Proteomics analysis of differentially expressed metastasis-associated proteins in adenoid cystic carcinoma cell lines of human salivary gland

Oral Oncology (2004) 40 400–408

http://intl.elsevierhealth.com/journals/oron/

Proteomics analysis of differentially expressed metastasis-associated proteins in adenoid cystic carcinoma cell lines of human salivary gland Jie Ana,1, Jun-yong Sunb,1, Quan Yuana, Hong-yu Tiana, Wei-liu Qiub, Wei Guob,2, Fu-kun Zhaoa,2,* a

Key Laboratory of Proteomics, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China b Department of Oral and Maxillofacial Surgery, Shanghai Ninth People’s Hospital, Shanghai Second Medical University, Shanghai 200011, China

Received 16 September 2003; accepted 28 September 2003

KEYWORDS

Summary Metastasis is the most insidious and life threatening aspect of cancers. However little is known about the molecular mechanisms of tumor metastasis. A poorly metastatic Acc-2 cell line and highly metastatic Acc-M cell line were selected as an experimental model to study on metastatic mechanisms and antimetastatic procedures. In the present study, two-dimensional gel electrophoresis and mass spectrometry are combined to approach the protein profiles associated with tumor metastasis between Acc-2 and Acc-M cell lines. Image analysis of silver stained 2dimensional gels revealed that 12 protein spots showed significantly quantitative and qualitative variations and mass spectrometry is utilized to further identify these spots. Of the identified proteins, transketolase, Dim1p, v-Ha-ras oncogene, type I collagen pro alpha, tumor necrosis factor (ligand) superfamily member 4, and pirin etc, have shown associations with distinct aspect of tumor metastasis to some extent. The dissimilar expression patterns of these 12 spots indicate the different roles they may play involved in tumor metastasis. c 2003 Elsevier Ltd. All rights reserved.

Proteome; Two-dimensional gel electrophoresis; Mass spectrometry; Acc cell; Metastasis




Introduction * Corresponding

author. Tel.: +86-21-54921155; fax: +86-21-

64331090. E-mail address: [email protected] (Fu-kun Zhao). 1 These authors contributed equally to this paper. 2 These authors are joint corresponding authors.




Metastasis is the most insidious and life threatening aspect of cancers. Despite improvements in early diagnosis, surgical techniques, general patient care, and local and systemic adjuvant therapies, most deaths of cancer patients result from

1368-8375/$ - see front matter c 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.oraloncology.2003.09.014

Proteomics analysis of differentially expressed metastasis the relentless growth of metastases that are resistant to conventional therapy.1 Adenoid cystic carcinoma (ACC), an aggressive, often indolent tumor, with a high incidence of distant metastasis (DM) to lung, which are responsible for a rather low long-term survival rate, accounts for approximately 10% of all neoplasms of the salivary glands.2 The value of annual chest films or other tests for the presence of distant metastases of an ACC seems rather questionable.3 So it is important to select reliable quantitative criteria for evaluating metastases of salivary adenoid cystic carcinoma. Relatively little has been reported about the factors that influence distant spread and subsequent survival of ACC because it is uncommon and more than a decade of observation may be required to appreciate the prolonged clinical course in some patients. Experimental model to study should be used in an experimental study on metastatic mechanisms and antimetastatic procedures. In our laboratory a cell line (Acc-2) was established from ACC in 1988,4 and after five repeated selections in vivo, combined with an in vitro cloning technique and analysis of platelet aggregation activity, a clone (Acc-M) highly metastatic to the lung was selected from Acc-2.5 The metastatic rate was 96% vs 18% for Acc-M and Acc-2 cell line. Some studies have been done about two cell lines,6–8 but metastasis is complex which is influenced by the local microenvironment, angiogenesis, stroma–tumor interactions, and elaboration of cytokines by the local tissue, and more significantly by its molecular phenotype.1 So understanding the molecular metastatic process and using that understanding we can develop marker panels, through which to predict presence and location of active metastatic disease, and where possible, to identify and develop therapeutic targets. However, traditional gene analysis could only study one or a few genes once, and is costly and time-consuming. A deeper understanding of mechanism will require expression analysis of multiple genes or proteins simultaneously. Because of this, the proteomics, based on the techniques of twodimensional electrophoresis (2-DE) and mass spectrometer (MS), offer a promising alternative way for its capability of resolving thousands of proteins once. Since the term “proteome” was proposed in 1995 as a concept of the protein complements expressed in a genome, a cell, or a tissue, proteomics has developed rapidly as a field to study proteomes.9 Compared to static genomic, proteomic analysis is necessary for an understanding of dynamic physiological phenomena to analyze systematically their nature and mechanism of action at the very site of occurrence.10 Re-

