One Bad Motor
|A Real Bad Motor: G680V Myosin|
f the mutants discussed on these pages, we claim to know the most about the G680V mutant. What do we claim to know and how did we come to know it? Here's the straight dope:
The G680V mutant was "captured" in our search for mutant motors that behaved moderately well at high temperatures, but were biologically inert at low temperatures. On the basis of this behavior, and upon discovering that the underlying mutation lay in a part of the motor that was known to be moving around during the cycle, we investigated it further.
he first noteworthy feature of the G680V motor came to light when Kathy Ruppel, then of Jim Spudich's lab (as was I), watched purified motors moving actin filaments. Actually, she watched moving actin filaments that had been treated so that they glowed under UV light, and the only thing that could've been moving them were G680V motors. They moved extremely slowly. No real surprise, given that they'd been doing a poor job in the cell! What was more striking was that they also forced slow movement on normal motors when the two were mixed together! This could only mean one of two things--the mutant motors were interfering with the normal ('wild type') motors by some magical means, or they were holding onto the actin filaments that the wild type motors were trying to move, and preventing motion--they were acting as "brakes."
his braking ability was quite exciting to us. It meant that whatever was wrong with the G680V motors, it involved holding tightly to actin. Assuming that the defect of the motors was in timing rather than in a more drastic misbehavior lead us to the hypothesis that the motors were stuck somewhere in the cycle when they were supposed to be holding or moving actin. Since this is the most mysterious part of the cycle, we plunged ahead with our quest: to figure out just when the pausing or sticking was occurring. The tools at our disposal: our knowledge of the motor's normal cycle and the methods of analysis we had adapted for our use.
ur first step was to examine the way the motor used fuel--measuring its ATPase (=ATP burning) activity. Briefly, Kathy found that the motor used very little ATP when idling, and less than wild type motors when actin was added. Nonetheless, the activation of the motor--the extent to which is revved up when actin was added-- wasn't so bad.In other words, it was stingy with its use of fuel, but apparently recognized when there was work to be done and proceeded slowly.
For the purposes of this discussion, I'll be using the linear version of the cycle introduced on the MotorCycle page:
ince the misbehaving G680V motors had tipped us off to their overzealous affinity for actin, we next investigated the actin-releasing behavior of the motor. Wild type motors "rest" with a tight grip on actin, as do G680V mutant motors. But in the presence of ATP, we detect a substantial difference. While the wild type motors spend most of their time off of actin in the presence of ATP (as indicated by the dotted brown line above) the G680V motors spend the majority of their time "dancing with actin". They do respond to ATP by increasing the amount of time they spend off of actin, but not nearly as much as wild type motors. We focused on 4 possibilities: 1) That a weak-binding state had been altered to bind strongly (dotted brown line had become a thick one), 2) that the motor was reluctant to bind ATP, causing it to overoccupy the 'rigor' (yellow, nucleotide-free) state, 3) that the motor was slow to initiate the stroke itself, or hung up during the stroke (blue purple and red states), and 4) that the motor was reluctant to let go of ADP when it had finished the stroke, and thus remained bound to actin (orange state).
he first two possibilities, that the motor had lost its taste for nucleotide (ATP) or failed to respond to its presence, weres addressed in an indirect fashion. The problem with trying to determine the effect of ATP binding on the motor is that the motor refuses to sit still--as soon as it gets ATP it breaks it, and is then in a position to proceed with the cycle. So we instead fed it 'ATPgammaS': a molecule much like ADP, but with a sulfur atom substituted for an oxygen, rendering it much harder for the myosin motor to "burn" or break into ADP and Pi.
o, what happened when we "fed" the motor ATPgammaS? We found that the motor gracefully released from actin. This indicates two things: first, that the motor responds to ATPgammaS in the normal way--by weakening its desire for actin. Second, we drew the provisional conclusion that the motor had not lost its liking for ATP (a conclusion which we supported by a more direct test that I won't describe here). Since ATP and ATPgammaS are very similar molecules, the fact that the G680V motors liked one implied that it liked the other. Indeed, by varying the amount of ATPgammaS we added to each "bottle" of motors and actin, we could tell just how much the motor liked ATPgammaS. We found that the G680V motors had a significantly greater affinity for ATPgammaS than wild-type. We could tell this because we could get most of the G680V motors to release from actin with only 1/40th as much ATPgammaS in solution than was required to get wild type motors to release. This ability to "suck" ATPgammaS out of the solution is an indication of the 'affinity' of the motor for the nucleotide.
o we had narrowed the problem down somewhat: the problem was not that the G680V motor "waited around" without a nucleotide (ATP or ADP), gripping tightly to actin. Parts of the cycle that we have ruled out are now hidden behind a pink box in the figure below:
e next investigated the possibility that the motor did its working steps, but refused to relinquish its grip on ADP at the end (orange step). This, too, would cause G680V motors to act as a "braking force" on filaments that neighboring wild type motors were trying to move. However, motors with or without ADP bound show the same basic behavior (i.e. strong, stable interaction with actin as indicated by the thick brown line) we could not assess the motor's preference for ADP using the actin release assay...or could we? We tried a variation on the assay that allowed us to reverse the "default" behavior of the motor from bound to released.
