Synthesis of aminal-type Lilium candidum alkaloids and lilaline; determination of their relative configuration by the concerted use of NMR spectroscopy and DFT conformational analysis

Article abstract of DOI:10.1016/j.tet.2020.131827

We hereby report the synthesis of six racemic alkaloids isolated from Lilium candidum L. Their common structural feature is a five-membered lactam ring which is, in the case of the flavonoid alkaloid lilaline, attached to the molecule's aromatic core, while in the case of the other five compounds, it is connected to the nitrogen atom of a pyrrolinone ring by an aminal function. The syntheses of these natural products were achieved via Mannich-type alkylations through cyclic N-acyliminium ions as intermediates. Besides the synthesis, the so far unexplored stereochemistry of these natural products was determined by a combination of NMR-based proton–proton distance measurements and theoretical conformational analyses carried out at the DFT level.

Full text of DOI:10.1016/j.tet.2020.131827

Tetrahedron 81 (2021) 131827
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis of aminal-type Lilium candidum alkaloids and lilaline;
determination of their relative configuration by the concerted use of
NMR spectroscopy and DFT conformational analysis
a
,
1
b
,
,
c
,
*
b
b
Sandor Nagy , Aron Szigetvari 1, Viktor Ilkei a , Balazs Kramos , Zoltan Beni ,
ꢀ
b
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
a
ꢀ
ꢀ
ꢀ
Csaba Szantay Jr. , Laszlo Hazai
a Department of Organic Chemistry and Technology, Budapest University of Technology and Economics, Budafoki út 8, Budapest, H-1111, Hungary
Spectroscopic Research Department, Gedeon Richter Plc., Gyomroi út 19-21, Budapest, H-1103, Hungary
Medicinal Chemistry Laboratory I, Gedeon Richter Plc., Gyomroi út 19-21, Budapest, H-1103, Hungary
b
€
}
c
€
}
a r t i c l e i n f o
a b s t r a c t
Article history:
We hereby report the synthesis of six racemic alkaloids isolated from Lilium candidum L. Their common
structural feature is a five-membered lactam ring which is, in the case of the flavonoid alkaloid lilaline,
attached to the molecule’s aromatic core, while in the case of the other five compounds, it is connected to
the nitrogen atom of a pyrrolinone ring by an aminal function. The syntheses of these natural products
were achieved via Mannich-type alkylations through cyclic N-acyliminium ions as intermediates. Besides
the synthesis, the so far unexplored stereochemistry of these natural products was determined by a
combination of NMR-based protoneproton distance measurements and theoretical conformational an-
alyses carried out at the DFT level.
Received 16 September 2020
Received in revised form
27 November 2020
Accepted 29 November 2020
Available online 6 December 2020
Keywords:
Iminium chemistry
Synthesis of lily alkaloids
Determination of relative configuration
Conformational analysis
1He1H distance measurements by NMR
© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
1. Introduction
Wiedhopf and co-workers [3]. Their initial structural conclusion
(the 4-Me analogue of 1 was postulated by the authors) was later
Lilium candidum L. e commonly known as the Madonna lily e is
a perennial plant in the true lily family. It has a high-growing leafy
stem and it bears white, strongly fragrant flowers in the summer.
The Madonna lily is native to the Balkans and to the Middle East, but
it has also become naturalized in other parts of the world [1]. In folk
medicine the extract of the plant is used for the treatment of burns,
ulcers, inflammation and for healing wounds [2].
Several interesting heterocyclic compounds have been isolated
from Lilium candidum L. (Fig. 1). Structurally, the largest subclass of
the isolated lily alkaloids contains an aminal function or a 3-
pyrrolin-2-one moiety (Fig. 1, highlighted in blue). Despite their
biological relevance, most members of this special family of lily
alkaloids have not yet been synthesized.
revised by Yakushijin et al. [4] and was further verified by two
distinct syntheses [5,6]. Jatropham (1) was found to show anticar-
cinogenic [3,7] and cell protecting [8] activity. Methyljatropham (2)
and ethyljatropham (3) were isolated as racemates from Lilium
hansonii [9] and from Lilium candidum [10], respectively. However,
the natural origin of these compounds is questionable since their
isolation involved extractions by the appropriate alcohols, meth-
anol or ethanol.
Continuing the phytochemical characterization of Lilium candi-
ꢀ
dum, Haladova and co-workers later isolated the pyrroline-
pyrrolidine alkaloids 4, 5 and 7 [11], the N-pyrrolidonyl derivative
of jatropham, 6 [12], and the jatropham dimers 8 and 10 [13].
Although most of these compounds contain two stereogenic cen-
ters, neither the isomeric composition, nor the stereochemistry of
the isolated products was discussed in these papers. Our literature
survey showed that the stereochemistry was only explored in the
case of compound 5, where an (R,R) relative configuration was
suggested on the basis of X-ray analysis [14]. No synthesis of these
dimeric pyrroline-alkaloids has been published so far, with the
Jatropham (1) was first isolated from Jatropha macrorhiza by
* Corresponding author. Medicinal Chemistry Laboratory I, Gedeon Richter Plc.,
€
}
Gyomroi út 19-21, Budapest, H-1103, Hungary.
E-mail address: viktor.ilkei@gmail.com (V. Ilkei).
These authors contributed equally.
1
https://doi.org/10.1016/j.tet.2020.131827
0040-4020/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Fig. 1. Pyrroline alkaloids and flavonoid constituents isolated from Lilium candidum L. (3-pyrrolin-2-one moieties are highlighted in blue).
exception of jatropham dimer 8, whose formation was observed by
Yakushijin et al. upon the acid treatment of an N-protected deriv-
ative of 1 [5].
enabled the efficient synthesis of the targeted lily alkaloids from
inexpensive, commercially available starting materials.
In addition, the 4-methylpyroglutamic acid derivative of
jatropham (9) [15] and jatropham glycosides 11 [9], 12 [15] and 13
[16] were isolated from Lilium hansonii, Lilium martagon and other
related lily species that belong to this structural class.
The flavonoid alkaloid lilaline (14), together with its flavonoid
core, kaempferol (15) and its methylsuccinyl derivative 16 (most
probably the biogenetic precursor of 14), were also isolated from
Lilium candidum [17e19]. The stereochemistry of 14 has not been
clarified in the literature either.
Continuing our investigations in the field of natural product
synthesis through iminium chemistry [20], we endeavored to
synthesize some of these interesting heterocyclic compounds,
focusing primarily on the pyrrolineepyrrolidine type dimers (4e8)
and the flavonoid alkaloid lilaline (14). In addition, the so far un-
determined stereochemistry of compounds 6e8 and 14 prompted
us to explore at least their relative configurations using NMR
spectroscopy complemented by molecular modeling.
The synthetic strategy we applied here is similar to the method
we recently implemented to effectively introduce a substituted 2-
pyrrolidon-5-yl group onto nucleophiles, such as heteroatoms or
activated aromatic rings [20]. The method involves the synthesis of
the appropriate N-acylaminocarbinol derivative, which is then
allowed to react with the nucleophile of our choice in a Mannich-
type alkylation reaction [21]. These reactions usually require acid
catalysis in order to enable the in situ conversion of the electrophile
precursor to an N-acyliminium ion that can attack the nucleophile
to form the desired compounds. The application of this strategy
2. Results and discussion
The retrosynthetic concept of the targeted compounds is out-
lined in Scheme 1. Five reactants are required for the synthesis of
the six alkaloids, demonstrating the versatility of these building
blocks in the synthesis of lily alkaloids. The building blocks are 3-
methyl-3-pyrrolin-2-one (17), 5-ethoxypyrrolidin-2-one (18), 5-
hydroxy-3-methylpyrrolidin-2-one (19), ( )-jatropham [5-
hydroxy-3-methyl-3-pyrrolin-2-one] (1) and the flavonoid
kaempferol (15) (Scheme 1), which is the only building block
available at a reasonable price.
The synthesis of the four pyrroline and pyrrolidine building
blocks was achieved in a few steps from cheap and easily accessible
starting materials. 3-Methyl-3-pyrrolin-2-one (17) was synthesized
in 4 steps from pyrrole (Scheme 2). In the first step the nitrogen
atom of pyrrole (20) was silylated with the bulky triisopropylsilyl
(TIPS) group to yield TIPS-pyrrole (21) [23]. This enabled for-
mylation in the subsequent Vilsmeier reaction in the C3 position
[23,24]. The work-up resulted in desilylation, yielding pyrrole-3-
carboxaldehyde (22), which was then converted to 3-
methylpyrrole (23) in a Wolff-Kishner reduction [25]. Finally,
oxidation of 23 b y aqueous hydrogen peroxide [26] gave the
desired lactam 17 in medium yield.
5-Ethoxypyrrolidin-2-one (18) was synthesized in a single step
from succinimide (24) via partial reduction by sodium borohydride
and a subsequent one-pot etherification under acidic conditions
(Scheme 3), following the procedure described earlier [22,20d].
2

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Scheme 1. Retrosynthetic concept of target compounds 4e8 and 14.
Scheme 2. Synthesis of 3-methyl-3-pyrrolin-2-one (17).
observed with the C-5 aminal carbon (80.0 ppm) proving the
structure of 1. In the case of the synthesis of the saturated N-acy-
laminocarbinol 19, 26 was first reduced by catalytic hydrogenation
to ( )-2-methylsuccinimide (27) [29], which was then further
reduced with DIBALH at 0 ꢀC to give an inseparable mixture of 5-
hydroxy-3-methylpyrrolidin-2-one (19) and its 2-hydroxy isomer
28. In contrast to the DIBALH reduction of 26 under similar con-
ditions, the regioselectivity of the reduction of 27 was much lower,
the resulting mixture contained 19 and 28 in an approximate ratio
of 2.7:1. The regioisomers were identified based on HMBC data. In
the case of 19, the methyl group showed correlation with the C-2
carbonyl group while no correlation to the C-5 aminal-type carbon
atom was observed. In contrast, in compound 28, the methyl group
gave correlation to the C-5 aminal-type carbon atom while no
correlation to the C-2 carbonyl group was detected. Despite the
presence of significant amounts (ca. 27%) of the isomeric N-acyla-
minocarbinol 28 in the obtained isomeric reagent mixture, it could
be utilized in the syntheses of the targeted lily alkaloids.
Scheme 3. Synthesis of 5-ethoxypyrrolidin-2-one (18).
The cyclic N-acylaminocarbinol reagents 1 and 19 were syn-
thesized from citraconic anhydride (25) in 2 and 3 steps, respec-
tively (Scheme 4). Citraconic anhydride (25) was first converted to
citraconimide (26) with hexamethyldisilazane (HMDS) in DMF at
100 ꢀC [27], which was reduced with diisobutyl aluminium hydride
(DIBALH) in THF at 0 ꢀC to yield ( )-jatropham (1) [28]. The
regioselectivity of the reduction was confirmed by NMR data. In the
HMBC spectrum the methyl group (1.83 ppm) showed correlation
to the C-2 carbonyl group (175.5 ppm) while no correlation was
3

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Scheme 4. Synthesis of the cyclic N-acylaminocarbinols 1 and 19.
With the four pyrroline-pyrrolidine building blocks in hand, we
synthesized the desired lily alkaloids by the reaction of the two
appropriate compounds in THF in the presence of catalytic amounts
of p-toluenesulfonic acid. The reactants were employed in equi-
molar amounts and the reactions proceeded to full conversion at
room temperature in 1 h. The reactions gave the targeted lily al-
kaloids 4e8 as the major products in medium to high yields
(Scheme 5). In the case of the synthesis of 5 and 7 the presence of
the isomeric N-acylaminocarbinol 28 in the reagent mixture beside
the major component 19 was not taken into account, however the
reactions still furnished the desired natural products in relatively
good yields. The explanation for this observation probably lies in
the preferential reaction of the nucleophile with 19 over its isomer
28, due to the aminal function being present in the sterically more
hindered site in the latter compound. And even though one of the
multiple minor spots visible beside the major product on the TLC of
the rather complex reaction mixtures could possibly belong to a
side product arising from 28, these multicomponent reaction
mixtures seemed to contain so many minor and trace products, that
no effort was made to isolate and characterize these substances.
In the case of 4 and 8, the precipitated crystalline products could
be isolated by a simple filtration, while in the case of 5 and 6 the
products were crystallized after evaporation of the solvent. The
pyrroline alkaloid 7 was purified by column chromatography.
Compound 4 was isolated as a single racemic compound, while in
the case of 5e8 due to the presence of multiple stereogenic centers
the synthesized lily alkaloids were all isolated as racemic mixtures
of diastereomers in different compositions (Scheme 5). These
mixtures were subjected to detailed NMR spectroscopic studies.
The mechanism of these aminal-forming Mannich-type alkyl-
ation reactions involves the in situ generation of cyclic N-acylimi-
nium ions from their N-acylaminocarbinol precursors via
protonation of the carbinol oxygen and the subsequent loss of an
ethanol (in the case of reagent 18) or water (in the case of 19 or 1)
molecule. This is followed by the electrophilic attack of the N-
acyliminium ion on the slightly nucleophilic lactam-nitrogen of the
other reactant. This mechanism and the possible side reactions are
presented through the example of the synthesis of lily alkaloid 7
(Scheme 6).
reactants as well. In this hypothetical side reaction 1 is converted
into the N-acyliminium ion 30, which in turn could attack the ni-
trogen atom of 19, giving the hypothetical “reverse alkylation”
product. Likewise, an N-acyliminium ion could also attack the ni-
trogen atom of its N-acylaminocarbinol precursor to yield a “self-
alkylation” product. Besides these side reactions, the formation of
larger adducts or oligomers composed by one or both reactants in
different ratios can be envisaged as well.
Despite these theoretical possibilities, a strong preference for
the formation of the expected products was observed in all cases.
Considering the formation of 7 as an example (Scheme 6), the
reason behind this preference e assuming that both reactive imi-
nium species (29 and 30) are present e most probably lies in the
reactivities of the similar species, meaning that 29 is a better
electrophile than 30 while 1 is a better nucleophile than 19. Sub-
sequently, the predominant reaction pathway turns out to be the
one leading to the formation of 7, resulting in the desired lily
alkaloid as the major product.
Despite the prevalence of the aforementioned reaction route,
some of the potentially arising side products could possibly be
present in the reaction mixtures. Although in most cases multiple
spots were visible on TLC beside the major products, due to the
rather complex reaction mixtures, no effort was made to isolate
minor or trace products.
The synthesis of lilaline (14) was carried out via the reaction of
kaempferol (15) and a slight excess (1.2 eq.) of 19 at reflux tem-
perature in 2 h in the presence of a catalytic amount of p-tolue-
nesulfonic acid and gave the target compound (14) in high yield
(Scheme 7). Even though the formation of the C6-isomer via C6-
amidoalkylation of kaempferol can also be hypothesized based on
our previous experience in the field [20c,d], only the C8-isomer
(lilaline, 14) could be isolated from the reaction mixture. This
high regioselectivity is in accordance with our earlier finding [20d],
showing that phenolic Mannich reactions [30] of other 5,7-
dihydroxylated flavonoids proceed with high regioselectivity as
well. Lilaline (14) was purified by preparative HPLC and was ob-
tained as a diastereomeric mixture. The ratio of the major trans
diastereomer and the minor cis isomer in the final product was
found to be approx. 3:1 (according to the 1H NMR data).
The saturated cyclic N-acylaminocarbinol 19 is transformed into
the N-acyliminium ion 29, which attacks the nitrogen atom of 1,
yielding the desired lily alkaloid 7 as the major product. However,
both reactants are N-acylaminocarbinol type compounds, thus,
theoretically the “reverse alkylation” is possible from the same
2.1. Determination of the relative configuration of compounds 6, 7
and 8
Firstly, the elemental composition and the constitution of all
4

