Breast cancer tissue (n=115) and normal tissue samples (n=30) were collected immediately after surgery and stored at −80°C. The clinical follow-up was routinely performed after surgery, and details were stored in a database. The median follow-up period was 120 months. The presence of tumour cells in the collected tissues was verified by a consultant pathologist, who examined H&E stained frozen sections. Patient clinical data is shown in Table I.

Frozen paired tissues from the cohort were cut into 6-μm sections, mounted, air dried and fixed in a mixture of 50% acetone, 50% methanol for 15 min. These sections were subsequently rehydrated in Optimax wash buffer (Sigma-Aldrich, Dorset, UK) before being blocked for 20 min in a wash buffer containing horse serum. The sections were then incubated for 1 h with an anti-DR3 primary antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) at a 1:100 concentration. The sections were subjected to extensive washes before being incubated with a biotinylated secondary antibody for 30 min. This was subsequently removed through further washes before adding Avidin Biotin Complex (ABC; Vector Laboratories, Inc., Nottingham, UK) for 30 min before further washes. The diaminobenzidine (DAB) chromagen (Vector Laboratories, Inc.) was added to the sections and incubated in the dark for 5 min before counterstaining in Mayer's haematoxylin, dehydrating in ascending grades of methanol, clearing in xylene and mounting. Sections were subsequently observed under the microscope before capturing representative images.

Quantitative polymerase chain reaction, utilising the Amplifluor technology, was used to detect transcript expression of DR3 in cDNA prepared from the cohort samples. This methodology has been previously reported (13). In brief, primer pairs were designed using the Beacon Designer programme, with one of the primers containing a Z sequence tag (Table II). This Z sequence is complementary to the Universal Z probe (Intergen, Inc., Oxford, UK) included in the reaction (10 pmol), alongside Hot-start Q-Master mix (Advanced Biotechnologies, Ltd., Surrey, UK), 10 pmol of specific forward primer, 1 pmol specific reverse primer containing Z sequence and cDNA from ~50 ng cohort tissue RNA. The reaction was prepared and carried out using an iCyclerIQ (Bio-Rad, Hemel Hempstead, UK) under the following conditions; 94°C for 12 min followed by 60 cycles of 94°C for 15 sec, 55°C for 40 sec and 72°C for 40 sec. Transcript expression was determined using an internal standard, amplified in conjunction with the test samples and samples were further normalised through calculation and standardisation in conjunction with cytokeratin-19 sample levels.

The MCF7 and MDA-MB-231 human breast cancer cell lines were obtained from the American Type Culture Collection (ATCC; Rockville, MD, USA) and cultured in DMEM/Ham's F12 with L-Glutamine medium (PAA Laboratories, Somerset, UK), supplemented with streptomycin, penicillin and 10% fetal calf serum (PAA Laboratories), and incubated at 37°C, 5% CO₂ and 95% humidity.

Anti-human DR3 hammerhead ribozyme transgenes were designed based on the secondary structure of DR3 mRNA generated using the Zuker RNA mFold programme (14). The ribozymes were subsequently synthesized and cloned into a pEF6/V5-His-TOPO plasmid vector (Invitrogen, Paisley, UK). Both DR3 ribozyme transgenes and empty pEF6 control plasmids were transfected, through electroporation with an Easyject Plus electroporator (EquiBio, Kent, UK), into MCF7 and MDA-MB-231 breast cancer cells, as previously reported (15–17). Following transfection, cells were subjected to a selective period through the addition of 5 μg/ml blasticidin to the culture medium and, following this period, were transferred into a maintenance medium (containing 0.5 μg/ml blasticidin). Cells were routinely tested to confirm stable transfection and knockdown of DR3 expression. MCF7 cells containing the ribozyme transgene or control pEF6 plasmid were designated MCF7^DR3KO and MCF7^pEF6 respectively. Similarly, MDA-MB-231 cells containing the ribozyme transgene or the control pEF6 plasmid were labelled MDA-MB-231^DR3KO and MDA-MB-231^pEF6 respectively. The unaltered wild-type cells were labelled MCF7^WT or MDA-MB-231^WT.