401

cently, more and more medical and clinical applications of tumor proteomic promised to identify new diagnostic markers and drug targets in different type of cancers such as liver cancer, leukemia, bladder cancer and lung cancer, etc.11 However, only a few studies have been involved in proteomics analysis of tumor metastasis. In this report, we describe protein expression alterations by 2-D gels between Acc-2 and Acc-M cell, a poorly and a highly metastatic cell lines. After image analysis by software, 12 proteins of interest were excised from the gels and identified by matrix assisted laser desorbtion ionization timeof-flight mass spectrometer (MALDI-TOF-MS). Their differential expression patterns imply distinct roles in ACC lung metastatic process. Our work is an initial step toward uncovering the molecular mechanisms of ACC lung metastasis, and maybe contribute to discover its diagnostic markers and therapeutic targets.

Materials and methods Cell culture and sample preparation Two cell lines (Acc-2 and Acc-M) used in the present study was recovered from frozen stock, and both cell lines were incubated in RPMI 1640 + 15% fetal cattle serum at 37
. The cell pellets are got by centrifuge at 1,000 rpm for 2 min and washed three times with ice clod PBS buffer (pH 7.2). The cell pellets were then solubilized in a lysis buffer containing 7 M urea, 2 M Thiourea, 4% CHAPS (w/v), 40 mM Tris-base, 40 mM DTT and 2% IPG (v/v) buffer (pH3–10NL),12 lysised for 1 h using a rocker under room temperature and centrifuged at 45,000 · g for 1 h at 4
C. The supernatant was collected and dispensed, then stored in 100 ll aliquots at )78
C. Protein concentrations were determined using the Bradford assay.13

Two-dimensional gel electrophoresis (2-DE) and multi-strips on one gel (MSOG) method 2-DE was performed mainly according to the method described by manufacture’s instructions14 €rg et al.12 IPG-IEF was run on an and modified by Go IPGphor isoelectric focusing system (Amersham Biosciences). After IEF, the IPG strips were immediately equilibrated for 2 · 15 min with gentle shaking in 10 ml equilibration solutions.12 SDS-PAGE

402 was run in an Ettan DALT twelve apparatus (Amersham Biosciences). After 2-DE, the gels were stained with silver nitrate as described for analytical gels15 and stained with Coomassie Brilliant Blue G250 as described by Neuhoff et al. for preparative gels.16 MSOG method, abbreviated from multi-IPG strips gels (length 6 13 cm) running simultaneously on one SDS polyacrylamide gels (260 · 200 · 1 mm, Ettan Dalt twelve system from Amersham Biosciences), is a improved method in our lab. In brief, after equilibration of the IPG strips in the equilibration buffer, two ends (acidic and basic ends) of each IPG strip were cut approximately 1 cm, respectively. Then, the two cut IPG strips were abreast inserted between the glass plates with a spatula and brought in close contact with the upper edge of one SDS gel. After the sealed agarose was cooled down, the second dimensional SDS-PAGE could be carried out.17

Image acquisition and analysis The silver stained and Coomassie Brilliant Blue stained 2-D gels were scanned with a D2000 Uniscan scanner (Tsinghua Uniscan) in a transmissive mode. Spot detection, quantification and matching were performed with Image Master 2D Elite Version 3.01 software according to the manufacturer’s directions (Amersham Biosciences).

