Protein phosphatase inhibitory activity of tautomycin photoaffinity probes evaluated at femto-molar level

Article abstract of DOI:10.1016/j.bmc.2007.11.034

Herein we describe the further improvement of our in-house developed firefly bioluminescence assay system for the determination of inhibition of protein phosphatase (PP). The advantage with the new system is higher sensitivity as well as being time and sample efficient. The inhibition activity of tautomycin with PP1γ was determined using the upgraded test system and K_i was found to be 4.5 nM, which compare favorably with the activity reported previously by others using different methods. The test system was then used in order to determine the activity of nine tautomycin (TTM) photoaffinity probes. One of the TTM photoaffinity probes (anti-10) was found to possess higher activity than the natural product itself with a K_i of 3.4 nM, while the remaining photoaffinity probes were found to possess K_i in the range of 8.0-213 nM.

Full text of DOI:10.1016/j.bmc.2007.11.034

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 1747–1755
Protein phosphatase inhibitory activity of tautomycin
photoaffinity probes evaluated at femto-molar level
Magne O. Sydnes,a Masaki Kuse,b Masakuni Kurono,a Aya Shimomura,a
Hiroshi Ohinata,c Akira Takaic and Minoru Isobea,d,
*
aLaboratory of Organic Chemistry, Graduate School of Bioagricultural Sciences, Nagoya University,
Furo-cho, Chikusa, Nagoya 464-8601, Japan
bChemical Instrument Division, RCMS, Nagoya University, Furo-cho, Chikusa, Nagoya 464-8602, Japan
cAsahikawa Medical College, Midorigaoka Higashi 2-1-1-1, Asahikawa 078-8510, Japan
dInstitute of Advanced research, Nagoya University, Furo-cho, Chikusa, Nagoya 464-8601, Japan
Received 12 October 2007; revised 6 November 2007; accepted 6 November 2007
Available online 17 November 2007
Abstract—Herein we describe the further improvement of our in-house developed firefly bioluminescence assay system for the deter-
mination of inhibition of protein phosphatase (PP). The advantage with the new system is higher sensitivity as well as being time and
sample efficient. The inhibition activity of tautomycin with PP1c was determined using the upgraded test system and Ki was found to
be 4.5 nM, which compare favorably with the activity reported previously by others using different methods. The test system was
then used in order to determine the activity of nine tautomycin (TTM) photoaffinity probes. One of the TTM photoaffinity probes
(anti-10) was found to possess higher activity than the natural product itself with a Ki of 3.4 nM, while the remaining photoaffinity
probes were found to possess Ki in the range of 8.0–213 nM.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
(CH3CN, 3% NaHCO3, pH 8) TTM (1) predominantly
existsas the diacid (TTMDA, 1b).10 However, underalka-
line conditions (MeOH, 20% CsCO3, pH 9) TTM (1) is
unstable and undergoes trans-esterification followed by
b-elimination resulting in the formation of acid 2 and
alcohol 3 (Scheme 1),2,10 neither of which shows any
inhibitory activity toward PP1 or PP2A.11 It was found
that the active inhibitor is the dicarboxylic acid form of
TTM (TTMDA, 1b),10,12 and indeed, the anhydride form
(1a) possesses no inhibitory activity toward PP1 or PP2A.
The hydroxyl groups at C22 and C30 are also essential for
biological activity9,12 as well as the hydrophobic spirok-
etal moiety (see red boxes in Scheme 1).9
Tautomycin (TTM, 1, Scheme 1) was isolated in 1987 by
Isono and co-workers from Streptomyces spiroverticilla-
tus found in a soil sample collected in Jiangsu Province,
China.1 Three years later the fully assigned structure of
TTM was reported to be as depicted in Scheme 1.2 Tauto-
mycin was found to have strong activity against Scleroti-
nia sclerotiorum1 as well as induce morphological
exchange in human leukemia cells K562.3 However, more
interesting is TTM’s ability to inhibit type 14 and type 2A5
protein phosphatase (PP1 and PP2A),6,7 and in particular
the fact that it inhibits PP1 more selectively than PP2A.8
The dissociation constant for the PP1-TTM interaction
compared to that of the PP2A-TTM interaction (viz.
PP1/PP2A ratio) is in the range of 0.01–0.03.
Based on computer calculations using Biograf and
NMRgraf package programs utilizing NOESY data,
we found that one of the stable conformations of
TTMDA (1b) is U-shaped in D2O and has a flexible
structure from carbon C20 to C70.10 Recently this folded
conformation was proven through special fluorescence-
quenching experiments of the excited states of two
chromophores attached at each end of TTM (1a).13
However, when TTMDA (1b) is bound to the active site
of PP1c, it is expected that it will change conformation
to the active one. A possible conformation of compound
As indicated in Scheme 1, TTM exists in a pH dependent
equilibrium between the dicarboxylic acid form1b and the
anhydride form 1a. Under mild alkaline conditions
Keywords: Tautomycin; Photoaffinity probe; Protein phosphatase 1;
Firefly bioluminescence assay; Luciferine phosphate.
