Functional analysis of TWIST1 domains regulating smooth muscle cell phenotype

Danielle Charissa Maramara Dy¹Thiel Lehman²Adrian Othon³Mitesh Rathod⁴William J Polacheck⁵Robert Wirka^1,6*

• ¹Department of Cell Biology & Physiology, University of North Carolina at Chapel Hill, Chapel Hill, United States
• ²McAllister Heart Institute, University of North Carolina at Chapel Hill, Chapel Hill, United States
• ³Department of Pharmacology, University of North Carolina at Chapel Hill, Chapel Hill, United States
• ⁴Department of Biomedical Engineering, University of North Carolina at Chapel Hill, Chapel Hill, United States
• ⁵University of North Carolina at Chapel Hill, Department of Biomedical Engineering, Chapel Hill, United States
• ⁶University of North Carolina at Chapel Hill Department of Medicine, Division of Cardiology, Chapel Hill, United States

TWIST1, a bHLH transcription factor, regulates mesenchymal specification, differentiation, proliferation and migration during development and in diseases such as cancer. More recently, genome-wide association studies have identified TWIST1 as a causal gene that increases risk for multiple vascular diseases, including atherosclerosis and hypertension. However, its molecular role in the vascular wall remains unclear. In this study, we interrogated how TWIST1 dimer composition and discrete TWIST1 domains affect SMC phenotype by expressing forced TWIST1 dimers or TWIST1 variants lacking specific domains, followed by bulk RNA sequencing and proliferation and migration assays in human coronary artery SMCs (HCASMCs). We found that TWIST1 homodimers had only modest transcriptomic effects but strongly promoted migration and proliferation - effects abolished by deletion of the TWIST1 N-terminus. Heterodimerization of TWIST1 with TCF3-endoded E proteins resulted in larger transcriptomic effects, promoting Rho/ROCK signaling and extracellular matrix production/organization, but had only modest effects on proliferation and no effect on migration. Deletion of the TWIST1 C-terminus resulted in a very large transcriptomic shift with predicted downregulation of angiotensin and Rho/ROCK signaling as well as ECM production/organization pathways, in a manner suggesting a dominant-negative effect on TWIST1-E12 function. Comparison with single-cell RNA-seq data from human endarterectomy samples placed the function of these TWIST1 variants in a disease context and showed that deletion of the C-terminal domain prevented a modulated SMC phenotype. These studies demonstrate that TWIST1 influences different aspects of SMC phenotype independently via discrete domains and dimer composition, and link TWIST1 to key signaling pathways that influence SMC phenotype during disease.