Discovery of glycine sulfonamides as dual inhibitors of sn-1-diacylglycerol lipase α and α/β-hydrolase domain 6

Article abstract of DOI:10.1021/jm500681z

sn-1-Diacylglycerol lipase α (DAGL-α) is the main enzyme responsible for the production of the endocannabinoid 2-arachidonoylglycerol in the central nervous system. Glycine sulfonamides have recently been identified by a high throughput screening campaign as a novel class of inhibitors for this enzyme. Here, we report on the first structure-activity relationship study of glycine sulfonamide inhibitors and their brain membrane proteome-wide selectivity on serine hydrolases with activity-based protein profiling (ABPP). We found that (i) DAGL-α tolerates a variety of biaryl substituents, (ii) the sulfonamide is required for inducing a specific orientation of the 2,2-dimethylchroman substituent, and (iii) a carboxylic acid is essential for its activity. ABPP revealed that the sulfonamide glycine inhibitors have at least three off-targets, including α/β-hydrolase domain 6 (ABHD6). Finally, we identified LEI-106 as a potent, dual DAGL-α/ABHD6 inhibitor, which makes this compound a potential lead for the discovery of new molecular therapies for diet-induced obesity and metabolic syndrome.

Full text of DOI:10.1021/jm500681z

Subscriber access provided by DALHOUSIE UNIV
Article
Discovery of Glycine Sulfonamides as Dual Inhibitors of
Sn-1 Diacylglycerol Lipase-# and #-#-Hydrolase Domain 6
Freek J. Janssen, Hui Deng, Marc P Baggelaar, Marco Allarà, Tom van der
Wel, Hans den Dulk, Alessia Ligresti, Annelot C. M. van der Esbroeck, Ross
McGuire, Vincenzo Di Marzo, Herman S. Overkleeft, and Mario van der Stelt
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm500681z • Publication Date (Web): 02 Jul 2014
Downloaded from http://pubs.acs.org on July 9, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Discovery of Glycine Sulfonamides as Dual
Inhibitors of Sn-1 Diacylglycerol Lipase-α
and α-β-Hydrolase Domain 6
1
1
1
2
1
Freek J. Janssen , Hui Deng , Marc P. Baggelaar , Marco Allarà , Tom van der Wel , Hans den
1
2
1
3
2
Dulk , Alessia Ligresti , Annelot C.M. van Esbroeck , Ross McGuire , Vincenzo Di Marzo ,
Herman S. Overkleeft & Mario van der Stelt
1
1*
1
Department of Bio-organic Synthesis, Leiden Institute of Chemistry, Leiden University, Leiden,
The Netherlands
2
Endocannabinoid Research Group, Pozzuoli, Italy
3
Bioaxis Research, Berghem, The Netherlands
*
Corresponding author
ABSTRACT
Sn-1 diacylglycerol lipase-α (DAGL-α) is the main enzyme responsible for the production of the
endocannabinoid 2-arachidonoylglycerol in the central nervous system. Glycine sulfonamides
have recently been identified by a high throughput screenings campaign as a novel class of
inhibitors for this enzyme. Here, we report on the first structure-activity relationship study of
glycine sulfonamide inhibitors and their brain membrane proteome-wide selectivity on serine
hydrolases with activity-based protein profiling (ABPP). We found that i) DAGL-α tolerates a
variety of biaryl substituents, ii) the sulfonamide is required for inducing a specific orientation of
the 2,2-dimethylchroman substituent and iii) a carboxylic acid is essential for its activity. ABPP
revealed that the sulfonamide glycine inhibitors have at least three off-targets, including α-β-
hydrolase domain 6 (ABHD6). Finally, we identified LEI-106 as a potent, dual DAGL-
1

α/ABHD6 inhibitor, which makes this compound a potential lead for the discovery of new
molecular therapies for diet-induced obesity and metabolic syndrome.
INTRODUCTION
Diacylglycerol lipases (DAGL) are responsible for the synthesis of the endogenous signaling
1
lipid 2-arachidonoylglycerol (2-AG) in the brain and immune system. 2-AG activates the
9
cannabinoid CB1 receptor, which is also the target of Δ -tetrahydrocannabinol, the main
psychoactive compound in marijuana. 2-AG functions as a retrograde messenger and modulates
synaptic plasticity at both GABAergic and glutamatergic synapses by inhibiting neurotransmitter
release in various regions of the central nervous system. It is involved in a broad array of
neurophysiological functions, such as adult neurogenesis, modulation of reward-related behavior,
2
, 3
memory formation, pain sensation and appetite. Activation of cannabinoid CB1 receptors via
4
2
-AG has been implicated in various disorders, including obesity and nicotine addiction. In
addition, 2-AG is a precursor for pro-inflammatory prostaglandins in animal models of
4
-7
neuroinflammation. As a consequence inhibitors of 2-AG biosynthesis may become useful for
8
, 9
therapeutic intervention in various human diseases.
Diacylglycerol lipases are intracellular, multi-domain integral membrane proteins that
produce 2-AG. Two homologous isoforms are known: sn-1 diacylglycerol lipase-α (DAGL-α)
and sn-1 diacylglycerol lipase-β (DAGL-β). The DAG lipases share extensive homology, but
differ in size: ~120 and ~70 kD for DAGL-α and DAGL-β respectively. DAG lipases employ the
typical Ser-His-Asp catalytic triad to specifically hydrolyze the ester bond of sn-1 acyl chains
from arachidonate-containing diacylglycerols. Studies with DAGL-KO mice have shown that
DAGL-α controls to a large extent the formation of 2-AG in the central nervous system, whereas
1
0
DAGL-β appears to partake in 2-AG production in the periphery during inflammation.
2

Selective chemical tools (i.e. inhibitors and activity-based probes) for DAGL-α and DAGL-β
can be used to obtain a better understanding of the physiological role of 2-AG. Such chemical
tools provide a critical counterpart of the DAGL-KO mice and allow the examination of acute
versus congenital inhibition. Currently, several different classes of DAGL-inhibitors have been
1
1
12-14
10, 15
reported. These are a) α-ketoheterocycles , b) fluorophosphonates
, c) beta-lactones
, d)
1
0
bis-oximino-carbamates , and e) 1,2,3-triazole ureas. The latter have recently been described as
non-brain-penetrable subtype selective DAGL-β inhibitors, which cross-react with α-β-hydrolase
1
6
domain 6 (ABHD6). This is an enzyme responsible for 2-AG hydrolysis and is involved in
1
7
obesity-related metabolic disorders. Most of the reported inhibitors are lipid-based molecules
with chemically reactive warheads that appear not to be sufficiently efficacious and selective
over other lipases and/or lack in vivo activity on DAGL. Thus, there is a need for new, subtype
selective DAGL-inhibitors, which can be used to study the function of DAGL-α and DAGL-β in
animal models of disease.
DAGLs belong to the class of serine hydrolases, which contains more than 200 members with
1
8
various physiological functions. In order to gain insight in the biological role of a specific
serine hydrolase, it is important to know the selectivity of the chemical tool that is used to study
the role of a particular enzyme. Activity-based protein profiling (ABPP) is often used to
characterize the selectivity and activity of serine hydrolase inhibitors in proteome wide screens
1
9
using tagged fluorophosphonates (FP)- as broad-spectrum activity-based probes. However,
DAGL-α does not react with such FP-probes. To overcome this shortcoming, we recently
developed a bodipy tagged tetrahydrolipstatin based β-lactone probe 1; 3-(5,5-difluoro-7-(4-(3-
(4-(4-((2S,3S)-2-((S)-2-((formyl-L-leucyl)oxy)tridecyl)-4-oxooxetan-3-yl)butyl)-1H-1,2,3-
4
4
-
triazol-1-yl)propoxy)phenyl)-1,3-dimethyl-5H-4λ ,5λ -dipyrrolo[1,2-c:2',1'-f][1,3,2]diazaborinin
3

