2-Carboxymethylendothal analogues as affinity probes for stabilized protein phosphatase 2A

Article abstract of DOI:10.1016/S0968-0896(99)00239-4

Endothal (1(diacid)) and [³H]cantharidic acid ([³H]CA) bind with high affinity to the catalytic subunit of protein phosphatase 2A (PP2A). PP2A in liver cytosol was greatly stabilized with 30% glycerol as a preliminary step in the pot

Full text of DOI:10.1016/S0968-0896(99)00239-4

Bioorganic & Medicinal Chemistry 7 (1999) 2937±2944
2-Carboxymethylendothal Analogues as Anity Probes for
Stabilized Protein Phosphatase 2A
Charles W. Laidley, a,y William G. Dauben, b,{ Zhen R. Guo, b,x
Joe Y. L. Lam b,{ and John E. Casida a,*
aEnvironmental Chemistry and Toxicology Laboratory, Department of Environmental Science, Policy and Management,
115 Wellman Hall, University of California, Berkeley, CA 94720-3112, USA
bDepartment of Chemistry, University of California, 115 Wellman Hall, Berkeley, CA 94720-1460, USA
Received 18 May 1999; accepted 26 July 1999
AbstractÐEndothal (1diacid) and [3H]cantharidic acid ([3H]CA) bind with high anity to the catalytic subunit of protein phospha-
tase 2A (PP2A). PP2A in liver cytosol was greatly stabilized with 30% glycerol as a preliminary step in the potential use of endo-
thal-type derivatives for anity chromatography. We report here the ®rst introduction of a functionalizable group into endothal
which allows retention of binding site anity (assayed as [3H]CA binding in mouse liver cytosol). 2-Carboxymethylendothal
anhydride (7) was prepared in two steps and 97% overall yield from cis-aconitic anhydride and furan. The potency of 7 was
retained on conversion to two 2-carboxymethyl esters but not to two 2-(n-alkylcarboxamidomethyl) analogues. # 1999 Elsevier
Science Ltd. All rights reserved.
Introduction
binding site. Microcystin±Sepharose anity chromato-
graphy for PP2A16 is controversial17 and okadaic acid-
Sepharose is not useful16 in recovering signi®cant PP2A
activity. An attractive alternative for anity chromato-
graphy is to use compounds based on the herbicide endo-
thal (1diacid) and its analogues 2 and 3 (Fig. 1) which
display non-covalent binding characteristics.18±25
The dynamic process of protein phosphorylation and
dephosphorylation by protein kinases and protein
phosphatases plays a major role in intracellular signal
transduction.1 Protein phosphatase 2A (PP2A) dephos-
phorylates serine and threonine residues in a broad range
of cellular phosphoproteins and is the intracellular target
for many drugs, toxins and tumor promoters.2±4 Natural
PP2A inhibitors include the exceptionally potent and
structurally complex microcystin-LR5,6 and okadaic
acid7,8 and the less active but much simpler palasonin (2)
and cantharidin (3) (Fig. 1).9-13 Microcystin14 and a pho-
toanity ligand based on okadaic acid15 bind covalently
making them valuable in characterization of the ligand
This study used the cantharidic acid radioligand
([3H]CA) (Fig. 1) binding assay to optimize the interac-
tion of new endothal analogues with PP2A. This proce-
dure was previously fully validated as being directly
related to its catalytic site and enzymatic activity.11 The
®rst goal of this study, as a prelude to anity chroma-
tography, was to determine the association rate and
saturation conditions of [3H]CA. Second, the enzyme
was stabilized so that equilibration of the anity probe
could be achieved. The third goal was to introduce a
functionalizable group into endothal which allowed
retention of binding site anity. This had not been
achieved before since substitution at the 2- and 3-car-
boxyl group in the endothal series or introduction of
substituents at the 1-, 4-, 5- or 6-position led to reduction in
or total loss of PP2A binding activity and toxicity.9±12,20±25
The new series examined here was the 2-substituted
endothal anhydride analogues other than the 2-methyl
compound (2). The fourth goal was to determine the
e€ect of derivatizing the functionalizable group on
Key words: Enzymes inhibitors; herbicides; natural products; sub-
stituent e€ects.
* Corresponding author. Tel.:+1-510-642-5424; fax:+1-510-642-
y6497; e-mail: ectl@nature.berkeley.edu
Present address: Department of Biology, Paci®c University, Forest
Grove, OR 97116, USA.
{
Dedicated to the memory of Professor William G. Dauben (1919±
1997) who made outstanding contributions to many areas of organic
chemistry including cantharidin synthesis and action.
Present address: Process R&D, Bristol-Myers Squibb Pharmaceu-
tical Research Institute, New Brunswick, NJ 08903-0191, USA.
x
{
Present address: Genetic Analysis R&D, PE Applied Biosystems,
Foster City, CA 94404, USA.
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00239-4

2938
C. W. Laidley et al. / Bioorg. Med. Chem. 7 (1999) 2937±2944
Scheme 1. Synthesis of 2-carboxymethylendothal anhydride [7] and
esters and amides thereof [8±11].
created by the carboxymethyl group. However, when
this reaction is carried out at 25ꢀC under 8 kbar for 4
days the conversion is 100% to give 98% yield of the
Diels±Alder adduct, dehydro-2-carboxymethylendothal
anhydride (dehydro-7), as the only product (no endo
isomer could be detected). Hydrogenation of dehydro-7
in THF over 10% Pd/C gave 7 in 99% yield. Com-
pound 7 was converted to esters 8 and 9 and amides 10
and 11 using dicyclohexylcarbodiimide (DCC) for cou-
pling the carboxylic acid and alcohol or primary amine.
