An autoimmolative spacer allows first-time incorporation of a unique solid-state fluorophore into a detection probe for acyl hydrolases

Article abstract of DOI:10.1002/chem.200902412

A study was conducted to demonstrate the use of an autoimmolative spacer to incorporate a unique solid-state fluorophore into a detection probe for acyl hydrolases. The molecule effectively detected a model peptidase and showed elevated robustness that allowed for the adaptation to diffusion-resistant sequence-specific detection of peptidases for maximal spatial control. A cyclization linker was selected that immolated through an energetically favorable five- or six-membered transition state, taking into account the acylated O,O-acetal moiety. A GABA model compound was prepared to whether a GABA-based three-component system fragmented reliably.

Full text of DOI:10.1002/chem.200902412

DOI: 10.1002/chem.200902412
An Autoimmolative Spacer Allows First-Time Incorporation of a Unique
Solid-State Fluorophore into a Detection Probe for Acyl Hydrolases
Xiao-bing Zhang,[a, b] Michael Waibel,[a] and Jens Hasserodt*[a]
The challenges and the promises for the field of advanced
chemical imaging have recently been assessed by a group of
experts.[1] The issues concerning optical modalities, and es-
pecially fluorescence-based techniques, have been identified
in “the need for greater spatial control” and in “greater ro-
bustness of fluorescent probes”. Such fluorogenic imaging
probes would be equally attractive for applications in vitro
and in vivo. Current probes are mostly based on water-solu-
ble fluorophores; probes that have previously found exten-
sive use in in vitro assays of purified proteins without any
need for spatial distinction. Localization of specific molecu-
lar entities in biological tissue is mostly done with the aid of
antibodies conjugated to a fluorophore (immunohistochem-
istry). Detecting a particular biochemical property in an in
situ context in the life sciences is always associated with the
kinetics of two processes: 1) diffusion of the probe toward
(and away from) the site where the activity is present and
2) the kinetics of signal generation—be it complex forma-
tion or enzymatic modification. These effects may more or
less cancel each other out and no particular choice of incu-
bation or readout time for the assay will give meaningful
imaging results.
ties, such as intense luminescence, large Stokes shifts, and
significant photostability.[2–4] HPQ thus distinguishes itself
greatly from common fluorescent dyes, such as water-soluble
fluorescein, which exhibits negligible emission in the solid
state due to a small Stokes shift and consequent self-quench-
ing (Figure S1 in the Supporting Information). The large
Stokes shift of HPQ contributes two additional, very wel-
come advantages: an increase in sensitivity and minimiza-
tion of background fluorescence. Both its strong fluores-
cence and insolubility are intimately associated with the in-
ternal hydrogen bond between the phenolic hydrogen and
the imine nitrogen. Should one hinder this interaction by at-
taching a polar, enzyme-susceptible group to this oxygen
atom, then a water-soluble imaging probe would be formed
that has the long-wavelength fluorescence of HPQ efficient-
ly eliminated. It does not come as a surprise that probes for
particular enzymes (b-glucuronidase and alkaline phospha-
tase) have been reported and commercialized.[5–8] However,
no functioning HPQ-based probe has ever been introduced
for the assaying or imaging of the huge classes of peptidases
and lipases. Both classes comprise prominent biomarkers for
various diseases, play an important role in proteome profil-
ing, and are also subjected to extensive modification by di-
rected evolution, which requires enormously efficient tests,
preferably in the bacterial colony assay format, in view of
the large libraries to be screened. We report herein, a mole-
cule that effectively detects a model peptidase, a molecule
that shows elevated robustness, and that should allow for
the adaptation to diffusion-resistant, se-
We have chosen a fluorophore, 2-(2’-hydroxyphenyl)-
4(3H)-quinazolinone (HPQ), that is completely insoluble in
aqueous media, but strongly fluorescent in the solid state
(Figure S1 in the Supporting Information). Fluorescent dyes
that exhibit excited-state intramolecular proton transfer
(ESIPT) have attracted great interest for several decades
because such compounds show good photophysical proper-
quence-specific detection of peptidases
for maximal spatial control.[9]
Many enzyme-responsive probes have
been designed in which the fluorophore
[a] Prof. Dr. X.-b. Zhang, M. Waibel, Prof. Dr. J. Hasserodt
Laboratoire de Chimie, Universitꢀ de Lyon—ENS
46 Allꢀe dꢁItalie, 69364 Lyon (France)
Fax : (+33)472-72-8860
is a phenol or an aniline that is directly
attached to the scissile bond.[10,11] Impor-
E-mail: jens.hasserodt@ens-lyon.fr
[b] Prof. Dr. X.-b. Zhang
tantly, phenolic esters, in particular, are prone to spontane-
ous hydrolysis (background signal generation). Also, we
have recently shown that an aniline-type derivative of HPQ
does not show any appreciable fluorescence.[12] HPQ esters
have an even higher tendency for spontaneous hydrolysis
State Key Laboratory of Chemo/Biosensing and Chemometrics
College of Chemistry and Chemical Engineering
Hunan University, Changsha, 410082 (P.R. China)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200902412.
