preparation of anti-ras mr 21,000 protein monoclonal...

[CANCER RESEARCH 48, 5503-5509, October 1, 1988)

Preparation of Anti-ras Mr 21,000 Protein Monoclonal Antibodies and

Immunohistochemical Analyses on Expression of ras Genes in HumanStomach and Thyroid Cancers1

Kuniko Yoshida, Kiyohiro Hamatani, Hiroshi Koide, Hiroaki Ikeda, Nori Nakamura, Mitoshi Akiyama,Hideo Tsuchiyama, Eiichi Nakayama, and Hiroshi ShikirSecond Department of Pathology [K. Y., H. T.] and Department of Oncology [H. K., H. 1., E. N., H. S.], Nagasaki University School of Medicine, 12-4 Sakamoto-cho,Nagasaki 852, Japan; Radiation Effect Research Foundation (Nagasaki) [K. H.], I-8-6 Nakagawa, Nagasaki 850, Japan; and Radiation Effect Research Foundation(Hiroshima) [N. N., M. A.], 5-2 Hijiyamakoen, Minami-ku, Hiroshima 732, Japan

ABSTRACT

Sixteen clones (RASK-1 to -16) of murine monoclonal antibodies wereraised against ras M, 21,000 protein (p21). The p21 produced by Esche-richia colt with inserted \-Ki-ru.v genes was used as immunogen. RASK-1 was found to be specific for Ki-ru.vp21, whereas RASK-2 to -16 reactedwith the p21s of Ki-, N-, and Ila-ras genes in both enzyme-linkedimmunosorbent and immunoblotting assays. Binding inhibition assayswith biotinylated monoclonal antibodies by enzyme-linked immunosorbent assay showed that the monoclonal antibodies of the 16 clonesincluded those binding to several mutually distinct sites on p21.

The expressions of ras p21 in human stomach and thyroid tissues wereexamined with RASK-3, which reacted with all the Ki-, N-, and I Ia-r</.vp21s immunohistochemically by the avidin-biotin peroxidase complexmethod. Formalin-fixed, paraffin-embedded tissues of 101 cases of stomach cancer, 53 cases of noncancerous stomach, 74 cases of cancer of thethyroid, and 59 cases of noncancerous thyroid were analyzed. In both thestomach and thyroid, cancer cells expressed p21 predominantly. Cells ofcases with various noncancerous disorders as well as certain types ofnormal cells were also p21 positive. These findings suggest that precaution is required in use of p21 as a cancer marker.

Expression of p21 was noted in moderately to well-differentiatedstomach cancer, intestinal metaplasia, and atypical hyperplasia. Thisfinding suggests that the appearance of p21 in stomach cancer may beinitiated before cytological transformation.

INTRODUCTION

Activation of ras genes results in either increased productionof the normal gene product or production of an aberrant geneproduct (1-8). These abnormal intracellular expressions of geneproducts may lead to transformation of cells of various speciesincluding humans (3-8). Two major reasons prompted us tostudy details of the profiles of expression of ras gene productsat the cellular level. The first was that precise information onthe cell types expressing p213 is essential in investigations of

the functional role of ras gene products in physiological conditions and in the sate of oncogenic processing. The second wasto answer the question of whether expression of ras p21 geneproducts is greater in certain types of human cancer, and if so,whether expression of ras genes is associated with cellulartransformation, and whether ras gene products are useful ascancer markers. Monoclonal antibodies that react with the ras

Received 9/2/87; revised 6/7/88; accepted 6/29/88.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' This work was partly supported by a Grant-in-Aid for Scientific Research

from the Ministry of Education, Science, and Culture, and by a grant from theMinistry of Health and Welfare of Japan.

2To whom requests for reprints should be addressed, at Department ofOncology, Nagasaki University School of Medicine, 12-4 Sakamoto-cho, Nagasaki 852, Japan.

3The abbreviations used are: p21. a protein with a molecular weight of 21.000;ELISA, enzyme-linked immunosorbent assay; IF. immunofluorescence; BSA,bovine serum albumin; ABC, avidin-biotin peroxidase complex: PBS, phosphate-buffered saline.

gene product p21 are particularly useful for this purpose, sincethey allow detection of p21 in both tissues and single cells (9-11).

This paper reports the production of anti-p21 monoclonalantibodies and analyzes p21 expression in human stomach andthyroid tissues.

MATERIALS AND METHODS

Antigen. KÂ¡rasp21, produced in Escherichia coli bearing the plasmidpHN 121 containing v-Ki-ros genes, was used as immunogen and kindlyprovided by Dr. Hirobumi Nakano, Tokyo Research Laboratories,Kyowa Hakko Kogyo Co., Tokyo, Japan (12). The specificities ofmonoclonal antibodies were examined using the following antigens: (a)Ki-ras p21 described above; (b) N-ras p21 and Ha-ras p21, producedby E. coli with inserted recombinant N-ros genes or Ha-ras genes (kindlyprovided by Dr. Toshikazu Matsui, Fujita-Gakuen Health UniversitySchool of Medicine, Aichi, Japan) (13); (c) extracts of rat kidney cellstransformed by Harvey or Kirsten murine sarcoma viruses (kindlyprovided by Dr. Takeo Tanaka, Hiroshima University School of Medicine, Hiroshima, Japan) (14); (d) and NIH/3T3 cells transformed withthe Ha-ras gene from the human urinary bladder cancer cell line EJ.

