Quantitative RT-PCR analysis of the expression of the EphB4 gene has shown that its expression is increased in 82% of tumour samples when compared with the matched normal tissue from the same patient. Using immunohistochemistry and Western analysis techniques with an EphB4-specific antibody, we also show that this receptor is expressed in the epithelial cells of the tumour tissue and either not at all, or in only low levels, in the normal tissue.

Primers EphB4F1 and EphB4R1 were designed to amplify a 466 bp product corresponding to bases 1377 – 1843 of the EphB4 cDNA sequence. Comparison with the EphB4 genomic sequence (Accession AY056047), shows that these primers span 4 introns and approximately 6 kb, allowing distinction of the cDNA and genomic products (a genomic product will not be synthesised under the conditions used in the RT-PCR analysis). However, when these primers were used in a PCR reaction using genomic DNA from six different patients as the template, bands of a similar size to the cDNA product were amplified (result not shown). The product from one of these patients was sequenced fully in both directions and it did not differ significantly from the EphB4 cDNA sequence. Because this suggested the presence of a pseudogene, further primers (EphB4F3 and EphB4R3) corresponding to bases 1819 – 3005 of the EphB4 cDNA sequence were designed to amplify a larger cDNA product of 1187 bp. As these did not amplify a product from the genomic template (primers span >10 kb), they were used for the RT-PCR analysis described later.

The expression of EphB4 was assessed for its suitability as a tumour specific marker using RT-PCR. RNA was extracted from ten colon cancer cell lines, reverse transcribed into cDNA then PCR-amplified using primers specific for the different genes. A primer set specific to CK19, a well characterised marker of colon epithelial cells [25-27], was used as a control for confirming integrity of the RT reaction. Transcripts corresponding to EphB4 and CK19 were amplified from all of the 10 cell lines (Figure ​(Figure11).

To determine if the increased expression of the EphB4 gene corresponded to an increase in the EphB4 protein we performed western analysis of colon and matched normal tissues using an EphB4-specific polyclonal antibody. The predicted size of wildtype EphB4 protein has been estimated by Bennet et al., (1994) to be 120 kD and in our experiment a band of this size was present in protein extracts from the tumour samples but low or absent in extracts from the normal mucosa (Figure ​(Figure4)4) [28]. This confirms that the increased expression in the tumour tissues does correspond to an increase amount of the protein itself. Additional signals corresponding to proteins of lower molecular weight were also visible specifically in the tumour samples. It is unknown at this time whether these are EphB4-specific cleavage products or not and this will be the subject of future investigation. Further samples will be need to be tested to determine if there is any association between the stage of tumour development and amount of protein present.

Colon cancer cell lines SW480, SW620, SW48, LIM1215, LIM2405, HT29, LIM1863, LIM2412 and LIM2099 were cultured in HEPES-buffered RPMI 1640 medium (Life Technologies, Grand Island, N.Y.), pH 7.4. The medium was supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, 160 μg/ml L-glutamine and 10% heat-inactivated foetal bovine serum (JRH Biosciences, Lenexa, KS) in 75 cm² vented tissue culture flasks at 37°C in a 5% CO₂ environment. Cells were collected at >90% confluency by trypsin digestion and centrifugation for 5 min at 1000 rpm, resuspended in phosphate buffered saline (PBS) and counted using a haemocytometer. The LIM cell lines were kindly provided by Dr R. Whitehead (Ludwig Institute for Cancer Research, Melbourne, Australia); the remainder were purchased from the American Type Culture Collection (ATCC, Rockville, MD).

RNA and protein were isolated sequentially from tissue samples and cell lines using Trizol™ solution (Sigma, St. Louis, MO). Growth medium was aspirated from the flask of growing cells (>90% confluency) and the cells washed with phosphate-buffered saline (PBS) before being directly lysed using 1 ml of the Trizol™ solution. After a five min incubation with gentle rocking, the cell lysate was removed to a 2 ml eppendorf tube. Tissue samples (0.5 – 1.0 g) were ground in liquid nitrogen using a mortar and pestle, then transferred to an eppendorf tube containing 1 ml Trizol™ solution. The RNA and protein fractions were extracted from each of these samples using the manufacturer's recommendations. Protein concentrations were determined using the DC Protein Assay Kit from Biorad (Sydney, NSW, Australia) following the manufacturer's protocol and using bovine serum albumin diluted from 0.2 mg/ml to 1.5 mg/ml to determine the standard curve.

