Introduction
Giant cell tumor of bone (GCTB) is one of the primary bone tumors and common in 20-40 year-old populations. The morbidity in China is up to 18.4%, which is 2-3 times the western countries [
2,
2]. The tissue source of GCTB has not yet been identified and its biological behavior is variable, and the progression of the disease is difficult to predict. Its recurrence is common after clinical therapy [
3,
4].
Some studies on the role of gene expression in giant cell tumors have been performed. The expression of ADAM12, a member of the disintegrin and metalloprotease family, has been examined in 20 giant cell tumors and found that ADAM12 was involved in the fusion process for mononuclear stromal cells in giant cell tumors [
5]. Some researchers emphasized higher expression of vascular endothelial growth factor (VEGF) and matrix metalloproteinase 9 in recurrent tumors [
6]. But, the limitation of these published studies is that the evaluations were usually confined to a few genes. This problem can be solved by the use of gene chip technology, which allows the assessment of a large number of genes simultaneously. Complementary DNA (cDNA) microarray has been a developed gene chip technology in recent years. The cDNA microarray utilizes the theory of gene chip technology of information integrating and processing in parallel, and mounts an enormous number of specific oligonucleotide or DNA fragment on a 645 square millimeter matrix, conducting high efficiency and low consumption in the process of assessing the biological information [
7]. After investigating the gene expression profiles in acute allograft rejection of renal transplantation, it was found that the early diagnosis was virtually impossible to make based on a single gene marker, but could be achieved by a set of gene markers [
8]. In this study, cDNA microarray technology was applied to differentiate the gene expression spectrum of GCTB, trying to explore the relationship between gene expression spectrum and the clinical presentation, and further provide evidences for gene diagnosis and potential gene therapy
Materials and methods
Experiment apparatus and software
The followings are the experiment apparatus and software used in the study: Eppendorf-desktop micro centrifuge, Eppendorf 5810R-freeze desktop centrifuge (Eppendorf AG, Germany), Gen III Microarray Spotter-array spotter, Generation III array scanner-array scanner (Amersham Pharmacia Biotech Ltd., UK), Shellab General Purpose Incubator—homothermic cross incubator (Shellab, USA), Bio-Rad Mini-Sub GT System—electrophoretic equipment (Bio-Rad, USA), DU520 UV/Vis Spectrophotometer — Ultraviolet Spectrophotometer (Beckman Coulter, USA), Vilber Gel Documentation System-Gel Documentation System, UV Crosslinker (Vilber Lourmat, France), MIDAS8064 — spot cDNA microarray (Chipscreen Biosciences Ltd., China), ImageQuant — image scanning software (Amersham Pharmacia Biotech Ltd., UK), Array Vision 6.0-image analytical software (Imaging Research Ltd.) and real-time fluorescence quantitative polymerase chain reaction (PCR) general system (ABI, USA).
Source of microarray
cDNA microarray with 8064 of target genes was provided by Shenzhen Chipscreen Biosciences Co. Ltd., China.
Experimental reagents
The following reagents were used in this study: Trizol Reagent (Life Technologies, Inc., USA), QIAgene RNeasy mini kit, QIAgene PCR Purification kit (QIAGEN, Inc., USA), Ambion messageAmp aRNA kit (Ambion, Inc., USA), CyScribe cDNA labeling kit (Amersham Pharmacia Biotech, Ltd., UK).
Specimen collection, grouping and RNA extraction
From 2002 to 2004, fresh specimens of 8 cases (GCTB group) from surgical excision were harvested from the Second Affiliated Hospital of Jinan University, China (Shenzhen People’s Hospital), the First Affiliated Hospital of Jinan University and the First Affiliated Hospital of the SUN YAT-SEN University (Proved to be GCTB by the presentation intra-operatively and by post-operative pathology). Three classifications were employed, namely Jaffe, Campanacci and Enneking. By the Jaffe classification, there were 6 cases in class I, 1 case in class II and 1 case in class III. By the Campanacci classification, there were 6 cases in class I, 1 case in class II and 1 case in class III. By the Enneking classification, there were 5 cases in G0T1-2M0, 2 cases in G1T1-2M0 and 1 case in G2T2M0. In terms of location, there were 5 cases in the distal femur, 2 cases in the proximal tibia and 1 case in the distal radius. Of the GCTB group, 5 were male and 3 female, aged 23-55 years, with the average age of 34.4. In the case of the control group, the normal bony callus was harvested from the fracture-union site of tibia or femur during operation of internal fixation removal. Of the control group, 2 were male and 2 were female, aged 23-50 years, with the average age of 30. All the specimens were dissected into equal pieces of 1 cm×1 cm×1 cm, preserved in liquid nitrogen within 30 minutes after excision. After a suitable volume of TRIZOL was added to the tissue samples, the tissues were automatically homogenized by a homogenizer into a complete homogenate. Then, trichlormethane, RW1, RPE solution, diethyl pyrocarbonate (DEPC) liquid were added in sequential order before centrifugalization. After standing for sedimentation using 3 mol/L sodium acetate and adding a suitable amount of DEPC liquid to dissolve the precipitation, RNA concentration and total amount of the sample was evaluated. Some samples were taken for agarose gel electrophoresis (AGE) to analyze the quality of RNA and the samples were kept on file and kept at -80°C (the concentration was 20 μg/mL, and its A260/A280 ratio should be not less than 1.8).
