S-Arachidonoyl-2-thioglycerol synthesis and use for fluorimetric and colorimetric assays of monoacylglycerol lipase

Article abstract of DOI:10.1016/j.bmc.2010.01.034

We report the first synthesis of 2-thioglycerol and S-arachidonoyl-2-thioglycerol (the thioester analog of the endocannabinoid 2-arachidonoylglycerol) in an eight or nine step procedure with a yield of ～25% and establish the use of this substrate for maleimide-based fluorescent and dithiobis(2-nitrobenzoic acid)-based colorimetric assays of human recombinant monoacylglycerol (MAG) lipase (hMAGL) and human brain membrane MAG hydrolase activity. Inhibitor structure-activity relationships observed here for hMAGL and 2-ATG correlate well (r² = 0.93, n = 9) with earlier findings for mouse brain MAG hydrolase with non-thiol substrates.

Full text of DOI:10.1016/j.bmc.2010.01.034

Bioorganic & Medicinal Chemistry 18 (2010) 1942–1947
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
S-Arachidonoyl-2-thioglycerol synthesis and use for fluorimetric
and colorimetric assays of monoacylglycerol lipase
a
b
b
John E. Casida a, , Alexander G. Gulevich , Richmond Sarpong , Eric M. Bunnelle
*
a Department of Nutritional Sciences and Toxicology, University of California, Berkeley, CA 94720, USA
b Department of Chemistry, University of California, Berkeley, CA 94720, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
We report the first synthesis of 2-thioglycerol and S-arachidonoyl-2-thioglycerol (the thioester analog of
the endocannabinoid 2-arachidonoylglycerol) in an eight or nine step procedure with a yield of ꢀ25% and
establish the use of this substrate for maleimide-based fluorescent and dithiobis(2-nitrobenzoic acid)-
based colorimetric assays of human recombinant monoacylglycerol (MAG) lipase (hMAGL) and human
brain membrane MAG hydrolase activity. Inhibitor structure–activity relationships observed here for
hMAGL and 2-ATG correlate well (r2 = 0.93, n = 9) with earlier findings for mouse brain MAG hydrolase
with non-thiol substrates.
Received 11 December 2009
Revised 13 January 2010
Accepted 14 January 2010
Available online 18 January 2010
Keywords:
Monoacylglycerol lipase
Endocannabinoid
Ó 2010 Elsevier Ltd. All rights reserved.
2-Arachidonoyl glycerol
S-Arachidonoyl-2-thioglycerol
1. Introduction
A rapid and selective high-throughput MAGL or 2-AG hydrolase
assay would greatly facilitate the discovery of new inhibitors and
The endocannabinoids 2-arachidonoylglycerol (2-AG)1 and N-
arachidonoyl ethanolamide (anandamide, AEA)2 (Fig. 1) bind to
and activate two G-protein coupled receptors, CB1 and CB2.3 CB1
receptors are the most abundant G-protein coupled receptors in
brain and are also expressed in many other parts of the body, while
CB2 receptors are almost exclusively in the immune system. After
serving their signaling roles 2-AG and AEA are taken up into the
cell where 2-AG is hydrolyzed primarily by monoacylglycerol li-
pase (MAGL) to arachidonic acid (AA) and glycerol4,5 and AEA by
fatty acid amide hydrolase (FAAH) to AA and ethanolamine6,7
(Fig. 1).
candidate therapeutic agents. MAGL in brain membranes mediates
about 85% of the total 2-AG hydrolase activity with two additional
serine hydrolases bringing the total to 98%.27,28 Outside of the
brain, substrate specificity and the contribution of MAGL to total
2-AG hydrolase activity differs, although in the majority of tissues
MAGL is most important. Selection of a preferred substrate for
MAGL assays depends on the enzyme source and preparation, the
analytical method (radiotracer, chromatographic, colorimetric or
The endocannabinoid system is a potential therapeutic target8–10
for appetite disorders,11,12 pain and inflammation,13–15 neurode-
generative diseases,16–19 and cancer.20 Although CB1 agonists such
as tetrahydrocannabinol have therapeutic potential, they are
associated with side effects, such as hypomotility, catalepsy, hypo-
thermia, and cognitive dysfunction. An alternative therapeutic
strategy is to block the degradation of 2-AG or AEA by inhibiting
MAGL or FAAH or both.21–23 While FAAH is moderately well under-
stood, with a defined three dimensional structure, knockout mice,
and a plethora of small molecule inhibitors, the tertiary structure
of MAGL is only recently defined,24,25 knockout mice don’t exist,
and only a few small-molecule inhibitors have been identified.26
Figure 1. Monoacylglycerol lipase (MAGL) hydrolyzes the principal endocannab-
inoid 2-arachidonoylglycerol (2-AG) or its thiol ester analog S-arachidonoyl-2-
thioglycerol (2-ATG) to arachidonic acid (AA) and glycerol or 2-thioglycerol (2-TG).
The second endocannabinoid anandamide (AEA) is hydrolyzed by fatty acid amide
hydrolase (FAAH) to AA and ethanolamine.
* Corresponding author. Tel.: +1 510 642 5424; fax: +1 510 642 6497.
E-mail address: ectl@berkeley.edu (J.E. Casida).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.01.034

J. E. Casida et al. / Bioorg. Med. Chem. 18 (2010) 1942–1947
1943
This paper describes the first synthesis of 2-TG and 2-ATG
(Scheme 2) and then establishes their use with MMBC and DTNB
for fluorimetric and colorimetric assays of MAGL and other en-
zymes involved in 2-AG hydrolysis. To our knowledge, 2-TG is
the only one of the mono-, di- and trithioglycerols not previously
reported.
