Sec62 Bridges the Gap from 3Q Amplification to Molecular Cell Biology in Non–Small Cell Lung Cancer

The American Journal of Pathology, Vol. 180, No. 2, February 2012 Copyright © 2012 American Society for Investigative Pathology. Published by Elsevier Inc. All rights reserved. DOI: 10.1016/j.ajpath.2011.10.039 Biomarkers, Genomics, Proteomics, and Gene Regulation

Sec62 Bridges the Gap from 3q Amplification to Molecular Cell Biology in Non–Small Cell Lung Cancer

Maximilian Linxweiler,* Johannes Linxweiler,* Lung cancer is the worldwide leading cause of cancer Monika Barth,* Julia Benedix,* Volker Jung,† mortality, resulting in approximately 1.4 million deaths each ‡ ‡ year.1 Despite much effort to develop new diagnostic,2–4 Yoo-Jin Kim, Rainer M. Bohle, 5,6 7 Richard Zimmermann,* and Markus Greiner* prognostic, and therapeutic strategies, lung cancer mortality remains high,8 with the 5-year survival rate remain- From the Department of Medical Biochemistry and Molecular ing markedly below 20%. However, during the last decade, Biology,* Saarland University, Homburg; and the Department of several key elements were identified that play crucial roles Urology and Pediatric Urology † and General and Surgical in the molecular carcinogenesis of lung cancer,9–11 provid- Pathology,‡ Saarland University Hospital, Homburg, Germany ing promising starting points for future research. Among these, an amplification of the chromosomal region 3q was identified as the most frequent genomic alteration in non–small cell lung cancers (NSCLCs).12,13 The molecular carcinogenesis of lung cancer has yet This amplification seems to be more characteristic of to be clearly elucidated. We investigated the possible squamous cell carcinoma (SCC) than of adenocarcinoma oncogenic function of SEC62 in lung cancer, which (AC)14,15 and has also been observed in numerous other was predicted based on our previous findings that tumor entities.16–19 Therefore, many studies focused on lung and thyroid cancer tissue samples exhibited in- identifying candidate oncogenes within this region, pro- creased Sec62 protein levels. The SEC62 gene locus is posing various genes, including PIK3CA, SOX2, RAP2B, at 3q26.2, and 3q amplification is reportedly the most FXR1, and EVI1,20–23 to be potential driver oncogenes on common genomic alteration in non–small cell lung 3q. However, many of these studies either were limited by cancer. We analyzed SEC62 mRNA and protein levels a low number of cases or lacked confirmation by func- in tissue samples from lung cancer patients by real- tional approaches. time quantitative PCR, Western blot, and IHC and We recently described a 3q gain in prostate cancer found significantly increased SEC62 mRNA and pro- and identified SEC62, a gene encoding for an endoplas- tein levels in tumors compared with tumor-free tissue mic reticulum (ER) transmembrane protein, as the most samples from the same patients. Correlation analyses commonly overexpressed gene within this region based revealed significantly higher Sec62 levels in tumors on increased mRNA levels.24 Further investigations re- with lymph node metastases compared with non- vealed higher Sec62 protein contents in prostate cancer metastatic tumors, as well as in poorly compared with cell lines and in prostate cancer tissue samples.25 In moderately differentiated tumors. On the basis of addition, we demonstrated that Sec62 had an essential these promising results, we examined the role of function in cell migration25 and in ER stress tolerance of Sec62 in cancer cell biology in vitro. Cell migration prostate cancer cells.26 In a multitumor TMA, we found assays with lung and thyroid cancer cells showed distinct stimulation of migration in SEC62-overex- Supported by a HOMFOR grant (R.Z.), grant FOR967 from the Deutsche pressing cells and inhibition of migration in Sec62- Forschungsgemeinschaft (R.Z.), and a research award of the Friends of depleted cells. Moreover, we found that SEC62 silenc- the Saarland University Hospital (M.G). ing sensitized the cells to thapsigargin-induced Accepted for publication October 25, 2011. endoplasmic reticulum stress. Thus, our results indicate M.L. and J.L. contributed equally to this work. that SEC62 represents a potential candidate oncogene in Supplemental material for this article can be found on http://ajp. the amplified 3q region in cases of non–small cell amjpathol.org; or at doi: 10.1016/j.ajpath.2011.10.039. lung cancer and harbors various functions in cancer Address reprint requests to and reprint requests: Markus Greiner, Ph.D., cell biology. (Am J Pathol 2012, 180:473–483; DOI: Department of Medical Biochemistry and Molecular Biology, Saarland Uni- 10.1016/j.ajpath.2011.10.039) versity, 66421 Homburg/Saar, Germany. E-mail: [email protected]. 473 474 Linxweiler et al AJP February 2012, Vol. 180, No. 2

