Proline-mediated formation of novel chroman-4-one tetrahydropyrimidines

Article abstract of DOI:10.1016/j.tet.2012.06.077

Novel tricyclic N-benzylated chroman-4-one tetrahydropyrimidine derivatives have been prepared through a multi-component reaction between various 2-substituted chroman-4-one derivatives, N-methylenebenzylamine and a catalytic amount of proline under mild reaction conditions. The tricyclic structure of 1a was determined by NMR spectroscopy and confirmed by X-ray crystallography. An additional product, 2a, was isolated from the reaction mixture and its structure and conformation were determined by a combination of theoretical (Monte Carlo conformational search) and NMR-based (NOE and ³J_HH couplings) conformational analysis. The NMR analysis revealed one preferred geometry for 1a and 2a in CHCl₃ solution.

Full text of DOI:10.1016/j.tet.2012.06.077

Tetrahedron 68 (2012) 7035e7040
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Proline-mediated formation of novel chroman-4-one tetrahydropyrimidines
a
a
b
a
a,c,
*
ꢀ
Maria Friden-Saxin , Tina Seifert , Lars Kristian Hansen , Morten Grøtli , Mate Erdelyi
,
,
Kristina Luthman a
*
a
€
Department of Chemistry and Molecular Biology, University of Gothenburg, SE-412 96 Goteborg, Sweden
b Department of Chemistry, University of Tromsø, NO-9037 Tromsø, Norway
c
€
Swedish NMR Centre, University of Gothenburg, SE-405 30 Goteborg, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 4 April 2012
Received in revised form 6 June 2012
Accepted 20 June 2012
Available online 28 June 2012
Novel tricyclic N-benzylated chroman-4-one tetrahydropyrimidine derivatives have been prepared
through a multi-component reaction between various 2-substituted chroman-4-one derivatives, N-
methylenebenzylamine and a catalytic amount of proline under mild reaction conditions. The tricyclic
structure of 1a was determined by NMR spectroscopy and confirmed by X-ray crystallography. An ad-
ditional product, 2a, was isolated from the reaction mixture and its structure and conformation were
determined by a combination of theoretical (Monte Carlo conformational search) and NMR-based (NOE
Keywords:
Multi-component reactions
Cyclization
Chroman-4-one
Proline
3
and JHH couplings) conformational analysis. The NMR analysis revealed one preferred geometry for 1a
and 2a in CHCl3 solution.
Ó 2012 Elsevier Ltd. All rights reserved.
b
-Turn
1. Introduction
position of the 2-alkyl-chroman-4-one scaffold. However, this re-
action did not result in the desired product, instead tricyclic chro-
Substituted chroman-4-ones are regarded as common struc-
tures for drug design.1 Along with related, higher order oxygen-
containing ring systems, they are frequently found in plants and
marine organisms2 and have been shown to possess antioxidant,3
antiviral4 and antibacterial5 activities. Protocols for the prepara-
tion of substituted chromones and chroman-4-ones have been
developed by our group with the most recent progress being the
incorporation of a carboxy functionality in the 6-position6,7 and an
amino group in the 3-position (Fig.1).7e9 Such chromone/chroman-
man-4-one derivatives were isolated. Herein, we report the
synthesis and conformational analysis of these novel chroman-4-
one tetrahydropyrimidines.
2. Results and discussion
As found in the present study a proline-mediated Mannich re-
action resulted in the unexpected formation of two novel tricyclic
derivatives 1 and 2 (Fig. 2). This fact indicates a yet unexplored
potential of the amino acid-catalyzed Mannich reaction. The new
compounds are expected to be of interest due to their high struc-
tural similarity to derivatives with significant pharmacological ac-
tivities.10,11 Examples of such compounds are the amine containing
tricyclic chroman-4-one analogues of 3, which exhibit anti-in-
flammatory12 and antiplatelet13 activities (Fig. 2) whereas the
Biginelli analogue 414 belongs to a type of structures shown to act as
calcium channel blockers.15
4-ones have been designed as potential b
-turn peptidomimetics.9
As an extension of these studies we explored the use of the
Mannich reaction to introduce an aminomethyl group in the 3-
2.1. Synthesis
Fig. 1. The 2,3,6,8-tetrasubstituted chromone system as a potential b-turn mimetic.
For the synthesis of the title compounds, the racemates of 8-
bromo-6-chloro-2-alkyl substituted chroman-4-ones 5aec
(Scheme 1) were used. They were synthesized via a microwave
promoted two-component reaction using commercially available
* Corresponding authors. Tel.: þ46 49 31 786 9031; fax: þ46 49 31 772 1394;
e-mail addresses: mate@chem.gu.se (M. Erdelyi), luthman@chem.gu.se
(K. Luthman).
0040-4020/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2012.06.077

ꢀ
M. Friden-Saxin et al. / Tetrahedron 68 (2012) 7035e7040
7036
Fig. 2. The novel tricyclic chroman-4-one derivatives 1 and 2 along with the struc-
turally related compounds 3 and 4, which were reported to exhibit promising phar-
macological activities.
