Novel activity-based probes for broad-spectrum profiling of retaining β-exoglucosidases in situ and in vivo

Article abstract of DOI:10.1002/anie.201207771

A high-end label: Cyclophellitol aziridine-type activity-based probes allow for ultra-sensitive visualization of mammalian β-glucosidases (GBA1, GBA2, GBA3, and LPH) as well as several non-mammalian β-glucosidases (see picture). These probes offer new ways to study β-exoglucosidases, and configurational isomers of the cyclophellitol aziridine core may give activity-based probes targeting other retaining glycosidase families. Copyright

Full text of DOI:10.1002/anie.201207771

Angewandte
Chemie
DOI: 10.1002/anie.201207771
Enzyme Labeling
Novel Activity-Based Probes for Broad-Spectrum Profiling of
Retaining b-Exoglucosidases In Situ and In Vivo**
Wouter W. Kallemeijn, Kah-Yee Li, Martin D. Witte, Andrꢀ R. A. Marques, Jan Aten,
Saskia Scheij, Jianbing Jiang, Lianne I. Willems, Tineke M. Voorn-Brouwer,
Cindy P. A. A. van Roomen, Roelof Ottenhoff, Rolf G. Boot, Hans van den Elst,
Marthe T. C. Walvoort, Bogdan I. Florea, Jeroen D. C. Codꢀe, Gijsbert A. van der Marel,
Johannes M. F. G. Aerts,* and Herman S. Overkleeft*
Retaining b-exoglucosidases are a broad class of glycosidases
widely found in nature. In humans, four members of this
family are known to date. Of these, the lysosomal enzyme
glucocerebrosidase (GBA1) and the non-lysosomal glucosyl-
ceramidase (GBA2) catalyze the hydrolysis of glucosylcer-
amide.[1–6] Deficiency of GBA1 forms the basis for the most
common inherited lysosomal storage disorder, Gaucher
disease.[2] Recently, mutations of GBA1 have been reported
to markedly increase the risk for Parkinsonism,[7–9] and
excessive degradation of glucosylceramide by GBA2 may
play a role in neuropathology.[4,10–12] The cytosolic broad-
specificity b-glucosidase (GBA3) is thought to be involved in
degrading xenobiotic b-glucosides.[13,14] The intestinal lactase/
phlorizin hydrolase (LPH) is able to hydrolyze lactose as well
as some hydrophobic b-glucosides including glucosylcera-
mide.[15] Deficient LPH activity underlies lactose intoler-
ance.[16]
have received growing interest in medicinal chemistry.[17,18]
An important issue is the target specificity of such com-
pounds, particularly given the limited knowledge of the
physiological role of GBA2/3. Until recently, few tools were
available to monitor active retaining b-exoglucosidases in
living cells, and the effect of potential inhibitors on these
cells.[19] Recently, fluorescent activity-based probes (ABPs)
have been developed that allow labeling of active GBA1
molecules.[20] These cyclophellitol epoxide-based ABPs 1 and
2 covalently bind to the nucleophile (residue E340) of GBA1
and allow visualization of active enzyme in living cells and
mice (Figure 1a,b). The BODIPY extension at C6 markedly
increases the affinity for binding to the GBA1 catalytic
pocket.[20] At the same time, GBA2, GBA3, and LPH do not
or only poorly bind ABPs 1 and 2. All the b-exoglucosidases
studied to date can, however, hydrolyze the artificial substrate
4-methylumbelliferyl-b-d-glucopyranoside, in which the agly-
con is located at the C1 position. Based on this we decided to
add a reporter group near this position in the ABP, and used
the aziridine-analogue of cyclophellitol as a scaffold to
develop broad-spectrum retaining b-exoglucosidase ABPs.
Cyclophellitol-aziridine has been found to be a potent
mechanism-based retaining b-exoglucosidase inhibitor.[21–24]
Herein, we reveal the development of aziridine ABPs 3 and
4 and demonstrate their merits as broad-spectrum retaining b-
exoglucosidase probes capable of tagging the human and
murine enzymes GBA, GBA2, GBA3, and LPH as well as
a wide variety of non-mammalian b-exoglucosidases.