1
1
2
-yl)propanoic acid (MB064), which is able to detect endogenous DAGL-α expressed in brain.
We used probe in conjunction with commercially available FP-probe,
1
a
carboxytetramethylrhodamine fluorophosphonate (TAMRA-FP), to assess the activity and
selectivity of novel DAGL-α inhibitors. For example, we discovered that α-ketoheterocycles are
a potent class of DAGL-α inhibitors and detected fatty acid amide hydrolase as the only off-
target of 1-(oxazolo[4,5-b]pyridin-2-yl)-6-phenylhexan-1-one (LEI-104) in mouse brain using
our competetive ABPP setting. This demonstrated the power of comparative ABPP to rapidly
11
identify and characterize novel hits.
Recently, researchers from Bristol-Myers Squibb disclosed a new reversible class of DAGL-
2
0, 21
α inhibitors.
glycine sulfonamide series as an initial lead for chemical optimization. Compounds 2a-c (figure
) were shown to be active on recombinant human DAGL and were selective over pancreatic
They performed a high throughput screening campaign and prioritized the
1
lipase and monoacylglycerol (MAG) lipase, the main enzyme in the central nervous system that
hydrolyzes 2-AG. Due to the lack of a obvious serine hydrolase warhead, we wanted to
investigate the structure-activity relationships of this novel chemotype, using 2c as a starting
point. In addition, a broader selectivity profile over the serine hydrolase family in native brain
proteome was not available.
4

2
0, 21
Figure 1. Glycine sulfonamides as reversible class of DAGL-α inhibitors.
Compounds 2a
(
IC = 0.5 µM), 2b (IC = 0.05 µM) and 2 c (IC = 0.02 µM) were initial leads for chemical
5
0
50
50
20, 21
optimization.
Here, we report the first structure-activity relationship study of glycine sulfonamide DAGL-
α inhibitors and assess their selectivity profile over brain membrane proteome using comparative
ABPP with 1 and TAMRA-FP. Chemical optimization of the sulfonamide series led to the
discovery of 15, which is a potent DAGL-α inhibitor. We confirmed that glycine sulfonamides
are selective over MAGL, but ABPP assays revealed that 15 is also a potent ABHD6 inhibitor.
Since ABHD6, a general lysophospholipid hydrolase, is strongly linked to high-fat-diet induced
1
7
obesity, hepatic steatosis and insulin resistance , this dual inhibition profile may be beneficial
for the treatment of obesity-related disorders.
RESULTS
To systematically investigate the structure-activity relationship of glycine sulfonamide series, we
developed a general synthetic scheme that allowed us to rapidly vary different parts of the
molecule. We used compound 2c, with a 2,2-dimethylchroman substituent as a starting point,
5

2
1
because this was one of the most potent inhibitors reported by BMS . First, we replaced the
biphenyl substituent by biarylethers (13-17), which are other well-known fatty acid tail bio-
isosters used in clinical candidates (e.g. N-Pyridazin-3-yl-4-(3-{[5-(trifluoromethyl)pyridin-2-
2
2
yl]oxy}benzylidene)piperidine-1-carboxamide, PF-04457845). To probe the size of the binding
site we synthesized analogs with the ether substituent at ortho (13), meta (14, 16) and para (15,
1
7) position of the first phenyl ring. Second, to reduce the polar surface area of the molecules,
we replaced the sulfonamide by an amide (19 vs 23; 20 vs 22 and 24 vs 25). Of note, we used
B
chlorophenyl substituents at R , because they are metabolically stable analogs of the tolyl
substituent in compound 2b and are commercially available as both a sulfonoyl chlorides and
acyl chlorides. Finally, since there is no obvious serine hydrolase warhead present in the starting
molecule 2c, we also determined the contribution of the carboxylic acid to its activity by
replacing it by ester (26), alcohol (27), carbamate (28) amide (29, 30), nitrile (32), amine (33)
and tetrazole (34) functionalities. To this end, we performed a reductive amination of glycine
methyl ester with benzaldehydes 3-7, followed by a (sulfon)amide coupling using different acyl
chlorides (4-chlorobenzoyl- and 2,6-dichlorobenzoyl chloride) and sulfonyl chlorides (4-
chlorophenyl-, 2,6-dichlorophenyl- or 2,2-dimethylchroman-6-sulfonyl chloride). Hydrolysis of
the glycine ester yielded final compounds 13-25.
6

Scheme 1. Syntheses of glycine sulfonamide derivatives 13-25.
a) Methyl glycinate, AcOH, NaCNBH , MeOH, 3Å molecular sieves. rt, 27-80%, 8: ortho, X =
3
CH, R = H, 9: meta X = CH, R = H, 10: para, X = CH, R = H, 11: meta X = N, R = CF , 12:
1
1
1
1
3
para X = N, R = CF
;
b) acylchloride or sulfonylchloride, Et N, DMAP or pyridine (cat.),
1
3
3
CH Cl , 0°C to rt. 25-94%; c) NaOH, THF, H O, rt, 33-92%, for the structures of final
2
2
2
compounds 13-25 see table 1.
To derivatize the carboxylic acid moiety we reduced methyl esters 14a and 26 and performed a
subsequent CDI coupling to yield carbamate 28 (see scheme 2A). Hydrolysis of methyl ester 20
followed by EDC coupling and deprotection provided amide 29. In addition, we performed
reductive aminations using 2-aminoacetamide and aldehydes 4-5, followed by sulfonylchloride
coupling to 30 and 31. Dehydration to nitriles 32 and 33a followed by reduction or cyclization
yielded amine 33 and tetrazole 34 respectively (see scheme 2B). All compounds were tested in a
colorimetric, biochemical assay using para-nitrophenylbutyrate as substrate and HEK293
1
1, 23
membranes overexpressing human DAGL-α, as previously described.
7

Scheme 2. Synthetic approach towards glycine sulfonamide derivatives 26-34.
a) LiBH , THF, -5°C, 65%; b) CDI, DIPEA, then piperidine, CH Cl , 0°C to rt, quant; c) NaOH,
4
2
2
THF, H O, rt, 90%; d) EDC, DIPEA, MeCN, then tert-butyl (2-(2-aminoethoxy)ethyl)carbamate,
2
rt, 21%; e) TFA, CH Cl , rt, 96%. f) 2-aminoacetamide, AcOH, NaCNBH , MeOH, 3Å
2
2
3
molecular sieves. rt, 50% (meta, 30a), 68% (para, 31a); g) 2,2-dimethylchroman-6-sulfonyl
chloride, Et N, pyridine (cat.), CH Cl , 0°C to rt, 42% (meta, 30), 63% (para, 31); h) pyridine,
3
2
2
TFAA, THF, 0°C to rt, 69% (meta, 32), 62% (para, 33a); i) borane THF complex, THF, 0°C,
4%; j) Et N.HCl, NaN , toluene, reflux, 16%;
5
3
3
8