179 of 187 lung cancer samples to be Sec62 positive, nia), 5 ␮g of the yielded total RNA was reverse tran- whereas none of 10 examined normal lung tissue sam- scribed using Superscript 2 RT (Invitrogen, Carlsbad, ples showed Sec62 immunoreactivity.25 On the basis of California) and Oligo(dT)12-18 primers (Invitrogen). The these promising results, in our present study we analyzed resultant cDNA was purified with the PCR Purification Kit the Sec62 mRNA and protein levels in 140 tissue samples (Qiagen, Hilden, Germany) and adjusted to a final con- from 70 NSCLC patients by quantitative PCR (qPCR), centration of 50 ng/␮L. RNA isolation from FFPE macro- Western blot, and immunohistochemistry (IHC). In addi- scopically dissected tumor tissue was performed with tion, we performed in vitro experiments to further examine three to four 10-␮m sections using the RNeasy FFPE kit the functional role of Sec62 in tumor cell biology. (Qiagen). RNA was analyzed for sufficient quality using Agilent RNA 6000 Nano Chips (Agilent Biotechnologies, Materials and Methods Waldbronn, Germany). According to previously published data, the RNA isolated from our samples was ap- 27–29 Cell Culture and Tissue Samples plicable for qPCR (see Supplemental Table S1 at http://ajp.amjpathol.org). Subsequently, reverse tran- Cell lines A549 [Deutsche Sammlung von Mikroorganis- scription was performed using the High Capacity RNA to men und Zellkulturen (German Collection of Microorgan- cDNA kit (Applied Biosystems, Foster City, California), isms and Cell Cultures) (DSMZ) number ACC 107], BC01 and the resultant cDNA was adjusted to a final concen- (kindly provided by G. Unteregger), BHT101 (DSMZ tration of 50 ng/␮L. Real-time qPCR was performed with number ACC 279), ML1 (DSMZ number ACC 464), and TaqMan Gene Expression Assays (Applied Biosystems) HEK293 (DSMZ number ACC 305) were cultured at 37°C for SEC62 (Hs00541815_s1), GUSB (␤-glucuronidase, in Dulbecco’s modified Eagle’s medium (Gibco Invitro- Hs99999908_m1), and ACTB (␤-actin, Hs00357333_g1) gene, Karlsruhe, Germany) containing 10% fetal bovine in a StepOne Plus 96-well system (Applied Biosystems) serum (FBS) (Biochrom, Berlin, Germany) and 1% peni- using 50 ng (fresh-frozen tissue) or 100 ng (FFPE tissue) cillin/streptomycin (PAA, Pasching, Austria) in a humidi- cDNA per well in a 20-␮L total reaction volume. Each fied environment with 5% CO2. The cell line H1299 sample was tested in triplicate. Gene expression was (American Type Culture Collection number CRL-5803D) calculated as SEC62 expression in the tumor relative to was cultured in RPMI 1640 medium (PAA) containing the that in the histologically tumor-free lung tissue of the ⌬⌬ same supplements. Stably transfected cell lines were same patient, using the CT method and GUSB (lung cultured in normal culture medium containing 1% G418 cancer) or ACTB (thyroid cancer) as an internal control.30 (Gibco Invitrogene). To ensure the use of a reliable internal control, we Tissue samples were obtained from 70 lung cancer tested 5 genes that had been previously reported as patients with pathologically confirmed AC or SCC, dis- appropriate internal controls31–33 for their stability in 48 of sected by a pathologist, sorted into cancerous and our fresh frozen lung tissue samples. In addition to GUSB healthy portions, snap frozen, and stored at Ϫ80°C. Ad- and ACTB, we tested POLR2A (RNA polymerase 2, ditional formalin-fixed, paraffin-embedded (FFPE) tissue Hs00172187_m1), GAPDH (glyceraldehyde-3-phosphate- samples of the same resectate were stored at room tem- dehydrogenase, Hs99999905_m1), and TBP (TATA box- perature. Tumor samples, as well as tumor-free samples, binding protein, Hs00427620_m1). GUSB proved to be the from 10 thyroid cancer patients were obtained from sur- best candidate for an internal control in lung tissue because gical FFPE resectates and also stored at room tempera- it showed the most stable expression level (see Supplemen- ture. SEC62 expression in the tumor tissue was com- tal Figure S134 at http://ajp.amjpathol.org). pared with histologically tumor-free tissue of the same patient. Only samples from patients who gave signed informed consent were used, following the guidelines of Protein Quantification of Sec62 in Lung Tissue the local ethics board. and Cell Culture Using Western Blot

Real-Time qPCR From each frozen sample, three 20-␮m sections were suspended in 300 ␮L of lysis buffer (20 mmol/L Tris-HCl, For each fresh-frozen tissue sample, 5-␮m cryostat sec- pH 7.4, 150 mmol/L NaCl, 1.5 mmol/L EDTA, 3% glycerol, tions were made and H&E stained before downstream 1% NP-40) and incubated on ice for 30 minutes. After analyses (qPCR and Western blot) to ensure tumor cell addition of 300 ␮Lof2ϫ sample buffer (60 mmol/L Tris- content above 80% and to exclude excessive contami- HCl, pH 6.8, 2% SDS, 10% glycerine, 5% ␤-mercapto- nation with inflammatory cells. For each FFPE sample, ethanol, 0.05% bromophenol blue), the samples were 3-␮m sections were prepared and H&E stained for the incubated for 5 minutes at 37°C, homogenized in the same purpose. In the cases of lung tissue, 20 (tumor) or presence of glass beads (Sigma Aldrich Chemie GmbH, 40 (tumor-free lung tissue) mg of fresh frozen tissue was Taufkirchen, Germany; diameter between 425 and 600 homogenized in liquid nitrogen under sterile conditions, ␮m) for 30 minutes, and again incubated for 10 minutes and consecutive total RNA isolation was performed using at 56°C. After centrifugation at 16,800 ϫ g for 10 minutes the illustra triplePrep kit (GE Healthcare, Chalfont St. at 4°C, 20, 25, and 30 ␮L of the supernatant were used Giles, England) according to the manufacturer’s instruc- for Western blot analysis using an affinity-purified anti- tions. After purification and concentration with the RNA body directed against the COOH-terminal undecapep- Clean & Concentrator kit (Zymo Research, Irvine, Califor- tide of human Sec62 protein and an affinity-purified poly- SEC62 as a Candidate Oncogene in NSCLC 475 AJP February 2012, Vol. 180, No. 2