*
Corresponding author. Tel.: +81 52 789 4109; fax: +81 52 789
4111; e-mail: isobem@agr.nagoya-u.ac.jp
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.11.034

1748
M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
OH
3'
O
1'
O
O
OH
3'
O
1'
OH
22
O
OH
18
H
1
O
O
10
14
3
6
> pH 7
< pH 3
24
20
O
HO
HO
O
O
O
H
O
OMe
5'
5'
O
Tautomycin (TTM, 1a)
Tautomycin diacid (TTMDA, 1b)
Not active
Inhibitor
> pH 9
OH
3'
O
O
O
OH
18
H
1
O
10
1'
14
OH
3
+
6
O
5'
24
22
20
HO
O
H
O
OMe
O
2
3
Scheme 1. Tautomycin (TTM, 1a) and tautomycin diacid (TTMDA, 1b). The red boxes indicate functional groups that are important for the
biological activity.
1b complexed with PP1c would be the one depicted in
Fig. 1. This model was generated so that each functional
group that is important for activity can have maximum
interaction with one of the amino acid residues Arg96,
Tyr134, Arg221, and Tyr272.14 In particular interaction
with Tyr272 seems to be important for activity. This fact
was highlighted by Lee et al. who showed that a muta-
tion in the protein at this amino acid lowered the activity
of tautomycin and calyculin A (another natural product
with activity toward PP1) by three orders of magnitude
compared with the natural protein.15
tein phosphatase 1c is one of totally three isoforms of
the PP1 catalytic subunit in mammalian cells with
PP1a and PP1b (also called PP1d) being the remaining
two isoforms.18
The principle for the test system is outlined in Scheme 2. In
the presence of PP1c, luciferin phosphate (4) is dephos-
phorylated to give luciferin (5), which reacts with lucifer-
ase, adenosine triphosphate sodium salt (ATP), and Mg2+
to yield oxyluciferin (6) and light. The light emitted is de-
tected by a single photon counter and converted to a dig-
ital signal, which can be integrated. However, with an
inhibitor present the conversion of luciferin phosphate
(4) to luciferin (5) will be hampered and the process even-
tually stops. As also shown in the Scheme, luciferin phos-
phate (4) does not react with luciferase, ATP, and Mg2+ to
give light, eliminating the possibility of a false result.
Previously we have reported that the in-house developed
firefly bioluminescence assay system is excellent for
determining the inhibition activity of PP2A16 and
PP1c17 by tautomycin and other inhibitors of PP, such
as okadaic acid, calyculin A, and microcystin-LR. Pro-
Figure 1. Proposed structure of PP1c complexed with TTMDA (1b) (TTMDA was placed in the active site of PP1c instead of calyculin A). The
model is generated and minimized using MacroModel 9.1 (force field MMFFs and OPLS2001, using water as solvent).

M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
1749
O
P
luciferase,
ATP, Mg2+
Solution A Solution B
PP1 - inhibitor complex
CO2H
HO
O
S
N
N
S
Luciferin phosphate
Immobilized-luciferase column
OH
hν
6 min
luciferin phosphate (4)
Light
Pump
Drain
Auto sampler
25 o
C
PP1γ
inhibitor
Singlephotoncounter
immobilized
luciferase,
ATP, Mg2+
CO2H
O
Data processer
HO
O
S
N
N
S
S
N
N
S
Solution A
inject
Enzyme solution (10 μL)
PP1γ 40 nM
luciferin (5)
oxyluciferin (6)
hν
Solution A
PP1 - inhibitor complex
Inhibitor solution (5 μL)
Inhibitor : (0 - 40 μM )
Scheme 2. Principle for the firefly bioluminescence system.
3
6 min
We have for some time been involved in work aiming at
elucidating the binding site of TTM with PP1c using
photolabeling followed by analysis using various MS
techniques. This has to date resulted in the synthesis
of various TTM photoaffinity probes.19 Recently we
also reported the synthesis of 13C-labeled tautomycin
and our efforts to study the interactions with PP1c by
Solution B
Luciferin phosphate (5 μL)
40 μM
Solution B
Luciferin phosphate
Figure 2. Schematic outline of the new assay protocol.
of the resulting solution was injected into the luciferase
column (see Figs. 2 and 3) with a flow rate of 0.5 mL/
min in order to measure the light yield. This procedure
was then repeated three times for each sample at each
concentration of inhibitor. The light intensity is directly
correlated to the total amount of luciferin present in the
sample, which is produced from luciferin phosphate
upon hydrolysis by PP1c. Based on the light yield for
each sample [value found by integrating the signals in
the PowerChrom chart (for an example of a Power-
Chrom chart see Figure S2 in SM)], the mean light yield
at each concentration of inhibitor was obtained. Plotting
this data gives the inhibitory curve (for an example of a
inhibitory curve, see Figure S3 in SM), which enables us
to obtain the IC50 value for the inhibitor tested. The Ki
value was then obtained by using the function (Eq. 1) as
outlined in Section 4.
13
using C NMR.20 However, the signals of the bound
form of 13C-labeled TTMDA with PP1c could not be
discerned due to line broadening occurring upon strong
binding with PP1c. The work described herein is part of
our ongoing effort toward the ultimate target of pin-
pointing the binding site of TTM with PP1c. Herein,
we disclose the modifications made to the biolumines-
cence assay system; we also report the biological activity
of our synthetic photoaffinity probes.
