Molecular Cancer Research
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Molecular Cancer Research 5, 331-339, April 1, 2007. doi: 10.1158/1541-7786.MCR-06-0367
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data, Olejniczak, et al.
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Olejniczak, E. T.
Right arrow Articles by Semizarov, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Olejniczak, E. T.
Right arrow Articles by Semizarov, D.

Cancer Genes and Genomics

Integrative Genomic Analysis of Small-Cell Lung Carcinoma Reveals Correlates of Sensitivity to Bcl-2 Antagonists and Uncovers Novel Chromosomal Gains

Edward T. Olejniczak1, Charles Van Sant1, Mark G. Anderson1, Gang Wang1, Stephen K. Tahir1, Guido Sauter2, Rick Lesniewski1 and Dimitri Semizarov1

1 Global Pharmaceutical Research and Development, Abbott Laboratories, Abbott Park, Illinois and 2 Department of Pathology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Requests for reprints: Dimitri Semizarov, Global Pharmaceutical Research and Development, Abbott Laboratories, 100 Abbott Park Road, Building AP-10, Department R4CD, Abbott Park, IL 60064. Phone: 847-936-0299. E-mail: dimitri.semizarov{at}abbott.com


   Abstract
Top
 Notes
 Abstract
Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Cancer is a highly heterogeneous disease in terms of the genetic profile and the response to therapeutics. An early identification of a genomic marker in drug discovery may help select patients that would respond to treatment in clinical trials. Here we suggest coupling compound screening with comparative genomic hybridization analysis of the model systems for early discovery of genomic biomarkers. A Bcl-2 antagonist, ABT-737, has recently been discovered and shown to induce regression of solid tumors, but its activity is limited to a fraction of small-cell lung carcinoma (SCLC) models tested. We used comparative genomic hybridization on high-density single-nucleotide polymorphism genotyping arrays to carry out a genome-wide analysis of 23 SCLC cell lines sensitive and resistant to ABT-737. The screen revealed a number of novel recurrent gene copy number abnormalities, which were also found in an independent data set of 19 SCLC tumors and confirmed by real-time quantitative PCR. A previously unknown amplification was identified on 18q and associated with the sensitivity of SCLC cell lines to ABT-737 and another Bcl-2 antagonist. The region of gain contains Bcl-2 and NOXA, two apoptosis-related genes. Expression microarray profiling showed that the genes residing in the amplified region of 18q are also overexpressed in the sensitive lines relative to the resistant lines. Fluorescence in situ hybridization analysis of tumors revealed that Bcl-2 gain is a frequent event in SCLC. Our findings suggest that 18q21-23 copy number will be a clinically relevant predictor for sensitivity of SCLC to Bcl-2 family inhibitors. The 18q21-23 genomic marker may have a broader application in cancer because Bcl-2 is associated with apoptosis evasion and chemoresistance. (Mol Cancer Res 2007;5(4):331–39)


   Introduction
Top
 Notes
 Abstract
 Introduction
Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Genetic heterogeneity of cancer is a factor complicating the development of efficacious cancer drugs. Cancers considered to be a single disease entity according to classic histopathologic classification often reveal multiple genomic subtypes when subjected to molecular profiling (1). In some cases, molecular classification proved to be more accurate than the classic pathology (2). The efficacy of targeted cancer drugs may correlate with the presence of a genomic feature, such as a gene amplification (3, 4) or a mutation (5, 6). For HER2 in breast cancer, detection of gene amplification provides superior prognostic and treatment selection information as compared with the immunohistochemical detection of the protein overexpression (7). It is therefore becoming clear that discovery of genomic stratification markers is an attractive approach that may improve the response rate of patients to targeted cancer therapeutics.

Lung malignancies are the leading cause of cancer mortality, which will result in ~160,000 deaths in the United States in 2006 (8). Small-cell lung carcinoma (SCLC) is a subtype of lung cancer, which represents ~20% of lung cancer cases (9). The remainder of lung cancer cases are non–small-cell lung carcinomas, a category that is composed of several common subtypes. In the past several years, there has been substantial progress in the development of targeted therapies for non–small-cell lung carcinoma, such as erlotinib and gefitinib (10). This progress has been facilitated by efforts in molecular profiling of non–small-cell lung carcinomas. Mutations and amplifications in the epidermal growth factor receptor (EGFR) kinase domain were shown to correlate with the response to erlotinib and gefitinib (6, 11-14). Unfortunately, no such progress has been achieved with SCLC, although genomic analysis of SCLC has been reported (15-18).

Comparative genomic hybridization (CGH; refs. 19-22) is a promising approach to genomic biomarker identification. Here we propose coupling genome-wide detection of gene copy number alterations by CGH with screening of therapeutic candidates to enable early discovery of genomic markers of drug sensitivity. We applied high-density genotyping arrays to screen a panel of SCLC lines used to screen a targeted SCLC agent. A gain was identified on 18q that correlates with sensitivity of SCLC cells to the drug. The clinical relevance of this gain has been verified by fluorescence in situ hybridization (FISH) analysis of SCLC tumors. The genes residing in the marker region were shown to be overexpressed in the sensitive lines. Our findings validate the methodology and reveal a mechanism-based marker of drug sensitivity for SCLC.


   Results
Top
 Notes
 Abstract
 Introduction
 Results
Discussion
 Materials and Methods
 Acknowledgements
 References
 
Identification of Recurrent Gene Copy Number Alterations in SCLC by CGH
A panel of 23 SCLC cell lines was profiled using high-density human single nucleotide polymorphism (SNP) genotyping arrays. The density of SNP coverage (~114,000 SNPs with a mean spacing of 23.6 kb) allowed for a very high resolution of the copy number scan. To validate the CGH method, we first sought to identify recurrent copy number abnormalities, as multiple aberrations have previously been defined for SCLC. Figure 1 presents a summary of DNA copy number alterations across the genome. The data are plotted as a frequency plot of copy number gains (≥3 copies) or losses (≤1.5 copies). Gains and losses with a frequency >25% are shown in red and green, respectively. The threshold for losses of 1.5 was chosen because of the apparent signal compression on 100K SNP arrays in regions of DNA losses. In numerous cases, we observed long regions with a copy number of ~1.5. Real-time quantitative PCR established that all these regions had a copy number of 1 (data not shown). This effect can be attributed to the smoothing algorithm used by the GTYPE program.


