A profile of differentially expressed genes in primary colorectal cancer using suppression subtractive hybridization.

As a step towards understanding the complex differences between normal cells and cancer cells, we have used suppression subtractive hybridization (SSH) to generate a profile of genes overexpressed in primary colorectal cancer (CRC). From a 35¿ omitted¿000 clone SSH-cDNA repertoire, we have screened 400 random clones by reverse Northern blotting, of which 45 clones were scored as overexpressed in tumor compared to matched normal mucosa. Sequencing showed 37 different genes and of these, 16 genes corresponded to known genes in the public databases. Twelve genes, including Smad5 and Fls353, have previously been shown to be overexpressed in CRC. A series of known genes which have not previously been reported to be overexpressed in cancer were also recovered: Hsc70, PBEF, ribophorin II and Ese-3B. The remaining 21 genes have as yet no functional annotation. These results show that SSH in conjunction with high throughput screening provides a very efficient means to produce a broad profile of genes differentially expressed in cancer. Some of the genes identified may provide novel points of therapeutic intervention.

Phage display of cDNA repertoires: the pVI display system and its applications for the selection of immunogenic ligands.

The selection of phage displayed cDNA repertoires on an immobilised target has been reported to be an efficient way to rapidly identify interacting partners. To date, however, only a few successful applications have been reported. Here, we present a review of the current status of the display and selection of cDNA libraries using phage. As an example, we report the construction of a set of phage display vectors suitable for cDNA display based on fusion to the minor bacteriophage coat protein 6 (pVI) of filamentous phage. We have evaluated these vectors through the display of the C(H)3 domain of human IgG and of the Escherichia coli alkaline phosphatase (PhoA) gene. Both the C(H)3 domain of IgG and PhoA are shown to be displayed on pVI, and PhoA is also shown to be enzymatically active. We have constructed primary colorectal tumor cDNA repertoires in these vectors and performed selections on both a monoclonal antibody to beta2 microglobulin (beta2M) and polyclonal antibody sera to human IgG. In both cases, relevant ligands were recovered from the phage displayed cDNA repertoire. These vectors may be used for selection of phage displayed cDNA libraries with polyclonal sera from patients. This will allow the identifying antigenic cDNA products in such diseases as cancer, viral/bacterial infections or autoimmune disease. Furthermore, by selections with other specific biomolecules, this display system may aid the identification of interacting partners in functional genomics.

A detailed analysis of K-ras point mutations in relation to tumor progression and survival in colorectal cancer patients.

Point mutations in codon 12, 13, and 61 of the K-ras gene are an early event in tumorigenesis of colorectal cancer, but the impact of number, type, and position of such mutations on the progression of adenomas as well as the clinical behaviour of colorectal carcinomas is not clearly established. A series of 35 adenomas and 117 carcinomas at various stages was subjected to single-strand conformation polymorphism (SSCP) to analyse type, position and number of exon-I K-ras point mutations and to relate the results with patients survival. From our data we conclude that the number of K-ras point mutated tumors shows a trend to increase with tumor progression. The number of multiple K-ras point mutations, however, significantly increases with stage. Most mutations occur in the 1st or 2nd base of codon 12, whereas point mutations in the 3rd base are rare. In adenomas mutations, particularly G-T transversions, in the K-ras gene could indicate a propensity to malignant transformation. G-A transitions and G-C transversions of the second base are associated with metastasized tumors. Regarding survival, patients with K-ras point mutated tumors did worse than their non-mutated counterparts. G-A transitions in the 1st and 2nd base and G-C transversions in the 2nd base were associated with a poor prognosis as compared with G-T transversions in both the 1st and 2nd base. Patient survival therefore is related to the occurrence and type, but not the location, of K-ras point mutations.

K-ras point mutations in routinely processed tissues: non-radioactive screening by single strand conformational polymorphism analysis.

AIMS: To develop a non-radioactive method to screen routinely fixed, paraffin wax embedded specimens for the occurrence of point mutations; to evaluate the single strand conformational polymorphism (SSCP) analysis technique for the detection of K-ras point mutations as a result of electrophoretic mobility shifts. METHODS: DNA was extracted from archival specimens of colon cancer and from established colon cancer cell lines with known point mutations. A K-ras gene fragment containing codons 12 and 13 of exon 1 was amplified with the polymerase chain reaction (PCR). Denatured DNA fragments were run on 10% polyacrylamide gels under non-denaturing conditions. After electrophoresis DNA was blotted and the single stranded DNA was detected using a digoxigenin labelled ras probe. The nature of the detected point mutations was identified and confirmed by sequencing and hybridisation with oligonucleotides using 32P labelling. RESULTS: Wild type and aberrant alleles were detected caused by mobility shifts after electrophoresis of the PCR products. Commonly occurring mutations in the K-ras gene--in the first two positions of codon 12--could easily be detected in DNA from archival paraffin wax embedded colon cancer tissue. In all the colon tumour samples studied wild type gene alleles were also found, presumably derived from normal cells in the specimen. CONCLUSIONS: The SSCP method permits rapid non-radioactive screening of adenomas or carcinomas for the occurrence of point mutations in the K-ras gene. But if a mutation is detected by an electrophoretic mobility shift, its identification requires confirmation by sequencing or oligonucleotide hybridisation.