clonal mouse anti–␤-actin antibody directed against the Immunofluorescence Staining N-terminus of human ␤-actin (Cell Signaling Technology, Danvers, MA). To quantify Sec62 in lysate from cultured Immunofluorescence double staining was performed on cells, 2 ϫ 105 cells were used for Western blot analysis, 3-␮m paraffin sections of tumor-free lung tissue, AC, and using the described primary antibodies and an anti- SSC. Deparaffinization, rehydration, and blocking of un- GAPDH antibody from rabbit (Santa Cruz Biotechnology specific binding sites were performed as previously de- FL-335). The primary antibodies were visualized with an scribed. The samples were incubated with specific anti- ECL Plex goat-anti-mouse IgG-Cy3 or goat-anti-rabbit IgG- bodies against Sec62 and Sec61␤ and visualized by Cy5 conjugate (GE Healthcare) and the Typhoon-Trio im- secondary antibodies labeled with different fluorescent aging system (GE Healthcare) and analyzed with Image dyes (anti-rabbit fluorescein isothiocyanate or anti-rabbit Quant TL software 7.0 (GE Healthcare). A ratio of Sec62 to Cy3; Dianova, Hamburg, Germany). ␤-actin or GAPDH was determined, respectively. The specificity of the anti-Sec62 antibody was confirmed previously under denaturing and native conditions.26 Silencing of Gene Expression by siRNA For gene silencing, cells were seeded at concentrations of 1.2 ϫ 105 cells per well in 12-well plates or 2.4 ϫ 105 IHC Analysis of Sec62 Protein Levels cells per well in six-well plates. The cells were transfected with SEC62 untranslated region directed small interfering Tissue samples from each of the 70 NSCLC cases (35 AC RNA (siRNA) (5=-CGUAAAGUGUAUUCUGUACdTdT-3=; cases and 35 SCC cases) were previously H&E stained Ambion, Austin, TX), SEC62 siRNA (5=-GGCUGUGGC- and evaluated by a pathologist; representative FFPE CAAGUAUCUUdTdT-3=; Ambion), or control siRNA (All- blocks of the primary tumor resectate were available from Stars Negative Control siRNA; Qiagen) using HiPerFect all 70 cases, and those of directly adjacent histologically Reagent (Qiagen) following the manufacturer’s instruc- tumor-free lung tissue were available from 30 cases. In tions. After 24 hours the medium was changed and the the same way, all 10 thyroid cancer samples were sub- cells were retransfected. Silencing was evaluated using jected to pathologic evaluation. Three of these were Western blot analysis. The maximum silencing effect was pathologically classified as follicular carcinomas and seen 72 hours after the first transfection. seven as papillary carcinomas. Sections (3 ␮m) were transferred onto Superfrost Ultra Plus Microscope Slides (Menzel-Gläser, Braunschweig, Germany) and dried in Plasmid Transfection the incubator at 37°C overnight. ϫ On deparaffinization, heat-induced epitope retrieval For overexpression of plasmid encoded SEC62, 1.2 5 ϫ 5 was performed by microwave treatment in 10 mmol/L 10 cells per well in 12-well plates or 2.4 10 cells per citrate buffer (pH 6.0), and nonspecific protein binding well in six-well plates were transfected with either an sites were blocked by incubation in blocking solution [80 IRES-GFP-SEC62 plasmid or with the IRES-GFP-LV plas- mL of 0.1M Tris-HCl, pH 7.2,3gofbovine serum albumin mid (LV plasmid) as a negative control using FuGene HD (Sigma Aldrich Chemie GmbH), and 20 mL of FBS Reagent (Qiagen) following the manufacturer’s instruc- (Sigma Aldrich Chemie GmbH)] for 30 minutes at room tion. Overexpression of the plasmid encoded gene was temperature. Subsequently, primary antibody incubation evaluated using Western blot analysis. The maximum was performed with a 1:200 solution (diluted in PBS/0.3% overexpression effect was seen 72 hours after plasmid BSA) of a specific Sec62 polyclonal rabbit antibody for 1 transfection. For stable transfection, cells were trans- hour at room temperature. Each staining series included fected as described, and after 72 hours, medium was a positive control and two negative controls (either with- changed to normal culture medium containing 1% G418, out primary antibody or with incubation in a 1:200 solution and cells were further cultured until a selection of the of the preimmune serum). Visualization was performed cultured cells was reached. Then cells were harvested, ␮ ␮ using the DAKO Real Detection System (DAKO, Glostrup, diluted to a density of 1 cell per 100 L, and 100 L per Denmark) according to the manufacturer’s instructions, well was seeded to a 96-well plate in medium containing and slides were counterstained with hematoxylin. Sec62 1% G418. Clones originating from a single cell were immunoreactivity was co-evaluated independently by a selected and analyzed with respect to Sec62 protein reference pathologist for lung diseases, using the Immu- content. noreactive Score (IRS) according to Remmele and 35 Stegner as a well-established and unbiased semiquan- Real-Time Cell Analysis titative validation system. In this system, staining intensity is classified as no staining (0), weak (1), intermediate (2), For the real-time analysis of cell proliferation, the xCEL- or strong (3) staining. The number of stained cells is Ligence SP system (Roche Diagnostics GmbH, Mann- classified as no stained cells (0), below 10% (1), 10% to heim, Germany) was used. Cells were pretreated with 50% (2), 51% to 80% (3), or more than 80% (4). The siRNA as described or left untreated and after 48 hours product of the scores for staining intensity and number of 1.0 ϫ 104 or 2.0 ϫ 104 cells were seeded into a 96-well stained cells is defined as the IRS. Scores of 0 to 2 were ePlate (Roche Diagnostics GmbH) according to the man- valued as negative, 3 to 4 as weak, 6 to 8 as moderate, ufacturer’s instructions. Cell proliferation was monitored and 9 to 12 as strong. for 72 hours. 476 Linxweiler et al AJP February 2012, Vol. 180, No. 2

Migration Potential Analysis Results Cells were pretreated with control siRNA or SEC62 siRNAs, SEC62 mRNA and Protein Levels Are ϫ 4 and a total of 2.5 10 cells were seeded in a 24-well BD Frequently Increased in NSCLC Falcon FluoroBlok system (BD, Franklin Lakes, NJ) in normal medium containing 0.5% FBS. The inserts were To examine the SEC62 expression pattern in NSCLC, we placed in medium with 10% FBS as a chemoattractant. analyzed SEC62 mRNA and protein levels by qPCR, West- After 72 hours, the cells were fixed with methanol and ern blot, and IHC in 140 tissue samples from 70 NSCLC stained with DAPI, and migrating cells were analyzed by patients. The analysis of SEC62 mRNA content revealed in fluorescence microscopy. 60.9% of all cases (P Ͻ 0.05) a more than 1.5-fold increase in SEC62 expression in tumor tissue relative to histologically tumor-free lung tissue of the same patients. We also found Statistical Analysis a distinct difference in SEC62 gene expression between the For statistical analysis, a Mann-Whitney U-test was used two histologic subtypes (Figure 1, A and C). Although with SPSS software, version 17.0 (IBM, Chicago, IL), and 87.1% of SCCs showed a significant SEC62 overexpression XLStat Pro (Addinsoft, New York, NY). P Ͻ 0.05 was (P Ͻ 0.01), this was true for only 36.4% of the analyzed ACs. considered statistically significant. As shown in Figure 1C, most SCCs were characterized by

Figure 1. Analysis of SEC62 mRNA and protein content using qPCR and Western blot. A: Total RNA of tumor tissue and tumor-free tissue derived from patients with confirmed cancer was isolated and reverse transcribed, and SEC62 mRNA content was analyzed by real-time qPCR. The box plots illustrate the obtained Sec62 CT values plotted against the examined tissue types as indicated. A lower CT value indicates an increase in the respective mRNA content. *P Ͻ 0.05, **P Ͻ 0.01, and ***P Ͻ 0.001. Each box represents the range from the first quartile to the third quartile. The median is indicated by a line, and the mean by a point. The lines outside the boxes represent the ranges of each group within the 1.5-fold interquartile range. B: The quantification of Sec62 protein level was performed using Western blot analysis. The amount of Sec62 and ␤-actin (loading control) was quantified and a ratio of Sec62/␤-actin was calculated. The mean values of these ratios were compared between tumor tissue and tumor-free tissue derived from patients with confirmed cancer. Sta- tistical analysis using a Mann-Whitney U-test showed a significant difference between tumor and tumor-free lung tissue. The data were pre- sented as described in A. C: The SEC62 mRNA content was normalized to GUSB as a reference gene, and an n-fold increase in tumor relative to tumor-free tissue (2-⌬⌬CT) was calculated for each patient. SCC and AC patients were grouped with respect to 2-⌬⌬CT value, and the number of patients in each group is shown in the upper panel. Relative Sec62 protein content in tumor compared with tumor-free tissue was calculated for each patient as rSec62 ϭ [Sec62(tumor)/␤- actin(tumor)]/[Sec62(lung)/␤-actin(lung)]. Pa- tients were grouped with respect to rSec62, and values for SCC (white bars) and AC (gray bars) are shown in the lower panel. SEC62 as a Candidate Oncogene in NSCLC 477 AJP February 2012, Vol. 180, No. 2 a more than fivefold increase in relative SEC62 expression number of SEC62 and PIK3CA (located on 3q) with that level, whereas most ACs had a lower than 1.5-fold relative of TOP2B (located on 3p) using GAPDH as a reference SEC62 expression. gene in tumor and normal lung tissue derived from five Next, we analyzed the Sec62 protein levels using patients with (all SCC patients) and five patients with- Western blot (Figure 1B) and found a significantly in- out (four AC patients and one SCC patient) SEC62 creased Sec62 content in tumor tissue compared with overexpression in the tumor by real-time qPCR.36 In histologically tumor-free tissue from the same patients tumor tissue with increased SEC62 mRNA, we found a (P Ͻ 0.0001). Significance was also reached for SCC twofold to fourfold copy number increase of the SEC62 (P Ͻ 0.0002) and AC (P Ͻ 0.02) separately, with a more gene (mean, 2.8-fold), whereas the non–SEC62-over- than 1.6-fold increased Sec62 level in 89% of SCCs and expressing tumors showed no copy number increase 71% of ACs. Corresponding to the qPCR data, most SCCs (mean, 1.2-fold). Also, for PIK3CA, we found a moder- had a relative Sec62 score higher than 3, whereas the ate copy number increase in the SEC62-overexpress- relative Sec62 scores for most ACs ranged from 1.5 to 3. ing tumors (mean, 1.7-fold), which could not be seen Moreover, Sec62 protein content was evaluated by within nonoverexpressing tumors (mean, 1.0-fold). For IHC again, showing a significantly elevated Sec62 pro- TOP2B, neither SEC62 overexpressing nor nonoverex- tein level in all lung tumors (P Ͻ 0.001) and in SCCs (P Ͻ pressing tumors showed an increased copy number, 0.001) and ACs (P Ͻ 0.001) separately (Figure 2). indicating the specificity of this phenomenon for the To investigate the linkage between SEC62 overex- 3q region (see Supplemental Figure S2 at http://ajp. pression and 3q amplification, we compared the copy amjpathol.org).