DAGL-α seems to have a large binding cavity, because ortho (13), meta (14, 16) and para (15,
A
1
7) substitution of the benzylamine moiety with arylethers was allowed at R (see table 1). Para
substitution was, however, preferred over ortho and meta substitution. The sulfonamides 13-21
and 24 were well tolerated, whereas amide derivatives 22, 23 and 25 proved inactive. Thus, a
B
specific orientation of the R substituent induced by the sulfonamide is required. The weak
activity of compound 19 indicated that the steric tolerance around the sulfonamide functionality
is limited, whereas a para-chlorophenyl group is allowed (compounds 18 and 22). This indicated
B
that both electron withdrawing and donating substituents (2b) are tolerated at R . Thus, the
activity in this subpocket seems to be more governed by lipophilic and steric interactions than
electronic interactions. In contrast, from the activity of compounds 26-34 it was clear that the
carboxylic acid is a crucial pharmacophoric feature, because methylester 26, alcohol 27, primary
amide 33 and tetrazole 34, a carboxylic acid bioisoster, were all inactive. Replacement of the
2
4
acid functionality by carbamate 28, which is a well-known warhead for serine hydrolases ,
resulted in loss of activity. It could be that the steric hinder posed by the piperidine moiety
caused the drop in activity, however, the primary amide (30) was also inactive. Moreover,
conversion of the acid moiety into substituted amide 29, as observed in the original hit 2b, was
detrimental for activity on DAGL-α (IC50 > 10 μM). All together, this suggests that the
carboxylate of the glycine forms an important ionic interaction with a basic amino acid in the
enzyme. Surprisingly, replacement of the carboxylic acid with a primary amine yielded
compound 33 with an IC of 512 ± 147 nM. Compound 15 proved the most potent glycine
50
sulfonamide derivative in the biochemical assay with an IC of 18 ± 6 nM and inhibited the
5
0
1
4
hydrolysis of [ C]-sn-1-oleoyl-2-arachidonoyl-glycerol, the natural substrate of DAGL-α, with a
10
K of 0.7 ± 0.08 μM.
i
9

To explain the observed structure-activity-relationships, we have docked 15 in a previously
reported homology model of DAGL-α (built from the crystal structure of the S146A mutant of
11
Thermomyces (Humicola) lanuginosa lipase in complex with oleic acid, PDB code 1GT6).
2
5
26
Docking was performed using the Yasara implementation of AutodockVina , generating 25
-1
high scoring poses within 1.3 kcal.mol of the calculated binding energy minimum. Two high
ranking binding poses (1 and 2) have been observed that place the carboxylic acid in very close
proximity to the catalytic serine (Ser472) or histidine (His650) residues. While in both poses pi-
cation interactions with the benzylphenoxy pharmacophore are observed, the interactions of the
key carboxylic acid moiety differ (figure 2A,B). In pose 1 a hydrogen bond between the
carboxylic acid with the catalytic serine (Ser472) is found, while in pose 2 the carboxylic acid is
in close proximity of the catalytic histidine (His650) and a neighbouring His471. This may
explain also the weak activity of the amine (compound 33), because a histidine can act both as an
acid and base functionality., The orientations of the two hydrophobic subsitutents differ as well
in the different binding poses. In pose 1 the 2,2-dimethylchroman substituent is in a hydrophobic
pocket lined by residues Leu410, Val526, Ile529 and Leu531 and the phenoxybenzyl group is
directed outside of the catalytic site into a broader cavity (figure 2C). In pose 5 the 2,2-
dimethylchroman substituent is in another hydrophobic pocket that is only accessible through its
perpendicular angle induced by the sulfonamide (figure 2D). In this pose the phenoxybenzyl
group is positioned in the pocket aligned by Leu410, Val526, Ile529 and Leu531. While these
binding modes were among the highest ranking docking poses and are in agreement with our
observed structure-activity relationships, a co-crystal structure of the glycine sulfonamide
inhibitors with DAGL-α is required to prove the proposed binding modes.
1
0

Figure 2. Two highly ranked binding poses of compound 15 in our DAGL-α homology
model. A) Binding mode 1 displaying key interactions with catalytic serine (Ser472) and
histidine (His650) and a neighbouring Tyr303. B) Binding mode 2 displaying key interactions
with the catalytic histidine (His650) and a neighbouring His471 C) Binding mode 1 in the
catalytic binding site cavity. D) Binding mode 2 in the catalytic binding site cavity.
Table 1. Structure-activity relationship overview.
BB
CC
R
R
R
R
N
N
AA
R
R
11

A
B
C
Entry
R
R
R
DAGL-α IC ±
5
0
SEM (nM)
*
O
1
1
1
3
4
5
185 ± 74
56 ± 11
18 ± 6
O
*
*
*
*
*
O
O
O
O
O
OH
OH
OH
OH
OH
O
S
S
S
S
S
*
O
O
O
O
O
O
*
*
O
O
O
O
O
*
O
N
(
LEI-106)
O
*
O
*
*
1
1
6
7
143 ± 60
65 ± 21
F
3
C
C
O
*
F
3
N
O
O
*
O
O
*
*
Cl
O
O
1
1
8
9
91 ± 20
O
*
*
OH
OH
S
*
O
Cl
1600 ± 500
O
S
*
Cl
O
*
*
Cl
O
O
2
2
0
1
66 ± 27
O
*
*
OH
OH
O
O
S
*
O
Cl
695 ± 335
O
S
*
Cl
O
*
*
Cl
O
O
4
2
2
2
3
> 10
*
*
*
OH
OH
O
O
C
O
Cl
4
> 10
*
C
Cl
O
O
F
F
3
C
C
*
*
Cl
Cl
O
O
2
2
4
5
302 ± 135
O
*
*
OH
OH
N
N
O
O
S
*
3
4
> 10
*
C
O
*
*
*
*
*
O
4
2
2
2
2
3
6
7
8
9
0
> 10
*
O
OMe
OH
O
O
O
S
*
O
O
O
O
O
*
4
> 10
O
O
O
O
O
S
S
S
S
*
O
O
4
> 10
*
O
N
O
*
O
4
> 10
*
O
N
H
NH
2
O
O
O
*
O
4
> 10
*
NH
2
O
*
1
2

