Tricyclic thienopyridine-pyrimidones/thienopyrimidine-pyrimidones as orally efficacious mGluR1 antagonists for neuropathic pain

Article abstract of DOI:10.1016/j.bmcl.2009.04.104

Introduction of small unsaturated alkylamino groups at the 4-position of the A-ring of the tricyclic framework (triazafluorenone) afforded extremely potent and selective mGluR1 antagonists with desirable properties. Compounds 11q and 11s are active in the

Full text of DOI:10.1016/j.bmcl.2009.04.104

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3199–3203
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Tricyclic thienopyridine–pyrimidones/thienopyrimidine–pyrimidones
as orally efficacious mGluR1 antagonists for neuropathic pain
a
a
a
a
a
T. K. Sasikumar a, , Li Qiang , Duane A. Burnett , William J. Greenlee , Cheng Li , Larry Heimark ,
*
Birendra Pramanik a, Mariagrazia Grilli b, Rosalia Bertorelli b, Gianluca Lozza b, Angelo Reggiani b
a Schering-Plough Research Institute, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
b Schering-Plough Research Institute, San Raffaele Science Park, Via Olgettina, 58, 20132 Milan, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Introduction of small unsaturated alkylamino groups at the 4-position of the A-ring of the tricyclic frame-
work (triazafluorenone) afforded extremely potent and selective mGluR1 antagonists with desirable
properties. Compounds 11q and 11s are active in the SNL pain model with ED50s 3.3 and 6.4 mg/kg
respectively. Metabolic outcome of propargyl amino moiety was studied.
Received 12 March 2009
Revised 21 April 2009
Accepted 23 April 2009
Available online 3 May 2009
Ó 2009 Published by Elsevier Ltd.
Keywords:
Tricyclic thienopyridine–pyrimidone
Thienopyrimidine–pyrimidone
Triazafluorenone
mGluR1 antagonist
Neuropathic pain
Glutamate is the major excitatory neurotransmitter in the cen-
tral nervous system and mediates its actions via activation of both
ionotropic and metabotropic receptor families. The metabotropic
glutamate receptors (mGluRs) form a family of eight subtypes
(mGlu1 to mGlu8) and are assigned to three groups based on their
structure, coupling to effector mechanisms and pharmacology.
Group I mGluRs (mGluR1 and mGluR5) are post synaptic receptors
while Group II (mGluR2 and mGluR3) and Group III (mGluR6,
mGluR7 and mGluR8) are located presynaptically. It has been
found that the Group I mGluRs play key roles in the central sensi-
tization of pain and other neurologic disorders.1–6 Over the last
decade, there have been tremendous advances in the development
of small molecules that selectively activate or inhibit specific mGlu
receptor subtypes. Given below are some of the recent examples
from literature (Fig. 1).7–11
Compound 1 was identified from our high throughput screening
as a potent mGluR1 antagonist that was active in an in vivo model
for pain (rat spinal nerve ligation, SNL).12 A recent paper describing
the SAR around the tricyclic frame work was published.13 They
have not disclosed any unsaturated alkyl substitutions on the A-
ring of the tricyclic structure. Our attempt to replace the dimethyl
amino moiety was exceptionally fruitful.14 We found that small
unsaturated alkyls gave extremely potent mGluR1 antagonists.
SAR studies involving A-ring and N-aryl modifications and some
in vivo results are discussed in this Letter (Fig. 2).
Compounds of this type were synthesized according to Scheme
1. Commercially available ethylcyano malonate was condensed
with dimethylformamide–dimethylacetal (DMF–DMA) followed
by acid catalyzed cyclization gave pyridine derivative 4 in good
yields. Upon reaction with POCl3 on 4 gave 5 which was cyclized
to a common intermediate 6. Heating a solution of 6 in DMF with
OH
N
N
N
N
O
O
O
NC
O
EM-TBPC
(-)-CPCCOEt
H
O
O
O
O
N
O
BAY36-7620
JNJ16259685
* Corresponding author. Tel.: +1 908 740 4373; fax: +1 908 740 7164.
E-mail addresses: sasikumartk@yahoo.com, thavalakulamgar.sasikumar@
spcorp.com (T.K. Sasikumar).
Figure 1. Noncompetitive antagonists of mGluR1.