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Scheme 5. Synthesis of lily alkaloids 4e8 and the diastereomeric composition of the synthesized products.
synthetic products were elucidated using high resolution mass
spectrometry and standard NMR spectroscopic methods (1H, 13C
NMR, 1He1H COSY, 1He1H NOESY or ROESY, 1He13C HSQC and
1He13C HMBC). Based on the spectral data complete 1H and 13C
NMR characterization of each stereoisomer of the expected prod-
ucts could be achieved. Regarding the stereochemistry, the relative
configuration of compounds 5 and 14 could easily be deduced on
the basis of Nuclear Overhauser Effect (NOE) data (protons or
functional groups that are situated on the same face of a ring are
likely to attain spatial proximity; the spatial proximity of protons
gives rise to strong peaks in the NOESY spectra, see Fig. 2). For both
5 and 14, the 3,5-trans isomers were found to be the major
products.
Determining the stereochemistry of 6, 7 and 8 was more chal-
lenging, because in these cases the chirality centers in the two rings
are connected by a freely rotating CeN bond (Fig. 3).
As is well known, for two protons whose distance varies in time
due to a fast conformational motion, the magnitude of the observed
1He1H NOE is a function of the conformational energy profile of the
system [31]. Thus, without knowledge of the pertinent
5

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Scheme 6. A plausible mechanism of aminal-forming Mannich-type alkylations and the possible side reactions presented through the example of the synthesis of lily alkaloid 7.
Scheme 7. Synthesis of the flavonoid alkaloid lilaline (14).
conformational profiles of the possible configurational isomers of
compounds 6, 7 and 8, and without knowing that there is a suitably
large difference between the profiles associated with the configu-
rational isomers, the relative configurations cannot be determined
reliably from the measured 1He1H NOEs. A comparison of the
corresponding selective inversion 1D NOESY spectra showed that
the different stereoisomers gave rise to the same resonances but
with substantially different intensities, which suggested that such a
marked configuration-dependent difference in the conformational
energy profiles exists, and so the relative configurations can be
determined if the NOE measurements are complemented by a
molecular modeling study. As shown by literature examples
[32e34] the QM-based calculation, followed by a sophisticated
statistical analysis of the calculated 1H and 13C NMR chemical shifts
with respect to the experimental ones might have been another
approach to assign the relative configurations of 6, 7, and 8. How-
ever, in these cases larger differences were obtained in the NOE
intensities than between the chemical shifts (see SI) of the di-
astereoisomers. This has prompted us to use the NOE derived dis-
tances as reference in our modeling study.
We note in passing that for compound 7 the determination of
the relative configuration of C-30 and C-5’ was straightforward from
routine 2D NOESY spectra.
Running selective inversion 1D NOESY experiments with
appropriate choice of the experimental parameters (application of
low-viscosity solvents, choice of relatively short mixing times
6

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
mixture of acetone-d6 and methanol-d4 [12], 7: acetone-d6 [11], and
8: methanol-d4 [13]), since their viscosity (acetone: 3.2∙10ꢁ4 Pa s,
methanol: 5.6∙10ꢁ4 Pa s at 300 K[37]) was suitable for the NOE
measurements. To maximize the signal-to-noise ratio of the 1D
NOESY spectra, which was a critical feature when acquiring the
spectra of the minor diastereomers, we chose the upper bound of
the mixing times (600 ms) that Butts et al. recommended to use
[35], since the NOE intensity is proportional to the mixing time in
the early section of the NOE buildup curve [31]. From the various
pulse sequences known for distance measurements [38], we chose
a single pulse field gradient spin-echo, zero-quantum filtered se-
lective inversion 1D NOESY sequence (selnogpzs.2), which was part
of the spectrometer’s standard pulse sequence library.
For the analysis of the stereoisomeric mixtures of the lily alka-
loids obtained in the syntheses we took advantage of the large
resolving power of our 800 MHz NMR spectrometer, instead of
resorting to the separation of these diastereomers (e.g. by prepar-
ative HPLC). This approach not only saved time and effort but
Fig. 2. Diagnostic NOE correlations used to determine the relative configuration of
compounds 5 and 14.
Fig. 3. Stereoisomers of compounds 6, 7 and 8.
relative to the T1 relaxation times) enable us to determine inter-
proton distances with high accuracy. Using the conventional no-
tation, let S denote the selectively inverted proton, and let I be any
other proton in the molecule on which we observe the NOE as a
result of having inverted S. According to Butts et al., in such an
experiment the NOE at spin I is to a very good approximation
proportional to the negative sixth power of the distance dI,S [31]
between I and S; the margin of error for the distance measurement
is ca. 4% [35]. Jones et al. were able to achieve the same margin of
error of interproton distance measurements in a flexible molecule
(4-propylaniline) that can adopt several well-defined low-energy
conformations [36].
ensured that for all isomers in a sample the 1D NOESY measure-
ments were carried out under exactly the same measurement
conditions (pH, temperature, molarity of the components, etc.). In
this way, any biases in distance measurements due to differences in
sample preparation and other experimental conditions could be
avoided, which was critical for a reliable comparison of the data.
The measured 1D NOESY intensities are reported in Tables 1e3;
the intensities are given relative to that of the selectively inverted
(ꢁ100%) resonance (in all cases that was the CH(OH) proton at
position 5 in the 1-alkyl-3-pyrrolin-2-one ring). The experimental
distances were determined by the following equation
!
For the interproton distance measurements of 6, 7 and 8, we
used solvents or solvent mixtures that were identical to the ones
used by the authors who provided the spectral characterization of
the lily alkaloids isolated from natural sources (6: equal volume
ꢁ6
hI;S
dI;S ¼ dRef ;S
,
hRef ;S
where I is the investigated proton, S is the selectively inverted
Table 1
Measured NOE intensities and estimated interproton distances in the stereoisomers of 6.
6a (major isomer)
Relative intensity
Estimated distance
6b (minor isomer)
Relative intensity
Estimated distance
Inverted proton
Observed protons
H-5 (
H-4 (
d
d
5.54)
6.67)
5.51)
ꢁ100%
1.063%
0.391%
0.896%
e
Inverted proton
Observed protons
H-5 (
H-4 (
d
d
5.48)
6.68)
5.73)
ꢁ100%
1.032%
0.340%
0.140%
e
2.69 Å (calib.)
3.18 Å
2.77 Å
2.69 Å (calib.)
3.24 Å
3.75 Å
H-5’ (
d
H-5’ (
d
Hb -4’ (
d
2.35)
Hb -4’ (
d
2.38)
The ratio of stereoisomers 6a:6b was ca. 79:21.
7

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Table 2
Measured NOE intensities and estimated proton-proton distances in the stereoisomers of 7.
7a (major „trans” isomer)
Relative intensity Estimated distance 7b (minor „trans” isomer)
Relative intensity Estimated distance
Inverted proton
Observed protons
H-5* (
H-4 (
H-5’ (
b -4’ (
d
5.57) ꢁ100%
6.63) 0.612%
5.42) 0.324%
2.54) 0.562%
e
Inverted proton H-5 (
Observed protons H-4 (
d
d
5.48)
6.64)
ꢁ100%
e
d
2.69 Å (calib.)
2.99 Å
2.73 Å
0.586%
2.69 Å (calib.)
3.33 Å
3.88 Å
d
H-5’ (
d
5.62) 0.163%
2.56) 0.065%
H
d
Hb -4’ (
d
7c (major „cis” isomer)
Inverted proton
Observed protons
Relative intensity Estimated distance 7d (minor „cis” isomer)
Relative intensity Estimated distance
H-5#
H-4 (
(
d
d
~5.63) (overlapped)
6.64) 0.298%
5.54) 0.164%
2.17) 0.393%
e
Inverted proton H-5#
Observed protons H-4 (
(
d
d
~5.64) (overlapped)
6.65) 0.137%
5.68) 0.040%
2.39) 0.027%
e
2.69 Å (calib.)
2.97 Å
2.57 Å
2.69 Å (calib.)
3.30 Å
H-5’ (
d
H-5’ (
d
Hb -4’ (
d
Hb -4’ (
d
3.53 Åþ
“Cis” and “trans” refer to the relative configuration in the pyrroline-2-one ring (C-30 and C-50).
The ratio of stereoisomers 7a:7b:7c:7d was ca. 53:35:7:5.
1
*Due to overlap of the signals in the H NMR spectrum we could not avoid inverting H-50 of 7c.
#
Due to spectral overlap we could only invert H-5 of 7c, H-5 of 7d and H-50 of 7b together.
þSince the intensity of the detected NOE signal is close to the detection limit, the error of quantitation might be larger than in the other cases.
Table 3
Measured NOE intensities and estimated proton-proton distances in the stereoisomers of 8.
8a (major isomer)
Relative intensity
Estimated distance
8b (minor isomer)
Relative intensity
Estimated distance
Inverted proton
Observed protons
H-5 (
H-4 (
H-5’ (
H-4’ (
d
d
5.40)
6.64)
5.96)
6.76)
ꢁ100%
1.187%
0.343%
0.747%
e
Inverted proton
Observed protons
H-5 (
H-4 (
H-5’ (
H-4’ (
d
d
5.25)
6.67)
6.11)
6.78)
ꢁ100%
1.172%
0.199%
0.133%
e
2.69 Å (calib.)
3.31 Å
2.91 Å
2.69 Å (calib.)
3.61 Å
3.87 Å
d
d
d
d
The ratio of stereoisomers 8a:8b was ca. 65:35 at the time of the 1D NOESY measurements.
proton, dRef,S is a reference distance (2.69 Å, calculated for H-5 e H-
the calculation of averaged distances using the Boltzmann pop-
ulations, the experimentally estimated distances could not be
reproduced well enough due to the poorly sampled conformational
space. Thus, a relaxed rotational scan around the C-5’ e N1 bond
was performed using a step size of 10ꢀ (except for compound 8,
where a step size of 5ꢀ was applied) to better sample the confor-
mational space. Since the rotation of the hydroxyl group could
significantly perturb the relative energies of the geometry, two
torsional scan calculations were finally carried out in each case.
These were started from geometries with “opposite” hydroxyl
group positions (the hydroxyl groups can be positioned as shown in
Fig. 4 and rotated by about 180ꢀ). The interproton distances were
calculated by averaging the appropriate values of all optimized
geometries (2 ꢂ 36 structures in each case, 2 ꢂ 72 for compound 8a
and 8b, according to their Boltzmann populations derived from the
relative potential energies. These ensemble-averaged distances are
listed in Table 4.
4 by molecular modeling) for calibration, and
h is the measured
relative signal intensity in the 1D NOESY spectrum. The decisive
interproton distances were H-5 e H-50 and H-5 e H-40, i.e. the
distances between the protons belonging to different rings.
A comparison of the estimated distances of the stereoisomers
showed that in the case of 6 and 7 the H-5 e Hb-40 distance, while in
the case of 8 the H-5 e H-4’ distance, is consistently shorter by
about 1 Å in the so-called “major” isomers than in the “minor”
isomers. One angstrom difference, which corresponds to ca. 30%, is
significant and can be used for the stereochemical assignment. In
order to interpret these substantial differences in the critical dis-
tances theoretical calculations were undertaken.
Firstly, conformational search was performed at the MM
(OPLS3) level for all stereoisomers of compounds 6e8 (only a single
enantiomer, in which the configuration of C-50 was arbitrarily taken
to be R, was considered in all cases). The lowest energy conformers
(using the default MacroModel setting, an MM energy window of
5 kcal/mol) were optimized at the QM level (B3LYP-D3/6-31þG**).
The global minima are shown in Fig. 4. However, if only the local
minima of the potential energy surface were taken into account in
By fixing the configuration of C-50, the configuration of the C-5
center affected mainly the H-5 e H-40 distance. This was found to
be significantly higher when the configurations of these two cen-
ters were identical (R) in the case of all three compounds.
8

