2-aroylbenzoyl serine proteases: Photoreversible inhibition or photoaffinity labeling?

Article abstract of DOI:10.1021/ja9842518

Phenyl esters of 2-benzoylbenzoates were determined to be inhibitors of the serine protease enzymes chymotrypsin and thrombin. Thus, p- guanidinophenyl 2-benzoylbenzoate (lb) inhibited thrombin while the corresponding p-nitrophenyl ester (1a) inhibited ch

Full text of DOI:10.1021/ja9842518

J. Am. Chem. Soc. 1999, 121, 2753-2761
2753
2-Aroylbenzoyl Serine Proteases: Photoreversible Inhibition or
Photoaffinity Labeling?
Paul B. Jones and Ned A. Porter*,†
Contribution from the Departments of Chemistry, Duke UniVersity, Durham, North Carolina 27708, and
Vanderbilt UniVersity, NashVille, Tennessee 37235
ReceiVed December 8, 1998
Abstract: Phenyl esters of 2-benzoylbenzoates were determined to be inhibitors of the serine protease enzymes
chymotrypsin and thrombin. Thus, p-guanidinophenyl 2-benzoylbenzoate (1b) inhibited thrombin while the
corresponding p-nitrophenyl ester (1a) inhibited chymotrypsin activity. Other p-nitrophenyl esters were prepared
that display activity as chymotrypsin inhibitors, three having methoxy group substitution on the benzoyl ring:
2-methoxy (2a), 2,5-dimethoxy (3a), and 2,4,5-trimethoxybenzoyl (4). After incubation with 1b, an acyl thrombin
was isolated that showed no lytic activity but which slowly regained activity in pH 7.4 buffer. Irradiation of
this acyl enzyme with 366 nm light led to an enzyme that showed no gain of lytic activity over time. Incubation
of chymotrypsin with 1a-3a and 4 led to acyl enzymes which showed no activity but which regained activity
slowly. Irradiation of these inactive acyl enzymes with 366 nm light led to a rapid increase in enzyme activity.
The formation of acylchymotrypsins that can be photochemically deacylated is suggested by these data.
Experiments that relate to the mechanism of enzyme acylation and the subsequent photochemistry of the
acylenzymes are reported.
In the last 30 years, knowledge of organic photochemistry
has been combined with a rapidly increasing knowledge of
biochemistry to mimic natural photobiological switches. Metal
binding,1,2 protein/ligand interactions,3 and enzyme activity4,5
have all been controlled photochemically. We have previously
reported a photoreversible serine protease inhibitor based on
the cinnamate chromophore.6-8 This compound inhibits a variety
of serine proteases, such as thrombin, Factor Xa, and chymot-
rypsin, by acylating the active site serine hydroxyl. A Z-E
photoisomerization of the acyl cinnamate results in deacylation
of the active site serine and formation of the native enzyme.
This photochemistry has been exploited in an immunoassay
strategy,9 used clinically in the annealment of skin grafts,10 and
used in the in vivo photoactivation of thrombin.11 This approach
Scheme 1
to control enzyme activity with light involves a photoremovable
alcohol protecting group strategy with the enzyme acting as the
alcohol.
More recently, we reported a photolabile alcohol protecting
group12 which used the photoreduction of 2-benzoyl benzoates
to create a nucleophile in proper position to lactonize and free
the alcohol (Scheme 1). These benzophenones underwent
1-electron reduction when irradiated in the presence of hydrogen
donors (e.g., 2-propanol) and 2-electron reduction when irradi-
ated in the presence of electron donors (e.g., cyclohexylamine).
Given the recent interest in enzyme photoactivation, we sought
to learn whether this benzophenone strategy would be applicable
to photochemical enzyme control. We report here the synthesis
of p-nitrophenyl and p-guanidinophenyl esters of 2-aroyl
benzoate esters and their study as chymotrypsin, thrombin, and
Factor Xa photoreversible inhibitors.
* Address correspondence to this author at Vanderbilt University.
E-mail: n.porter@vanderbilt.edu.
† Vanderbilt University.
(1) Blank, M.; Soo, L. M.; Wasserman, N. H.; Erlanger, B. F. Science
1981, 214, 70-72.
(2) Takeshita, M.; Irie, M. Tetrahedron Lett. 1998, 39, 613-616.
(3) (a) Willner, I.; Rubin, S.; Wonner, J.; Effenberger, F.; Bauerle, P. J.
Am. Chem. Soc. 1992, 114, 3150-3151. (b) for a review see: Willner, I.;
Rubin, S.. Angew. Chem., Int. Ed. Engl. 1996, 35, 367-385.
(4) Willner, I.; Rubin, S.; Riklin, A. J. Am. Chem. Soc. 1991, 113, 3321-
3325.
(5) For a review, see: Porter, N. A.; Bruhnke, J. D.; Koenigs, P.
Biological Applications of Photochemical Switches; Morrison, H., Ed.; John
Wiley and Sons: New York, 1993; Vol. II, Chapter 4, pp 197-242.
(6) (a) Turner, A. D. Ph.D. Thesis, Duke University, 1986. (b) Turner,
A. D.; Monroe, D. M.; Roberts, H. R.; Porter, N. A.; Pizzo, S. V.
Biochemistry 1986, 25, 4929. (c) Turner, A. D.; Pizzo, S. V.; Rozakis, G.
W.; Porter, N. A. J. Am. Chem. Soc. 1987, 109, 1274. (d) Turner, A. D.;
Pizzo, S. V.; Rozakis, G. W.; Porter, N. A. J. Am. Chem. Soc. 1988, 110,
244.
Results and Discussion
Synthesis. A series of p-nitrophenyl esters, 1a-3a and 4-7,
were prepared by dicyclohexylcarbodiimide (DCC) coupling of
p-nitrophenol and the corresponding 2-aroylbenzoic acid. These
esters, shown in Figure 1, were prepared as potential chymot-
rypsin inhibitors. The analogous p-guanidinophenyl esters,
(7) Bruhnke, J. D. Ph.D. Thesis Duke University, 1990.
(8) Koenigs, P. Ph.D. Thesis Duke University, 1993.
(9) Porter, N. A.; Bush, K. A.; Kinter, K. S. J. Photochem. Photobiol.,
B 1997, 38, 61.
(11) Arroyo, J. G.; Jones, P. B.; Porter, N. A.; Hatchell, D. L. Thromb.
Haemostasis 1997, 78, 791.
(12) Jones, P. B.; Pollastri, M. P.; Porter, N. A. J. Org. Chem. 1996, 61,
9455.
(10) Porter, N. A.; Bruhnke, J. D. U.S. Patent 5,114,851, 1993.
10.1021/ja9842518 CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/12/1999

2754 J. Am. Chem. Soc., Vol. 121, No. 12, 1999
Jones and Porter
Scheme 3
Figure 1. p-Nitrophenyl 2-aroylbenzoates used for chymotrypsin
inhibition experiments and p-guanidinophenyl 2-aroylbenzoates used
for thrombin and Factor Xa inhibition experiments.
Scheme 2
methanolic HCl to give amine 15. This amine was acylated by
BCNHS (niotin caproyl N-hydroxysuccinimide) to give 16.
Hydrolysis of the methyl ester with methanolic LiOH gave
carboxylic acid 17. DCC coupling with p-nitrophenol gave 7
(biotin tether) in 23% yield.
Thrombin and Factor Xa Inhibition. None of the p-
guanidinophenyl esters shown in Figure 1 (1b-3b) were found
to be Factor Xa inhibitors. Even when 100 equiv of compound
was added to buffered solutions of Factor Xa, no change in
Factor Xa activity, as measured by chromogenic assay,15 was
observed. Likewise, only one of these compounds was found
to inhibit thrombin. Enzyme activity vs time plots for the
inhibition of thrombin by 20 equiv of compounds 1b-3b are
shown in Figure 2a. Only compound 1b (Phenyl) displays any
tendency toward thrombin inhibition.
Photolysis of a solution of thrombin inhibited by 20 equiv
of 1b (phenyl) did not result in any photoactivation. In fact,
photolysis of the inhibited enzyme resulted in a loss of activity.
Thrombin inhibited by 1b (phenyl) and kept in the dark slowly
regains lytic activity as the enzyme hydrolyzes the ester. On
the other hand, if this solution is photolyzed for 60 min early
in the inhibition, no increase in enzyme activity is observed
during photolysis. Nor does this photolyzed solution eVer regain
enzyme activity (Figure 2b). This suggests that irradiation of
the inhibited thrombin leads only to photoaffinity labeling of
the active site. The labeled thrombin is unable to hydrolyze in
the normal serine protease fashion.
