N-Me-pAB-Glu-γ-Glu-γ-Tyr(3-NO2): An internally quenched fluorogenic γ-glutamyl hydrolase substrate

Article abstract of DOI:10.1016/S0960-894X(01)00103-2

A γ-glutamyl tripeptide containing an internally quenched fluorophore has been synthesized and shown to be a substrate for recombinant rat γ-glutamyl hydrolase. HPLC, LC-MS, and fluorescence spectra support the conclusion that selective hydrolysis occurs

Full text of DOI:10.1016/S0960-894X(01)00103-2

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1561–1564
N-Me-pAB-Glu-ꢀ-Glu-ꢀ-Tyr(3-NO2): An Internally Quenched
Fluorogenic ꢀ-Glutamyl Hydrolase Substrate
Jessica J. Pankuch and James K. Coward*
Departments of Chemistry and Medicinal Chemistry, University of Michigan, Ann Arbor, MI 48109-1055, USA
Received 18 December 2000; accepted 15January 2001
Abstract—A g-glutamyl tripeptide containing an internally quenched fluorophore has been synthesized and shown to be a substrate
for recombinant rat g-glutamyl hydrolase. HPLC, LC–MS, and fluorescence spectra support the conclusion that selective hydrolysis
occurs at the penultimate peptide bond. Preliminary data indicate that hydrolysis of this substrate can be monitored continuously
to yield steady-state kinetic data # 2001 Elsevier Science Ltd. All rights reserved.
g-Glutamyl hydrolase (GH, EC 3.4.19.9) catalyzes the
hydrolysis of folylpoly-g-glutamates, the intracellular
forms of folate cofactors involved in one-carbon meta-
bolism and thus provides one means of regulating cel-
lular levels of folates.1 Rat and human forms of GH
have been cloned and overexpressed. Site-directed
mutagenesis and calculations revealing high structural
homology with the glutaminase domain of carbamoyl
phosphate synthetase indicate that GH is a cysteine
protease, in contrast to the zinc metalloprotease, gluta-
mate carboxypeptidase II (GCPII, ‘conjugase,’ EC
3.4.17.21), which cleaves dietary folylpoly-g-gluta-
mates.2
3 are hydrolyzed more slowly than 1 by a factor of at
least 25(J. J. Pankuch and J. K. Coward, unpublished
results). These data were obtained using a non-
continuous end-point HPLC assay.6 A convenient con-
tinuous spectral assay was desired to obtain more
accurate kinetic measurements for both the fluoro-
glutamate-containing substrates and the physiological
glutamate-containing substrates. It has been shown that
g-glutamyl peptides lacking the pteridine heterocycle
present in folates and methotrexate are excellent GH
substrates (e.g., 4).6 We now report the synthesis of a
similar g-glutamyl peptide, 5, containing a fluorescence
donor at the N-terminus that is intramolecularly quen-
ched by 3-nitrotyrosine (Tyr(3-NO2)) at the C-termi-
nus.7 GH-catalyzed hydrolysis of this peptide results in
a time-dependent increase in fluorescence, thereby
allowing for the continuous monitoring of the reaction
by fluorescence spectroscopy. This is the first internally
quenched fluorogenic substrate reported for GH and
should enable detailed mechanistic studies as well as
more efficient inhibitor screening.