No attempt was made to resolve the diastereomeric
acids, esters and amides 7±11.
Figure 1. Structures of variously-substituted exo,exo-7-oxabicy-
clo[2.2.1]heptane-2,3-dicarboxylic acids (1diacid±3diacid) and anhydrides
(1±11) and of radioligand [3H]CA (asterisks designate labeled positions)
for the active site of protein phosphatase 2A. Palasonin is compound 2
and cantharidin is 3. One structure gives the numbering system.
Results and Discussion
Enzyme stability and stabilization
Temperature e€ects on association and dissociation. The
rates of [3H]CA association and dissociation are highly
temperature-dependent (Fig. 2). There is negligible
ligand association/dissociation at 4ꢀC and even at 21ꢀC
ligand association appears incomplete at 8 h and dis-
sociation only reaches the t1/2 value after 24 h. Although
steady-state binding appears to have been nearly
attained after 8 h at 37ꢀC the maximal binding levels
remain similar to those obtained at 21ꢀC. Dissociation
rates at 37ꢀC are identical using 1diacid, pyrophosphate
or MnCl2 as the competing ligand (Fig. 2) consistent
with earlier observations4 that they act competitively
rather than allosterically at the [3H]CA binding site.
Further, the apparent rate of ligand dissociation seen at
37ꢀC (Fig. 2) is confounded by a parallel loss of binding
activity under these conditions (Fig. 3) leading to further
examination of binding site stability and stabilization.
inhibitory potency. These are preliminary steps in using
the endothal-type series as improved anity probes and
possibly for anity chromatography.
Chemistry
Compounds studied here from syntheses reported ear-
lier are: 1diacid, 2diacid, 3diacid (CA), 2, 3, 4 and 6;18±21 the
individual enantiomers of 2, i.e. (+)-2 and ( )-2.24
Compound 5 was obtained by high pressure Diels±
Alder reaction between 1-cyclopentene-1,2-dicarboxylic
anhydride and furan under 8 kbar and 3 days to give a
pair of exo and endo dehydro-2,3-trimethyleneendothal
anhydride isomers in the ratio of 13:12 in 99% yield.
The two isomers were separated by fractional recrys-
tallization and the pure exo isomer was hydrogenated
with 10% Pd/C in THF to give 5.21
Stability of binding site. Although increasing the tem-
perature facilitates ligand binding, prolonged exposure
of cytosol to temperatures above 4ꢀC results in loss of
[3H]CA binding activity. Radioligand binding in 20%
cytosol is essentially stable for 24 h at 4ꢀC but declines
at 21 and 37ꢀC with t1=2 values of 16 and 2.6 h, respec-
tively (Fig. 3). Similar studies using more dilute (0.5%)
cytosol in imidazole bu€er establish the same trends
(data not shown).
Carboxymethylendothal anhydride (7) was the key tar-
get compound for introducing a functionalizable group
at the 2-position (Scheme 1). The [4+2] Diels±Alder
reaction between furan and the commercially available
cis-aconitic anhydride will not occur at atmospheric
pressure due to the aromaticity of the furan and the
sterically hindered environment in the transition state

C. W. Laidley et al. / Bioorg. Med. Chem. 7 (1999) 2937±2944
2939
Figure 2. Temperature e€ects on [3H]CA association and dissociation. Association kinetics determined by adding 5 nM [3H]CA alone or with 10mM
unlabeled CA (for nonspeci®c binding) to cytosol in pH 7.0 imidazole/EDTA/EGTA bu€er and incubating for up to 8 h. Dissociation kinetics determined
by preincubating cytosol with 5 nM [3H]CA for equilibration with dissociation initiated by 1diacid (10 mM) and monitored for 24h. The same curve was
obtained for dissociation of bound [3H]CA at 37ꢀC by 1diacid (10 mM) (illustrated) and by pyrophosphate (100 mM) or MnCl2 (5mM) (not shown).
(1 mM). There is some enhancement in stability with 2-
mercaptoethanol (12 mM), dithiothreitol (5 mM),
MnCl2 (50 mM), phenylmethanesulfonyl ¯uoride (1 mM)
and 2-phenyl-4H-1,3,2-benzodioxaphosphorin 2-oxide
(PSCP)26 (10 mM) (data not shown) whereas glycerol
(30%) gives remarkably increased binding site stability.
The imidazole/glycerol medium, designed from these
observations, extends the binding activity not only at
21ꢀC but also at 37ꢀC where 10, 20 and 30% glycerol
enhance binding by 40, 100 and 160%, respectively,
relative to bu€er without glycerol (Fig. 4A). Subsequent
supplementation of bu€er containing 30% glycerol with
PSCP, 2-mercaptoethanol and Mn2+ results in a further
increase in binding activity. The apparent rate of
[3H]CA association was not a€ected by the glycerol
concentration (Fig. 4B) and accurate rates of ligand
association/dissociation could now be considered for
the ®rst time. Re-examination of binding kinetics in this
new `stabilized' bu€er media yielded complete stabiliza-
tion after 6±8 h at 37ꢀC (association t1=2 of 67 min) and
90% dissociation after 33 h (t1=2 of 114 min) (Fig. 5).