Cells were grown to confluence in a 25 cm² tissue culture flask before being subjected to RNA extraction using TRI Reagent (Sigma-Aldrich) in accordance with the supplied protocol. Following extraction, the RNA was quantified using a spectrophotometer (WPA UV 1101; Biotech Photometer, Cambridge, UK) and RNA levels throughout the samples were standardised to 250 ng for use in reverse transcription. An enhanced avian reverse transcriptase-PCR-100 kit with anchored oligo(dT) primers was used to carry out reverse transcription in accordance with the supplied protocol (Sigma-Aldrich). The quality of the cDNA generated was tested using GAPDH primers before probing for the expression of DR3 using specific primers. PCR primers were designed using Beacon Designer (Palo Alto, CA, USA) and synthesized by Invitrogen. Full primer sequences are outlined in Table II. PCR was performed in a T-Cy Thermocycler (Creacon Technologies, Ltd., CD Emmen, The Netherlands) using REDTaq^® ReadyMix™ PCR Reaction Mix (Sigma-Aldrich). PCR was conducted under the following reaction conditions: initial denaturing at 94°C for 5 min, 34–36 cycles of denaturing at 94°C for 40 sec, annealing at 55°C for 40 sec and extension at 72°C for 1 min and final extension at 72°C for 10 min. Following PCR, products were loaded on a 0.8% agarose gel, separated electrophoretically, stained in ethidium bromide and visualised under ultraviolet light.

Cancer cells were lysed with an SDS lysis buffer on a rotor wheel before being spun at 13,000 × g to remove insolubles. Following lysis, protein levels were quantified using the Bio-Rad DC Protein Assay kit (Bio-Rad Laboratories, Hercules, CA, USA), standardised to 2 mg/ml and diluted with Laemmli 2× concentrate sample buffer (Sigma-Aldrich). Samples were boiled for 5 min before loading onto a 12% acrylamide gel and being subjected to electrophoretic separation. Once sufficient separation had occurred the proteins were blotted onto a Hybond-C extra nitrocellulose membrane (Amersham Biosciences UK, Ltd., Buckinghamshire, UK), blocked in 10% milk and subjected to specific antibody probing. GAPDH was used as an internal control. Proteins were probed using the respective primary antibodies at a concentration of 1:250 and specific peroxidase conjugated secondary antibodies at a concentration of 1:1,000. Monoclonal rabbit anti-DR3 (SC-7909) and anti-GAPDH (SC-32233) were obtained from Santa Cruz Biotechnology, Inc. Peroxidase-conjugated anti-mouse and rabbit IgG secondary antibodies were purchased from Sigma-Aldrich. Protein bands were visualized using a Supersignal™ West Dura system and documented using a gel documentation system (UVItec, Cambridge, UK).

Twenty thousand cells/well were seeded in glass chamber slides (Lab-Tek; Fisher Scientific UK, Ltd., Loughborough, UK). Following overnight incubation, cells were fixed and then permeabilized with 0.1% Triton X-100 for 5 min in TBS. Following a blocking with SuperSensitive™ Wash Buffer (BioGenex, CA, USA) containing horse serum, the primary antibody was added at a concentration of 1:100 and incubated for 60 min. After extensive washing, the peroxidase conjugated specific secondary antibody was added at a concentration of 1:100. This step was followed by further washes before staining the sections using a DAB kit (Vector Laboratories), counterstaining in Mayer's haematoxylin for 2 min, rinsing in tap water and mounting. Slides were subsequently visualised under the microscope and representative images were captured.

An in vitro growth assay was used to examine the growth rates of MCF7 and MDA-MB-231 control cells and those transfected with the ribozyme transgenes. Cells were seeded, at a density of 3,000 cells/well, into 96-well plates before being incubated for 4 h, 1, 3 or 5 days. Plates were then fixed in 4% formaldehyde (v/v) and stained with 0.5% (w/v) crystal violet. Crystal violet stain was subsequently extracted using 10% acetic acid (v/v) and cell density was determined using an ELx800 spectrophotometer (Bio-Tek Instruments, Inc., Winooski, VT, USA) to measure the absorbance at 540 nm. Percentage increase in cell density, compared to the 4 h reference plate, was used as a measure of cell growth.

An in vitro Matrigel invasion assay was used to assess the invasiveness of control and transfected breast cancer cells. This assay has previously been described by our group (18). Briefly, 15,000 control or transfected cells were seeded into transwell inserts (Becton-Dickinson Labware, Franklin Lakes, NJ, USA), containing 8.0 μm pores, which had previously been coated with 50 μg/insert of Matrigel Matrix Basement Membrane (BD Biosciences, Oxford, UK) and incubated for 3 days. Following this incubation period, cells that had invaded through the Matrigel membrane and migrated to the underside of the transwell insert were fixed in 4% formaldehyde (v/v) and stained with 0.5% (w/v) crystal violet. The number of invaded cells in several representative fields was quantified by counting cells/field under ×20 objective magnification.