Figure 1
View larger version (47K):
[in this window]
[in a new window]

  		FIGURE 1. Recurrent gene copy number abnormalities in a panel of SCLC cell lines. Y axis, frequency of copy number gains (≥3 copies) or losses (≤1.5 copies). Arrow, centromere. The p and q arms are shown to the left and to the right of the centromere, respectively. Gains and losses present in ≥25% of the lines are shown in red and green, respectively.

 
The SCLC lines possess multiple recurrent genome alterations, with long regions of frequent gains on almost every chromosome (Fig. 1). Recurrent gains of genetic material (frequency ≥ 25%) are observed on 1p, 1q, 3q, 5p, 6p, 7p, 7q, 8q, 10p, 11q, 12q, and 13q and throughout chromosomes 14, 18, 19, and 20. These hotspots of genomic instability in SCLC contain known oncogenes related to deregulation of cell growth, such as EGFR (26.1% of cell lines), FGFR1 (52.2%), c-myc (56%), and N-myc (39%). The most frequent losses were found on 2q, 3p, 4p, 4q, 5q, 10, 13, 14, 16p, and 21p. These regions of recurrent loss contain known tumor suppressor genes, such as PTEN (34%), APC (21%), and FHIT (30%).

A previously reported pioneering study of 70 lung tumors with 100K SNP arrays (18) included 19 primary SCLC tumors. However, recurrent aberrations in SCLC have not been specifically identified. Here, we obtained the data set by Zhao et al. (18) to validate our findings of recurrent aberrations in SCLC lines. To select only aberrations representative of tumorigenesis in SCLC, we analyzed our data set and the data set of Zhao et al. (18) concurrently using our CGH analysis program. Only abnormalities present in both data sets at a frequency of ≥40% were considered recurrent and selected for further analysis. Supplementary Table S1 summarizes the alterations recurrent in both data sets and lists the location, length, and frequency of change for each region, as well as all the genes residing in the regions.

All of the SCLC aberrations reported to date were present in our data set, thus validating our CGH method. In addition, our data set also contained a number of novel recurrent copy number changes. Table 1 and Supplementary Table S1 (worksheet 2) summarize all copy number abnormalities present in ≥40% of our lines as well as in ≥40% of SCLC tumors from the data set of Zhao et al. (18) and not previously reported in the literature. The list of novel aberrations includes gains of 2q, 6p, 7p, 9q, 11p, 11q, 12p, 12q, 13q, 14q, 17q, 18q, 20p, 20q, 21q, and 22q and losses of 10q21.1.


View this table:
[in this window]
[in a new window]

  		Table 1. Summary of Novel Recurrent Chromosomal Aberrations in SCLC

 
Real-time quantitative PCR was used to confirm these novel recurrent abnormalities. Each locus of interest was screened in three representative lines. The results (Supplementary Fig. S1) show a 94.8% agreement with lines that have amplifications above the 3.0 copy number threshold. Of all 59 gained loci examined, 53 were confirmed in all lines evaluated, in which the remaining loci were confirmed in two of the three lines. The copy number derived from quantitative PCR was slightly higher than that from CGH. This may be attributed to possible signal compression by 100K SNP arrays, as reported by others (18). The results for the four CGH-derived heterozygous deletions were inconclusive, reflecting the difficulty of detecting single copy loss by quantitative PCR. Overall, the quantitative PCR data validate the CGH results and confirm the existence of the novel copy number gains.

Most of the novel aberrations are relatively short (70 kb-3.6 Mb). It is therefore possible that they have not been previously detected because of the lower resolution of the previously used techniques (metaphase CGH, BAC array CGH, and FISH). The mean spacing between the SNPs on the 100K SNP array used in this study is 23.6 kb, thus permitting identification of very short regions of gains and losses. It is also possible that some of the newly detected recurrent copy number changes represent copy number polymorphisms (23). However, this is only a remote possibility because the copy number is determined relative to a panel of 110 normal individuals (24).

It is noteworthy that the novel regions on 7p22.1, 12q14.2, and 20q13.32 contain members of the Ras family [i.e., RAC1 (gains in 69% of lines and 54% of tumors), RASSF3 (65% of lines and 70% of tumors), and RAB22A (42% of lines and 84% of tumors)]. Other novel regions include antiapoptotic genes, such as BI-1, FAIM-2 (12q13.12; gained in 73% of lines and 58% of tumors), and RFC2 (7q11.23; gained in 71% of lines and 60% of tumors). The known oncogene ABL1 on 9q34.1 was amplified in 72% of lines and 54% of tumors, indicating a potential role in tumorigenesis in SCLC.

Identification of Gene Amplifications/Deletions Correlating with the Sensitivity to the Bcl-2 Inhibitor
Members of the Bcl-2 protein family are central regulators of programmed cell death (25). The family members that inhibit apoptosis are overexpressed in cancers and contribute to tumorigenesis (26). A small-molecule inhibitor of Bcl-2, Bcl-XL, and Bcl-w has recently been discovered at Abbott Laboratories and shown to induce regression of solid tumors (27). This compound (ABT-737) displayed selective potency against SCLC and lymphoma cells (27). Here, we tested the compound against our panel of 23 SCLC lines as described in ref. 27 (Supplementary Fig. S2). Sensitive (10 nmol/L < EC50 < 1 µmol/L) and resistant (EC50 > 10 µmol/L) lines were identified. The sensitive group consisted of NCI-H889, NCI-H1963, NCI-H1417, NCI-H146, NCI-H187, DMS53, NCI-H510, NCI-H209, NCI-H526, NCI-H211, NCI-H345, and NCI-H524, and the resistant group included NCI-H82, NCI-H196, SW1271, and H69AR. A structurally distinct Bcl-2 family antagonist yielded a very close inhibition profile for the same 23 cell lines (data not shown), indicating that the observed response is mechanism based.