Human Tissues. Paraffin blocks of 10% formalin-fixed tissues fromthe stomach and thyroid, removed at surgery, were obtained from theDepartment of Pathology, the Japanese Red Cross, Nagasaki AtomicBomb Hospital. The specimens were from 101 cases of stomach cancer,8 cases of atypical hyperplasia of the stomach, 15 cases of hyperplasticpolyp of the stomach, 13 cases of gastric ulcer, 3 cases of gastritis, 14cases of normal stomach, 5 cases of chronic thyroiditis, 2 cases oflymphocytic thyroiditis, 12 cases of follicular adenoma, 2 cases ofoxyphilic adenoma, 3 cases of Graves' disease, 2 cases of nodular goiter,

1 case of cyst, 1 case of adenomatous goiter, and 1 case of normalthyroid.

Similar blocks of thyroid tissues from autopsied cases were obtainedfrom the Radiation Effect Research Foundations in Nagasaki andHiroshima. These tissues were from 74 cases of thyroid cancer (papillary carcinoma, 59 cases; follicular carcinoma, 13 cases; medullarycarcinoma, 1 case; HÃ¼rthlecell carcinoma. 1 case) and 30 cases withnormal thyroid. Forty-eight cases of thyroid cancer specimens wereobtained from atomic bomb survivors. Doses of radiation by atomicbombs were based on DS86 (15). All of 30 cases with normal thyroidwere nonirradiated individuals.

Preparation of Anti-ru.v p21 Monoclonal Antibodies. BALB/c xC57BL/6 F! mice were immunized 4 times with v-Ki-ras p21 at 2-wkintervals; the first time s.c. with 50 Â¿igof v-Ki-ras p21 and completeFreund adjuvant; the second time s.c. with 100 /ig and incompleteFreund adjuvant; and the third and fourth times i.p. with 100 fig and200 Mg*respectively. Three days after the last immunization, the spleensof the immunized mice were removed, and the spleen cells were fusedwith NS-1 cells as described previously (16). Culture fluids of thehybridomas were assayed for the presence of anti-ras p21 antibody byELISA and IF assays. Limiting dilution was carried out 2 or 3 timesto obtain monoclonality.

Purification and Biotinylation of Monoclonal Antibodies. Protein Abinding IgG antibody was prepared from ascites with an Affigel ProteinA MAPS-II kit (Bio-Rad Lab., Richmond, CA). For preparation of

5503

Research. on November 9, 2018. © 1988 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

ANTI-ros p21 MONOCLONAL ANTIBODIES

biotinylated antibodies, antibodies were dialyzed against 0.1 M Na-HCOj (1 mg/ml) and then mixed with jV-hydroxysuccinimide biotin (1mg/ml of dimethyl sulfoxide) at room temperature for 4 h.

ELISA. Target protein (50 tig/ml) was fixed to 96-well immunoplates(Nunc, Roskilde, Denmark) and used as the solid phase. It was blockedby incubation for l h at room temperature with 5% BSA, and thenmonoclonal antibodies were introduced and the plates were incubatedfor 45 min at room temperature. After blocking again with 5% BSA,goat anti-mouse IgG antibodies conjugated with peroxidase (Medicaland Biological Laboratory Co., Nagoya, Japan) were introduced as thesecond antibody. o-Phenylenediamine was used as substrate, and ab-sorbance at 492 nm was measured with an automatic ELISA reader(Model SLT210; SLT-Labinstruments, Salzburg, Austria).

Immunohistochemical Staining of Tissues. The ABC method was usedwith Vectastain ABC kits (Vector Laboratories, Inc., Burlingame, CA).For this, 4-Minsections were deparaffinized and immersed in methanolcontaining 0.3% H2O2 for 30 min to eliminate endogenous peroxidaseactivity. The sections were then washed with PBS, normal horse serumwas added, and the sections were incubated for 20 min at roomtemperature. The slides were then incubated at room temperature for40 min with anti-p21 monoclonal antibody (10 Mg/ml) (IgG2a). Theywere then washed with PBS and incubated at room temperature for 30min after addition of biotinylated horse anti-mouse IgG antibodies.The slides were then washed with PBS again and incubated withperoxidase-labeled avidin-biotin complexes for 60 min. After washingwith PBS, the slides were treated with 0.05% diaminobenzidine in 0.05M Tris-HCl (pH 7.2) containing 0.01% H2O2 for 5 min, washed, andcounterstained with hematoxylin. Anti-human platelet monoclonal antibody HPL-2 (IgGl) was used as control antibody at a concentrationof 10 Mg/ml.

Immunoblotting. Extracted protein was separated on 12% sodiumdodecyl sulfate-polyacrylamide gel and transferred electrophoreticallyto nitrocellulose membranes. The membranes were then incubated withmonoclonal antibodies for l h at room temperature. The ABC methodwas used for detection of antibodies with 4-chloro-l-naphthol as substrate.

