Immunohistochemical Detection of K-sam Protein in Stomach ...Science, Sports and Culture of Japan; and by the Uehara Memorial Foundation, the Research Foundation for Cancer and Cardiovascular

Vol. 2, 1373-1381, August 1996 Clinical Cancer Research 1373

Immunohistochemical Detection of K-sam Protein in

Stomach Cancer1

Yutaka Hattori, Hiroshi Itoh, Shinya Uchino,

Kouichi Hosokawa, Atsushi Ochiai, Yoshinori Ino,

Hideshi Ishii, Hiromi Sakamoto,

Naohito Yamaguchi, Kazuyoshi Yanagihara,

Setsuo Hirohashi, Takashi Sugimura, and

Masaaki Terada2

Divisions of Genetics [Y. H., H. It., H. Is., H. S., T. S., M. T.],

Pathology [S. U., A. 0.. Y. I., S. H.], and Cancer Information and

Epidemiology [N. Y.]. National Cancer Center Research Institute. and

National Cancer Center Hospital [K. H.], Tokyo 104; and Department

of Pathology, Research Institute for Nuclear Medicine and Biology.

Hiroshima University. Hiroshima 734 [K. Y.J, Japan.

ABSTRACTThe K-sam gene, originally isolated as an amplified

gene from the stomach cancer cell line KATO-Ill, is

characterized by its preferential amplification in the un-

differentiated type (diffuse type) of stomach cancer and

encodes one of the receptors for heparin-binding growth

factors or fibroblast growth factors. The K-sam gene has

been isolated by different methods and has been desig-

nated BEK, TK14, and Cek2. The receptor for keratino-

cyte growth factor was also found to be encoded by the

same gene. To examine the expression of the K-sam pro-

tein in stomach cancer, polycbonal antibody pKl-2 was

raised against the extracellular domain of the gene prod-

uct. This antibody detected K-sam proteins by Western

blot and flow cytometry analyses in stomach cancer cell

lines KATO-III and HSC39, in which the K-sam gene is

amplified and overexpressed. By immunohistochemical

analysis, 20 of 38 cases of the undifferentiated type of

advanced stomach cancer were K-sam positive, whereas

none of 1 1 cases of the differentiated or intestinal type

revealed K-sam staining. The K-sam product was ob-

served predominantly in diffusely infiltrative lesions. In

one autopsy case, the K-sam protein was detected only

focally in the primary tumor, whereas markedly in-

Received 1/22/96; revised 4/22/96; accepted 4/25/96.

I Supported in part by a Grant-in-Aid for the 2nd Term Comprehensive

10-Year Strategy for Cancer Control from the Ministry of Health andWelfare of Japan; by Grants-in-Aid for Cancer Research from theMinistry of Health and Welfare and from the Ministry of Education.

Science, Sports and Culture of Japan; and by the Uehara Memorial

Foundation, the Research Foundation for Cancer and Cardiovascular

Diseases (Osaka, Japan). and the Bristol-Myers Squibb Foundation.

H. It., S. U., and H. Is. were awardees of a Research Resident Fellow-ship from the Foundation for Promotion of Cancer Research.2 To whom requests for reprints should be addressed. at Genetics Divi-sion, National Cancer Center Research Institute, 1-I, Tsukiji 5-chome.

Chuo-ku, Tokyo 104. Japan. Phone: 81-3-3542-251 1. extension 4400;

Fax: 81-3-3541-2685.

creased staining for the K-sam product was detected dif-

fusely in the metastasized tumor in the lymph node and

liver. These results suggest that K-sam overexpression is

associated with the malignant phenotype of the undiffer-

entiated type of stomach cancer, such as infiltrative

growth and metastasis.

INTRODUCTION

Although stomach cancer is the most common malignant

disease in the world after lung cancer, information on the

genetic changes of stomach cancer has just begun to be accu-

mulated (1, 2). Little has been known about the correlation

between elinicopathobogical characteristics and genetic informa-

tion. Stomach cancers are histopathobogically classified into two

types: undifferentiated (diffuse type) and differentiated (intesti-

nal type). These two types are also divergent in terms of bio-

logical behavior (3). For example, undifferentiated carcinoma is

more frequently observed in the younger generation than is the

differentiated type, demonstrates diffusely infiltrative growth.

readily causes peritoneal dissemination, and is thought to be

unfavorable in prognosis (4). From the clinicopathobogical point

of view, it is important to clarify the molecular mechanism that

determines the biological features of the undifferentiated type of

stomach cancer.

The K-sam gene was originally identified as an amplified

gene in a stomach cancer cell line, KAIO-III, by the in-gel

DNA renaturation method (5, 6). This gene is preferentially

amplified in the undifferentiated type of stomach cancer, includ-

ing poorly differentiated adenocarcinoma, signet-ring cell car-

cinoma, and mucinous adenocarcinoma (6). K-sam is also

known to encode one of the heparin-binding growth factor

receptors or FGFRs3 (7). The K-sam gene has been isolated by

different methods and has been designated BEK, TKI4, and

Cek2 (8, 9). The receptor for KGF was also found to be encoded

by the same gene (10). The K-sam protein showed selective

affinity for KGF or basic FGF by mutually exclusive alternative

splicing of exons in the ligand-binding domain (I 1). The K-sam

gene is capable of transforming NIH3I3 cells, and the trans-

formed cells are tumorigenic in nude mice (12). Interestingly,

truncation of the COOH terminus of the K-sam gene product

potentiates transforming activity and is predominantly ex-

pressed in stomach cancer cell lines of undifferentiated type

(12).