1b-3b, were prepared as potential thrombin or Factor Xa
inhibitors. These compounds were prepared by DCC coupling
of the carboxylic acid with p-N,N′-bis-BOC-guanidinophenol
(8). Subsequent removal of the BOC groups gave the indicated
esters.
Scheme 2 shows the synthesis of a 13C-labeled p-nitrophenyl
ester. Methyl 2-(2,5-dimethoxybenzoyl)benzoate (9) was pre-
pared by Friedel-Crafts acylation of dimethoxybenzene.12 The
o-methoxy group was selectively cleaved by aluminum chloride
to give phenol 10. The phenol was alkylated by 13CH3I to give
9-13C. Hydrolysis of the ester gave the carboxylic acid (11-
13C), which was coupled to p-nitrophenol using DCC to obtain
5 (2*,5-(OMe)2).13
The synthesis of a biotinylated benzophenone inhibitor is
shown in Scheme 3. Methyl p-toluoylbenzoate14 (12) was
prepared by esterification of p-toluoylbenzoic acid (13). The
ester was brominated with 3,3-dimethyl-N,N′-dibromohydantoin
and AIBN in CCl4 to give 14. The bromide was displaced by
hexamethylenetetramine and the salt hydrolyzed by refluxing
(13) For both the inhibitors and corresponding acyl enzymes, the
compound number will be followed by the substitution pattern on the
aromatic ring opposite that containing the carboxyl group. For example,
p-nitrophenyl ester 3a would appear as 3a (2,5-(OMe)2) and acyl-
chymotrypsin as Chymo-1 (phenyl).
Chymotrypsin Inhibition. Unlike the case with thrombin
and Factor Xa, the compounds shown in Figure 1 were found
(15) Erlanger, B. F.; Kokowski, N.; Cohen, W. Arch. Biochem. Biophys.
1961, 95, 271.
(14) Bhatt, M. V.; Rao, G. V.; Rao, K. S. J. Org. Chem. 1979, 44, 984.

2-Aroylbenzoyl Serine Proteases
J. Am. Chem. Soc., Vol. 121, No. 12, 1999 2755
Figure 3. (a) Inhibition of chymotrypsin (3-10 µM) by 1 equiv of
1a (phenyl), 2a (2-OMe), or 3a (2,5-(OMe)2). Phosphate buffered saline
was used as the solvent ([KH2PO4] ) 0.05 M, [NaCl] ) 0.15 M). This
graph indicates a long time course for the inhibition. Compare to panel
b, which shows the same inhibition over a shorter time course. (b)
Inhibition of chymotrypsin (3-10 µM) by 1 equiv of 1a (phenyl), 2a
(2-OMe), or 3a (2,5-(OMe)2). Phosphate buffered saline was used as
the solvent ([KH2PO4] ) 0.05 M, [NaCl] ) 0.15 M). This graph
indicates a short time course for the inhibition.
Figure 2. (a) Inhibition of thrombin by 20 equiv of 1b (phenyl), 2b
(2-OMe), or 3b (2,5-(OMe)2) in the dark. Conditions: pH 7.4 Tris buffer
(50 mM), 150 mM NaCl, [Thrombin] ) 0.9 µM. Inhibitors were added
to aqueous thrombin as solutions in methanol (v/v < 5%). (b) Inhibition
and photolysis of thrombin by 20 equiv of 1b (phenyl) in pH 7.4 Tris
buffer (50 mM), 150 mM NaCl. The curve designated “dark” was
inhibited and the enzyme activity followed over time while the sample
was kept in the dark. The curve designated “photolyzed” was treated
exactly as for the “dark” sample with the exception that the sample
was irradiated (1000 W Hg/Xe arc lamp, grating monochromoter used
to isolate λ ) 366 nm) for 60 min while enzyme activity was zero
(800-860 min). [Thrombin] ) 0.9 µM.
to be excellent chymotrypsin inhibitors. Figure 3 shows the
inhibition of chymotrypsin by three of these compounds, 1b-
3b. Inhibition is rapid and, in the case of 3b (2,5-(OMe)2),
indefinite. The acyl enzymes can be isolated by size-exclusion
chromatography (SEC), shown in Figure 4. In addition to these
three compounds, 4 (2,4,5-(OMe)3) inhibits chymotrypsin. The
inhibition is slower than that for 1b-3b and the acyl enzyme
hydrolyzes more rapidly than for any of these three inhibitors.
Compound 6 (2,4,6-(CH3)3) inhibits chymotrypsin very slowly.
Only after 4 days is enzyme activity reduced to less than 10%
of the initial value when 25 equiv of 6 is added to chymotrypsin.
The fact that 6 is so sluggish as an inhibitor suggests that, in
forming the acyl enzyme, the benzophenone functionality
becomes planar. With two ortho substituents, 6 would resist
this. Scheme 4 shows the formation of acyl enzymes.
Figure 4. Isolation of acyl-chymotrypsin Chymo-3 (2,5-(OMe)2) by
size-exclusion chromatography (Sephadex G-25), pH 6.0 PBS eluent.
Chymotrypsin was inhibited by excess 3a in pH 6.6 PBS ([chymot-
rypsin] ) 8 µM, [KH2PO4] ) 50 mM, [NaCl] ) 150 mM).
Observation of enzyme activity vs time for solutions of acyl
enzyme kept in the dark allow the calculation of a pseudo-first-
order rate constant for hydrolysis of the acyl enzyme. The rate
constants and half-lives for the acyl enzymes are tabulated in
Table 1. At pH 6.0 Chymo-3 (2,5-(OMe)2) is stable indefinitely.
After 10 days less than 8% of the acyl enzyme hydrolyzes.
Because of the very slow nature of this reaction it proved
impossible to accurately measure the rate constant, but it could
be estimated and an upper limit projected. Even at pH 7.4 this
acyl enzyme has a half-life of over 3 weeks. The least stable

2756 J. Am. Chem. Soc., Vol. 121, No. 12, 1999
Jones and Porter
Scheme 4
phenone so that intramolecular hydrogen abstraction, observed
by Wagner in the analogous non-ester-containing benzophe-
nones, does not occur.19
Furthermore, 1,6-hydrogen abstraction does not explain the
photoactivation of Chymo-1 (phenyl). If Chymo-2-4 (methoxy
containing) were to photoactivate by 1,6-hydrogen abstraction
followed by lactonization, a second mechanism would still be
required to explain the photoactivation of Chymo-1 (phenyl)
since this pathway is not available for this substrate. Any
mechanism that explains photoactivation of Chymo-1 (phenyl)
is most likely to depend on hydrogen or electron transfer from
the enzyme to which it is boundsessentially an intramolecular
reaction. The hydrogen abstraction from o-alkoxy groups in aryl
ketones is known not to be competitive with hydrogen or
Table 1. Half-lives and Rate Constants for Hydrolysis of
Acyl-Chymotryspins
acyl enzyme
pH
khydrolysis
half-life
Chymo-3 (2,5-(OMe)2)
Chymo-3 (2,5-(OMe)2)
Chymo-2 (2-OMe)
Chymo-4 (2,4,5-(OMe)3)
Chymo-1 (phenyl)
7.4
6.0
6.0
6.0
6.0
6.0
3.4 × 10-7 s-1
0.006 × 10-7 s-1 a
4.81 × 10-6 s-1
7.83 × 10-5 s-1
1.11 × 10-5 s-1
4.06 × 10-5 s-1
23.8 days
4 yearsa
40 h
2.5 h
17 h
Chymo-7 (biotinylated)
4.7 h
19
electron transfer from solvent.
a Estimated.
There are reports of a benzophenone-containing chymotrypsin
inhibitor that photolabels the active site and leaves the labeled
enzyme with attenuated lytic activity.20,21 On the basis of this
precedent, the apparent photoactivation may be no more than a
photolabeling of the active site. To answer this question, two
labeling experiments were performed.
acyl-chymotrypsin [Chymo-4 (2,4,5-(OMe)3)] has a half-life of
only 2.5 h at pH 6.0.
Acyl-Chymotrypsin Photoactivation. Irradiation of most of
the acyl-chymotrypsins leads to a sharp increase in lytic activity.
Figure 5 shows enzyme activity vs time of photolysis for acyl-
chymotrypsins Chymo-1-3. In the case of Chymo-2 (2-OMe)
and Chymo-3 (2,5-(OMe)2) enzyme activity increases very
quickly when the acyl enzyme is irradiated with 366 nm light.
