First, we analysed ADAM33 expression in breast tumour cell lines by RT-PCR and western blotting. We also used 5-aza-2'-deoxycytidine (5azadCR) treatment and DNA bisulphite sequencing to study the promoter methylation of ADAM33 in breast tumour cell lines. We evaluated ADAM33 methylation in primary tumour samples by methylation specific PCR (MSP). Finally, ADAM33 promoter hypermethylation was correlated with clinicopathological data using the chi-square test and Fisher's exact test.

The expression analysis of ADAM33 in breast tumour cell lines by RT-PCR revealed gene silencing in 65% of tumour cell lines. The corresponding lack of ADAM33 protein was confirmed by western blotting. We also used 5-aza-2'-deoxycytidine (5-aza-dCR) demethylation and bisulphite sequencing methodologies to confirm that gene silencing is due to ADAM33 promoter hypermethylation. Using MSP, we detected ADAM33 promoter hypermethylation in 40% of primary breast tumour samples. The correlation between methylation pattern and patient's clinicopathological data was not significantly associated with histological grade; tumour stage (TNM); tumour size; ER, PR or ERBB2 status; lymph node status; metastasis or recurrence. Methylation frequency in invasive lobular carcinoma (ILC) was 76.2% compared with 25.5% in invasive ductal carcinoma (IDC), and this difference was statistically significant (p = 0.0002).

Total RNA was kindly provided by Dr Michael O'Hare from the Ludwig Institute for Cancer Research (University College, England) from the following breast tumour cell lines: MDA-MB-134, MDA-MB-415, MDA-MB-175, MDA-MB436, MDA-MB-435, MDA-MB-468, MDA-MB456, BT-20, ZR-75-30, ZR-75-1, CAMA-1, GI101, 734B, CAL51, MCF7, SK-B-7; SK-BR-5, SK-BR-3, PMC42 and DU4475. The following breast cell lines, all obtained from the Ludwig Institute for Cancer Research (São Paulo, Brazil) were cultured in this study: HB4a (normal epithelial immortalised) [11], MDA-MB-231, MDA-MB-436, MDA-MB-435, MCF7 and PMC42. The cell lines were cultured in RPMI 1640 medium (GIBCO/Invitrogen Life Technologies, USA) containing 10% foetal bovine serum (complemented with 0.2 mM glutamine, 40 μg/mL garamycin and 10 μg/mL insulin) at 37°C in a humidified incubator with 5% CO₂.

Frozen tissue samples were dissolved in lysis buffer for subsequent DNA isolation using the phenol/chloroform protocol. They were then subjected to sodium bisulphite treatment using the EpiTect^® Bisulfite Kit (Qiagen). For total RNA isolation, the TRIzol Reagent (Life Technologies, USA) was used according to the protocol supplied by the manufacturer.

ADAM33 mRNA expression analysis was performed by applying RT-PCR to an RNA panel of 20 breast tumour cell lines and normal breast. HB4a was included as a normal cell line control. Reverse transcription reactions were performed using 500 ng of DNA-free RNA, an oligo(dT)_12–18 primer and Superscript II Reverse Transcriptase (Gibco, BRL). PCR was performed using ADAM33-specific primers. The sense primer was 5' CAT GAC ACC TTC ATG CTG, and the anti-sense primer was 5' ATC TTG GCA TCT GGA CTT G. The PCR was performed in a volume of 20 μl containing 1× PCR buffer (Invitrogen), 1.5 mM MgCl₂ (Invitrogen), 200 μM dNTPs (Gibco, BRL), 0.30 μM of each primer and 1 U of Taq Platinum (Invitrogen). The PCR conditions were as follows: 95°C for 12 min, 94°C for 45 s, 60°C for 45 s, 72°C for 1 min and a final extension of 72°C for 5 min. Primers for GAPDH were as follows: sense, 5' CTG CAC CAC CAA CTG CTT A; anti-sense, 5' CAT GAC GGC AGG TCA GGT C. PCR conditions were as follows: 95°C for 12 min, 94°C for 45 s, 63°C for 45 s, 72°C for 1 min and a final extension of 72°C for 5 min. PCR products were resolved on 1% agarose gels and visualised by ethidium bromide staining. Hybridisations were carried out as previously described [12] using ADAM33 or GAPDH cDNA probes fragments corresponding to nucleotides 2191–2515 and 486–778, respectively. They were amplified using ADAM33 RT-PCR and GAPDH RT-PCR primers, purified from agarose gels and labelled with ³²P.

