European Archives of Oto-Rhino-Laryngology

, Volume 269, Issue 6, pp 1653–1663

Gene expression analysis of SCC tumor cells in muscle tissue

Authors

    • Department of RadiologyPhilipps University Marburg
  • Esther L. Yuh
    • Department of Radiology, Lucas CenterStanford University
  • Mykhaylo Burbelko
    • Department of RadiologyPhilipps University Marburg
  • Andreas Kiessling
    • Department of RadiologyPhilipps University Marburg
  • Mark D. Bednarski
    • Department of Radiology, Lucas CenterStanford University
  • Silke Steinbach
    • Department of Otolaryngology Head and Neck SurgeryPhilipps University Marburg
Article

DOI: 10.1007/s00405-011-1799-0

Cite this article as:
Hundt, W., Yuh, E.L., Burbelko, M. et al. Eur Arch Otorhinolaryngol (2012) 269: 1653. doi:10.1007/s00405-011-1799-0
  • 120 Views

Abstract

The purpose of this study was to evaluate microarray technology of HNSCC cells in muscle tissue. 200 SCCVII tumor cells were injected intramuscularly into the right flank of ten C3H/Km mice each. One week later the animals were killed and the tissue taken out. Histology (H&E staining) and microarray of the tissue were performed. Histology showed a few tumor cells between the muscle fibers. Microarray technology showed different gene expression pattern of the muscle tissue with SCCVII cells in comparison with normal muscle tissue. Only those genes showing a fold change difference of 5 or higher were considered. Gene expression analysis revealed changes in the expression levels of SCCVII cells in muscle tissue in 220 genes. Significant gene expression differences between SCCVII cells in muscle tissue and pure muscle tissue could be seen.

Keywords

SCCVII tumor cellsMuscle tissueHistologyGene expression

Introduction

Head and neck malignancies account for 6% of all cancers diagnosed in the United States and result in an estimated 14,000 deaths annually [1]. Although improvements in local control and survival have been achieved with the use of combined modality therapies, 5-year survival rates for head and neck cancers have not improved significantly over the past 20 years [2, 3]. Local–regional relapse after definitive therapy is a major cause of morbidity and mortality in patients with head and neck squamous cell carcinoma (HNSCC). The local relapse is in most cases due to micrometastases either in lymph nodes or the surrounding muscle tissue [4]. Unfortunately, current routine clinical and pathological methods of detecting metastasis are suboptimal for identifying the presence of micrometastases and may lead to the understanding of many patients with head and neck squamous cell carcinoma (HNSCC) [5, 6]. The limitations of routine pathology for detecting micrometastatic disease have made it necessary to explore molecular means of diagnosis that can detect disease through tissue sampling. Molecular detection of HNSCC cells in a background of surrounding muscle tissue demands highly specific and sensitive biomarkers. Ideally, these biomarkers would be abundantly yet exclusively expressed in squamous epithelium, whereas having negligible expression in muscle tissue. One method for the molecular detection of these biomarkers that has shown promise in recent studies is microarray technology. High-throughput gene expression profiling techniques offer a unique sensitive mechanism for interrogating transcriptome-wide levels of gene expression. It provides the ability to perform qualitative and quantitative analysis for biomarkers with great sensitivity and from minute amounts of starting material. Because this technology is sensitive, it offers the potential to improve clinical decision making [710]. Squamous cell carcinoma (SCCVII) is a syngeneic tumor cell line of the C3H mice and has been used extensively as a model for human head and neck cancers. This tumor models exhibits characteristics that are similar to human head and neck carcinomas [11]. In this study, we applied gene expression microarray technology to muscle tissue after injection of squamous cell carcinoma tumor cells. We evaluated whether functional genomics can be used as method to detect few tumor cells in muscle tissue.

Materials and methods

All animal experiments were performed in compliance with institutional animal care committee guidelines and with the approval of the animal care committee.

Cell implantation into the muscle tissue

Ten C3H/Km mice aged 10–12 weeks were anesthetized with intraperitoneal Nembutal (58 mg/kg) and their right flank was shaved and prepped with isopropyl alcohol. An average of 200 tumor cells (mouse squamous cell carcinoma VII—SCCVII) in Hanks’ solution were injected intramuscular of the right flank region of each mouse with a 27 G needle. The total volume of injection was 100 μl. One week later the animals were euthenized by CO2 asphyxiation and the muscle tissue taken out. As control muscle tissue of the contralateral side of five animals was used and muscle tissue after injection of 100 μl saline in order to investigate tissue changes due to muscle tissue.

Histology

In order to show SCCVII cells in muscle tissue Hematoxylin and Eosin (H&E) staining was performed. For the H&E staining the tissue samples were preserved in 10% formalin solution for 96 h. Afterwards they were embedded in paraffin, sectioned, stained with hematoxylin and eosin, and then mounted on the glass slides.

