Identification of 12Cysβ on tubulin as the binding site of tubulyzine

Article abstract of DOI:10.1016/j.bmc.2005.09.069

We have undertaken quantitative binding site studies in order to identify the binding site of the known microtubule destabilizing agents, the tubulyzines, in the tubulin dimer. Two different approaches were employed that utilized the tubulyzines and their

Full text of DOI:10.1016/j.bmc.2005.09.069

Bioorganic & Medicinal Chemistry 14 (2006) 1169–1175
Identification of 12Cysb on tubulin as the binding site of tubulyzine
Yeoun Jin Kim,a Dan L. Sackett,b Matthieu Schapira,c, Daniel P. Walsh,d
Jaeki Min,d Lewis K. Pannella and Young-Tae Changd,*
aNational Institute of Diabetes, Digestive, Kidney Diseases, National Institutes of Health, U.S. Department of Health and
Human Services, Bethesda, MD 20892, USA
bNational Institute of Child Health and Human Development, National Institutes of Health, U.S. Department of Health and
Human Services, Bethesda, MD 20892, USA
cDepartment of Pharmacology, New York University Medical Center, New York, NY 10016, USA
dDepartment of Chemistry, New York University New York, NY 10003, USA
Received 9 September 2005; revised 24 September 2005; accepted 26 September 2005
Available online 2 November 2005
Abstract—We have undertaken quantitative binding site studies in order to identify the binding site of the known microtubule desta-
bilizing agents, the tubulyzines, in the tubulin dimer. Two different approaches were employed that utilized the tubulyzines and their
derivatives. The first approach was based on a chemical affinity labeling method using tubulyzine affinity derivatives, and the second
approach employed the mass spectrometric measurement of the differential reactivity of cysteines using the tubulyzines and mono-
bromobimane. Based on overlapping data from these two approaches, we propose that the tubulyzines bind at the guanosine-50-
triphosphate binding site of b-tubulin. Interestingly, we also show that the tubulyzinesÕ binding to tubulin induces a conformational
change in tubulin that prevents further interaction of the 239Cysb with other reagents.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
drugs have prompted the need for the development of
improved therapeutics.1–4 A purine based microtubule
inhibitor, myoseverin, has demonstrated unique and
promising properties, the most interesting of which is
low cytotoxicity.5,6 Despite the promise of myoseverin,
its moderate activity remains to be improved. However,
due to synthetic restraints stemming from its purine
scaffold, large scale diversity generation and scale-up
are difficult. Therefore, we designed a library of tubulin
targeted myoseverin derivatives around the triazine scaf-
fold. Its threefold symmetry, low cost, and synthetic
flexibility allowed for the generation of diverse, highly
pure libraries. We previously identified the microtubule
destabilizing entities, the tubulyzines, which were shown
to target tubulin and possess an activity greater than
that of myoseverin itself.7 Figure 1 shows the structures
of myoseverin and its related compounds tubulyzine A
and tubulyzine B.
Tubulin, the building block of microtubules, is one of
the proven targets for anticancer drugs. Thus far, several
agents have been developed that either stabilize or desta-
bilize microtubules, such as taxol and the vinca alka-
loids, respectively. Although these agents are currently
being used in cancer treatment, problems associated
with the intrinsic toxicity and solubility of the known
Abbreviations: MS-DRC, mass spectrometric measurement of the dif-
ferential reactivity of cysteines; mBrB, monobromobimane; GTP, gu-
anosine-50-triphosphate; rt, room temperature; LC–MS, liquid
chromatography–mass spectrometry; THF, tetrahydrofuran; DIPEA,
diisopropylethylamine; EtOAc, ethyl acetate; TEA, triethylamine; T-
LC, thin-layer chromatography; PIPES, piperazine-N,N-bis(ethane-
sulfonic acid); TCEP, tris(2-carboxyethyl)phosphine hydrochloride;
Tn, tubulyzine affinity derivatives; ESI, electrospray ionization; HP-
LC, high-performance liquid chromatography; CID, collision induced
dissociation; MS, mass spectrometry; SIC, single ion current; TIC,
total ion current; FWHM, full width at half maximum; GDP, guan-
osine di phosphate; TA, tubulyzine A; TB, tubulyzine B.