To identify potential genomic correlates of the sensitivity of SCLC cells to ABT-737, we developed a bioinformatics approach that identifies chromosomal aberrations that discriminate between the sensitive and resistant groups. Our program tested for statistical significance using Fisher's exact test to determine if a SNP shows preferential gain/loss in the sensitive or resistant group. The copy number thresholds for amplifications and deletions were set at 2.8 and 1.5, respectively. Contiguous regions of three or more probe sets (SNPs) with low table and two-sided P values were subjected to further analysis. Two regions (on chromosomes 18 and 19) were identified with P < 0.02 (Supplementary Table S2). One region on chromosome 18q was of particular interest because of high copy numbers in the sensitive cell lines. The increased copy number region starts at position 45,704,096 and ends at position 74,199,087, and spans chromosomal bands from 18q21.1 through 18q23, with the section having the lowest P values (<0.02) located between 18q21.31 and 18q22.3.

We applied real-time quantitative PCR to validate this region as a potential stratification marker derived by CGH. Two different primer sets run in triplicate were used to evaluate seven loci starting at 48 Mb (18q21.1) and ending at 74 Mb (18q23) within chromosome 18. The quantitative PCR results (Fig. 2 ) indicate segregation between the sensitive and resistant lines based on the copy number of the test locus (P < 0.0001, ANOVA), thus confirming the CGH data. The sensitive lines carry an amplification of the region under consideration (3-7 copies), whereas the resistant lines display a normal copy number.


Figure 2
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 2. Real-time quantitative PCR analysis of the copy number for the 18q21-23 marker region in the SCLC cell lines sensitive and resistant to ABT-737.

 
Notably, the Bcl-2 gene (P = 0.015, Fisher's exact test), the target of ABT-737, is located within this discriminant region on 18q21.3, suggesting that the sensitivity of a cell line to the drug may be determined by the amplification status of Bcl-2. Figure 3 illustrates the relationship between the Bcl-2 gene copy number and the sensitivity of the SCLC cell lines. The cell lines are arranged in the order of decreasing sensitivity to the drug, as determined by EC50 values (Supplementary Fig. S2; ref. 27). The copy number for each cell line is presented as the mean ± SD of the copy numbers for 17 SNPs within the Bcl-2 gene. It is clear from the plot that the sensitivity of the SCLC cell lines correlates with the Bcl-2 copy number (R = 0.71, P < 0.0002). The most sensitive lines (H889, H1963, H1417, and H146) have the highest Bcl-2 copy number (4 or 5 copies).


Figure 3
View larger version (34K):
[in this window]
[in a new window]

  		FIGURE 3. Correlation between the Bcl-2 gene copy number and the sensitivity of SCLC cell lines to ABT-737 (R = 0.71, P < 0.0002). The cell lines are arranged in the order of decreasing sensitivity to ABT-737 (increasing EC50).

 
Another apoptosis-related gene (NOXA), whose product promotes degradation of Mcl-1 (28), is located next to Bcl-2 and has a similar copy number profile. There are two outliers in this data set, H187 and H526, which are sensitive but have a normal copy number of the Bcl-2 gene. Their sensitivity to ABT-737 may be attributed to an extra copy of the Bcl-w gene on 14q11.2 (data not shown), which is also a target of the drug. Other genes located in the 18q marker region are listed in Supplementary Table S2. None of these genes seems to be linked to oncogenic processes, making Bcl-2 and NOXA the main candidates for amplification targets in this amplicon. Thus, we established a correlation between the amplification of Bcl-2 and NOXA on 18q21.3 and the sensitivity of SCLC cell lines to ABT-737. This observation is consistent with the mechanism of action of the drug and suggests that the single-agent sensitivity of a cell line to the drug may be determined by the amplification status of 18q21.3.

Analysis of the Bcl-2 Gene Copy Number in SCLC Tumors
To determine the clinical relevance of the 18q21 gain, we analyzed the Bcl-2 copy number in SCLC tumors by FISH. Our analysis included tumors from 62 patients arrayed on a tissue microarray. As can be seen in Fig. 4 , low-level gains of the Bcl-2 gene are present in 40% (25 of 62) of the patients and high-level gains are observed in 8% (5 of 62) of the tumors. This finding is consistent with our data from the SCLC cells, as most changes in the cell lines were also low-level gains. The percentage of lines carrying the aberration was also similar. Thus, we showed that an amplification of the Bcl-2 gene is a frequent event in SCLC, implying that the correlation between the Bcl-2 copy number and the sensitivity of SCLC cells to Bcl-2 family inhibitors can be exploited in clinical trials.


Figure 4
View larger version (14K):
[in this window]
[in a new window]

  		FIGURE 4. Bcl-2 gene amplification in SCLC tumors as determined by FISH.