Microarray analysis

For microarray analysis tissue samples of the muscle tissue with SCCVII cells, muscle tissue after injection of 100 μl saline and only muscle tissue was used. The tissue samples for microarray analysis were deep frozen at a temperature of −80°C. In total ten muscle-tissue samples with SCCVII cells and five muscle tissue samples after injection of saline and five pure muscle tissue samples were analyzed. The total RNA was isolated using TRIzol Reagent® (GibcoBRL Life Technologies, Rockville, MD) and the double-stranded cDNA created by using the SuperScript Choice system (Life Technologies). In further steps the cDNA was extracted and precipitated. Biotinilated cRNA was synthesized using Enzo Bio Array High Yield RNA Transcript Labeling Kit (Enzo Diagnostics Inc., Farmingdale, NY). After incubation the labeled cRNA was cleaned up according to the RNeasy Mini kit protocol (Qiagen). The cRNA was fragmented and hybridized on the murine Genome U74Av2 set array. The chips were washed and stained with streptavidin phycoerythrin (SAPE; Molecular Probes, Eugene, OR). To amplify staining, streptavidin phycoerythrin solution was added twice with an antistreptavidin biotinylated antibody (Vector Laboratories, Burlingame, CA) staining step in between. The probe array was scanned on a Hewlett-Packard confocal microscope scanner (Hewlett Packard Gene Array Scanner; Hewlett Packard Corporation, Palo Alto, CA) at the excitation wavelength of 488 nm. The amount of light emitted at 570 nm was proportional to bound target at each location on the probe array. After hybridization and scanning, the microarray images were analyzed using Microarray Suite 4.0® (Affymetrix Inc. Santa Clara, CA) and Gene spring 4.0® (Silicon Genetics, Inc., Redwood City, CA) software. All samples were prepared as described and hybridized onto the Affymetrix Murine Genome U74Av2 Set array, which represents nearly 36,000 full-length murine genes and EST sequences. Each chip contains 16–20 oligonucleotide probe pairs per gene or cDNA clone.

Analysis of microarray data

Preprocessing of microarray data

Preprocessing of the Affymetrix arrays was carried out using GeneData Refiner 3.06 software to correct for variations in hybridization intensity because of gradient effects, dust specks, or scratches. Gene expression intensity for each array was scaled to an arbitrary value of 1,500 intensity units to allow comparisons across all arrays. Each tissue sample was analyzed once, getting one result of fold change by comparing the SCCVII cells in muscle tissue samples with pure muscle tissue. The mean value and standard deviation of each analyzed tissue sample group were calculated. The FC calculations include a series of statistical parameters considering background and noise intensity within each gene chip. Expression intensity values for each gene were derived using Refiner by applying the Microarray Suite 5.0 algorithm. The software calculates an average of the two images, defines the probe cells, and computes intensity for each cell.

Statistical analysis

Genes differentially expressed between the SCCVII cells containing muscle tissue compared with normal muscle tissue were identified using a Satterthwaite t test to robustly estimate significance despite unequal variance among groups. Genes that merited additional investigation met the following criteria: p < 0.001, absolute value of the difference in mean expression between the two groups of samples (Δ) > 1,500 intensity units, and fold difference in mean expression ≥5.0.

Results

Histology

In the muscle tissue after injection of SCCVII tumor cells no tumor cell nests could be found; however, only a few single SCCVII cells (arrow) between the muscle fibers could be detected; otherwise, histology showed normal muscular tissue structure (H&E) (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00405-011-1799-0/MediaObjects/405_2011_1799_Fig1_HTML.jpg
Fig. 1

H&E staining of muscle tissue after injection of tumor cells. A few tumor cells in the tissue could be found (arrow)

Microarray analysis

Gene expression analysis revealed changes in the expression levels of SCCVII cells containing muscle tissue compared with pure muscle tissue in 220 genes being different. The up-regulated genes in SCCVII cells containing muscle tissue compared with pure muscle tissue are related to different functional groups. They belong to immune response (Table 1), protein binding (Table 2), receptor activity (Table 3), membrane function (Table 4), cell matrix (Table 5), cell growth (Table 6), cell core and nucleotide activity (Table 7), calcium binding (Table 8), enzyme activity (Table 9), and lipid metabolism (Table 10).
Table 1

Genes related to immune response in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Immune response

