Structure activity relationship of 3-nitro-2-(trifluoromethyl)-2H-chromene derivatives as P2Y6 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2021.128008

Various 6-alkynyl analogues of a known 3-nitro-2-(trifluoromethyl)-2H-chromene antagonist 3 of the G_q-coupled P2Y₆ receptor (P2Y₆R) were synthesized using a Sonogashira reaction to replace a 6-iodo group. The analogues were tested in a functional assay consisting of inhibition of calcium mobilization in P2Y₆R-expressing astrocytoma cells elicited by native P2Y₆R agonist UDP. 6-Ethynyl and 6-cyano groups were installed, and the alkynes were extended through both alkyl and aryl spacers. The most potent antagonists, with IC₅₀ of ~1 μM, were found to be trialkylsilyl-ethynyl 7 and 8 (3–5 fold greater affinity than reference 3), t-butyl prop-2-yn-1-ylcarbamate 14 and p-carboxyphenyl-ethynyl 16 derivatives, and 3 and 8 displayed surmountable antagonism of UDP-induced production of inositol phosphates. Other chain-extended terminal carboxylate derivatives were less potent than the corresponding methyl ester derivatives. Thus, the 6 position in this chromene series is suitable for derivatization with flexibility of substitution, even with sterically extended chains, without losing P2Y₆R affinity. However, a 3-carboxylic acid or 3-ester substitution did not serve as a nitro bioisostere, as the affinity was eliminated. These compounds provide additional ligand tools for the underexplored P2Y₆R, which is a target for inflammatory, neurodegenerative and metabolic diseases.

Full text of DOI:10.1016/j.bmcl.2021.128008

Bioorg. Med. Chem. Lett. 41 (2021) 128008
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure activity relationship of 3-nitro-2-(trifluoromethyl)-2H-chromene
derivatives as P2Y6 receptor antagonists
Young-Hwan Jung, Shanu Jain, Varun Gopinatth, Ngan B. Phung, Zhan-Guo Gao,
Kenneth A. Jacobson*
Molecular Recognition Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health,
Bethesda, MD 20892, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Various 6-alkynyl analogues of a known 3-nitro-2-(trifluoromethyl)-2H-chromene antagonist 3 of the Gq-coupled
P2Y6 receptor (P2Y6R) were synthesized using a Sonogashira reaction to replace a 6-iodo group. The analogues
were tested in a functional assay consisting of inhibition of calcium mobilization in P2Y6R-expressing astrocy-
toma cells elicited by native P2Y6R agonist UDP. 6-Ethynyl and 6-cyano groups were installed, and the alkynes
were extended through both alkyl and aryl spacers. The most potent antagonists, with IC50 of ~1 µM, were found
to be trialkylsilyl-ethynyl 7 and 8 (3–5 fold greater affinity than reference 3), t-butyl prop-2-yn-1-ylcarbamate 14
and p-carboxyphenyl-ethynyl 16 derivatives, and 3 and 8 displayed surmountable antagonism of UDP-induced
production of inositol phosphates. Other chain-extended terminal carboxylate derivatives were less potent
than the corresponding methyl ester derivatives. Thus, the 6 position in this chromene series is suitable for
derivatization with flexibility of substitution, even with sterically extended chains, without losing P2Y6R affinity.
However, a 3-carboxylic acid or 3-ester substitution did not serve as a nitro bioisostere, as the affinity was
eliminated. These compounds provide additional ligand tools for the underexplored P2Y6R, which is a target for
inflammatory, neurodegenerative and metabolic diseases.
P2Y receptor
Antagonist
Purinergic
Calcium mobilization
Structure activity relationship
The P2Y6 receptor (P2Y6R) is activated by UDP and is localized in
immune cells as well as small intestine, blood, heart, blood vessels and
the central nervous system.1,2 P2Y6R is a Gq protein-coupled member of
the purinergic metabotropic P2Y receptor family and is associated with
proinflammatory effects.2 In microglial cells, the P2Y6R induces
phagocytosis of debris such as amyloid proteins, and its agonists have
been explored for treatment of Alzheimer’s disease and brain
ischemia.3–5 A P2Y6R antagonist was shown to reduce chronic neuro-
pathic pain by counteracting UDP-induced microglial and inflammatory
responses.6 Furthermore, a P2Y6R agonist prodrug was found to be
beneficial in an in vivo asthma model.7 P2Y6R antagonists have been
explored for cancer treatment with mixed results.8,9 The vaso-
constricting effects of P2Y6R agonists have also been explored, sug-
gesting a role for antagonists in treating hypertension.10 The P2Y6R is
proposed to form a heterodimer with the angiotensin (AT) II type-1 re-
ceptor (AT1R), leading to an undesired effect of P2Y6R agonists to pro-
mote vascular inflammation. A recent study supported the concept of
using P2Y6R antagonists for treatment of obesity and diabetes based on
the mouse phenotype on a high fat diet when the receptor is knocked out
in adipocytes.11 UDP analogues with increased affinity have been re-
ported, including those with substituted uracil rings and with rigid,
bicyclic substitutions of the ribose ring to achieve a receptor-preferred
conformation.12,13 The latter ribose modification is termed South (S)-
methanocarba (bicyclo[3.1.0]hexane), which maintains a constrained
(S) conformation. A UDP analogue with an isomeric North (N)-meth-
anocarba modification was completely inactive at the human (h) P2Y6R.
