Smoke & Mirrors
|Nature of Knowledge: Intro|
irtually all the knowledge accumulated in the fields of Molecular Biology and Genetics (and a ton of other fields, for that matter) is built up from inferences. Very little of what we "know" is based on direct observation. In order to give you a feel for this and to better capture the challenge of trying to learn about something too small to see and too fast to "catch in the act," this section will introduce you to the techniques that underlie our amazingly detailed (and simultaneously, vastly incomplete!) knowledge of the structure and behaviors of the myosin motor. FYI, the mutant hunting section details how we use macroscopic behavior of Dictyostlium colonies to identify cells with perturbed or refurbished myosins.
Myosin structures: How do we know what they look like?
ominating this site are "pictures" of the myosin motor, frozen in one or two states. How is it that Ivan Rayment's group has been able to deduce the structure of the motor? How did they get it to "sit still" long enough to have it's picture taken? Frankly, I don't fully understand the process, because it backs all the way up into the wave nature of light, which is something I'm not sure I believe in. But here's the basics as understood by me: First, those structures are based on not one myosin molecule, but millions or billions examined simultaneously. For the tricks they use to work, these millions of molecules must all be sitting in exactly the same position, arrayed in a 3-dimensional grid, or crystal. And they must all be in the same "state" or shape! How are they persuaded to do this? By some very rude treatment indeed! Getting them all to do the same thing: the motors are deprived of fuel, or fed fuel molecules that they cannot consume. By doing this, a large populaion of motors are generated that are all "stuck" at the same point in their cycles. Next, the stuck motors, along with salts or other chemicals are slowly dehydrated. As they run out of "room" in the solution, they are forced to pack very tightly to each other. With a lot of trial and error and a little magic, conditions will be found in which the molecules "discover" a uniform, repeating way of packing next to each other--the process of crystallization. This is the same thing that is happening when you make rock candy by letting the water evaporate from a sugar solution, except that it is much easier to make a crystal of sugar, with its 24 atoms, than myosin, with its many thousands of atoms.
uppose all this went swimmingly well, and a comparatively vast array of identical myosins in an orderly crystal results. Now what? Unfortunately, here comes the wave nature of light. An ordered beam of light passing through a matrix of identical particles will exit in such a way as to form "spots" on a film or wall on the other side of the crystal. This phenomenon results from the ability of waves to interfere with one another. The uniform spacing of the objects in the beam's path results in a pattern of interference and enhancement of the waves making up the beam. I have seen this demonstrated, so I can vouch for the truth of it, but I shall have to insert a link to another site for a compelling explanation. Suffice it to say that the position and brightness of the resulting spots contains information about the matrix through which the light has passed. Since the light has been obeying the Laws of Physics all along, the mathematically inclined can deduce certain things about what the light must have passed through. Given a perfect enough crystal, along with highly similar crystals in which specific alterations of the molecule in question are introduced (such as by putting a very heavy atom into a specific location in every molecule), and a clever enough crystallographer, under ideal circumstances enough can be deduced about the molecules making up the crystal that an image, such as the ones shown at this site, can be constructed.
t is important to consider the limitations of this technique! First, only molecules that are willing to pack into a crystal can be examined in this way. For technical reasons, actin cannot currently be included in myosin crystals, so all states where the motor is engaged in moving actin or even holding onto it are inaccessible. Second, the technique requires us to arrest motors, so we cannot watch them cycle. Third, the motors must all be frozen into a single state--so any state that we cannot currently force the motor into is inaccessible. And fourth, the motors are forced to form crystals under conditions unlike those they normally work under--so their final shape could differ significantly from their "real" one. Nonetheless, this is a hugely important technique and has provided us with our best look at the motor at rest!
Checking under the hood: how we acquire myosin motors
n the DataDungeon section, I've shown the data we have taken on how well the motor is doing certain tasks. How can we tell it's doing anything if we can't directly observe it? The first step is to get a pure "bottle of motors"--a test tube containing only myosin. These days, this is not so hard. Essentially, we take advantage of certain properties of myosin motors to selectively recover them from the vat of molecules that is a Dicty cell. Briefly, we first selectively capture molecules that are attached to the "skin" of cell itself, discarding all the stuff floating around on the inside. Since myosin is mostly stuck to the skin, we retain most of it while discarding lots of irrelevant stuff. Then we add ATP, which makes the motors periodically "let go" of actin, keep the loose stuff, and again discard the rest.
e next use a trick to add another property to myosin motors. We've genetically engineered six consecutive histidine residues (residue = amino acid after it's incorporated into proteins) onto the tail end of the motors we work with. This arrangement has a very special property: histidines contain nitrogen atoms that can partially "satisfy" a nickel atom's desire for partners; if there are enough histidines close together, several of them can combine to completely satisfy a nickel atom, and a tight association is formed. The relationship can be disrupted, however, by "tempting" the nickel with a very large number of free histidine molecules. By setting our conditions carefully, we can selectively recover molecules that stick to nickel under one condition, and release them by 'competing' with excess free histidine. By passing material from opened up Dicty cells through these four "sieves"--cell attachment, ATP-induced release, nickel attachment and histidine competition--we end up with a pretty pure bottleful of myosin motors!
