Impact of Fibroblast-Derived SPARC on Invasiveness of Colorectal Cancer Cells.
Daniel DrevFelix HarpainAndrea BeerAnton StiftElisabeth S GruberMartin KlimpfingerSabine ThalhammerAndrea RetiLukas KennerMichael M BergmannBrigitte MarianPublished in: Cancers (2019)
Secreted protein acidic and rich in cysteine (SPARC) is a matricellular protein modulating cell-matrix interactions and was found up-regulated in tumor stroma. To explore the effect of high stromal SPARC on colorectal cancer (CRC) cell behavior and clinical outcome, this study determined SPARC expression in patients suffering from stage II and III CRC using a publicly available mRNA data set and immunohistochemistry of tissue microarray sections. Moreover, in vitro co-culture models using CRC cell lines together with colon-associated fibroblasts were established to determine the effect of fibroblast-derived SPARC on cancer cells. In 466 patient samples, high SPARC mRNA was associated with a shorter disease-free survival. In 99 patients of the tissue microarray cohort, high stromal SPARC in the primary tumor was an independent predictor of shorter survival in patients with relapse (27 cases; HR = 4574, p = 0.004). In CRC cell lines, SPARC suppressed phosphorylation of focal adhesion kinase and stimulated cell migration. Colon-associated fibroblasts increased migration velocity by 30% and doubled track-length in SPARC-dependent manner. In a 3D co-culture system, fibroblast-derived SPARC enhanced tumor cell invasion. Taken together, stromal SPARC had a pro-metastatic impact in vitro and was a characteristic of aggressive tumors with poor prognosis in CRC patients.
Keyphrases
- poor prognosis
- end stage renal disease
- ejection fraction
- free survival
- chronic kidney disease
- prognostic factors
- peritoneal dialysis
- bone marrow
- small cell lung cancer
- cell migration
- binding protein
- small molecule
- stem cells
- transcription factor
- signaling pathway
- machine learning
- escherichia coli
- electronic health record
- tyrosine kinase
- case report
- deep learning
- big data
- blood flow
- artificial intelligence