MALDI-TOF peptide mass fingerprinting (PMF) analysis and database searching Proteins for MALDI peptide mass fingerprinting analysis were taken from preparative gels stained for 24 h with a Coomassie Brilliant Blue G250 staining method modified by Neuhoff et al.16 Proteins of interest were excised from the gel by sucking with pipette tips. Each slice was cut into small pieces and was placed into 0.65 ml siliconized tubes. An in-gel digestion was performed by the method described by Shevchenko et al.18 Protein samples digested by trypsin (sequencinggrade modified trypsin, Promega) were analyzed in the delayed-extraction linear positive mode of the Voyager-DE STR MALDI-TOF-MS (PerSeptive Biosystems, Framinham, MA). Identifications of proteins by PMF were performed with the searching programs ProFound. Monoisotopic masses were used and a mass tolerance of 100 ppm was allowed.

J. An et al.

Result Proteomics analysis of differentially expressed proteins between Acc-2 cell and Acc-M cell by 2-DE To study the factors involved in the different metastatic rate between Acc-2 and Acc-M cell lines, we extracted the total proteins of these two cell lines and used large-format, high-resolution 2D gel to monitor changes in the abundance of proteins (Fig. 1). The pH range for the first dimension is pH3–10NL. The second dimension SDS-polyacrylamide gel separation ranges from 10 to 100 kDa. We obtained high resolution (over 2800 spots/gel) and high reproducibility (>95%) 2-D gels. For each group, we integrated six 2-DE silver staining gels and used Image Master 2D Elite Version 3.01 software to analyze the 2-DE image. Through image analysis, we found that 12 protein spots showed significantly differential expression patterns between two cell lines (p < 0:01). These 12 spots showed dissimilar expression patterns and they could be classified into four groups. Spot 1, 3, 4, 5, 7, 8, 9, 11 show a similar expression mode. All were expressed at a low expression level in Acc-2 cells and their expressions were upregulated in Acc-M cells. Spot 2 and 12 belong to another group, whose expressions were low in Acc-M cell and have a upregulated expression in Acc-2 cell. Spot 6 is expressed in Acc2 and not detected in Acc-M cell, while spot 10 is expressed only in Acc-M cell and not detectable in Acc-2 cell. The distinct expression patterns of these 12 spots reveal the different roles they may play in the distant metastasis of ACC.

Further Quantitative comparison of differentially expressed proteins between ACC-2 cell and ACC-M cell by MSOG method Owing to inherently complicated nature of silver staining procedures, spot intensities may easily vary by as much as 20% from batch to batch.19 For comparative proteomic research, the main task is to find the reproducible differential expressed proteins and then identify them. When the diversities of background of samples exist or the intensities of spots vary, it will be hard to accurately quantitate these differentially expressed proteins. And the possibility of taking artifactual spots as real spots is increased.

Proteomics analysis of differentially expressed metastasis

403

Figure 1 Comparison of 2-DE patterns between Acc-2 (A) and Acc-M (B)cell lines. Each 180 lg proteins of the samples were subjected to 2-DE system with cup-loading method (first dimension, IPG strip, pH3–10NL, 24 cm; second dimension, 12.5% SDS-PAGE, 260 · 200 · 1 mm). Proteins were visualized by silver nitrate staining. The 12 indicated proteins have a significantly differential expression between two cell lines.

For this reason, we developed a new IPG-Dalt procedure that multi-IPG strip gels (length 6 13 cm)were run simultaneously on one SDS polyacylamide gel (MSOG). This method not only improves the reproducibility and matching efficiency of 2-DE, but also achieves higher resolution power comparing to the former method called one IPG strip gel running on one gel such as using SE 600 system. Furthermore, the final 2-D patterns showed the artifactual errors could be diminished to minimum.17 With our MSOG method, further comparative analysis was performed between Acc-2 cell and

Acc-M cell, and the protein expression profile was shown in Fig. 2. Fig. 2 indicates the comparative 2DE patterns visualized by silver nitrate staining and over 1900 proteins were detected by image analysis. Fig. 3 shows the magnified comparison patterns of spot6 (A, B) and spot 8(C, D) with different 2-DE methods. Through gel matching, the 12 differentially expressed proteins obtained in the previous method showed the same expression patterns as in MSOG method. The result indicated that the differential expression patterns of 12 proteins obtained previously between Acc-2 and Acc-M cell were convinced.

