Figs. 1 and 2 display the gel localization of the proteins identified in the present study. Twelve protein spots, differentially expressed in both tumour cell lines (i.e. SM5 and SM7) are labelled using white numbering. Protein spots that were uniquely differentially expressed in SM5 or SM7 cells are labelled using black numbering. Protein identifications are listed in Tables 1, 2 and 3. Of the 47 identified spots 5, i.e. 472, 521, 707/730, 1216 and 1470 were ambiguously determined, since more than one protein were present in each spot. Thus, for these 5 spots it could not be unequivocally determined which of the proteins that were differentially expressed. Western blotting was used to verify several of the changes seen.

The commonly regulated proteins are given in Table 1. We identified up-regulation of rho GDP-dissociation inhibitor 2 (1236), proteasome subunit α type 3 (1062), transforming acidic coiled-coil containing protein 3 (131), mitochondrial ornithine aminotransferase (653) and epidermal fatty acid binding protein (1468).

Among the down regulated proteins, we identified adenylosuccinate synthetase (622), tubulin β-3 chain (830/845), a 25 kDa actin fragment (1123), proteasome subunit β type 9 (1284), basic cofilin-1 variant (1357), acidic cofilin-1 variant (1363/1367) and glia maturation factor γ, (1382).

Tables 2 and 3 show the obtained peptides from uniquely regulated proteins. Unambiguously identified up-regulated proteins in tumour cells SM5 included: vinculin, serine/threonine-protein phosphatase 2A 65 kDa regulatory subunit A α isoform, prolyl 4-hydroxylase α-1 subunit, α enolase, ERC-55, eukaryotic translation initiation factor 3 subunit 4, eukaryotic translation initiation factor 2 subunit 1, triosephosphate isomerase, nucleoside diphosphate kinase B and in tumour cells SM7: 40 kDa peptidyl-prolyl cis-trans isomerase, acyl-protein thioesterase 1, and nudC domain-containing protein 2. Spots 298, 300, 302, 309 and 311 contained bovine albumin while spot 308 contained bovine α 2HS-glycoportein or fetuin. Spots 521, 472 and 1470 contained more than one protein.

Unambiguously identified down-regulated proteins in tumour cells SM5 included: poly(rC)-binding protein 1, proteasome inhibitor PI31 subunit, histone H4, ATP synthase α chain and in the SM7 tumour cell line: coronin-1A, adenosine deaminase, heterogeneous nuclear ribonucleoprotein A/B, 4-hydroxyphenyl puruvate dioxygenase, eukaryotic translation initiation factor 3 subunit 2, peroxiredoxin-2, ADP-ribosylation factor 1, ubiquitin-conjugating enzyme E2 N, actin-like protein 3 and transforming protein RhoA. Spots 707/730 and 1216 contained more than one protein.

To independently confirm 2D-PAGE data, 1D Western Blot analysis was used to assess the differential expression of selected protein immunoreactive. Ideally, a single strong band on a Western blot reflects an antibody possessing high affinity with specific reaction against the antigen. The data obtained by Western Blot analysis is presented in Fig. 3.

Fig. 3A depicts Western blot analysis for three commonly regulated antigens: i.e. epidermal fatty acid binding protein (FABPE), proteasome subunit β type 9 (PSB9) and cofilin-1 (COF1). Each produced strong bands in the cells where the proteins were present in high amounts. Furthermore the data are consistent with and confirm 2D-PAGE data. That is, that FABPE is more than 2-fold up-regulated in both of the tumour cells and that PSB9 and COF1 are both more than 2-fold down-regulated in tumour cells.

Fig. 3B and 3C depict Western blot data that verifies the differential regulation of selected uniquely regulated proteins. That is, eukaryotic translation initiation factor 3 subunit 4 (IF34) and triosephosphate isomerase (TPIS) were found to be more than 2-fold up-regulated in the SM5 cell line but in fact also in the SM7 cell line. Similarly, down-regulation in tumour cells SM7 was confirmed in case of the proteins coronin-1A (COR1A, 10-fold), heterogeneous nuclear ribonucleoprotein A/B (ROAA, 9-fold), peroxiredoxin-2 (PRDX2, 1.5-fold), ubiquitin-conjugating enzyme E2 N (UBE2N, >10-fold) and actin-like protein 3 (ARP3, 3.1-fold). All of these proteins were also down-regulated in the tumour SM5.

