A semisynthetic metalloenzyme based on a protein cavity that catalyzes the enantioselective hydrolysis of ester and amide substrates

Article abstract of DOI:10.1021/ja970820k

In an effort to prepare selective and efficient catalysts for ester and amide hydrolysis, we are designing systems that position a coordinated metal ion within a defined protein cavity. Here, the preparation of a protein-1,10-phenanthroline conjugate and the hydrolytic chemistry catalyzed by this construct are described. Iodoacetamido-1,10-phenanthroline was used to modify a unique cysteine residue in ALBP (adipocyte lipid binding protein) to produce the conjugate ALBP-Phen. The resulting material was characterized by electrospray mass spectrometry, UV/vis and fluorescence spectroscopy, gel filtration chromatography, and thiol titration. The stability of ALBP-Phen was evaluated by guanidine hydrochloride denaturation experiments, and the ability of the conjugate to bind Cu(II) was demonstrated by fluorescence spectroscopy. ALBP-Phen-Cu(II) catalyzes the enantioselective hydrolysis of several unactivated amino acid esters under mild conditions (pH 6.1, 25°C) at rates 32-280-fold above the background rate in buffered aqueous solution. In 24 h incubations 0.70 to 7.6 turnovers were observed with enantiomeric excesses ranging from 31% ee to 86% ee. ALBP-Phen-Cu(II) also promotes the hydrolysis of an aryl amide substrate under more vigorous conditions (pH 6.1, 37°C) at a rate 1.6 x 10⁴-fold above the background rate. The kinetics of this amide hydrolysis reaction fit the Michaelis-Menten relationship characteristic of enzymatic processes. The rate enhancements for ester and amide hydrolysis reported here are 10²-10³ lower than those observed for free Cu(II) but comparable to those previously reported for Cu(II) complexes.

Full text of DOI:10.1021/ja970820k

J. Am. Chem. Soc. 1997, 119, 11643-11652
11643
A Semisynthetic Metalloenzyme Based on a Protein Cavity
That Catalyzes the Enantioselective Hydrolysis of Ester and
Amide Substrates
Ronald R. Davies and Mark D. Distefano*
Contribution from the Department of Chemistry, UniVersity of Minnesota,
Minneapolis, Minnesota 55455
X
ReceiVed March 13, 1997
Abstract: In an effort to prepare selective and efficient catalysts for ester and amide hydrolysis, we are designing
systems that position a coordinated metal ion within a defined protein cavity. Here, the preparation of a protein-
1
1
,10-phenanthroline conjugate and the hydrolytic chemistry catalyzed by this construct are described. Iodoacetamido-
,10-phenanthroline was used to modify a unique cysteine residue in ALBP (adipocyte lipid binding protein) to
produce the conjugate ALBP-Phen. The resulting material was characterized by electrospray mass spectrometry,
UV/vis and fluorescence spectroscopy, gel filtration chromatography, and thiol titration. The stability of ALBP-
Phen was evaluated by guanidine hydrochloride denaturation experiments, and the ability of the conjugate to bind
Cu(II) was demonstrated by fluorescence spectroscopy. ALBP-Phen-Cu(II) catalyzes the enantioselective hydrolysis
of several unactivated amino acid esters under mild conditions (pH 6.1, 25 °C) at rates 32-280-fold above the
background rate in buffered aqueous solution. In 24 h incubations 0.70 to 7.6 turnovers were observed with
enantiomeric excesses ranging from 31% ee to 86% ee. ALBP-Phen-Cu(II) also promotes the hydrolysis of an aryl
4
amide substrate under more vigorous conditions (pH 6.1, 37 °C) at a rate 1.6 × 10 -fold above the background rate.
The kinetics of this amide hydrolysis reaction fit the Michaelis-Menten relationship characteristic of enzymatic
2
3
processes. The rate enhancements for ester and amide hydrolysis reported here are 10 -10 lower than those observed
for free Cu(II) but comparable to those previously reported for Cu(II) complexes.
Introduction
advantage of a substrate binding site present in the starting
enzyme, their specificities are often limited to structures that
bind to the native enzyme. This is a significant impediment
for the development of catalysts with new specificities.
Our design for the development of new catalysts is based on
Enzymes perform chemical reactions with high specificity
and rate enhancement in aqueous media at low temperatures
and neutral pH. These features have made these catalysts
attractive for a variety of purposes in medicine and industry.
However, the use of naturally occurring enzymes is restricted
by their inherent specificity. To circumvent this limitation, the
development of artificial enzymes has received considerable
attention. One approach for the design of new enzymes is to
modify a known protein or enzyme at a defined site with a
the use of a protein cavity that can accommodate a variety of
12
substrates.
By introducing catalytic functionality into the
constrained environment of the cavity it should be possible to
impart selectivity to chemical reactions occurring within.
Different specificities and reactivities could then be optimized
by mutagenesis of particular cavity residues. A key requirement
for constructing this type of system is the availability of a
suitable protein cavity. In this regard, adipocyte lipid binding
cofactor or new functional group to create a semisynthetic
system with novel properties.1
-10
A number of such constructs
have been prepared resulting in a new generation of “enzyme-
protein (ALBP) is an excellent candidate. ALBP is a small,
like” catalysts.11 However, since most of these systems take
3
1
31 amino acid, protein that possesses a 600 Å cavity formed
from two orthogonal planes of antiparallel â-sheet secondary
structure. This large cavity size provides ample space for both
X
Abstract published in AdVance ACS Abstracts, November 1, 1997.
13
(1) Collins, A. N.; Sheldrake, G. N.; Crosby, J. Chirality in Industry:
catalyst and putative substrate and effectively sequesters them
from the exterior. Several crystal structures of ALBP with
different ligands bound within the cavity have been solved as
well as a number of additional structures of closely related fatty
The Commercial Manufacture and Applications of Optically ActiVe
Compounds; Collins, A. N., Sheldrake, G. N., Crosby, J., Eds.; John Wiley
&
Sons: Chichester, 1992.
2) (a) Kaiser, E. T.; Lawrence, D. S. Science 1984, 226, 505-511. (b)
Levine, H. L.; Nakagawa, Y.; Kaiser, E. T. Biochem. Biophys. Res. Comm.
(
14
1
977, 76, 64-70.
acid binding proteins. Recently, a structure of a similar protein
complexed with palmitate was determined using NMR meth-
(
3) Corey, D. R.; Schultz, P. G. Science 1987, 238, 1401-1403.
4) (a) Polgar, L.; Bender, M. L. J. Am. Chem. Soc.1966, 88, 3153-
(
1
5
ods. All of these tertiary structures are quite similar despite
their highly divergent primary sequences. This is important
3
154. (b)Neet, K. E.; Koshland, D. E. Proc. Natl. Acad. Sci. U.S.A. 1966,
6, 1606-1609.
5
(5) Wu, Z. P.; Hilvert, D. J. Amer. Chem. Soc. 1989, 111, 4513-4514.
(6) Planas, A.; Kirsch, J. F. Biochemistry 1991, 30, 8268-8273.
(7) Wuttke, D. S.; Gray, H. B.; Fisher, S. L.; Imperiali, B. J. Am. Chem.
(12) Kuang, H.; Brown, M. L.; Davies, R. R.; Young, E. C.; Distefano,
M. D. J. Am. Chem. Soc. 1996, 118, 10702-10706.
(13) Banaszak, L.; Winter, N.; Xu, Z.; Bernlohr, D. A.; Cowan, S.; Jones,
T. A. AdV. Protein Chem. 1994, 45, 89-151.
(14) (a) Sacchettini, J. C.; Meininger, T. A.; Lowe, J. B.; Gordon, J. L.;
Banaszak, L. J. J. Biol. Chem. 1987, 262, 5428-5430. (b) Sacchettini, J.