RESULTS

Production of \nti-p2I Monoclonal Antibodies. Culture su-pernatants of hybridomas prepared as described in "Materialsand Methods" were screened by ELISA. Hybridomas which

gave supernatants that reacted with Ki-ras p21 but not withlysates of E. coli that did not have an inserted v-Ki-ras genewere picked up, and monoclonal cells were obtained by thelimiting dilution 2 or 3 times. In this way, 16 clones of monoclonal antibodies, RASK-1 to RASK-16, were obtained, andtheir anti-p21 specificities were analyzed by ELISA and im-munoblotting assays with p21 of the Ki-ras, N-ras, and Ha-rasgenes. Fig. 1 shows examples of immunoblotting assays withtwo clones of antibodies, and Table 1 summarizes results obtained by ELISA and immunoblotting assays. The antibody ofRASK-1 was specific for Ki-ras p21, whereas those of the other15 clones reacted with the p21s of all members of the ras family.

Multiple Binding Sites of Monoclonal Antibodies on p21 Molecules. We used antibody binding inhibition assays to determinewhether the p21 antibodies of all our clones bound to the samesite on p21 molecules. Anti-p21 monoclonal antibodies werebiotinylated, and their binding to Ki-ras p21 that had beenpretreated with other nonbiotinylated anti-p21 monoclonal antibodies was examined. Table 2 shows an example of inhibitionassays with four biotinylated monoclonal antibodies, RASK-3to RASK-6. Preincubation of p21 on the solid phase with 6unlabeled clones, RASK-1 to RASK-6, resulted in inhibitionsof the binding of subsequently added biotinylated antibodies insome combinations of antibodies, but not in others. There weremultiple clones that bound to mutually distinct sites on p21.

i l Â»II

21.5K

14.4K

.â€ž.,

RASK-1 RASK-4Fig. 1. Immunoblotting assays with two anti-p21 monoclonal antibodies,

RASK-1 and RASK-4. RASK-1 reacted with Ki-ras p21 but not with p21s of N-or Ha-ras genes, whereas RASK-4 reacted with the p21 s of N-, Ki-, and Ha-rasgenes. jY-ros, N-ros p21 produced by E. coli; Ki-ras, Ki-ras p21 produced by E.coli; E. coli, extract of E. coli with no inserted ras gene; Ha-NRK, extract of ratkidney cells transformed by Harvey murine sarcoma viruses; Ki-NRK, extract ofrat kidney cells transformed by Kirsten murine sarcoma virus. The ABC methodwas used.

Table 1 Reactivity ofanti-p21 monoclonal antibodies with the ras p21 family:summary of analyses by ELISA and immunoblotting assays

MonoclonalantibodyKi-rasP21N-rasHa-rasRASK-1

++RASK-20 ++ + +RASK-3 \RASK-16 I

* The reactivities of RASK-2 with N-ras p21 and Ha-ras p21 were weaker thanthat with Ki-ras p21.

Table 2 Binding inhibition assaysRASK-3 to -6 were biotinylated, and their bindings to Ki-ras p21 that had

been pretreated with other nonbiotinylated anti-p21 monoclonal antibodies (inhibitors) were examined.

Biotinylatedmonoclonal antibody

Inhibitor"RASK-1

RASK-2RASK-3RASK-4RASK-5RASK-6RASK-30.60*

0.3800.180.510.53RASK-40.66

0.530.390.190.650.60RASK-50.78

0.450.4400.080.59RASK-60.34

0.040.030.030.070.01

None 0.64 0.70 0.78 0.32" Nonbiotinylated monoclonal antibodies were preincubated with Ki-ras p21.* Absorbance at 492 nm.

Immunohistochemical Studies on Stomach and Thyroid Tissues with Anti-p21 Monoclonal Antibodies. The expressions ofras p21 in the stomach and thyroid were analyzed inuminoliis-tochemically with one of the anti-ros p21 monoclonal antibodies, that of RASK-3, which reacts with p21 of all the Ki-, N-,and Ha-ras genes.

Expression of p21 in the Stomach. Use of a wide range ofantibody concentration, 10 ng/ml to 100 Mg/ml, was initiallyattempted for immunohistochemistry. Based on consistencyand intensity of reactivities, a concentration of 10 ng/m\ waschosen. Use of a lower concentration of antibody sometimesgave inconsistent results, and use of a higher concentrationsometimes gave nonspecific stainings.