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Fig. 4. Lowest energy conformers of the selected enantiomers optimized at the B3LYP-D3/6-31þG** level.
Table 4
Experimentally estimated and calculated proton-proton distances for the selected enantiomers.
Compound
Isomer
major
Configuration (30) e 5’ e 5
R e S
Distance
Exp. Dist [Å]
Calc. Dist.[Å]
Deviation (Calc. e Exp.) [Å]
Relative error (Deviation/Calc.)
6a
H-5 e H-40
H-5 e H-50
H-5 e H-40
H-5 e H-50
H-5 e H-40
H-5 e H-50
H-5 e H-40
H-5 e H-50
H-5 e H-40
H-5 e H-50
H-5 e H-40
H-5 e H-50
H-5 e H-40
H-5 e H-50
H-5 e H-40
H-5 e H-50
2.77
3.18
3.75
3.24
2.73
2.99
3.88
3.33
2.57
2.97
3.53
3.30
2.91
3.31
3.87
3.61
2.40
3.34
3.55
3.54
2.41
3.32
3.54
3.50
2.29
3.41
2.92
3.52
2.82
3.53
3.68
3.68
ꢁ0.37
0.16
ꢁ15%
þ5%
6b
7a
minor
R e R
ꢁ0.20
ꢁ6%
0.30
þ8%
major trans
minor trans
major cis
minor cis
major
R e R e S
R e R e R
S e R e S
S e R e R
R e S
ꢁ0.32
0.33
ꢁ13%
þ10%
ꢁ10%
þ5%
7b
7c
ꢁ0.34
0.17
ꢁ0.28
0.44
ꢁ12%
þ13%
ꢁ21%
þ6%
7d
8aa
8ba
ꢁ0.61
0.22
ꢁ0.09
0.22
ꢁ3%
þ6%
minor
R e R
ꢁ0.19
ꢁ5%
0.07
þ2%
a
Step size in the rotational scan was 5ꢀ.
Qualitatively, this can be explained by comparing the appropriate
geometries of the lowest energy conformers of the epimeric pairs,
as presented in Fig. 4. If the relative position of the pyrroline and
pyrrolidine rings of the steroisomers is the same, H-5 occupies
“opposite” sides in the stereoisomers, which renders H-5 signifi-
cantly closer to H-40 in the 5 S,50R isomers. On this basis the relative
configurations of the stereoisomers (Table 4) could already be
assigned with high confidence.
A comparison of the ensemble-averaged (and experimentally
estimated) distances showed that the H-5 e H-50 distances are
systematically overestimated (ca. 0.2 Å, 5e13%), while the H-5 e H-
4’ distances are systematically underestimated (ca. 0.3 Å, 6e21%) by
the calculations.
most probably due to two factors. First, the conformational space
sampling of the molecules is still not accurate enough. (This is
supported by the significantly lower (2e6%) deviations obtained in
the case of 8a and 8b, where twice as many conformers (a step-size
of 5ꢀ instead of 10ꢀ in the torsional scan) were taken into account
for the distance calculation.) Secondly, due to the H-bond forming
capability, a protic and polar solvent such as methanol may
significantly affect the conformational space of the studied mole-
cules as well. This however, cannot be taken into account by the
continuum solvent models (in our case the Poisson Boltzmann
Finite element method (PBF) based calculations). The importance of
solvent models when dealing with the accurate DFT calculations of
chemical shifts has been highlighted recently in the literature
[39,40].
From the modeling perspective these systematic deviations are
9

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
From an experimental perspective, the application of different
NMR solvents (used in order to obtain NMR data directly compa-
rable to the ones reported in the literature), the omission of the so-
called three-spin effect (the presence of a third, nearby proton can
compound.
A similar comparison to the reported NMR data [12,13] in the
case of 6 (Table 6) and 8 (Table 7) led to the conclusion that the
compounds isolated from natural sources have the same relative
configurations as our major synthetic stereoisomers, (5S*,50R*)-5-
hydroxy-3-methyl-1-(20-oxopyrrolidin-50-yl)-3-pyrrolin-2-one
(6a), and (5S*,50R*)-5-hydroxy-3-methyl-1-(30-methyl-20-oxo-30-
pyrrolin-50-yl)-3-pyrrolin-2-one (8a). The measured 1H NMR
chemical shifts and coupling constants and the values reported in
the literature differed from each other by less than 0.01 ppm and
0.2 Hz, respectively. The other synthetic diastereomers (6b and 8b)
are new compounds.
corrupt the h
f dꢁI,S6 rule), as well as the integration errors caused by
low-level components of the mixtures, may all contribute to the
observed deviations between the theoretical and experimental
distance values.
Nevertheless, the relative configurations of the C-5/C-50 centers
in the different stereoisomers of the compounds could be confi-
dently determined on the basis of interproton distance measure-
ments and conformational analysis. Since the deviation between
the calculated and experimentally determined distances was of
systematic nature, the difference obtained for the H-5 e H-40 dis-
tance in the case of the epimeric species could be unambiguously
interpreted (compounds 6 and 8). In the case of 7 (possessing three
stereogenic centers), the problem was divided into two sub-
problems: the relative configuration of C-30 and C-50 (cis and trans)
were known from 2D NOESY data prior to conformational analysis;
subsequently, the decision between 7a and 7b, as well as between
7c and 7d, i.e. the relative configuration of C-5 and C-50, could be
assessed with the aid of the calculations.
In the case of compound 7 however, ambiguities were found
between the reported [11] and observed spectral data (Table 8). In
particular the chemical shifts of H-4, C(5)-OH and H-10 reported by
ꢀ
Haladova et al. were ca. 0.4 ppm downfield of those observed for
any of the synthetic isomers. It should be noted however that the
reported chemical shifts are inconsistent with the data reported for
close structural analogues in the same publication as well. Due to
this discrepancy, in the case of compound 7 the relative configu-
ration of the real lily alkaloid can only be given speculatively. By
accepting the plausible reasoning that the inconsistency in the
NMR data arises simply from the use of a different NMR solvent, a
tentative assignment can be given. Based on the reported similar
coupling constants, an identical relative configuration of C-30 and
C-50 can be suggested in the natural product as it was determined in
lilaline 14. Assuming that the formation of 7 is governed by a
similar metabolic pathway that leads to 6 in the plants, the relative
configuration of C-5 and C-50 is likely to be the same. On this basis
the naturally occurring alkaloid can most probably be described by
the same relative stereochemistry as our main synthetic stereo-
isomer 7a (Fig. 5): rel-(5S,30R,50R)-5-hydroxy-3-methyl-1-(30-
methyl-20-oxopyrrolidin-50-yl)-3-pyrrolin-2-one. The other di-
astereomers of 7 are new synthetic derivates; they have not been
reported in the literature.
To the best of our knowledge, our method of choice for the
determination of the relative configuration of small molecules is
new, or at least uncommon in the sense that we took into account
the full conformational space of the investigated molecule, rather
than trying to describe the conformational space with several
selected low-energy conformers.
In this way we were able to calculate the ensemble-averaged
NOEs (and interproton distances) with a higher accuracy, and we
could reliably decide among the diastereomers of the studied
molecules.
2.2. Assigning the relative configuration of lily alkaloids
The application of identical solvents (or solvent mixtures)
enabled the direct comparison of our NMR data (1H chemical shifts
and coupling constants) of the synthetic products with those re-
ported for lily alkaloids isolated from natural sources [11e14,17].
With regard to 4 and 5 the measured 1H NMR chemical shifts
agreed well with the literature data [11]. A comparison of the
1He1H coupling constants (cf. Ref. [17]) reported for lilaline and
determined for the two diasteromers of 14 (Table 5) showed that
the natural product possesses the same relative configuration as
our major synthetic product, (300R*,500S*)-14 (Fig. 5) with the
functional groups of the pyrrolinone ring having a trans arrange-
ment; the coupling constants in our dataset and the dataset in the
literature differed from each other by less than 0.2 Hz. The
(300R*,500R*)-14 diastereomer, which shows a completely different
coupling pattern in the 1H NMR spectrum, is a new synthetic
3. Conclusion
In the present work we accomplished the synthesis of five
racemic pyrroline alkaloids as well as the flavonoid alkaloid lilaline,
which were all isolated earlier from Lilium candidum L. These
compounds were synthesized from cheap, easily accessible starting
materials via Mannich-type alkylations employing cyclic N-acyla-
minocarbinols as electrophile precursors. The synthesized lily al-
kaloids were isolated as racemates or mixtures of diastereomers in
different compositions. The relative configuration of some lily al-
kaloids was determined by the concerted use of NMR-based
interproton distance measurements and conformational analyses
at the DFT level. To the best of our knowledge, our method of choice
for the determination of the relative configuration of flexible
Table 5
Comparison of the literature 1H NMR data for 14 with our data for 14a and 14b.
1H NMR data
Lilium candidum L.
Synthetic products
literature data [17]
(300R*,500S*)-14 ¼ trans-14 (major, 76%)
(300R*,500R*)-14 ¼ cis-14 (minor, 24%)
Position
H-6
dH (ppm) (apparent multiplicity; estimated J coupling constants in Hz)
6.24 (s)
6.26 (s)
6.25 (s)
7.98 (m)
6.85 (m)
7.83 (br)
2.48 (ddq; 11.5, 8.4, 7.0)
1.05 (d; 7.0)
1.94 (ddd; 11.9, 11.5, 10.0)
2.34 (ddd; 11.9, 8.4, 6.6)
5.26 (dd; 10.0, 6.6)
H-20, H-60
H-30, H-50
NH-100
8.01 (d; 9.0)
7.95 (m)
6.91 (d; 9.0)
6.89 (m)
e
7.67 (br)
H-300
C(300)-CH3
H2-400
2.77 (dqd; 9.7, 7.4, 4.9)
1.29 (d; 7.4)
2.17 (ddd; 12.9, 8.9, 4.9)
2.57 (ddd; 12.9, 9.7, 5.9)
5.56 (dd; 8.9, 5.9)
2.54 (dqd; 9.6, 7.4, 5.0)
1.13 (d; 7.4 Hz)
1.99 (ddd; 12.7, 8.8, 5.0)
2.38 (ddd; 12.7, 9.6, 5.9)
5.35 (dd; 8.8, 5.9)
H-500
10

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
Fig. 5. Relative configurations of lily alkaloids determined by NMR and molecular modeling.
Table 6
Comparison of the literature 1H NMR data for 6 with our data for 6a and 6b.
1H NMR data in acetone-d6: methanol-d4 ¼ 1:1
Lilium candidum L.
Synthetic products
literature data [12]
(5S*,50R*)-6 ¼ 6a (major, 79%)
(5R*,50R*)-6 ¼ 6b (minor, 21%)
Position
C(3)-CH3
H-4
H-5
H2-30
dH (ppm) (apparent multiplicity; estimated J coupling constants in Hz)
1.83 (dd; 1.8, 1.4)
6.68 (dq; 1.8, 1.8)
5.53 (dq; 1.8, 1.4)
2.29 (m)
1.82 (t; 1.6)
6.67 (qui; 1.7)
5.54 (m)
1.83 (t; 1.6)
6.68 (qui; 1.7)
5.48 (m)
2.30 (m)
2.27 (m)
2.75 (m)
2.34 (m)
2.75 (m)
2.35 (m)
2.69 (m)
2.38 (m)
H2-40
2.57 (m)
5.51 (dd; 8.7, 2.4)
2.57 (m)
5.51 (dd; 8.7, 2.4)
2.51 (m)
5.73 (dd; 8.5, 2.5)
H-50
qui: quintet.
Table 7
Comparison of the literature 1H NMR data for 8 with our data for 8a and 8b.
1H NMR data in methanol-d4
Lilium candidum L.
Synthetic products
literature data [13]
(5S*,50R*)-8 ¼ 8a (major, 75%)
(5R*,50R*)-8 ¼ 8b (minor, 25%)
Position
C(3)-CH3
H-4
dH (ppm) (apparent multiplicity; estimated J coupling constants in Hz)
1.85 (dd; 1.7, 1.4)
1.85 (dd; 1.7, 1.5)
6.64 (quid; 1.7, 0.3)
5.40 (qui; 1.5)
1.90 (t; 1.7)
1.86 (dd; 1.7, 1.4)
6.67 (quid; 1.8, 0.5)
5.25 (qui; 1.3)
1.88 (t; 1.8)
6.78 (qui; 1.8)
6.11 (qui; 1.8)
6.64 (dqd; 1.8, 1.7, 0.4)
5.40 (dqd; 1.8, 1.4, 0.3)
1.91 (dd; 1.7, 1.7)
H-5
C(30)-CH3
H-40
6.76 (dq; 2.0, 1.7)
5.96 (dqdd; 2.0, 1.7, 0.4, 0.3)
6.76 (qui; 1.8)
5.96 (qui; 1.8)
H-50
qui: quintet.
Table 8
Comparison of the literature 1H NMR data for 7 with our data for 7ae7d.
1H NMR data in acetone- Lilium candidum L.
Synthetic products
d6
literature data [11]
rel-(5S,30R,50R)-7 ¼ 7a
rel-(5R,30R,50R)-7 ¼ 7b (35%) rel-(5S,30S,50R)-7 ¼ 7c
rel-(5R,30S,50R)-7 ¼ 7d
(53%)
(7%)
(5%)
Position
C(3)-CH3
H-4
dH (ppm) (apparent multiplicity; estimated J coupling constants in Hz)
1.77 (dd; 1.8, 1.3)
6.20 (qd; 1.8, 1.8)
5.54 (ddq; 10.1, 1.8, 1.3)
4.83 (d; 10.1)
1.77 (t; 1.5)
6.63 (qui; 1.7)
5.58 (dqui; 10.0, 1.5)
5.24 (d; 10.0)
1.77 (t; 1.5)
z1.78 (om)
z6.64 (om)
5.63 (om)
5.36 (d; 9.4)
e
z1.78 (om)
z6.64 (om)
5.64 (om)
5.33 (d; 9.6)
e
z6.64 (qui; 1.8)
5.50 (dqui; 10.0, 1.4)
5.22 (d; 10.0)
H-5
C(5)-OH
NH-10
6.86 (br s; 1.4, 0.8)
2.85 (ddq; 9.5, 8.7, 7.2)
1.12 (d; 7.2)
~7.30 (br)
2.92 (ddq; 9.1, 8.9, 7.3)
1.133 (d; 7.3)
~7.38 (br)
2.88 (ddq; 9.2, 8.9, 7.2)
1.127 (d; 7.2)
H-30
2.51 (om)
1.22 (d; 7.1)
2.49 (om)
1.20 (d; 6.9)
C(30)-CH3
H2-40
2.10 (ddd; 13.7, 9.5, 8.5)
2.54 (dddd; 13.7, 8.7, 1.4,
0.8)
2.11 (ddd; 13.6, 8.9, 8.6)
2.54 (ddd; 13.6, 9.1, 1.2)
2.07 (ddd; 13.3, 9.2, 8.6)
2.55 (dddd; 13.3, 8.9, 1.2,
0.4)
2.17 (ddd; 12.7, 9.1, 7.2) 2.39 (ddd; 12.4, 9.6, 7.4)
2.58 (ddd; 12.7, 9.3, 7.6) 2.51 (om)
H-50
5.40 (ddd; 8.5, 1.4, 1.4)
5.45 (ddd; 8.6, 1.4, 1.2)
5.62 (ddd; 8.6, 1.4, 1.2)
5.57 (dd; 7.6, 7.2)
5.68 (t; 7.4)
qui: quintet.
om: overlapping multiplet.
11

