E. H. Steen1, M. Fahrenholtz4, M. Kang3, L. Wadhwa4, J. Grande-Allen1,3, S. Balaji1, S. Keswani1,2 1Baylor College Of Medicine,Department Of Surgery,Houston, TX, USA 2Texas Children’s Hospital,Department Of Pediatric Surgery,Houston, TX, USA 3Rice University,Department Of Bioengineering,Houston, TX, USA 4Texas Children’s Hospital,Department Of Surgery, Congenital Heart Surgery Service,Houston, TX, USA
Introduction: Discrete subaortic stenosis (DSS) is characterized by the formation of fibromembranous tissue in the left ventricular outflow tract (LVOT), leading to cardiac dysfunction. Surgical resection has a high rate of recurrence which characterizes an aggressive phenotype. The pathogenesis of DSS is theorized to be due to altered LVOT geometry, producing shear forces that stimulate a fibrotic response. We used computational modeling of patient echocardiograms (ECHO) to develop a novel parallel flow bioreactor to mimic DSS shear forces. We hypothesize that increased shear force on cardiac endothelial cells promote endothelial-fibroblast crosstalk via CD31, a known mechanosensory molecule, and via inflammatory cytokines to stimulate fibrotic tissue formation.
Methods: Human DSS tissues (n=7; 3 aggressive/recurrent DSS, 4 primary DSS) were compared using special stains and ICC to develop a histology-based scoring system, then compared to the patient’s preoperative ECHO reports to define a correlation between ECHO and aggressive histologic phenotype. In a parallel plate flow chamber, co-cultures of endocardial endothelial cells (EEC) and cardiac fibroblasts (CF) isolated from porcine LVOT were subjected to static, low, and high shear conditions modeled on patient ECHO data, then analyzed by ICC (CD31, VE-cadherin) and PCR. To test the role of CD31 and EEC, CD31 signaling was inhibited using a Src inhibitor in high shear conditions and similarly analyzed.
Results: Histological analysis reveals notable differences in aggressive membranes, characterized by immature collagen with increased matrix turnover and by increased cellularity compared to non-aggressive samples. These parameters were developed into a scoring system with aggressive samples having higher scores, which are commensurately associated with high preoperative mean LVOT gradients (55± 5?mmHg by ECHO), an accepted predictor of DSS recurrence. In vitro, high shear conditions causes de-localization of mechanosensitive CD31 from VE-cadherin-rich cellular junctions. PCR array reveals upregulation of pro-inflammatory CSF1 and CSF3 in the high shear condition (4.2 and 4.7fold vs low shear) and upregulation of chemoattractants CCL3, CCL4, and CCL5 (15.3, 15.6, 7.6fold). Inhibition of CD31 signaling attenuates the effect of high shear (3.5fold decrease in CCL3 and CCL4, 4fold decrease in CCL5, and 2fold decrease in CSF3).
Conclusion: Our data suggest that increased shear forces change the expression profile of mechanosensitive proteins and cytokines in EEC in part due to a CD31-dependent signaling mechanism. Additionally, we developed a histology-based scoring system that correlated with increased shear force on ECHO to define an aggressive/recurrent DSS phenotype. This work implicates the role of shear forces in the development of DSS and may help predict patients susceptible to recurrence and elucidate molecular therapeutics for DSS and other disease processes characterized by fibrosis.