The cell-matrix adhesive potential of control and transfected cells was examined using an in vitro Matrigel adhesion assay. This assay was adapted from a previously reported method (19). In brief, 45,000 control or transfected cells/well were seeded into a 96-well plate that had been pre-coated with 5 μg of Matrigel artificial basement membrane. This plate was subsequently incubated for 45 min before being subjected to vigorous washing in BSS. Following washings, any remaining adherent cells were then fixed in 4% formaldehyde (v/v) and stained with 0.5% (w/v) crystal violet. Adherent cells were counted in several random fields under ×20 objective magnification and quantified to determine adhesive capacity.

Control and transfected cell migration rates were examined using an in vitro migration/wounding assay. This technique was modified from a previously described method reported by our group (20). Cell migration rates were calculated based on the capacity to close an artificial wound in the monolayer. In brief, cells were cultured in a 24-well plate to form a monolayer; this monolayer was subsequently wounded through scratching with a blunted needle. The closure of this wound, through the migration of cells at the wound fronts, was subsequently tracked and recorded over 90 min using a time-lapsed video system (Panasonic UK, Ltd., London, UK). Cellular migration was then calculated at set time points, in reference to the initial time point, using the ImageJ analysis software.

Experimental procedures were repeated independently a minimum of 3 times. Statistical comparisons were made between cells transfected with the DR3 ribozyme transgenes (MCF7^DR3KO and MDA-MB-231^DR3KO) and the respective control cell line containing the empty pEF6 plasmid (MCF7^pEF6 and MDA-MB-231^pEF6). Data was statistically analysed using a two-sample, two-tailed t-test or two way ANOVA and the Minitab 14 software package. A p-value <0.05 was taken to indicate statistically significant differences. Patient survival was analysed using the Kaplan-Meier method (SPAW18).

DR3 transcript levels were examined in the breast cohort using quantitative PCR. A decreased level of DR3 expression was revealed in breast tumours (n=115) (854±473 copies/50 ng RNA), compared to normal tissues (n=30) (24975±12176 copies/50 ng RNA, p=0.057) (Fig. 1A). Immunohistochemical staining (IHC) showed that DR3 staining in normal mammary epithelial cells (Fig. 1B) is largely nuclear staining. Fig. 1B confirmed reduced levels of DR3 staining in tumour tissues (shown are grade 1 and 3 tumours) compared to normal tissues (Fig. 1B).

Further analysis of DR3 transcript levels against the clinical aspect demonstrated that decreased DR3 expression was correlated with poor disease prognosis over the 10 year follow-up period (Fig. 2A). DR3 levels were found to be elevated in patients who remained disease free (1165±669 copies/50 ng RNA), when compared to those with metastasis (16.1±16 copies/50 ng RNA, p=0.090), with local recurrence (2.69±2.1 copies/50 ng RNA, p=0.086) and those who succumbed to breast cancer (2.27±1.2 copies/50 ng RNA, p=0.086) with close to significant levels being reached. A similar trend was observed in respect to Nottingham prognostic index (NPI) (Fig. 2C) with DR3 expression levels being higher in NPI-1 (1497±915 copies/50 ng RNA) compared to NPI-2 (300±292 copies/50 ng RNA, p=0.22) and NPI-3 (0.110±0.072 copies/50 ng RNA, p=0.11). By contrast, DR3 expression levels were found to be elevated in the poorly differentiated, higher grade 2 (264±190 copies/50 ng RNA, n=39, p=0.17) and grade 3 (1629±993 copies/50 ng RNA, n=54, p=0.11) tumours when compared to grade 1 tumours (0.026±0.026 copies/50 ng RNA, n=20) (Fig. 2B). No significant correlations were observed between DR3 expression and TNM classification, although again the expression levels of DR3 tended to be lower in the higher TNM classification tumours (Fig. 2D).

Low levels of DR3 transcripts were associated with a significantly shorter overall survival compared with those who had tumours with high levels [127.8 (115.1–140.5) months vs. 139.2 (130–148.4) months, p=0.038] (Fig. 2E). There was a trend of decreased disease free survival for patients with low DR3 transcript, although this failed to reach statistical significance (p=0.095) (Fig. 2F).