Binding activity under these conditions remained stable
Figure 3. Temperature e€ects on stability of [3H]CA binding site of
for >24 h (see control line) (Fig. 5). The similarity in
PP2A. Cytosol (20% fresh weight equivalent in pH 7.0 imidazole/
EDTA/EGTA bu€er) was incubated at 4, 21 and 37ꢀC then diluted
40-fold to 0.5% cytosol in the same bu€er and assayed for [3H]CA
binding activity for an additional 3 h at 37ꢀC. Cytosol maintained at
0ꢀC was used as the control (i.e. 100% binding activity).
association and dissociation rates in the presence or
absence of glycerol suggests that this stabilizer acts to
increase the number of binding sites from 1.1 to
2.7 pmol/mg protein without altering the binding kinetics
under the standard assay conditions.
Stabilization of binding site. Several candidate protein
stabilizing agents were tested for possibly prolonging
[3H]CA binding activities for cytosol in imidazole bu€er
with incubations for 14h at 21 or 37ꢀC. Little or no stabili-
zation or enhanced binding is achieved with standard pro-
tease inhibitors (0.3 mM aprotonin, 20mM leupeptin, and
1 mM pepstatin, tested individually), benzamidine (1 mM),
EDTA (1 mM), EGTA (1 mM) and N-methylmaleimide
Earlier studies in the absence of stabilizers were made
with incubations at 37ꢀC without recognizing that
binding site loss occurred to a major extent leading to
an underestimate of maximum binding at equilibrium.
These values in the imidazole/glycerol bu€er are now
260% of those obtained without stabilization indicating
that earlier estimates of PP2A concentration of cytosol

2940
C. W. Laidley et al. / Bioorg. Med. Chem. 7 (1999) 2937±2944
Figure 4. Glycerol enhancement of [3H]CA binding at equilibrium (A) without a€ecting the association rate (B). Association determined as in Fig. 2
at 37ꢀC but in pH 7.0 imidazole/glycerol bu€er with 0, 10, 20 or 30% glycerol.
Figure 5. [3H]CA association and dissociation in pH 7.0 imidazole/
Figure 6. Competition curves showing the potency of six endothal
glycerol bu€er with 30% glycerol at 37ꢀC. Association and dissociation
analogues for inhibition of [3H]CA binding at pH 7.25. Cytosol in pH
determined as in Fig. 2 but with dissociation initiated by 1diacid (10 mM).
7.25 imidazole/glycerol bu€er (see Experimental) incubated with 5 nM
[3H]CA alone or with competitor for 12 h at 37ꢀC using unlabeled CA
(1 mM) to correct for nonspeci®c binding. Binding density at 100% is
2.7 pmol/mg protein. Data can be compared with results at pH 6.0 in
by this procedure were only about 40% of the actual
value. Although the mechanism by which glycerol leads
to increased binding activity is not known, glycerol is
thought to stabilize protein±protein interactions27 and is
used for storing commercially-puri®ed PP2A.28
the same experiment (Table 1). The potency order refers to com-
pounds designated as in Fig. 1.
further increased with the -CH2CH2CH2- linkage (5).
The thioanhydride of 1 (compound 6) is fourfold less
potent than 1diacid itself. The potency order of 1diacid
<2diacid <3 at pH 7.25 indicates that introduction of 2-
or 2,3-substituents at the endo position may achieve the
desired goal. The functionalizable carboxymethyl group
was then introduced endo at C-2 of 1 to obtain 7 with an
IC50 of 45±47 nM at pH 7.25 and 6.0 which is a potency
increase of 3.1-fold at pH 7.25 and decrease of 1.7-fold
at pH 6.0.
Inhibitor structure±activity and pH relationships
Selection of carboxymethylendothal (7). The 2- and 3-
substituents greatly in¯uence potency for inhibition of
[3H]CA binding in imidazole/glycerol at pH 7.25 with
12 h incubation (Fig. 6, Table 1). Relative to endothal
(1diacid),
a 2-methyl substituent (2diacid) increases
potency by sevenfold at pH 7.25 and threefold at pH
6.0. Anhydride (‹)-2 is similar in potency to (+)-2 and
( )-2, so the presence of the 2-methyl substituent is
more important than its stereochemistry. A second
methyl substituent in the anhydride series increases the
potency a further fourfold (3 versus 2) and this activity
level is maintained with the -CH2SCH2- bridge (4) and
Activity of esters and amides (8±11) of carboxymethylen-
dothal anhydride (7). The ®nding that 7 is of similar
potency to 2 at pH 7.25, i.e. 2-carboxymethyl versus 2-
methyl, led to the preparation of esters and amides as
models for linking 7 to anity matrices. Esters 8 and 9
show similar potency to the parent acid 7, whereas

C. W. Laidley et al. / Bioorg. Med. Chem. 7 (1999) 2937±2944
2941
amides 10 and 11 are considerably less active. On this
basis, esters 8 and 9 may hydrolyze to compound 7, a
possibility not directly examined in the present study.
than at 7.25. Compounds 3, 5 and 7 and the four esters
and amides (8±11) show much less di€erence (0.8±1.5-
fold) in potency with pH (Table 1). A more detailed
examination established that the binding anities of 3
and 7 are not greatly in¯uenced by the pH whereas
those of 1diacid, 2 and 6 are highly dependent on the
assay pH (Fig. 7). The mechanism by which the pH
di€erentially a€ects the inhibitory potency is not known
but probably involves optimal interactions of the car-
boxylate substituent directly (1diacid and 2diacid) or after
hydrolysis10,21,23 with the ligand binding site.