In the present paper, we studied the expression of K-sam

protein in stomach cancer. We also immunohistoehemicalby

examined the presence of the K-sam product in surgical and

3 The abbreviations used are: FGFR, fibroblast growth factor receptor:FGF, fibrobbast growth factor: KGF, keratinocyte growth factor: KGFR.keratinocyte growth factor receptor; AMeX, acetone, methyl benzoate,

and xylene.

on April 2, 2020. © 1996 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

A)

1 2 3 4 56

116_ 116_

84

67 -

B)1 2 3

� 28Sj! �

1 8S-

S

7 8 9 10kDa

205-

67 � : . �

55 -

p1(1 -2 antibody

‘_1 6#{176}#{149}i��’�� 2 � � � �Fluorescence intensity

1374 Detection of K-sam Protein in Stomach Cancer

kDa205 -

kDa205-

84-

67��

55_�

pNl -1 antibodypKl -2 antibody

Fig. I Expression of the K-sam protein in NIH3T3 transfeetants and human stomach cancer cell lines. A, Western blot analysis with the anti-human

K-sam polyclonal antibody pKl-2 (Lanes 1-4 and 7-11) or the anti-human N-sam polyclonal antibody pNl-l (Lanes 5 and 6). Lane 1,

pcDNAlneo/NIH3T3; Lane 2, N-sam/NIH3T3; Lane 3, K-sam-11C3 (truncated COOH terminus)/NIH3T3. <, Mr 125,000 K-sam protein. The

smaller-sized band is likely a degradation product; Lane 4, K-sam-IICI (intact COOH terminus)/NIH3T3. �-, Mr 135,000 K-sam protein; Lane 5,pcDNAlneo/NIH3T3; Lane 6, N-sam/NIH3T3. i-, Mr 145,000 N-sam protein; Lane 7, KATO-IlI cells. �, Mr 125,000135,000 K-sam protein; Lane8, HSC39 cells. 4-, Mr 145,000 and 130,000 K-sam proteins; Lane 9, MKN1 cells; Lane 10, immunogen absorption test of KATO-Ill cells; Lane 11,immunogen absorption test of HSC39 cells. kDa, molecular weight in thousands. B, RNA blot analysis. Ten p.g of total RNAs were hybridized tothe K-sam-specific probe RAO.7 as described previously (7). Lane 1, KATO-IlI; Lane 2; HSC39; Lane 3, MKN1. <1, 3.5-kb major K-sam transcriptsin KATO-ifi cells; 4-, 4.7-kb and 4.0-kb K-sam transcripts in HSC39 cells.

,g0 __________ �__________ __________

�[��0h/3T3 � � I ______� i6#{176}#{149}�1�2 #{149}i�6�� i6#{176}161 i62 i6� i� � i6#{176}i61 i62 i6� i�

Fluorescence intensity Fluorescence intensity Fluorescence intensity

Fig. 2 Flow cytometry assay of NIH3T3 transfeetants and stomach cancer cell lines with pKl-2. KATO-Ill and HSC39 were K-sam positive, and

MKN1 was K-sam negative. Cells were reacted with or without pKl-2. The reaction was developed with anti-rabbit IgG labeled with FITC. Solid

lines, cells incubated with pKl-2; dotted lines, cells incubated without pKl-2.

postmortem samples of stomach cancer. It was found that

K-sam is preferentially expressed in the undifferentiated

type, and that the expression is associated with malignant

phenotypes such as infiltrative growth and remote metastasis.

PATIENTS AND METHODS

Cell Culture. Stomach cancer cell lines KAIO-Ill,

HSC39, and MKN1 were cultured in RPMI 1640 supplemented

with 10% FCS as described previously (13, 14). For establish-

ment of K-sam and N-sam transfectants to NIH3I3 cells,

K-sam-IIC l/KGFR, K-sam-IIC3IKGFR with truncated COOH

terminus, and N-samIFGFR1 eDNAs were cloned into

pcDNAlneo expression vector (Invitrogen, San Diego, CA) and

were transfected by the calcium phosphate method as described

previously (12, 15, 16). After two weeks of eoeulture with

G4l8, independent colonies were isolated and were used for



A) B) C)

�! . 4 �

. C. .*

�- i 0.� � I,. S _

4. .�

5,

�� S

.�,�. C

�.,5___..#{248}. #{149}� �;, �. � .i#{234}.�i. �

Clinical Cancer Research 1375

D)t �..‘ .. �: ‘ �. :- 0 ‘ � � .�

� $�.:;�:*!%� �

;�) *� � � � :?��,! 4 � �I4�

? � � �

�;:��) � � .� �tc..,,. C.”

t� ‘#{188}’

�Fig. 3 Immunoeytoehemieal staining of K-sam protein in K-sam transfeetants to NIH3T3 cells and KATO-III cells. A, peDNAlneo/NIH3T3; B,K-sam-IIC3 (truncated COOH terminus)/NIH3T3; C, K-sam-IICI (intact COOH terminus)/NIH3T3 cells; D, KATO-Ill.

further analysis. For Western blotting and flow cytometry, cub-

ture mediums were changed one day before the experiments to

obtain logarithmically growing cells.

Patients and Histological Characteristics. Surgical

specimens from 49 Japanese patients with advanced gastric

cancer were used in this study. A!! of the patients were admitted

to and operated on at the National Cancer Center Hospital in

Tokyo between May 1990 and October 1991. Pathological di-

agnoses and clinical staging were based on the Japanese

Classification of Gastric Carcinoma by the Japanese Re-

search Society for Gastric Cancer ( I 7). Based on the Japa-

nese Classification of Gastric Carcinoma or on Borrman’s

classification ( I 8), tumors were macroscopically divided into

four types: type 1 , pobypoid tumors; type 2, ulcerated carci-

nomas with sharply demarcated and raised margins; type 3.

ulcerated carcinomas infiltrating the surrounding wall; and

type 4, diffusely infiltrating carcinomas without marked ul-

ceration. Surgical and autopsy specimens were fixed by ac-

etone and were embedded in paraffin by the AMeX method

described previously (19).