12 SCC (n=35) AC (n=35) lung (n=30)

10 negave 1 (3%) 5 (14%) 27 (90%) Figure 2. IHC analyses reveal a significant and 8 specific increase of Sec62 protein content in NSCLC tissue. A: IHC and immunofluorescence 6 weak 7 (20%) 13 (37%) 3 (10%) double staining were performed on 3-␮m FFPE IRS tissue sections. The number of patients classified IRS 4 to a specific IRS is indicated by the number of moderate 7 (20%) 12 (35%) 0 (0%) circles and plotted against examined histologic 2 subtypes. IRSs were calculated according to the method of Remmele and Stegner35 (the portion strong 20 (57%) 5 (14%) 0 (0%) 0 of tumor cells showing a positive reaction times the intensity of Sec62 staining). Statistical analy- lung AC SCC sis using a Mann-Whitney U-test revealed a sig- (n=30) (n=35) (n=35) nificant difference among the individual groups. *** *** *P Ͻ 0.05, **P Ͻ 0.01, and ***P Ͻ 0.001. B: On *** the basis of the IRS, samples were subdivided SCC AC peripheral lung bronchial epithelium into groups showing negative (IRS, 0-2), weak C (IRS, 3-4), moderate (IRS, 6-8), or strong (IRS, 9-12) staining. C: Representative images (original magnification, ϫ100) of the Sec62-immunostained and corresponding H&E-stained tissue Sec62 sections for SCC, AC, normal peripheral lung tissue, and normal bronchial epithelium, respectively, are shown. The strong (SCC) or weaker (AC) red staining indicates the Sec62-positive staining. We note that in the bronchial epithe-

HE lium the plasma cells in all our experiments showed strong Sec62 staining (red stained cells) and therefore served as an internal positive control for the immunostaining reaction. D: Paraffin FITC (Sec62) Cy3 (Sec61ß) DAPI (nuclei) merged sections (3 ␮m thick) of tumor-free lung tissue, D AC, and SCC were incubated with specific antibodies against Sec62 and Sec61␤ and visualized by secondary antibodies labeled with different fluorescent dyes. Fluorescein isothiocyanate in the case of Sec62 leads to a green staining (first tumor-free column) and Cy3 in the case of Sec61␤ leads to a red staining (second column). DNA staining with DAPI indicates the cell nuclei (third column). In the fourth column a yellow staining of the cytoplasm indicates a similar content of Sec62 and Sec61␤, whereas the green staining in the bottom row indicates a dominating Sec62 signal. SCC AC 478 Linxweiler et al AJP February 2012, Vol. 180, No. 2

To ensure that our findings were based on a specific Furthermore, poorly differentiated tumors (G3) compared increase of SEC62 mRNA and protein level and not on a with moderately differentiated tumors (G2) were charac- general expansion of the ER, we analyzed the expression terized by higher Sec62 protein levels. Again, this was levels of different subunits of the Sec61 complex, the key true for lung tumors in general (P Ͻ 0.005) and for SCCs element of the ER protein translocation machinery, using separately (P Ͻ 0.001) but not for ACs. Corresponding Western blot, qPCR, and immunofluorescence double correlation analyses of the IHC and qPCR data showed staining (Figure 2D; see also Supplemental Figure S3 at similar tendencies. http://ajp.amjpathol.org). We detected no significant el- evation of the SEC61 expressions in tumor compared with tumor-free samples. Although Sec62 and Sec61␤ Elevated SEC62 mRNA and Protein Levels showed almost equivalent immunofluorescence signals Might Also Be Characteristic Features of in tumor-free lung tissue, the Sec62 signal dominated Thyroid Cancer clearly in ACs and even more in SCCs. This finding indicates that the observed increase in Sec62 protein con- In our previous multitumor TMA, a positive Sec62 reaction tent is not due to a global ER expansion. In summary, was observed in 92% of follicular and papillary thyroid increased SEC62 mRNA and protein levels are common carcinomas, whereas only 19% of thyroid adenomas and 25 in NSCLC, with emphasis on SCC. 14% of normal thyroid tissue were Sec62 positive. Mul- titumor TMAs are sensitive to sampling errors because of the small sizes of tissue specimens; therefore, in the Increase in Sec62 Protein Content Correlates present study we confirmed these results with a detailed with the Occurrence of Lymph Node analysis of 10 thyroid cancer cases (seven with papillary Metastases and Dedifferentiation of Tumor Cells and three with follicular differentiation) using qPCR and IHC (Figure 4). We found no correlation between Sec62 protein and Four cases showed an increased SEC62 mRNA level, mRNA levels and tumor size (T staging) or with patient and six cases showed an elevated Sec62 protein content age or sex. Although not statistically significant, we ob- in cancerous samples compared with tumor-free tissue served a trend toward higher Sec62 levels in advanced samples from the same patient (overexpression Ͼ1.5 for cases by comparing the qPCR and Western blot data qPCR or difference Ͼ2 in IRS for IHC). As in lung cancer, with the clinical stage of disease (data not shown). A elevated Sec62 protein content seemed to correlate with statistically significant correlation was observed between advanced disease because five of the six tumors with increased Sec62 protein level and the appearance of elevated Sec62 protein levels showed characteristic fea- lymph node metastases (Nϩ), compared with nonmetas- tures of invasive growth, including lymph node metasta- tasized tumors (N0), for lung tumors as a whole (P Ͻ ses, invasion of lymph and blood vessels, or excessive 0.05) and for SCCs separately (P Ͻ 0.005; Figure 3). infiltration of the thyroid capsule; these features were not