404

J. An et al.

Figure 2 Comparison of 2-DE patterns between Acc-2 (left) and Acc-M (right) cell lines with MSOG 2-DE method. Each 80 lg proteins of the samples were subjected to 2-DE system with cup-loading method (first dimension, IPG strip, pH3–10NL, 13 cm; second dimension, 12.5% SDS-PAGE, 260 · 200 · 1 mm). Proteins were visualized by silver nitrate staining. The 12 indicated proteins have the same significantly differential expression patterns between two cell lines as in Fig. 1.

Identification of proteins by MALDI-TOF-MS and peptide mass fingerprintings (PMF)

Figure 3 The magnified comparison maps of spot 6 (A, B) and spot 8 (C, D) in the 2-DE patterns between Acc-2 (left) and Acc-M (right) cell lines. Spot 6 (A) was significantly downregulated in Acc-M compared with Acc-2 cell line, in contrast spot 8 (C) was significantly upregulated in Acc-M compared with Acc-2 cell line. The comparison results were further convinced by using MSOG method (B, D).

Based on the comparative analysis of protein expression patterns between Acc-2 and Acc-M cell lines, the 12 differentially expressed proteins were taken from preparative gels stained with a Coomassie Brilliant Blue G250 staining method for MALDI peptide mass fingerprinting analysis. All of them were specifically digested by trypsin and got peptide masses maps from PMF analysis. Table 1 shows the proteins identified by MALDI-TOF-MS and Fig. 4 shows the peptide mass fingerprints of pirin and modulator recognition factor 2. From database searching, spot 3 and spot 5 were characterized as hypothetical proteins, implicating that they might be unknown proteins. In the identified protein candidates, 3-hydroxy-3-methylglutaryl-coenzyme A synthase 1, transketolase, modulator recognition factor 2, Dim1p homolog, splicing factor (arginine/ serine-rich 9) and v-Ha-ras oncogene were all lowly expressed in the poorly metastatic Acc-2 cell and significantly upregulated in highly metastatic AccM cell, while type I collagen pro alpha and tumor necrosis factor (ligand) superfamily member 4 have a high expression in Acc-2 cell and a low expression in Acc-M cell. Pirin (spot 6) just appears in Acc-2 cell and is not detectable in Acc-M cell, while

Proteomics analysis of differentially expressed metastasis

405

Table 1 The list of proteins identified by MALDI-TOF-MS Spot no.a Probabilityb

Sequence coveragec (%)

pId

Mwe (kD)

Relative Expres- Acc. no.g sion level of (NCBIr) proteinf

Protein descriptionh

Acc-M/Acc-2 (n-fold) 1

1.0e + 000

13

5.2

57.85

>3.0

NP_002121

2

2.4e ) 001

19

5.9

31.89

<0.3

AAB27856

3 4

3.5e ) 001 1.0e + 000

16 28

5.8 9.1

42.76 43.22

>4.0 >7.0

XP_028560 AAH08615

5 6 7

4.8e ) 001 9.6e ) 001 1.0e + 000

16 16 30

5.4 6.4 8.5

29.31 32.21 27.96

>5.0 0 >7.0

XP_039443 NP_003653 S27963

8 9

7.6e ) 001 1.0e ) 001

46 12

4.8 8.9

11.0 25.64

>8.0 >4.0

AAB81951 NP_003760

10

4.2e ) 001

16

9.4

37.02

1

NP_038463

11

7.6e ) 002

19

4.8

19.02

>4.0

AAH06499

12

4.3e ) 001

23

7.0

21.26

<0.3

NP_003317

3-Hydroxy-3-methylglutaryl-Coenzyme A synthase 1 Type I collagen pro alpha 1(I) chain propeptide Hypothetical protein Transketolase (Wernicke–Korsak-off syndrome) Hypothetical protein Pirin Modulator recognition factor 2 Dim Ip homolog Splicing factor, arginine/serine-rich 9 Retinal homeobox protein v-Ha-ras Harvey rat sarcoma viral oncogene homolog Tumor necrosis factor (ligand) superfamily, member 4

