Good afternoon everyone. I will begin by presenting to you my proposed research project which should hopefully take no more than 15-20 minutes.
As we all know, cells divide. Beginning with Walther Flemming’s pioneering studies on mitosis and all throughout the past century biologists have continually endeavored to understand this ever complex process. As you may have inferred simply by looking at this illustration, cell division requires quite a bit of both spatial and temporal regulation. And considering how important it is that the genome be duplicated and segregated into two daughter cells without error the cell has evolved many regulatory mechanisms that spatially and temporally control progress through mitosis. I am interested in studying how one of these regulators, the kinase Aurora-B, achieves proper control over cell division…
Aurora-B localizes at different locations throughout the different phases of mitosis. Most notably, Aurora-B, shown here in green, localizes to the inner centromere throughout metaphase where it mainly functions in correcting improper kinetochore-MT attachments. At the onset of anaphase, Aurora-B re-locates to the spindle midzone or central spindle where it plays a role in midzone assembly and organization in addition to regulating abscission timing to complete cytokinesis. Aurora-B is part of the evolutionarily conserved chromosome passenger complex, consisting of surivivin, borealin, and the inner centromere protein (INCENP). Mammalian and xenopus CPCs are thought to have an additional member, telophase-disc 60 or TD-60. The initial step in activating Aurora-B is the binding of INCENP’s C-terminal hydrophobic motif known as the INbox to a hydrophobic pocket in the kinase’s N-lobe. Then Aurora-B must engage in auto-phosphorylation in trans to become fully active. This is known as auto-activation and simply means that to be fully active the Aurora-B’s have to phosphorylate each other after the initial activation by INCENP.
The CPC has been shown to directly bind MTs both in vivo and in vitro. And recently the coiled-coil domain of INCENP was shown to be the CPC’s MT targeting and binding domain.I think that this microtubule binding has a purpose. I want to know how does the CPC bind to MTs and what is the functional significance of the CPC’s interaction with MTs?
The CPC’s interaction with MTs has been proposed to be the method through which a spatial gradient of Aurora-B activity is established and maintained at the spindle midzone. What is likely taking place is the following: The CPC transfers to the spindle midzone during anaphase. There aurora-B’s activity stimulates MT assembly. Then in-turn MTs stimulate or accelerate aurora-B auto-activation. So, through this positive feedback loop a gradient of Aurora-B activity is established and maintained at the midzone.
This brings us back to our first question: How does Aurora-B interact with the MT lattice? To answer this question…(next slide)
I have used the single-molecule TIRF assay illustrated here with a minimal construct of the CPC consisting of INCENP’s coiled-coil domain and INbox bound to Aurora-B which has a GFP tag at it’s C-terminus. This construct will be referred to as CCA for convenience sake. Briefly, fluorescently labeled MTs are immobilized to a glass surface using antibodies. The surrounding surface is coated with blocking co-polymers to block non-specific adsorption. Then my fluorescently labeled construct is introduced and single molecules are visualized. This is a great technique because of its adequate signal-to-background we are able to look at single gfp-tagged molecules as illustrated in this figure compared to epi-fluorescence.
My initial TIRF experiment revealed that the CCA-GFP construct binds the MT lattice transiently and undergoes what appears to be an unbiased random walk. My preliminary rough measurements estimate the diffusion coefficient between 0.2 and 0.5 um2/s and a mean residence time on the lattice between 0.8 and 0.9s before correcting for photobleaching.
As a result, my hypothesis is that the transient transition from 3D to 1D CPC diffusion after binding the MT lattice is the means by which MTs serve to accelerate Aurora-B’s kinase activity.
As you can imagine my hypothesis will depend on the diffusion coefficient and the off-rate constant of my construct. So, to get these I will build trajectories that quantify the CCA-GFP’s random walk using particle tracking software. Briefly, I will track a single-molecule in time across each frame until it disappears. Then I’ll plot the average of the squared root mean displacements of multiple molecules as a function of time and the slope of the resulting linear fit will give me my diffusion coefficient. Then I’ll plot the times each molecule stays on the lattice and the resulting exponential fit will give me the mean residence time spent on the MT lattice the inverse of which will give me my off-rate constant. I should mention that the mean residence time derived here will have to be corrected for photobleaching before determining the off-rate constant.
Now there’s a second fundamental question to be answered. Are MTs increasing the rate of Aurora-B’s substrate targeting and phosphorylation? This is a particularly important question when considering Aurora’B’s MT-bound substrates at the spindle midzone some of which require the CPC for stable MT localization.