Only a few series of P2Y6R antagonists have been reported,35,14–16
and none have achieved nM affinity. A uridylyl phosphosulfate deriva-
tive weakly antagonized P2Y6R with IC50 112 µM.16 A bicyclic diketo-
piperazine scaffold was also found to inhibit P2Y6R effects with IC50
~
10 µM, but it was not receptor subtype selective.17 Thus, there is a
pressing need for more potent, competitive and drug-like antagonists.
Aryl di-isothiocyanate derivatives are moderately potent antagonists,
but because of the requirement of the chemically reactive -NCS groups
are likely irreversibly receptor-bound antagonists, also consistent with
the insurmountable antagonism observed.14 Widely used P2Y6R
* Corresponding author at: Molecular Recognition Section, Bldg. 8A, Rm. B1A-19, NIH, NIDDK, LBC, Bethesda, MD 20892-0810, USA.
E-mail address: kennethj@niddk.nih.gov (K.A. Jacobson).
https://doi.org/10.1016/j.bmcl.2021.128008
Received 26 February 2021; Received in revised form 24 March 2021; Accepted 25 March 2021
Available online 6 April 2021
0960-894X/Published by Elsevier Ltd.

Y.-H. Jung et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 128008
antagonist MRS2578 1 (Chart 1) having two phenyl 3-isothiocyanato
groups inhibited P2Y6R signaling in two species homologues with IC50
values of 37 nM (h) or 98 nM (rat, r), with selectivity compared to other
P2YRs. Curiously, the potency and selectivity were highly dependent on
the length of the alkyl chain and the positions of the symmetric aryl
isothiocyanate groups. A shorter symmetric homologue bearing phenyl
4-isothiocyanato groups, compound 2, only inhibited hP2Y6R (although
selectively), but not rP2Y6R.14 A step forward in the search for poten-
tially competitive antagonists was achieved several years ago, when a
hit from library screening, 3-nitro-2-(trifluoromethyl)-2H-chromene 3,
was reported by Ito et al. to be a selective P2Y6R antagonist of µM af-
finity.15 At a 30 µM concentration, it displayed no antagonism of cal-
cium mobilization induced by the native agonists of P2Y1R (10 nM
ADP), P2Y2R (1 µM UTP), P2Y4R (100 nM UTP) and P2Y11R (10 µM
ATP) expressed in 1321N1 astrocytoma cells. Its 6-bromo analogue 4
was also a P2Y6R antagonist of moderate affinity, but 2-phenyl substi-
tution reduced the affinity by an order of magnitude. The binding sites
on the P2Y6R of both of these chemical classes, isothiocyanates and
chromenes, are unexplored, using either mutagenesis or molecular
modeling. In this study, we have chosen to empirically probe the
structure activity relationship (SAR) of the chromene series using the 6
position as the site for installation of alkynes through a Sonogashira
reaction.
the alkynyl group, differing by up to 2 orders of magnitude. The bromo
derivative 4 displayed an IC50 of 3.49 µM, although it was reported as 4-
fold more potent in Ito et al.15 Pharmacological procedures are
described in the Supporting information.
