A combination of two electrophoretical approaches for detailed proteome-based characterization of SCLC subtypes.

Context: Small cell lung cancers (SCLC) are heterogeneous and tumours differ in growth characteristics and treatment resistance. Objective: To get insight into the underlying protein profiles responsible for this heterogeneity, two subtypes of SCLC cells mutually differing in chemo resistance properties and growth characteristics are analysed. Materials and methods: Two different electrophoresis approaches in combination with mass spectrometry were used to detect differences between the SCLC cell lines GLC1 and GLC1M13: IEF/SDS-PAGE as well as cetyltrimethylammonium bromide (CTAB)-SDS-PAGE. Results: Altogether 60 non redundant differentially expressed proteins were found of which 5 were verified by Western Blot analysis. Discussion: Most of these proteins identified are involved in processes of tumour progression. Therefore, these proteins are interesting candidates for further functional analysis. Conclusion: Additional CTAB-SDS page is a complementary method to IEF-SDS page revealing a complete new subset of proteins differentially expressed between GLC1 and GLC1 M13 cells SCLC subtypes.

Keywords: Cancer; cells; chemoresistance; DIGE; lung; SCLC

Introduction

Lung cancer is one of the most common cancer types accounting for approximately 17% of all cancer deaths. Worldwide, the annual mortality for lung cancer is more than 1.2 million. (Shibuya et al., [56]). Most patients have a poor prognosis, the median survival time being less than 20 months (Etzel et al., [19]). The occurrence of lung cancer is greatly influenced by risk factors, the most important one being tobacco smoking which is responsible for about 70% of all lung cancer (Danaei et al., [15]). Besides applying a prevention strategy, the development of tools for early diagnosis is desirable to decrease lung cancer related mortality. Great efforts have been made to establish radiological based screenings (Pastorino, [42]) and also in the field of biomarker discovery some work has been done (Schneider, [53]; Poschmann et al., [46]), but progress in these fields is quite modest. One reason for the difficulties in finding sensitive and specific tumour markers is the heterogeneity of lung tumours, resulting in different approaches for their classification. Normally they are classified in two major subtypes: non small cell lung carcinoma (NSCLC) and small cell lung carcinoma (SCLC). According to histological criteria NSCLCs could be further subdivided into adenocarcinoma, squamous cell carcinoma, large cell carcinoma and mixed types (Brambilla, [5]). But this classification suffers from high variability within the tumour stages and therefore the two main categories SCLC and NSCLC are still the gold standard for treatment decisions and prognosis. NSCLC is not chemosensitive and commonly treated by surgery, whereas SCLC tumours exhibit an aggressive phenotype susceptible to chemo- and radiotherapy. Historically lung tumours have also been divided into neuroendocrine (including typical and atypical carcinoid, large cell neuroendocrine carcinoma and SCLC) and not neuroendocrine tumours whereas the majority of neuroendocrine tumours are SCLCs (Travis et al., [59]; Franks & Galvin, [21]).

SCLCs represent 15 to 20% of all lung cancer cases (Yesner, [68]). Cell lines derived from these tumours could be classified in "classic" and "variant" forms. Here the classification is done according to morphological and biochemical properties of the cells. The classical type of cells grows in vitro as tightly packed floating cellular aggregates; exhibit a long doubling time and low colony-forming efficiency. In contrast the variant form differs either in their biochemical or their morphological properties and resembles undifferentiated carcinoma cells growing in vitro as loosely attached floating aggregates. The variant form exhibits often a relatively short doubling time and high colony forming efficiency. This phenotype might reflect changes that had occurred in the parental tumours from which they were derived. These changes might lead to an increased radio-resistance and more aggressive behaviour of the tumours (Gazdar et al., [24]).

Therefore it is interesting to investigate the cellular changes leading to the two different phenotypes on a molecular level. By a better understanding of tumour biology, including their genomic and proteomic features, it might be possible to distinguish better between different tumour subtypes or influence tumours' behaviour in a targeted fashion.

As SCLC tumours in most cases are not removed by surgery, the analysis of primary material is an issue (Inoue et al., [32]). Therefore we chose a cell model of variant and classic SCLC cell lines and compared their proteomes by 2D-differential in gel electrophoresis (2D-DIGE) analysis, a very suitable method even for small sample amounts (Poschmann et al., [46], [45]).

In our study we compared the protein expression profiles of the SCLC cell line GLC1, classified as a variant SCLC cell line and its sub-clone: GLC-1-M13, classified as a classic phenotype. We evaluate if our proteome-based analysis, which is a combination of two electrophoretical approaches, can be used to perform a detailed biochemical characterisation and identification of proteins expressed in specific lung cancer subtypes. Furthermore, upon further study of these differentially expressed proteins involvement in growth characteristics and treatment resistance pathways can be clarified.

GLC1 cells exhibit a phenotype of a poorly differentiated tumour cell and GLC-1-M13 cells show a simple epithelium-like behaviour (de Leij et al., 1985). The GLC1 cell line was established from a pleural effusion of a 59-year-old man presenting a tumour in the hilum of the left lung with sub-pleural metastases in the left lung and bone marrow. Clone GLC1 M13 was established by limited dilution of GLC1 cells in conditioned serum-free medium and exhibited morphological differences from the parent cell line. GLC1 cells grow in loosely packed floating aggregates whereas GLC1 M13 cells were tightly packed. Both cell lines induce tumours after s.c. injection in nude mice (Brouwer et al., [10]; de Leij et al., [16]).

Material and methods

Cell culture

The human small cell lung cancer cell lines GLC1 M13 and GLC1 (de Leij et al., 1985) were cultured in RPMI-1640 (Invitrogen, Karlsruhe, Germany) supplemented with 10% foetal bovine serum in the presence of gentamycin at 37°C in a humidified chamber with 5% CO2. GLC1 cells were washed with PBS and harvested at 80% confluence 1 day after passaging; GLC1 M13 cells were harvested 1 day after passaging and also washed using PBS.

Experimental design 2 D DIGE

For both applied methods, 2D-IEF/SDS-PAGE and 2D-CTAB/SDS-PAGE cells, as well as each different sample, from independent culture dishes or flasks were used. Cells from a total of 10 dishes (GLC1) and 10 flasks (GLC1 M13) were harvested and used for sample preparation. Four independent samples of each cell line were used for IEF/SDS-PAGE and six independent samples of each cell line prepared for CTAB/SDS-PAGE. As three different samples: the internal standard as well as one samples of each cell line were co-separated in a 2D-Gel a total of four IEF/SDS-PAGE gels and six CTAB/SDS-PAGE gels were carried out and analysed.

Sample preparation and protein labelling for 2 D DIGE

For 2D-IEF/SDS-PAGE cell pellets (∼100 mg wet weight) were homogenized with 2.4 µl/µg cell pellet lysis buffer (30 mM TrisHCl; 2 M thiourea; 7 M urea; 4% CHAPS, pH 8.5). Following sonication (6 × 10 s pulses on ice) and removal of cell debris (centrifugation at 16 000 g) protein concentrations were determined using the Bio-Rad Protein Assay (Bio-Rad Laboratories, Munich, Germany).

An enrichment of membrane proteins has been performed for samples separated by CTAB/SDS-PAGE: about 60 mg cells/sample were homogenized using a glass/Teflon homogenizer in 0.5 ml of homogenization buffer (10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 200 mM sucrose, Complete protease inhibitor (Roche Diagnostics, Mannheim, Germany)). After centrifugation at 10 000 g the resulting pellet was homogenized with 0.5 ml homogenization buffer and the sample again subjected to centrifugation at 10 000 g. The supernatants were combined and proteins harvested by ultracentrifugation at 100 000 g for 1.15 h (Optima TLX Ultracentrifuge, Beckman Coulter, Krefeld, Germany). Using 25 mM sodium carbonate the pellet was washed and again suspected to ultracentrifugation. The resulting pellet has been solubilised using 60 µl CTAB buffer (30 mM Tris; 5 M Urea, 6% CTAB, 2% CHAPS pH 8.5, 0.00005% Pyronin Y) for at least 1 h at 4 °C. A determination of protein concentration was carried out using the method of Popov et al. ([44]).