Methods
Labeled probe
RNA sample and second round primer, Spike RNA were placed at 70°C for combined degeneration and cooling down for 5 min, respectively. It reacted with the mixture of 0.1 mol/L DTT, dye-dCTP, dNTP for 2 hours at 42°C. It was purified by QIA quick PCR Purification Kit after the reaction. Buffer PB was added, and the mixture was stirred well, and then gradually centrifuged to 57 μL.
Pre-hybridization or pre-processing
The inserted glass slide of film hanger was placed into the staining jar which was added with 200-250 mL prehybridization buffer solution pre-heated to 55°C. Then, the jar was placed on the swing bed at room temperature and stirred slowly for 20-30 min and taken out.
Probe preparation and hybridization
The indicated probe (Cy3/Cy5) was mixed well and divided equally (40 pmol each). After dehydration, in the proper order, the reactants were added as follows: hybridization buffer solution 7.5 μL, formamide 15 μL, and sterilization water was added to reach 30 μL of the total volume. After it was fully dissolved and mixed well, it was slightly centrifuged, then put into 95°C water and bathed for 5 min, and immediately cooled down on an ice bath for 1 min, oscillated until mixed well and then, gently centrifuged. After the prepared probe was laid on the sample side of cDNA microarray and a glass cover slip put on, the glass slide was placed into the incubator at 42°C for 16 hours.
Washing steps
After hybridization, the glass slide was taken out and placed into the film hanger, and then put into the staining jar which was filled with ionic water, slipping off vertically and naturally. After the hanger was taken out, the glass slide was rapidly put into another washing-staining jar which was pre-filled with 200-250 mL of 0.1% SSC (saline sodium citrate) and pre-heated to 55°C and slowly shaken for 20 min on the swing bed at room temperature. The washing step was repeated and then, the glass slide was taken immediately out from the film hanger and blown slightly to dry completely with the buffer jar.
Scanning and data processing
A completely dry glass was placed onto the scanning film hanger and scanned with the Amersham Pharmacia Gen III. The images were converted to digital signal based on its fluorescence intensity and data analysis was carried out and processed by Image Quant.
Hybridization signal analysis and standardization
Hybridization signal intensity was measured by every fluorescence pathway of scanned data, including the actual value (sVOL) which was the scanned value (Vol-RFU) of each gene after reducing the background value (Bkgd). It could preliminary estimate the expression level of genes on each sample. The efficacy of each gene data was also based on that value for determination. After the fluorescence signal had been unified by the LOWESS method, the ratio represented a variation in the degree of gene expression between the experimental group (tumor group) and the control group (larger than 1 represented the up-regulated multiple of the gene expression, while less than 1 was the relative down-regulation of the gene expression, and represented by the negative value of its reciprocal). If the fluorescence signal of the both pathways met the effective signal standard, at least 2-fold changes relative to the control were considered to be threshold. If effective hybridization signals were just in one of the pathways, 3-fold changes relative to the control was taken as threshold [
8].
Hybridization and error detection control
The gene array hybridized with fluorescence indicator was whole range of cDNA expression microarray containing a set of 8064 human genes. It had 7488 cloning fragments, 7458 of which were gene expression fragments. The length was above 500 base pairs (average length was 1.4 kb). Other 30 fragments were blank carrier fragments, including 120 of ex-reference standard genes and 132 of in-reference standard genes. Three hundred and eighty four genes were negative and positive control genes. Every control gene on array was repeated 12 times for sampling.
