Neoplastic disease
Immunohistochemical Validation of Spontaneously Arising Canine Osteosarcoma as a Model for Human Osteosarcoma

https://doi.org/10.1016/j.jcpa.2017.07.005Get rights and content

Summary

Osteosarcoma (OS) originates from bone-forming mesenchymal cells and represents one of the primary bone tumours. It is the most common primary bone tumour in dogs and man. The characterization of an appropriate natural disease animal model to study human OS is essential to elucidate the pathogenesis of the disease. This study aimed to validate canine OS as a model for the human disease by evaluating immunohistochemically the expression of markers known to be important in human OS. The immunohistochemical panel included vimentin, alkaline phosphatase (ALP), desmin, S100, neuron-specific enolase (NSE), runt-related transcription factor 2 (Runx2) and bone morphogenetic protein 4 (BMP4). Immunohistochemistry was conducted on formalin-fixed, paraffin wax-embedded tissue sections from 59 dogs with confirmed primary OS. Vimentin, ALP, Runx2 and BMP4 were highly expressed by all tumours, while desmin, S100 and NSE were expressed variably. The findings were similar to those described previously for human OS and suggest that canine OS may represent a useful model for the study of the human disease.

Introduction

Osteosarcoma (OS) is a malignancy that originates from bone-forming mesenchymal cells (Brodey and Riser, 1969) and is the most commonly diagnosed primary bone tumour in both man and dogs (Brodey and Riser, 1969, Mirabello et al., 2011). OS occurs frequently in children, young and old adults, taller people and in young, large and giant breed dogs (Owen, 1984, Mirabello et al., 2011).

Worldwide, human OS has a bimodal age distribution with the highest incidence in children and young adults (0–24 years) and a second peak in the elderly (>60 years) (Mirabello et al., 2009a). People with localized OS have 5-year survival rates of around 65%, while patients with metastatic OS have a worse prognosis (5-year survival rates of 20%) (Li and Ye, 2014). In dogs treated with amputation and adjuvant chemotherapy, the 2-year survival rate for localized OS is approximately 20% (Bielack and Carrle, 2008).

In the USA, fewer than 1,000 new cases of human OS are diagnosed annually (Dorfman and Czerniak, 1995). The incidence of OS in dogs is significantly higher (>8,000 cases/year) (Jongeward, 1985). Canine OS represents 85–98% of all canine bone cancers (Dernell et al., 2007). Similar to human OS, canine OS is also known to have a bimodal age distribution of 18–24 months and 7–9 years. Interestingly, the incidence is significantly higher in older dogs compared with that occurring in younger dogs (Dernell et al., 2007). In both man and dogs, OS mostly affects the ends of long bones near the metaphyseal regions (Owen et al., 1975, Mirabello et al., 2011, Guil-Luna et al., 2015). The molecular nature of OS is not yet fully understood (Bayani et al., 2003, Lau et al., 2004, Man et al., 2004, Gorlick, 2009) because of the complexity of OS genetics and an inability to elucidate related pathways (Gorlick, 2009).

The use of animal models in cancer research has a long history (Hewitt, 1978, Herberman, 1983). Such models are useful for the study of the biology of cancers and to discover new therapeutic targets, but defining the most suitable animal models is often difficult. Animal models commonly used to study human OS are the dog, mouse and rat (Guijarro et al., 2014).

People and dogs share a similar basis of global gene expression signatures (Paoloni et al., 2009), environmental conditions (Owen et al., 1975, Guil-Luna et al., 2015), physiological characteristics (Gordon and Khanna, 2010), OS biological behaviours (Owen, 1969), histological morphology, molecular targets, tumour progression and response to conventional treatments (Guil-Luna et al., 2015). In addition, canine OS is 10–100 times more common than human OS (Mueller et al., 2007, Paoloni and Khanna, 2008), it is a naturally occurring tumour (Brodey, 1979, Vail and MacEwen, 2000) and metastasis initiates in a shorter period than in human OS (Owen, 1967), which offers better opportunities to investigate the effects of potential therapies (Dernell et al., 2007, Mirabello et al., 2009b). Furthermore, dogs have shorter lifespans and faster OS progression than people (Owen, 1967, Gordon and Khanna, 2010). The latter characteristic provides an opportunity to study the response to different therapies that could improve human survival time (Guil-Luna et al., 2015). All of these aspects of canine OS make it an attractive animal model for study of human OS (Owen et al., 1975, Paoloni et al., 2009, Morello et al., 2011).

Detection of specific tumour markers in OS is important for accurate diagnosis, prediction of prognosis, metastases and response to chemotherapy (Selvarajah and Kirpensteijn, 2010). Despite the aetiology of OS being complex, some tumour markers are expressed at low levels or are overexpressed in canine and human OS (Trieb et al., 2003, Barger et al., 2005, Won et al., 2009). Tumour markers that are generally expressed in human OS include vimentin, S100, osteonectin, osteocalcin, smooth muscle actin (SMA), CD99 and neuron-specific enolase (NSE) (Trihia and Valavanis, 2012). In human OS, vimentin (Rosenberg, 1995), alkaline phosphatase (ALP) (Rosenberg, 1995), runt-related transcription factor 2 (Runx2) (Andela et al., 2005) and bone morphogenetic protein 4 (BMP4) (Rosier and Bukata, 2007) are highly expressed. In addition, desmin (Coffin and Belchis, 2006), S100 (Coffin and Belchis, 2006) and NSE (Rosenberg, 1995) are weakly expressed in some human OSs.