We have developed fluoroglutamate-containing g-gluta-
myl peptides as mechanistic probes for GH (Table 1). In
earlier work investigating the effect of electron-deficient
peptide bonds on the rate of GH-catalyzed hydrolysis,
we observed that a methotrexate (4-amino-4-deoxy-
10-methylpteroyl glutamate, AMPteGlu) derivative
AMPte(2RS,4RS)-4-fluoroglutamyl-g-glutamate
was
hydrolyzed at a surprisingly slow rate by the hog kidney
enzyme.3 Using the recombinant enzymes, we have
extended this observation with g-glutamyl peptides
containing either (2S,4S)-4-fluoroglutamate4 (2) or
(2RS)-4,4-difluoroglutamate (3).5 Preliminary studies
were conducted comparing the hydrolysis of AMPte-
Glu-g-Glu (1) versus the two analogous fluoroglutamate-
containing peptides. The kinetic data indicate that 2 and
Our approach to the synthesis of the desired peptide is
outlined in Scheme 1. This modular approach was cho-
sen to allow for the ultimate incorporation of a wide
variety of both natural and unnatural amino acids, e.g.,
fluoroglutamates, into this type of peptide. Thus, it was
envisioned that a differentially protected g-glutamyl
dipeptide could be coupled to the N-terminal fluoro-
phore, N-methyl-p-aminobenzoic acid (N-Me-pABA,
10) using one of many possible peptide coupling meth-
ods. Subsequent selective deprotection of the C-terminal
g-carboxyl group should allow for incorporation of a
*Corresponding author. Fax: +1-734-647-4865; e-mail: jkcoward@
umich.edu
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00103-2

1562
J. J. Pankuch, J. K. Coward / Bioorg. Med. Chem. Lett. 11 (2001) 1561–1564
suitably protected 3-nitrotyrosine (12) as the chromophore
designed to quench the fluorescence of the N-Me-pAB
moiety by intramolecular fluorescence resonance energy
transfer (FRET). The differentially protected dipeptide
utilizing tert-butyl (OtBu) and trimethyl silyl ethyl
(OTMSE) esters, Z-Glu(a-OtBu)-g-Glu(a-OtBu)(g-
OTMSE) (9), was synthesized from Z-Glu(a-OtBu)-g-
OH (8) and H-Glu(a-OtBu)(g-OTMSE), the latter
compounds derived from the protected oxazolidinone
6.8 Removal of the N-terminal Z group (H2/Pd–C) fol-
lowed by coupling with N-Me-pABA (EDC, HOBt,
DMAP) led to the fluorophore-containing peptide, 11.
Selective removal of the TMSE ester was effected with
TBAF, and then coupling (EDC, HOBt) of the free
g-COOH to H-Tyr(3-NO2)-OtBu (12) led to the fully
protected tripeptide, 13. Acid hydrolysis with TFA
removed all tert-butyl esters to provide the desired final
product, 5, after purification by HPLC.9 All inter-
mediates were completely characterized by NMR (1H,
13C, DEPT) including two-dimensional spectra (COSY,
C,H-HETCOR) for selected intermediates.10 In addi-
tion, the final fluorogenic peptide, 5, was further
characterized by LC–MS (see below).
Table 1. Reaction catalyzed by g-glutamyl hydrolase
Compds
R1
X
H
Y
H
R2
1
OH
In order to ascertain if peptide 5 is a GH substrate, a
three-hour incubation of 5 (100 mM) in the presence of
rat GH (rGH)11 was carried out under standard condi-
tions and the products were analyzed by HPLC.6,12
Peptide 5 (tR=16.3 min) was completely hydrolyzed
under these conditions and two UV-absorbing products
were detected. Spectroscopic analysis of each chroma-
tographic peak (diode array) showed that the early
eluting peak contained the N-Me-pAB chromophore
(lmax=287 nm, tR=5.2 min) while the late eluting peak
contained the Tyr(3-NO2) chromophore (lmax=278
nm, tR=7.2 min). Subsequent analysis of the product
mixture by LC–MS13 (Hitachi HPLC components,
Finnigan LCQ) clearly identified the products as N-Me-
pAB-Glu-OH (tR=9.05min) and H-Glu- g-Tyr(3-NO2)-
OH (tR=9.37 min) (Fig. 1A,C). Analysis of the same
2
H
F
F
F
OH
3a
OH
OH
4
5
H
H
H
H
H
H
aCompound 3 was synthesized as a mixture of racemic diastereomers.
Scheme 1.

J. J. Pankuch, J. K. Coward / Bioorg. Med. Chem. Lett. 11 (2001) 1561–1564
1563
samples by fluorescence spectroscopy showed a 7-fold
increase in fluorescence upon GH-catalyzed hydrolysis
of 5. In addition, HPLC with fluorescence detection
showed that the early peak, identified initially as N-Me-
pAB-Glu based on an identical absorption spectra (diode
array) as an authentic synthetic standard, contained the
expected fluorophore (lex=290 nm, lem=365nm) (Fig.
