RAI3 antigen expression and purification
The RAI3 cDNA was cloned in pGEX2a as an in-frame fusion with an N-terminal GST tag and a C-terminal hexahistidine tag as described previously . The N-terminal fusion abolishes the toxicity observed when eukaryotic membrane proteins are expressed in E. coli by promoting the direction of protein to inclusion bodies which results in high-level production . The construct was introduced into Escherichia coli BL21, and 1 l of LB medium containing 100 mg/l ampicillin was inoculated with 10 ml of an overnight culture containing 50 mg/l carbenicillin and shaken at 37°C. When the culture reached an A600 of 0.80, protein expression was induced by adding 0.1 mM of isopropyl-β-D-thiogalactoside for 4 h. Cells were harvested by centrifugation, resuspended in 10 ml ice-cold 20 mM Tris/HCl pH 7.5 and frozen at -20°C. After thawing to 4°C and adding EDTA and dithiothreitol (DTT) each to a final concentration of 1 mM, the cell suspension was stirred for 10 min, disrupted by repeated passage through a high-pressure homogeniser (Emulsiflex C5, Avestin Inc. Canada), supplemented to a final concentration of 50 mM EDTA, 10 mg/ml Triton X-100, and stirred for a further 30 min. Insoluble protein was pelleted by centrifugation for 45 min at 48400 g. The pellet was washed in 20 mM sodium phosphate, 150 mM NaCl (pH 7.0) containing 1 mM DTT, centrifuged as above and finally resuspended in 10 ml of the same buffer. This suspension, referred to as "inclusion bodies", although it includes the bacterial membranes, was stored at -20°C.
Inclusion bodies were thawed on ice, pelleted by centrifugation for 30 min at 46500 g at 4°C, resuspended in 5 ml of 25 mM Tris/HCl pH 8.5, 150 mM NaCl and incubated at room temperature under reducing conditions (5 mM DTT) for 1 h. Ten volumes of 10 mg/ml lauroyl sarcosine (LS) in 25 mM Tris/HCl (pH 8.5), 250 mM NaCl were added and the sample was sonicated for 1 min using a tip sonicator on ice. After stirring at room temperature for 2 h and at 4°C for 1 h, the C-terminal GST tag was cleaved with 10 U thrombin per ml of inclusion bodies for 1 h at 4°C. To abort cleavage, we added 200 μM phenylmethylsulphonyl fluoride and an additional 10 mg/ml LS. The insoluble fraction was removed by centrifugation at 38500 g, 4°C. For RAI3 purification, the supernatant was incubated overnight at 4°C with 1.5 ml Ni-NTA Superflow (Qiagen, Hilden, Germany) per ml of inclusion bodies equilibrated in 20 mg/ml LS, 25 mM Tris/HCl (pH 8.5), 250 mM NaCl.
The Ni-matrix was packed into a column and washed at a flow rate of 2 cm/h with (i) three column volumes (CV) of 20 mM Tris/HCl (pH 8.5), 250 mM NaCl, 10 mM β-mercaptoethanol (βME), 10 mg/ml n-tetradecylphosphocholine (Fos14, Anatrace, Maumee, OH), (ii) 10 CV of 20 mM Tris/HCl (pH 8.5), 250 mM NaCl, 10 mM βME, 1 mg/ml Fos14, 0.2 mg/ml Folch Lipid Fraction I (Sigma) and (iii) one CV of 20 mM Tris/HCl (pH 7.5), 250 mM NaCl, 1 mM glutathione (GSH), 0.1 mg/ml Fos14. Protein was eluted in 20 mM Tris, 250 mM NaCl, 1 mM GSH, 300 mM imidazole, 0.1 mg/ml Fos14 adjusted to pH 7.5. Protein-containing fractions were pooled and DTT was added to 10 mM. Monomeric RAI3 was isolated by applying the concentrated pooled fraction to a Superdex 200 HiLoad XK 16/60 gel filtration column (GE Healthcare, Freiburg, Germany) run in 20 mM HEPES/NaOH (pH 7.0), 200 mM NaCl, 0.25 mg/ml Fos14, 1 mM DTT at 1 ml/min. Monomeric RAI3-containing fractions were pooled, and concentrated to 1 mg/ml.
