Cancer Therapy Vol 4, 223-230, 2006

Analysis of cytoplasmic polypeptides expression in gastric cancer and correlation with pathologic parameters

Research Article

Panayiota K. Stroumbouli^1,*, Andreas Ch. Lazaris¹, Efstratios S. Patsouris¹, Anastasios Kalofoutis², Aphrodite Nonni¹, Sofia Tseleni-Balafouta¹

¹Department of Pathology,

²Department of Biochemistry, School of Medicine, National and Kapodistrian University of Athens, 75 Mikras Assias Str., GR-115 27 Athens, Greece

__________________________________________________________________________________

*Correspondence: Panayiota Stroumbouli, PhD, 42 Dodekanisou str., GR-15234 Vrilissia, Greece; Tel: 2108101161; Fax: 2108102690; E-mail: ny11@tellas.gr

Key words: 2D eletrophoresis, cytoplasmic polypeptides, gastric cancer, pathologic parameters, prognostic markers

Abbreviations: 3-(3-cholamidotropyl) dimethylammonio)-1-prop-anesulphonate, (CHAPS) dithiothreitol, (DTT); ethyleneldinitrinotetraacetic acid, (EDTA); isoelectric focusing, (IEF); isoelectric point, (PI); molecular weight, (MW); phenylmethylsulfonyl fluoride, (PMSF); sodium dodecylsulfate, (SDS); two-dimensional electrophoresis, (2D-electrophoresis)

Received: 13 March 2006; Revised: 26 May 2006

Accepted: 1 June 2006; electronically published: July 2006

Summary

Although there have been many studies including analysis of prognostic factors in gastric cancer, there are currently very few markers that are clinically in use. The aim of this study is to relate polypeptide expression with pathologic characteristics of patients with gastric cancer in order to determinate potential prognostic markers. Using high-resolution two-dimensional electrophoresis (2D-electrophoresis), 154 stomach tissues were examined. Seventy-seven of the specimens were histologically diagnosed as primary adenocarcinomas. The other 77 specimens represented mirror biopsies from each patient, obtained far from the tumor area, from endoscopically and histologically normal gastric mucosa. Significant polypeptide expression differences had been noted comparing each cancer tissue sample with the respective mirror biopsy of the same patient. More specifically, 24 cytoplasmic polypeptides had been studied. Six of them were detected in all mirror biopsy tissues and not detected in the majority of gastric cancer tissues ( Pa, Pb, Pc, Pk, Pl, Pm), 14 polypeptides were over-expressed ( P2, P3, P4, P5, P7, P8, P9, P10, P11, P13, P15, P16, P17, P18) and 4 were under-expressed ( P1, P6, P12, P14) in cancer tissues. Tumor samples were classified in groups according to pathologic prognostic variables. Polypeptides’ densities were normalized and the results were evaluated statistically. Statistical analysis indicated that several polypeptides were associated with pathologic characteristics. Polypeptides P4 (29/6.0), P16 (17/6.9) showed an extremely significant density increase (p<0.0001) in tumor samples with 7-15 metastatic lymph nodes and in those samples with tumor invasion to the muscular layer. Polypeptide Pc (46/2.6) was detected only in poorly differentiated tumor samples and so may serve as a marker of poor prognosis. Polypeptides P4, P16, Pc may have prognostic value and their further analysis may provide useful information to the direction of developing markers of stomach cancer.

I. Introduction

Although gastric cancer mortality has declined markedly around the world the last twenty years, gastric cancer still remains the second most common cancer next to lung cancer worldwide (Roder, 2002). The results of gastric cancer treatment have been markedly improved and considerable progress in diagnostic methods is being reported. The identification of prognostic factors is important for the establishment of therapeutic strategies. Of the many factors relevant to survival, depth of invasion, lymph node metastasis and degree of differentiation have been considered major prognostic factors in gastric cancer (Ryu et al, 2003). There have been many studies including analysis of prognostic factors but there are currently very few markers that are clinically in use.