O
*
*
*
N
4
3
3
3
2
3
4
*
> 10
O
O
O
O
S
S
S
*
O
O
O
*
*
NH
N
2
512 ± 147
O
O
O
*
4
N
> 10
N
N
O
H
*
To determine the activity and selectivity of the glycine sulfonamide series on
endogenously expressed DAGL-α and other serine hydrolases in brain membrane proteome, we
incubated the three most potent inhibitors 14, 15 and 17 for 20 min with mouse brain membrane
homogenates (20 μg protein), followed by labeling of the serine hydrolases by 1 or FP-probe
(figure 3). The non-selective lipase inhibitor, tetrahydrolipstatin (THL) was used as a positive
control. All compounds were able to block the labeling of DAGL-α in the mouse brain proteome
in a concentration dependent manner. Complete blockade was observed at 20 μM of inhibitors
1
4, 15 and 17. The order of potency was similar to that found in the biochemical assay.
Compound 15 inhibited DAGL-α labeling with an IC of 124 ± 13 nM (n = 3), which is ~2.5
5
0
times more potent than compound 14 and ~6 times more potent than compound 17 (figure 3A
and B). No reduction of MAGL (protein bands 11 and 12, figure 3C) labeling at 20 μM was
observed for any of the three compounds, which is in line with the reported lack of activity of
this chemical series in a biochemical assay. However, compounds 14 and 15 partially prevented
labeling of an unknown off-target protein by TAMRA-FP (protein 4 in figure 3C) and all three
compounds reduced labeling of two other proteins (labeled 8 and 9 in figure 3D) by 1. We
1
3

identified protein 9 as ABHD6. Hence, three off-targets for 14 and 15 and two off-targets for 17
were identified in the current setting.
Figure 3: Activity based protein profiling assay with TAMRA-FP and 1. A) Concentration
dependent inhibition of DAGL-α in the mouse brain membrane proteome by 14, 15 and 17. B)
Dose–response curves for DAGL-α inhibition by 14 (black, ---■---, IC = 355 ± 54 nM), 15
5
0
(
grey ▬●▬, IC = 124 ± 13 nM) and 17 (black, ▬▲▬, IC = 743 ± 72 nM) as measured by
50 50
competitive ABPP with ABP 1 (SEM, n = 3). C and D) Competitive ABPP with the DAGL-α
inhibitors 14, 15 and 17 (20 μM) using ABP 1 and carboxytetramethylrhodamine
fluorophosphonate (TAMRA-FP) in mouse brain membrane proteome. The quantified data of the
selectivity screening can be found in the supplementary information.
Since ABHD6 is a hydrolase that is capable of metabolizing 2-AG, the natural product of
2
7
DAGL-α, we set up a natural substrate-based assay for ABHD6. This assay employs ABHD6
to convert 2-AG into glycerol, which is coupled to the oxidation of commercially available
fluorogenic Amplifu™Red via a multi-enzyme cascade (figure 4A). We found that membrane
1
4

preparations of HEK293T cells overexpressing human ABHD6 are able to hydrolyze 2-AG with
an apparent K of 25.4 ± 3.3 μM and a V of 5.94 ± 0.28 nmol/min/mg protein. Pre-incubation
m
max
of the three inhibitors dose-dependently inhibited ABHD6 with a K of 0.8, 2.7 and 18.3 μM for
i
compound 15, 17 and 14, respectively (figure 4C). This indicates that the glycine sulfonamides
are a dual class of inhibitors for DAGL-α and ABHD6.
Figure 4. Biochemical hABHD6 activity assay and dose response analysis. A) Enzymatic
activity of hABHD6 was measured in a natural substrate based activity assay as previously
2
7
described.
B) Normalized and corrected fluorescence measurements over time. C)
Concentration response analysis for ABHD6 inhibition by 14 (black, ---■---, K = 18.3 ± 1.8 μM
i
)
, 15 (grey ▬●▬, K = 0.8 ± 0.1 μM) and 17 (black, ▬▲▬, K = 2.7 ± 0.4 μM) as measured by
i
i
resorufin fluorescence (SEM, n = 2, N = 2).
1
5

DISCUSSION & CONCLUSION
We have reported the first structure-activity-relationship study of non-covalent small molecule
inhibitors of DAGL-α based on the glycine sulfonamide scaffold. We found that i) the enzyme
tolerates a variety of biaryl substituents, ii) the sulfonamide is required for inducing a specific
orientation of a 2,2-dimethylchroman substituent and iii) a carboxylic acid is essential for its
activity. Analysis of potential binding modes of the most potent inhibitor 15 in a homology
model of DAGL-α indicated that the carboxylic acid may interact with the catalytic Ser472 or
His650 and that the two hydrophobic substituents may address different hydrophobic binding
pockets.
In contrast to a previous report, which claimed that glycine sulfonamide were selective
inhibitors for DAGL-α, our proteome-wide ABPP assay indicated that the glycine sulfonamides
do interact with other proteins. We found that compound 15 prevented labeling of at least three
other unknown proteins targeted by 1 and TAMRA-FP. We identified one off-target as ABHD6.
Using a recently developed activity assay for ABHD6 that employs 2-AG as natural substrate,
we discovered that 15 prevented the hydrolysis of 2-AG with a K of 0.8 ± 0.1 μM, which is very
i
similar as for DAGL-α (K = 0.7 ± 0.08 μM) . Thus, compound 15 is a dual inhibitor of DAGL-
i
α/ ABHD6. Of note, it seems that ABHD6 is a common off-target for DAGL-α inhibitors in
1
6
12-14
10, 15
general, because also the triazole ureas , fluorophosphonates
and beta-lactones
cross-
react with this enzyme. To investigate DAGL biology it is necessary to identify compounds that
do not cross-react with ABHD6 and the two other unknown off-targets.
Recently it has been reported that ABHD6 is linked to metabolic disorders and is a
regulator of de novo lipogenesis in the liver by acting as a general lysophospholipid hydrolase
1
6

using lysophosphatidylglycerol, lysophosphatidylethanolamine, lysophosphatic acid and
lysophosphatidyl serine as substrates. Peripheral knock-down of ABHD6 protected mice from
high-fat-diet induced obesity, hepatic steatosis and insulin resistance. In vivo inhibition of
ABHD6 with a small molecule resulted in reduced glucose intolerance and induced weight loss,
thereby providing rationale for the development of small molecule inhibitors of ABHD6 for the
1
7
treatment of obesity-induced metabolic disorders. Since inhibition of DAGL-α is also strongly
1
4
linked with reduced body weight and resistance against high-fat diet induced obesity, via
4
, 28, 29
reduced cannabinoid CB1 receptor activation,
it can be envisioned that dual inhibition of
both ABHD6 and DAGL-α may provide additional therapeutical benefits. Since both DAGL-α
and ABHD6 are highly expressed in the brain, it would be of interest to develop compounds with
restricted access to the central nervous system to prevent adverse neuropsychiatric side effects as
previously observed with the cannabinoid CB1 receptor antagonist Rimonabant. The glycine
sulfonamide series, as exemplified by 15, may be an interesting lead series to provide such tool
2
compounds, because their polar surface area is quite large (> 70 Å ).
In summary, we have provided here the first structure-activity relationship of the glycine
sulfonamide series of DAGL-α inhibitors and discovered that this class of compounds is not
selective over ABHD6. This provided important insights in the direction for optimization of 15,
LEI-106, as a selective DAGL-α tool compound. In addition, this dual DAGL-α/ ABHD6
inhibition profile might provide a new lead for the treatment of diet-induced obesity and
metabolic syndrome.
EXPERIMENTAL SECTION
Experimental procedures computational chemistry
1
1
Docking in DAGL-α homology model. Docking was performed as previously reported.
1
7

