The archetypes: despite the fact that we have hundreds of mutants of the myosin motor, a striking number of relationships exist among these mutants. It's as if there aren't 100 things that can go wrong with the motor, but only a half dozen or so. There are many WAYS each thing can go wrong, and each mutant is a different way--but many mutants represent a single family. Returning to the car engine analogy, you can imagine, for example, that any of the cylinders could shrink, preventing the pistons from moving. That's 4-8 changes (depending on your cylinder count), but ONE result--jammed pistons, broken engine. The cool thing is that by looking at 6 such changes, you would learn that they are all part of the same process--the cylinders constitute a functional family. That's what we are seeing with our mutants, except that the information is far more valuable, because the myosin motor is a protein machine, so it's not conveniently divided into hardware pieces like pistons--we need to learn to perceive the parts AND guess what they're doing.

In order to delve deeper into the issue of mutant families and their relationships, we've selected some representatives--which we term 'archetypes' since they (are intended to) encapsulate the key features of a group. Shown below are the archetypes we are currently working with:

In the figure, underlines are used to indicate cold-sensitive mutants found in our original screen and arrows are used to indicate genetic suppression--i.e. combining a mutant from a gray circle with another mutant that is in a gray circle joined to the first by an arrow results in a functional motor--despite possessing 2 defects, either of which alone renders the motor useless. Given that all our cs-cs suppression interactions have been discovered serendipitously, this is an INCREDIBLE amount of 'order' found in our cs mutants. Could it be that there really are only 4-8 'states' of the motor, and cold sensitive mutants 'mess up' the motor in one of these states? And that a motor that's stuck in state I can be fixed by ALSO altering State II, somehow making the progression I => II easier? That's what we believe, and intend to prove.

Here's the rationale for the groupings (stolen from my NIH grant submission...). In general, a family (grouping enclosed by a gray circle) is chosen if 1 or more of 3 criteria are met:

1. Family members can be suppressed by SEVERAL common secondary mutations. For example, Erik Misner found that E531Qcs, P536Rcs and R562Hcs have at least EIGHT common suppressors! Thus they form a family, represented by archetypes ÆWrinkle (included for other reasons including sharing some of the suppressors) and P536Rcs
2. Family members can suppress another archetype. T189I and V192F are members of a larger group (termed 'the Cluster' elsewhere on this site) including L176F, G182C, C240Ccs, G240Vcs, and L453F all of whom repair the defect caused by G680Vcs.
3. Family members have common biochemistry. Again, the 'Cluster' mutants all show enhanced basal ATPase (the 'drooling' of the motor during which it burns energy but does no useful work, akin to your car idling at an intersection), again uniting them.

It's worth pointing out that several families meet more than one of the criteria

T189I, V192Fcs: Both mutants are suppressors of G680Vcs (and therefore, formally in a 'different' group from it) and have similar biochemical properties as each other (enhanced basal ATPase, poor enhancement of ATPase by actin). T189I is being crystallized.

Y494Kcs, W501Lcs: To date, all suppressors that restore function to one mutant restore function to the other (Y118F, E150K, M486I, P650T, I656F ). Biochemical data from Taro Uyeda suggest their properties are similar, though not identical. They are also in direct contact with each other (and G691) in extant structures. N483S, a suppressor of G680V, appears to be a member of this family.

E531Qcs, P536Rcs, ÆWrinkle: The latter is a deletion of a conserved element in surface helix 510-534; it is discussed more completely in Aim 3. All 3 mutants share a host of suppressors (M486I, E586Kcs, L596S). All three also show a failure to accelerate ATPase in the presence of actin. E531Qcs and P536Rcs appear to differ somewhat in their actin interaction defect; the former appears to fail to initiate strong binding, while the latter may arrest at the strong binding point (Giese and Spudich, 1997).

G691Vcs: While the G691Ccs mutant was found in the initial cs screen (and as a suppressor of P536Rcs), it has only a weak phenotype in the plaque expansion assay. We have shown that G691Vcs is suppressible. It shows overlapping biochemical properties with T189I and V192Fcs, but to date we have not identified common suppressors. Further, it demonstrates actin enhancement of ATPase despite high basal ATPase, while T189I and V192Fcs do not.

G680Vcs: The best characterized of the mutants; it shows defects consistent with overoccupancy of a state just following strong actin binding and preceding the stroke itself. Suppressed by T189I, V192Fcs, G240C,Vcs.

L222F, L596S: Both are common suppressors of E531Qcs, P536Rcs and R562Hcs. Both can be suppressed by V534A and V534G.

Note: divisions between groups are sometimes uncertain in that some are based on a 'no reason to group them together' rationale. At present, the G680Vcs and L222F & L596S groups may 'actually' constitute a single group, as may the T189IG680V & V192Fcs and G691Vcs groups.

Bruce Patterson http://research.biology.arizona.edu/myosin