2. Results and discussion
2.1. Firefly bioluminescence system
The first generation firefly bioluminescence assay with
PP1c and inhibitors was operated by measuring the total
light production during the dephosphorylation reaction
(0–15 min).17 The data obtained was then plotted as
light intensity vs. time for each concentration of inhibi-
tors [see Figure S1 in supplementary material (SM)]. The
slope of these curves corresponds to the activity of lucif-
erase at the tested concentration. The relative activities
were then derived by comparing the absence of inhibi-
tors to the presence of inhibitor at the different concen-
tration of inhibitors. This data was then plotted in order
to afford the IC50 value for the inhibitors.
Tautomycin (TTMDA, 1b) was measured to have a Ki
of 4.5 nM with our new assay protocol (Table 1). This
value matched well with the data reported previously
by others using different test methods [Ki = 0.425 nM,8
For our advanced assay system we have further devel-
oped a new protocol for the measurement of the PP1c
activity (see Fig. 2). Protein phosphatase 1c (10 lL of
a 40 nM solution) and inhibitors with various concen-
tration (5 lL of solutions with concentration 0–40 lM,
which equals a test concentration of 0–10 lM after addi-
tion of PP1c and luciferin phosphate solution) were pre-
mixed in sample vials to promote the formation of
PP1c-inhibitor complex. These vials containing the
resulting 15 lL of sample mixtures were caped and
placed in the sample-stage rack in the auto-sampler.
Luciferin phosphate (5 lL of a 40 nM solution) was then
injected into the first vial and mixed by applying a 2 lL
air purge. After 6 min of reaction time an aliquot (1 lL)
Figure 3. Photo of the column containing biotinylated luciferase
placed in front of the mirror in the single photon counter. The
immobilized luciferase is in the middle of the column (in the center of
the red box).

1750
M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
Table 1. Ki values for the PP1c inhibitory activity of photoaffinity
probes syn-7–anti-11a
5
4
3
2
1
0
y = 0.225x + 0.017
Compound
Ki (nM)
SDb
TTMDA (1b)
syn-7
anti-7
4.5
32.5
8.0
0.1
2.3
0.2
4.1
6.1
0.3
1.7
0.4
0.3
1.2
syn-8
anti-8
49.7
213
8.4
syn-9
anti-9
17.3
3.4
anti-10
syn-11c
anti-11c
23.9
19.6
a Each measurement was repeated three times.
b SD = standard deviation.
c The assay was conducted under the same way except the omission of
dithiothreitol (DTT) in the Tris buffer in order to protect the S–S
bonds (see Section 4 for details).
Ki = 0.16 nM,21 IC50 = 0.19 nM,22 and IC50 = 22–
32 nM]23 and represents a significant improvement from
the old assay system (IC50 = 610 nM).17 Thus the results
acquired with the new assay protocol were confirmed to
be consistent with the results obtained previously by
others. One of the merits with the new method is time-
saving due to the simplified assay protocol. Since the
reaction time between PP1c and luciferin phosphate is
fixed to 6 min, there is no longer a need to plot the light
intensity versus time. By just measuring the amount of
luciferin produced from the phosphatase reaction mix-
tures, the inhibitory curve can easily be obtained. The
new method also has the advantage that we can measure
the activity of many inhibitors at once, since the auto-
sampler has 196 wells. With the high accuracy of the
sample loader, we have been able to decrease the total
sample usage to half of the volume needed for the previ-
ous method. Furthermore, as evident from the calibra-
tion data, the system can detect concentrations of
0.2 fmol/lL (0.2 nM) luciferin (see Figure S4 in SM)
and shows a linear calibration curve with increasing
concentration of luciferin (see Fig. 4). The system also
has a very good reproducibility with only a 0.7% fluctu-
ation for every integrated light yield at the same concen-
tration of luciferin.
0
0.5
1.0
1.5
2.0
Concentration fmol/μL (nM)
Figure 4. Calibration curve for luciferin at concentrations 0.2–
2.0 fmol/lL (0.2–2.0 nM).
irine (syn-8 and anti-8), and azide (syn-9-anti-11) (see
Fig. 5). The synthesis of these compounds has recently
been reported.19 Tautomycin and the photoaffinity
probes are stored as the anhydride and must therefore
be converted to the active diacid form prior to testing.
This was accomplished by treating the anhydride in ace-
tonitrile/water 4:1 with base (NaHCO3 aq solution)
according to our previously reported method (see Sec-
tion 4 for details).19a After a simple purification by
HPLC, we obtained the active diacid form of the photo-
affinity probes, viz. compounds syn-7–anti-11, which
were subjected to activity testing using our improved
bioluminescence assay system.
The assaying of the TTM photoaffinity probes was con-
ducted in two rounds spread out in time with two differ-
ent lots of protein using TTMDA (1b) as internal
standard in both rounds.24 The results from these anal-
yses are summarized in Table 1 and show that photo-
affinity probe anti-10 has a Ki of 3.4 nM, which
implements that this compound has higher activity than
the natural product itself, making it an exceptional good
inhibitor of PP1c (for a proposed structure of PP1c
complexed with photoaffinity probe anti-10, see
Fig. 6). Compounds anti-7 and syn-9 also possess strong
activity toward PP1c with Ki in the same order of mag-
nitude as TTMDA (1b) (Ki = 8.0 and 8.4 nM, respec-
tively). Photoaffinity probes syn-7, syn-8, anti-9, syn-
11, and anti-11 possesses moderate activity toward
PP1c, while compound anti-8 is a rather weak inhibitor.