 
Expression of the Genes Located in the Amplified Region on Chromosome 18
The 18q amplification observed in the sensitive SCLC cell lines is biologically relevant if it leads to higher expression of the genes located in the amplified region. To determine the relative expression of the 18q21-23 genes in the sensitive and resistant SCLC cell lines, we profiled the lines with expression microarrays. In Fig. 5 , the 12 most sensitive cell lines are shown on the left and the 4 resistant lines are grouped on the right (in the order of decreasing sensitivity). Each row of the heatmap corresponds to a gene and each column corresponds to a cell line, whereas the color of each square reflects the relative expression of the gene in a particular cell line. The heatmap covers all genes located in the marker region on 18q21-23 and present on the U133A microarray. A continuum of expression levels is observed between highly sensitive, moderately sensitive, and resistant lines, with the expression levels decreasing with the sensitivity. Because the Bcl-2 gene did not display a clear pattern of decreasing expression with decreasing sensitivity, we examined the expression of the Bcl-2 protein in the sensitive and resistant cell lines. The Bcl-2 protein expression pattern was consistent with the amplification pattern in the sensitive lines and the normal copy number in the resistant lines (29), suggesting that the lower expression in some of the sensitive lines shown by the expression microarray is likely to be an artifact of the microarray probe hybridization. Thus, the gene expression data provide further support for the selection of the 18q21-23 amplification as a stratification biomarker in SCLC. Additionally, this finding implies a significant degree of correlation between gene amplification and gene overexpression.


Figure 5
View larger version (52K):
[in this window]
[in a new window]

  		FIGURE 5. Expression of genes of the 18q21-23 marker region in the SCLC cell lines sensitive and resistant to ABT-737. The lines are arranged in the order of decreasing sensitivity to ABT-737. The gene names are shown on the right. The following genes also belong to the 18q21-23 marker region but are not present on the expression microarray: CCDC11, MAPK4, MRO, STARD6, RAB27B, SE57-1, ONECUT2, ALPK2, SEC11L3, RAX, CPLX4, CCBE1, SDCCAG3, MC4R, CDH20, RNF152, KIAA1468, PHLPP, SERPINB12, SERPINB13, SERPINB4, SERPINB3, SERPINB11, SERPINB7, SERPINB2, SERPINB10, CDH19, TXNDC10, DOK6, CD226, RTTN, SOCS6, GTSCR1, CBLN2, NETO1, FBXO15, CNDP1, ZNF407, ZADH2, SDCCAG3, and GALR1.

 

   Discussion
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
Materials and Methods
 Acknowledgements
 References
 
Classification of cancers based on histopathology does not address the genetic heterogeneity of the disease. The response rates to novel cancer therapeutics can be improved by selecting patients based on the presence or absence of a genomic stratification marker. An early identification of such markers in the drug discovery process can provide a path for more successful clinical trials. Because early discovery is almost always done in cancer cell lines, genomic profiling of the model systems coupled with screening of candidate compounds represents a potentially very powerful strategy in genomic biomarker identification. In this study, we sought to identify genomic biomarkers of sensitivity to a Bcl-2 family inhibitor (ABT-737) through CGH analysis of 23 SCLC cell lines used to test Bcl-2 inhibitors in vitro. The compound is a selective and potent inhibitor of Bcl-2 proteins that exhibits single-agent mechanism-based killing of cells and induces regression of solid tumors (27). We used high-density SNP genotyping arrays for CGH analysis because of their high resolution and reproducibility. Their use in CGH analysis has recently been pioneered in a study of 70 lung cancers (18).

To validate our CGH method, we initially focused on recurrent gene copy number abnormalities in SCLC. A number of recurrent chromosomal aberrations in SCLC have previously been identified by conventional cytogenetics and array CGH. In particular, loss on chromosome 3p is considered the most frequent chromosomal aberration in SCLC (30-34). The fragile histidine triad gene (FHIT), the von Hippel Landau (VHL), and the protein-tyrosine phosphatase-{gamma} gene were suggested as candidate SCLC tumor suppressor genes on 3p (35, 36). Another common region of chromosomal loss is located on 5q (31, 33, 37). Losses on chromosomes 13 (31) and 17p are also common in SCLC, with the frequently lost regions overlapping at 17p13 (34). Studies by CGH have identified additional aberrations in SCLC including losses of 4q, 10q, and 16q and gains of 8q and 19q (36, 38). Frequent gains have been detected on 1p22-32 (MYCL), 2p23-25 (MYCN), 8q24 (MYC), 5p, 1q24, and Xq26, as well as losses on 17p13 (TP53), 13q14 (RBI), 3p, 22q12-13, 10q26, and 16p11.2 (39). Recurrent amplifications of regions containing known oncogenes c-myc and N-myc have been identified in SCLC (40-43). Using a submegabase-resolution tiling CGH array with overlapping BAC clones, Coe et al. (16) have detected a gain of a 350-kbp gene-specific region on 7p22.3 as the most frequent event in a panel of SCLC cell lines. Genome-wide analysis of 5 SCLC cell lines and 19 patient tumors used 100K SNP genotyping arrays similar to the arrays used in this study (18). High-level amplifications have been identified on 8q12-13 and homozygous deletions have been detected on 3q25 and 9p23-24.1.

In our study, we detected all the aforementioned aberrations, thus validating our CGH method. We also identified several novel recurrent copy number abnormalities (i.e., gains of 2q, 6p, 7p, 9q, 11p, 11q, 12p, 12q, 13q, 14q, 17q, 18q, 20p, 20q, 21q, and 22q and losses of 10q21.1). We only focused on frequent aberrations present in multiple cell lines (frequency > 40%) because they are more likely to reflect the oncogenic processes in SCLC. To identify aberrations most relevant to SCLC tumors, we selected only the abnormalities that were also present in a previously published data set on 19 SCLC tumors (18). The novel gains present in both cell lines and tumors were confirmed by quantitative PCR. The final list (Table 1; Supplementary Table S1, worksheet 2) thus contains verified novel recurrent copy number gains characteristic of SCLC. Additional studies will be required to determine the functional role of the genes residing in these regions in the pathogenesis of SCLC.