ow did we achieve such a thing? We relied again on our buddy ATPgammaS. Recall that we had demonstrated that this nucleotide caused release of the motor from actin. So if we flooded the motors with ATPgammaS, most of the would be in an actin releasing state most of the time. So if ADP snuck into a motors nucleotide binding site, its behavior would change to strong binding to actin, which we could readily detect. We therefore introduced an amount of ATPgammaS sufficient to release most of the G680V mutant motors from actin. We then introduced increasing amounts of ADP into the mixture until we could detect motors binding to actin--evidence that they were loading ADP on board despite the presence of ATPgammaS. This 'competition' assay allowed us to figure out the relative affinity of the motor for ADP vs. ATPgammaS. Our finding: the relative affinities for ADP and ATPgammaS were the same for the mutant motor and the wild type motor. Since we knew that the mutant G680V motors had a 40-fold higher affinity for ATPgammaS, it must also have a 40-fold higher affinity for ADP. What did this mean? Simply this: since the G680V mutant motor exhibited greater enthusiasm for ADP than the wild-type motor, it was quite plausible that its problem arose from the fact that it held on to ADP overenthusiastically, "freezing" it in a strong actin-binding state. This could account for the ability of the G680V motor to interfere with the hard-working wild type motors. Below is the resulting view of the G680V cycle, with the candidate step boxed in green.
o we have restricted the problem points to a subset of the cycle and generated a candidate change, but still don't know what's really wrong. Notice that the hypothetical strength of actin binding (thickness of brown line) changes between the states that we are considering as candidates for the flawed behavior of G680V motors. One way of determining just how strongly the motor feels about actin is to "distract it" with a salt, in this case KCl. In the figure below, I've added dashed lines to help you consider how competition with salt would require greater strength of actin binding. If the brown line is not wider than the dashed lines, then myosin's desire for actin would be overwhelmed by its attentiveness to salt, and we would not observe binding. Note that according to this interpretation, the "empty" ('rigor'), stroking and ADP states of the motor would be sufficiently strong to remain bound in the presence of salt, while the "attaching" and Pi release states would not.
Since the excessive stickiness of the G680V motors is defeated by salt whereas its strong binding interactions are not we conclude that the G680V motor must be held up in some intermediate state of actin binding!!! To restate, this experiment tells us that motors that are freed by salt are NOT in the ADP or nucleotide free states. So I have added to the figure pink boxes over all the strong actin binding states, since we know we're looking for a weak or intermediate strength state, and have already ruled out the weak ones!
he extremely cool thing is that we can squeeze some more information out of the data. First, we can begin thinking about exactly what's wrong with the motors. We believe that they are screwing up somewhere around the stroking part of the cycle. By reexamining the ATPase data, we can extract something else: recall that the mutant motors are slower in their "idling" ATPase than wild type ones. From the work of others, we know that the slowest step in the cycle in the absence of actin is the release of phosphate. This is sensible, given the model that actin binding causes a "gateway" to open up. In the absence of actin binding, the only way for the Pi to escape is for the gateway to "jiggle" open. If the G680V motors have stiffer gateways, we can explain both their lower basal ATPase (stiffer gates would "jiggle" open less often) and the 'cold-sensitive' nature of the motors. If we presume that the gates are stiff enough that the binding of actin doesn't fully overpower them, then the temperature is also important. The greater the random motion of the gateway, the easier it would be for actin binding to tip the balance in favor of opening.
hile I'm storytelling, let me throw in one final interpretive discourse. The intermediate binding strength we observe for the G680V motors in the presence of actin is novel. Unlike wild type motors, which are mostly unbound in the presence of ATP anyway, the G680V motors are spending a lot of their time in a state that shows up as "bound" in the absence of salt. In the presence of salt, however, most of them let go. This is in contrast to both G680V motors and wild type motors in the presence of ADP or in the absence of nucleotide: here, both motors remain stuck in even in the presence of salt. So the G680V motors in the presence of ATP are sticking to actin less well than "fully attached" motors. One exciting possibility is that the G680V motors initiate binding to actin, which is supposed to trigger the opening of the gateway. But since the gateway is stiff, the motors wait--and fail to progress on to a more complete binding to actin. Thus they are sensitive to the presence of salt, whereas fully attached motors are not. So the resulting view of the cycle is shown below. Note that it is just like the wild type cycle, except that the blue triangle of "Bind Actin" has elongated into a long rectangle as the motor "waits" for the gates to open!
Notice how this new step now makes a major contribution to the total length of the cycle (accounting for the fact that the G680V motor takes twice as long as the wild type motor to finish a cycle in the presence of actin) and that a major part of the cycle time is now spent in the intermediate strength of actin binding--neatly accounting for all the key observations! Note: I shrank this figure to make it fit; in "real life" this cycle should be twice as wide as the original!
hile this is an entertaining story, note that there is no direct evidence for any of it--it is largely inferential. This is fine for a starting point, and we're proceeding from it. One path we're taking is the genetic route of trying to figure out the problem by letting nature come up with solutions. This work is providing some exciting supporting evidence for the above hypothesis. For a detailed discussion, try the 2 Wrong = Right page. In the meantime, here's key tidbit: changes that fix the G680V motor apparently do so by making the gateway easier to open!