TGF-beta1 mediated expression and accumulation of collagen IV by bronchial airway epithelium in asthmatics and smokers: an in silico experiment

Fibrosis is a common complication of asthma and other peribronchiolar lung diseases leading to death of many individuals in the United States. Much traditional bench work research has been done to detect and cure asthma. However, to narrow the attempts made in order to develop more effective therapies for fibrosis of the basement membrane, several computational methods have been employed. We employ an in silico computational approach for simulating an in situ environment and analyzing the effect of a range of TGF-β1concentrations on the proliferation of collagen IV levels in the extracellular matrix (ECM) and collagen IV gene expression levels. Three studies were performed using the Transcriptome to Reactome (TTR(TM)) deterministic modeling method: 1) asthma vs. healthy control; 2) smoker vs. healthy control and 3) one week placebo vs. fluticasone. The primary datasets were the bronchial airway epithelial transcriptome samples of 118 human subjects, accessed from NCBI GEO GSE4302 profile. A software, COmplex PAthway SImulator (COPASI ®), and TGF-β1 Signaling Reactome model were used to perform biosimulations and generate secondary data sets for each of these transcriptome samples by subjecting them to 15 different concentrations of TGF-β1 in silico. Half maximal effective concentration (EC50) values were determined to assess sensitivity differences. Endogenous pathophysiological and maximum dose response values for each individual transcriptome sample were derived from the dose response curves and were subjected to statistical analysis. There was a significant difference in levels of collagen IV species and gene expression fluxes of asthmatics at baseline versus healthy controls and at 10pg/ml concentration of TGF-β1. Also, the means of maximum responses of smokers versus healthy controls showed statistically significant differences for collagen IV levels (p<0.001) and gene expression fluxes (p<0.05). However, the hypotheses that increase in TGF-β1 signaling: 1) leads to increase in collagen IV production in asthmatics vs. healthy controls; 2) smokers vs. healthy controls and; 3) corticosteroid drug fluticasone normalizes the expression levels of collagen IV at the transcription level and as extra cellular matrix protein concentration were refuted. This study showed no sensitivity differences to TGF-β1 concentrations when EC50 and maximum response levels were analyzed for collagen IV levels and gene expression values. No changes in levels of collagen IV in fluticasone vs. placebo controls was seen due to short duration (1 week) of fluticasone drug exposure; this result is consistent with numerous clinical reports in human patients. Therefore, we conclude that the TTR(TM) method reliably reproduces in vivo observations and reveals novel details that would be impossible to perform in situ in humans. Further, the TTR(TM) biosimulation method can be a powerful tool to perform fully empirical in silico experiments.