For protein labelling stock cyanine dyes (GE-Healthcare, Freiburg, Germany) were diluted in anhydrous DMF (Sigma-Aldrich) to 400 pmol/µl, and 400 pmol dye was added to 50 µg protein lysate. The sample was briefly centrifuged and left on ice for 30 min in the dark. The labeling reaction was stopped by adding 1 µl of 10 mM L-lysine per 400 pmol dye. After vortexing and centrifugation, the sample was left on ice for 10 min.

IEF/SDS-PAGE

Carrier ampholyte based IEF was performed in a self-made IEF chamber using tube gels (20 cm × 1.5 mm) as described elsewhere (Klose & Kobalz, [34]). Briefly, after running a 21.25 h voltage gradient, the ejected tube gels were incubated in equilibration buffer (125 mM Tris, 40% (w/v) glycerol, 3% (w/v) SDS, 65 mM DTT, pH 6.8) for 10 min. The second dimension was performed in a Desaphor VA 300 system using polyacrylamide gels (15.2% total acrylamide, 1.3% bisacrylamide) (Klose & Kobalz, [34]). Therefore, the IEF tube gels were placed onto the polyacrylamide gels (20 cm × 30 cm × 1.5 mm) and fixed using 1.0% (w/v) agarose containing 0.01% (w/v) bromophenol blue dye.

CTAB/SDS-PAGE

The used two dimensional CTAB/SDS-PAGE separation of proteins has been previously described (Helling et al., [30]) and carried out with some slight modifications. A 1st dimensional protein separation was performed in 1.5 mm tube gels with a 10 mm stacking gel and a 210 mm separating gel. The electrode buffer solution comprised of 150 mM glycine, 0.25% (w/v) CTAB and 0.29% (w/v) H3PO4. To ensure sufficient solubilization of the proteins under these conditions, the electrophoretic separation was performed at 30°C in a water thermostat heated self-made chamber. For electrophoretic protein separation according to Klose (Klose & Kobalz, [34]), a constant voltage of 50 V was initially applied for 10 min, followed by 100 V for 20 min and finally 400 V was applied until the dye front of the pyronin Y reached a distance of 5 mm from the end of the gel.

Scanning and image analysis

After 2-DE, the gels were left between the glass plates and images were acquired using the Typhoon 9400 scanner (GE Healthcare, Freiburg, Germany). Excitation wavelengths and emission filters were chosen specific for each of the CyDyes according to the Typhoon user guide. Before image analysis with the DeCyder 2D 6.5 software (GE Healthcare) the images were cropped with ImageQuantTM software (GE Healthcare, Freiburg, Germany). The intra-gel spot detection and quantification were performed using the Differential In-gel Analysis (DIA) mode of the DeCyder software. The estimated number of spots was set to 3000. An exclusion filter was applied to remove spots with a slope greater than 1.6.

The software algorithm first applies a normalization procedure resulting in normalized spot volumes for each spot map. Based on these normal volumes, standardization was performed by building ratios between the Cy5 (Cy3) channel and Cy2 channel (internal standard) of each spot triple. The resulting normalized and standardized spot volumes were used for further calculations: the mean volumes of matched spots were calculated for each group (GLC1, GLC1 M13) and provided the basis for building spot volume ratios. If this number was less than one the negative reciprocal is listed (e.g. 0.5 is reported as −2 fold change). To identify protein spots exhibiting differential abundance between GLC1 and GLC1M3 cells, besides calculating conventional Student t-tests on log transformed data the Significance Analysis of Micorarrays (SAM) algorithm (Tusher et al., [60]) implemented in Perseus version 1.2.7.4 (Max Planck Institute of Biochemistry, Planegg, Germany) was used (false discovery rate threshold: 0.05, the constant S0 was 0.8). The algorithm accounts both for the change of protein abundance and standard deviation of measurements. Only protein spots found in a minimum of four independent gels were considered for statistical analysis.

In-gel tryptic digestion and protein identification using MALDI-TOF-MS

Directly after gel scanning the spots of interest were manually isolated from the preparative gel and in-gel digested with trypsin (Promega, Mannheim, Germany) in 10 mM ammonium bicarbonate buffer (pH 7.8) at 37°C overnight (Schaefer et al., [52], [51]).

For MALDI MS analyses, tryptic peptides were extracted from the gel matrix (Schaefer et al., [51]) and prepared on the MALDI target using the AnchorChipTM technology (Bruker Daltonics, Bremen, Germany) according to the manufacturer's instructions with α-cyano-4-hydroxycinnamic acid as MALDI-matrix. MALDI-TOF-MS analyses were performed on an UltraFlexTM II (Bruker Daltonics) instrument according to the manufacturer's instructions. The instrument was equipped with a ScoutTM MTP MALDI target. The spectra were acquired in the positive ion mode according to the settings given by the manufacturer. For external calibration, a peptide standard (m/z 757.399, 1296.684, 1619.822, 2093.086 and 3147.471) was used. The MALDI-PMF spectra were processed using the FlexAnalysis™ 2.4 software (Bruker Daltonics) and converted in the.xml format. For peak detection, the spectra were subjected to an internal recalibration using 13 different mono-isotopic masses from autolysis products of trypsin and fragments of keratins ranging from m/z 842.509 – 2825.406. Following parameters were applied: snap peak detection algorithm, signal to noise threshold of 6, maximal number of peaks 100, quality factor threshold 50 and baseline subtraction TopHat. The generated mass lists were subsequently sent to ProteinScapeTM 1.3 (Bruker Daltonics, Bremen, Germany), triggering database searches using ProFound (Version 2002.03.01, Proteometrics LLC) (Zhang & Chait, [70]) and MASCOT (Version 2.3.02, Matrix Science, London, UK) (Perkins et al., [43]).

The following search parameters were selected: fixed cysteine modification with propionamide, variable modification due to methionine oxidation, one maximal missed cleavage sites in case of incomplete trypsin hydrolysis and no details about 2-DE derived protein mass and pI. Using the Score booster function of ProteinScapeTM the mass lists were recalibrated and background masses removed using a list containing 44 masses occurring in a minimum of 10% of generated peak lists (Supplemental Table 1). The database searches were run with a mass tolerance of 40 ppm using UniProt's human complete proteome set (downloaded on 26.10.2012) containing 68.109 protein entries. The used database is a composite database consisting of the UniProtKB entries and a duplicate of the same database, in which the amino acid sequence of each protein entry was randomly shuffled (Stephan et al., [57]). Proteins reaching Profound score >1.5 or Mascot score >64 were considered as identified. Using these criteria one decoy database entry was found by the search engines indicating high confidence of protein identifications. If several database entries of homologues proteins matched these criteria only the entry with the highest score was reported.

Table 1. Differentially abundant protein spots revealed by comparing GLC1 and GLC1 M13 cells using a 2D-DIGE approach. The t-test and fold change values for individual spots result from the analysis of the 2D-DIGE images and are an indicator for differential regulation. Fold change values were calculated from the mean values of the normalized and standardized spot volumes of each group (GLC1, GLC1 M13). Positive values indicate a higher spot abundance in GLC1 M13 cells whereas negative values indicate a higher spot abundance in GLC1 cells. Furthermore, the results of database searches of MS-experiment data are given: the accession numbers of UniProtKB and proteins' GeneBank names as well as sequence coverage information of identified peptides, the score values of the used search algorithms (Profound, Mascot) as well as a combined score referred as Meta Score. (A) Results from IEF/SDS-PAGE analysis and (B) results from CTAB/SDS-PAGE analysis. N indicates the number of gel matches for the corresponding spot.