The key finding of this study is the discovery of a potential genomic stratification marker in SCLC, an amplification on 18q harboring the Bcl-2 and NOXA genes. Overexpression of Bcl-2 in SCLC has previously been shown (44, 45). However, to our knowledge, amplification of 18q21 and, in particular, of the Bcl-2 gene has not been reported in SCLC. We showed that an amplification of a region on 18q correlates with the sensitivity of SCLC cell lines to two novel anticancer drugs targeting Bcl-2 family members. The region contains one of the drug targets, Bcl-2, thus providing a likely explanation for the differential sensitivity of Bcl-2 amplified lines. Additionally, the marker region contains NOXA, a gene regulating apoptosis by promoting degradation of Mcl-1 (28). Amplification of NOXA would enhance degradation of Mcl-1 and thus reduce the function of this antiapoptotic protein not targeted by ABT-737. We have previously shown that Mcl-1 functionally compensates for Bcl-2 and contributes to apoptosis resistance in SCLC cells, whereas NOXA overexpression sensitizes resistant cells to ABT-737 (29). Additionally, two recent studies have shown that overexpression of Mcl-1 induces resistance to ABT-737, whereas neutralization of Mcl-1 by NOXA sensitizes cells to ABT-737, establishing Mcl-1 as a major factor in cellular resistance to ABT-737 (46, 47). In our experiments, there was no correlation between the Mcl-1 copy number and the sensitivity to ABT-737 (a gain of one copy was observed in 4 of the 23 lines profiled), suggesting an alternative mechanism of regulation of Mcl-1 expression. Thus, we observed concurrent gains of Bcl-2 and NOXA in the sensitive but not in resistant lines, which is consistent with the findings that increased levels of NOXA sensitize cells to Bcl-2 antagonists by inactivating Mcl-1.

To investigate the 18q21-23 gain further as a genomic marker, we showed that the genes residing in the region are also overexpressed in the sensitive lines relative to the resistant lines. Next we sought to establish the clinical relevance of the 18q.21-23 gain by studying the copy number of the Bcl-2 gene in SCLC tumors. It was shown that Bcl-2 is amplified in 48% of the patients. To our knowledge, the Bcl-2 gain has not been previously explored in SCLC. Taken together, our data suggest that the 18q21-23 copy number will be a clinically relevant predictor for sensitivity of SCLC to Bcl-2 antagonists as they enter clinical trials. The size of the subgroup of cell lines and patients carrying the 18q21-23 marker (~40%) is comparable to the fraction of lung cancer patients carrying the EGFR amplification (48, 49) or breast cancer patients displaying the Her-2 gain (50, 51), implying that a significant population of SCLC patients would benefit from the Bcl-2 inhibitor, if the correlation seen in vitro reproduces in the clinic. The Bcl-2 gene copy number in clinical samples can be determined by FISH using commercially available probes. In many cases, detection of gene amplification provides superior prognostic and treatment selection information as compared with immunohistochemistry (7). It is noteworthy that the protein product of another gene in the 18q marker region, pro-GRP, is elevated in the serum of some SCLC patients, thus raising the possibility that the levels of pro-GRP in plasma may be used to stratify SCLC patients (52).

It is also possible that the 18q marker may have a broader application in SCLC (and possibly other cancers) as Bcl-2 amplification may correlate with sensitivity to other anticancer therapies. Indeed, Bcl-2 is a key protein regulating cell survival and its amplification may affect the sensitivity of cells to drugs. Indeed, overexpression of Bcl-2 in lung cancer cells induced resistance to cytotoxic agents (53). Furthermore, overexpression of Bcl-2 has been associated with resistance to therapy in ovarian (54, 55), cervical (56, 57), and gastric (58) cancers.

In summary, we used a genome-wide CGH screen to detect novel recurrent amplifications in SCLC and identify a genomic stratification marker for a targeted cancer drug. This work illustrates the potential benefits of CGH analysis of preclinical model systems as an approach to early codiscovery of targeted therapeutics and genomic diagnostics. As Bcl-2 antagonists are entering clinical trials, the discovered sensitivity marker will be used to predict patient response.


   Materials and Methods
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
Acknowledgements
 References
 
Cell Culture and CGH
The following SCLC cell lines were obtained from American Type Culture Collection (Manassas, VA): NCI-H889, NCI-H1963, NCI-H1417, NCI-H146, NCI-H187, DMS53, NCI-H510, NCI-H1209, NCI-H526, NCI-H211, NCI-H345, NCI-H524, NCI-H69, NCI-H748, DMS79, NCI-H711, SHP77, NCI-446, NCI-H1048, NCI-H82, NCI-H196, SW1271, and NCI-H69AR. All cells were cultured in the recommended media at 37°C. Genomic DNA was isolated using a DNAeasy kit (Qiagen) and run on 100K SNP genotyping array sets (Affymetrix). The arrays were run according to the manufacturer's protocol. The raw microarray data files have been loaded into Gene Expression Omnibus (accession no. GSE7068) and Array Express (accession no. E-MEXP-1008). The data were processed using the GTYPE software (Affymetrix, Santa Clara, CA) to create copy number (.cnt) files containing information on the inferred copy number for each probe set (SNP). The .cnt files contained combined information from both arrays in the set. These files were converted into .txt files and loaded into an internally developed software package for further analysis.

This internal program was used for the graphical display and analysis of multiple copy number files. The data were displayed as a histogram of copy number versus SNPs ordered sequentially along the chromosome. For each SNP, the predicted cytogenetic band, as well as any genes between this and the next adjacent SNP, was reported. The gene coordinates and cytogenetic band positions were inferred from the build 35 of the human genome. Data summaries and plots can also be exported to other desktop software. When a region is selected on the histogram, a summary file can be exported that contains the coordinates of all probe sets (SNPs) with the corresponding copy numbers, cytogenetic bands, gene IDs, names, and the coordinates of all the genes residing in the region.