Experimental procedures biochemistry
Cloning procedures. Full length human hDAGL-α and hABHD6 cDNA was purchased from
Biosource and cloned into mammalian expression vector pcDNA3.1, containing genes for
ampicillin and neomycin resistance. The empty vector was used as a negative control (mock). All
plasmids were grown in XL-10 Z-competent cells and prepped (Maxi Prep, Qiagen). The
sequences were confirmed by sequence analysis at the Leiden Genome Technology Centre.
Cell culture and membrane preparation. HEK293T cells were grown in DMEM with stable
glutamine and phenolred (PAA or Sigma) with 10% New Born Calf serum, penicillin and
streptomycin. Cells were passaged every 2-3 days by resuspension in medium and seeding to the
appropriate confluence. Membranes were prepared from transiently transfected HEK293T cells.
7
One day prior to transfection 10 cells were seeded in a 15 cm petri dish. Cells were transfected
by the addition of a 3:1 mixture of polyethyleneimine (60µg) and plasmid DNA (20µg) in 2 mL
serum free medium. The medium was refreshed after 24 hours, and after 72 h the cells were
harvested by suspending them in 20 mL medium. The suspension was centrifuged for 10 min at
o
1
000 rpm, and the supernatant was removed. The cell pellet was stored at -80 C until use.
Cell pellets were thawed on ice and suspended in lysis buffer A (20 mM Hepes, pH 7.2, 2 mM
DTT, 0.25 M sucrose, 1 mM MgCl , 1x Cocktail (Roche cOmplete EDTA free), 25U/mL
2
Benzonase). For hABHD6 membrane preperations, the protease inhibitor cocktail was not added.
The suspension was homogenized by polytrone (3 × 7 sec) and incubated for 30 min on ice. The
o
suspension was subjected to ultracentrifugation (93.000 × g, 30 min, 4 C, Beckman Coulter,
Type Ti70 rotor) to yield the cytosolic fraction in the supernatant and the membrane fraction as a
pellet. The pellet was resuspended in lysis buffer B (20 mM Hepes, pH 7.2, 2 mM DTT, 1x
Cocktail (Roche cOmplete EDTA free)). The protein concentration was determined with Quick
1
8

Start Bradford assay (Biorad). The protein fractions were diluted to a total protein concentration
o
of 1 mg/mL and stored in small aliquots at -80 C until use.
Biochemical DAGL-α activity assay. The biochemical hDAGL-α activity assay was performed
1
1
as previously reported.
Biochemical ABHD6 activity assay. The biochemical hABHD6 activity assay is based on the
production of glycerol by 2-arachidonoyl glycerol (2-AG) hydrolysis by membrane preparations
2
7
from HEK293T cells transiently transfected with hABHD6, as previously reported. The
produced glycerol is coupled to the oxidation of commercially available Amplifu™Red via a
multi-enzyme cascade, resulting in a fluorescent signal from the dye resorufin. Assay reagents
and enzymes (glycerol kinase from Cellulomonas sp., glycerol-1-phosphate oxidase from
Streptococcus thermophilus, horse radish peroxidase from Horseradish) were purchased from
Sigma Aldrich.
The assay was performed in HEMNB buffer (50 mM Hepes pH 7.4, 1 mM EDTA, 5 mM MgCl2,
1
00 mM NaCl, 0.5% (w/v) BSA) in black, flat bottom Greiner 96-wells plates. Final protein
concentration of membrane preparations of HEK293T cells overexpressing hABHD6 was 0.04
µg/µL. Inhibitors were added from 40x concentrated DMSO stocks. After 30 min. incubation,
assay mix containing glycerol kinase (GK), glycerol-1-phosphate oxidase (GPO), horse radish
peroxidase (HRP), adenosine triphosphate (ATP), Amplifu™Red and 2-arachidonoyl glycerol
(
2-AG) was added and fluorescence was measured in 5 min. intervals for 75 min. (λ = 535 nm,
ex
λem = 595 nm) at 21°C on a plate reader (TECAN GENios microplate reader). Final assay
concentrations: 0.2 U/mL GK, GPO and HRP, 0.125 mM ATP, 10 µM Amplifu™Red, 25 µM 2-
AG, 5% DMSO, 1% ACN in a final volume of 200 µL. Slopes were determined in the linear
1
9

interval of t = 10 to t = 35 min. All experiments were performed at N = 2, n = 2 for experimental
measurements and N = 2, n = 4 for controls.
1
1
The Z’-factor for each measurement was calculated as previously reported, but with ABHD6
and DMSO as a positive control and with slope values determined in the interval of t = 10 to t =
3
5 min. Plates with Z’≥ 0.6 were accepted for further analysis. Fluorescence values were
corrected for the average fluorescence of the negative control. The average, standard deviation
SD) and standard error of mean (SEM) were calculated and normalized to the corrected positive
control. Data was exported to Graphpad Prism 5.0 for the calculation of the IC using a non-
(
5
0
linear dose-response analysis.
For K and V determination, the concentration of 2-AG was varied and slopes of corrected
M
max
fluorescence in time were exported to GraphPad Prism 5.0 (Michaelis-Menten analysis). The
calculated Vmax value in RFU/min was then converted to a value in nmol converted
glycerol/min/mg protein according to a standard curve with the rate of increase in fluorescence
as a function of converted glycerol.
Preparation of mouse brain membrane proteome. Mouse brains were isolated according to
guidelines approved by the ethical committee of Leiden University (DEC#13191). Mouse brains
were thawed on ice and homogenized by polytrone (3 × 7 sec.) in pH 7.2 lysis buffer A (20 mM
Hepes, 2 mM DTT, 1 mM MgCl , 25 U/mL Benzonase) and incubated for 15 minutes on ice,
2
o
followed by low speed spin (2500 × g, 3 min. at 4 C) to remove debris. The supernatant was
o
subjected to ultracentrifugation (100.000 × g, 45 min. 4 C, Beckman Coulter, Type Ti70 rotor)
to yield the cytosolic fraction in the supernatant and the membrane fraction as a pellet. The pellet
was resuspended in lysis buffer B (20 mM Hepes, 2 mM DTT). The total protein concentration
2
0

was determined with Quick Start Bradford assay (Biorad). Membranes were stored in small
o
aliquots at -80 C until use.
Activity based protein profiling activity assay. IC50 determination of inhibitors against
endogenously expressed DAGL-α in the mouse brain membrane proteome. Inhibitors were
incubated at the indicated concentrations (total volume 20 μL) for 30 min at rt, prior to
incubation with 1 for 10 min at rt. The reaction was quenched with 10 μL standard 3 × SDS page
TM
sample buffer, and resolved on 10 % SDS-page. The gels were scanned with a ChemiDoc MP
sytem (Cy 3 settings; 605/50 filter). The percentage activity remaining was determined by
measuring the integrated optical intensity of the bands using Image Lab 4.1. software. IC50
values were determined from a dose-response curve generated using Prism software (GraphPad).
Activity based protein profiling selectivity assay. For gel based ABPP experiments mouse
brain proteome (2.0 mg/mL, 20 μL) was preincubated for 30 min with vehicle (0.5 μL DMSO) or
inhibitor in 0.5 μL DMSO at rt and subsequently treated with 250 nM (final concentration) ABP
1
or 500 nM (final concentration) TAMRA-FP (from Thermo Fischer Scientific) for 15 minutes
at rt. The reactions were quenched with 10 μL standard 3 × SDS page sample buffer. The
samples were directly loaded and resolved on SDS page gel (10 % acrylamide). The gels were
TM
scanned with a ChemiDoc MP sytem (Cy 3 settings; 605/50 filter) and analyzed using Image
Lab 4.1.
Quantification of protein bands for the determination of selectivity of 14, 15 and 17 The
percentage activity remaining was determined by measuring the integrated optical intensity of
the bands using Image Lab 4.1 software. This activity was corrected for the total protein loading
TM
per lane as determined by coomassie stain and imaging with a ChemiDoc MP sytem, followed
2
1