1B). These data provide convincing evidence that the
GH-catalyzed hydrolysis of 5 proceeds as shown in
Table 1.
Further studies of the hydrolysis of 5 by fluorescence
spectroscopy demonstrated that the cleavage reaction
can be monitored continuously. Incubation of 5 (2.5, 5,
10 mM) with rat GH (3 nM) resulted in a time-depen-
dent increase in fluorescence and an increase in max-
imum fluorescence with increasing concentration of 5
(Fig. 2). The successful synthesis of 5 and the demon-
stration that GH-catalyzed hydrolysis of 5 can be mon-
itored continuously in a spectral assay permits the first
detailed kinetic investigation of this key enzyme in
folate biochemistry. In addition, the spectral assay will
greatly facilitate the evaluation of mechanism-based
inhibitors of GH, the synthesis of which is currently
underway in our laboratory.
Acknowledgements
Figure 1. LC–MS analysis of cleavage products produced by extensive
incubation with rGH. 50 mM N-Me-pAB-Glu-g-Glu-g-Tyr(3-NO2) (5)
was incubated with 3 nM rGH in 100 mL of 50 mM sodium acetate, 50
mM b-mercaptoethanol, pH 6.0 at 37 ꢀC for 3 h. (A) HPLC elution
profile by UV (289 nm) where the peak at 7.38 is buffer, and the peaks
at 9.05and 9.37 min are N-Me-pAB-Glu and Glu-g-Tyr(3-NO2),
respectively. The arrow at 14.27 min indicates where N-Me-pAB-Glu-
g-Glu-g-Tyr(3-NO2) (5) elutes. (B) HPLC elution profile by fluores-
cence (lex=290, lem=365) showing that only the peak at 9.00 min is
fluorescent. (C) Mass spectrum of peptide fragment N-Me-pAB-Glu
(left) showing m/z 281.0 (M+H)+ and Glu-g-Tyr(3-NO2) (right)
showing (M+H)+ m/z 356.1. The minor (M+D)+ peaks are due to
the presence of deuterium resulting from exposure of 5 to D2O in prior
1H NMR experiments.
This research was supported by a grant from the
National Cancer Institute (CA28097). JJP is a trainee of
the Michigan Chemistry–Biology Interface Training
Program and was supported in part by a grant from the
National Institutes of General Medical Sciences
(GM08597). We thank Drs. John Galivan and Thomas
Ryan for generously providing the recombinant GH
used in this research and for many stimulating discus-
sions. We thank Matthew Alexander for helpful discus-
sions concerning the synthesis of 5, David Rosen for his
work on the fluorescence assay, and Prof. Kevin Rice
and Dr. Donald McKenzie for LC–MS analysis of the
products obtained from GH-catalyzed hydrolysis of 5.
References and Notes
1. Galivan, J.; Ryan, T. J.; Chave, K.; Rhee, M.; Yao, R.;
Yin, D. Pharmacol. Ther. 2000, 85, 207.
2. Chave, K. J.; Auger, I. E.; Galivan, J.; Ryan, T. J. J. Biol.
Chem. 2000, 275, 40365.
3. Licato, N. J.; Coward, J. K.; Nimec, Z.; Galivan, J.; Bola-
nowska, W. E.; McGuire, J. J. J. Med. Chem. 1990, 33, 1022.
4. Hart, B. P.; Licato, N. J.; Bolanowska, W. E.; McGuire,
J. J.; Coward, J. K. J. Med. Chem. 1996, 39, 5 6.
5. Tsukamoto, T.; Coward, J. K. J. Org. Chem. 1996, 61,
2497.
6. Wang, Y.; Nimec, Z.; Ryan, T. J.; Dias, J. A.; Galivan, J.
Biochim. Biophys. Acta 1993, 1164, 227.