Compared to other tubulin targeting compounds, the
unique characteristics of myoseverin and tubulyzine
may either come from targeting a different binding site
on tubulin or from the existence of cellular targets other
than tubulin. To address the former possibility, we
designed mass spectrometric approaches for identifying
the tubulyzinesÕ binding site. Two strategies were
Keywords: Tublin; Tubulyzine; Binding site; MS-DRC.
*
Corresponding author. Tel.: +1 212 998 8491; fax: +1 212 260
7905; e-mail: yt.chang@nyu.edu
Present address: Aptanomics, 181-203 Ave Jean Jaures, 69001 Lyon,
France.
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.09.069

1170
Y. J. Kim et al. / Bioorg. Med. Chem. 14 (2006) 1169–1175
c
a
b
O
O
O
HN
HN
HN
6
N
N
N
N
N
N
N
2
HN
9
N
HN
N
NH
HN
N
NH
O
O
O
Myoseverin
Tubulyzine A (TA)
Tubulyzine B (TB)
Figure 1. Molecular structures of (a) myoseverin, (b) tubulyzine A, and (c) tubulyzine B.
employed: (1) a chemical affinity method and (2) the
mass spectrometric measurement of the differential reac-
tivity of cysteine (MS-DRC).
binding site. In this approach, the binding site of a
non-covalent drug can be identified by measuring the
differential reaction yield of each cysteine toward mono-
bromobimane (mBrB) before and after drug treatment.
The reactivity of a specific cysteine close to the drug
binding site will decrease after drug treatment due to
the occupancy of the drug. The great advantage of this
method is that it measures the reactivity change of each
cysteine toward an alkylating agent, instead of monitor-
ing the drug itself. Thus, no modification of the drug is
required. In this manner, the random reaction of a
chemical affinity label with an amino acid, which is
not related to the original binding site, can be overcome.
We applied this method, which also uses mass spectro-
metric ion intensity measurements, to measure the reac-
tivity and to identify the tubulyzine binding site.
For the chemical affinity method, multiple tubulyzine
affinity derivatives with different arm lengths or different
reaction moieties were synthesized to optimize the reac-
tion efficiency. Previous studies demonstrated that addi-
tion of similar arms with reactive ends to myoseverin
allowed specific labeling of the binding site on tubulin.5
Therefore, the affinity derivatives of tubulyzine repre-
sent similar modifications to those already validated.
Figure 2 shows two examples of these derivatives. After
binding to their specific site on tubulin, the tubulyzine
derivatives can covalently attach to any cysteines in
the vicinity. Therefore, the binding site can be deter-
mined by identifying tubulyzine modified peptides after
tryptic digestion. To identify the binding site, we em-
ployed mass spectrometric peptide sequencing. Since
non-specific binding is possible in this approach and
may generate a false positive, we compared modification
yields across all the cysteines in the tubulin dimer, rather
than identifying only cysteine-modified peptides. The
modification (reaction) yield was calculated by the mass
spectrometric intensity of each cysteine-containing pep-
tide and its modified form. The relative reaction rate
of each cysteine will reveal the most reactive cysteine
toward the tubulyzine derivatives.
2. Results and discussion
2.1. Identification of the tubulyzine binding site using the
chemical affinity labeling method
Tubulin contains 20 cysteines (12 in a-tubulin, 8 in
b-tubulin).11 Table 1 shows 14 cysteine(s)-containing
peptides that can be generated from the tryptic digestion
of the tubulin dimer. Cysteine-containing peptides and
their Tn (T6 or T7) modified forms were identified using
mass spectrometric sequencing. Figure 3 shows an
example of the peptide sequencing before and after T6
modification. Once a tubulyzine derivative bound to
On the other hand, the MS-DRC is an indirect foot-
printing method for localizing a non-covalent drug
O
NH
Cl
O
Cl
O
NH
O
O
HN
HN
N
N
N
N
N
N
N
H
N
N
N
H
H
H
O
O
O
O
T6
T7
Figure 2. Tubulyzine derivatives with differing linker arm lengths for the chemical affinity study.

