Expression of the Candidate Tumor Suppressor Gene Hsrbc Is Frequently Lost in Primary Lung Cancers with and Without DNA Methylation

Home , Treatment of lung cancer

Expression of the candidate tumor suppressor gene hSRBC is frequently lost in primary lung cancers with and without DNA methylation

Sabine Zo¨ chbauer-Mu¨ ller*,1,2, Kwun M Fong3, Joseph Geradts4, Xie Xu5, Sonja Seidl1, Adelheid End-Pfu¨ tzenreuter6, Gyo¨ rgy Lang6,7, Gerwin Heller1,2, Christoph C Zielinski1,2, Adi F Gazdar8,9 and John D Minna8,10,11

1Clinical Division of Oncology, Department of Medicine I, University Hospital, Wa¨hringer Gu¨rtel 18-20, Vienna 1090, Austria; 2Center of Excellence in Clinical and Experimental Oncology, University Hospital, Vienna, Austria; 3The Prince Charles Hospital, Brisbane, Australia; 4Roswell Park Cancer Institute, Buffalo, NY, USA; 5Sagres Discovery, Davis, CA, USA; 6Department of Cardiothoracic Surgery, University Hospital, Vienna, Austria; 7Department of Thoracic Surgery, Kora´nyi National Institute for Pulmonology, Budapest, Hungary; 8Hamon Center for Therapeutic Oncology Research, The University of Texas Southwestern Medical Center, Dallas, TX, USA; 9Department of Pathology, The University of Texas Southwestern Medical Center, Dallas, TX, USA; 10Department of Internal Medicine, The University of Texas Southwestern Medical Center, Dallas, TX, USA; 11Department of Pharmacology, The University of Texas Southwestern Medical Center, Dallas, TX, USA