High-resolution two-dimensional electrophoresis (2D-electrophoresis) is widely used in comparative studies of protein expression levels between healthy and diseased states with the purpose of developing diagnostic markers. This method can readily separate a number of proteins based on differences in their molecular weight (MW) and isoelectric point (PI) properties (Nagase et al, 1991). High-resolution 2D-electrophoresis combines first dimension isoelectric focusing (IEF) with the conventional sodium dodecylsulfate (SDS) electrophoresis in the second dimension (Whilson, 1977).

Whether 2D-electrophoresis is followed by computational image analysis and protein identification with tumor samples could also lead to the defining of cancer–specific protein markers which will be the basis for developing new methods for the early diagnosis of cancer (Kim et al, 1998).

2D-electrophoresis has also been applied in the study of gastrointestinal track cancers (Isoda et al, 1990). More specifically, in relation to colon cancer, five major polypeptide spots were detected in tumor samples which were not detected in normal specimens. Researchers (Okuda, 1989; Tracy et al, 1982) detected two hight-molecular weight polypeptides in colon cancer tissues which were absent in normal mucosa.

In relation to gastric cancer, 140 proteins were identified in cancer stomach tissues, 7 proteins were over-expressed and 7 were under-expressed in stomach cancer (Ryu et al, 2003). An acid proteinase was detected in gastric tumor samples and was not detected in normal specimens (Aoki, 1994). Other researchers, using 2D-electrophoresis, were studied two proteins, annexin I and thioredoxin, which were over-expressed in cancer cells (Sinha et al, 1998).

The aim of this study is to relate polypeptide expression with pathologic characteristics of patients with gastric cancer in order to determine their prognostic impact.

II. Materials and methods

A total of 154 stomach tissues deriving from 77 gastric cancer patients were surgically obtained at the Surgical Department of the "Hippocration" Hospital of Athens, over a period of 6 years (1994-2000). Seventy-seven of the specimens were histologically diagnosed as primary adenocarcinomas. The other 77 specimens represented mirror biopsies from each patient, obtained far from the tumor area, from endoscopically and histologically normal gastric mucosa.

A. Preperation of stomach tissue sambles

Fresh specimens were collected directly by the surgeon in the operating room, rinsed immediately in isotonic saline to remove external blood contamination and were frozen instantly in liquid nitrogen. Samples were stored at -70° C until further use. On the day of analysis, frozen tissues were sliced with a scalpel, washed in a hogenasation buffer consisting of 10mM Tris HCL, pH 7.4, 1.5mM ethyleneldinitrinotetraacetic acid (EDTA), 0.5 mM dithiothreitol (DTT), 0.2 mM phenylmethylsulfonyl fluoride (PMSF) as a protease inhibitor and homogenized at 0°C using a Virtis homogenizer for 3x12 sec. Homogenates were centrifuged for 1h at 105.000xg and the supernatant fraction was collected. This fraction referred hitherto as cytosolic fraction was lyophilized to dryness immediately after preparation and stored at -70° C prior to electrophoresis.

2D-PAGE electrophoresis analysis of cytosolic polypeptides was performed according to the procedure of O’ Farrel in 1975) and modified by Hochstrasser and colleagues in 1988 using the BioRad 2D-PAGE system (BioRad Lab Inc., USA). Lyophilised protein samples were dissolved in sample solution (9M urea, 4 % w/v 3-(3-cholamidotropyl) dimethylammonio) -1-prop-anesulphonate (CHAPS), 1% DTT, ampholytes 3.5-10 (1%, 4-7 (4%). Polypeptide content was determined using the Ramagli and Rodriguez (1985) modification of the Bradford protein assay (Bradford, 1976).

B. First dimension (IEF)

First dimension (IEF) was performed in a 180 x 1.5 mm tube gel consisting of 30% acrylamide solution, containing 10 M urea and a 4:1 ratio of 4-7 and 3.5-10 carrier ampholytes (final concentration 5 %) 75 μg of polypeptide were loaded onto the basis end of each IEF gel. The cathode chamber was filled with freshly degassed NaOH solution (0.1M) and the proteins were focused 700V for the approximately 19h. Each sample was electrofocused in dublicate with and without internal 2D-SDS-PAGE. Standards (Bio Rad Laboratories) and pI calibration kit (BDH, labs Supplies, Poole, England). The gels were gently extruded from the glass tubes and stored in extended form on Parafilm strips at –70° C until use. The IEF gels were allowed to soak in equilibration buffer (0.04M Tris-HCL, 3.2 SDS, 0.034M DTT) for 5 min before being loaded on at the second dimensional gel and sealed with a small amount of overlay agarose (0.025M Tris base, 0.192M glycine 0.1% SDS, 0.5% agarose).