Y. J. Kim et al. / Bioorg. Med. Chem. 14 (2006) 1169–1175
1171
Table 1. Cysteine-containing peptides resulting from the trypsin digestion of the ab-tubulin heterodimer
Residue
Sequence
Mw
Probing Cys
b: 3–19
EIVHIQAGQCGNQIGAK
1764.89
3212.48
4478.06
2650.31
1007.40
970.49
12Cys
127, 129Cys
201, 211Cys
239Cys
b: 123–154
b: 175–213
b: 217–241
b: 298–306
b: 351–359
a: 3–40
a: 125–156
a: 167–214
a: 281–304
a: 305–308
a: 312–320
a: 340–352
a: 374–390
ESESCDCLQGFQLTHSLGGGTGSGMGTLLISK
VSDTVVEPYNATLSVHQLVENTDETYCIDNEALYDICFR
LTTPTYGDLNHLVSATMSGVTTCLR
NIMMAACDPR
303Cys
354Cys
TAVCDIPPR
ECISIHVGQAGVQIGNACWELYCLEHGIQPDSQMPSDK
LADQCTGLQGFLVFHSFGGGTGSGFTSLLMER
LEFSIYPAPQVSTAVVEPYNSILTTHTTLEHSDCAFMVDNEAIYDICR
AYHEQLSVAEITNACFEPANQMVK
CDPR
4124.89
3332.60
5402.55
2692.26
489.20
4, 20, 25Cys
129Cys
200, 213Cys
295Cys
305Cys
YMACCLLYR
TIQFVDWCPTGFK
AVCMLSNTTAIAEAWAR
1134.50
1540.74
1806.88
315, 316Cys
347Cys
376Cys
a
y1 y2
y3 y4
y5 y6 y7 y8 y9 y10 y11 y12 y13 y14 y15
M+H
100
%
R
A
W
A
E
A
I
A
T
T
N
S
L
M
C
VA
Mw= 1806.88 Da
0
b
100
%
*
LM[C-T6]VA
Mw= 2312.16 Da
R
A
W
A
E
A
I
A
T
T
N
S
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
m/z
Figure 3. MS/MS spectra of one of the cysteine-containing peptides (a) and its T6-modified form (b). Intensities of the y-series ions of the T6
modified peptide were reduced with a strong p-methoxybenzyl ion at 121.1 Da indicated with an asterisk.
the peptides, the intensity of the y- and b-series ions in
the MS/MS spectrum decreased dramatically along
with the increasing intensity of the p-methoxybenzyl
ion released from tubulyzine during fragmentation.
The partial loss of the y-series ions and the reduction
of the total fragment ion signals were due to the
stable p-methoxybenzyl ion (121.1 Da) that neutralizes
other protonated peptide fragments. Although this
event decreases the coverage of amino acid sequencing,
the strong peak at 121.1 Da provides a unique signa-
ture for tubulyzine modification, which confirms
identification.
unmodified peptides. Therefore, we added all the possi-
ble charge states together in the calculation. The reac-
tion yield of each cysteine toward Tn was calculated
from Eq. 1. Figure 4 shows the relative reaction yield
of each cysteine toward the tubulyzine derivatives T6
and T7, which are 5–10 times weaker than tubulyzines,
but still active tubulin inhibitors. We found that nine
peptides were modified in the T6 experiment and seven
peptides were modified in the T7 experiment. In both
data sets, the reactivity profiles of the cysteines (across
14 peptides) look very similar, and the most reactive cys-
teine in b-tubulin was 12Cys in both the T6 and T7
experiments. However, in a-tubulin, 376Cys shows reac-
tivity similar to 12Cysb.