The assay system described herein offers advantages
over other methods used to determine PP1 inhibition
activity. The 32P-labeled glycogen phosphorylase meth-
od, a commonly used method for such assaying,21 uses
radioactive phosphorous atom and therefore generates
radioactive waste that requires special treatment post
assaying. Compared with the p-nitrophenyl phosphate
(pNPP) method, another commonly used method, the
bioluminescence assay is much more sample efficient.9
The pNPP method normally requires a sample concen-
tration of 200–20000 lM while our assay system only re-
quires a sample concentration of 10 lM.
From the test results it seems that both the benzophe-
non and azide containing photoaffinity probes fit well
in the active site of PP1c while the methoxy group
associated with the diazirine photolabeling unit pre-
vents the photoaffinity probe to adapt an ideal position
for binding. Furthermore, from the azide containing
2.2. Analysis of the photoaffinity probes of tautomycin
The photoaffinity probes of tautomycin (syn-7–anti-11)
analyzed in this study possess three different photolabel-
ing units, namely benzophenone (syn-7 and anti-7), diaz-

M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
1751
O
O
OH
O
OH
O
OH
H
O
HO
HO
O
O
H
N
OMe
O
O
H
N
syn-7
anti-7
O
N
F3C
N
O
OH
O
OH
O
OH
H
O
HO
HO
O
O
H
N
OMe
O
OMe
H
N
O
N
H
O
syn-8
anti-8
O
F
O
OH
O
OH
O
OH
H
O
N3
HO
HO
O
O
H
N
OMe
O
H
N
syn-9, n = 1
anti-9, n = 1
anti-10, n = 2
O
N
O
n
H
O
F
O
O
OH
O
OH
O
OH
H
O
N3
HO
HO
O
O
H
N
OMe
O
H
N
S
S
N
H
O
syn-11
anti-11
O
Figure 5. Tautomycin with photoaffinity probes.
Figure 6. Proposed structure of PP1c complexed with photoaffinity probe anti-10 (compound anti-10 was placed in the active site of PP1c instead of
calyculin A). The model is generated and minimized using MacroModel 9.1 (force field MMFFs and OPLS2001, usign water as solvent).
photoaffinity probes we see that the length of the linker
also has some influence on the activity with the three-
carbon spacer being the ideal length. The high activity
for compound anti-10 is thought to be due to the ben-
eficial hydrogen bonding with Arg132 (see Fig. 6), an
interaction which is not possible for the remaining
photoaffinity probes.
Interestingly, compound anti-7 possessed higher activity
toward PP1c than photoaffinity probe syn-7, while the

1752
M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
Figure 7. Preliminary results obtained by MALDI–TOF–MS analysis from photolabeling experiments utilizing photoaffinity probe anti-7. (a)
Successful photolabeling of PP1c. (b) Control experiment.
result was opposite for the two isomers of compound 8
with the syn isomer being the most active of the two. A
similar phenomenon is also seen for photoaffinity probes
9 and 11; however, the difference in Ki is much smaller in
this case. The reason for this phenomenon is most likely
connected with the spacial constraints present in the ac-
tive site of PP1c.
with the new assay system matched well with the data re-
ported in the literature. The nine TTM photoaffinity
probes were found to possess activity toward PP1c in
the range of 3.4–213 nM. In particular, compound
anti-10 was found to be a very good inhibitor of PP1c
with activity greater than the natural product
(Ki = 3.4 nM). With these active photoaffinity probes
in hand, work is now focused on using them in order
to pinpoint the binding site of TTM with PP1c. Results
from these studies will be reported in due course.
Despite the rather low activity for photoaffinity probe
anti-8, we still anticipate that it will be useful for phot-
olabeling studies. It could even be possible that the com-
pounds with lower activity toward PP1c could be the
best substrates for photolabeling due to closer interac-
tion with the protein surface.
4. Experimental
4.1. General experimental
With these active photoaffinity probes in hand, work is
now focused on utilizing them for photolabeling studies.
Preliminary results from these studies show that after
10 min irradiation (using a high-pressure mercury lamp)
of a solution containing the complex between TTMDA
(1b) and photoaffinity probe anti-7 resulted in an in-
creased mass of the protein (Fig. 7a). On the other hand,
when the control experiment was conducted (Fig. 7b),
which implies irradiating a sample of PP1c, TTMDA
(1b), and compound anti-7, no mass increase was ob-
served. The latter result demonstrates that the labeling
of PP1c shown in Figure 7a is not due to random label-
ing of the protein, but the result of specific labeling after
the photoaffinity probe has been bound to the active
site. These encouraging results are now the basis for
our further efforts toward establishing the specific point
that has been labeled.
