Synthesis of spiro[chroman-2,4′-piperidin]-4-one derivatives as acetyl-CoA carboxylase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.11.099

Various spiro[chroman-2,4′-piperidin]-4-one derivatives (38a-m and 43a-j) have been designed, synthesized and evaluated for in vitro acetyl-CoA carboxylase (ACC) inhibitory activity. Several compounds have shown ACC inhibitory activity in low nanomolar ra

Full text of DOI:10.1016/j.bmcl.2008.11.099

Bioorganic & Medicinal Chemistry Letters 19 (2009) 949–953
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis of spiro[chroman-2,40-piperidin]-4-one derivatives as acetyl-CoA
carboxylase inhibitors
Pundlik Shinde a, Sanjay K. Srivastava a, Rajendra Odedara a, Davinder Tuli a, Siralee Munshi b,
Jitendra Patel b, Shitalkumar P. Zambad c, Rajesh Sonawane d, Ramesh C. Gupta a,
,
*
Vijay Chauthaiwale e, Chaitanya Dutt f
a Medicinal Chemistry, Torrent Research Centre, Torrent Pharmaceuticals Ltd, PO Bhat, Gandhinagar 382428, Gujarat, India
b Cellular and Molecular Biology, Torrent Research Centre, Torrent Pharmaceuticals Ltd, PO Bhat, Gandhinagar 382428, Gujarat, India
c Pharmacology, Torrent Research Centre, Torrent Pharmaceuticals Ltd, PO Bhat, Gandhinagar 382428, Gujarat, India
d Analytical Development, Torrent Research Centre, Torrent Pharmaceuticals Ltd, PO Bhat, Gandhinagar 382428, Gujarat, India
e Discovery Research, Torrent Research Centre, Torrent Pharmaceuticals Ltd, PO Bhat, Gandhinagar 382428, Gujarat, India
f Torrent Research Centre, Torrent Pharmaceuticals Ltd, PO Bhat, Gandhinagar 382428, Gujarat, India
a r t i c l e i n f o
a b s t r a c t
Various spiro[chroman-2,40-piperidin]-4-one derivatives (38a–m and 43a–j) have been designed, synthe-
sized and evaluated for in vitro acetyl-CoA carboxylase (ACC) inhibitory activity. Several compounds have
shown ACC inhibitory activity in low nanomolar range. Compound 38j reduced the respiratory quotient
(RQ) in C57BL/6J mice indicating increase in whole body fat oxidation even in the presence of high car-
bohydrate diet. Structure–activity relationship (SAR) has been discussed.
Article history:
Received 6 September 2008
Revised 19 November 2008
Accepted 24 November 2008
Available online 3 December 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Acetyl-CoA carboxylase inhibitors
Spirochromanones
Acetyl-CoA carboxylase (ACC), a biotin-dependent enzyme, cat-
alyzes the ATP dependent carboxylation of acetyl-CoA to form mal-
onyl-CoA. ACC plays an important role in energy balance by
controlling malonyl-CoA synthesis.1–4 Malonyl-CoA participates
in two opposing pathways: a substrate for fatty acid synthesis
and a regulator of fatty acid oxidation. ACCs are involved in con-
trolling the switch between carbohydrate and fatty acid utilization
in liver and skeletal muscle and also regulating insulin sensitivity
in the liver, skeletal muscle and adipose tissue.2,3 Inhibition of
ACCs, with its resultant inhibition of fatty acid synthesis and stim-
ulation of fatty acid oxidation, has the potential to favorably affect
the multitude of risk factors associated with metabolic syndrome,
obesity, and type-2 diabetes.1
Bispiperidylcarboxamide derivative I (CP-640186, Fig. 1), hav-
ing a tricyclic hydrophobic core, showed inhibition of both ACC1
and ACC2 with IC50 of ꢀ55 nM. Compound I has shown to increase
fatty acid oxidation in vitro in C2C12 cells and in ob/ob mice. It also
reduced fatty acid synthesis in HepG2 cells and in SD rats.1
Recently, several spirochromanone derivatives II and III (Fig. 1),
containing a bicyclic ring hydrophobic core, reported as potent
ACC inhibitors. In vivo experiment of spirochromanones resulted
in reduced body weight gain, fat mass gain and hepatic triglyceride
content in C57BL/6J mice whereas reduced plasma glucose levels in
KKAy mice. It is also reported that in humans with a BMI P 30,
effective compounds resulted in loss of body weight or an
improvement in insulin levels5 (structures are not disclosed).