The 6-iodo derivative 5 had an IC50 similar to the known 6-bromo
compound 4. Due to the synthetic accessibility, we prepared a series
of 6-ethynyl derivatives that tended to preserve the P2Y6R affinity. The
simplest substitution in the 6-ethynyl derivative 6 increased the hP2Y6R
antagonist by 2-fold compared to the lead compound 3. Trimethylsilyl 7
and triethylsilyl 8 derivatives were found to be among the most potent
hP2Y6R antagonists in this functional assay with IC50 < 1 µM. Therefore,
we replaced the Si atom with C in t-butyl derivative 9 and examined
closely related alkynes (c-Pr 10 and hydroxymethyl 11 derivatives). The
silicon-to-carbon switch of an ibuprofen analogue, conceptually similar
to compound 7 compared to 9, was exploited and was suggested as a
general modification in medicinal chemistry.20 However, here the car-
bon analogue 9 was roughly an order of magnitude weaker than the
silicon derivative 7.
An acylated propargylamine moiety, present in acetyl 12 and Boc 14
derivatives, mostly preserved the P2Y6R affinity. A neutral Boc-amino
derivative 14 was equipotent to the ethynyl compound 6. Therefore,
there appeared to be no immediate steric limitation to extending the
group at the 6 position. Thus, a propargylamine moiety was included as
a spacer in subsequent derivatives, including 12–14 and 18–22. Ter-
minal methyl ester groups (15, 18, 20, 23 and 24) or carboxylates (13,
16, 19 and 21) were present in various chain elongated derivatives.
However, no terminal amino derivatives were included. The attempted
synthesis of an unacylated propargylamine derivative (not shown)
related to 14 failed, because this chemical series is unstable to strong
acidic (i.e. Boc deprotection of 14) and basic conditions needed for
removal of common protecting groups (i.e. phthaloyl in 22). The lack of
steric constraints in this region of the chromene scaffold was indicated
by the substitution of the alkyne with a phthaloylamino group in 22
leading to slightly higher affinity than the N-acetyl analogue 12. Curi-
ously, a benzoic acid derivative 16 was equipotent to 14, but its methyl
ester derivative 15 was ~6-fold less potent. Therefore, an early prefer-
ence for an anionic group at the terminal position of the chain appeared,
at least with a p-substituted phenylethynyl group in 16. Initially, this
seemed consistent with a hypothetical proximity to the highly positively
charged extracellular loop regions typical of P2YRs. However, in the
case of extended aliphatic amino acid derivatives, a terminal carbox-
ylate group led to much weaker interactions with the receptor. For
example, methyl ester 18 displayed roughly an order of magnitude
higher affinity than its corresponding carboxylate 19, and methyl ester
20 had 13-fold higher affinity than its carboxylate 21. Similarly, anionic
succinate derivative 13 was 30-fold weaker in its interaction with the
P2Y6R than its neutral, truncated N-acetyl derivative 12. The presence of
a water-solubilizing, terminal urea group in 17 lowered affinity.
A 6-cyano derivative 25 had 4-fold lower affinity than the corre-
sponding 6-ethynyl derivative 6. An attempt to identify a nitro
The synthesis of 3-nitro-2-(trifluoromethyl)-2H-chromene de-
rivatives based on 3 and 4 was performed as shown in Scheme 1 (pro-
cedures in Supporting information). The newly synthesized analogues
tested for hP2Y6R antagonism (5–25) are shown in Table 1. Substitution
of the 6 position of the 3-nitro-2-(trifluoromethyl)-2H-chromene scaf-
fold was chosen as the principal site of derivatization. The 6-bromo
analogue 4 was previously reported as similarly potent as the 6-H
compound,15 suggesting that other halo atoms might be suitable at
this position. A 6-iodo group was selected as the first novel halo deriv-
ative (5) in our series. Subsequently, a Sonogashira reaction was used to
replace a 6-iodo group with various alkynes (7–24). The TMS group was
removed from 7 to provide alkyne 6. A variety of functionalized ter-
minal alkyne intermediates (Schemes 2A and B) were utilized to prepare
these 2H-chromene derivatives. For example, t-butyl prop-2-yn-1-
ylcarbamate reacted with 6-iodo derivative 5 to yield the Boc-amino
derivative 14. Also, a 6-cyano derivative 25 was prepared from 30. To
introduce carboxylate groups at the 3 position of 5, the 3-ethyl ester 26
was synthesized, followed by hydrolysis with LiOH to yield 27 (Scheme
2C).18 The preparation of tetramethylrhodamine derivative 43 by a
Sonogashira reaction using the alkyne derivatized fluorophore is shown
in Supporting information (Scheme S1).