by determination of the integrated optical intensity per lane by using Image Lab 4.1. The
intensity of the protein bands from the protein samples treated with vehicle was set to 100%.
Experimental procedures chemistry
General remarks
All reactions were performed using oven or flame-dried glassware and dry solvents. Reagents
were purchased from Sigma Aldrich, Acros and Merck and used without further purification
unless noted otherwise. All moisture sensitive reactions were performed under an argon
atmosphere. Traces of water were removed from starting compounds by co-evaporation with
toluene.
1
13
1
H- and C-NMR spectra were recorded on a Bruker AV 400 MHz spectrometer at 400.2 ( H)
1
3
and 100.6 ( C) MHz using CDCl or CD OD as solvent, unless stated otherwise. Chemical shift
3
3
values are reported in ppm with tetramethylsilane or solvent resonance as the internal standard
1
13
1
13
(
CDCl : δ 7.26 for H, δ 77.16 for C, CD OD: δ 3.31 for H, δ 49.00 for C). Data are reported
3
3
as follows: chemical shifts (δ), multiplicity (s = singlet, d = doublet, dd = double doublet, td =
triple doublet, t = triplet, q = quartet, quintet = quint, br = broad, m = multiplet), coupling
constants J (Hz), and integration. HPLC purification was performed on a preparative LC-MS
system (Agilent 1200 series) with an Agilent 6130 Quadruple MS detector. High-resolution mass
spectra (HRMS) were recorded on a Thermo Scientific LTQ Orbitrap XL. Liquid
chromatography was performed on a Finnigan Surveyor LC-MS system, equipped with a C18
column. Melting points were measured on a Stuart SMP30 melting point apparatus. Flash
chromatography was performed using SiliCycle silica gel type SiliaFlash P60 (230 – 400 mesh).
TLC analysis was performed on Merck silica gel 60/Kieselguhr F254, 0.25 mm. Compounds
2
2

were visualized using either Seebach’s reagent (a mixture of phosphomolybdic acid (25 g),
cerium (IV) sulfate (7.5 g), H O (500 mL) and H SO (25 mL)) or a KMnO stain (K CO (40
2
2
4
4
2
3
g), KMnO (6 g) and H O (600 mL)).
4
2
Methyl (2-phenoxybenzyl)glycinate 8
To a strirred solution of 2-phenoxybenzaldehyde (3, 150 µL, 0.87 mmol) in MeOH (50 mL) with
Å MS was added methyl glycinate (377.6 mg, 3.01 mmol) and AcOH (260 µL, 4.54 mmol) at
3
rt, the reaction mixture was stirred for 15 minutes. Lastly was added NaCNBH (151.4 mg, 2.41
3
mmol) and the reaction mixture was stirred for 2 hours at rt. After completion the reaction was
diluted with EtOAc and washed with NaHCO , water and brine, dried, concentrated in vacuo and
3
purified by flash chromatography to yield methyl (2-phenoxybenzyl)glycinate (141.9 mg, 0.52
1
mmol, 60 %) as a white solid. H NMR (CDCl , 400 MHz): δ 7.41 (d, J = 7.3 Hz, 1H), 7.32 (t, J
3
=
7.6 Hz, 2H), 7.22 (t, J = 7.7 Hz, 1H), 7.16 – 7.05 (m, 2H), 6.96 (d, J = 8.0 Hz, 2H), 6.88 (d, J =
1
3
8
.0 Hz, 1H), 3.88 (s, 2H), 3.68 (s, 3H), 3.45 (s, 2H), 1.92 (bs, 1H). C NMR (CDCl , 101 MHz):
3
δ 170.51, 157.54, 155.07, 130.54, 129.89(2C), 128.76, 125.85, 123.91, 123.12, 119.17,
18.24(2C), 54.19, 51.97, 49.87.
1
Methyl (3-phenoxybenzyl)glycinate 9
The title compound was synthesized from 3-phenoxybenzaldehyde (4, 100 µl, 0.58 mmol)
according to the procedures described for compound 8. This yielded methyl (3-
1
phenoxybenzyl)glycinate (125.7 mg, 0.46 mmol, 80 %) as a white solid. H NMR (CDCl , 400
3
MHz): δ 7.39 – 7.27 (m, 3H), 7.13 – 7.04 (m, 2H), 7.04 – 6.97 (m, 3H), 6.90 (ddd, J = 8.2, 2.5,
1
3
1
.0 Hz, 1H), 3.79 (s, 2H), 3.73 (s, 3H), 3.42 (s, 2H). C NMR (CDCl , 101 MHz): δ 170.73,
3
2
3

1
5
57.76, 157.16, 133.93, 129.88(2C), 129.84, 123.36, 123.14, 119.04(2C), 118.73, 117.71, 99.91,
3
0
3.09, 51.74, 50.02. Spectroscopic data are in agreement with those reported in literature.
Methyl (4-phenoxybenzyl)glycinate 10
The title compound was synthesized from 4-phenoxybenzaldehyde (5, 100 mg, 0.50 mmol)
according to the procedures described for compound 8. This yielded methyl (4-
1
phenoxybenzyl)glycinate (240 mg, 0.88 mmol, 44 %) as a white solid. H NMR (400 MHz,
CDCl ): δ 7.38 – 7.22 (m, 4H), 7.08 (t, J = 7.4 Hz, 1H), 6.98 (ddd, J = 10.1, 7.2, 1.5 Hz, 4H),
3
1
3
3
1
4
.77 (s, 2H), 3.72 (s, 3H), 3.42 (s, 2H), 2.18 (s, 1H); C NMR (101 MHz, CDCl ): δ 172.87,
3
57.37, 156.38, 134.38, 129.75(2C), 129.72(2C), 123.18, 118.97(2C), 118.76(2C), 52.68, 51.82,
9.85.
3
-((5-(Trifluoromethyl)pyridin-2-yl)oxy)benzaldehyde 11a.
The title compound was synthesized from 3-hydroxybenzaldehyde (366 mg, 3 mmol) and 2-
chloro-5-(trifluoromethyl)pyridine (363 mg, 2 mmol) according to the procedures described for
compound 12a. This yielded 3-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzaldehyde (135 mg, 0.5
1
mmol, 30%) after flash chromatography as a white solid. H NMR (CDCl 400 MHz): δ 10.02
3
,
(s, 1H), 8.45 – 8.40 (m, 1H), 7.96 (dd, J = 8.6, 2.3 Hz, 2H), 7.77-7.79 (m, 1H), 7.71 – 7.67 (m,
1
H), 7.62 (t, J = 7.8 Hz, 1H), 7.43-7.46 (m, 1H), 7.10 (d, 1H, J = 8.8 Hz). Spectroscopic data are
3
1
in agreement with those reported in literature.
Methyl (3-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzyl)glycinate 11.
The title compound was synthesized from 3-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzaldehyde
(6, 183mg, 0.68mmol) according to the procedures described for compound 8. This yielded
2
4