Recently, the human SRBC (hSRBC)gene, a candidate findings suggest that hSRBC is a candidate TSG involved tumor suppressor gene (TSG), has been mapped to the in lung cancer pathogenesis, where expression is fre- chromosomal region 11p15.5–p15.4 where frequent allele quently inactivated by methylation and other mechanisms. loss has been described in lung cancer. Aberrant methyla- Oncogene (2005) 24, 6249–6255. doi:10.1038/sj.onc.1208775; tion (referred to as methylation)of the promoter region of published online 6 June 2005 TSGs has been identified as an important mechanism for gene silencing. Loss of hSRBC protein expression occurs Keywords: tumor suppressor gene; lung cancer; hSRBC; frequently in lung cancer cell lines and sodium bisulfite epigenetic; methylation-specific PCR sequencing of the promoter region of hSRBC in several lung cancer cell lines suggested that methylation plays an important role in inactivating hSRBC. To determine the methylation status of hSRBC in a large collection of Introduction primary lung cancer samples, corresponding nonmalignant lung tissues and lung cancer cell lines (N ¼ 52), we Frequent allele loss at chromosomal regions is a strong designed primers for a methylation-specific PCR assay. evidence for the presence of one or more tumor Methylation was detected in 41% of primary non-small- suppressor genes (TSG) in these regions. Chromosome cell lung cancers (NSCLC)( N ¼ 107)and in 80% of 11p15.5–p15.4has been recognized as such a region primary small-cell lung cancers (SCLC)( N ¼ 5), but was since allele loss at this region was frequently observed in seen only in 4% of corresponding nonmalignant lung a variety of malignant diseases including lung cancer tissues (N ¼ 103). In all, 79% of lung cancer cell lines (Vandamme et al., 1992; Winqvist et al., 1995; Baffa were methylated and the frequency of hSRBC methylation et al., 1996; Tran and Newsham, 1996). Recently, the was significantly higher in SCLC (100%)than in NSCLC human SRBC (hSRBC) gene has been mapped to this (58%)cell lines. Normal hSRBC protein expression was chromosomal region (Xu et al., 2001). A cDNA clone detected in only 18% of primary NSCLCs (N ¼ 93)by encoding human SRBC (serum deprivation response immunostaining and a significant association between loss factor (sdr)-related gene product that binds to c-kinase) of protein expression and methylation was found. hSRBC was isolated when a normal mammary epithelial cDNA re-expression was observed after treatment of lung cancer library was screened by the yeast two-hybrid method cells with the demethylating agent 5-aza-20-deoxycytidine. using a recombinant protein containing amino acids 1– In addition, 45% of the 76 hSRBC immunostaining- 304of the breast cancer susceptibility gene BRCA1 as negative NSCLCs did not have hSRBC promoter the ‘bait’ (Xu et al., 2001). The hSRBC gene encodes an methylation, indicating that other mechanisms of hSRBC open-reading frame of 261 amino acids with a leucine expression silencing also exist. Both hSRBC immuno- zipper-like motif in its amino-terminal region (GenBank staining and methylation results did not correlate with accession number AF408198). As a BRCA1-interacting clinicopathological characteristics of these patients. Our protein, hSRBC may also be involved in DNA damage response and participate in BRCA1-mediated tumor suppression pathways. hSRBC protein expression is *Correspondence: S Zo¨ chbauer-Mu¨ ller; E-mail: [email protected] frequently lost in breast and lung cancer cell lines Received 25 November 2003; revised and accepted 5 April 2005; compared to normal mammary and lung epithelial cells, published online 6 June 2005 suggesting that hSRBC is a candidate TSG involved in hSRBC methylation in lung cancers SZo¨chbauer-Mu¨ller et al 6250 the pathogenesis of breast and lung carcinomas (Xu et al., 2001). This loss was associated with tumor cell line acquired promoter methylation. Aberrant methylation of normally unmethylated CpG islands located in or near the promoter region of cancer-related genes has been associated with transcriptional inactivation of these gene in human cancer (Jones and Baylin, 2002). It is becoming increasingly apparent that aberrant methylation (referred to as methylation) of the promoter region of genes is the main mechanism for gene silencing in tumors (Jones and Baylin, 2002). Bisulfite genomic sequencing of several breast and lung cancer cell lines revealed that the promoter region of hSRBC is frequently methylated in these cell lines. Moreover, hSRBC expression in the breast cancer cell line MCF7 was partially restored by treatment with the deacetylase inhibitor trichostatin A, whereas treatment with both trichostatin A and, additionally, the demethylation reagent 5-aza-20-deoxycytidine (5-aza-dC) induced hSRBC expression in MCF7 cells (Xu et al., 2001). Thus, there is a strong evidence that tumor acquired promoter region methylation is an important mechanism for inactivating hSRBC in some human tumors. To further investigate the role of hSRBC in cancer pathogenesis, we developed a methylation-specific PCR (MSP) assay, which is a very suitable method to investigate methylation in a large number of primary tumor samples. Using MSP, we analysed the methylation status of the promoter region of hSRBC in a total of 267 samples consisting of primary lung carcinomas, corresponding nonmalignant lung tissues and lung cancer cell lines. In addition, immunostaining of hSRBC was performed in primary lung carcinomas and lung cancer cell lines, and these results were compared with the methylation pattern of hSRBC. Furthermore, the Figure 1 Examples of hSRBC immunostaining in primary results from both the methylation analysis and the NSCLCs. (a) Positive staining with distinct membrane hSRBC immunostaining were compared with clinicopathologi- reactivity in all areas of the tumor (patient #76). (b) No hSRBC cal characteristics of these patients. staining in tumor with preserved reactivity in admixed mesen- chymal cells (patient #74)