The crystal structure of the carboxytransferase (CT) domain of
yeast ACC in complex with compound I has been described by
Zhang et al.6 Compound I is bound at the active site of the CT do-
main, at the interface between the two monomers of the CT dimer.
The structure suggests that the active site of CT domain is mostly
hydrophobic in nature with few hydrogen bonding potentials. From
the structure, it appears that a big hydrophobic pocket is formed by
the side chains of Ile2033 and Lys1764 and can accommodate a
bulkier lipophilic group like tricyclic ring systems or teraryl ring
systems. There can be a presence of hydrogen bond acceptor like
carbonyl oxygen next to the bulkier hydrophobic group that can
make hydrogen bonding with the main chain of Glu2026. There is
another hydrophobic groove on the surface of the active site and
hydrogen bond forming site formed by the main chain amide of
Gly1958. Based on this structural information, we designed and
synthesized several spirochromanones linked to various mono-,
bi-, and tricyclic hydrophobes and evaluated for ACC inhibitory
activity. Herein, we report the synthesis, in vitro ACC inhibitory
activity and SAR of various amide (38a–m) and urea (43a–j) deriv-
atives of spiro[chroman-2,40-piperidin]-4-one. The in vivo efficacy
profile of one of the potent derivative (38j) has also been discussed.
Synthesis of amide derivatives of various unsubstituted, 6-
substituted, 7-substituted, and 6,7-disubstituted spiro[chroman-
* Corresponding author. Tel.: +91 79 23969100; fax: +91 79 23969135.
E-mail address: rameshgupta@torrentpharma.com (R.C. Gupta).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.11.099

950
P. Shinde et al. / Bioorg. Med. Chem. Lett. 19 (2009) 949–953
Figure 1. Structures of previously reported compounds as ACC inhibitors.
Scheme 1. Synthesis of compounds (38a–m). Reagents and conditions: (i) pyrrolidine, EtOH, reflux, 24–36 h; (ii) MeCN–HCl, 1–2 h; (iii) 2,6-diphenylisonicotinic acid, EDCI,
HOBt, DIEA, THF, 8–12 h, 10–88%; (iv) Br-(CH2)3-COOEt, NaOAc, H2O, 12 h; (v) NaOH, MeOH–H2O, 18 h; (vi) EDCI, HOBt, DMAP, THF, 8–12 h, 82%; (vii) MeI, NaH, THF, 4–6 h,
68–70%; (viii) R1-H, reflux; (ix) Pd(Ph3)2Cl2, CuI, Et3N, THF, 18 h; (x) hydrazine hydrate, EtOH, reflux, 4 h; (xi) CH3COCl, Et3N, THF, 3 h.
2,40-piperidin]-4-ones (38a–m) are described in Scheme 1. Appro-
priate 2-hydroxyacetophenones 1–7 were treated, separately, with
N-BOC-4-piperidone to provide respective spiro[chroman-2,40-
piperidin]-4-one derivatives 8–14 using the similar method as re-
ported in the literature.7 The removal of Boc group in 8–11,
through acidic hydrolysis yielded respective amine derivatives as
their hydrochloride salts 17–20. The coupling reaction of com-
pounds 17–20 with 2,6-diphenylisonicotinic acid in the presence
of diisopropylethylamine (DIEA), 3-ethyl-1-(N,N-dimethyl) amino-
propylcarbodiimide (EDCI), and hydroxybenzotriazole (HOBt) to
yield corresponding amide derivatives 38a–d. The N-alkylation of
6-amino spirochroman 14 with ethyl 4-bromobutanoate in the
presence of sodium acetate followed by its basic hydrolysis with
sodium hydroxide yielded compound 15. Intramolecular cycliza-
tion of 15 was carried out in the presence of EDCI and HOBt to give
its pyrrolidinone derivative and further acidic hydrolysis resulted
into compound 16. Compound 38j was prepared by the reaction
of 16 with 2,6-diphenylisonicotinic acid using the same coupling
condition as described for 38a. Introduction of gem-dimethyl in
compounds 8 and 9 were carried out in the presence of methyl io-
dide using sodium hydride as a base and THF as a solvent to afford
compounds 34 and 35, respectively. Deprotection of compounds 34
and 35 using acetonitrile-hydrogen chloride followed by coupling
with 2,6-diphenylisonicotinic acid, using the same condition as de-
scribed for compounds 38a–c, yielded compounds 38l and 38m,
respectively. Sonogashira coupling8 of 13 with (S)-2-(1-methyl-
prop-2-ynyl)-isoindole-1,3-dione (39)9 yielded compound 31,
which was then refluxed in the mixture of hydrazine hydrate and

P. Shinde et al. / Bioorg. Med. Chem. Lett. 19 (2009) 949–953
951
Scheme 2. Synthesis of compounds (43a–j).