Once we identified all the cysteine-containing peptides
and their Tn-modified forms, the ion intensities were
quantified. Since modification of the Tn to the peptide
introduces an additional protonation site at the triazine
ring, the charge state of the Tn-modified peptides in the
mass spectra is higher than that of of the corresponding
Although the reactivity seen was assumed to be predom-
inantly driven by the specific binding of tubulyzine, a
general alkylating reaction as a result of non-specific
binding could possibly contribute to the final reaction

1172
Y. J. Kim et al. / Bioorg. Med. Chem. 14 (2006) 1169–1175
Figure 4. Relative reactivity of cysteines toward (A) T6 and (B) T7. In both cases, 12Cys in the b and 376Cys in the a-tubulin showed the greatest
reactivity.
yield. Thus, the reaction of Tn to cysteines which are not
near the original tubulyzine binding site remains a
possible problem. This is one of the drawbacks of the
traditional chemical affinity labeling approach.12
However, it is notable in this regard that most of the
peptides labeled by T6 and T7 contain cysteine residues
previously shown to be the most reactive to a number of
general sulfhydryl reactive probes, including iodoaceta-
mide and N-ethylmaleimide. These residues are 239
and 354 in b-tubulin, and 315, 316, 347, and 376 in
a-tubulin.11 Thus, the major site of the reaction of T6
and T7 with tubulin, aside from the non-specifically
reactive residues, is at 12Cysb.
Primarily, any reactivity reduction is due to the occu-
pancy of a drug that blocks the alkylation reaction.
Therefore, a cysteine that has differential reactivity
before and after drug treatment can be considered as a
probable drug-binding site. Previous MS-DRC experi-
ments demonstrated two categories of known tubulin-
binding agents: (1) those agents for which the largest
reduction of reactivity occurred at 239Cysb, such as col-
chicine, podophyllotoxin, and 2-methoxyestradiol, and
(2) those agents for which the largest reduction of reac-
tivity occurred at 12Cysb, such as GDP,13 and crypto-
phycin and vincristine (unpublished data). In no case
did any of these tubulin targeting agents show a dual de-
crease at both 12Cysb and 239Cysb as was the case for
tubulyzine.
2.2. Identification of the tubulyzine binding site using the
MS-DRC method
The second source of reactivity reduction could possibly
arise from a conformational change in the target pro-
tein. Previous studies have suggested that colchicine
binding at tubulin causes a conformational change in
tubulin14 and the MS-DRC study demonstrated that
pretreatment with colchicine decreased the reactivity at
239Cysb and increased the reactivity at 12Cysb. This in-
creased reactivity is due to enhanced vulnerability after
the conformational change. Therefore, the dual-site
reactivity changes seen after tubulyzine treatment may
demonstrate a possible conformational change in
tubulin.
Since the MS-DRC method does not require any struc-
tural modification to the drugs, a series of tubulyzines
(TA and TB) were used directly in this study. Figure 5
shows the reaction yield change of each cysteine toward
mBrB before and after TA treatment. Noticeable reac-
tivity changes at 12Cysb and 239Cysb were found in this
data set.
Interestingly, a similar MS-DRC result was found when
comparing the free dimer form of tubulin with the
microtubule form.13 The reactivity of the cysteines in
the free dimer form was reduced upon formation of
the microtubule. The largest reduction of reactivity
was shown at both Cys12 and Cys239 in b-tubulin.
The reduction at Cys12b is due to the occupancy of
the GDP site as well as its burial in the polymer form,
and the reduction at Cys239b may possibly be explained
by the restricted accessibility of mBrB toward Cys239b
due to the more rigid structure of tubulin in the polymer
form.
Figure 5. Relative reactivities of each cysteine toward mBrB with and
without TA pretreatment. The reactivity of mBrB decreased at 12Cys
and 239Cys in b-tubulin after TA treatment.
Cys239b is a very reactive cysteine that binds with vari-
ous tubulin targeting agents including T13067 developed

Y. J. Kim et al. / Bioorg. Med. Chem. 14 (2006) 1169–1175
1173
by Tularik Inc.15 Since the Cys239b is not located on the
surface of tubulin, its distinctive reactivity is intrigu-
ing.9,16 One possibility is that its higher reactivity may
be due to the higher accessibility and/or a lower pKa
of its thiol.11 However, it has been speculated that tubu-
linÕs dynamic structural change (semi-folding and
unfolding equilibrium) may allow the agent access to
the inside of tubulin. Therefore, once it forms a rigid
structure in the microtubule form, tubulin may no long-
er open up, and the accessibility toward 239Cysb can be
reduced. This hypothesis explains the large reactivity
reduction at Cys239b seen following microtubule
formation.
structural information for the rational design of future
tubulin targeting compounds and second generation
tubulyzines.
