Development of tubulin-polymerization inhibitors based on the thalidomide skeleton

Article abstract of DOI:10.1248/cpb.55.944

We synthesized a series of compounds based on the potent tubulin-polymerization inhibitor 5-hydroxy-2-(2,6-diisopropylphenyl)-1H- isoindole-1,3-dione [5HPP-33 (3)], which is structurally derived from thalidomide (1), and investigated their inhibitory effe

Full text of DOI:10.1248/cpb.55.944

944
Notes
Chem. Pharm. Bull. 55(6) 944—949 (2007)
Vol. 55, No. 6
Development of Tubulin-Polymerization Inhibitors Based on the
Thalidomide Skeleton
Hiroshi AOYAMA,* Tomomi NOGUCHI, Takashi MISAWA, Takanori NAKAMURA, Hiroyuki MIYACHI,
Yuichi HASHIMOTO, and Hisayoshi KOBAYASHI
Institute of Molecular & Cellular Biosciences, The University of Tokyo; 1–1–1 Yayoi, Bunkyo-ku, Tokyo 113–0032, Japan.
Received February 19, 2007; accepted April 2, 2007; published online April 4, 2007
We synthesized a series of compounds based on the potent tubulin-polymerization inhibitor 5-hydroxy-2-
(2,6-diisopropylphenyl)-1H-isoindole-1,3-dione [5HPP-33 (3)], which is structurally derived from thalidomide
(1), and investigated their inhibitory effects on tubulin polymerization. Direct interaction between 5HPP-33 (3)
and a,b-tubulin heterodimer protein was demonstrated by means of a surface plasmon resonance study.
Key words thalidomide; phthalimide skeleton; tubulin polymerization inhibitor; surface plasmon resonance
Thalidomide (1) (Fig. 1) was developed and marketed in
We also reported that 5HPP-33 (3) exerts cell growth-in-
the late 1950’s as a sedative/hypnotic drug, but subsequently hibitory activity by causing cell cycle arrest and inducing
had to be withdrawn from sale due to its teratogenicity.1—6) apoptosis of human myelocytic cells IM9.11) Additionally, we
However, it was subsequently identified as an effective agent recently reported that fluorinated phthalimide analogs, FPP-
for the treatment of multiple myeloma (MM), AIDS, leprosy, 33 (4), 4FPP-33 (5), and 4,7FPP-33 (6), exhibit potent tubu-
and various cancers.1—5,7,8) The United States Food and Drug lin polymerization-inhibitory activity (Fig. 2).17) However,
Administration (FDA) approved it for the treatment of ery- whether these tubulin polymerization inhibitors, including
thema nodosum in leprosy (ENL) in 1998. In addition, offi- 5HPP-33 (3), bind/interact with a,b-tubulin heterodimer
cial approval for the use of thalidomide (1) to treat MM has protein remains unclear.
been applied for in Japan.
Therefore, we designed a hybrid compound of 5HPP-33
Thalidomide (1) has been discovered to have various (3) and FPP-33 (4), i.e., 5HFPP-33 (9), for which potent
biological activities, such as inhibition of tumor necrosis tubulin polymerization-inhibitory activity is expected. In this
factor-a (TNF-a) production, and anti-inflammatory, anti- paper, we describe (1) structural development studies of
angiogenic, and cyclooxygenase (COX)-inhibitory activi- 5HPP-33 (3) and its fluorinated derivatives based on tubulin
ties.1—5,9,10) In the course of our studies of thalidomide polymerization-inhibitory activity, (2) analysis of the tubulin
metabolites, we found that 5-hydroxythalidomide (2) (Fig. 1) polymerization-inhibitory activity of 5HPP-33 (3), and (3) an
exhibited tubulin polymerization-inhibitory activity, though experimental analysis of the interaction between 5HPP-33
thalidomide (1) itself does not.11—15) Inhibition of tubulin (3) and a,b-tubulin heterodimer protein by means of a sur-
function, including tubulin polymerization, is considered to face plasmon resonance (SPR) study.
be one of the molecular mechanisms of anti-tumor agents
such as vinblastine and paclitaxel. Therefore, it was sus- Results and Discussion
pected that the efficacy of thalidomide (1) for the treatment
Chemistry The tubulin-polymerization inhibitor, 5HPP-
of MM might be due to the tubulin polymerization-inhibitory 33 (3), was prepared as previously reported.11) Compounds
activity of its metabolite, 5-hydroxythalidomide (2), at least 7a—h and 9 were synthesized by usual organic synthetic
in part. During our structural development studies of 5-hy- methods or reported methods (Charts 1, 2). Briefly, conden-
droxythalidomide (2) based on tubulin polymerization-in- sation of 5-carboxyphthalic anhydride with neat 2,6-diiso-
hibitory activity, we obtained 5HPP-33 (3), which exhibits propylaniline gave TCA-33 (7a), then condensation with am-
potent inhibitory activity, comparable with that of rhizoxin or monia, methylamine, and dimethylamine in the presence of
colchicine.11) However, Li et al. recently reported that 5HPP- N-ethyl-Nꢀ-(3-dimetylaminopropyl)carbodiimide (EDCI), 1-
33 (3) is not a tubulin polymerization inhibitor,16) but a tubu- hydroxybenzotriazole (HOBt), and N,N-diisopropylethyl-
lin polymerization enhancer, which prompted us to re-exam- amine (DIPEA) afforded TAD-33 (7b), TDM-33 (7c), and
ine the effects elicited by 5HPP-33 (3) on tubulin polymer- TDMM-33 (7d), respectively. Reduction of 7a with
ization/depolymerization, which led us to conclude that BH3·THF gave the 5-hydroxymethyl analog TOL-33 (7e),
5HPP-33 (3) acts a tubulin plolymerization inhibitor, at least
in our experimental conditions.