Results the carcinoma (Figure 2). In other cases, most of the hSRBC expression by RT–PCR and immunostaining staining in the positive areas was localized to the periphery of tumor nests. Samples with a heterogeneous hSRBC expression was investigated in 18 lung cancer staining pattern were combined with the group of cell lines by RT–PCR. The expression of this gene was samples with complete loss of protein expression observed only in eight (44%) lung cancer cell lines. Eight resulting in 77 (76%) samples negative for hSRBC of 13 (62%) non-small-cell lung cancer (NSCLC) cell immunostaining. Six samples demonstrated only cyto- lines expressed hSRBC, while none of the five (0%) plasmatic staining and were excluded from the analysis. small cell cancer (SCLC) cell lines expressed this gene. Only two of 11 (18%) lung cancer cell lines expressed In 101 primary NSCLC samples and in 11 lung cancer hSRBC protein by immunostaining (Figure 3). Interest- cell lines, hSRBC protein expression was investigated by ingly, expression was observed in two of six (33%) immunostaining. In all, 18 (18%) primary NSCLCs NSCLC cell lines, but in none of the five (0%) demonstrated a membranous staining pattern in the investigated SCLC cell lines. neoplastic cells and were scored as positive for hSRBC protein expression (Figure 1). However, complete loss of Frequency of hSRBC methylation by MSP hSRBC protein expression was observed in 62 (61%) (Figure 1) and a heterogeneous staining pattern was We performed genomic sequencing of the hSRBC observed in 15 (15%) of the primary NSCLCs (Figure 2). 50CpG island (GenBank accession number AF408198) In some cases with a heterogeneous pattern, the areas of after bisulfite treatment of DNA from five lung cancer positive staining were randomly distributed throughout cell lines, which did not express hSRBC by Western

Oncogene hSRBC methylation in lung cancers SZo¨chbauer-Mu¨ller et al 6251

Figure 2 Heterogeneous hSRBC expression pattern in the Figure 3 Examples of hSRBC immunostaining in lung cancer cell primary NSCLC sample from patient #123. (a) Area of hSRBC lines. (a) hSRBC is expressed in cells of cell line NCI-H1264and ( b) expression in the tumor sample and (b) area with loss of hSRBC hSRBC expression is lost in cells of cell line NCI-H661 expression in tumor cells of the same sample while the admixed stroma is positive immunoblotting as described by Xu et al. (2001). The normal cells. By contrast, the lung cancer cell lines methylation pattern of the 50CpG island of the hSRBC represent pure populations of tumor cells and we found gene indicated that the region between nts 3520 and these tumor lines to have either methylated or un- 3866 was highly methylated. In addition, bisulfite methylated hSRBC alleles. genomic sequencing was also performed in several primary tumor samples, confirming the results in tumor Comparison of hSRBC methylation with hSRBC cell lines. An example is shown in Figure 4. Based on expression by immunostaining these results, we designed a set of PCR primers that A statistical significant correlation between methylation distinguish between methylated and unmethylated DNA of hSRBC and loss of hSRBC protein expression was sequences in the 50 region of the hSRBC gene. Using found for both primary NSCLCs (P ¼ 0.00003) and lung these primers in MSP assay, we determined the cancer cell lines (P ¼ 0.01) (Table 2). All cases with frequency of hSRBC methylation in primary NSCLC methylation exhibited loss of hSRBC protein expres- samples and corresponding nonmalignant lung tissues, sion. Of interest, there were 34/76 (45%) of tumors that primary SCLC specimens and lung cancer cell lines. The had lost hSRBC expression but were not methylated results are summarized in Table 1 and examples for (Table 2). This indicates that there are other mechan- MSP are shown in Figure 5. hSRBC promoter region isms besides tumor acquired promoter methylation methylation was frequent in primary lung tumors and leading to loss of hSRBC expression. rare in nonmalignant lung specimens. In addition, the unmethylated form of hSRBC was found in 100% of hSRBC re-expression after treatment with 5-aza-dC nonmalignant lung specimens and in 100% of primary tumors, which had been only macroscopically dissected, Loss of hSRBC expression was found in the lung cancer and thus all samples had some contamination with cell line NCI-H1993 by RT–PCR. In addition, this cell

Oncogene hSRBC methylation in lung cancers SZo¨chbauer-Mu¨ller et al 6252

Figure 4 Representative part of the 50CpG island sequence of the hSRBC gene in lung cancer sample #34. Methylated cytosines in CpG sites are indicated with an arrow