ethanol to yield amino derivative 32. N-acetylation of 32 followed
Table 1
In vitro ACC activity of amide derivatives of various spiro[chroman-2,40-piperidin]-4-
by Boc deprotection yielded compound 33. Compound 38k was
prepared by coupling of 33 with 2,6-diphenylisonicotinic acid
using the same amide coupling reagent as described earlier.
The 7-bromo group in 12 was replaced, by reacting various sec-
ondary cyclic amines such as morpholine, N-methylpiperazine,
piperidine, pyrrolidine, and 2,6-dimethylmorpholine resulting to
corresponding 7-amino spirochroman derivatives 21–25. The
acidic hydrolysis of 21–25 yielded compounds 26–30. As described
earlier for amide formation, compounds 26–30 were coupled, sep-
arately, with 2,6-diphenylisonicotinic acid to yield corresponding
7-substituted spiro[chroman-2,40-piperidin]-4-one derivatives
38e–38i.
ones 38a–m
O
5
8
4
6
7
3
R
3'
4'
2'
1'
2
R1
O
O
N
1
5'
6'
38a-m
N
Compound
R
R1
IC50 (nM)
Synthesis of urea derivatives of various spiro[chroman-2,40-
piperidin]-4-ones 43a–j have been described in Scheme 2. Com-
pounds 17, 18, 16, 20, 26 were treated with phenoxazine-10-car-
bonyl chloride (40) in the presence of DIEA in solvent THF to
give 43a, 43b, 43c, 43d, and 43e, respectively. Similarly, reaction
of phenothiazine-10-carbonyl chloride (41) with 17, 18, and 20
yielded corresponding urea derivatives 43f–h and treatment of
10,11-dihydro-dibenzo[b,f]azepine-5-carbonyl chloride (42) with
18 and 20 furnished 43i and 43j, respectively.
38a
38b
38c
38d
38e
38f
38g
38h
38i
38j
38k
38l
38m
I1,6
H
H
H
1200
6
2
6
26
27
36
61
59
NHCO(CH2)2CH3
Morpholine
Morpholine
H
H
H
H
H
Methyl
H
Morpholine
(N-Methyl)piperazine
Piperidine
Pyrrolidine
(2,6-Dimethyl)morpholine
(2-oxo)pyrrolidine
C„CCH(CH3)NHCOCH3
H
H
H
H
H
—
—
—
14
61
All the synthesized spiro[chroman-2,40-piperidin]-4-one deriv-
atives 38a–m and 43a–j were screened for their in vitro inhibitory
activity against rat skeletal muscle ACC enzyme. Compounds
44%a
83%a
55
N(CH3)CO(CH2)2CH3
—
—
—
II5
ꢀ100%b
showing >90% ACC inhibition at 10 lM concentration, were sub-
III5
ꢀ100%b
jected for IC50 evaluation.10 The data are summarized in Tables 1
and 2. A number of synthesized compounds showed ACC activity
in nanomolar range and structure–activity relationship is de-
scribed below.
a
Inhibition at 10
Inhibition at 1
lM.
b
lM/L; —, not applicable.