The synthesis of ABPs 3 and 4 is described in the
Supporting Information (Scheme S1). Whereas epoxides
1 and 2 are selective nanomolar GBA1 inhibitors, aziridines
3 and 4 inhibit GBA1, GBA2, and GBA3 in the low
nanomolar to high picomolar range (IC50 values are given in
Figure S1) Recombinant human GBA1 (Imiglucerase, Gen-
zyme) was used to further study the potency of 3. GBA1 is
a retaining b-exoglucosidase in which E340 acts as a nucleo-
phile and E235 as an acid/base during catalysis.[25] Pre-
incubation of GBA1 with the irreversible inhibitors condur-
itol b-epoxide (CBE), (azido-)cyclophellitol, and ABP 2
completely blocked covalent labeling by ABP 3 (Figure 2a).
As expected, pre-incubation with non-fluorescent ABP 8 (see
Scheme S1 for the formula of 8) also blocked labeling by 3, as
did high concentrations of the competitive inhibitor AMP-
DNM.[4] These findings confirm that covalent labeling of
GBA1 with 3 involves the active site, and in particular E340.
In recent years, small molecules that interact with b-
glucosidases, either as inhibitors or as molecular chaperones,
[*] W. W. Kallemeijn,[+] A. R. A. Marques, S. Scheij, T. M. Voorn-Brouwer,
C. P. A. A. van Roomen, R. Ottenhoff, Dr. R. G. Boot,
Prof. Dr. J. M. F. G. Aerts
Department of Medicinal Biochemistry, Academic Medical Center
Meibergdreef 15, 1105 AZ, Amsterdam (The Netherlands)
E-mail: j.m.aerts@amc.uva.nl
K.-Y. Li,[+] Dr. M. D. Witte, J. Jiang, L. I. Willems, H. van den Elst,
M. T. C. Walvoort, Dr. B. I. Florea, Dr. J. D. C. Codꢀe,
Prof. Dr. G. A. van der Marel, Prof. Dr. H. S. Overkleeft
Leiden Institute of Chemistry and the Netherlands Proteomics
Centre
P.O. box 9502, 2300 RA Leiden (The Netherlands)
E-mail: h.s.overkleeft@chem.leidenuniv.nl
Dr. J. Aten
Department of Pathology
Meibergdreef 15, 1105 AZ Amsterdam (The Netherlands)
[+] These authors contributed equally to this work.
[**] This work was supported by The Netherlands Organization for
Scientific Research (NWO-CW) and The Netherlands Genomics
Initiative. We thank K. T. Senel for technical assistance.
Supporting information for this article (experimental details) is
available on the WWW under http://dx.doi.org/10.1002/anie.
201207771. Details of synthesis, purification and analysis of ABPs 3
and 4 and intermediates are provided, as are full details of enzyme
activity, molecular cloning, site-directed mutagenesis, and fluores-
cence microscopy. Full gels are also provided in the Supporting
Information.
Angew. Chem. Int. Ed. 2012, 51, 12529 –12533
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
12529

.
Angewandte
Communications
Complete irreversible dena-
turation of GBA1 by expo-
sure to high temperature and
1% (w/v) SDS abolishes
labeling by ABP 3 to GBA1
(Figure 2a). Thus aziridine
ABP 3 behaves like the epox-
ide ABPs 1/2 towards label-
ing and sensitivity of GBA1
inhibition (see Figure S3 for
labeling sensitivity).
We had found previously
that labeling with ABPs
1 and 2 coincides exactly
with the pH profile of enzy-
matic activity towards 4-
methylumbelliferyl-b-d-glu-
copyranoside (Figure 2c). In
contrast, the cyclophellitol
aziridine-based ABP
3 is
able to label GBA1, also at
high pH conditions where it
is inactive towards the artifi-
cial substrate 4-methylum-
belliferyl-b-d-glucopyrano-
side (Figure 2b,c). It is intri-
guing that ABP 3 is able to
label GBA1 at high pH
values under which the
enzyme is inactive, this sug-
gests that labeling by this
Figure 1. a) Cyclophellitol-epoxide type ABPs 1 and 2 and broad-spectrum cyclophellitol aziridine ABPs 3 and
4. b,c) Mechanism of irreversible inhibition of retaining b-exo-glucosidases by b) ABP 1 or 2 and c) ABP 3 or
4.