Table 1 Frequency of hSRBC methylation in primary lung carcino- sion and any of the clinicopathological characteristics mas, corresponding nonmalignant lung specimens and lung cancer cell was observed. However, a statistically signiﬁcant asso- lines ciation between hSRBC methylation and loss of hetero- No. tested hSRBC methylated zygosity at the MYCL locus at chromosome 1p was detected (P ¼ 0.005). Moreover, a correlation between Primary NSCLCs 107 44 (41%) methylation of hSRBC and methylation of the RARb Primary SCLCs 5 4(80%) Corresponding nonmalignant lung 103 4(4%) (P ¼ 0.04) was found. No correlation between hSRBC methylation and other investigated molecular abnorm- Lung cancer cell lines 52 41 (79%) alities was observed. NSCLCs 2414(58%) SCLCs 2424(100%) Carcinoid 1 1 (100%) NSCLC, not speciﬁed 3 2 (67%) Discussion

We studied the methylation status of the 50CpG island line was hSRBC methylated, making it a suitable of hSRBC in primary lung cancers, adjacent nonmalig- candidate for treatment with the demethylating agent nant tissues and lung cancer cell lines. We first 5-aza-dC. Re-expression of hSRBC was seen after sequenced sodium bisulfite-treated DNA from lung treatment of NCI-H1993 cells with 5-aza-dC, confirming cancer cell lines in the hSRBC 50CpG island region as the role of methylation in regulating hSRBC expression reported by Xu et al. (2001). Based on these results, we (Figure 6). designed a set of primers for MSP. Using this primer set, we found a high percentage of primary lung cancer samples and lung cancer cell lines methylated for Comparison of hSRBC methylation and loss of protein expression with clinicopathological characteristics and hSRBC. Interestingly, both primary SCLC samples molecular abnormalities and SCLC cell lines were methylated in a higher percentage compared to primary NSCLC samples and Both the methylation results and data about loss of NSCLC cell lines. In concordance, loss of hSRBC hSRBC protein expression from primary NSCLC expression was more frequently observed in SCLC samples were compared with clinicopathological char- compared to NSCLC cancer cell lines. While the acteristics from these patients including sex, age, comparison of hSRBC methylation between SCLC histology, TNM classification of the tumors, smoking and NSCLC cell lines was statistically significant history and overall survival of the patients. In addition, (P ¼ 0.0004), the comparison between primary SCLC a comparison of our results with certain molecular samples and primary NSCLC specimens did not reach abnormalities that had been investigated previously was statistical significance because of the small number of performed. These abnormalities included K-ras codon primary SCLC samples. Different methylation frequen- 12, and p53 exons 5–8 mutations, allele loss at 1p cies between SCLC and NSCLC samples also have been (MYCL), 3p21 (D3S1029), 3p25.3–26.2 (D3S1038), 8p reported for the genes p16, adenomatous polyposis coli (LPL), 9p (IFNA, D9S126), 13q (RB, D13S260) and 18q (APC), H-cadherin (CDH13)andRASSF1A (Dam- (DCC), and the methylation status of the genes retinoic mann et al., 2000, 2001; Burbee et al., 2001; Toyooka acid receptor b-2 gene (RARb), RAS association domain et al., 2001). In three of four nonmalignant lung samples family 1A (RASSF1A), FHIT, tissue inhibitor of where hSRBC methylation was detected, the corres- metalloproteinase-3 (TIMP-3), p16INK4a (p16), O6- ponding tumor was also methylated for this gene, methylguanine-DNA-methyltransferase (MGMT), suggesting that there was some contamination with death-associated protein kinase (DAPK), E-cadherin adjacent malignant cells. Another explanation for this (ECAD), p14ARF (p14), BLU and glutathione S-transfer- finding might be that hSRBC methylation occurs ase P1 (GSTP1). No significant correlation between already in premalignant changes as it has been demons- hSRBC methylation or loss of hSRBC protein expres- trated for other genes (Kersting et al., 2000; Belinsky

