Methods
Expression and purification. A previously described baculovirus construct [1] was used
to produce recombinant human -tubulin with a C-terminal myc-His6 tag. Sf9 cells at
2x106 cells/ml were infected with amounts of the -tubulin baculovirus optimized by
small scale pilot infections. Cells were harvested 48 hrs. post-infection, flash-frozen in
liquid nitrogen, and stored at –80°C. Infections were carried out at the National Cell
Culture Center (Minneapolis, MN).
All protein purification and subsequent work was performed at 4°C. A cell pellet
derived from 1 liter of cells was resuspended in 9 volumes of lysis buffer (50 mM KPO4
pH 8.0, 150 mM KCl, 1 mM MgCl2, 0.25 M GTP, 5 mM ME, Complete EDTA-free
Protease Inhibitor Cocktail Tablet (1 tablet/50 ml) (Roche)), lysed by manual dounce
homogenization, and centrifuged for 10 minutes at 12,000 RPM in an SS34 rotor.
Glycerol, imidazole, and KCl were added directly to the supernatant to final
concentrations of 10%, 5 mM, and 500 mM, respectively. The adjusted lysate was
clarified by centrifugation for 30 minutes at 35,000 RPM in a Type 45 rotor and
subsequently mixed with 0.5 ml Ni-NTA Superflow resin (Qiagen) for 1 hour. Beads
were harvested and washed with 10 volumes of 50 mM KPO4 pH 8.0, 250 mM KCl, 1
mM MgCl2, 10% glycerol, 25 mM imidazole, 0.25 M GTP, 5 mM ME, and 10
volumes of 50 mM K-MES pH 6.6, 500 mM KCl, 5mM MgSO4, 10% glycerol, 25 mM
imidazole, 0.25 M GTP, 5 mM ME. Protein was eluted in 50 mM K-MES pH 6.6, 500
mM KCl, 5 mM MgSO4, 10% glycerol, 250 mM imidazole, 1 M GTP, 5 mM ME.
Fractions containing recombinant -tubulin were pooled and further purified by gel
filtration using a HiLoad 16/60 Superdex 200 column (Amersham Biosciences) preequilibrated with gel filtration buffer (50 mM K-MES pH 6.6, 500 mM KCl, 5 mM
MgSO4, 1 mM K-EGTA, 1 M GTP, 1 mM DTT). Peak fractions were identified by A280
and confirmed by SDS-PAGE. Protein concentration was determined by Bradford assay.
When necessary, protein was concentrated using Ultrafree Biomax microconcentrators
(Millipore).
GTP/GDP binding experiments. GTP binding to -tubulin was measured using dilution
series GTP ranging in concentration from 16 nM to 4.8 M doped with a constant
fraction of [-32P] GTP (400 Ci/mmol; 10mCi/ml). -tubulin was prepared by diluting
2M protein 1:30 in 50 mM MES 6.6, 5 mM MgSO4, 1 mM EGTA (final [KCl]=16 mM
KCl, final [-tubulin]=~66 nM) and aliquoting as before. 4 l of each 32P-doped GTP
dilution was added to 60 l -tubulin. Final [GTP] ranged from 1 nM to 300 nM.
Solutions were mixed and incubated for 1 hour on ice. Bound nucleotide was
crosslinked to -tubulin for 30 seconds of illumination with a UV lamp (312nm handheld
UV lamp fitted with an 8W bulb, Fisher Scientific). Protein was separated from unbound
nucleotide by 12% SDS-PAGE, and 32P labeling was measured using a Phosphorimager
(Storm 840, Molecular Dynamics). Gel band intensities were quantified and backgroundcorrected with ImageQuant software (Molecular Dynamics). All crosslinking
experiments were performed in triplicate. To obtain dissociation constants, a hyperbolic
binding curve was fit to the band intensities.
GDP competition efficiency was determined using a GDP dilution series ranging in
from 8 M to 4800 M. 2 l of each concentration of GDP was added to 60 l -tubulin,
yielding final [GDP] from 0.25 M to 25 M. 2 l 625 nM -32P GTP was then added to
each -tubulin sample (final [GTP]=19.5 nM). Solutions were mixed, incubated for 1
hour on ice, and analyzed as above. To obtain inhibition constants, a competitive
inhibition model was fit to the band intensities. The Ki calculation assumes that GTP and
GDP have similar photocrosslinking efficiencies. The discrepancy between our calculated
GDP/tubulin affinity and the previously reported value [2] may be explained if the phosphate helps to better orient the purine ring for more efficient photocrosslinking or if
the -phosphate becomes crosslinked to His139.
1.
2.
Murphy, S.M., L. Urbani, and T. Stearns, The mammalian -tubulin complex contains
homologues of the yeast spindle pole body components spc97p and spc98p. J Cell Biol,
1998. 141(3): p. 663-74.
Zeeberg, B., Caplow, M., Determination of free and bound microtubular protein and
guanine nucleotide under equilibrium conditions. Biochemistry, 1979. 18: p. 3880-3886.