E€ect of pH on structure±activity relationships. The
endothal derivatives di€er in the e€ect of pH on their
inhibitory activity. The pH optimum of [3H]CA binding
in imidazole/glycerol bu€er with a 12 h incubation at
37ꢀC is broad, with little di€erence in the range of 6.0 to
7.25 (Fig. 7). In contrast, [3H]6 binding is optimal at pH
6.0 with <10% of this activity at pH 7.4.12,21 The IC50
values of endothal analogues were determined at two
pH values (6.0 and 7.25) (Table 1) which in themselves
have little e€ect on [3H]CA binding in the absence of
inhibitors (Fig. 7). This comparison recon®rms the pH
e€ect on 6 binding noted above since we ®nd a sixfold
Candidate anity probes
The goal of discovering an endothal analogue combin-
ing a functionalizable substituent with reasonably high
potency in the [3H]CA binding assay was achieved with
7. It was accordingly converted to two esters and two
amides (Table 1). Potent esters 8 and 9 may hydrolyze
during assay (despite attempted inhibition of esterases
by PSCP)26 and act as 7. Amides 9 and 10, with N-butyl
and N-heptyl substituents, respectively, have reduced
anity or cleave more slowly. In comparison with
microcystin±Sepharose anity chromatography16,17 the
use of a carboxymethylendothal derivative potentially
provides greater reversibility and ease of PP2A dis-
placement. Anity chromatography with an endothal
analogue would require several hours equilibration time
at 37ꢀC based on the association kinetics of [3H]CA in
imidazole/glycerol bu€er at pH 7.0 (Fig. 5). In conclu-
sion, it is clear that the endo 2-position (either alone or
in combination with a 3-substituent) is the site for deri-
vatization to retain binding site activity and further
structure±activity studies are required to achieve con-
jugates of adequate potency and stability for anity
chromatography.
greater potency at pH 6.0 than at 7.25. Moreover, 1diacid
2diacid, the enantiomers of 2, and 4 follow a similar pat-
,
tern with two to ®vefold greater potency at pH 6.0
Table 1. Structure±activity relationships for endothal analogues and
carboxymethylendothal derivatives as inhibitors of [3H]CA binding at
pH 7.25 and pH 6.0 in imidazole/glycerol bu€er with 12 h incubation
at 37ꢀC
No.a
IC50 (nM) at indicated pHb
IC50 ratio
7.25/6.0
7.25
6.0
Endothal analogues
1diacid
(‹)-2
(+)-2
( )-2
(‹)-2diacid
3
4
5
6
139‹14
31‹5
26‹1
37‹3
19‹2
8.0‹1.2
10‹1
3.0‹0.4
608‹6
28‹4
12‹1
9.7‹0.7
14‹1
9.4‹0.8
9.9‹2.2
2.7‹0.0
2.4‹0.5
104‹3
5.0
2.6
2.7
2.6
2.0
0.8
3.7
1.3
5.9
(‹)-Carboxymethylendothal derivatives
7
8
45‹3
42‹4
47‹5
31‹6
63‹12
168‹18
216‹15
1.0
1.4
0.8
1.5
1.0
Experimental
Chemicals
9
10
11
48‹10
251‹32
206‹21
General. 1H NMR (400 MHz) and 13C NMR
(100.6 MHz) spectra were determined for CDCl3 or
THF-d8 solutions calibrated with Me4Si. IR spectra
a
For structures see Fig. 1.
Data are the means ‹SE of three experiments.
b
Figure 7. E€ect of pH on [3H]CA binding (control in A) and inhibition by endothal analogues. A and B: experimental conditions the same as Fig. 6
except pH varied from 5.5 to 8.5 and competitors included at levels that inhibited 42±64% at pH 7.0. C: data for 1diacid and 3 presented as %
inhibition at each pH value.

2942
C. W. Laidley et al. / Bioorg. Med. Chem. 7 (1999) 2937±2944
were recorded using KBr pellets for solid samples.
Melting points (mp) determined in capillary tubes are
uncorrected. Et2O and THF were distilled over CaH2.
Flash chromatography utilized 230±400 mesh silica gel
at medium pressure. The high pressure apparatus and
the general procedures for high pressure reactions have
been described.24
under hydrogen (3 L in¯ated balloon) for 20h and ®ltered.