3-bromo-5-chloro-2-hydroxyacetophenone and an aldehyde in the
presence of N,N-diisopropylamine (DIPA) in ethanol.8 As
L-proline is
known to efficiently catalyze asymmetric Mannich reactions16e18 it
was chosen for the corresponding reactions of derivatives 5aec.
Interestingly, reacting 5a with an excess of N-methylenebenzyl-
Fig. 3. The crystal structure of the tricyclic derivative 1a.
experiments using additional secondary amine sources such as
DIPA and pyrrolidine, which were shown to mainly react as nu-
cleophiles leading to ring opening of the chroman-4-one ring
(according to 1H NMR spectroscopic analysis of the crude reaction
amine19 (5 equiv) and a catalytic amount of
L-proline (0.3 equiv) in
DMSO at 50 ꢀC for 48 h afforded instead the novel tricyclic de-
rivative 1a in 52% yield (Scheme 1). The chroman-4-one 5b
substituted with a phenethyl group in the 2-position and 5c with
a considerably smaller 2-methyl substituent were also examined
(Scheme 1). Applying the identical reaction conditions, an excess of
N-methylenebenzylamine in DMSO in the presence of a catalytic
mixture). In addition, upon heating a mixture of L-proline and
chroman-4-one 1a at 50 ꢀC no enamine formation was observable
by 1H NMR spectroscopy. Hence, in the proposed mechanism the
enol of 5aec attacks the preformed N-methylenebenzylamine
providing the Mannich product as an intermediate. The subsequent
nucleophilic attack of the newly formed amino function on a sec-
ond N-methylenebenzylamine gives the aminal of which one
amino group attacks the carbonyl functionality in the chroman-4-
one. Subsequent dehydration provides the tetrahydropyrimidine
ring and thus the final product.
amount of -proline, the products 1b and 1c were formed but in
L
lower yields (26% and 15%, respectively) as compared to 1a. An
attempt to synthesize a derivative with a 2-phenyl substituent was
unsuccessful.
Scheme 1. The L-proline catalyzed formation of tricyclic derivatives 1aec.
The structure of 1a was determined by HMBC, HSQC and NOESY-
based NMR spectroscopic investigation and was confirmed by X-
ray crystallography (Fig. 3). Remarkably, the bulky substituent in
the 5-position prefers to adopt an axial orientation.
Compounds 1aec are formed via a Biginelli-type mechanism.15
Proline is suggested to catalyze the enolization of the chroman-4-
one (Scheme 2) instead of mediating enamine formation, which
Scheme 2. Proposed mechanism for the synthesis of derivatives 1aec.
was previously proposed for L
-proline.16 This conclusion is based on

ꢀ
M. Friden-Saxin et al. / Tetrahedron 68 (2012) 7035e7040
7037
In an attempt to optimize the yield of the tricyclic derivatives
1aec a series of reaction conditions were examined. The use of
smaller amounts of N-methylenebenzylamine, shorter reactions
times or higher temperatures (20 min or 2 h at 80, 120 or 150 ꢀC
under microwave irradiation) resulted in lower conversions. Simi-
lar observations were made upon variation of the chiral catalysts
(sarcosine,
L-pipecolic acid), the use of achiral catalysts (glycine,
DIPA, DIPEA or pyrrolidine), racemic catalyst (D/L-proline) or alter-
ation of solvents (THF or DMF). Neither the change of substrate
structure by removal of substituents or by introduction of electron
donating (OMe) or electron withdrawing (NO2 or Cl) groups in the
6-position of the chroman-4-one resulted in improved yields.
Further attempts on reacting 5a with electrophiles such as N-
methylene p-anisidine imine provided only traces of the Mannich
product along with numerous impurities. Using dibenzyl imine as
the electrophile resulted only in recovered starting material.
The isolated yields of derivatives 1a-c were moderate due to the
competing formation of additional heterocyclic products. For ex-
ample, the synthesis of 1a also yielded 2a in 7% isolated yield
(Scheme 3). As expected the formation of analogous products was
detected also in the synthesis of 1b and 1c (2b and 2c in 26% and
23% yields, respectively). Compound 2a was found to have identical
molecular weight to 1a, but showed a different 1H NMR spectrum
and chromatographic behaviour. Therefore additional HMBC and
NOESY-based NMR spectroscopic investigations were performed,
as described in detail below.
Scheme 4. A proposed mechanism for the formation of compounds 2aec.
Scheme 3. The L-proline-mediated formation of tricyclic derivatives 1aec and 2aec.
generated structures was performed using the OPLS-2005 all atom
force field23 and the Born solvation model for chloroform,24 as
implemented in the MacroModel program (v. 9.7).25
The mechanism for the formation of compounds 2aec is sug-
gested to occur via a nucleophilic attack by benzylamine on the
chroman-4-one ring system as shown in Scheme 4. Benzylamine is
most likely formed by partial hydrolysis of N-methylenebenzyl-
amine. However, using dry DMSO as the solvent and molecular
sieves (4 A) or MgSO4 as drying agents did not prevent the de-
composition of N-methylenebenzylamine and hence the formation
of the heterocyclic products 2aec.