0960-894X/$ - see front matter Ó 2009 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2009.04.104

3200
T. K. Sasikumar et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3199–3203
R2
R3
R2
R3
S
N
N
N
OMe
N
N
C
N
NH2
X
N
N
a-c
A
N
B
S
R1
Y
N
N
O
N
R1
N
OMe
N
N
S
S
O
O
O
O
12
13 (a-z)
1
R1 = Aryl
R
2 = unsaturated alkyls
R3 = H or Me
Scheme 2. Reagents and conditions: (a) DMF–DMA, DMF, 100 °C; (b) 4-Cl-aniline
(R1 = Cl), HOAc, toluene, 110 °C; (c) R2R3NH, DMSO.
X = C or N; Y = C or N
Figure 2. General SAR plans.
Table 1
mGluR1 receptor binding for compounds 1, 11a–i
DMF–DMA afforded compound 7 which was cyclized to the tricy-
clic compound 8 in good yields. During this course of our studies,
we found that heating a solution of 6 with desired amine in tri-
ethylorthoformate in the presence of acetic acid gave the cyclized
compound in very high yields.15 Demethylation of 8 followed by
triflation of the phenol afforded 10 in good yields. Intermediate
10 was reacted with unsaturated alkylamines to afford the final
compounds in excellent yields. We used allyl, methallyl and prop-
argyl amines in this current SAR study. Similar chemistry was em-
ployed in the construction of thienopyrimidine–pyrimidone
analogs as shown in Scheme 2. The known amino ester 1216 was
converted to the final targets in three steps with good overall
yields. The final step is the displacement of methoxy group in
the pyrimidine nucleus with appropriate amine in polar solvents
such as DMSO to get compounds 13a–z.
The mGluR1 inhibitory potencies of the newly synthesized tri-
cyclic compounds are summarized in Table 1. Initially we turned
our attention to the allyl substitution on the A-ring of the pyridine
nucleus. Human mGluR1 IC50 and rat Ki are shown throughout in
this manuscript. Allylamino compounds 11a–i are generally well
tolerated as shown in Table 1. Compound 11a showed human
IC50 value of 2 nM with a rat Ki of 1.1 nM. Other substitution on
the aromatic ring such as 4-Me, 4-Br, and 4-F are very well toler-
ated. However the 4-chloro substitution (compound 11c) gave only
moderate activity on the human receptor but showed excellent po-
tency in the rat assay. Similar results were obtained with the 4-Br
derivative (11i). Disubstitution on the aromatic ring is well toler-
ated in this scaffold. N-Methylated compounds such as 11g and
11h are also tolerated. It has been found that all these tricyclic
derivatives are inactive in the mGluR5 assay.
R2
N
N
R1
N
N
S
O
11 (a-i)
Compd
R1
R2
h-mGluR1 IC50 (nM)
r-mGluR1 Kia (nM)
a
1
9.5
2
11
7.9
1.1
2.7
1.1
0.7
14
7.7
5.4
4.1
3.2
11a
11b
11c
11d
11e
11f
11g
11h
11i
4-MeO-Ph
4-Me-Ph
4-Cl-Ph
4-Br-Ph
4-F-3-MeO-Ph
4-F-Ph
4-Cl-Ph
4-MeO-Ph
4-Br-Ph
H
H
H
H
H
H
60
12.6
6.8
4.4
1.1
6.9
66
Me
Me
Me
a
The IC50 and Ki data are an average of at least three measurements, performed
on human mGlu1/5 and rat mGlu1 receptors, respectively. The standard error was
10%, and variability was less than twofold from assay to assay. h-mGluR5
IC50 > 3 lM.
exhibited subnanomolar activity in rat mGluR1 assay. Methallyl
substitution was also very effective in the pyrimidine series (Table
3). Simple phenyl substitution (compound 13f)17 provided good
human affinity but a moderate rat Ki value. It was interesting to
note that the 4-pyridyl analog displayed poor activity at both the
human and rat receptors. Unlike in the case of pyridine analogs,
disubstitution in the pyrimidine series is very well tolerated. For
example, compound 13l, showed human IC50 of 8.9 nM and rat Ki
of 2.3 nM.
Next, we studied the effect of substitution on the allyl group.