Amongst amide derivatives 38a–m, the unsubstituted
spiro[chroman-2,40-piperidin]-4-one derivative 38a was initially
found active (IC50 1200 nM) against ACC enzyme (Table 1). ACC
inhibitory activity was at least 19-fold improved upon introducing
substituent at either position-6 or -7 in spiro[chroman-2,40-piper-
idin]-4-one except in compound 38l. We found that butanamide
(38b) and morpholine (38d) substituent at position-6 in

952
P. Shinde et al. / Bioorg. Med. Chem. Lett. 19 (2009) 949–953
Table 2
In vitro ACC activity of urea derivatives of various spiro[chroman-2,40-piperidin]-4-
ones 43a–j
O
5
8
6
7
4
3
R
3'
4'
2'
1'
2
R1
O
O
N
1
5'
6'
N
43a-j
Y
Compound
R
R1
Y
IC50 (nM)
43a
43b
43c
43d
43e
43f
43g
43h
43i
43j
I1,6
H
H
H
H
H
O
O
O
O
O
S
S
S
51%b
101
236
NHCO(CH2)2CH3
(2-Oxo)pyrrolidine
Morpholine
H
99
208
Figure 2. Effect of compound 38j on RQ in C57BL/6J mice through ip route.
Morpholine
H
H
H
H
H
H
—
—
—
16%a
48%a
64%a
4%a
NHCO(CH2)2CH3
Morpholine
NHCO(CH2)2CH3
Morpholine
—
—
—
have shown ACC inhibitory activity in nanomolar range (Table 2).
For example, unsubstituted spiro[chroman-2,40-piperidin]-4-one
derivative 43a was found active (51% ACC inhibition at 5 lM).
(CH2)2
(CH2)2
7%a
55
II5
ꢀ100%c
The activity was several fold increased upon incorporating substi-
tuent at position-6 in spiro[chroman-2,40-piperidin]-4-one in 43a
such as compounds 43b and 43d, having butanamide and morpho-
line substituent, respectively, have shown IC50 ꢀ 100 nM. The
activity was, however, >2-fold lowered either upon introducing
2-pyrrolidinone substituent at position-6 (compound 43c). The
compounds 43c and 43e, were found still more potent than 43a.
Interestingly, none of the phenothiazine (43f–h) and iminodiben-
zyl (43i–j) derivatives showed significant ACC inhibitory activity
III5
ꢀ100%c
a
Inhibition at 10
l
M.
M.
M/L; —, not applicable.
b
c
Inhibition at 5
Inhibition at 1
l
l
spiro[chroman-2,40-piperidin]-4-one (IC50 6 nM) elicited 200-fold
better activity in comparison to parent compound 38a. The
spiro[chroman-2,40-piperidin]-4-one derivative having pyrrolidi-
none substitutent (38j) was found 85-fold more potent (IC50
14 nM) as compared to parent compound 38a. Compound 38k pos-
sessing alkyne substituent linked to spiro[chroman-2,40-piperi-
din]-4-one through C–C bond, elicited relatively lesser activity
(IC50 61 nM) than compounds 38b, 38d, and 38j. However, com-
pound 38k was still 28-fold more potent than parent compound
38a. These findings suggested that morpholine and butanamide
at position-6 in spiro[chroman-2,40-piperidin]-4-one were better
substituents.
Introduction of morpholine substituent (38e) at position-7 in
spiro[chroman-2,40-piperidin]-4-one, activity (IC50 26 nM) was 4-
fold lowered in comparison to its positional isomer 38d. No major
change in the activity was observed upon replacement of morpho-
line with other six-membered rings such as N-methylpiperazine
(38f) or piperidine (38g). Replacement of morpholine with five-
membered rings such as pyrrolidine (38h), inhibitory activity
(IC50 61 nM) was further 2-fold reduced when compared to com-
pound 38e. After introducing dimethyl group in morpholine ring
in compound 38e, resulting compound 38i showed 2-fold less
activity than parent compound 38e. The 7-substituted spiro[chro-
man-2,40-piperidin]-4-one derivatives 38e–i exhibited lesser activ-
ity than 6-substuited derivatives 38b, 38d, 38j, and 38k but still
found to be at least 19-fold more active than unsubstituted spiro-
chromanone 38a. The ACC inhibitory activity did not improve even
after introducing gem-dimethyl group in unsubstituted or 6-
substituted spiro[chroman-2,40-piperidin]-4-one derivatives, such
as compounds 38l and 38m, which showed 44% and 82% ACC inhi-
as they exhibited 4–64% inhibition at 10 lM concentration as
shown in Table 2. It appeared that upon replacement of oxygen
atom in phenoxazine hydrophobe by sulfur or di-methylene, activ-
ity is lowered.