5504



ANTI-roi p21 MONOCLONAL ANTIBODIES

The expression of p21 was studied in 101 cases of stomachcancer and 53 cases of noncancerous stomach. In many cancertissues, cancer cells gave a strongly positive reaction, whereasmorphologically normal epithelial cells showed scarcely anyreaction. Parietal cells and intestinal metaplasia, however, oftengave a positive reaction, though chief cells and mucous cellsgave an essentially negative reaction. Smooth muscle and ganglion cells consistently gave a positive reaction. In individualcells showing a positive staining reaction, the cytoplasm wasalways stained. Examples of immunohistochemical staining areshown in Fig. 2. In cases of stomach cancer, the expressions ofp21 in cancer cells and morphologically normal epithelial cellswere evaluated on the same slide. The 101 cases were classifiedinto 5 groups on the basis of their ratios of p21-positive cellsto the total cells counted, and results are summarized in Table3. More than 80% of the cancer cells in 46 cases and 50 to 80%of those in 24 cases expressed p21. But in all cases, less than50% of the morphologically normal cells expressed p21. Staining of these p21-positive epithelial cells was in general weakerthan that of cancer cells. In 87 cases (86%), normal epithelialcells were virtually p21 negative. Correlation of p21 expressionwith histological types of stomach cancer was analyzed (Table4). In general, expression of p21 seems to be more dominant inmore differentiated types of cancer. The detailed distributionof p21-positive cells was studied in two cases of moderately

differentiated stomach cancer. Totals of 112 and 84 sections ofeach case were prepared, and p21 expression was evaluated. Asshown in Fig. 3, all cancerous areas strongly expressed p21. In

regions of normal epithelium, however, there were sparse clusters of weakly positive normal epithelial cells, and these clustersshowed some tendency to surround the cancerous areas.

Expression of p21 was also examined in 53 cases of noncancerous stomach, and results are summarized in Table 5. In 7 of8 cases of atypical hyperplasia, more than 50% of the epithelialcells were positive, and in 2 of 15 cases of hyperplastic polyp,5 to 50% of the cells were p21 positive. In 3 cases of gastriculcer, epithelial cells expressed p21. These 3 cases of gastriculcer showed dominant regenerative patterns of the epitheliumwith healing of the ulcer, and the same regenerating epithelialcells were p21 positive.

Expression of p21 in Thyroid. The expression of p21 wasanalyzed in 74 cases of thyroid cancer and 59 cases of noncancerous thyroid (Table 6). Expression of p21 was evaluated inthe same way as in stomach cancer. More than 50% of thecancer cells express p21 in all cases of thyroid cancer, and morethan 80% of cancer cells express p21 in 69 cases (93%). Morphologically normal thyroid follicular cells were, however, alsooften p21 positive, though their intensity of staining was ingeneral weaker than that of cancerous areas. Some cells werealso p21 positive in noncancerous thyroid, though their frequency was less than in cancer tissue (Table 7). Correlation ofp21 expression with histological types of thyroid cancer wasalso analyzed (Table 8). No obvious correlation was observed.Since 48 cases with thyroid cancer were atomic bomb survivors,correlation of radiation doses with expression of p21 in thyroidtissues were analyzed. As shown in Table 9, no clear association

^v*i*.

^ <^*fÂ£ p â€¢â€¢â€¢â€¢":â€¢''â€¢â€¢'â€¢'>''V'-" ' â€¢;-â€¢'â€¢'"' -:

â€¢.:â€¢'.''â€¢/J"'- ;;." .'--'-."".-â€¢ .---.'

f â€¢ â€¢J â€¢*,**** â€¢" â€¢â€¢ ~*â€¢^- .-^vÃ³e v--"J ..-â€¢.r --â€¢-'.â€¢ .â€¢â€¢â€¢.,' .... -- .' . - ; â€¢.;"-' J/ri

--, ' - _ ;>

B* . DFig. 2. Examples of immunohistochemical staining of stomach and thyroid cancers by the ABC method. Each tissue was examined with RASK-3 and control

monoclonal antibody. HPL-2. A, stomach cancer (RASK-3); B. stomach cancer (HPL-2); C thyroid cancer (RASK-3); D, thyroid cancer (HPL-2). x 342.

5505




Table 3 Expressions ofp21 in formalin-fixed tissues of 101 cases of stomachcancer determined by the ABC method

>of positive cells"

Tissue >80 50-80 20-50 5-20 <5Cancer*NoncancerJ46

(46)"

024(24)

014(14) 7(7)KD 7(7)16(16)87 (86)Â°Proportion of p21-positive cells.* Morphologically cancerous parts.c Numbers in parentheses, percentage of total cases.'' Morphologically normal epithelium.

Table 4 Expression of ras p21 in stomach cancer: correlation with histolÃ³gica!types of cancer

Histologicaltype-No,

ofcasesPositive*Partially positive"Negative''

Papillary 10 9(90)' 1(10) 0

TubularWell differentiated 11 10(91) 1(9) 0Moderately differentiated 31 24(77) 2(6) 5(16)

Poorly differentiated 37 21(57) 7(19) 9(24)Murinoli-, 6 6(100) O OSignet-ring cell 60 4 (67) 2 (33)

Total 101 70 15 16Â°Based on criteria of the general rules for the gastric cancer study by the

Japanese Research Society for Gastric Cancer (34).4 More than 50% of the cells were p21 positive.c Five to 50% of the cells were p21 positive.d Less than 5% of the cells were p21 positive.' Numbers in parentheses, percentage.

was observed in either cancerous or noncancerous cells intissues.