Methallyl derivatives generally afforded active analogs with 2–3-
fold less mGluR1 affinity as shown in Table 2. Disubstitution on
the aromatic ring (compounds 11n, 11o and 11p) generally de-
creases the human inhibitory potency. Compounds 11k and 11l
Replacement of the dimethylamino group of our lead com-
pound with propargylamine afforded an extremely potent com-
pound 11q with human IC50 0f 0.9 nM and a rat Ki of 0.6 nM as
NC
CN
OMe
OMe
OMe
OMe
N
NC
CN
NH2
N
CN
Cl
CN
OH
e
a
c
d
b
MeO
O
O
N
EtO
N
N
S
N
S
N
N
O
O
7
6
2
4
5
3
g
f
R2
R3
S
N
OMe
OTf
OH
N
N
N
h
j
i
N
N
O
N
N
R1
R1
R1
N
N
N
N
R1
S
S
S
O
O
O
11 (a-x)
10
9
8
Scheme 1. Reagents and conditions: (a) DMF–DMA, MeOH, 80 °C; (b) HOAc (80%), 130 °C; (c) POCl3, Et3N, 100 °C; (d) HSCH2CO2Me, NaOMe, DMF, 80 °C; (e) DMF–DMA, DMF,
100 °C; (f) 4-Cl-aniline (R1 = Cl), HOAc, toluene, 110 °C; (g) 4-Cl-aniline (R1 = Cl), HOAc, CH(OEt)3, 160 °C; (h) BBr3, CH2Cl2; (i) NPhTf2, CH2Cl2; (j) R2R3NH, DMSO.

T. K. Sasikumar et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3199–3203
3201
Table 2
Table 4
mGluR1 receptor binding for compounds 11j–p
mGluR1 receptor binding for compounds 11q–x
R2
N
NH
N
N
R1
N
R1
N
N
S
N
S
O
O
11 (q-x)
11 j-p
Compd
R1
R2
h-mGluR1 IC50 (nM)
r-mGluR1 Kia (nM)
a
Compd
R1
h-mGluR1
IC50 (nM)
r-mGluR1
a
Kia (nM)
11q
11r
11s
11t
11u
11v
11w
11x
4-Cl-Ph
H
H
H
H
0.9
1.5
2.1
1.8
12.5
3.8
0.6
1.6
3.5
0.7
28
18
27
8
4-Me-Ph
4-MeO-Ph
4-Br-Ph
3-F-4-MeO-Ph
4-Cl-Ph
11j
11k
11l
11m
11n
11o
11p
4-Cl-Ph
4-Me-Ph
4-MeO-Ph
4-Br-Ph
3-F-4-MeO-Ph
10.6
16.4
10.4
43
103
196
241
1.2
0.2
0.4
1.5
13
H
Me
Me
Me
4-MeO-Ph
4-Br-Ph
10.8
1.9
3-(2,3-Dihydrobenzo[b] [1,4]Dioxin-6-yl)
3-(Benzo[d][1,3]dioxol-5-yl)
13
14
a
The IC50 and Ki data are an average of at least three measurements, performed
on human mGlu1/5 and rat mGlu1 receptors, respectively. The standard error was
10%, and variability was less than twofold from assay to assay. h-mGluR5
a
The IC50 and Ki data are an average of at least three measurements, performed
on human mGlu1/5 and rat mGlu1 receptors, respectively. The standard error was
10%, and variability was less than twofold from assay to assay. h-mGluR5
IC50 > 3 lM.
IC50 > 3 lM.