The ®ltrate was evaporated to give a colorless oil which
crystallized out upon addition of CH2Cl2 to give 7 (1.22 g,
5.4 mmol) in 99% yield. A sample was recrystallized from
hexane±EtOAc: mp 224±225ꢀC; IR nmax 3067 (broad),
3003, 1860, 1777, 1714 cm 1; 1H NMR (THF-d8) d 1.59±
1.92 (m, 4H), 2.74, 3.05 (AB quartet, J=18 Hz, 2H), 2.99
(s, 1H), 4.66 (d, J=4.7Hz, 1H), 4.80 (d, J=5.4 Hz, 1H);
13C NMR (THF-d8) d 25.5, 28.2, 35.8, 55.6, 57.4, 82.2,
83.8, 172.2, 173.1, 176.3. Anal. calcd for C10H10O6: C,
53.10; H, 4.46. Found: C, 52.72; H, 4.47.
exo-2,3-Trimethyleneendothal anhydride (5). A solution
of 1-cyclopentene-1,2-dicarboxylic anhydride (1.5 g,
10.9 mmol), furan (0.81 g, 12.0mmol), and CH2Cl2
(1.5 mL) was pressurized to 8 kbar for 3 days at 25ꢀC. The
solvent was evaporated under reduced pressure to give a
pair of exo and endo dehydro-2,3-trimethyleneendothal
anhydride isomers as a white solid (2.24 g, 99% yield). 1H
NMR showed exo to endo in the ratio of 13:12. This mix-
ture of the two isomers was recrystallized from hex-
ane:EtOAc (6:4). The crystals collected were identi®ed as
endo-dehydro-5: mp 127±128ꢀC; IR nmax 2968, 2948, 1841,
1778, 1444, 1241 cm 1; 1H NMR (CDCl3) d 1.33±1.41 (m,
2H), 1.89±2.04 (m, 2H), 2.17±2.23 (m, 2H), 5.24 (t,
J=0.95 Hz, 2H), 6.66 (t, J=1.03 Hz, 2H); 13C NMR
(CDCl3) d 31.2, 31.6, 69.4, 83.9, 137.0, 174.4. The mother
liquor was evaporated to dryness and recrystallized ®ve
times from hexane:EtOAc (7:3) to give pure exo-dehydro-
5: mp 133±134ꢀC; IR nmax 2974, 2943, 1847, 1771, 1252,
1199 cm 1; 1H NMR (CDCl3) d 1.76±1.87 (m, 3H), 2.11±
2.13 (m, 1H), 2.47±2.52 (m, 2H), 4.98 (s, 2H), 6.62 (t,
J=0.82 Hz, 2H); 13C NMR (CDCl3) d 30.0, 33.6, 68.1,
833.4, 136.6, 171.8.
2-(Cyclohexylmethoxycarboxymethyl)endothal anhydride
(8). To a solution of 7 (1.0 g, 4.4 mmol) in THF (30 mL)
was added DCC (0.91 g, 4.4 mmol) in THF (5 mL) drop-
wise at room temperature, and the mixture was stirred for
5 min and cooled in an ice-bath. Cyclohexylmethanol
(0.51 g, 4.5 mmol) in THF (5mL) was added dropwise,
and the mixture was stirred for 3 h at 0ꢀC. The white pre-
cipitate was ®ltered, and the ®ltrate was evaporated to
give a white solid which was recrystallized from 40%
EtOAc in hexanes to give 8 (1.28 g, 4.0 mmol) in 90%
yield; mp 125±127ꢀC; IR nmax 2924, 2842, 1846, 1778,
1
1721, 1445, 1408, 1357, 1268, 1206, 1144, 998, 958 cm
;
1H NMR (CDCl3) d 0.90±0.96 (m, 2H), 1.15±1.25 (m,
2H), 1.56±1.61 (m, 2H), 1.64±1.73 (m, 6H), 1.79±1.83 (m,
2H), 1.90±1.94 (m, 1H), 2.60, 3.12 (AB quartet, J=18 Hz,
2H), 2.84 (s, 1H), 3.12 (d, J=17.77Hz, 1H), 3.90 (m, 2H),
4.78 (s, 1H), 4.95 (d, J=5.4Hz, 1H); 13C NMR (CDCl3) d
24.7, 25.5, 26.2, 27.7, 29.5, 35.7, 36.8, 55.0, 55.3, 71.2, 81.0,
82.6, 170.3, 170.8, 175.0. HRMS calcd for C17H22O6:
322.1416. Found: 322.1419.
To a solution of exo-dehydro-5 (0.51 g, 2.5 mmol) in THF
(20 mL) was added 10% Pd/C (0.15 g). The resulting mix-
ture was stirred under hydrogen (3 L in¯ated balloon) for
20h and the reaction mixture was ®ltered. The ®ltrate was
evaporated to give 5 as a white solid (0.50 g, 98%). A
sample was recrystallized from hexane:EtOAc (6:4) to give
5 as white crystals; mp 171±172ꢀC; IR nmax 2983, 2900,
1859, 1779 cm 1; 1H NMR (CDCl3) d 1.40 (m, 1H), 1.66±
1.82 (m, 4H), 1.86±1.95 (m, 3H), 2.49±2.54 (m, 2H), 4.59
(dd, J=2.31, J=2.28 Hz, 2H); 13C NMR (CDCl3) d 25.9,
26.4, 35.8, 70.3, 81.9, 172.6. HRMS calcd for C11H12O4:
208.0736. Found: 208.0736.
2-(4-Methoxybenzyloxycarboxymethyl)endothal anhydride
(9). To a solution of 7 (0.80 g, 3.5 mmol) in THF (35 mL)
was added DCC (0.72 g, 3.5 mmol) in THF (5 mL) drop-
wise at room temperature, and the mixture was stirred for
5 min and cooled in an ice-bath. 4-Methoxybenzyl alcohol
(0.49 g, 3.5 mmol) in THF (10 mL) was added dropwise,
and the mixture was stirred for 1 h at 0ꢀC, then overnight
at room temperature. The white precipitate was ®ltered,
and the ®ltrate was evaporated to give a white solid which
was recrystallized from 40% EtOAc in hexanes to give 9
(1.10 g, 3.2 mmol) in 91% yield; mp 135±136ꢀC; IR nmax
2926, 2849, 1850, 1779, 1724, 1518, 1406, 1357, 1253,
1139, 998, 909, 821 cm 1; 1H NMR (CDCl3) d 1.53±1.57
(m, 1H), 1.76±1.82 (m, 2H), 1.90±1.95 (m, 1H), 3.13, 2.60
(AB quartet, 2H, J=17.81Hz), 2.82 (s, 1H), 3.81 (s, 3H),
4.78 (d, 1H, J=4.17 Hz), 4.93 (d, 1H, J=5.39 Hz), 5.06,
5.05 (AB quartet, 2H, J=11.77Hz), 6.90 (m, 2H), 7.24±
7.31 (m, 2H); 13C NMR (CDCl3) d 24.7, 27.7, 35.8, 55.0,
55.3, 56.3, 67.7, 81.0, 82.6, 114.1, 126.7, 130.6, 160.0,
170.1, 170.8, 175.0. HRMS calcd for C18H18O7: 346.1053.
Found: 346.1059.
Dehydro-2-carboxymethylendothal anhydride (dehydro-7).
A solution of cis-aconitic anhydride (1.0g, 6.4 mmol),
furan (0.48 g, 7.1 mmol), and dry THF (2mL) was pres-
surized to 8 kbar for 4 days at 25ꢀC. The solvent was eva-
porated under reduced pressure to give dehydro-7 as an oil
(1.41 g, 98%) and 1H NMR showed 100% conversion and
>95% purity. The oily product was stirred with anhy-
drous CH2Cl2, ®ltered and then evaporated to give white
crystals of dehydro-7 (1.22 g, 5.4 mmol) in 85% yield: mp
122±123ꢀC; 1H NMR (THF-d8) d 2.21, 3.21 (AB quartet,
J=18 Hz, 2H), 2.94 (s, 3H), 5.09 (m, 1H), 5.26 (m, 1H),
6.59 (dd, J=1.73, 5.75Hz, 1H), 6.65 (dd, J=1.77, 5.76Hz,
1H); 13C NMR (THF-d8) d 37.8, 52.8, 58.4, 84.1, 85.1,
136.5, 138.7, 171.2, 172.9, 175.0. Anal. calcd for C10H8O6:
C, 53.58; H, 3.60. Found: C, 53.24; H, 3.74.
2-(n-Butylcarboxamidomethyl)endothal anhydride (10).
To a solution of 7 (1.0 g, 4.4 mmol) in THF (30 mL)
was added DCC (0.908 g, 4.4 mmol) in THF (5 mL)
dropwise at room temperature. The mixture was stirred
for 5 min and then cooled in an ice-bath. n-Butylamine
(0.323 g, 4.4 mmol) in THF (5 mL) was added dropwise,
and the mixture was stirred for 5 h at 0ꢀC. The white
precipitate was ®ltered, and the ®ltrate was evaporated
2-Carboxymethylendothal anhydride (7). To a solution
of dehydro-7 (1.22 g, 5.4 mmol) in dry THF (30 mL) was
added 10% Pd/C (0.8 g), and the mixture was stirred

C. W. Laidley et al. / Bioorg. Med. Chem. 7 (1999) 2937±2944
2943
to give a white solid 10 (1.15 g, 4.1 mmol) in 93% yield. IR
nmax 3175 (broad), 2967, 1775, 1702, 1407, 1210, 1138,
1051, 990; H NMR (CDCl3) d 0.86 (t, 3H, J=7.23 Hz),
1.49 (m, 2H), 1.57 (m, 2H), 1.64 (m, 2H), 1.80 (m, 2H),
2.60 (s, 1H), 2.81 (d, 1H, J=18.26Hz), 2.94 (d, 1H,
J=18.41Hz) 3.42 (m, 2H), 4.43 (s, 1H), 5.05 (d, 1H,
J=4.47 Hz); 13C NMR (CDCl3) d 13.6, 19.9, 24.9, 27.8,
31.4, 37.0, 39.6, 55.3, 56.8, 80.9, 82.7, 168.8, 171.5,
176.0. HRMS calcd for C14H19O5N: 281.1263. Found:
281.1268.
rinsed (3Â1 mL) with imidazole bu€er at 5ꢀC, using a
Wallac TomTec cell harvester (Harvester 96 Mach II,
Gaithersburg, MD). Filtermats were dried overnight
and impregnated with MeltiLex melt-on scintillator
sheets (Wallac) and the bound radioactivity was quan-
ti®ed using a Wallac Betaplate (model 1205) ¯atbed
counter.4
1
Association and dissociation kinetics for [3H]CA binding
site. Association rates were determined at 4, 21 and 37ꢀC
by mixing cytosol with 5 nM [3H]CA (total binding) and
10mM unlabeled CA (non-speci®c binding) and incubat-
ing for periods of 5 min to 8 h at each temperature. Ligand
dissociation rates were determined by preincubating
cytosol in imidazole/EDTA/EGTA bu€er with 5 nM
[3H]CA for 6 h at 37ꢀC followed by the addition of either
1diacid (10 mM), pyrophosphate (100mM) or MnCl2
(5 mM) for periods ranging from 1 to 24h at temperatures
of 4, 21 and 37ꢀC. The kinetics of [3H]CA binding were re-
examined in imidazole/glycerol bu€er to determine if sta-
bilization was associated with reduced rates of ligand
association and dissociation.
