[CANCERRESEARCH55, 237-241,January15,19951

Advances in Brief

bcl-2 and p53 Oncoprotein Expression during Colorectal Tumorigenesis

Frank A. Sinicrope,1 San Bao Ruan, Karen R. Cleary, L. Clifton Stephens, J. Jack Lee, and Bernard Levin

Departments of Gastrointestinal Medical Oncology and Digestive Diseases (F. S., B. Li and Biomathematics (J. J. Li and the Divisions of Pathology (K. R. C.J and VeterinaryMedicine and Surgery (L C. 5.1, The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030

Abstract

Apoptosls or programmed cell death represents a mechanism by whichcells possessing DNA damage can be deleted. The bcl-2 proto-oncogene isa known inhibitorof apoptosisthat may allow the accumulationandpropagation of cells containing genetic alterations. To determine if andwhen the bcl-2 gene is activated during colorectal tumorigenesis and ItsrelatIonship to p53, we analyzed normal mucosa, hyperplastic and dysplastic epltbellal polyps, and carcinomas for the expression ofthese markera using Immunohistochemistry. Whereas bcl-2 staining was restiicted tobasal eplthelial cells In normal and hyperplastic mucosa, bcl-2 expressionwasdetectedin parabasalandsuperficialregIonsIndysplasticpolypsandcarclnoma& An inverse correlation was found between bcl-2 and p53expression In adenomas, suggesting that these markers may regulate acommon cell death pathway. Furthermore, carcinomas with a high percentage of bcl-2-posltlvecells were significantlymore likely to have lowrates of spontaneous apoptosis, as determined histologically, than thosecancers with low or absent bcl-2 expression. Abnormal actIvatIon of thebcl-2 gene appears to be an early event In colorectal tumorigenesis thatcan InhibIt apoptosis in vivo and may facilitate tumor progression.

Introduction

Colorectal carcinoma is the second leading cause of cancer mortality in the United States, accounting for an estimated 56,000 deathsand 149,000 new cases in 1994 (1). Colorectal cancer is believed toresult from a series of genetic alterations that disorder normal mechanisms controlling cell growth (2, 3). Although chromosomal andDNA abnormalitiesare frequentevents in colorectal cancers (2—4),factors which augment the acquisition and proliferation of cells contaming genetic damage are poorly understood. Apoptosis or programmed cell death appears to be an important mechanism by whichtumors delete cells that have sustained carcinogen-induced DNAdamage (5, 6). The bcl-2 proto-oncogene is a known inhibitor ofapoptosis and may thereforeallow an accumulationof genetic alterations that become propagated by cell division and potentially contribute to neoplastic development. The bcl-2 gene encodes a Mr25,000 protein that localizes to the mitochondrial membrane, nuclearenvelope, and endoplasmic reticulum (7, 8). First described in follicular lymphoma, overexpression of the bcl-2 gene was shown to resultfrom a chromosomal 14;18 translocation (9). Extension of cell survival by bcl-2 has been demonstrated in hematopoietic and epithelialcell lines (10—13)and in transfection experiments (10). Moreover, intransgenic mice that overexpress bcl-2 in hematopoietic cells andlymphoidtissues, lymphocytesdisplayextendedsurvivalandprogressto high-grade lymphoma (14). Whetherbcl-2 contributesto the developmentor progressionof epithelialneoplasms,however, is largelyunknown. Furthermore, no data exist concerning the relationshipbetween bcl-2 expression and spontaneous apoptotic rates in epithelialcancers.

Received 9/6/94; accepted 12/2/94.The costs of publicationof this articlewere defrayedin partby the paymentof page

charges.This articlemust thereforebe herebymarkedadvertisementin accordancewith18 U.S.C. Section 1734 solely to indicatethis fact.

1 To whom requests for reprints should be addressed, at Department of Gastrointestinal

Oncology and Digestive Diseases, 1515 Holcombe Blvd., Box 78, Houston,TX 77030.

237

The p53 tumor suppressor gene is the most commonly mutatedgene in humancancer(15) and is a frequentabnormalityin colorectalcancers (15—17).bcl-2 has been shown to inhibit apoptosis triggeredby wild-type p53 (18) and mutantp53 also appears to inhibit apoptosis(19). An inverse relationshipbetween bcl-2 and p53 has been obsewed in human breast cancers and suggests that these oncoproteinsmay interact in the regulation of apoptosis (20). Colorectal neoplasiaprovides an ideal model to study tumor progression given that mostcarcinomas appear to arise from adenomas (2, 3) and tumors atvarious stages of development can be analyzed. Therefore, we arialyzed bcl-2 expression in normal,hyperplastic,dysplastic, and malignant colorectal epithelia and determined the timing of bcl-2 activationand its relationship to p53 alterations.

