Background Cardiac fibrosis occurs when quiescent cardiac fibroblasts (CF) transdifferentiate into myofibroblasts in response to cardiac stress. Whether changes in the mechanical environment, as occurs in scarring, are as important as biochemical signals, or even a key factor in the propagation of fibrosis in the diseased heart is currently unknown. Furthermore, the lack of a suitable model hinders the study of the different cellular states, as well as the mechanisms at play, necessary for the identification of new anti-fibrotic drugs.

Objective Developing a dual engineered connective tissue (ECT) model to test the hypothesis that the distinctive geometry and stiffness conferred by the architecture of the mould modulates the human CF phenotype. We intend to implement the ECT model for mechanistic studies of cellular transformation, development of disease modelling and screening of anti-fibrotic drugs.

Methods and Results To investigate the mechanical responsiveness of human CF, ECT were prepared in moulds with two flexible poles (non-uniform, low-stress) or a central rod (uniform, high-stress) and cultured for 5 days. This resulted in ECT with stiffnesses resembling the healthy and fibrotic myocardium. The scar-like character of the high stress ECT became also evident by a lower strain resistance and a heart failure-resembling ECM gene signature. A progressive increase in pole stiffness in the low-stress model resulted in increased tissue stiffness, however, plateauing at a significantly lower level than in the high-stress ECT. The additional use of TGF-β1 was not sufficient to overcome the plateau. The TGF-β1 effect, demonstrated by enhanced ECT contraction, was surprisingly 2-fold, regardless of pole stiffness. Furthermore, we were able to identify important relationships between cell and tissue characteristics, as well as between biomechanical tissue parameters by implementing cells from ICM and non-ICM patients. To further evaluate the mechanical response of CF under low stress, the culture duration was extended. In long-term cultures, a second wave of tissue contraction accompanied by an increase in ECT stiffness was observed, indicating that in the non-uniform model the cells did not achieve mechanical homeostasis with their environment. Addition of elastic fluorescent beads to the ECT revealed a higher stress level in the pole compared to the arm region in conjunction with a higher smooth muscle-actin expression.

Conclusions CF autonomously remodel their environment according to the mechanical situation. As long as high stress areas are present, the process propagates. We hypothesized that paratensile signalling might be a relevant mechanism. Moreover, TGF-β1 was not sufficient to achieve comparable effects to the high mechanical stress in our model, underlining the dominant role of the mechanical environment. Thus, the ECT model allows estimating the different CF phenotypes of the healthy and diseased heart, providing information about disease-related processes and the factors involved, thereby supporting the identification of anti-fibrotic drugs.