Based on Western blots data, 2 of the 11 proteins did not display concordance with 2D-PAGE data, i.e. serine/threonine-protein phsophatase 2A 65 kDa regulatory subunit A α isoform (2AAA) and ADP-ribosylation factor (ARF). In case of ADP-ribosylation factor, the apparent discrepancy between Western and 2D-PAGE data may be due to the fact that six isoforms are known of this protein 12. It is highly unlikely that the six isoforms migrate in the same spot on the 2D gel. Generally, the 2D gel will be able to distinguish the isoforms as seen with cofilin-1 (1357 and 1363) and coronin-1A (453, 460, and 467) although the peptides obtained by MS do not contain sufficient information to determine which of the variants 1 to 5 is identified. Only variant 6 can be ruled out. The 1D Western blot, however, generally recognize all the isoforms in one band as seen with cofilin-1 and coronin-1A. The changed expression level of one isoform determined by 2D-PAGE may then be obscured if other isoforms do not change similarly or change in the opposite direction giving no major overall change of the different isoforms taken as a whole on the Western blot. The same may be true for serine/threonine-protein phsophatase 2A 65 kDa regulatory subunit A α isoform (2AAA) 13. Thus, generally our study shows, as others 1415, that the differential regulation of the proteins seen by 2D-PAGE could be quantitatively confirmed by Western blotting in 8 of 11 cases. These eight were epidermal fatty acid binding protein (FABPE), proteasome subunit β type 9 (PSB9), cofilin-1 (COF1), eukaryotic translation initiation factor 3 subunit 4 (IF34), triosephsophate isomerase (TPIS), coronin-1A (COR1A), heterogeneous nuclear ribonucleoprotein A/B (ROAA) and ubiquitin-conjugating enzyme E2 N (UBE2N).

With respect to the five spots containing more than one protein we tried antibodies against UV excision repair protein RAD23 homolog B (472), galectin-7 (1470) and Ras related protein Rab-11A (spot 1216). Unfortunately, the quality of the results was insufficient for firm conclusions.

Each of the identified proteins may be functionally classified based on the Gene Ontology annotation system 20. This system is a structured, organized vocabulary that systematically describes each protein with respect to cellular component, biological process and molecular function. By using the DAVID Bioinformatics Resource 21 we performed a functional annotation clustering of the proteins involved in the transformation. With the highest classification stringency, we found seven clusters for SM5 and eleven clusters for SM7 as shown in Tables 4 and 5. The clusters are listed with the most enriched at top. For example, the 3 proteins PURA2, NKDB and ATPA of the 24 proteins analyzed with the cell line SM5 belong to the cluster purine nucleotide biosynthetic process. The p value at 0.0039 indicates that this function is significantly over represented among the 24 proteins analyzed. Similarly, proteins involved in cellular process, glucose metabolic process, translation factor activity (nucleic acid binding) and the actin cytoskeleton were all significant. Some of the proteins belong to more than one category since they possess multiple functions. The SM7 cell line was in some respects similar to SM5 but in other respects different since functions that were over represented by the transformation included GTP binding, cytoskeleton, cellular process, ubiquitin-dependent protein catabolic process, intracellular transport and GTPase activity. The partly different functions of the two cells may partly explain the different physiology of the two cell lines where SM7 is 3–5 times more tumourigenic compared with SM5 10.

Oncogenesis is not a term within the Gene Ontology classification. However, some of the commonly regulated proteins are known from other studies also to be involved in carcinogenesis and will be briefly described.

In SM5 cells we identified two proteins that were markedly present and clearly originated from bovine plasma, fetuin (308) and albumin (298, 300, 302, 309 and 311). Both proteins were detected from SM5 cells but were only vaguely seen in SM7 cells. The likely differences between the isoforms of BSA are various types of modifications that may occur, e.g. following oxidation 35. These modifications include glycosylations at a number of Lys residues. Since the cells were treated similarly, it is unlikely that the presence of them in one of the cell lines is due to contamination obtained during the preparation of the cells. Endothelial cells have previously been shown to take up fetuin 36. Recently, it was shown that fibroblasts are also able to perform this internalization and that the internalization is down-regulated by hypoxia 37. The present study indicates that especially transformed thymocytes (SM5) may have gained the ability to perform this internalisation. We, however, cannot exclude that the proteins also could adhere to the surface of the cells rather than being internalized.

Using one of the early mouse oncogene arrays, Atlas mouse array I (Clontech) encompassing 588 oncogenes, we have previously identified several differentially regulated transcripts in the two tumour cell lines 89. Only a few overlaps were found between the identified set of proteins differentially expressed here and the previously identified differentially regulated transcripts. These included nucleoside diphosphate kinase B, UV excision repair protein RAD23 homolog B, and ubiquitin regulating enzyme E2.