792
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 792 – 795

COMMUNICATION
than regular phenol esters, which we have observed our-
selves.[13] Previously, others have postulated that a self-cata-
lytic effect (anchimeric assistance by the imine) is responsi-
ble for this property while studying a different, albeit relat-
ed, molecular system.[14] We were thus forced to come up
with a substituent other than acyl on the phenolic oxygen of
HPQ. Acylated O,O-acetals have been used for a long time
in the design of prodrugs of phosphorylated species to
render them electroneutral for better bioavailability.[15,16]
More recently, they were employed to devise detection
probes incorporating soluble phenolic fluorophores.[17–19]
Here, they still showed a residual background hydrolysis,
except for a pivaloyl O,O-acetal group, which was complete-
ly immune to it.[20] We thus selected this solution in our
target design.
compound (S13, Scheme S1 in the Supporting Information)
with a carbobenzyloxy (Cbz) group on its amino terminus,
instead of a peptidase substrate moiety, and mimicked the
action of an enzyme by hydrogenative removal of this Cbz
group. When S13 was hydrogenated with 5% Pd/C in
MeOH under ambient pressure and temperature for 1 h,
TLC analysis indicated the release of HPQ to be complete.
By using acidic conditions (10% acetic acid in MeOH), the
presence of the nonimmolated intermediate S14 on a TLC
plate could be confirmed under a UV lamp (365 nm) be-
cause it can be easily distinguished from the strongly fluo-
rescent HPQ. However, even evaporation of the sample
without any heating did not allow us to isolate S14, which
simply fragmented too easily (Scheme S1 in the Supporting
Information).
Communicating the event of peptide-bond cleavage rapid-
ly to the part of the molecule that can liberate the fluoro-
phore requires employment of a self-immolative linker.
Such linkers (also known as traceless linkers in solid-phase
organic synthesis) have found wide-spread application in the
field of prodrug activation,[21] since the first such spacers
were introduced in the early 1980s.[22] In recent years there
has been increasing interest in the design of three-compo-
nent enzyme substrates (trigger–linker–fluorophore) by
using self-immolative spacers as linkers.[23] Such three-com-
ponent substrates display unique selectivity and limited in-
terference with environmental effects because these systems
unmask their intense fluorescence only by a specific enzy-
matic cleavage.[24–31] According to the mechanism of active-
compound release, the linkers can be classified as cyclization
or elimination linkers.[21] One of the first and most frequent-
ly employed self-immolative spacers is p-aminobenzyloxy-
carbonyl (PABC). Very recently, this linker has been em-
ployed in the design of two three-component substrates for
proteases containing soluble phenolic fluorophores (e.g., 7-
hydroxycoumarine) to circumvent the above-mentioned
problems of two-component systems.[31,32]
With these encouraging results we equipped our two re-
lated linkers with a leucine residue to test our probe candi-
dates in a real-world situation by using microsomal leucine
aminopeptidase (EC 3.4.11.2) as an efficient model enzyme
(Scheme 1). Thus, target 1 was synthesized by preparing 3
Taking into account our acylated O,O-acetal moiety, we
opted instead for a cyclization linker that immolates via an
energetically favorable five- or six-membered transition
state, thus setting a limit to the distance of the fluorogenic
group and the scissile carboxamide bond. We identified the
synthetically more accessible targets 1 and S13 (Scheme S1
in the Supporting Information) that contain a g-aminobuty-
ric acid (GABA) derived linker and target 2, which features
a 2,2-dimethyl GABA unit. Compound 2 can be predicted
to immolate more rapidly due to the Thorpe–Ingold
effect.[33]
Scheme 1. Synthetic strategy to access target compounds 1 and 2 by em-
ploying phase-transfer conditions. Instead of 9, a product of identical
À
mass was formed from a coupling with the N H site of HPQ; TFA=tri-
fluoroacetic acid, OSu=succinate ester.