The expression of DR3 was examined in MCF7 and MDA-MB-231 cells using RT-PCR (Fig. 3A). MCF7 and MDA-MB-231 cells were found to express modest levels of the DR3 transcript and were subsequently transfected with anti-DR3 ribozyme transgenes. RT-PCR demonstrated that DR3 mRNA expression was successfully knocked down in MCF7^DR3KO and MDA-MB-231^DR3KO cells by the ribozyme transgene in comparison to the level of expression in wild-type cells (MCF7^WT and MDA-MB-231^WT) and in empty plasmid control cells (MCF7^pEF6 and MDA-MB-231^pEF6) (Fig. 3B and C).

Additionally, western blotting was used to probe for DR3 protein levels in both the control and transfected cell lines. Similar to the trends seen at the mRNA level, DR3 protein was found to be expressed in all of the control cell lines (MCF7^WT, MDA-MB-231^WT, MCF7^pEF6 and MDA-MB-231^pEF6), and the expression of DR3 protein exhibited a dramatic reduction in the ribozyme transgene transfected cell lines (MCF7^DR3KO and MDA-MB-231^DR3KO) (Fig. 3D and E). Immunocytochemical (ICC) staining was similarly used to detect protein levels of DR3 in these cells. ICC staining indicated a decreased protein level of DR3 in cells transfected with the ribozyme transgene (Fig. 3F and G).

The effect of DR3 on the growth of MCF7 and MDA-MB-231 cells was examined using an in vitro tumour cell growth assay. Transfection with the DR3 ribozyme transgene did not appear to impact the cell growth rate of either MCF7 or MDA-MB-231 cells. Knockdown of DR3 resulted in no significant difference in growth rate between the MCF7^DR3KO and MCF7^pEF6 cell lines over the 1-, 3- and 5-day incubation periods (Day 1, 67.3±38.9 vs. 55.99±31.75, p=0.536; Day 3, 366.0±37.6 vs. 381.2±91.6, p=0.86; Day 5, 645.0±57.5 vs. 728.4±92.6, p=0.486) (Fig. 4A). A similar trend was seen between the MDA-MB-231^DR3KO and MDA-MB-231^pEF6 cell lines over the 1-, 3- and 5-day incubation period (Day 1, 136.0±81.8 vs. 153.8±97.4, p=0.980; Day 3, 306.8±78.9 vs. 339.8±96.9, p=0.798; Day 5, 1131.4±220.2 vs. 958.5±223.5, p=0.596) (Fig. 4B).

The capacity of MCF7 and MDA-MB-231 breast cancer cells to adhere to an artificial Matrigel basement membrane was examined using an in vitro Matrigel adhesion assay. Transfection with the DR3 ribozyme transgene did not affect the adhesive properties of either MCF7 or MDA-MB-231 cells to an artificial Matrigel basement membrane over a 40-min incubation period (Fig. 4C and D). No significant difference in adhesive capacity was seen between cells containing the ribozyme transgene and their respective pEF6 control cells (MCF7^DR3KO vs. MCF7^pEF6, 91.0±7.7 vs. 98.8±18.3, p=0.714; MDA-MB-231^DR3KO vs. MDA-MB-231^pEF6, 118.6±10.1 vs. 132.5±21.4, p=0.637).

No significant difference in cellular invasion was found between the MDA-MB-231^DR3KO and MDA-MB-231^pEF6 cell lines over the 3-day incubation period (128.85±1.98 vs. 112.15±6.43, respectively, p=0.111) (Fig. 5B). Similarly, no significant differences were seen between the invasive capacity of MCF7^DR3KO and the control MCF7^pEF6 cell lines (110.1±15.0 vs. 120.1±2.93, respectively, p=0.688) (Fig. 5A).

The effect of DR3 expression on cell migration was assessed using a migration/wounding assay. Knockdown of DR3 negatively impacted the migration of MCF7 cells (Fig. 5C). The rate of cell migration in MCF7^DR3KO was significantly lower than that in MCF7 cells containing the empty pEF6 control plasmid after 90 min (p=0.011). Similarly, migratory rates of MDA-MB-231^DR3KO tended to be reduced compared to the pEF6 control cell line after 90 min, although this did not reach statistical significance (p=0.065) (Fig. 5D).