Western Blotting and Dot-Blot Analysis. The anti-hu-

man K-sam antibody pKl-2 was raised by the inoculation into

rabbits of the synthetic oligopeptide conjugated with bovine

thyrogbobulin, consisting of residues 23-38 of the K-sam type II

product P23SFSLVED1TLEPEDA38, which corresponds to just

downstream of the signal peptide in the extracellular domain of

K-sam protein (7). Immunoaffinity purification of the antiserum

was carried out by protein A and immunogen synthetic oh-

gopeptide. In the same way, the anti-human N-samIFGFR1

polyclonal antibody pNl-l was raised by the inoculation of

synthetic oligopeptide conjugated with bovine thyroglobuhin,

18TGEEVEVQDSVPADSGLYA’#{176}#{176}, and was purified by im-

munoaffinity column (20).

For Western blot analyses, cell bysates from cell lines

and paraffin-embedded sections of surgical and autopsy spec-

imens were prepared by a radioimmunoprecipitation assay

buffer [RIPA buffer: 50 mrvi Iris (pH 7.4), 150 m�i sodium

chloride, 1% NP4O, 0.1% SDS, 0.5% sodium deoxycholate,

1 ms� phenylmethylsulfonylfluoride, and 0.2 units/mI aproti-

nm]. Cell lysates (50 sag) were eleetrophoresed in 7.5%



.�o ,,,

�

--‘�,f�:.� .5 ,rfr,�.

4�.5- � -,


A)

. _e�, � �. C.�:. #{149}� � �, � -C ) ‘:� � � �r�.i#{149} �

. - C ‘.: , .,

a #{149}� #{149} �3#{149} �,.C.t,. #{149},.. ;. -. �_.;‘ � .�. ,, .‘*

‘�;� ‘ �r � � .rk�.. .‘..�

�;. : � � #{149}� �?eC� :. .�; S�*4� �., � #{149}� � � #{149}‘�e. � �. ,�� ‘�ra......;...,.. � . .

0 � � ... I,

��lC% , #{149}�..s � #{149}:. � .. #{149}��.., .5 #{149}.* ._ . S 1�.e.b��;; � � � #{149}1

‘�#{149} _ � .. ‘V’.. ,�� :1 i.� �‘�: � .,� � .�

�#{149} � ., � #{149}:. . � � #{149}. . � � � �

� .. -�.‘..-. �.� � � � .. .

.� � � � �.;.#{149}.� . �

..�.. � . -‘.�__#{149}�S�_ � � - ..� -�

,.#{149} #{149}1 C , � �

S � �, � %.; . �.�.,. �r ‘��:#{149}� � #{149}� .�. ,. . .. � ,‘-r, �‘,: ‘ : � �_. -

S �;.� .. . ,�

B).... �,. .

. - � #{149}�. #{149}�#{149}‘ ,* . � � 5, � , . �. S

.. -.�*. S � � . � -.

,.,�5. S � � . .

- . .‘...-, . , SS�_ � #{149}� t�- - . . a .

-i#{149}_. , S� � � � , . .� .5..,

‘tti�as;%’*�..� 5. � �, �

� , S #{149}� ‘ ‘�#{149} ‘ �

#{149}� � : ‘ : �:.‘#{149}_#{149}-,_�‘t�

n� - ;‘#{149}.. . � : “� � � � /‘. � .4.�

,. �

�.. , .5.. .::: .‘.�,,,. ... ,..) ,�/‘3�?’.

/ � �- $1� �

��15a� � �

‘�#{149}�1�’- ‘,I ‘‘�; .,. . . .5 .,4�.I:. :‘� �S, ‘ � . .:�.

: .� -� :�: ‘ � �

S , ‘ � S #{149}� � � :� �

b � �:r, ‘. S � #{149} ,‘ � �3’ .� .5 .: .� � ..�#{149},,. -�

,,#{149}.t� I- �

( .7

Fig. 4 Immunohistoehemical

\�. characterization of the anti-hu-

man K-sam antibody in stom-

ach cancer tissues. Patient is a

68-year-old female with poorly

differentiated adenocarcinoma,

stage IV (A-C, X 100). Stomach

cancer cells diffusely infiltrated

the submucosa, muscle layer,

and subserosa. Seirrhous reac-

tion was observed. A, K-sam

immunostaining with pKl-2

antibody; B, K-sam immuno-staining after treatment of

pKI-2 antibody with antigen

ohigopeptide (0. 1 ig/ml); C,stained with H&E; D, nonean-

cerous portion of the stomach

(X40). Note the weak staining

in the fundic glands.

SDS-polyacrylamide gel under reducing conditions and were

transferred to polyvinylidene difluoride membrane (Immo-

bilon; Millipore, Bedford, MA), using blotting apparatus

(Trans-Blot CELL; Bio-Rad, Richmond, CA). Filters were

incubated with pKl-2 at a dilution of 1:1000 with or without

immunogen oligopeptide (0. 1 p.g/ml) for 1 h at room tem-

perature. The detection of K-sam protein was done by the

enhanced chemiluminescence system (Amersham, Bucking-

hamshire, United Kingdom). The peroxidase-conjugated don-

key anti-rabbit IgG was used as the secondary antibody at a

dilution of 1:5000.

DNA purification from the paraffin-embedded sections of

these tumors was also described elsewhere (21). DNAs (10 jig)

were blotted on Nitroplusmembrane (Micron Separations, West-

borough, MA) and were hybridized to the K-sam-specific probe

RAO.7 under high-stringency conditions as described previously

(7).