C. Two D- SDS-PAGE

Two- D- SDS-PAGE electrophoresis was performed at constant current of 35 mA/gel at 8° C. Gels were fixed in ethanol water (30%) and acetic acid (10%). Proteins were visualized on 2D gel preparations by silver staining with Sigma Silver Stain Kit (Sigma Chemicals Co.,St. Louis, USA). For increased sensitivity the staining prosedure was repeated. Gels were photographed and scanned, using a GS-700 BioRad Imaging Densitometer.

The analysis was carried out by BioRad PDQuest-2D software, based on a modification of the grid system used by Narayan et al, 1986. Spots were numbered in descending order of molecular mass and were identified by their location relative to (known) protein markers (ie, albumin, actin, transferrin). Spot densities were normalized using the above software, prior to performing comparisons between different gels. The statistical analysis was performed using student's-t-test with the package STATA (STATA Co, USA). Tumor samples were classified in groups according to pathologic prognostic variables (Table 1).

III. Results

Two-dimensional electrophoresis of the stomach cancer tissue produced about 1500 spots for each sample, every spot representing a cytosolic polypeptide.

In this study we compared each cancer tissue sample with the respective mirror biopsy of the same patient in order to evaluate any differential polypeptide expression between them. While most of the polypeptides were present in both cancerous tissues and mirror biopsy tissues, several qualitative and quantitative polypeptide differences were noted between cancerous and noncancerous samples.

As far as qualitative differences are concerned, the comparison between stomach cancer tissue and mirror biopsy of each patient showed six polypeptides in the acid area of PI between 3.0-5.0 and MW 40-46 kDa which were detected in all mirror biopsy tissues (Figure 1a) and not detected in the majority of adenocarcinomas (Figure 1b). These polypeptides were identified as follows (MW/PI): Pa (46/3.3), Pb (46/2.8), Pc (46/2.6), Pk (40/4.2), Pl (40/4.0), Pm (40/3.8).

In addition to the above significant qualitative differences, quantitative differences in polypeptide concentration were also observed, highlighting 18 polypeptides significantly altered in malignant specimens (Figure 1a, 1b).

Table 1. Classification of tumor samples according to histological characteristics of prognostic potential

Histological characteristics	n
Depth of tumor invasion
mucosa and submucosa layers	34
muscular layer	43
Lymph nodes metastasis
negative lymph nodes	12
1-6 positive lymph nodes	21
7-15 positive lymph nodes	44
Lauren's classification
Intestinal type	62
Diffuse type	15
Degree of differentiaiton
Well differentiated	9
Moderately differentiated	30
Poorly differentiated	38

Figure 1. Electrophoretic patterns of cytosolic polypeptides from stomach adenocarcinoma (A) and the respective mirror biopsy (B). Missing spots are illustrated with open circles (Pa, Pb, Pc, Pk, Pl, Pm spots) (A). All quantitative differences are indicated with small arrowheads in both figures (Alb: albumin, Actin: actin, Tran: transferin).

Fourteen polypeptides were significantly increased in tumor samples (ie, P2, P3, P4, P5, P7, P8, P9, P10, P11, P13, P15, P16, P17, P18) while four polypeptides (ie, P1, P6, P12, P14) were decreased in malignant specimens.

Statistical analysis of the polypeptide densities showed highly significant differences (p<0.001) with regard to polypeptides Pa, Pb, Pc, Pl qualitative expression in normal tissues compared to the respective densities in adenocarcinomas as well as with regard to polypeptides P9 and P10, as far as their quantitative expression is concerned. All other polypeptides showed a merely significant difference in their quantitative expression (p<0.05).