(A) Spot ID	t-test value	Fold change	t-test difference	Accession	Protein	Gene	Seq. Cov. [%]	Profound Z value	Mascot score	Meta score
121	0.033	1.58	−0.653394	Q92598	Heat shock protein 105 kDa	HSPH1	18.64	2.21	115	98.21
131	0.0048	1.79	−0.855679	P22314	Ubiquitin-like modifier-activating enzyme 1	UBA1	22.11	2.40	179	98.77
174	0.006	2.20	−1.18722	O95757	Heat shock 70 kDa protein 4L	HSPA4L	13.58	2.36	115	98.31
238	0.028	1.80	−0.89475	Q99798	Aconitate hydratase, mitochondrial	ACO2	19.48	2.37	177	98.74
268	0.000058	−3.81	1.95099	Q13409	Cytoplasmic dynein 1 intermediate chain 2	DYNC1I2	21.47	2.39	103	98.25
274	0.003	2.74	−1.45301	Q92945	Far upstream element-binding protein 2	KHSRP	31.08	2.06	109	98.07
316	0.00058	1.67	−0.753303	P17812	CTP synthase 1	CTPS1	26.73	2.38	180	98.77
450	0.00066	1.68	−0.757109	P14866	Heterogeneous nuclear ribonucleoprotein L	HNRNPL	23.42	1.54	65.5	94.26
485	0.0045	1.65	−0.74223	Q03252	Lamin-B2	LMNB2	53.00	2.42	451	100.64
495	0.0027	−2.85	1.51018	Q6UYC3	Lamin A/C	LMNA	21.00	2.18	150	98.43
497	0.0055	−1.66	0.753611	P35520	Cystathionine beta-synthase	CBS	50.09	2.31	208	98.91
497	0.0055	−1.66	0.753611	P31939	Bifunctional purine biosynthesis protein PURH	ATIC	21.28	−0.37	78.4	47.95
502	0.00063	3.94	−2.06913	P17987	T-complex protein 1 subunit alpha	TCP1	39.02	2.29	272	99.34
503	0.00044	4.06	−2.04966	Q13177	Serine/threonine-protein kinase PAK 2	PAK2	43.32	2.24	130	98.33
503	0.00044	4.06	−2.04966	P17987	T-complex protein 1 subunit alpha	TCP1	28.95	−0.40	68.7	45.76
644	0.0078	1.56	−0.644553	B4DT35	Nucleoporin p54	NUP54	58.60	2.16	208	98.81
644	0.0078	1.56	−0.644553	P40227	T-complex protein 1 subunit zeta	CCT6A	35.02	0.04	77.6	49.16
669	0.0017	4.52	−2.21759	O43852	Calumenin	CALU	35.23	2.10	109	98.10
679	0.0095	1.60	−0.659009	Q12849	G-rich sequence factor 1	GRSF1	42.91	2.29	120	98.30
686	0.0069	1.58	−0.663983	Q02790	Peptidyl-prolyl cis-trans isomerase FKBP4	FKBP4	30.93	1.90	108	97.96
686	0.0069	1.58	−0.663983	F8VVB9	Tubulin alpha-1B chain (Fragment)	TUBA1B	43.31	0.48	73.8	61.51
686	0.0069	1.58	−0.663983	Q71U36	Tubulin alpha-1A chain	TUBA1A	27.71	0.46	69.3	59.89
686	0.0069	1.58	−0.663983	Q9BQE3	Tubulin alpha-1C chain	TUBA1C	27.83	0.44	69.4	59.31
709	0.0012	2.05	−1.05342	P21281	V-type proton ATPase subunit B, brain isoform	ATP6V1B2	41.68	2.37	162	98.64
721	0.0055	1.69	−0.777649	P06576	ATP synthase subunit beta, mitochondrial	ATP5B	69.56	2.40	263	99.35
744	0.00014	3.44	−1.8001	P05787	Keratin, type II cytoskeletal 8	KRT8	28.57	2.37	188	98.82
746	0.0001	6.48	−2.74878	F8VXB4	Keratin, type II cytoskeletal 8	KRT8	22.50	2.36	125	98.38
764	0.001	1.90	−0.940855	A8K092	ATP synthase subunit alpha	ATP5A1	28.82	2.12	105	98.08
775	0.0037	2.03	−1.01819	P25705	ATP synthase subunit alpha, mitochondrial	ATP5A1	42.49	2.25	129	98.33
796	0.001	−1.53	0.620369	Q86UY0	TXNDC5 protein	TXNDC5	46.11	2.15	167	98.53
796	0.001	−1.53	0.620369	P63261	Actin, cytoplasmic 2	ACTG1	27.20	−0.10	67.5	45.64
815	0.000014	−2.47	1.30653	P06733	Alpha-enolase	ENO1	55.29	2.40	235	99.16
826	0.00038	−1.57	0.660319	Q9UQ80	Proliferation-associated protein 2G4	PA2G4	32.99	2.33	93.8	98.15
857	0.00015	5.79	−2.61568	P09104	Gamma-enolase	ENO2	46.08	2.35	212	98.97
862	0.00026	−1.68	0.739364	P06733	Alpha-enolase	ENO1	34.10	2.27	101	98.16
865	0.000048	5.28	−2.44942	B7Z2X9	Enolase	ENO2	24.80	2.26	95.5	98.11
866	0.007	−1.57	0.63123	P06733	Alpha-enolase	ENO1	48.38	2.37	163	98.65
868	0.0013	−1.72	0.787951	P06733	Alpha-enolase	ENO1	52.30	2.38	158	98.62
888	0.000048	−1.90	0.927113	Q15293	Reticulocalbin-1	RCN1	46.22	2.31	123	98.33
917	0.0055	−1.59	0.661034	P06733	Alpha-enolase	ENO1	41.93	2.16	123	98.23
952	0.0011	2.17	−1.14786	P63261	Actin, cytoplasmic 2	ACTG1	25.86	2.29	107	98.21
952	0.0011	2.17	−1.14786	P60709	Actin, cytoplasmic 1	ACTB	25.86	2.27	107	98.20
976	0.0027	2.03	−1.05405	P60709	Actin, cytoplasmic 1	ACTB	54.13	2.39	180	98.78
976	0.0027	2.03	−1.05405	P63261	Actin, cytoplasmic 2	ACTG1	54.13	2.39	180	98.78
976	0.0027	2.03	−1.05405	P68032	Actin, alpha cardiac muscle 1	ACTC1	27.58	0.91	87.7	86.10
976	0.0027	2.03	−1.05405	P68133	Actin, alpha skeletal muscle	ACTA1	27.58	0.89	87.7	85.00
976	0.0027	2.03	−1.05405	P63267	Actin, gamma-enteric smooth muscle	ACTG2	24.73	0.74	75.2	75.62
976	0.0027	2.03	−1.05405	P62736	Actin, aortic smooth muscle	ACTA2	24.66	0.72	75.1	74.50
987	0.0022	1.91	−0.954241	P60709	Actin, cytoplasmic 1	ACTB	58.66	2.30	185	98.75
987	0.0022	1.91	−0.954241	P63261	Actin, cytoplasmic 2	ACTG1	58.66	2.30	185	98.75
987	0.0022	1.91	−0.954241	P68032	Actin, alpha cardiac muscle 1	ACTC1	37.93	0.88	108	84.59
987	0.0022	1.91	−0.954241	P62736	Actin, aortic smooth muscle	ACTA2	42.97	0.77	124	78.65
987	0.0022	1.91	−0.954241	P68133	Actin, alpha skeletal muscle	ACTA1	31.83	0.67	92.3	72.93
987	0.0022	1.91	−0.954241	P63267	Actin, gamma-enteric smooth muscle	ACTG2	31.91	0.59	92.5	68.54
1066	0.012	1.94	−0.97915	P60709	Actin, cytoplasmic 1	ACTB	34.93	2.27	94.6	98.11
1066	0.012	1.94	−0.97915	P63261	Actin, cytoplasmic 2	ACTG1	34.93	2.27	94.6	98.11
1069	0.00068	−1.56	0.637715	O14979	Heterogeneous nuclear ribonucleoprotein D-like	HNRPDL	15.95	2.31	77	97.36
1183	0.00039	−2.00	1.0104	Q13347	Eukaryotic translation initiation factor 3 subunit I	EIF3I	25.53	2.42	78.7	97.82
1187	0.00012	−1.98	0.991703	C9J9K3	40S ribosomal protein SA (Fragment)	RPSA	38.63	2.38	105	98.26
1187	0.00012	−1.98	0.991703	A6NE09	Protein RPSAP58	RPSAP58	29.49	1.57	86.7	97.59
1207	0.0007	−2.23	1.18036	O76003	Glutaredoxin-3	GLRX3	43.58	2.26	121	98.29
1212	0.00011	−1.88	0.907804	Q6FI81	Anamorsin	CIAPIN1	37.50	2.32	112	98.26
1215	0.00013	−2.25	1.16543	Q13347	Eukaryotic translation initiation factor 3 subunit I	EIF3I	24.92	1.51	71.1	95.50
1215	0.00013	−2.25	1.16543	Q6FI81	Anamorsin	CIAPIN1	31.41	0.72	75.4	74.56
1326	0.0011	−1.99	1.00846	P29692	Elongation factor 1-delta	EEF1D	53.38	2.38	174	98.73
1335	0.009	1.95	−0.959166	P40926	Malate dehydrogenase, mitochondrial	MDH2	47.04	2.32	165	98.63
1338	0.00017	1.68	−0.74853	P29692	Elongation factor 1-delta	EEF1D	35.23	2.18	111	98.16
1343	0.00034	1.99	−0.986318	P12004	Proliferating cell nuclear antigen	PCNA	33.33	2.38	105	98.26
1470	0.0076	2.52	−1.47011	P08758	Annexin A5	ANXA5	18.43	1.40	77.7	94.48
1516	0.000047	1.98	−0.991669	A8MVD5	Tubulin-folding cofactor B	TBCB	35.98	2.11	91	97.98
1607	0.00049	1.82	−0.863653	P61981	14-3-3 protein gamma	YWHAG	42.91	2.16	123	98.23
1675	0.00038	1.68	−0.757938	P22676	Calretinin	CALB2	37.26	2.40	129	98.43
1696	0.003	−1.78	0.823488	F8W1A4	Adenylate kinase C	AK2	52.58	2.19	119	98.23
1696	0.003	−1.78	0.823488	P09661	U2 small nuclear ribonucleoprotein A'	SNRPA1	33.33	−0.03	67.2	45.61
1699	0.000052	−1.89	0.920371	P09661	U2 small nuclear ribonucleoprotein A'	SNRPA1	49.41	2.41	191	98.86
1734	0.005	2.21	−1.15537	P22676	Calretinin	CALB2	26.19	1.72	76.1	96.77
1833	0.0000071	2.51	−1.32991	P09936	Ubiquitin carboxyl-terminal hydrolase isozyme L1	UCHL1	44.84	2.31	103	98.20
2014	0.000098	1.64	−0.712496	P32119	Peroxiredoxin-2	PRDX2	36.86	2.37	113	98.30
2026	0.0000025	2.25	−1.17096	Q9UI15	Transgelin-3	TAGLN3	61.30	2.29	124	98.33
2688	0.013	−1.84	0.847047	P61604	10 kDa heat shock protein, mitochondrial	HSPE1	52.94	2.27	127	98.33
2711	0.017	−2.01	1.0435	Q13162	Peroxiredoxin-4	PRDX4	38.74	2.22	94.1	98.08
(B)
Spot ID	t-test value	Fold change	t-test difference	Accession	Protein	Gene	Seq. cov. [%]	Profound Z value	Mascot score	Meta score	N
175	0.00170	−1.6	0.63	G5E9M5	Interleukin enhancer binding factor 3, 90 kDa, isoform CRA_b	ILF3	20.60	2.16	111.0	98.15	4
284	0.00140	−1.5	0.59	P14625	Endoplasmin	HSP90B1	17.18	2.39	88.1	98.15	6
343	0.00130	−1.5	0.58	Q13200	26S proteasome non-ATPase regulatory subunit 2	PSMD2	24.55	2.29	150.0	98.50	6
355	0.00065	−1.9	0.95	O00571	ATP-dependent RNA helicase DDX3X	DDX3X	21.75	2.21	98.1	98.10	6
365	0.00031	−2.3	1.23	P15311	Ezrin	EZR	21.50	−0.15	79.3	48.27	5
377	0.00720	−1.6	0.77	Q13283	Ras GTPase-activating protein-binding protein 1	G3BP1	38.62	0.21	77.1	54.15	6
537	0.00097	2.1	−1.04	P36578	60S ribosomal protein L4	RPL4	33.02	1.55	59.5	92.39	6
563	0.00210	2.1	−1.03	P38159	RNA-binding motif protein, X chromosome	RBMX	14.32	1.70	102.0	97.78	5
851	0.00035	−1.7	0.81	P13010	X-ray repair cross-complementing protein 5	XRCC5	12.97	2.28	54.5	86.46	6
856	0.00260	−1.7	0.79	P55072	Transitional endoplasmic reticulum ATPase	VCP	10.91	2.37	104.0	98.24	6
858	0.00016	−1.5	0.59	Q15029	116 kDa U5 small nuclear ribonucleoprotein component	EFTUD2	18.82	1.84	84.1	97.75	6
859	0.01100	1.6	−0.58	Q9Y3A5	Ribosome maturation protein SBDS	SBDS	20.40	0.95	67.2	85.37	5
860	0.00049	1.8	−0.87	P22626	Heterogeneous nuclear ribonucleoproteins A2/B1	HNRNPA2B1	60.62	2.42	214.0	99.03	6