Flow Cytometry. Cells were treated with or without

pKl-2 antibody at a 1 : 100 dilution in PBS with 2% FCS for 30

mm on ice. Cells were then incubated with fluorescein-conju-

gated anti-rabbit IgG (Cappel, West Chester, PA) at a 1:50

dilution in PBS with 2% FCS for 30 mm on ice. K-sam or

N-sam expression in stained cells was analyzed by using the

Becton Dickinson FACSCalibur (San Jose, CA).




Table 1 K-sam immunoreaetivity and histologica I type

No. (%) of patients

P

K-sam positive K-sam negative

(n20) (n=29)

Histological type’�

Undifferentiated type

Porl

Por2

Sig

Mue

Differentiated type

Pap

Tubl

Tub2

20 (53) 18 (47)

2(29) 5(71)

13(59) 9(41)

2 (50) 2 (50)

3 (60) 2 (40)

0 (0) 1 1 (100)

0(0) 2(100)

0(0) 2(100)

0(0) 7(100)

P = 0.0018

a Porl, poorly differentiated adenoearcinoma. solid type; Por2,

poorly differentiated adenocareinoma. nonsolid type: Sig, signet-ring

cell carcinoma: Mue, mucinous adenocarcinoma; Pap. papillary adeno-

carcinoma; Tub 1 , well differentiated tubular adenocarcinoma; Tub2,

moderately differentiated tubular adenoeareinoma.

Immunohistochemical Staining. Immunocytoehemical

and immunohistochemical analyses were performed as follows:

5-jim-thick AMeX sections of stomach cancer cell line and

surgical or autopsy specimens were cut and deparaffinated.

NIH3I3 transfeetants were cultured in chamber slides (Nune,

Roskilde, Denmark) and were washed with PBS. After blocking

endogenous peroxidase activity and nonspecific binding, the

anti-human K-sam antibody pKl-2 was applied to the sections

at a 1 :250 dilution with or without immunogen oligopeptide (0.1

jig/ml) overnight at 4#{176}C.Subsequently, biotinybated goat anti-

rabbit immunogbobulin (Vector Laboratories, Inc., Burlingame,

CA) was added at a 1 :200 dilution, and the sections were

incubated for 1 h at room temperature. An avidin-biotin-horse-

radish peroxidase complex (Vector Laboratories, Inc.) was then

applied at a dilution of 1 : 100 and was incubated for 1 h. Finally,

0.02% diaminobenzidine (DAKO Corp., Carpinteria, CA) and

0.02% H2O7 were used to visualize immunoreactivity, and

counterstaining was performed with methyl-green.

Statistical Analyses. Chinicopathobogical features of K-

sam product-positive and -negative eases of stomach cancer

were compared by Fisher’s exact test for dichotomous variables

or by Wileoxon’s rank-sum test for polychotomous variables.

Overall survival was defined as the time from surgical operation

to death. Survival distributions were estimated by the Kaplan-

Meier method, using the Statistical Analysis System (SAS)

procedure (22). The significance of the difference in survival

rates was tested by use of the log-rank test.

RESULTS

Characterization of the Anti-K-sam Antibody pKl-2.

The specificity of this antibody was examined by Western blot

analysis in NIH3T3 cells transfected with K-sam eDNAs. As

shown in Fig. 1A, K-sam-IIC3 (KGFR with a truncated COOH

terminus)INIH3I3 cells demonstrated a Mr 125,000 protein

(Lane 3), and K-sam-IICI (KGFR with an intact COOH termi-

nus)INIH3T3 cells demonstrated a Mr 135,000 protein (Lane 4).

These results are consistent with those reported previously (23).

At least four members of the samIFGFR family have been

Table 2 K-sam immunoreactivity and elinicopathological

characteristics of patients with undifferentiated stomach cancer

No. of patients

K-sap,, K-sampositive negative

(Ii = 20) (,i = 18) P’

Age ± SD 53.4 ± 3.3 55.2 ± 3.0

Sex (male:female) 9: 1 1 7: 1 1 0.75Peritoneal dissemination

+ 7 3 0.20

- 13 15Lymph node metastasis”

nO 2 3 0.32

nI 3 8

n2 lb 2

n3 2 0

n4 2 5

Stage I + 2 3 8 0.091

3 5 3

4 12 7

Macroscopic observation’

Borrman type 1 , 2, 3 6 12 0.050

type4 14 6Depth of tumor infiltration”

mp I 3 0.25

55 3 4

se 10 7

Si 6 4

Type of infiltration”

Infiltrative (lNF--�) 16 9 0.087

Noninfiltrative (INF-a + -�3) 4 9

Stromal reaction

Seirrhous type’ 17 14 0.69

Nonseirrhous type 3 4

Curativity of surgical resection

Curative resection 9 1 1 0.35

Noncurative resection I 1 7

a Clinicopathological features of K-sam product-positive and -neg-

ative eases of stomach cancer were compared by Fisher’s exact test for

dichotomous variables or Wileoxo&s rank-sum test for polychotomous

variables.

h No., no. lymph node metastases: nl-n4, group 1-4 lymph node

metastasis.

(. Macroscopic classification based on the Japanese Classification

of Gastric Carcinoma by Japanese Research Society for Gastric Cancer

(17) or by Borrman (18). Type 1, polypoid tumors; type 2. ulcerated

carcinomas with sharply demarcated and raised margins; type 3. ulcer-

ated carcinomas infiltrating the surrounding wall: type 4. diffuselyinfiltrating carcinomas without marked ulceration.

(, Mp. muscularis propria; ss. subserosa: se, tumor invasion ob-served on the serosal surface and the peritoneal cavity; si. tumor inva-

sion observed in adjacent structures.

e IFN--y, infiltrating growth without a distinct border from sur-

rounding tissue; lFN-�3, between IFN-a and IFN-y; IFN-cs, expanding

growth without a distinct border.