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
compounds is uncommon in the sense that it is based on mapping
the full conformational space in order to obtain reliable results,
rather than arbitrarily restricting the conformational space to
several representative conformers.
2867, 1460, 1185, 1081, 1045 cmꢁ1.1H NMR (500 MHz, CDCl3)
d
(ppm) 1.10 (18H; d; J ¼ 7.5 Hz; CH3); 1.45 (3H; spt; J ¼ 7.5 Hz;
CH(CH3)2); 6.30e6.33 (2H; m; H-3, H-4); 6.77e6.82 (2H; m; H-2, H-
5). 13C NMR (126 MHz, CDCl3)
(ppm) 11.7 (CH(CH3)2); 17.8 (CH3);
110.0 (C-3, C-4); 124.0 (C-2, C-5). Lit. 1H NMR (500 MHz, CDCl3)
d
4. Experimental section
d
(ppm) 1.09 (18 H; d; J ¼ 7.4 Hz; CH3); 1.45 (3H; spt; J ¼ 7.4 Hz;
CH(CH3)2); 6.32 (2H; t*; H-3, H-4); 6.80 (2H; t*; H-2, H-5); *should
be interpreted as apparent multiplicity; no coupling constants were
reported [23].
Melting points were measured on a SANYO Gallenkamp appa-
ratus and are uncorrected. IR spectra were recorded on a Bruker FT-
IR instrument and a PerkinElmer Spectrum 100 FT-IR Spectrometer
equipped with Universal ATR (diamond/ZnSe) accessory. NMR
measurements (1H, 13C, COSY, 1D NOESY, 1D ROESY, 2D ROESY,
HSQC, HMBC, 1,1-ADEQUATE) were performed on Varian VNMRS
400 MHz (equipped with 5 mm OneNMR 15Ne31 P/{1He19F} PFG
Probe), Varian VNMRS 500 MHz (equipped with 1H{13C,15N} 5 mm
PFG Triple Resonance 13C Enhanced Cold Probe) and Varian VNMRS
800 MHz (equipped with 1H{13C,15N} Triple Resonance 13C
Enhanced Salt Tolerant Cold Probe), Bruker Avance III HDX
500 MHz (equipped with 1H {13C,15N} 5 mm TCI CryoProbe), and
Bruker Avance III HDX 800 MHz (equipped with 1H/19F {13C,15N}
5 mm TCI CryoProbe) spectrometers. 1H and 13C chemical shifts are
given on the delta scale in parts per million (ppm) relative to tet-
ramethylsilane (TMS). 1H multiplicities are given as d (doublet), t
(triplet), q (quartet), spt (septet) and their combinations or as m
(multiplet); br corresponds to broad. Coupling constants are given
in Hz in descending order. NMR spectra were processed using
VnmrJ 2.2 Revision C (Varian, Inc., Palo Alto, CA, USA), Bruker
TopSpin 3.5 pl 6 (Bruker Corporation, Billerica, MA, USA) and ACD/
Spectrus Processor version 2017.1.3 (Advanced Chemistry Devel-
opment, Inc., Toronto, ON, Canada); Daisy plugin of TopSpin was
used for spin system simulations. HRMS and MS analyses were
performed on a Finnigan MAT 95 XP and a Thermo LTQ FT Ultra as
well as a Thermo LTQ XL (Thermo Fisher Scientific, Bremen, Ger-
many) system. The ionization method was EI operated in positive
ion mode on a Finnigan MAT 95 XP. The electron energy was 70 eV
and the source temperature was set to 220 ꢀC. The ionization
method was ESI operated in positive ion mode on the other two
systems. For the CID experiment helium was used as the collision
gas, and normalized collision energy (expressed in percentage),
which is a measure of the amplitude of the resonance excitation RF
voltage applied to the endcaps of the linear ion trap, was used to
induce fragmentation. The protonated molecular ion peaks were
fragmented by CID at a normalized collision energy of 35%. Data
acquisition and analysis were accomplished with Xcalibur software
version 2.0 (Thermo Fisher Scientific). TLC was carried out on TLC
Silica gel 60 F254 on 20 ꢂ 20 cm aluminium sheets (Merck), and
preparative TLC was carried out using Silica gel 60 PF254þ366
(Merck) coated glass plates. Column chromatography was per-
formed using Silica gel 60 (0.063e0.200 mm) (Merck). Kaempferol
was purchased from Xi’an Lyphar Biotech Co., Ltd (Xi’an, Shaanxi,
China) and used without further purification.
4.2. Pyrrole-3-carboxaldehyde (22)
Phosphoryl chloride (8.27 mL, 13.6 g, 88.7 mmol) was cooled to
0
ꢀC with stirring, then DMF (6.86 mL, 6.50 g, 88.9 mmol) was
slowly added at 0 ꢀC. To this mixture a solution of 1-(triisopro-
pylsilyl)pyrrole (21) (16.5 g, 73.8 mmol) in acetonitrile (60 mL) was
added over 30 min. The reaction mixture was let to warm up to
room temperature and was stirred for 2 h, then it was again cooled
to 0 ꢀC. The mixture was quenched with 2 M NaOH solution until
pH ¼ 9. The mixture was extracted with diethyl ether (3 ꢂ 30 mL).
The ether solution was dried (Na2SO4), then the solvent was
removed under reduced pressure. The crude product was purified
by column chromatography, using hexane:EtOAc (1:4) as the
eluent, giving the title compound 22 as a pale brown crystalline
solid (4.56 g, 65%), mp 65 ꢀC (lit.: 68 ꢀC) [23]; Rf
(hexane:EtOAc ¼ 1:4) 0.63; nmax (KBr) 3258, 3116, 1653, 1505, 1435,
1416, 1402, 1377, 1297, 1091, 1061, 1052 cmꢁ1
CDCl3)
7.47e7.48 (1H; m; H-2); 9.20e9.54 (1H; br m; NH); 9.82 (1H; ~s;
CHO). 13C NMR (201 MHz, CDCl3)
(ppm) 107.5 (C-4); 120.7 (C-5);
;
1H NMR (800 MHz,
d
(ppm) 6.69e6.70 (1H; m; H-4); 6.85e6.87 (1H; m; H-5);
d
126.7 (C-3); 127.4 (C-2); 186.1 (CHO). EI-HRMS: M ¼ 95.0353
(C5H5ON; calc.: 95.0366); EI-MS (rel. int. %): 95(100); 66(16).
4.3. 3-Methylpyrrole (23)
To a solution of KOH (10.8 g, 192 mmol) in ethylene glycol
(72 mL) a solution of pyrrole-3-carboxaldehyde (22) (4.56 g,
47.9 mmol) was added in hydrazine hydrate (2.33 mL, 2.40 g,
48.0 mmol). The flask was fitted for distillation, and the mixture
was slowly heated to 200 ꢀC with vigorous stirring. The pure frac-
tions of the product were collected at 140e145 ꢀC. The distillate
was extracted with diethyl ether (3 ꢂ 30 mL). The ether solution
was dried (Na2SO4), then the solvent was removed under reduced
pressure, giving the title compound 23 as a colorless liquid (2.84 g,
73%), bp 143e145 ꢀC (lit.: 142e143 ꢀC) [41]; Rf
(hexane:EtOAc ¼ 4:1) 0.60; nmax (neat film) 3387, 3091, 2924, 2867,
4.1. 1-(Triisopropylsilyl)pyrrole (21)
To a solution of pyrrole (5.17 mL, 5.00 g, 74.5 mmol) in dry THF
(100 mL) at ꢁ75 ꢀC, n-butyllithium solution (33.0 mL, 2.5 M soln. In
hexanes, 82.5 mmol) was slowly added under argon. The solution
was stirred at ꢁ75 ꢀC for 10 min, then triisopropylsilyl chloride
(16.0 mL, 14.4 g, 74.8 mmol) was added over 10 min. After the
addition, the mixture was let to warm up to room temperature,
then it was quenched with water (100 mL). The mixture was
extracted with diethyl ether (3 ꢂ 50 mL). The ether solution was
dried (Na2SO4), then the solvent was removed under reduced
pressure, giving the title compound 21 as a colorless oil (16.5 g,
99%), Rf (hexane:EtOAc ¼ 1:4) 0.95; nmax (neat film) 2946, 2892,
2744, 1601, 1560, 1485, 1429, 1254, 1137, 1059 cmꢁ1
(800 MHz; CDCl3)
4); 6.59e6.61 (1H; m; H-2); 6.74e6.76 (1H; m; H-5); 7.82e8.18 (1H;
br m; NH). 13C NMR (201 MHz; CDCl3)
(ppm) 11.7 (CH3); 109.6 (C-
;
1H NMR
d
(ppm) 2.19 (3H; s; CH3); 6.12e6.13 (1H; m; H-
d
4); 115.5 (C-2); 117.7 (C-5); 118.7 (C-3). EI-HRMS: M ¼ 81.0571
(C5H7N; calc.: 81.0573); EI-MS (rel. int. %): 81(65); 80(100); 53(21).
12