Table 3
Table 5
mGluR1 receptor binding for compounds 13a–m
mGluR1 receptor binding for compounds 13n–z
NH
NH
N
N
N
N
R1
N
R1
N
N
S
N
O
S
O
13 (n-z)
13 (a-m)
Compd R1
h-mGluR1 IC50
(nM)
r-mGluR1 Kia
(nM)
a
Compd
R1
h-mGluR1
IC50 (nM)
r-mGluR1
a
Kia (nM)
13n
13o
13p
13q
13r
13s
13t
4-Cl-Ph
4-Me-Ph
4-MeO-Ph
4-Br-Ph
3-F-4-MeO-Ph
4-F-Ph
2-F-4-MeO-Ph
3-(2,3-Dihydrobenzo[b]
[1,4]dioxin-6-yl)
3-(Benzo[d][1,3]dioxol-5-yl)
3-(Benzo[d]thiazol-5-yl)
3-(Benzo[d]thiazol-6-yl)
3-(Benzofuran-5-yl)
3.0
3.9
13
6.8
212
8.2
85
4.4
8.3
25
29
77
39
32
160
13a
13b
13c
13d
13e
13f
13g
13h
13i
4-Cl-Ph
4-Me-Ph
4-MeO-Ph
4-Br-Ph
3-F-4-MeO-Ph
Ph
4-Py
4-F-Ph
3-Cl-Ph
7.7
12.4
9.6
14.5
72
4.6
955
10.7
53
16
17
8.9
33
2.2
2.9
7.5
3.8
68
25
1000
38
62
43
49
2.3
64
13u
106
13v
13w
13x
13y
13z
137
45
209
21
997
65
107
63
13j
13k
13l
3-(Benzo[d]thiazol-5-yl)
3-(Benzo[d]thiazol-6-yl)
3-(Benzo[b]thiophen-5-yl)
13m
3-(2,3-Dihydrobenzo[b] [1,4]dioxin-6-yl)
3-(Benzo[b]thiophen-5-yl)
219
69
a
a
The IC50 and Ki data are an average of at least three measurements, performed
on human mGlu1/5 and rat mGlu1 receptors, respectively. The standard error was
10%, and variability was less than twofold from assay to assay. h-mGluR5
The IC50 and Ki data are an average of at least three measurements, performed
on human mGlu1/5 and rat mGlu1 receptors, respectively. The standard error was
10%, and variability was less than 2-fold from assay to assay. h-mGluR5 IC50 > 3 lM.
IC50 > 3 lM.
shown in Table 4. 4-Me, 4-MeO, and 4-Br substitution on the N-aryl
group afforded similarly highly potent compounds. N-Methylation
lowered affinity 4–5-fold in all cases except 4-BrPh derivative
(11x).
Similar results were obtained in the A-ring pyrimidine series as
evidenced from the Table 5. Generally, mono substitution at the
para substitution of the N-aryl ring is generally well tolerated.
Compounds 13n, 13o, 13q and 13s attained single digit nanomolar
at the human mGluR1 assay. However disubstituted compounds
such as 13x and 13z exhibited a large decrease in potency as evi-
denced from Table 5.
of these types of molecules. Data for representative compounds
11q and 11s are shown in Table 6.
Compounds 11q and 11s showed moderate rat AUC with high
brain to plasma ratio as demonstrated in Table 6. These com-
pounds were inactive in the hERG assay. It has been found that
compounds 11q and 11s were active in the in vivo pain model
(rat spinal nerve ligation, SNL)12 with ED50 values of 3.3 and
6.4 mg/kg, respectively, when dosed orally. The metabolic pathway
in rat for high clearance compound 11q was further investigated
using microsome incubation. The metabolic details are shown in
Figure 3. The metabolism of acetylenic compounds commonly used
in the formulation of pharmaceuticals have been investigated pre-
viously using 13C NMR and mass spectrometry.19,20 Compound 11q
Having achieved excellent potency against mGluR1 receptors,
we shifted our attention toward measuring the pharmacokinetics

3202
T. K. Sasikumar et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3199–3203
OH
NH
NH
NH2
S
N
N
N
[O]
O
N
N
N
+
H
N
N
N
Cl
Cl
Cl
S
S
O
O
O
(11q) C18H11ClN4OS (RT = 14.1 Min)
(14) C18H11ClN4O2S (RT = 13.0 Min)
(15) C15H9ClN4OS (RT = 11.3 Min)
(16)
Mol. Wt.: 366.82
Mol. Wt.: 382.82
Mol. Wt.: 328.78
OH
O
OH
NH
OH
H2O
H
NH
N
N
N
N
N
N
S
Cl
Cl
S
O
O
(18)
(17) C18H13ClN4O3S (RT = 3.9 Min)
Mol. Wt.: 400.84
Figure 3. Metabolism of propargyl group (Mass analysis of rat bile samples, 0–24 h).