The potency of the 3-nitro-2-(trifluoromethyl)–2H-chromene de-
rivatives was determined in an assay of calcium mobilization in hP2Y6R-
expressing 1321N1 astrocytoma cells,12–14 in a 96-well format. A fixed
concentration (100 nM, 2.4XEC50) of the native agonist UDP was used,
and the IC50 of each chromene derivative determined (Fig. 1). There is a
wide spread of IC50 values depending on the distal functionalization of
Chart 1. Structures of reported selective P2Y6R antagonists.14,15
2

Y.-H. Jung et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 128008
Scheme 1. Synthesis of compounds 5–25. Reagents and Conditions: A. (a) TEA, DCM, rt, 5 h, 65–77%; (b) PdCl2(PPh3)2, CuI, TEA, THF, rt, 12 h, 47–96%; (c) K2CO3,
MeOH, rt, 1.5 h, 93%. B. (a) TEA, DCM, rt, 5 h, 65%.
bioisostere in the form of a carboxyl group present in anionic derivative
27 and its ethyl ester 26 failed. Sterically bulky fluorescent tetrame-
thylrhodamine derivative 43 only weakly bound to the P2Y6R. Although
some of the compounds with improved affinity, e.g. 8, were more hy-
drophobic than the lead compound 3, the inhibitory potency was not a
function of hydrophobicity. Carboxylate 16 was both more polar and
more potent than 15. The cLogP values (ChemDraw, v. 19) are in pa-
rentheses: 3 (3.00), 8 (3.00), 15 (5.60), 16 (5.38) and 23 (2.43).
Compound 3 has been shown to inhibit UDP-induced calcium
mobilization in 1321N1 cells expressing the human P2Y6R in an insur-
mountable manner.15 In testing the generality of this characteristic
within the chemical series, we also found that increasing concentrations
of the triethylsilyl analogue 8 right-shifted the UDP activation curve in
P2Y6R expressing 1321N1 astrocytoma cells with a reduction in the
maximal effect, as well (Fig. 2). Since calcium mobilization is not an
assay under equilibrium conditions, it is not possible to conclusively
determine whether 3 and 8 are surmountable or insurmountable an-
tagonists. Thus, we further examined the ability of 3 and 8 to inhibit
UDP-induced IP-1 accumulation (Gq/11-mediated production of inositol
phosphates).43 Fig. 3 shows that at the concentration of 10 µM, 3 and 8
shifted the concentration–response curve to the right 4.6- and 10.9-fold,
respectively, without affecting the maximum effect of UDP. Thus, both 3
and 8 are surmountable antagonists, and 8 is more potent than 3 in both
functional assays. Unlike 3 and 8, MRS2578 1 has been demonstrated to
be an insurmountable antagonist in a similar assay of inositol
phosphates.14
Off-target activity of selected, relatively potent analogues (5, 7, 8,
12, 14 and 22) was determined at 46 different receptors and channels by
the Psychoactive Drug Screening Program (PDSP) at University of North
Carolina.21 A few significant off-target interactions were found for these
derivatives in the µM range, particularly at some biogenic amine re-
ceptors. They included: 5 (receptor, Ki in µM or % inhibition at 10 µM):
adrenergic α2A, 4.8; dopamine D5, 1.6; GABAA 65%; σ1, 2.1; 7: α2A, 4.7;
D5, 4.2; σ1, 1.4; 8: α2A; 9: D5, 2.1; 12: α2A, 2.2; α2B, 6.7; D5, 1.1; σ1, 4.4;
14: α2A, 0.94 ± 0.33; D3, 1.7 ± 1.0; D5, 1.25 ± 0.16; 16: D1, 1.1; D5, 2.1;
opioid µ, 4.2; α2A, 1.8; α2B, 1.0; α2C, 2.6; 22: α2A, 0.65; α2B, 4.7; D5, 1.4.
Thus, t-butyl derivative 9 displayed only one weak off-target activity.
Other receptors reported <50% inhibition at 10 µM, and most of these
interactions displayed a negligible % inhibition in the primary
comprehensive screen, as shown for compound 7 (Supporting Infor-
mation). Thus, these compounds are not promiscuous receptor binders.
Nevertheless, the relatively few off-target interactions in the µM range
may in some cases interfere with the use of these compounds as phar-
macological probes, and the need for additional SAR study to eliminate
3

Y.-H. Jung et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 128008
these activities is suggested.
Table 1
Inhibition of UDP-induced Ca2+ mobilization in a hP2Y6R-expressing 1321N1
human astrocytoma cell line.a
The stability of a representative P2Y6R antagonist toward nucleo-
philes was examined. There was no evidence of reaction of compound 7
with 2-phenylethanethiol as a nucleophile during aqueous incubation in
the presence of LiOH as a base (Supporting Information).22
The discovery of drug-like antagonists would enable the proof of
concept of various therapeutic applications of blocking this receptor.