The mixed torsions/low mode method was employed and con-
formations within 21 kJ/mol from the global minimum were kept.
This protocol yielded two distinct conformational families for 1a and
2a. Because of the uncertainty and force field dependency of the
energies of computationally derived conformations, the predicted
populations of theoretical ensembles do not necessarily represent
the molecular structure present in solution,20,26,27 yet allow their
determination when utilized in combination with experimental
data.20,22 Accordingly, the set of structures derived by the confor-
mational analysis was evaluated in a subsequent NAMFIS (NMR
analysis of molecular flexibility in solution) analysis.28 Distances
were determined by acquisition of NOE-buildups with five mixing
times (100, 150, 200, 250, and 400 ms) using the initial rate ap-
proximation,29 whereas scalar couplings were obtained from stan-
dard 1H and P.E. COSY30 spectra. Despite the few available protons on
the tricyclic backbone of 1a and 2a, a sufficient number of NOEs were
observed for description of the orientation of their flexible fragments
(Fig. 4). As enantiomeric mixtures yield a single set of NMR signals,
the conformational analysis of 1a and 2a was carried out without
chiral separation. Assignment of the diastereotopic CH2 protons was
based on relative NOE intensities and corresponded to the expected
ꢁ
2.2. Conformational analysis of 1a and 2a
Small molecules encompassing flexible bonds commonly exist
in solution as a mixture of rapidly interconverting conformers.20
Their solution structure usually cannot be correctly represented
by a single averaged structure, but is preferably described as the
probability-weighted ensemble of several conformations present in
solution. The determination of such ensembles is possible, although
not yet extensively carried out, by deconvolution of their time-
averaged spectroscopic data.21,22 For elucidation of the structure
of 2a as well as the available conformational space of the tricyclic
backbones of 1 and 2 a combined computational and NMR spec-
troscopic approach was utilized. A Monte Carlo conformational
search followed by molecular mechanics minimization of the

ꢀ
M. Friden-Saxin et al. / Tetrahedron 68 (2012) 7035e7040
7038
Fig. 4. NOE correlations observed in NMR spectra of 1a and 2a in chloroform. Two and
six additional J-couplings, respectively, are described in Supplementary data.
distances, when starting the identification from the chiral centres
(Fig. 4). The solution ensembles were determined by identification of
the geometries truly present in solution from the theoretically pre-
dicted conformational pool using experimental selection criteria.
Hence, 8 and 11 possibly time-averaged NMR-derived distances and
dihedral angles were used to deconvolute the conformational pool of
1a and 2a using the NAMFIS protocol. Details of the analysis, in-
cluding the comparison of the observed and the calculated distances
are given in Supplementary data.
The analysis of 1a indicated one preferred conformation in so-
lution. This geometry showed dihedral angles corresponding to
those observed in the solid state by X-ray analysis including also the
axially positioned 5-substituent (Fig. 5a).
Fig. 6. Alignment of a modified structure of the tricyclic derivative 6 and a type VIII
b
-turn (F(iþ1)¼ꢂ60ꢀ, j(iþ1)¼ꢂ30ꢀ, F(iþ2)¼ꢂ120ꢀ and j(iþ2)¼120ꢀ).31
proline-mediated enolization mechanism. Depending on the
identity of the 2-substituent of the chroman-4-one products were
obtained in varying yields. Combined NMR spectroscopic and the-
oretical conformational analysis of 1a and 2a revealed the presence
of a single geometry in solution, which for 1a was revealed to be
identical to its solid state (X-ray) structure. The obtained products
are of considerable interest due to their potential pharmaceutical
applicability. Their use as potential scaffolds for type VIII
peptidomimetics will be further explored.
b-turn
Compound 2a was revealed to also prefer a single conformation
(Fig. 5b) in which the large 4-substituent is equatorially oriented.
Thus, the refinement revealed that only one of the computationally
predicted conformational families exists in solution for 1a and 2a. It
should be underlined that application of the NAMFIS protocol en-
sures that the identified geometries are the real solution structures
and therefore are pharmacologically relevant.
4. Experimental section
4.1. General
Given our aim to use chroman-4-one/chromone scaffolds as
novel mimetics of bioactive peptides it was especially interesting to
find that the tricyclic cores of 1a and 2a efficiently mimic a native
Commercially available chemicals were used without prior pu-
rification. The reactions were monitored by thin-layer chromatog-
raphy (TLC) on silica plated aluminium sheets (Silica gel 60 F254, E.
Merck). Flash chromatography was performed on silica gel 60
(0.040e0.063 mm, manually or using a Biotage SP4 Flashþ in-
strument). Microwave reactions were carried out using a Biotage
InitiatorÔ Sixty with fixed hold time modus in 2e5 mL or
10e20 mL capped microwave vials. IR was recorded with a ChiralIR-
2xÔ from BioTools. Every compound was dissolved in 0.5 mL CDCl3.