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
4.6. Citraconimide (26)
To a solution of citraconic anhydride (25) (12.0 g, 107 mmol) in
DMF (100 mL) at 100 ꢀC hexamethyldisilazane (34.2 mL, 160 mmol)
was added. The mixture was stirred at 100 ꢀC for 1 h, then it was
cooled to room temperature. The solvent was evaporated under
reduced pressure. The crude product was recrystallized from a
mixture of hexane and ethyl acetate to give the title compound 26
as a white crystalline solid (8.33 g, 70%), mp 104e105 ꢀC (lit.:
104e105 ꢀC) [3]; Rf (DCM:MeOH ¼ 10:1) 0.71; nmax (KBr) 3271,
3098, 2674, 1844, 1778, 1764, 1712, 1634, 1342, 1284, 1175,
4.4. 3-Methyl-3-pyrrolin-2-one (17)
30% Hydrogen peroxide (6.90 mL, 67.6 mmol) was added to 3-
methylpyrrole (23) (2.84 g, 35.0 mmol) at room temperature, and
the mixture was vigorously stirred for 24 h. The mixture was
extracted with diethyl ether (6 ꢂ 20 mL), the ether solution was
dried (Na2SO4), then the solvent was evaporated under reduced
pressure. The crude product was recrystallized from acetone, giving
the title compound 17 as a pale yellow crystalline solid (1.80 g, 53%),
mp 87e89 ꢀC (lit.: 95e96 ꢀC) [26]; Rf (acetone) 0.55; nmax (KBr)
3215, 3068, 2971, 2923, 2855, 1683, 1640, 1447, 1363, 1231, 1170,
1090 cmꢁ1
J ¼ 1.8 Hz; C(3)-CH3); 6.50 (1H; q; J ¼ 1.8 Hz; H-4); 10.72 (1H; br s;
NH-1); 13C NMR (126 MHz; DMSO‑d6)
(ppm) 10.2 (C(3)-CH3);
; d (ppm) 1.94 (3H; d;
1H NMR (500 MHz; DMSO‑d6)
d
1077 cmꢁ1
m; C(3)-CH3); 3.75e3.79 (2H; m; H2-5); 6.84e6.88 (1H; m; H-4);
8.01e8.23 (1H; br; NH-1). 13C NMR (101 MHz; DMSO‑d6)
(ppm)
; d (ppm) 1.71e1.75 (3H;
1H NMR (400 MHz; DMSO‑d6)
128.0 (C-4); 145.9 (C-3); 172.2 (C-5); 173.1 (C-2); HRMS:
M ¼ 111.0311 (C5H5O2N; delta ¼ ꢁ8.1 ppm); HR-EI-MS (rel. int. %):
111(100); 83(8); 68(39).
d
10.7 (C(3)-CH3); 45.6 (C-5); 134.0 (C-3); 138.6 (C-4); 173.8 (C-2). EI-
HRMS: M ¼ 97.0519 (C5H7ON; calc.: 97.0528); EI-MS (rel. int. %):
97(100); 82(11); 78(5); 69(15); 68(17); 54(4).
4.7. ( )-Jatropham [5-hydroxy-3-methyl-3-pyrrolin-2-one] (1)
To a solution of citraconimide (26) (1.00 g, 9.00 mmol) in THF
(50 mL) under nitrogen at ꢁ78 ꢀC diisobutyl aluminium hydride
solution (10.7 mL, 16.0 mmol, 25 wt % in toluene) was added via a
syringe. The mixture was let to warm up to 0 ꢀC and it was stirred at
this temperature for a further 90 min, then it was quenched with a
mixture of methanol (25 mL) and water (25 mL). The solvents were
evaporated under reduced pressure, then the crude product was
triturated in ethyl acetate. The suspension was filtered, the filtrate
was dried (Na2SO4) and evaporated under reduced pressure to give
the title compound 1 as a white crystalline solid (930 mg, 91%), mp
115e117 ꢀC (lit.: 115e118 ꢀC) [5]; Rf (DCM:MeOH ¼ 10:1) 0.42; nmax
(KBr) 3244, 2985, 2931, 1688, 1650, 1409, 1306, 1281, 1216,
4.5. 5-Ethoxypyrrolidin-2-one (18)
To a solution of succinimide (24) (7.16 g, 72.3 mmol) in ethanol
(300 mL) at 0 ꢀC was added sodium borohydride (4.00 g, 106 mmol)
in one portion. The reaction mixture was stirred at 0 ꢀC for 4 h,
during which time every 15 min five drops of 2 M ethanolic
hydrogen chloride solution were added. Then the reaction mixture
was acidified to pH ¼ 3 with 2 M ethanolic hydrogen chloride so-
lution over 30 min, after which it was stirred at 10 ꢀC for 90 min.
Then the reaction mixture was neutralised (pH ¼ 7) with 5%
ethanolic potassium hydroxide solution and evaporated in vacuo to
give a syrupy solid, which was suspended in chloroform (80 mL),
filtered, and the precipitate washed with chloroform (3 ꢂ 20 mL).
The filtrate was evaporated in vacuo to give a colourless oil, which
was dissolved in dichloromethane (80 mL) and washed with water
(3 ꢂ 10 mL). The aqueous phase was extracted with dichloro-
methane (6 ꢂ 20 mL), then the organic phases were unified, dried
(MgSO4) and the solvent evaporated in vacuo to give the title
compound 18 (5.79 g, 62%) as a colourless oil, which crystallized on
standing, giving white crystals, mp 54e58 ꢀC (lit.: 48e53 ꢀC) [22];
Rf (acetone) 0.72; nmax (KBr) 3200, 2978, 1707, 1689, 1668, 1457,
1058 cmꢁ1 1H NMR (500 MHz; CD3OD)
; d (ppm) 1.83 (3H; dd;
J ¼ 1.7; 1.4 Hz; C(3)-CH3); 5.44 (1H; dq; 1.7; 1.4 Hz; H-5); 6.63 (1H;
~qui; J ¼ 1.7 Hz; H-4); 13C NMR (126 MHz; CD3OD)
d (ppm) 10.6
(C(3)-CH3); 80.0 (C-5); 136.9 (C-3); 143.1 (C-4); 175.5 (C-2); HRMS:
MþH ¼ 114.05495 (delta ¼ ꢁ0.04 ppm; C5H8O2N); HR-ESI-MS-MS
(CID ¼ 55%, rel. int. %): 97(100).
1282, 1250, 1067, 986 cmꢁ1; 1H NMR (400 MHz; DMSO‑d6)
d (ppm)
1.78e1.89 (1H; m; Hx-4); 1.95e2.05 (1H; m; Hx-3); 2.11e2.30 (2H;
m; Hy-3, Hy-4); 3.27e3.35 (1H; m), 3.44e3.54 (1H; m) (OCH2);
4.83e4.89 (1H; m; H-5); 8.62 (1H; s; NH-1); 13C NMR (101 MHz;
4.8. ( )-2-Methylsuccinimide (27)
DMSO‑d6)
d
(ppm) 15.1 (CH3); 27.7 (C-4); 28.1 (C-3); 61.6 (OCH2);
To a solution of citraconimide (26) (4.50 g, 40.5 mmol) in
ethanol (50 mL) 10% palladised charcoal (2.00 g) was added. The
suspension was hydrogenated under atmospheric pressure for 4 h.
85.0 (C-5); 177.4 (C-2); ESI-MS: MþH ¼ 130 (C6H12O2N); ESI-MS-
MS (CID ¼ 35%, rel. int. %): 84(100).
13

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
After completion of the reaction (confirmed with TLC) the mixture
was filtered through a pad of Celite, the Celite layer was washed
with ethanol (2 ꢂ 20 mL), and the filtrate was evaporated under
reduced pressure, giving the title compound 27 as a white crys-
talline solid (4.55 g, 99%), mp 57e59 ꢀC (lit.: 65 ꢀC) [42]; Rf
(DCM:MeOH ¼ 10:1) 0.52; nmax (KBr) 3201, 3071, 2766, 1763, 1707,
J ¼ 7.4 Hz; C(5)-OH); 8.17 (1H; br; NH-1); 13C NMR (201 MHz;
DMSO‑d6)
d (ppm) 13.9 (C(4)-CH3); 34.3 (C-4); 35.0 (C-3); 79.9 (C-
5); 176.8 (C-2).
1352, 1206, 1176, 1037 cmꢁ1; 1H NMR (400 MHz; DMSO‑d6)
d (ppm)
1.16 (3H; d; J ¼ 7.2 Hz; C(3)-CH3); 2.26 (1H; m; 2J ¼ ꢁ17.7 Hz (Ha-4),
3J ¼ 5.1 Hz (H-3); Hb-4); 2.78 (1H; m; 3J ¼ 9.2 Hz (H-3); Ha-4); 2.81
(1H; m; 3J ¼ 7.3 Hz (C(3)-CH3); Ha-4); 10.94e11.10 (1H; br; NH-1);
13C NMR (101 MHz; DMSO‑d6)
37.1 (C-4); 178.0 (C-5); 182.2 (C-2); HRMS:
d
(ppm) 15.7 (C(3)-CH3); 35.4 (C-3);
113.0461
4.10. 3-Methyl-1-(20-oxopyrrolidin-50-yl)-3-pyrrolin-2-one (4)
M
¼
(delta ¼ ꢁ14.2 ppm; C5H7NO2); HR-EI-MS (rel. int. %): 113(100);
To a solution of 3-methyl-3-pyrrolin-2-one (17) (106 mg,
1.09 mmol) in THF (2 mL) 5-ethoxypyrrolidin-2-one (18) (141 mg,
1.09 mmol) and p-toluenesulfonic acid monohydrate (13.0 mg,
70(77).
68.3 mmol) were added. The mixture was stirred at room temper-
ature for 1 h, after which a white crystalline solid precipitated. The
mixture was cooled to 0 ꢀC and the product was filtered. The
crystals were washed with cold acetone and dried by suction, giv-
ing the title compound 4 as a white crystalline solid (107 mg, 54%),
mp 164e166 ꢀC (lit.: 167e169 ꢀC) [11]; Rf (acetone) 0.38; nmax (KBr)
3189, 3096, 2916, 1707, 1672, 1462, 1378, 1271, 1240, 1190, 1174,
1094 cmꢁ1 1H NMR (500 MHz; acetone-d6)
; d (ppm) 1.79 (3H; m;
4.9. 5-Hydroxy-3-methylpyrrolidin-2-one (19)
5J ¼ 2.1 Hz (Ha-5), 5J ¼ 2.0 Hz (Hb-5), 4J ¼ 1.7 Hz (H-4); C(3)-CH3);
2
3
2.08 (1H; m; J ¼ ꢁ13.7 Hz (Ha-40), J ¼ 9.8 Hz (Hb-30), 3J ¼ 4.6 Hz
(Ha-30), 3J ¼ 3.3 Hz (H-50); Hb-40); 2.22 (1H; m; 2J ¼ ꢁ17.1 Hz (Hb-30),
3J ¼ 10.2 Hz (Ha-40); Ha-30); 2.44 (1H; m; 3J ¼ 7.7 Hz (Ha-40); Hb-30);
2.50 (1H; m; 3J ¼ 7.6 Hz (H-50); Ha-40); 3.91 (1H; m; 2J ¼ ꢁ19.4 Hz
(Hb-5); 3J ¼ 1.4 Hz (H-4); Ha-5); 3.98 (1H; m; 3J ¼ 2.4 Hz (H-4); Hb-
5); 5.80 (1H; m; H-50); 6.88 (1H; m; H-4); 7.20 (1H; br; NH-10); 13C
To
a solution of ( )-2-methylsuccinimide (27) (1.00 g,
8.84 mmol) in THF (50 mL) at ꢁ78 ꢀC diisobutyl aluminium hydride
solution (10.7 mL, 16.0 mmol, 25% in toluene) was added via a sy-
ringe. The mixture was let to warm up to 0 ꢀC and was stirred at this
temperature for a further 90 min. Then it was quenched with a
mixture of methanol (25 mL) and water (25 mL). The solvents were
evaporated under reduced pressure, then the crude product was
triturated in ethyl acetate. The suspension was filtered, the filtrate
was dried (Na2SO4) and evaporated under reduced pressure to give
the title compound 19 (contaminated with 27% of the minor isomer
28) as a colorless oil (737 mg, 72%), Rf (DCM:MeOH ¼ 10:1) 0.54;
nmax (neat film) 3268, 2967, 2935,1664,1458,1335,1254,1107,1054,
1000 cmꢁ1. HRMS: 2MþNa ¼ 253.11569 (delta ¼ ꢁ0.74 ppm;
NMR (126 MHz; acetone-d6)
d
(ppm) 11.1 (C(3)-CH3); 26.7 (C-40);
29.6 (C-30); 46.4 (C-5); 62.5 (C-50); 135.2 (C-3); 137.6 (C-4); 172.0
(C-2); 177.0 (C-20); Lit. 1H NMR [11]: 1.78 (3H; ddd; 5J ¼ 2.0 Hz (Hx-
5), 5J ¼ 2.0 Hz (Hy-5), 4J ¼ 1.8 Hz (H-4); C(3)-CH3); 2.02e2.56 (4H;
m; H2-30, H2-40); 3.90 (1H; m; 2J ¼ 19.3 Hz (Hy-5), 3J ¼ 2.0 Hz (H-4);
Hx-5); 3.96 (1H; m; 3J ¼ 2.0 Hz (H-4); Hy-5); 5.79 (1H; m; 3J ¼ 7.8 Hz
3
3
(Hx-40), J ¼ 3.3 Hz (Hy-40), J ¼ 1.3 Hz (NH-10), 5J ¼ 0.6 Hz (H-4),
4J ¼ 0.5 Hz (Hx-5), 4J ¼ 0.5 Hz (Hy-5); H-50); 6.86 (1H; ddqd; H-4);
7.07 (br s, 1H, NH-10); ESI-MS: MþH ¼ 181 (C9H13O2N2); ESI-MS-MS
(CID ¼ 55%) (rel. int. %): 164(26); 98(100).
C
10H18O4N2Na); HR-ESI-MS-MS (CID ¼ 35%, rel. int. %): 235(47);
233(11); 138(100).
((3S*,5S*)-19): 1H NMR (800 MHz; DMSO‑d6)
d (ppm) 1.00 (3H;
d; J ¼ 7.2 Hz; C(3)-CH3); 1.76 (1H; ddd; J ¼ 13.2; 9.5; 6.1 Hz; Hb-4);
1.99 (1H; dd; J ¼ 13.2; 8.2 Hz; Ha-4); 2.44e2.49 (1H; m; H-3); 4.97
(1H; br d; J ¼ 6.1 Hz; H-5); 5.64 (1H; br; C(5)-OH); 8.18 (1H; br; NH-
1); 13C NMR (201 MHz; DMSO‑d6)
d (ppm) 15.7 (C(3)-CH3); 32.9 (C-
3); 39.2 (C-4); 76.1 (C-5); 179.1 (C-2).
((3S*,5R*)-19): 1H NMR (500 MHz; DMSO‑d6)
d (ppm) 1.09 (3H;
d; J ¼ 7.3 Hz; C(3)-CH3); 1.35 (1H; ddd; J ¼ 13.2; 6.8; 4.0 Hz; Hb-4);
2.20 (1H; dqd; J ¼ 9.2; 7.3; 6.8 Hz; H-3); 2.43e2.47 (1H; m; Ha-4);
5.05 (1H; dd; J ¼ 6.3; 4.0 Hz; H-5); 5.74 (1H; d; J ¼ 6.9 Hz; C(5)-OH);
8.08 (1H; br; NH-1); 13C NMR (201 MHz; DMSO‑d6)
d
(ppm) 17.0
4.11. 3-Methyl-1-(30-methyl-20-oxopyrrolidin-50-yl)-3-pyrrolin-2-
(C(3)-CH3); 35.0 (C-3); 38.6 (C-4); 77.1 (C-5); 178.1 (C-2).
one (5)
((4S*,5R*)-28): 1H NMR (800 MHz; DMSO‑d6)
d (ppm) 0.98 (3H;
d; J ¼ 7.1 Hz; C(4)-CH3); 1.64 (1H; dd; J ¼ 16.6, 4.1 Hz; Hb-3);
2.01e2.06 (1H; m; H-4); 2.47 (1H; dd; J ¼ 16.6, 8.3 Hz; Ha-3); 4.60
(1H; ddd; J ¼ 6.9, ~2.1, ~1.1 Hz; H-5); 5.74 (1H; d; J ¼ 6.9 Hz; C(5)-
To a solution of 3-methyl-3-pyrrolin-2-one (17) (50.0 mg,
515
(60.0 mg, 521
(7.00 mg, 36.8 m
m
mol) in THF (2 mL) 5-hydroxy-3-methylpyrrolidin-2-one (19)
mol) and p-toluenesulfonic acid monohydrate
mol) were added. The mixture was stirred at room
m
OH); 8.11 (1H; br; NH-1); 13C NMR (201 MHz; DMSO‑d6)
d
(ppm)
18.4 (C(4)-CH3); 36.8 (C-3); 37.7 (C-4); 85.0 (C-5); 175.7 (C-2).
temperature for 1 h, after which the mixture turned opaque. The
solvent was removed under reduced pressure and the product was
recrystallized using a mixture of hexane and ethyl acetate. The
product was filtered and the crystals were dried by suction, giving
the title compound 5 as a white crystalline solid (87.0 mg, 87%),
((4S*,5S*)-28): 1H NMR (800 MHz; DMSO‑d6)
d (ppm) 0.95 (3H;
d; J ¼ 6.9 Hz; C(4)-CH3); 1.90 (1H; dd; J ¼ 16.2, 10.6 Hz; Hb-3); 2.04
(1H; dd; J ¼ 16.2, 8.4 Hz; Ha-3); 2.31 (1H; ddqd; J ¼ 10.6, 8.4, 6.9,
5.4 Hz; H-4); 4.84 (1H; ddd; J ¼ 7.4, 5.4, 1.2 Hz; H-5); 5.43 (1H; d;
14