The performance and analysis of two-dimensional gels have been described in detail previously 10. Briefly, silver stained gels were scanned in the transmissive mode on a GS-710 Imaging Densitometer (Bio-Rad) using Quantity One. The 16-bit gray scale TIFF files were imported into Melanie II. Initially, protein spots were automatically defined and quantified. Spot intensities were expressed as relative volumes in percentages (%VOL) by integrating the optical density of each pixel in the spot area (VOL) and dividing with the sum of volumes of all spots detected. About 10–12 of the spots were used as landmarks. One of the gels was selected as reference gel and used to align and match the other gels. The quality of the match was critically evaluated and necessary editions and corrections done manually. Comparisons were performed between two groups of gels by selecting spots that differed more than two-fold between the groups. These data were exported to Excel for further statistical analysis using Student's t-test as previously described 10.

Protein identification by LC-MS/MS was performed as previsously described 38. Briefly, previously run gels 10 containing protein spots selected for identification were re-hydrated in water. The cellophane sheets were peeled off so the protein spots could be excised from the gels. Proteins were in-gel digested with trypsin. Gel pieces were first dehydrated in acetonitrile, then dried and proteins reduced for 1 h at 56°C in 10 mM dithiotreitol (DTT) and 100 mM NH₄HCO₃. The solution was exchanged with 55 mM iodoacetamide in 100 mM NH₄HCO₃ for 45 min. Then it was washed in 100 mM NH₄HCO₃, dehydrated in acetonitrile, rehydrated in 100 mM NH₄HCO₃, dehydrated in acetonitrile, dried and swelled in digestion buffer (50 mM NH₄HCO₃, 5 mM CaCl₂ and 12.5 ng/μl trypsin Gold (mass spectrometry grade; Promega, Madison, WI, USA). Digestion was performed overnight at 37°C and the peptides were extracted by 1 change of 20 mM NH₄HCO₃ and 3 changes of 5% formic acid in 50% acetonitrile. The sample was finally dried and the peptides resuspended in 12 μl of buffer A (water/acetonitrile/formic acid, 97.7/2/0.3, V/V/V). The peptides were separated on an inert nano LC system composed of a FAMOS micro autosampler, a Switchos micro column switching module and an Ultimate micro pump from LC Packings (San Francisco, CA) before MS analysis. Of the in-gel digested samples 5 μl was preconcentrated and desalted on a 300 μm inner diameter × 5 mm Nano-Precolumn (LC Packings) packed with 5 μm C18 PepMap100 material. A 75 μm inner diameter × 15 cm Nano column packed with 3 μm C18 PepMap100 material was used to separate the peptides. Elution from the column was made with a gradient by mixing decreasing volumes of buffer A with increasing volumes of buffer B (water/acetonitrile/formic acid, 9.7/90/0.3, V/V/V). The peptides were eluted into the nano electrospray ion source of the quadrupole time-of-flight Q-TOF Ultima mass spectrometer (Micromass, Manchester, UK). MS survey scans were acquired using MassLynx 4 SP4 (Waters) at a rate of 1 per second from m/z 400–2000. The instrument was operated in a data-dependent MS to MS/MS switching mode. Doubly, triply and quadruply charged peptide ions detected in MS survey scans triggered a switch to MS/MS for obtaining peptide fragmentation spectra with an interval of m/z values 50–2000. Raw data were processed using ProteinLynx GlobalServer 2.1 (Waters). Processing parameters were as follows: Background Subtract: Normal, Background Threshold: 35%, Background Polynomial: 5, Smoothing Type: Savitzky-Golay, Smoothing Iterations: 2, Smoothing Window: 2 channels, Deisotoping Type: Normal, Deisotoping Threshold: 1%. The processed data were used to search the total part and the mouse fraction of the Swiss-Prot database (releases 49.7, 54.0 and 54.4) using the on-line version of the Mascot MS/MS Ion Search facility (Matrix Science, Ltd., http://www.matrixscience.com) 39. Searching was performed with doubly and triply charged ions with 2 missed cleavages, a peptide tolerance of 20, 50 or 100 ppm, one variable modification either Carbamidomethyl-C or N-terminal acetyl and an MS/MS tolerance of 0.05 or 0.08 Da. Contaminating peptides, including keratins and trypsin were disregarded. At least one 'bold red' peptide was required in the search. Peptides for proteins with scores giving a less than 5% probability that the observed match was a random event are reported.