High-resolution mass spectral analysis (Q-TOF-MS) was performed
at Stenhagen Analyslab AB, Gothenburg, Sweden. Elemental anal-
yses were performed at Kolbe Mikroanalytisches Laboratorium,
type VIII b-turn (Fig. 6; see the Experimental section for details on
the calculations). Their Cl- and Br-substitutions provide possibili-
ties for selective functionalization through Pd-mediated cross-
coupling reactions,6 and thereby allow broad applicability. How-
ever, any further investigations in this direction are outside the
scope of the present study.
3. Conclusions
A one-pot proline catalyzed synthetic route to novel chroman-4-
€
Mulheim and der Ruhr, Germany.
one tetrahydropyrimidine derivatives 1aec and 2aec has been
developed. The reactions are proposed to proceed through an L-
NMR spectra were recorded on JEOL GX-270 (400 MHz) or
Varian Unity Innova (800 MHz) spectrometer. Assignments of sig-
nals of derivatives 1a and 2a were made using HMBC, HSQC and
NOESY spectra. The samples were dissolved in CDCl3. Chemical
shifts are reported in parts per million with the solvent residual
peak as internal standard: CDCl3 [CHCl3 dH 7.26, CDCl3 dC 77.0]. The
NOE buildup studies were performed on 0.5 mmol/dm3 solutions at
mixing times of 100, 150, 200 and 250 and 400 ms. Distances were
ꢁ
calculated with a reference distance of 1.78 A for geminal protons.
NOE peak intensities were calculated using normalization of both
cross-peaks with both diagonal peaks according to ([xpeak1ꢁx-
peak2])/([diagpeak1ꢁdiagpeak2])0.5. Five mixing times yielding
Fig. 5. (a) The solution structure of the core of 1a (yellow) overlapped with its X-ray
derived conformation (green). (b) The solution conformation of the tricyclic core of 2a,
as identified by NAMFIS analysis.

ꢀ
M. Friden-Saxin et al. / Tetrahedron 68 (2012) 7035e7040
7039
a linear (r2>0.95) initial NOE rate were used to estimate the sij
sref sij
(1/6), where rij
is the distance between protons i and j and sij is the normalized
intensity obtained from NOE experiments. The 3JHH couplings were
derived from E.COSY experiments.
4-one 5a (0.100 g, 0.179 mmol, 1 equiv) was dissolved in DMSO
(2.5 mL). N-Methylenebenzylamine19 (0.106 g, 0.895 mmol,
5 equiv) and L-proline (6.18 mg, 0.054 mmol, 0.3 equiv) were
buildup rates according to the equation rij¼rref
(
/ )
added. The reaction mixture was stirred at 50 ꢀC for 48 h. The re-
action was quenched with NH4Cl (satd, aq) followed by the addition
of EtOAc. The phases were separated and the aqueous phase was
extracted with EtOAc. The combined organic phases were washed
three times with brine, dried over anhydrous MgSO4, filtered and
the solvent was removed under vacuum. Purification by flash
chromatography using a gradient of heptane/toluene (20%) gave 1a
(72 mg, 52%) and 2a (10 mg, 7%) as orange oils.
The computer based studies were performed using the Macro-
Model program (v. 9.7)25 as implemented in Maestro (v. 9.0). The
conformational search for 1a and 2a was performed using the
OPLS-2005 force field and the Born solvation model for chloroform.
The number of torsional rotations were restricted to 36 and the cut
ꢁ
off was set to 0.5 A. The conformational search was performed
using the mixed torsions/low mode method with 5000 steps. The
minimization method used was PRCG (Polak-Ribiere Conjugate
Gradient) with a maximum of 500 iterations. Conformations within
21 kJ/mol from the global minimum were retained. This resulted in
2000 (1a) and 1351 (2a) conformations, respectively. The confor-
mations were subsequently re-minimized using TNCG (truncated
Newton conjugate gradient) using the same criteria as described
above. The repeated minimization gave 162 (1a) and 349 (2a)
conformations, respectively, where conformations within 21 kJ/mol
from the global minimum were retained.
4.3.1.1. Compound 1a. Rf¼0.4 (5% EtOAc/toluene); IR 3160, 2362,
2337, 1822, 1794, 1453, 1375 cmꢂ1 1H NMR (400 MHz)
; d 8.03 (d,
J¼8.2 Hz, 1H), 7.75 (d, J¼8.2 Hz, 2H), 7.51e7.16 (m, 18H), 4.68e4.61
(m, 1H), 4.03e3.92 (m, 2H), 3.59e3.47 (m, 3H), 3.32 (d, J¼11.1 Hz,
1H), 3.19 (d, J¼16.9 Hz, 1H), 3.09e2.85 (m, 3H), 2.33 (s, 3H),
2.19e2.06 (m, 1H), 1.93e1.82 (m, 1H); 13C NMR (100 MHz)
d 148.2,
144.6, 138.2, 137.7, 135.3, 135.2, 134.7, 131.2, 130.8, 129.7 (2C), 128.8,
128.4, 128.3, 128.2, 127.3, 127.1, 126.6, 126.5, 124.6, 123.0, 122.7,
121.9, 121.4, 119.8, 119.4, 113.7, 111.3, 77.3, 67.3, 59.0, 54.4, 52.8, 32.8,
21.4, 20.8. Anal. Calcd for C42H37BrClN3O3S$EtOAc: C, 63.70; H, 5.23;
N, 5.06. Found: C, 64.10; H, 5.60, N, 5.06.