2. Results and discussion
2.1. Sources of chemicals and synthesis
AEA, 1-ATG and JZL184 were from Cayman Chemical Co. (Ann
Arbor, MI), DTNB and 1- and 2-OG were from Sigma (St. Louis,
MO) and MMBC was from Berry and Associates (Dexter, MI). Sev-
eral of the OP inhibitors were prepared in this laboratory (Supple-
mentary data).
Scheme 1. Fluorimetric and colorimetric analysis of 2-TG from enzymatic hydro-
lysis of 2-ATG.
fluorimetric) and the number of samples (a few samples versus ra-
pid-throughput screens). The endocannabinoid 2-AG itself is obvi-
ously the most relevant substrate. Many types of 2-AG hydrolysis
assays have been used involving chromatographic and radiotracer
techniques (Supplementary data). 1-Acylglycerol and other esters
also have been employed as alternative MAGL substrates including
1-[14C]oleoylglycerol (radiotracer); 1-(p-nitrobenzofurazanacyl)-
glycerol and S-arachidonoyl-1-TG (1-ATG) (colorimetric); and 7-
hydroxycoumarinyl arachidonate (fluorimetric) (Supplementary
data). The colorimetric and fluorimetric assays with 1-acylglycerol
esters allow kinetic monitoring of the reactions and high-through-
put screens but unless recombinant or purified preparations are
used they may measure an undefined balance of enzymes in addi-
tion to MAGL.
A synthesis of 2-TG started with selective protection of the pri-
mary hydroxyl groups of glycerol with tert-butyldimethylsilylchlo-
ride (TBSCl) and imidazole in tetrahydrofuran (THF) to give 1
(Scheme 2). Under these conditions, glycerol was only sparingly
soluble in THF, whereas 1 was very soluble which therefore might
be expected to give triprotected glycerol. However, the high yield
of diprotected glycerol suggested that the steric requirements for
functionalization of the secondary hydroxyl slowed its silylation
significantly, an observation confirmed in subsequent steps. Initial
attempts at activation of the secondary hydroxyl with toluenesul-
fonyl chloride or benzenesulfonyl chloride failed to give the de-
sired product even after prolonged reaction time. Only with
methanesulfonyl chloride (MsCl) as the electrophile did the reac-
tion proceed at a useful rate under ambient conditions to give 2.
Attempts to displace the secondary hydroxyl group with halogens
(such as potassium iodide and sodium bromide) and sulfur nucle-
ophiles (such as potassium thioacetate, potassium thiocyanate, and
potassium hydrogen sulfide) universally failed at this point, pre-
sumably because of the steric demands of the TBS-ethers. Cleavage
of the TBS-ethers of mesylate 2 with methanol and acidic Amber-
lystÒ 15 resin gave diol 3 in good yield. Concerns that this com-
pound could readily decompose to an epoxide proved not to be
the case. In fact, 3 was robust enough to be subjected to substitu-
tion conditions with potassium thiocyanate in acetonitrile at re-
flux to give 4. Although 4 could be characterized by NMR
spectroscopy, it proved to be unstable which complicated efforts
to reduce it to 2-TG. Therefore, 4 was immediately protected with
tert-butylsilyl (TBS) groups to give 5 in moderate yield over two
steps. Isothiocyanate 5 was successfully reduced using lithium
aluminum hydride, and the TBS-ethers were cleaved to give 2-
TG in 58% yield over two steps. Intermediate 6 was not isolated
An obvious alternative to 2-AG for MAGL assays is its thiolester
S-arachidonoyl-2-thioglycerol (2-ATG). It is the closest mimic of
2-AG with minimal changes in steric and electronic properties.
The hydrolysis product 2-thioglycerol (2-TG) should be easily
determined by fluorimetric assay with methyl maleimidobenzo-
chromenecarboxylate (MMBC)29 or colorimetric analysis with
5,50-dithiobis(2-nitrobenzoic acid) (DTNB)30 (Scheme 1) allowing
rapid-throughput screens. This potential is readily evident from
the extensive use of acetylthiocholine in place of acetylcholine in
studying acetylcholinesterase and its inhibitors.31,32 There is also
the possibility that 2-ATG may allow histochemical localization
of the cytological distribution of MAGL activity based on derivati-
zation of liberated 2-TG as has been used so successfully for cholin-
esterase localization based on thiocholine release.32,33 The MAGL
active site has the typical Ser-Asp-His catalytic triad but in contrast
to most a
/b hydrolases it also has a sensitive and accessible cys34
providing the possibility of inhibition and fluorescence labeling
with a thiol-reactive dye.35
Scheme 2. Synthesis of 2-TG and 2-ATG.

1944
J. E. Casida et al. / Bioorg. Med. Chem. 18 (2010) 1942–1947
Figure 3. Typical inhibition curves for IDFP and CPO with hMAGL.
Figure 2. Fluorimetric (MMBC) and colorimetric (DTNB) assays of 2-ATG hydrolysis
by hMAGL (A and B) and human brain membranes (C). Background activity is
shown as empty vector (EV) control for hMAGL with MMBC (A), no protein for
hMAGL with DTNB (B), and chlorpyrifos oxon (CPO) (100
with MMBC.
lM) for brain membranes
because, under the workup conditions for the reduction, partial
TBS-ether cleavage occurred.