Figure 3. Sec62 protein level significantly correlates with the occurrence of lymph node metastases and dedifferentiation of tumor cells. The correlation between Sec62 protein levels in the tumor samples with clinical and pathologic data of the included patients was analyzed. *P Ͻ 0.05, **P Ͻ 0.01, and ***P Ͻ 0.001. A: For NSCLC as a whole (P Ͻ 0.05) and SCC separately (P Ͻ 0.005), we found a significantly higher Sec62 protein level, calculated as rSec62 ϭ [Sec62(tumor)/␤-actin(tumor)]/[Sec62(lung)/␤-actin(lung)], in tumors showing lymph node metastases (Nϩ) compared with in nonmetastasized tumors (NϪ). B: Poorly differentiated tumors (G3) showed a markedly higher Sec62 protein content compared with moderately differentiated tumors (G2). Again, this was true of NSCLC as a whole (P Ͻ 0.005) and SCC (P Ͻ 0.001). SEC62 as a Candidate Oncogene in NSCLC 479 AJP February 2012, Vol. 180, No. 2

AB Figure 4. Analysis of SEC62 mRNA and protein content in thyroid cancer tissue by qPCR and IHC. A: IHC was performed on 3-␮m FFPE tissue sections. For each examined case, the IRS was calculated according to the method of Remmele and Stegner35 (the portion of tumor cells showing a positive reaction times the intensity of ∆∆ Sec62 staining). The IRSs of Sec62 immunostained tumor samples (gray bar) and corresponding tumor-free thyroid tissue samples from the same patient (white bar) are depicted. B: Total RNA of tumor tissue and tumor-free tissue derived from 10 patients with confirmed cancer was isolated and reverse transcribed, and SEC62 mRNA content was analyzed by real-time qPCR. C The SEC62 mRNA content was normalized to ACTB as a reference gene, and an n-fold increase in tumor relative to tumor-free tissue (2- ⌬⌬CT) was calculated for each patient. Each sample was analyzed in triplicates. C: Representative images (original magnification, ϫ100) of the Sec62-immunostained and corresponding H&E- stained tumor-containing and tumor-free tissue sections are shown for one case with strong (case 10) and one case with moderate (case 6) Sec62 protein content increase in the tumor, respectively.

exhibited by any of the tumors with normal Sec62 protein levels (G2 and D4 cells) compared with 100% in LV cells. levels. In particular, the most advanced thyroid tumor Indeed, cells with an increased Sec62 protein level according to the mentioned characteristics showed a showed stronger migration in a 72-hour period compared more than fivefold overexpression of SEC62 mRNA and with the control cells. The migration of G2 cells (731% strong staining of all cancer cells in IHC, whereas the Sec62) was stimulated most, followed by D4 cells (350% tumor-free tissue sample almost lacked Sec62 immuno- Sec62) (Figure 5). Analyzing the proliferation behavior of reactivity (Figure 4). Although not as consistent as in the same cells using the xCelligence system, we found NSCLC, these data suggest a pathogenetic role of Sec62 no significant difference. Thus, these results indicate that in the molecular carcinogenesis of a subset of follicular migration is markedly stimulated by SEC62 overexpres- and papillary thyroid carcinomas, which together ac- sion, as was suggested by our finding of a higher Sec62 count for more than 80% of thyroid cancers. level in metastasized compared with nonmetastasized NSCLCs. Moreover, real-time proliferation data clearly showed Migration Potential and ER Stress Tolerance of that Sec62 is essential for the ER stress tolerance of lung Lung and Thyroid Cancer Cells Strongly and thyroid cancer cells because RNA interference– Depend on Sec62 Protein Level based Sec62 depletion led to a markedly reduced proliferation rate up to cell death under the condition of In a next step, we wanted to find out how cancer cells thapsigargin-induced ER stress (see Supplemental Fig- may profit from an increased Sec62 protein level. On the ure S4 at http://ajp.amjpathol.org). basis of the correlation data and previous reports of ER stress being a common finding in cancer cells,37,38 we performed functional cell culture experiments to investi- Discussion gate the functions of Sec62 in cell migration and ER stress tolerance. SEC62 Is a Promising Candidate Oncogene Our previous findings in prostate cancer cell lines within the Frequently Amplified Terminal 3q demonstrated that silencing of SEC62 mRNA led to dis- Region in NSCLCs tinct inhibition of the migration potential.26 We found that SEC62 silencing also inhibited the capability to migrate in The qPCR, IHC, and Western blot analyses of lung can- lung (BC01) and thyroid (BHT101 and ML1) cancer cell cer tissue samples showed significantly increased Sec62 lines without affecting cell proliferation (Figure 5). After protein (P Ͻ 0.0001) and mRNA levels (P Ͻ 0.05) in tumor demonstrating the essential function of SEC62 in cell tissue compared with tumor-free lung tissue from 70 lung migration, we next wanted to test whether an overexpres- cancer patients. Multiple reports described 3q amplifica- sion of SEC62 would stimulate migration of an otherwise tion, particularly within the region from 3q22 to 3qter, as nonmigrating cell. We used human embryonic kidney the most common genomic alteration in NSCLC.12,13 The cells (HEK293) stably transfected with either LV plasmid SEC62 gene is located at 3q26.2, and the results of our as a negative control or an IRES-GFP-SEC62 plasmid and study suggest that SEC62 is a promising candidate on- selected two cell populations with different Sec62 protein cogene within the 3q area. This hypothesis is supported 480 Linxweiler et al AJP February 2012, Vol. 180, No. 2

Figure 5. Effects of silencing and overexpressing the SEC62 gene on migration and proliferation in BC01, BHT101, ML1, and HEK293 cells. BC01, BHT101, and ML1 cells were transfected with SEC62 untranslated region siRNA, SEC62 siRNA, or control siRNA in 24-well plates. After 24 hours transfection was repeated, and after another 24 hours the cells were used for the migration and proliferation assay. For the overexpression experiments, HEK293 cells stably transfected with LV plasmid (control) or IRES- GFP-SEC62 plasmid (two clones with different increase in Sec62 protein content, termed D4 or G2) were used. Migration was tested in the BD Falcon FluoroBlok system in 24-well inserts. Af- ter 72 hours, cells were fixed with methanol, stained with DAPI, and analyzed by fluorescence microscopy. The proliferation of cells was monitored using the xCELLigence system. A: Ef- fects of SEC62 silencing and overexpression on migration. B: Real-time proliferation analysis using the xCELLigence system. C: The amount of Sec62 was measured after SDS-PAGE and West- ern blotting of an aliquot taken from the cells that were seeded on the FluoroBlok inserts and the ePlate96. GAPDH was used as a loading control. All experiments were performed in triplicate.