Fig. 1. Structures of Thalidomide (1) and Its Metabolite 5-Hydroxy- Fig. 2. Potent Tubulin Polymerization Inhibitors with a Phthalimide Skele-
thalidomide (2)
ton Derived from Thalidomide (1)
∗ To whom correspondence should be addressed. e-mail: aoyama@iam.u-tokyo.ac.jp
© 2007 Pharmaceutical Society of Japan

June 2007
945
Reagents and conditions: (a) 2,6-diisopropylaniline, 200 °C, 2 h (32%); (b) NHR1R2, EDCI, HOBt, DIPEA, RT, 10 h (8—18%); (c) BH3·THF, THF, RT, overnight (62%); (d)
MnO2, CH2Cl2, RT, 24 h (60%); (e) HONH2·HCl, pyridine, EtOH, RT, 4 h (99%); (f) 10% Pd/C, H2 (0.3 MPa), 2 N HCl, EtOH, RT, 1.5 h, (9%).
Chart 1
Reagents and conditions: (a) KOH, H2O, 90 °C, 9 h (85%); (b) Ac2O, 100 °C, 10 min; (c) 2,6-diisopropylaniline, pyridine, 120 °C, 4 h (19%).
Chart 2
which was oxidized with MnO2 to afford the 5-formyl analog
TAL-33 (7f). Condensation of 7f with hydroxylamine hy-
drochloride in the presence of pyridine gave the 5-hydroxy-
imino analog TOX-33 (7g), whose hydroxylamino group was
reduced with hydrogen gas over Pd/C to afford the corre-
sponding 5-aminomethyl analog TA-33 (7h) (Chart 1). The
fluorinated compound 5HFPP-33 (9) was synthesized from
4,5,6,7-tetrafluorophthalic anhydride. Replacement of fluo-
rine with a hydroxyl group at the 5-position of 4,5,6,7-tetra-
fluorophthalic anhydride was performed by treatment with
KOH aqueous solution to give 5-hydroxy-4,6,7-trifluoro-
phthalic acid (8).18,19) Compound 8 was dehydrated with
Ac2O, followed by condensation with 2,6-diisopropylaniline
in pyridine to give 5HFPP-33 (9) (Chart 2).
Fig. 3. Tubulin Polymerization-Inhibitory Activity of Compounds 7a—h,
Inhibitory Effects of the Prepared Compounds on 9, and 3
Tubulin Polymerization We examined the tubulin poly-
merization-inhibitory activity of the prepared compounds centration of the test compounds was 50 mM, except for
(7a—h, 9) using the method previously described.20,21) Mi- 5HFPP-33 (9, 20 mM). As shown in Fig. 3, although TDMM-
crotubule protein was prepared from porcine brain.22,23) 33 (7d), TOL-33 (7e), and TA-33 (7h) exhibited little or no
Tubulin polymerization was followed by means of turbidity tubulin polymerization-inhibitory activity, other compounds
measurements at 37 °C in microtubule assemble buffer con- showed moderate to high activity. The inhibitory activity of
taining 100 mM 2-morpholinoethanesulfonic acid (MES), the amide compounds 7b—d was influenced by the number
1 mM EGTA, 0.5 mM MgCl2, 1 mM 2-mercaptoethanol, and of methyl groups on the nitrogen atom of amide group. The
1 mM GTP (pH 6.5). Although the measured turbidity inhibitory activity of the compounds decreased in the order
(Abs.400) values of polymerized tubulin differed from experi- of non-substituted compound TAD-33 (7b; ca. 85%)ꢁmono-
ment to experiment, the results (calculated IC50 values of the substituted compound TDM-33 (7c; ca. 60%)ꢁdi-substituted
compounds) were basically reproducible. A typical set of compound TDMM-33 (7d; 0%). The compounds bearing hy-
data is presented in Figs. 3 and 4. In these figures, the con- droxymethyl and aminomethyl moieties, TOL-33 (7e) and

946
Vol. 55, No. 6
Fig. 4. Dose-Dependence of Inhibition of Tubulin Polymerization by
TAD-33 (7b), TOX-33 (7g), and 5HFPP-33 (9)
Fig. 5. Effects of Potent Tubulin Polymerization Inhibitors with a Phthal-
imide Skeleton on Tubulin Depolymerization
TA-33 (7h), exhibited low inhibitory effects on tubulin poly-
merization, amounting to about 10% and 15%, respectively. tubulin of our novel polymerization inhibitors, we performed
Compounds substituted with a carbamoyl [TAD-33 (7b)], a binding competition studies using commercially available
formyl [TAL-33 (7f)], or a hydroxyimino group [TOX-33 radio-labeled colchicine and vinblastine, but no binding com-
(7g)], showed more potent inhibitory activity, amounting to petition was observed (data not shown). The results sug-
about 85%, 80%, and 75%, respectively.
gested that these molecules do not bind to the site at which
As expected, 5HFPP-33 (9), which is a tri-fluorinated de- colchicine or vinblastine binds.