We attempted a correlation between the above mentioned polypeptide expression with clinicopathologic characteristics such as presence of lymph node metastasis, depth of tumor invasion, degree of differentiation, Lauren’s histologic type classification, in order to determine any prognostic value for the examined polypeptides.

As far as lymph nodes metastasis and depth of invasion is concerned, statistical analysis (student’s t-test) showed that polypeptides P2, P4, P10, P16, P17 were over-expressed in tumor samples with metastasis in 7-15 lymph nodes (Figure 2, Table 2) and polypeptides P3, P4, P10, P16, P17 were over-expressed in tumor samples with invasion to the muscular layer (Figure 3, Table 3). Moreover, polypeptides P4, P16 showed an extremely significant density increase (p<0.0001) in both above mentioned tumor groups while all other polypeptides showed a merely significant increase in their expression in the above tumor groups (p<0.05).

Table 2. Correlation between polypeptide densities and lymph nodes metastasis

		Lymph nodes metastasis
	Negative lymph nodes*	Positive lymph nodes
P		1-6 lymph nodes*	7-15 lymph nodes*	p- value	C%
P2	0.64±0.02	0.967±0.23	1.251±0.14	0.039	110
P4	1.32±0.44	2.093±0.85	2.73±1.37	0.0076	90
P10	0.32±0.04	1.093±0.57	1.63±1.02	0.045	190
P16	1.08±0.63	1.92±0.94	2.53±1.187	0.0052	80
P17	0.47±0.05	0.83±0.46	2.84±1.06	0.015	240

P: polypeptide, * mean density of the respective tumor samples ± statistical error, C:% change

Figure 2. Electrophoretic patterns of cytosolic polypeptides from tumor sample with metastasis in 7-15 lymph nodes where polypeptides P2, P4, P10, P16, P17 were over-expressed (A) and tumor sample with metastasis in 1-6 lymph nodes (B).

Table 3. Correlation between polypeptide densities and depth of tumor invasion

P	Depth of tumor invasion
	mucosa or submucosa layer*	muscular layer*	p- value	C %
P3	0.79±0.37	1.63±0.404	0.032	110
P4	2.15±2.61	3.93±2.31	0.0029	90
P10	1.12±1.47	1.38±1.025	0.027	30
P16	1.67±.0.12	3.18±1.45	0.007	90
P17	0.95±0.045	1.605±0.34	0.041	80

P: polypeptide, *: mean density of tumor samples ± statistical error, C: % change

Figure 3. Electrophoretic patterns of cytosolic polypeptides from tumor sample with invasion to muscular layer, where polypeptides P3, P4, P10, P16, P17 were over-expressed (A) and tumor sample with tumor invasion to mucosa or submucosa layer (B).

With regard to tumor grade, statistical analysis of the polypeptide expression (student’s t-test) have provided the following results: i) the majority of polypeptides densities were over-expressed in moderately differentiated tumors samples, except for the polypeptides (molecular weight/isoelectric point) P9(25/7.3), P11(22/6.7), P13(20/5.6); the latter were significantly increased in poorly differentiated tumor samples (p<0.05) ii) polypeptides: Pa (46/3.3), Pb(46/2.8), Pc (46/2.6), were not detected in well differentiated tumor samples; moreover, the expression of polypeptides Pa, Pb showed a significant increase in moderately and poorly differentiated tumor samples (p<0.05). iii) polypeptide Pc was detected only in poorly differentiated tumor samples (Figure 4, Table 4). With regard to Lauren’s classification of histologic type, no significant differences were noticed. Multivariate statistical analysis indicates polypeptides P4, P16 as independent prognostic factors.

IV. Discussion

In the present study, using 2D-electrophoresis, we observed 6 cytosolic polypeptides being detected in all mirror biopsy tissues and not detected in the majority of gastric cancer tissues, as well as 14 cytosolic polypeptides being over-expressed and 4 being under-expressed in cancer tissues. Our results are in agreement with those of other researchers who have also observed several polypeptides to be significantly altered in human gastric cancer tissues, by comparison to their mirror biopsies (Ryu et al, 2003). As it has been reported, changes in cytosolic polypeptides expressed in tumors of gastrointestinal tract may occur as a result of the expression of silent genes, the arrest of immature cells at an early stage of their normal differentiation or the presence of an unusual cell-cycle in actively proliferating groups of cells (Natly et al, 1998).