Statistical analysis
SPSS 11.5 statistical package was applied to deal with the experimental data between the GCTB group and control group. Group comparison was analyzed by analysis of variance (ANOVA). A P<0.05 was considered statistically significant.
Results
Follow up results
Concerning the follow-up cases, focused observation was performed in 8 cases, self or variant bone transplantation performed in 4 cases, and 4 cases underwent tumor edge surgical excision or artificial joint replacement. We are still following all these cases up to now. The cases were followed up for 32 months on the average. One case was reported as a relapse, without pulmonary metastasis.
Gene expression difference of giant cell tumor of bone in patients
The expression levels of more than 26% of 8000 genes in the GCTB group were significantly different compared with those of the control group (P<0.01). It shows that apart from possible experimental error, there would be a biological source of the difference (individual difference and physiopathology specific difference). Differentially expressed genes were mainly cytokine and receptor, transcription-related, apoptosis-related, growth differentiation-related, signal transduction-related, cell cycle-related, electron transport- and oxidation-related, transferase- related and nuclease-related genes. The expressions of 47 genes in the GCTB group were increased more than 5-fold compared with those in the control group, with 25 gene expression up-regulated significantly and 22 gene expression down-regulated. We performed preliminary analysis on the functional categories of these genes. Its functional categories mainly included extracellular matrix regulation, tumorigenesis related, cell proliferation and apoptosis regulation gene, cytokine genoid (Tables 1-3).
Typical up-regulated expression genes
We found that 12 genes were extracellular matrix genes related to the formation and transition of osteocytes. Among these, the expressions of matrix metalloproteinase 13 (MMP13), dermatan sulphate mucoprotein3 (DSPG3), skin pontine protein (DPT), matrix metalloproteinase 12 (MMP12) in GCTB group were increased more than 10-fold in comparison with the control group; 6 were tumor related genes, the differential expressions of hypophysin tumor transforming gene 1 (PTTG1), ornithoerythroblast leukemia virus tumorigenesis gene 3 (ERBB3), bone marrow nucleus differentiation antigen (MNDA), pim-2 oncogene (PIM2) were above 10-folds between CCTB group and the control group. Seven genes were cytokine and receptor genoid. Among these, the differential expressions of small inducible cytokine A3 (SCYA3), tumor necrosis factor receptor family 11B (TNFRSF11B), small inducible A3 like cytokine 1 (SCYA3L1), small inducible cytokine (SCYB6) were increased above 10-fold.
Typical down-regulated expression genes
Eight genes were extracellular matrix genes. Among these, the expressions of matrix metalloproteinase 3 (MMP3), cartilage oligomer matrix protein (COMP), collagen of type α1X (COL10A1), matrix metalloproteinase 7 (MMP7), type I collagen α2 (COL1A2) were decreased more than 10-fold compared with the control. Six genes were tumor-related genes and among these, the differential expressions of prostaglandin E synthase (PTGES), protein tyrosine kinase (TYRO3), hyaluronic acid glucosaminidase 1 (HYAL1), E74 like factor 3 (ELF3) between GCTB and the control group were above 10-fold. Seven genes were ranging cytokine and receptor genoid, and the differential expressions of small inducible cytokine D1 (SCYD1), small inducible cytokine B14 (SCYB14), γ interferon 1 (IFNGR1), tumor necrosis factor and α-induced protein 2 (TNFAIP2) were above 10-fold.
Discussion
The development and occurrence of tumor are complicated, multistage biological processes. When various tumor related genes, which participate in regulating cell proliferation, differentiation and apoptosis of fundamental vital process, develop mutations, resorting, deletions, amplifications and then are abnormally activated or inactivated successively or simultaneously, the normal-controlled proliferation, differentiation and apoptosis are disturbed. Then, the cells proliferate infinitely, and the differentiation or apoptosis of cell is interfered, leading to the formation of tumors. A variety of genes are involved in tumorigenesis which are dynamic and non-balanced. What urgently needs study is not just one gene or several genes but the whole genome ranging from normal status to tumor-formation, involving thousands of gene expression and its dynamic change in different stages. Thus, the gene expression data profile renders it possible to produce a full-view of tumor growth and a more comprehensive tumor-model related to gene expression.