Two duplicate (from the same manufacturing batch) Atlas Cancer cDNA arrays (Clontech, Palo Alto, CA) were differentially screened with cDNA probes made using mRNA extracted from tumour and matched normal tissue from two different patients and the supplied control mRNA. Total RNA (25 μg) was DNase treated (Life Technologies, Bethesda, MD) then re-extracted using Trizol™ solution. mRNA was prepared from these samples using the Oligotex mRNA purification kit (Qiagen, Germany) following the manufacturer's protocol. cDNA probes were synthesised using the method supplied with the cDNA array filters. These ³²P-labelled probes were hybridised to different arrays in 15 ml ExpressHyb™ solution (Clontech) overnight at 68°C, then washed at high stringency in the recommended wash solutions. Blots were exposed to phosphorimager plates overnight, then visualised using a Fuji-BAS phosphorimager. The blots were re-hybridised with a new probe after removal of the previous probe by incubation in boiling 0.1% SDS for 10 min and confirmation by phosphorimaging that there was no residual signal retained.

Total RNA (2 μg) was reverse transcribed at 37°C using 3 μl pD(N)₆ primers (Life Technologies), 200 μM each deoxyribonucleoside triphosphate (dNTP) (Pharmacia, Uppsala, Sweden) and 200 U MMLV reverse transcriptase (Life Technologies) in a reaction volume of 30 μl. Primers specific to either EphB4 (EphB4F1 5'- GAG AGG TAC CTC CTG CAG TGT C and EphB4R1 5'- CCA TGT CCG ATG AGA TAC TGT CCG-3' or EphB4 F2 5'-CGG ACA GTA TCT CAT CGG ACA TG-3' and EphB4 R2 5'-GCT TGG CCT GGG ACT TCA TGT G-3') or CK19 (CK19 F 5'-GAC TAC AGC CAC TAC TAC ACG ACC and CK19 R 5'-AGC CGC GAC TTG ATG TCC ATG AGC C), were used in a PCR reaction carried out in a Perkin Elmer thermocycler with the following conditions: 1.5 mM MgCl₂, 200 μM each dNTP, 50 ng of each primer, and 0.5 units of Tth polymerase in 1 × PCR buffer (10 × PCR buffer is 670 mM Tris-HCl pH 8.8, 166 mM [NH₄]₂SO₄, 4.5% Triton X-100, 2 mg/mL Gelatin) (Biotech International, Perth, Australia). Cycling conditions included an initial denaturation at 94°C for 5 min, followed by 30 cycles of 94°C for 30 sec, 60°C (EphB4 set 1) or 63°C (EphB4 set 2) or 68°C (CK19) for 30 sec, and 72°C for 30 sec, with a final extension of 72°C for 7 min. The QuantumRNA™ 18S Internal Standards kit from Ambion (Austin, TX) was used to calculate the relative abundance of transcripts corresponding to EphB4. One microlitre of a 3:7 ratio of 18S primer to 18S competimer (product of 489 bp) was added to each reaction tube and the annealing temperature increased to 62°C as recommended in the Ambion protocol. PCR products were separated by electrophoresis through 1.5% agarose gels and the results analysed using the Kodak Digital Science ID 2.0.2 program (Eastman Kodak Company, New Haven, CT, USA). The level of EphB4 expression in each sample was calculated relative to the amount of 18S product within that sample (expressed as a ratio). The relative expression of tumour compared to normal was compared using the GraphPad Instat™ program (http://www.graphpad.com/instat3/instat.htm).

Protein samples (20 μg) extracted from tumour and matched normal tissue were analyzed in duplicate using 6% SDS-PAGE gels [24]. One of these gels was stained with coomassie blue to visualise the proteins and compare loading of samples and the second was used for transfer to Immobilon-P PVDF membrane (Millipore Corporation, Bedford, MA). EphB4 antigens were detected using a 1:1000 dilution of the rabbit polyclonal antibody EphB4 (H-200) (Santa Cruz Biotechnology, Santa Cruz, CA) raised to human EphB4. After an overnight incubation at 4°C, the primary antibody was detected using the Lumi-Light^PLUS Western Blotting kit (Roche Biochemicals, Mannhein, Germany) following the manufacturer's recommendation, and between 10s and 30s exposure to Hyperfilm™ ECL™ (Amersham Pharmacia Biotech, UK).