C.; Gordon, J. I.; Banaszak, L. J. J. Mol. Biol. 1989, 208, 327-339. (c)
Xu, Z.; Bernlohr, D. A.; Banaszak, L. J. Biochemistry 1992, 31, 3484-
3492. (d) LaLonde, J. M.; Bernlohr, D. A.; Banaszak, L. J. Biochemistry
1994, 33, 4885-4895.
Soc. 1993, 115, 8455-8456.
(
(
34.
8) Imperiali, B.; Roy, R. S. J. Am. Chem. Soc. 1994, 116, 12083-12084.
9) Suckling, C. J.; Zhu, L. M. Bioorg. Med. Chem. Lett. 1993, 3, 531-
5
(
10) Germanas, J. P.; Kaiser, E. T. Biopolymers 1990, 29, 39-43.
(11) (a) Bell, I. M.; Hilvert, D. New Biocatalysts Via Chemical Modifica-
tion; Bell, I. M., Hilvert, D., Eds.; John Wiley & Sons, Ltd.: Chichester,
994; pp 73-88. (b) Schultz, P. G. Science 1988, 240, 426-433.
1
S0002-7863(97)00820-2 CCC: $14.00 © 1997 American Chemical Society

11644 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
DaVies and Distefano
Figure 1. Stereoview of model of ALBP-Phen.
because it suggests that most of the cavity residues are amenable
to substitution. Indeed, a number of mutagenesis studies have
been carried out with ALBP and related fatty acid binding
catalyzes the enantioselective hydrolysis of several unactiVated
amino acid esters; these include methyl, ethyl, and isopropyl
esters of the amino acids alanine, tyrosine, and serine. Enan-
tioselectivities as high as 86% ee and catalysis with as many as
7.6 turnovers in 24 h have been obtained. The hydrolysis of
an aryl amide substrate and an aryl ester are also described.
16
proteins resulting in very little change in the overall structure.
Transition metal complexes catalyze a diverse array of
chemical reactions. The ligand system 1,10-phenanthroline has
been attached to a number of proteins to perform oxidative
chemistry.17 Cupric ions, Cu(II)-phenanthroline, and related
Results
metal-ligand systems have also been shown to promote the
hydrolysis of esters and amides resulting in rate enhancements
Preparation of ALBP-Phen. ALBP-Phen (2) was prepared
by reacting iodoacetamido 1,10-phenanthroline (1) with ALBP
at ambient temperature for 48 h as shown below. The extent
of reaction with Cys117 was quantitated by thiol titration with
Ellman’s reagent under denaturing conditions. Using this
5
6
18-20
1
0 -10 -fold above the background rate.
Based on these
precedents, we decided to investigate the hydrolytic chemistry
catalyzed by copper phenanthroline within the ALBP cavity.
2
1
In the work described here, we have attached a phenanthroline
ligand to a unique cysteine residue present in the interior of
ALBP in order to position a Cu(II) ion within the protein cavity.
The resulting semisynthetic metalloprotein, ALBP-Phen-Cu(II)
(
15) Hodsdon, M. E.; Ponder, J. W.; Cistola, D. P. J. Mol. Biol. 1996,
64, 585-602.
16) (a) Sha, R. S. C. D. K.; Xu, Z.; Banaszak, L. J.; Bernlohr, D. A. J.
Biol. Chem. 1993, 268, 7885-7892. (b) Jiang, N.; Frieden, C. Biochemistry
993, 32, 11015-11021. (c) Frieden, C.; Jiang, N.; Cistola, D. P.
2
(
1
Biochemistry 1995, 34, 2724-2730. (d) Kim, K.; Cistola, D. P.; Frieden,
C. Biochemistry 1996, 35, 7553-7558. (e) Cistola, D. P.; Kim, K.; Rogl,
H.; Frieden, C. Biochemistry 1996, 35, 7559-7565. (f) Stump, D. G.; Chytil,
F. J. Biol. Chem. 1991, 266, 4622-4630.
method, alkylation efficiencies of 85-95% were typically
obtained. A model for ALBP-Phen, based on the X-ray crystal
structure of an oleic acid-ALBP complex, is shown in Figure
(
17) (a) Sigman, D. S.; Chen, C.-H. B.; Gorin, M. B. Nature 1993, 363,
74-475. (b) Pan, C. Q.; Feng, J. A.; Finkel, S. E.; Landgraf, R.; Sigman,
D.; Johnson, R. C. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 1721-1725.
c)) Ebright, R. H.; Ebright, Y. W.; Pendergrast, P. S.; Gunasekera, A. Proc.
2
2
4
1.
Characterization of ALBP-Phen. Following preparation,
(
ALBP-Phen was purified by gel filtration chromatography to
remove unincorporated 1, and the resulting protein was char-
acterized by a number of spectroscopic and physical methods
sumarized in Table 1. Figure 2 shows the UV/vis spectra of
ALBP (A) and ALBP-Phen (B) at equal concentrations. These
spectra show that attachment of the phenanthroline moiety to
Natl. Acad. Sci. U.S.A. 1990, 87, 2882-2886. (d) Francois, J. C.; Behmoaras,
T. S.; Chassignol, M.; Thuong, N. T.; Sun, J. S.; Helene, C. Biochemistry
1
988, 27, 2272-2276. (e) Sigman, D. S. Acc. Chem Res. 1986, 19, 180-
1
86. (f) Sigman, D. S.; Bruice, T. W.; Mazumder, A.; Sutton, C. L. Acc.
Chem Res. 1993, 26, 98-104.
(
18) (a) Kroll, H. J. Am. Chem. Soc. 1952, 74, 2036-2039. (b) Bender,
M. L.; Turnquist, B. W. J. Amer. Chem. Soc. 1957, 79, 1889-1893. (c)
Meriwether, L.; Westheimer, F. H. J. Am. Chem. Soc. 1956, 78, 5119-
5
123.
(21) For examples of other artificial metalloproteins and metalloenzymes,
see: (a) Sasaki, T.; Kaiser, T. E. Biopolymers 1990, 29, 79-88. (b) Handel,
T.; DeGrado, W. F. J. Am. Chem. Soc. 1990, 112, 6710-6712. (c) Regan,
L.; Clarke, N. D. Biochemistry 1990, 29, 10878-10883. (d) Braxton, S.;
Wells, J. A. Biochemistry 1992, 31, 7796-7801. (e) Higaki, J. N.; Fletterick,
R. J.; Craik, C. S. Trends Biochem. Sci. 1992, 17, 100-104. (f) Mack, D.
P.; Dervan, P. B. Biochemistry 1992, 31, 9399-9405. (g) Ghadiri, M. R.;
Soures, C.; Choi, C. J. Am. Chem. Soc. 1992, 114, 4000-4002. (h)
Desjarlais, J. R.; Berg, J. M. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 2256-
2260. (i) Pessi, A.; Bianchi, E.; Crameri, A.; Venturini, S.; Tramotano, A.;
Sollazzo, M. Nature 1993, 362, 367-369. (j) McGrath, M. E.; Haymore,
B. L.; Summers, N. L.; Craik, C. S.; Fletterick, R. J. Biochemistry 1993,
32, 1914-1919. (k) Wade, W. S.; Koh, J. S.; Han, N.; Hoekstra, D. M.;
Lerner, R. A. J. Am. Chem. Soc. 1993, 115, 4449-4456. (l) Cuenoud, B.;
Schepartz, A. Science 1993, 259, 510-513. (m) Bryson, J. W.; Betz, S. F.;
Lu, H. S.; Suich, D. J.; Zhou, H. X.; O’Neil, K. T.; DeGrado, W. F. Science
1995, 270, 935-941.
(19) (a) Sayre, L. M.; Reddy, K. V.; Jacobson, A. R.; Tang, W. Inorg.
Chem. 1992, 31, 935-937. (b) Reddy, K. V.; Jacobson, A. R.; Kung, J. I.;
Sayre, L. M. Inorg. Chem. 1991, 30, 3520-3525. (c) Weijnen, J. G. J.;
Koudijs, A.; Engbersen, J. F. J. J. Chem. Soc., Perkin Trans. 2 1991, 1121-
1
1
125. (d) Weijnen, J. G. J.; Koudijs, A.; Engbersen, J. F. J Org. Chem.