a

Spot no. were defined according to spot positions in 2-D gel indicated as in Fig. 1. Probability: the possibility that the observed match is a random event. c Sequence coverage: percent of identified sequence to the complete sequence of the known protein. d pI: theoretical isoelectric point of the matching protein. e Mw: theoretical molecular weight of the matching protein in kDa. f Relative expression level of proteins: the ratio of average volume of each spot in Acc-M cell to Acc-2 cell, Average volume of each spot was calculated using 6 gel with Image Master 2D Elite Version 3.01 software. g Acc. no.: NCBInr database accession number. h Protein description: name of each matched protein in NCBInr database. b

retinal homeobox protein is just detected in Acc-M cell and does not appear in Acc-2 cell.

Discussion Interests have been growing over the past years to discover the molecular mechanisms of tumor metastasis. However, due to the complexity of the tumor metastasis, relatively little is known about the key factors of this mysterious phenomena. In the present study, a poorly metastatic Acc-2 cell line and a highly metastatic Acc-M cell line were selected as an experimental model to study on metastasis mechanism and antimetastasis procedure. With comparative proteomic analysis, 12

proteins showing differential expression patterns between Acc-2 and Acc-M cell lines were separated and identified. Intriguingly, most identified proteins have shown associations with distinct aspect of tumor metastasis such as nucleic acid ribose synthesis, cell proliferation, tumor growth, carcinogenesis, angiogenesis, mitosis and apoptosis, etc. Transketolase (TKT) is a thiamine diphosphatedependent enzyme linking the nonoxidative branch of the pentose phosphate pathway (PPP) to the glycolytic pathway. The PPP generates sugar phosphates for intermediary biosynthesis and nucleic acid synthesis and NADPH for reductive biosynthesis.20 Metabolic control analysis indicates that stimulators of transketolase enzyme synthesis

406

J. An et al.

Voyager Spec #1=>BC=>NF0.7[BP = 1474.3,1957] 1474.27

100

1956.7

90 80

% Intensity

70 60

1811.99

50 1649.17

40 30

898.46

20

935.39 736.36 718.40 864.35 908.34 778.20 986.08

10 0

700

1678.22

1349.25

1054.48

1523.15 1231.31

1125.27 1151.38 1181.85

1915.23

2316.46 2091.20

1398.14

1539.12 1724.55 1868.36 2097.84 1376.31 1371.94 1532.13 1686.47 1968.80 2124.95

1260

1820

2345.04 2491.80 2541.02

2788.72 2746.61

2380

(A)

2903.25

3222.68 3100.37

3355.48 0 3500

2940

Mass (m/z)

Voyager Spec #1=>BC=>NF0.7[BP = 1368.4,917] 1368.44

100

917.1

90 80

2694.13 1475.41

70

% Intensity

60

1529.52 1812.32

1349.54

50

1890.20

40 30

729.39

10

2324.06

1231.49

1283.35 1079.70 741.27 723.57 873.35 1124.57 1288.63 814.49 1070.50 1264.33 998.26 767.15 1272.36

20

2577.06

1694.54

707.48

1687.59 1481.46 1567.39

1756.89 1728.31

1931.04 1914.12 1925.74 1940.56

2091.30 2340.77 2157.30 2303.59 2107.08

2512.64 2544.55

2704.08 2756.16 2771.08

3222.86 2925.43

3085.62 2937.81 3091.46

0 700

(B)

1260

1820

2380

2940

3286.03 3295.41 0 3500

Mass (m/z)

Figure 4 Identification of differentially expressed proteins in 2-DE gel with MALDI-TOF peptide mass fingerprinting and database searching. (A) The MALDI peptide mass fingerprints of pirin. (B) The MALDI peptide mass fingerprints of modulator recognition factor 2.