Gene Name

Acc. #

FC ± SD

Complement component 1, q subcomponent, c polypeptide

X66295

108 ± 14*

Complement component 1, q subcomponent, alpha polypeptide

X58861

103 ± 14*

Major histocompatibility locus class III regions Hsc70t gene

AF109905

25 ± 3.1*

Granulin

D16195

25 ± 3.5*

Mouse MHC (Qa) Q2k gene for class I antigen, exons 1–3

X58609

17 ± 1.4*

Histocompatibility 2, K region locus 2

M27134

17 ± 1.4*

Major histocompatibility locus class III regions Hsc70t gene

AF109905

12 ± 3*

Major histocompatibility complex region

AF110520

12 ± 2.5*

Mouse mRNA with a set 1 repetitive element for a class I major histocompatibility complex(MHC) antigen

X00246

11 ± 1.2*

CD14 antigen

X13333

11 ± 1.0*

Histocompatibility 2. T region locus 23

Y00629

9.5 ± 3.9*

Mouse Q4 class I MHC gene (exon 5)

X16202

9.4 ± 2.5*

Mouse heat shock protein 86 mRNA

J04633

9.3 ± 5.3*

Synthetic D(b) mRNA for D(b) glycoprotein (alpha chain) and beta(2) microglobulin

X52490

8.9 ± 2.7*

Histocompatibility 2. D region locus 1

M69069

8.4 ± 2.9*

Beta-2 microglobulin

X01838

7.8 ± 2.0*

Small inducible cytokine A9

U49513

7.6 ± 4.1*

Protease (prosome. macropain) 28 subunit. alpha

AB007136

7.5 ± 6.8*

Mus musculus H-2K gene for MHC class I antigen H-2K (allele b). Exons 1 to 3 and joined CDS

V00746

6.6 ± 3.1*

Mus musculus mSTI1 mRNA

U27830

5.7 ± 4.5*

Antigen identified by monoclonal antibodies 4F2

X14309

5.2 ± 3.8*

Table 2

Genes related to protein binding in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Protein binding

Gene Name

Acc. #

FC ± SD

S100 calcium-binding protein A4

M36579

108 ± 7.6*

Calpactin I light chain

M16465

24 ± 2.2*

S100 calcium-binding protein A6

X66449

23 ± 2.9*

Lipocortin 1

M69260

21 ± 1.7*

HSP47 mRNA

X60676

21 ± 3.1*

Procollagen C-proteinase enhancer protein

X57337

20 ± 1.8*

Mac-2 antigen

X16834

20 ± 3.3*

Mouse mRNA for I-E(beta-b) gene

X00958

17 ± 3.5*

Mouse calpactin I heavy chain (p36) mRNA

M14044

17 ± 1.8*

Procollagen, type VI, alpha 2

Z18272

14 ± 1.0*

Mouse mRNA for 14-3-3 zeta

D83037

14 ± 2.5*

Lamin A

D49733

11 ± 5*

Immunosuperfamily protein Bl2 mRNA

AF061260

10 ± 4.1*

Heparan sulfate proteoglycan 1, cell surface-associated (fibroglycan)

U00674

10 ± 4*

Interferon activated gene 204

M31419

10 ± 3.1*

Thioredoxin

X77585

7.2 ± 5.3*

Mus musculus connexin 43 (alpha-1 gap junction) mRNA

M63801

7.0 ± 4.3*

M.musculus (129/Sv) Ccte mRNA for CCT (chaperonin containing TCP-1) epsilon subunit

Z31555

6.8 ± 5.8

Prothymosin beta 4

U38967

6.7 ± 5.3

Mus musculus WW domain binding protein 5 mRNA

U92454

5.9 ± 5.4

Mouse G protein beta 2 subunit mRNA

U34960

5.7 ± 4.9

Table 3

Genes related to receptor activity in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Receptor activity

Gene Name

Acc. #

FC ± SD

mRNA for 4F2/CD98 light chain

AB017189

17 ± 2*

Peptidylprolyl isomerase C-associated protein

X67809

14 ± 1.4*

Mannose receptor, C type 1

Z11974

14 ± 1.2*

Vascular cell adhesion molecule 1

M84487

13 ± 1.9*

Mouse Ia-associated invariant chain (Ii) mRNA fragment

X00496

12 ± 2.3*

Solute carrier family 4 (anion exchanger), member 2

J04036

11 ± 5*

Interleukin 10 receptor, beta

U53696

11 ± 4*

Mouse interleukin-4 receptor (secreted form) mRNA

M27960

9.9 ± 6.4*

Benzodiazepine receptor peripheral

D21207

8.9 ± 7.7*

Mus musculus connexin 43 (alpha-1 gap junction) mRNA

M63801

7.0 ± 4.3*

Low density lipoprotein receptor related protein

X67469

6.9 ± 3.5*

Cd63 antigen

D16432

6.7 ± 5.1*

Ferritin light chain 1

L39879

5.7 ± 4.5*

Mus musculus endobrevin mRNA

AF053724

5.4 ± 4.9*

Table 4

Genes related to the cell membrane in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Membrane