3. Experimental procedures
3.1. Synthesis of tubulyzine affinity molecules
Unless otherwise noted, materials and solvents were
obtained from commercial suppliers (Acros and
Aldrich) and were used without further purification.
Analytical thin-layer chromatography was conducted
on SAI F254 precoated silica gel plates (250 lm layer
thickness). Chromatography was performed on Sorbent
Technologies silica gel, 60 (63–200 mesh). All
compounds were identified by an LC–MS at 250 nm
(Agilent Technology, HP1100) using a C18 column
(20 · 4.0 mm) with a gradient of 5–95% CH3CN–H2O
(containing 0.1% acetic acid) as an eluent over 4 min.
Therefore, the dual reduction of reactivity using tubuly-
zine may indicate that the tubulyzines bind to 12Cysb
and their binding inhibits the opening of the tubulin
structure. This then leads to the reactivity reduction in
239Cysb. Additionally, 12Cysb is known to be a part
of the exchangeable GTP-binding site14,17 and has been
suggested as playing an important role in the structure
and function of tubulin.18 Figure 6 shows the computa-
tional modeling of T6 resting at a GDP-binding pocket
of b-tubulin.
1H NMR spectra were measured on
AV-400 MHz spectrometer.
a Bruker
3.1.1. Synthesis of 2-chloro-4,6-di(4-methoxybenzylami-
no)-1,3,5-triazine. This procedure was adapted from US
Patent # 6,169,086.8 A suspension of cyanuric chloride
(925 mg, 5.0 mmol) in acetone (10 mL) was cooled on
ice and treated dropwise with 4-methoxybenzylamine
(1.29 mL, 9.87 mmol). After the mixture was stirred at
that temperature for 10 min, 10 mL of a 1 N sodium
hydroxide aqueous solution was added thereto dropwise
at rt, followed by stirring at rt for 15 h. The precipitate
was collected by filtration and washed with H2O, diethyl
ether, and hexane to obtain the product as a white solid.
Sixty to seventy percent yield liquid chromatography–
mass spectrometry (LC–MS) (m/z) calculated for
C19H20ClN5O2: 385.13. Found: 386.1 [M+H]+.
The 12Cysb binding site result from MS-DRC overlaps
with the previous result from the chemical affinity exper-
iment. The reactivity change seen at 376Cysa was not
significant enough to consider in this study. The results
obtained following pretreatment with TB were the same
(results not shown).
Using the previously identified tubulyzines and their
chemical affinity derivatives, we undertook a two
pronged approach to identify the binding site of tubulin.
Based on our mass spectrometric binding studies using
both chemical affinity and MS-DRC, we conclude that
the binding site of tubulyzine on the tubulin dimer is
12Cysb. This information illuminates the nature of the
tubulyzinesÕ activity, while also providing important
3.1.2. Synthesis of 2-(1,6-hexanediamine)-4,6-di(4-meth-
oxybenzylamino)-1,3,5-triazine. To a solution of 2-chlo-
ro-4,6-di(4-methoxybenzylamino)-1,3,5-triazine (100 mg,
0.25 mmol) and 1,6-hexanediamine (189 lL, 1.3 mmol)
in THF (10 mL) was added DIPEA (270 lL,
1.55 mmol). The solution was heated at 70 ꢁC for 3 h
in a heating block and the solvent was removed by vac-
uum filtration. The compound was isolated by column
chromatography in 50/50 ethyl acetate (EtOAc)/MeOH
with 1% triethylamine (TEA) Rf = 0.267. LC–MS (m/z)
calculated for C25H35N7O2: 465.29. Found: 466.1
[M+H]+.
3.1.3. Synthesis of 2-(4,7,10-trioxa-1,13-tridecan-
ediamine)-4,6-di(4-methoxybenzylamino)-1,3, 5-triazine.
To a solution of 2-chloro-4,6-di(4-methoxybenzylami-
no)-1,3,5-triazine (100 mg, 0.259 mmol) and of 4,7,10-
trioxa-1,13-tridecanediamine (285 lL, 1.3 mmol) in
THF (10 mL) was added DIPEA (270 lL, 1.554 mmol).