All that remained was to arachidonoylate 2-TG (Scheme 2). It
was envisioned that 2-TG addition into arachidonoyl chloride
would proceed with selective S-acylation under carefully selected
conditions. Although it was anticipated that the separation of 2-
ATG and AA might be difficult, an excess of arachidonoyl chloride
was initially used to avoid possible unreacted 2-TG as an impurity.
However, under this condition a significant impurity resulted from
an additional O-arachidonoylation. Because this over-arachidonoy-
lated impurity was observed even when an excess of 2-TG was
used, the relative solubility of 2-TG was considered as the cause.
The rate of dissolution of the thiol is likely slower than over acyl-
ation of 2-ATG, which would explain the observation of ester peaks
in the spectra of the crude reaction product. 2-TG is only sparingly
soluble in chloroform, dichloromethane, and diethyl ether and is
insoluble in benzene and toluene but has reasonable solubility in
THF. Indeed, optimized reaction conditions for the formation of
2-ATG employed THF as the solvent and a two-fold excess of 2-
TG which was removed by a water wash in the workup enabling
product isolation without identifiable impurities.
Figure 4. Correlation of IC50 values for hMAGL assayed with 2-ATG (this study) to
published values for mouse brain membrane MAG hydrolase assayed with 1-OG or
2-AG. Inhibitor structures and IC50 values are given in Supplementary data.
nate (OSF) and one carbamate (JZL184) proposed to phosphorylate,
sulfonylate and carbamoylate, respectively, serine in the MAGL ac-
tive site.23,27,28,31 They range from the highly potent IDFP to the
moderately potent CPO and JZL184 to the low potency DFP with
IC50 values of 0.07–0.76, 8–34 and 27,000–45,000 nM, respectively.
The inhibitor specificity was found to be the same for hMAGL as-
sayed with 2-ATG compared with published IC50 data27,36 for the
same set of compounds with mouse brain membrane enzyme
sources assayed by different procedures (Supplementary data).
Importantly, despite differences in procedures and investigators
there is a close agreement between the assay methods in the inhib-
itor structure–activity relationships for organophosphate, sulfo-
nate and carbamate chemotypes (r2 = 0.93, n = 9) (Fig. 4).
N-Arachidonoyl maleimide (NAM) is a potent inhibitor of
MAGL,34 presumably by derivatizing a critical cysteine in the active
site, and produces cannabinoid effects in vivo.37 We confirm the
potency of NAM relative to N-ethylmaleimide with hMAGL using
the 2-ATG hydrolase assay (Supplementary data) but kinetic
measurements in this system are potentially complicated by the
presence of two maleimide-based derivatizing agents, the inhibitor
NAM and fluorescent reagent MMBC.
1-ATG is available for use as a purified MAGL substrate (Supple-
mentary data) and 1- and 2-oleoylglycerol (1- and 2-OG) are
proposed to be hydrolyzed by the same enzyme in brain mem-
branes.38–40 Direct comparison here of hMAGL with 1-ATG and
2-ATG gave identical hydrolysis rates (Fig. 5). The OP sensitivities
to IDFP and CPO. DPD and paraoxon were also very similar for
1-ATG and 2-ATG hydrolysis by hMAGL and human brain mem-
brane preparations, although brain ATG hydrolase was a little less
sensitive (Supplementary data) perhaps due to other competing
phosphorylation sites. Thus, MAGL hydrolyzes 1-ATG and 2-ATG
equally well and with similar or the same inhibitor specificity
suggesting that the 1- or 2-ester can be used interchangeably for
assays with hMAGL or brain MAG hydrolase. However, this may
2.2. Biological evaluation
2-ATG was found to be an excellent substrate for human recom-
binant MAGL (hMAGL) in microplate assays measuring 2-TG
liberation with the MMBC fluorescent probe (Fig. 2A) or DTNB
colorimetric reagent (Fig. 2B). The fluorescent procedure was about
10-fold more sensitive than the colorimetric assay based on the
level of enzyme activity readily quantitated. The MMBC assay
was also equally applicable for brain membranes on considering
the CPO-sensitive 2-TG liberation as MAGL activity23,36 (Fig. 2C).
Further optimization would be required for DTNB assays using
brain membranes.
Inhibitors were assayed with 10 min preincubation with
hMAGL or brain membranes before substrate addition resulting
in typical inhibition curves for isopropyl dodecylfluorophospho-
nate (IDFP) and CPO shown in Figure 3. To further evaluate the
method, nine inhibitors were used to compare hMAGL assayed
with 2-ATG with earlier data for mouse brain membranes assayed
by a number of procedures (Fig. 4) (Supplementary data). These
inhibitors were seven organophosphorus compounds, one sulfo-

J. E. Casida et al. / Bioorg. Med. Chem. 18 (2010) 1942–1947
1945
than 90% was established by 1H and 13C NMR spectra (Supplemen-
tary data).