Human Caco-2 cells transfected with c-Ha-Ras as a model for endocrine differentiation in the large intestine.

Endocrine cells occur in approximately 30% of all colorectal adenocarcinomas, and this feature appears to correlate with a relatively poor prognosis. To study the factors regulating endocrine differentiation in colorectal cancer, which may bear resemblance to the regulation of endocrine differentiation in normal intestinal mucosa, models in which differentiation can be manipulated are essential. However, endocrine features in colorectal cancer cell lines are scarce and are almost exclusively observed in xenografts, presumably as a result of differentiation induction by stromal components. We attempted to demonstrate endocrine differentiation in the colonic adenocarcinoma cell line Caco-2, which is frequently used as a model for enterocytic differentiation. In vitro endocrine tumor cells were not encountered. In vivo studies were cumbersome, because of the low take rate of Caco-2 cells. We did manage to establish nude mouse xenografts of Caco-2 cells by inoculating cells in collagen gel and by suppressing natural killer cell activity. In an attempt to induce a better take rate and to investigate the effect of Ras oncoprotein overexpression on endocrine differentiation, Caco-2 cells were transfected with a point-mutated c-Ha-Ras gene. The cell line Caco-2 EJ6, generated from these experiments, could be xenografted in nude mice with a high take rate, yielding a moderately well differentiated adenocarcinoma, morphologically identical to the tumors derived from untransfected Caco-2 cells. The xenografts displayed goblet cell, enterocytic, Paneth cell and endocrine differentiation. In vitro endocrine differentiation was observed neither under standard conditions nor with extracellular matrix components as differentiation inducers. We conclude that the Caco-2 cell line and its c-Ha-Ras transfected subline Caco-2 EJ6 in vivo display endocrine differentiation. Ras overexpression does not enhance endocrine differentiation. Due to its favorable growth properties in vivo, Caco-2 EJ6 is a suitable model for studies on endocrine differentiation in colorectal cancer.

Extracellular matrix components induce endocrine differentiation in vitro in NCI-H716 cells.

Endocrine cells occur in +/- 30% of colorectal adenocarcinomas. The significance of this phenomenon in terms of tumor behavior is still controversial. Endocrine differentiation in colorectal cancer cell lines is almost confined to tumor xenografts in vivo, suggesting that endocrine differentiation might be regulated by epithelial-stromal interactions. This hypothesis was studied in the cecal adenocarcinoma-derived cell line NCI-H716 by comparing the expression of chromogranin A protein and messenger RNA in vivo and in vitro and by attempts to induce differentiation in vitro. We found that chromogranin A expression, which was strongest in vivo, could be significantly enhanced in vitro by culturing tumor cells in the presence of native extracellular matrix, on fibroblast feeder layers, and in a defined medium with basic fibroblast growth factor. The results suggest that the extracellular matrix induces endocrine differentiation through factors (e.g., basic fibroblast-growth factor) that may be produced by stromal cells and after secretion bind to the extracellular matrix.

NCI-H716 cells as a model for endocrine differentiation in colorectal cancer.

In colonic neoplasms, endocrine differentiation is encountered not only in carcinoid tumors but also in adenocarcinomas, where endocrine cells may represent a distinct line of differentiation in the tumor. The significance of endocrine differentiation in colorectal cancer is not well established, partly because of the paucity of tumor cell lines which can serve as a model for studying endocrine differentiation. In this report we describe the properties of NCI-H716 cells, a cell line derived from a poorly differentiated adenocarcinoma of the caecum, under various in vitro conditions and as xenografts in athymic mice. Phenotypical properties were immunohistochemically assessed using a panel of differentiation related antibodies, and also by Northern blot analysis and by electron microscopy. Receptors for biogenic amines and peptide hormones were analyzed by ligand binding assay. These studies show that: 1. NCI-H716 cells can be undifferentiated, or show endocrine, mucin-producing or "amphicrine" properties. 2. Endocrine differentiation of NCI-H716 cells preferentially occurs in xenografts in athymic mice, which suggests that mesenchymal elements induce endocrine differentiation. 3. NCI-H716 cells express large amounts of high affinity receptors for gastrin, serotonin and somatostatin and these substances can regulate growth. Thus, NCI-H716 cells form a suitable model for the study of endocrine differentiation in intestinal epithelium and of auto- or paracrine growth regulation in intestinal neoplasia.

Southern and dot blot analysis of DNA from formalin-fixed, paraffin-embedded tissue samples from colonic carcinomas.

Two extraction methods for the isolation of DNA from formalin-fixed, paraffin-embedded tissue samples from colonic carcinomas were compared. The processed DNAs were compared with DNAs from fresh specimens of the same tumors. The two extraction methods gave similar results. Formalin-fixation and paraffin-embedding irreversibly denatured DNA and consequently decreased the extraction yield and interfered with the quantitative measurement of DNA. Southern blot and dot blot analysis of processed and native DNA was performed using a c-myc and an actin probe. The results show that for Southern analysis processed DNA can be used but, due to the generation of random breaks, the restriction fragments have to be small. Furthermore, the fixation-induced crosslinking of DNA appears to hamper hybridization. For these reasons processed DNA can be analyzed better by dot blot rather than Southern blot hybridization.