ABP does not require enzyme-based protonation. The
acylated nitrogen in the aziridine structure, a reasonably
good electrophile, appears to not need protonation, thereby
eliminating the need for acid/base catalysis. To test this, we
substituted the acid/base E235 and nucleophile E340 in myc/
His-tagged GBA1 separately to both glutamine and glycine
by site-directed mutagenesis, giving recombinant GBA1 with
E235G, E235Q, E340G, and E340Q mutations. As expected,
neither ABP labeled GBA1 when the nucleophile was
mutated to either a glycine or glutamine, but ABP 3 was
able to label GBA1 when only the acid/base was mutated.
Interestingly, significant labeling of E235Q by 3 was observed,
indicating that the glutamine activates the acylated aziridine
through hydrogen bonding (Figure S5). These findings offer
an explanation for the different pH dependence of labeling of
GBA1 by the cyclophellitol epoxide-type ABP 2 and cyclo-
phellitol aziridine-type ABP 3.
We next studied the ability of ABP 3 to label GBA1 in
living cells. For this purpose, we incubated cultured HepG2
cells with a suboptimal amount (5 nm) of ABP 2 for two
hours. After washing and incubation with 5 mm a-cyclo-
dextrin to remove any free ABP 2, cells were washed and then
labeled with 10 nm of ABP 3 for 16 hours. Finally, fixed cells
were examined by confocal-laser scanning microscopy. As
shown in Figure 2d, incubation with both ABP 2 and 3
resulted in a similar staining of lysosomes.[20]
Figure 2. In vitro labeling of recombinant GBA1. a) Labeling with 3
was blocked in the presence of various inhibitors or by denaturing the
protein (full gel shown in Figure S2). b) Effect of pH value on labeling
of GBA1 using 2 or 3 (full gel shown in Figure S4). c) Plot showing the
quantification of in-gel fluorescence from (b) of GBA1 labeled with
~
ABP 2 ( ), 3 (&), and compared with GBA1 enzymatic activity profile
towards 4-methylumbelliferyl-b-d-glucopyranoside (*). Values have
been normalized to the conditions at pH 5, error bars indicate the
standard deviation. d) In situ labeling of GBA1. Representative spectral
imaging micrographs of cells labeled with ABPs 2 (red) and 3 (green),
and nuclei stained with DAPI (blue). Regions with overlapping signal
of 2 and 3 are yellow. Scale bar=25 mm (complete micrograph analysis
is shown in Figure S6).
We next overexpressed the human b-exoglucosidases
GBA2, GBA3, and LPH in COS-7 cells. Then, lysates of the
12530 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12529 –12533

Angewandte
Chemie
cells were incubated with 100 nm ABP 3 as well as 2 at
different pH-values (Figure S7). Figure 3a shows the labeling
at optimal pH values for each b-glucosidase. ABP 2 specif-
ically labels GBA1 in the various cell lysates whereas
different concentrations of the ABPs 3 or 2 at the optimal pH
determined for each enzyme and with the required additives
(Figure S11, S12). Labeling was analyzed following gel
electrophoresis (Figure 3c). ABP 2 labels extremely well
human GBA1, but not the other glucosidases tested,
with the exception of the b-glucosidases of R. etli,
B. fragilis, and Agrobacterium sp. Less-intense labeling
was observed for the b-glucosidase from C. thermocel-
lum and Thermotoga maritime GH1. ABP 3 labeled
the enzymes mentioned above more avidly, plus also b-
glucosidase from P. dulcis and A. fumigatis. Dimin-
ished labeling was observed for SP1599 and b-gluco-
sidases from A. niger and A. fumigatus. Neither 2 nor 3
reacted with the xylanases and cellulase. No clear
correlation between labeling of b-glucosidases and
their classification in the glycosidase protein families
(CAZy) was observed (Table S2). At present, it is not
clear how important the region of the binding pocket
with the aglycon is for labeling by ABP 3. A more
systematic survey of the structural requirements for
covalent labeling, including studies with crystals of
various retaining b-exoglucosidases with and without
bound 3, may help in answering this question.
Figure 3. In vitro labeling of multiple retaining b-exoglucosidases. a) In vitro
fluorescent labeling of endogenous GBA1 (*) and recombinant GBA2 ( ),
GBA3 (&), and LPH (*) in homogenates of COS-7 cells using 2 and 3.