The solution was heated at 70 ꢁC for 3 h in a heating
block, The solvent was removed by vacuum filtration.
The compound was isolated by silica column chroma-
tography in 50/50 EtOAc/MeOH with 1% TEA.
Rf = 0.4 LC–MS (m/z) calculated for C29H43N7O5:
569.33. Found: 570.2 [M+H]+.
Figure 6. Predicted conformation of T6 docked at the C12 site of the
tubulin GDP binding pocket. Docked T6 (white) superimposed with
the crystal structure of bound GDP (green).9 The model predicts that a
methoxy-phenyl moiety of T6 (top left) occupies a side cavity next to
the GDP binding pocket generated by side-chain rearrangements of
the tubulin surface residues. This pharmacophoric feature could guide
further optimization of high affinity ligands. The molecular surface
representation of tubulin is colored by its electrostatic potential (red,
electronegative; gray, hydrophobic). T6/GDP color coding is carbon,
white/green; oxygen, red; and nitrogen, blue. Cys12 is colored as violet.

1174
Y. J. Kim et al. / Bioorg. Med. Chem. 14 (2006) 1169–1175
3.1.4. Synthesis of T6. To a solution of 2-(1,6-hexanedi-
amine) 4,6-di(4-methoxybenzylamino)-1,3,5-triazine
3.3. High-performance liquid chromatography and MS
A CapLC-coupled Q-TOF2 mass spectrometer
(50 mg, 107 lmol) in THF (5 mL) was added chloro-
acetyl chloride (12.9 lL, 162 lmol) with pyridine
(12.9 lL) in an ice bath with stirring and allowed to
rise to rt after the addition. The reaction proceeded
for 60 min and was monitored by TLC. The com-
pound was purified by prep-TLC in EtOAc.
Rf = 0.24; 1H NMR (CDCl3) d (ppm) 7.220 (d,
J = 8.0, 4H), 6.829 (d, J = 8.4, 4H) 4.499 (s, 2H),
4.132 (s, 2H), 3.760 (s, 6H), 3.589 (m, 2H), 3.299
(m, 2H), 1.539–1.738 (m, 4H), 1.256–1.432 (m, 4H).
LC–MS (m/z) calculated for C26H34ClN7O3: 541.26.
Found: 542.26 [M+H]+.
(Waters, Beverly, MA, USA) equipped with an electro-
spray ionization (ESI) source was used for HPLC–MS
analysis. A 100-min high-performance liquid chroma-
tography (HPLC) method was used with a gradient
5% solvent A (0.2% formic acid in aqueous solution)
to 90% solvent B (0.2% formic acid in acetonitrile) gra-
˚
dient. A C18 column (300 A, 0.32 · 150 mm, Microtech
Scientific, Vista, CA, USA) and a C18 precolumn
(Waters) were used for the separation. A flow rate of
6 lL/min was used in the HPLC pump, and a 6:1 split-
ter resulted in a 1 mL/min flow rate at the electrospray.
Each sample was subjected to HPLC–MS twice. A tan-
dem mass spectrometry (MS/MS) mode was used in the
first run to identify all peptides using collision-induced
dissociation (CID)-based peptide sequencing. Once all
the peptides and modified peptides were identified, a
subsequent full-scan MS mode was used in the second
run to quantify the modified and unmodified cysteine-
containing peptide pairs.
3.1.5. Synthesis of T7. To a solution of 2-(4,7,10-trioxa-
1,13-tridecanediamine)-4,6-di(4-methoxybenzylamino)-
1,3,5-triazine (50 mg, 87 lmol) in THF (5 mL) was
added chloroacetyl chloride (10.5 lL, 132 lmol) with
pyridine (10.5 lL) in an ice bath with stirring and al-
lowed to rise to rt after the addition. The reaction
proceeded for 30 min and was monitored by TLC.