P2Y6R antagonists have been considered in the context of possible
anticancer, cerebroprotective, antihypertensive, anti-inflammatory,
antidiabetic and anti-ischemic activity, or for use in heart failure and
chronic pain.6,23–27 For example, a heterodimer of two GPCRs, AT1R and
P2Y6R, was disrupted by MRS2578, a P2Y6R-selective inhibitor, to
reduce the risk of angiotensin-induced hypertension in an animal
model.23 P2Y6R activation also promotes the appearance of membrane
protrusions and migration of lung and colon cancer cells.28 Thus, there is
sufficient justification to discover novel P2Y6R antagonists. However,
curiously the deletion of P2Y6R promoted a more inflammatory
phenotype in alveolar macrophages.29 Also, the mouse P2Y6R knockout
increased inflammation and intestinal recruitment of Th17/Th1 lym-
phocytes in the inflammatory bowel disease (dextran sodium sulfate,
DSS) model.30 Recently, Salem et al. showed that activation of the
colonic epithelial cell P2Y6R induced intestinal inflammation, which
was reduced by native nucleoside triphosphate diphosphohydrolase 8
(NTPDase8) through cleaving endogenous UDP.31 Consistently,
MRS2578 protected against intestinal inflammation in colitis induced by
DSS.
Compound
R =
IC50, hP2Y6R (µM,
mean ± SEM)c
Known compounds
3, TIM-38b
H
2.91 ± 1.21
3.49 ± 1.54
4, Br-TIM-
Br
38b
New compounds
5
I
4.00 ± 1.59
1.33 ± 0.31
0.785 ± 0.058
0.604 ± 0.239
6.59 ± 1.95
4.87 ± 1.96
7.31 ± 1.03
7.67 ± 2.22
121 ± 72
–
–
6
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
CH
–
7d
C-Si(CH )
–
–
–
3
3
8d
C-Si(CH CH )
–
–
–
2
3
3
9d
C-C(CH )
–
–
–
3
3
–
–
10
11
12
13
14d
15
16d
17
18
19
20
21
22
C-(c-Pr)
–
–
–
C-CH OH
–
2
–
–
CCH -NH-COCH
–
2
3
–
–
CCH -NH-CO(CH ) -CO H
–
2
2
2
2
–
–
CCH -NH-Boc
1.20 ± 0.08
7.13 ± 3.14
1.09 ± 0.42
10.6 ± 1.3
–
2
–
–
C-Ph-p-CO CH
–
2
3
–
–
C-Ph-p-CO H
–
2
The effect of pharmacological inhibition and complete genetic
deletion of the receptor might not be equivalent, but there is still justi-
fication to test newly-reported P2Y6R antagonists in various disease
models.32 In a high-fat diet model of obesity, the knockout of P2Y6R in
both whole body and selectively in adipocytes, but not in skeletal muscle
alone, improved metabolic parameters.11 The adipocyte knockout mice
resembled the whole-body knockout of P2Y6R, and therefore systemic
antagonism is predicted to be beneficial in diabetes. Recent studies also
suggest the application of P2Y6R antagonists for chronic pain treat-
ment,6,27,33 although an earlier study showed no protective effect by
MRS2578 (10 mg/kg, i.p.) in the mouse spared nerve injury model, even
though the antagonist was bioavailable over at least 30 min.34 MRS2578
administration (i.t.) relieved neuropathic pain in the rat chronic
constriction injury (CCI) model or following UDP (i.t.) administration.24
P2Y6R antagonists might also prove useful in treating persistent bladder
storage symptoms in male benign prostatic hyperplasia (BPH).35 P2Y6R
activation led to bladder overactivity via increased mucosal ATP release,
an effect blocked by blocked by 50 nM MRS2578.