Gene Name

Acc. #

FC ± SD

Retinoic acid-inducible E3 protein mRNA,

U29539

12 ± 2.0*

Plp2 mRNA for proteolipid protein 2

AB031292

11 ± 3*

Mus musculus mRNA for myeloid-associated differentiation protein

AJ001616

9.5 ± 8.2*

Mus musculus mRNA for peroxisomal integral membrane protein PMP34

AJ006341

8.5 ± 6.5*

Histocompatibility 2. Blastocyst

U21906

7.8 ± 1.2*

TPA-regulated locus

M23568

5.9 ± 4.8*

Integral membrane protein 2 B

U76253

5.0 ± 2.9*

Table 5

Genes related to the cell matrix in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Cell matrix

Gene Name

Acc. #

FC ± SD

Cytokeratin (endoB) gene

M22832

68 ± 12*

mRNA for PGI (biglycan)

X53928

29 ± 2.7*

Cofilin 1

D00472

24 ± 2.6*

Procollagen, type XVIII, alpha 1

U03715

24 ± 4*

Mouse fibronectin (FN) mRNA

M18194

21 ± 2.1*

mbh1 gene for Myc basic motif homologue-1 (mbh1)

X54511

20 ± 2.5*

Transforming growth factor, beta induced, 68 kDa

L19932

20 ± 2.5*

Mouse tropomyosin isoform 2 mRNA

M22479

18 ± 2.0*

Mouse procollagen type V alpha 2 (Col5a-2) mRNA

L02918

13 ± 1.1*

Type VI collagen alpha 3 subunit mRNA

AF064749

11 ± 3.1*

Procollagen, type VI, alpha 1

X66405

11 ± 1.0*

Mouse alpha-tubulin isotype M-alpha-7 mRNA

M13443

9.5 ± 2.0*

Tubulin, beta 2

M28739

9.4 ± 7.8*

Tubulin alpha1

M28729

7.6 ± 1.1*

Tubulin alpha 2

M28727

6.6 ± 1.3*

Mouse cytoskeletal gamma-actin mRNA

M21495

6.3 ± 1.6*

Procollagen, type III, alpha 1

X52046

6.3 ± 7.8*

Procollagen, type I, alpha 2

X58251

5.9 ± 9.5*

Tubulin alpha1

M28729

5.8 ± 1.3*

Procollagen, type IV, alpha 2

X04647

5.2 ± 3.6*

Procollagen, type IV, alpha 1

M15832

5.6 ± 2.1*

Table 6

Genes related to cell growth in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Cell growth

Gene Name

Acc. #

FC ± SD

Epithelial membrane protein 3

U87948

29 ± 2.7*

Thymic shared antigen-1 (TSA-1) gene

U47737

29 ± 2.5*

Epithelial membrane protein 1

X98471

16 ± 1.5*

Cyclin B2

X66032

12 ± 1.9*

Stromal cell-derived factor 1

L12029

11 ± 1.1*

F52 mRNA for a novel protein

X61399

10 ± 3.1*

Mouse mRNA for SCID complementing gene 2

D78188

9.6 ± 6.3*

House mouse; Musculus domesticus adult thymus mRNA for Mer protein

D73368

9.1 ± 6.8*

Platelet derived growth factor. alpha

M29464

6.5 ± 1.2*

M.musculus mRNA for centrin gene

Y12474

6.0 ± 4.4*

Epithelial membrane protein 3

U87948

5.8 ± 1.3*

Mus musculus WD40-repeat type I transmembrane protein A72.5 mRNA

U67327

5.6 ± 4.1*

Hypoxia inducible factor 1

Y09085

5.8 ± 3.1*

Mouse cytoplasmic beta-actin mRNA

M12481

5.1 ± 1.5*

Mouse mRNA for cysteine-rich glycoprotein SPARC

X04017

5.1 ± 4.3*

Table 7

Genes related to cell core and nucleotide activity in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Cell core and nucleotide activity

Gene Name

Acc. #

FC ± SD

Nsp-like 1 protein (Nspl1) gene, complete cds; tRNA-Sec gene, complete sequence; and FosB protein (Fosb) gene

AF093624

28 ± 6.9*

Mouse mRNA for beta-tubulin (isotype Mbeta 5)