such as thiamin (vitamin B-1) support a high rate of nucleic acid ribose synthesis necessary for tumor cell survival, chemotherapy resistance, and proliferation. and also indicates that transketolase inhibitor drugs will have the opposite effect on tumor cells.21;22 Comin-Anduix et al.23 reported that there is an in vivo tumour growth control coefficient of 0.9 for the thiamine–transketolase complex in mice with Ehrlich’s ascites tumour, and that transketolase enzyme and thiamine clearly determine cell proliferation in the Ehrlich’s ascites tumour model. This high control coefficient allows us to predict that in advanced tumours, which are commonly thiamine deficient, supplementation of thiamine could significantly increase tumour

growth through transketolase activation. In the present study, transketolase has a high expression level in highly metastatic Acc-M cell line compared with pooly metastatic Acc-2 cell line. This result also suggests that transketolase maybe enhance tumor cell proliferation and tumor growth, further enhance tumor metastasis through metabolic pathway. All of these may have important implications in the nutrition and future treatment of patients with cancer. Limited administration of thiamine and concomitant treatment with transketolase inhibitors will be a more rational approach to treat cancer and avoid metastasis.24 The Ras superfamily of GTPases act as important regulatory switches to co-ordinate extracellular

Proteomics analysis of differentially expressed metastasis stimuli with activation of intracellular signaling pathways and appropriate biological responses. Ras proteins can activate at least three downstream signaling cascades mediated by the Raf-MEKextracellular signal-regulated kinase (ERK) family, phosphatidylinositol-3 (PI3) kinase, and Ral-specific guanine nucleotide exchange factors (RalGEFs).25 The Ras branch of this superfamily includes H-, Kand N-Ras, which are commonly mutated in particular human cancers. Ichikawa et al.26 reported that v-H-ras transfection can result in the development of metastatic ability in rat mammary cancer cells and the reason may be that the acquisition of metastatic ability following v-Ha-ras transfection involves loss of metastasis suppressor gene function. The ability of Ras to induce vascular endothelial growth factor has suggested that Ras might affect metastatic ability through induction of angiogenesis27 and treatment involving the targeting of ras oncogene could inhibit solid tumor growth by suppressing tumor-associated angiogenesis. Functional studies utilizing both in vitro and in vivo models demonstrate that Ras signaling can regulate a variety of endpoints relevant to breast cancer progression, including anchorage dependent and independent growth, tumorigenesis, steroid sensitivity and invasion.28 It is also reported that some tumor cell lines transfected with v-H-ras has potential ability of lung metastasis29;30 and these studies are identical to our results. In the present study the expression of v-H-ras protein is low in poorly metastic Acc-2 cell line and upregulated to a high level in Acc-M cell line which has a highly metastatic potential to lung. All of these indicate that H-ras is one of the key factors associated with metastasis and maybe play a key role in the procession of tumor metasatsis. These findings have far-reaching implications for the prevention of tumour growth, invasion and metastasis. Therefore, analysis of the processing and signaling mechanisms of the Ras superfamily will identify potential targets for therapeutic intervention. Pirin is a novel highly conserved nuclear protein and an interactor of nuclear factor I/CCAAT box transcription factor (NFI/CTF1), which is known to stimulate adenovirus DNA replication and RNA polymerase II-driven transcription.31 The expression of pirin was lowered after inhibition of MEK1 and signalling via the protein kinase Raf-MEK-ERK pathway is of major importance for transformation by oncogenes. It was reported that pirin has an increased expression in RAS and c-JUN transformed cell.32 Human pirin, a nuclear factor reported to interact with the human oncogene Bcl-3, stabilizes the formation of quaternary complexes between Bcl-3, the antiapoptotic transcription factor NF-

407

kappaB and its DNA target sequences in vitro. As a novel Bcl-3 interactors like Jab1, Tip60 and Bard1, pirin also binds to other transcription factors including c-Jun, nuclear factor I (NFI), HIV-1 Tat or the tumor suppressor and PolII holoenzyme component Brca1, respectively.33 In the present study, pirin and Ras maybe function through Raf-MEK-ERK signal pathway,25 however pirin maybe function through mediating NF-kappa B activation just like tumor necrosis factor (ligand) superfamily.34 Whether pirin play its role through Raf-MEK-ERK pathway cooperating with H-Ras or through mediating NF-kappa B activation cooperating with TNF ligand superfamily is still unclear and needed to be elucidated further, but all of the studies showed that pirin maybe play a very important role in tumor metastasis. In conclusion, the proteins we identified may be involved in the adenoid cystic carcinoma lung metastasis through different mechanisms. However, further studies are still needed to elucidate the real roles of these differentially expressed proteins. Our work is an initial step toward uncovering the molecular mechanism of tumor metastasis, and maybe contribute to discover diagnostic markers and therapeutic targets.