rivative of 5HPP-33 (3), was found to be the most potent in-
Recently, Li et al. reported the effect of 5HPP-33 (3) on
hibitor of tubulin polymerization among our compounds. As tubulin polymerization/depolymerization.16) They found that
shown in Fig. 4, 5HFPP-33 (9) was also a potent inhibitor, 5HPP-33 (3) does not inhibit tubulin polymerization, but en-
with the inhibition amounting to 82% at 7.5 mM. Its IC50 hances the polymerization and inhibits cold-induced tubulin
value was calculated to be about 5 mM, whereas those of depolymerization. The reason for this conflict between the
TAD-33 (7b) and TOX-33 (7g) were about 16 mM and 11 mM, results of Li et al. and our findings is not yet clear. However,
respectively. These results indicate that compounds substi- there were a number of differences between the experimental
tuted with a protic or a carbon–heteroatom double bond conditions in the two studies: (1) Li et al. used purified a,b-
(CꢂO or CꢂN)-containing functional group at the 5-posi- tubulin heterodimer protein, while we used microtubule pro-
tion showed moderate to high tubulin polymerization-in- tein, i.e., microtubule-associated proteins (MAPs)-containing
hibitory activity. In addition, introduction of fluorine atoms tubulin fraction, (2) the incubation buffer of Li et al. con-
on the phthalimide skeleton enhanced the inhibitory potency. tained a high concentration of DMSO (12% v/v), while ours
Next, we investigated the tubulin depolymerization-in- contained less than 2% v/v DMSO, and (3) the incubation
hibitory activity of TOX-33 (7g) and 5HFPP-33 (9). Depoly- buffer of Li et al. contained a high concentration of gluta-
merization of tubulin was induced by cooling of the tubulin mate (0.4 M), while ours did not contain glutamate.
fraction polymerized at 37 to 0 °C, or by addition of 4 mM
Therefore, we first compared the polymerization profile of
CaCl2 at 37 °C, and was followed by means of turbidity purified tubulin and MAPs-containing tubulin (Fig. 6A). We
measurements as described previously.20—22) As shown in found that purified tubulin without MAPs hardly polymerizes
Fig. 5, once-polymerized tubulin was depolymerized by cool- in the absence of a high concentration of DMSO (Fig. 6A,
ing at 0 °C (Fig. 5, black circles). The tubulin depolymerized line 4). The observed increase of turbidity was ca. 0.025
by cooling could be re-polymerized again by warming at OD400/1—2 mg protein/ml (Li et al. measured the turbidity at
37 °C. The addition of 15 mM TOX-33 (7g) or 10 mM 5HFPP- OD351, and the increase was 0.04—0.09 judging from their
33 (9) to polymerized tubulin at 37 °C did not affect the cool- data16)). The turbidity increase at the same concentration of
ing-induced depolymerization step (Fig. 5, white squares and MAPs-containing tubulin fraction generally reached 0.24
white circles, respectively). As expected, the tubulin depoly- OD400 (this value is defined as 100% in Figs. 6A—C). The
merized by cooling did not re-polymerize so effectively upon inability of purified tubulin protein to polymerize/assemble
being warmed in the presence of TOX-33 (7g) or 5HFPP-33 has been already reported.24) Addition of DMSO at the final
(9) (at least, they did not enhance the re-polynerization) (Fig. concentration of 12% v/v to purified tubulin resulted in rapid
5, white squares and white circles, respectively). Thus, TOX- and efficient polymerization, as indicated by the turbidity in-
33 (7g) and 5HFPP-33 (9) showed no inhibitory effect on crease (in Fig. 6A, line 2). This phenomenon has also been
tubulin depolymerization, and they themselves do not induce reported by Himes et al.24,25) Interestingly, addition of 12%
tubulin depolymerization, at least in our system. Our findings v/v DMSO moderately inhibited the polymerization of
suggest that TOX-33 (7g) and 5HFPP-33 (9) bind only to MAPs-containing tubulin (Fig. 6A. line 3 vs. line 1). Our ten-
a,b-tubulin heterodimer protein to inhibit tubulin polymer- tative interpretation/speculation is that MAPs are mandatory
ization, but not to polymerized tubulin. These profiles of the for tubulin polymerization, as reported,17) but that DMSO can
effects of the compounds on tubulin polymerization/depoly- partially and functionally substitute for MAPs as regards
merization are quite similar with those that we previously ob- both interaction with tubulin and function as a tubulin poly-
served for 5HPP-33 (3).17) To examine the binding site(s) on merization helper.24,25) This effect of 12% v/v DMSO on pu-

June 2007
947
Fig. 6. Effects of 5HPP-33 (3) and Additives on Tubulin Polymerization
(A) Comparison of purified tubulin and MAPs-containing tubulin, and effect of DMSO. Line 1: MAPs-containing tubulin. Line 2: purified tubulin in the presence of 12% v/v
DMSO. Line 3: MAPs-containing tubulin in the presence of 12% v/v DMSO. Line 4: purified tubulin. (B) Effects of 5HPP-33 (3) and paclitaxel on polymerization of purified tubu-
lin in the presence or absence of DMSO. Line 5: 5 mM paclitaxel and 12% v/v DMSO. Line 6: 12% v/v DMSO. Line 7: 5 mM paclitaxel. Line 8: 40 mM 5HPP-33 (3) and 12% v/v
DMSO. Line 9: no additive. Line 10: 40 mM 5HPP-33 (3). (C) Effects of 5HPP-33 (3) and 5HFPP-33 (9) on polymerization of purified tubulin in the presence of 0.4 M glutamate
and in the presence or absence of DMSO. Line 11: 12% v/v DMSO. Line 12: 20 mM 5HPP-33 (3) and 12% v/v DMSO. Line 13: 40 mM 5HPP-33 (3) and 12% v/v DMSO. Line 14:
no additive. Line 15: 40 mM 5HFPP-33 (9).
rified tubulin polymerization was also observed in the pres-
ence of 0.4 M glutamate (Fig. 6C, line 11). Addition of 0.4 M
glutamate alone did not enhance the polymerization of puri-
fied tubulin (Fig. 6C, line 14).