Specificity of Immunohistochemical Staining. Antibody binding inhibition assays with biotinylated RASK-3 were used toconfirm that tissue staining with RASK-3 was specific. Prein-cubation of sections with nonbiotinylated RASK-3, but not withcontrol monoclonal antibody, HPL-2, reduced the reactivitiesof biotinylated RASK-3 with stomach and thyroid cancer cells(data not shown).

DISCUSSION

The 16 clones of anti-p21 monoclonal antibodies producedin this work show differences in reactivity: one clone, RASK-1,reacted only with Ki-ras p21, not with other N-ras or Ha-rosp21, whereas the other 15 clones reacted with p21s of all theKi-, N-, and Ha-ras gene family. The antibodies produced bythese clones were, however, composed of several groups whichbound to distinct sites of molecules, and thus the binding ofantibody of one clone does not interfere with the binding ofothers. These characteristics of our antibodies allowed us toexamine the expressions of total ras p21 and Ki-ras only, andalso to apply multiple antibodies for various serological analyses such as sandwich assays for detection of p21. Since allanti-p21 antibodies could detect p21 in 10% formalin-fixedtissues, we could use them to examine expression of p21 inlarge numbers and various types of tissues fixed with 10%formalin and embedded in paraffin wax.

In most cases of stomach cancer and in all cases of thyroidcancer, cancer cells were strongly p21 positive. In many cases,a variety of normal cells, such as parietal, smooth muscle, andganglion cells, and also thyroid follicular cells were also foundto express raÃp21. However, normal epithelial cells, chief cells,and mucous cells of the stomach were mostly p21 negative.Thor et al. reported the similar findings and suggested thatmany of the cell types with enhanced ras p21 expression havebeen associated with ion exchange functions and/or producehormone products which effect ion exchange mechanisms in

BFig. 3. Expression of p21 in stomach cancer. In two cases of stomach cancer,

112 (A) and 84 sections (li) were made for each stomach, and the expression ofp21 in all epithelial cells in all sections was examined. RASK-3 was used for theABC method. Black areas, areas with cancer and strong p21 expression; asterisks,areas with p21-positive normal epithelial cells; white areas in rectangles,with p21-negative normal epithelial cells.

Table 5 Expression ofp21 in formalin-fixed tissues of 53 cases of noncancerousstomach determined by the ABC method

TissueAtypical

hyperplasiaHyperplastic

polypUlcerGastritis

and othersNo.

ofcases8IS1317Positive"7

(88)*02(15)0Partially

positive"02(13)1(8)3(18)Negative"1(13)13(87)10(77)14

(82)1As described for Table 4.' Numbers in parentheses, percentage.

Table 6 Expression ofp21 in formalin-fixed tissues of 74 cases of thyroid cancerdetermined by the ABC method

% of positive cells

Tissue >80 50-80 20-50 5-20 <5

CancerNoncancer*69

(93)Â°

2(3)5(7) 6(8)020 (28)0 24 (33)020 (28)Â°Numbers in parentheses, percentage.* The total number of cases is 72, because in 2 cases no noncancerous parts

were observed.

5506



ANTl-ras p21 MONOCLONAL ANTIBODIES

Table 7 Expression of pii in formalin-fixed tissues of 59 cases of noncancerousthyroid determined by the ABC method

TissueAdenomaThyroiditis

and othergoitersNo.

ofcases1414Positive"8

(57)'4(29)Partially

positive"5(36)8(57)Negative"1(7)2(14)

No primary disease ofthe thyroid

31 5(16) 19(61) 7(23)

' As described for Table 4.* Numbers in parentheses, percentage.

Table 8 Expression of ras p21 in thyroid cancer: no correlation with histologicaltypes of cancer

HistologicaltypePapular)'

FollicularMedullaryHÃ¼rthlecellNo.

ofcases59

1311%

of positivecells">8054

(92)Â°

13(100)1 (100)1 (100)50-805(8)

000<500

000

Total 74 69" Proportion of p21-positive cells.* Numbers in parentheses, percentage.

Table 9 No correlation of radiation doses to expression ofp2I in cancerous andnoncancerous cells in thyroid cancer

CancerouscellsNoncancerouscells'Dose

(rads)0<100>100UnknownTotal0<100>100UnknownTotal%

of positivecells>8023

(88)Â°13(100)20(100)13(87)6901(8)1(5)0250-803(12)002(13)54(16)01(5)1(7)620-50000006(24)3(23)7(37)4(27)205-20000007(28)6(46)5(26)6(40)24<5000008(32)3(23)5(26)4(27)20Total261320IS742513191572

odologies, way of evaluation, and also antibodies. In situ hybridization to clarify this issue is ongoing in our laboratory.