Acknowledgements This work was supported by the National Natural Science Foundation of China (Grant No. 39990600) and the Foundation of Chinese Academy of Sciences (Grant No. KSCX 2-2-06) and Shanghai Sciences of Developing Foundation (Grant No. ODJC14018).

References 1. Lance A. Liotta, Elise C. Kohn. Cancer biology. In: Holland F, editor. Cancer medicine, 5th ed. Harcout Asia Pet. Ltd, 2001, p. 121. 2. Spiro RH. Distant metastasis in adenoid cystic carcinoma of salivary origin. Am J Surg 1997;174(5):495–8. 3. Van der Wal JE, Becking AG, Snow GB, Van der Wal I. Distant metastases of adenoid cystic carcinoma of the salivary glands and the value of diagnostic examinations during follow-up. Head Neck 2002;24(8):779–83. 4. He RG, Qiu WL, Zhou XJ. The establishment of Acc-2 and Acc-3 and their morphological observation. J West China Stromatol 1988;6(1):1–4. 5. Guan XF, Qiu WL, He RG, Zhou XJ. Selection of adenoid cystic carcinoma cell clone highly metastatic to the lung: an experimental study. Int J Oral Maxillofac Surg 1997;26(2): 116–9. 6. Shi H, He RG, Zhou Z. Effects of laminin on the invasive and metastatic behaviors of human salivary adenoid cystic

408

7.

8.

9.

10.

11. 12.

13. 14.

15.

16.

17.

18.

19. 20.

21.

carcinoma cell lines. Zhonghua Kou Qiang Yi Xue Za Zhi 1999;34(1):25–7. Sun CX, He RG, Cheung LK, Zhang ZY, Chen WT, Liu XK, et al. The biological behaviour of human adenoid cystic carcinoma cells transduced with interleukin-2-gene. Int J Oral Maxillofac Surg 2002;31(6):650–6. Huang D, Chen WT, Zhang ZY, Zhang P, He RG, Zhou XJ, et al. Identification of genes with consistent expression alteration pattern in ACC-2 and ACC-M cells by cDNA array. Chin Med J (Engl) 2003;116(3):448–52. Wasinger VC, Cordwell SJ, Cerpa-Poljak A, Yan JX, Gooley AA, Wilkins MR, et al. Progress with gene-product mapping of the Mollicutes: Mycoplasma genitalium. Electrophoresis 1995;16:1090–4. Martzen MR, McCraith SM, Spinelli SL, Torres FM, Fields S, Grayhack EJ, et al. A biochemical genomics approach for identifying genes by the activity of their products. Science 1999;286:1153–5. Herrmann PC, Liotta LA, Petricoin EF. Cancer proteomics: the state of the art. Dis Markers 2001;17:49–57. €rg A, Obermaier C, Boguth G, Harder A, Scheibe B, Go Wildgruber R, et al. The current state of two-dimensional electrophoresis with immobilized pH gradients. Electrophoresis 2000;21:1037–53. Kruger NJ. In: Walker JM, editor. The Protein protocols handbook. Totowa, NJ: Humana Press Inc.; 1996. p. 15–20. Berkelman T, Stenstedt T. 2-D electrophoresis: using immobilized pH gradients; principles and methods. Uppsala, Sweden: Amersham Biosciences; 1998. Yan JX, Wait R, Berkelman T, Harry RA, Westbrook JA, Wheeler CH, et al. A modified silver staining protocol for visualization of proteins compatible with matrix-assisted laser desorption/ionization and electrospray ionizationmass spectrometry. Electophoresis 2000;21:3666–72. Neuhoff V, Arold N, Taube D, Ehrhardt W. Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis 1988;9:255–62. Yuan Q, An J, Liu DG, Zhao FK. Multi-strips on one gel method to improve the reproducibility, resolution power and high-throughput of two-dimensional electrophoresis. Acta Biochim Biophys Sinica 2003;35(7):611–8. Shevchenko A, Wilm M, Vorm O, Mann M. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal Chem 1996;68:850–8. Quadroni M, James P. Proteomics and automation. Electrophoresis 1999;20:664–77. Xu Z-P, Wawrousek Eric F, Piatigorsky J. Transketolase haploinsufficiency reduces adipose tissue and female fertility in mice. Mol Cell Biol 2002:6142–7. Boren J, Montoya AR, de Atauri P, Comin-Anduix B, Cortes A, Centelles JJ, et al. Metabolic control analysis aimed at the ribose synthesis pathways of tumor cells: a new strategy for