Tautomycin (1) used for this study was purified accord-
ing to our previously published procedure25 and lucif-
erin phosphate (4) used for the assay was freshly
prepared using our well-established method.26,27 Unless
otherwise mentioned, chemical reagents were of analyt-
ical grade. Immobilized biotinylated luciferase (BLU-Y)
on avidine-beads (conjugated form)28 was kindly pro-
vided by Kikkoman Company and was kept in the
refrigerator prior to use. The immobilized luciferase re-
tains its activity for a prolonged period of time,29 and
after more than 70 h at 25 ꢁC it still has an activity great-
er than 60%. Protein phosphatase type 1c expressed in
Escherichia coli was donated by Dr. P.T.W. Cohen30
and was purified according to our previously reported
method.20 Protein phosphatase 1c was then stored in
buffer containing 20 mM Tris (hydroxymethyl) ami-
nomethan hydrochloric acid (Tris–HCl), 4 mM ethylen-
diamine tetraacetic acid (EDTA), 2 mM dithiothreitol
(DTT), and 50% (v/v) glycerol (pH 7.4 at 25 ꢁC) at
ꢀ20 ꢁC until use. After measuring the concentration of
PP1c by using the titration method described by Takai
and Mieskes,31 PP1c was diluted with the Tris–HCl buf-
fer to give a 40 nM solution used for the assay. Tauto-
mycin and the photoaffinity probes are stored as the
anhydride and must therefore be converted to the active
diacid form prior to testing. The progress of this reac-
tion was monitored by HPLC using a Develosil C30-
3. Conclusion
We have succeeded in improving the performance of our
in-house developed firefly bioluminescence assay system.
The new assay system has higher sensitivity [0.2 fmol/lL
(0.2 nM)] and is time and sample efficient compared to
the old version providing us with a powerful analytical
method for the evaluation of our photoaffinity probes.
The inhibitory activity of TTMDA with PP1c measured

M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
1753
UG-5 column (4.6 · 250 mm id), CH3CN/H2O 4:1 with
a flow rate of 1.0 mL/min, detected at 254 nm.
was kept flowing into the column with a flow rate of
0.5 ml/min during the luminescence assay.
4.2. Instrumentation
4.2.7. Procedure for calibration of the single photon
counter. Calibration of the photon counter was performed
with the reaction between immobilized luciferase and
luciferin (0–20 nM solution in Tris–HCl buffer). Each
aliquot (1 lL) of luciferin solution was injected into the
column every 2 min by using the auto-sampler (JASCO
AS-1559) with a flow rate of 0.5 mL/min. The resulting
light emission from the column was detected by the single
photon counter. The data was accumulated with a HP
53131ꢂ digital counter with high S/N ratio and was then
converted to analogue signals with an ADC R6142ꢂ volt-
age/current generator. The resulting current was then
sampled with the PowerChormꢂ software and transferred
to a Macintosh LC 630 with a SCSI cable. Total light yield
for each concentration of luciferin was integrated using
the PowerChorm software on a Macintosh LC 630. The
HP 53131ꢂ digital counter was controlled with the Lab-
VIEW program modified by JASCO. The dynamic range
of the photon counter was adjusted to 0.01 mV/count for
measurement with the LabVIEW program.
The luminescence detector consisted of a JASCO PU-
980 pump, JASCO AS-1559ꢂ auto-sampler, JASCO sin-
gle photon counter equipped with a HP53131 digital
counter and an ADC R6142ꢂ voltage/current generator,
and PowerChromꢂ software for digital sampling of the
photon counting. The counted photons were integrated
using the PowerChrom software on a Macintosh LC
630. The mobile phase was kept in a bottle cooled in
an ice-bath and was heated to 25 ꢁC through a 20 cm
long loop placed in a water bath (25 ꢁC) before flowing
into the luciferase column (see Figure S5 in SM).
4.2.1. Preparation of 4-(2-hydroxyethyl)-1-piperazinee-
thanesulfonic acid (Hepes) buffer. 40 mM Hepes, 2 mM
MgCl2Æ6H2O, and 0.4 mM EDTA were dissolved in dis-
tilled water and pH was adjusted to 7.76 (at 25 ꢁC) by
adding 1 M NaOH aq solution. The resulting solution
was kept cool in an ice-bath.
4.2.2. Preparation of mobile phase. ATP (0.1% w/v) and
DTT (0.01 mM) were added to the Hepes buffer and the
resulting solution was kept in a bottle cooled in an ice-
bath.
4.2.8. Method for converting tautomycin (1) and photo-
affinity probe anhydrides (syn-7–anti-11) to diacid. An
aqueous solution of NaHCO3 (18 lL of a 20 mg/mL
solution) was added to a stirred solution of the relevant
anhydride (0.0025 mmol) in CH3CN/water 4:1 (0.3 mL)
at room temperature. The resulting reaction mixture was
stirred for 3–4 h before being neutralized with HCl
(0.1 N aq solution) and purified by HPLC [Develosil
C30-UG-5 column (4.6 · 250 mm id), CH3CN/H2O
4:1, 1.0 mL/min, 254 nm] followed by concentration of
the relevant fractions in vacuo. The resulting residue
was then dissolved in DMSO/CH3CN 4:1 in order to ob-
tain a 4 mM stock solution of the inhibitor.
4.2.3. Preparation of buffer for activation of immobilized
luciferase. ATP (0.5% w/v) and DTT (0.01 mM) were
added to the Hepes buffer and the resulting solution
was kept in a bottle cooled in an ice-bath.
4.2.4. Preparation of Tris (hydroxymethyl) aminome-
than–HCl (Tris–HCl) buffer. 40 mM Tris, 30 mM
MgCl2Æ6H2O, and 20 mM KCl were dissolved in dis-
tilled water and pH was adjusted to 8.43 (at 25 ꢁC) by
adding 1 M HCl aq solution. The resulting solution
was kept cool in an ice-bath.