Materials and Methods

Tissue Samples.Formalin-fixed,paraffin-embeddedtumorblockswereobtained from the Pathology Department at the University of Texas M. D.Anderson Cancer Center. Nonconsecutive patient specimens were identifiedretrospectively and 4—6-p@m-thicksections were cut from their respectivetissue blocks. Biopsy specimens from 18 adenomas, 15 hyperplastic polyps,and 6 combined colorectal polyps were obtained. Within the same patient, theconcurrent or synchronous presence of a colorectal cancer and a mucosal polypwere noted in 3 of 15 patients with hyperplastic polyps and 7 of 18 adenomas.Adenomatous polyps are dysplastic lesions which represent neoplastic epithehal proliferation and are believed to precede frank malignancy (2). Combinedpolyps contain both hyperplastic and adenomatous elements (21). Twenty-onesurgically resected and untreated primary colorectal adenocarcinomas werealso examined. Most of the cases were Dukes' stage C tumors, i.e., invasion ofthe muscularis propria with regional lymph node involvement (22). Histologically normal mucosa was analyzed from the proximal or distal resectionmargins of these cases. The diagnosis of carcinoma and an assessment ofhistological grade were determined by examination of hematoxylin and eosinstained sections from the same block of each case.

Five of 15 hyperplastic polyps, 14 of 18 adenomas, and 2 combined polypswere 1 cm in size (Tables 1 and2). The adenomaswere tubularor tubulovillous and none had pure villous histology. Three adenomas had focal highgrade dysplasia and the remainder had low-grade dysplasia. Synchronouscolorectal carcinomas were present in 3 of 15 hyperplastic polyps and 7 of 18adenomas.Of the carcinomas,3 were well, 12 were moderate,and 5 werepoorly differentiated (Table 2).

Apoptotic IndeL The percentage of apoptotic cells was determined bymicroscopic examination of hematoxylin and eosin-stained sections at X400.Coded slides were scored by an examiner who was blinded to the immunostaining results. For each slide, five fields of nonnecrotic areas of the carcinoma were examined and 100 nuclei in each field were scored as normal orapoptotic in appearance.Morphologicalfeatures used to identify apoptoticnuclei included overall shrinkage and homogeneous dark basophiia as described previously (23).

Immunohlstochemistry. An immunoperoxidase method using an avidinbiotinylated horseradish peroxidase complex (Vectastain Elite ABC; VectorLaboratories, Burlingame, CA) was used to detect bcl-2 and p53 proteins indeparaffinized tissue sections (24). The primary anti-bcl-2 mAb (Dako Corporation,Carpenteria,CA) is a murineantihumanmAb, subclassIgGi, thatrecognizes a cytoplasmic epitope of bcl-2 (25). The anti-p53 mAbs DOl and1801(OncogeneScience,Inc.,Manhasset,NY)are murineantilsumanmAbs,subclass IgG2A and IgOl, respectively, that recognize an amino-terminal

on July 10, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table 1 Relationship between bcl-2 and p53 immunostaining and clinicopathologicalfeatures of colorectal po!vps

Colorectal polyps have been grouped according to histological type. The results ofbcl-2 and p53 staining are related to polyp type and clinical features.

Tumor stage and histopathological features of colorectal adenocarcinomas arerelatedtobcl-2 and p53 staining and the spontaneous apoptotic index of eachtumor.bcl-2

p53CancerDukes' Histological weighted weightedApoptoticPatient

location stage° grade score score index(%)40

Distal B2 Moderate 0 04.041Distal C2 Moderate 8 01.242DiStal D Moderate 0 128.043Proximal C2 Moderate 9 00.244Proximal C2 Moderate 12 122.045Distal B2 Moderate 41246Distal C2 Moderate 6 00.247Distal C2 Moderate 8 120.448Rectum C2 Moderate 12449Distal C2 Poor 8 00.650Proximal C2 Moderate 12 9051Distal C2 Moderate 8 120.452

Distal D Poor 0 03.453Proximal C2 Well 3 11.054Proximal C2 Moderate 0 02.855Proximal C2 Poor 0 120.656Rectum C2 Well 0 05.857Distal C2 Poor 8 00.858Proximal C2 Moderate 12 80.459

Proximal C2 Poor 0 02.260Proximal Cl Well 12 0 7.0

bcl-2 AND p53 EXPRESSION IN COLORECTAL NEOPLASIA

Statistics. The association between bcl-2 and p53 staining, and with theapoptotic index, was analyzed using the Pearson's correlation coefficient. Therelationship between oncoprotein staining and polyp size and the presence orabsence of synchronous cancers was determined using the Wilcoxon sum ranktest. Lastly, the Kruskal-Wallis one-way ANOVA was used to analyze therelationship between bcl-2 and p53 staining and histological grade. Statistical

significance was defined as P 0.05.

