Biology Reference
In-Depth Information
(
Fig. 21.1
A, labeled). Changing the pf number can change the helical pitch of the
microtubule. 15-pf microtubules, for example, have a 4-start helix and thus no seam
(
Sui & Downing, 2010
).
For 13-pf microtubules, the pfs run parallel to the long axis of the microtubule
(see
Fig. 21.1
A and C). Adding or removing pfs, however, introduces a supertwist
(
Fig. 21.1
D).
Chretien and Wade (1991)
developed a theoretical explanation of the
supertwist, the “surface lattice accommodation model,” shown in
Fig. 21.1
C and D,
which was confirmed by EM. This supertwist creates a moir´ pattern specific to each
microtubule type in EM (
Langford, 1980
), which is the basis for most EM measure-
ments of pf number. Although the pf number and supertwist of a microtubule is pre-
sumably established during the nucleation process, the pf number may change along
the length of a microtubule during growth (
Chretien, Metoz, Verde, Karsenti, &
Wade, 1992
), thereby introducing a “lattice defect.” Collectively, “microtubule ar-
chitecture” is the combination of lateral curvature, longitudinal curvature, pitch,
supertwist, and defects that describe an individual microtubule lattice.
Importantly, MAPs are sensitive to microtubule architecture, and thus different
microtubule types can be used to reveal the intrinsic properties of MAPs. For exam-
ple, the supertwist of non-13-pf microtubules was used to demonstrate that kinesin-1
travels on the path of single pfs, due to the fact that surface-immobilized kinesins
caused non-13-pf microtubules to rotate (
Ray, Meyhofer, Milligan, & Howard,
1993
). The rotation of kinesin-1 around a supertwisted microtubule lattice allows
for an indirect measurement of pf number by TIRF microscopy in which the 3D path
of a quantum dot-labeled kinesin is tracked (
Nitzsche, Ruhnow, & Diez, 2008
). It
was suggested that end-binding proteins bind specifically to the microtubule seam
(
Sandblad et al., 2006
), although further experiments have cast doubt on this model
(
Maurer, Fourniol, Bohner, Moores, & Surrey, 2012
) (see the
chapter 23
in this
volume). Finally, lattice defects are thought to be the preferred site of interaction
for severing enzymes with microtubules (
Davis, Odde, Block, & Gross, 2002;
Diaz-Valencia et al., 2011
). It has not been possible, however, to visualize pf
transitions and severing enzymes concurrently.
What is needed is a suite of assays that visualize microtubule architecture, micro-
tubule dynamics, and MAP activity concurrently by fluorescence. Investigating mi-
crotubule architecture by fluorescence will also allow us to gain insight into the
nature of spontaneous nucleation and microtubule polymerization. After all, why
do microtubules not form a more perfect lattice? We do not understand, for example,
the mechanistic basis for pf number distributions. The pf number distributions mea-
sured in the literature vary widely; in our hands, day-to-day variability is common-
place. What is needed is an assay through which the basis of pf number variability
could be investigated.
In this chapter, we describe methods we have developed for assaying microtubule
architecture by fluorescence microscopy. First, we describe how to use controlled
growth conditions to generate microtubules of predetermined architecture. These
“control” microtubules can be used to determine whether a MAP binds preferentially
to one type or another. Second, we describe how to use GFP-labeled doublecortin