992, 57, 7258-7265. (e) Nakon, R.; Rechani, P. R.; Angelici, R. J. J.
Am. Chem. Soc. 1974, 96, 2117-2120.
(20) (a) Brown, R. S.; Zamkanei, M.; Cocho, J. L. J. Am. Chem.
Soc.1984, 106, 5222-5228. (b) Schepartz, A. R. B. J. Am. Chem. Soc.
1
987, 109, 1814-1826. (c) Hay, R. W.; Basak, A. K. Chem. Soc., Dalton
Trans. 1982, 1819-1823. (d) Hay, R. W.; Banerjee, P. J. Chem. Soc., Dalton
Trans. 1980, 362-365. (e) Hay, R. W.; Banerjee, P. K. J. Inorg. Biochem.
1
3
1
981, 14, 147-154. (f) Burgeson, I. E.; Kostic, N. M. Inorg. Chem. 1991,
0, 4299-4301. (g) Linkletter, B.; Chin, J. Angew. Chem., Int. Ed. Engl.
995, 34, 472-474. (h) Buckingham, D. A.; Foster, D. M.; Sargeson, A.
M. J. Am. Chem. Soc. 1969, 91, 4102-4110. (i) Suh, J. Acc. Chem Res.
1
992, 25, 273-278. (j) Chin, J. Acc. Chem Res. 1991, 24, 145-152.
(22) Kraulis, P. J. J. Appl. Crystallogr. 1991, 24, 946-956.

A Semisynthetic Metalloenzyme
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11645
Table 1. Properties of ALBP and ALBP-Phen
property
ALBP
ALBP-Phen
titratable thiols (from DTNB titration)
molecular mass (from electrospray MS)
0.95 thiol/mol
0.10 thiol/mol
14,541.59 ( 1.75 Da
14777.04 ( 2.78 Da
-
1
-1
-1
-1
λ
max, ꢀ (from UV/vis spectroscopy)
molecular mass (from gel filtration)
280 nm, 17 mM ‚cm
274 nm, 29 mM ‚cm
19 600 Da
18 900 Da
1.2 M
denaturation midpoint ([guanidine‚HCl])
1.4 M
Figure 2. UV/vis spectra of ALBP, ALBP-Phen, and phenanthroline:
Figure 4. Gel filtration molecular mass analysis of ALBP-Phen: (b)
standards, (4) ALBP, and (3) ALBP-Phen.
(A) spectrum of ALBP (6.5 µM); (B) spectrum of ALBP-Phen (6.5
µM); (C) spectrum of 1,10-phenanthroline, and (D) difference spectrum
of ALBP-Phen minus ALBP.
Figure 5. Guanidine hydrochloride denaturation profiles for ALBP
and ALBP-Phen obtained by fluorescence spectroscopy: (b) ALBP
and (O) ALBP-Phen.
Figure 3. Fluorescence emission spectra of ALBP and ALBP-Phen:
(
A) spectrum of ALBP (1.0 µM) and (B) spectrum of ALBP-Phen (1.0
µM). Spectra were obtained by excitation at 275 nm.
parable folded structures; no large perturbations of the ALBP
structure occur upon alkylation with 1. It should be noted that
the masses for both proteins obtained by gel filtration chroma-
tography differ from those determined by mass spectrometry
and denaturing electrophoresis (data not shown). These dif-
ferences most likely reflect the extended, largely â-sheet
secondary structure of ALBP that differs from the more compact
globular proteins typically used to standardize gel filtration
chromatography experiments. As shown in Figure 5, denatur-
ation experiments monitored by fluorescence spectroscopy using
guanidinium chloride as a denaturant were perfomed to estimate
the effect of phenanthroline attachment on protein stability.
ALBP and ALBP-Phen gave values of 1.4 and 1.2 M,
respectively, for their denauration midpoints indicating that
conjugation of ALBP with phenanthroline does not result in
any large destabilization of the protein. Finally, it should also
be noted that titration of ALBP-Phen with the fluorescent fatty
acid analogs cis-paranaric acid and 12-anthroloxyoleic acid
indicate that the conjugate lacks the ability to bind fatty acids.
These results are similar to those observed by Bernlohr and co-
workers who have previously reported that modification of
Cys117 with a number of agents greatly attenuates fatty acid
ALBP results in a shift in λmax from 280 to 274 nm accompanied
by a substantial increase in extinction. Subtraction of these two
spectra leads to a difference spectrum (D) that is similar to the
spectrum of unsubstituted 1,10-phenanthroline (C). Figure 3
shows fluorescence emission spectra for ALBP (A) and ALBP-
Phen (B). An emission maximum at 406 nm appears only in
the spectrum of ALBP-Phen and can be attributed to the
phenanthroline fluorophore. It is also interesting to note that
the tryptophan fluorescence centered at 340 nm in ALBP is
substantially reduced in ALBP-Phen; this is likely due to
quenching by the phenanthroline. Additional evidence for
phenanthroline attachment derives from electron spray ionization
mass spectrometry. Spectra obtained from samples of ALBP
and ALBP-Phen gave masses of 14 541.59 ((1.75) Da and
1
4 777.04 ((2.78) Da, respectively. These results give an
observed mass difference of 235.45 Da that is in close agreement
with the calculated mass difference of 235 Da attributable to
the phenanthroline moiety. Gel filtration chromatography was
performed on samples of ALBP and ALBP-Phen under non-
denaturing conditions, and the apparent masses of the proteins
were estimated by comparing their elution volumes with those
from a set of standard proteins (Figure 4). ALBP and ALBP-
Phen gave masses of 19 600 and 18 900 Da. The similarity of
these values suggests that ALBP and ALBP-Phen have com-
1
4c,d
binding to ALBP.
Copper Binding to ALBP-Phen. The binding of Cu(II) to
ALBP-Phen was first examined by fluorescence measurements.
As shown in Figure 6, addition of Cu(II) to a solution containing

11646 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
DaVies and Distefano
Figure 6. 1,10-Phenanthroline fluorescence quenching by Cu(II).
Fluorescence emission spectra of (A) 1,10-phenanthroline and (B)
1
,10-phenanthroline in the presence of 2 equivs of Cu(II).
Figure 8. Fluorescence of 1,10-phenanthroline-Cu(II) in the presence
of picolinic acid. (A) Emission spectra of 1,10-phenanthroline (1),
1
,10-phenanthroline-Cu(II) (2), and 1,10-phenanthroline-Cu(II) in the
presence of 4 equivs of picolinic acid (3). (B) Fluorescence at 362 nm
of 1,10-phenanthroline-Cu(II) in the absence (b) and presence (O) of
picolinate. The fluorescence of 1,10-phenanthroline is indicated by (9),
and the fluorescence of 1,10-phenanthroline-Cu(II) following the
addition of EDTA is given by (0).
Figure 7. Phenanthroline fluorescence quenching by Cu(II) in
ALBP-Phen. Fluorescence emission spectra of (A) ALBP-Phen and
(B) ALBP-Phen in the presence of 40 equivs of Cu(II).
phenanthroline (A) resulted in substantial quenching of the
phenanthroline fluorescence (B). Figure 7 shows a similar
experiment with ALBP-Phen. Addition of Cu(II) to a sample
of ALBP-Phen (A) resulted in quenching of the phenanthroline
fluorescence (B); several equivs of Cu(II) were necessary to
drive this reaction to completion. This may reflect the fact that
the ALBP cavity is lined with several cationic residues that are
involved in fatty acid binding which may repel the entry of the
additional cations. Treatment of ALBP-Phen with Cu(II)
followed by gel filtration to remove unbound metal ions resulted
in a sample whose phenanthroline fluorescence remained
quenched upon prolonged storage and suggested that Cu(II)
remains bound and is not subject to loss over several weeks.
