Genes influenced by the non-muscle isoform of Myosin light chain kinase impact human cancer prognosis.

TitleGenes influenced by the non-muscle isoform of Myosin light chain kinase impact human cancer prognosis.
Publication TypeJournal Article
Year of Publication2014
AuthorsZhou T, Wang T, Garcia JGN
JournalPLoS One
Volume9
Issue4
Paginatione94325
Date Published2014
ISSN Number1932-6203
KeywordsAnimals, Cluster Analysis, Disease Models, Animal, Gene Expression Profiling, Gene Expression Regulation, Humans, Mice, Mice, Knockout, Myosin-Light-Chain Kinase, Neoplasm Grading, Neoplasm Metastasis, Neoplasm Staging, Neoplasms, Prognosis, Transcriptome, Tumor Burden
Abstract

<p>The multifunctional non-muscle isoform of myosin light chain kinase (nmMLCK) is critical to the rapid dynamic coordination of the cytoskeleton involved in cancer cell proliferation and migration. We identified 45 nmMLCK-influenced genes by bioinformatic filtering of genome-wide expression in wild type and nmMLCK knockout (KO) mice exposed to preclinical models of murine acute inflammatory lung injury, pathologies that are well established to include nmMLCK as an essential participant. To determine whether these nmMLCK-influenced genes were relevant to human cancers, the 45 mouse genes were matched to 38 distinct human orthologs (M38 signature) (GeneCards definition) and underwent Kaplan-Meier survival analysis in training and validation cohorts. These studies revealed that in training cohorts, the M38 signature successfully identified cancer patients with poor overall survival in breast cancer (P<0.001), colon cancer (P<0.001), glioma (P<0.001), and lung cancer (P<0.001). In validation cohorts, the M38 signature demonstrated significantly reduced overall survival for high-score patients of breast cancer (P = 0.002), colon cancer (P = 0.035), glioma (P = 0.023), and lung cancer (P = 0.023). The association between M38 risk score and overall survival was confirmed by univariate Cox proportional hazard analysis of overall survival in the both training and validation cohorts. This study, providing a novel prognostic cancer gene signature derived from a murine model of nmMLCK-associated lung inflammation, strongly supports nmMLCK-involved pathways in tumor growth and progression in human cancers and nmMLCK as an attractive candidate molecular target in both inflammatory and neoplastic processes.</p>

DOI10.1371/journal.pone.0094325
Alternate JournalPLoS ONE
PubMed ID24714365
PubMed Central IDPMC3979809
Grant ListHL91899 / HL / NHLBI NIH HHS / United States