J Seirrhous type, tumor with abundant stroma.

reported, including FGFR1IN-sam, FGFR2/K-samIKGFR,

FGFR3/sam3, and FGFR4 (7-10, 20, 24). The comparison of

deduced whole amino acid sequences showed that N-sam was

the molecule most closely related to K-sam. The amino acid

sequence of the immunogen ohigopeptide of K-sam was 3 1 %

homologous to the corresponding amino acid sequence of N-

sam, 0% homologous to that of sam3/FGFR3, and 6% homol-

ogous to that of FGFR4. To examine the cross-reaction of



C)A)

� �--- -�--.-.-#{149}.

. I � :

�

. :� ‘ #{149} � - .- I.’ � �-S 5’ � t,

4�::� � : �.

‘; �

� � e�;

Fig. 5 Immunohistochemical detection of K-sam protein in stomach cancer cells. A, a 50-year-old male with poorly differentiated adenocareinoma,

stage IV. Tumor cells that invaded the lymphatic duets were predominantly stained. B, same ease as in Fig. 2. C, a 5 1-year-old female with poorly

differentiated adenocareinoma, stage IV. Tumor cells diffusely infiltrated the muscle layer and subserosa. Scirrhous reaction was clear. A, X200; B,X200; C, X200.


? ‘ � �p�.sI._ #{149}� -. #{149}�

� �� . : � � � � � :. ‘.1. - .-.�r’ ‘ ‘� �. �.... S

-S.,,, Q� ‘�: � - #{149}.

. . ., r -�.. -

. .b- �. ;‘

�‘�:.- �

pKl-2 with the N-sam product, Western blot analysis was

conducted with pKl-2. On N-sam-transfected N1H313 cell

lysates, no N-sam signal was detected, although the N-sam gene

product was overexpressed in these transfectants (Fig. 1A, Lanes

2 and 6). In KATO-III cells, a broad, strong band with a Mr Of

125,000-135,000 was observed by Western blotting (Fig. 1A,

Lane 7). The strong signal was also detected by immunopre-

cipitation of the metabolically labeled KAIO-HI lysates (23)

and was likely to correspond to the translational product of the

3.5-kb mRNA detected by Northern blot analysis (Fig. 1B, Lane

1). The K-sam 4.7-kb and 4.0-kb mRNAs were also overex-

pressed in the human stomach cancer cell line HSC39 (Fig. 1B,

Lane 2). In this cell line, two bands (Mr 145,000 and Mr

130,000) of proteins were observed by Western blot analysis

(Fig. 1A, Lane 8). The Mr l25,000135,000 strong signal in

KATO-Ill cells and the Mr 145,000 and Mr 130,000 signals in

the HSC39 cells were all completely absorbed by incubation

with immunogen oligopeptides (Fig. lA, Lanes 10 and 11). The

different sizes of the K-sam products were probably generated

by the alternative splicing of the K-sam transcripts (12, 25) and

by the genomic rearrangement in the course of K-sam gene

amplification. In MKN1 cells from differentiated adenocarci-

noma, there was no K-sam mRNA or protein (Fig. 1A, Lane 9;

Fig. lB. Lane 3).

The peptide antiserum was directed against the amino

terminus of the receptor. Therefore, it did not distinguish K-

sam-II/KGFR from the K-sam-IIBEK isoforms. We have shown

previously that the amount of the K-sam-IJBEK type of mRNAs

was significantly less than that of K-sam-ILIKGFR in all the

stomach cancer cell lines examined (15). Furthermore, all of the

24 K-sam eDNA clones obtained from the eDNA library of

KAIO-III cells contained K-sam-IIIKGFR-type cDNAs (12).

These results suggest that the K-sam/FGFR2 proteins detected

by this antibody in stomach cancer cells are very likely to be

mainly K-sam-IIJKGFR proteins, whereas the amounts of K-

sam-I/BEK proteins detected are likely to be small.

Expression of the K-sam protein was also examined by

flow cytometry with pKl-2 antibody (Fig. 2). The intensity of

FIIC staining was consistent with the results obtained by West-

era and RNA blotting. High K-sam expression was observed in

K-sam-transfected NIH3I3 cells, KAIO-Ill, and HSC39, but no

K-sam signal was detected in NJH3T3 cells transfected with

vector or MKN1 cells.

Immunocytochemical Stainings of K-sam Products.

Results of the immunocytochemical staining of NIH3T3 trans-

fectants and the stomach cancer cell line KATO-Ill with pKl-2

are shown in Fig. 3. The cytoplasm and membrane of NIH3T3

transfectants of K-sam-IIC3IKGFR with truncated COOH ter-

minals and K-sam-IIC1/KGFRs with intact COOH terminals

were clearly stained, but in-vector transfectants demonstrated no

K-sam staining. In KATO-IJI cells, the periphery of the cells

was clearly stained, indicating the presence of K-sam products

on the cell membrane. The specificity of immunocytochemical

staining was also confirmed by immunogen absorption test and

by negative staining without pKl-2 incubation (data not

shown).

Expression of K-sam Protein in Surgical Specimens of

Gastric Cancer. The staining pattern with pKl-2 was com-

pared between cancerous and noncancerous samples (Fig. 4). In

a slice-section from a noncancerous area, only the fundic glands

in the stomach wall were weakly stained in the cytoplasm (Fig.