Western blotting to detect ADAM33 protein was performed on the HB4a normal breast cell line (positive for ADAM33 mRNA expression) as well as on the MDA-MB-231 and MDA-MB-435 (both negative for ADAM33 mRNA expression) tumour cell lines. Protein lysates were obtained from approximately 1.2 × 10⁷ cells in lysis buffer (50 mM Tris-HCl, pH 7.4, 0.5% Triton X-100 and 0.2% sodium deoxycholate) containing protease inhibitors (Complete, Roche). Protein samples (100 μg) were resolved by one-dimensional 7.5% SDS-PAGE. Molecular mass was estimated by comparison with Rainbow Molecular Weight Markers (RPN 756, Amersham). Proteins were transferred to PVDF membranes by electroblotting. A commercial polyclonal antibody (Sigma) against the catalytic domain of ADAM33 was used for immunodetection. Horseradish peroxidase-conjugated anti-rabbit secondary antibody was used to detect the binding of primary antibody. HSP70 was detected on the blots as a control of protein integrity.

The cell lines MDA-MB-231, MDA-MB-435, and MDA-MB-436 were analysed using this technique. Cells (10⁶) were incubated with 1 μM of 5-aza-dCR (Sigma Aldrich, Deisenhein, Germany) or left untreated. Every day the medium was changed and no significant cell death was observed. After seven days of treatment, a reduction in cell number was observed, and total RNA was extracted. The expression of ADAM33 in breast tumour cells was analysed by RT-PCR using the housekeeping gene GAPDH as an internal control. PCR products were visualised on a silver-stained 8% polyacrylamide gel. MM 10 bp (Invitrogen) were used as molecular weight markers.

Identification of the ADAM33 CpG island was accomplished using the human genome sequence corresponding to the promoter region of the transcription start site (TSS) of the ADAM33 gene. We identified the RefSeq number based on the GenBank accession and submitted the gene sequence to the Blat Search Genome at the UCSC Genome bioinformatics website http://genome.ucsc.edu. We selected 2000 bp of sequence extending from the 5'-upstream region to 1000 bp downstream region of the TSS. The sequence was submitted for analysis to the CpGPLOT program from the European Bioinformatics Institute website http://www.ebi.ac.uk/emboss/cpgplot. Typical CpG islands were defined as ≥ 200 bp of sequence that had a C+G content of ≥ 50% and a value of >0.6 for the ratio (CpG observed)/(CpG expected) [13]. The selected region of -472 to +389 was amplified from bisulphite-treated DNA samples using a nested-PCR amplification protocol. The first set of primers included a sense primer 5' AGG GAG TTA TGT TTT TTG TTT TGT TAG and an anti-sense primer 5' ATT ACC TAA ACC TTC CTA TCC TTA. PCR products were used as templates for the nested PCR. The second set of primers included a sense primer 5' GGG TTA GTT TAA GTA TAT TTG AG and an anti-sense primer 5' ACA CCC AAT ACA AAT AAA TAA CC. The PCR conditions were as follows: one round of 95°C for 12 min, 94°C for 3 min, 48°C for 3 min, 72°C for 2 min; five cycles of 94°C 3 min, 50°C for 3 min, 72°C for 2 min; and 35 cycles of 94°C for 1 min, 52°C for 1 min, and 72°C. Different annealing temperatures (55°C, 57°C and 59°C) were used for the nested reaction. Amplified products were purified using the Qiaquick Gel Extraction Kit (Qiagen) and cloned into a pCR2.1 cloning vector (Invitrogen). Eight clones were sequenced for each cell line using the vector's universal and/or reverse primers. DNA sequencing reactions were performed using Big Dye Terminator technology (Applied Biosystems) on an ABI 377 sequencer (Applied Biosystems) according to the manufacturer's instructions. The 100% of methylation was obtained if cytosine in the CpG dinucleotides was present in the eight sequenced clones. The methylation percentage for each tumour cell lines (global methylation pattern) was calculated by dividing the number of methylated CpG dinucleotides by the total number of CpGs analysed.