Acyl-chymotrypsin Chymo-1 (phenyl) photoactivates much
slower and to a lesser extent than do Chymo-2 (2-OMe) or
Chymo-3 (2,5-(OMe)2). Acyl-chymotrypsin Chymo-4 (2,4,5-
(OMe)3) photoactivates similarly to Chymo-2 (2-OMe) and
Chymo-3 (2,5-(OMe)2) with approximately 70-80% return of
lytic activity, although this activation required longer photolysis
times than did Chymo-2 and Chymo-3.
Acyl-chymotrypsin Chymo-6 (2,4,6-(CH3)3) did not photo-
activate to a significant extent. Two hours of photolysis led to
less than a 10% return of lytic activity. The lack of photoacti-
vation of Chymo-6 (2,4,6-(CH3)3) is probably due to photo-
enolization of this substrate.16,17 The presence of ortho alkyl
groups in aryl ketones is known to significantly diminish
reactivity toward hydrogen or electron donors. The pattern of
photoactivation observed for the other acyl-chymotrypsins is
less explicable. It is interesting that two of the acyl enzymes,
Chymo-2 and Chymo-3, photoactivate so much more efficiently
than the other acyl enzymes. The two that photoactivate the
best have the possibility of an intramolecular 1,6-hydrogen
abstraction from the o-methoxy group to the phenone carbonyl.18
Examination of the photochemistry of the corresponding methyl
esters of 2 and 5 revealed no tendency toward 1,6-hydrogen
abstraction upon photolysis. One compound, designed to be
more susceptible to 1,6-hydrogen abstraction, methyl 2-(2-
benzyloxy-5-methoxybenzoyl)benzoate also did not undergo 1,6-
hydrogen abstraction upon photolysis in degassed benzene.
Presumably, the o-carboxyl group alters the reactivity of the
The first experiment used 13C-labeled inhibitor, 3 (2*,5-
(OMe)2), in order that the reaction could be monitored by NMR.
Chymotrypsin was inhibited by 1 equiv of the labeled inhibitor.
Once lytic activity had reached zero, the solution was irradiated
using a 1000 W Hg/Xe arc lamp and grating monochromator
to isolate 366 nm light. The solution regained about 70% of
the initial enzyme activity. Spectra were obtained (13C) for both
the acyl enzyme and the photolyzed solution. The acyl enzyme
displays a signal at δ 61.80. After photoactivation, two signals
appearsa strong one at δ 59.45 and a weaker signal at δ 62.28.
The signal at δ 59.45 appears as a quartet in a non-decoupled
spectrum with a C-H coupling constant of 145.3 Hz. The
chemical shift and coupling constant of this signal is very nearly
identical with that of 2-(2,5-dimethoxybenzoyl)benzoic acid
obtained under similar conditions. For this compound in pH
6.6 deuterated phosphate buffered saline (PBS) and a concentra-
tion of chymotrypsin of approximately 600 µM, the chemical
shift is δ 59.34 and the coupling constant is 145.2 Hz.
Furthermore, when a solution of acyl-chymotrypsin Chymo-3
(2,5-(OMe)2) is photoactivated and then extracted with ethyl
acetate, the organic layer contains a product with m/z 286. This
corresponds to the molecular ion of 2-(2,5-dimethoxybenzoyl)
benzoic acid. Taken together these data indicate that the product
of acyl-chymotrypsin photoactivation is the corresponding
carboxylic acid.
The labeling experiment provides no data regarding the yield
of the acid or active enzyme. That is, the chromogenic assay
used to assess enzyme activity does not indicate if fully active
enzyme is generated. The photoactivation may give, for
example, fully active enzyme in 60% yield or a quantitative
yield of enzyme that has only 60% activity. Affinity-labeled
enzyme resulting from photolysis may be proteolytically active,
(16) (a) Wagner, P. J.; Chen, C.-P. J. Am. Chem. Soc. 1976, 98, 239. (b)
Wagner, P. J. Pure Appl. Chem. 1977, 49, 259.
(17) (a) Das, P. K.; Encinas, M. V.; Small, R. D., Jr.; Scaiano, J. C. J.
Am. Chem. Soc. 1979, 101, 6965. (b) Wagner, P. J.; Wirz, J.; Haag, R.
HelV. Chim. Acta 1977, 60, 2595.
(18) For a review, see: Wagner, P. J.; Park, B. S. Organic Photochem-
istry; Padwa, A., Ed.; Marcel Dekker: New York, 1991; Vol. XI, Chapter
4, pp 227-366.
(19) Wagner, P. J.; Giri, B. P.; Scaiano, J. C.; Ward, D. L.; Gabe, E.;
Lee, F. L. J. Am. Chem. Soc. 1985, 107, 5483.
(20) Campbell, P.; Gioannini, T. L. Photochem. Photobiol. 1979, 29,
883.
(21) Gioannini, T. L.; Campbell, P. Biochem. Biophys. Res. Commun.
1980, 96, 106.

2-Aroylbenzoyl Serine Proteases
J. Am. Chem. Soc., Vol. 121, No. 12, 1999 2757
Figure 5. Photoactivation of acyl-chymotrypsins Chymo-1 (phenyl), Chymo-2 (2-OMe), Chymo-3 (2,5-(OMe)2), and Chymo-4 (2,4,5-(OMe)3).
Conditions: pH 6.0 PBS, 1000 W Hg/Xe arc lamp, λ ) 366 nm, [chymotrypsin] ) 5-10 µM.
or it may have partial activity. A second labeling experiment
was therefore performed to address this question.
Study of a biotinylated inhibitor, 7, allows for the separation
of photoaffinity-labeled enzyme from unlabeled enzyme using
affinity chromatography. The avidin/biotin interaction is one
of the strongest known non-covalent interactions.22 The KD for
dissociation of biotin from avidin is 10-15 M, and it is almost
completely unaffected by changes in pH, organic solvent, or
denaturing agents. This is the case for avidin in its tetrameric
state. In monomeric form the interaction with biotin is still strong
but reversible enough for use as an affinity chromatographic
reagent.23-25 Following photoactivation of the biotinylated acyl-
chymotrypsin Chymo-7, any enzyme that has been photoaffinity
labeled will be retained by an avidin column, while enzyme
that is not labeled will not be retained.
When 20 equiv of inhibitor 7 (biotinylated) is added to
chymotrypsin at pH 6.6, enzyme activity is reduced to <5% in
a few minutes. The acyl enzyme Chymo-7 (biotinylated) was
isolated by SEC using pH 6.0 PBS as an eluent. Hydrolysis of
Chymo-7 (biotinylated) at pH 6.0 occurred with a rate constant
of 4.1 × 10-5 s-1, giving this acyl enzyme a half-life of 4.7 h.
Photolysis of Chymo-7 (biotinylated) led to a 72% increase in
enzyme activity after 90 min. Both the acyl enzyme and the
crude photolysis mixture were chromatographed over im-
mobilized monomeric avidin. The avidin chromatograms for
these two solutions are shown, respectively, in parts a and b of
Figure 6.
Figure 6a shows that Chymo-7 (biotinylated) is bound to the
avidin column. Thus, uninhibited chymotrypsin elutes from the
avidin column with pH 6.0 PBS while the acyl enzyme elutes
only when 2 mM biotin is included in the eluent. This shows
that the biotin on the inhibitor is free to associate with avidin
Figure 6. (a) Avidin affinity chromatography of Chymo-7 (biotinylated
acyl-chymotrypsin). Eluent: fractions 1-6, pH 6.0 PBS; fractions
7-14, pH 6.0 PBS containing 2 mM biotin. Fraction volume was 2
mL. Nonbiotinylated material elutes in fractions 1-6, while biotinylated
outside of the enzyme-binding pocket. All of the enzyme activity
material elutes in fractions 7-14. (b) Avidin affinity chromatography
of the solution placed on the column elutes in fraction 2.
Fractions 8 and 9 display no enzyme activity, and these fractions
undergo photoactivation and dark hydrolysis exactly like the
acyl enzyme isolated by SEC.
of photoactivated Chymo-7 (biotinylated acyl-chymotrypsin). Eluent:
fractions 1-7, pH 6.0 PBS; fractions 8-14, pH 6.0 PBS containing 2
mM biotin. Fraction volume was 2 mL. Nonbiotinylated material elutes
in fractions 1-7, while biotinylated material elutes in fractions 8-14.
Figure 6b shows that most of the enzyme is not bound to
biotin following photoactivation. Fractions 2-4, eluting with
only pH 6.0 PBS and containing only unbiotinylated material,
accounts for >95% of the active enzyme placed on the avidin
column. Fractions 9-11, eluting with pH 6.0 PBS containing
2-mM biotin and all biotinylated material, contain both labeled
enzyme and the biotinylated byproduct of photoactivation. These
fractions contain only very slight lytic activity (0.02-0.03 µM).