Results

Anti-bcl-2Immunostaining

Normal Mucosa. In histologically normal mucosa, bcl-2 stainingwas observed in basal epithelial cells of the colonic crypts overlyingthe muscularis mucosae but was absent in superficial portions of themucosa. bcl-2 staining in basal epithelial cells was intense (3+) andlocalized to the cytoplasm, and in some cases nuclear membranestaining was also observed (Fig. 1B). Intense (3+) immunostaining ofthe nuclear membrane of lymphocytes in the lamina propria wasobserved in normal sections and in infiltrating lymphocytes withintumor stroma (Fig. 1C). Similar to other reports (26), neurons dem

onstrated intense positivity for bcl-2. In addition, we consistentlyobserved anti-bcl-2 immunoreactivity in smooth muscle cells of thearterial wall and the muscularis mucosae and muscularis propria of thebowel (Fig. 1B). Anti-bcl-2 immunoreactivity in smooth muscle cells

was not observed in the negative control slides wherein the primarymAb was omitted. As with neurons, smooth muscle cells are longlived postmitotic cells in which bcl-2 expression may also function toextend their survival.

Hyperplastic Polyps. As in normal mucosa, bcl-2 staining wasrestricted to basal epithelial cells of the lower colonic crypts. Many of

the basal crypts were histologically normal but staining of basal cellsin hyperplastic crypts was also observed (Fig. 1). More superficialintestinal cells and mucin cells were uniformly negative for bcl-2 inthe 15 pure hyperplastic polyps and in the hyperplastic portion ofcombined polyps (Fig. 1C). In the combined polyps, positive bcl-2staining highlighted dysplastic glands and allowed them to be easilydistinguished from negatively stained hyperplastic glands.

Adenomas. Dysplasticcells stainedpositivelyforbcl-2proteinin 17of 24 (71%) adenomas and combined polyps Fig. 1D). In 10 of 17(59%)

PatientPolyp typePolyplocationPolypsize

l cmSynchronouscarcinomabcl-2weighted

scorer'p53weighted

score'@1HyperplasticDistal+—002HyperplasticDistal——013HyperplasticDistal——014HyperplasticDistal+—005HyperplasticProximal—+006HyperplasticDistal——007HyperplasticDistal—+008HyperplasticDistal+—009HyperplasticDistal——0010HyperplasticProximal——001

1HyperplasticDistal——0012HyperplasticDistal——0013HyperplasticRectum+-0214HyperplasticDistal++0015HyperplasticDistal——0016CombinedProximal——3017CombinedProximal——0318CombinedDistal——9619CombinedRectum+—4020CombinedProximal——0021CombinedProximal——0022―AdenomaProximal—+81223AdenomaRectum++0224cAdenomaRectum++01225AdenomaRectum++01226AdenomaProximal+—12027AdenomaDistal+—6028AdenomaDistal+—8329AdenomaDistal+—12030'AdenomaRectum+—6031AdenomaRectum+—9232AdenomaDistal+—12033AdenomaDistal++6034AdenomaProximal——4035AdenomaDistal+—12336AdenomaProximal++0937AdenomaDistal+—12338AdenomaProximal——4039bAdenomaDistal—+120

a Weighted score represents the product of staining intensity and percentage of tumor

cell positivity (see “Materialsand Methods―).b Adenomas obtained from the same patient.C Focus of severe mucosal dysplasia.

Table2 Relationshipbetweenbcl-2 and p53 immunostainingand clinicopathologicalfeaturesofadenocarcinomasepitope of the human p53 molecule. Anti-p53 mAbs DOl and 1801 were

compared in this study and greater than 95% agreement was observed. Allreported p53 data were obtained using mAb DOl. For all mAbs, staining wasdeveloped by immersing slides in 0.06% 3,3'-diaminobenzidine tetrahydrochloride (Sigma Chemical Co., St. Louis, MO), and the slides were counterstained in hematoxylin (Richard Allen, Inc., Richland, MI) and 0.3% ammoniawater and then mounted using an aqueous medium.

Controls. Formalin-fixed, paraffin-embedded sections of normal humanlymph node served as a positive control for bcl-2 in all slide runs. A colorectalneoplasm with a high level of nuclear p53 immunoreactivity was used as apositive control for this oncoprotein. Negative control slides were processedwith each slide run and excluded the primary antibody but included all other

steps of the procedure.ScoringMethods. Positivetumorcellswerequantifiedby twoindependent

observers, expressed as the percentage of the total number of tumor cells, andassigned to one of 5 categories: 0, 5%; 1, 5—25%;2, 25—50%;3, 50—75%;and 4, 75%. For both oncoproteins, 5% positive cells was used as thecutoff to define negative tumors. The intensity of bcl-2 and p53 immunostaining was scored as: 1+, weak; 2+, moderate; and 3+, intense. The percentageof positivity of tumor cells and staining intensity were multiplied to produce aweighted score for each tumor specimen. For bcl-2, intensity was judgedrelative to the basal epithelial cells or to infiltrating lymphocytes which also

served as internal positive controls and were arbitrarily designated as 3+. ForPS3, intensity was judged relative to an intensely stained p53-positive controlcolorectal neoplasm.

a Adenocarcinomas were staged according to the Astler-Coller modification of the

Dukes' classification (21).