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
mp 168e172 ꢀC (lit.: 172e174 ꢀC) [11]; Rf (DCM:MeOH ¼ 10:1) 0.52;
nmax (KBr) 3193, 3094, 2978, 1709, 1677, 1650, 1454, 1382, 1238,
1695, 1661, 1460, 1417, 1381, 1296, 1281, 1238, 1091, 1079,
1031 cmꢁ1; HRMS: MþH ¼ 197.09201 (delta ¼ ꢁ0.30 ppm;
C9H13O3N2); HR-ESI-MS-MS (CID ¼ 35%, rel. int. %): 179(100).
1181 cmꢁ1
;
HRMS: MþH
10H15O2N2); HR-ESI-MS-MS (CID ¼ 55%, rel. int. %): 178(54);
138(8); 98(100).
¼
195.11283 (delta
¼
0.13 ppm;
C
4.15. (5S*,50R*)-5-hydroxy-3-methyl-1-(20-oxopyrrolidin-50-yl)-3-
pyrrolin-2-one (6a)
4.12. (30R*,50R*)-1-(30-methyl-20-oxo-50-pyrrolidinyl)-3-methyl-3-
pyrrolin-2-one (5a)
1H NMR (800 MHz; acetone-d6: methanol-d4 ¼ 1:1 (v/v))
d
(ppm) 1.82 (3H; m; C(3)-CH3); 2.30 (1H; ddd; J ¼ 16.9,10.3, 3.4 Hz;
1H NMR (500 MHz; acetone-d6)
d
(ppm) 1.14 (3H; d; J ¼ 7.2 Hz;
Ha-30); 2.35 (1H; dddd; J ¼ 13.9, 9.8, 3.4, 2.4 Hz; Hb-40); 2.57 (1H;
dddd; J ¼ 13.9, 10.3, 8.7, 8.3 Hz; Ha-40); 2.75 (1H; ddd; 16.9, 9.8, 8.3;
Hb-30); 5.51 (1H; dd; J ¼ 8.7, 2.4 Hz; H-50); 5.53e5.55 (1H; m; H-5);
6.67e6.68 (1H; m; H-5); Lit. 1H NMR [12]: 1.83 (3H; dd; J ¼ 1.8,
1.4 Hz; C(3)-CH3); 2.29 (1H; m; Hx-3); 2.34 (1H; m; Hx-4); 2.57 (1H;
m; Hy-4); 2.75 (1H; m; Hy-3); 5.51 (1H; dd; J ¼ 8.7, 2.4 Hz; H-50);
5.53 (1H; dq; J ¼ 1.8, 1.4 Hz; H-5); 6.68 (1H; dq; J ¼ 1.8, 1.8 Hz; H-4);
C(30)-CH3); 1.78e1.79 (3H; m; C (3)-CH3); 2.14 (1H; ddd; J ¼ 13.7,
9.5, 8.2 Hz; Ha-40); 2.32 (1H; ddd; J ¼ 13.7, 8.7, 1.5 Hz; Hb-40); 2.64
(1H; ddq; J ¼ 9.5, 8.7, 7.2 Hz; H-30); 3.87e4.00 (2H; m; H2-5);
5.69e5.73 (1H; m; H-50); 6.85e6.87 (1H; m; H-4); 7.23 (1H; br; NH-
10); 13C NMR (126 MHz; acetone-d6)
d (ppm) 11.1 (C(3)-CH3); 16.6
(C(30)-CH3); 35.1 (C-30); 35.8 (C-40); 46.7 (C-5); 60.3 (C-50); 135.2 (C-
3); 137.4 (C-4); 171.9 (C-2); 179.5 (C-20); Lit. 1H NMR [11]: 1.14 (3H;
d; J ¼ 7.1 Hz; C(30)-CH3); 1.78 (3H; ddd; J ¼ 2.0, 2.0, 1.8 Hz; C(3)-
CH3); 2.13 (1H; ddd; J ¼ 13.7, 9.5, 8.1 Hz; Hx-40); 2.32 (1H; dddd;
J ¼ 13.7, 8.7, 1.9, 0.7 Hz; Hy-40); 2.63 (1H; ddqd; J ¼ 9.5, 8.7, 7.1,
0.3 Hz; H-30); 3.90, 3.96 (2H; m; 2J ¼ 19.3 Hz, 4J ¼ 2.0 Hz (H-4),
5J ¼ 2.0 Hz (C (3)-CH3); H2-5); 5.70 (1H; dddddd; J ¼ 8.1, 1.9, 1.3, 0.6,
0.5, 0.5 Hz; H-50); 6.85 (1H; ddqd; J ¼ 2.0, 2.0, 1.8, 0.6 Hz; H-4); 7.12
(1H; br s; J ¼ 1.3, 0.7, 0.3 Hz; NH-10).
13C NMR (126 MHz; acetone-d6: methanol-d4 ¼ 1:1 (v/v))
d (ppm)
10.6 (C(3)-CH3); 26.2 (C-40); 30.6 (C-30); 63.7 (C-50); 82.3 (C-5);
136.5 (C-3); 141.6 (C-4); 171.7 (C-2); 180.4 (C-20).
4.16. (5R*,50R*)-5-hydroxy-3-methyl-1-(20-oxopyrrolidin-50-yl)-3-
pyrrolin-2-one (6b)
1H NMR (800 MHz; acetone-d6: methanol-d4 ¼ 1:1 (v/v))
4.13. (30S*,50R*)-1-(30-methyl-20-oxo-50-pyrrolidinyl)-3-methyl-3-
pyrrolin-2-one (5b)
d
(ppm) 1.83 (3H; m; C(3)-CH3); 2.27 (1H; ddd; J ¼ 17.1, 10.2, 3.6 Hz;
Ha-30); 2.38 (1H; dddd; J ¼ 13.6, 9.8, 3.5, 2.5 Hz; Hb-40); 2.51 (1H;
dddd; J ¼ 13.6, 10.2, 8.5, 8.3 Hz; Ha-40); 2.69 (1H; ddd; J ¼ 17.1, 9.8,
8.3 Hz; Hb-30); 5.47e5.49 (1H; m; H-5); 5.73 (1H; dd; J ¼ 8.5, 2.5 Hz;
H-50); 6.68e6.69 (H-4). 13C NMR (126 MHz; acetone-d6: methanol-
1H NMR (500 MHz; acetone-d6)
d
(ppm) 1.18 (3H; d; J ¼ 7.1 Hz;
C(30)-CH3); 1.73 (1H; ddd; J ¼ 12.9, 9.4, 7.4 Hz; Hb-40); 1.79e1.80
(3H; m; C(3)-CH3); 2.46e2.51 (1H; m; H-30); 2.56e2.61 (1H; m; Ha-
40); 3.87e4.00 (2H; m; H2-5); 5.80 (1H; ~t; J ¼ 7.4 Hz; H-50);
6.88e6.90 (1H; m; H-4); 7.08 (1H; br; NH-10); 13C NMR (126 MHz;
d4 ¼ 1:1 (v/v))
d
(ppm) 10.6 (C(3)-CH3); 27.1 (C-40); 30.1 (C-30); 63.5
(C-50); 80.9 (C-5); 136.0 (C-3); 141.7 (C-4); 171.9 (C-2); 180.6 (C-20).
acetone-d6)
d
(ppm) 11.1 (C(3)-CH3); 16.3 (C(30)-CH3); 34.8 (C-40);
36.5 (C-30); 46.0 (C-5); 60.7 (C-50); 135.2 (C-3); 137.8 (C-4); 172.4
(C-2); 178.4 (C-20).
4.17. 5-Hydroxy-3-methyl-1-(30-methyl-20-oxopyrrolidin-50-yl)-3-
pyrrolin-2-one (7)
4.14. 5-Hydroxy-3-methyl-1-(20-oxopyrrolidin-50-yl)-3-pyrrolin-2-
one (6)
To a solution of jatropham 1 (67.5 mg, 597
5-hydroxy-3-methylpyrrolidin-2-one (19) (69.0 mg, 599
p-toluenesulfonic acid monohydrate (7.50 mg, 39.4 mol) were
m
mol) in THF (10 mL)
mmol) and
m
To a solution of jatropham 1 (90.0 mg, 796
5-ethoxypyrrolidin-2-one (18) (103 mg, 800
sulfonic acid monohydrate (10.0 mg, 52.6
m
mol) in THF (5 mL)
added. The mixture was stirred at room temperature for 1 h, after
which it turned opaque. The solvent was removed under reduced
pressure and the crude product was purified by column chroma-
tography (100 g silica) using DCM:MeOH (10:1) as the eluent. The
fractions of the major product (Rf ¼ 0.40) were combined and the
solvent was evaporated, giving the title compound 7 as a white
crystalline solid (54.0 mg, 43%), mp 162e164 ꢀC (lit.: 169e171 ꢀC)
[11]; Rf (DCM:MeOH ¼ 10:1) 0.40; nmax (KBr) 3246, 3077, 2923,
1680, 1661, 1384, 1286, 1239, 1078 cmꢁ1; HRMS: MþH ¼ 211.10758
(delta ¼ ꢁ0.7 ppm; C10H15O3N2); HR-ESI-MS-MS (CID ¼ 35%; rel.
int. %): 193 (100).
m
mol) and p-toluene-
mol) were added. The
m
mixture was stirred at room temperature for 1 h, after which a
white crystalline solid precipitated. The solvent was removed un-
der reduced pressure and the crude product was triturated in ethyl
acetate. The suspension was cooled to 0e5 ꢀC and the product was
filtered. The crystals were washed with cold ethyl acetate and were
dried by suction, giving the title compound 6 as a white crystalline
solid (95.0 mg, 61%), mp 174e178 ꢀC (lit.: 176e178 ꢀC) [12]; Rf
(DCM:MeOH ¼ 10:1) 0.35; nmax (KBr) 3209, 2998, 2977, 2924, 1718,
15

