Little is known about how molecular motors bind to their vesicular cargo. We have now shown that myosin Va, an actin-based vesicle motor, binds to one of its cargoes, the melanosome, by interacting with a receptor-protein complex containing Rab27a and melanophilin, a postulated Rab27a effector. Rab27a binds to the melanosome first and then recruits melanophilin, which in turn recruits myosin Va. Melanophilin creates this link by binding to Rab27a in a GTP-dependent fashion through its amino terminus, and to myosin Va through its carboxy terminus. This latter interaction, similar to the ability of myosin Va to colocalize with melanosomes and influence their distribution in vivo, is absolutely dependent on the presence of exon-F, an alternatively spliced exon in the myosin Va tail. These results have provided the first molecular description of an organelle motor for an actin-based motor, illustrated how alternate exon usage can be used to specify cargo, and further expanded the functional repertiore of Rab GTPases and their effectors. Regulation of actin filament dynamics by actin binding proteins is critical for the motility, shape, and structure of cells. The Dictyostelium protein CARMIL (for Capping protein, ARp2/3 complex, Myosin I Linker) appears to serve as a scaffold that links three different actin binding proteins: capping protein (CP), the major terminator of actin filament growth, the Arp2/3 complex, the major nucleator of actin filament growth, and myosin I, a ubiquitous barbed-end-directed motor protein. In vivo CARMIL co-localizes with CP, Arp2/3, and myosin I in dynamic actin-rich extensions like the leading edge of migrating cells and dorsal macropinocytic structures. Furthermore, in Dictyostelium cells lacking CARMIL, pinocytosis and chemotactic aggregation are severely inhibited and the cellular content of F-actin is decreased. These observations suggest that CARMIL, which is also present in vertebrate cells, is physiologically important. To characterize CARMIL biochemically we have now purified its homolog from Acanthamoeba. The most striking observation regarding the purification is that CP copurifies extensively with CARMIL. Separation of CARMIL and CP was achieved by gel filtration at low pH. Gel filtration studies also suggested that CARMIL must either have an elongated shape or exist in multimers, since elution volumes reflected a much higher molecular weight than CARMIL?s sequence implies. Consistent with these observations, chemical crosslinking and analytical ultra centrifugation showed that purified CARMIL is in a monomer-dimer equilibrium with an association constant of 10?6M?1, that CP promotes dimerization, and that the CARMIL dimer is very asymmetric. Purified CP rebinds to CARMIL with a stochiometry of one CP per CARMIL dimer, and with an affinity of 400 nM or tighter. Given the cellular concentrations of CARMIL and CP (2 and 1 micromolar, respectively), binding of CP to CARMIL could significantly affect the growth of actin filaments in vivo. Importantly, in vitro experiments showed that the CP-CARMIL complex can still cap the barbed ends of actin filaments. Together, these results indicate that CARMIL is a bonafide CP interacter, and that their interaction does not serve to simply sequester CP. Given that CARMIL also interacts with myosin I and Arp2/3, we suggest that the CP-CARMIL complex may be translocated to the barbed end by myosin I, and through its ability to recruit Arp2/3, initiate polymerization of a new filament off of a capped filament end.
Showing the most recent 10 out of 25 publications
