A modular approach to the bisbenzylisoquinoline alkaloids tetrandrine and isotetrandrine

Article abstract of DOI:10.1039/d0ob00078g

An efficient racemic total synthesis of the bisbenzylisoquinoline alkaloids tetrandrine and isotetrandrine in four different routes is reported herein. Key steps of the synthesis include N-acyl Pictet-Spengler condensations to access the tetrahydroisoquin

Full text of DOI:10.1039/d0ob00078g

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
A modular approach to the bisbenzylisoquinoline
alkaloids tetrandrine and isotetrandrine†‡
Cite this: DOI: 10.1039/d0ob00078g
a
Ramona Schütz, a Maximilian Meixner,b Iris Antes*b and Franz Bracher
*
An efficient racemic total synthesis of the bisbenzylisoquinoline alkaloids tetrandrine and isotetrandrine in
four different routes is reported herein. Key steps of the synthesis include N-acyl Pictet–Spengler con-
densations to access the tetrahydroisoquinoline moieties, as well as copper-catalyzed Ullmann couplings
for diaryl ether formation. Starting from commercially available building blocks tetrandrine and isotetran-
drine are accessed in 12 steps. Depending on the sequence of the four central condensation steps, equi-
molar mixtures of both diastereomers or predominantly tetrandrine or its diastereomer isotetrandrine are
obtained. Through computational analysis we were able to rationalize the differences in the observed dia-
stereomeric specificities.
Received 14th January 2020,
Accepted 13th February 2020
DOI: 10.1039/d0ob00078g
rsc.li/obc
reduced by pharmacological inhibition of TPC1 and TPC2
in vitro and in vivo using tetrandrine (1). Another important
Introduction
Tetrandrine (1) is a bisbenzylisoquinoline alkaloid isolated pharmacological target of tetrandrine (1) is P-glycoprotein
from Stephania tetrandra (Menispermaceae), a climbing plant (P-gp), a universal efflux pump for xenobiotics responsible for
native to Asia.1 This herb has been used in Chinese and multidrug resistance of tumors. Overexpression of P-gp in
Japanese traditional medicine for treating a variety of diseases cancer cells is a key factor of drug resistance towards a variety
such as tuberculosis, malaria, asthma, hyperglycaemia, hyper- of structurally different antitumor agents.12 Among a series of
tension and cancer.2 First isolated in 1928 by Kondo and isoquinoline-type alkaloids, tetrandrine (1) was identified as
Yano,3,4 tetrandrine (1) was identified as one of the active an outstanding inhibitor of P-gp.13,14 Under the name CBT-1™
agents of the herb. Far beyond the traditional use the pharma- this alkaloid has advanced to clinical studies.15
cological effects of tetrandrine (1) in its pure form have been
Previous efforts in antitumor drug discovery evolved a
subject of numerous studies. Most importantly, tetrandrine (1) number of semi-synthetic tetrandrine derivatives. Hereby,
has been identified as an antagonist of calcium channels.5,6 structural modifications started from this compound, available
Thereby two-pore channels (TPC), voltage gated calcium chan- from plant sources and mainly focused on introduction of sub-
nels located on lysosomal membranes,7 emerged as an inter- stituents on C-516–18 as well as C-1419–21 by electrophilic substi-
esting pharmacological target recently, since they play a role in tution reactions on the electron-rich aromatic rings, further on
the pathomechanism of diseases such as Ebola virus infection N-alkylation at N-2′ to give quaternary ammonium salts.22
and cancer.8–10 Hereby, tetrandrine (1) was shown to be a
Besides its heterogeneous pharmacological profile also
potent inhibitor of Ebola virus entry in vivo as well as in vitro possible toxic side effects were reported for tetrandrine (1).23
in sub-micromolar concentrations (IC50 value of 55 nM)
through inhibiting TPC channels (Fig. 1).8
TPC channels are also involved in tumor metastasis.9
Nguyen et al.9 demonstrated that cancer cell migration can be
aDepartment of Pharmacy – Center for Drug Research, Ludwig-Maximilians
University Munich, Butenandtstr. 5-13, 81377 Munich, Germany.
E-mail: Franz.Bracher@cup.uni-muenchen.de
bTUM School of Life Sciences, Technical University Munich, Emil-Erlenmeyer-Forum
8, 85354 Freising, Germany
†In memory of Prof. Dr Peter Pachaly, a pioneer in the field of bisbenzylisoqui-
noline alkaloids, who passed away on January 10, 2019.