238

on July 10, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

A@ ‘@a@@‘. . S. ,,,,@

..@ j

. ‘@%.:.1@::.. . -

. ‘.@@

--@@

- .—.‘-..‘.-@

-@ .:@:i.:@@x :@

..@ •.@•

@ . ,- .@-c •@r:@@. , . . .@ . i..,.- ‘.@.

D@@@

, ,. . I.-.., -2@ •-@‘ ‘ - . ‘@

#@4;1 fi@cE@ ‘-@

. : ,@‘@@ ., . .,@

p

@@@ ;@JAt@d@;

Fig. 1. A, section of colorectal carcinoma stained immunohistochemically but excluding the anti-bcl-2mAb, i.e., negative control. X 63. B,normal human colonic mucosa stainedfor the bcl-2 protein demonstratesbcl-2 positivity (brown reaction product) in basal epithelial cells of thelower colonic crypts and absence ofthe bcl-2 protein in superficialcrypts. X 100. In C, hyperplastic cobrectal epithelial cells stain negativelyfor the bcl-2 protein in contrast to lymphocytes in the lamina propria, whichdisplay intense (3+) bcl-2 positivity. X 200. In D, dysplastic cells inadenomatous cobonic glands show diffuse positivity for bcl-2. X 100. In E,malignantcobonicglandsdemonstratediffuse cytoplasmic immunostainingfor bcl-2 protein in the adenocarcinoma cells. x 200. F, p53-positivenuclear staining in epitheial cells of ahyperplastic coborectalpolyp. X 150.

,,,:@@

-@@: ‘@

I,

•.1

,_ a―

....o,

aIr@c.•., . .

S .“@@!@ : ‘ -

“I? ‘ •.@ ‘@.ta@:@ S@,‘ 4 ‘ P@: ‘@ •

@;@ ...@@

.— . - .,@ - .., @l @; . rn@ • •‘@

, .

@, I J@@@

cases (2 adenomas from same patient), greater than 75% of dysplasticcells stained positive for bcl-2. Furthermore, 6 of these 10 (60%) caseshad the highest possible weighted score (Fable 1). Anti-bcl-2 immunostaming was localized to the cytoplasm in all cases. Severe epithelialdysplasia was present in 3 cases, 1 of which was bcl-2 negative and theother two of which were bcl-2 positive (Table 1).

Adenocarcinomas. In 14 of 21 (67%) cases, bcl-2 staining wasdetected in carcinoma cells in the malignant glands (Table 2). Diffuse(75% of cells), homogeneousbcl-2 immunostainingwas observedin 11 of 14 (78%) positive cases and the highest possible weightedscore was noted in 5 of 14 (36%). Localization of staining waspredominantly cytoplasmic and in several cases also involved thenuclear membrane (Fig. 1E). Furthermore, cytoplasmic immunoreactivity was most prominent in the apical portion of the tumor cells(Fig. 1E).

Anti-p53 Immunostaining

Normal Mucosa. Normal mucosa was uniformly negative for fluclear p53 immunostaining.

Hyperplastic Polyps. Nuclear p53 staining was detected in 3 of15 (20%) pure hyperplastic polyps but was not observed in hyperplastic tissue in the combined polyps. In the 3 positve cases, p53staining was weak (1+) and focal, involving <25% of cells (2cases) and 25—50% of cells (1 case) (Fig. iF). The latter p53-positive case was a large ( 1 cm) rectal hyperplastic polyp and theother two were <1-cm polyps from the distal colon. Positively

stained cells in these hyperplastic polyps were not dysplastic andnone of the cases were from patients with synchronous colorectalcancers.

Adenomas. Nuclear p53 staining was detected in 9 of 24 (37%)adenomas, including 2 of 6 combined polyps with positive staining inthe dysplastic cells. Excluding combined polyps, 14 of 18 (78%)adenomas were 1 cm in size and 5 of 9 (56%) p53-positive caseswere from patients with synchronous colorectal cancers (Table 1). In4 of these 5 cases, 50% of dysplastic cells stained positive formutant p53 protein. In contrast, focal p53 staining was detected inp53-positive adenomas from patients without synchronous cancers.

Adenocarcinomas. Consistentwith otherstudies (16, 17), nuclearp53 positivity was found in 10 of 21 (48%) colorectal carcinomas(Table 2). The highest possible weighted score was observed in 6(60%) of these cases and p53 staining was detected in 75% of tumorcells in 7 (70%) of cases. Immunostaining for p53 was not significantly related to histological grade.

bcl-2 and p53 Stainingand ClinicopathologicalFeatures. Adenomas from patients with synchronous colorectal cancers had significantly less bcl-2 staining compared with those from patients withoutsynchronous cancers; the converse was noted for p53 staining (Table3). Importantly, a significant inverse correlation was found betweenbcl-2 and p53 in immunostaining in the adenomas but not in the

carcinomas (Table 4). No association was observed between the bcl-2or p53 weighted score and adenoma size, tumor location, or histological grade of the carcinomas.