Paclitaxel (5 mM) showed potent polymerization-enhancing
effects on purified tubulin (Fig. 6B, line 7), which is in good
accordance with the results reported by Li et al.,16) though its
effect is less potent than that of 12% v/v DMSO (Fig. 6B,
line 6). The effects of paclitaxel and DMSO seemed to be ad-
ditive (Fig. 6B, line 5). However, 5HPP-33 (3) seemed to in-
hibit polymerization of purified tubulin in the presence of
12% v/v DMSO (Fig. 6, line 8), and dose-dependency for the
inhibition of polymerization of purified tubulin was observed
Fig. 7. SPR Measurement of Tubulin-5HPP-33 (3) Interaction
in the presence of 12% v/v DMSO and 0.4 M glutamate (Fig.
6C, lines 12, 13). No depolymerization-inhibitory activity of
5HPP-33 (3) was observed under any of the conditions de-
5HPP-33 (3) was injected at the time indicated by arrows “A” (100 mM), “B”
(200 mM) and “C” (400 mM). Washing procedures were started at the times indicated
by arrows “W”.
scribed above (data not shown). Although precise analysis is
still required, the effect of DMSO and/or MAPs cannot be incubation with a,b-tubulin heterodimer protein. After im-
excluded.
mobilization of the protein, unreacted succinimide terminals
a b
Binding Analysis of 5HPP-33 (3) with , -Tubulin Het- on the sensor chip were blocked with ethanolamine. When
erodimer Although 5HPP-33 (3) strongly inhibited tubulin 5HPP-33 (3) was injected onto the a,b-tubulin heterodimer
polymerization under our experimental conditions (especially protein-bearing sensor chip, a distinct, dose-dependent in-
for MAPs-containing tubulin), the presence or absence of in- crease of the degree of difference line was detected (Fig. 7,
teraction between 5HPP-33 (3) and a,b-tubulin heterodimer arrows A, B, C). Washing the sensor chip with running
protein remained unclear (vide supra). Thus, to examine this buffer caused disappearance of the increase (Fig. 7, arrow
issue, we conducted surface plasmon resonance (SPR) meas- W), suggesting dissociation of 5HPP-33 (3) from the tubulin-
urements using a SPR670 instrument (Nippon Laser & Elec- bearing chip. These results indicate that 5HPP-33 (3)
tronics Lab.). For the analysis, the sensor chip was treated binds/interacts directly and reversibly with a,b-tubulin het-
with 4,4-dithiodibutyric acid (DDA) to introduce a linker erodimer protein. From the SPR analysis, the binding con-
segment for connection to tubulin protein. Immobilization of stant of 5HPP-33 (3) with a,b-tubulin heterodimer protein
ꢄ1
a,b-tubulin heterodimer protein onto the sensor chip with was calculated to be 4.5ꢃ106 M under our experimental
the sulfanylbutyric acid-modified surface was carried out by conditions. The calculated value is in good agreement with
treatment of the modified sensor chip with N-hydroxysuccin- the IC50 value of 5HPP-33 (3) for inhibition (6.9—
imide and EDCI in 100 mM MES buffer solution, followed by 7.9ꢃ106 Mꢄ1 11,17) of tubulin polymerization.
)

948
Vol. 55, No. 6
luted by adding ethyl acetate, washed successively with dil. HCl, saturated
aqueous NaHCO3, and brine, dried over MgSO4, and evaporated under
reduced pressure. The residue was purified by silica-gel column chroma-
tography (hexane/ethyl acetateꢂ1/1) to give TDMM-33 (7d) (20.0 mg,
0.053 mmol, 19%) as a white powder; mp 134 °C. 1H-NMR (CDCl3) d: 1.15
Conclusion
We examined the tubulin polymerization-inhibitory activ-
ity of a series of phthalimide analogs (7a—h, 9). Analogs
bearing a substituent with protic character (TAD-33; 7b) or
having a carbon–heteroatom double bond-containing func- (12H, d, Jꢂ6.8 Hz), 2.68 (2H, hept, Jꢂ6.8 Hz), 3.05 (3H, s), 3.17 (3H, s),
tional group (TAL-33; 7f, TOX-33; 7g) at the 5-position
7.29 (2H, d, Jꢂ8.1 Hz), 7.46 (1H, t, Jꢂ8.1 Hz), 7.85 (1H, dd, Jꢂ1.3,
7.7 Hz), 7.98 (1H, d, Jꢂ1.3 Hz), 8.00 (1H, d, Jꢂ7.7 Hz). FAB-MS m/z:
379.2017 (Calcd for C23H27N2O3: 379.2022).
5-Hydroxymethyl-2-(2,6-diisopropylphenyl)-1H-isoindole-1,3-dione
showed potent inhibitory activity. 5HFPP-33 (9), which is a
fluorinated derivative of 5HPP-33 (3), also exhibited potent
activity. In SPR experiments, we observed direct binding/in-
teraction of 5HPP-33 (3) with a,b-tubulin heterodimer pro-
tein. Further structural development studies, investigation of
the molecular mechanisms involved, and studies to identify
the binding site(s) are in progress.