from commercial starting materials by a simple peptide cou-
pling reaction. The free acid 4, thus obtained, had to be
transformed into the chloromethyl ester. The classic proce-
dure[34] of treatment of the free acid with paraformaldehyde
and ZnCl2 in dry toluene did not result in the formation of
any 6, even when the reaction was heated to 558C for 24 h,
as determined by TLC and LC–MS analysis. A more recent
strategy,[35] carried out by using the commercial reagent
chloromethyl chlorosulfate and phase-transfer conditions
gave 6 in excellent yield (88%) in only 1 h. Iodination prior
to coupling with a phenol has been reported to be superior
to direct coupling of the chloride. Several strategies were
To test whether a GABA-based three-component system
fragments reliably, we first synthesized a GABA model
Chem. Eur. J. 2010, 16, 792 – 795
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
793

J. Hasserodt et al.
described on how to couple iodomethyl esters with a
phenol: standard single-phase conditions with acetonitrile or
acetone as the solvent and K2CO3 as the base at room tem-
perature,[18] strong single-phase conditions with DMF/THF
as the solvent and NaH as the base at low temperature,[17]
and a biphasic system in the presence of tetrabutylammoni-
um hydrogen sulfate as the phase-transfer catalyst.[35] We
À
were worried about interference of the N H acidic site on
HPQ and opted for the mildest procedure, namely, that em-
ploying biphasic conditions. This gave 8 in low yield, the hy-
drogenation of which necessitated the use of 10% Pd/C, but
then gave 1 in quantitative yield in 2 h.
Scheme 2. Isomer-free access to substrate 2; TsOH=para-toluene sulfon-
ic acid.
Target 2, containing two methyl groups on the a carbon
of the linker, required us to exhaustively methylate pyrroli-
din-2-one according to a literature procedure (Scheme 1).[36]
The resulting a,a-dimethyl-g-lactam can be opened by heat-
ing in concentrated hydrochloric acid. The strong tendency
of this GABA derivative to recyclize makes a special proce-
dure necessary, that is, the activation of Cbz-Leucine as a
succinate ester and its smooth conversion with the obtained
hydrochloride in the presence of triethylamine to give the
free acid 5 in 57% yield. Although this acid can be easily
transformed into 7 with the established biphasic conditions,
the corresponding iodide of 7 could not be coupled with
HPQ, even though we tried all of the strategies described
above. In fact, TLC and mass analysis indicated the exclu-
sive formation of a compound resulting from coupling of the
À
N H site of HPQ. This compound was less polar and pos-
sessed the same mass as 9. It emitted a green fluorescence
under a UV lamp (365 nm) because the phenolic hydroxyl
was still intact. In contrast, compound 9 was more polar
than HPQ itself and showed absolutely no fluorescence
under a UV lamp (365 nm). Similar byproducts were also
found in the synthesis of 8.
Figure 1. Emission spectra (lEx =365 nm) of probe 2 (20 mm) before (a)
and after (b) incubation with leucyl aminopeptidase for 1 h at 258C in
Tris-HCl.
To avoid this unwelcome complication, we turned to an-
other strategy, which was to introduce the HPQ unit only
upon completion of the trigger–linker–phenol unit. Coupling
of the iodide of 7 with salicylaldehyde thus furnished alde-
hyde 10 by employing monophasic conditions and a strong
base to overcome the weak acidity of salicylaldehyde, which
exhibits an internal hydrogen bond. Condensation of this al-
dehyde with anthranilamide, followed by oxidation with 2,3-
dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) and catalytic
hydrogenation gave target 2 without any byproducts and in
satisfactory yield (65%; Scheme 2).