Although the above experiments suggested that Cu(II)
remained bound to ALBP-Phen for prolonged periods of time
in the absence of competing ligands free in solution, it was also
important to analyze the stability of this complex in the presence
of exogenous ligands. Thus the phenanthroline-Cu(II) and
ALBP-Phen-Cu(II) complexes were examined in the presence
of EDTA, picolinic acid, and alanine. Additon of EDTA to
solutions of phenanthroline-Cu(II) and ALBP-Phen-Cu(II)
resulted in the return of phenanthroline fluorescence indicating
that Cu(II) was being removed by this powerful chelating agent
Figure 9. Fluorescence of 1,10-phenanthroline-Cu(II) in the presence
of alanine. (A) Emission spectra of 1,10-phenanthroline (1), 1,10-
phenanthroline-Cu(II) (2), and 1,10-phenanthroline-Cu(II) in the pres-
ence of excess EDTA (3). (B) Fluorescence at 362 nm of 1,10-
phenanthroline-Cu(II) in the absence (b) and presence (m) of alanine.
The fluorescence of 1,10-phenanthroline is indicated by (9), and the
fluorescence of 1,10-phenanthroline-Cu(II) following the addition of
EDTA is given by (0).
that titration of phenanthroline-Cu(II) with alanine results in
only a small return in phenanthroline fluorescence even after
the addition of several equivalents. Thus, alanine is less
effective than picolinic acid at removing the Cu(II) from
phenanthroline.
Experiments similar to those described above were also
performed with samples of ALBP-Phen-Cu(II). However, in
these cases, significantly different results were obtained. Ti-
tration of ALBP-Phen-Cu(II) with picolinic acid, shown in
Figure 10A, results in a decrease in the protein fluorescence at
336 nm, but no change in the phenanthroline fluorescence at
406 nm. In fact, Figure 10B shows that the 406 nm fluorescence
does not increase even after the addition of 4 equivs of picolinate
although there is an immediate increase upon addition of EDTA.
The decrease in protein fluorescence at 336 nm suggests that
picolinic acid is binding to the protein although the nature of
this complex remains undefined; picolinate may or may not be
(data not shown). In contrast, addition of picolinate or alanine
gave significantly different results. As shown in Figure 8A, in
PIPES buffer (pH 6.1), the fluorescence of phenanthroline is
quenched by Cu(II), but this effect can be partially reversed by
the addition of picolinate. As shown in Figure 8B, titration of
phenanthroline-Cu(II) with up to 4 equivs of picolinate results
in the return of approximately half of the phenanthroline
fluorescence. These results are in contrast to those obtained
with EDTA shown in Figure 9A. Clearly, picolinate is less
effective than EDTA at removing the Cu(II). Figure 9B shows

A Semisynthetic Metalloenzyme
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11647
Figure 10. Fluorescence of ALBP-Phen-Cu(II) in the presence of
picolinic acid. (A) Emission spectra of ALBP-Phen-Cu(II) (1), ALBP-
Phen-Cu(II) in the presence of 4 equivs of picolinic acid (2), and ALBP-
Phen-Cu(II) in the presence of excess EDTA (3). (B) Fluorescence at
Figure 11. Fluorescence of ALBP-Phen-Cu(II) in the presence of
alanine. (A) Emission spectra of ALBP-Phen-Cu(II) (1), ALBP-Phen-
Cu(II) in the presence of 4 equivs of alanine (2), and ALBP-Phen-Cu-
(II) in the presence of excess EDTA (3). (B) Fluorescence at 336 nm
(b) and 406 nm (O) of ALBP-Phen-Cu(II) in the presence of alanine.
The fluorescence at 336 nm (9) and 406 nm (0) of ALBP-Phen-Cu-
(II) upon the addition of excess EDTA.
3
36 nm (b) and 406 nm (O) of ALBP-Phen-Cu(II) in the presence of
picolinate. The fluorescence at 336 nm (9) and 406 nm (0) of ALBP-
Phen-Cu(II) upon the addition of excess EDTA.
bound to Cu(II). Nevertheless, it is clear that Cu(II) remains
bound to the phenanthroline in the presence of excess picolinate.
Similar results were obtained upon titration of ALBP-Phen-Cu-
(II) with alanine. As shown in Figure 11A,B, a small diminution
in the 336 nm fluorescence is observed upon addition of alanine,
but no increase in 406 nm fluorescence occurs. Thus, in contrast
to phenanthroline-Cu(II), ALBP-Phen-Cu(II) does not appear
to release Cu(II) in the presence of picolinate or alanine.
Activated Ester Hydrolysis. With a well characterized
conjugate in hand, the ability of ALBP-Phen-Cu(II) to hydrolyze
an activated ester substrate was examined; the bidentate ligand/
substrate picolinic acid nitrophenyl ester (PNPE) was employed
in this study since the production of nitrophenol (5) could be
followed spectrophotometrically. Hydrolysis reactions were
Figure 12. Hydrolysis of PNPE (3) by ALBP-Phen-Cu(II) and Cu-
(II). Concentration of p-nitrophenol (5) produced by 10 µM ALBP-
Phen-Cu(II) (O), 10 µM Cu(II) (b) or buffer alone (4).
acid substrates were studied next. These compounds also
allowed the enantioselectivity of hydrolysis catalyzed by ALBP-
Phen-Cu(II) to be assessed; data for these experiments are
performed with excess 3 in PIPES buffer, pH 6.1, at 25 °C.
Incubation of ALBP-Phen-Cu(II) with 3 resulted in the hy-
drolysis of of at least 2 equivs of 3 in 2 h. As shown in Figure
12, the rate of hydrolysis for the ALBP-Phen-Cu(II)-containing
reaction is 5-fold greater than the background hydrolysis rate
and is linear over 6 h (6 turnovers, data not shown). In contrast,
Figure 12 also shows that reactions containing free Cu(II) give
biphasic progress curves.1 In the early part of the reaction, 1
equiv of 3 is hydrolyzed rapidly. This is followed by a decrease
in the hydrolysis rate to that of the background rate. The rapid
phase of the reaction is likely due to the hydrolysis of 3 bound
to free Cu(II), while the slow phase is attribuable to hydrolysis
of 3 promoted by a picolinate-Cu(II) complex or simply
background hydrolysis. Thus, these experiments demonstrate
that ALBP-Phen-Cu(II), but not free Cu(II), can promote the
catalytic hydrolysis of PNPE with a modest (5-fold) increase
in rate relative to background. However, beacause of the high
rates for the background reaction, this substrate was not studied
further.
8b
tabulated in Table 2. Hydrolysis reactions were performed with
racemic amino acid esters in PIPES buffer, pH 6.1, at 25 °C
and monitored by derivatizing the amino acid products with
o-phthaldialdehyde and N-acetyl-L-cysteine. The resulting di-
astereomers were separated by reversed phase HPLC and
quantified by fluorescence detection. A chromatogram show-
ing the separation of the two diastereomeric alanine derivatives
is shown in Figure 13A. Reaction of ALBP-Phen-Cu(II) with
23
(
23) (a) Nimura, N.; Kinoshita, T. J. Chromatogr. 1986, 352, 169-177.
b) Buck, R. H.; Krummen, K. J. Chromatogr. 1987, 387, 255-265. (c)
Koh, J. T. L. D.; Breslow, R. J. Am. Chem. Soc. 1994, 116, 11234-11240.