4.2. Synthesis of chroman-4-ones 5aec
4.2.1. 8-Bromo-6-chloro-2-(2-(1-tosyl-1H-indol-3-yl)ethyl)-chro-
4.3.1.2. Compound 2a. Rf¼0.5 (5% EtOAc/toluene); IR 3159,
man-4-one (5a). The chroman-4-one was synthesized according
3022, 2988, 2363, 2333, 1792, 1470, 1457, 1377 cmꢂ1 1H NMR
;
8
ꢀ
to the procedure reported by Friden-Saxin et al. To an ethanolic
(400 MHz)
d
7.99 (d, J¼8.3 Hz, 1H), 7.74 (d, J¼8.3 Hz, 2H), 7.44 (d,
solution (2.5 mL) of 30-bromo-50-chloro-20-hydroxyacetoph
enone (0.250 g, 1.002 mmol, 1 equiv) 3-(1-tosyl-1H-indol-3-yl)
propanal (0.146 mL, 1.10 mmol, 1.1 equiv) and DIPA (0.154 mL,
1.10 mmol, 1.1 equiv) were added. The reaction was run in a mi-
crowave cavity for 1 h at 170 ꢀC. The reaction mixture was di-
luted with Et2O and the phases were separated. The organic
phase was washed with NaOH (aq, 1%), HCl (aq, 0.1 M), water and
brine. The organic phase was dried over anhydrous MgSO4, fil-
tered and concentrated under vacuum. The obtained crude
product was purified by flash chromatography using EtOAc/
heptane (2.5%) yielding 5a as an orange oil (0.31 g, 88%) as
previously reported.8
J¼7.4 Hz, 1H), 7.39e7.13 (m, 17H), 4.74 (d, J¼13.5 Hz, 1H), 4.55 (d,
J¼13.5 Hz, 1H), 4.13e4.02 (m, 2H), 3.95 (d, J¼13.6 Hz, 1H), 3.85 (d,
J¼11.8 Hz, 1H), 3.57e3.44 (m, 2H), 3.21 (br s, 1H), 2.88e2.70 (m,
2H), 2.34 (s, 3H), 2.16e2.01 (m, 1H), 2.00e1.85 (m, 1H); 13C NMR
(100 MHz)
d 150.6, 144.7, 138.5, 138.1, 136.5, 135.3, 135.2, 131.2,
130.9, 129.8, 128.9, 128.5, 128.4, 128.0, 127.4, 127.2, 126.8, 126.6,
124.6, 123.2, 123.1, 123.0, 122.5, 121.9, 119.3, 117.0, 113.8, 110.8, 67.0,
64.7, 58.5, 57.1, 54.6, 30.5, 21.5, 19.8. Anal. Calcd for
42H37BrClN3O3S$H2O: C, 63.28; H, 4.93; N, 5.27. Found: C, 63.21; H,
4.99; N, 5.20.
C
4.3.2. 1,3-Dibenzyl-7-bromo-9-chloro-5-phenethyl-2,3,4,5-
tetrahydro-1H-chromeno[4,3-d]pyrimidine (1b) and 1,3-dibenzyl-7-
bromo-9-chloro-4-phenethyl-2,3,4,5-tetrahydro-1H-chromeno[4,3-d]
pyrimidine (2b). Chroman-4-one 5b (0.100 g, 0.275 mmol) was
reacted with N-methylenebenzylamine19 (0.163 g, 1.37 mmol) and
4.2.2. 8-Bromo-6-chloro-2-phenethylchroman-4-one
(5b). 30-
Bromo-50-chloro-20-hydroxyacetophenone (1.38 g, 5.54 mmol) was
reacted with 3-phenylpropanal (2.00 g, 6.10 mmol) and DIPA
(0.856 mL, 6.10 mmol) in ethanol (15 mL) following the general
procedure. Purification by flash chromatography using toluene/
heptane (50%) afforded 5b as a yellow solid (2.28 g, 74%) as pre-
viously reported.8
L
-proline (9.44 mg, 0.082 mmol) according to the general pro-
cedure. Purification by flash chromatography using EtOAc/heptane
(5:95/4:6) followed by toluene/heptane (3:7/1:1) as eluents
gave 1b (42 mg, 26%) and 2b (42 mg, 26%) as yellow oils.