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
4.18. rel-(5S,30R,50R)-5-hydroxy-3-methyl-1-(30-methyl-20-
oxopyrrolidin-50-yl)-3-pyrrolin-2-one (7a)
4.22. 5-Hydroxy-3-methyl-1-(30-methyl-20-oxo-30-pyrrolin-50-yl)-
3-pyrrolin-2-one (8)
1H NMR (800 MHz; acetone-d6)
d
(ppm) 1.133 (3H; d; J ¼ 7.3 Hz;
To a solution of jatropham (1) (100 mg, 884
p-toluenesulfonic acid monohydrate (11.0 mg, 57.8
m
mol) in THF (5 mL)
mol) were
C(30)-CH3); 1.77 (3H; m; C(3)-CH3); 2.11 (1H; ddd; J ¼ 13.6, 8.9,
8.6 Hz; Ha-40); 2.54 (1H; ddd; J ¼ 13.6, 9.1, 1.2 Hz; Hb-40); 2.92 (1H;
ddq; 9.1, 8.9, 7.3 Hz; H-30); 5.24 (1H; d; J ¼ 10.0 Hz; C(5)-OH); 5.45
(1H; ddd; J ¼ 8.6, 1.4, 1.2 Hz; H-50); 5.57e5.59 (1H; m; H-5);
6.62e6.65 (1H; m; H-4); 7.30 (1H; br; NH-10); Lit. 1H NMR [11]: 1.12
(3H; d; J ¼ 7.2 Hz; C(30)-CH3); 1.77 (3H; dd; J ¼ 1.8, 1.3 Hz; C(3)-
CH3); 2.10 (1H; ddd; J ¼ 13.7, 9.5, 8.5 Hz; Hx-40); 2.54 (1H; dddd;
J ¼ 13.7, 8.7, 1.4, 0.8 Hz; Hy-40); 2.85 (1H; ddq; J ¼ 9.5, 8.7, 7.2 Hz; H-
30); 4.83 (1H; d; J ¼ 10.1 Hz; C(5)-OH); 5.40 (1H; ddd; J ¼ 8.5, 1.4,
1.4 Hz; H-50); 5.54 (1H; ddq; J ¼ 10.1, 1.8, 1.3 Hz; H-5); 6.20 (1H; qd;
J ¼ 1.8, 1.8 Hz; H-4); 6.86 (1H; br s; J ¼ 1.4 Hz, 0.8 Hz; NH-10); 13C
m
added. The mixture was stirred at room temperature for 1 h, after
which a white crystalline solid precipitated. The mixture was
evaporated under reduced pressure to ca. Half volume. The sus-
pension was cooled to 0e5 ꢀC and the product was filtered. The
crystals were washed with cold THF and were dried by suction,
giving the title compound 8 as a white crystalline solid (47.0 mg,
51%), mp 192e194 ꢀC (lit.: 194e198 ꢀC) [13]; Rf (DCM:MeOH ¼ 10:1)
0.45; nmax (KBr) 3326, 3222, 3073, 2920, 1706, 1678, 1661, 1645,
1402, 1291, 1098, 1075 cmꢁ1
;
HRMS: MþH
¼
209.09197
(delta ¼ ꢁ0.47 ppm; C10H13O3N2); HR-ESI-MS-MS (CID ¼ 35%, rel.
NMR (201 MHz; acetone-d6)
d
(ppm) 10.64 (C(3)-CH3); 16.5 (C(30)-
int. %): 191(100); 163(1).
CH3); 35.12 (C-40); 35.7 (C-30); 60.9 (C-50); 81.9 (C-5); 136.2 (C-3);
140.8 (C-4); 170.6 (C-2); 180.7 (C-20).
4.23. (5S*,50R*)-5-hydroxy-3-methyl-1-(30-methyl-20-oxo-30-
pyrrolin-50-yl)-3-pyrrolin-2-one (8a)
4.19. rel-(5R,30R,50R)-5-hydroxy-3-methyl-1-(30-methyl-20-
oxopyrrolidin-50-yl)-3-pyrrolin-2-one (7b)
1H NMR (500 MHz; methanol-d4)
d
(ppm) 1.85 (3H; dd; J ¼ 1.7,
1.4 Hz; C(3)-CH3); 1.90 (3H; t; J ¼ 1.7 Hz; C(30)-CH3); 5.39e5.41 (1H;
m; H-5); 5.95e5.97 (1H; m; H-50); 6.63e6.65 (1H; m; H-4);
6.76e6.77 (1H; m; H-40); Lit. 1H NMR [13]: 1.85 (3H; dd; J ¼ 1.7,
1.4 Hz; C(3)-CH3); 1.91 (3H; dd; J ¼ 1.7, 1.7 Hz; C(3)-CH3); 5.40 (1H;
dqd; J ¼ 1.8,1.4, 0.3 Hz; H-5); 5.96 (1H; dqdd; J ¼ 2.0,1.7, 0.4, 0.3 Hz;
H-50); 6.64 (1H; dqd; J ¼ 1.8, 1.7, 0.4 Hz; H-4); 6.76 (1H; dq; J ¼ 2.0,
1H NMR (800 MHz; acetone-d6)
d
(ppm) 1.127 (3H; d; J ¼ 7.2 Hz;
C(30)-CH3); 1.77 (3H; m; C(3)-CH3); 2.07 (1H; ddd; J ¼ 13.3, 9.2,
8.6 Hz; Ha-40); 2.55 (1H; dddd; J ¼ 13.3, 8.9, 1.2, 0.4 Hz; Hb-40); 2.88
(1H; ddq; J ¼ 9.2, 8.9, 7.3 Hz; H-30); 5.22 (1H; d; J ¼ 10.0 Hz; C(5)-
OH); 5.49e5.51 (1H; m; H-5); 5.62 (1H; ddd; J ¼ 8.6, 1.4, 1.2 Hz; H-
50); 6.62e6.65 (1H; m; H-4); 7.38 (1H; br; NH-10); 13C NMR
1.7 Hz; H-40); 13C NMR (126 MHz; methanol-d4)
d (ppm) 10.70,
10.76 (C(3)-CH3, C(30)-CH3); 64.0 (C-50); 82.5 (C-5); 136.3 (C-3);
139.0 (C-30); 139.4 (C-40); 142.4 (C-4); 172.6 (C-2); 175.5 (C-20).
(201 MHz; acetone-d6)
d
(ppm) 10.63 (C(3)-CH3); 16.4 (C(30)-CH3);
35.13 (C-30); 36.2 (C-40); 60.7 (C-50); 80.5 (C-5); 135.8 (C-3); 141.1
(C-4); 170.9 (C-2); 181.0 (C-20).
4.24. (5R*,50R*)-5-hydroxy-3-methyl-1-(30-methyl-20-oxo-30-
pyrrolin-50-yl)-3-pyrrolin-2-one (8b)
4.20. rel-(5S,30S,50R)-5-hydroxy-3-methyl-1-(30-methyl-20-
oxopyrrolidin-50-yl)-3-pyrrolin-2-one (7c)
1H NMR (500 MHz; methanol-d4)
d
(ppm) 1.86 (3H; dd; J ¼ 1.7,
1.4 Hz; C(3)-CH3); 1.88 (3H; t; J ¼ 1.7 Hz; C(30)-CH3); 5.24e5.26 (1H;
1H NMR partial assignment (800 MHz; acetone-d6)
d (ppm) 1.22
m; H-5); 6.10e6.12 (1H; m; H-50); 6.66e6.68 (1H; m; H-4);
(3H; d; J ¼ 7.1 Hz; C(30)-CH3); 1.77 (3H; m; C(3)-CH3); 2.17 (1H; ddd;
J ¼ 12.7, 9.1, 7.2 Hz; Hb-40); 2.49e2.53 (1H; m; H-30); 2.58 (1H; ddd;
J ¼ 12.7, 9.3, 7.6 Hz; Ha-40); 5.36 (1H; d; J ¼ 9.4 Hz; C(5)-OH); 5.57
(1H; dd; J ¼ 7.6, 7.2 Hz; H-50); 5.62e5.65 (1H; m; H-5); 6.62e6.65
(1H; m; H-4); 13C NMR partial assignment (201 MHz; acetone-d6)
6.77e6.79 (1H; m; H-40); 13C NMR (126 MHz; methanol-d4)
d (ppm)
10.67, 10.70 (C(3)-CH3, C(30)-CH3); 63.3 (C-50); 79.9 (C-5); 136.1 (C-
3); 136.9 (C-30); 141.7 (C-40); 142.8 (C-4); 173.3 (C-2); 176.1 (C-20).
d
(ppm) 15.9 (C(30)-CH3); 33.8 (C-40); 36.6 (C-30); 61.68 (C-50); 81.5
(C-5); 179.0 (C-20).
4.21. rel-(5R,30S,50R)-5-hydroxy-3-methyl-1-(30-methyl-20-
oxopyrrolidin-50-yl)-3-pyrrolin-2-one (7d)
1H NMR partial assignment (800 MHz; acetone-d6)
d (ppm) 1.20
(3H; d; J ¼ 6.9 Hz; C(30)-CH3); 1.77 (3H; m; C(3)-CH3); 2.39 (1H;
ddd; J ¼ 12.4, 9.6, 7.4 Hz; Hb-40); 2.49e2.53 (2H; m; H-30, Ha-40);
5.33 (1H; d; J ¼ 9.6 Hz; C(5)-OH); 5.63e5.66 (1H; m; H-5); 5.68 (1H;
t; J ¼ 7.4 Hz; H-50); 6.62e6.65 (1H; m; H-4); 13C NMR partial
4.25. Lilaline [8-(300-methyl-200-oxopyrrolidin-500-yl)kaempferol]
(14)
To a refluxing solution of kaempferol (15) (205 mg, 716
mmol) in
assignment (201 MHz; acetone-d6)
d (ppm) 10.6 (C(3)-CH3); 16.0
nitromethane (10 mL) 5-hydroxy-3-methylpyrrolidin-2-one (19)
(C(30)-CH3); 34.2 (C-40); 36.7 (C-30); 61.68 (C-50); 80.6 (C-5); 179.4
(C-20).
(99.0 mg, 860
(15.0 mg, 78.9
m
m
mol) and p-toluenesulfonic acid monohydrate
mol) were added. After 2 h yellow crystals
16

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
precipitated.
The
reaction
was
monitored
by
TLC
Ref. [17].
(DCM:MeOH ¼ 10:1). After completion of the reaction, the mixture
was cooled to 0 ꢀC and the crystalline product was filtered. The
crystals were washed with cold nitromethane and were dried un-
der an infrared lamp. The crude product weighed 272 mg (99%), a
100 mg sample of which was purified by preparative HPLC to yield
the pure C8 isomer 14 (the title compound) in the form of a yellow
crystalline solid (83.0 mg, 82%), mp 245e248 ꢀC (lit.: 247 ꢀC) [17]; Rf
(DCM:MeOH ¼ 10:1) 0.38; nmax (KBr) 3159, 2810, 2731, 2665, 1631,
1596,1400,1349 cmꢁ1; HRMS: MþH ¼ 384.10803 (delta ¼ 0.7 ppm;
4.28. Modeling
MacroModel [43] was used for an initial conformational search
and starting structure generation for the QM calculations. An OPLS3
force field [44] and water solvent were used (there is no methanol
on the list of accessible solvents, but it was useable for starting
structure generation.). The lowest energy structures in a 5 kcal/mol
energy window were optimized at the B3LYP-D3/6-31þG** level
[45e47] and the solution phase energy was determined using PBF
methanol implicit solvation model [48,49]. In order to better
sample the conformational space, the two lowest energy con-
formers having the hydrogen of the hydroxyl group rotated by
about 180ꢀ were selected for a relaxed torsional scan around the C-
5’ e N-1 bond in all cases. The torsional scans were carried out at
the B3LYP-D3/6-31þG** level in gas phase with a step size of 10ꢀ
(except 8a and 8 b where a step size of 5ꢀ was applied) and the
whole 0e360ꢀ interval was sampled. The effect of the solvent was
taken into account in the course of single point calculations at the
B3LYP-D3/6-31þG** level using PBF methanol solvation model.
Further details are given in the SI. The Jaguar software was used for
QM calculations [50]. Frequency calculations were carried out for
the optimized global minima and no imaginary frequency was
obtained. The effective proton-proton distances comparable to the
distances derived from the NOE intensities were calculated with
the following formula containing the Boltzmann factor and dꢁ6
averaging [31].
C
20H18O7N). HR-ESI-MS-MS (CID ¼ 35%; rel. int. %): 367(100);
339(24).
4.26
(300S*,500R*)-8-(300-Methyl-200-oxopyrrolidin-500-yl)kaempferol
(ppm) 1.13 (3H; d;
(trans-14): 1H NMR (500 MHz; DMSO‑d6)
d
J ¼ 7.4 Hz; C(300)-CH3); 1.99 (1H; ddd; J ¼ 12.7, 8.8, 5.0 Hz; Hx-400);
2.38 (1H; ddd; J ¼ 12.7, 9.6, 5.9 Hz; Hy-400); 2.54 (1H; dqd; J ¼ 9.6,
7.4, 5.0 Hz; H-300); 5.35 (1H; dd; J ¼ 8.8; 5.9 Hz; H-500); 6.26 (1H; s;
H-6); 6.87e6.91 (2H; m; H-30, H-50); 7.62e7.72 (1H; br; NH-100);
7.93e7.98 (2H; m; H-20, H-60);* 13C NMR (126 MHz; DMSO‑d6)
d
(ppm) 16.9 (C(300)-CH3); 34.2 (C-400); 36.2 (C-300); 45.2 (C-500); 98.8
(C-6); 101.9 (C-10); 106.2 (C-8); 115.4 (C-30, C-50); 121.7 (C-10);
129.4 (C-20, C-60); 135.2 (C-3); 146.4 (C-2); 154.2 (C-9); 159.2 (C-40);
159.4 (C-5); 165.0 (C-7); 175.6 (C-4); 179.3 (C-200). Lit. 1H NMR [17]:
(300 MHz; CD3OD)
d
(ppm) 1.29 (3H; d; J ¼ 7.4 Hz; C(300)-CH3); 2.17
(1H; ddd; J ¼ 12.9, 8.9, 4.9 Hz; Hx-400); 2.57 (1H; ddd; J ¼ 12.9, 9.7,
5.9 Hz, Hy-400); 2.77 (1H; dqd; J ¼ 9.7, 7.4, 4.9 Hz; H-300); 5.56 (1H;
dd; J ¼ 8.9, 5.9 Hz; H-500); 6.24 (1H; s; H-6); 6.91 (2H; d**;
8
9
20
1
3
ꢁ1=6
>
>
>
<
>
>
>
=
E
RT
J ¼ 9.0 Hz; H-30, H-50); 8.01 (2H; d**; J ¼ 9.0 Hz; H-20, H-60); Lit. 13
C
X
i
6B
6B
6B
4@
C
C
C
A
7
7
7
eꢁ
P
1
dx;i
d
(ppm) 17.5 (C(300)-CH3); 35.9 (C-400);
dx;effective
¼
E
NMR [17]: (76 MHz; CD3OD)
6
j
>
5>
>
>
RT
i
jeꢁ
38.5 (C-300); 48.1 (C-500); 99.3 (C-6); 104.8 (C-10); 106.8 (C-8); 116.5
(C-30, C-50); 123.6 (C-10); 130.8 (C-20, C-60); 137.1 (C-3); 148.5 (C-2);
155.9 (C-5); 160.7 (C-40); 161.8 (C-9); 163.7 (C-7); 177.5 (C-4); 183.9
(C-200).
>
:
>
;
where Ei and Ej is the calculated solution-phase energy of the
conformer i and j.
4.27
Declaration of competing interest
(300S*,500S*)-8-(300-Methyl-200-oxopyrrolidin-500-yl)kaempferol
The authors declare no conflict of interest.
Acknowledgments
(cis-14): 1H NMR (500 MHz; DMSO‑d6)
d (ppm) 1.05 (3H; d;
J ¼ 7.0 Hz; C(300)-CH3); 1.94 (1H; ddd; J ¼ 11.9, 11.5, 10.0 Hz; Hx-400);
2.34 (1H; ddd; J ¼ 11.9, 8.4, 6.6 Hz; Hy-400); 2.48 (1H; ddq; J ¼ 11.5,
8.4, 7.0 Hz; H-300); 5.26 (1H; dd; J ¼ 10.0, 6.6 Hz; H-500); 6.25 (1H; s;
H-6); 6.83e6.87 (2H; m; H-30, H-50); 7.79e7.87 (1H; br; NH-100);
7.96e8.00 (2H; m; H-20, H-60);* 13C NMR (126 MHz; DMSO‑d6)
The authors are grateful to Gedeon Richter Plc. for financial
ꢀ
ꢀ ꢀ
assistance. The authors wish to thank dr. Miklos Dekany (Gedeon
Richter Plc., Spectroscopic Research Department, MS laboratory)
and Barbara Herczeg (ComInnex, Inc.) for their contribution to the
results presented in this manuscript.
d
(ppm) 15.0 (C(300)-CH3); 35.3 (C-400); 36.6 (C-300); 45.8 (C-500); 98.6
(C-6); 102.1 (C-10); 104.5 (C-8); 115.3 (C-30, C-50); 121.7 (C-10);
129.5 (C-20, C-60); 135.2 (C-3); 146.3 (C-2); 154.7 (C-9); 159.2 (C-40);
159.5 (C-5); 165.2 (C-7); 175.5 (C-4); 178.2 (C-200).
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tet.2020.131827.
References
[1] (For information on lilies see:). (a) H.J. Elwes, A Monograph of the Genus
Lilium, Taylor and Francis, London, 1880;
(b) H.B.D. Woodcock, W.T. Stearn, Lilies of the World: Their Cultivation and
Classification, Country Life, London, 1950;
(c) P.M. Synge, Lilies: A Revision of Elwes’ Monograph of the Genus Lilium and
its Supplements, Universe Books, New York, 1980;
(d) E.A. McRae, Lilies: A Guide for Growers and Collectors, Timber Press,
Oregon, 1998.
*The OH protons of the lilaline molecules were not observed as
distinct resonances because the sample contained formic acid and
water; the OH protons experienced fast chemical exchange and
gave a common broad resonance at 4 ppm. **Actually, these reso-
nances are AA’XX’ multiplets, but were described as doublets in
[2] (For more information on the medicinal aspects of lilies see:). (a) M. Grieve,
A Modern Herbal, vols. 1e2, Dover Publications, New York, 1971;
(b) D. Brown, Encyclopedia of Herbs and Their Uses, Dorling Kindersley,
17