X04663

21 ± 2.2*

Dynamin

L31397

18 ± 1.7*

Proliferating cell nuclear antigen

X57800

16 ± 1.0*

Mouse histone H2A.1 gene

M33988

15 ± 1.7*

Adenylate kinase isozyme 2

AB020202

15 ± 1.5*

RA70

AB014485

12 ± 1.1*

Sec61 mRNA

AB032902

12 ± 1.1*

eIF3 p66

AB012580

12 ± 1.1*

Small nuclear ribonucleoprotein D1

M58558

10 ± 4.1*

Mouse alpha-tubulin isotype M-alpha-6 mRNA

M13441

10 ± 4.1*

High mobility group protein 14

X53476

10 ± 3.5*

Murine mRNA for replacement variant histone H3.3

X13605

9.6 ± 7.9*

M.musculus rhoC mRNA

X80638

9.4 ± 6.9*

M.musculus mRNA for poly(A)-binding protein

X65553

9.3 ± 4.5*

House mouse; Musculus domesticus adult thymus mRNA for Mer protein

D73368

9.2 ± 6.7*

Ngfi-A binding protein 1

U47008

8.0 ± 2.0*

Interferon regulatory factor 1

M21065

7.8 ± 3.6*

Mouse ERp99 mRNA encoding an endoplasmic reticulum transmembrane protein

J03297

7.7 ± 2.0*

Lamin A

D49733

7.5 ± 6.7*

Mus musculus mRNA for sid23p

AB025406

7.3 ± 2.4*

Mus musculus histone H2A.Z (H2A.Z) mRNA

U70494

7.2 ± 2.0*

Mus musculus high mobility group protein homolog HMG4 (Hmg4) mRNA

AF022465

6.7 ± 4.5*

Nucleolin

X07699

6.3 ± 4.2*

Mus musculus fertilization antigen-1 mRNA

U95114

6.2 ± 2.8*

M.musculus mRNA for small nuclear ribonucleoprotein E

X65704

6.3 ± 2.7*

FBJ osteosarcoma oncogene

V00727

6.2 ± 1.3*

M.musculus mRNA for TIF1 beta protein

AI847564

6.1 ± 1.9*

Mouse mRNA for hepatoma-derived growth factor (HDGF)

D63707

6.1 ± 1.1*

Mus musculus rbm3 mRNA

AB016424

5.8 ± 4.6*

Mouse nucleolar protein N038 mRNA

M33212

5.3 ± 2.3*

Mus musculus TAFII30 gene for mTAFII30 protein. exons 1-5

AJ249987

5.2 ± 1.0*

Mus musculus mRNA for 49 kDa zinc finger protein

AB013357

5.2 ± 2.8*

Table 8

Genes related to calcium binding in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Calcium binding

Gene Name

Acc. #

FC ± SD

Endothelial monocyte-activating polypeptide I mRNA

U41341

29 ± 3.9*

Follistatin-like

M91380

15 ± 1.1*

Matrix gamma-carboxyglutamate (gla) protein

D00613

12 ± 1.0*

Mouse mRNA for annexin V

D63423

11 ± 1.0*

Fibrillin 1

L29454

9.5 ± 2.2*

Calumenin

U81829

5.8 ± 3.6*

Table 9

Genes related to enzyme activity in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Enzym activity

Gene Name

Acc. #

FC ± SD

Cathepsin S

AJ223208

54 ± 6.1*

Mouse lysozyme M gene

M21050

43 ± 5.5*

TYRO protein tyrosine kinase-binding protein

AF024637

31 ± 3.7*

Colony-stimulating factor 1 receptor

X06368

27 ± 3.2*

CPP32 apoptotic protease mRNA

U63720

26 ± 2.5*

Cystatin B

U59807

19 ± 1.5*

Large multifunctional protease 7

U22033

18 ± 2.2*

Antioxidant enzyme AOE372 mRNA

U96746

16 ± 1.7*

Mouse cytochrome beta-558 mRNA, 3 end

M31775

16 ± 1.6*

Mouse MHC class I D-region cell surface antigen (D2d) gene

M27034

15 ± 1.6*

Fumarylacetoacetate hydrolase

Z11774

15 ± 1.3*

Legumain

AJ000990

15 ± 1.1*

Putative steroid dehydrogenase (KIK-I) mRNA

AF064635

14 ± 2.3*

Pigment epithelium-derived factor (PEDF) mRNA

AF036164

12 ± 1.3*

Spermidine/spermine N1-acetyl transferase

L10244

12 ± 1.0*

Protein-tyrosine phosphatase mRNA

AF013490

11 ± 6*

Protein inhibitor of nitric oxide synthase (PIN) mRNA

AF020185

11 ± 1.3*

Exostoses (multiple) 1

X96639

11 ± 1.2*

Alpha-mannosidase II

X61172

10 ± 3*

Mus musculus PAF acetylhydrolase mRNA

U34277

9.1 ± 1.0*

Mouse 3 mRNA for beta-galactoside-specific lectin (14 kDa)