(TOL-33, 7e) To a solution of TCA-33 (7a) (877 mg, 2.50 mmol) in THF
(10 ml) was added BH3·THF (1.0 M THF solution, 3.0 ml, 3.00 mmol) at
0 °C, and the mixture was stirred at ambient temperature overnight. The re-
action was quenched by adding diluted HCl. The mixture was extracted with
ethyl acetate, washed with brine, dried over MgSO4, and evaporated under
reduced pressure. The residue was purified by silica-gel column chromatog-
raphy (hexane/ethyl acetateꢂ1/1) to give TOL-33 (7e) (357 mg, 1.06 mmol,
42%) as a white powder; mp 136 °C. 1H-NMR (CDCl3) d: 1.15 (6H, d,
Experimental
General 1H-NMR (500 MHz) and 13C-NMR (125 MHz) spectra were Jꢂ6.8 Hz), 1.16 (6H, d, Jꢂ6.8 Hz), 2.15 (1H, t, Jꢂ4.7 Hz), 2.71 (2H, hept,
recorded on a JEOL JNM-a 500 spectrometer. Mass spectra were obtained Jꢂ6.8 Hz), 4.89 (2H, d, Jꢂ4.7 Hz), 7.29 (2H, d, Jꢂ7.7 Hz), 7.46 (1H, t,
on JEOL JMA-HX 110 spectrometer with m-nitrobenzyl alcohol. Melting
points determined on Yanaco MP-J3 micro melting point apparatus are un-
corrected. Flash column chromatography was performed on silica gel 60
Kanto Kagaku (40—100 mm).
Jꢂ7.7 Hz), 7.81 (1H, d, Jꢂ7.5 Hz), 7.95 (1H, d, Jꢂ7.5 Hz), 7.99 (1H, s).
FAB-MS m/z: 338.1711 (Calcd for C21H24NO3: 338.1756).
5-Formyl-2-(2,6-diisopropylphenyl)-1H-isoindole-1,3-dione (TAL-33,
7f) To a solution of TOL-33 (7e) (357 mg, 1.06 mmol) in CH2Cl2 (10 ml)
5-Carboxy-2-(2,6-diisopropylphenyl)-1H-isoindole-1,3-dione (TCA- was added MnO2 (1.00 g, 11.5 mmol), and the mixture was stirred at ambi-
33, 7a) A mixture of trimellitic anhydride (192 mg, 1.00 mmol) and 2,6-di-
isopropylaniline (192 mg, 1.00 mmol) was stirred at 200 °C for 2 h, then
cooled to room temperature. The crude product was purified by silica-gel
column chromatography (hexane/ethyl acetateꢂ1/8) and recrystallized from
ent temperature overnight. The reaction mixture was filtered through a pad
of Celite, which was then washed with CH2Cl2, and the combined filtrate
and washing was concentrated under reduced pressure. The residue was pu-
rified by silica-gel column chromatography (hexane/ethyl acetateꢂ3/1) to
give TAL-33 (7f) (310 mg, 0.924 mmol, 87%) as a white powder; mp
hexane and ethyl acetate to give TCA-33 (7a) (113 mg, 0.320 mmol, 32%) as
1
a white powder; mp 234 °C (sublimation). H-NMR (CDCl3) d: 1.17 (12H, 153 °C. 1H-NMR (CDCl3) d: 1.17 (12H, d, Jꢂ6.8 Hz), 2.68 (2H, hept,
d, Jꢂ6.8 Hz), 2.69 (2H, hept, Jꢂ6.8 Hz), 7.31 (2H, d, Jꢂ7.7 Hz), 7.49 (1H,
t, Jꢂ7.7 Hz), 8.10 (1H, d, Jꢂ7.9 Hz), 8.59 (1H, d, Jꢂ7.9 Hz), 8.71 (1H, s).
Jꢂ6.8 Hz), 7.31 (2H, d, Jꢂ7.7 Hz), 7.48 (1H, t, Jꢂ7.7 Hz), 8.15 (1H, d,
Jꢂ7.7 Hz), 8.35 (1H, d, Jꢂ7.7 Hz), 8.47 (1H, s), 10.22 (1H, s). FAB-MS
Anal. Calcd for C21H21NO4: C, 71.78; H, 6.02; N, 3.99. Found: C, 71.52; H, m/z: 336.1602 (Calcd for C21H22NO3: 336.1600).
6.15; N, 3.89.