Oncogene hSRBC methylation in lung cancers SZo¨chbauer-Mu¨ller et al 6253 a TU 6 TU 21 TU 34 TU 45 TU 82 TU 91 TU 98

347 bp product

M M M M M

b TU 6 TU 21 TU 34 TU 45 TU 82 TU 91 DNA ladder 347 bp product

UUUU U U U Figure 5 Examples of hSRBC MSP in primary NSCLCs. (a) The amplified product represents the methylated form of hSRBC with a product size of 347 bp (M). Samples whose lanes do not show a band are not methylated. (b) To confirm the presence and integrity of DNA, MSP was also performed with the primer pair specific for the unmethylated form of hSRBC and all primary NSCLC samples tested positive (U) because of contamination with normal cells

Table 2 Correlation between methylation of hSRBC and loss of 5-aza-dC hSRBC protein expression in primary NSCLCs and in lung cancer cell -+H2O lines 500 bp hSRBC hSRBC methylation DNA Primary NSCLCs ladder hSRBC immunostaining À + 5-aza-dC À 344276 -+H2O +17017P ¼ 0.00003 GAPDH Total 51 42 93 366 bp DNA Lung cancer cell lines ladder hSRBC immunostaining À + À 189 Figure 6 Re-expression of hSRBC after treatment of lung cancer +202P ¼ 0.01 cells with 5-aza-dC. The lung cancer cell line NCI-H1993 that lack Total 3 8 11 hSRBC expression was grown in the presence ( þ lanes) and absence (Àlanes) of 0.5 mM 5-aza-dC for 6 days. Afterwards, RT– PCR was performed. GAPDH was used as a control et al., 2002; Soria et al., 2002; Zo¨ chbauer-Mu¨ ller et al., The role of histone deacetylation in silencing hSRBC 2003). has not been determined yet and needs to be investigated We compared hSRBC methylation and loss of in future studies. expression of hSRBC by immunostaining in primary We did not find any significant correlation between NSCLC samples and lung cancer cell lines. A significant hSRBC methylation status and clinicopathological correlation between hSRBC methylation and loss of characteristics including sex, age, histology, TNM expression by immunostaining for both primary classification of the tumors, smoking history and overall NSCLC samples and lung cancer cell lines was observed. survival of the NSCLC patients. However, we observed The heterogeneity of hSRBC immunostaining in 15 a strong correlation between hSRBC methylation and primary NSCLC samples may be explained by tumor allele loss at the MYCL locus at chromosome 1p, which heterogeneity for methylation. Every case with hSRBC needs to be further evaluated. methylation exhibited loss of hSRBC protein expres- In conclusion, we found frequent methylation of sion, which is an evidence that methylation plays a hSRBC associated with loss of hSRBC expression in major role in inactivating hSRBC. However, 45% of primary lung cancers and lung cancer cell lines. More- primary NSCLCs lacked hSRBC expression that did not over, we were able to show that hSRBC methylation is show hSRBC methylation, and thus other mechanisms reversible with 5-aza-dC. Our results support the besides methylation are responsible for silencing hSRBC hypothesis that inactivation of hSRBC is involved in expression. Although Xu et al. (2001) detected hSRBC- the pathogenesis of lung cancer also providing evidence coding region mutations in a few ovarian and lung that hSRBC is functioning as a TSG. However, the cancer cell lines, they do not seem to play an important function of hSRBC needs to be determined. role in inactivating hSRBC. However, mutations as well as deletions may be responsible for the inactivation of hSRBC in some cases. Recently, DNA methylation and chromatin structure has been linked. Methylated DNA Materials and methods recruits methyl binding proteins, which attract a chromatin-remodeling complex along with proteins that Tissue samples modify histones by deacetylating them, thus closing Tissues were collected after obtaining appropriate Institutional down DNA to transcription (Jones and Baylin, 2002). Review Board permission and informed consent from the