J. An et al.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

antitumor drug development. Mol Biol Rep 2002;29(1–2): 7–12. Cascante M, Centelles JJ, Veech RL, Lee WN, Boros LG. Role of thiamine (vitamin B-1) and transketolase in tumor cell proliferation. Nutr Cancer 2000;36(2):150–4. Comin-Anduix B, Boren J, Martinez S, Moro C, Centelles JJ, Trebukhina R, et al. The effect of thiamine supplementation on tumour proliferation: a metabolic control analysis study. Eur J Biochem 2001;268(15):4177–82. Boros LG, Brandes JL, Lee WN, Cascante M, Puigjaner J, Revesz E, et al. Thiamine supplementation to cancer patients: a double edged sword. Anticancer Res 1998; 18(1B):595–602. Ward Y, Wang W, Woodhouse E, Linnoila I, Liotta L, Kelly K. Signal pathways which promote invasion and metastasis: critical and distinct contributions of extracellular signalregulated kinase and Ral-specific guanine exchange factor pathways. Mol Cell Biol 2001;21(17):5958–69. Ichikawa T, Ichikawa Y, Isaacs JT. Genetic factors and suppression of metastatic ability of v-Ha-ras-transfected rat mammary cancer cells. Proc Natl Acad Sci USA 1992;89(5): 1607–10. Larcher F, Robles AI, Duran H, Murillas R, Quintanilla M, Cano A, et al. Up-regulation of vascular endothelial growth factor/vascular permeability factor in mouse skin carcinogenesis correlates with malignant progression state and activated H-ras expression levels. Cancer Res 1996;56(23): 5391–6. Malaney S, Daly RJ. The ras signaling pathway in mammary tumorigenesis and metastasis. J Mammary Gland Biol Neoplasia 2001;6(1):101–13. Endo Y, Seiki M, Uchida H, Noguchi M, Kida Y, Sato H, et al. Experimental metastasis of oncogene-transformed NIH 3T3 cells in chick embryo. Jpn J Cancer Res 1992;83(3):274–80. Virone A, Monier R, Zerial A, Lavelle F, Feunteun J. Metastatic phenotype of murine tumor cells expressing different cooperating oncogenes. Int J Cancer 1992;51(5): 798–804. Zeng Q, Li X, Bartlam M, Wang G, Pang H, Rao Z. Purification, crystallization and preliminary X-ray analysis of human pirin. Acta Crystallogr D Biol Crystallogr 2003;59(8):1496–8. Bergman AC, Alaiya AA, Wendler W, Binetruy B, Shoshan M, Sakaguchi K, et al. Protein kinase-dependent overexpression of the nuclear protein pirin in c-JUN and RAS transformed fibroblasts. Cell Mol Life Sci 1999;55(3):467–71. Dechend R, Hirano F, Lehmann K, Heissmeyer V, Ansieau S, Wulczyn FG, et al. The Bcl-3 oncoprotein acts as a bridging factor between NF-kappaB/Rel and nuclear co-regulators. Oncogene 1999;18(22):3316–23. Han S, Yoon K, Lee K, Kim K, Jang H, Lee NK, et al. TNFrelated weak inducer of apoptosis receptor, a TNF receptor superfamily member, activates NF-kappa B through TNF receptor-associated factors. Biochem Biophys Res Commun 2003;305(4):789–96.