Based on in vitro ACC inhibitory activity and permeability data,
compound 38j was selected for in vivo studies. Although com-
pounds 38b–d are more potent than 38j, but all three having low
permeability as compared to 38j. Compound 38j showed moderate
permeability (PAMPA, Pe 26.76 ꢁ 10ꢂ6 cm/s at pH 6.8). As inhibi-
tion of ACC would be expected to promote fat oxidation, we exam-
ined effect of compound 38j11 on respiratory quotient (RQ) in
C57BL/6J mice,12 which reflects contribution of carbohydrate and
fat oxidation to total energy expenditure. The RQ is the ratio of
CO2 production to O2 consumption. The value of RQ reduced, when
there is shift from carbohydrate to fat metabolism. Compound 38j
was administered intraperitoneally (ip) at two doses, that is,
30 mg/kg and 60 mg/kg (Fig. 2). The compound 38j dose depen-
dently decreased the RQ indicating increase in whole body fat oxi-
dation even in the presence of high carbohydrate diet.
In conclusion, we have found that spirochromanones linked to
hydrophobic core through amidic linkage were, in general, superior
to those having urea linkage. The spiro[chroman-2,40-piperidin]-4-
one attached to diphenylisonicotinoyl hydrophobic core showed
potential for providing potent ACC inhibitory activity. In particular,
substitution at position-6 in aromatic ring of spiro[chroman-2,40-
piperidin]-4-one through C–N bond, has played crucial role for
eliciting activity. These results suggest that further derivatives at
position-6 and -7 of spiro[chroman-2,40-piperidin]-4-one may
yield compounds with better in vivo activity.
bition at 10 lM concentration, respectively. It was interesting to
note that the 6,7-disubstituted derivative such as 7-methyl-6-mor-
pholino-spiro[chroman-2,40-piperidin]-4-one derivative 38c was
found most potent (IC50 2 nM) compound. Compound 38c has
shown 3-fold more activity than its structurally closer congener
38d.
Acknowledgments
The authors wish to offer their deep thanks to Dr. Shikha Kumar
for her support in preparing the manuscript, and are also grateful
to the Analytical department for their help in generating spectro-
scopic data.
Amongst urea derivatives of spiro[chroman-2,40-piperidin]-4-
one derivatives 43a–j, only compounds containing phenoxazines

P. Shinde et al. / Bioorg. Med. Chem. Lett. 19 (2009) 949–953
and terminated by addition of 25
953
ll
of 10% perchloric acid. In case of
References and notes
compounds 38a and 43a, where assay buffer contained 50 mM HEPES (pH
7.5), 10 mM MgCl2, 10 mM tripotassium citrate, dithiothreitol (DTT) 1 mM and
0.100% BSA, 2 mM ATP, 0.0625 mM acetyl CoA and 7 mM KHCO3. The
supernatants analyzed by HPLC. The HPLC method used column packed with
octadecyl silane, embedded with polar substituents, on silica support as
stationary phase and ammonium phosphate buffer and acetonitrile as mobile
phase.
1. Harwood, H. J., Jr.; Petras, S. F.; Shelly, L. D.; Zaccaro, L. M.; Perry, D. A.;
Makowski, M. R.; Hargrove, D. M.; Martin, K. A.; Tracey, W. R.; Chapman, J. G.;
Magee, W. P.; Dalvie, D. K.; Soliman, V. F.; Martin, W. H.; Mularski, C. J.;
Eisenbeis, S. A. J. Biol. Chem. 2003, 278, 37099.
2. McGarry, J. D.; Brown, N. F. Eur. J. Biochem. 1997, 244, 1.
3. Abu-Elheiga, L.; Matzuk, M. M.; Abo-Hashema, K. A.; Wakil, S. J. Science 2001,
291, 2613.
4. Harwood, H. J., Jr. Curr. Opin. Investig. Drugs 2004, 5, 283.
5. Takeru, Y.; Hideki J.; Kenji, N.; Koji, Y.; Tomoharu, I.; Mitsuru, O.; Hideaki, I.; Jun
S.; Jun, K.; Lihu, Y. WO/2007/011809.