Oncogene hSRBC methylation in lung cancers SZo¨chbauer-Mu¨ller et al 6254 patients. From NSCLC patients, tumor samples (N ¼ 107) and external controls, respectively (Xu et al., 2001). A staining corresponding normal lung tissues (N ¼ 103) were obtained reaction was scored as positive if there was distinct membrane surgically in the Prince Charles Hospital in Brisbane, reactivity in all areas of the tumor. A tumor was scored as Australia. This cohort of patients had been investigated negative if there was complete or, less commonly, partial loss previously for various genetic abnormalities including LOH of membrane staining with preserved reactivity in adjacent at multiple chromosomal sites, K-ras codon 12 mutations, p53 benign cells. exons 5–8 mutations and methylation analysis of multiple genes (Fong et al., 1995a, b, 1996; Zo¨ chbauer-Mu¨ ller et al., 2001a, b; Burbee et al., 2001; Agathanggelou et al., 2003). Methylation-specific PCR There were 76 males and 31 females, age 28–81 (mean 61) years DNA was prepared from tissue samples, cell lines by standard at diagnosis. In all, 61 patients had stage I, 21 stage II, 24stage methods and bisulfite modification of genomic DNA was IIIA and one patient stage IIIB disease. Histological subtypes performed as reported by Herman et al. (1996). Treatment of included 45 adenocarcinomas, 43 squamous cell carcinomas, genomic DNA with sodium bisulfite converts unmethylated 11 adenosquamous carcinomas, four large-cell carcinomas, but not methylated cytosines to uracil, which are then three atypical carcinoids and one typical carcinoid. In all, 98 converted to thymidine during the subsequent PCR step giving patients were ever smokers (consisting of current and former sequence differences between methylated and unmethylated smokers) with mean exposure of 31 pack-years, and nine were DNA. PCR primers that distinguish between methylated and never smokers. Survival data of 5 or more years were available unmethylated DNA were designed. Primer sequences were for most patients. SCLC samples (N ¼ 5) were obtained from determined based on the sequence data of the 50CpG island of patients who underwent a surgical procedure because of the gene as described in the ‘Results’ section. Primer sequences suspect lung lesions in the University Hospital in Vienna, for the methylated hSRBC reaction were 50-GTT TCG GGT Austria. There were five males with an age of 54–76 (mean 63) TTT GAT AGT TCG CG-30 (forward) and 50-CCT TCC GCT years at diagnosis. ATC CCG CGC CG-30 (reverse), and primer sequences for the unmethylated hSRBC reaction were 50-GTT TTG GGT TTT Lung cancer cell lines GAT AGT TTG TG-30 (forward) and 50-CCT TCC ACT ATC CCA CAC CA-30 (reverse). The PCR mixture has been In all, 52 lung cancer cell lines (24NSCLC, 24SCLC, one reported previously (Zo¨ chbauer-Mu¨ ller et al., 2001a, b). carcinoid and three not further specified NSCLC cell lines) Amplification was carried out in a 9700 Perkin-Elmer Thermal which were generated by us and which have been previously Cycler. DNA from peripheral blood lymphocytes treated with described were analysed for the methylation status of hSRBC SssI methyltransferase (New England BioLabs Inc., Beverly, (Phelps et al., 1996; Wistuba et al., 1999). In 18 and in 11 lung MA, USA) was used as a positive control for methylated cancer cell lines, expression analysis of hSRBC was performed alleles. Negative control samples without DNA were included by RT–PCR and immunostaining, respectively. for each set of PCR. PCR products were analysed on 2% agarose gels and visualized under UV illumination. The PCR hSRBC expression by RT–PCR and 5-aza-dC treatment of lung reactions for all samples demonstrating methylation were cancer cell line NCI-H1993 repeated to confirm these results. For RT–PCR, total cellular RNA was isolated using TRIZOL (Gibco BRL, Carlsbad, CA, USA) according to the manu- Other molecular markers facturer’s instructions. Total RNA (2 mg) was reverse tran- scribed using the Superscript II RNase HÀ Reverse Available molecular markers from previous studies in the 107 Transcriptase kit (Gibco BRL, Carlsbad, CA, USA). Primer NSCLC patients included K-ras codon 12, and p53 exons 5–8 sequences for the RT–PCR are 50-AGC TCC ACG TTC TGC mutations and allele loss at 1p (MYCL), 3p21 (D3S1029), TCT TCA-30(forward) and 50-GGC GTG AGT GCT ACA 3p25.3–26.2 (D3S1038), 8p (LPL), 9p (IFNA, D9S126), 13q TTC TGA-30(reverse). Primers for glyceraldehyde-3-phos- (RB, D13S260) and 18q (DCC) (Fong et al., 1995a, b, 1996). In phate dehydrogenase were used to confirm RNA integrity addition, data about the methylation status of the genes (Virmani et al., 2000). For gene reactivation, NCI-H1993 cells RARb, RASSF1A, FHIT, TIMP-3, p16, MGMT, DAPK, (2 Â 105/ml) were seeded in 10% RPMI and treated with ECAD, p14, BLU and GSTP1 were available from these 0.5 mM 5-aza-dC (Sigma Chemical Co., St Louis, MO, USA) patients (Burbee et al., 2001; Zo¨ chbauer-Mu¨ ller et al., 2001a, b; for 6 days. Gene reactivation was tested by RT–PCR. Agathanggelou et al., 2003).