Stereoselective Ester Hydrolysis. To obtain a lower rate
for background hydrolysis, unactivated alkyl esters of amino
(

11648 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
DaVies and Distefano
Table 2. Kinetic Data for the Hydrolysis of Amino Acid Esters Catalyzed by ALBP-Phen-Cu(II)
temp
(°C)
turnovers
(n)
k
cat(D)
k
cat(L)
k
cat(T)
k
uncat
k
k
cat(T)
/
(M 1‚s-1
-
)
(M-1‚s-1)
(M-1‚s-1)
(M-1‚s-1)
substrate
ee (%)
uncat
2
2
1
3
3
1
Ala-OMe, 3a
Ala-OEt, 3b
Ala-OiPr, 3c
Tyr-OMe, 5a
Tyr-OMe, 5a
Tyr-OMe, 5a
Tyr-OEt, 5b
Ser-OMe, 7
PMNA, 9
25
25
25
25
15
4
25
25
37
5.5
3.0
1.3
7.6
1.3
0.70
0.70
2.3
0.7
67 ( 6.0
40 ( 5.0
86 ( 9.0
39 ( 4.0
79 ( 2.0
86 ( 8.0
31 ( 11
43 ( 1.3
8.5 ((2.9) × 10
8.3 ((2.1) × 10
8.8 ((5.3) × 10
4.2 ((1.6) × 10
5.1 × 10
7.5 ((1.5) × 10
68
3
3
3
3
2
2
2
3
1
1
1
1
1
1
1
1.9 ((0.43) × 10
1.2 ((0.38) × 10
4.9 ((0.48) × 10
1.1 ((0.03) × 10
6.0 ((0.03) × 10
4.3 ((0.71) × 10
6.2 ((0.19) × 10
2.8 × 10
2.7 ((0.23) × 10
0.93 ((2.3) × 10
2.5 ((0.31) × 10
1.9 ((0.31) × 10
1.5 ((0.08) × 10
2.1 ((0.11) × 10
4.4 ((0.34) × 10
110
130
280
61
44
32
48
1.6 × 10
3
1.2 × 10
3
2
3
2.1 ((0.34) × 10
1.2 ((0.18) × 10
7.0 × 10
3
1.2 × 10
1
2
4.4 ((2.7) × 10
6.4 × 10
2
2
2.3 ((0.71) × 10
1.5 ((0.02) × 10
6.6 × 10
3
3
2.1 × 10
2
-2
4
3.6 × 10
2.3 × 10
KD
ALBP-Phen-Cu(II) + ester {\ } ALBP-Phen-Cu(II)‚ester
(1)
kcat(T)
-
ALBP-Phen-Cu(II)‚ester + OH
8
ALBP-Phen-Cu(II) + acid (2)
In this mechanism, an ester substrate binds to ALBP-Phen-Cu-
-
(
II) to form a complex that then reacts with OH to form the
acid product. Using a rapid equilibrium approach, the following
equation for the rate can be written
d[acid]cat kcat(T)[ALBP-Phen-Cu(II)][ester][OH-]
)
(3)
Figure 13. Hydrolysis of a racemic mixture of alanine isopropyl ester
is selective for the L enantiomer. Panel A: Reversed-phase HPLC
chromatogram showing the separation of diastereomeric derivatives
prepared from a racemic mixture of alanine. Panel B: Rates of
hydrolysis of the two enantiomers of alanine isopropyl ester catalyzed
by ALBP-Phen-Cu(II).
dt
K + [ester]
D
This expression can be simplified when [ester] . KD. Given
the concentration of ester present in these reactions (1 mM)
and the known dissociation constants for Cu(II) amino acid ester
complexes (ca. 100 µM), this is a reasonable assumption.24 Thus
(
3) can be reduced as follows:
alanine ethyl ester (6b) over a 24 h period gave a linear rate
for the production of alanine; after 24 h the conjugate had
catalyzed the hydrolysis of 3 equivs of the ester substrate.
Interestingly, modest stereoselectivity was observed in this
reaction. The L enantiomer was hydrolyzed 2.3 times faster
than the D isomer. Better selectivity was obtained with the
bulkier substrate alanine isopropyl ester (6c). In this case the
L enantiomer was hydrolyzed 13 times faster (see Figure 13B)
than the D isomer with 1.3 equivs of alanine produced in 24 h.
The corresponding L enantiomer of alanine methyl ester (6a)
was hydrolyzed 5 times faster than the D ester and produced
d[acid]cat
)
kcat(T)[ALBP-Phen-Cu(II)][OH-]
(4)
dt
The term kcat(T) can be viewed as the sum of the rate constants
for the hydrolysis of the two enantiomeric amino acid esters
kcat(T) ) kcat(D) + kcat(L)
(5)
The uncatalyzed background reaction is shown below
kuncat
5
.5 equivs of alanine in 24 h. Other amino acid esters can be
-
ester + OH
8 acid
(6)
hydrolyzed by ALBP-Phen as well. The L isomer of tyrosine
ethyl ester (8b) was hydrolyzed 1.9 times more rapidly than
the D enantiomer; in this case 0.7 equivs were hydrolyzed in
An equation for the background reaction rate can also be written
2
4 h. The L enantiomer of tyrosine methyl ester (8a) underwent
d[acid]
)
kuncat[ester][OH-]
(7)
hydrolysis 2.3 times faster than the D enantiomer giving 7.6
equivs of tyrosine (9) in 24 h. This selectivity increased at lower
temperature. At 15 °C, the L enantiomer of 8a hydrolyzed 8.6
times faster than the D isomer, and at 4 °C the selectivity for
the L enantiomer increased to 14-fold above that for the D
enantiomer. However, this increase in selectivity was ac-
companied by a decrease in overall conversion. At 15 °C, 1.3
equivs of 8a were hydrolyzed, and at 4 °C only 0.7 equivs were
hydrolyzed. Serine methyl ester (9a) was also hydrolyzed by
ALBP-Phen, but with the opposite stereoselectivity. In this case,
dt
Using eqs 4, 5, and 7, the second order rate constants for amino
acid ester hydrolysis in the presence and absence of ALBP-
Phen-Cu(II) were calculated and are listed in Table 2. Values
4
3
-1 -1
for kcat(T) range from 6.4 × 10 -7.0 × 10 M ‚s , while kuncat
-
1
-1
values vary from 9.3 to 75 M ‚s . Calculation of the ratio
kcat(T)/kuncat for the various amino acid ester hydrolysis reactions
gives values ranging from 32 to 280 and provides an indication
of the relative rate enhancement resulting from the ALBP-Phen-
Cu(II) catalyst.
2
.3 equivs of 9a were hydrolyzed in 24 h by the conjugate with
the D enantiomer reacting 2.4 times more rapidly than the L
isomer. Interestingly, all attempts to hydrolyze valine esters
were unsuccessful with no reaction being observed for either
the methyl, ethyl, or isopropyl esters.
The rates for ester hydrolysis reactions catalyzed by ALBP-
Phen-Cu(II) were next analyzed using the kinetic scheme shown
below
Amide Hydrolysis. The ability of ALBP-Phen-Cu(II) to
hydrolyze amide bonds was examined next using the substrate
picolinic acid nitromethyl anilide (PMNA, 12). This compound
produces N-methyl-4-nitroaniline (MNA, 13) upon hydrolysis
which can be monitored spectrophotometrically at 400 nm.
(24) See: Nakon, R.; Rechani, P. R.; Angelici, R. J. J. Am. Chem. Soc.
1974, 96, 2117-2120, and references therein.

A Semisynthetic Metalloenzyme
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11649
Discussion
The conjugate ALBP-Phen was prepared by alkylation of a
unique cysteine residue (Cys117) from ALBP with iodoaceto-
mido-1,10-phenanthroline. Thiol titration data confirms that
Cys117 has been modified. UV/vis spectroscopy, fluorescence
measurements, and electrospray mass spectrometry indicate that
a covalently bound phenanthroline moiety is present. Gel
filtration and guanidine denaturation experiments suggest that
the overall ALBP structure remains intact. Fluorescence and
UV/Vis measurements are consistent with Cu(II) bound to the
phenanthroline. Taken together, these data provide compelling
evidence for the preparation of the desired conjugate, ALBP-
Phen.
The conjugate ALBP-Phen-Cu(II) provides a metal binding
site within a well defined protein cavity. This sequestered Cu-
(II) ion catalyzes the hydrolysis of unactiVated esters and amides
at neutral pH. A possible mechanism for ALBP-Phen-Cu(II)
promoted ester hydrolysis is shown below. This mechanism is
presented only to give a picture of what may be occurring with
ALBP-Phen-Cu(II); alternative mechanisms can be envisioned.