3.1.2. 2,2,3,3,9,9,10,10-Octamethyl-4,8-dioxa-3,9-disilaundecan-
6-ol (1)
A white suspension of glycerol (12.3 g, 134 mmol), imidazole
(40.0 mL, 294 mmol), and TBSCl (46.3 mL, 267 mmol) in THF
(134 mL) was stirred for 15 h at rt, then poured into water
(600 mL) and extracted with diethyl ether (5 ꢁ 150 mL). The com-
bined ether extract was washed with brine (100 mL), dried over
magnesium sulfate and concentrated in vacuo to give alcohol 1
(40.7 g, 127 mmol, 95% yield) as a colorless oil that was used with-
out further purification. Rf 0.66 (4:1 hexanes/ethyl acetate); 1H
NMR (600 MHz, CDCl3) d 3.75–3.60 (m, 5H), 2.47 (s, 1H), 0.89 (s,
18H), 0.06 (s, 12H); 13C NMR (150 MHz, CDCl3) d 71.8, 63.4, 25.9,
18.3, ꢂ5.43, ꢂ5.44; IR (film) mmax 3473, 2955, 2930, 2885, 2858,
Figure 5. Comparison of 1-ATG and 2-ATG hydrolysis by hMAGL.
not be the case with peripheral tissues where MAGL is a less
important contributor to 2-AG (and therefore presumably 1-ATG
or 2-ATG) hydrolysis.23,27,41,42
1472, 1256, 1098, 1054, 1006, 837, 777 cmꢂ1
.
The endocannabinoids 2-AG and AEA, lacking a free thiol substi-
tuent, do not interfere with MMBC assays of 2-ATG hydrolysis. This
allows a convenient assay of 2-AG analogs as competitors for the
esteratic site of hMAGL (Supplementary data). At equimolar levels
1-AG and 2-AG inhibit 2-ATG hydrolysis by 29–31% and with 10-
fold molar excess the inhibition for both is 50–54%, that is, once
again the 1- and 2-acylglycerol analogs compete equally well.
The oleoylglycerols compete much less favorably than the arachi-
donoylglycerols but 1- and 2-OG are equally effective. AEA is not
an inhibitor/competitor under these assay conditions.
In conclusion, the new substrate 2-ATG has several advantages
over 1-ATG, 1-OG and 2-OG used earlier for MAGL and MAG hydro-
lase assays (Supplementary data). 2-AG but not 1-AG is the natu-
rally-occurring endocannabinoid so findings with 2-ATG are
more relevant than those with 1-ATG and 1- and 2-ATG are hydro-
lyzed faster than 1- and 2-OG by hMAGL. Importantly, MAGL is the
main 2-AG hydrolase in brain but in peripheral tissues other
hydrolases are of increased importance. Therefore, 2-ATG is pre-
ferred as the closest analog of 2-AG and in contrast to 1-AG, 2-
AG, 1-OG or 2-OG its hydrolysis is easily and quickly assayed by
fluorimetric and colorimetric procedures described here.
3.1.3. 2,2,3,3,9,9,10,10-Octamethyl-4,8-dioxa-3,9-disilaundecan-
6-yl methanesulfonate (2)
A colorless solution of 1 (10.0 g, 31.2 mmol) in dichloromethane
(156 mL) was cooled to 0 °C. Triethylamine (6.58 mL, 46.8 mmol)
was added followed by MsCl (2.92 mL, 37.4 mmol). The solution
was stirred at 0 °C for 30 min, after which TLC indicated the reac-
tion was finished. The dichloromethane was diluted to 400 mL,
and washed with water (100 mL), saturated aqueous cupric sulfate
(100 mL) and brine (100 mL) before drying with magnesium sul-
fate and concentrating to yield 2 (12.1 g, 30.3 mmol, 97% yield)
as a yellow oil, used without additional purification. Alternatively,
it was further purified by column chromatography (8:1 hexanes/
ethyl acetate) to yield a colorless oil. Rf 0.59 (4:1 hexanes/ethyl
acetate); 1H NMR (500 MHz, CDCl3) d 4.59 (p, J = 5.3 Hz, 1H), 3.82
(dd, J = 10.9, 4.5 Hz, 2H), 3.80 (dd, J = 10.9, 5.5 Hz, 2H), 3.06 (s,
3H), 0.89 (s, 18H), 0.07 (s, 6H), 0.07 (s, 6H); 13C NMR (150 MHz,
CDCl3) d 83.6, 62.3, 38.3, 25.8, 18.2, ꢂ5.47, ꢂ5.52; IR (film) mmax
2955, 2930, 2885, 2858, 1472, 1359, 1258, 1179, 1105, 939, 837,
779 cmꢂ1; MS (FAB+), m/z 399 (M+H+), 341 ([MꢂC4H9]+), 303
([MꢂOMs]+); HRMS (CI+) calcd for [C16H39O5SSi2]+: m/z 399.2057,
found: 399.2065.
3. Experimental section
3.1. Synthesis
3.1.4. 1,3-Dihydroxypropan-2-yl methanesulfonate (3)
To a colorless solution of 2 (12 g, 30.1 mmol) in methanol
(60 mL) was added AmberlystÒ 15 (1.8 g). The resulting suspension
was stirred for 6 h at rt, and then filtered and concentrated in va-
cuo to a semisolid. The concentrate was sonicated with diethyl
ether (20 mL) to solidify the product. After the product had settled,
the ether was decanted away, and the crude product was washed
twice more with ether (2 ꢁ 20 mL). The resulting solid was dried
in vacuo to give 3 (4.69 g, 27.5 mmol, 92% yield) as an off-white so-
lid. 1H NMR (500 MHz, CD3OD) d 4.65 (tt, J = 5.9, 4.3 Hz, 1H), 3.81
(dd, J = 12.3, 4.3 Hz, 2H), 3.77 (dd, J = 12.3, 6.0 Hz, 2H), 3.18 (s,
3H); 13C NMR (150 MHz, CDCl3) d 85.7, 62.4, 38.7; IR (KBr) mmax
3302, 3026, 2942, 1736, 1613, 1490, 1340, 1172, 1108, 1093,
1010, 986, 927, 566, 524 cmꢂ1; MS (FAB+), m/z 171 (M+H+); MS
(ESI+), m/z 193 (M+Na+); HRMS (ESI+) calcd for [C5H10O5SNa]+: m/
z 193.0141, found: 193.0141.
