Sea of Hypotheses

As the CamShaft turns...


ne totally excellent conjecture has recently arisen from a slew of genetic data. It has to do with conformations of the CamShaft, and is thus central to the myosin motor mission. Below is the mad tale of what I think this means and exactly what I expect to find in our quest to nail down the findings.


o, who's the lucky suppressor? M486I. And who are the lonely mutants made whole by this unsavory change? L453F, E531Q, P536R and delWrinkle (replacement of the sequence G515-G524 with a single alanine). So, how do we attempt to unravel this story? First, let's review what we know about each of the perpetrators.

L453F: A suppressor of the G680V mutation. L453F mutants (without accompanying G680V mutations) show enhanced basal ATPase rates--they let Pi 'slip out' (presumable past the GateKeepers of the BackDoor) in the absence of actin binding. So we believe they favor the 'doors open' conformation of myosin. And we believe this effect comes from putting 'pressure' on the camshaft from 'above' (this story is told throughout these pages, but the specific conjecture can be found in the CamShaft Rules section and the DataDungeon).

E531Q and P536R: These two mutants are deemed 'actin binding mutants' on the basis of their location in the crystal structure and efforts to guess the actin-myosin interaction by combining electron microscopic images with crystal information. Biochemical analysis of these mutations performed by Kim Giese and Jim Spudich indicate that these mutants do have defects during actin interaction, but a clear picture has not yet emerged. Erik Misner's genetic data (to appear in a month or two on these pages) demonstrates that both mutations can be suppressed by the same suppressors.

delWrinkle: this mutation was made in order to investigate the Quivering Chin Hypothesis. All we know at this point is that deleting the Wrinkle (amino acids 519-524) yields a non-functional myosin. AND we know that 2 mutations that restore function to the deletion also suppress E531Q and P536R (including M486I); another pair of suppressors are strikingly similar.

M486I: We know little about this mutation, but can make a bevy of interesting observations. First, we know it's position in the crystal structure. Here is where the plot first thickens. Residue 486 is 2/3 of the way down the CamShaft. But here's the kicker. You may recall that the CamShaft is bent in the Dictyostelium vanadate structure, but straight in the chicken structure and the Dictyostelium beryllium fluoride structure. In case you forgot, feast your eyes on the image below.

Comparison of CamShaft structures (whale 'facing' right). Dicty vanadate structure is shown in color (vanadate is pinkish), chicken structure in gray. Dicty M486 is colored magenta. Note how the helices diverge just after position 486.



ut wait! There's more! In all 4 cases to date, the change at position 486 that gives rise to suppression is M->I (methionine[Met] > isoleucine[Ile]). Can we integrate this is to our wild conjecture? The structures of isoleucine and methionine are both full of carbons and hydrogens. Thus both are chemically similar, hydrophobic ("water-fearing"; preferring to associate with similar residues rather than with water) elements. However, where Met is a long, oft-bent cylinder, Ile is a forked or "branched" molecule. The location of the branch has important consequences for the two residue's willingness to participate in a-helix structures. Wherease the slender Met fits nicely iinto an alpha helix, the elbow of the isoleucine discomfits its neighbors sufficiently to locally weaken the helix. Thus we might (and do!) predict that an Ile at position 486 would favor the 'bent' helical structure more than Met would.


s there more? You betcha! You may ask yourself 'what other amino acids have the property of disgruntling helices? There are a couple forms to the answer. Valine (Val) and Threonine (Thr) have similar branched structures to Ile. Glycine (Gly) has no sidechain whatsoever, and is thus not all that excited about being in a helix (although it doesn't actively object, it's desire to explore other possibilities is detrimental to the helix). Proline is 'the helix-breaker', as it exists in a specific conformation incompatible with helix membership. So I would guess that Val and Thr can also suppress L453F, E531Q, P536R and delWrinkle. The absence of Val and Thr in our suppressor collection is therefore perplexing to me.


here's one more thing: what is the identity of this residue in other myosins? Now we get a little weird. In myosins most similar to Dictyostelium's (primarily those from muscle tissues), the residue is always a Met. However, in virtually all other myosins, this position is occuppied by...Ile, Val or Thr!!! How do I squirm out of this conumdrum? Quite easily, actually. Recall that our starting assumption in playing this game is that myosin has several favorite shapes, and that the cycling of the motor is progression from one state to another. Myosin's tendency to be in any shape is the result of

  • Favorable internal interactions of the shape
  • Unfavorable internal interactions of the shape
  • Favorable or unfavorable interactions of the shape with actin and/or ATP
  • Competition with other possible shapes
Presence of Met or Ile at the bending point of the helix would be expected to influence myosin's preference, but not in a vacuum--other positions are also contributing to the decision. So my interpretation would be that muscle myosins have chosen a 'stiff' helix, and use other factors to cause bending at the appropriate times, whereas non-muscle type myosins have opted for a more bendable helix, compensated perhaps by a lesser tendency of the rest of the molecule to favor bending.

What do we intend to do about it?


ttack position 486 directly! The above thinking yields several direct predictions about what residues will and will not act as suppressors when substituted for Met at position 486. Val and Thr should suppress, and Gly might. The other 16 amino acids, on the other hand, ought not (although Pro is tough to predict--it will force bending at the appropriate place, but this might be too detrimental to myosin when it wants the helix to be straight. Further, the bend may not be of the proper angle). I have ordered the tools required to alter position 486 either randomly or specifically to Glycine. These random alterations will be combined with each of the starter mutations (E531Q, P536R, L453F and delWrinkle). Suppressors will be sought amongst the resulting double mutants. Then we'll sequence the lucky winners and see if we find Val, Ile and Thr--and no others (excepting perhaps Gly and maybe maybe Pro). This work should take a month or two to complete; results will be posted here!


What Happened


ake your best guess and click HERE to see how you came out against the House!


Previous Hypothesis Sea of Hypotheses Main Page  
NavTable Comments Help!


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