User:Andy Maloney/Notebook/Lab Notebook of Andy Maloney/2009/04/13/Kinesin & Microtubules

Disclaimer

These are my thoughts on the paper I've read. My thoughts and interpretations may not be right, so make sure to read the paper yourself before evaluating my notes.

Previous info

This is from the other day when I tried to review this article. I'm going to continue with reviewing it because I can see that there is a lot of information in the paper.

Paper

Equilibrium and Transition between Single- and Double-Headed Binding of Kinesin as Revealed by Single-Molecule Mechanics

Review

I'll try again with this paper. My previous attempt at reviewing it was terrible and I said some rather nasty things. I'm pretty sure I was in an all around bad mood that day and I apologize for being so harsh. The paper has generated a lot of questions that I think are good ones. Perhaps more knowledge in this subject will teach me the answers. Or better yet, some of my questions may generate papers themselves.

Introduction

• Tubulin subunits are approximately 8 nm in length. This is why kinesin takes 8 nm steps. It steps between tubulin subunits.
• Each step requires 1 ATP of fuel.
• The length a kinesin will walk (run length) along a microtubule depends on

The ionic strength of the buffer.

Temperature.

The amount of fixing the microtubules.

Question: Does the run length depend on pH? What about if you change the water to heavy water?^{SJK 00:29, 12 April 2009 (EDT)}

00:29, 12 April 2009 (EDT)
For the first question, I would have to guess definitely, since pH would affect the rate constants in the kinetic pathway differently. But I don't know the reference for the data off the top of my head. For the second question, I also guess, yes, and I hope that it is dramatic! It should be a very fun set of experiments to carry out. And I can't point you to data because I haven't found any kinesin activity papers in heavy water (just those few myosin papers (PMID 11779555)). BTW: I more often use "processivity" as a synonym for "run length."

• I still need to commit to memory what ATP hydrolysis is.
• Kinesin likes to walk in the "+" direction along a microtubule.
• They claim that kinesin primarily binds only one head to microtubules if ATP is not present in solution but ADP and AMP-PNP are.
• They claim that kinesin primarily binds both head to microtubules if just AMP-PNP is present and no ATP.

Question: So they are talking about kinesin motility in two different buffers. What is the current model for kinesin motility?^{SJK 00:42, 12 April 2009 (EDT)}

00:42, 12 April 2009 (EDT)
Excellent question. A tremendous amount is known about the kinetic pathway of dimeric kinesin "walking" on MTs. Rate constants for ATP binding, hydrolysis, inorganic phosphate (Pi) release, ADP release, rear head unbinding, front head binding, etc. etc. There are data for all of these. Unfortunately, in the past 10 minutes I couldn't find a good diagram in any of the very good papers I was looking at. It will take us a very long time to learn (or re-learn) what is know about this.

• They claim that the average unbinding force depends on the loading direction, i.e. in the +/- directions of the microtubules.
• They claim that the apparent ratio between single and double head binding depends on the loading rate.
• They claim to be able to resolve when a double head binding turns into a single head binding on microtubules. Just for the AMP-PNP buffer. They can do it by looking at the elastic modulus of kinesin during loading. They claim the transition rate is 1 1/s.
• Complete kinesin detachment from microtubules can occur for both single and double head binding.
• They claim that the double headed binding state of kinesin is predominant under equilibrium in the absence of applying a load for both their buffers.
• They claim that the lower the loading rate, the higher the probability for single headed unbinding.

Materials & Methods

• Their buffers are made up of

2 mM MgCl₂

80 mM PIPES-KOH

1 mM EGTA

pH 6.8

4.5 mg/mL Glucose

0.22 mg/mL Glucose oxidase

0.036 mg/mL Catalase

and they differ by the inclusion of either

Buffer 1	Buffer 2
Apyrase - an ATP scavenger & ATP	AMP-PNP and no ATP

• I thought personally that it was cool that you can purchase microtubules that are fluorescently labeled differently by their "+" or "-" ends. Not sure how you would go about doing this though.
• Question: Why would they use EGTA to chelate out Mg when they use MgCl₂ as their primary source of ions?
  • Steve Koch 00:13, 14 April 2009 (EDT): The presence of EGTA is a weird part of kinesin motility buffers. I learned about this a few years ago, but I'm hazy now. It has to do with differential affinity for Mg++ versus Ca++. I think EGTA has a much higher affinity for Ca++ than for Mg++. Ca++ is destabalizing to MTs and a very bad thing to have around. So, that's the rationale, I think. But I see what you're saying, since EGTA can chelate two Mg++ ions (right?). Maybe they are not mentioning that they also have Mg++ in their ATP or AMP-PNP buffers?
• They note that kinesin is unstable in ATP free buffers with no microtubules present.

Question: I don't understand why then they call their first buffer "nucleotide free". If kinesin is unstable in ATP free buffers, then how are they able to do the experiments?

Question: Does Amylase take a long time to scavenge all the ATP? If it does then there is ATP in solution when they were doing these experiments.

Question: Does Amylase act very quickly to get rid of ATP? If it does, then that means kinesin has residual energy stores available so it can continue to stay bound to the microtubules.

Question: If kinesin does not have energy stores to stay on the microtubules, then this means that kinesin wants to bind to the microtubules and it is the ATP that helps remove a kinesin head from the microtubules so it can take a step. Which way is it? Is it the ATP that "glues" kinesin to microtubules or is it the ATP that removes kinesin from the microtubules?

Question: Why would kinesin be stable when it is attached to a microtubule when there is no ATP, but isn't when there are no microtubules and no ATP? Is it stable when there is ATP and no microtubules?

Experiment

• This is what they did.

Attached kinesin to microspheres in a 1:1 ratio.

Attached and fixed microtubules to coverslips with Taxol and DTT.