3.1.1. General experimental procedures
All air- or moisture-sensitive reactions were conducted in
flame-dried glassware under an atmosphere of nitrogen using
dry, deoxygenated solvents. THF and diethyl ether were distilled
over sodium/benzophenone ketyl; dichloromethane and methanol
were distilled over calcium hydride. TLC was performed using E.
Merck Silica Gel 60 F254 precoated plates (0.25 mm) and visual-
ized by UV and anisaldehyde or ceric ammonium molybdate stain.
Fisher silica gel 240–400 mesh (particle size 0.032–0.063) was
used for flash chromatography. 1H and 13C NMR spectra were re-
corded on a Bruker AV-600 (at 600 MHz and 150 MHz, respec-
tively), Bruker AV-500 (at 500 MHz and 125 MHz, respectively),
or a Bruker DRX-500 (at 500 MHz and 125 MHz, respectively) spec-
trometer for samples in chloroform-d or methanol-d4 at 23 °C.
Chemical shifts were referenced to the residual solvent peak,
which was set at d = 7.26 for 1H NMR and d = 77.0 for 13C NMR,
for CDCl3; and d = 3.35 for 1H NMR and d = 49.3 for 13C NMR, for
CH3OD. IR spectra were recorded on a Nicolet MAGNA-IR 850 spec-
trometer. Low and high resolution mass spectral data were ob-
tained on a VG 70-Se Micromass spectrometer for FAB, a VG
Prospec Micromass spectrometer for EI, or a ThermoFisher Scien-
tific LTQ Orbitrap XL for ESI. For all compounds, a purity of greater
3.1.5. 2-Thiocyanatopropane-1,3-diol (4)
Potassium thiocyanate (4.57 g, 47.0 mmol) was added to a solu-
tion of 3 (2.00 g, 11.8 mmol) in acetonitrile (118 mL) at rt. The
reaction mixture was heated to 90 °C for 24 h, then filtered and
concentrated to give a yellow oil that was used immediately with-
out purification because, upon standing, the product decomposed
to an insoluble polymer, with an odor of hydrogen cyanide. 1H
NMR and 13C NMR were consistent with 4. Due to the instability
of the compound, IR and mass spectral data were not recorded

1946
J. E. Casida et al. / Bioorg. Med. Chem. 18 (2010) 1942–1947
for this intermediate. 1H NMR (600 MHz, CD3OD) d 3.92 (dd,
J = 11.8, 5.7 Hz, 2H), 3.86 (dd, J = 11.8, 6.0 Hz, 2H), 3.46 (p,
J = 5.8 Hz, 1H); 13C NMR (150 MHz, CD3OD) d 62.6, 55.5.
27.2, 25.8, 25.64, 25.63, 25.60, 24.8, 22.6, 14.1; IR (film) mmax
3012, 2927, 2857, 1820, 1753, 1454, 1035 cmꢂ1. The crude arach-
idonoyl chloride was dissolved in THF (5 mL) and added dropwise
to an ice-cold solution of 2-TG (44.8 mg, 0.41 mmol) and triethyl-
3.1.6. 2,2,3,3,9,9,10,10-Octamethyl-6-thiocyanato-4,8-dioxa-
3,9-disilaundecane (5)
amine (58.2 lL, 0.41 mmol) in THF (15 mL). After 1 h, the reaction
solution was diluted with hexanes (20 mL) and filtered through
glass wool. The filtrate was concentrated, redissolved in chloro-
form (10 mL), washed with water, then dried over magnesium sul-
fate and concentrated to give 2-ATG (80 mg, 0.20 mmol, 98% yield)
as a viscous yellow oil. Although several attempts were made (ESI,
EI, FAB), mass spectrometric data could not be acquired on this
compound. 1H NMR (500 MHz, CDCl3) d 5.44–5.30 (m, 8H), 3.95
(dd, J = 11.3, 4.5 Hz, 2H), 3.84 (dd, J = 11.3, 5.8 Hz, 2H), 3.76
(tt, J = 5.8, 4.4 Hz, 1H), 2.85–2.77 (m, 6H), 2.60 (dd, J = 7.8, 7.3 Hz,
2H), 2.46 (br s, 2H), 2.11 (q, J = 7.3 Hz, 2H), 2.05 (q, J = 6.9 Hz, 2H),
1.75 (p, J = 7.4 Hz, 2H), 1.38–1.21 (m, 6H), 0.88 (t, J = 7.0 Hz, 3H);
13C NMR (125 MHz, CDCl3) d 198.8, 130.5, 129.1, 128.59, 128.57,
128.3, 128.0, 127.8, 127.5, 63.8, 47.3, 43.7, 31.5, 29.3, 27.2, 26.3,
To a yellow solution of 4 (1.56 g, 11.75 mmol) in THF (117 mL)
was added TBSCl (4.07 mL, 23.5 mmol) followed by imidazole
(3.52 mL, 25.8 mmol). The reaction mixture was stirred at rt for
15 h, and then it was poured over water (200 mL) and extracted
with diethyl ether (4 ꢁ 100 mL). The combined ether extract was
washed with brine (25 mL), dried over magnesium sulfate and con-
centrated in vacuo. The crude yellow oil was purified by column
chromatography (200 mL silica, 8:1 hexanes/ethyl acetate) to yield
5 (2.59 g, 7.15 mmol, 61% yield over two steps) as a colorless oil. Rf
0.46 (9:1 hexanes/ethyl acetate); 1H NMR (600 MHz, CDCl3) d 3.91
(dd, J = 10.5, 4.9 Hz, 2H), 3.88 (dd, J = 10.5, 6.1 Hz, 2H), 3.40 (tt,
J = 5.9, 5.0 Hz, 1H), 0.90 (s, 18H), 0.08 (s, 12H); 13C NMR
(150 MHz, CDCl3) d 112.2, 61.6, 53.5, 25.7, 18.2, ꢂ5.5, ꢂ5.6; IR
(film) mmax 2955, 2930, 2885, 2858, 2156, 1471, 1390, 1362,
1305, 1256, 1120, 1028, 1006, 838, 778, 694, 669 cm-1; MS
(FAB+), m/z 362 (M+H+); HRMS (FAB+) calcd for [C16H36NO2SSi2]+:
m/z 362.2005, found: 362.2000.