Measuring fuel use: ATPases
The Idling motor: 'basal' ATPase
aving large numbers of motors allows us to detect motor action in ways that aren't sensitive enough to detect the workings of a single motor! The first assay we perform on motors is to determine their rate of ATP burning. Motors are "fed" ATP molecules and given the chance to use them for a certain period of time. Then the motors are destroyed, and the amount of phosphate liberated by myosin action is measured (ATP is broken into ADP and Pi (free phosphate) by myosin). We achieve this by another trick: Pi has the unique property that it can form a complex with a dye named Malachite Green. ATP and ADP cannot do this even though they contain bound Pi. When Pi complexes with Malachite Green, the properties of the dye are changed, altering its color. By determining the color of a solution of Malachite Green and myosin-treated ATP, we can deduce the amount of motor activity that went on. By a conceptually similar method, we can infer how much myosin was in the bottle, so we can backtrack and figure out how much damage each molecule did. This, essentially, is the basis for the "basal" ATPase that we report--the rate myosin chews up ATP in the absence of actin.
Revving the engine: 'actin-activated' ATPase
he point of the myosin motor is not to run quietly, but to rev up and move actin filaments when it's told to. We measure this behavior in a way very similar to the basal ATPase, but of course, we provide the motor with actin. The actin is acquired through a series of steps that focus on its unique properties, much like we took advantage of myosin's properties above. We then add our purified actin to our bottle of motors, toss in some ATP, and measure Pi production as above! Having both numbers, the 'basal' and 'actin-activated' ATPase, lets us determine how much actin accelerates the working cycle of the motor. Examples of this type of data can be found in the DataDungeon.
Holding on & letting go: actin binding assays
he other property of myosin that we examine extensively is its ability to interact with actin. As you may recall, in the course of motor events, myosin spends much of its time getting ready, and only a small fraction of its time holding and moving actin. We can assess how strong it is holding on to actin, and how much of its time is spent doing so, in the following way.
he critical aspect of the assay is that one of actin's forms is as a filament--a hugely long stack of individual actin molecules that join together end to end. This array is so large that it will sink to the bottom of a tube of water, albeit very slowly. We accelerate this process by putting the tube at the edge of a very rapidly spinning disk. When we spin the tube at rates of 100,000 revolutions per minute, the forces within it quickly bring the giant actin filaments to the side of the tube. So far, so what? Here's the kicker: a single motor is nowhere near as large as an actin filament, so under the conditions described above, motors will not end up on the side of the tube. That is--unless the motor attaches to actin! Herein lies the essence of our assay. By putting the myosin motors in the same tube as the actin and spinning them very rapidly, we can separate the motors into two groups: those that were attached to the actin when we started spinning, and those that were not. Attached motors will be dragged to the side of the tube by the actin; detached ones remain floating. In the "experiment" shown here, the red motors are spending most of their time attached to filaments, the green ones are mostly detached. When we spin down the (brown) actin filaments, the red motors come with them, while the green ones remain floating in the solution. We then pour off the liquid and harvest any proteins that are on the side of the tube and assess the amount of myosin there.
s a final trick, we can perform this assay with other components present as well. For example, we can add large amounts of ATP. In the presence of ATP and actin, the myosin motors will cycle--breaking ATP, releasing Pi/binding to actin, stroking, releasing ADP, binding ATP and letting go of actin again. If we perform our hard spin while this action is occurring, we can tell what fraction of myosin motors are attached to actin at a given moment in time! Obviously, this experiment has an infinite number of variations, several of which we actually perform. We assess the fraction of motors bound in the presence of ATP, ADP, and ATPgammaS (a mimic of ATP that myosin can bind but cannot break). Also, we evaluate how many motors remain bound if we weaken their grasp on actin by making the solution salty. These techniques give us insights into how the motor is performing at different points in its cycle. This concept is developed more fully in the One Bad Motor: the G680V mutant section; other examples of this type of data can be found in the DataDungeon.
Squeezing & stroking: motors at work
ll the assays described above dissect out a specific motor function. We also have a way of checking the motor's ability to do work. Specifically, we measure the ability of the motor to change the shape of a cell. As with the other assays described above, we don't actually look at a single cell. Instead, we follow the ability of millions of cells to scatter a beam of light, using a protocol developed by Ed Kuczmarski's lab. Briefly, we first prepare cells by using a mild detergent to 'poke holes' in them, allowing all their innards to ooze out. The shell or 'ghost' that is left behind contains, among other things, a meshwork of actin and myosin. When we add ATP to these 'ghosts', the motors activate and tug on the actins, causing the 'ghosts' to shrink. As the ghosts become smaller, they become more dense, and thus less transparent to light. We can detect the decrease in light passing through the bottle of ghosts. By measuring the rate of change in light passage and the extent of change, we can assess the speed of motor function, and perhaps get an idea of the amount of force the motors are exerting.