4.2.9. Method for sample preparation. The 4 mM stock
solution ofthe inhibitors was dilutedwith Tris–HCl buffer
in order to obtain samples with the following concentra-
tions: 40.0 lM, 20.0 lM, 12.0 lM, 8.0 lM, 4.0 lM,
2.0 lM, 400 nM, 200 nM, 80 nM, 40 nM, 4.0 nM,
0.4 nM, 0.04 nM, and 0.0 nM. The sample solutions
(5 lL) together with the PP1c solution (10 lL of a 40 nM
solution) were then put into sample vials and mixed prior
to loading the vials into the sample-stage rack.
4.2.5. Preparation of Tris buffer for dilution of PP1c
(Tris-B). Bovine serum albumin (BSA)32 (0.15% w/v)
and DTT (4.0 mM) were added to the Tris–HCl buffer
and the resulting solution was kept cool in an ice-bath.
For photoaffinity probes syn-11, and anti-11 the buffer
was prepared without DTT.33
4.2.6. Preparation and mounting of luciferase column.
The immobilized luciferase was packed as a 3 mm long
band at the center of a transparent polytetrafluoroethyl-
ene tube (1.07 mm id · 15 cm) and both sides of the
luciferase band were capped with Teflon wool. A poly-
tetrafluoroethylene tube (1.07 mm od · 7.2 cm) was then
inserted into each end until it was parallel with the outer
tube (see Figure S6 in SM). The luciferase column was
then attached in front of the single photon counter (sin-
gle photon counter is beneficial for measuring weak light
emission with high sensitivity) and the luciferase column
was activated by flowing the activation buffer through
for 1 h with a flow rate of 0.5 mL/min. Immediately after
the activation of the luciferase column was finished the
flow of mobile phase was started. The mobile phase
4.2.10. Activity test. Luciferin phosphate (5 lL of a 40 lM
solution)34 was injected into the sample vial followed by a
2 lL air purge. After 6 min 1 lL of the resulting mixture
(final amount of sample 1.0 · 10ꢀ11–0.0 mol, final
amount of PP1c 2.0 · 10ꢀ8 mol, final amount of luciferin
phosphate 1.0 · 10ꢀ11 mol) was injected into the lucifer-
ase column with a flow rate of 0.5 mL/min. The resulting
emitted light was then recorded as described in the proce-
dure for calibration of the single photon counter (vide su-
pra). This procedure was repeated three times for each
sample at each concentration of inhibitors.
4.2.11. Dose–inhibition analysis; determination of Ki. In
the previous experiments with tight-binding phospha-
tase inhibitors including okadaic acid, microcystins,

1754
M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
and calyculin A as well as TTMDA, we have shown that
the steady-state dose–inhibition relationships at a fixed
total enzyme concentration, Et, is well described by a
theoretical function
Scientists (19780087). We also thank Dr. Kiyoshi Isono
(ex-Riken Institute) and Professor Petricia Cohen (Uni-
versity of Dundee) for kindly providing a sample of tau-
tomycin and PP1c expressed in E. coli, respectively.
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
Et ꢀ It ꢀ Ki þ ðEt ꢀ It ꢀ KiÞ þ 4EtKi
Supplementary data
/ðItÞ ¼
ꢁ /ð0Þ
ð1Þ
2Et
Figures S1–S6 and Supplementary material Figures 1
and 2 can be found in the supplementary material. Sup-
plementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmc.2007.11.034.
where /(It) denotes the enzyme activity at a given inhib-
itor concentration, It, and Ki stands for the dissociation
constant.8 This general model function takes into ac-
count the reduction of the free inhibitor concentration
as a result of binding with the enzyme. In the present
experiments the Ki values were estimated with the stan-
dard errors by non-linear least-squares regression of Eq.
1 to (It, /(It)) data obtained by dose–inhibition experi-
ments. In the regression calculations, we used as the
weight the reciprocal of the square of the standard error
at each value of It.
References and notes
1. Cheng, X.-C.; Kihara, T.; Kusakabe, H.; Magae, J.;
Kobayashi, Y.; Fang, R.-P.; Ni, Z.-F.; Shen, Y.-C.; Ko,
K.; Yamaguchi, I.; Isono, K. J. Antibiot. 1987, 40, 907–
909; For a review regarding tautomycin and other maleic
anhydride containing natural products, see: Chen, X.;
Zheng, Y.; Shen, Y. Chem. Rev. 2007, 107, 1777–1830.
2. Ubukata, M.; Cheng, X.-C.; Isono, K. J. Chem. Soc.
Chem. Commun. 1990, 244–246; Cheng, X.-C.; Ubukata,
M.; Isono, K. J. Antibiot. 1990, 43, 809–819; Ubukata, M.;
Cheng, X.-C.; Isobe, M.; Isono, K. J. Chem. Soc., Perkin
Trans. 1 1993, 617–624.
4.2.12. Protein modeling. Docking simulations were per-
formed following the method of Chamberlin et al.22
Tautomycin and photoaffinity probes were overlaid on
the bound conformation of calyculin A (from X-ray
structure) and minimized by using MacroModel 9.1.