–
–
C-Ph-p-CONH(CH ) -NHCONH
–
2
2
2
–
–
CCH -NH-CO(CH ) -CO CH
5.23 ± 1.74
69.1 ± 20.7
5.90 ± 1.38
77.9 ± 60.7
4.65 ± 1.17
–
2
2
3
2
3
–
–
CCH -NH-CO(CH ) -CO H
–
2
2
3
2
–
–
CCH -NH-COCH C(CH )2 CH2-CO2CH3
–
2
2
3
–
–
CCH -NH-COCH C(CH )2 CH2-CO2H
–
2
2
3
23
24
6.33 ± 0.90
6.84 ± 0.57
However, tractable leads for novel high affinity P2Y6R antagonists
are lacking. Furthermore, there is no X-ray or cryo-EM structure of the
P2Y6R. It is apparent that the structures of the P2Y6R and other P2YR
already determined are sufficiently divergent to make an in silico
screening approach to identify novel chemotypes as P2Y6R antagonists
uncertain. It would be most efficient for receptor modeling to already
have diverse classes of competitive antagonists as a modeling test set,
but such is lacking. Although both P2Y1R and P2Y6R are nucleoside-5′-
diphosphate receptors of the Gq-coupled subfamily, and models of the
P2Y6R based on the X-ray structure of Gq-coupled P2Y1R (and on the
chemokine CXCR4 receptor) have been reported,13,36 there appear to be
major structural differences affecting ligand SAR. For example, it was
not possible to derive P2Y6R antagonists by generalizing from bisphos-
phates P2Y1R antagonists, such as MRS2179,14 to the uridine family.
The corresponding 2′-hydroxy- or 2′-deoxy-uridine-3,5-bisphosphates
did not antagonize the P2Y6R.37
–
CH3CH2
H
–
25
26
27
43
C
N
5.33 ± 1.83
>100
–
>100
34.6 ± 8.9
Measured in Ca2+ assays in whole hP2Y6R-expressing 1321N1 cells (gift of T.
K. Harden, Univ. of North Carolina, Chapel Hill, NC).19,12–14 UDP (100 nM) was
used as agonist.
a
Thus, we chose to explore a reported lead despite its unfavorable
physicochemical characteristics, e.g. 3-nitro-2-(trifluoromethyl)-2H-
chromene derivatives.15 The related chromone scaffold has also been
applied to diverse drug classes,38 and several chromenes were found to
inhibit ecto-nucleotidase CD73.39 One of the disadvantages of com-
pound 3 as a lead molecule is that the analogues are racemic. Another
Compound reported by Ito et al., 2017.15 IC50 values shown were deter-
b
mined here.
c
d
3–4 independent determinations, each in triplicate.
7, MRS4695; 8, MRS4774; 9, MRS4773; 14, MRS4706; 16, MRS4656.
4

Y.-H. Jung et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 128008
Scheme 2. Synthesis various functionalized
alkyne intermediates and 3-carboxy de-
rivatives 26 and 27. A. (a) succinic anhy-
dride, 4-DMAP, DCM, rt,
1 h, 87%; (b)
glutaric acid monomethyl ester chloride, 4-
DMAP, DCM, rt, 5 h, 72%; (c) KOH, MeOH:
water
= 2:1, rt, 0.5 h, 68–98%; (d) 5-
methoxy-3,3-dimethyl-5-oxopentanoic acid,
HATU, 4-DMAP, DCM, rt, 12 h, 77%. B. (a)
(2-amino-ethyl)-urea hydrochloride, DIPEA,
HATU, DMF, 0 ◦C, 1 h, 71%; (b) glycine
methyl ester hydrochloride for 40 or L-
alanine methyl ester hydrochloride for 41,
HATU, 4-DMAP, DCM, rt, 12 h, 84–87%. C.
(a) K2CO3, DMF, 80 ◦C, 12 h, 46%; (b) LiOH,
water:MeOH = 1:1, 50 ◦C, 5 h, 85%.
Fig. 1. Antagonism by representative antagonists (silyl derivative 7, benzoic acid derivative 16, and extended ester 20 and corresponding carboxylate 21) of UDP-
induced Ca2+ mobilization in P2Y6R-expressing 1321N1 cells.19,12–14 100 nM UDP was used as agonist.
disadvantage is the presence of a nitro group in 3, which is a structural
alert in medicinal chemistry,40 although nitro-bearing drugs and drug
candidates have been developed.41 Although various functional groups
have been found to serve as nitro group bioisoteres at various targets,42
replacement with a 3-carboxylate or 3-ester failed to preserve affinity in
the present series. Nevertheless, we have established that the SAR of 3
can be explored, leading to a family of alkynyl congeners with variable
µM P2Y6R affinity. A site on the scaffold for chain extension has been
identified: the 6 position, which is suitable for derivatization with
flexibility of substitution, even with sterically extended chains, without
5

Y.-H. Jung et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 128008
Fig. 2. Antagonism by fixed concentrations of compound 8 of concentration response curves for UDP-induced Ca2+ mobilization in P2Y6R-expressing 1321N1 cells
(representative curves). EC50 values (µM, mean ± SEM, N = 3) for UDP are: 0.287 ± 0.036 (1 µM); 0.59 ± 0.27 (3 µM); 1.03 ± 0.56 (6 µM); 2.37 ± 1.28 (9 µM). EC50
value for UDP alone was 41.6 nM ± 7.0 nM.