methyl (3-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzyl)glycinate (84 mg, 0.25 mmol, 37 %) as a
1
colorless oil. H NMR (CDCl , 400 MHz): δ 8.43 (s, 1H), 7.86 – 7.89 (m, 1H), 7.36 – 7.39 (m,
3
1
4
H), 7.14 – 7.23 (m, 2H), 6.98 – 7.01 (m, 2H), 3.86 (s, 2H), 3.72 (s, 3H), 3.45 (s, 2H).
-((5-(Trifluoromethyl)pyridin-2-yl)oxy)benzaldehyde 12a.
To a strirred solution of 4-hydroxybenzaldehyde (366 mg, 3 mmol) in DMSO (30 mL) was
added K CO (360 mg, 6 mmol) at room temperture. After 30 minutes of stirring was added
2
3
dropwise a solution of 2-fluoro-5-(trifluoromethyl)pyridine (0.241 mL, 2 mmol) in DMSO (2
mL). After full completion (24 h) HCl (1 M) was added. The mixture was poured into water, and
the product was extracted with EtOAc, washed with water and brine, dried, concentrated in
vacuo and purified by flash chromatography to yield 4-((5-(trifluoromethyl)pyridin-2-
1
yl)oxy)benzaldehyde (366 mg, 1.37 mmol, 70 %) as a white solid. H NMR (CDCl , 400 MHz):
3
1
3
δ 10.01 (s, 1H), 8.45 (s, 1H), 7.95 – 7.99 (m, 3H), 7.31 – 7.35 (m, 2H), 7.11 – 7.13 (m, 1H). C
NMR (101 MHz, CDCl ): δ 190.94, 164.87, 158.25, 145.49 (q, J = 4.3 Hz), 137.24 (q, J = 3.2
3
Hz), 133.58, 131.69(2C), 123.68 (q, J = 272.6 Hz), 122.50 (q, J = 33.5 Hz), 121.92(2C), 112.24.
Methyl (4-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzyl)glycinate 12.
The title compound was synthesized from 4-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzaldehyde
(7, 267 mg, 1 mmol) according to the procedures described for compound 8. This yielded methyl
(4-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzyl)glycinate (215 mg, 0.63 mmol, 63 %) as a
1
colorless oil. H NMR (CDCl , 400 MHz): δ 8.44 (s, 1H), 7.88-7.91 (m, 1H), 7.40 (d, 2H, J = 8.4
3
Hz), 7.10-7.13(m, 2H), 7.00 (d, 1H, J = 8.8 Hz), 3.83 (s, 2H), 3.75 (s, 3H), 3.46 (s, 2H).
Methyl N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(2-phenoxybenzyl)glycinate 13a
2
5

To a stirred solution of methyl (2-phenoxybenzyl)glycinate (8, 63.9 mg, 0.24 mmol) in CH Cl
2
2
(50 mL) was added 2,2-dimethylchromane-6-sulfonyl chloride (96 mg, 0.368 mmol),
triethylamine (0.1 mL, 0.717 mmol) and 1 drop of pyridine (or 4-dimethylaminopyridine) at rt.
After completion (2 h), the reaction mixture was diluted with EtOAc, washed with sat. NaHCO3,
brine and purified using column chromatography to yield methyl N-((2,2-dimethylchroman-6-
1
yl)sulfonyl)-N-(2-phenoxybenzyl)glycinate (96.0 mg, 0.19 mmol, 82 %) as an oil. H NMR
(
CDCl , 400 MHz): δ 7.57 – 7.46 (m, 3H), 7.33 – 7.27 (m, 2H), 7.25 – 7.19 (m, 1H), 7.14 – 7.04
3
(m, 2H), 6.87 – 6.74 (m, 4H), 4.55 (s, 2H), 4.02 (s, 2H), 3.52 (s, 3H), 2.74 (t, J = 6.7 Hz, 2H),
1
3
1
1
1
.80 (t, J = 6.7 Hz, 2H), 1.34 (s, 6H). C NMR (CDCl , 101 MHz): δ 169.68, 157.96, 156.85,
3
55.46, 131.49, 130.43(2C), 129.95, 129.57, 129.51, 127.15, 126.53, 123.89, 123.52, 121.31,
18.57(2C), 118.26, 117.79, 75.73, 52.06, 48.00, 46.26, 32.38, 27.01(2C), 22.45.
N-((2,2-Dimethylchroman-6-yl)sulfonyl)-N-(2-phenoxybenzyl)glycine 13
To
a
strirred
solution
of
methyl
N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(2-
phenoxybenzyl)glycinate (13a, 60 mg, 0.121 mmol) in THF (1 mL): methanol (1 mL) was
added aqueous NaOH (2 mL, 4.00 mmol) at rt. After completion (30 mins), to the reaction was
added Amberlight 150 H resin until neutral, the mixture was filtered, concentrated in vacuo,
coevaporated with toluene and purified by flash chromatography to yield N-((2,2-
dimethylchroman-6-yl)sulfonyl)-N-(2-phenoxybenzyl)glycine (25.3 mg, 0.053 mmol, 43% yield)
as a white solid. mp 155-157 °C. HRMS (ESI+) m/z: calculated for C H NO S ([M + H]),
2
6
28
6
1
4
2
2
82.1632, found: 482.1631. H NMR (CDCl , 400 MHz): δ 7.72 – 7.35 (m, 3H), 7.33 – 7.27 (m,
3
H), 7.23 (t, J = 7.5 Hz, 1H), 7.17 – 6.98 (m, 2H), 6.94 – 6.65 (m, 4H), 4.54 (s, 2H), 4.08 (s,
13
H), 2.71 (t, J = 6.6 Hz, 2H), 1.78 (t, J = 6.4 Hz, 2H), 1.33 (s, 6H). C NMR (CDCl , 101
3
MHz): δ 173.82, 158.16, 156.62, 155.63, 131.64, 130.46, 130.02(2C), 129.76, 129.62, 127.20,
2
6