Overlaid image of each cut out gel is shown at right panel of each set.
b) Labeling of endogenously expressed b-glucosidases in homogenates of
murine liver, duodenum, kidney, and brain using 2 and 3. Overlaid image of
We next investigated the ability of ABP 3 to label
retaining b-glucosidases in living mice. For this pur-
pose, we injected intravenously mice with 10, 100, or
1000 pmol (about 0.2 mgkgÀ1, 2 mgkgÀ1, and 20 mgkgÀ1,
respectively) of ABP 1 or 3. After two hours, animals
^
each cut out gel is shown at right panel of each set. Equal fluorescence in both
channels is yellow in the overlaid image. c) Analysis of binding efficacy in vitro
for non-mammalian glycosidases for different concentrations of ABP 2 and 3 at
the corresponding optimal pH and buffer conditions A) Homo sapiens GBA1;
B) R. etli CFN42 GBA; C) B. fragilis NCTC GBA; D) Agrobacteria sp. GBA;
E) C. thermocellum GBA; F) Th. Maritime GH1; G) A. niger GBA; H) Sp1599;
I) P. dulcis GBA; J) A. fumigates GBA; K) A. niger cellulase C1184; L) T. viride
xylanase; M) T. lanuginosus xylanase). See also Figures S9–S12.
were deeply anesthetized with FFM mix before
perfusion with PBS, and then various tissues were
isolated. Homogenates of tissues were prepared,
aliquots (1 mg) were subjected to gel electrophoresis,
and fluorescently labeled proteins were visualized
(Figure 4a–d).
In brain (Figure 4a), no proteins were labeled by
treatment of mice with ABP 3 or 1. In contrast, in vitro
incubation of a brain homogenate of a vehicle-treated
cyclophellitol aziridine-type ABP 3 is able to, in addition,
label all other b-glucosidases: GBA2 MW= 102 kDa, GBA3
MW= 54 kDa, and LPH MW is a doublet of 210 kDa and
160 kDa.
mouse shows labeling of GBA1 with 1000 pmol ABP 1 or 3.
Also, GBA2 and GBA3 in brain homogenates are labeled
with ABP 3. In duodenum (Figure 4b), a prominent labeling
of LPH had occurred in ABP 3 and 1 treated mice. It should,
however, be noted that in animals treated with a low dose of 3
(but not with 1), LPH was maximally labeled. This indicates
a marked difference in affinity for in vivo labeling of LPH by
ABP 3 versus 1. The duodenum-derived image (Figure 4b)
reveals a complex labeling pattern compared to the other
tissues (also observed in the duodenum extracts, Figure 2b).
Likely the added bands stem from LPH that has been
partially processed by digestive proteases, although it cannot
be excluded that a-glucosidases are targeted by our ABPs. In
kidney (Figure 4c), ABP 1 treatment only resulted in a dose-
dependent labeling of GBA1. ABP 3 treatment resulted, in
addition, in dose-dependent labeling of GBA3. GBA2 was
already maximally labeled in animals treated with the lowest
dose of ABP 3. Finally, in liver (Figure 4d), ABP 1 treated
mice showed only dose-dependent labeling of GBA1, whilst
in ABP 3 treated mice additional labeling of GBA2 occurred.
The ability of 3 to profile multiple retaining b-exogluco-
sidases in situ and in vivo and in their enzymatically active
To test the specificity of ABP 3, we incubated homoge-
nates of various mouse tissues with 100 nm probe 3 for
30 minutes. An identical labeling was performed with 2.
Figure 3b shows that 2 specifically labels GBA1 in the tissue
homogenates, with the exception of faint labeling of LPH in
the duodenum. In sharp contrast, ABP 3 not only labeled
GBA1 in all tissue homogenates, but also GBA2 in the liver,
kidney, and brain, GBA3 in kidney and brain, and LPH in the
duodenum. No other proteins were modified by ABP 7,
illustrating the high degree of specificity. We confirmed
labeling of GBA1 and GBA2 by cyclophellitol aziridine-type
probes by labeling mouse spleen homogenate and HEK cell
lysate with biotin aziridine ABP 4, followed by streptavidin
pull-down and mass-spectrometry based identification of the
captured proteins (Figure S8).