This demonstrates that most of the 2-aroyl benzoate inhibitor
acts as a true photoreversible inhibitor and not as a photoaffinity
label. Also, because the biotin-containing material contains such
(22) Green, N. M. AdV. Protein Chem. 1975, 29, 85.
(23) Green, N. M.; Toms, E. J. Biochem. J. 1973, 133, 687.
(24) Guchait, R. B.; Polakis, S. E.; Dimroth, P.; Stoll, E.; Moss, J.; Lane,
M. D. J. Biol. Chem. 1974, 249, 6633.
(25) Henrickson, K. P.; Allen, S. H. G.; Maloy, W. L. Anal. Biochem.
1979, 94, 366.

2758 J. Am. Chem. Soc., Vol. 121, No. 12, 1999
Jones and Porter
Scheme 5
groups increase the rate of this reaction. According to these
reports, the rates of intermolecular hydrogen abstraction for the
acyl-chymotrypsins discussed above should follow as Chymo-1
(phenyl) > Chymo-2 (2-OMe) > Chymo-3 (2,5-(OMe)2) >
Chymo-4 (2,4,5-(OMe)3). As hydrogen atom abstraction be-
comes faster, photoaffinity labeling should increase at the
expense of photoactivation. Therefore, the better hydrogen atom
abstractor should photoactivate in the lowest yield. This is what
is observed, although more data, with a variety of substituents
is needed to say with any certainty that the ability of the
benzophenone to abstract hydrogen is related to efficiency of
photoactivation.
Conclusion
The above experiments demonstrate that p-guanidinophenyl
2-aroylbenzoyl benzoates do not inhibit either thrombin or
Factor Xa or the inhibited enzyme does not photoactivate. The
analogous p-nitrophenyl 2-aroylbenzoates serve as effective
photoreversible inhibitors of chymotrypsin. Acyl-chymotrypsins
are produced that are isolable by size-exclusion chromatography.
These inactive enzymes are stable to hydrolysis for times varying
from several hours to months. Irradiation of the acyl-chymot-
rypsin with 366 nm light produces a sharp increase in enzymatic
activity. Up to 80% of preinhibition activity is returned within
minutes of photolysis. Labeling experiments with 13C indicate
that some of the benzophenone inhibitor acts as a photoaffinity
labeling reagent. The same experiments show that the product
of the photoactivation is the carboxylic acid corresponding to
the p-nitrophenyl ester used as inhibitor. Biotin-labeling experi-
ments confirm that most of the enzyme is photoactivated and
any photoaffinity-labeled enzyme is proteolytically inactive. The
mechanism of the photoactivation remains unknown.
a small amount of enzyme activity, all of the recovered enzyme
activity upon photolysis must be due to enzyme that is not
photoaffinity labeled. These data suggest that the recovery of
the enzyme activity described in the experiments with inhibitors
2 and 3 results from fully active enzyme and that the remainder
of the enzyme is lost to photoaffinity labeling.
Experimental Section
General Synthetic Procedures. Melting points are uncorrected.
Chemicals for which a synthesis is not reported were purchased
commercially. Reactions were carried out in flame-dried glassware
under argon. Gas chromatography was performed on a Hewlett-Packard
5890A gas chromatograph, with flame ionization detection, coupled
to a Hewlett-Packard 3393A integrator (15 m SPB-5, 0.20 mm i.d.).
Photolysis was carried out with a 450 W medium-pressure mercury
lamp (Hanovia), 1000 W mercury/xenon arc lamp (Oriel), or a handheld
These experiments do not account for the mechanism of
carboxylic acid production in the photoactivation step. Scheme
5 shows two possibilities. The first possible mechanism shown
involves an electron transfer. The excited inhibitor abstracts an
electron from the protein to give a radical ion pair. The ketyl
radical anion lactonizes to give the active site serine hydroxyl
and a phthalide radical. Back electron transfer returns the
enzyme to its “normal” state and leaves a phthalide cation.
Trapping of the cation by water produces the carboxylic acid.
The back electron transfer would become more favorable with
the addition of electron-donating groups to the benzophenone.
This may explain the increase in enzyme yield versus photo-
affinity labeling for these acyl-chymotrypsins.
Path B in Scheme 5 shows a simple increase in hydrolysis
rate results from a change in geometry of the benzophenone.
The fact that 6 (2,4,6-(CH3)3) inhibits so slowly suggests that
in the acylation of the serine hydroxyl the benzophenone
becomes planar. If this were true, the change in conformation
(planar to tetrahedral)26 of the benzophenone upon excitation
could explain the photoactivation.
1
4 W mercury lamp (Spectronics) as indicated. All H and 13C nuclear
magnetic resonance (NMR) spectra were recorded using a Varian
INOVA spectrometer at 300 MHz for 1H and 75.4 MHz for 13C unless
otherwise noted. Low-resolution mass spectra (MS) were recorded using
a Hewlett-Packard GC/MS system 5988A mass spectrometer by
chemical ionization (CI) at 70 eV with CH4/NH3 as reagent gas. Fast
atom bombardment (FAB) mass spectra were recorded using a JEOL
JMS-SX102A mass spectrometer with xenon as the fast atom. Infrared
spectra were recorded using a Bomem MB-100 infrared spectrometer
and a NaCl chamber cell, NaCl window, or KBr disk. Elemental
analysis was performed by Atlantic Microlabs (Atlanta, GA). Sephadex
G-25 was purchased from Sigma Chemical Co. Immunopure Im-
mobilized Monomeric Avidin was purchased in kit form from Pierce
(catalog no. 20227).
p-Nitrophenyl 2-Benzoylbenzoate (1a). To a solution of 2-ben-
zoylbenzoic acid (1.2 g, 5 mmol), p-nitrophenol (0.834 g, 6 mmol),
and DMAP (0.122 g, 1 mmol) in pyridine (50 mL) was added DCC
(1.23 g, 6 mmol), and the resulting solution was stirred at 25 °C under
argon for 60 h. The reaction mixture was chilled and filtered. The filtrate
was poured into 1 M HCl, and the solution was extracted 2 × 100 mL
of EtOAc. The combined organic layers were washed with 2 × 100
mL of 1 M HCl, 2 × 100 mL of saturated sodium bicarbonate, and 1
× 50 mL of brine, dried (MgSO4), and concentrated in vacuo. The
crude oil was purified via flash column chromatography (SiO2, CH2-
Cl2) to give a white solid (0.539 g, 1.55 mmol, 31%), mp 86-87 °C:
Furthermore, the observed efficiencies of photoactivation
versus photoaffinity labeling fit this model. Wagner has studied
the effect of substitution on excited-state aryl ketone hydrogen
abstraction in detail.27 This work indicates that electron-donating
substituents on benzophenones slows the rate of intermolecular
hydrogen abstraction upon photolysis. Electron-withdrawing
(26) Turro, N. J. Modern Molecular Photochemistry, University Science
Books: Mill Valley, CA, 1991; pp 38-151, 153-195.
(27) Wagner, P. J.; Truman, R. J.; Scaiano, J. C. J. Am. Chem. Soc.
1985, 107, 7093.

2-Aroylbenzoyl Serine Proteases
J. Am. Chem. Soc., Vol. 121, No. 12, 1999 2759
1H (CDCl3, ppm) 8.17 (d, 1H, J ) 16.0 Hz), 8.14 (d, 2H, J ) 9.3 Hz),
7.77 (d, 2H, J ) 7.1 Hz), 7.74 (td, 1H, J ) 1.4, 7.5 Hz), 7.67 (td, 1H,
J ) 1.4, 7.6 Hz), 7.55 (dt, 2H, J ) 7.5, 16.0 Hz), 7.44 (t, 2H, J ) 7.9
Hz), 6.99 (d, 2H, J ) 9.2 Hz); 13C (CDCl3, ppm) 196.17, 163.42,
154.91, 141.87, 136.85, 133.51, 133.39, 130.83, 130.10, 129.52, 128.71,
128.39, 127.96, 125.07, 122.09. Anal. Calcd for C20H13NO5: C, 69.16;
H, 3.77; N, 4.03; found: C, 69.18; H, 3.81; N, 3.99.