239

on July 10, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Adenomabcl-2 weighted score―@53 weightedscore―Synchronous

carcinomaWith

Without3.7 8.8P=0.046.70.45P=0.02Adenoma

size1 cm

<1 cm6.8 7.0P=0.913.33.0P=0.52a

Mean value of product of staining intensity and turnor cell positivity.

Table 4 Correlation between bcl-2 and p53 staininga

The correlationbetween bcb-2and p53 immunostalningresults in premnomas and colorectal carcinomas was determined as shown below.alignant

ads

Pearson's correlationn coefficient(r)PAdenomas

18 —0.58Carcinomas 21 0.170.010.45a

Weighted score values.

bct-2 AND p53 EXPRESSIONIN COL0REC@ALNEOPLASIA

Table 3 Relationship between bcl-2 and p53 staining and clinicopathological variablesThe results of bcb-2and p53 staining in adenomatous polyps are related to the presence

or absence of a synchronous coborectalcarcinoma and polyp size using the Wilcoxon sumrank test.

mutant p53 in premalignant polyps. Similarly, an inverse correlationbetween bcl-2 and p53 staining has been reported in human breastcancers (20), and studies in cancer cell lines indicate that mutant p.53can down-regulate bcl-2 expression (29, 30). A potential molecularbasis for this effect may involve a p-53-dependent negative responseelement on the bcl-2 gene through which p53 can directly or indirectlytranscnptionally down-regulate bcl-2 expression (31). Furthermore,p53-deficient mice exhibit increases in bcl-2 expression in severaltissues suggesting that loss of p53 alone is sufficient to result inelevated bcl-2 expression, at least some tissues in vivo (29).

Coexpression of the bcl-2 and p53 proteins was detected in 29% ofadenomas and 38% of colorectal cancers, respectively. In humannon-Hodgkin's lymphomas, the presence of both bcl-2 and p53 oncoproteins was associated with a worse prognosis than with eitheroncogene alone (32). bcl-2 (11, 12) and mutantp53 (19) may enhancegenetic instability by inhibiting apoptosis, thereby interfering withDNA repairprocesses and allowing furtherdivision of tumor cellscontaining genetic alterations necessary for tumor progression. Inhibition of apoptosis by bcl-2 may also increase the likelihood of cellsacquiring new mutations in p53 or other oncogenes including c-mycand ras which are frequent abnormalities in adenomas and colorectalcarcinomas (16, 33).

Spontaneous apoptosis is a feature of some human cancers and mayregulate cell growth (5, 6). Given that the relationship between theapoptotic rate and bcl-2 expression in epithelial cancers has not beenexamined, we quantitated the proportion of apoptotic nuclei histologically and determined its relationship to bcl-2 staining. We found thatthe spontaneous apoptotic rate in colorectal carcinomas ranged from 0to 7%. Moreover, our results indicate that carcinomas with a highpercentage of cells expressing bcl-2 were significantly more likely tohave low AIs than those with low or absent bcl-2. This result supportsa functional role for bcl-2 in vivo as an inhibitor of apoptosis incolorectal cancers.

Available evidence indicates that hyperplastic polyps do not haveneoplastic potential (25). However, we detected p53 alterations in 3hyperplastic polyps, one of which was 1 cm in size, which indicatesthat some hyperplastic polyps may have a propensity toward neoplasia. This result is consistent with data from patients with head andneck carcinoma, whereby p53 expression has been detected in hyperplastic epithelia and in histologically normal epithelia adjacent tocarcinomas (34). In adenomas and carcinomas, p53 staining wasdetected in 37 and 48% of tumors, respectively. A significant increasein p53 staining was observed in adenomas from patients with synchronous cancers and 78% of adenomas in this study were 1 cm insize. These factors may have contributed to the relatively high frequency of p53 expression observed in the adenomas. Our results areconsistent with a report by Pignatelli et a!. (35) whereby nuclear p53staining was detected in 42% of formalin-fixed, paraffin-embeddedadenomas and in a separate study (36), p53 staining was observed in41% of adenomas from patients with multiple lesions and synchronous cancers. Furthermore, our results show a progressive increase inp53 alterations during colorectal tumorigenesis. Although p.53 mutation is thought to be a late event in colorectal tumorigenesis, our

Apoptotic Index. The mean Al2 for the carcinomas (n = 19) was2.16 ±0.56 (SE) (range, 0.2—7.0)(Table 3). A significant inversecorrelation was observed between the Ms and the percentage of cellspositive for bcl-2 (r = —0.47,P = 0.04). The correlation between A!and bcl-2 intensity was of borderline significance (r = —0.45,P = 0.05); however, the correlationwith the weighted score was notstatistically significant (r = —0.39,P = 0.10). Neither the percentageof cells positive, nor the intensity, nor the weighted score for p53staining were significantly associated with the A!.