RAI3 reconstitution into liposomes
Purified RAI3 was dialysed against 100 volumes of 100 mM boric acid/NaOH (pH 9.5), 0.1 mg/ml Fos14 at 4°C prior to reconstitution. A lipid blend was obtained by dissolving Folch Lipid Fraction I, phosphatidyl ethanolamine from sheep brain (Sigma) and cholesterol at a ratio of 40:32:28 in chloroform. After removal of the chloroform under vacuum over night the lipid was first resuspended in water at 10 mg/ml and then partially dissolved in CHAPS by incubating for 1 h at room temperature with a CHAPS to lipid ratio of 3:1 and a lipid concentration of 7.5 mg/ml. CHAPS and SDS at final concentrations of 5 and 0.1 mg/ml respectively were added to the dialysed RAI3. After 10 min, 70 mM KCl was added and the RAI3 solution was merged with 1.3 times the volume of the lipid/CHAPS mixture. This mixture was rotated end-over-end over night at 4°C. Detergent was then extracted by adding an equal volume of fresh Calbiosorb beads, which had been thoroughly washed with water and drained, followed by end-over-end rotation for 3 h at 4°C. Beads were removed using a fritted plastic column and proteoliposomes were harvested by centrifugation for 30 min at 100,000 g, 4°C. The resulting pellet was resuspended in 20 mM HEPES/NaOH (pH 7.0), 150 mM NaCl at a theoretical protein concentration of 1 mg/ml. Aliquots were frozen in liquid nitrogen and stored at -80°C. BALB/c mice were immunised using the RAI3 antigen reconstituted in liposomes. All additional in vitro experiments were also performed with the soluble protein prior to liposome reconstitution.
Production of monoclonal antibodies
Six-week-old female BALB/c mice were immunised subcutaneously with 100 μg RAI3 liposomes and 60 μl GERBU Adjuvant MM (GERBU Biochemicals, Gaiberg, Germany) as a priming injection following three booster injections with 50 μg RAI3 mixed with 40 μl adjuvant at 10-day intervals. Two days prior to fusion, a final booster injection was given comprising 50 μg RAI3 without adjuvant. On each day of immunisation a 10-μl blood sample was taken to monitor the specific antibody titre in the serum. The immune response was analysed by direct ELISA using plates coated with recombinant RAI3 protein (50 ng/well) and a HRP-conjugated anti-mouse IgG antibody (Sigma-Aldrich, Munich, Germany; 1:5,000) using a serial dilution of mouse serum.
On day 45, the spleen was dissected and spleen cells were fused with murine myeloma SP2-IL6 cells (ATCC: CRL-2016 ) at a ratio of 1:1 in the presence of PEG 1500 (Boehringer, Mannheim, Germany). Fused cells were cultured and selected in RPMI 1640 (Gibco Invitrogen, Carlsbad, USA) supplemented with 20% bovine calf serum (Biochrom AG, Berlin, Germany), 100 U/ml penicillin, 100 μg/ml streptomycin (Gibco Invitrogen), 100 μM hypoxanthine, 16 μM thymidine, and 0.4 μM aminopterin (Sigma-Aldrich), 80 U/ml murine interleukin-6 (Calbiochen, San Diego, USA) and 50 μM βME (Gibco Invitrogen) in 96-well microtitre plates. The selection medium was refreshed after one week. After 12 days, culture supernatants were screened for antibodies binding to RAI3 liposomes in a direct ELISA. Hybridomas in wells generating the highest ELISA signals were used for limiting-dilution cloning in HT medium (RPMI 1640 supplemented with 20% bovine calf serum, 100 U/ml penicillin, 100 μg/ml streptomycin, 100 μM hypoxanthine, 16 μM thymidine, 80 U/ml interleukin-6 and 50 μM βME). Monoclonal hybridomas were cultured in standard growth medium (RPMI 1640 supplemented with 10% bovine calf serum, 100 U/ml penicillin, 100 μg/ml streptomycin). Prior to purification, hybridomas were adapted to serum-free hybridoma culture medium ISF-1 (Biochrom AG). Antibodies were purified from culture supernatants by Protein-G affinity chromatography (ÄKTAFPLC systems, Amersham Biosciences, Freiburg, Germany). Isotypes were determined using a BD Bioscience (Heidelberg, Germany) isotyping kit (Heidelberg, Germany). Clone 24 2.3 was selected for immunohistochemical analysis.
Direct ELISA for the analysis of hybridoma supernatants
Hybridoma supernatants containing monoclonal anti-RAI3 antibodies were analysed by direct ELISA. High-binding microtitre plates (Greiner Bio-One, Frickenhausen, Germany) were coated overnight at 4°C with varying amounts of recombinant RAI3 (0.2–200 ng/well) in 100 mM carbonate buffer (pH 9.5), or with 50 ng/well of the negative control proteins bovine serum albumin (BSA), an unrelated human protein (CD30 ligand), and a control GPCR (GPR30) produced under the same conditions as RAI3. Uncoated wells were used as an additional negative control. After blocking with 2% non-fat dry milk, 50 μl hybridoma supernatant was added directly to each well and incubated at room temperature for 1 h. Plates were washed three times with phosphate buffered saline containing 0.05% Tween20 (PBST). The binding of anti-RAI3 antibodies was detected with a HRP-conjugated anti-mouse antibody (Sigma-Aldrich; 1:5,000) and 2'-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) (ABTS) (Roche Diagnostics, Mannheim, Germany) as the substrate. Absorption was measured at 405 nm using an ELISA reader (Biotek Instruments, Bad Friedrichshall, Germany). Each measurement as performed in triplicate.