‡Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0ob00078g
Fig. 1 Tetrandrine (1) (1S,1’S) and isotetrandrine (2) (1R,1’S)
Numbering of the skeleton according to the convention established by
Shamma.11
–
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Oxidative metabolism involving the methoxy group at C-12 and further, by using N-alkoxycarbonyl residues in the N-acyl-
results in the formation of a highly reactive quinone methide arylethylamine building blocks, the required N-methyl groups
prone to reaction with bio-nucleophiles, and suspected to be should be available in only one additional step by lithium
responsible for pulmonary damage in animal models.24,25 alanate reduction of the carbamate groups. This should lead
Furthermore, after continuous administration of tetrandrine (1) to a significant reduction of the required number of steps.
a pathological change of liver tissues was observed in dogs.26
Alternatively, hydrolysis of the carbamate groups would
All of these data make tetrandrine (1) a very interesting lead provide secondary amino groups (N-2-H, N-2′-H), which in turn
compound for further development as a chemical tool for could subsequently be N-alkylated to give various tertiary
pharmacological studies and as a drug candidate. Systematic amino groups. Further, ώ-alkoxystyrenes (instead of unstable
structure modifications should pave the way for increasing arylacetaldehydes) were envisaged as building blocks for the
selectivity and decreasing toxicity. Since options for semi-syn- N-acyl Pictet–Spengler cyclizations following Comins’
thetic variations starting from tetrandrine (1) from plant approach.30 This should reduce the number of steps further
sources are limited and likely to be already exploited we aimed compared to the original method, since ώ-alkoxystyrenes are
to develop a new synthetic route allowing to flexibly implement available from the corresponding substituted benzaldehydes
diverse structure variations.
in one single step, whereas construction of arylacetic acids for
The first published total synthesis of tetrandrine (1), its Bischler–Napieralski reactions requires a number of steps.
enantiomer phaeanthine, and its diastereomer isotetrandrine Further, we aimed at exploring modern alternatives such as
(2) by Inubushi and coworkers, first published in 1968,27,28 Buchwald–Hartwig coupling31 or Chan–Evans–Lam reaction32
implies more than 20 steps. Key steps are two copper-mediated for the construction of the diaryl ether moieties.
Ullmann couplings for the formation of the diaryl ether moi-
eties and two Bischler–Napieralski reactions for the construc- tetrandrine (1) (and its isomers) were explored: In route 1 both
tion of the two benzylisoquinoline units. The final steps of this 1-benzyltetrahydroisoquinoline moieties meaning both
Based on these considerations several alternative routes to
–
total synthesis are an intramolecular Bischler–Napieralski cycli- stereocenters at C-1 and C-1′ in consequence – were to be built
zation, followed by reduction of the resulting dihydroisoquino- up early in intermolecular N-acyl Pictet–Spengler reactions,
line intermediate and N-methylation. Very recently, a similar and the central step for the construction of the macrocycle
approach using the same key reactions was published in a should be an intramolecular diaryl ether synthesis at a late
Chinese patent.29 This approach includes two successive stage. In contrast, in route 2 a late intramolecular N-acyl
Ullmann couplings of racemic benzyltetrahydroisoquinoline Pictet–Spengler cyclization should provide the macrocycle
units as the final steps, and consequently lacks stereoselectivity. under construction of the second stereocenter. For each route
Hereafter, we present a new 12 step total synthesis of racemic an additional variation was designed, considering that either
tetrandrine (rac-1) and its diastereomer isotetrandrine (rac-2).
the benzyltetrahydroisoquinoline moiety consisting of rings A–
C or the one consisting of rings A′–C′ can be constructed first.
Besides the general feasibility of both approaches, the stereo-
chemical outcome of all four conceivable variants was of high
interest, as high diastereoselectivity of the whole protocol was
Results and discussion
In contrast to the published total syntheses of tetrandrine (1), desirable.
our retrosynthetic approach aimed at utilizing N-acyl Pictet–
The sequence of the Pictet–Spengler- and Ullmann-type
Spengler cyclizations instead of Bischler–Napieralski protocols. steps of the four variants is depicted in Fig. 2. In detail, in the
This approach would provide tetrahydroisoquinolines directly, variants 1a and 1b the retrosynthetic approach for the syn-
Fig. 2 Retrosynthetic analysis of the envisaged routes to tetrandrine precursor 15.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 Route 1a: Synthesis of the benzyltetrahydroisoquinoline 8.