by the results of other studies revealing 3q amplification Functional Approaches Reveal a Crucial Role of to be particularly characteristic of SCC compared with SEC62 Overexpression in Cancer Cell Biology AC of the lung.14,15 In addition, a comparison of SEC62 overexpressing with nonoverexpressing tumors revealed To find out how Sec62 promotes the neoplastic transfor- a twofold to fourfold copy number increase for SEC62 in mation of human cells, we performed different cell culture the overexpressing tumors but not in the nonoverex- experiments using various cell lines. On the basis of our pressing tumors, indicating a link between 3q amplifica- previous studies with prostate cancer cells, we examined tion and SEC62 overexpression. We also found SEC62 the effects of SEC62 silencing on migration and ER stress mRNA and protein levels to be elevated more often and tolerance of lung and thyroid cancer cell lines. As previ- to a higher degree in SCC than in AC. Furthermore, not ously reported for PC3 cells,26 we found a distinct inhi- only is 3q amplification described in NSCLC, but it also bition of migration potential and ER stress tolerance in seems to be a common feature in various human cancer different human lung and thyroid cancer cell lines after entities, predominantly SCCs,16–19 suggesting that 3q RNA interference–based reduction of the Sec62 protein harbors protooncogenes. level. Moreover, we showed that in agreement with our On the basis of the data yielded by quantitative anal- finding of increased Sec62 content in metastasized tu- ysis of SEC62 mRNA and protein levels in tissue samples, we performed correlation analyses between the Sec62 mors, overexpression of SEC62 clearly stimulated the concentrations and clinical data sets of the study pa- migrative potential of HEK293 cells. tients. We found that significantly increased Sec62 con- On the basis of these results, tumor cells may benefit tent correlated with tumors with lymph node metastases from an increased Sec62 level in terms of a pronounced compared with nonmetastasized tumors, as well as with capability to migrate and thus to metastasize and of a poorly differentiated tumors (G3) compared with moder- higher efficiency to cope with ER stress, which exists at a ately differentiated tumors (G2). These correlations were persistently high level in tumor tissue due to elevated more evident for SCCs than for ACs. Hence, SEC62 over- proliferation and hypoxic metabolic conditions.37–39 The expression does not just seem to be a coincident phe- exact mechanisms by which Sec62 influences these mo- nomenon correlated with the cancerous phenotype but, lecular processes remain unknown. Because Sec62 is on the contrary, appears to play a functional role in the suggested to play a role in protein translocation into the molecular carcinogenesis of lung cancer. ER, one can speculate that the inhibition of cell migration SEC62 as a Candidate Oncogene in NSCLC 481 AJP February 2012, Vol. 180, No. 2 in dependency on lowered Sec62 protein content might toxic effects.44 Because we have shown that increased be due to a reduced biogenesis of specific membrane SEC62 mRNA and protein levels are more frequent in proteins or secretory proteins being involved in cell mi- SCC than in AC, Sec62-targeted therapy would espe- gration. In addition, SEC62 silencing in HeLa cells in our cially be appropriate for the treatment of SCC patients. previous study led to an elevated basal level of cytosolic Almost all recently developed molecular therapies (eg, calcium and to a considerably higher increase of the the inhibition of vascular endothelial growth factor by cytosolic calcium concentration after treatment with thap- bevacizumab,45 epidermal growth factor receptor by ge- sigargin, compared with cells with normal Sec62 levels.25 fitinib,7 and anaplastic lymphoma kinase by crizotinib46) Thus, we suggest that besides migration, Sec62 may are limited to treatment of AC cases, emphasizing the influence various other cellular processes, including ER urgent need for new strategies for the treatment of SCC stress tolerance, apoptosis, and unfolded protein re- patients. In addition, for determining a personalized ther- sponse induction by affecting cellular calcium homeosta- apeutic approach, monitoring Sec62 protein levels could sis. Previously, we induced ER stress in prostate cancer be a powerful predictive marker. Patients with a low cells using either thapsigargin, a SERCA inhibitor, or tu- Sec62 protein level—in the case of lung cancer, mostly nicamycin, which blocks the formation of N-linked glyco- AC patients—would benefit from a therapy using thapsi- proteins. Results showed that Sec62 depletion resulted in gargin or analog agents, which are considered for cancer higher sensitivity to thapsigargin- but not tunicamycin- therapy in recent reports,47,48 whereas tumor cells with induced ER stress.25 Thus, the protective effect of Sec62 high Sec62 protein levels are protected against such seems to be limited to ER stress induced by changes in therapeutics. Interestingly, in this context, several groups the cellular calcium level. Alternatively, Sec62 overpro- reported that a fusion protein of SubA, a bacterial toxin duction in tumor cells may allow for facilitated transloca- that cleaves BiP as the key element of unfolded protein tion of certain precursor proteins into the ER and, there- response, and the epidermal growth factor forms an ef- fore, stimulate cellular activities, such as cell migration fective killer of epidermal growth factor receptor–positive and ER stress response. This proliferative advantage of tumor cells,49,50 which are predominantly found in ade- Sec62 overproduction might be attenuated by drugs, nocarcinoma of the lung. Sec62, as an indicator for ER such as apratoxin A, that inhibit the co-translational trans- stress tolerance of tumor cells, could be useful as a location of proteins into the ER.40 Future research will be predictive marker for such therapeutic approaches and needed to explore the exact mechanisms by which possibly as a therapeutic target. Sec62 influences these processes. All things considered, and with regard to the entire field of molecular oncology, this study has identified one more player in the complex network of tumor cell biology. Sec62 as a Prognostic Marker and Therapeutic We characterized the functional role of Sec62 within car- Target in NSCLC Patients cinogenesis and cancer cell biology and evaluated its relevance as a potential prognostic marker and therapeu- The conclusions of this study indicate that Sec62 is a tic target in lung cancer. promising new candidate for a prognostic marker in NSCLC patients. Other studies showed that 3q amplification is characteristic of the transition from severe dys- Acknowledgments plasia to invasive SCC of the uterine cervix.41,42 Although the data in our study were not statistically significant, we We thank Anika Müller, Eva Schmitt, and Helga Angeli for observed a trend of higher Sec62 levels in more ad- excellent technical assistance. vanced cases by comparing the qPCR and Western blot data with the Union for International Cancer Control staging of the examined tumors. To determine whether Sec62 References can actually serve as a prognostic marker in NSCLC patients, further studies should primarily include larger 1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM: Estimates populations and the clinical follow-up of the patients in- of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J cluded in this study. Cancer 2010, 127:2893–2917 Beyond that, Sec62 might also serve as a therapeutic 2. Lo Russo G, Franchi F, Seminara P: Is CEA better than CYFRA 21-1 target for the treatment of NSCLC patients. The applica- in the monitoring of squamous cell lung cancer progression? Med Princ Pract 2011, 20:200 tion of small RNAs in cancer therapy has been intensively 3. Nisman B, Biran H, Ramu N, Heching N, Barak V, Peretz T: The studied,43 and it is possible that the RNA interference– diagnostic and prognostic value of ProGRP in lung cancer. Antican- based decrease in Sec62 protein level, as described in cer Res 2009, 29:4827–4832 this study, could be used as a therapeutic tool. A reduc- 4. Carpagnano GE, Foschino-Barbaro MP, Resta O, Gramiccioni E, Carpagnano F: Endothelin-1 is increased in the breath condensate of tion in the Sec62 content of tumor cells could inhibit their patients with non-small-cell lung cancer. Oncology 2004, 66:180–184 capability to migrate and thus to metastasize, as well as 5. Paez JG, Jänne PA, Lee JC, Tracy S, Greulich H, Gabriel S, Herman limit their proliferation rate via enhanced sensitization to P, Kaye FJ, Lindeman N, Boggon TJ, Naoki K, Sasaki H, Fujii Y, Eck ER stress. In the case of lung cancer, therapeutic small MJ, Sellers WR, Johnson BE, Meyerson M: EGFR mutations in lung RNA could be applied as an aerosol, thus avoiding re- cancer: correlation with clinical response to gefitinib therapy. Science 2004, 304:1497–1500 ported delivery problems, such as an inefficient siRNA 6. Potti A, Mukherjee S, Petersen R, Dressman HK, Bild A, Koontz J, uptake by the cancer cells and insufficient selectivity of Kratzke R, Watson MA, Kelley M, Ginsburg GS, West M, Harpole DH 482 Linxweiler et al AJP February 2012, Vol. 180, No. 2