Coated exposed glass with Casein.

Introduced kinesin microspheres to the microtubules.

Grab a microsphere with an optical tweezers setup.

Bring the microsphere close to a microtubule and hold it there for 30 seconds. Note: I'm skeptical about this but, it seems that they were taking data with what they thought was an attached kinesin molecule. Personally I would have put a lot more microspheres into solution and just let nature do its thing. Once I saw a microsphere moving along a microtubule, I would have then grabbed a hold of it with the OT. That way, there is no ambiguity about the kinesin being attached to the microtubule.

They moved the microsphere along the microtubule and measured the deflection of the light from the microsphere. Measuring deflection implies that a load or force is being exerted on the kinesin/microtubule system.

They measured unbinding forces in the above manner for both buffers used.

• So many questions...

Question 1: They claim that the kinesin/microsphere system can stay in the trap center for quite a long run without any deflection of the laser beam, i.e. no load on the system. They attribute this to the fact that a kinesin molecule can rotate. Since position detectors do not measure rotation, they were unable to detect rotation and thus they assume the kinesin molecule is attached to the microtubule. What if the kinesin molecule was never attached and then suddenly attached while they moved it down a microtubule? ^{SJK 13:09, 13 April 2009 (EDT)}

13:09, 13 April 2009 (EDT)
There are so many good questions here that I'm overwhelmed! Instead of trying to keep up, I'm going to put in notes when I think of them. (1) You will love the "limping kinesin" stuff out of the Block Lab. Do a Pub Med search for Fehr Asbury Block, should produce two papers. I'd read the Science paper first (PMID 14657506 ). It addresses your questions about the ability of kinesin to deal with the torsional strain--they make the coiled-coil shorter and shorter and see more and more limping. (2) Bead rotation: Check out the "optical torque wrench" paper by Arthur LaPorta (he was a postdoc when I was a grad student). PMID 15169392 . You'll love his optics setup as well as seeing that you are onto something with bead rotation.

Question 2: Can you measure rotation of the bead? I'm assuming that if a bead rotates in an electric field, the field will acquire a phase. You can detect phase changes pretty easily. Maybe we should look into this.

• Amylase buffer.
  • They saw that in the Amylase buffer, kinesin only either detached or attached once during the run along the microtubule.
  • They saw a shift in the data taken for loading rates. For small loading rates, kinesin detachment data occurred in a "unimodal" manner. However, when they increased the loading rates, detachment occurred in a "bimodal" manner. Question: This is crazy! I can't believe it! Why would it do something like this? What ever it is, it is super cool and I have no clue as to why there is a shift. There has to be some sort of change going on for the data to be represented by two distinct curves.
  • Their data shows that unbinding forces are about 45% greater for detachment when loading was done in the "-" direction of microtubules. Note: Wow! Again wow! This means that either conformationally or electronically, the kinesin molecules (if attached to microtubules) will stay on them and only come off in one direction. Biology kicks ass if this is true. Question: Why would this make sense evolutionarily speaking?
• AMP-PNP buffer
  • In the AMP-PNP buffer, kinesin detached and reattached several times.
  • They saw that in their force curves, the initial detachment of kinesin took twice as much force to detach than subsequent detachments.
  • Reattachment of kinesin after detachment took about 2 - 3 seconds.
  • Again they saw a shift in their detachment data from "bimodal" to "unimodal"
• I'm not convinced how they measured the elastic modulus of kinesin. I just don't know enough to comment about it but they say that they throw away 60% of their data to interpret an elastic modulus. They even say that the elastic modulus changes for their two different buffers. Not sure why this would be the case.

Discussion

• Wow! People think that there are two types of binding for kinesin heads; weak and strong. That's just weird.
  • Steve Koch 00:09, 14 April 2009 (EDT): This could be the weak ionic association I was talking with you about today (the "E-loop")...this may be a good parallel with site-specific DNA-binding proteins (such as EcoRI), and the difference between tight site-specific binding mode, and loose non-specific ionic association with the DNA backbone.
  • Steve Koch 00:09, 14 April 2009 (EDT): Whoah: something just occurred to me. It's quite possible that the MTs will blow to smithereens in high osmolyte concentration. That will suck! Or it's also possible they'll be more stable. That will not suck. In any case, we'll probably want to have someone study the stability effects of osmotic pressure--maybe a good undergrad project?
• They go through their model analysis in this section.

My final analysis

I can't say I agree with their results. Here's why I don't agree.

First, in their AMP-PNP buffer, they were able to move the kinesin molecules without any load exerted on them. Since AMP-PNP is similar to ATP, it will bind to the kinesin molecules like an ATP will. However, AMP-PNP cannot allow the kinesin to make a power stroke. This means that the kinesin molecules, if attached, will not walk along the microtubules. Their data shows that they obtained kinesin "walking" along the microtubules with no load being exerted on them from the OT.

AMP-PNP data is unusually similar to their data from the Amylase buffer. In the Amylase buffer, I can be inclined to believe that their kinesin can walk along a microtubule because the buffer contains residual ATP molecules. However, since kinesin cannot make power strokes with just AMP-PNP in solution, one should not get the characteristic curves they got. They should have seen in the AMP-PNP buffer, a continual load being placed on the kinesin because it should have been dragged along the microtubules if it was attached.

The similarity between the data sets for both buffers inclines me to believe that they never had kinesin attached to microtubules. What I think happens is that their kinesin never is strictly attached to microtubules and at some point in time, their trap is brought close enough for the kinesin to attach to the microtubules and get stuck to it. This is why the kinesin stalls and they can get force measurements.

Their force measurements are good. I just disagree with their interpretation of their results and they are not convincing me with their discussion of their results. I think I should revisit this paper later when I have more knowledge of the subject to see if my mind changes.