Moderately to well-differentiated types of stomach cancerhave been considered to originate from epithelial cells of thestomach. There is also circumstantial evidence that these typesof cancer may appear via intestinal metaplasia (24-26). Interestingly, our analyses showed that p21 expression is marked inmoderately to well-differentiated cancer, intestinal metaplasia,and atypical hyperplasia, but not in normal epithelial cells andhyperplastic polyps, a part of which was also the case in thereport by Ohuchi et al. (23). These results may indicate thatexpression of p21 in epithelial cells of the stomach has increasedas a consequence of cellular changes to premalignant status,such as intestinal metaplasia and atypical hyperplasia. Alternatively, all these p21-positive gastric epithelial cells sharemorphological and other cellular characteristics with p21 -positive epithelial cells of the intestine, as shown in Fig. 4. Fromthe pattern of p21 distribution observed in the present study,we may illustrate pathways of cellular changes as shown in Fig.5. If this is the case, moderately to well-differentiated cancer,atypical hyperplasia, and intestinal metaplasia start to expressp21 as a consequence of their metaplastic changes to intestinalepithelial cells. Expression of p21 by these differentiated cancers is therefore prÃ©existentto cytological transformation. Thisinterpretation may support the view of Chesa et al. that p21 isrelated to cellular differentiation rather than to the maintenanceof transformed phenotypes (18).

Reservations are necessary in the interpretations of other

Â°Numbers in parentheses, percentage.* As described for Table 6.

the effector organ to maintain homeostasis within the normalphysiological range (17). Chesa et al. also reported from im-munohistochemical analyses with anti-ras p21 monoclonal antibodies that, in many cell lineages, well-differentiated cellsusually express p21 more than poorly differentiated cells (18).p21 is known to have GTP binding capacity and thus is relatedto cellular regulation of adenyl cyclase like other members ofthe Protein G family (19-21). In highly differentiated cells withspecified functions, p21 may be dominantly expressed becauseit participated in critical cellular functions.

A crucial question is whether dominance of p21 expressionin cancer cells is a consequence of cellular transformation. Instomach cancer, expression of p21 was apparently higher incancer cells than in normal epithelial cells. This tendency wasmarked in moderately to well-differentiated cancers. Similarresults were reported by Noguchi et al. (22), who studied 96cases of stomach cancer with the anti-p21 monoclonal antibody,RAP-5, produced by Hand et al. (9). The results, however, werenot necessarily confirmed by Ohuchi et al. (23), who analyzed44 cases of stomach cancer with the same RAP-5 antibodies.They observed essentially no correlation of histological typesof stomach cancer with p21 expression by immunohistochemi-cal analyses, and the observation was also confirmed by direct-binding liquid competitive radioimmunoassay and //; situ hybridization. Reasons for this discrepancy are unclear. Possibilities may include differences of races for sample donors, meth-

5507

Fig. 4. Expression of p21 in intestinal metaplasia and intestinal epithelialcells. KASK i was used for the ABC method. A, intestinal metaplasia; B, epithelialcells of normal small intestine, x 342.




Stomach epithelial cells

Fig. 5. Pathways of cellular changes andexpression of p21.

HyperplastiC polyp

Intestinal metaplasia

nAt ypicai

hyperplasia |

\Moderatelyto welldifferentiatedstomach cancer

(p2T) !

Intestinal epitheliacells

I ._ ._ . _ J

types of stomach cancer and thyroid cancer in the same way.Our monoclonal antibodies reacted with products of both pro-

tooncogenes and activated genes. The pattern of p21 expressionby activated ras genes might be unique and different from thatby protooncogenes, which cannot be distinguished with currently available anti-p21 monoclonal antibodies including ours.

Detailed analysis showed no dominant clustering of p21-positive normal epithelial cells. Cells showing weak expressionof p21 seemed to surround cancerous areas, though the reasonfor this is unknown.

In our study, thyroid cancer cells in all cases highly expressedp21. Our samples mainly consisted of papillary and follicularcarcinoma, and no difference of p21 expression between thetwo of them was observed. Whether undifferentiated types ofthyroid cancer express p21 has to be determined. Johnson etal. reported a lower frequency of immunohistochemically positive p21 expression in thyroid cancers than ours (27). The dataof their analyses on benign and normal thyroid tissues, however,seem to be essentially compatible with ours. Whether the discrepancy of results between ours and theirs is due to differencesof antibodies, methodologies, way of evaluations, or specimensis not clear. A possibility of irradiation to enhance expressionof p21 in thyroid tissues seems to be unlikely, since even 26cases of totally nonirradiated cases of thyroid cancer were alsoall p21 positive (Table 9). Though no correlation of radiationdoses with p21 expression in cancerous or noncancerous cellsin cases of thyroid cancer was observed, we need further studiesto clarify if it is the case in either benign or normal thyroids.

There are reports of p21 expression observed by the immu-nohistochemical staining in cancers of the colon (19, 28-30),mammary gland (9, 17, 31), prostate (32), urinary bladder (33),thyroid (27), and stomach (22, 23). In general, expression ofp21 in cancer cells was found to be greater than that in benignor normal cells. There are, however, reports of no significantdifference between the expressions of p21 in cancer cells andnormal cells in the colon (29,30). These results and ours suggestthat caution is required in the use of p21 as a cancer marker.

Thus we conclude that further investigations are needed ongreater numbers and types of tissues with antibodies that arespecific for the Ki-, N-, and Ha-ras gene products, respectively,and for point mutation sites of the proteins.