Discussion

To determineif the bcl-2 oncoproteinis involved in the colorectalcancer sequence (2, 3), and the timing ofbcl-2 activation, we analyzedthe expression of this protein in normal, hyperplastic, dysplastic, andmalignant colorectal epithelia. Our results indicate that basal epithelialcells of the normal colonic crypts uniformly express the bcl-2 protein.Given that the crypt cell population arises from basally located stemcells (27), the immunolocalization of bcl-2 suggests that it protectsstem cells from apoptosis. Similarly, prostatic basal epithelial cells, incontrast to secretory cells, express bcl-2 and are resistant to apoptosisinduced by androgen withdrawal (28). Therefore, protection fromapoptosis in basal cells of complex epithelia may be a consequence ofbcl-2 expression. Colonocytes in the superficial portion of the mucosadid not express bcl-2 and following terminal differentiation, will besloughed from the mucosal surface. Extruded enterocytes have beenshown to bear ultrastructural features of apoptosis, suggesting thatthese enterocytes undergo programmed cell death (27). It appears,therefore, that loss of bcl-2 expression may be a normal event duringcolonic epithelial cell differentiation.

As in normal mucosa, bcl-2 staining was restricted to basal epithehal cells, i.e., the stem cell compartment, in hyperplastic coloniccrypts. However, a topographic change in bcl-2 expression was foundin adenomas and carcinomas. In this regard, 71 and 67% of dysplasticand carcinoma cells, respectively, abnormally expressed the bcl-2oncoprotein in parabasal and superficial mucosal layers. Furthermore,a significant increase in the percentage of bcl-2 and p53-positive cellswas observed during progression from premalignant polyps to carcinomas. The finding of a high level of diffuse homogenous bcl-2

expression in dysplastic and malignant cells, in contrast to nonneoplastic cells, suggests that abnormal bcl-2 gene activation is an earlyevent in neoplastic development or progression.

Interestingly, bcl-2 and p53 staining were inversely correlated inthe adenomas but not in the carcinomas. p53 nuclear staining wasincreased in adenomas from patients with compared with those without synchronous colorectal cancers and the converse was the case forbcl-2. This result suggests a potential down-regulation of bcl-2 by

2 The abbreviation used is: Al, spontaneous apoptotic index.

240

on July 10, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

bcl-2AND p53 EXPRESSIONIN C0LOREC@ALNEOPLASIA

Wang,Y., Szekely, L., Okan,I., Klein,G., andWiman,R. G. Wild-typep53-triggeredapoptosis is inhibited by bcl-2 in a v-myc-induced T-cell lymphoma cell line.Oncogene, 8: 3427—3431,1993.

19. Lotem, J., and Sachs, L Regulation by bcl-2, c-myc, and p53 of susceptibility toinduction of apoptosis by best shock and cancer chemotherapy compounds in differentiation-competent and -defective myeboidleukemic cells. Cell Growth & Differ., 3:41—47,1993.

20. Silvestri, R., Venerom, S., Daldone, M. 0., Benini, E., Boracchi, P., Mezzetti, M., DiFronzo, G., Rilke, F., and Veronesi, U. The bcl-2 protein; a prognostic indicatorstrongly related to p53 protein in lymph node-negative breast cancer patients. J. Natl.Cancer Inst., 86: 499—504,1994.

21. Lewin, K. J., Riddell, R. H., and Weinstein, W. M. Gastrointestinal pathology and itsclinical implications, pp. 1288—1295.New York: Igaku-Shoin, 1992.

22. Fisher, E. R, Sass, R., Palekar, A., Fisher, B., and Wolmark, N. Dukes' classificationrevisited. Finding from the National Surgical Adjuvant Breast and Bowel Projects(Protocol R-01). Cancer (Phila.), 64: 2354—2360, 1989.

23. Stephens, L C., Ang, K. K., Schultheiss, T. E., Milas, L, and Meyn, R. E. Apoptosisin irradiated murine tumors. Radiat. Res., 127: 308—316,1991.

24. Hsu, S. M., Raine, L., and Fanger, H. A comparative study of the peroxidaseantiperoxidase method and an avidin-biotin complex method for studying polypeptidehormones with radioimmunoassay antibodies. Am. J. Clin. Pathol., 75: 734—739,1981.

25. Pezzella, F., Tso, A. G. D., Cordell, J. L, Pulford, K. A. F., Gatter, K. C., and Mason,D. Y. Expressionof the bcl-2 oncogene proteinis not specific for the 14;18 chromosomal translocation. Am. J. Pathol., 137: 225—232,1990.

26. Hockenbery, D., Zutter, M., Hickey, W., Nahm, M., and Korsmeyer, S. J. Bcl-2proteinis topographicallyrestrictedin tissues characterizedby apoptoticcell death.Proc. Natl. Acad. Sci. USA, 88: 6961—6965,1991.