A pilot study was conducted to investigate if the tumour markers currently used to classify human OS are applicable to canine OS, using five samples of canine OS tissue. It was found that similar to human OS, this panel of antibodies (vimentin, CK7, desmin, neurofilament protein, CD45, NSE, CD57 and S100) could be used to differentiate canine OS from other canine tumours. This was the first demonstration of the presence of CK7, desmin and S100 in canine OS (H. J. Gunn, RMIT University, Melbourne, Australia; unpublished data). This panel of antibodies was modified for the purpose of cost benefit and the size of some of the canine OS tissue samples available for the present study. Neurofilament protein and CD45 were excluded because they are not expressed in human OS. CK7 and CD57 were also excluded because they are not mesenchymal cell markers. Antibodies specific for ALP, Runx2 and BMP4 were added.

Demonstrating differences or similarities in expression profiles of specific protein markers in canine and human OS could help in therapeutic and comparative histopathological studies to improve our understanding of this tumour. Therefore, the aim of this study was to validate canine OS as a model for human OS by investigating and analyzing the expression of the selected marker proteins in spontaneously arising canine OS. The expression profiles were then compared with those previously obtained for human OS (Rosenberg, 1995, Andela et al., 2005, Coffin and Belchis, 2006, Rosier and Bukata, 2007).

Section snippets

Tumour Samples

This study was conducted using 59 formalin-fixed, paraffin wax-embedded (FFPE) primary canine OS tissues collected between 1991 and 2015. These tissue blocks originated from the archival collection of the Pathology Unit, School of Veterinary Sciences, University of Bristol, UK. The OS tumours were histopathologically confirmed according to the veterinary classifications of domestic animals (Thompson and Dittmer, 2017).

An application was submitted to RMIT University Animal Ethics Committee.

Clinical Data

Data were gathered from 59 dogs with primary OS. Their ages ranged from 1 to 13 years with a mean of 8 years. Slightly more dogs were female (52%). Fifty dogs (89%) were of pure breed, five (9%) were of mixed breed and four were of unknown breed. Most dogs (93%) were large or medium breeds where breed was known. The descending order of frequency of the most common breeds was Rottweiler (18%), Labrador Retriever (11%), Golden Retriever (9%) and Dobermann (7%). OSs were mostly located in the

Discussion

This study aimed to validate canine OS as a model for human OS using immunohistochemical labelling. The data revealed that vimentin, ALP, Runx2 and BMP4 were highly expressed in all canine OS tissue samples, while desmin, S100 and NSE were expressed variably. These observations were similar to those for human OS (Rosenberg, 1995, Andela et al., 2005, Coffin and Belchis, 2006, Rosier and Bukata, 2007). Statistical analysis demonstrated no significant differences in the expression levels of

Acknowledgments

The authors acknowledge funding provided by the Ministry of Higher Education, Sultanate of Oman, and by the School of Health and Biomedical Sciences, RMIT University. We thank A. Barimah for his assistance.

References (93)

  • E. Mornet et al.

    Structural evidence for a functional role of human tissue nonspecific alkaline phosphatase in bone mineralization

    Journal of Biological Chemistry

    (2001)
  • L.N. Owen

    Neoplasms in older dogs with particular reference to management. I

    British Veterinary Journal

    (1984)
  • D. Paulin et al.

    Desmin: a major intermediate filament protein essential for the structural integrity and function of muscle

    Experimental Cell Research

    (2004)
  • J.W. Said et al.

    Immunoreactive neuron-specific enolase, bombesin, and chromogranin as markers for neuroendocrine lung tumors

    Human Pathology

    (1985)
  • G.T. Selvarajah et al.

    Prognostic and predictive biomarkers of canine osteosarcoma

    Veterinary Journal

    (2010)
  • K. Trieb et al.

    Survivin expression in human osteosarcoma is a marker for survival

    European Journal of Surgical Oncology

    (2003)
  • M. van der Deen et al.

    Genomic promoter occupancy of runt-related transcription factor RUNX2 in osteosarcoma cells identifies genes involved in cell adhesion and motility

    Journal of Biological Chemistry

    (2012)
  • M. van der Deen et al.

    MicroRNA-34c inversely couples the biological functions of the runt-related transcription factor RUNX2 and the tumor suppressor p53 in osteosarcoma

    Journal of Biological Chemistry

    (2013)
  • X. Wu et al.

    Analysis of gene expression changes associated with human carcinoma-associated fibroblasts in non-small cell lung carcinoma

    Biological Research

    (2017)
  • G. Xi et al.

    Therapeutic potential for bone morphogenetic protein 4 in human malignant glioma

    Neoplasia

    (2017)
  • H. Yoshikawa et al.