To study the rate of substrate phosphorylation I want to do in vitro kinase assays with a twist. Instead of using radioactive ATP and phospho-imaging, I want to use this FRET sensor in fluorimeter-based in vitro kinase assays. As is illustrated, loss of FRET will denote Aurora-B phosphorylation of the sensor. By adding different localization sequences I can have a sensor that mimics a stationary bound MT substrate, a substrate diffusing along the MT lattice and a soluble substrate. I believe this will be a more quantifiable way of assessing the rate at which Aurora-B is phosphorylating its various substrates, especially at the midzone.
In the absence of MTs I expect to see a significant auto-activation lag and perhaps below certain kinase concentrations little to no activity. In the presence of MTs I expect to see a diminished or loss of auto-activation lag. And with a microtubule bound substrate I expect to see an acceleration in the rate of phosphorylation. So, for a MT-bound substrate I’d expect to see a loss or decrease in auto-activation lag and an increased rate of substrate phosphorylation. And for a soluble substrate in the presence of MTs I’d expect to see a loss or decreased auto-activation lag but with slower rates of substrate phosphorylation.
I believe that by 1D lattice diffusion MTs increase the local concentration of CPCs which in turn increases the rate of auto-activation. To rule out MT induced allosteric shifts in the CPC being the cause of this I will replace INCENP’s coiled-coil domain with KIF1A’s diffusive MT-binding domain, the K-loop. If indeed increased auto-activation is due to this 1D lattice diffusion then I should see similar kinase kinetics when using this construct of incenp with Aurora-B in my assays.Kinesin 3 family Monomeric kinesin with positively charged k-loop that diffuses along MT-lattice by electrostatically interacting with negatively charged E-hooks.
As a negative control I will replace INCENP’s coiled-coil domain with TAU3’s MT-binding domain which should bind the MT lattice in a stationary manner.
Roles suggested for TD-60: key role in integrating the kinetochore into the mitotic spindle and a global role in mitotic spindle formation. I’d like to know the nature of TD-60’s interaction with MTs and exactly how TD-60 is enhancing Aurora-B’s activity?
Ph.D. Qualifying Exam Presentation (McGill University, Department of Biology))
Single-Molecule Studies of Aurora-B Kinase and the Chromosome Passenger Complex Michael G. Noujaim Ph.D. Qualifying Exam January 20th, 2011
Aurora-B Kinase is Part of the Chromosomal Passenger Complex (CPC)TD-60 Ruchaud, S. et al. Nature Reviews Molecular Cell Biology, 2007. 8: p. 798-812.
The CPC Binds to Microtubules Ruchaud, S. et al. Nature Reviews Molecular Cell Biology, 2007. 8: p. 798-812. Kang et al. JCB 2001. 155(5): p. 763-774.
Microtubules Activate Aurora-B S E Rosasco-Nitcher et al. Science 2008;319:469-472
Active Aurora-B Creates a Spatial Phosphorylation Gradient Auto-activation Midzone MTs Aurora-B MT assembly Positive Feedback Loop Fuller, B. G. et al. Nature 2008. 453: p. 1132-1137.
How does Aurora-B interact withthe microtubule lattice?
Single-Molecule TIRF 2 μm TIRF Epi-fluorescence In real-time Gell et al., 2010
CCA-GFP transiently binds to microtubules and undergoes 1s an unbiased random walk Time (s) 1.3 μm In real-time 0.8 μm Position (μm)
A “reduction in the dimensionality” of theCPC’s diffusion accelerates Aurora-B’s kinaseactivity. Target Target >> 1D 3D
Characterization of CCA-GFP Lattice Diffusion Displacement squared <x2> (μm2) Events e(-t/<τ>) => koff= 1/τ Slope = 2D (μm2/s) Time (s) Time (s)
Are microtubules increasing the rate ofAurora-B’s auto-activation, substratetargeting and phosphorylation?
Rate of Substrate Phosphorylation Fuller, B. G. et al. Nature 2008. 453: p. 1132-1137.
Rate of Substrate Phosphorylation CFP:YFP Slope = phosphorylation CFP:YFP rate Auto-activation lag Increase in slope Decreased phos. rate Time (min)CFP:YFP CFP:YFP Decrease in auto-activation lag Decrease in auto-activation lag Time (min) Increase in slope Loss of auto-activation lag Time (min) Time (min)
Reduction in Dimensionality or Allostery? INCENP 1 Coiled-Coil Domain 873 INbox TSS INCENPK-loop 1 873 INbox TSS Motor K-Loop1 1690 K-Loop KIF1A
Reduction in Dimensionality or Allostery? INCENP1 Coiled-Coil Domain INCENPTau-MT 873 INbox TSS 1 873 Tau MT-binding INbox TSS 1 352 MT-binding Tau
How Does TD-60 Enhance Aurora-B’s Activity? TD-60 Aurora-B Microtubule