aqueous KOH, 2 × 50 mL of water, and 1 × 50 mL of brine. The
organic layer was dried over MgSO4 and concentrated in vacuo. The
crude oil was purified via flash column chromatography (SiO2, 2%
acetone in chloroform) to give a white foam (1.346 g, 2.17 mmol, 73%),
mp 82-85 °C: IR (CHCl3, cm-1) 3451, 3277, 3155, 2988, 2936, 1725,
1710, 1691, 1654, 1640, 1630, 1547; 1H (CDCl3, ppm) 11.61 (bs, 1H),
10.34 (bs, 1H), 8.02 (dd, 1H, J ) 0.8, 7.8 Hz), 7.68 (dd, 1H, J ) 1.6,
7.2 Hz), 7.59 (td, 1H, J ) 1.6, 7.6 Hz), 7.53 (td, 1H, J ) 1.6,7.6 Hz),
7.44 (m, 4H), 6.96 (td, 1H, J ) 0.8, 7.4 Hz), 6.88 (d, 1H, J ) 8.4 Hz),
6.82 (d, 2H, J ) 9.2 Hz), 3.82 (s, 3H), 1.50 (s, 9H), 1.45 (s, 9H); 13C
(CDCl3, ppm) 195.28, 172.39, 165.27, 159.07, 153.45, 147.54, 144.09,
134.26, 132.27, 131.63, 129.93, 129.49, 128.83, 127.84, 126.97, 123.17,
121.76, 120.55, 112.21, 55.62, 28.07; HRMS calcd for [C32H35N3O8
+ H]+ 590.2504, found 590.2484.
p-Guanidinophenyl 2-(2-Methoxybenzoyl)benzoate (2b). Dry HCl
gas was bubbled through a solution of p-N,N′-bis-BOC-guanidinophenyl
2-(2-methoxybenzoyl)benzoate (0.103 g, 0.17 mmol) in 5 mL of EtOAc
for 5 h. The solution was chilled to 0 °C and the precipitate collected,
washed with ether (100 mL), and dried in vacuo to give the desired
product as a hygroscopic white powder (0.0364 g, 0.086 mmol, 50%):
1H (CD3OD, ppm) 8.05 (dd, 1H, J ) 1.2, 7.2 Hz), 7.69 (m, 3H), 7.57
(td, 1H, J ) 1.6, 8.8 Hz), 7.42 (dd, 1H, J ) 1.6,7.2 Hz), 7.28 (d, 2H,
J ) 8.8 Hz), 7.06 (m, 4H), 3.57 (s, 3H); 13C (CD3OD, ppm) 197.44,
187.49, 177.52, 174.26, 167.14, 160.71, 158.19, 151.12, 145.29, 136.06,
133.70, 132.44, 131.07, 130.88, 128.91, 128.08, 124.13, 121.68, 113.67,
56.13; HRMS calcd for [C22H19N3O4 + H]+ 390.1454, found 390.1445.
2-(2,5-Dimethoxybenzoyl) p-(Nitrophenyl)benzoate (3a). A mix-
ture of 2-(2,5-dimethoxybenzoyl)benzoic acid29 (200 mg, 0.7 mmol),
dicyclohexylcarbodiimide (173 mg, 0.84 mmol, 1.2 equiv), p-nitro-
phenol (117 mg, 0.84 mmol, 1.2 equiv), and 10 mol % DMAP in 7
mL of pyridine was stirred at room temperature for 48 h. The reaction
mixture was then poured into a separatory funnel containing 50 mL of
1 M HCl and 50 mL of EtOAc. The organic layer was then washed 2
× 50 mL of 1 M HCl, 4 × 50 mL of saturated sodium carbonate
solution (until the aqueous wash was no longer yellow), 1 × 50 mL of
distilled water, and 1 × 50 mL of brine, dried over MgSO4 and
concentrated in vacuo. The residue was dissolved in 1 mL of
dichloromethane and cooled to -10 °C. The mixture was then passed
through a filter syringe and concentrated to give a yellow paste. The
paste was dissolved in 3 mL of diethyl ether and allowed to sit for 2
h. At this time yellow crystals were observed in the flask. These were
collected and shown to be the desired product (63.8 mg, 0.16 mmol,
23%), mp 109-110 °C: IR (CHCl3) 1780, 1740, 1630 cm-1; 1H NMR
(CDCl3, ppm) 8.22 (d, 2H, J ) 9 Hz), 8.06 (d, 1H, J ) 8.1 Hz), 7.63
(m, 2H), 7.44 (d, 1H, J ) 6.6 Hz), 7.32 (d, 1H, J ) 3.3 Hz), 7.17 (d,
2H, J ) 9.0 Hz), 7.07 (d, 1H, J ) 3.0 Hz), 6.88 (d, 1H, J ) 9.0 Hz),
3.77 (s, 3H), 3.49 (s, 3H); 13C NMR (CDCl3, ppm) 194.8, 164.7, 155.3,
153.5, 153.4, 145.3, 144.1, 135.9, 132.7, 130.0, 128.0, 127.8, 126.9,
125.6, 125.0, 122.2, 121.4, 114.7, 114.0, 56.2, 55.8; GC/CIMS (CH4/
NH3(g)) m/z 408 (MH+). Anal. Calcd for C22H17NO7: C, 64.86; H,
4.21; N, 3.44. Found: C, 64.79; H, 4.26; N, 3.48.
p-N,N′-Bis-BOC-guanidinophenyl 2-Benzoylbenzoate. To a solu-
tion of 2-benzoylbenzoic acid (0.678 g, 3.0 mmol), p-N,N′-bis-BOC-
guanidinophenol (1.05 g, 3.0 mmol), and DMAP (0.037 g, 0.3 mmol)
in pyridine (5 mL) was added DCC (1 M in pyridine, 6 mL, 6.0 mmol),
and the resulting solution was stirred at 25 °C under argon for 48 h.
The reaction mixture was chilled and filtered. The filtrate was poured
into 1 M HCl, and the solution was extracted 3 × 100 mL of EtOAc.
The combined organic layers were washed with 2 × 100 mL of 1 M
of HCl, 2 × 100 mL of 10% aqueous KOH, 2 × 50 mL of water, and
1 × 50 mL of brine. The organic layer was dried over MgSO4 and
concentrated in vacuo. The crude oil was purified via flash column
chromatography (SiO2, 3% acetone in chloroform) to give a white solid
(0.850 g, 1.52 mmol, 51%), mp 150-160 °C (gas evolution): IR
(CHCl3, cm-1) 3555, 3161, 2996, 2930, 1787, 1737, 1710, 1691, 1677,
1
1658, 1640, 1564; H (CDCl3, ppm) 11.59 (bs, 1H), 10.28 (bs, 1H),
8.17 (d, 1H, J ) 8.0 Hz), 7.73 (d, 2H, J ) 8.4 Hz), 7.70 (td, 1H, J )
1.2, 7.6 Hz), 7.62 (td, 1H, J ) 1.2, 7.6 Hz), 7.53 (t, 1H, J ) 7.6 Hz),
7.44 (m, 4H), 6.72 (d, 1H, J ) 8.8 Hz), 1.49 (s, 9H), 1.45 (s, 9H); 13
C
(CDCl3, ppm) 209.70, 173.70, 164.49, 153.42, 144.91, 141.91, 141.84,
137.12, 133.22, 132.90, 130.90, 130.69, 129.81, 129.47, 128.59, 128.09,
122.99, 121.46, 78.15, 28.14, 28.04; HRMS calcd for [C31H33N3O7 +
H]+ 560.2398, found 560.2390.
p-Guanidinophenyl 2-Benzoylbenzoate Hydrochloride (1b). To
a solution of p-N,N′-bis-BOC-guanidinophenyl 2-benzoylbenzoate
(0.158 g, 0.28 mmol) in CHCl3 (10 mL) was added trifluoroacetic acid
(5 mL). The resulting yellow solution was stirred at 25 °C for 24 h.
The solvent was removed in vacuo and the residue dissolved in 1 M
HCl in ether and stirred for 2 h. Solvent was removed in vacuo to give
a chalky, white powder (0.095 g, 0.24 mmol, 86%): 1H (CD3OD, ppm)
8.21 (d, 1H, J ) 8.0 Hz), 7.83 (td, 1H, J ) 1.6, 7.6 Hz), 7.76 (m, 3H),
7.62 (tt, 1H, J ) 1.6 Hz, 7.2 Hz), 7.50 (m, 3H), 7.25 (d, 2H, J ) 9.2
Hz), 6.97 (d, 2H, J ) 9.2 Hz); 13C (CD3OD, ppm) 198.36, 166.26,
158.20, 150.91, 143.08, 138.32, 134.73, 134.46, 134.00, 131.70, 131.46,
130.63, 129.85, 129.42, 128.09, 124.06; HRMS calcd for [C21H17N3O3
+ H]+ 360.1349, found 360.1342.