    Bone morphogenetic proteins in bone tumors

    Journal of Orthopaedic Science

    (2004)
  • K. Zhou et al.

    Bone morphogenetic protein 4 is overexpressed in and promotes migration and invasion of drug-resistant cancer cells

    International Journal of Biological Macromolecules

    (2017)
  • N.N. Ali et al.

    Constitutive expression of non-bone/liver/kidney alkaline phosphatase in human osteosarcoma cell lines

    Journal of Bone and Mineral Research

    (1996)
  • V.B. Andela et al.

    An immunohistochemical analysis to evaluate an inverse correlation between Runx2/Cbfa1 and NF kappa B in human osteosarcoma

    Journal of Clinical Pathology

    (2005)
  • A. Barger et al.

    Use of alkaline phosphatase staining to differentiate canine osteosarcoma from other vimentin-positive tumors

    Veterinary Pathology

    (2005)
  • J. Bayani et al.

    Spectral karyotyping identifies recurrent complex rearrangements of chromosomes 8, 17, and 20 in osteosarcomas

    Genes, Chromosomes and Cancer

    (2003)
  • S.S. Bielack et al.

    State-of-the-art approach in selective curable tumors: bone sarcoma

    Annals of Oncology

    (2008)
  • R.S. Brodey

    The use of naturally occurring cancer in domestic animals for research into human cancer: general considerations and a review of canine skeletal osteosarcoma

    Yale Journal of Biology and Medicine

    (1979)
  • R.S. Brodey et al.

    Canine osteosarcoma: a clinicopathologic study of 194 cases

    Clinical Orthopaedics and Related Research

    (1969)
  • L. Carlsson et al.

    Desmin-related myopathies in mice and man

    Acta Physiologica Scandinavica

    (2001)
  • T. Chano et al.

    Analysis of the presence of osteocalcin, S-100 protein, and proliferating cell nuclear antigen in cells of various types of osteosarcomas

    European Journal of Histochemistry

    (1996)
  • H. Chen et al.

    S100 protein family in human cancer

    American Journal of Cancer Research

    (2014)
  • P. Dias et al.

    Evaluation of desmin as a diagnostic and prognostic marker of childhood rhabdomyosarcomas and embryonal sarcomas

    British Journal of Cancer

    (1987)
  • F. Dong et al.

    Chimaphilin inhibits human osteosarcoma cells invasion and metastasis through suppressing the TGF-β1-induced epithelial-to-mesenchymal transition markers via PI-3K/Akt, ERK1/2 and Smad signaling pathways

    Canadian Journal of Physiology and Pharmacology

    (2017)
  • H.D. Dorfman et al.

    Bone cancers

    Cancer

    (1995)
  • F. Eckert et al.

    Coexpression of cytokeratin and vimentin intermediate filaments in benign and malignant sweat gland tumors

    Journal of Cutaneous Pathology

    (1994)
  • Z. Gao et al.

    The application of immunohistochemistry in the diagnosis of bone tumors and tumor-like lesions

    Skeletal Radiology

    (2005)
  • C. Gebhard et al.

    Morphological and immunohistochemical characterization of canine osteosarcoma spheroid cell cultures

    Anatomia, Histologia, Embryologia

    (2016)
  • R.D. Goldman et al.

    The function of intermediate filaments in cell shape and cytoskeletal integrity

    Journal of Cell Biology

    (1996)
  • I.K. Gordon et al.

    Modeling opportunities in comparative oncology for drug development

    Institute for Laboratory Animal Research Journal

    (2010)
  • R. Gorlick

    Current concepts on the molecular biology of osteosarcoma

    Cancer Treatment and Research

    (2009)
  • M.V. Guijarro et al.

    Animal models in osteosarcoma

    Frontiers in Oncology

    (2014)
  • S. Guil-Luna et al.

    Comparative aspects between canine and human mesenchymal tumors

    Sarcoma Research International

    (2015)
  • T. Hasegawa et al.

    Histological and immunohistochemical diversities, and proliferative activity and grading in osteosarcomas

    Cancer Detection and Prevention

    (1997)
  • R.B. Herberman

    Counterpoint: animal tumor models and their relevance to human tumor immunology

    Journal of Biological Response Modifiers

    (1983)
  • A. Javed et al.

    Impaired intranuclear trafficking of Runx2 (AML3/CBFA1) transcription factors in breast cancer cells inhibits osteolysis in vivo

    Proceedings of the National Academy of Sciences of the USA

    (2005)
  • Cited by (20)

    • Non-rodent animal models of osteosarcoma: A review

      2021, Cancer Treatment and Research Communications
      Citation Excerpt :

      Al-Khan and colleagues sought to validate canine osteosarcoma as a model by immunohistochemically evaluating the expression of osteosarcoma-prominent markers in canine tissue. Markers such as vimentin, ALP, Runx2, and BMP-4 were all highly expressed by tumors, similar to human osteosarcoma [55]. Another study by Roy and team investigated membrane-associated proteomics of metastatic and non-metastatic variants of the human and canine osteosarcoma.

    View all citing articles on Scopus
    View full text