5-(Hydroxyimino)methyl-2-(2,6-diisopropylphenyl)-1H-isoindole-1,3-
5-Carbamoyl-2-(2,6-diisopropylphenyl)-1H-isoindole-1,3-dione (TAD- dione (TOX-33, 7g) To a solution of TAL-33 (7f) (40.0 mg, 0.119 mmol)
33, 7b) To a solution of TCA-33 (7a) (100 mg, 0.285 mmol) in DMF
in a mixture of EtOH (15 ml) and pyridine (8 ml) was added hydroxylamine
(20 ml) were added 1-ethyl-3-(3ꢀ-N,N-dimethylaminopropyl)carbodiimide, hydrochloride (8.1 mg, 0.117 mmol). The mixture was stirred at 90 °C for
hydrochloride (EDCI) (110 mg, 0.574 mmol), 1-hydroxybenzotriazole 4 h, then cooled to room temperature, diluted with ethyl acetate, and washed
(HOBt) (81.0 mg, 0.599 mmol), diisopropylethylamine (151 mg, 1.17 mmol),
and ammonium chloride (30.0 mg, 0.561 mmol). The mixture was stirred at
successively with 2 N HCl, saturated aqueous NaHCO3, H2O, and brine. The
organic layer was dried over MgSO4 and concentrated under reduced
ambient temperature overnight, diluted by adding ethyl acetate, washed suc- pressure. The residue was purified by silica-gel column chromatography
cessively with dil. HCl, saturated aqueous NaHCO3, and brine, dried over (hexane/ethyl acetateꢂ5/1) to give TOX-33 (7g) (35.0 mg, 0.100 mmol,
MgSO4, and evaporated under reduced pressure. The residue was purified by 84%) as a white powder; mp 215 °C (sublimation). 1H-NMR (CDCl3) d:
silica-gel column chromatography (hexane/ethyl acetateꢂ1/1) to give TAD- 1.16 (12H, d, Jꢂ6.8 Hz), 2.70 (2H, hept, Jꢂ6.8 Hz), 7.29 (2H, d, Jꢂ7.7 Hz),
33 (7b) (8.0 mg, 0.023 mmol, 8%) as a white powder; mp 181 °C (sublima- 7.46 (1H, t, Jꢂ7.7 Hz), 7.56 (1H, s), 7.99 (1H, d, Jꢂ7.9 Hz), 8.00 (1H, d,
tion). 1H-NMR (CDCl3) d: 1.10 (12H, d, Jꢂ6.8 Hz), 2.61 (2H, hept,
Jꢂ6.8 Hz), 5.69 (1H, bs), 6.13 (1H, bs), 7.23 (2H, d, Jꢂ7.8 Hz), 7.41 (1H, t,
Jꢂ7.8 Hz), 8.00 (1H, d, Jꢂ7.7 Hz), 8.27 (1H, s), 8.28 (1H, d, Jꢂ7.7 Hz).
Jꢂ7.9 Hz), 8.21 (1H, s), 8.27 (1H, s). FAB-MS m/z: 351.1731 (Calcd for
C21H23N2O3: 351.1709).
5-Aminomethyl-2-(2,6-diisopropylphenyl)-1H-isoindole-1,3-dione
Anal. Calcd for C21H23N2O3: C, 71.98; H, 6.33; N, 7.99. Found: C, 71.86; H, (TA-33, 7h) To a solution of TOX-33 (7g) (25.0 mg, 0.071 mmol) in a
6.48; N, 7.76.
mixture of EtOH (20 ml) and 2 N HCl (1 ml) was added 10% Pd–C (10 mg),
and the mixture was stirred at ambient temperature for 1.5 h under an H2
atmosphere (0.3 MPa). The reaction mixture was filtered through a pad of
5-(N-Methylamino)carbonyl-2-(2,6-diisopropylphenyl)-1H-isoindole-
1,3-dione (TDM-33, 7c) To
a solution of TCA-33 (7a) (100 mg,
0.285 mmol) in DMF (20 ml) were added 1-ethyl-3-(3ꢀ-N,N-dimethylamino- Celite and the filtrate was concentrated under reduced pressure. The residue
propyl)carbodiimide, hydrochloride (EDCI) (110 mg, 0.574 mmol), 1-hy- was taken up in Et2O and filtered. The filtrate was concentrated under re-
droxybenzotriazole (HOBt) (81 mg, 0.599 mmol), diisopropylethylamine duced pressure, and the residue was purified by recrystallization (hexane and
(151 mg, 1.17 mmol), and methylamine (40 wt% in water, 50.0 mg, ethyl acetate) to give TA-33 (7h) (8.0 mg, 0.024 mmol, 33%) as a white
1
0.644 mmol). The mixture was stirred at ambient temperature overnight, di- powder; mp 170 °C. H-NMR (D2O) d: 1.01 (12H, d, Jꢂ6.8 Hz), 2.64 (2H,
luted by adding ethyl acetate, washed successively with dil. HCl, saturated hept, Jꢂ6.8 Hz), 4.32 (2H, s), 7.35 (2H, d, Jꢂ8.1 Hz), 7.50 (1H, t,
aqueous NaHCO3, and brine, dried over MgSO4, and evaporated under re- Jꢂ8.1 Hz), 7.89 (1H, d, Jꢂ7.9 Hz), 7.95 (1H, s), 7.96 (1H, d, Jꢂ7.9 Hz).
duced pressure. The residue was purified by silica-gel column chromatogra- FAB-MS m/z: 337.1951 (Calcd for C21H25N2O2: 337.1916).
phy (hexane/ethyl acetateꢂ1/1) to give TDM-33 (7c) (8.0 mg, 0.022 mmol,
4-Hydroxy-3,5,6-trifluorophthalic Acid (8)18,19) To a solution of 85%
8%) as a white powder; mp 163 °C. 1H-NMR (CDCl3) d: 1.16 (6H, d, KOH (2.0 g, 30.3 mmol) in H2O (10 ml) was added 3,4,5,6-tetrafluoro-
Jꢂ6.8 Hz), 1.17 (6H, d, Jꢂ6.8 Hz), 2.68 (2H, hept, Jꢂ6.8 Hz), 3.00 (3H, d, phthalic anhydride (1.0 g, 4.54 mmol) at 90 °C. The reaction mixture was
Jꢂ5.1 Hz), 6.43 (1H, bs), 7.31 (2H, d, Jꢂ7.7 Hz), 7.47 (1H, t, Jꢂ7.7 Hz), stirred for 9 h at the same temperature, then cooled to room temperature.
8.04 (1H, d, Jꢂ7.7 Hz), 8.32 (1H, d, Jꢂ7.7 Hz), 8.33 (1H, s). FAB-MS m/z: The reaction was quenched by adding conc. HCl to ca. pH 2. The resulting
365.1842 (Calcd for C22H25N2O3: 365.1865).