The photophysical properties of the proposed fluorogenic
probes 1 and 2 were evaluated under simulated physiologi-
cal conditions (Tris-HCl, pH 7.4, Tris=tris(hydroxymeth-
ACHTUNGTRENNUNGyl)aminomethane). As predicted, profluorescent substrates 1
and 2 emitted no fluorescence at 495 nm (maximum emis-
sion wavelength for free HPQ). However, after incubation
of 2 with leucyl aminopeptidase for 1 h, intense fluorescence
at 495 nm was observed (Figure 1). This initial result sup-
ported the conclusion that 2 is an effective substrate for the
enzyme and releases insoluble HPQ. The kinetics of the
transformation of 1 and 2 were followed with a spectro-
fluorimeter (Figure 2). The maximum fluorescence intensity
Figure 2. Fluorescence assay (lEx =365 nm, lEm =495 nm) for the cleavage
of substrates 1 and 2 with leucyl aminopeptidase in Tris-HCl (25 mm,
pH 7.4), substrate (38 mm), leucyl aminopeptidase (0.02 UmLÀ1), 258C.
With enzyme: a) substrate 2, b) substrate 1. Without enzyme (bovine
serum albumin (0.1 mgmLÀ1) was added): c) substrate 2, d) substrate 1.
794
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 792 – 795

Detection Probe for Acyl Hydrolases
COMMUNICATION
[5] Z. J. Huang, E. Terpetschnig, W. M. You, R. P. Haugland, Anal. Bio-
chem. 1992, 207, 32–39.
[6] J. J. Naleway, C. M. J. Fox, D. Robinhold, E. Terpetschnig, N. A.
Olson, R. P. Haugland, Tetrahedron Lett. 1994, 35, 8569–8572.
[7] M. J. Zhou, R. H. Upson, Z. Diwu, R. P. Haugland, J. Biochem. Bio-
phys. Methods 1996, 33, 197–205.
[8] Z. Diwu, Y. X. Lu, R. H. Upson, M. J. Zhou, D. H. Klaubert, R. P.
Haugland, Tetrahedron 1997, 53, 7159–7164.
[9] X. Zhang, M. Waibel, J. Hasserodt, Eur. Pat. Appl. 09305579.6,
2009.
attests to the quantitative conversion of 1 and 2 into free
HPQ. As expected, the highly preorganized spacer in 2 led
to a superior rate of conversion with the maximum intensity
reached after about 35 min at the indicated assay conditions.
But substrate 1 also showed acceptable kinetics with the
maximum fluorescence reached after about 150 min. More
importantly, substrate 2 showed a lower background in
signal generation (Figure 2, curve c) than that of substrate 1
(Figure 2, curve d), in congruence with the reported behav-
ior for a pivaloyl O,O-acetal.[20] The fact that a residual
background can still be observed for 2 may be explained by
the presence of the free N terminus. The observed kinetic
parameters (apparent Michaelis constant, KM,app and appar-
ent maximal reaction rate, Vmax,app) for probe 2 with leucyl
aminopeptidase at 258C were determined as KM,app =98 mm
[10] J. L. Reymond, V. S. Fluxa, N. Maillard, Chem. Commun. 2009, 34–
46.
[11] Enzyme Assays: High-throughput Screening, Genetic Selection and
Fingerprinting (Ed.: J.-L Reymond) Wiley-VCH, Weinheim, 2006.
[12] M. Waibel, PhD Thesis, Universitꢀ de Lyon—ENS (France), 2008.
[13] M. Waibel, X. Zhang, J. Hasserodt, Synthesis 2008, 318–324.
[14] J. J. Zamet, M. J. Haddadin, Issidori. Ch, J. Chem. Soc. Perkin Trans.
1 1974, 1687–1691.
[15] R. Niemi, J. Vepsalainen, H. Taipale, T. Jarvinen, J. Med. Chem.
1999, 42, 5053–5058.
and
V
max,app =5.12ꢃ10À8 msÀ1 (25 mm Tris-HCl, pH 7.4,
258C). The KM,app value is in the same range as that of Leu-
7-amido-4-methylcoumarin (KM =39.6 mm) at comparable
assay conditions,[37] whereas its Vmax,app value is three orders
of magnitude larger than that of the latter (Vmax =7.2ꢃ
10À11 msÀ1).
In conclusion, we have been able to employ the unique
solid-state fluorophore HPQ in the design of a water-soluble
nonfluorescent probe for a particular peptidase. The synthe-
sis developed opens up the possibility to prepare a large va-
riety of probes for exo-(amino)peptidases, but we do not
rule out the possibility that sequences may be found that in-
clude modifications of the GABA spacer to make the probe
susceptible to selected endoproteases as well. Many applica-
tions in the life sciences (e.g., high-throughput screening as
well as in vivo imaging) can be envisaged in which such
probes assay or image peptidases in a true off/on mode
without suffering from diffusion-related signal dilution.