Measuring binding affinity
The binding affinity of purified monoclonal antibodies was estimated in a competitive ELISA . Briefly, high-binding microtitre plates were coated overnight at 4°C with 50 μl of the 1 μg/ml protein solutions described above. After blocking, a 1 nM antibody solution was pre-incubated with a serial dilution of RAI3 antigen in the molar range 2 × 10-8–2 × 10-10M until equilibrium was achieved. The antibody-antigen mixture was then applied to the coated microtiter plate in triplicate, to measure the unbound antibodies remaining in solution. The immobilised antibodies were detected and quantified as described for the direct ELISA. Kd values were calculated from the absorption data as described elsewhere .
Transfection of mammalian cells
RAI3 template DNA was kindly provided by M-Fold Biotech, Tübingen, Germany. The DNA was amplified with gene-specific primers containing appropriate restriction sites for cloning in pMS RAI3 III and pMS GFP-RAI3 MH, allowing RAI3 to be expressed under the CMV promoter with and without C-terminal GFP. The primers were NheI Kozak 5' (5'-cat tcg agG CTA G cc acc ATG GCT ACA ACA GTC CCT G-3'), NotI 3' (5'-act tgt caG CGG CCG CGC TGC CCT CTT TCT TTA CTT C-3') and SfiI 5' (5'-cat tcg agG GCC CAG CCG GCC ATG GCT ACA ACA GTC CCT G-3'). Restriction sites are underlined, small roman letters are anchor sequences and italic letters show the Kozak consensus. Mock expression vectors with and without GFP were generated as controls. HEK 293T cells (ATCC: CRL-11268) were cultured in RPMI 1640 supplemented with 10% bovine calf serum, 100 U/ml penicillin and 100 μg/ml streptomycin and transfected using Roti-fect (Carl Roth GmbH, Karlsruhe, Germany) according to the manufacturer's protocol. Transfected cells were selected by supplementing the medium with 100 μg/ml zeocin (Invitrogen).
Western blot analysis of cell lysates
RAI3-transfected and mock-transfected HEK293T cells were harvested and agitated for 30 min at 4°C in a non-denaturing lysis buffer (20 mM Tris HCl (pH 8), 137 mM NaCl, 10% glycerol, 2 mM EDTA, 1% Triton X-100) containing a freshly added protease inhibitor cocktail tablet (Roche Diagnostics). The insoluble fraction was separated by centrifugation for 20 min at 12,000 × g. Cell lysates containing approximately 10 μg total protein were separated by sodium dodecylsulphate polyacrylamide gel electrophoresis (SDS-PAGE) on a 12% polyacrylamide gel and blotted to nitrocellulose membranes (Whatman Schleicher & Schuell, Dassel, Germany). Membranes were blocked for 1 h at room temperature in PBST containing 2% non-fat dry milk, washed three times in PBST and incubated with anti-RAI3 antibodies (100 ng/ml in PBST) overnight at 4°C. After three PBST washes, the membranes were incubated with HRP-conjugated anti-mouse antibody (Sigma-Aldrich; 1:5000 in PBST) for 1 h at room temperature, followed by three further washes. Specific binding was detected with ECL Western Blotting substrate (Pierce, Rockford, USA) and visualised in a LAS-3000 imager (Fujifilm). The blot was subsequently stripped by incubation in stripping buffer (62.5 mM Tris (pH 6.8), 2% SDS, 100 mM βME) for 45 min at 50°C. For loading control, the blot was reprobed with an HRP-conjugated anti-β-actin antibody (Sigma-Aldrich; 1:2000) and detected with the ECL system as described above.
The specificity of the anti-RAI3 antibody (Mab 24 2.3) was further analysed by immunocytochemistry in HEK293T cells. Briefly, HEK293T cells transiently transfected with the expression vector encoding RAI3-GFP were seeded at 2.5 × 104 cells per well on a multi-well chamber slide (Nunc) and incubated overnight at 37°C to allow cells to attach. Cells were fixed in 4% paraformaldehyde, washed with PBS and subsequently permeabilised and blocked in PBS containing 0.5% Triton X-100 and 3% BSA. Following a washing step, the primary anti-RAI3 antibody Mab 24 2.3 was applied in a 1:100 dilution and incubated on the cells for 1 h at room temperature. Primary antibody binding was detected indirectly using a 1:2000 dilution of AlexaFluor-546 conjugated anti-mouse secondary antibody (A546; Invitrogen) which was incubated on the cells in the dark for 1 h at room temperature followed by a further washing step. The primary antibody was omitted in control cells. The cell nucleus was stained post fixation using a 20 μM Draq5 solution (Biostatus) in PBS followed by a final washing and a second fixation step. Images of the cells were acquired by confocal microscopy.