p-Nitrophenyl 2-(2-Methoxybenzoyl)benzoate (2a). To a solution
of 2-(2-methoxybenzoyl)benzoic acid28 (1.27 g, 4.7 mmol), p-nitro-
phenol (0.784 g, 5.64 mmol), and DMAP (0.122 g, 1 mmol) in pyridine
(40 mL) was added DCC (1.16 g, 5.64 mmol). The solution was stirred
at 25 °C under argon for 60 h. The mixture was poured into 1 M HCl,
and the resulting solution was extracted 2 × 100 mL of EtOAc. The
combined organic layers were washed 2 × 100 mL of 1 M HCl, 2 ×
100 mL of saturated sodium bicarbonate, 1 × 50 mL of brine, dried,
and concentrated in vacuo. The resulting oil was dissolved in ap-
proximately 5 mL of ether. After 2 days the desired product was
obtained as white needles (0.539 g, 1.43 mmol, 30%), mp 80-82 °C:
1H (CDCl3, ppm) 8.19 (d, 2H, J ) 9.2 Hz), 8.03 (m, 1H), 7.71 (dd,
1H, J ) 1.8 Hz, 7.7 Hz), 7.62 (m, 2H), 7.49 (m, 2H), 7.13 (d, 2H, J
) 9.2 Hz), 7.01 (m, 1H), 6.93 (d, 1H, J ) 8.4 Hz), 3.57 (s, 3H); 13C
(CDCl3, ppm) 195.04, 159.02, 155.29, 143.80, 134.47, 132.60, 131.58,
130.01, 129.95, 129.84, 128.27, 128.17, 126.72, 125.05, 122.03, 120.69,
112.17, 55.60. Anal. Calcd for C21H15NO6: C, 66.84; H, 4.01; N, 3.71;.
Found: C, 66.96; H, 4.06; N, 3.73.
Methyl 2-(2-Hydroxy-5-methoxybenzoyl)benzoate (10). Methyl
2-(2,5-dimethoxybenzoyl)benzoate12 (1.595 g, 5.32 mmol) was dis-
solved in CH2Cl2 (20 mL). Aluminum chloride (5 equiv, 4.4 g, 26.6
mmol) was added in one portion. The color quickly went from colorless
to dark red to yellow-green upon addition of the aluminum chloride.
After being stirred for 8 h at ambient temperature, the reaction mixture
was partitioned between ice water and EtOAc (200 mL each). The
aqueous layer was brought to pH 2 with addition of 1 M HCl, and the
layers were separated. The aqueous layer was extracted twice more
with 150 mL EtOAc. The combined organic layers were washed with
150 mL of brine, dried over magnesium sulfate, and concentrated in
vacuo. The yellow residue was chromatographed (Silica gel, 2% acetone
in chloroform) to give the desired compound as a yellow solid (1.296
g, 4.53 mmol, 85%), mp 71-72 °C: IR (Nujol, cm-1) 3400-2800
p-N,N′-Bis-BOC-guanidinophenyl 2-(2-Methoxybenzoyl)ben-
zoate. To a solution of 2-(2-methoxybenzoyl)benzoic acid27 (0.853 g,
2.98 mmol), p-N,N′-bis-BOC-guanidinophenol (1.29 g, 3.80 mmol),
and DMAP (0.037 g, 0.3 mmol) in pyridine (5 mL) was added DCC
(1 M in pyridine, 12 mL, 12.0 mmol), and the resulting solution was
stirred at 25 °C under argon for 48 h. The reaction mixture was chilled
and filtered. The filtrate was poured into 1 M HCl, and the solution
was extracted 3 × 100 mL of EtOAc. The combined organic layers
were washed with 2 × 100 mL of 1 M HCl, 2 × 100 mL of 10%
1
(bs), 2920 (ss), 1740, 1640, 1620, 1615, 1480, 1460, 1400; H NMR
(CDCl3, ppm) 11.52 (s, 1H), 8.07 (dd, 1H, J ) 1.2, 8.0 Hz), 7.65 (td,
1H, J ) 1.2, 7.4 Hz), 7.58 (td, 1H, J ) 1.2, 7.8 Hz), 7.37 (dd, 1H, J
) 1.6, 7.6 Hz), 7.08 (dd, 1H, J ) 3.0, 8.8 Hz), 6.98 (d, 1H, J ) 9.2
(28) Russell, J.; Thomson, R. H. J. Chem. Soc. 1962, 3379.
(29) Lagodzinski, K. Chem. Ber. 1895, 28, 117.

2760 J. Am. Chem. Soc., Vol. 121, No. 12, 1999
Jones and Porter
Hz), 6.51 (d, 1H, J ) 1.6 Hz), 3.69 (s, 3H), 3.58 (s, 3H); 13C NMR
(CDCl3, ppm) 202.34, 166.00, 156.87, 151.58, 140.03, 132.52, 130.34,
129.89, 128.72, 127.35, 123.86, 119.70, 119.09, 115.26, 55.78, 52.41;
HRMS calcd for [C16H14O5]+ 286.0841, found 286.0832.
vacuo to give the desired product (0.050 g, 73%): 1H (CD3OD, ppm)
8.06 (dd, 1H, J ) 0.8, 7.2 Hz), 7.71 (td, 1H, J ) 1.2, 7.6 Hz), 7.66 (td,
1H, J ) 1.6, 7.6 Hz), 7.42 (dd, 1H, J ) 1.2, 7.2 Hz), 7.28 (m, 3H),
7.30 (d, 1H, J ) 7.2 Hz), 7.16 (dd, 1H, J ) 3.2, 9.2 Hz), 7.05 (m, 3H),
3.75 (s, 3H), 3.49 (s, 3H); 13C (CD3OD, ppm) 197.06, 182.27, 167.06,
158.20, 156.00, 151.14, 145.45, 133.75, 130.96, 130.82, 129.81, 128.76,
128.12, 124.13, 122.09, 116.02, 115.45, 56.70, 56.23; HRMS calcd
for [C23H21N3O5 + H]+ 420.1559, found 420.1596.
p-Nitrophenyl 2-(2,4,5-Trimethoxybenzoyl)benzoate (4). The fol-
lowing reagents were dissolved in dry CH2Cl2:2-(2,4,5-trimethoxyben-
zoyl)benzoic acid (0.612 g, 2 mmol), p-nitrophenol (0.417 g, 3 mmol),
and DMAP (0.122 g, 1 mmol). To this solution was added DCC (0.618
g, 3 mmol) in one portion. The reaction mixture was then stirred, under
argon, at 25 °C for 48 h. The solution was then washed 3 × 50 mL of
1 M NaOH, dried (MgSO4), and concentrated in vacuo. The residue
was flash chromatographed (SiO2, 50% EtOAc in hexane) to afford a
yellow solid, which was recrystallized from ether/hexane to afford the
desired ester as white crystals (0.211 g, 0.48 mmol, 24%), mp 114-
15 °C: 1H (CDCl3, ppm) 3.47 (s, 3H), 3.86 (s, 3H), 3.94 (s, 3H), 6.42
(s, 1H), 7.18 (d, 2H, J ) 9.2 Hz), 7.37 (dd, 1H, J ) 1.1, 7.5 Hz), 7.49
(s, 1H), 7.56 (td, 1H, J ) 1.3, 7.6 Hz), 7.66 (td, 1H, J ) 1.3, 7.5 Hz),
8.11 (dd, 1H, J ) 1.1, 7.6 Hz), 8.22 (d, 2H, J ) 9.2Hz); 13C (CDCl3,
ppm) 193.66, 155.70, 155.31, 154.87, 146.10, 145.25, 143.53, 132.98,
130.12, 128.79, 126.91, 126.81, 125.11, 122.24, 118.06, 112.56, 112.52,
96.93, 56.38, 56.30, 56.18. Anal. Calcd. for C23H19NO8: C, 63.16; H,
4.38; N, 3.20. Found: C, 63.08; H, 4.43; N, 3.21.
Methyl 2-(2-13C-,5-Dimethoxybenzoyl)benzoate (9-13C). Tetrabu-
tylammonium flouride (1 M in THF, 25 mL, 25 mmol) was concentrated
in vacuo. To the flask containing the TBAF was added methyl 2-(2-
hydroxy-5-methoxybenzoyl)benzoate (1.77 g, 6.2 mmol) in dissolved
in DMF (8 mL). Once this solution was homogeneous iodomethane
(1.00 g, 7 mmol) was added via syringe. The initial red color began
dissipating instantly. After 5 h the color had become pale yellow. The
reaction mixture was poured into 50 mL of 2 M NaOH and extracted
3 × 50 mL of EtOAc. The combined organic layers were washed 2 ×
100 mL of 2 M NaOH, 2 × 100 mL of water, and 1 × 100 mL of
brine, dried over magnesium sulfate, and concentrated in vacuo to give
a tan solid. This solid was chromatographed (Silica gel, 20-35% EtOAc
in hexanes) to give the desired product as a white solid (1.442 g, 4.81
mmol, 78%): 1H (CDCl3, ppm) 7.89 (d, 1H, J ) 7.2 Hz), 7.58 (t, 1H,
J ) 7.2 Hz), 7.48 (t, 1H, J ) 7.2 Hz), 7.43 (d, 1H, J ) 7.6 Hz), 6.88
(d, 1H, J ) 8.8 Hz), 6.82 (d, 1H, J ) 2.8 Hz), 6.80 (s, 1H), 6.62 (d,
1H, J ) 2.8 Hz), 3.85 (s, 3H), 3.65 (s, 3H). 3.50 (d, 3H, (13C-1H) J
) 148 Hz). All other spectra matched those of 9.