2-(n-Heptylcarboxamidomethyl)endothal anhydride (11).
To a solution of 7 (0.86 g, 3.8 mmol) in THF (30 mL) was
added DCC (0.78g, 3.8 mmol) in THF (5mL) dropwise at
room temperature. The mixture was stirred for 5 min and
then cooled in an ice-bath. n-Heptylamine (0.461g,
4.0 mmol) in THF (5 mL) was added dropwise, and the
mixture was stirred for 5 h at 0ꢀC, then stirred at room
temperature overnight. The white precipitate was ®ltered,
and the ®ltrate was evaporated to give a white solid 11
(1.17 g, 3.6 mmol) in 95% yield. IR nmax 3187 (broad),
2931, 1774, 1702, 1407, 1181; 1H NMR (CDCl3) d 0.85 (t,
3H, J=7.05Hz), 1.24 (m, 8H), 1.53 (m, 4H), 1.81 (m, 2H),
2.61 (s, 1H), 2.82 (d, 1H, J=18.31Hz), 2.94 (d, 1H,
J=18.02Hz), 3.41 (m, 2H), 4.44 (s, 1H), 5.06 (d, 1H,
J=4.44 Hz); 13C NMR (CDCl3) d 14.0, 22.6, 26.4, 26.8,
27.4, 28.8, 31.7, 38.2, 39.1, 40.1, 55.9, 60.7, 78.3, 81.8,
174.0, 174.8, 179.0. HRMS calcd for C17H25O5N:
323.1733. Found: 323.1733.
Stability and stabilization of [3H]CA binding site.
[3H]CA association and equilibrium binding is best
achieved at 37ꢀC under conditions where stability must
be considered. For stability studies, 20% cytosol was
incubated at 4, 21 or 37ꢀC for periods ranging from 1 to
24 h then diluted 40-fold and assayed for binding activ-
ity. Various potential stabilizing factors were examined
including speci®c protease inhibitors (detailed earlier),
thiol protective and derivatizing agents, PSCP, divalent
ion chelators, and Mn2+, a potential cofactor.4 Com-
pounds were initially tested for stabilizing e€ects on
[3H]CA binding activity by overnight (14 h) incubations
at 21 and 37ꢀC, followed by dilution in imidazole/gly-
cerol bu€er and assay of [3H]CA binding activity.
Protein phosphatase 2A
Assay of [3H]CA binding site of PP2A. The procedure
requires the radioligand, a PP2A source, and an appro-
priate bu€er. [3H]3 (34 Ci/mmol) was repuri®ed by silica
gel chromatography with CH2Cl2 as required to maintain
a radiochemical purity of >95%.4 It was diluted to 25nM
in 50mM imidazole bu€er at pH 7.4 and held for at least
1 h at 25ꢀC to allow hydrolysis to [3H]CA for use in bind-
ing assays.4 To prepare the PP2A source, mouse liver was
homogenized, using polytron and Potter±Elvehjem pro-
cedures, in 4 volumes of ice-cold 50mM imidazole bu€er
at pH 7.0 (20% fresh weight equivalent) and the homo-
genate was centrifuged at 15,000Âg for 15min, and the
resulting supernatant at 105,000Âg for 60min to obtain
the cytosol which was stored at 70ꢀC. Two other sup-
plemented bu€ers were also used for comparison. The
imidazole/EDTA/EGTA bu€er at pH 7.0 consisted of
50mM imidazole, 1 mM each of EDTA and EGTA,
10mM N-methylmaleimide and 10mM PSCP. The imida-
zole/glycerol bu€er at pH 6.0, 7.0 or 7.25 (or as speci®ed)
contained 50 mM imidazole, 50 mM NaCl, 12 mM 2-
mercaptoethanol, 10 mM PSCP, 10 mM MnCl2 and
30% (v/v) glycerol. Ligand binding assays were con-
ducted in 96-microwell plates as previously described.4
Cytosol was diluted 20-fold in imidazole bu€er (or imi-
dazole/EDTA/EGTA or imidazole/glycerol bu€er) and
incubated with 5 nM [3H]CA in the absence (total bind-
ing tubes) and presence of 10 mM unlabeled CA (non-
speci®c binding tubes). Speci®c binding was calculated
as the di€erence between total and nonspeci®c binding.
Assays were terminated by ®ltration through glass-®ber
®lters presoaked in 0.3% polyethylenimine, and rapidly
Structure±activity and pH relationships for inhibitors of
[3H]CA binding. Incubations consisted of [3H]CA (5 nM
®nal concentration) and unlabeled competitor each
added in dimethyl sulfoxide (5 mL) to 0.5% cytosol
(20 mg protein, Bradford method)29 in imidazole/glycerol
bu€er or other bu€er as speci®ed (200mL). IC50 values
were determined by four parameter logistic curve ®tting.