ꢀ
ꢀ
S. Nagy, A. Szigetvari, V. Ilkei et al.
Tetrahedron 81 (2021) 131827
London, 1995;
(d) B.E. Maryanoff, H. Zhang, J.H. Cohen, I.J. Turchi, C.A. Maryanoff, Chem. Rev.
104 (2004) 1431e1628;
(e) J. Royer, M. Bonin, L. Micouin, Chem. Rev. 104 (2004) 2311e2352;
(f) A. Yazici, S.G. Pyne, Synthesis (2009) 339e368;
(c) R. Chiej, Encyclopedia of Medicinal Plants (Macdonald Encyclopedias),
MacDonald, London, 1984;
(d) N. Al-Bayati, J. Pharmaceut. Sci. Res. 10 (2018), 2014e2016;
(e) J. Patocka, Z. Biomed Navratilova, J. Sci.
&
Tech. Res. 18 (2019)
(g) A. Yazici, S.G. Pyne, Synthesis (2009) 513e541.
13859e13862;
[22] J.C. Hubert, J.B.P.A. Wijnberg, W.N. Speckamp, Tetrahedron 31 (1975)
1437e1441.
[23] B.L. Bray, P.H. Mathies, R. Naef, D.R. Solas, T.T. Tidwell, D.R. Artis,
J.M. Muchowski, J. Org. Chem. 55 (1990) 6317e6328.
[24] I.M. Downie, M.J. Earle, H. Heaney, K.F. Shuhaibar, Tetrahedron 49 (1993)
4015e4034.
[25] D.O.A. Garrido, G. Buldain, M.I. Ojea, B. Frydman, J. Org. Chem. 53 (1988)
403e407.
(f) P. Wang, J. Li, F.A.K. Attia, W. Kang, J. Wei, Z. Liu, C. Li, Food Sci. Hum. Well. 8
(2019) 330e336;
(g) S. Momtaz, M. Dibaj, A. Abdollahi, G. Amin, R. Bahramsoltani, M. Abdollahi,
P. Mahdaviani, A.H. Abdolghaffari, J. Med. Plants 19 (2020) 109e118;
(h) M. Zaccai, L. Yarmolinsky, B. Khalfin, A. Budovsky, J. Gorelick, A. Dahan,
S. Ben-Shabat, Plants 9 (2020) 959.
[3] R.M. Wiedhopf, E.R. Trumbull, J.R. Cole, J. Pharmacol. Sci. 62 (1973)
1206e1207.
[4] K. Yakushijin, M. Kozuka, Y. Ito, R. Suzuki, H. Furukawa, Heterocycles 14
(1980) 1073e1076.
[26] G.P. Gardini, V. Bocchi, Gazz. Chim. Ital. 102 (1972) 91e101.
[27] J. Bessho, Y. Shimotsu, S. Mizumoto, N. Mase, H. Yoda, K. Takabe, Heterocycles
63 (2004) 1013e1016.
[5] K. Yakushijin, R. Suzuki, R. Hattori, H. Furukawa, Heterocycles 16 (1981)
1157e1160.
[28] (a) N. Mase, T. Nishi, M. Hiyoshi, K. Ichihara, J. Bessho, H. Yoda, K. Takabe,
J. Chem. Soc., Perkin Trans. 1 (2002) 707e709;
[6] T. Nagasaka, S. Esumi, N. Ozawa, Y. Kosugi, F. Hamaguchi, Heterocycles 16
(1981), 1987e1992.
[7] A. Vachalkova, E. Eisenreichova, M. Haladova, P. Mucaji, B. Jozova, L. Novotný,
Neoplasma 47 (2000) 313e318.
(b) N. Mase, T. Nishi, Y. Takamori, H. Yoda, K. Takabe, Tetrahedron: Asym-
metry 10 (1999) 4469e4471.
[29] P. Kuehne, M. Hesse, Tetrahedron 49 (1993) 4575e4580.
[30] (For relevant studies on regioselective phenolic Mannich reactions see:). (a)
M. Tramontini, L. Angiolini, Mannich Bases e Chemistry and Uses, CRC Press,
Boca Raton, 1994, pp. 32e38;
ꢀ
ꢀ
ꢀ
ꢀ
ꢁ
ꢁ
ꢀ
[8] G. Jovtchev, S. Gateva, A. Stankov, Environ. Toxicol. 31 (2016) 751e764.
[9] H. Shimomura, Y. Sashida, Y. Mimaki, Y. Minegishi, Phytochemistry 26 (1987)
582e583.
(b) Y. Omura, Y. Taruno, Y. Irisa, M. Morimoto, H. Saimoto, Y. Shigemasa,
Tetrahedron Lett. 42 (2001) 7273e7275.
ꢀ
ꢀ
ꢁ
ꢀ
[10] E. Eisenreichova, M. Haladova, A. Buckova, K. Ubik, D. Uhrín, Chem. Pap. 45
(1991) 709e711.
[31] D. Neuhaus, M.P. Williamson, The Nuclear Overhauser Effect in Structural and
Conformational Analysis, second ed., Wiley, New York, 2000.
[32] S.G. Smith, J.M. Goodman, J. Org. Chem. 74 (2009) 4597e4607.
[33] N. Grimblat, M.M. Zanardi, A.M. Sarotti, J. Org. Chem. 80 (2015) 12526e12534.
[34] G. Tarazona, G. Benedit, R. Fernandez, M. Perez, J. Rodríguez, C. Jimenez,
C. Cuevas, J. Nat. Prod. 81 (2018) 343e348.
ꢀ
ꢀ
ꢁ
ꢀ
[11] M. Haladova, E. Eisenreichova, A. Buckova, J. Tomko, D. Uhrín, Collect. Czech
Chem. Commun. 53 (1988) 157e160.
ꢀ
ꢀ
ꢁ
ꢀ
[12] E. Eisenreichova, M. Haladova, A. Buckova, J. Tomko, D. Uhrín, K. Ubik,
ꢀ
ꢀ
ꢀ
Phytochemistry 31 (1992) 1084e1085.
ꢀ
ꢀ
ꢁ
ꢀ
[13] M. Haladova, E. Eisenreichova, A. Buckova, J. Tomko, D. Uhrín, K. Ubik, Collect.
Czech Chem. Commun. 56 (1991) 436e438.
[14] F. Pavelcík, Acta Crystallogr. C45 (1989) 956e957.
[35] C.P. Butts, C.R. Jones, E.C. Towers, J.L. Flynn, L. Appleby, N. Barron, J. Org.
Biomol. Chem. 9 (2011) 177e184.
ꢁ
[15] K. Ori, Y. Mimaki, K. Mito, Y. Sashida, T. Nikaido, T. Ohmoto, A. Masuko,
Phytochemistry 31 (1992) 2767e2775.
[16] T. Satou, Y. Mimaki, M. Kuroda, Y. Sashida, Y. Hatakeyama, Phytochemistry 41
(1996) 1225e1230.
[36] C.R. Jones, C.P. Butts, J.N. Harvey, Beilstein J. Org. Chem. 7 (2011) 145e150.
[37] Engineering ToolBox, Dynamic viscosity of common liquids, Accessed on 18
Oct 2019, https://www.engineeringtoolbox.com/absolute-viscosity-liquids-d_
1259.html, 2008.
[38] S. Boros, Z. Gaspari, G. Batta, Annu. Rep. NMR Spectrosc. 94 (2018) 1e39.
[39] S.H. Mari, P.C. Varras, A. Wahab, I.M. Choudhary, M.G. Siskos,
I.P. Gerothanassis, Molecules 24 (2019) 2290.
[40] L.B. Casabianca, Magn. Reson. Chem. 58 (2020) 611e624.
[41] R.E. Lancaster Jr., C.A. VanderWerf, J. Org. Chem. 23 (1958) 1208e1209.
[42] M. Hall, C. Stueckler, H. Ehammer, E. Pointner, G. Oberdorfer, K. Gruber,
B. Hauer, R. Stuermer, W. Kroutil, P. Macheroux, K. Faber, Adv. Synth. Catal.
350 (2008) 411e418.
ꢁ
ꢀ
ꢀ
ꢀ
[17] I. Masterova, D. Uhrín, J. Tomko, Phytochemistry 26 (1987) 1844e1845.
ꢀ
ꢁ
ꢀ
ꢁ
ꢀ
ꢀ
[18] (a) E. Eisenreichova, I. Masterova, A. Buckova, M. Haladova, J. Tomko,
ꢁ
Ceskoslov. Farm. 34 (1985) 408e409;
(b) E. Eisenreichova, M. Haladova, P. Mucaji, D. Grancai, Acta Fac. Pharm. Univ.
ꢀ
ꢀ
ꢁ
ꢁ
Comenianae 51 (2004) 27e37.
[19] A. Buckova, E. Eisenreichova, M. Haladova, D. Uhrín, J. Tomko, Phytochemistry
ꢁ
ꢀ
ꢀ
ꢀ
27 (1988) 1914e1915.
ꢀ
ꢀ
[20] (For our previous works in the field see:). (a) V. Ilkei, K. Farago, Z. Santa,
ꢀ ꢀ
ꢀ
ꢀ
€
€
M. Dekany, L. Hazai, C. Szantay Jr., C. Szantay, G. Kalaus, Int. J. Org Chem. 4
[43] Schrodinger Release 2018-2, MacroModel, Schrodinger, LLC, New York, NY,
2018.
(2014) 309e313;
ꢀ
ꢀ
ꢀ
(b) V. Ilkei, P. Bana, F. Toth, A. Pallo, T. Holczbauer, M. Czugler, Z. Santa,
[44] E. Harder, W. Damm, J. Maple, C. Wu, M. Reboul, J.Y. Xiang, L. Wang,
D. Lupyan, M.K. Dahlgren, J.L. Knight, J.W. Kaus, D.S. Cerutti, G. Krilov,
W.L. Jorgensen, R. Abel, R.A. Friesner, J. Chem. Theor. Comput. 12 (2016)
281e296.
ꢀ
ꢀ ꢀ
ꢀ
ꢀ
ꢀ
M. Dekany, A. Szigetvari, L. Hazai, C. Szantay Jr., C. Szantay, G. Kalaus, Tetra-
hedron 71 (2015) 9579e9586;
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ ꢀ
ꢀ
(c) V. Ilkei, A. Spaits, A. Prechl, A. Szigetvari, Z. Beni, M. Dekany, C. Szantay Jr.,
ꢀ
ꢀ
€
€
J. Müller, A. Konczol, A. Szappanos, A. Mandi, S. Antus, A. Martins, A. Hunyadi,
[45] A.D. Becke, J. Chem. Phys. 98 (1993) 5648e5652.
€
ꢀ
G.T. Balogh, G. Kalaus, H. Bolcskei, L. Hazai, T. Kurtan, Beilstein J. Org. Chem. 12
[46] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785e789.
[47] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 132 (2010) 154104.
[48] D.J. Tannor, B. Marten, R. Murphy, R.A. Friesner, D. Sitkoff, A. Nicholls,
M. Ringnalda, W.A. Goddard III, B. Honig, J. Am. Chem. Soc. 116 (1994)
11875e11882.
(2016) 2523e2534;
ꢀ
€
€
ꢀ
(d) V. Ilkei, A. Spaits, A. Prechl, J. Müller, A. Konczol, S. Levai, E. Riethmüller,
ꢀ
ꢀ
ꢀ
ꢀ ꢀ
A. Szigetvari, Z. Beni, M. Dekany, A. Martins, A. Hunyadi, S. Antus,
ꢀ
€
C. Szantay Jr., G.T. Balogh, G. Kalaus, H. Bolcskei, L. Hazai, Tetrahedron 73
(2017) 1503e1510.
[49] B. Marten, K. Kim, C. Cortis, R.A. Friesner, R.B. Murphy, M.N. Ringnalda,
D. Sitkoff, B. Honig, J. Phys. Chem. 100 (1996) 11775e11788.
[50] A.D. Bochevarov, E. Harder, T.F. Hughes, J.R. Greenwood, D.A. Braden,
D.M. Philipp, D. Rinaldo, M.D. Halls, J. Zhang, R.A. Friesner, Int. J. Quant. Chem.
[21] (For reviews on the chemistry of N-acyliminium ions and the synthesis of
their precursors see:). (a) W.N. Speckamp, H. Hiemstra, Tetrahedron 41 (1985)
4367e4416;
€
€
(b) W.N. Speckamp, M.J. Moolenaar, Tetrahedron 56 (2000) 3817e3856;
(c) C.M. Marson, ARKIVOC 1e16 (2001);
113 (2013) 2110e2142 (Schrodinger Release 2018-1: Jaguar, Schrodinger,
LLC, New York, NY, 2018.).
18