Microtubules as regulators of transverse (t)-tubule homeostasis

The rapid, synchronous rise in Ca2+ responsible for cardiac contraction, is achieved by the presence of transverse (t)-tubules(1). Cardiac microtubules promote delivery of L-type Ca2+ channels to the t-tubules aided by the tubule-forming protein BIN1(2). In heart failure (HF), microtubule remodelling is associated with t-tubule loss(3). Beneficial autophagy is also facilitated by microtubules and upregulating autophagy prevents t-tubule loss in culture(4). Therefore we aimed to investigate if microtubules are important for t-tubule homeostasis, and the involvement of autophagy in this process.

Neonatal rat ventricular myocytes (NRVMs) were co-transfected with BIN1 and EMTB-3xGFP to drive tubule formation and label microtubules respectively. 48 hours pre t-tubule development, or 48 hours post t-tubule formation, nocodazole or EHNA (microtubule dynein motor inhibitor) was added to NRVM, to depolymerise microtubules or inhibit microtubule trafficking respectively, and NRVMs cultured further for 48 hours. A subset of nocodazole-treated NRVMs were transduced with an LC3-GFP-mCherry reporter to measure autophagosomes and autolysosomes, giving a read out of autophagic flux The autophagy-mediating microtubule-associated protein 1S (MAP1S) was knocked out in C57BL mice to impair autophagy, and t-tubule networks analysed.

Depolymerization of microtubules with nocodazole, either prior to or following t-tubule formation, decreased t-tubule density and length, and increased fragmentation (p<0.01-<0.0001, N=12). Inhibiting dynein-based microtubule trafficking with EHNA attenuated t-tubule development by decreasing tubule density and length (p=0.0049, and p<0.0001, N=10). EHNA decreased pre-existing tubule density and length (p=0.0001 and p=0.007, N=12) suggesting a role for dynein in both t-tubule elongation and maintenance. In time-matched experiments in LC3-GFP-mCherry transduced NRVMs microtubule depletion with nocodazole caused autophagosome aggregation, indicative of impaired autophagic flux (p=0.005, N=6). Consistently, an increased abundance of the p62 substrate was observed with nocodazole treatment (p=0.003, N=4 litters), further suggesting impaired autophagic flux when microtubules are disrupted. Lastly, the knockdown of MAP1S (-/-), which impairs autophagy and increases mortality, reduced t-tubule network density and length in MAP1S-/- mice compared to WT littermates, suggesting microtubule-mediated autophagy regulates t-tubule integrity (p<0.0001, N=6).

Our data suggests intact, dynamic microtubules and dynein are important for t-tubule development and homeostasis. Similarly, intact microtubules are integral for autophagic flux, as removing microtubules ensues autophagy impairment. As microtubules critically regulate autophagy via MAP1S, and removing microtubule-mediated autophagic flux reduces t-tubule turnover, this data suggests microtubules may mediate t-tubule turnover via autophagy.

Microtubules mediate t-tubule growth