3. Magae, J.; Watanabe, C.; Osada, H.; Cheng, X.-C.; Isono,
K. J. Antibiot. 1988, 41, 932–937.
4. For the X-ray crystal structure of PP1, see: Egloff, M.-P.;
Cohen, P. T. W.; Reinemer, P.; Barford, D. J. Mol. Biol.
1995, 254, 942–959.
5. For a recent review regarding the 3D structure of PP2A,
see: Mumby, M. ACS Chem. Biol. 2007, 2, 99–103.
6. (a) Cohen, P. Annu. Rev. Biochem. 1989, 58, 453–508; (b)
Shenolikar, S.; Nairn, A. C. Adv. Second Messenger
Phosphoprotein Res. 1991, 23, 1–121.
7. For reviews regarding inhibition of PP1 and PP2A by
naturally occurring toxins, see for example: (a) McClus-
key, A.; Sim, A. T. R.; Sakoff, J. A. J. Med. Chem. 2002,
45, 1151–1175; (b) Dawson, J. F.; Homes, C. F. B. Front.
Biosci. 1999, 4, d646–d658; (c) Sheppeck, J. E.; Gauss, C.-
M.; Chamberlin, A. R. Bioorg. Med. Chem. 1997, 5, 1739–
1750.
8. Takai, A.; Sasaki, K.; Nagai, H.; Mieskes, G.; Isobe, M.;
Isono, K.; Yasumoto, T. Biochem. J. 1995, 306, 657–665.
9. Takai, A.; Tsuboi, K.; Koyasu, M.; Isobe, M. Biochem. J.
2000, 350, 81–88.
10. Sugiyama, Y.; Ohtani, I. I.; Isobe, M.; Takai, A.;
Ubukata, M.; Isono, K. Bioorg. Med. Chem. Lett. 1996,
6, 3–8.
11. Nishiyama, U.; Ubukata, M.; Magae, J.; Kataoka, T.;
Erdo¨di, F.; Hartshorne, D. J.; Isono, K.; Nagai, K.;
Osada, H. Biosci. Biotech. Biochem. 1996, 60, 103–107.
12. Liu, W.; Sheppeck, J. E.; Colby, D. A.; Huang, H.-B.;
Nairn, A. C.; Chamberlin, A. R. Bioorg. Med. Chem. Lett.
2003, 13, 1597–1600.
4.2.12.1. Photoaffinity labeling. A solution of photo-
affinity probe anti-7 (141 lM) and PP1c (11 lM) in
Tris–HCl buffer (10 mM) containing 0.4% DMSO and
3.4% CH3CN was incubated at 0 ꢁC for 30 min and then
at 25 ꢁC for 30 min. Five microliters of the resulting
solution was loaded into a glass capillary tube (Ring-
capsꢂ 5/10 lL) and the tube was sealed with a gas bur-
ner. The sample was irradiated with a high-pressure
mercury lamp for 10 min at room temperature and the
resulting product mixture was subjected to MALDI–
TOF–MS analysis.
4.2.12.2. Control experiment. A solution of TTMDA
(1b) (632 lM), photoaffinity probe anti-7 (141 lM),
and PP1c (11 lM) in Tris–HCl buffer (10 mM) contain-
ing 0.4% DMSO and 3.4% CH3CN was incubated at
0 ꢁC for 30 min and then at 25 ꢁC for 30 min. Five
microliters of the resulting solution was loaded into a
glass capillary tube (Ringcapsꢂ 5/10 lL) and the tube
was sealed with a gas burner. The sample was irradiated
with a high-pressure mercury lamp for 10 min at room
temperature and the resulting product mixture was sub-
jected to MALDI–TOF–MS analysis.
Acknowledgments
13. Kurono, M.; Isobe, M. Chem. Lett. 2004, 33, 452–453.
14. The X-ray crystal structures of PP1 complexed with
microcystin-LR, okadaic acid, and calyculin A are all
known. The crystallographic data suggests that these
inhibitors interact with five common amino acid residues.
The polar functional groups, viz. carboxylic or phosphoric
acid, in these inhibitors interact with Arg96 and Tyr272,
respectively. For the X-ray crystal structure of microcy-
stin-LR bound to PP1, see: (a) Goldberg, J.; Huang, H.;
Kwon, Y.; Greengard, P.; Nairn, A. C.; Kuriyan, J.
Research was supported by funding from Grant-in-Aid
for Specially Promoted Research (16002007) from the
Ministry of Education, Culture, Sports, Science and
Technology (MEXT), Japan. M.O.S. is grateful for the
provision of an Inoue Fellowship from the Inoue Foun-
dation for Science (July 2006-June 2007) and JSPS for a
JSPS Postdoctoral Fellowship for Foreign Researchers
(July 2007–June 2009). M.K. is the grateful recipient
of a JSPS Grant-in-Aid for Encouragement of Young

M. O. Sydnes et al. / Bioorg. Med. Chem. 16 (2008) 1747–1755
1755
Nature 1995, 376, 745–753; For the X-ray crystal structure
of okadaic acid bound to PP1, see: (b) Maynes, J. T.;
Bateman, K. S.; Cherney, M. M.; Das, A. K.; Luu, H. A.;
Holmes, C. F. B.; James, M. N. G. J. Biol. Chem. 2001,
276, 44078–44082; For the X-ray crystal structure of
calyculin A bound to PP1, see: (c) Kita, A.; Matsunaga, S.;
Takai, A.; Kataiwa, H.; Wakimoto, T.; Fusetani, N.;
Isobe, M.; Miki, K. Structure 2002, 10, 715–724.