JR., Nevins JR: A genomic strategy to refine prognosis in eraly-stage 25. Greiner M, Kreutzer B, Lang S, Jung V, Cavalié A, Unteregger G, non-small-cell lung cancer. N Engl J Med 2006, 355:570–580 Zimmermann R, Wullich B: Sec62 protein level is crucial for the 7. Maemondo M, Inoue A, Kobayashi K, Sugawara S, Oizumi S, Isobe H, ER-Stress tolerance of prostate cancer. Prostate 2011, 71:1074–1083 Gemma A, Harada M, Yoshizawa H, Kinoshita I, Fujita Y, Okinaga S, 26. Greiner M, Kreutzer B, Jung V, Grobholz R, Hasenfus A, Stöhr RF, Hirano H, Yoshimori K, Harada T, Ogura T, Ando M, Miyazawa H, Tornillo L, Dudek J, Stöckle M, Unteregger G, Kamradt J, Wullich B, Tanaka T, Saljo Y, Hagiwara K, Morita S, Nukiwa T: Gefitinib or Zimmermann R: Silencing of the Sec62 gene inhibits migratory and chemotherapy for non-small-cell lung cancer with mutated EGFR. invasive potential of various tumor cells. Int J Cancer 2011, 128:2284– N Engl J Med 2010, 362:2380–2388 2295 8. Bray FI, Weiderpass E: Lung cancer mortality in 36 European 27. Ribeiro-Silva A, Zhang H, Jeffrey SS: RNA extraction from ten year old countries: secular trends and birth cohort patterns by sex and region formalin-fixed paraffin-embedded breast cancer samples: a compar- 1970–2007. Int J Cancer 2009, 126:1454–1466 ison of column purification and magnetic bead-based technologies. 9. Herbst RS, Heymach JV, Lippman SM: Lung cancer. N Engl J Med BMC Mol Biol 2007, 8:118 2008, 359:1367–1380 28. Reis PP, Waldron L, Goswami RS, Xu W, Xuan Y, Perez-Ordonez B, 10. Sato T, Murakumo Y, Hagiwara S, Jijiwa S, Suzuki C, Yatabe Y, Gullane P, Irish J, Jurisica I, Kamel-Reid S: mRNA transcript quanti- Takahashi M: High-level expression of CD109 is frequently detected fication in archival samples using multiplexed, color-coded probes. in lung squamous cell carcinoma. Pathol Int 2007, 57:719–724 BMC Biotechnol 2011, 11:46 11. Moran C: Importance of molecular features of non-small cell lung 29. Zhao H, Bertos NR, Park M: RNA isolation from FFPE tissues: comparing cancer for choice of treatment. Am J Pathol 2011, 178:1940–1948 the performances of three commercial kits. Biochemica 2008, 4:33–34 12. Dehan E, Ben-Dor A, Liao W, Lipson D, Frimer H, Rienstein S, Siman- 30. Schmittgen TD, Livak KJ: Analyzing real-time PCR data by the com-