7. Knight, C. G. Meth. Enzymol. 1995, 248, 18.
8. Itoh, M. Chem. Pharm. Bull. 1969, 17, 1679.
Figure 2. Progress curves for the rat GH-catalyzed hydrolysis of N-
Me-pAB-Glu-g-Glu-g-Tyr(3-NO2) (5) (A, 10 mM; B, 5 mM; C, 2.5
mM). GH in storage buffer (25mM sodium acetate (pH 6.0), 50 mM b-
mercaptoethanol, 1 mM b-octylglucoside, 1 mM EDTA, 0.1 M
sodium chloride) was preincubated at 37 ꢀC for 30 min. An aliquot (10
mL) of the enzyme solution was added to assay buffer (37 ꢀC) consist-
ing of 50 mM sodium acetate (pH 6.0) and 50 mM b-mercaptoethanol
and the indicated final concentration of 5 in a total volume of 100 mL.

1564
J. J. Pankuch, J. K. Coward / Bioorg. Med. Chem. Lett. 11 (2001) 1561–1564
9. HPLC: Vydac C18 column, gradient conditions were 3%
solvent A (acetonitrile) for 1 min (solvent B 0.1 M NH4OAc,
pH 5.5) followed by an increase to 25% solvent A over 25 min
at 1 mL/min.
NMR (CDCl3, 300 MHz) d 7.67–7.64 (d, 2H), 7.13–7.10 (d,
1H), 6.98–6.96 (d, 1H), 6.53–6.50 (d, 2H), 4.58 (m, 1H), 4.47–
4.45(m, 1H), 4.15–4.10 (m, 2H), 2.80 (s, 3H), 2.40–1.96 (m,
8H+EtOAc), 1.44–1.41 (d, 18H), 0.97–0.91 (t, 2H), 0.00 (s,
9H); 13C NMR (CDCl3, 75MHz) d 172.9, 172.3, 171.6, 171.1,
170.9, 167.3, 152.2, 128.8 (CH), 121.2 (CH), 111.2 (CH), 82.1,
62.7 (CH2), 52.6 (CH), 52.2 (CH), 32.5 (CH2), 30.4 (CH2), 30.1
(CH3), 28.6 (CH2), 27.9 (CH3), 27.4 (CH2), 17.2 (CH2), À1.6
(CH3). COSY and C,H-HETCOR spectra were also collected.
12: 1H NMR (CD3OD, 300 MHz) d 7.82 (s, 1H), 7.39–7.35(m,
1H), 6.99–6.96 (d, 1H), 3.55–3.50 (m, 1H), 2.85–2.83 (m, 2H),
1.27 (s, 9H); 13C NMR (CD3OD, 75MHz) d 174.4, 154.4,
139.1 (CH), 135.7, 129.8, 126.7 (CH), 121.4 (CH), 82.9, 56.8
(CH), 40.1 (CH2), 28.2 (CH3). 13: 1H NMR (CDCl3, 300
MHz) d 10.40 (s, 1H), 7.81–7.80 (d, 1H), 7.67–7.64 (d, 2H),
7.38–7.35(m, 2H), 7.00–6.97 (d, 1H), 6.89–6.86 (d, 1H), 6.69–
6.67 (d, 1H), 6.54–6.51 (d, 2H), 4.70–4.65 (m, 2H), 4.44 (m,
1H), 2.99–2.92 (m, 2H), 2.87 (s, 3H), 2.41–1.84 (m, 11H),
1.50–1.38 (m, 31H); 13C NMR (CDCl3, 75MHz) d 172.3,
171.8, 171.8, 170.9, 170.2, 167.4, 153.9, 152.2, 138.9 (CH),
133.0, 129.2, 128.9 (CH), 125.2 (CH), 121.1, 119.7 (CH), 111.3
(CH), 82.6, 82.3, 53.5 (CH), 52.2 (CH), 52.0 (CH), 36.6 (CH2),
32.4 (CH2), 32.0 (CH2), 30.1 (CH3), 29.6 (CH2), 29.3 (CH2),
29.2 (CH2), 28.0 (CH3), 27.9 (CH3), 0.4 (CH3). COSY and
C,H-HETCOR spectra were also obtained.