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository (Vizcaino, [63]) with the dataset identifier PXD000132 and DOI 10.6019/PXD000132.

Verification of differentially regulated proteins using Western Blotting

To verify the differences in expression found by the 2D-DIGE approach, proteins from 4 individual dishes per cell line were subjected to Western Blot analysis. Cells were harvested and lysates prepared as described for IEF/SDS 2D-PAGE sample preparation. 20 µg of proteins were mixed with one quarter volume sample buffer (0.03% w/v Coomassie, 6% w/v DTT, 30% w/v glycerin, 12% w/v SDS, 150 mM Tris/HCl, pH 7) and loaded on 4–20% gradient Tris/Glycine gels (Anamed, Groß-Bieberau, Germany) and blotted on PVDF membrane (GE-Healthcare, Munich, Germany).

For protein detection following primary antibodies known to recognize the human antigens were used: anti-CIAPIN1 (SantaCruz sc-49601, polyclonal goat antibody, 1:20), anti-XRCC5 (Cell Signalling #2180, monoclonal rabbit antibody, 1:200), anti-HSPE1 (Abcam ab13528, polyclonal rabbit antibody, 1:200), anti-UCHL1 (SantaCruz sc-23852, polyclonal goat antibody, 1:20) and anti-SBDS (SantaCruz sc-49257, polyclonal goat antibody, 1:20).