4.2.3. 8-Bromo-6-chloro-2-methylchroman-4-one (5c). 30-Bromo-
50-chloro-20-hydroxyacetophenone (1.51 g, 6.05 mmol), acetalde-
hyde (0.373 mL, 6.61 mmol) and DIPA (0.932 mL, 6.61 mmol) were
reacted following the general procedure. Purification by flash
chromatography using EtOAc/heptane (1:9) gave 5c (0.47 g, 28%) as
a yellow oil. Rf¼0.77 (5% EtOAc/heptane); IR 3158, 2906, 2362,
4.3.2.1. Compound 1b. Rf¼0.7 (10% EtOAc/hexane); IR 3159,
3020, 2364, 2337, 1689, 1454 cmꢂ1; 1H NMR (400 MHz)
d 7.39e7.15
(m, 17H), 4.68e4.61 (m, 1H), 4.03e3.92 (m, 2H), 3.55e3.48 (m, 3H),
3.26e3.12 (m, 2H), 3.00e2.91 (m, 2H), 2.83e2.75 (m, 1H),
2.09e2.00 (m, 1H), 1.82e1.74 (m, 1H); 13C NMR (100 MHz)
d 148.4,
141.2, 138.3, 137.7, 134.6, 131.2, 128.8, 128.6, 128.5, 128.5, 128.4,
128.3, 127.3, 127.2, 126.5, 126.0, 122.7, 121.4, 119.9, 111.4, 77.5, 67.4,
59.1, 54.5, 52.8, 35.2, 31.7; HRMS (Q-TOF-MS) [MþH]þ, calcd for
2332, 1819 cmꢂ1
;
1H NMR (400 MHz)
d
7.80 (d, J¼2.6 Hz, 1H), 7.70
(d, J¼2.6 Hz, 1H), 4.72e4.62 (m, 1H), 2.78e2.65 (m, 2H), 1.59 (d,
J¼6.2 Hz, 3H); 13C NMR (100 MHz)
d
190.4,156.7, 138.4,126.8,125.8,
C
33H31BrClN2O: 585.1308, found: 585.1320.
122.1, 112.5, 75.4, 43.7, 20.7; HRMS (Q-TOF-MS) [MþH]þ, calcd for
C
10H9BrClO2: 274.9474, found: 274.9429.
4.3.2.2. Compound 2b. Rf¼0.6 (10% EtOAc/hexane); IR 3156,
3083, 3067, 3027, 2931, 2362, 2335, 1700 cmꢂ1; 1H NMR (400 MHz)
4.3. Synthesis of the tricyclic derivatives 1aec and 2aec
d
7.39e7.13 (m, 17H), 4.82 (d, J¼13.4 Hz, 1H), 4.65 (d, J¼13.4 Hz, 1H),
4.06e3.91 (m, 3H), 3.80 (d, J¼11.6 Hz, 1H), 3.51e3.40 (m, 2H),
3.24e3.20 (m, 1H), 2.91e2.68 (m, 2H), 2.16e2.05 (m, 1H), 1.97e1.86
4.3.1. 1,3-Dibenzyl-7-bromo-9-chloro-5-(2-(1-tosyl-1H-indol-3-yl)
ethyl)-2,3,4,5-tetrahydro-1H-chromeno[4,3-d]pyrimidine (1a) and
1,3-dibenzyl-7-bromo-9-chloro-4-(2-(1-tosyl-1H-indol-3-yl)ethyl)-
2,3,4,5-tetrahydro-1H-chromeno[4,3-d]pyrimidine (2a). Chroman-
(m, 1H); 13C NMR (100 MHz)
d 150.5, 142.3, 138.7, 138.1, 136.3, 131.1,
128.9, 128.5, 128.42 (2C), 128.40, 128.1, 127.2, 127.1, 126.7, 125.8,
123.1, 121.8, 117.6, 110.7, 67.0, 64.7, 58.7, 56.8, 54.6, 33.0, 30.7; HRMS

ꢀ
M. Friden-Saxin et al. / Tetrahedron 68 (2012) 7035e7040
7040
(Q-TOF-MS) [MþH]þ, calcd for C33H31BrClN2O: 585.1308, found:
Supplementary data. Crystallographic data (excluding structure fac-
tors) for the structure 1a in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publica-
tion CCDC 838242. Copies of the data can be obtained, free of charge,
on application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax:
þ44 (0)1223 336033 or e-mail: deposit@ccdc.cam.ac.uk].
585.1309.
4.3.3. 1,3-Dibenzyl-7-bromo-9-chloro-5-methyl-2,3,4,5-tetrahydro-
1H-chromeno[4,3-d]pyrimidine (1c) and 1,3-dibenzyl-7-bromo-9-
chloro-4-methyl-2,3,4,5-tetrahydro-1H-chromeno[4,3-d]pyrimidine
(2c). Chroman-4-one 5c (0.100 g, 0.363 mmol) was reacted with N-
methylenebenzylamine19 (0.216 g, 1.82 mmol) and
L-proline
Acknowledgements
(12.5 mg, 0.109 mmol) according to the general procedure. Purifi-
cation by flash chromatography using EtOAc/heptane (5/40%)
followed by toluene as the eluent afforded 1c (27 mg, 15%) and 2c
(41 mg, 23%) as yellow oils.
We thank the Swedish Research Council (Project # 2010-4868
and # 2007-4407) and the Department of Chemistry and Molecular
Biology, University of Gothenburg, for financial support.