25.62, 25.61 (2C), 25.3, 22.6, 14.1; IR (film) mmax 3368, 3012, 2955,
2930, 2857, 1691, 1680, 1454, 1074, 1028, 984 cmꢂ1
.
3.2. MAGL and 2-ATG hydrolase assays
The hMAGL construct was generated by PCR with primers 50-
GCTCTCGAGGCCGCCATGCCAGAGGAAAGTTCC-30 and 50-AGCTGAA
TTCTCAGGGTGGGGACGCAGTTCCTG-30. PCR products were sub-
cloned into the pMSCVpuro vector (Clontech) by using XhoI and
EcoRI restriction sites and generating retrovirus using the Ampho-
Pack-293 Cell Line (Clontech). hMAGL overexpression MUM2C
cells and the corresponding empty vector (EV) control cells were
provided by Daniel Nomura and Benjamin Cravatt (Scripps Re-
search Institute, La Jolla, CA). The cells were cultured to 80% conflu-
ence by Ann Fischer (Department of Molecular and Cell Biology,
University of California, Berkeley) in standard RPMI 10% fetal calf
serum, glutamine medium, washed with 1X phosphate buffered
saline, pH 7.4, harvested in 50 mM Tris, pH 7.4 and stored at
ꢂ80 °C. Homogenates were prepared of the MUM2C cells in
50 mM Tris pH 7.4 and human brain (20% w/v) in 50 mM Tris,
0.2 mM EDTA pH 7.4 at 4 °C. Then the cell homogenate or brain
supernatant fraction (10,000 g for 10 min) was sonicated and cen-
trifuged at 100,000 g for 60 min to obtain the membrane fraction
used for protein determination43 and ATG hydrolysis assays.
3.1.7. 2-Thioglycerol (2-TG)
To an ice-cooled solution of 5 (8.6 g, 23.8 mmol) in THF (238 mL)
was added lithium aluminum hydride (0.81 g, 23.8 mmol). The reac-
tion mixture was stirred at rt overnight, then treated sequentially
with water (0.8 mL), 2 M sodium hydroxide (1.6 mL), and water
(2.4 mL). The reaction mixture was stirred 1 h, at which point a
white precipitate had formed, then acetic acid (4 mL) was added
and the mixture was stirred 2 h, filtered through Celite and concen-
trated under high vacuum, removing all remaining water and acetic
acid. The crude product (a mixture of 6, 2-TG, and the mono-TBS
ether) was dissolved in methanol (300 mL) and stirred, then Amber-
lystÒ 15 beads (2 g) were added. After 24 h, the reaction mixture was
filtered through Celite and concentrated. The concentrate was sus-
pended in ethyl acetate (30 mL) and filtered. The ethyl acetate was
removed by rotary evaporation, and dichloromethane (15 mL) was
added to the viscous opaque oil. The resulting biphasic mixture
was cooled to ꢂ78 °C, and the dichloromethane decanted away from
a sticky semisolid. After warming to room temperature, the sticky
solid had become a viscous oil and the residual solvent was removed
under high vacuum. 2-TG (1.50 g, 13.9 mmol, 58% yield over two
steps) was recovered as a viscous colorless oil, with an odor charac-
teristic of a thiol. It was also possible to purify 2-TG by distillation
under high vacuum, but on small scale, the above procedure was
more convenient and higher yielding. 1H NMR (600 MHz, CDCl3) d
3.84 (dd, J = 11.2, 5.2 Hz, 2H), 3.77 (dd, J = 11.2, 6.0 Hz, 2H), 3.06
(dtt, J = 10.0, 6.0, 5.2 Hz, 1H), 2.24 (s, 2H), 1.50 (d, J = 10.0 Hz, 1H);
13C NMR (150 MHz, CDCl3) d 66.0, 43.8; IR (film) mmax 3356, 2928,
2877, 1643, 1462, 1077, 1029 cmꢂ1; MS (EI+), m/z 108 (M+), 90
([MꢂH2O]+); HRMS (EI+) calcd for [C3H8O2S]+: m/z 108.0245, found:
108.0242.