We next extended our investigation to non-mammalian
glucosidases that are commercially available. About one
microgram of each individual enzyme was incubated with
Angew. Chem. Int. Ed. 2012, 51, 12529 –12533
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12531

.
Angewandte
Communications
Keywords: activity-based probes ·
.
fluorescent probes · glucocerebrosidase ·
glycosidases · proteomics
[1] R. O. Brady, J. N. Kanfer, R. M. Brad-
ley, D. Shapiro, J. Clin. Invest. 1966, 45,
1112 – 1115.
[2] G. A. Grabowski, Lancet 2008, 372,
1263 – 1271.
[3] S. Van Weely, M. Brandsma, A. Strij-
land, J. M. Tager, J. M. Aerts, Biochim.
Biophys. Acta Mol. Basis Dis. 1993,
1181, 55 – 62.
[4] H. S. Overkleeft, G. H. Renkema, J.
Neele, P. Vianello, I. O. Hung, A. Strij-
land,
A. M.
van der Burg,
G. J.
Koomen, U. K. Pandit, J. M. Aerts, J.
Biol. Chem. 1998, 273, 26522 – 26527.
[5] Y. Yildiz, H. Matern, B. Thompson,
J. C. Allegood, R. L. Warren, D. M.
Ramirez, R. E. Hammer, F. K. Hamra,
S. Matern, D. W. Russell, J. Clin. Invest.
2006, 116, 2985 – 2994.
[6] R. G. Boot, M. Verhoek, W. Donker-
Koopman, A. Strijland, J. Van Marle,
H. S. Overkleeft, T. Wennekes, J. M.
Aerts, J. Biol. Chem. 2007, 282, 1305 –
1312.
Figure 4. In vivo labeling of endogenous retaining b-exoglucosidases in mice using 1 and 3.
Fluorescently labeled proteins were analyzed in tissue homogenates of a) brain, b) duodenum,
c) kidney, and d) liver. Marked are human recombinant GBA1 (*), endogenous GBA1 (*),
^
GBA2 ( ), GBA3 (&), LPH and its degradation fragments ( ). Full gels are shown in Figure S13.
*
CBB=Coomassie Brilliant Blue staining.
[7] O. Goker-Alpan, G. Lopez, J. Vithayathil, J. Davis, M. Hallett, E.
Sidransky, Arch. Neurol. 2008, 65, 1353 – 1357.
[8] W. Westbroek, A. M. Gustafson, E. Sidransky, Trends Mol. Med.
2011, 17, 485 – 493.
[9] T. Shachar, C. Lo Bianco, A. Recchia, C. Wiessner, A. Raas-
Rothschild, A. H. Futerman, Mov. Disord. 2011, 26, 1593 – 1604.
[10] J. M. Aerts, C. Hollak, R. Boot, A. Groener, Philos. Trans. R.
Soc. London Ser. B 2003, 358, 905 – 914.
[11] K. M. Ashe, D. Bangaari, L. Li, M. A. Cabrera-Salazar, S. D.
Bercury, J. B. Nietupski, C. G. Cooper, J. M. Aerts, E. R. Lee,
D. P. Copeland, S. H. Cheng, R. K. Scheule, J. Marhall, PLoS
One 2011, 6, e21758.
form has great potential. It allows investigations on a variety
of enzymes such as GBA2, GBA3, and LPH. Further
investigations of GBA1 and GBA2 using ABPs are of
particular interest because of the observed link between
deficiency in GBA1 and the risk for the development of
Parkinsonism.[7–9] It is conceivable that impaired degradation
of glucosylceramide in GBA-deficient lysosomes is compen-
sated by non-lysosomal hydrolysis by GBA2.[10] GBA2 may
thus play an important role in pathophysiological processes
associated with GBA1 deficiency. The new ABPs 3 and 4 are
elegant tools to study GBA2 in this respect. ABPs 3 and 4, and
to a lesser extent 1 and 2, also may be used to study LPH,
whose deficiency causes lactose intolerance.
[12] B. Nietupski, J. J. Pacheco, W. L. Chuang, K. Maratea, L. Li, J.
Foley, K. M. Ashe, C. G. Cooper, J. M. Aerts, D. P. Copeland,
R. K. Scheule, S. H. Cheng, J. Marshall, Mol. Genet. Metab. 2012,
105, 621 – 628.