X15986

8.7 ± 2.3*

Mouse mRNA for prothymosin alpha

X56135

8.5 ± 1.5*

Cathepsin C

U74683

8.3 ± 7.1*

Fibroblast growth factor inducible 13

U42383

8.3 ± 3.4*

Mus musculus Pontin52 mRNA

AF100694

8.1 ± 2.1*

M.musculus fibrillarin mRNA

Z22593

7.3 ± 2.3*

Mus musculus ATPase inhibitor (IF1) mRNA

AF002718

6.9 ± 2.2*

Thymoma viral proto-oncogene

X65687

6.7 ± 2.2*

Perlecan (heparan sulfate proteoglycan 2)

M77174

6.7 ± 2.2*

Mus musculus Nedd5 mRNA for septin

D49382

6.1 ± 2.2*

Mus musculus GDP-dissociation inhibitor mRNA. preferentially expressed in hematopoietic cells

L07918

6.0 ± 4.1*

Mus musculus RNA polymerase II largest subunit (RPB1) mRNA

U37500

5.8 ± 1.4*

Mus musculus mRNA for sid2895p

AB025405

5.7 ± 1.1*

Mus musculus integrin-binding protein kinase mRNA

U94479

5.7 ± 2.2*

Phosphofructokinase. liver. B-type

J03928

5.6 ± 2.0*

Mus musculus beta prime coatomer protein mRNA

AF043120

5.4 ± 1.4*

Mouse mRNA for beta-1.4-galactosyltransferase

D37790

5.1 ± 2.6*

Table 10

Genes related to lipid metabolism in tumor cells are significantly* higher up-regulated compared with corresponding genes in pure muscle tissue

Lipid metabolism

Gene Name

Acc. #

FC ± SD

Apolipoprotein E

D00466

16 ± 4.8*

Phospholipid transfer protein

U28960

11 ± 1.4*

Annexin III

AJ001633

11 ± 2.1*

Adipose differentiation related protein

M93275

5.2 ± 1.5*

In the group of immune response the highest up-regulated gene was the Complement component 1, q subcomponent, c polypeptide and the Complement component 1, q subcomponent, alpha polypeptide with a fold difference of 108 ± 14 and 103 ± 14, respectively. In the group of protein binding the highest specific up-regulated genes were the S100 calcium-binding protein A4, Calpactin I light chain, and the S100 calcium-binding protein A6 with a fold difference of 108 ± 7.6, 24 ± 2.2, and 23 ± 2.9, respectively. Different receptors were up-regulated in the fold different range between 17 ± 2 and 11 ± 4, for example, the mRNA for 4F2/CD98 light chain receptor, Peptidylprolyl isomerase C-associated protein or the Mannose receptor, C type 1. Two genes having an important membrane function were up-regulated, the Retinoic acid-inducible E3 protein mRNA and the Plp2 mRNA for proteolipid protein 2. The highest up-regulated cell matrix gene was the Cytokeratin (endoB) gene with a fold difference of 68 ± 12. Genes related to cell growth were the Epithelial membrane protein 3, thymic shared antigen-1 (TSA-1) gene, and Epithelial membrane protein 1 with a fold difference of 29 ± 2.7, 29 ± 2.5 and 16 ± 1.5, respectively. In the group of cell core and nucleotide activity the Nsp-like 1 protein (Nspl1) gene (28 ± 6.9 fold difference), Mouse mRNA for beta-tubulin (isotype Mbeta 5) (21 ± 2.2 fold difference), and Dynamin (18 ± 1.7 fold difference) were up-regulated. An important function in calcium binding is the Endothelial monocyte-activating polypeptide I mRNA (29 ± 3.9 fold difference). Up-regulated genes with a high enzyme activity were Cathepsin S, Mouse lysozyme M gene, TYRO protein tyrosine kinase-binding protein, Colony-stimulating factor 1 receptor, and CPP32 apoptotic protease mRNA in a fold difference between 54 ± 6.1 and 26 ± 2.5. Active in the lipid metabolism were Apolipoprotein E, Phospholipid transfer protein, and Annexin III.

Comparing the gene expression profile of the muscle tissue after injection of 100 μl saline and the contralateral muscle tissue no significant gene expression differences could be observed.