For ester substrates, several amino acid esters including alanine,
Figure 14. Rate versus substrate concentration profile of ALBP-Phen-
Cu(II) promoted PMNA hydrolysis. Saturation kinetics is observed with
ALBP-Phen-Cu(II) (O), while the background rate of PMNA hydrolysis
(b) increases linearly with increasing concentrations of PMNA.
Figure 15. Double reciprocal analysis of ALBP-Phen-Cu(II) promoted
PMNA (12) hydrolysis. Plot of the reciprocal of the initial rate of NMA
serine, and tyrosine can be hydrolyzed. However, the â-branched
amino acid valine ethyl ester is not a substrate. In all cases the
(13) formation (1/v) versus the reciprocal of [PMNA] (1/[S]) catalyzed
by ALBP-Phen-Cu(II).
hydrolytic reactions are stereoselective and produce the
L
isomers (except for serine) with modest selectivity. The
selectivities obtained depend on the amino acid and type of
alkoxy group present in the substrate; the enantioselectivity for
alanine isopropyl ester hydrolysis is much greater than that for
the corresponding ethyl ester (40% versus 86% ee). Efficient
hydrolysis of tyrosine methyl ester occurs at temperatures as
low as 4 °C which results in a concomitant increase in selectivity
(
39% ee at 25 °C, 86% ee at 4 °C). The second order rate
ALBP-Phen-Cu(II) was incubated with PMNA in PIPES buffer,
pH 6.1, at 37 °C, and the rate of N-methyl-4-nitroaniline
production was determined. These experiments demonstrated
that ALBP-Phen-Cu(II) was capable of hydrolyzing an aryl
amide bond. It should be noted that only low levels of
background hydrolysis were observed using ALBP instead of
ALBP-Phen-Cu(II) as a catalyst. Analysis of initial rate data
at various PMNA concentrations, illustrated in Figure 14, shows
that ALBP-Phen-Cu(II) catalyzed PMNA hydrolysis exhibits
saturation kinetics, while the background rate of PMNA
hydrolysis (which is quite low) increases linearly with increasing
concentrations of PMNA up to 200 µM; thus, this saturation
phenomenon is not due to a solubility effect. The ALBP-Phen-
Cu(II) catalyzed PMNA hydrolysis is consistant with the rate
law given below
2
constants for ester hydrolysis (kcat(T)) vary from 6.4 × 10 -7.0
3
-1 -1
×
10 M ‚s (Table 2) and are comparable to those observed
by Angelici and co-workers for Cu(II) promoted glycine methyl
ester hydrolysis. Specifically, the values for kcat(T) reported here
are less than those obtained by Nakon et al. with uncomplexed
4
-1 -1
Cu(II) (kcat(T) ) 7.6 × 10 M ‚s ) but higher than those
2
obtained with Cu(II)‚(nitrilotriacetic acid) (kcat(T) ) 4.6 × 10
-
1
-1 19e
M ‚s ).
The attenuated hydrolysis rate constants obtained
in the presence of ligands such as ALBP-Phen probably reflect
the decreased Lewis acidity of the phenanthroline-coordinated
Cu(II) relative to that of free Cu(II).
PMNA, an amide containing substrate, is also hydrolyzed by
ALBP-Phen-Cu(II) although this reaction requires higher tem-
peratures (37 °C) and longer incubation times (several days).
The initial rates of hydrolysis exhibit saturation kinetics
consistent with enzymatic behavior. Analysis of these rate data
-
d[MNA]cat
k [ALBP-Phen-Cu(II)][PMNA][OH ]
cat
-1 -1
gave values for KD (70 µM) and kcat (360 M ‚s ). The value
)
(8)
dt
K + [PMNA]
D
of kcat obtained with ALBP-Phen-Cu(II) is 67-fold lower than
the corresponding value obtained with Cu(II)‚(2,2′-bipyridine)
by Sayre and co-workers at 40 °C in EtOH/H2O although there
A double reciprocal plot of the rate data showing this “enzyme-
like” behavior is given in Figure 15. Further analysis of these
4
19a
is still considerable rate acceleration (kcat/kuncat ) 1.6 × 10 ).
-
1
-1
data gave values of 70 µM for KD and 360 M ‚s for kcat.
This decrease in rate may reflect a nonoptimal geometry for
PMNA binding within the ALBP cavity due to steric interac-
tions. It is also interesting to compare the KD value of 70 µM
-
2
-1 -1
Using a published value for kuncat of 2.3 × 10 M ‚s , a
4
19b
value of 1.6 × 10 was calculated for the ratio kcat/kuncat.

11650 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
DaVies and Distefano
obtained with ALBP-Phen-Cu(II) to the value for KD of “greater
than 10 mM” estimated by Reddy et al. for Cu(II)‚(2,2′-
bipyridine) catalyzed PMNA hydrolysis.1 Clearly, the hy-
drophobic cavity of the ALBP protein is enhancing the binding
of PMNA to the bound Cu(II) although not to the same extent
that it does for fatty acids since the KD values for these
physiologically relavent ligands are below 1 µM.
This is not surprising given the fact that the protein does not
contribute additional ligands or functional groups that could
participate in general acid/base chemistry; the cavity is primarily
a hydrophobic, shape-selective pocket that can only control the
reaction specificity. Efforts to introduce additional side chains
via mutagenesis to modulate the reactivity and specificity are
currently underway.
9b
An important issue related to the chemistry reported here
concerns the question of whether or not our results can be
attributed to the presence of small amounts of free Cu(II) in
our reactions. We believe that this is not the case for three
reasons. First, fluorescence experiments with ALBP-Phen-Cu-
The selectivity and catalytic properties of ALBP-Phen clearly
demonstrate that covalent modification of the ALBP cavity can
be used to generate a metal ion-containing hydrolytic catalyst
with enzyme-like features. Although the efficiency and selec-
tivity exhibited by this construct are significantly less than those
of naturally occurring enzymes, the ALBP cavity portends to
be a useful system for the development of new catalysts.
Several other conjugates have already been prepared which
perform reactions with selectivites as high as 95% ee.12 Mutants
have been prepared that alter the catalytic efficiency and
selectivity of transformations performed in the cavity.25 Finally,
crystals of several ALBP conjugates, including ALBP-Phen,
suitible for X-ray analysis have been prepared and are being
studied to provide a structural framework for understanding the
selectivities of reactions that have been carried out within this
(II) showed that the addition of picolinate or alanine did not
lead to measurable decomplexation of Cu(II) even in the
presence of excess ligand after 24 h of incubation. However,
it is interesting to note that, in similar experiments, picolinate
and alanine (to a small extent) were able to remove Cu(II) from
free phenanthroline. Thus, the cupric ion present in ALBP-
Phen-Cu(II) is clearly in a sequestered environment more
characteristic of protein-bound metal ions. Second, the hy-
drolysis of amino acid esters reported here occured in an
enantioselective manner. This would not be possible if free
Cu(II) was the actual catalyst. For the achiral substrates (3 and
2
6
constrained protein cavity.
1
2), a third line of reasoning can be considered. Initial
experiments with picolinic acid nitrophenyl ester (3) showed
that while free Cu(II) promoted the hydrolysis of 1 equiv of
substrate more rapidly than ALBP-Phen-Cu(II), Cu(II) did not
effect catalytic hydrolysis; the rate of hydrolysis of 3 after one
turnover was close to or equal to that of the background reaction.