Immunostaining Statistics For immunohistochemical protein expression analysis, we used Statistical analysis was performed using w2 test for differences a previously described mouse monoclonal antibody developed between groups and t-tests between means. Overall survival in one of our laboratories (Xu et al., 2001). Deparaffinized and was calculated using Kaplan–Meier log-rank testing. rehydrated formalin-fixed tissue sections were reacted with primary antibody at a concentration of 0.2 mg/ml for 2 h, following a 20 min antigen retrieval step in subboiling 0.1 M Acknowledgements EDTA buffer (pH 8.0) and a 20 min blocking step in 0.3% This work was supported by grants from the Austrian H2O2. The antigen detection reaction utilized the mouse Science Foundation (J1658-MED, J1860-MED), the Austrian Envision kit from Dako (Carpinteria, CA, USA). Diamino- Federal Ministry of Education, Science and Culture (GZ benzidine and hematoxylin were used as chromogen and 200.062/2-VI/1/2002), the Medical-Scientific Fund of the counterstain, respectively. Nonspecific mouse IgG was used in Mayor of the Federal Capital Vienna (project number lieu of the hSRBC-specific antibody in negative control 2089), Lung Cancer SPORE P50 CA70907, Department of reactions. Non-neoplastic lung tissue and a cell block of lung Defense Grant DAMD170110422 and The Susan G Komen cancer cell line H2141 were used as positive and negative Foundation.

Oncogene hSRBC methylation in lung cancers SZo¨chbauer-Mu¨ller et al 6255 References

