Purpose: Transverse tubules (T-Tubules, TTs) form a complex network of continuous membrane invaginations which are essential to ensure effective EC coupling in ventricular myocytes (VMs). Resolution limited microscopic studies suggested a loss of TTs as a leading mechanism in late stages of heart failure (HF). However, TTs have dimensions below the resolution limit of conventional light microscopy. Therefore, we employed superresolution microscopy and novel quantitative image analysis strategies to directly quantify TT remodeling in different stages during HF development.
Methods: TTs were investigated in VMs of two different murine HF models. Analyses were performed at an early (1 week), an intermediate (4 wks) and a late (8 wks) stage of HF development both after myocardial infarction (MI) and transverse aortic constriction (TAC). Stimulated emission depletion (STED) microscopy was applied to visualize di-8-ANEPPS stained TTs in living VMs. STED live cell imaging was complemented by imaging of immunostained fixed VMs, by protein biochemical analyses and by multimodal imaging of TTs and Ca2+ transients.
Results: Already 1 week after MI and TAC, analysis of STED images revealed significant changes on the level of the TT network. A strong increase of TT density was accompanied by significantly more axial TT elements and a higher branching complexity compared to Sham control groups. 4 wks after MI and TAC these changes were observed to almost the same extent, whereas a tendency towards a decreasing TT density occured 8 wks after the interventions. In addition, the size of individual TT cross-sections increased significantly after both interventions, starting early and progressing during HF development. Analysis of Caveolin-3 (Cav3) immunostains revealed a strong increase of Cav3 positive axial structures after MI and TAC suggesting a potential role in the development of the increasing amount of axial TT structures. This was supported by an increased Cav3 protein expression determined by Western blot analysis. Simultaneous imaging of TTs and Ca2+ transients suggested that TT remodeling might directly lead to a delayed and heterogeneous intracellular Ca2+ release.
Conclusions: We identified significant proliferative TT remodeling mechanisms in two independent mouse models (MI and TAC), already in early stages of HF development. Our data suggest that TT remodeling might directly lead to defective Ca2+ release units and accordingly to dyssynchronous intracellular Ca2+ release. This might in turn increase the risk of delayed afterdepolarizations, lead to action potential prolongation and arrhythmias.