Discussion

Reported here are the results of a study to examine the use of microarray technology to investigate gene expression changes in muscle tissue with a few SCCVII tumor cells. The sensitivity of pathological analysis by H&E staining for the detection of small tumor deposits in muscle tissue or only single tumor cells has been improved by the addition of immunohistochemical staining, which has been demonstrated to up-stage many patients who were classified as having no clinically measurable metastatic disease [12]. The invasion of tumor cells into surrounding muscle tissue is of clinical relevance. In a study of Diaz [4], it could be seen that patients with T1- or T2-sized tumors had only a 78 and 66% 5-year survival, respectively. Muscle invasion, Stensen’s duct involvement, and extracapsular spread of involved lymph nodes were all associated with decreased survival (p < 0.05), Because of this important issue it is of clinical relevance to find a more sensitive way of discovering these tumor cells with molecular markers. Early studies focused on the detection of clonal genetic changes that were specific for HNSCC tumor cells, such as mutations in p53 [13]. In recent years, researchers have shifted focus from tumor-specific markers toward tissue-specific markers, as they seek to take advantage of the differential gene expression of HNSCC cells and other tissues [14, 15]. DNA microarray analysis of human tumor specimens to identify tumor cell-related genes has been reported in several types of cancers [1618]. Roepman et al. [19] identified 102 genes in primary tumors as an expression profile for prediction of lymph node metastases from primary HNSCC tumors. Chung et al. [20] also used DNA microarray to classify HNSCC and predict lymph node metastasis. Due to heterogeneity, the HNSCC cells may utilize different gene products to achieve similar functions. Therefore, it is hard to validate expression of a large number of genes at the protein level in tissue specimens or to validate their biological relationship and functional pathways in metastasis. On the other hand, DNA microarray analysis using cell lines and animal models would allow further testing of the identified metastatic gene profile and its underlying mechanisms of metastasis. Similarly, other research groups have also reported use of cell lines from breast cancer animal models for DNA microarray analysis [21]. DNA microarray data mining analysis provided important information for understanding the biological behaviors of metastatic HNSCC cells. The gene expression pattern of tumor cells containing muscle tissue in our experiment was completely different compared with the gene expression pattern of pure muscle tissue. In our experiment we chose a fold difference of five in order to select only highly up-regulated different genes to get a very high specificity for the up-regulated genes. In other studies, genes were considered as significant if they have a fold difference of two [22, 23]. The up-regulated genes in SCCVII tumor cells containing muscle tissue found in our experiment belong to different physiological functional groups. The genes are important for immune response, protein binding, receptor activity, membrane function, cell matrix, cell growth, cell core, calcium binding, enzyme activity, lipid metabolism, and nucleotide activity. In the group of immune response the highest up-regulated genes were the Complement component 1, q subcomponent, c polypeptide, and the Complement component 1, q subcomponent, alpha polypeptide. C1q is the target recognition protein of the classical complement pathway that is crucial for the clearance of pathogens and apoptotic cells [24]. It is involved in a number of immunological processes such as phagocytosis of bacteria, neutralization of retroviruses, cell adhesion, modulation of dendritic cells (DC), B cells and fibroblasts, and maintenance of immune tolerance via clearance of apoptotic cells [25]. In the group of protein binding the highest specific up-regulated genes were the S100 calcium-binding protein A4, Calpactin I light chain, and the S100 calcium-binding protein A6. It could be shown that the calcium-binding protein S100A promotes metastasis in several experimental animal models, and S100A4 and A6 protein expression is associated with patient outcome in a number of tumor types. These proteins are localized in the nucleus, cytoplasm, and extracellular space and possess a wide range of biological functions, such as regulation of angiogenesis, cell survival, motility, and invasion [26]. Different receptors were up-regulated, for example, the mRNA for 4F2/CD98 light chain receptor and the Mannose receptor, C type 1. The human 4F2 cell surface antigen is a 120-kDa disulfide-linked heterodimer which is composed of an 80- to 90-kDa glycosylated heavy chain (4F2HC) and a 35- to 40-kDa nonglycosylated light chain (4F2LC). The precise function of the 4F2 molecule remains unknown. However, a role for 4F2 in the regulation of cell growth and activation has been suggested by the finding that 4F2 is expressed at low levels in most quiescent cells in vivo, but it is expressed at high levels on all dividing human tissue culture cells and most, if not all, malignant human cells [27]. The Mannose receptor, C type 1 mediates uptake of glycoconjugates carrying mannose in an end position [28]. Interestingly, this receptor uses a different binding domain to take up blood-borne collagen alpha chains, [29] and contributes to the adhesion of cancer cells to LSECs [30]. Two genes having an important membrane function were up-regulated in the Retinoic acid-inducible E3 protein mRNA and the Plp2 mRNA for proteolipid protein 2. Retinoic acid-inducible gene-I (RIG-I) is an intracellular pattern recognition receptor that plays important roles during innate immune responses. The mechanisms and signaling molecules that participate in the downstream events that follow activation of RIG-I are incompletely characterized. In addition, the factors that define intracellular availability of RIG-I and determine the steady-state levels of this protein are only partially understood, but are likely to play a major role during innate immune responses [31]. ThePlp2 mRNA for proteolipid protein 2 is a protein up-regulated in tumors, especially in oligodendrogliomas and important for the development of the tumors [32]. The highest up-regulated cell matrix gene was the Cytokeratin (endoB). Cytokeratin (CK) is a cytoskeletal intermediate filament protein. At present, there are 20 subtypes expressed in various types of human epithelial cells. The CK isotype depends on the cell type and the localization of CK in the cytoplasm [33]. The different CK types depend on whether they are acidic or basic types, as well as molecular weight, and the distribution of the various CKs differ between the several kinds of epithelia throughout the body. Genes related to cell growth were the Epithelial membrane protein 3, thymic shared antigen-1 (TSA-1) gene, and Epithelial membrane protein 1. Thymic-shared antigen-1 (TSA-1) was initially defined as a cell surface molecule expressed on both immature thymocytes and thymic stromal cells. TSA-1 has a role as a cell-surface adhesion molecule in T cells, with an associated signal-transduction role, possibly in conjunction with the TcR complex [34]. Epithelial membrane proteins are expressed in many tissues, and functions in cell growth, differentiation, and apoptosis has been reported. EMP-1 is highly up-regulated during squamous differentiation and in certain tumors, and a role in tumorigenesis has been proposed. EMP-1 is highly up-regulated during squamous cell differentiation and in certain tumors, and a role in tumorigenesis has been proposed [35]. In the group of cell core and nucleotide activity the highest up-regulated gene was the Nsp-like 1 protein (Nspl1), Mouse mRNA for beta-tubulin (isotype Mbeta 5) and Dynamin were up-regulated. Nsp-like 1 protein (Nspl1) [36] contributes to integrin and receptor tyrosine kinase signaling. Beta-tubulin [37] is a microtubule protein. Microtubules are multifunctional cytoskeletal proteins involved in many essential cellular roles, including maintenance of cell shape, intracellular transport, and in mitosis, forming mitotic spindles to ensure proper chromosome segregation and cell division. Dynamin is a large, GTP-binding mechanoenzyme that interacts with lipids, and undergoes conformational changes in response to nucleotide hydrolysis [38]. An important function in calcium binding is the Endothelial monocyte-activating polypeptide I mRNA, which has a procoagulant activity [39]. The highest up-regulated gene having enzyme activity was Cathepsin S. Cathepsin S is a lysosomal cysteine protease that is synthesized as an inactive precursor (36 kDa) and then activated in the acidic environment of lysosomes by proteolytic cleavage of its propeptide. Cathepsins have been found to participate in apoptosis; they are also known to play a role in the promotion of tumors during cancer progression [40]. Active in the lipid metabolism was the highly up-regulated Apolipoprotein E. APO E and its gene product, the APO E protein, are involved in cholesterol transport, lipid metabolism, and protein synthesis, by mediating the binding of the low-density lipoprotein (LDL) receptor, and the APO E receptor of lipid particles to specific lipoprotein receptors. APO E is also involved in other numerous functions, including tissue repair, immune response, and regulation, as well as cell growth and differentiation [41].