4.3.3.1. Compound 1c. Rf¼0.5 (10% EtOAc/toluene); IR 3161,
Supplementary data
3081, 3030, 2363, 2338, 1692, 1457 cmꢂ1 1H NMR (400 MHz)
;
d
7.40e7.18 (m, 12H), 4.88 (q, J¼6.5 Hz, 1H), 4.01e3.91 (m, 2H),
NMR spectra of compounds 1aec, 2aec and 5c; results from
NAMFIS calculations of 1a and 2a; crystal data and refinement of
1a. Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.tet.2012.06.077.
3.56e3.47 (m, 3H), 3.34 (d, J¼11.0 Hz, 1H), 3.18 (d, J¼16.8 Hz, 1H),
3.06 (d, J¼16.8 Hz, 1H), 1.39 (d, J¼6.5 Hz, 3H); 13C NMR (100 MHz)
d
148.7, 138.3, 134.3, 131.1, 128.9, 128.5 (2C), 128.4 (2C), 128.3, 127.4,
127.1, 126.3, 121.3, 111.3, 99.9, 74.7, 67.3, 59.1, 54.5, 52.6, 19.4. Anal.
Calcd for C26H24BrClN2O$0.5H2O: C, 61.86; H, 4.99; N, 5.55. Found:
C, 61.91; H, 5.24; N, 5.49.
References and notes
1. Horton, D. A.; Bourne, G. T.; Smythe, M. L. Chem. Rev. 2003, 103, 893e930.
2. Hong, B. C.; Kotame, P.; Tsai, C. W.; Liao, J. H. Org. Lett. 2010, 12, 776e779.
3. Lee, H.; Lee, K.; Jung, J. K.; Cho, J.; Theodorakis, E. A. Bioorg. Med. Chem. Lett.
2005, 15, 2745e2748.
4. Yu, D. L.; Suzuki, M.; Xie, L.; Morris-Natschke, S. L.; Lee, K. H. Med. Res. Rev.
2003, 23, 322e345.
4.3.3.2. Compound 2c. Rf¼0.7 (5% EtOAc/toluene); IR 3159,
3074, 3028, 2973, 2930, 2826, 2359, 2336,1651,1559 cmꢂ1; 1H NMR
(400 MHz)
d
7.51e6.99 (m, 12H), 4.83 (d, J¼13.3 Hz, 1H), 4.64 (d,
J¼13.3 Hz, 1H), 3.93e3.84 (m, 2H), 3.76 (d, J¼13.6 Hz, 1H), 3.67 (d,
5. Cottiglia, F.; Dhanapal, B.; Sticher, O.; Heilmann, J. J. Nat. Prod. 2004, 67, 537e541.
J¼11.4 Hz, 1H), 3.53 (d, J¼13.6 Hz, 1H), 3.41e3.26 (m, 2H), 1.26 (d,
ꢀ
ꢀ
6. Dahlen, K.; Wallen, E. A.; Grøtli, M.; Luthman, K. J. Org. Chem. 2006, 71,
J¼6.5 Hz, 3H); 13C NMR (100 MHz)
d 150.4, 138.7, 138.3, 135.1, 131.1,
6863e6871.
128.9, 128.5, 128.4, 127.2, 127.1, 126.7, 123.4, 121.8, 120.6, 110.5, 99.9,
ꢀ
7. Ankner, T.; Friden-Saxin, M.; Pemberton, N.; Seifert, T.; Grøtli, M.; Luthman, K.;
67.6, 63.1, 55.3, 54.3, 54.2, 16.0. HRMS (Q-TOF-MS) [MþH]þ, calcd
Hilmersson, G. Org. Lett. 2010, 12, 2210e2213.
ꢀ
77
8. Friden-Saxin, M.; Pemberton, N.; Andersson, K. D.; Dyrager, C.; Friberg, A.;
for C
H BrClN2O: 495.0839, found: 495.0836.
26 25
Grøtli, M.; Luthman, K. J. Org. Chem. 2009, 74, 2755e2759.
ꢀ
ꢀ
9. Wallen, E. A. A.; Dahlen, K.; Grøtli, M.; Luthman, K. Org. Lett. 2007, 9, 389e391.
10. Skrabek, R. Q.; Galimova, L.; Ethans, K.; Perry, D. J. Pain 2008, 9, 164e173.
11. Simon-Levert, A.; Arrault, A.; Bontemps-Subielos, N.; Canal, C.; Banaigs, B. J. Nat.
Prod. 2005, 68, 1412e1415.
4.4. Conformational analysis of general type VIII
derivatives 6 and 2a
b-turns and
12. Bruno, O.; Schenone, S.; Ranise, A.; Bondavalli, F.; Filippelli, W.; Falcone, G.;
For evaluation of the potential applicability of these novel ring
systems as peptidomimetics their most stable conformations were
Motola, G.; Mazzeo, F. Farmaco 1999, 54, 95e100.
13. Bruno, O.; Schenone, S.; Ranise, A.; Bondavalli, F.; Barocelli, E.; Ballabeni, V.;
Chiavarini, M.; Bertoni, S.; Tognolini, M.; Impicciatore, M. Bioorg. Med. Chem.
2001, 9, 629e636.