substitution of the alkyne with a phthaloylamino group in 22 or phenyl
groups preserves affinity. We were unsuccessful in an effort to find a
bioisostere of the 3-nitro group or to maintain affinity in a fluorescent
analogue. These identification of position 6 in this chromene series as a
versatile site for derivatization will provide additional ligand tools for
the underexplored P2Y6R, which was shown to be involved in various
animal models of disease.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgments
We thank the NIDDK Intramural Research Program (ZIADK31116)
Fig. 3. Antagonism of inhibit UDP-induced IP-1 accumulation by compounds 3
for support. We thank Bryan L. Roth, National Institute of Mental
and 8 (10 µM) in P2Y6R-expressing 1321N1 cells (N = 3). EC50 values (µM,
Health’s Psychoactive Drug Screening Program (Univ. North Carolina at
mean ± SEM) for UDP (conc. 8) are: 0.295 ± 0.088 (UDP alone); 1.35 ± 0.45
Chapel Hill, Contract # HHSN-271-2008-00025-C) for screening data.
(3); 3.21 ± 0.86 (8).
Appendix A. Supplementary data
losing P2Y6R affinity.
In conclusion, substitution of 3-nitro-2-(trifluoromethyl)-2H-chro-
Supplementary data to this article can be found online at https://doi.
menes with linear alkynes and extending amide chains provided
org/10.1016/j.bmcl.2021.128008.
considerable P2Y6R affinity in some analogues. Relatively potent
trimethylsilyl-ethynyl 7 and triethylsilyl-ethynyl 8 derivatives displayed
References
3-fold greater antagonistic affinity (IC50 < 1 µM) than reference chro-
mene 3. There are other derivatives also which may be worth consid-
[1] von Kügelgen I, Hoffmann K. Neuropharmacology. 2015;104:50–61.
ering, for further screening. The surmountable antagonism by 3 and 8 of
UDP-induced production of inositol phosphates is encouraging for the
future utility of this compound series. P2YR ligands often benefit from
the presence of anionic groups, consistent with the negative charge on
the native nucleotide ligands. However, only one terminal carboxylate
derivative, a carboxy-phenyl-ethynyl analogue 16 retained comparable
affinity, while extended carboxylates were considerably weaker in
binding. Thus, we have expanded to range of analogues of 3 that block a
P2Y6R functional effect with a consistent SAR pattern observed for 6-
alkynyl derivatives. Although we do not have a three-dimensional
model of the P2Y6R binding of this series, the empirical SAR explora-
tion suggested that the binding region on the receptor surrounding the
extended 6-alkynyl groups allows much steric freedom of substitution,
possibly suggestive of facing the extracellular loops. For example,
[2] Jacobson KA, Delicado EG, Gachet C, et al. Br J Pharmacol. 2020;177:2413–2433.
[3] Koizumi S, Shigemoto-Mogam Y, Nasu-Tada K, et al. Nature. 2007;446:1091–1095.
[4] Haydon P, Lee J, Dong J, et al. Uridine diphosphate derivatives, compositions and
methods for treating neurodegenerative disorders. WO 2013049686 A1.
[5] Wen R-X, Shen H, Huang S-X, et al. CNS Neurosci Ther. 2020;26:416–429.
[6] Bian J, Zhang Y, Liu Y, et al. Pain Res Manage. 2019, 2612534. https://doi.org/
10.1155/2019/2612534.
[7] Chetty A, Sharda A, Warburton R, et al. J Asthma Allerg. 2018;11:159–171.
[8] Placet M, Arguin G, Molle CM, et al. Biochim Biophys Acta (BBA) - Mol Basis Dis.
2018;1864:1539–1551.
[9] Wan H, Xie R, Xu J, et al. Sci Rep. 2017;7:2459. https://doi.org/10.1038/s41598-
017-02562-x.
[10] X. Du S. Zhang Q. Xiang et al. BioMed Res International. 2020;1983940:https://
doi.org/10.1155/2020/1983940.
[11] Jain S, Pydi SP, Toti KS, et al. Proc Natl Acad Sci USA. 2020;117(48):30763–30774.