Normal C57BL/6 mice (wt) were purchased from Bomholtgaard, Ry, Denmark. The mice were kept under conventional conditions at the animal facility at the Panum Institute. The thymus was pressed through a metal net and single cell suspensions were recovered on a foetal calf serum gradient allowing clumps and debris to settle. By haemocytometer examination more than 99% of the cells are lymphoid. Thymocytes were prepared from three mice and pooled. They were then washed extensively in PBS and lyophillized. C57BL/6J-Trp53tm1Tyj mice deficient for the p53 gene were purchased from the Jackson Laboratory, USA. Two spontaneously developing thymic lymphomas, SM5 and SM7, were explanted, in vitro cultured, and established as cell lines growing in RPMI-1640 culture medium supplemented with 10% fetal calf serum and 50 μM 2-mercaptoethanol. Cells were washed extensively in PBS, harvested with a rubber police man and lyophilized. Protein concentration was determined using a non-interfering assay (NI Protein Assay, Geno Technology Inc., St Louis, MO, USA). For 1D gel electrophoresis equal amounts of proteins (either 5, 10 or 30 μg) from each of the cell preparations, wt, SM5 and SM7 were added to each lane of a 10–20% Tris-Glycine gel (Invitrogen). After electrophoresis the separated proteins were transferred to a nitrocellulose membrane. For immunodetection, the membrane was incubated overnight at 4°C in PBS containing 0.05% Tween-20 and 5% skim milk, washed three times in PBS with Tween-20, and incubated with the appropriate dilution of the primary antibody for at least 1 h at room temperature in PBS with Tween-20 and 5% skim milk. After three washes in PBS with Tween-20 the blot was incubated for 1 h with peroxidase-conjugated secondary antibody in PBS with Tween-20. Finally, after five washes in PBS with Tween-20, the blot was developed using enhanced chemiluminescence (Amersham Biosciences, Piscataway, NJ, USA). The bands were visualized on film (Kodak), scanned with a GS-710 Imaging Densitometer from Bio-Rad and the intensities of the bands were evaluated using the Quantity One software package.

Several antibodies were obtained from commercial suppliers. The following rabbit antibodies were used: anti-cofilin (Abcam), dilution 1:10,000 (9 μg/ml); anti-ubiquitin conjugating enzyme E2 N (Calbiochem), dilution 1:10 (50 μg/ml); anti-proteasome 20S LMP2 (Abcam), dilution 1:100 (1 μg/ml); anti-peroxiredoxin 2 (Abcam), dilution 1:2,000 (0.5 μg/ml); anti-transforming acidic coiled coil containing protein 3 (Biolegend), dilution 1:5; anti-NM23-H1/H2 (Alexis) dilution 1:100 (10 μg/ml); anti-RAD23B (Abgent), dilution 1:10 (25 μg/ml), anti-cyclophillin 40 PPID (Abcam), dilution 1:10; anti-galectin 7 (Abcam), dilution 1:30 (33 μg/ml); anti-adenosine deaminase (ADA) (Nordic Biosite), dilution 1:20; anti-Rab11a (Nordic Biosite), dilution 1:60 (2 μg/ml); anti-Ly-GDI (Santa Cruz Biotechnology, INC.), dilution 1:50 (4 μg/ml); anti-serine/threonine protein phosphatase 2a/a (Sigma), dilution 1:10 (10 μg/ml). The following mouse antibodies were used: anti-ADP ribosylation factor (Abcam), dilution 1:50; anti-proteasome 20S α 3 (Abcam), dilution 1:25; anti-Actin like protein 3 (Arp3) (BD Biosciences Pharmingen), dilution 1:100 (2.5 μg/ml); anti-hnRNP-A2/B1 (ImmuQuest, Ltd.) dilution 1:500 (2 μg/ml); anti-vinculin (Abcam), dilution 1:10 (1 mg/ml); anti-Rho GDP-dissociation inhibitor 2 (Rho-GDI2) (Alexis) dilution 1:10 (50 μg/ml). The following goat antibodies were used: anti-coronin 1 (Everest Biotech), dilution 1:100 (5 μg/ml); anti-FABP5 (R & D Systems), dilution 1:10 (20 μg/ml); anti-triosephosphate isomerase (TPI1) (Everest Biotech, Ltd.), dilution 1:1,000 (0.5 μg/ml); anti-eIF3β (Santa Cruz Biotechnology, Inc.), dilution 1:5 (40 μg/ml). The following chicken antibody was used: anti-eukaryotic translation initiation factor 3, subunit 4 δ, 44 kDa (EIF3S4), (GenWay Biotech, Inc.), dilution 1:1,000 (1 μg/ml). Secondary HRP conjugated antibodies were purchased from DAKO (Denmark): anti-rabbit PO217, dilution 1:5,000 (0.26 μg/ml); anti-mouse PO260, dilution 1:5,000 (0.26 μg/ml); anti-goat, PO449, dilution 1:5,000 (0.1 μg/ml) and from Abcam (UK) anti-chicken (ab16349), dilution 1:2,500. Imaging Kodak films were scanned in the transmissive mode on a GS-710 Calibrated Imaging Densitometer from Bio-Rad using the Quantity One software package. Each band was thereby designated a value proportional to the protein concentration in the band. The quality of the antibody reactions was evaluated on the basis of the Western blots where a single strong band indicates a high-affinity specific antibody. Several antibodies were found to give Western blots of insufficient quality in our system either because reactions were to weak or because of the presence of several bands indicating that the antibody possesses low-affinity and/or unspecific reactivity in the given system.