1
2
26.06, 123.88, 123.71, 121.49, 118.71(2C), 118.18, 117.85, 75.83, 48.15, 46.90, 32.31,
7.00(2C), 22.42. Purity >95% as determined by LC/MS.
Methyl N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(3-phenoxybenzyl)glycinate 14a
The title compound was synthesized from methyl (3-phenoxybenzyl)glycinate (9, 37.5 mg, 0.14
mmol) according to the procedures described for compound 13a. This yielded methyl N-((2,2-
dimethylchroman-6-yl)sulfonyl)-N-(3-phenoxybenzyl)glycinate (64.5 mg, 0.13 mmol, 94 %) as
1
an oil. H NMR (CDCl 400 MHz) δ 7.62 – 7.51 (m, 2H), 7.33 (t, J = 8.4, 7.5 Hz, 2H), 7.29 –
3
,
7
3
.23 (m, 1H), 7.11 (t, J = 7.4 Hz, 1H), 7.03 – 6.94 (m, 3H), 6.94 – 6.80 (m, 3H), 4.44 (s, 2H),
1
3
.92 (s, 2H), 3.57 (s, 3H), 2.80 (t, J = 6.7 Hz, 2H), 1.83 (t, J = 6.7 Hz, 2H), 1.35 (s, 6H). C
NMR (CDCl 101 MHz) δ 169.45, 158.14, 157.67, 156.99, 137.43, 130.15, 129.96, 129.92(2C),
3,
1
5
29.64, 127.15, 123.58, 123.44, 121.45, 119.06, 119.02(2C), 118.39, 117.95, 75.82, 52.13,
1.25, 46.95, 32.36, 26.99(2C), 22.48.
N-((2,2-Dimethylchroman-6-yl)sulfonyl)-N-(3-phenoxybenzyl)glycine 14
The title compound was synthesized from methyl N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(3-
phenoxybenzyl)glycinate (14a, 52.4 mg, 0.11 mmol) according to the procedures described for
compound
13.
This
yielded
N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(3-
phenoxybenzyl)glycine (27.9 mg, 0.06 mmol, 55 %) as a white solid after lyophilization. mp
1
4
7
71-173 °C. HRMS (ESI+) m/z: calculated for C H NO S ([M + H]), 482.1632, found:
26 28 6
1
82.1630. H NMR (CDCl , 400 MHz): δ 7.70 – 7.52 (m, 2H), 7.33 (t, J = 7.9 Hz, 2H), 7.25 –
3
.29 (m, 1H), 7.15 – 7.06 (m, 1H), 7.05 – 6.87 (m, 4H), 6.87 – 6.77 (m, 2H), 4.42 (s, 2H), 3.94
13
(
s, 2H), 2.78 (t, J = 6.6 Hz, 2H), 1.82 (t, J = 6.7 Hz, 2H), 1.34 (s, 6H). C NMR (CDCl , 101
3
MHz): δ 173.84, 158.34, 157.82, 156.91, 137.14, 130.28, 129.95(2C), 129.67, 129.56, 127.18,
2
7

1
2
23.66, 123.39, 121.63, 119.08(2C), 118.97, 118.51, 118.02, 75.93, 51.40, 46.87, 32.32,
7.00(2C), 22.47. Purity >95% as determined by LC/MS. NMR spectra in supplementary
information.
Methyl N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(4-phenoxybenzyl)glycinate 15a
See compound 26.
N-((2,2-Dimethylchroman-6-yl)sulfonyl)-N-(4-phenoxybenzyl)glycine 15
The title compound was synthesized from methyl N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(4-
phenoxybenzyl)glycinate (15a, 62 mg, 0.125 mmol) according to the procedures described for
compound
13.
This
yielded
N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(4-
phenoxybenzyl)glycine (54 mg, 0.11 mmol, 90%) as a white solid. mp 192-194 °C. HRMS
1
(
ESI+) m/z: calculated for C H NO S ([M + H]) 482.1632, found 482.1631. H NMR (400
2
6
28
6
MHz, DMSO): δ 7.58 (d, J = 2.2 Hz, 1H), 7.52 (dd, J = 8.6, 2.3 Hz, 1H), 7.40 (dd, J = 8.5, 7.5
Hz, 2H), 7.23 (d, J = 8.6 Hz, 2H), 7.15 (t, J = 7.4 Hz, 1H), 7.04 – 6.97 (m, 2H), 6.93 (d, J = 8.6
Hz, 2H), 6.85 (d, J = 8.6 Hz, 1H), 4.37 (s, 2H), 3.84 (s, 2H), 2.80 (t, J = 6.6 Hz, 2H), 1.81 (t, J =
1
3
6
1
1
.7 Hz, 2H), 1.31 (s, 6H); C NMR (101 MHz, DMSO): δ 170.47, 157.70, 157.04, 156.54,
31.36, 130.66, 130.61(2C), 130.51(2C), 129.63, 127.05, 123.94, 122.03, 119.02(2C),
18.89(2C), 117.66, 76.13, 51.14, 48.15, 32.00, 27.02(2C), 22.11. Purity >95% as determined by
LC/MS. NMR spectra in supplementary information.
Methyl
N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(3-((5-(trifluoromethyl)pyridin-2-
yl)oxy)benzyl)glycinate 16a
2
8

The title compound was synthesized from methyl (3-((5-(trifluoromethyl)pyridin-2-
yl)oxy)benzyl)glycinate (11, 34 mg, 0.1 mmol) according to the procedures described for
compound 13a. This yielded methyl N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(3-((5-
(trifluoromethyl)pyridin-2-yl)oxy)benzyl)glycinate (19 mg, 0.03 mmol, 35 %) as a white solid.
1
H NMR (CDCl , 400 MHz): δ 8.34 (s, 1H), 7.82-7.85 (m, 1H), 7.50-7.52 (m, 2H), 7.28-7.32(m,
3
1
H), 7.07-7.09 (m, 1H), 7.00 (d, 2H, J = 8.0 Hz), 6.94 (d, 1H, J = 8.8 Hz), 6.76 (d, 1H, J = 8.4
Hz), 4.43 (s, 2H), 3.89 (s, 2H), 3.62 (s, 3H), 2.73 (t, 2H, J = 6.4 Hz), 1.75 (t, 2H, J = 6.8 Hz),
.26 (s, 6H).
1
N-((2,2-Dimethylchroman-6-yl)sulfonyl)-N-(3-((5-(trifluoromethyl)pyridin-2-
yl)oxy)benzyl)glycine 16
The title compound was synthesized from methyl N-((2,2-dimethylchroman-6-yl)sulfonyl)-N-(3-
((5-(trifluoromethyl)pyridin-2-yl)oxy)benzyl)glycinate (16a, 9.0 mg, 0.016 mmol) according to
the procedures described for compound 13. This yielded N-((2,2-dimethylchroman-6-
yl)sulfonyl)-N-(3-((5-(trifluoromethyl)pyridin-2-yl)oxy)benzyl)glycine (5.6 mg, 0.010 mmol,
6
3% yield) as a white solid after lyophilization. mp 149-151 °C. HRMS (ESI+) calculated for
1
C H F N O S ([M + H]). 551.1458, found: 551.1456. H NMR (CDCl , 400 MHz): δ 8.47 (s,
26
25
3
2
6
3
1
H), 7.92 (d, J = 7.2 Hz, 1H), 7.58 – 7.62 (m, 2H), 7.30 – 7.32 (m, 1H), 7.03 – 7.11 (m, 4H),
.85 (d, J = 8.4 Hz, 1H)，4.47 (s, 2H), 3.96 (s, 2H), 2.82 (t, J = 6.4 Hz, 2H), 1.85 (t, J = 6.8 Hz,
6
1
3
2
4
2
3
H), 1.28 (s, 6H). C NMR (151 MHz, CDCl ): δ 171.96, 165.66, 158.33, 153.54, 145.43(q, J =
3
.1 Hz), 137.51, 137.11(q, J = 3.2 Hz), 130.29, 129.72, 129.60, 127.22, 125.87, 123.73 (q, J =
71.5 Hz), 121.83, 121.65 (q, J = 33.4 Hz), 121.60, 121.46, 118.01, 111.83, 75.91, 51.27, 46.83,
2.32, 26.99(2C), 22.47. Purity >95% as determined by LC/MS.
2
9