After membrane blocking the membrane was incubated with the primary antibody in TBS (including 5% non-fat dry milk, 0.05% Tween20 for CIAPIN1, UCHL1 and SBDS and for 2% BSA and 3% non-fat dry milk for HSPE1 and XRCC5). After washing a fluorophore labelled secondary antibody was applied (anti-rabbit IR680, Li-Cor Biosciences, 1:30 000; anti goat IR680, Li-Cor Biosciences, 1:15 000). The membrane was washed, dried and fluorescence signals acquired using the Odyssey infrared fluorescence detection system (Li-Cor Biosciences). As typical housekeeping proteins like actin came up in our study to be differentially in abundance between the two cell lines, a Ponceau post-staining of the membranes was used as a loading control.

Fluorescence signals were quantified using ImageQuant TL Software (GE-Healthcare, Munich, Germany) using "rolling-ball" background correction. The pixel-intensities per area were calculated and used for statistical analysis. Mann and Whitney U-tests were performed using the Statistica8 software package (StatSoft, Hamburg, Germany) and OriginPro 8G (Origin Lab, Northampton, MA, USA) was used for data visualization.

Results

Identification of proteins differentially expressed between variant and classic SCLC cells

As variant and classic SCLC cell lines show a different behaviour in regard to growth characteristics and chemo resistance we are interested in identifying the molecular factors determining these phenotypes. Therefore, we used the cell lines GLC1 (variant) and GLC1 M13 (classic) as a model system. The cell line GLC1 M13 is a sub-clone of the GLC1 cell line, obtained through selection of cells cultivated in serum starved medium and limited dilution (Brouwer et al., [10]; de Leij et al., [16]). To obtain a more comprehensive view of the proteomic differences two methods were applied. IEF/SDS-PAGE was considered to analyse whole cell lysates, whereas CTAB/SDS-PAGE were chosen to separate purified membrane fractions of the cells.

Analysis of the whole cell lysate using IEF/SDS-PAGE

As a first step we carried out a DIGE analysis of the whole cell lysate running four biological replicates of each cell line. Proteins from GLC1 cells were co-separated with an internal standard (pool of all samples) labelled with Cy2. As in a same-same pilot experiment no dye specific protein labelling was observed we decided not to label one cell line with alternating dyes (dye swop). The image analysis of four gels (4 GLC1 gel images (Cy3), 4 GLC1 M13 gel images (Cy5) and 4 internal standard images) revealed an average number of 2700 detected protein spots per gel image (Figure 1; Supplemental Figure 1). The subsequent quantitative analysis showed 109 protein spots to be differentially regulated (t-test < 0.05, FDR < 5%, SAM S0 = 0.8). Examples of differentially expressed spots are given in Figure 3. All differentially expressed protein spots were subjected to tryptic digestion and 59 proteins were identified using MALDI-MS (Figure 1B, Table 1A). The proteins provided the basis for compiling a list containing 47 non-redundant candidate proteins.

Graph: Figure 1. (A) Example of 2D-DIGE IEF/SDS-PAGE. Lysates of GLC-1 and GLC-1-M13 were labelled with different fluorescent dyes and separated in two dimensions. For the differential analysis 4 independent samples per cell line have been used. More abundant spots in GLC1 M13 protein lysat separations appear in red, more abundant spots in GLC1 protein separations appear in green. (B) Differentially abundant protein spots identified by MALDI-MS were marked by arrows.

Analysis of purified membrane fractions using CTAB/SDS-PAGE

IEF/SDS-PAGE is a broadly applied method for the quantitative analysis of complex protein mixtures. Due to the IEF step this technique has some limitations particularly in analysing hydrophobic proteins as, for example, membrane proteins (Santoni et al., [49]; Galeva & Altermann, [23]). To obtain a more comprehensive coverage of the proteomes of the classic and variant SCLC cells we also analysed a purified membrane fraction of GLC1 and GLC1 M13 cells by CTAB/SDS-PAGE. In contrast to IEF CTAB as a cationic detergent offers the opportunity to keep hydrophobic proteins solubilized during separation. As the first step we enriched membrane and membrane associated proteins by ultracentrifugation and a sodium carbonate washing and subsequently applied CTAB/SDS-PAGE for protein separation (Navarre et al., [41]; Helling et al., [30]).

As expected, most of the resolved proteins show different separation behaviours in both dimensions, resulting in a widely spread diagonal spot pattern. Using this approach we resolve about 900 protein spots in one gel (Figure 2, Supplemental Figure 2). The image analysis revealed 62 proteins to be differentially expressed between the two cell lines (t-test < 0.05, FDR < 5%, SAM S0 = 0.8). Out of these protein spots 13 non-redundant proteins were identified via MALDI-MS (Figure 2B and Figure 3, Table 1B).

Graph: Figure 2. (A) Example of 2D-DIGE CTAB/SDS-PAGE. More abundant spots in GLC1 M13 protein lysat separations appear in red, more abundant spots in GLC1 protein separations appear in green. For the differential analysis six independent samples per cell line have been used. (B) Differentially abundant protein spots identified by MALDI-MS are marked by arrows.

Graph: Figure 3. Examples of differentially regulated protein spots. The circles represent normalized and standardized spot volumes of individual samples whereas the triangles mark the group means. The standard deviation of each group is given by the error bars.

Molecular functions of differentially abundant proteins

Comparing the proteins identified from IEF/SDS and CTBA/SDS-PAGE experiments we found completely complementary candidate proteins representing a broader spectrum of protein classes potentially involved in the biology of GLC1 cell line subtypes. For example, the gene ontology category proteolysis and nucleic acid binding are 3.7 and 2.1 fold enriched (p-value 0.006 and 0.002 respectively, Fisher exact test) in proteins with differential abundance from the CTAB-SDS PAGE experiments. Most of the found proteins can be associated with gene ontology terms revealing their potential role in mediating tumour cell characteristics. Commonly found terms are, for example, metabolic process (60% of proteins), regulation of biological process (43% of proteins), nucleic acid metabolic process (27% of proteins), response to stress (15% of proteins), regulation of programmed cell death (11.7% of proteins), cell cycle process (10% of proteins) and chemotaxis (8% of proteins). Unexpectedly we did not find an enrichment of membrane proteins in proteins identified from CTAB-SDS experiments. Instead proteins associated with the gene ontology cellular compartment category nuclear part were twofold enriched (p-value = 0.002, Fisher exact test).

Verification of differentially expressed proteins using Western Blotting

For verification of protein abundance differences between GLC1 and GLC1 M13 cells found by 2D-DIGE, we applied Western Blot analysis. Therefore, we choose proteins from the IEF/SDS-PAGE approach (UCHL1 gene product, CIAPIN1, HSPE1) as well as proteins found by the CTAB/SDS-PAGE experiment (SBDS, XRCC5) on basis of antibody availability. The verification includes both, proteins with higher abundancy in GLC1 (CIAPIN1, HSPE1, XRCC5) as well as in GLC1 M13 cells (UCHL1 gene product, SBDS). A negative fold change means that the protein is more abundant in variant GLC1 cells; proteins exhibiting a positive fold change are more abundant in GLC1 M13 cells. For all proteins tested, the semi-quantitative Western Blot analysis confirmed the differences in protein abundancy between the two analysed cell lines as shown by 2D-DIGE (Figure 4): UCHL1, fold change 3.6, p = 0.02 (2D-DIGE: fold change 2.5, p = 7.1*10−6); CIAPIN1 fold change −1.9, p = 0.04 (2D-DIGE: fold change −1.9, p = 0.0001); HSPE1, fold change −1.6, p = 0.02 (2D-DIGE: fold change −1.8, p = 0.013); SBDS, fold change 2.5, p = 0.02 (2D-DIGE: fold change 1.6, p = 0.01); XRCC5, fold change −1.8, p = 0.02 (2D-DIGE: fold change −1.7, p = 0.0004).