5-(N,N-Dimethylamino)carbonyl-2-(2,6-diisopropylphenyl)-1H-isoin-
dole-1,3-dione (TDMM-33, 7d) To a solution of TCA-33 (7a) (100 mg,
acidic solution was extracted with ethyl acetate, and the organic solution was
washed with brine, dried over MgSO4, and concentrated under reduced pres-
sure to give 8 (0.912 g, 85%) as a white powder. 1H-NMR (DMSO-d6) d:
0.285 mmol) in DMF (20 ml) were added 1-ethyl-3-(3ꢀ-N,N-dimethylamino- 11.89 (2H, bs). 13C-NMR (CD3OD) d: 111.5 (d, JC–Fꢂ13.0 Hz), 120.1 (d,
propyl)carbodiimide, hydrochloride (EDCI) (110 mg, 0.574 mmol), 1-hy- 2JC–Fꢂ16.6 Hz), 140.2 (dd, 2JC–Fꢂ14.8, 14.8 Hz), 143.2 (dd, 1JC–Fꢂ247.8 Hz,
droxybenzotriazole (HOBt) (81 mg, 0.599 mmol), diisopropylethylamine 2JC–Fꢂ11.1 Hz), 146.9 (d, 1JC–Fꢂ246.0 Hz), 147.8 (dd, 1JC–Fꢂ253.4 Hz,
(151 mg, 1.17 mmol), and dimethylamine (40 wt% in water, 62.5 mg, 2JC–Fꢂ11.1 Hz), 165.4 (s), 166.4 (s). FAB-MS m/z: 332 (MꢅHꢅ).
2
0.555 mmol). The mixture was stirred at ambient temperature overnight, di-
5-Hydroxy-4,6,7-trifluoro-2-(2,6-diisopropylphenyl)-1H-isoindole-1,3-

June 2007
949
dione (5HFPP-33, 9) A solution of 4-hydroxy-3,5,6-trifluorophthalic acid ethanolamine (i.e., sensor chip without tubulin protein) did not afford in-
(8) (1.00 g, 4.24 mmol) in Ac2O (1.0 ml) was stirred at 100 °C for 10 min crease of the degree of difference line, eliminating the possibility that 5HPP-
and the solvent was removed under reduced pressure. The residue was dis- 33 itself gives false positives.
solved in pyridine (25 ml), and to this solution was added 2,6-diisopropyl-
aniline (752 mg, 4.24 mmol). The mixture was stirred at 120 °C for 4 h, then
cooled to room temperature, diluted with ethyl acetate, and washed with 1 N
Acknowledgments The work described in this paper was partially sup-
ported by Grants-in Aid for Scientific Research from The Ministry of Edu-
HCl. The organic layer was dried over MgSO4 and concentrated under re- cation, Culture, Sports, Science and Technology, Japan, and the Japan Soci-
duced pressure. The residue was purified by silica-gel column chromatogra- ety for the promotion of Science. The authors are grateful to Mr. Yoshio
phy (ethyl acetate) and then recrystallized (hexane and ethyl acetate) to give Kondo (Moritex Corporation) for helpful discussions and technical assis-
5HFPP-33 (9) (303 mg, 0.803 mmol, 19%) as a white powder; mp 199 °C.
tance with the SPR analysis.
1H-NMR (CDCl3) d: 1.16 (12H, d, Jꢂ6.8 Hz), 2.66 (2H, hept, Jꢂ6.8 Hz),
7.27 (2H, d, Jꢂ7.7 Hz), 7.41 (1H, t, Jꢂ7.7 Hz). Anal. Calcd for References and Notes
C20H19F3NO3: C, 63.66; H, 4.81; N, 3.71. Found: C, 63.31; H, 4.74; N, 3.60.
Tubulin Polymerization Measurement Microtubule protein was pre-
pared from porcine brain as described.22,23) A typical run was as follows:
fresh brains were cooled on ice, washed with aqueous solution containing
100 mM 2-morpholinoethanesulfonic acid (MES), 1 mM EGTA, 0.5 mM
MgCl2, pH 6.5, and homogenized in 0.5 ml/g of MES buffer (100 mM MES,
1 mM EGTA, 0.5 mM MgCl2, 1 mM 2-mercaptoethanol, and 1 mM GTP, pH
6.5) under ice-cooling. After centrifugation (100000 g, 30 min, 4 °C), the su-
pernatant was mixed with an equal volume of glycerol buffer (MES buffer
and 8 M glycerol, pH 6.5), and the mixture was warmed at 37 °C for 30 min
to polymerize tubulin. The polymerized microtubule protein was collected
by centrifugation (100000 g, 45 min, 25 °C) and resuspended in MES buffer,
and the suspension was chilled (0 °C, 30 min) to allow depolymerization.
After centrifugation of the suspension (100000 g, 60 min, 4 °C), the super-
1) Bartlett J. B., Dredge K., Dalglish A. G., Nat. Rev. Cancer, 4, 314—
322 (2004).
2) Bartlett J. B., Tozer A., Stirling D., Zeldis J. B., Brit. J. Cancer, 93,
613—619 (2005).
3) Hashimoto Y., Bioorg. Med. Chem., 10, 461—479 (2002).
4) Hashimoto Y., Curr. Med. Chem., 5, 163—178 (1998).
5) Hashimoto Y., Tanatani A., Nagasawa K., Miyachi H., Drugs Future,
29, 383—391 (2004).
6) Hales B. F., Nat. Med., 5, 489—490 (1999).
7) Calbrese L., Fleisher A. B., Am. J. Med., 108, 487—491 (2000).