27. Lipkin, M., Sherlock, P., and Bell, B. M. Generation time of epithelial cells in thehuman colon. Nature (Lond.), 195: 175—177,1962.

28. English, H. F., Kyprianou, N., and lasses, J. T. Relationship between DNA fragmentation and apoptosis in the programmed cell death in the rat prostate followingcastration. Prostate, 15: 233—250,1989.

29. Miyashita, T. Krajewshi, S., Krajewshi, M., Wang, H-G., Un, H-K, Hoffman, B.,Lieberman, D., and Reed, J. C. Tumor suppressor p53 is a regulator of bcl-2 and baxgene expression in vitro and in vivo. Oncogene, 9: 1799—1805,1994.

30. Haldar, S., Negrini, M., Monna, M., Sabbioni, S., and Croce, C. Down-regulation ofbcl-2 by p53 in breast cancer cells. Cancer Res., 54: 2095—2097,1994.

31. Miyashuta, T. Harigai, M., Hanada, M., and Reed, J. Identification ofa p53 dependentnegative response element in the bcl-2 gene. Cancer Res., 54: 3131—3135,1994.

32. Piris, M. A., Pezella, F., Martinez-Montero, 3. C., Orradre, J. L., Villuendas, R.,Sanchez-Beato, M., Cuena, R., Cniz, M. A., Martinez, B., Garrido, M. C., Garter, K.,Aiello, A., Delia, D., Giardini, R., and Rilke, F. p53 and bcl-2 expression inhigh-grade B-cell lymphomas: correlation with survival time. Br. J. Cancer, 69:337—341,1994.

33. Stewart, J., Evans, G., Watson, J., and Sikora, K. Detection of the c-myc oncogeneproduct in cobonicpolyps and carcinomas. Br. J. Cancer, 53: 1—6,1986.

34. Shin, D. M., Kim, J., Ro, J. Y., Hittelman, J., Roth, J. A., Hong, W. K., and Hittelman,W. N. Activationof p53 gene expression in premalignantlesions duringhead andneck tumorigenesis. Cancer Rca., 54: 321—326,1994.

35. Pignatelli, M., Stamp, 0. W. H., Kafiri, F., Lane, D., and Bodmer, W. F. Overexpression of p53 nuclear oncoprotein in coborectal adenomas. Int. J. Cancer, 50:683—688,1992.

36. Darmon, E.,@ K. R., and Wargovich, M. J. Immunohistochemical analysis ofp53 overexpression in human cobonic tumors. Cancer Detect. Prey., 18: 187—195,1994.

37. Endicott, J. A., and Ling, V. The biochemistry of P-glycoprotein-mediated multidrugresistance. Annu. Rev. Biochem., 38: 137—171,1989.

38. Miyashita, T., and Reed, J. bcl-2 gene transfer increases relative resistance of 549.1and WEHI7.2 lymphoid cells to cell death and DNA fragmentation induced byglucocorticoids and multiple chemotherapeutic drugs. Cancer Res., 52: 5407—5411,1992.

39. Walton, M. I., Whysong, D., O'Connor, P. M., Hockenberry, D., Korsmeyer, S. J.,and Kohn, R. W. Constitutive expression of human bcl-2 modulates nitrogen mustardand camptothecin-induced apoptosis Cancer Res., 53: 1853-1861, 1993.

40. Hsu, B., Mann, M. C., Brisbay, S., McConnel, K., and McDonnell, T. J. Expressionof bcl-2 gene confers multidrug resistance to chemotherapy-induced cell death.Cancer Bull., 46: 125—129,1994.

41. Fisher, T. C., Milner, A. E., Gregory, C. D., Jackman, A. Z., Wynne-Aheme, 0.,Hartley, J. A., Dive, C. D., and Hickman, J. A. bcl-2 modulation of apoptosis inducedby anticancer drugs: resistance to thymidylate stress is independent of classicalresistance pathways. Cancer Res., 53: 3321—3327,1993.

42. Lowe, S. W., Ruley, H. E., Jacks, T., and Housman, D. E. p53-dependent apoptosismodulates the cytotoxicity of anticancer agents. Cell, 74: 957—967,1993.

241

results and those of others (35, 36) suggest thatp53 alterations may be 18.more common in premalignant polyps than previously recognized.

Surgery remains the only potentially curative therapy for colorectalcancer; however, nearly 50% of patients will eventually developmetastatic disease. The intrinsic resistance of these tumors rendersthem poorly responsive to current anticancer therapies (37). Recently,studies have shown that cells expressing either bcl-2 or mutantp53 areresistant to inducers of apoptosis, including radiation and severalDNA-damaging anticancer drugs, including fluorouracil (38—42).The mechanism(s) by which bcl-2 and p53 confer drug resistanceappears to be distinct from classical resistance pathways (41). Moreover, the frequent expression of bcl-2 and p53 proteins in colorectalcancers, as shown in this study, suggests a genetic basis for clinicaldrug resistance which may account for the ineffectiveness of anticancer treatments. Analysis of tumors for expression of these oncoproteins may be of value in predict therapeutic response and prognosis.Furthermore, bcl-2 and p53 represent potential targets of new pharmacological agents.