The category and richness of mRNA can be directly detected by cDNA microarray technology, which is a powerful tool to study gene expression. In this study, cDNA microarray technology was applied to explore the mechanism of tumor development of GCTB, so as to find a grading method of GCTB at a molecular level, and explore the relationship between the gene expression spectrum and the clinical presentation and furthermore provide evidence for the potential antisense gene therapy. Typical specimens of GCTB were selected, using normal human bone tissue as control. A cDNA microarray with the whole range of the human gene was applied.
In this study, gene expression profile of giant cell tumor and normal bone and a different diagram of gene expression were successfully obtained, including preliminary screening of significant differential expression of specific functional genome. The significantly up-regulating gene expression was consistent with its function which had been ever known. Whereas TNFRSF11B is a bone protecting gene and can promote osteoclast activity which plays an important rolle in osteoporosis pathogenesis, it was mainly expressed by lymphocyte, osteoclast of marrow stromal or prosperous osteogenesis [
9].
In vivo, hypercalcemia can be caused by injection with osteoprotegerin [
10]. Osteoprotegerin ligand (OPGL) can strengthen the function of lymphatic dendritic cells (DCs) to promote T cell pulsing so as to produce osteoclast cytokine [
11]. Also, it has recently been discovered that interferon blocks osteoclast differentiation
via affecting osteoprotegerin ligand (OPGL) signal transduction [
12]. In the study, the gene significant up-regulating presentation matches the expectation of the experiment. Through the screening of microarray, it was discovered that tyrosine protein kinase played an important role in the occurrence and development of GCTB. Transmembrane receptor protein tyrosine protein kinase signal pathway (ERBB3), tyrosine protein kinase Src family member, negative regulating cell migration and adherence (FGR), protein tyrosine kinase (TYRO3) as the screening gene of differential expression, are all related to tyrosine kinase. This result matches with the study of Guenther R [
13], applying an Affymetrix microarray screening on the differential expression gene of GCTB, discovering that among the primary and recurrent giant cell tumor, most of the differential expression genes have relationship with the activity of receptor tyrosine kinase. Further study on these tyrosine kinase related genes of GCTB has advantages for illustrating tumorigenesis and the mechanism of development.
After long-term clinical observation, the Jaffe grading based on the quantitation of giant cell tumors and the characteristics of cytohistology of tumor stroma and biological behavior of tumor did not match. Although it reflects the invasiveness of the tumor cell to a certain degree, the value for indicating therapeutic prognosis is limited because of the inherent subjectivity of the inspector and the disparity of preoperative paracentesis and pathologic biopsy, especially around the boundary of the class II and class III which is ambiguous. There is a certain degree of recurrence and metastasis rate among the patients of class I and class II [
14]. The Campanacci standard of radiologic classification is not related to tumor recurrence [
15]. GCTB of some patients show extremely strong invasiveness by radiology, while its pathologic Jaffe grading is class I [
16]. The Enneking musculoskeletal system referring to three aspects of histology, tumor size and distant metastasis, and is subjected to the performance of a surgical boundary treatment. Typical treatment method is tumor curettage by using autogenous iliac bone, only attaining intracystic surgical boundary. Its recurrence rate is up to 40%-60% [
3]. Therefore, clinical doctors tend to integrate these three methods to analyze the attribute and lesion degree to direct the suitable choice of therapeutic regimen. However, the value of this kind of integration also has its limitations. It has been reported in many cases that the integrated analysis was certified as typically benign GCTB, but lung metastasis occurred in a widespread fashion in a half year post-operation and was the cause of death [
17]. It is found that the current GCTB grading method and classification rule cannot completely reflect the malignancy, metastasis and recurrence rate. cDNA microarray technology can be applied to perform cluster analysis of gene expression for benign and malignant of GCTB, so as to establish a quantitative and molecular grading standard of GCTB. After the analysis of the typical recurrence, it is supposed that GCTB with an increase of 10-fold up-regulated gene MMP13, TNFRSF11B and an increase of 10-fold down-regulated gene SCYD1, MMP3 possesses aggravating invasiveness and poor prognosis which is more reliable as an indicator for a radical operation procedure. Verification on this supposition would be an important direction of our progressive study.
Conclusion
In conclusion, some specific gene expressions can be screened with the employment of cDNA microarray technology. It provides important clues for studying the pathogenetic mechanism and gene therapy of GCTB. Based on the screening of the differential expression gene, the study of functional identification of the specific gene cluster of GCTB will continue and new grading and therapeutic approaches can be developed.
Higher Education Press and Springer-Verlag Berlin Heidelberg
PDF
(99KB)