This observation is consistent with results reported by other
investigators and can be attributed to the lower reactivity of
the picolinate-Cu(II) complex whose Lewis acidity has been
attenuated by the negatively charged picolinate ligand. Thus,
the catalytic hydrolysis of 3 cannot be accounted for by the
presence of free Cu(II). Similar reasoning can be used to rule
out the involvement of free Cu(II) in the amide hydrolysis
reactions. Although catalytic hydrolysis of 12 was not observed
with ALBP-Phen-Cu(II), it was possible to detect the hydrolysis
of 0.7 equivs of substrate. Since decomplexation of 70% of
ALBP-Phen-Cu(II) would have been detected in our fluores-
cence experiments, this hydrolysis could not result from a large
amount of free Cu(II). However, if this reaction involved a
trace of free Cu(II) acting in a catalytic fashion, the active
species would have to be a picolinate-Cu(II) complex. As noted
above, such a complex is actually less reactive in ester hydrolysis
than ALBP-Phen-Cu(II).
Experimental Section
General Instrumentation and Materials. Fluorescence measure-
ments were performed with a Perkin Elmer Luminescence Spectrometer
LS 50B, and UV data were obtained using a Hewlett Packard Diode
27
Array Spectrophotometer 8452A. Iodoacetamido-1,10-phenanthroline
(
1), picolinic acid p-nitrophenyl ester (3), and picolinic acid methyl-
19a
nitroanilide (12) were synthesized according to published procedures.
Amino acid esters were purchased from Aldrich or were prepared from
the corresponding amino acids by standard methods. E. coli JM105/
pMON4 was a generous gift from Dr. D. Bernlohr, Department of
Biochemistry, University of Minnesota, Minneapolis, MN, 55455.
Purification of ALBP. Cultures of E. coli JM105/pMON4 were
grown in LB media (4 × 1L) to an OD600 of 0.8, induced with nalidixic
acid, and grown for an additional 4 h. The cells were then harvested
by centrifugation, frozen in N (l), and stored at -80 °C until
2
purification. The frozen cell paste (21 g) was resuspended in 50 mL
of buffer A (25 mM imidazole-HCl, pH 7.0, 50 mM NaCl, 5 mM
EDTA, 0.1 mM PMSF, and 1 mM 2-mercaptoethanol). The cells were
then lysed by sonication (0 °C, 5 min), and the cellular debris was
removed by centrifugation. Protamine sulfate (5 mL, 5% in Buffer A)
was added dropwise to the supernatant at 4 °C, and the cloudy mixture
was stirred for another 45 min. The precipitate was removed by
centrifugation. The supernatant was titrated to pH 5.0 with sodium
acetate (2.0 M, pH 4.9), and the mixture was allowed to stir overnight
at 4 °C. The supernatant was clarified by centrifugation and was
concentrated to 7 mL by ultrafiltration (YM3 filters, Amicon). The
concentrated protein solution was fractionated by chromatography
employing a Sephadex G-75 Superfine column (5 cm × 100 cm)
equilibrated in buffer B (12.5 mM HEPES, pH 7.2, and 0.25 M NaCl)
and eluted at 0.5 mL/min. The ALBP eluted after approximately 39
h. The fractions containing the protein were identified by SDS PAGE
and concentrated by ultrafiltration. Approximately 75 mg of protein
As a final note, it should be pointed out that the hydrolysis
rates described here are reported relative to the rates of
background hydrolysis in the absence of Cu(II). This was done
in lieu of reporting the rates relative to free Cu(II) or phenan-
throline-Cu(II) since the former is not present under catalytic
conditions (after one turnover it gains a ligand), whereas the
latter undergoes ligand exchange with the reaction products to
varying extents. Moreover, since one of the goals of our work
in this area is to develop selective protein-based catalysts, the
ratio of the rate of the protein-catalyzed reaction versus the
background rate is of particular importance. The enantioselec-
tivity that can be obtained in a kinetic resolution reaction is, in
part, controlled by this ratio. However, it is still useful to note
that the rate enhancements for ester and amide hydrolysis
2
8
was obtained as determined by the Bradford protein assay and UV
-
1
-1 29
(ꢀ278 ) 15 500 M cm ). The purified ALBP was shown to be
30
active by its ability to bind fluorescent fatty acid analogs.
(25) Kuang, H.; Davies, R. R.; Distefano, M. D. Bioorg. Med. Chem.
Lett. 1997, 7, 2055-2060.
(26) Ory, J.; Mazhary, A.; Kuang, H.; Davies, R. R.; Distefano, M. D.;
Banaszak, L. J. Submitted.
27) Chen, C. B.; Sigman, D. S. Proc. Natl. Acad. Sci. U.S.A. 1986, 83,
2
3
reported here are 10 -10 lower than those observed for free
(
Cu(II) (under single turnover conditions) but comparable to
7147-7150.
1
9e
(28) Bradford, M. M. Anal. Biochem. 1976, 72, 248-254.
those previously reported for Cu(II) complexes.
Thus,
(
(
29) Buelt, M. K.; Bernlohr, D. A. Biochemistry 1990, 29, 7408-7413.
introduction of the phenanthroline-Cu(II) moiety into ALBP
does not substantially alter the reactivity of the metal complex.
30) Sha, R. S.; Kane, C. D.; Xu, Z.; Banaszak, L. J.; Bernlohr, D. A. J.
Biol. Chem. 1993, 268, 7885-7892.

A Semisynthetic Metalloenzyme
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11651
Synthesis of ALBP-Phen. A solution of iodoacetamido-1,10-
phenanthroline (100 mM) in DMF (128 µL) was added to a solution
of purified ALBP (53 µM, 21.5 mL) in HEPES buffer (12.5 mM, pH
parison between ALBP and ALBP-Phen, the data were scaled from 0
to 1 using the relationship Frel ) (F - Fmin)/(Fmax - Fmin) where Frel
is the relative fluorescence, and Fmax and Fmin are the maximum and
minimum fluorescence values observed for the particular protein,
respectively.
Amino Acid Ester Hydrolysis Reactions. Hydrolysis reactions
contained racemic amino acid ester 6a, 6b, 6c, 8a, 8b, or 10 (1 mM)
and ALBP-Phen-Cu(II) (25 µM) in PIPES buffer (10 mM, pH 6.1) in
a total reaction volume of 52 µL and were performed in microcentrifuge
tubes at room temperature or in a thermostated water bath. At given
time points, 5.0 µL aliquots of the reaction mixture were removed and
7
.2), and the mixture was allowed to react at room temperature, in the
dark, for approximately 48 h. The reaction was monitored by titration
with 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) in 5.2 M guanidine
hydrochloride as described by Riddles et al. to determine the extent of
3
1
alkylation of Cys117
.
Conjugation reactions were typically 85% to
95% complete based on this assay. Following conjugation, the solution
was concentrated 5-fold by ultrafiltration and purified by gel filtration
chromatography (Biogel P6-DG, Biorad) in Tris‚HCl (50 mM, pH 7.4)
to remove the excess phenanthroline reagent. The fractions containing
the purified ALBP-Phen were combined and concentrated, and the
concentration of the final product was determined by protein assay.
quenched with 17 µL of H O and 3.0 µL of 100 mM EDTA and frozen
2
at -20 °C until analysis.
Derivatization. Immediately before analysis, reaction samples were
thawed, and 8.0 µL of the resulting material was derivatized with 8.0
µL of a reagent consisting of N-acetyl-L-cysteine dissolved in “Incom-
plete o-Phthaldialdehyde Reagent Solution” (Sigma no. P7914) at a
concentration of 5 mg/mL; these reagents convert enantiomeric amines
into diastereomeric isoindoles. The derivatization reaction was allowed
to proceed for 5 min at which time the sample was immediately
analyzed by HPLC.
Metallation of ALBP-Phen. CuSO
0 mM, pH 7.4) was added to a solution of ALBP-Phen (1.5 mL, 150
4
(1.5 mL, 10 mM, in Tris‚HCl,
5
µM) in the same buffer and allowed to react for 1 h at ambient
temperature. Copper binding was confirmed by fluorescence spec-
troscopy (as described below), and the unbound copper was removed
by gel filtration chromatography (Biogel P6-DG, Biorad) in PIPES
buffer (10 mM, pH 6.1).