Agathanggelou A, Dallol A, Zochbauer-Muller S, Morrissey Soria JC, Rodriguez M, Liu DD, Lee JJ, Hong WK and Mao C, Honorio S, Hesson L, Martinsson T, Fong KM, Kuo MJ, L. (2002). Cancer Res., 62, 351–355. Yuen PW, Maher ER, Minna JD and Latif F. (2003). Toyooka S, Toyooka KO, Maruyama R, Virmani AK, Girard Oncogene, 22, 1580–1588. L, Miyajima K, Harada K, Ariyoshi Y, Takahashi T, Sugio Baffa R, Negrini M, Mandes B, Rugge M, Ranzani GN, K, Brambilla E, Gilcrease M, Minna JD and Gazdar AF. Hirohashi S and Croce CM. (1996). Cancer Res., 56, (2001). Mol. Cancer Therap, 1, 61–67. 268–272. Tran YK and Newsham IF. (1996). Cancer Res., 56, Belinsky SA, Palmisano WA, Gilliland FD, Crooks LA, 2916–2921. Divine KK, Winters SA, Grimes MJ, Harms HJ, Tellez CS, Vandamme B, Lissens W, Amfo K, De Sutter P, Bourgain C, Smith TM, Moots PP, Lechner JF, Stidley CA and Crowell Vamos E and De Greve J. (1992). Cancer Res., 52, 6646– RE. (2002). Cancer Res., 62, 2370–2377. 6652. Burbee DG, Forgacs E, Zo¨ chbauer-Mu¨ ller S, Shivakumar L, Virmani AK, Rathi A, Zochbauer-Muller S, Sacchi N, Fong KM, Gao B, Randle D, Kondo M, Virmani A, Bader Fukuyama Y, Bryant D, Maitra A, Heda S, Fong KM, S, Sekido Y, Latif F, Milchgrub S, Toyooka S, Gazdar AF, Thunnissen F, Minna JD and Gazdar AF. (2000). J. Natl. Lerman MI, Zabarovsky E, White M and Minna JD. (2001). Cancer Inst., 92, 1303–1307. J. Natl. Cancer Inst., 93, 691–699. Winqvist R, Hampton GM, Mannermaa A, Blanco G, Dammann R, Li C, Yoon JH, Chin PL, Bates S and Pfeifer Alavaikko M, Kiviniemi H, Taskinen PJ, Evans GA, Wright GP. (2000). Nat. Genet., 25, 315–319. FA, Newsham I. (1995). Cancer Res., 55, 2660–2664. Dammann R, Takahashi T and Pfeifer GP. (2001). Oncogene, Wistuba II , Bryant D, Behrens C, Milchgrub S, Virmani AK, 20, 3563–3567. Ashfaq R, Minna JD and Gazdar AF. (1999). Clin Cancer Fong KM, Kida Y, Zimmerman PV and Smith PJ. (1996). Br. Res., 5, 991–1000. J. Cancer, 74, 1975–1978. Xu XL, Wu LC, Du F, Davis A, Peyton M, Tomizawa Y, Fong KM, Zimmerman PV and Smith PJ. (1995a). Cancer Maitra A, Tomlinson G, Gazdar AF, Weissman BE, Res., 55, 28–30. Bowcock AM, Baer R and Minna JD. (2001). Cancer Res., Fong KM, Zimmerman PV and Smith PJ. (1995b). Cancer 61, 7943–7949. Res., 55, 220–223. Zo¨ chbauer-Mu¨ ller S, Fong KM, Maitra A, Lam S, Geradts J, Herman JG, Graff JR, Myo¨ ha¨ nen S, Nelkin BD and Baylin Ashfaq R, Virmani AK, Milchgrub S, Gazdar AF and SB. (1996). Proc. Natl. Acad. Sci. USA, 93, 9821–9826. Minna JD. (2001a). Cancer Res., 61, 3581–3585. Jones PA and Baylin SB. (2002). Nat. Rev. Genet., 3, Zo¨ chbauer-Mu¨ ller S, Fong KM, Virmani AK, Geradts J, 415–428. Gazdar AF and Minna JD. (2001b). Cancer Res., 61, Kersting M, Friedl C, Kraus A, Behn M, Pankow W and 249–255. Schuermann M. (2000). J. Clin. Oncol., 18, 3221–3229. Zo¨ chbauer-Mu¨ ller S, Lam S, Toyooka S, Virmani AK, Phelps RM, Johnson BE, Ihde DC, Gazdar AF, Carbone DP, Toyooka KO, Seidl S, Minna JD and Gazdar AF. (2003). McClintock PR, Linnoila RI, Matthews MJ, Bunn Jr PA, Int. J. Cancer, 107, 612–616. Carney D, Minna JD and Mulshine JL. (1996). J. Cell Biochem. Suppl., 24, 32–91.

Oncogene