Communications
that BTB-1 might act in a noncompetitive, uncompetitive, or
mixed-competition manner. Detailed analyses of the Michae-
lis–Menten kinetics identified the uncompetitive mode of
inhibition as the best-fitting model, thereby implying that
BTB-1 binds only to the complex formed by Mt-bound
Kif18A. If this applies, then BTB-1 would not be expected to
inhibit the basal, microtubule-independent ATPase activity of
Kif18A. To test this, we used an enzyme-coupled assay to
measure the basal ATPase activity of recombinant His-
Kif18Amotor before the addition of DMSO/BTB-1 (phase I),
after the addition of DMSO/BTB-1 (phase II), or the
stimulated ATPase activity of His-Kif18Amotor upon addition
of Mts in the presence of DMSO or BTB-1 (phase III). To be
able to measure the low, Mt-independent ATPase activity of
Kif18A, we had to use high concentrations of His-Kif18Amotor
and, accordingly, of BTB-1. Before the addition of BTB-1
(Figure 3d, blue line) or DMSO (Figure 3d, red line), His-
Kif18Amotor hydrolyzed ATP at a rate of approximately
0.12 sÀ1 (Figure 3d, phase I). ATP hydrolysis was mediated
by His-Kif18Amotor as no ATP turnover was detectable in
identically treated samples lacking recombinant motor pro-
tein (Figure 3d, violet (DMSO) and black (BTB-1) lines).
Intriguingly, addition of 100 mm BTB-1 did not affect the
ATPase activity of His-Kif18Amotor in the absence of Mts
(Figure 3d and 3 f, phase II: 0.12 sÀ1 (BTB-1) and 0.11 sÀ1
(DMSO)). The transient increase in absorption observed
upon BTB-1 addition was unrelated to Kif18A as the same
effect was observed for the BTB-1 treated sample lacking
Kif18Amotor (Figure 3d, black line, phase II). Importantly,
BTB-1 significantly inhibited the Mt-stimulated ATPase
activity of His-Kif18Amotor (Figure 3d and 3 f, phase III:
0.13 sÀ1 (BTB-1) and 0.3 sÀ1 (DMSO)) corroborating our
model that the inhibitory activity of BTB-1 depends on the
formation of complexes between Kif18A and Mts. Monastrol,
consistent with previous reports,[6] inhibited both the basal
and Mt-stimulated ATPase activity of Eg5 (Figure 3e,g).
Collectively, these data imply that BTB-1 inhibits Kif18A in
an ATP-competitive, Mt-uncompetitive manner.
Figure 2. a) Recombinant His-Kif18AFL was adsorbed to the glass sur-
face and incubated with rhodamine-labeled Mts. Upon ATP hydrolysis,
Kif18A is able to induce Mts gliding. Top: schematic representation of
the assay (red=Mt, blue/black structure=Kif18A, light blue=glass
surface, Pi =phosphate). Bottom: Fluorescence images of Kif18A-
mediated microtubule movement (arrows indicate the microtubule tip
at each time point; scale bar=5 mm). b) Representative kymographs
of Mt gliding assay performed in the presence of DMSO, after flushing
in 100 mm BTB-1, and after wash-out of BTB-1 (see text for details and
Supporting Information, Movies S1–S3). c) Quantification of Mt motil-
ity (n=10 Mts, averages of three independent experiments and
standard deviations are shown). d) Representative kymographs of His-
Kif18AFL-mediated movement of a Mt in the presence of DMSO or
100 mm monastrol (see Supporting Information, Movie S4). e) Quan-
tification of Mt motility (n=10 Mts, averages of three independent
experiments and standard deviations are shown).
Kinesins hydrolyze ATP as they walk along microtubules,
this shows ATP to be a genuine substrate and microtubules as
pseudosubstrates that are not turned over by the enzyme. We
investigated whether BTB-1 competes with ATP for Kif18A
binding. Specifically, we determined the rate of Mt-stimulated
ATP hydrolysis by His-Kif18Amotor in the presence of
saturating concentrations of Mts, varying concentrations of
ATP and BTB-1, and fitted each set of data to the Michaelis–
Menten equation (see Supporting Information). As can be
derived from Figure 3a, BTB-1 increased the Km value for
ATP while not significantly affecting the Vmax value. ATPgS, a
non-hydrolyzable ATP-competitive analogue, similarly
affected Km and Vmax (Figure 3b) implying that BTB-1
inhibits Kif18A in an ATP-competitive manner.
We tested whether BTB-1 affects the mitotic progression
of HeLa cells. RNA-interference (RNAi)-mediated depletion
of Kif18A causes severe defects in spindle morphology
accompanied by decisive failures in chromosome congression
resulting in the accumulation of cells at an early stage of
mitosis.[2,7] Notably, HeLa cells treated with BTB-1 accumu-
lated in mitosis in a dose-dependent manner (Figure 4c).
Immunofluorescence images revealed that spindle structures
were severely compromised in BTB-1 treated cells (Fig-
ure 4b). Yet, elongated spindles observed upon RNAi-
mediated depletion of Kif18A[2] were not detectable in
BTB-1 treated cells. However, the dual-functionality of
Kif18A—it can move along Mts and depolymerize them at
the tips—complicates the interpretation of phenotypes
caused by unequal approaches, that is, removal of Kif18A
from the cellular context by RNAi versus inhibition of its
ATPase activity by BTB-1. Thus, further research efforts are
required to unambiguously evaluate Kif18A as the relevant
target of BTB-1 in cells.
Next, we analyzed how microtubules affect the inhibitory
effect of BTB-1 on Kif18A. To this end, the rate of His-
Kif18Amotor-mediated ATP hydrolysis in the presence of
saturating ATP concentrations and varying concentrations
of Mts and BTB-1 was quantified (Figure 3c). BTB-1 affected
1
both the K = value for Mts (for Mts as pseudosubstrates the
In summary, we report herein the discovery of the first
inhibitor of Kif18A, BTB-1, identified by a protein-based,
2
1
term K = instead of Km is used) and the Vmax value indicating
2
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 9072 –9076