Methylation-specific PCR (MSP) was performed as previously described [14]. MSP primers for the methylated condition (M) included a sense primer 5'GTT TGA GGT TGT ATC GGG TA and an anti-sense primer 5'ACT CGC AAC TCC GAC TCC G. For the unmethylated condition (U), a sense primer 5' GTT TGA GGT TGT ATT GGG TA and an anti-sense primer 5' ACT CAC AAC TCC AAC TCC A were used. The PCR protocol was one round of 95°C for 10 min; 35 cycles of 94°C for 45 s and either 66°C for methylated condition (M) or 64°C for unmethylated condition (U) for 15 s, followed by 72°C for 45 s; and a final extension of 72°C for 5 min. PCR products were run on 8% polyacrylamide gels and silver stained.

To investigate the mechanism of ADAM33 transcriptional silencing, we treated MDA-MB-435, MDA-MB-436 and MDA-MB-231 breast tumour cell lines with the demethylating agent 5'-aza-2'deoxycytidine (5-azad-CR). The expression of ADAM33 was restored upon treatment in all cell lines, as detected by RT-PCR (Figure (Figure1C1C).

Two putative islands were found in the ADAM33 promoter region. One of the CpG islands is located between -1012 and -748. This island has 264 bp, proximal to the cut-off for a CpG island [13]. For this reason, we decided to study the region from -421 to +324, which contains part of the first exon. This region possesses promoter activity, as previously described by others [6], as well as more than 40 transcription factor-binding sites predicted by the TFSEARCH software (data not shown). Sodium bisulphite sequencing was carried out on a region containing 77 CpG dinucleotides (-472 to +389), 71 of which were located within one of the CpG islands (-421 to +324) of the TSS (Figure (Figure2A).2A). We analysed the methylation pattern of eight independent alleles (eight clones) from normal HB4a, as well as the breast tumour cell lines PMC42, MDA-MB-231, MDA-MB-435, MDA-MB-436 and MCF7. Methylation of more than 80% of the CpG dinucleotides in this region was detected in the breast tumour cell lines analysed except PMC42 (Figure (Figure3).3). All cell lines that exhibited hypermethylation showed down-regulation of ADAM33 expression (Figure (Figure1A).1A). We observed that there are some dinucleotides that are likely to be more important than others in transcriptional regulation. Comparing Figure Figure1A1A with Figure Figure3,3, it is clear that ADAM33 transcripts from cell lines with a lower methylation density were expressed at higher levels than in cells with a higher methylation density. However, MCF7 has its ADAM33 promoter region highly methylated (methylation percentage = 86.1%), and transcription occurs to a smaller extent (Figure (Figure1A).1A). These results indicate that there are crucial CpGs that must be methylated in order to completely silence gene transcription. These results also indicate that the methylation density is critical to the down-regulation of expression. In other words, hypermethylation may not be the only factor of critical importance to transcription. These dinucleotides may be important for the binding of the transcription machinery, mainly because they are located near the TSS. The methylation seems to initiate mostly at dinucleotides 1–3 and 74–77 and to spread towards the promoter region. These results may therefore support the spreading model proposed by Turker [18]. Moreover, methyl-binding proteins physically associate with methylated CpG and attract repressive complexes such as histone deacetylases (HDAC) [19]. Different histone tail modifications may exist. It is possible that the level of gene expression for a given allele can vary from extremely high to undetectable. Alternatively, repression by the methyl-CpG-binding proteins can occur via mechanisms that do not involve histones [19]. Another possibility is that the biallelic inactivation of tumour suppressor genes involves DNA methylation, deletion or point mutation. Furthermore, the sequence of inactivating events can occur in any order [20].