Computational Model Predicts Paracrine and Intracellular Drivers of Fibroblast Phenotype After Myocardial Infarction

Abstract

2The fibroblast is a key mediator of wound healing in the heart and other organs, yet how 2 3 it integrates multiple time-dependent paracrine signals to control extracellular matrix 2 4 synthesis has been difficult to study in vivo. Here, we extended a computational model to 2 5 simulate the dynamics of fibroblast signaling and fibrosis after myocardial infarction in 2 6 response to time-dependent data for nine paracrine stimuli. This computational model 2 7was validated against dynamic collagen expression and collagen area fraction data from 2 8 post-infarction rat hearts. The model predicted that while many features of the fibroblast 2 9 phenotype at inflammatory or maturation phases of healing could be recapitulated by 3 0 single static paracrine stimuli (interleukin-1 and angiotensin-II, respectively), mimicking 3 1 of the proliferative phase required paired stimuli (e.g. TGFβ and angiotensin-II). Virtual 3 2 overexpression screens with static cytokine pairs and after myocardial infarction 3 3 predicted phase-specific regulators of collagen expression. Several regulators increased 3 4 (Smad3) or decreased (Smad7, protein kinase G) collagen expression specifically in the 3 5 proliferative phase. NADPH oxidase overexpression sustained collagen expression from 3 6 proliferative to maturation phases, driven by TGFβ and endothelin positive feedback 3 7 loops. Interleukin-1 overexpression suppressed collagen via NFκB and BAMBI (BMP 3 8 and activin membrane-bound inhibitor) incoherent feedforward loops, but it then later 3 9 sustained collagen expression due to the TGFβ positive feedback loop. These model-4 0 based predictions reveal network mechanisms by which the dynamics of paracrine stimuli 4 1 and interacting signaling pathways drive the progression of fibroblast phenotypes and 4 2 fibrosis after myocardial infarction. 4 3 4 4 4 5Wound healing is a complex process that involves a dynamic interplay between 4 6 inflammatory and proliferative signaling. This process is especially important following 4 7 injury to the heart, where cardiomyocytes are unable to regenerate. Scar formation and 4 8 the preservation of viable heart muscle are important for continued cardiac function[1]. 4 9 8 1 understanding how fibroblasts respond during the different phases of wound healing 8 2 could identify mechanisms by which fibrosis develops in other organs. 8 3 Myocardial infarct healing is notoriously difficult to investigate because it 8 4 involves many dynamic and interacting signaling processes. Fibroblasts are particularly 8 5 difficult to study in situ during wound healing because they can differentiate from many 8 6 different cell types and there is no clear consensus on fibroblast markers[20].8 7Computational modeling has been a useful method for investigating complex dynamic 8 8 processes in many areas of biology. Although models have been constructed to study the 8 9 wound healing process post-MI[21], no such model has yet been applied to study 9 0 5 fibroblast intracellular signaling and phenotypic changes during m...

The Cell Surface Receptors Ror1/2 Control Cardiac Myofibroblast Differentiation

“…RNA was then isolated, and next-generation RNA sequencing was performed. Many genes associated with a manually curated network of fibrosis and fibroblasts 50, 51 were differentially regulated between control and Ror1/2-KO fibroblasts ( Figure 3A ). As expected, cells were largely void of Ror1 and Ror2 transcripts, and the planar cell polarity gene Ptk7, Vangl1, Vangl2, Prickle1, Dvl1, Dvl2, Dvl3 ) were generally reduced in the Ror1/2-KO fibroblasts.…”

Fibroblast Mechanotransduction Network Predicts Targets for Mechano-Adaptive Infarct Therapies

“…This capability is particularly important in the context of an infarcted left ventricle wherein fibroblasts in the infarct scar are subjected to heightened tensile stretches compared to the fibroblasts in the remote, non-infarcted myocardium that are subjected to normal myocardial tensions (Torres et al, 2018). We investigated this capability by simulating the expression of model outputs in response to valsartan treatment in both low-and high-tension contexts, while setting the model cytokine inputs to experimentally matched postinfarct levels as previously described (Zeigler et al, 2020). We compared model output expressions to a study by Ramirez and colleagues, who observed significant differences in fibrosis-related gene expression between infarct and remote zones and negligible effects of valsartan treatment alone using a mouse model (Ramirez et al, 2014).…”

“…All negative control simulations were conducted using input reaction weights of 0.1, and all input reaction weights and drug perturbation levels for in vitro studies were chosen to maximize the dynamic range of network responses. Input reaction weights for in vivo studies were interpolated from experimental data of cytokine levels post-MI as previously described (Zeigler et al, 2020) using a time point of 4 weeks post-infarction, and tension reaction weights for control and infarct zones were set to 0.1 and 0.6, respectively. All input reaction weights and perturbation conditions can be found in Table 1.…”

“…Computational approaches for investigating cell signaling have shown promise for discovering mechanisms-of-action underlying pathological states and drug treatments. Previous models representing cardiac cell signaling as a network of interactions have proven successful in generating detailed insight into cardiomyocyte contraction and hypertrophy as well as fibroblast chemotransduction and various other biological systems (Saez-Rodriguez et al, 2015;Saucerman and McCulloch, 2004;Tan et al, 2017;Zeigler et al, 2020). Previous models have simulated cardiac and adventitial fibroblast activation and pro-fibrotic activity in response to biochemical and biomechanical cues, although limited exploration of mechanotransduction pathways in these models still leave unanswered questions about the role of tension in regional matrix turnover (Wang et al, 2020;Zeigler et al, 2016).…”

Network model‐based screen for FDA‐approved drugs affecting cardiac fibrosis

“…We build on our previous literature-validated model of cardiac fibroblast signaling, 27 which has predicted network mechanisms underlying fibroblast phenotype in vitro and during post-MI wound healing. 28 In this study, we create a framework for integrating this fibroblast network model with the drug-target DrugBank database 19 to perform an in silico screen for drugs that are profibrotic or antifibrotic. We validate the predicted drug effects on fibrosis with in vivo studies from the literature, and we further validate the predicted effect of Galunisertib on fibroblast phenotype in vitro.…”