sky D, Krupsky M, Yaron P, Friedman E, Rechavi G, Periman M, parative CT method. Nat Protoc 2008, 6:1101–1108 Aviram-Goldring A, Izraeli S, Bittner M, Yakhini Z, Kaminski N: Chro- 31. Liu DW, Chen ST, Liu HP: Choice of endogenous control for gene mosomal aberrations and gene expression profiles in non-small cell expression in nonsmall cell lung cancer. Eur Respir J 2005, 26:1002– lung cancer. Lung Cancer 2007, 56:175–184 1008 13. Massion PP, Taflan PM, Jamshedur Rahman SM, Yildiz P, Shyr Y, 32. Saviozzi S, Cordero F, Lo Iacono M, Novello S, Scagliotti GV, Calog- Edgerton ME, Westfall MD, Roberts JR, Pietenpol JA, Carbone DP, ero RA: Selection of suitable reference genes for accurate normal- Gonzalez AL: Significance of p63 amplification and overexpression in ization of gene expression profile studies in non-small cell lung can- lung cancer development and prognosis. Cancer Res 2003, 63: cer. BMC Cancer 2006, 6:200 7113–7121 33. Zhang F, Wang ZM, Liu HY, Bai Y, Wei S, Li Y, Wang M, Chen J, Zhou 14. Björkvist AM, Husgafvel-Pursiainen K, Anttila S, Karjalainen A, Tam- QH: Application of RT-PCR in formalin-fixed and paraffin-embedded milehto L, Mattson K, Vainio H, Knuutila S: DNA gains in 3q occur lung cancer tissues. Acta Pharmacol Sin 2010, 31:111–117 frequently in squamous cell carcinoma of the lung, but not in adeno- 34. Ginzinger DG: Gene quantification using real-time quantitative PCR: carcinoma. Genes Chrom Cancer 1998, 22:79–82 an emerging technology hits the mainstream. Exp Hematol 2002, 15. Chujo M, Noguchi T, Miura T, Arinaga M, Uchida Y, Tagawa Y: 30:503–512 Comparative genomic hybridization analysis detected frequent over- 35. Remmele W, Stegner HE: Recommendation for uniform definition of an representation of chromosome 3q in squamous cell carcinoma of the immunoreactive score (IRS) for immunohistochemical estrogen receptor lung. Lung Cancer 2002, 38:23–29 detection (ER-ICA) in breast cancer tissue. Pathologe 1987, 8:138–140 16. Caraway NP, Khanna A, Dawlett M, Guo M, Guo N, Lin E, Katz RL: 36. Kindich R, Florl AR, Jung V, Engers R, Müller M, Schulz WA, Wullich B: Gain of the 3q26 region in cervicovaginal liquid-based pap prepara- Application of a modified real-time PCR technique for relative gene copy tions is associated with squamous intraepithelial lesions and squa- number quantification to the determination of a relationship between mous cell carcinoma. Gynecol Oncol 2008, 110:37–42 NKX3.1 loss and MYC gain in prostate cancer. Clin Chem 2005, 51:649– 17. Mori Y, Sato F, Selaru FM, Olaru A, Perry K, Kimos MC, Tamura G, 652 Matsubara N,Wang S, Xu Y, Yin J, Zou TT, Leggett B, Young J, 37. Lee AS, Hendershot LM: ER stress and cancer. Cancer Biol Ther Nukiwa T, Stine OC, Abraham JM, Shibata D, Meltzer SJ: Instabilo- 2006, 5:721–722 typing reveals unique mutational spectra in microsatellite-unstable 38. Koumenis C: ER stress, hypoxia tolerance and tumor progression. gastric cancers. Cancer Res 2002, 62:3641–3645 Curr Mol Med 2006, 6:55–69 18. Chang YC, Yeh KT, Liu TC, Chang JG: Molecular cytogenetic char- 39. Bi M, Naczki C, Koritzimsky M, Fels D, Blais J, Hu N, Harding H, acterization of esophageal cancer detected by comparative genomic Novoa I, Varia M, Raleigh J, Scheuner D, Kaufman RJ, Bell J, Ron D, hybridization. J Clin Lab Anal 2010, 24:167–174 Wouters BG, Koumenis C: ER stress-regulated translation increases 19. Sheu JJC, Lee CH, Ko JY, Tsao GSW, Wu CC, Fang CY, Tsai FJ, Hua tolerance to extreme hypoxia and promotes tumor growth. EMBO J CH, Chen CL, Chen JY: Chromosome 3p12.3-p14.2 and 3q26.2- 2005, 24:3470–3481 q26.32 are genomic markers for prognosis of advanced nasopharyn- 40. Liu Y, Law BK, and Luesch H: Apratoxin A reversibly inhibits the geal carcinoma. Cancer Epidemiol Biomarkers Prev 2009, 18:2709– secretory pathway by preventing cotranslational translocation. Mol 2716 Pharmacol 2009, 76:91–104 20. McCaughan F, Pole JCM, Bankier AT, Konfortov BA, Carroll B, Falzon 41. Heselmeyer K, Schröck E, Du Manoir S, Blegen H, Shah K, Steinbeck M, Rabbitts TH, George PJ, Dear PH, Rabbitts PH: Progressive 3q R, Auer G, Ried T: Gain of chromosome 3q defines the transition from amplification consistently targets SOX2 in preinvasive squamous lung severe dysplasia to invasive carcinoma of the uterine cervix. Proc cancer. Am J Respir Crit Care Med 2010, 182:83–91 Natl Acad SciUSA1996, 93:479–484 21. Comtesse N, Keller A, Diesinger I, Bauer C, Kayser K, Huwer H, 42. Jalali GR, Herzog TJ, Dziura B, Walat R, Kilpatrick MW: Amplification Lenhof HP, Meese E: Frequent overexpression of the genes FXR1. of the chromosome 3q26 region shows high negative predictive value CLAMP1 and EIF4G located on amplicon 3q26-27 in squamous cell for non-malignant transformation of LSIL cytologic finding. Am J carcinoma of the lung Int J Cancer 2007, 120:2538–2544 Obstet Gynecol 2010, 202:581.e1-5 22. Massion PP, Kuo WL, Stokoe D, Olshen AB, Treseler PA, Chin K, 43. Christie RJ, Nishiyama N, Kataoka K: Delivering the code: polyplex Chen C, Polikoff D, Jain AN, Finkel D, Albertson DG, Jablons DM, carriers for deoxyribonucleic acid and ribonucleic acid interference Gray JW: Genomic copy number analysis of non-small cell lung therapies. Endocrinology 2009, 151:466–473 cancer using array comparative genomic hybridization: implications 44. Bonetta L: RNA-based therapeutics: ready for delivery? Cell 2009, of the phosphatidylinositol 3-kinase pathway. Cancer Res 2002, 62: 136:581–584 3636–3640 45. Crinò L, Dansin E, Garrido P, Griesinger F, Laskin J, Pavlakis N, 23. Qian J, Massion PP: Role of chromosome 3q amplification in lung Stroikovski D, Thatcher N, Tsai CM, Wu YL, Zhou C: Safety and cancer. J Thorac Oncol 2008, 3:212–215 efficacy of first-line bevacizumab-based therapy in advanced non- 24. Jung V, Kindich R, Kamradt J, Jung M, Müller M, Schulz WA, Engers squamous non-small cell lung cancer (SAiL. MO19390): a phase 4 R, Unteregger G, Stöckle M, Zimmermann R, Wullich B: Genomic and study Lancet Oncol 2010, 11:733–740 expression analysis of the 3q25-q26 amplification unit reveals 46. Kwak EL, Bang YJ, Camidge DR, Shaw AT, Solomon B, Maki RG, Ou TLOC1/SEC62 as a probable target gene in prostate cancer. Mol SH, Dezube BJ, Janne PA, Costa DB, Varella-Garcia M, Kim WH, Cancer Res 2006, 4:169–176 Lynch TJ, Fidias P, Stubbs H, Engelman JA, Sequist LV, Tan W, SEC62 as a Candidate Oncogene in NSCLC 483 AJP February 2012, Vol. 180, No. 2

Gandhi L, Mino-Kenudson M, Wei GC, Shreeve SM, Ratain MJ, phobicity and orientation of side chains for inactivation of sarcoplasmic Settleman J, Christensen JG, Haber DA, Wilner K, Salgia R, Shapiro reticulum Ca2ϩ-ATPase with thapsigargin and thapsigargin analogs. GI, Clark JW, Iafrate AJ: Anaplastic lymphoma kinase inhibition in J Biol Chem 2010, 285:28883–28892 non-small-cell lung cancer. N Engl J Med 2010, 363:1693–1703 49. Backer JM, Krivoshein AV, Hamby CV, Pizzonia J, Gilbert KS, Ray YS, 47. Ghantous A, Gali-Muhtasib H, Vuorela H, Salina NA, Darwiche N: Paton AW, Paton JC, Backer MV: Chaperone-targeting cytotoxin and What made sesquiterpene lactones reach cancer clinical trials? Drug endoplasmic reticulum stress-inducing drug synergize to kill cancer Discov Today 2010, 15:668–678 cells. Neoplasia 2009, 11:1165–1173 48. Winther AM, Liu H, Sonntag Y, Oelsen C, le Maire M, Soehoel H, 50. Backer MV, Backer JM, Chinnaiyan P: Targeting the unfolded protein Olsen CE, Christensen SB, Nissen P, Moller JV: Critical roles of hydro- response in cancer therapy. Methods Enzymol 2011, 491:37–56