4D). The signals were absorbed by incubation with immunogen

oligopeptide, suggesting K-sam-specific staining. In contrast,

tumors were stained more intensely than noneancerous cells

(Fig. 4A). The tumor cells were judged as K-sam-staining pos-

itive when the following conditions were fulfilled: (a) the tumor

cells were stained much more strongly than the noncancerous

portion of the same sample; and (b) the staining was absorbed

by immunogen oligopeptides (Fig. 4B). A total of 49 AMeX

samples of advanced stomach cancer (including 38 samples of

undifferentiated type and 1 1 samples of differentiated type)

obtained at surgical resection were studied. Table 1 summarizes

the results. Of the 49 samples, 20 were positive, and all of these

positive samples were diagnosed as undifferentiated type, in-

eluding 15 cases of poorly differentiated adenocareinoma, 2

eases of signet-ring cell carcinoma, and 3 cases of mucinous

adenocarcinoma. Eighteen undifferentiated-type samples were

negative. In contrast, none of the 1 1 cases of differentiated



I

B) la lb 2 3

�( �‘‘.

-

1’

\�T

_.:#{149}� �

�.; � 5_�’ . 5, .

Fig. 6 Postmortem specimens of a stomach cancer patient with mucinous adenocareinoma. A, dot-blot analysis of K-sam with RA 0.7 probe. 1.

primary stomach tumors; 2, metastasized tumors in the liver; 3, metastasized tumors in the para-aortie lymph nodes. B, immunohistoehemical staining

of K-sam. Numbers of the photographs correspond to 1-3 of A. Ia. primary site of tumors in the stomach. Tumor cells were not clearly stained with

pKl-2 antibody. lb. small group of stomach cancer cells that invaded the stomach serosa were selectively stained with K-sam antibody. 2, liver

metastasis; 3, para-aortie lymph node metastasis. la, X 100; Ib, X400; 2, X 100; 3, X 100.

-0�-- K-sam negative (n=1 8)

*�kQt� � �sam positive (n=20)

-C-- K-sam negative (n=7)

-0’--’ K-sam positive (n=12)

5i�i0 idoo i�oo 2000 (days)

Survival

Fig. 7 Kaplan-Meier survival analysis of patients with K-sam immunohistochemicab staining-positive and -negative cells of undifferentiated type of

stomach cancer in the primary lesion. Left, patients in all stages (log-rank, P = 0.155): right, patients in stage IV (log-rank. P = 0.102).

Survival


A)12 3

S� . ; ;5, . � . ; .� . �

, (

.�.

�, C..

1 � � � %�_‘#{149}‘

‘��:; ‘�‘�‘

S -1:�

S “ : , � “ #{149}� � ‘�L�4� �. ,.. .. . . I” �

..‘: #{149} �‘�‘‘�‘ “ �. .). � �.

.,‘ .. ..; �

;, �:;; � ;;.. �

.. + .. ._t,

p #{149}%

0.8Ce>

� 0.6

0>.

:B 04Ce

.00

0� 0.2

0.8Ce>

� 0.6

0

:a 0.4Ce

.00

a. 0.2

adenocarcinoma showed positive staining, although 8 of them

were in advanced stage IV. The representative cases are shown

in Fig. 5. In case 1, tumor cells that invaded the lymphatic duet

were predominantly stained. In ease 2, the K-sam gene was

found to be amplified by dot-blot hybridization of DNAs iso-

bated from a paraffin-embedded section (data not shown). In

ease 3, tumor cells that invaded diffusely into the muscle layer

were strongly stained. The presence of K-sam protein was

confirmed by Western blot analysis (data not shown).

The correlation of K-sam-positive staining with various

elinicopathological characteristics of patients with the undiffer-

entiated type of stomach cancer is summarized in Table 2.

Macroscopically, Borrman type IV or diffusely infiltrative stom-

ach cancer was frequently stained compared to other types (P =

0.050). The infiltrative growth pattern examined by microscopic

observation was classified into three types in the Japanese

Classification of Gastric Carcinoma by the Japanese Research

Society for Gastric Cancer (17). Namely, in IFN-a, tumor cells




grow, constructing a tumor nest, and tumors are bordered clearly

from the noncancerous area. IFN43 has characteristics between

IFN-a and IFN--y. In IFN-’y, tumors demonstrate an infiltrative

growth pattern, and the difficulty is to show a clear margin

between the tumor mass and the noncancerous portion. The

P-value of association with infiltrative growth pattern IFN--y

was 0.087.

Seven cases of postmortem AMeX specimens of stomach

cancer and their metastasized tumor samples were available for

analyses of the expression of K-sam protein (Fig. 6). In one

case, K-sam products were found focally only in the primary

tumor, which infiltrated deeply to the stomach subserosa. How-

ever, in this case, gene amplification was undetectable in the

primary stomach tumor by dot-blot analysis. In contrast, metas-

tasized tumors of the liver and para-aortie lymph nodes of the

same patients were shown to be K-sam-product positive with the

amplified gene.

To elucidate the clinical relevance of K-sam immunohis-

tochemical positivity, we compared the survival of 20 K-sam

product-positive cases and 18 patients with the undifferentiated

type of stomach cancer by Kaplan-Meier methods (Fig. 7). The

results showed that there was no statistically significant differ-

ence in survival between patients in all stages with K-sam-

positive stomach cancer and those with K-sam-negative stom-

ach cancer (P = 0. 155). Among the patients in stage IV, the

difference in survival between K-sam-positive and -negative

cases (Fig. 7) was not statistically significant either (P = 0.102).

DISCUSSION

We reported previously that amplification of the K-sam

gene preferentially occurs in the undifferentiated type of stom-

ach cancer, whereas c-erbB-2 amplification occurs in the dif-

ferentiated type (6, 26). In the present study, anti-human K-sam

polyclonal antibody was raised and characterized to study the

expression of K-sam protein in stomach cancer. Western blot

and flow cytometry analyses using NIH3I3 cells transfected

with K-sam and N-sam eDNAs revealed that this antibody

specifically detected K-sam gene products and did not cross-

react with N-sam protein, which is the most homologous to the

K-sam protein. Both Western blotting and flow cytometry assay

clearly demonstrated the expression of K-sam protein in the

stomach cancer cell lines KATO-Ill and HSC39 with K-sam

amplification. The sizes of K-sam protein were different in these

two cell lines, and the difference was probably caused by

various patterns of alternative splicing (12, 25), by genomic

rearrangements during the process of the gene amplification (7),

or by a different degree of glycosylation of the K-sam protein.