The compound was purified by prep-TLC in EA/
10% MeOH. Rf = 0.53; 1H NMR (CDCl3) d (ppm)
7.205–7.216 (d, J = 8.0, 4H) 6.816–6.838 (d, J = 8.8,
4H), 4.483 (s, 4H), 4.082 (s, 2H), 3.782 (s, 6H),
3.515–3.658 (m, 12H), 1.762–1.856 (m, 4H). LC–MS
(m/z) calculated for C31H44ClN7O6: 645.3. Found:
646.4 [M+H]+.
3.4. Reaction for differential reaction yield study
Tubulin was incubated with 50 lM unmodified tubuly-
zine for 20 min at 37 ꢁC to form the protein–ligand com-
plex. The solution was then incubated with 50 lM mBrB
and processed as above.
3.2. Reaction of tubulin with tubulyzine derivatives
3.4.1. Data analysis. Tryptic digestion generates 14
cysteine-containing peptides from the major form of
brain tubulin dimer as shown in Table 1. Each peptide
and its modified entity were identified based on sequence
information obtained by the MS/MS experiment. The
retention time and molecular weight of each identified
peptide were then used for identification in the liquid
chromatography LC–MS data of the second run. A sin-
gle ion current (SIC) chromatogram of each ion pair
(cysteine-containing peptide ion and modified cysteine-
containing peptide ion) was reconstituted to extract a
selected signal that originates from the particular pep-
tide of interest from the total ion current (TIC) chro-
matogram, as shown in Figure 2. The m/z extraction
range was determined by taking the full width at half
maximum (FWHM) of the most abundant isotope peak
(m/z 0.5 FWHM) of each ion. Extracting only the sin-
gle isotope peak enhances the accuracy of quantitation
by eliminating the involvement of highly overlapped
peaks. Each peak area of the SIC chromatogram
constructed by this method was integrated using built-
in software (MassLynx 4.0, Waters Corporation) to
calculate the intensity of each peak. Reactivity yields
were calculated by dividing the intensity of the modified
peptide by the overall intensity:
Rat brain tubulin was purified from microtubule pro-
tein19 by differential polymerization as previously de-
scribed.20 Ten micromole rat brain tubulin in PIPES
buffer
(50 mM
piperazine-N,N-bis(ethanesulfonic
acid), 0.5 mM MgCl2, pH 7.0) was incubated with
50 lM T6. After incubation, the reaction was
quenched by the addition of acetic acid 0.2% (v/v).
The protein was immediately precipitated by the
addition of 9 vol of ꢀ20 ꢁC acetone. The sample
was held at 20 ꢁC for 15 min, centrifuged at
15,000g for 5 min, and all excess reagents were
removed. The precipitated tubulin was resolubilized
using sonication in 100 mM ammonium bicarbon-
ate buffer solution, pH 7.7, with a bath sonifier.
Disulfide bonds, which might be formed after
destruction of the tertiary structure of tubulin, were
reduced by 1 Tris(2-carboxyethyl)phosphine hydro-
chloride (TCEP). Additionally, 10% acetonitrile was
added for extended denaturation. The protein was
digested with sequence-grade modified trypsin (20:1
w/w) for 15 h at 37 ꢁC. Then, 5 mM TCEP was add-
ed to ensure complete reduction of the disulfide
bonds after digestion. 2-Mercaptoethanol has been
widely used as a quenching agent for sulfhydryl reac-
tion, but it was avoided in this study since 2-mercap-
toethanol binds not only to the tubulyzine affinity
derivatives (Tn) but also to cysteines resulting in a
76 Da molecular weight increase and retention time
change during the HPLC separation, which may lead
to less accurate quantification due to the splitting
and complexity of peaks.
Reactivity ¼ 100 ꢁ ½IM=ðIU þ IMÞꢂ;
where IU and IM are intensities of the unmodified pep-
ð1Þ
tide and modified peptide, respectively.
3.4.2. Computational study. T6 was docked to the guano-
sine diphosphate (GDP) binding pocket of tubulin with
ICM (Molsoft LLC, San Diego, CA). Hydrogen was

Y. J. Kim et al. / Bioorg. Med. Chem. 14 (2006) 1169–1175
1175
6. Perez, O. D.; Chang, Y. T.; Rosania, G.; Sutherlin, D.;
Schultz, P. G. Chem. Biol. 2002, 9, 475.