8) Richardson P., Hideshima T., Anderson K., Annu. Rev. Med., 53,
629—657 (2002).
9) D’Amato R. J., Loughnan M. S., Flynn E., Folkman J., Proc. Natl.
Acad. Sci. U.S.A., 91, 4082—4085 (1994).
natant was collected. The same polymerization/depolymerization process 10) Bauer K. S., Dixon S. C., Figg W. D., Biochem. Pharmacol., 55,
was repeated, then an equal volume of glycerol buffer was added to the su-
pernatant solution, and the mixture was stored at ꢄ80 °C. The microtubule
protein thus obtained was used as MAPs-containing tubulin for polymeriza-
tion/depolymerization assay after the polymerization/depolymerization
1827—1834 (1998).
11) Inatsuki S., Noguchi T., Miyachi H., Oda S., Iguchi T., Kizaki M.,
Hashimoto Y., Kobayashi H., Bioorg. Med. Chem. Lett., 15, 321—325
(2005).
process described above was once more repeated. This tubulin fraction was 12) Noguchi Y., Shinji C., Kobayashi H., Makishima M., Miyachi H.,
further purified by phosphocellulose column chromatography22) and its pu-
Hashimoto Y., Biol. Pharm. Bull., 28, 563—564 (2005).
rity was evaluated by polyacrylamide gel electrophoresis. It was then used as 13) Noguchi T., Miyachi H., Katayama R., Naito M., Hashimoto Y.,
purified tubulin for polymerization/depolymerization assay. Tubulin poly- Bioorg. Med. Chem. Lett., 15, 3212—3215 (2005).
merization was followed by means of turbidity measurements at 37 °C in mi- 14) Noguchi T., Fujimoto H., Sano H., Miyajima A., Miyachi H.,
crotubule assemble buffer containing 100 mM MES, 1 mM EGTA, 0.5 mM
MgCl2, 1 mM 2-mercaptoethanol, and 1 mM GTP (pH 6.5), in the presence or
absence of 12% v/v DMSO and/or 0.4 M glutamate. Although the measured
Hashimoto Y., Bioorg. Med. Chem. Lett., 15, 5509—5513 (2005).
15) Nakamura T., Noguchi T., Kobayashi H., Miyachi H., Hashimoto Y.,
Chem. Pharm. Bull., 54, 1709—1714 (2006).
turbidity (Abs.400) values of polymerized tubulin differed from experiment to 16) Li P. K., Pandit B., Sackett D. L., Hu Z., Zink J., Zhi J., Freeman D.,
experiment, the results (calculated IC50 values of the compounds) were basi-
cally reproducible.
Robey R. W., Werbovetz K., Lewis A., Li C., Mol. Cancer Ther., 5,
450—456 (2006).
Surface Plasmon Resonance Interaction of 5HPP-33 (3) and a,b-tubu-
lin heterodimer protein was monitored by surface plasmon resonance analy-
sis using a SPR670 instrument (Nippon Laser & Electronics Lab.). Prepara-
tion of the sensor chip on which a,b-tubulin heterodimer protein was immo-
bilized was carried out as follows. (1) Immobilization of sulfanylbutyric acid
on the sensor chip by soaking the chip in 1 mM dithiodibutyric acid (DDA)
ethanol solution at 4 °C overnight. (2) Activation of carboxyl groups with N-
hydroxysuccimide (NHS). This operation was carried out using EDCI
17) Yanagawa T., Noguchi T., Miyachi H., Kobayashi H., Hashimoto Y.,
Bioorg. Med. Chem. Lett., 16, 4748—4751 (2006).
18) Ataka K., Oku M., EP Patent 0271275 (1988).
19) Yoshida M., Miyata K., Shibabuchi H., Imai Y., JP Patent 03232838
(1991).
20) Li Y., Kobayashi H., Hashimoto Y., Shirai R., Hirata A., Hayashi K.,
Hamada Y., Shioiri T., Iwasaki S., Chem. Biol. Interact., 93, 175—183
(1994).
(25 mg) in Milli Q water (1 ml) and NHS (15 mg) in 1,4-dioxane (10 ml) for 21) Li Y., Kobayashi H., Tokiwa Y., Hashimoto Y., Iwasaki S., Biochem.
30 min at room temperature. (3) Immobilization of a,b-tubulin heterodimer
protein was performed by injection of the protein (50 mg/ml in 100 mM MES
buffer solution, 60 mM) on the sensor chip. (4) Blocking of unreacted surface
was performed by treatment of the tip with 1 M ethanolamine solution in
Pharmacol., 143, 219—224 (1992).
22) Takahashi M., Iwasaki S., Kobayashi H., Okuda S., Murai T., Sato Y.,
J. Antibiot., 40, 66—72 (1987).
23) Willam R. C., Lee J. C., Methods Enzymol., 85, 376—385 (1982).
Milli Q water (60 ml). 5HPP-33 (3) was dissolved at the concentration of 24) Himes R. H., Burton P. R., Gaito J. M., J. Biol. Chem., 252, 6222—
100, 200 or 400 mM in 100 mM MES buffer (containing 1—2% v/v DMSO). 6228 (1977), and references therein.
This solution was injected at 25 °C at a flow rate of 15 ml/min onto the sur- 25) Chakrabarti G., Kim S., Gupta M. L., Jr., Barton J. S., Himes R. H.,
face of the sensor chip upon which a,b-tubulin heterodimer protein was im-
mobilized. Using the sensor chip which was directly blocked with
Biochemistry, 38, 3067—3072 (1999).