[12] Maruoka H, Barrett MO, Ko H, et al. J Med Chem. 2010;53:4488–4501.
[13] Toti KS, Jain S, Ciancetta A, et al. Med Chem Commun. 2017;8:1897–1908.
[14] Mamedova L, Joshi BV, Gao ZG, et al. Biochem Pharmacol. 2004;67:1763–1770.
[15] Ito M, Egashira S, Yoshida K, et al. Life Sci. 2017;180:137–142.
6

Y.-H. Jung et al.
Bioorganic & Medicinal Chemistry Letters 41 (2021) 128008
[16] Meltzer D, Ethan O, Arguin G, et al. Bioorg Med Chem. 2015;23(17):5764–5773.
[17] Besada P, Mamedova L, Thomas CJ, et al. Org Biomol Chem. 2005;3:2016–2025.
[18] Wang JL, Aston K, Limburg D, et al. Bioorg Med Chem Lett. 2010;20(23):
7164–7168.
[29] Nagai J, Balestrieri B, Fanning LB, et al. J Clin Invest. 2019;129(12):5169–5186.
[30] Salem M, El Azreq MA, Pelletier J, et al. Biochimica et Biophysica Acta (BBA) - Mol
Basis Dis. 2019;1865(10):2595–2605.
[31] Salem M, Lecka J, Pelletier J, et al. Gut. Published Online First: 15 January 2021.
doi: 10.1136/gutjnl-2020-320937.
[19] Robaye B, Boeynaems JM, Communi D. Eur J Pharmacol. 1997;329(2):231–236.
[20] Kleemiss F, Justies A, Duvinage D, et al. J Med Chem. 2020;63(21):12614–12622.
[21] Besnard J, Ruda GF, Setola V, et al. Nature. 2012;492:215–220.
[22] A solution of 7 in ethanol (1.0 mM) was treated with 2-phenylethanethiol (1 eq)
and LiOH (1 eq), and the reaction mixture was stirred at 22 ◦C for 1 h. After
warming to 37 ◦C, the mixture was stirred for an additional 1 h. The stability of
compound 7 in the reaction mixture was analyzed by TLC and HPLC, both after 1 h
at rt and after the additional 1 h at 37 ◦C. There was no evidence of reaction of
compound 7 with 2-phenylethanethiol as a nucleophile.
[32] Vieira RP, Müller T, Grimm M, et al. Am J Respir Crit Care Med. 2011;184(2):
215–223.
[33] Wang Z, Zhao W, Shen X, et al. J Cell Biochem. 2019;120:17123–17130.
[34] Neumann A, Müller CE, Namasivayam V. WIREs Comput Mol Sci. 2020;10, e1464.
[35] Syhr KMJ, Kallenborn-Gerhardt W, Lu R, et al. Pharmacol Biochem Behav. 2014;
124:389–395.
˜
[36] Silva I, Ferreirinha F, Magalhaes-Cardoso MT, et al. J Urol. 2015;194(4):
1146–1154.
[23] Sunggip C, Nishimura A, Shimoda K, et al. Pharmacol Res. 2017;120:51–59.
[24] Anwar S, Pons V, Rivest S. Cells. 2020;9:1595.
[37] Besada P, Shin DH, Costanzi S, et al. J Med Chem. 2006;49:5532–5543.
[38] Gaspar A, Matos MJ, Garrido J, et al. Chem Rev. 2014;114(9):4960–4992.
[39] Ripphausen P, Freundlieb M, Brunschweiger A, et al. J Med Chem. 2012;55:
6576–6581.
[25] Huang D, Yang J, Liu XH, et al. J Clin Neurosci. 2018;56:156–162.
[26] Zhou M, Wang W, Li Y, et al. Drug Discovery Today. 2020;25(3):568–573.
[27] Shimoda K, Nishimura A, Sunggip C, et al. Sci Rep. 2020;10:13926. https://doi.
org/10.1038/s41598-020-70956-5.
[40] Kalgutkar AS. J Med Chem. 2020;63(12):6276–6302.
[41] Nepali K, Lee HY, Liou JP. J Med Chem. 2019;62(6):2851–2893.
[42] Meanwell NA. J Med Chem. 2011;54:2529–2591.
[43] Gao ZG, Jacobson KA. Mol Pharmacol. 2017;92:613–626.
´
[28] Girard M, Dagenais Bellefeuille S, Eiselt E, et al. J Cell Physiol. 2020;235:
9676–9690.
7