To facilitate identification of recurrent aberrations, the frequency of copy number change is calculated and plotted for each probe set (SNP) after the user defines thresholds for gains (e.g., ≥3 copies) or losses (e.g., ≤1.5 copies). The program also enables identification of copy number abnormalities that correlate with a predefined class label within a data set. The cell lines were classified into sensitive and resistant and a threshold was set for gains (≥2.8 copies) and losses (≤1.5 copies). Fisher's exact test was used to identify aberrations associated with the sensitivity of cell lines to a Bcl-2 inhibitor. For each SNP, a 2 x 2 contingency table was constructed for testing the significance of an increase or a decrease in copy number in the two groups. A minimum of three contiguous SNPs meeting the P value threshold were considered to be one region.

Real-time Quantitative PCR
Primers were designed using Vector NTI (Invitrogen, Carlsbad, CA), purchased from IDT (Coralville, IA), and tested to ensure amplification of single discrete bands (Supplementary Table S3). Real-time PCR was conducted on an iCycler (Bio-Rad, Richmond, CA) using SYBR Green quantitative PCR supermix UDG (Invitrogen). Each reaction was run in triplicate and contained 10 ng of purified genomic DNA and 300 nmol/L of each primer in a final volume of 50 µL. The cycling conditions were as follows: 95°C for 3 min; 35 cycles of 95°C for 10 s; 57°C for 45 s. Melting curves were done to ensure that only a single amplicon was produced and samples were run on a 4% agarose gel (Invitrogen) to confirm specificity. Data analysis was done with the linear regression software DART-PCR v1.0 (59) using raw thermocycler values. Normalization of sample input was conducted with geometric averaging software GeNorm v3.3 (60) to GAPDH, ß2-microglobulin, YWHAZ, RPL13a, and PLP-1. The copy number for each locus was determined by establishing the normalized quantitative PCR output for the sample, dividing this value by the normalized quantitative PCR output for a control genomic DNA (Clontech, Palo Alto, CA), and multiplying the resulting value by 2. Each quantitative PCR copy number estimate is the average value for two independent primer sets (mean coefficient of variation, 11.5%).

FISH
A tissue microarray containing primary SCLC tumors from 62 patients was analyzed by FISH using a dual-color FISH probe targeting 18q21 (LSI Bcl-2 Break-apart probe, Abbott Molecular, Des Plains, IL). The slides were deparaffinized for 10 min in xylol, rinsed in 95% ethanol, air-dried; incubated in a Pretreatment Solution for 15 min at 80°C, rinsed in water; incubated in a Protease Buffer for 2.5 to 5 h, rinsed; dehydrated in 70%, 80%, and 95% ethanol, and air-dried. The probe (10 µL) was applied onto the slide, and the slide was covered, sealed, heated to 72°C for 5 min, and hybridized overnight at 37°C in a wet chamber. The slides were then washed with 2x SSC containing 0.3% NP40 (pH 7-7.5) for 2 min at 75°C, rinsed with water, air-dried, mounted with a 4',6-diamidino-2-phenylindole solution and a 24 x 50 mm coverslip, and examined under an epifluorescent microscope. For each tissue sample, the number of red and green FISH signals (both corresponding to the Bcl-2 locus) was estimated in at least 20 cells. An average copy number per spot was then calculated based on the minimal and maximal numbers of Bcl-2 FISH signals per cell nucleus in each tissue spot. Copy number groups were built based on the absolute gene copy numbers according to the following criteria: 1 to 2 signals, average copy number <2.5; 3 to 4 signals, average copy number ≥2.5 and <4.5; 5 to 6 signals, average copy number ≥4.5 and <6.5; and 7 to 10 signals, average copy number ≥6.5.

Expression Microarray Analysis
Total RNA was purified on RNeasy columns (Qiagen, Valencia, CA). Labeled cRNA was prepared according to the microarray manufacturer's protocol and hybridized to human U133A 2.0 arrays (Affymetrix). Data files were loaded into the Rosetta Resolver software for analysis and the intensity values for all probe sets were normalized using the Resolver Experimental Definition. The intensity values for the probe sets corresponding to genes within the amplified regions were normalized across each gene and compared in heatmaps using Spotfire software.


   Acknowledgements
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
References
 
We thank Drs. Barbara Weir and Matthew Meyerson (Dana-Farber Cancer Institute, Boston, MA) for providing the CGH data for tumors; Ravindra Mamadipaka and Dr. Jeremy Packer for help in writing the CGH data analysis program; Lisa Roberts for help with CGH experiments; Dr. John Wass for statistical support; and Drs. Steven Elmore, Stephen Fesik, and Christin Tse for critical reading of the manuscript.


   Notes
Top
 Notes
Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/).

Received 10/30/06; revised 2/20/07; accepted 2/22/07.


