffman, B., Iro, A., Guillouf, C., Liebermann,D. A., and el-Houseini, M. E. Induction of p21WAF-1/CIP1 during differ-entiation. Oncogene,9: 3389–3396, 1994.21. el-Deiry, W. S., Tokino, T., Waldman, T., Oliner, J. D., Velculescu,V. E., Burrell, M., Hill, D. E., Healy, E., Rees, J. L., Hamilton, S. R.,Kinzler, K. W., and Vogelstein, B. Topological control of p21WAF1/CIP1

expression in normal and neoplastic tissues. Cancer Res.,55: 2910–2919, 1995.22. Kitamura, K., Kuwano, H., Matsuda, H., Mori, M., and Sugimachi,K. Secondary early carcinoma in a reconstructed gastric tube aftersubtotal esophagectomy for early carcinoma of the esophagus. Hepato-gastroenterology,40: 44–47, 1993.23. Ohashi, K., Nemoto, T., Eishi, Y., Matsuno, A., Nakamura, K., andHirokawa, K. Expression of the cyclin dependent kinase inhibitorp21WAF1/CIP1 in oesophageal squamous cell carcinomas. VirchowsArch., 430: 389–395, 1997.24. Nadal, A., Jares, P., Cazorla, M., Fernandez, P. L., Sanjuan, X.,Hernandez, L., Pinyol, M., Aldea, M., Mallofre, C., Muntane, J.,Traserra, J., Campo, E., and Cardesa, A. p21WAF1/Cip1 expression isassociated with cell differentiation but not with p53 mutations in squa-mous cell carcinomas of the larynx. J. Pathol.,183: 156–163, 1997.25. Yang, G., Zhang, Z., Liao, J., Seril, D., Wang, L., Goldstein, S., andYang, C. S. Immunohistochemical studies on Waf1p21, p16, pRb and




p53 in human esophageal carcinomas and neighboring epithelia from ahigh-risk area in northern China. Int. J. Cancer,72: 746–751, 1997.

26. Sarbia, M., Stahl, M., zur Hausen, A., Zimmermann, K., Wang, L.,Fink, U., Heep, H., Dutkowski, P., Willers, R., Muller, W., Seeber, S.,and Gabbert, H. E. Expression of p21WAF1 predicts outcome of esoph-ageal cancer patients treated by surgery alone or by combined therapymodalities. Clin. Cancer Res.,4: 2615–2623, 1998.

27. Seta, T., Imazeki, F., Yokosuka, O., Saisho, H., Suzuki, T., Koide,Y., and Isono, K. Expression of p53 and p21WAF1/CIP1proteins in gastricand esophageal cancers: comparison with mutations of thep53 gene.Dig. Dis. Sci.,43: 279–289, 1998.

28. van Oijen, M. G., Tilanus, M. G., Medema, R. H., and Slootweg,P. J. Expression of p21Waf1/Cip1 in head and neck cancer in relation toproliferation, differentiation, p53 status and cyclin D1 expression.J. Oral Pathol. Med.,27: 367–375, 1998.

29. Miettinen, M. Keratin immunohistochemistry: update of applica-tions and pitfalls. Pathol. Annu.,28: 113–143, 1993.

30. Lam, K. Y., Loke, S. L., Shen, X. C., and Ma, L. T. Cytokeratinexpression in non-neoplastic oesophageal epithelium and squamous cellcarcinoma of the oesophagus. Virchows Arch.,426: 345–349, 1995.

31. Watanabe, H., Jass, J. R., and Sobin, L. H. Histological typing ofoesophageal and gastric tumors in collaboration with pathologist in 8countries.In: WHO International Classification of Tumours, Ed. 2,pp.11–18. Berlin: Springer-Verlag, 1990.

32. Marchetti, A., Doglioni, C., Barbareschi, M., Buttitta, F., Pellegrini,S., Bertacca, G., Chella, A., Merlo, G., Angeletti, C. A., Dalla Palma, P.,and Bevilacqua, G. p21 RNA and protein expression in non-small celllung carcinomas: evidence of p53-independent expression and associa-tion with tumoral differentiation. Oncogene,12: 1319–1324, 1996.

33. Doglioni, C., Pelosio, P., Laurino, L., Macri, E., Meggiolaro, E.,Favretti, F., and Barbareschi, M. p21/WAF1/CIP1 expression in normalmucosa and in adenomas and adenocarcinomas of the colon: its rela-tionship with differentiation. J. Pathol.,179: 248–253, 1996.34. Erber, R., Klein, W., Andl, T., Enders, C., Born, A. I., Conradt, C.,Bartek, J., and Bosch, F. X. Aberrant p21CIP1/WAF1protein accumulationin head-and-neck cancer. Int. J. Cancer,74: 383–389, 1997.35. Sheikh, M. S., Li, X. S., Chen, J. C., Shao, Z. M., Ordonez, J. V.,and Fontana, J. A. Mechanisms of regulation ofWAF1/Cip1 geneexpression in human breast carcinoma: role of p53-dependent andindependent signal transduction pathways. Oncogene,9: 3407–3415,1994.36. Cox, L. S. Multiple pathways control cell growth and trans-formation: overlapping and independent activities of p53 andp21Cip1/WAF1/Sdi1. J. Pathol.,183: 134–140, 1997.37. Wynford-Thomas, D. p53 in tumour pathology: can we trust im-munocytochemistry? J. Pathol.,166: 329–330, 1992.38. Bennett, W. P., Hollstein, M. C., Metcalf, R. A., Welsh, J. A.,He, A., Zhu, S. M., Kusters, I., Resau, J. H., Trump, B. F., Lane,D. P., and Harris, C. C.p53 mutation and protein accumulationduring multistage human esophageal carcinogenesis. Cancer Res.,52: 6092– 6097, 1992.39. Wagata, T., Shibagaki, I., Imamura, M., Shimada, Y.,Toguchida, J., Yandell, D. W., Ikegata, M., Tobe, T., and Ishizaki, K.Loss of 17p, mutation of thep53 gene, and overexpression of p53protein in esophageal squamous cell carcinoma. Cancer Res.,53:846 – 850, 1993.40. Dix, B., Robbins, P., Carello, S., House, A., and Iacopetta, B.Comparison ofp53 gene mutation and protein overexpression in colo-rectal carcinoma. Br. J. Cancer,70: 585–590, 1994.




2000;6:541-550. Clin Cancer Res Yasuhiro Shirakawa, Yoshio Naomoto, Masashi Kimura, et al. Esophageal Squamous Dysplasia

Expression inWAF1/CIP1Topological Analysis of p21

Updated version

http://clincancerres.aacrjournals.org/content/6/2/541

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/6/2/541.full#ref-list-1

This article cites 37 articles, 13 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/6/2/541.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] 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://clincancerres.aacrjournals.org/content/6/2/541To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/content/6/2/541.full#ref-list-1

http://clincancerres.aacrjournals.org/content/6/2/541.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Documents

Topological Analysis of p21 Expression in Esophageal Squamous Dysplasia€¦ · Yasuhiro Shirakawa,1 Yoshio Naomoto, Masashi Kimura, Ryuichi Kawashima, Tomoki Yamatsuji, Takahiko