7. Moon, H. S.; Jacobson, E. M.; Khersonsky, S. M.;
Luzung, M. R.; Walsh, D. P.; Xiong, W.; Lee, J. W.;
Parikh, P. B.; Lam, J. C.; Kang, T. W.; Rosania, G. R.;
Schier, A. F.; Chang, Y. T. J. Am. Chem. Soc. 2002, 124,
11608.
8. Ejima, A.; Ohsuki, S.; Ohki, H.; Naito, H. U.S. Patent
6,169,086, 2001.;
9. Lowe, J.; Li, H.; Downing, K. H.; Nogales, E. J. Mol.
Biol. 2001, 313, 1045.
added to the crystal structure of the tubulin–GDP com-
plex (PDB code 1jff)9 and GDP was removed. T6 was
placed at random in the GDP pocket, and the energy
of the system was extensively minimized by a Biased
Probability Monte Carlo simulation (the conformation-
al space was sampled by 2,570,000 Monte Carlo steps) in
the internal coordinates, space10 with a flexible ligand
and flexible receptor side chains within 12 A of bound
GDP. Three independent runs with differing starting
conformations converged towards the final model.
˚
10. Totrov, M.; Abagyan, R. Proteins Suppl. 1997, 1,
215.
Acknowledgments
11. Britto, P. J.; Knipling, L.; Wolff, J. J. Biol. Chem. 2002,
277, 29018.
12. Bai, R.; Pei, X. F.; Boye, O.; Getahun, Z.; Grover, S.;
Bekisz, J.; Nguyen, N. Y.; Brossi, A.; Hamel, E. J. Biol.
Chem. 1996, 271, 12639.
13. Kim, Y. J.; Pannell, L. K.; Sackett, D. L. Anal. Biochem.
2004, 332, 376.
14. Shivanna, B. D.; Mejillano, M. R.; Williams, T. D.;
Himes, R. H. J. Biol. Chem. 1993, 268, 127.
15. Shan, B.; Medina, J. C.; Santha, E.; Frankmoelle, W. P.;
Chou, T. C.; Learned, R. M.; Narbut, M. R.; Stott, D.;
Wu, P.; Jaen, J. C.; Rosen, T.; Timmermans, P. B.;
Beckmann, H. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
5686.
Y.T.C. is grateful for financial support from NIH (R01
CA096912) and thanks the National Science Foundation
for equipment grants for the NMR (MRI-0116222) and
the Capillary LC-Ion Trap Mass spectrometer (CHE-
0234863). Components of this work were conducted in a
Shared Instrumentation Facility constructed with sup-
port from Research Facilities Improvement Grant C06
RR-16572 from the NCRR/NIH.
References and notes
16. Nogales, E. Annu. Rev. Biophys. Biomol. Struct. 2001, 30,
397.
17. Luduena, R. F.; Roach, M. C. Pharmacol. Ther. 1991, 49,
133.
18. Gupta, M. L., Jr.; Bode, C. J.; Dougherty, C. A.;
Marquez, R. T.; Himes, R. H. Cell Motil. Cytoskeleton
2001, 49, 67.
1. Vilpo, J. A.; Koski, T.; Vilpo, L. M. Eur. J. Haemato.
2000, 65, 370.
2. Smith, C. D.; Zhang, X. J. Biol. Chem. 1996, 271, 6192.
3. Poirier, V. J.; Hershey, A. E.; Burgess, K. E.; Phillips, B.;
Turek, M. M.; Forrest, L. J.; Beaver, L.; Vail, D. M. J.
Vet. Intern. Med. 2004, 18, 219.
4. Tinwell, H.; Ashby, J. Carcinogenesis 1994, 15, 1499.
5. Rosania, G. R.; Chang, Y. T.; Perez, O.; Sutherlin, D.;
Dong, H.; Lockhart, D. J.; Schultz, P. G. Nat. Biotechnol.
2000, 18, 304.
19. Sackett, D. L.; Knipling, L.; Wolff, J. Protein Expr. Purif.
1991, 2, 390.
20. Wolff, J.; Sackett, D. L.; Knipling, L. Protein Sci. 1996, 5,
2020.