We have evaluated our 2D- data based on a recent 2D-map of human stomach tissue (18) in an effort to provide potential identifications. The polypeptides

Table 4. Correlation between polypeptide densities and degree of differentation

Degree of differentiation
P	Poorly differentiated	Moderately differentiated	Well differentiated	p-value
Pa	1.36±0.64	1.8±0.69	-	0.015
Pb	1.23±0.99	1.49±0.05	-	0.042
Pc	0.82±0.02	-	-	0.003
P2	1.93±0.44	2.89±0.32	Ο.65±0.21	0.041
P3	0.95±0.44	2.12±1.55	0.45±0.07	0.032
P4	1.02±0.88	1.45±0.65	0.92±0.51	0.006
P9	3.44±2.11	1.04±0.06	0.77±0.33	0.022
P10	2.38±1.98	4.14±2.36	0.78±0.42	0.055
P11	2.48±1.67	2.02±1.38	0.89±0.63	0.011
P13	1.64±0.55	1.48±0.81	1.55±0.73	0.028
P16	1.36±0.57	3.02±2.01	0.28±0.02	0.051
P17	2.13±1.69	2.57±1.11	0.34±0.05	0.048

P: polypeptide, *: mean density of the respective tumor samples ± statistical error

Figure 4. Electrophoretic patterns of cytosolic polypeptides from tumor sample poorly differentiated where polypeptide Pc is detected (A) and tumor sample moderately differentiated where polypeptide Pc is absent (B).

(MW/PI): P5 (27/6.2), P6 (26/6.0), P18 (15/5.3) correspond to the antioxidants glutathione-S-transferase, peroxiredoxin-2 and thioredoxin, respectively. These antioxidants protect cells from oxidative damage caused by various oxidative stimuli and have been related with chemoresistance of tumor cells, especially to the oxidative stress that anticancer drugs produce (Chung et al, 2001).

We also showed that several polypeptides’ expression is associated with the most significant prognostic factors in gastric cancer.

As other researchers have reported, the presence of lymph node metastasis, the depth of tumor invasion and the tumor degree of classification have prognostic significance in gastric cancer (Kim and Jung, 1987; Adachi et al, 1994) while Lauren’s classification seems to have no significant value (Fiocca et al, 2001).

Apart from the clinical and pathological prognostic factors in gastric cancer, there is a great research interest in new prognostic factors such as growth factors and receptors, oncogenes and suppressor genes, cell adhesion molecules (Ryu et al, 2003).

The polypeptides P4(29/6.0), P16(17/6.9) showed an extremely significant density increase (p<0.0001) in tumor samples with 7-15 metastatic lymph nodes and tumor invasion to the muscular layer and may serve as prognostic factors. Multivariate statistical analysis reinforces the above results.

Taking into account their presence in a small number of tumor samples, polypeptides Pa (46/3.3), Pb (46/2.8), which were not detected in well differentiated tumor samples, could be further investigated as potential markers for the biological aggressiveness because their expression showed a significant increase in moderately and poorly differentiated tumor samples polypeptides. Polypeptide Pc (46/2.6) was detected only in poorly differentiated tumor samples and so were, though to a lesser extend, polypeptides P9(25/7.3), P11(22/6.7), P13(20/5.6); the latter polypeptides may be associated with poor prognosis of gastric cancer because of their significant density increase in poorly differentiated tumor samples.

In conclusion, we propose that polypeptides P4, P16, Pc may be significant biomarkers of poor prognosis. Therefore, their further molecular identification may provide useful information to the direction of developing markers for stomach cancer.

References

Aoki T, Takasaki T, Morikawa J, Yano T, Watabe H (1994) Electrophoretic analysis of a gastric cancer-associated acid proteinase using a highly sensitive detection system. Biol Pharm Bull 17, 1358-6

Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248-54.