15. Zhang, L.; Zhang, Z.; Long, F.; Lee, E. Y. C. Biochemistry
1996, 35, 1606–1611.
16. (a) Isobe, M.; Takahashi, H.; Usami, K.; Hattori, M.;
Nishigohri, Y. Pure Appl. Chem. 1994, 66, 765–772; (b)
Isobe, M.; Sugiyama, Y.; Ito, T.; Ohtani, I. I.; Toya, Y.;
Nishigohri, Y.; Takai, A. Biosci. Biotech. Biochem. 1995,
59, 2235–2238.
17. Sugiyama, Y.; Fujimoto, K.; Ohtani, I. I.; Takai, A.;
Isobe, M. Biosci. Biotech. Biochem. 1996, 60, 1260–1264.
18. (a) Fresu, M.; Bianchi, M.; Parsons, J. T.; Villa-Moruzzi,
E. Biochem. J. 2001, 358, 407–414; (b) Berndt, N.;
Campbell, D. G.; Caudwell, F. B.; Cohen, P.; da Cruz e
Silva, E. F.; da Cruz e Silva, O. B.; Cohen, P. T. W. FEBS
Lett. 1987, 223, 340–346; (c) Dombradi, V.; Axton, J. M.;
Brewis, N. D.; da Cruz e Silva, E. F.; Alphey, L.; Cohen,
P. T. W. Eur. J. Biochem. 1990, 194, 739–745; (d) Barker,
H. M.; Jones, T. A.; da Cruz e Silva, E. F.; Spurr, N. K.;
Sheer, D.; Cohen, P. T. W. Genomics 1990, 7, 159–166; (e)
Sasaki, K.; Shima, H.; Kitagawa, Y.; Irino, S.; Sugimura,
J.; Nagao, M. Jpn. J. Cancer Res. 1990, 81, 1272–1280.
19. (a) Kurono, M.; Shimomura, A.; Isobe, M. Tetrahedron
2004, 60, 1773–1780; (b) Sydnes, M. O.; Isobe, M.
Tetrahedron 2007, 63, 2593–2603.
24. Photoaffinity probes syn-7–anti-8 was tested with one lot
of protein in 2003 and compounds syn-9-anti-11 was tested
with a different lot of protein in 2007. The measured Ki for
TTMDA (1b), which was used as internal standard in the
two rounds of assaying, was found to be different in the
two sets of analysis most likely due to difference in specific
activity of the protein in the two lots. However, for the
ease of comparison of the activity of the photoaffinity
probes, we adjusted the Ki for the last round of analysis by
using the Ki of TTMDA, which is an internal standard in
the analysis.
25. Kurono, M.; Isobe, M. Tetrahedron 2003, 59, 9609–
9617.
26. Toya, Y.; Takagi, M.; Kondo, T.; Nakata, H.; Isobe,
M.; Goto, T. Bull. Chem. Soc. Jpn. 1992, 65, 2604–
2610.
27. Luciferin phosphate is commercially available.
28. Biotinylated luciferase (BLU-Y) on avidine-beads (conju-
gated form) is commercially available (http://www.kikko-
man.co.jp/bio/j/rinsyou/pdf/57_BLU-Y.pdf).
29. Kajiyama, N.; Nakano, E. Biosci. Biotechnol. Biochem.
1994, 58, 1170–1171.
30. Alessi, D. R.; Street, A. J.; Cohen, P.; Cohen, P. T. Eur.
J. Biochem. 1993, 213, 1055–1066.
31. Takai, A.; Mieskes, G. Biochem. J. 1991, 275, 233–239.
32. BSA is necessary in order to avoid PP1c to adsorb on the
sample vials.
33. We have previously shown that the disulfide bond
incorporated in the photoaffinity probes syn-11 and anti-
11 is cleaved in the presence of excess amount of DTT (see
Ref. 19b). DTT was therefore omitted in the Tris buffer
used to dilute PP1c during assaying of syn-11 and anti-11
in order to avoid the possibility of the disulfide bond being
cleaved.
34. The luciferin phosphate solution should be freshly made
prior to use. If the luciferine phosphate solution is stored
for a prolonged period of time, prior to use it is slowly
dephosphorylated resulting in high background signals.
Luciferin phosphate (solid) is stable for a long period of
time when stored at ꢀ30 ꢁC.
20. Isobe, M.; Kurono, M.; Tsuboi, K.; Takai, A. Chem.
Asian J. 2007, 2, 377–385.
21. MacKintosh, C.; Klumpp, S. FEBS Lett. 1990, 277, 137–
140.
22. Colby, D. A.; Liu, W.; Sheppeck, J. E.; Huang, H.-B.;
Nairn, A. C.; Chamberlin, A. R. Bioorg. Med. Chem. Lett.
2003, 13, 1601–1605.
23. Hori, M.; Magae, J.; Han, Y.-G.; Hartshorne, D. J.;
Karaki, H. FEBS Lett. 1991, 285, 145–148.