   References
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 

  1. Ramaswamy S, Golub TR. DNA microarrays in clinical oncology. J Clin Oncol 2002;20:1932–41.[Abstract/Free Full Text]
  2. Armstrong SA, Staunton JE, Silverman LB, et al. MLL translocations specify a distinct gene expression profile that distinguishes a unique leukemia. Nat Genet 2002;30:41–7.[CrossRef][Medline]
  3. Cobleigh MA, Vogel CL, Tripathy D, et al. Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J Clin Oncol 1999;17:2639–48.[Abstract/Free Full Text]
  4. Vogel CL, Cobleigh MA, Tripathy D, et al. Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2-overexpressing metastatic breast cancer. J Clin Oncol 2002;20:719–26.[Abstract/Free Full Text]
  5. Ji H, Zhao X, Yuza Y, et al. Epidermal growth factor receptor variant III mutations in lung tumorigenesis and sensitivity to tyrosine kinase inhibitors. Proc Natl Acad Sci U S A 2006;103:7817–22.[Abstract/Free Full Text]
  6. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129–39.[Abstract/Free Full Text]
  7. Pauletti G, Dandekar S, Rong H, et al. Assessment of methods for tissue-based detection of the HER-2/neu alteration in human breast cancer: a direct comparison of fluorescence in situ hybridization and immunohistochemistry. J Clin Oncol 2000;18:3651–64.[Abstract/Free Full Text]
  8. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. CA Cancer J Clin 2006;56:106–30.[Abstract/Free Full Text]
  9. Al-Ajam M, Seymour A, Mooty M, Leaf A. Ten years of disease-free survival between two diagnoses of small-cell lung cancer: a case report and a literature review. Med Oncol 2005;22:89–97.[CrossRef][Medline]
  10. Wakeling AE, Guy SP, Woodburn JR, et al. ZD1839 (Iressa): an orally active inhibitor of epidermal growth factor signaling with potential for cancer therapy. Cancer Res 2002;62:5749–54.[Abstract/Free Full Text]
  11. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497–500.[Abstract/Free Full Text]
  12. Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 2004;101:13306–11.[Abstract/Free Full Text]
  13. Takano T, Ohe Y, Sakamoto H, et al. Epidermal growth factor receptor gene mutations and increased copy numbers predict gefitinib sensitivity in patients with recurrent non-small-cell lung cancer. J Clin Oncol 2005;23:6829–37.[Abstract/Free Full Text]
  14. Tsao MS, Sakurada A, Cutz JC, et al. Erlotinib in lung cancer—molecular and clinical predictors of outcome. N Engl J Med 2005;353:133–44.[Abstract/Free Full Text]
  15. Ashman JN, Brigham J, Cowen ME, et al. Chromosomal alterations in small cell lung cancer revealed by multicolour fluorescence in situ hybridization. Int J Cancer 2002;102:230–6.[Medline]
  16. Coe BP, Lee EH, Chi B, et al. Gain of a region on 7p22.3, containing MAD1L1, is the most frequent event in small-cell lung cancer cell lines. Genes Chromosomes Cancer 2006;45:11–9.[CrossRef][Medline]
  17. Kim YH, Girard L, Giacomini CP, et al. Combined microarray analysis of small cell lung cancer reveals altered apoptotic balance and distinct expression signatures of MYC family gene amplification. Oncogene 2006;25:130–8.[Medline]
  18. Zhao X, Weir BA, LaFramboise T, et al. Homozygous deletions and chromosome amplifications in human lung carcinomas revealed by single nucleotide polymorphism array analysis. Cancer Res 2005;65:5561–70.[Abstract/Free Full Text]
  19. Kallioniemi A, Kallioniemi OP, Sudar D, et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 1992;258:818–21.[Abstract/Free Full Text]
  20. Pinkel D, Segraves R, Sudar D, et al. High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 1998;20:207–11.[CrossRef][Medline]
  21. Pinkel D, Albertson DG. Array comparative genomic hybridization and its applications in cancer. Nat Genet 2005;37 Suppl:S11–7.[CrossRef][Medline]
  22. Pollack JR, Perou CM, Alizadeh AA, et al. Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nat Genet 1999;23:41–6.[Medline]
  23. Sebat J, Lakshmi B, Troge J, et al. Large-scale copy number polymorphism in the human genome. Science 2004;305:525–8.[Abstract/Free Full Text]
  24. Huang J, Wei W, Zhang J, et al. Whole genome DNA copy number changes identified by high density oligonucleotide arrays. Hum Genomics 2004;1:287–99.[Medline]
  25. Danial NN, Korsmeyer SJ. Cell death: critical control points. Cell 2004;116:205–19.[CrossRef][Medline]
  26. Kirkin V, Joos S, Zornig M. The role of Bcl-2 family members in tumorigenesis. Biochim Biophys Acta 2004;1644:229–49.[Medline]
  27. Oltersdorf T, Elmore SW, Shoemaker AR, et al. An inhibitor of Bcl-2 family proteins induces regression of solid tumours. Nature 2005;435:677–81.[CrossRef][Medline]
  28. Willis SN, Chen L, Dewson G, et al. Proapoptotic Bak is sequestered by Mcl-1 and Bcl-xL, but not Bcl-2, until displaced by BH3-only proteins. Genes Dev 2005;19:1294–305.[Abstract/Free Full Text]
  29. Tahir SK, Yang X, Anderson MG, et al. Influence of Bcl-2 family members on the cellular response of small cell lung cancer cell lines to ABT-737. Cancer Res 2007;67:1176–83.[Abstract/Free Full Text]
  30. Whang-Peng J, Kao-Shan CS, Lee EC, et al. Specific chromosome defect associated with human small-cell lung cancer; deletion 3p(14-23). Science 1982;215:181–2.[Abstract/Free Full Text]
  31. Miura I, Graziano SL, Cheng JQ, Doyle LA, Testa JR. Chromosome alterations in human small cell lung cancer: frequent involvement of 5q. Cancer Res 1992;52:1322–8.[Abstract/Free Full Text]
  32. Sozzi G, Bertoglio MG, Borrello MG, et al. Chromosomal abnormalities in a primary small cell lung cancer. Cancer Genet Cytogenet 1987;27:45–50.[Medline]
  33. Petersen I, Langreck H, Wolf G, et al. Small-cell lung cancer is characterized by a high incidence of deletions on chromosomes 3p, 4q, 5q, 10q, 13q and 17p. Br J Cancer 1997;75:79–86.[Medline]