Graph: Figure 4. Verification of differentially expressed proteins using Western blot analysis. (A) Western Blot analysis was performed using the Odyssey system and near-infrared absorbing dye-labelled secondary antibodies (anti UCHL1 gene product, anti SBDS, anti CIAPIN1, anti HSPE1, anti XRCC5). (B) Band quantification verified the differences in protein abundances in GLC1 and GLC1 M13 cells (*p < 0.05, U-test). The height of the boxes represent the mean of the relative band intensities the standard deviation is given by the error bars. Four individual samples per cell line were analysed.

Discussion

SCLC tumours exhibit an aggressive phenotype and are normally treated by chemo- and/or radiotherapy (Chen et al., [13]). To learn more about the processes involved, chemo- and radio-sensitivity cell lines are commonly used as model systems because tissue specimens of SCLC tumours are normally not available, since surgery and resection leading to appropriate sample material are routinely not performed (Inoue et al., [32]).

Classic and variant SCLC cells can not only be distinguished and categorized by their functional features. It has been shown that in classic SCLC intermediate filament proteins like cytokeratin 8 and 18 are clearly expressed, while no neurofilaments nor vimentin was shown. Variant SCLCs on the other hand, contain vimentin but do not express cytokeratin 8 and 18, whereas the majority of the variant SCLC cell lines contain neurofilaments (Broers et al., [9], [6], [7], [8]). In this study, using our approach, we confirmed the high expression of cytokeratin 8 in classic GLC1 M13 cells, supporting the feasibility of our technology and the models used. It has been shown that SCLC cancer cell lines are a good representation of fresh tumour tissue: most of them can be grouped with fresh tumour tissue in a hierarchical cluster analysis as assessed after cDNA microarray analysis (Virtanen et al., [62]).

The use of the model cell system of classic and variant GLC1 cell lines has multiple advantages. First the two cell lines exhibiting different growth characteristics and chemo resistance have been characterized and extensively studied before (Broers et al., [9], [6]; de Leij et al., [16]; Verbeeck et al., [61]). Second, the GLC1 M13 cell line has been established as a subclone of the GLC1 cell line (de Leij et al., [16]) and therefore is derived not only from the same patient, but from the same parental cell line, reducing biological variation and allowing to focus on the functional aspects in a proteome study. Third, the production of cells can be scaled up making it possible to analyse also a sub-fraction of the proteome.

Next to the classical IEF/SDS-2D PAGE analysing the whole cell lysate of the two cell lines we isolate a sub-fraction of the cellular proteomes by ultracentrifugation and carried out a subsequent analysis by CTAB/SDS-2D PAGE studying the membrane fraction of the proteome. CTAB/SDS-2D PAGE is known as a method to separate/enrich proteins which are under-represented in IEF/SDS 2D-PAGE gels (Navarre et al., [41]; Helling et al., [30]). Therefore, we expected other differentially expressed proteins using the CTB/SDS-2D PAGE analysis. As there is no overlap at all between the proteins resulting from both analyses, we clearly can state that the two approaches are highly complementary to one another.

We found differentially expressed proteins involved in metabolism, stress response, apoptosis and chemotaxis, all processes typically affected in tumour cells resulting in different growth profiles, aggressiveness and chemo-resistance. Differences in protein expression might be the reason for altered phenotypes.

Here, we highlight some examples of found proteins which might have an influence on pathways leading to characteristics of the different cell lines.

Resistance of variant cells to apoptosis

Cytokine-induced apoptosis inhibitor 1 (CIAPIN1) is a protein known to mediate multidrug resistance in leukaemia cells (Li et al., [37]). In line with this result, over-expression of CIAPIN1 in a gastric cancer cell line leads to a lower sensitivity for chemo-therapeutica (Hao et al., [27]). The molecule has been found to be over-expressed in hepatocellular carcinomas and a knockdown of the molecule lead to a lower proliferation rate in vitro and in xenograph mice model (Li et al., [36]). CIAPIN1 knockout mice die in late gestation, erythroid colony formation in response to cytokines was severely disrupted; erythoid cells initiated apoptosis during terminal maturation. This might be mediated through an impaired expression of Bcl-xL and Jak2 (Shibayama et al., [55]). In our study we found CIAPIN1 highly expressed in variant GLC1 cells suggesting that this protein might inhibit apoptotic pathways leading to chemo-resistance and a higher proliferation rate in our cell model.

Another protein identified in our study which might contribute to the inhibition of apoptosis in variant SCLC cells is heat shock protein E1 (HSPE1). This protein is known to be involved in protein folding in collaboration with HSPD1, but also might participate in apoptosis signalling. HSPE1 over-expression in cardiac muscle cells increases the abundance of anti-apoptotic Bcl-2 and Bcl-xl and reduces the protein content of the pro-apoptotic Bax (Bcl-2associated X protein) (Shan et al., [54]; Czarnecka et al., [14]). For several cancer types (endometrial, colorectal, exocervical prostate, ovarian and colorectal cancer) HSPE1 has been found, to be over-expressed and even suggested as a biomarker (Cappello et al., [11], [12]; Yang et al., [67]; Akyol et al., [1]; Melle et al., [40]; Czarnecka et al., [14]; DeSouza et al., [17]; Dube et al., [18]).

Repair of DNA damage

We found XRCC5 as protein which might determine the characteristics of SCLC cells. This protein has been described to recognize and bind broken ends of double-strand DNA and acts as an alignment factor to promote DNA end-joining, a process which might be regulated by BCL2 (Feldmann et al., [20]; Wang et al., [64]). Tumours expressing XRCC5 are more resistant to irradiation as shown for lung cancer cells, cervical tumours and rectal carcinomas (Harima et al., [28]; Komuro et al., [35]; Guo et al., [26]), for NSCLC XRCC5 expression correlated with shorter patient survival (Saviozzi et al., [50]). Moreover polymorphisms in XRCC5 gens might contribute to tumour susceptibility in breast cancer, myeloma and bladder cancer and NSCLC (Hayden et al., [29]; Wang et al., [65]; Willems et al., [66]; Yin et al., [69]). Our results suggest also a potential role of XRCC5 in SCLC tumours, it might mediate the increased radio-resistance of variant SCLC tumour cells.

A highly conserved protein known to be involved in the maintenance of genomic stability during mitosis is SBDS (Boocock et al., [4]; Austin et al., [2]; Rujkijyanont et al., [48]). As we found this protein to be higher expressed in the classic GLC1 M13 cells we speculate if this protein in SCLCs cell might contribute to the prevention of genomic instability in classic SCLC cells and thereby contradicting the development of resistance to chemo- and radiotherapy.

Protein degradation

UCHL1 is a well-known gene whose gene product is known to be involved in the removal of ubiquitin from ubiquitinylated proteins and thereby preventing their degradation by the proteasome in neurons (Gong & Leznik, [25]). The role of UCHL1 in the context of cancer development is discussed controversially and seems to be cancer type specific. UCHL1 promotor methylation has been described to be elevated in tumours of colorectal cancer (Fukutomi et al., [22]), oesophageal tumours (Mandelker et al., [39]), head and neck squamous cell carcinoma (Tokumaru et al., [58]) and renal carcinoma (Kagara et al., [33]). For head and neck squamous carcinoma and renal carcinoma the UCHL1 gene product has been suggested to exhibit features of a tumour suppressor (Kagara et al., [33]; Tokumaru et al., [58]), whereas for lung carcinomas it might act as an oncogene as the over-expression characterizes more advanced stages of the disease (Rehm et al., [47]; Bittencourt Rosas et al., [3]; Hibi et al., [31]). In lung cancer patients UCHL1 gene product auto-antibodies have been found (Madoz-Gurpide et al., [38]). Our data suggest a role for UCHL1 in SCLC cells as a tumour suppressor because we found this protein to be highly expressed in the classic GLC1 M13 cell line. Further studies are needed to clarify the role of UCHL1 in the development and progression of SCLC.

