Tetraspanin Expression Profile in Soft Tissue Sarcomas

Research Article

J Dis Markers. 2015;2(2): 1024.

Tetraspanin Expression Profile in Soft Tissue Sarcomas

Katia C Carvalho¹*, Isabela W Cunha², Rafael M Rocha³, Marcilei E Buim³, Mariana Maschietto³,Yukie S Kuwabara³, Luiz F L Reis4 and Fernando A Soares2,3

¹Laboratório de Ginecologia Estrutural e Molecular (LIM-58), Disciplina de Ginecologia, Hospital da Clinicas daFaculdade de Medicina da Universidade de Sao Paulo,Brazil

²Departamento de Anatomia Patológica, AC Camargo Cancer Center, Sao Paulo, Brazil

³Centro Internacional de Pesquisas (CIPE), AC Camargo Cancer Center, São Paulo, Brazil

4Hospital Sirio-Libanes, São Paulo, SP, Brasil

*Corresponding author: Katia Candido Carvalho,Obstetrics and Gynecology Department - FMUSP, Sao Paulo University, Cerqueira Cesar - Sao Paulo/SP, Brazil.Cep. 01246-903

Received: April 14, 2015; Accepted: May 27, 2015; Published: May 29, 2015


Sarcomas are rare malignant tumors rich in stromal and extracellular material; however, the features involved in their adhesion and cell migration are unclear. Moreover, little is known about molecular alterations, which could contribute to their malignant transformation and biology. We have previously observed a lack of correlation in tetraspanin expression module on sarcomas and a poorer prognosis, using a cDNA microarray platform. Tetraspanins are proteins, which are involved in several cellular processes and their profile, or role in sarcomas is unknown. Our aim was to validate cDNA microarray data and to evaluate the correlation of tetraspanins expression profile with clinical pathological parameters. Protein and transcript expression were evaluated by immunohistochemistry (227 samples) and quantitative real time - PCR (47 samples). Results showed a significant downregulation of CD81 and CD37 transcripts in high grade and metastatic tumors. CD63 showed the highest frequency of cytoplasm protein expression, showing a trend of association with poor survival (p=0,053). CD9 protein was observed, mainly in the membrane of leiomyosarcomas and pleomorphic sarcomas. Taken together, our results indicate that CD9, CD81 and CD82 does appear to play an important role in soft tissue sarcomas prognosis and their higher expression correlates with less aggressive forms of behavior. The role of CD63 protein needs further investigation, due to its high frequency and intensity of expression in STS.

Keywords: Sarcomas; Tetraspanins; qRT-PCR; Immunohistochemistry; Prognosis


Sarcomas are malignant tumors of the connective tissue and can be classified in different histological types [1]. In their early stages, patients are asymptomatic, because soft tissues are relatively elastic. This enables tumor growth without presenting alterations or sequelae [2]. Although these neoplasms are relatively rare, patients present high rates of mortality and morbidity [3]. Generally, sarcomas are histologically classified according to their differentiated tissue types (eg. leiomiosarcoma, liposarcoma, osteosarcomas and others) [1,2]. Diagnosis is based on the histological type and tumor grade [2,4]. Whilst tumor size and histological features are the best prognostic factors available for mesenchymal tumors, very little is known about their molecular alterations, which could contribute to our understanding of the cells origins, malignant transformations, and tumor biology of sarcomas [4]. Several previous works have shown an aberrant tetraspanin (TSPAN) expression in cancer, but do not associate these protein families with sarcomas’ clinical or pathological features.

These proteins are integral membrane proteins, which are characterized by the presence of four transmembrane domains. They belong to a large family, in which there are at least 33 members, expressed all cell types except for red blood cells [5]. Countless studies have been done which demonstrates the involvement of TSPAN in several physiological processes. Including tissue differentiation, cell proliferation, cell-matrix adhesion, cell migration, viral-induced syncytium formation, fusion processes, signal transduction and cellular activation [6]. Experimental and clinical studies have shown the importance of TSPAN in tumor biology and behavior [7,8].