UV/vis and Fluorescence Spectroscopy. UV/vis spectra of ALBP
and ALBP-Phen were obtained using 6.5 µM protein determined by a
Bradford protein assay in 25 mM Tris, pH 7.4. UV data for 1,10-
phenanthroline was obtained in the same buffer. Fluorescence spectra
for ALBP and ALBP-Phen were obtained in 50 mM Tris, pH 7.4. To
follow ALBP-Phen metalation, the fluorescence of a solution of ALBP-
Phen (1.0 mL, 1.0 µM) in Tris‚HCl (50 mM, pH 7.4) was obtained,
HPLC Analysis. All reaction mixtures were separated by reversed
phase HPLC using a Rainin Microsorb-MV C18 column (5 µm, 4.6 ×
250 mm). Chromatography was performed by injecting a 10 µL sample
and eluting with a flow rate of 0.75 mL/min using a gradient of solvent
A (80 mM sodium citrate and 20 mM sodium phosphate, pH 6.8,
containing 10% MeOH) and solvent B (MeOH). Detection of the
isoindole derivatives was achieved using a Beckman 6300A fluores-
cence detector (356 nm excitation filter and 450 nm emission filter),
and integration was performed with a HP 3393A integrator or System
Gold software. Response factors for the diastereomeric derivatives were
obtained from calibration curves generated by derivatization and
analysis of standard solutions of racemic amino acids. For alanine (7),
an elution profile consisting of isocratic elution with 80% solvent A
for 5 min followed by a 10 min linear gradient from 20% to 40% solvent
B and further elution with 60% solvent A was employed. Using this
system, the retention times for the D-alanine and L-alanine derivatives
were 22.5 and 23.8 min, respectively. For tyrosine (9), an elution profile
consisting of isocratic elution with 70% solvent A for 5 min followed
by a 5 min linear gradient from 30% to 35% solvent B and further
isocratic elution with 65% solvent A giving retention times for the
L-tyrosine and D-tyrosine derivatives of 21.2 and 25.1 min, respectively.
Serine (11) was separated using isocratic conditions (10% solvent B)
with a flow rate of 0.65 mL/min. Under these conditions, the D-serine
and L-serine derivatives eluted with retention times of 15.8 and 17.0
min, respectively.
and CuSO
4
(1.0 mM in Tris buffer) was titrated into the solution (40
µL total) until there was no further change in the phenanthroline
fluorescence (λex ) 275 nm, λem ) 406 nm). The excess copper was
then removed by gel filtration (NAP 10 column, Pharmacia), and the
fluorescence spectrum was again obtained. A similar procedure was
performed with 1,10-phenanthroline; however, only 2 equivs of CuSO
4
were needed to completely quench the fluorescence. Ligand competi-
tion experiments between phenanthroline-Cu(II) and picolinate or
alanine were performed using phenanthroline (1.0 µM), Cu(OAc) (1.0
2
µM), and 1.0-4.0 µM picolinate or alanine in PIPES buffer (10 mM,
pH 6.1) at 25 °C. Experiments with the protein conjugate contained
5
.0 µM ALBP-Phen-Cu(II) and 5.0-20 µM picolinate or alanine in
PIPES buffer (10 mM, pH 6.1) at 25 °C. Fluorescence emission spectra
from 300-450 nm were obtained by excitation at 275 nm (5.0 nm
slits for both monochrometers, 100 nm/min scan speed). After titration,
samples were incubated at 25 °C in the dark, spectra were recorded,
and EDTA was added to a final concentration of 500 µM. A final
spectrum was recorded after EDTA addition.
Electrospray Mass Spectrometry. Samples for mass spectrometry
were obtained by extensive dialysis of solutions of ALBP and ALBP-
Phen in 200 mM HEPES (1 mL, 1.1 mg/mL) against H
2
O/MeOH/HOAc
PNPE (3) Hydrolysis. The rate of hydrolysis of picolinic acid
p-nitrophenyl ester (3) was determined by measuring the increase in
absorbance at 316 nm due to the formation of p-nitrophenol (5). The
(
89/10/1, v/v/v), (2 × 1 L). Spectra were obtained using a PE SCIEX
reactions contained ALBP-Phen-Cu(II) or CuSO (10 µM) and PNPE
4
(100 µM) in PIPES buffer (10 mM, pH 6.1) and were performed in
quartz cuvettes at 25 °C. Absorbance values were converted to
concentrations using a a standard curve of p-nitrophenol (5) obtained
in PIPES.
API III MS/MS instrument, and the masses of ALBP and ALBP-Phen
were determined by analysis of the intensity versus m/z data using
HyperMass software.
Gel Filtration Chromatography. Gel filtration chromatography
was performed using an FPLC system (Pharmacia) and a Superose 12
HR 10/30 column. Samples (100 µL) were eluted with 25 mM HEPES,
pH 7.5, and 150 mM NaCl at a flow rate of 0.5 mL/min at 4 °C. The
column was calibrated using bovine lung aprotinin (6500 g/mol), horse
heart cytochrome c (12 400 g/mol), bovine erythrocyte carbonic
anhydrase (29 000 g/mol), and bovine serum albumin (66 000 g/mol)
as standards.
Protein Denaturation Experiments. The denaturation of ALBP
and ALBP-Phen was followed spectrofluormetrically by monitoring
the emission maximum upon excitation at 280 nm (5.5 nm excitation
slitwidth, 7.5 nm emission slitwidth) in a cuvette thermostated to 25
PMNA (12) Hydrolysis. The rate of hydrolysis of picolinic acid
nitromethyl anilide 12 (PMNA) was determined by measuring the
increase in absorbance at 400 nm due to the formation of N-methyl-
4-nitroaniline (13). The reactions contained ALBP-Phen-Cu(II) (4.8
µM) and PMNA (20 µM-2 mM) in PIPES buffer (10 mM, pH 6.1)
and were performed in microcentrifuge tubes in a 37 °C water bath.
At various times the solutions were removed from the water bath and
equilibrated to room temperature and the absorbances at 400 nm were
measured. The rate of background hydrolysis of PMNA was deter-
mined from similar experiments performed in the absence of ALBP-
Phen. Additional control experiments used ALBP or 4.8 µM CuSO
4
.
°
C. Samples contained ALBP or ALBP-Phen (1.0 µM), Tris‚HCl (50
Absorbance values for N-methyl-4-nitroaniline in PIPES were converted
-
1
-1
mM, pH 7.4), EDTA (0.1 mM), and guanidinium chloride (0-5.0 M).
Data were fitted to a two state model, previously used for lipid binding
proteins, described by Frieden and co-workers. To facilitate com-
to concentrations using a value of ꢀ400 ) 17 600 M ‚cm calculated
from a standard curve obtained in PIPES. All reactions were performed
in duplicate and the data averaged.
3
2
(
31) Riddles, P. W.; Blakeley, R. L.; Zerner, B. Methods Enzymol. 1983,
(32) Ropson, I. J.; Gordon, J. I.; Frieden, C. Biochemistry 1990, 29,
9591-9599.
9
1, 49-59.

11652 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
DaVies and Distefano
Acknowledgment. The authors thank M. Silberglitt and S.
ligand is sequestered within the protein cavity in a fashion
similar to the model shown in Figure 1.
2
6
Flaim for modeling ALBP-Phen, C. Lambert for performing
the denaturation experiments, L. Banaszak for providing the
X-ray coordinates of ALBP, and D. Bernlohr for the plasmid
pMON4. This work was supported by grants from the National
Science Foundation (NSF-CHE-9506793) and the Petroleum
Research Fund (PRF#28140-G4). R.R.D. was supported by a
National Institutes of Health training grant (NIH 2T32GM08347-
6).
Note Added in Proof. X-ray crystallographic analysis of
the structure of ALBP-Phen indicates that the phenanthroline
Supporting Information Available: Stereoviews of the
ALBP-Phen model constructed with Insight Software and
additional UV data for titrations of phenanthroline-Cu(II) with
picolinate and alanine (5 pages). See any current masthead for
ordering and Internet access instructions.
0
JA970820K