10. Selected 1H and 13C NMR spectroscopic data. 5: 1H
NMR (D2O, 500 MHz) d 7.84 (s, 1H), 7.62–7.58 (d, 2H), 7.44
(s, 1H), 6.99–6.96 (d, 1H), 6.68–6.67 (d, 2H), 4.39–4.34 (m,
2H), 3.95(m, 1H), 3.17–3.12 (m, 2H), 2.87–2.84 (d, 1H), 2.75
(s, 3H), 2.38–1.70 (m, integration obscured by NH4OAc peak,
1
d 1.97). 7: H NMR (CDCl3, 300 MHz) d 7.34–7.31 (m, 5H),
5.43–5.40 (d, 1H, NH), 5.09 (s, 2H), 4.29–4.24 (t, 1H), 4.18–
4.10 (t, 2H), 2.35–1.92 (m, 4H), 1.45 (s, 9H), 0.99–0.94 (t, 2H),
0.00 (s, 8H); 13C NMR (CDCl3, 75MHz) d 173.0, 171.1,
156.0, 136.4, 128.6 (CH), 128.2 (CH), 82.5, 67.0 (CH2), 63.0
(CH2), 54.0 (CH), 30.8 (CH2), 28.1 (CH3), 17.4 (CH2), À1.4
(CH3). Purification of 7 by silica gel chromatography led to a
product which was slightly contaminated by Z-Glu(OtBu)2,
resulting from incomplete TMSE protection of the g-CO2H of
6 and/or partial hydrolysis of the intermediate g-TMSE ester
1
prior to reaction with isobutylene. 8: H NMR (CDCl3, 300
MHz) d 10.71 (s, 1H), 7.37–7.28 (m, 5H), 5.43–5.41 (d, 1H),
5.11 (s, 2H), 4.34–4.28 (m, 1H), 2.48–2.21 (m, 2H), 2.21–2.17
(m, 1H), 1.98–1.91 (m, 1H), 1.48 (s, 9H); 13C NMR (CDCl3,
75MHz) d 178.2, 171.1, 156.2, 136.3, 128.7 (CH), 128.4 (CH),
128.3 (CH), 82.9, 67.3 (CH2), 53.8 (CH), 31.8 (CH2), 30.1
(CH2), 28.1 (CH3). 9: 1H NMR (CDCl3, 300 MHz) d 7.36–7.31
(m, 5H), 6.43 (d, 1H), 5.58 (d, 1H), 5.12 (s, 2H), 4.49–4.48 (d,
1H), 4.24 (d, 1H), 4.18–4.15(m, 2H), 2.35–2.28 (m, 6H), 1.95
(m, 2H) 1.46–1.44 (d, 18H), 0.99–0.96 (m, 2H), 0.00 (s, 9H);
13C NMR (CDCl3, 75MHz) d 173.1, 172.0, 171.1, 156.3,
136.3, 128.6 (CH), 128.2 (CH), 82.5, 67.1 (CH2), 63.0 (CH2),
54.0 (CH), 53.5, 52.3 (CH), 32.3 (CH2), 30.5(CH 2), 28.9
(CH2), 28.1 (CH3), 27.7 (CH2), 17.3 (CH2), À1.4 (CH3).
11. Yao, R.; Nimec, Z.; Ryan, T. J.; Galivan, J. J. Biol. Chem.
1996, 271, 8525.
12. HPLC: Vydac C18 column, gradient conditions were 3%
solvent A (acetonitrile) for 3 min (solvent B 0.1 M NaOAc, pH
5.5) followed by an increase to 4% solvent A over 5 min, fol-
lowed by an increase to 95% solvent A over 5 min at 1 mL/min.
13. HPLC: Vydac C18 column, gradient conditions were 3–
50% solvent B (solvent A: H2O, 0.1% AcOH, 0.02% TFA;
solvent B: CH3CN, 0.1% AcOH, 0.02% TFA).
1
COSY and C,H-HETCOR spectra were also collected. 11: H