Methyl 2-(2-Benzyloxy-5-methoxybenzoyl)benzoate. To a solution
of methyl 2-(2-hydroxy-5-methoxybenzoyl)benzoate (0.1442 g, 0.5
mmol) in DMF (6 mL) were added K2CO3 (0.345 g, 2.5 mmol) and
18-crown-6 (0.013 g, 0.05 mmol) in several portions. After the orange
solution had stirred for 15 min, benzyl bromide (0.260 g, 0.2 mL, 1.5
mmol) was added. The solution stirred for 36 h at 25 ° C under argon.
Over the course of the reaction the color faded to pale yellow. The
reaction mixture was poured into EtOAc (100 mL), and the organic
layer was washed with 3 × 50 mL of 10% aqueous KOH, 2 × 50 mL
of water, and brine. The organic layer was dried over MgSO4 and
concentrated in vacuo. The crude oil was chromatographed (SiO2, 5%
acetone in chloroform. Rf ) 0.2) and eluted as a yellow band to give
the desired product as a yellow oil (0.200 g, 0.5 mmol, 100%): 1H
NMR (CDCl3, ppm) 7.66 (m, 1H), 7.42 (dd, 1H, J ) 0.8, 10.0 Hz),
7.39 (m, 1H), 7.28 (m, 2H), 7.20 (m, 3H), 7.02 (dd, 1H, J ) 3.2, 8.8
Hz), 4.72 (s, 2H), 3.81 (s, 3H), 3.62 (s, 3H); 13C NMR (CDCl3, ppm)
195.51, 167.13, 153.62, 152.58, 144.19, 136.01, 131.59, 129.37, 128.94,
128.16, 127.64, 127.49, 127.28, 126.79, 121.01, 114.92, 114.53, 71.07,
55.83, 52.03; HRMS calcd for [C23H20O5]+ 376.1311, found 376.1295.
p-N,N′-Bis-BOC-guanidinophenyl 2-(2,5-Dimethoxybenzoyl)ben-
zoate. To a solution of 2-(2,5-dimethoxybenzoyl)benzoic acid29 (0.853
g, 2.98 mmol), p-N,N′-bis-BOC-guanidinophenol (1.29 g, 3.80 mmol),
and DMAP (0.037 g, 0.3 mmol) in pyridine (5 mL) was added DCC
(1 M in pyridine, 12 mL, 12.0 mmol), and the resulting solution was
stirred at 25 °C under argon for 48 h. The reaction mixture was chilled
and filtered. The filtrate was poured into 1 M HCl, and the solution
was extracted 3 × 100 mL of EtOAc. The combined organic layers
were washed with 2 × 100 mL of 1 M HCl, 2 × 100 mL of 10%
aqueous KOH, 2 × 50 mL of water, and 1 × 50 mL of brine. The
organic layer was dried over MgSO4 and concentrated in vacuo. The
crude oil was purified via flash column chromatography (SiO2, 2%
acetone in chloroform) to give a white foam (1.346 g, 2.17 mmol, 73%),
mp 147-49 °C (gas evolution): IR (CHCl3, cm-1) 3572, 3154, 2993,
2916, 1776, 1725, 1710, 1691, 1677, 1657, 1640, 1630, 1565. 1H
(CDCl3, ppm) 11.60 (bs, 1H), 10.30 (bs, 1H), 8.04 (dd, 1H, J ) 1.2,
7.6 Hz), 7.59 (td, 1H, J ) 1.2, 7.6 Hz), 7.52 (td, 1H, J ) 1.6, 7.6 Hz),
7.48 (d, 2H, J ) 8.8 Hz), 7.38 (dd, 1H, J ) 1.2, 7.2 Hz), 7.30 (d, 1H,
J ) 7.2 Hz), 7.02 (dd, 1H, J ) 3.2, 9.2 Hz), 6.84 (m, 3H), 3.73 (s,
3H), 3.44 (s, 3H), 1.50 (s, 9H), 1.45 (s, 9H); 13C (CDCl3, ppm) 195.03,
165.18, 153.62, 153.51, 147.29, 144.42, 134.32, 132.29, 129.87, 129.29,
128.60, 127.45, 127.22, 123.10, 121.65, 121.16, 114.69, 114.11, 83.90,
79.90, 56.30, 55.84, 28.14, 28.07; HRMS calcd for [C33H37N3O9 +
H]+ 620.2605, found 620.2607.
p-N,N′-Bis-BOC-guanidinophenol (8). To a solution of p-ami-
nophenol (6.54 g, 60 mmol), N,N′-bis-BOC-thiourea30 (16.5 g, 60
mmol), and pyridine (15.8 g, 16.2 mL, 200 mmol) in DMF (150 mL)
was added HgCl2 (19.01 g, 70 mmol). The resulting thick, yellow
solution was stirred for 48 h at 25 °C under argon. The solution was
partitioned between chloroform (500 mL) and water (400 mL). The
aqueous layer was extracted twice more with chloroform (500 mL total),
and the combined organic layers were washed with 2 × 500 mL of
saturated sodium bicarbonate, 2 × 500 mL of water, and brine (200
mL). The organic layer was dried over MgSO4 and concentrated in
vacuo to give a thick red oil. The desired product was obtained as a
pale tan powder (6.75 g, 19.9 mmol, 32%) from hot 2-propanol/water,
mp gas evolution between 180 and 210 °C, decomposition at 210 °C:
IR (CHCl3, cm-1) 3429, 3202, 3265, 2993, 2929, 1790, 1723, 1635,
1
1515, 1479, 1411, 1370; H (CDCl3, ppm) 11.58 (bs, 1H), 9.92 (bs,
1H), 6.99 (d, 2H, J ) 8.8 Hz), 6.54 (d, 2H, J ) 8.8 Hz), 6.82 (d, 2H,
J ) 9.2 Hz), 3.82 (s, 3H), 1.50 (s, 9H), 1.45 (s, 9H); 13C (CDCl3, ppm)
155.77, 155.34, 153.15, 126.93, 126.26, 125.87, 116.13, 83.77, 80.00,
28.10, 27.97; HRMS calcd for [C17H25N3O5]+ 351.1794, found 351.1806.
p-Nitrophenyl 2-(2,4,6-Trimethylbenzoyl)benzoate (6). To a solu-
tion of 2-(2,4,6-trimethylbenzoyl)benzoic acid31 (1.07 g, 4.0 mmol),
p-nitrophenol (0.695 g, 5.0 mmol), and DMAP (0.050 g, 0.4 mmol) in
pyridine (4 mL) was added DCC (1 M in pyridine, 8 mL, 8 mmol).
The reaction was stirred at 25 °C for 72 h, chilled, and filtered. The
filtrate was poured into a separatory funnel containing 50 mL of 1 M
HCl and 50 mL of EtOAc. The organic layer was then washed 2 × 50
mL of 1 M HCl, 4 × 50 mL of saturated sodium carbonate solution
(until the aqueous wash was no longer yellow), 1 × 50 mL of distilled
water, 1 × 50 mL of brine, dried over MgSO4, and concentrated in
vacuo. The residue was dissolved in 1 mL of dichloromethane and
cooled to -10 °C. The mixture was then passed through a filter syringe
and concentrated to give a yellow paste. The paste was dissolved in 3
mL of diethyl ether and allowed to sit for 2 h. At this time yellow
crystals were observed in the flask. These were collected and shown
to be the desired product (0.388 g, 1.0 mmol, 25%), mp 99.5-100 °C:
IR (CHCl3, cm-1) 3154, 2994, 2923, 2865, 1754, 1710, 1665, 1526;
1H (CDCl3, ppm) 8.27 (d, 2H, J ) 9.2 Hz), 7.77 (dd, 1H, J ) 0.8, 7.6
Hz), 7.66 (td, 1H, J ) 1.2, 7.6 Hz), 7.51 (m, 3H), 7.42 (dd, 1H, J )
1.2, 7.6 Hz), 6.88 (s, 2H), 2.31 (s, 3H), 2.13 (s, 6H); 13C (CDCl3, ppm)
p-Guanidinophenyl 2-(2,5-Dimethoxybenzoyl)benzoate Hydro-
chloride (3b). Gaseous hydrogen chloride was bubbled through a
solution of p-N,N′-bis-BOC-guanidinophenyl 2-(2,5-dimethoxybenzoyl)-
benzoate (0.093 g, 0.15 mmol) in EtOAc (10 mL) for 3 h. The cloudy
yellow solution was cooled to 0 °C and the precipitate collected. The
white solid was washed with >25 mL of cold EtOAc and dried in
(30) Iwanowicz, E. J.; Poss, M. A.; Lin, J. Synth. Commun. 1993, 23,
1443.