In this study we showed that using a combination of two different 2D gel based separation methods highly complementary results can be obtained. A combination of both methods revealed differentially expressed proteins in specific lung cancer subtypes, which could be successfully validated by Western blot analysis. We conclude that these methods might be used for the identification and characterization of potential proteins differentially expressed in specific lung cancer subtypes. These identified proteins might be involved in treatment resistance pathways of specific lung cancer subtypes, and determine specific growth characteristics thus having important clinical implications. Further investigations of other SCLC cell lines and in vivo experiments have to be carried out to characterize their individual roles in more detail.

Declaration of interest

This work was supported by a grant from the European Commission (LCVAC, COOP-CT-2004-512855) and the Ministry of Innovation, Science, Research and Technology (MIWFT) of North Rhine-Westphalia.

Acknowledgements

The authors thank Kathy Pfeiffer for excellent technical assistance.

1 Akyol S, Gercel-Taylor C, Reynolds LC, Taylor DD. (2006). HSP-10 in ovarian cancer: expression and suppression of T-cell signaling. Gyneco Oncol, 101, 481–6 2 Austin KM, Gupta ML, Coats SA, et al. (2008). Mitotic spindle destabilization and genomic instability in Shwachman-Diamond syndrome. J Clin Inv, 118, 1511–18 3 Bittencourt Rosas SL, Caballero OL, Dong SM, et al. (2001). Methylation status in the promoter region of the human PGP9.5 gene in cancer and normal tissues. Cancer Lett, 170, 73–9 4 Boocock GR, Marit MR, Rommens JM. (2006). Phylogeny, sequence conservation, and functional complementation of the SBDS protein family. Genomics, 87, 758–71 5 Brambilla E, Travis WD, Colby TV, Corrin B, Shimosato Y. (2001). The new World Health Organization classification of lung tumours. Eur Respir J, 18, 1059–68 6 Broers JL, Carney DN, Klein Rot M, et al. (1986). Intermediate filament proteins in classic and variant types of small cell lung carcinoma cell lines: a biochemical and immunochemical analysis using a panel of monoclonal and polyclonal antibodies. J Cell Sci, 83, 37–60 7 Broers JL, Ramaekers FC, Rot MK, et al. (1988a). Cytokeratins in different types of human lung cancer as monitored by chain-specific monoclonal antibodies. Cancer Res, 48, 3221–9 8 Broers JL, Rot MK, Oostendorp T, Bepler G, et al. (1988b). Spontaneous changes in intermediate filament protein expression patterns in lung cancer cell lines. J Cell Sci, 91 (Pt 1), 91–108 9 Broers JLV, Carney DN, Ley LD, et al. (1985). Differential expression of intermediate filament proteins distinguishes classic from variant small-cell lung cancer cell lines. PNAS, 82, 4409–13 Brouwer M, De Ley L, Feltkamp CA, et al. (1984). Serum-dependent "cannibalism" and autodestruction in cultures of human small cell carcinoma of the lung. Cancer Res, 44, 2947–51 Cappello F, Bellafiore M, David S, Anzalone R, Zummo G. (2003). Ten kilodalton heat shock protein (HSP10) is overexpressed during carcinogenesis of large bowel and uterine exocervix. Cancer Lett, 196, 35–41 Cappello F, Czarnecka AM, La Rocca G, et al. (2007). Hsp60 and Hspl0 as antitumor molecular agents. Cancer Biol Therapy, 6, 487–9 Chen YT, Feng B, Chen LB. (2012). Update of research on drug resistance in small cell lung cancer chemotherapy. Asian Pac J Cancer Prev, 13, 3577–81 Czarnecka AM, Campanella C, Zummo G, Cappello F. (2006). Heat shock protein 10 and signal transduction: a "capsula eburnea" of carcinogenesis?. Cell Stress & Chaperones, 11, 287–94 Danaei G, Vander Hoorn S, Lopez AD, Murray CJ, Ezzati M. (2005). Causes of cancer in the world: comparative risk assessment of nine behavioural and environmental risk factors. Lancet, 366, 1784–93 De Leij L, Postmus PE, Buys CH, et al. (1985). Characterization of three new variant type cell lines derived from small cell carcinoma of the lung. Cancer Res, 45, 6024–33 Desouza LV, Grigull J, Ghanny S, et al. (2007). Endometrial carcinoma biomarker discovery and verification using differentially tagged clinical samples with multidimensional liquid chromatography and tandem mass spectrometry. Mol Cell Prot, 6, 1170–82 Dube V, Grigull J, Desouza LV, et al. (2007). Verification of endometrial tissue biomarkers previously discovered using mass spectrometry-based proteomics by means of immunohistochemistry in a tissue microarray format. J Prot Res, 6, 2648–55 Etzel CJ, Lu M, Merriman K, et al. (2006). An epidemiologic study of early onset lung cancer. Lung Cancer (Amsterdam, The Netherlands), 52, 129–34 Feldmann E, Schmiemann V, Goedecke W, et al. (2000). DNA double-strand break repair in cell-free extracts from Ku80-deficient cells: implications for Ku serving as an alignment factor in non-homologous DNA end joining. Nucleic Acids Res, 28, 2585–96 Franks TJ, Galvin JR. (2008). Lung tumors with neuroendocrine morphology: essential radiologic and pathologic features. Arch Pathol Lab Med, 132, 1055–61 Fukutomi S, Seki N, Koda K, Miyazaki M. (2007). Identification of methylation-silenced genes in colorectal cancer cell lines: genomic screening using oligonucleotide arrays. Scand J Gastroent, 42, 1486–94 Galeva N, Altermann M. (2002). Comparison of one-dimensional and two-dimensional gel electrophoresis as a separation tool for proteomic analysis of rat liver microsomes: cytochromes P450 and other membrane proteins. Proteomics, 2, 713–22 Gazdar AF, Carney DN, Nau MM, Minna JD. (1985). Characterization of variant subclasses of cell lines derived from small cell lung cancer having distinctive biochemical, morphological, and growth properties. Cancer Res, 45, 2924–30 Gong B, Leznik E. (2007). The role of ubiquitin C-terminal hydrolase L1 in neurodegenerative disorders. Drug News Perspect, 20, 365–70 Guo WF, Lin RX, Huang J, et al. (2005). Identification of differentially expressed genes contributing to radioresistance in lung cancer cells using microarray analysis. Rad Res, 164, 27–35 Hao Z, Li X, Qiao T, et al. (2006). CIAPIN1 confers multidrug resistance by upregulating the expression of MDR-1 and MRP-1 in gastric cancer cells. Cancer Biol Therapy, 5, 261–6 Harima Y, Sawada S, Miyazaki Y, et al. (2003). Expression of Ku80 in cervical cancer correlates with response to radiotherapy and survival. Am J Clin Onc, 26, e80–5 Hayden PJ, Tewari P, Morris DW, et al. (2007). Variation in DNA repair genes XRCC3, XRCC4, XRCC5 and susceptibility to myeloma. Human Mol Gen, 16, 3117–27 Helling S, Schmitt E, Joppich C, et al. (2006). 