Chung YM, Yoo YD, Park JK, Kim YT, Kim HJ (2001) Increased expression of peroxiredoxin II confers resistance to cisplatin. Anticancer Res 21, 1129-33.

Fiocca R, Luinetti O, Villani L, Mastracci L, Quilici P, Grillo F, Ranzani GN (2001) Molecular mechanisms involved in the pathogenesis of gastric carcinoma: interactions between genetic alterations, cellular phenotype and cancer histotype. Hepatogastroenterology 48, 1523-30.

Ha GH, Lee SU, Kang DG, Ha NY, Kim SH, Kim J, Bae JM, Kim JW, Lee CW (2002) Proteome analysis of human stomach tissue: separation of soluble proteins by two-dimensional polyacrylamide gel electrophoresis and identification by mass spectrometry. Electrophoresis 23, 2513-24

Hochstrasser DF, Harrington MG, Hochstrasser AC, Miller MJ, Merril CR (1988) Methods for increasing the resolution of two-dimensional protein electrophoresis. Anal Biochem 173, 424-35.

Isoda N, Kajii E, Ikemoto S, Kimura K (1990) Two-dimensional polyacrylamide gel electrophoretic pattern of duodenal tumour proteins. J Chromatogr 534, 47-55.

Kennedy JA, Kirk SJ, McCrory DC, Halliday MI, Barclay GR, Rowlands BJ (1994) Modulation of immune function and weight loss by L-arginine in obstructive jaundice in the rat. Br J Surg 81, 1199-201

Kim JP, Jung SE (1987) Patients with gastric cancer and their prognosis in accordance with number of lymph node metastases. Scand J Gastroenterol 22, 33-5.

Kim JP, Lee JH, Kim SJ, Yu HJ, Yang HK (1998) Clinicopathologic characteristics and prognostic factors in 10 783 patients with gastric cancer. Gastric Cancer 1, 125-133.

Nagase T, Sugiyama T, Kawata S, Tarui S, Deutsch HF, Taniguchi N (1991) Analyses of polypeptides in the liver of a novel mutant (LEC rats) to hereditary hepatitis and hepatoma by two-dimensional gel electrophoresis: identification of P29/6.8 as carbonic anhydrase III and triosephosphate isomerase. Comp Biochem Physiol B 99, 193-201.

Nalty TJ, Taylor CW, Yeoman LC (1988) Variation in cytosolic protein expression between human colon tumors that differ with regard to differentiation class. Clin Chem 34, 71-5.

Narayan RK, Heydorn WE, Creed GJ, Jacobowitz DM (1986) Protein patterns in various malignant human brain tumors by two-dimensional gel electrophoresis. Cancer Res 46, 4685-94.

O'Farrell PH (1975) High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250, 4007-21.

Okuda K (1989) Biological and immunological studies on human tumor-associated cellular proteins detected by two-dimensional gel electrophoresis. Nippon Geka Gakkai Zasshi 90, 1697-705.

Ramagli LS, Rodriguez LV (1985) Quantitation of microgram amounts of protein in two-dimentional polyacrylamide gel electrophoresis sample buffer. Electrophoresis 6, 559-563.

Roder DM (2002) The epidemiology of gastric cancer. Gastric Cancer 5 (suppl 1), 5-11.

Ryu JW, Kim HJ, Lee YS, Myong NH, Hwang CH, Lee GS, Yom HC (2003) The proteomics approach to find biomarkers in gastric cancer. J Korean Med Sci 18, 505-9.

Sinha P, Hutter G, Kottgen E, Dietel M, Schadendorf D, Lage H (1998) Increased expression of annexin I and thioredoxin detected by two-dimensional gel electrophoresis of drug resistant human stomach cancer cells. J Biochem Biophys Methods 37, 105-16.

Tracy RP, Wold LE, Currie RM, Young DS (1982) Patterns for normal colon mucosa and colon adenocarcinoma compared by two-dimensional gel electrophoresis. Clin Chem 28, 915-9

Wilson DL, Hall ME, Stone GC, Rubin RW (1977) Some improvements in two-dimensional gel electrophoresis of proteins. Protein mapping of eukaryotic tissue extracts. Anal Biochemistry 83, 33-44.