ACKNOWLEDGMENTS

We are indebted to Dr. O. Takahara for providing pathological tissuespecimens. We also thank Y. Abe, M. Ohkohchi, and Y. Fujita forexpert technical assistance. The mice used for immunization and forproduction of ascites of monoclonal antibodies were maintained in theLaboratory Animal Center for Biomedicai Research, Nagasaki University School of Medicine.

REFERENCES

1. Der, C. J., Krontiris, T. G., and Cooper, G. M. Transforming genes of humanbladder and lung carcinoma cell lines are homologous to the rus genes ofHarvey and Kirsten sarcoma viruses. Proc. Nati. Acad. Sci. USA, 79: 3637-3640, 1982.

2. Shimizu, K., Goldfarb, M., Suard, Y., Perucho, M., Li, Y., Kamata, T.,Feramisco, J., Stavnezer, E., Fogh, J., and Wigler, M. H. Three humantransforming genes are related to the viral ras oncogenes. Proc. Nati. Acad.Sci. USA, 80: 2112-2116, 1983.

3. Schwab, M., Alitalo, K., Varmus, H. E., and Bishop, J. M. A cellularoncogene (c-KÃ¬-ras)is amplified, overexpressed, and located within kary-otypic abnormalities in mouse adrenocortical tumour cells. Nature (Lond.),303:497-501,1983.

4. Chang, E. H., Furth, M. E., Scolnick, E. M., and Lowy, D. R. Tumorigenictransformation of mammalian cells induced by a normal human gene homologous to the oncogene of Harvey murine sarcoma virus. Nature (Lond.),297: 479-483, 1982.

5. DeBortoli, M. E., Abou-Issa, H., Haley, B. E., and Cho-Chung, Y. S.Amplified expression of p21 ras protein in hormone-dependent mammarycarcinomas of humans and rodents. Biochem. Biophys. Res. Commun., 127:699-706, 1985.

6. Tabin, C. J., Bradley, S. M., Bargmann, C. L, Weinberg, R. A., Papageorge,A. G., Scolnick, E. M., Dhar, R., Lowy, D. R., and Chang, E. H. Mechanismof activation of human oncogene. Nature (Lond.), 300: 143-149, 1982.

7. Reddy, E. P., Reynolds, R. K., Santos, E., and Barbacid, M. A point mutationis responsible for the acquisition of transforming properties of the T24human bladder carcinoma oncogene. Nature (Lond.), 300:149-152, 1982.

8. Taparowsky, E., Suard, Y., Fasano, O., Shimizu, K., Goldfarb, M., andWigler, M. Activation of the T24 bladder carcinoma transforming gene islinked to a single amino acid change. Nature (Lond.), 300: 762-765, 1982.

9. Hand, P. H., Thor, A., Wunderlich, D., Muraro, R., Caruso, A., and Schlom,J. Monoclonal antibodies of predefined specificity detect activated ras geneexpression in human mammary and colon carcinomas. Proc. Nati. Acad. Sci.USA, 81: 5227-5231, 1984.

10. Furth, M. E., Davis, L. J., Fleurdelys, B., and Scolnick, E. M. Monoclonalantibodies to the p21 products of the transforming gene of Harvey murinesarcoma virus and of the cellular ras gene family. J. Virol., 43: 294-304,1982.

11. Kuzumaki, N., Oda, A., Yamagiwa, S., Taniguchi, N., Kobayashi, H., andOikawa, T. Establishment of four mouse hybridoma cell lines producingmonoclonal antibodies reactive with ras oncogene product p21. J. Nati.Cancer Inst., 77:1273-1279, 1986.

12. Tamaoki, T., Mizukami, T., Perucho, M., and Nakano, H. Expression ofintact Ki-ros p21 protein in Escherichia coli. Biochem. Biophys. Res. Commun., 132: 126-133, 1985.

5508




13. Matsui, T.. Hirano, M., Naoe, T., Yamada, K., and Kurosawa, Y. Productionof chimeric protein of human N-ros and v-Ha-ras in Escherichia coli. Gene,52:215-223, 1987.

14. Tanaka, T., Slaman, D. J., Halt Â¡lora.H., and Cline, M. J. Expression of p21ras oncoproteins in human cancers. Cancer Res., 46: 1465-1470, 1986.

15. Woolson, W. A., Egbert, S. D., and Gritzner, M. L. Dosimetry system 1986.In: W. C. Roesch (ed), US-Japan Joint Reassessment of Atomic BombRadiation Dosimetri in Hiroshima and Nagasaki, Final Repon, Vol. 1, pp.405-431. Hiroshima: The Radition Effects Research Foundation, 1987.

16. Kato, K., Naoe, T., Hiraiwa, A., Namikawa, R., Suzuki, S., Yamada, K., andShiku, H. Production and analysis of HH10 monoclonal antibodies reactiveto immature hematopoietic cells and their use for monitoring acute leukemiacells. Jpn. J. Cancer Res. (Gann), 76: 524-531, 1985.