6. Zhang, H.; Tweel, B.; Li, J.; Tong, L. Structure 2004, 12, 1683.
7. Yang, L.; Morriello, G.; Prendergast, K.; Cheng, K.; Jacks, T.; Chan, W. W.;
Schleim, K. D.; Smith, R. G.; Patchett, A. A. Bioorg. Med. Chem. Lett. 1998, 8, 107.
8. Clark, R. F.; Zhang, T.; Wang, X.; Wang, R.; Zhang, X.; Camp, H. S.; Beutel, B. A.;
Sham, H. L.; Gu, Y. G. Bioorg. Med. Chem. Lett. 2007, 17, 1961.
9. Beutel, B. A.; Camp, H. S.; Clark, R. F.; Gu, Y.-G.; Hansen, T. M.; Keyes, R. F.; Li, Q.;
Liu, G.; Sham, H. L.; Wang, X.; Weitzberg, M.; Xin, Z.; Xu, X.; Zhand, T. US Patent
0,178,400, 2006.
10. ACC enzyme was isolated from rat skeletal muscle by streptavidin affinity
chromatography. Briefly, ammonium sulfate precipitated skeletal muscle
homogenate was dialyzed over-night. Dialyzed sample was passed through
streptavidin column for protein binding. Protein was eluted by higher
concentration of biotin (5 mM).1,13 Protein concentration was estimated by
Bradford Method. The enzyme activity was measured by the loss of acetyl CoA
and/or the appearance of malonyl CoA using HPLC.9,13,14 Assay buffer contained
50 mM HEPES (pH 7.5), 20 mM MgCl2, 10 mM tripotassium citrate, and 0.075%
Bovine Serum Albumin (BSA), 2 mM ATP, 1 mM acetyl-CoA, 17.6 mM NaHCO3
11. 38j: Rf = 0.6 (ethylacetate/hexane, 75:25); IR (KBr): 3358, 3064, 2956, 1694,
1624, 1546, 1488, 1440, 1223, 972, 692 cmꢂ1 1H NMR (400 MHz,CDCl3): d
;
8.26–8.15 (m, 5H), 7.68–7.65 (m, 3H), 7.55–7.46 (m, 6H), 7.07 (d, 1H,
J = 9.2 Hz), 4.62 (d, 1H, J = 13.2 Hz), 3.88 (t, 2H, J = 6.8 Hz), 3.62–3.56 (m, 2H),
3.38–3.32 (m, 1H), 2.79 (s, 2H), 2.63 (t, 2H, J = 8.0 Hz), 2.28–2.15 (m, 3H), 2.09–
2.05 (m, 1H) 1.84–1.76 (m, 1H), 1.67–1.66 (m, 1H); ESMS (relative intensity)
558 (M+1) (100%); HPLC purity 98.77%.
12. C57BL/6J mice (7–8 weeks old) fed with normal chaw diet (carbohydrate 64%,
fat 7%, and protein 20%) were divided in following three groups: Group I:
vehicle, (n = 8); Group II: compound 38j, 30 mg/kg, (n = 8); and Group III:
compound 38j, 60 mg/kg, (n = 8). The mice were treated with the respective
treatment as mentioned above intraperitoneally, once a day for 8 days. On 7th
day of treatment the mice were kept in the oxymax cages for 24 h to
acclimatize and stabilize. After 24 h acclimatization, the mice were given the
respective treatment and then kept in oxymax chamber immediately for
recording O2 consumption and CO2 production. Initial 30-min observation was
not been considered in the analysis. This is the time taken by mice to stabilize
in the oxymax cages after injection. For determining RQ, the average of initial
8 h reading was calculated and then compared statistically using Student’s t
test.
13. Tipper, J. P.; Witters, L. A. Biochim. Biophys. Acta 1982, 715, 162.
14. Levert, K. L.; Waldrop, G. L.; Stephens, J. M. J. Biol. Chem. 2002, 277, 16347.
and 0.25
lg ACC enzyme (specific activity of 0.8–1 lmol/mg/min). The reaction
was carried out at 37 °C temperature for 30 min in presence of test compound