  34. Testa JR, Liu Z, Feder M, et al. Advances in the analysis of chromosome alterations in human lung carcinomas. Cancer Genet Cytogenet 1997;95:20–32.[CrossRef][Medline]
  35. Sozzi G, Veronese ML, Negrini M, et al. The FHIT gene 3p14.2 is abnormal in lung cancer. Cell 1996;85:17–26.[CrossRef][Medline]
  36. Balsara BR, Testa JR. Chromosomal imbalances in human lung cancer. Oncogene 2002;21:6877–83.[CrossRef][Medline]
  37. Morstyn G, Brown J, Novak U, et al. Heterogeneous cytogenetic abnormalities in small cell lung cancer cell lines. Cancer Res 1987;47:3322–7.[Abstract/Free Full Text]
  38. Ried T, Petersen I, Holtgreve-Grez H, et al. Mapping of multiple DNA gains and losses in primary small cell lung carcinomas by comparative genomic hybridization. Cancer Res 1994;54:1801–6.[Abstract/Free Full Text]
  39. Levin NA, Brzoska P, Gupta N, et al. Identification of frequent novel genetic alterations in small cell lung carcinoma. Cancer Res 1994;54:5086–91.[Abstract/Free Full Text]
  40. Nau MM, Brooks BJ, Jr., Carney DN, et al. Human small-cell lung cancers show amplification and expression of the N-myc gene. Proc Natl Acad Sci U S A 1986;83:1092–6.[Abstract/Free Full Text]
  41. Wong AJ, Ruppert JM, Eggleston J, et al. Gene amplification of c-myc and N-myc in small cell carcinoma of the lung. Science 1986;233:461–4.[Abstract/Free Full Text]
  42. Brennan J, O'Connor T, Makuch RW, et al. myc family DNA amplification in 107 tumors and tumor cell lines from patients with small cell lung cancer treated with different combination chemotherapy regimens. Cancer Res 1991;51:1708–12.[Abstract/Free Full Text]
  43. Johnson BE, Makuch RW, Simmons AD, et al. myc family DNA amplification in small cell lung cancer patients' tumors and corresponding cell lines. Cancer Res 1988;48:5163–6.[Abstract/Free Full Text]
  44. Jiang SX, Sato Y, Kuwao S, Kameya T. Expression of bcl-2 oncogene protein is prevalent in small cell lung carcinomas. J Pathol 1995;177:135–8.[CrossRef][Medline]
  45. Yan JJ, Chen FF, Tsai YC, Jin YT. Immunohistochemical detection of Bcl-2 protein in small cell carcinomas. Oncology 1996;53:6–11.[CrossRef][Medline]
  46. van Delft MF, Wei AH, Mason KD, et al. The BH3 mimetic ABT-737 targets selective Bcl-2 proteins and efficiently induces apoptosis via Bak/Bax if Mcl-1 is neutralized. Cancer Cell 2006;10:389–99.[CrossRef][Medline]
  47. Konopleva M, Contractor R, Tsao T, et al. Mechanisms of apoptosis sensitivity and resistance to the BH3 mimetic ABT-737 in acute myeloid leukemia. Cancer Cell 2006;10:375–88.[CrossRef][Medline]
  48. Hirsch FR, Varella-Garcia M, Bunn PA, Jr., et al. Epidermal growth factor receptor in non-small-cell lung carcinomas: correlation between gene copy number and protein expression and impact on prognosis. J Clin Oncol 2003;21:3798–807.[Abstract/Free Full Text]
  49. Hirsch FR, Varella-Garcia M, McCoy J, et al. Increased epidermal growth factor receptor gene copy number detected by fluorescence in situ hybridization associates with increased sensitivity to gefitinib in patients with bronchioloalveolar carcinoma subtypes: a Southwest Oncology Group Study. J Clin Oncol 2005;23:6838–45.[Abstract/Free Full Text]
  50. Slamon DJ, Godolphin W, Jones LA, et al. Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science 1989;244:707–12.[Abstract/Free Full Text]
  51. Slamon DJ, Clark GM, Wong SG, et al. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987;235:177–82.[Abstract/Free Full Text]
  52. Yonemori K, Sumi M, Fujimoto N, et al. Pro-gastrin-releasing peptide as a factor predicting the incidence of brain metastasis in patients with small cell lung carcinoma with limited disease receiving prophylactic cranial irradiation. Cancer 2005;104:811–6.[CrossRef][Medline]
  53. Ohmori T, Podack ER, Nishio K, et al. Apoptosis of lung cancer cells caused by some anti-cancer agents (MMC, CPT-11, ADM) is inhibited by bcl-2. Biochem Biophys Res Commun 1993;192:30–6.[CrossRef][Medline]
  54. Mano Y, Kikuchi Y, Yamamoto K, et al. Bcl-2 as a predictor of chemosensitivity and prognosis in primary epithelial ovarian cancer. Eur J Cancer 1999;35:1214–9.[CrossRef][Medline]
  55. Kassim SK, Ali HS, Sallam MM, et al. Increased bcl-2 expression is associated with primary resistance to chemotherapy in human epithelial ovarian cancer. Clin Biochem 1999;32:333–8.[CrossRef][Medline]
  56. Harima Y, Harima K, Shikata N, et al. Bax and Bcl-2 expressions predict response to radiotherapy in human cervical cancer. J Cancer Res Clin Oncol 1998;124:503–10.[CrossRef][Medline]
  57. Pillai MR, Jayaprakash PG, Nair MK. bcl-2 immunoreactivity but not p53 accumulation associated with tumour response to radiotherapy in cervical carcinoma. J Cancer Res Clin Oncol 1999;125:55–60.[CrossRef][Medline]
  58. Nakata B, Muguruma K, Hirakawa K, et al. Predictive value of Bcl-2 and Bax protein expression for chemotherapeutic effect in gastric cancer. A pilot study. Oncology 1998;55:543–7.[CrossRef][Medline]
  59. Peirson SN, Butler JN, Foster RG. Experimental validation of novel and conventional approaches to quantitative real-time PCR data analysis. Nucleic Acids Res 2003;31:e73.[Abstract/Free Full Text]
  60. Vandesompele J, De Preter K, Pattyn F, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 2002;3:RESEARCH0034.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data, Olejniczak, et al.
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Olejniczak, E. T.
Right arrow Articles by Semizarov, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Olejniczak, E. T.
Right arrow Articles by Semizarov, D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Cell Growth & Differentiation