COMMUNICATION
Electrophilic aromatic substituted luciferins as bioluminescent probes
for glutathione S-transferase assays{
Wenhui Zhou,*a John W. Shultz,*b Nancy Murphy,b Erika M. Hawkins,b Laurent Bernad,a Troy Good,a
Leonard Moothart,a Susan Frackman,b Dieter H. Klaubert,a Robert F. Bulleitb and Keith V. Woodb
Received (in Cambridge, MA, USA) 25th July 2006, Accepted 31st August 2006
First published as an Advance Article on the web 26th September 2006
DOI: 10.1039/b610682j
significantly higher pKa values of the 6-hydroxy group of luciferin
and quinolinyl luciferin (8.5 and 9.4)8 than the pKa 7.219 of
o-nitrophenol (Scheme 1). We anticipated both the reactivity and
isozyme selectivity could be modulated by the effects of
substituents in the nitrophenyl ring or by converting the
nucleofugic group (ether bond) to a better leaving group.
New highly sensitive latent bioluminescent luciferin substrates
were designed and synthesized for monitoring mammalian
glutathione S-transferase (GST) and Schistosoma japonicum
enzyme activities.
The glutathione S-transferases (GSTs) represent a major group of
detoxification enzymes that catalyze the nucleophilic addition of
the tripeptide glutathione (GSH) to many xenobiotics and
endogenous electrophiles.1 The level of expression of GSTs is a
crucial factor in determining the susceptibility to cancer
chemotherapy. Three classes of GST isozymes, Alpha (A), Mu
(M) and Pi (P) are often found over-expressed in drug-resistant
tumor or tumor cell lines.1,2 Thus, there continues to be a need for
highly sensitive probes for monitoring mammalian GST isozyme
activity in tumor cells to examine anti-cancer drug resistance.
Currently, only a few methods for assaying GST activity are
available,3,4 1-chloro-2,4-dinitrobenzene (CDNB)3 being the most
prevalent method, but its use is often limited by low sensitivity
(absorbance), high background (non-enzymatic GSH conjugation)
and lack of selectivity for isozymes. The generic advantages of low
background and high sensitivity found in luciferase-coupled
bioluminescent assays5 led us to search for bioluminogenic GST
substrates as a possible approach to a highly sensitive and selective
mammalian GST assay.
Initial attempts at the synthesis of o-nitrophenyl luciferin ether
compounds via a two-step reaction sequence, that is, nucleophilic
substitution of 1-chloro-2-nitro-benzene derivatives with
6-hydroxy-2-cyanobenzothiazole, followed by condensation with
D-cysteine were unsuccessful due to the unexpected nucleophilic
reaction of sulfur in the benzothiazole ring with 1-chloro-2-nitro-
benzene. However, a similar method employed with 6-hydroxy-2-
cyanoquinoline resulted in the target o-nitroaromatic quinolinyl
luciferin ether compounds 2a, 2b and 2c (Scheme 1). The activities
of mammalian isozymes A1-1, M1-1 and P1-1 with compounds
2a–2c as substrates and GSH as a co-substrate were examined in a
two-step assay format. The resulting luminescence with these
compounds in the presence of GSH and GST enzyme above GSH
control signal indicated that quinolinyl luciferin was released by
GST-catalyzed nucleophilic substitution. The products from GST/
GSH reactions were analyzed by LC-MS, and only free luciferin
and GSH-nitrophenyl adduct were observed with no o-nitrophenol
or luciferin–GSH adduct detectable, confirming that the GSH only
attacked the ether carbon site of the o-nitrophenyl ring. GSH
control (non-enzymatic) for compounds 2a–2c yielded very low net
signals with net signal-to-background ratios ,0.1. Compounds 2b
and 2c exhibited reactivity with isozyme A1-1 and M1-1 with
signal-to-background ratios of 14 and 11 for A1-1, 46 and 170 for
M1-1, respectively, whereas none of the compounds showed
appreciable reactivity toward isozyme P1-1. Given that compound
2a showed lack of reactivity toward any GST isozyme and
We previously demonstrated that chemical modification of the
6-hydroxyl group of luciferin is an effective means to approach
bioluminescent assays for enzymes of interest.5 Relying on the
observation that various electrophilic nitrophenyl compounds are
detoxified by GSTs6 with a catalytic reactivity depending on the
electrophilicity of the substrate to the GSH anion, we designed a
series of electrophilic o-nitrophenyl luciferin or quinolinyl luciferin
ether derivatives. Relying on the hypothesis that the delocalization
of the negative charge on the phenoxide oxygen is correlated to the
pKa values of the phenols,7 it seemed reasonable to use the pKa as
a simple indicator of the electrophilicity at the phenol carbon site.
We therefore envisioned free luciferin would be liberated by GSH
attack on the ether carbon site of the o-nitrophenyl ring due to the
aPromega Biosciences Inc, 277 Granada Drive, San Luis Obispo,
CA 93401, USA. E-mail: wenhui.zhou@promega.com;
Fax: 001 8055431531; Tel: 001 8055448524
bPromega Corporation, 2800 Woods Hollow Road, Madison,
WI 53711-5399, USA. E-mail: john.shultz@promega.com;
Fax: 001 6082772601; Tel: 001 6082744330
{ Electronic supplementary information (ESI) available: Synthetic proce-
dures and characterizations for compounds 2a–c, 3a–3c, 4a–4c and 5;
protocols for GST enzyme assays and determination of kinetic parameters
for GST isozymes with compound 3c; LC-MS spectra for analyzing the
products from the GST reactions. See DOI: 10.1039/b610682j
Scheme 1 The release of luciferin from o-nitroaromatic luciferin ethers
by nucleophilic reaction with GSH catalyzed by GST and syntheses of
o-nitroaromatic quinolinyl luciferin ethers.
4620 | Chem. Commun., 2006, 4620–4622
This journal is ß The Royal Society of Chemistry 2006