Regulation Of Smooth and Nonmuscle Myosin

National Heart, Lung, And Blood Institute

Investigators

Linked publications & trials

Abstract

Nonmuscle myosin 2 (NM2) molecules carry out a wide variety of functions within cells. There are three NM2 heavy chain genes in mammals. We are expressing full length NM2 proteins and fragments of these myosins in the baculovirus Sf9 system. The find that NM2A moves actin filaments the fastest and NM2C, the slowest of the three paralogs. We are studying their structure both at the level of the single molecule folded autoinhibited state as well as the structure of myosin filaments. We are particularly interested in how phosphorylation of both the heavy chain and light chain affects filament formation and activity. We use a single filament motility assay system wherein we can image the movement of fluorescently labeled myosin filaments over actin filaments fixed to the surface. We are examining the copolymerization of NM2A and NM2B in vitro and wish to understand the dynamics of filament assembly. Optical trapping studies reveal that NM2A and NM2B are not processive as single molecules. Bipolar filaments of NM2B containing about 30 myosin molecules move processively along actin filaments attached to the surface. In collaboration with others, we have obtained high resolution structures of both NM2A and NM2B bound to actin in the presence and absence of ADP using cryo-electron microscopy and have crystallized NM2A in the presence of ATP analogs. This will give us the ability to observe high resolution structures of myosin in all four cross-bridge states. We also have solved the structure of the folded, autoinhibited form (also called the “off state”)of nonmuscle myosin 2B at less than 4 Angstrom resolution. This has allowed us to determine the structural basis for this 10S structure and to speculate on why phosphorylation activates the myosin. We have also obtained a high resolution structure of the NM2B molecule in the folded, off state. We are using micropatterning in conjunction with TIRF microscopy to create higher order actin structures to examine their interaction with NM2 paralogs. One of these involves creating parallel stripes of various separation on a microscope coverslip which can be coated with formin, an actin nucleator. Upon adding G-actin the formin nucleates polymerization of actin in which the barbed in is bound to the stripe. In the middle of the patterns between stripes there is an antiparallel array of actin filaments. When NM2B filaments are added to this system they bind to the actin filaments and begin to generate tension which aligns the actin filaments. The bipolar myosin filaments near the center area between stripes encounter actin filaments of opposite polarity which causes the myosin filaments to effectively stall where they are likely generating isometric tension on the actin filaments within the stripes. In these cases the myosin filaments remain bound to the actin filaments for a long time, likely because the kinetic cycle has been slowed by force dependent processes. We are using a myosin that had a FRET-based force sensor embedded into the S2 region of the myosin. Preliminary data show that as the actin filaments become tensed a portion of the myosin molecules in a filament become force bearing. When NM2A is added to this system the actin filaments are rapidly shredded. This suggests that an in vivo function for this myosin, which is abundantly present in the actin arcs just behind the Arp2/3 branched actin network at the leading edges of some cells, might be to shred actin filaments to promote depolymerization creating more actin monomers that can feed the protrusive machinery at the leading edge. We have expressed a myosin with a built in FRET-based force sensor and see that the myosin filaments adopt a tensed state when interacting with the actin filaments of opposite polarity. In addition, the same force sensor was built into the formin used to polymerize the actin from the pattern’s stripes. After creation of the actin network the force sensor shows no tension in the formin, but this changes upon addition of myosin. We are using fluorescently-labeled myosin molecules to examine the exchange of myosin regulatory light chain using TIRF microscopy. Preliminary data show that the RLC exchanges in a slow time scale between heavy chains of different myosins. In collaboration with Mike Ostap’s lab at the University of Pennsylvania, we are using optical trapping to examine the force dependencies of the kinetics for NM2A and NM2B to continue studies carried out at NIH. We are also using TIRF microscopy to examine the exchange of myosins between filaments. Early findings show that filaments composed of NM2A exchange more rapidly than those of NM2B. NM2A is phosphorylated by Protein Kinase C at a single site near the end of the tail. We have created non-phosphorylatable versions of this myosin by substituting the appropriate serine residue with and alanine and have created a phosphomimetic version by replacing this residue with an aspartate. We will examine how this phosphorylation controls filament assembly. The nonmuscle myosin 2 paralogs can be alternatively spliced at a region, termed loop 2, which is near the actin binding region of the myosin motor domain. We previously published that a heavy meromyosin-like construct of NM2C containing this splice no longer required phosphorylation of the regulatory light chain (RLC) for its enzymatic activity to be activated by actin. In other words, this myosin was constitutively active. We have revisited this question in the context of the full length NM2C and find that the alternatively spliced myosin is largely inactive in the absence of RLC phosphorylation. It is now known that the “off” state involves an asymmetric interaction between the two motor domains and that this state is stabilized by three sections of the myosin rod. An explanation for the lack of regulation of the truncated HMM vs the regulation of the full length NM2C could be that the other two segments of the tail of the full length molecule is necessary to stabilize the motor-motor interactions necessary to suppress the enzymatic activity. We are also very interested in the interaction between NM2 paralogs and the closely related myosin 18A (M18A) which we have shown is a pseudo-enzyme that does not hydrolyze ATP and only interacts with actin weakly. It also does not form filaments on its own, but will copolymerize with NM2 paralogs. M18A has an alternatively-spliced form with an additional amino acid sequence at the N-terminus that is proposed to be an ATP-insensitive actin binding domain. We have expressed this form of M18A and will examine how it affects the mechanical and motile properties of NM2/M18A cofilaments.