Characterization of the metastatic phenotype of a panel of established osteosarcoma cells

Abstract

Osteosarcoma (OS) is the most common bone tumor in pediatric patients. Metastasis is a major cause of mortality and morbidity. The rarity of this disease coupled with the challenges of drug development for metastatic cancers have slowed the delivery of improvements in long-term outcomes for these patients. In this study, we collected 18 OS cell lines, confirmed their expression of bone markers and complex karyotypes, and characterized their in vivo tumorgenicity and metastatic potential. Since prior reports included conflicting descriptions of the metastatic and in vivo phenotypes of these models, there was a need for a comparative assessment of metastatic phenotypes using identical procedures in the hands of a single investigative group. We expect that this single characterization will accelerate the study of this metastatic cancer. Using these models we evaluated the expression of six previously reported metastasis-related OS genes. Ezrin was the only gene consistently differentially expressed in all the pairs of high/low metastatic OS cells. We then used a subtractive gene expression approach of the high and low human metastatic cells to identify novel genes that may be involved in OS metastasis. PHLDA1 (pleckstrin homology-like domain, family A) was identified as one of the genes more highly expressed in the high metastatic compared to low metastatic cells. Knocking down PHLDA1 with siRNA or shRNA resulted in down regulation of the activities of MAPKs (ERK1/2), c-Jun N-terminal kinases (JNK), and p38 mitogen-activated protein kinases (MAPKs). Reducing the expression of PHLDA1 also delayed OS metastasis progression in mouse xenograft models.

Keywords: DHT; breast cancer; estrogen receptor-α; integrin αvβ3; resveratrol.

Figure 1. The metastatic potentials of OS cell lines and the expression of metastatic related genes in high/low metastatic OS cell lines

a. Kaplan-Meier Survival curves of human and mouse OS cell lines in experimental metastasis experiments. b. Summary of the median survival days in the clonally related OS cell lines. c. Western blot analysis of the expression of previously reported metastasis related genes in the pairs of high/low metastatic OS cell lines.

Figure 2. High PHLDA1/TDAG51 expression is associated with high metastatic potential in OS cell lines and lower survival probability in OS patients

a. Expression array heat map shows the genes selected by p < 0.01 and >2 fold change between high and low metastatic OS cells. Four probes for the PHLDA1/TDAG51 gene are highlighted. b. Western blot analysis shows the protein expression of TDAG51 in pairs of high/low metastasis OS cell lines. Green label: low met cells: red label: high met cells. c. Comparison of PHLDA1/TDAG51 expression with RNA Seq on HOS and MNNG cells in vitro, as well as one day in PuMA. d. Kaplan-Meier Survival estimation shows pediatric OS patients with high TDAG51 expression (red line) (n = 18) had lower survival probability than the patients with low TDAG51 (blue line) (n = 16).

Figure 3. Suppression of TDAG51 expression reduced pulmonary metastasis of OS cells

a. Western blot analysis of TDAG51 knockdown in MG63.3/GFP cells with 4 siRNA probes. b. Assessment of early metastatic progression of TDAG51 knockdown MG63.3/GFP cells (siRNA #5 was used for knockdown). Fluorescent microscope images of lungs (n = 5 in each group) 17 days post injection of OS cells, in which green fluorescent events represent tumor cells (bar, 2 mm). c. Quantification of total fluorescent green tumor cells in the lungs. There is statistically significant reduction in tumor cells following TDAG51 knock down compared to the control group (p = 0.0001).

Figure 4. Suppression of TDAG51 expression improved overall survival in a mouse model of OS pulmonary metastasis and reduced the kinase activities of MAPK signaling pathway members

a. Western blot analysis of TDAG51 knockdown in MG63.3/GFP cells with shRNA. b. Experimental metastasis experiments were performed by tail vein injection of highly metastatic OS cells (MG63.3/GFP) transfected with TDAG51 siRNA or control siRNA. Tumor bearing mice with TDAG51 shRNA expressing cells had a significantly longer survival rate compared to control mice (p = 0.026). c. Kinase array assay using cell lysates from MG63.3/GFP cells stably expressing TDAG51 shRNA or control shRNA. Knock down of TDAG51 expression significantly reduced phosphorylation of kinases in the MAP kinase signal transduction pathway.