2-D differential membrane proteome analysis of scarce protein samples. Proteomics, 6, 4506–13 Hibi K, Westra WH, Borges M, et al. (1999). PGP9.5 as a candidate tumor marker for non-small-cell lung cancer. Am J Pathol, 155, 711–15 Inoue M, Sawabata N, Okumura M. (2012). Surgical intervention for small-cell lung cancer: what is the surgical role? Gen Thorac Cardiovasc Surg, 60, 401–5 Kagara I, Enokida H, Kawakami K, et al. (2008). CpG hypermethylation of the UCHL1 gene promoter is associated with pathogenesis and poor prognosis in renal cell carcinoma. J Urology, 180, 343–51 Klose J, Kobalz U. (1995). Two-dimensional electrophoresis of proteins: an updated protocol and implications for a functional analysis of the genome. Electrophoresis, 16, 1034–59 Komuro Y, Watanabe T, Tsurita G, et al. (2005). Evaluating the combination of molecular prognostic factors in tumor radiosensitivity in rectal cancer. Hepato-gastroenterology, 52, 666–71 Li X, Hao Z, Pan Y, et al. (2008). Adenovirus-delivered CIAPIN1 small interfering RNA inhibits HCC growth in vitro and in vivo. Carcinogenesis, 29, 1587–93 Li X, Hong L, Zhao Y, et al. (2007). A new apoptosis inhibitor, CIAPIN1 (cytokine-induced apoptosis inhibitor 1), mediates multidrug resistance in leukemia cells by regulating MDR-1, Bcl-2, and Bax. Biochem Cell Biol (Biochimie et biologie cellulaire), 85, 741–50 Madoz-Gurpide J, Kuick R, Wang H, et al. (2008). Integral protein microarrays for the identification of lung cancer antigens in sera that induce a humoral immune response. Molec Cell Proteomics, 7, 268–81 Mandelker DL, Yamashita K, Tokumaru Y, et al. (2005). PGP9.5 promoter methylation is an independent prognostic factor for esophageal squamous cell carcinoma. Cancer Res, 65, 4963–8 Melle C, Bogumil R, Ernst G, et al. (2006). Detection and identification of heat shock protein 10 as a biomarker in colorectal cancer by protein profiling. Proteomics, 6, 2600–8 Navarre C, Degand H, Bennett KL, et al. (2002). Subproteomics: identification of plasma membrane proteins from the yeast Saccharomyces cerevisiae. Proteomics, 2, 1706–14 Pastorino, U. (2006). Early detection of lung cancer. Respiration, 73, 5–13 Perkins DN, Pappin DJ, Creasy DM, Cottrell JS. (1999). Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis, 20, 3551–67 Popov N, Schmitt M, Schulzeck S, Matthies H. (1975). [Reliable micromethod for determination of the protein content in tissue homogenates]. Acta Biol Med Ger, 34, 1441–6 Poschmann G, Sitek B, Sipos B, Stuhler K. (2012). Application of saturation labeling in lung cancer proteomics. Methods Mol Biol, 854, 253–67 Poschmann G, Sitek B, Sipos B, et al. (2009). Identification of proteomic differences between squamous cell carcinoma of the lung and bronchial epithelium. Mol Cell Proteomics, 8, 1105–16 Rehm H, Wiedenmann B, Betz H. (1986). Molecular characterization of synaptophysin, a major calcium-binding protein of the synaptic vesicle membrane. Embo J, 5, 535–41 Rujkijyanont P, Watanabe K, Ambekar C, et al. (2008). SBDS-deficient cells undergo accelerated apoptosis through the Fas-pathway. Haematologica, 93, 363–71 Santoni V, Molloy M, Rabilloud T. (2000). Membrane proteins and proteomics: un amour impossible? Electrophoresis, 21, 1054–70 Saviozzi S, Ceppi P, Novello S, et al. (2009). Non-small cell lung cancer exhibits transcript overexpression of genes associated with homologous recombination and DNA replication pathways. Cancer Res, 69, 3390–96 Schaefer H, Chamrad DC, Marcus K, et al. (2005). Tryptic transpeptidation products observed in proteome analysis by liquid chromatography-tandem mass spectrometry. Proteomics, 5, 846–52 Schaefer H, Chervet JP, Bunse C, et al. (2004). A peptide preconcentration approach for nano-high-performance liquid chromatography to diminish memory effects. Proteomics, 4, 2541–4 Schneider J. (2006). Tumor markers in detection of lung cancer. Adv Clin Chem, 42, 1–41 Shan YX, Liu TJ, Su HF, et al. (2003). Hsp10 and Hsp60 modulate Bcl-2 family and mitochondria apoptosis signaling induced by doxorubicin in cardiac muscle cells. J Mol Cell Cardiol, 35, 1135–43 Shibayama H, Takai E, Matsumura I, et al. (2004). Identification of a cytokine-induced antiapoptotic molecule anamorsin essential for definitive hematopoiesis. J Exp Med, 199, 581–92 Shibuya K, Mathers CD, Boschi-Pinto C, et al. (2002). Global and regional estimates of cancer mortality and incidence by site: II. Results for the global burden of disease 2000. BMC Cancer, 2, 37 Stephan C, Reidegeld KA, Hamacher M, et al. (2006). Automated reprocessing pipeline for searching heterogeneous mass spectrometric data of the HUPO Brain Proteome Project pilot phase. Proteomics, 6, 5015–29 Tokumaru Y, Yamashita K, Kim MS, et al. (2008). The role of PGP9.5 as a tumor suppressor gene in human cancer. Int J Cancer, 123, 753–9 Travis WD, Linnoila RI, Tsokos MG, et al. (1991). Neuroendocrine tumors of the lung with proposed criteria for large-cell neuroendocrine carcinoma. An ultrastructural, immunohistochemical, and flow cytometric study of 35 cases. Am J Surg Pathol, 15, 529–53 Tusher VG, Tibshirani R, Chu G. (2001). Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA, 98, 5116–21 Verbeeck MA, Elands JP, De Leij LF, et al. (1992). Expression of the vasopressin and gastrin-releasing peptide genes in small cell lung carcinoma cell lines. Pathobiology, 60, 136–42 Virtanen, C, Ishikawa, Y, Honjoh, D, et al. (2002). Integrated classification of lung tumors and cell lines by expression profiling. Proc Natl Acad Sci USA, 99, 12357–62 Vizcaino JA, Cote RG, Csordas A, et al. (2013). The Proteomics Identifications (PRIDE) database and associated tools: status in 2013. Nucleic Acids Res, 41, D1063–9 Wang, Q, Gao, F, May, WS, et al. (2008a). Bcl2 negatively regulates DNA double-strand-break repair through a nonhomologous end-joining pathway. Molecular Cell, 29, 488–98 Wang S, Wang M, Yin S, et al. (2008b). A novel variable number of tandem repeats (VNTR) polymorphism containing Sp1 binding elements in the promoter of XRCC5 is a risk factor for human bladder cancer. Mutation Res, 638, 26–36 Willems, P, Claes, K, Baeyens, A, et al. (2008). Polymorphisms in nonhomologous end-joining genes associated with breast cancer risk and chromosomal radiosensitivity. Genes, Chromosomes & Cancer, 47, 137–48 Yang EC, Guo J, Diehl G, et al. (2004). Protein expression profiling of endometrial malignancies reveals a new tumor marker: chaperonin 10. J Proteome Res, 3, 636–43 Yesner R. (1983). Small cell tumors of the lung. Am J Surgical Pathol, 7, 775–85 Yin, M, Liao, Z, Liu, Z, et al. (2012). Genetic variants of the nonhomologous end joining gene LIG4 and severe radiation pneumonitis in nonsmall cell lung cancer patients treated with definitive radiotherapy. Cancer, 118, 528–35 Zhang W, Chait BT. (2000). ProFound: an expert system for protein identification using mass spectrometric peptide mapping information. Analyt Chem, 72, 2482–9

By Gereon Poschmann; Anna Lendzian; Julian Uszkoreit; Martin Eisenacher; Ann Vander Borght; Frans C. Ramaekers; Helmut Erich Meyer and Kai Stühler

Reported by Author; Author; Author; Author; Author; Author; Author; Author