Data presented in ®gures represent typical single experi-
ments. The e€ect of pH on [3H]CA binding and inhibitor
potency was examined in imidazole/glycerol bu€er adjus-
ted to pHs ranging from 5.5 to 8.5 in 0.5 pH increments,
with the pH con®rmed by direct measurement at the end
of the incubation period.
Acknowledgements
The project described was supported by grant P01
ES00049 to J.E.C. from the National Institute of
Environmental Health Sciences, National Institutes of
Health, and by Elf-Atochem North America, Inc. (Phi-
ladelphia, PA). Research in the Department of Chem-
istry was supported by National Science Foundation
Grant No. 8618303 to W.G.D. Advice and assistance
were provided by Gary Quistad, Phillip Je€eries and

2944
C. W. Laidley et al. / Bioorg. Med. Chem. 7 (1999) 2937±2944
Mahmoud Mahajna of the Environmental Chemistry
and Toxicology Laboratory.
14. MacKintosh, R. W.; Dalby, K. N.; Campbell, D. G.; Cohen,
P. T. W.; Cohen, P.; MacKintosh, C. FEBS Lett. 1995, 371, 236.
15. Nishiwaki, S.; Fujiki, H.; Suganuma, M.; Ojika, M.;
Yamada, K.; Sugimura, T. Biochem. Biophys. Res. Commun.
1990, 170, 1359.
16. Nishiwaki, S.; Fujiki, H.; Suganuma, M.; Nishiwaki-Mat-
sushima, R.; Sugimura, T. FEBS Lett. 1991, 279, 115.
17. Moorhead, G.; MacKintosh, R. W.; Morrice, N.; Galla-
gher, T.; MacKintosh, C. FEBS Lett. 1994, 356, 46.
18. Dauben, W. G.; Kessel, C. R.; Takemura, K. H. J. Am.
Chem. Soc. 1980, 102, 6893.
19. Dauben, W. G.; Gerdes, J. M.; Smith, D. B. J. Org. Chem.
1985, 50, 2576.
20. Matsuzawa, M.; Graziano, M. J.; Casida, J. E. J. Agric.
Food Chem. 1987, 35, 823.
21. Kawamura, N.; Li, Y.-M.; Engel, J. L.; Dauben, W. G.;
Casida, J. E. Chem. Res. Toxicol. 1990, 3, 318.
References
1. Cohen, P. Annu. Rev. Biochem 1989, 58, 453.
2. MacKintosh, C.; MacKintosh, R. W. Adv. Protein Phos-
phatases 1994, 8, 343.
3. MacKintosh, C.; MacKintosh, R. W. Trends Biochem. Sci.
1994, 19, 444.
4. Laidley, C. W.; Cohen, E.; Casida, J. E. J. Pharmacol. Exp.
Ther. 1997, 280, 1152.
5. Carmichael, W. W. In Handbook of Natural Toxins, Vol. 3,
Marine Toxins and Venoms; Tu, A., Ed.; Marcel Dekker: New
York, 1998; pp 121±147.
6. MacKintosh, C.; Beattie, K. A.; Klumpp, S.; Cohen, P.;
Codd, G. A. FEBS Lett. 1990, 264, 187.
7. Tachibana, K.; Scheuer, P. J.; Tsukitani, Y.; Kikuchi, H.;
Van Engen, D.; Clardy, J.; Gopichand, Y.; Schmitz, F. J. J.
Am. Chem. Soc. 1981, 103, 2469.
8. Bialojan, C.; Takai, A. Biochem. J. 1988, 256, 283.
9. Graziano, M. J.; Waterhouse, A. L.; Casida, J. E. Biochem.
Biophys. Res. Commun. 1987, 149, 79.
22. Erdodi, F.; Toth, B.; Hirano, K.; Hirano, M.; Hartshorne,
D. J.; Gergely, P. Am. J. Physiol. 1995, 269, C1176.
23. McCluskey, A.; Taylor, C.; Quinn, R. J.; Suganuma, M.;
Fujiki, H. Bioorg. Med. Chem. Lett. 1996, 6, 1025.
24. Dauben, W. G.; Lam, J. Y. L.; Guo, Z. R. J. Org. Chem.
1996, 61, 4816.
25. Sodeoka, M.; Baba, Y.; Kobayashi, S.; Hirukawa, N.
Bioorg. Med. Chem. Lett. 1997, 7, 1833.
10. Graziano, M. J.; Pessah, I. N.; Matsuzawa, M.; Casida, J.
E. Mol. Pharmacol. 1988, 33, 706.
11. Li, Y.-M.; Casida, J. E. Proc. Natl. Acad. Sci. USA 1992,
89, 11867.
12. Li, Y.-M.; MacKintosh, C.; Casida, J. E. Biochem. Phar-
macol 1993, 46, 1435.
13. Honkanen, R. E. FEBS Lett. 1993, 330, 283.
26. Casida, J. E.; Eto, M.; Baron, R. L. Nature 1961, 191, 1396.
27. Dignam, J. D. In Methods in Enzymology, Vol. 182, Pre-
paration of Extracts from Higher Eukaryotes; Deutscher, M.
P., Ed.; Academic Press: San Diego, 1990; pp 194±203.
28. Calbiochem/Novabiochem. Technical Data Sheet for Pro-
tein Phosphatase 2A, Bovine Kidney, 1996.
29. Bradford, M. M. Anal. Biochem. 1976, 72, 248.