Abstract 233: Clustering Transcriptional Network is Useful to Characterize Heart Failure Models
Background: It remains to be challenging in the post-genomic era to systematically identify gene regulation pattern during the progression of heart disease in order to establish personalized medicine. In this study, we aimed to identify the unique combination of transcription factor binding sites in the upstream sequences of co-regulated genes in different models of heart failure.
Methods and Results: We used two different models of heart failure in mice (6-week-old), including pressure overload by banding of the transverse aorta (TAC) and neurohormonal activation by angiotensin II infusion (2.0 mg/kg/day for 15 days, AngII) (n=5 each). One week after the operation, the expression profiling by gene-chip microarray (Affymetrix 430A 2.0) was performed in the left ventricle in both models. We extracted the genes that displayed >2-fold increase in the heart failure groups and a 5 standard deviation excess from the mean expression level in the controls. The number of the genes extracted was 511 in the TAC and 13 in the AngII model, respectively. To annotate the information of transcription factor binding sites, we collected the promoter sequences of 18,218 murine genes from DBTSS (http://dbtss.hgc.jp) and motifs of transcription factor binding sites retrieved from the TRANSFAC database. We identified significantly frequent transcription factor binding sites in the promoter regions of the co-regulated genes in the heart failure models (P <0.05, binomial probability). GATA binding sites were noted in both models of heart failure. Eight binding sites, including SP-1 and AP4, were noted in the TAC model, whereas HNF4 and Evil were identified in the AngII model. Finally, we directly injected a firefly luciferase vector plasmid containing each selected binding site into the left ventricle in both models of heart failure (n=10 each). We were able to confirm the enhancement of the transcriptional activity in both models by using the in vivo imaging system (IVIS) in living mice during the progression of heart failure.
Conclusion: Our novel approach is useful to identify the unique combination of transcription factors, in order to characterize different models of heart failure.