17. Thor, A., Ohuchi, N., Hand, P. H., Callahan, R., Weeks, M. O., Theillet, C,Lidereau, R., Escot, C., Page, D. L., Vilasi, V., and Schlom, J. Biology ofdisease, ras gene alterations, and enhanced levels of mi p21 expression in aspectrum of benign and malignant human mammary tissues. Lab. Invest.,55:603-615, 1986.

18. Chesa, P. G., Rettig, W. J., Melamed, M. R., Old, L. J., and Niman, H. L.Expression of p21"" in normal and malignant human tissues: lack of associ

ation with proliferation and malignancy. Proc. Nati. Acad. Sci. USA, 84:3234-3238, 1987.

19. Bishop, J. M. Viral oncogenes. Cell, 42: 23-38, 1985.20. Shih, T. Y., Papageorge, A. G., Stokes, P. E., Weeks, M. O., and Scolnick,

E. M. Guanine nucleotide-binding and autophosphorylating activities associated with the p2l"* protein of Harvey murine sarcoma virus. Nature (Lond.),2*7:686-691, 1980.

21. Sweet, R. W., Yokoyama, S., Ramata, T., Feramisco, J. R., Rosenberg, M.,and Gross, M. The product of ras is a GTPase and the T24 oncogenic mutantis deficient in this activity. Nature (Lond.), 311: 273-275, 1984.

22. Noguchi, M., Hirohashi, S., Shimosato, Y., Thor, A., Schlom, J., Tsunokawa,Y., Terada, M., and Sugimura, T. Histologie demonstration of antigensreactive with anti-p21 ras monoclonal antibody (PAP-5) in human stomachcancers. J. Nati. Cancer Inst., 77: 379-385, 1986.

23. Ohuchi, N., Hand, P. H., Merlo, G., Fujita, J., Mariani-Costantini, R., Thor,A., Nose, M., Callahan, R., and Schlom, J. Enhanced expression of c-Ha-

ras p21 in human stomach adenocarcinomas defined by immunoassays usingmonoclonal antibodies and in situ hybridization. Cancer Res., 47: 1413-1420, 1987.

24. Pagnini, C. A., and Rugge, M. Gastric cancer: problems in histogenesis.Histopathology, 7:699-706, 1983.

25. Sugimura, T.. Matsukura, N., and Sato. S. Intestinal metaplasia of thestomach as a precancerous stage. IARC Sci. Pubi., 39: 515-530, 1982.

26. lida. F., and Kusama, J. Gastric carcinoma and intestinal metaplasia. Significance of types of intestinal metaplasia upon development of gastric carcinoma. Cancer (Phila.), 50: 2854-2858, 1982.

27. Johnson, T. L., Lloyd, R. V., and Thor, A. Expression of ras oncogen p21antigen in normal and proliferative thyroid tissues. Am. J. Pathol., /27: 60-65, 1987.

28. Thor, A., Hand, P. H., Wunderlich, D., Caruso, A., Muraro, R., and Schlom,J. Monoclonal antibodies define differential ras gene expression in malignantand benign colonie disease. Nature (Lond.), ill: 562-565. 1984.

29. Ken. I. B., Lee, F. D.,Quintanilla,M.,and Balmain, A. Immunocytochemicaldemonstration of p21 ras family oncogene product in normal mucosa and inpremalignant and malignant tumours of the colorectum. Br. J. Cancer, 52:695-700,1985.

30. Williams, A. R. W., Piris, J., Spandidos, D. A., and Wyllie, A. H. Inuminohistochemical detection of the ras oncogen p21 product in an experimentaltumour and in human colorectal neoplasma. Br. J. Cancer, 52: 687-693,1985.

31. Ohuchi, N., Thor, A., Page, D. L., Hand, P. H., Halter, S. A., and Schlom,J. Expression of the 21,000 molecular weight ras protein in a spectrum ofbenign and malignant human mammary tissues. Cancer Res., 46:2511-2519,1986.

32. Viola, M. V., Fromowitz, F., Gravez, S., Deb, S., Finkel, G., Lundy, J.,Hand, P. H., Thor, A., and Schlom, J. Expression of ras oncogene p21 inprostate cancer. N. Engl. J. Med., 314: 133-137, 1986.

33. Viola, M. V., Fromowitz, F., Oravez, S., Deb, S., and Schlom, J. ras oncogenep21 expression is increased in premalignant lesions and high grade bladdercarcinoma. J. Exp. Med., 161: 1213-1218, 1985.

34. Japanese Research Society for Gastric Cancer. The General Rules for theGastric Cancer Study, Ed. 11, pp. 46-57. Tokyo: Ranchara Publishing Co.,1985.

5509



1988;48:5503-5509. Cancer Res Kuniko Yoshida, Kiyohiro Hamatani, Hiroshi Koide, et al.

Genes in Human Stomach and Thyroid Cancersrasof Antibodies and Immunohistochemical Analyses on Expression

21,000 Protein MonoclonalrM rasPreparation of Anti-

Updated version

http://cancerres.aacrjournals.org/content/48/19/5503

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

[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://cancerres.aacrjournals.org/content/48/19/5503To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



preparation of anti-ras mr 21,000 protein monoclonal...

Documents