Laser scanning confocal microscopy
Imaging was carried out by laser scanning confocal microscopy using 488, 561 and 635 nm excitation laser lines and tunable emission filters adjusted to cover emission ranges between 540/75, 600/40 and 685/70 nm for RAI3-GFP, AlexaFluor-546 and Draq5, respectively. Images of cells labelled with the three fluorophores were acquired using a 40× PlanApo air objective (0.85 numerical aperture) and sequential acquisition to prevent measuring "cross-talk" between the three emission signals. The images were pseudo coloured in Adobe Photoshop.
RAI3 protein expression in breast cancer patients was assessed using a previously-described tissue microarray (TMA)  comprising 157 breast cancer specimens and 44 normal breast tissue samples. The TMA contained one tissue core from non-selected, formalin-fixed and paraffin-embedded primary breast cancer from patients aged 25–82 (median age of 56), diagnosed between 1994 and 2002 at the Institute of Pathology, University of Regensburg, Germany. An experienced surgical pathologist (A.H.) evaluated H&E-stained slides of all specimens prior to construction of the TMA in order to identify representative tumour areas. All tumours were graded according to Elston and Ellis . Clinical follow-up data, provided by the Central Tumour Registry, Regensburg, Germany, were available for all 157 breast cancer patients with a median follow-up period of 78 months (range 0–148 months). All patients gave informed consent for retention and analysis of their tissue for research purposes and the Institutional Review Board of the participating centre approved the study.
Breast cancer cDNA dot blot hybridisation
The breast cancer profiling array (CPA) (BD Clontech, Heidelberg, Germany) contains 50 pairs of cDNAs generated from matched cancerous and normal breast tissue samples of individual patients and three breast cancer lymph node metastasis specimens, spotted on a nylon membrane . Hybridisations using 25 ng of a gene-specific32P-labelled cDNA probe prepared from a RAI3 Unigene cDNA clone were carried out according to the manufacturer's recommendations. This hybridisation probe included base pair 550 to 2835 of the RAI3 cDNA deposited under accession number NM_003979. The tumour/normal intensity ratio was calculated using a STORM-860 phosphorimager (Molecular Dynamics, Sunnyvale, CA, USA) and normalised against the background.
Immunohistochemical characterisation of the tissue microarray
Immunohistochemical studies of HER2 expression were carried out using the avidin-biotin peroxidase method with a 3,3'-diaminobenzidine (DAB) chromogen in a NEXES immunostainer (Ventana, Tucson, AZ, USA) following antigen retrieval by microwaving for 30 min. Primary anti-HER2 (DAKO, Hamburg, Germany; 1:400), anti-ER and anti-PR (Novocastra, Newcastle Upon Tyne, UK; 1:20) antibodies were detected using the ChemMate detection kit (DAKO). A surgical pathologist (A.H.) performed a blinded evaluation of the TMA slides without clinical data. Non-interpretable results were caused by a lack of tumour tissue and the presence of necrosis or crush artefacts. HER2 expression was scored according to the DAKO HercepTest. For the evaluation of ER and PR, a semi-quantitative immunoreactivity score (IRS) was used as described by Remmele and Stegner .
The tissue microarray was stained using the Advance Kit (DAKO K5007) according to the manufacturer's instructions. Breast carcinomas were used as positive controls. After paraffin removal and rehydration, the tissue samples were boiled in a microwave oven for 30 min in 10 mM sodium citrate buffer (pH 7.6). Endogenous peroxidase was blocked by peroxidase blocking solution (DAKO S 2023) for 10 min. The anti-RAI3 antibody Mab 24 2.3, was selected for immunohistochemical application. It was applied at a dilution of 1:50 for 3–4 h at 4°C and detected with DAB. In negative controls the primary antibody was omitted. As an additional negative control, a competitive immunohistochemical analysis was performed by pre-incubation of the anti-RAI3 antibody Mab 24 2.3 with 200-fold molar excess of RAI3. Slides were counterstained with haematoxylin and after dehydration mounted in Vitro-Clud (Langenbrinck, Emmendingen, Germany). An experienced pathologist (N.B.) scored the immunohistochemical staining intensity according to the scoring system suggested devised by Remmele and Stegner .
SPSS version 14.0 (SPSS Inc, Chicago, IL) was used for statistical evaluation. Differences were considered statistically significant at P < 0.05. For analysis of the CPA, a Kolmogorov-Smirnov test was applied to test for normal value distribution, followed by a two-sided paired t-test to analyse differences in normal and tumour expression. A statistical association between clinicopathological and molecular parameters was tested, using two-sided Fisher's exact test. Recurrence-free (RFS) and overall survival (OS) were calculated according to the Kaplan-Meier equation.