Enzyme activities and their inhibition were assayed with 20 lg
protein43 in 50 mM Tris pH 7.4 for hMAGL or in Tris pH 7.4 with
0.2 mM EDTA for brain membrane 2-ATG hydrolase in a final vol-
ume of 200
were added in dimethyl sulfoxide (DMSO) (1
for 10 min prior to addition of the substrate (1-ATG or 2-ATG,
L of 10 mM stock solution in DMSO) (50 M final concentra-
tion). In fluorimetric determinations, after 10 min for substrate
hydrolysis, MMBC was added in DMSO (2 L of 5 mM stock solu-
tion) (50 M final concentration). Liberated 2-TG was measured
by adding 150 L of the final reaction mixture to an opaque micro-
l
L at 25 °C. For inhibition assays, the test compounds
lL) and incubated
1
l
l
l
l
l
plate for fluorescence measurements of the 2-TG-MMBC derivative
with excitation at 379 nm and readings at 513 nm using the Spec-
traMax M2 microplate reader (Molecular Devices, Sunnyvale, CA).
This standard procedure was modified to monitor the rate of
hydrolysis by taking aliquots at 0–120 min from incubation mix-
tures on a larger scale and adding MMBC to determine the 2-TG-
MMBC derivative as above. Colorimetric assays followed the same
procedure as fluorimetric determinations with the following
3.1.8. S-Arachidonoyl-2-thioglycerol (2-ATG)
To a solution of arachidonic acid (63 mg, 0.21 mmol) in anhy-
drous dichloromethane (10 mL) was added oxalyl chloride (36 lL,
0.41 mmol) in dichloromethane (1.1 mL) followed by N,N-dimeth-
ylformamide (1 mg) in dichloromethane (0.02 mL). The solution
was stirred for 1 h at rt, and concentrated to give crude arachido-
noyl chloride. 1H NMR (600 MHz, CDCl3) d 5.48–5.30 (m, 8H),
2.89 (t, J = 7.3 Hz, 2H), 2.86–2.79 (m, 6H), 2.15 (q, J = 7.4 Hz, 2H),
2.06 (q, J = 7.2 Hz, 2H), 1.79 (p, J = 7.3 Hz, 2H), 1.38–1.25 (m, 6H),
0.88 (t, J = 7.0 Hz, 3H); 13C NMR (150 MHz, CDCl3) d 173.7, 130.5,
129.7, 128.6, 128.4, 127.9, 127.8, 127.7, 127.5, 46.4, 31.5, 29.3,
exceptions: DTNB at 10 mM stock and 100
lM final concentrations
replaced MMBC at 5 mM and 50 M, respectively; in monitoring
l
the rate of hydrolysis the absorbance was recorded continuously
throughout the incubation period; a clear microplate was used
for absorption measurements of the 2-TG-DTNB derivative at

J. E. Casida et al. / Bioorg. Med. Chem. 18 (2010) 1942–1947
1947
11. Di Marzo, V.; Matias, I. Nat. Neurosci. 2005, 8, 585.
12. Vastag, B. J. JAMA 2003, 289, 1763.
13. Martin, B. R.; Lichtman, A. H. Neurobiol. Dis. 1998, 5, 447.
14. Hohmann, A. G.; Suplita, R. L.; Bolton, N. M.; Neely, M. H.; Fegley, D.; Mangieri,
R.; Krey, J. F.; Walker, J. M.; Holmes, P. V.; Crystal, J. D.; Duranti, A.; Tontini, A.;
Mor, M.; Tarzia, G.; Piomelli, D. Nature 2005, 435, 1108.
15. Buchweitz, J. P.; Karmaus, P. W.; Williams, K. J.; Harkema, J. R.; Kaminski, N. E. J.
Leukoc. Biol. 2008, 83, 785.
412 nm with the VersaMax microplate reader (Molecular Devices).
In both the MMBC and DTNB assays there was 25–30% 2-ATG sub-
strate hydrolysis at 10 min with the standard 20 lg protein level
for hMAGL and brain membranes. Results are given as the mean
of two experiments with different enzyme preparations or as the
mean standard deviation of three or more replicates in the same
experiment.
16. Papa, S. M. Exp. Neurol. 2008, 211,
334.
17. Koppel, J.; Davies, P. J. Alzheimers Dis. 2008, 15, 495.
18. Pazos, M. R.; Sagredo, O.; Fernandez-Ruiz, J. Curr. Pharm. Des. 2008, 14, 2317.
19. Croxford, J. L. CNS Drugs 2003, 17, 179.
Acknowledgments
20. Guzmán, M. Nat. Rev. Cancer 2003, 3, 745.
21. Casida, J. E.; Quistad, G. B. Chem. Res. Toxicol. 2004, 17, 983.
22. Saario, S. M.; Laitinen, J. T. Basic Clin. Pharmacol. Toxicol. 2007, 101, 287.
23. Nomura, D. K.; Blankman, J. L.; Simon, G. M.; Fujioka, K.; Issa, R. S.; Ward, A. M.;
Cravatt, B. F.; Casida, J. E. Nat. Chem. Biol. 2008, 4, 373.
24. Labar, G.; Bauvois, C.; Borel, F.; Ferrer, J.-L.; Johan, W.; Lambert, D. M.
ChemBioChem 2009. doi:10.1002/cbic.200900621.
25. Bertrand, T.; Augé, F.; Houtmann, J.; Rak, A.; Vallée, F.; Mikol, V.; Berne, P. F.;
Michot, N.; Cheuret, D.; Hoornaert, C.; Mathieu, M. J. Mol. Biol. 2009.
doi:10.1016/j.jmb.2009.11.060.