(31) Bargellini, G. Gazz. Chim. Ital. 1914, 44I, 193.
(32) Storrs, R. W.; Tropper, F. D.; Li, H. Y.; Song, C. K.; Kuniyoshi, J.
K.; Sipkins, D. A.; Li, K. C. P.; Bednarski, M. D. J. Am. Chem. Soc. 1995,
117, 7301.

2-Aroylbenzoyl Serine Proteases
J. Am. Chem. Soc., Vol. 121, No. 12, 1999 2761
199.65, 167.27, 155.79, 145.47, 139.71, 138.04, 135.37, 135.28, 132.73,
132.28, 131.29, 130.76, 129.33, 128.79, 125.24, 122.49, 21.18, 19.64;
HRMS calcd for [C23H19NO5]+ 389.1263, found 389.1259.
Inhibitions. Inhibitions were performed by adding an excess of
inhibitor dissolved in organic solvent (typically methanol, acetonitrile,
or DMSO) to solutions of enzyme. The addition was performed in such
a manner that the volume percentage of organic solvent never exceeded
3%. The activity of the resulting solutions was then assayed (as
described above) as a function of time. The activity of the preinhibited
enzyme solution was assigned a value of 100%. All inhibitions were
performed in the dark at ambient temperature.
Enzyme Assays and Inhibition Measurements. Enzyme solutions
were assayed for protease activity by a chromogenic method. Substrates
were purchased from Chromogenix and diluted to a concentration of
1.00 mM with deionized water. These substrates are non-natural
peptides that contain an enzyme recognition sequence and a terminal
p-nitroanilide. The protease acts to catalyze the hydrolysis of the
p-nitroanilide, releasing p-nitroaniline, which can be measured by its
UV-vis absorbance.
A polystyrene cuvette containing 1 mL of pH 8.3 Tris buffer (30
mM Tris, 3 mM CaCl2, and 400 mM NaCl) and a set amount of the
chromogenic substrate solution was prepared and the solution mixed
thoroughly. The cuvette was angled and 20 mL of enzyme solution
was added to the side of the cuvette. The enzyme and substrate were
mixed by shaking the capped cuvette three times in as rapid manner as
possible. The rate of change of the solution’s absorbance at 402 nm
(λmax p-nitroaniline (pNA) ꢀ ) 10 400) was measured for periods
ranging between 15 and 120 s. Michaelis-Menten kinetics describe
the rate of pNA formation.
r-Chymotrypsin. R-Chymotrypsin (C-4129) was purchased from
Sigma Chemical Co. The activity of R-chymotrypsin was measured
by chromogenic assay using S-2586 (MeO-Suc-Arg-Pro-Tyr-pNA -
HCl; Km ) 32.6 mM, kcat ) 26.77 s-1)7 in pH 8.3 buffer (30 mM Tris,
3 mM CaCl2, and 400 mM NaCl) at room temperature. A polystyrene
cuvette containing 1 mL of the pH 8.3 buffer and 150 mL of S-2586
(1.00 mM in H2O) was prepared. The cuvette was angled, and 20 mL
of chymotrypsin (1-5 mM) in pH 7.4 buffer (50 mM Tris and 150
mM NaCl) was added to the side of the cuvette. The enzyme and
substrate were mixed by shaking the capped cuvette rapidly. The rate
of change of the solution’s absorbance at 402 nm (λmax p-nitroaniline
(pNA) ꢀ ) 10 400) was measured for 70 s. Michaelis-Menten kinetics
describe the rate of pNA formation as described above.
Thrombin. Human thrombin (T-7009) was purchased from Sigma
Chemical Co. The activity of thrombin was measured by chromogenic
assay using S-2238 (H-D-Phe-Pip-Arg-pNA - 2HCl, Km) 27.8 mM,
kcat ) 56.6 s-1)7 in pH 8.3 buffer (30 mM Tris, 3 mM CaCl2, and 400
mM NaCl) at room temperature. A polystyrene cuvette containing 1
mL of the pH 8.3 buffer and 150 mL of the chromogenic substrate
(1.00 mM in H2O) was prepared and mixed thoroughly. The cuvette
was angled, and 20 mL of enzyme solution was added to the side of
the cuvette. The enzyme and substrate were mixed by shaking the
capped cuvette rapidly three times before being placed in a UV-vis
spectrophotometer. The rate of change of the solution’s absorbance at
402 nm was measured for periods ranging between 15 and 120 s.
Michaelis-Menten kinetics describe the rate of pNA formation as
described above.
Acyl Enzyme Purification. In cases where a stable acyl enzyme
was formed, as determined by plot of enzyme activity versus time
following addition of an inhibitor, the acyl enzyme was isolated by
gel filtration. The gel column was prepared by suspending Sephadex
G-25 in the desired buffer and pouring into a plastic column. Ten
milligrams of BSA was eluted through the column to block any
nonspecific protein binding sites in the column. Eluent was collected
until A280 was less than 0.005. The acyl enzyme solution was then loaded
onto the top of the column and eluent passed through the column as
fractions were collected. Each fraction was monitored by UV-vis
spectroscopy for absorbance at 280 nm. Each fraction was ap-
proximately 1.0 mL and was collected into disposable polystyrene
cuvettes. Fractions containing significant quantities of protein were
collected and combined.
Photoactivation. Samples of acyl enzyme were irradiated by one
of the two methods described below.
Method 1: Samples to be irradiated were placed either in plastic
pipet tips or acrylic cuvettes and irradiated with a 4 W 366 nm lamp
(Spectronics Corp., Westbury, NY). The gain of enzymatic activity was
measured by chromogenic assay (as described above) on aliquots
removed from the sample over time or, in the case of pipet tip
photoactivations, the entire sample was assayed.
Method 2: Samples were placed in a polystyrene UV-vis cuvette
and irradiated by a 1000 W Hg/Xe arc lamp, equipped with grating
monochromoter, at a distance of 100 cm. The monochromoter was set
to 366 nm with a 20 nm bandwidth. At specified times, 20 µL of the
sample was removed and assayed.
Avidin Affinity Chromatography.22-24 Samples containing poten-
tially biotinylated protein were purified via affinity chromatography.
The column was prepared by first blocking nonspecific or nonreversible
binding sites by eluting 3 × 2 mL of pH 7.2 PBS containing 2 mM
biotin through the column. Next, pH 2.8 glycine (0.1 M) buffer was
eluted through the column (3 × 4 mL). The buffer in which the sample
was dissolved was then passed through the column (4 × 4 mL). The
sample was loaded onto the column in a 2 mL volume. Fractions were
collected by stepwise addition of 2 mL eluent onto the column. UV
absorbance was monitored. After at least five fractions were collected
the eluent was switched to pH 6.0 PBS containing 2 mM biotin. Once
all of the protein was washed off of the column it was regenerated by
elution with glycine buffer.
Factor Xa. Bovine Factor Xa (F-2027) was purchased from Sigma
Chemical Co. The activity of Factor Xa was measured by chromogenic
assay using S-2222 (Bz-Ile-Glu-Gly-Arg-pNA - HCl, Km) 30.0 mM,
kcat ) 100 s-1)7 in pH 8.3 buffer (30 mM Tris, 3 mM CaCl2, and 400
mM NaCl) at room temperature. A polystyrene cuvette containing 1
mL of the pH 8.3 buffer and 150 mL of the chromogenic substrate
(1.00 mM in H2O) was prepared and mixed thoroughly. The cuvette
was angled and 20 mL of enzyme solution was added to the side of
the cuvette. The enzyme and substrate were mixed by shaking the
capped cuvette rapidly three times before being placed in a UV-vis
spectrophotometer. The rate of change of the solution’s absorbance at
402 nm was measured for periods ranging between 15 and 120 s.
Michaelis-Menten kinetics describe the rate of pNA formation as
described above.
Acknowledgment. This research was supported by a grant
from NIH (HL 17921). P.B.J. acknowledges support from an
NIH Biological Training Grant.
Supporting Information Available: Experimental proce-
dures and characterization data for compounds 7 and 14-17
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
JA9842518