Table 6
PK Profile and in vivo activity of selected compounds
R1
R2
NH
N
α
NH
N
N
N
N
S
R
N
O
S
O
19; R1 = Me, R2 = H; h-IC50 = 314 nM
20; R1 = Me, R2 = Me; h-IC50 = 171 nM
21; R1 = Et, R2 = Et; h-IC50 = 358 nM
Parameters
R = Cl (11q)
R = OMe (11s)
Human mGluR1 IC50 (nM)
Rat mGluR1 Ki (nM)
Human mGluR5 IC50 (nM)
0.9
0.6
2.1
3.5
>3000
3.3 mg/kg po
>3000
6.4 mg/kg po
12
Figure 4. Effect of a-alkylation on human mGluR1 potency.
Rat SNL ED50
Caco-2 permeability
Efflux substrate
430 nm/s
No
427
67
690 nm/s
No
682
257
2.5
Rat PK, (10 mg/kg), AUC (ng.h/mL)18
Brain conc. @ 6 h (ng/g)
Brain/Plasma
seen in compounds 11q and 11s. These types of compounds show
moderate PK and high brain plasma ratio. It has also been found
that these compounds are inactive in the hERG assay. Further data
will be published elsewhere.
1.8
Clearance (mL/min/kg)
Bioavailability (%)
30
16
22
34
Acknowledgments
We thank Drs. Deen Tulshian, Julius Matasi and Peter Korakas
for helpful discussions. We also thank Lisa Broske’s group for ani-
mal dosing and Sam Wainhaus’ group for PK sample analysis.
was incubated with rat liver microsome for 0–24 h and the metab-
olites were identified by LC–MS methodology. The oxidation of the
carbon adjacent to the amine lead to the intermediate 14 followed
by degradation to the final molecules 15 and 16. LC–MS peak at
13 min with an observed mass of 383 is attributed to the interme-
diate 14. LC–MS analysis showed a peak at 3.9 min. with a mass of
401 could be water addition product (17) to intermediate 14 as
shown in Figure 3. The parent amino compound 15 displayed in
LC–MS at 11.3 min. This retention time of the amine 15 was inde-
pendently confirmed via analysis of pure synthetic sample. Com-
pound 15 was found to be the major metabolite in the rat bile.
The side product of this whole sequence, compound 16, is a known
glutathione scavenger described in the literature.19
In order to avoid the metabolic issues, we introduced methyl
groups adjacent to the amino group. Unfortunately monomethyla-
tion and dimethylation of our lead compounds afforded 200–300-
fold less active compounds (19–21) as shown in Figure 4. Identical
results were obtained for compounds with different substitution
on the right hand side aromatic ring.
References and notes
1. Higgins, G. A.; Miczek, K. A. Psychopharmacology 2005, 179, 1.
2. Kew, J. N. C.; Kemp, J. A. Psychopharmacology 2005, 179, 4.
3. Steckler, T.; Lavreysen, H.; Oliveira, A. M.; Aerts, N.; Van Craenendonck, H.;
Prickaerts, J.; Megens, A.; Lesage, A. S. J. Psychopharmacology 2005, 179, 198.
4. Niswender, C. M.; Jones, C. K.; Conn, P. J. Curr. Top. Med. Chem. 2005, 5, 847.
5. Kunishima, N.; Shimada, Y.; Tsuji, Y.; Sato, T.; Yamamoto, M.; Kumasaka, T.;
Nakanishi, S.; Jingami, H.; Morikawa, K. Nature 2000, 407, 971.
6. Constantino, G.; Macchiarulo, A.; Pellicciari, R. Bioorg. Med. Chem 2001, 9, 847.
7. Layton, M. E. Curr. Top. Med. Chem 2005, 5, 859.
8. Ott, D.; Floersheim, P.; Inderbitzin, W.; Stoehr, N.; Francotte, E.; Lecis, G.;
Richert, P.; Rihs, G.; Flor, P. J.; Kuhn, R.; Gasparini, F. J. Med. Chem. 2000, 43,
4428.
9. Malherbe, P.; Kratochwil, N.; Knoflach, F.; Zenner, M.-T.; Kew, J. N. C.;
Kratzeisen, C.; Maerki, H. P.; Adam, G.; Mutel, V. J. Biol. Chem. 2003, 278, 8340.