14. Wang, Z. T.; Xu, L. W.; Xia, C. G.; Wang, H. Q. Tetrahedron Lett. 2004, 45, 7951e7953.
15. Kappe, C. O. Tetrahedron 1993, 49, 6937e6963.
16. Dalko, P. I.; Moisan, L. Angew. Chem., Int. Ed. 2001, 40, 3726e3748.
17. Verkade, J. M. M.; van Hemert, L. J. C.; Quaedflieg, P. J. L. M.; Rutjes, F. P. J. T.
Chem. Soc. Rev. 2008, 37, 29e41.
compared to that of various
initial analysis indicated that the tricyclic ring systems of 1a and 2a
mimic a type VIII -turn. The conformational analyses were made
b-turn conformations of peptides. An
b
on simplified structures of 1 denoted 6 (Fig. 6) and 2 (not shown)
with methyl groups as substituents in the 5- and 7-positions,
mimicking the possible orientation of attached chains. One nitro-
gen atom in the tetrahydropyrimidine ring was acetylated to cor-
respond to the N-terminus of a peptide and the C-terminal was
methylamidated. A low energy conformation of the novel ring
system in 6 obtained from the NAMFIS calculations was aligned
18. Gaunt, M. J.; Johansson, C. C. C.; McNally, A.; Vo, N. T. Drug Discov. Today 2007,
12, 8e27.
ꢀ
19. Axen, A.; Grennberg, H.; Gogoll, A. J. Chem. Res. 1998, 712e713.
20. Nevins, N.; Cicero, D.; Snyder, J. P. J. Org. Chem. 1999, 64, 3979e3986.
21. Erdelyi, M.; d’Auvergne, E.; Navarro-Vazquez, A.; Leonov, A.; Griesinger, C.
Chem.dEur. J. 2011, 17, 9368e9376.
22. Erdelyi, M.; Pfeiffer, B.; Hauenstein, K.; Fohrer, J.; Gertsch, J.; Altmann, K. H.;
Carlomagno, T. J. Med. Chem. 2008, 51, 1469e1473.
23. Kaminski, G. A.; Friesner, R. A.; Tirado-Rives, J.; Jorgensen, W. L. J. Phys. Chem.
2001, 105, 6474e6487.
24. Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T. J. Am. Chem. Soc. 1990,
112, 6127e6129.
25. MacroModel v. 9.7; Schrodinger: 2007; www.schrodinger.com.
26. Agrafiotis, D. K.; Gibbs, A. C.; Zhu, F. Q.; Izrailev, S.; Martin, E. J. Chem. Inf. Model.
2007, 47, 1067e1086.
27. Bonnet, P.; Agrafiotis, D. K.; Zhu, F. Q.; Martin, E. J. Chem. Inf. Model. 2009, 49,
2242e2259.
with an energy minimized type VIII
b-turn with the amino acid
sequence AceGly-Ala-Ala-GlyeNHMe as shown in Fig. 6.31,32 The
dihedral angles of the
b-turn VIII was constrained according to
previously defined angles (F(iþ1)¼ꢂ60ꢀ, j(iþ1)¼ꢂ30ꢀ, F(iþ2)¼ꢂ120ꢀ
and j(iþ2)¼120ꢀ).31 The type VIII
b-turn of the peptide was energy
minimized using OPLS-2005 as the force field in a simulated water
environment. The PRCG method was used with 500 iterations.
28. Cicero, D. O.; Barbato, G.; Bazzo, R. J. Am. Chem. Soc. 1995, 117, 1027e1033.
29. Stott, K.; Keeler, J.; Van, Q. N.; Shaka, A. J. J. Magn. Reson. 1997, 125, 302e324.
30. Griesinger, C.; Sorensen, O. W.; Ernst, R. R. J. Magn. Reson. 1987, 75, 474e492.
31. Hutchinson, E. G.; Thornton, J. M. Protein Sci. 1994, 3, 2207e2216.
32. Floquet, N.; Hery-Huynh, S.; Dauchez, M.; Derreumaux, P.; Tamburro, A. M.;
Alix, A. J. P. Biopolymers 2004, 76, 266e280.
33. CrystalClear Software User’s Guide; Rigaku Corporation, Molecular Structure
Corporation, (c): 2000; www.rigaku.com.
34. Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.; Burla, M. C.; Polidori,
G.; Camalli, M. J. Appl. Crystallogr. 1994, 27, 435e435.
4.5. X-ray diffraction
Crystals of 1a were obtained by recrystallization from EtOAc/
hexane (1:1). The crystal structure determination was carried out on
a Rigaku Saturn CCD area detector with graphite monochromated Mo
c and q
scans.33 The data were cor-
ꢁ
Ka
radiation (l¼0.71073 A) using
rected for Lorentz and polarization effects. The structure was solved
by direct methods and was refined by full-matrix least-squares on
F.34,35 Hydrogen atoms were refined using the riding model. The
crystal data and experimental parameters are summarized in
35. Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.; de Gelder, R.; Israel,
R.; Smits, J. M. M. The DIRDIF-99 Program System; University of Nijmegen: The
Netherlands, 1999.