Studies have analyzed human tumors from different primary sites and shown the importance of TSPAN in cancer progression, metastasis and prognosis. The downregulation of these proteins, is in general, associated with the development of metastasis or the loss of cell adhesion and neoplastic cell invasion [9,10]. However, a pro-tumoral role has also been identified with some TSPAN. The transfection of CD151 into cancer cells demonstrated an enhancement of the cell motility, invasion and metastatic potential in vitro of human colon cancer cells, glioblastoma and fibrosarcoma [11]. Research also demonstrated that a spliced variant of CD82, obtained by the excision of exon 7, was associated with an increase in cell motility, loss of cell adhesion, tumor growth and metastasis occurrence [12].

Although the importance of TSPAN in a wide range of epithelial tumors is well known, their role in soft tissue sarcomas (STS) has not been investigated. In a preliminary study, with 73 histologically different soft tissue tumors in a cDNA microarray platform, (containing 4.608 ORESTES sequences); we found a lack (change) of linear correlations between TSPAN gene pairs when considering tumor characteristics and gene expression profiles. CD9, CD37, CD63, CD81 and CD82 presented significant differences in their fold of expression. We aimed our investigation therefore, at the gene and protein expression of CD9, CD37, CD63, CD81 and CD82 and to evaluate the correlation of their expression profile with clinical pathological parameters.

Material and Methods

Tissue samples

Tissue samples were obtained from a total of 227 patients undergoing operative procedures on primary tumor sites, between 1973-2006. All tissue samples were obtained from the Tumor Bank and Paraffin Block archive, in the Department of Anatomic Pathology of the A.C. Camargo Cancer Center, São Paulo, Brazil. Hematoxylin and eosin stained slides were reviewed by two pathologists (IWC and FAS) and classified according to the criteria of 2013 World Health Organization (WHO). Among the available systems for sarcomas graduation, we used the French Fédération Nationale des Centres de Lutte Contre le Câncer (FNCLCC) and the National Cancer Institute (NCI). The Institutional Review Boards of the A.C. Camargo Cancer Center (Study Number 1031/08) approved this study and preoperative informed consents were obtained from all patients.

The follow-up of records of patients, for a period of more than sixty months, and all clinical and pathological data was obtained from patient archives. Sex, ethnicity, age, treatment, tumor size and staging, metastasis, recurrence, survival and other clinical features, were evaluated. For purposes of statistical analysis, we grouped intermediate and high histological grade tumors, and moderate and high protein expressions. When data analysis was performed separately, no significant differences were discovered, probably due the low number of samples for some histological types of tumors.

cDNA microarray, mathematical analysis and quantitative real time (qRT)-PCR

The cDNA microarray analysis was performed as described previously [13]. To analyze the relationship of TSPANs between 73 different frozen samples, we mapped a microarray platform containing 4.608 ORESTES sequences (Ludwig Institute chip [13,14] in order to find the presence of these genes on the chip. Detailed descriptions are available at Gene Expression Omnibus data repository under accession number GPL1930, and the accession number for raw data is GSE14541 (https://www.ncbi.nlm.nih.gov/ projects/geo). We performed a relevance networks analysis, which enables the visualization of the correlations between molecules under different conditions. A Fisher’s Z transformation was used to obtain the difference in significance between the correlations undergoing different analyses [15]. The initial chosen criteria for evaluation were: 1) metastatic versus non-metastatic tumors (after a five-year period of patient follow-up), 2) lower versus higher histological grade tumors (at diagnosis stage), and 3) patient status (death due to cancer or alive, free from cancer disease, after a five-year period of follow-up).

47 frozen samples of STS (previously evaluated by cDNA microarray) were investigated for the CD9, CD37, CD63, CD81 and CD82 gene expression; by qRT- PCR (Table 1). RNA was obtained using the Trizol method (Life Technologies, Carlsbad, CA) and firststrand cDNA was synthesized with High-Capacity cDNA Reverse Transcription Kit (Life Technologies). The synthesis was performed from 2 μg of total RNA samples, following manufacturer suggestions. The qRT-PCR reactions were performed using applied biosystems 7900HT Fast Real-Time PCR System (Life Technologies).