Immunohistochemical detection revealed that even non-

cancerous tissues were stained weakly by pKl-2 antibody.

These signals were believed to represent K-sam proteins be-

cause (a) the staining was absorbed by incubation with immuno-

gen oligopeptides; (b) pKl-2 antibody did not cross-react with

other FGFR molecules; and (c) signals detected in Western

blotting were considered to be the K-sam proteins. K-sam!

KGFR is thought to be expressed selectively in epithelial cells

and mediates proliferative signals in response to KGF stimula-

tion ( 1 1 , 27). Therefore, it is reasonable that epithelial cells of

noncancerous stomach mucosa were stained with anti-human

K-sam antibody. Immunohistochemical analyses clearly showed

that K-sam overexpression was predominantly observed in the

undifferentiated type of stomach cancer, but not at all in differ-

entiated type (P = 0.0018). Although the overexpression of

K-sam protein was not always accompanied by gene amplifica-

tion, the present results were consistent with our previous data,

in which K-sam gene amplification was observed preferentially

in the undifferentiated type (6). K-sam overexpression was also

significantly associated with the maeroseopieally, diffusely in-

filtrative lesion Borrman type 4 (P = 0.050; Table 2). Also, in

microscopic observation, an infiltrative tumor growth pattern

tends to be more often observed in K-sam-positive eases (P =

0.081). Additionally, in one postmortem ease, K-sam-positive

cells existed focally only in the primary site but occupied the

entire tumor cell in metastatie sites (Fig. 6). It is likely that

tumor cells with an overexpression of K-sam product may have

a growth advantage, may elonally evolute with more infiltrative

capacity, and may have increased potential to metastasize. In

fact, we often found cases in which tumor cells that invaded the

lymphatic duct were significantly stained with pKl-2 antibody

(Fig. SA, ease 1). It is possible that the putative K-sam ligand

KGF may be expressed in the interstitial tissues in stomach or in

metastatic lesions, and that tumor cells with the K-sam!KGFR

overwhelmingly proliferate by paracrine mechanism (27). Many

of the eases with overexpression of the K-sam protein caused a

seirrhous reaction. In these cases, interstitial tissues were highly

abundant by microscopic observation and may produce KGF. In

vitro study also showed that overexpression of the K-sam gene

can transform NIH3T3 cells and can facilitate tumorigenesis

( 1 2). Therefore, these results suggest that K-sam overexpression

is likely one of the determinants of a biologically malignant

tendency of the undifferentiated type of stomach cancer rather

than a simple tumor marker.

It will be worthwhile to further study whether or not

K-sam-positive and -negative eases in the early stage take

different clinical courses. These findings also suggest that K-

sam may be a target molecule for the new approach of bio-

therapy of stomach cancer; for example, gene therapy or immu-

notherapy. Further investigation will be needed to elucidate the

more precise clinical relevance of studying the overexpression

of K-sam gene products in stomach cancer.

ACKNOWLEDGMENTS

We thank Yuko Yamauchi for technical assistance.

REFERENCES

1. Parkin, D. M., Pisani, P. E., and Ferlay, J. Estimates of the worldwide

incidence of 18 major cancers in 1985. Int. J. Cancer, 54: 594-606,

1993.

2. Hirohashi, S., and Sugimura, T. Genetic alterations in human gastric

cancer. Cancer Cells (Cold Spring Harbor), 3: 49-52, 1991.

3. Sugano, H., Nakamura, K., and Kato, Y. Pathological studies of

human gastric cancer. Aeta Pathol. Jpn., 32: 329-347, 1982.

4. Maruyama, K. Results of surgery correlated with staging. In: P. E.Preece, A. Cusehieri, and J. M. Wellwood (eds.), Cancer of the Stom-

ach, pp. 145-163. Orlando, FL: Grune & Stratton, 1986.

5. Nakatani, H., Tahara, E., Yoshida, T., Sakamoto, H., Suzuki, T.,

Watanabe, H., Sekiguehi, M., Kaneko, Y., Sakurai, M., Terada, M., and

Sugimura, T. Detection of amplified DNA sequences in gastric cancers




by a DNA renaturation method in gel. Jpn. J. Cancer Res.. 77: 849-853.

1986.

6. Nakatani, H., Sakamoto, H., Yoshida, T., Yokota, J., Tahara, E.,Sugimura. T., and Terada, M. Isolation of an amplified DNA sequencein stomach cancer. Jpn. J. Cancer Res., 81: 707-710, 1990.

7. Hattori, Y., Odagiri, H., Nakatani, H., Miyagawa, K., Naito, K.,

Sakamoto, H., Katoh, 0., Yoshida, T., Sugimura, T., and Terada, M.

K-sam, an amplified gene in stomach cancer, is a member of the

heparin-binding growth factor receptor genes. Proc. Nail Acad. Sei.

USA, 87: 5983-5987, 1990.

8. Jaye, M., Schlessinger, J., and Dionne, C. A. Fibrobbast growth factor

receptor tyrosine kinases: molecular analysis and signal transduction.

Biochim. Biophys. Acta, 1135: 185-199, 1992.

9. Johnson, D. E., and Williams, L. T. Structural and functional diver-sity in the FGF receptor multigene family. Adv. Cancer Res.. 60: 1-41,1993.

10. Miki, T., Fleming, T. P., Bottaro, D. P., Rubin, J. S., Ron. D., and

Aaronson, S. A. Expression eDNA cloning of the KGF receptor by

creation of a transforming autocrine loop. Science (Washington DC).