10. Carroll, F. Y.; Stolle, A.; Beart, P. M.; Voerste, A.; Brabet, I.; Mauler, F.; Joly, C.;
Antonicek, H.; Bockaert, J.; Müller, T.; Pin, J. P.; Prézeau, L. Mol. Pharmacol. 2001,
59, 965.
11. Lavreysen, H.; Wouters, R.; Bischoff, F.; Pereira, S. N.; Langlois, X.; Blokland, S.;
Somers, M.; Dillen, L.; Lesage, A. S. J. Neuropharmacology 2004, 47, 961.
12. Varty, G. B.; Grilli, M.; Forlani, A.; Fredduzzi, S.; Grzelak, M. E.; Guthrie, D. H.;
Hodgson, R. A.; Sherry, X. L.; Nicolussi, E.; Pond, A. J.; Parker, E. M.; Hunter, J. C.;
Higgins, G. A.; Reggiani, A.; Bertorelli, R. Psychopharmacology 2005, 179, 207.
In summary, we have achieved a large number of single digit
nanomolar mGluR1 antagonists in the tricyclic series. This excel-
lent potency sometimes extrapolates to good in vivo efficacy as

T. K. Sasikumar et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3199–3203
3203
13. Recently, scientists from Abbott Laboratories disclosed similar mGluR1
antagonists based on the tricyclic triazafluorenone core structure. See:
Zheng, G. Z.; Bhatia, P.; Daanen, J.; Kolasa, T.; Patel, M.; Latshaw, S.;
Kouhen, O. F. E.; Chang, R.; Uchic, M. E.; Miller, L.; Nakane, M.; Lehto, S.
G.; Honore, M. P.; Moreland, R. B.; Brioni, J. D.; Stewart, A. O. J. Med.
Chem. 2005, 48, 7374.
14. Our efforts in the modification of tricyclic structure is being published in the
following patents, WO 2006/002051 and US 2006/167029. Also see: Wu, W-L.;
Burnett, D. A.; Domalski, D.; Greenlee, W. J.; Li, C.; Bertorelli, R.; Fredduzzi, S.;
Lozza, G.; Veltri, A.; Reggiani, A. J. Med. Chem 2007, 50, 5550. Biology materials
and methods were described in this Letter.
15. Typical one step synthesis of tricyclic compound is described as follows: Compound
6 (2 g, 0.0084 mol) was suspended in 15 mL triethyl orthoformate and treated
with acetic acid (2 mL) and 4-chloroaniline (2 g, 0.0156 mol). The contents
were heated in a sealed tube at 160 °C for 16 h. The reaction mixture was
cooled to room temperature and the precipitated product was washed several
times with ether. The product was dried in vacuo to get 2.5 g of compound 8
(R1 = Cl) as white solid. 1H NMR (CDCl3): d 8.65 (d, 1H), 8.29 (s, 1H), 7.56 (d,
2H), 7.40 (d, 2H), 6.93 (d, 1H), 4.17 (s, 3H). Mass Spectrum (M+1): m/z calcd for
C16H11ClN3O2S+ = 344.03, found m/z = 344.2.
16. Clark, J.; Shahhet, M. S. J. Heterocycl. Chem. 1993, 30, 1065.
17. Spectral data for compound 13f: 1H NMR (CDCl3): d 8.62 (s, 1H), 8.27 (s, 1H),
7.91 (s, 1H), 7.58 (m, 3H), 7.45 (m, 2H), 4.99 (s, 1H), 4.93 (s, 1H), 4.32 (d, 2H),
1.86 (s, 3H). Mass Spectrum (M+1): m/z calcd for C18H16N5OS+ = 350.11, found
m/z = 350.2.
18. Korfmacher, W. A.; Cox, K. A.; Ng, K. J.; Veals, J.; Hsien, Y.; Wainhaus, S.; Broske,
L.; Prelusky, D.; Nomeir, A.; White, R. E. Rapid Commun. Mass. Spectrom. 2001,
15, 335.
19. Banijamali, A. R.; Xu, Y.; DeMatteo, V.; Strunk, R. J.; Sumner, S. J. J. Agric. Food
Chem. 2000, 48, 4693.
20. DeMaster, E. G.; Sumner, H. W.; Kaplan, E.; Shirota, F. N.; Nagasawa, H. T.
Toxicol. Pharmacol. 1982, 65, 390.