In conclusion, we have developed new potent broad-
spectrum activity-based probes to monitor retaining b-gluco-
sidases in vitro, in situ, and in vivo. Our new cyclophellitol
aziridine-based ABPs, together with our previously reported
cyclophellitol probes, are a highly useful pair of probes for the
study of human b-glucosidases in relation to disease states.
Modification of the acyl aziridine scaffold, for instance
through a strategy by which broad-spectrum serine hydrolase
inactivators were turned into selective inhibitors,[26] may lead
to selective irreversible inhibitors of GBA2/3 and LPH for
subsequent biological studies on these enzymes. Finally, the
design of ABPs 3/4, with the reporter pointing towards the
area normally occupied by the substrate aglycon, suggests that
configurational isomers of 3 could be designed that may result
in effective activity-based probes that are able to label
a variety of retaining glycosidases.
[13] V. Gopalan, A. Pastuszyn, W. R. Galey, Jr., R. H. Glew, J. Biol.
Chem. 1992, 267, 14027 – 14032.
[14] N. Dekker, T. Voorn-Brouwer, M. Verhoek, T. Wennekes, R. S.
Narayan, D. Speijer, C. E. Hollak, H. S. Overkleeft, R. G. Boot,
J. M. Aerts, Blood Cells Mol. Dis. 2011, 46, 19 – 26.
[15] J. C. Arribas, A. G. Herrero, M. Martin-Lomas, F. J. Canada, S.
He, S. G. Withers, Eur. J. Biochem. 2000, 267, 6996 – 7005.
[16] M. Lloyd, G. Mevissen, M. Fischer, W. Olsen, D. Goodspeed, M.
Genini, W. Boll, G. Semenza, N. Mantei, J. Clin. Invest. 1992, 89,
524 – 529.
[17] T. Wennekes, R. J. B. H. N. van den Berg, R. G. Boot, G. A.
van der Marel, H. S. Overkleeft, J. M. F. G. Aerts, Angew. Chem.
2009, 121, 9006 – 9028; Angew. Chem. Int. Ed. 2009, 48, 8848 –
8869.
[18] A. Trapero, P. Gonzꢀlez-Bulnes, T. D. Butters, A. Llebaria, J.
Med. Chem. 2012, 55, 4479 – 4488.
[19] B. P. Rempel, M. B. Tropak, D. J. Mahuran, S. G. Withers,
Angew. Chem. 2011, 123, 10565 – 10567; Angew. Chem. Int. Ed.
2011, 50, 10381 – 10383.
[20] M. D. Witte, W. W. Kallemeijn, J. Aten, K. Y. Li, A. Strijland,
W. E. Donker-Koopman, A. M. van den Nieuwendijk, B. Bleijl-
evens, G. Kramer, B. I. Florea, B. Hooibrink, C. E. Hollak, R.
Received: September 26, 2012
Published online: November 8, 2012
12532 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12529 –12533

Angewandte
Chemie
Ottenhoff, R. G. Boot, G. A. van der Marel, H. S. Overkleeft,
J. M. Aerts, Nat. Chem. Biol. 2010, 6, 907 – 913.
[21] M. Nakata, C. Chong, Y. Nitawa, K. Toshima, K. Tasuma, J.
Antibiot. 1993, 46, 1919 – 1924.
[24] K. Tatsuta, Pure Appl. Chem. 1996, 68, 1341 – 1346.
[25] L. Premkumar, A. R. Sawkar, S. Boldin-Adamsky, L. Toker, I.
Silman, J. W. Kelly, A. H. Futerman, J. L. Sussman, J. Biol.
Chem. 2005, 280, 23815.
[22] G. Caron, S. G. Withers, Biochem. Biophys. Res. Commun. 1989,
163, 495 – 499.
[23] M. K. Tong, B. Ganem, J. Am. Chem. Soc. 1988, 110, 312 – 314.
[26] A. Adibekian, B. R. Martin, C. Wang, K.-L. Hsu, D. A.
Bachovchin, S. Niessen, H. Hoover, B. F. Cravatt, Nat. Chem.
Biol. 2011, 7, 469 – 478.
Angew. Chem. Int. Ed. 2012, 51, 12529 –12533
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12533