251: 72-75, 1991.

1 1. Miki, T., Bottaro, D. P., Fleming, T. P., Smith, C. L., Burgess,W. H., Chen, A. M-L., and Aaronson, S. A. Determination of bigand-binding specificity by alternative splicing: two distinct growth factor

receptors encoded by a single gene. Proc. NatI. Acad. Sci. USA, 89:

246-250, 1992.

12. Itoh, H., Hattori, Y., Sakamoto, H., Ishii, H., Kishi, T., Sasaki, H.,

Yoshida, T., Koono, M., Sugimura, T., and Terada, M. Preferentialalternative splicing in cancer generates a K-sam messenger RNA withhigher transforming activity. Cancer Res., 54.’ 3237-3241, 1994.

13. Sekiguehi, M., Sakakibara, K., and Fujii, G. Establishment of

cultured cell lines derived from a human gastric carcinoma. Jpn. J. Exp.

Med., 48: 61-68, 1978.

14. Yanagihara, K., Seyama. T., Tsumuraya, M.. Kamada. N.. and

Yokoro. K. Establishment and characterization of human signet-ring cellgastric carcinoma cell lines with amplification of the e-invc oncogene.Cancer Res., 51: 381-386, 1991.

15. Ishii, H., Hattori, Y., Itoh, H., Kishi, T., Yoshida, T., Sakamoto, H.,

Oh, H.. Yoshida, S., Sugimura, T., and Terada, M. Preferential expres-sion of the third immunoglobulin-hike domain of K-sam product pro-

vides keratinocyte growth factor-dependent growth in carcinoma cell

lines. Cancer Res., 54: 5 18-522, 1994.

16. Ishii, H., Yoshida, T., Oh, H., Yoshida, S., and Terada, M. Atruncated K-sam product lacking the distal earboxyl-terminal portionprovides a reduced level of autophosphorylation and greater resistance

against induction of differentiation. Mol. Cell. Biol., 15: 3664-3671,

1995.

17. Japanese Research Society for Gastric Cancer. Part I (clinical,

surgical, and conclusive findings) and II (histological findings). In:

Japanese Classification ofGastrie Carcinoma. 1st English ed., pp. 2-65.Tokyo: Kanehara, 1995.

18. Borrman, R. Gesehwulste des Magens und Duodenums. In: F.

Henske and 0. Lubarseh (eds.), Handbuch der Speziellen Pathologis-

chen Anatomie und Histologie, pp. 864-871. Berlin: Julius Springer,

1926.

19. Sato, Y., Mukai, K., Watanabe, S., Goto, M., and Shimosato, Y.The AMeX method: a simplified technique of tissue-processing andparaffin-embedding with improved preservation of antigens for immu-nostaining. Am. J. Pathol., 125: 431-435, 1986.

20. Hattori, Y., Odagiri, H., Katoh, 0., Sakamoto, H., Morita, T.,

Shimotohno, K., Tobinai, K., Sugimura, T., and Terada, M. K-sam-

related gene, N-sam, encodes fibroblast growth factor receptor and is

expressed in T-lymphocytie tumors. Cancer Res., 52: 3367-3371, 1992.

21. Tsuda, H., Shimosato, Y., Upton, M. P., Yokota, J., Terada, M.,

Ohira, M., Sugimura. T., and Hirohashi, S. Retrospective study on

amplification of N-myc and e-myc genes in pediatric solid tumors and its

association with prognosis and tumor differentiation. Lab. Invest., 59:

321-327, 1988.

22. Kaplan, E. L., and Meier, P. Nonparametrie estimation from ineom-

plete observations. J. Am. Stat. Assoc., 53: 457-481, 1958.

23. Kishi, T., Yoshida, T., and Terada, M. A soluble form of K-sam/FGFR2 protein in the culture medium of human gastric cancer cells.

Biochem. Biophys. Res. Commun., 202: 1387-1394, 1994.

24. Katoh, 0., Hattori. Y., Sasaki, H., Sakamoto, H., Fujimoto. K.,

Fujii, T.. Sugimura, T.. and Terada, M. Isolation of the complementary

DNA encoding a mouse heparin-binding growth factor receptor with theuse of a unique kinase insert sequence. Cancer Res., 53: 1 136-1 141,

1993.

25. Katoh, M.. Hattori, Y., Sasaki, H., Tanaka, M., Sugano, K., Yazaki,

Y.. Sugimura. T., and Terada, M. K-sam gene encodes secreted as wellas transmembrane receptor tyrosine kinase. Proc. Natl. Acad. Sci. USA,

89: 2960-2964, 1992.

26. Yokota, J., Yamamoto, T., Miyajima, N., Toyoshima, K., Nomura,

N., Sakamoto, H., Yoshida, T., Terada, M., and Sugimura, T. Genetic

alterations of the e-erbB-2 oncogene occur frequently in tubular adeno-

carcinoma of the stomach and are often accompanied by amplification

of the v-erbA homologue. Oncogene, 2: 283-287, 1988.

27. Finch, P. W., Rubin, J. S., Miki, T., Ron, D., and Aaronson, S. A.

Human KGF is FGF-related with properties of a paracrine effector ofepithelial cell growth. Science (Washington DC), 245: 752-755,

1989.



1996;2:1373-1381. Clin Cancer Res Y Hattori, H Itoh, S Uchino, et al. cancer.Immunohistochemical detection of K-sam protein in stomach

Updated version

http://clincancerres.aacrjournals.org/content/2/8/1373

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://clincancerres.aacrjournals.org/content/2/8/1373To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



Documents

Immunohistochemical Detection of K-sam Protein in Stomach ...Science, Sports and Culture of Japan; and by the Uehara Memorial Foundation, the Research Foundation for Cancer and Cardiovascular