Colonic epithelium reactive monoclonal antibodies. Identification and immunohistochemical localization of the target epitopes.

We have produced a small library of colonic mucosa and colorectal carcinoma reactive monoclonal antibodies (MoAbs) by immunizations with extracts of human colon cancer tissue and a human colon cancer cell line. Hybridoma supernatants were tested on (normal and neoplastic) human tissues by immunoperoxidase methods to evaluate organ or tissue specificity. Initial biochemical characterization of the target antigens was performed by gelpermeation chromatography, Western blotting and competition assays. Based upon the immunoreactivity patterns and the characteristics of the antigen four groups of MoAbs could be distinguished. The first group concerns the antibodies PARLAM 3, 9 and 10. These antibodies react with an 87 kDa protein moiety in high molecular weight (2-5 x 10(6) Da) glycoproteins. In intestinal and colon mucosa these antibodies showed diffuse binding with goblet cells. In colon carcinoma decreased reactivity with these MoAbs was found. The second group consists of antibodies PARLAM 8, 12 and 13. These antibodies react with large (greater than 5 x 10(6) Da) glycoproteins, most likely with carbohydrate epitopes. By immunohistochemistry in normal colon mucosa the antibodies all show granular supranuclear reactivity with goblet cells. These antibodies show increased reactivity with colon adenomas and adenocarcinomas. A third group is formed by PARLAM 2, which also reacts with a large (greater than 5 x 10(6) Da) glycoprotein, showing a granular distribution in goblet cells. In colon carcinomas more extensive expression is found than in normal colonic mucosa. Finally, the fourth group consists of PARLAM 11, which also reacts with a large (greater than 5 x 10(6) Da) glycoprotein, located in the brush border of colonic columnar cells. These antibodies might be useful tools for the analysis of the expression of mucin related glycoproteins in normal, preneoplastic and neoplastic colon mucosa.

The establishment and characterization of two new cell lines derived from a single human colonic adenocarcinoma.

Two cell lines with different in vitro growth characteristics were established from a single mucinous colonic adenocarcinoma. Epithelial cells of the line 5583-E demonstrated anchorage-dependent growth while those of line 5583-S were anchorage-independent and grew as multicellular floating spheroids. Both cell lines shared common characteristics with respect to the expression of differentiation markers (secretory component, carcinoembryonic antigen), mucins and karyotype (trisomy 12 and 14, marker chromosome) but also showed consistent differences. In nude mice 5583-S cells formed moderately differentiated mucinous adenocarcinomas with high carcinoembryonic antigen and mucin production, whereas 5583-E xenografts were poorly differentiated and almost entirely failed to produce carcinoembryonic antigen and mucins. The plating efficiency of 5583-E cells appeared to be greater and doubling time shorter than those of 5583-S cells. Furthermore, 5583-E cells showed an extra isochromosome, 1q. The cell lines were genotypically and phenotypically stable over a period of 2 years. Our results reemphasize that multiple cell lines with heterogeneous phenotypic and genotypic characteristics can be obtained from a single primary tumor.

CEA-specificity of CEA-reactive monoclonal antibodies. Immunochemical and immunocytochemical studies.

Four carcinoembryonic antigen (CEA) reactive monoclonal antibodies Parlam 1, 4, 5 and 6 were studied in respect to reactivity with CEA and its cross-reacting antigens (NCA-1, NCA-2, BGP). In immunochemical studies (ELISA and SDS-PAGE followed by immunoblotting) Parlam 4 did not react with these crossreacting antigens, whereas Parlam 1, 5, and 6 demonstrated variable reactivity with these antigens. As expected, by immunocytochemistry Parlam 1, 5, and 6 stained bile canaliculi in the liver (due to crossreactivity with BGP), pneumocytes and splenic tissue (NCA-1) to an extent comparable with the results in biochemical tests. In contrast with the immunochemical observations, however, Parlam 4 showed slight but distinct reactivity with splenic granulocytes (NCA-1) and hepatic bile canaliculi (BGP). Relative epitope specificity of the monoclonal antibodies was tested in blocking experiments. These showed that Parlam 1 and 5 detect the same epitope on CEA as they block each others binding completely. In other combinations monoclonal antibodies block each others binding only partially, indicating that they detect different or partially overlapping epitopes. These results suggest that CEA specific epitopes may partly overlap with epitopes on crossreacting antigens. In this context, we propose that on crossreacting antigens epitopes exist that are structurally similar to epitopes on CEA or that some epitopes on CEA may consist of a spatial configuration which involves CEA specific as well as non-CEA specific structures. The antibody will show the highest affinity towards CEA, as this molecule contains the uniquely matching or complete epitope and lower affinity towards the crossreacting antigen with an imperfectly matching or only partially available epitope.