26. Evans, M. J.; Cravatt, B. F. Chem. Rev. 2006, 106, 3279.
27. Blankman, J. L.; Simon, G. M.; Cravatt, B. F. Chem. Biol. 2007, 14, 1347.
28. Long, J. Z.; Li, W.; Booker, L.; Burston, J. J.; Kinsey, S. G.; Schlosburg, J. E.; Pavón,
F. J.; Serrano, A. M.; Selley, D. E.; Parsons, L. H.; Lichtman, A. H.; Cravatt, B. F.
Nat. Chem. Biol. 2009, 5, 37.
This work was supported in part by the William Muriece Hos-
kins Chair in Chemical and Molecular Entomology (J.E.C. and
A.G.G.), a DuPont Young Professor Award (R.S.) and a Chevron Fel-
lowship (E.M.B.). We acknowledge our laboratory colleagues Jeff
Chen, Alan Huang and Ken Juacalla for technical assistance. We
thank Astar Winoto (Department of Molecular and Cell Biology,
University of California, Berkeley) for the use of the fluorescent
microplate reader.
Supplementary data
29. Yang, J. R.; Langmuir, M. E. J. Heterocycl. Chem. 1991, 28, 1177.
30. Ellman, G. Arch. Biochem. Biophys. 1959, 82, 70.
31. Aldridge, W. N.; Reiner, E. Enzyme Inhibitors as Substrates: Interactions of
Esterases with Esters of Organophosphorus and Carbamic Acids; North Holland:
Amsterdam, 1972. 328p.
32. Cholinesterases and Cholinesterase Inhibitors; Giacobini, E., Ed.; Martin Dunitz:
London, 2000. 270p.
33. Koelle, G. B. In Cholinesterases and Anticholinesterase Agents; Koelle, G. B., Ed.;
Springer: Berlin, 1963; pp 187–298.
Supplementary data (1H and 13C NMR and IR spectra for inter-
mediates and end products: substrates and analytical methods
for MAGL assays; inhibitor sensitivities of hMAGL and 2-ATG
hydrolase with 1-ATG and 2-ATG as substrates; structures of MAGL
inhibitors studied) associated with this article can be found, in the
online version, at doi:10.1016/j.bmc.2010.01.034.
34. Saario, S. M.; Salo, O. M. H.; Nevalainen, T.; Poso, A.; Laitenen, J. T.; Järvinen, T.;
Niemi, R. Chem. Biol. 2005, 12, 649.
35. Tyagarajan, K.; Pretzer, E.; Wiktorowicz, J. E. Electrophoresis 2003, 24, 2348.
36. Quistad, G. B.; Klintenberg, R.; Caboni, P.; Liang, S. N.; Casida, J. E. Toxicol. Appl.
Pharmacol. 2006, 211, 78.
37. Burston, J. J.; Sim-Selley, L. J.; Harloe, J. P.; Mahadevan, A.; Razdan, R. K.; Selley,
D. E.; Wiley, J. L. J. Pharmacol. Exp. Ther. 2008, 327, 546.
38. Karlsson, M.; Contreras, J. A.; Hellman, U.; Tornqvist, H.; Holm, C. J. Biol. Chem.
1997, 272, 27218.
39. Ghafouri, N.; Tiger, G.; Razdan, R. K.; Mahadevan, A.; Pertwee, R. G.; Martin, B.
R.; Fowler, C. J. Brit. J. Pharmacol. 2004, 143, 774.
40. Saario, S. M.; Savinainen, J. R.; Laitenen, J. T.; Järvinen, T.; Niemi, R. Biochem.
Pharmacol. 2004, 67, 1381.
41. Long, J. Z.; Nomura, D. K.; Cravatt, B. F. Chem. Biol. 2009, 16, 744.
42. Nomura, D. K.; Hudak, C. S.; Ward, A. M.; Burston, J. J.; Issa, R. S.; Fisher, K. J.;
Abood, M. E.; Wiley, J. L.; Lichtman, A. H.; Casida, J. E. Bioorg. Med. Chem. Lett.
2008, 18, 5875.
References and notes
1. Sugiura, T.; Kondo, S.; Sukagawa, A.; Nakane, S.; Shinoda, A.; Itoh, K.;
Yamashita, A.; Waku, K. Biochem. Biophys. Res. Commun. 1995, 215, 89.
2. Devane, W. A.; Hanuš, L.; Breuer, A.; Pertwee, R. G.; Stevenson, L. A.; Griffin, G.;
Gibson, D.; Mandelbaum, A.; Etinger, A.; Mechoulam, R. Science 1992, 258,
1946.
3. Piomelli, D. Nat. Rev. Neurosci. 2003, 4, 873.
4. Dinh, T. P.; Carpenter, D.; Leslie, F. M.; Freund, T. F.; Katona, I.; Sensi, S. L.;
Kathuria, S.; Piomelli, D. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 10819.
5. Dinh, T. P.; Kathuria, S.; Piomelli, D. Mol. Pharmacol. 2004, 66, 1260.
6. Cravatt, B. F.; Demarest, K.; Patricelli, M. P.; Bracey, M. H.; Giang, D. K.; Martin,
B. R.; Lichtman, A. H. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 9371.
7. McKinney, M. K.; Cravatt, B. F. Annu. Rev. Biochem. 2005, 74, 411.
8. DiMarzo, V.; Bifulco, M.; De Petrocellis, L. Nat. Rev. Drug Disc. 2004, 3, 771.
9. Pertwee, R. G. AAPS J. 2005, 7, E625.
43. Bradford, M. M. Anal. Biochem. 1976, 7, 248.
10. Mackie, K. Annu. Rev. Pharmacol. Toxicol. 2006, 46, 101.