References

1. Boring, C. C., Squires, T. S., and Tong, T. Cancer Statistics, 1994. CA Cancer J.Clin., 44: 7—26,1994.

2. Fearon, E., and Vogelstein, B. A genetic model for colorectal tumorigenesis. Cell, 61:759—767,1990.

3. Vogebstein,B., Fearon, E. R., Hamilton, S. R., Kern, S. E., Preisinger, A. C., Leppert,M., Nakamura,Y., White, R., Smits, A. M. M., and Bce, J. L Genetic alterationsduring colorectal tumor development. N. EngI. J. Med., 319: 525—532,1988.

4. Offerhaus, 0. J. A., Dc Feyter, E. P., Cornelisse, C. J., Tersmette, K. W., Floyd, J.,Kern, S. E., Vogelstein, B., and Hamilton, S. R. The relationship of DNA aneupboidyto molecular genetic alterations in coborectal carcinoma. Gastroenterobogy, 102:1612—1619,1992.

5. Wylie, A. H. The biology ofcell death in tumors. Anticancer Res., 5: 131—136,1985.6. Sarraf,C. E., andBowen, I. D. Proportionof mitoticandapoptoticcells in a rangeof

untreated experimental tumors. Cell Tissue Kinet., 21: 45—49,1988.7. Krajewski, S., Tanaka, S., Takayama, S., Schibler, M. J., Fenton, W., and Reed, J. C.

Investigation of the subcellular distribution of the bcl-2 oncoprotein: residence in thenuclear envelope, endoplasmic reticulum, and outer mitochondrial membranes. Cancer Res., 53: 4701—4714,1993.

8. delong, D., Prins, F. A., Mason, D. Y., Reed, J. C., wan Ommen, G. B., and Kluin,P. M. Subcellularlocalizationof the bcl-2 proteinin malignantandnormallymphoidcells. Cancer Rca., 54: 256-260, 1994.

9. Tsujimoto, Y., and Croce, C. M. Analysis of the structure, transcripts and proteinproductsofbcl-2, thegene involvedin humanfollicularlymphoma.Proc.Nail. Acad.Sd. USA, 83: 5214—5218, 1986.

10. Vaux, D. L., Cory, S., and Adams, J. M. bcl-2 gene promotes hematopoietic cellsurvival and cooperates with c-myc to immortalize pre-B cells. Nature (Lond.), 355:440—442,1988.

11. Hockenbery,D., Nunez, G., Milhiman,C., Schreiber,R. D., and Korsmeyer,S. J.bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death.Nature (Lond.), 349: 254—256,1991.

12. Korsmeyer,S. bcl-2 initiatesa new categoryof oncogenes: regulatorsof cell death.Blood, 80: 879—886,1992.

13. Nunez, G., London, L, Hockenbery, D., Alexander, M., McKearn, J. P., andKorsmeyer,S. J. Deregulatedbcl-2 gene expressionselectively prolongssurvivalofgrowth factor-deprived hemopoietic cell lines. J. Immunol., 144: 3602—3610,1990.

14. McDonnell, T. J., and Korsmeyer,S. J. Progressionfrom lymphoidhyperplasiatohigh-grade malignant lymphoma in mice transgenic for the (14;18). Nature (Lond.),9: 254—256,1991.

15. Levine, A.. J., Perry,M. E., Chang,A., Silver, A., Dittmer,D., Wu, M., andWelsh,D. The 1993 WalterHubertLecture:the role of the p53 tumor-suppressorgene intumorigenesis. Br. J. Cancer, 69: 409—416,1994.

16. Bell, S. M., Scott, N., Cross,D., Sagar,P., Lewis, F. A., Blair, 0. E., Taylor,G. R.,Dixon, M. F., and Quirke, P. Prognostic value of p53 overexpression and c-ki-rasgene mutationsin coborectalcancer.Gastroenterobogy,104: 57—64,1993.

17. Hamelin, R., Laurent-Puig, P., Olschwang, S., Jego, N., Asselain, B., Remvikos, Y.,Girodet,J., Salmon,R. J., andThomas,G. Associationofp53 mutationswithshortsurvival in coborectalcancer. Gastroenterology, 106: 42—48,1994.

on July 10, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1995;55:237-241. Cancer Res   Frank A. Sinicrope, San Bao Ruan, Karen R. Cleary, et al.   Tumorigenesis

and p53 Oncoprotein Expression during Colorectalbcl-2

  Updated version

  http://cancerres.aacrjournals.org/content/55/2/237

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/55/2/237To request permission to re-use all or part of this article, use this link

on July 10, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from