The different expressed genes play a crucial role in the development, differentiation, and functioning of tumor tissue and because they display remarkable tissue-specificity [42, 43], different pattern of gene expression are ideal for use as tissue classifiers. For example, the EMP-1 is highly up-regulated in squamous cell differentiation. This helps to differentiate between a squamous cell tumor from different tumor types. However, further studies that include more tumor types in this type of experiment are needed. The different gene expression patterns in our study hold potential for assisting in the determination of primary tumor site for small metastases of unknown origin. Our demonstration of this highly discriminatory assay for the detection of small tumor deposits or only a few tumor cells not detected by histology will hopefully supply the pilot data needed to incorporate this technique into a clinically relevant application that will improve our ability to properly stage patients with metastatic HNSCC. The further clinical relevance of this study might be that it is helpful to apply microarray technology to surgical margins for molecular analysis. Several muscle samples of the surrounding muscle tissue should be taken in order to assess in a more sensitive way any possibility of micrometastasis in deeper tissue layer, however, with the risk of missing micrometastasis nests.

In summary, our HNSCC metastatic animal model plus DNA microarray analysis provided valuable information on the unique biological behaviors of SCCHN cells. We have identified EMP-1 of the metastatic cells and put forward putative molecular bases leading to these phenotypes. We have started a new study translating our results into human application in order to bring this method closer into the clinic as valuable tool to detect micrometastasis.

Acknowledgment

The authors gratefully acknowledge Corrine Davis for reading the histological slides and Pauline Chu for technical assistance for histology

Conflict of interest

None to declare.

Copyright information

© Springer-Verlag 2011