[16] S. J. Hecker, M. D. Erion, J. Med. Chem. 2008, 51, 2328–2345.
[17] N. Bensel, M. T. Reymond, J. L. Reymond, Chem. Eur. J. 2001, 7,
4604–4612.
[18] O. Y. Hui, R. T. Borchardt, T. J. Siahaan, Tetrahedron Lett. 2002, 43,
577–579.
[19] J. D. Thomas, K. B. Sloan, Int. J. Pharm. 2008, 346, 80–88.
[20] E. Leroy, N. Bensel, J. L. Reymond, Bioorg. Med. Chem. Lett. 2003,
13, 2105–2108.
[21] F. Kratz, I. A. Muller, C. Ryppa, A. Warnecke, ChemMedChem
2008, 3, 20–53.
[22] P. L. Carl, P. K. Chakravarty, J. A. Katzenellenbogen, J. Med. Chem.
1981, 24, 479–480.
[23] D. X. Shan, M. G. Nicolaou, R. T. Borchardt, B. H. Wang, J. Pharm.
Sci. 1997, 86, 765–767.
[24] S. S. Chandran, K. A. Dickson, R. T. Raines, J. Am. Chem. Soc.
2005, 127, 1652–1653.
[25] G. B. Jones, C. F. Crasto, J. E. Mathews, L. F. Xie, M. O. Mitchell, A.
El-Shafey, A. V. DꢁAmico, G. J. Bubley, Bioorg. Med. Chem. 2006,
14, 418–425.
[26] L. D. Lavis, T. Y. Chao, R. T. Raines, ACS Chem. Biol. 2006, 1, 252–
260.
[27] L. D. Lavis, T. Y. Chao, R. T. Raines, ChemBioChem 2006, 7, 1151–
1154.
[28] N. H. Ho, R. Weissleder, C. H. Tung, ChemBioChem 2007, 8, 560–
566.
Acknowledgements
X.Z. gratefully acknowledges financial support from the French Research
Ministry (visiting professorship to the ENS) and from the National Natu-
ral Science Foundation of China (Grants 20505008, 20975034). M.W.
thanks the French Research Ministry for a doctoral fellowship for Euro-
pean mobility.
[29] E. Danieli, D. Shabat, Bioorg. Med. Chem. 2007, 15, 7318–7324.
[30] R. Sicart, M.-P. Collin, J.-L. Reymond, Biotechnol. J. 2007, 2, 221–
231.
[31] Y. Meyer, J. A. Richard, M. Massonneau, P. Y. Renard, A. Romieu,
Org. Lett. 2008, 10, 1517–1520.
[32] J. A. Richard, Y. Meyer, V. Jolivel, M. Massonneau, R. Dumeunier,
D. Vaudry, H. Vaudry, P. Y. Renard, A. Romieu, Bioconjugate
Chem. 2008, 19, 1707–1718.
[33] T. C. Bruice, U. K. Pandit, J. Am. Chem. Soc. 1960, 82, 5858–5865.
[34] R. P. Iyer, D. Yu, N. H. Ho, S. Agrawal, Synth. Commun. 1995, 25,
2739–2749.
Keywords: autoimmolative spacers · diagnostic agents ·
fluorescent probes · peptidases · prodrugs
[35] J. D. Thomas, K. B. Sloan, Tetrahedron Lett. 2007, 48, 109–112.
[36] Y. Ahn, G. I. Cardenas, J. Yang, D. Romo, Org. Lett. 2001, 3, 751–
754.
[37] S. K. Grant, J. G. Sklar, R. T. Cummings, J. Biomol. Screening 2002,
7, 531–540.
[1] National Research Council, Visualizing Chemistry—The Progress
and Promise of Advanced Chemical Imaging, The National Academ-
ies Press, Washington D.C., 2006.
[2] A. Heller, D. L. Williams, J. Phys. Chem. 1970, 74, 4473–4480.
[3] J. Catalan, F. Fabero, M. S. Guijarro, R. M. Claramunt, M. D. S.
Maria, M. D. Focesfoces, F. H. Cano, J. Elguero, R. Sastre, J. Am.
Chem. Soc. 1990, 112, 747–759.
Received: October 28, 2009
[4] L. M. Tolbert, K. M. Solntsev, Acc. Chem. Res. 2002, 35, 19–27.
Published online: November 26, 2009
Chem. Eur. J. 2010, 16, 792 – 795
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
795