Structure And Function Of Unconventional Myosins and CARMIL Proteins

National Heart, Lung, And Blood Institute

Investigators

Linked publications, trials & patents

Abstract

Recent solution studies have shown that the isolated CAH3 domain from mouse and Acanthamoeba CARMIL rapidly and potently restores actin polymerization when added to actin filaments previously capped with Capping Protein (CP). To demonstrate this putative uncapping activity directly, we observed single, CP-capped actin filaments before and after the addition of the CAH3 domain from mouse CARMIL-1 (mCAH3) using TIRF microscopy. Addition of mCAH3 rapidly restored the polymerization of individual capped filaments, consistent with uncapping. To verify uncapping, filaments were capped with recombinant mouse CP tagged with mGFP. Restoration of polymerization upon mCAH3 addition was immediately preceded by the complete dissociation of mGFP-CP from the filament end, confirming the CAH3-driven uncapping mechanism. Quantitative analyses showed that the percentage of uncapped actin filaments increased with increasing concentration of mCAH3 added, reaching a maximum of 90% at 250 nM mCAH3. Moreover, the time interval between mCAH3 addition and uncapping decreased with increasing concentration of mCAH3, with the average half-life of CP at the barbed end decreasing from 30 minutes without mCAH3 to 10 seconds with saturating amounts of mCAH3. mCAH3 containing a single point mutation (R993E) that abrogates its tight binding to free CP was totally devoid of uncapping activity in these TIRF-based assays. Finally, using mCAH3 tagged with mGFP, we obtained direct evidence that the complex of CAH3 and CP has a small but measurable affinity for the barbed end, as inferred from kinetic modeling. We conclude that the isolated CAH3 domain of CARMIL, and presumably the intact molecule as well, possesses the ability to uncap CP-capped actin filaments. This activity may drive, along with de novo nucleation and filament severing, the generation of free barbed ends in vivo, and may be responsible at least in part for the short half-life of CP at the barbed end inside cells (JCB, 2006). Our results contrast with a recent report that PIP2, which was also thought from solution studies to uncap CP-capped filaments, does not appear to do so when examined by TIRF microscopy (JBC, 2007).[unreadable] The tubulovesicular contractile vacuole (CV) complex in Dictyostelium exhibits extensive association with and motility along the actin-rich cortex. Here we show that the type V myosin myoJ targets to CV membranes, and that myoJ null cells exhibit increased sensitivity to hypo osmotic conditions, a dramatic loss of CV membranes from the cortex, and an inability of CV bladders to undergo tubulation following water discharge. Complementation of myoJ- cells with GJP-tagged myoJ fully rescues the defects in cortical association of CV membranes and motility of CV tubules. Complementation with versions of myoJ that either walk very slowly or take shorter steps results in rescue of cortical CV membrane distribution in both cases, but rescue of tubulation only in the case of the step size mutant, and the tubules in this case move at half their normal speed. Finally, a steady state accumulation of CV membranes around the MTOC seen in myoJ- cells forced us to visualize CV membrane dynamics in way that could highlight possible microtubule-dependent movements. These images revealed that CV tubules move not only on cortical actin in the plane of the membrane, but bi directionally along microtubules between the cortex and the MTOC as well. Therefore, in addition to myoJs role in driving the cortical association and motility of CV membranes, it cooperates with plus and minus end-directed microtubule motors to drive the proper distribution and dynamics of the CV complex in Dictyostelium.