thesis of the macrocyclic skeleton consists of two alternating (TfOH) as a stronger acidic catalyst35 and obtained 8 in a com-
N-acyl Pictet–Spengler condensations and C–O couplings. parable yield of 49%. But due to easier handling we preferred
Finally the macrocycle is generated in an intramolecular using TFA in the following procedures. The N-acyl Pictet–
Ullmann coupling (“U-2”) of a seco-bisbenzylisoquinoline Spengler condensation in this step delivers exclusively the
intermediate, connecting rings C and C′. In route 2, both desired regioisomer 8. There was no by-product found in
diaryl ethers are constructed early (operations “U-1” and which cyclization took place at the para position of bromine.
“U-2”), and the macrocycle is generated in an intramolecular
Next, the first C–O coupling was performed in order to
N-acyl Pictet–Spengler condensation, whereby in route 2a ring obtain diaryl ether 11 (Scheme 2). The required phenol 10 was
B′ is constructed in this late operation (“P-S-2”), whereas in prepared from commercially available 3-methoxytyramine (9;
route 2b this final connection is performed for construction of itself readily available from the aromatic aldehyde vanilline)
ring B (“P-S-1”).
and ethyl chloroformate. For the coupling of 8 with 10 several
In route 1a the highly substituted 1-benzyltetrahydroisoqui- reaction conditions (Table 1) were explored testing different
noline 8, which covers rings A–C of tetrandrine (1), was pre- ligand/catalyst systems. Palladium-catalyzed Buchwald–
pared from the commercially available aldehydes 5-bromovera- Hartwig cross-coupling reactions (Table 1, entries 1–3) using
traldehyde (3) and O-benzylisovanilline. In a Henry reaction 3 different phosphine ligands36 or Pd-PEPPSI™-IPr as catalyst
was condensed with nitromethane33 to give β-nitrostyrene 4 in failed to furnish the desired diaryl ether 11. Next a set of cata-
60% yield. Reduction with zinc/HCl33 gave the corresponding lytic systems for Ullmann-type C–O couplings was explored
arylethylamine 5, and subsequent reaction with ethyl chloro- employing different combinations of copper(I) halides, ligands
formate led to carbamate 6 in 70% yield over both steps. and bases (entries 4–8). Using the combination of copper(I)
Known enol ether 734 was synthesized in a Wittig olefination iodide, N,N-dimethylglycine and caesium carbonate in
of O-benzylisovanilline in 87% yield. The following N-acyl dioxane37 we were able to detect traces of the desired product
Pictet–Spengler condensation30 with carbamate 6, catalyzed by 11 (entry 4) by LC-MS. Increasing catalyst and ligand to stoi-
trifluoroacetic acid (TFA), gave racemic tetrahydroisoquinoline chiometric amounts did not improve the yields. Employing
8 in 47% yield (Scheme 1). This moderate yield might be due copper(I) bromide-dimethylsulfide complex with caesium car-
to the sterically demanding bromine substituent. We tried to bonate in pyridine38 without any additional ligand (entry 9) we
enhance the yield by using trifluoromethanesulfonic acid were able to isolate minor amounts (2%) of the desired diaryl
Scheme 2 Route 1a: First Ullmann coupling.
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Conditions for Ullmann-type C–O coupling
Equiv.
phenol 10 (%)
Yield
Entry Conditions
1
2
3
4
5
6
7
8
Pd2(dba)3 (1.5 mol%), tBuXPhos (2.0 mol%), K3PO4 (2.0 equiv.), toluene, 100 °C, 48 h (ref. 36)
Pd(AcO)2 (5.0 mol%), Me4tButylXphos (7.0 mol%), K3PO4 (2.0 equiv.), toluene, 100 °C, 48 h (ref. 36)
PEPPSI-IPr (2.0 mol%), K3PO4 (2.0 equiv.), toluene, 100 °C, 48 h
CuI (0.1 equiv.), N,N-dimethylglycine (0.3 equiv.), Cs2CO3 (2.0 equiv.), dioxane, 105 °C, 72 h, pressure tube (ref. 37)
CuI (1.0 equiv.), N,N-dimethylglycine (3.0 equiv.), Cs2CO3 (2.0 equiv.), dioxane, 105 °C, 72 h, pressure tube (ref. 37)
1.2
1.2
1.2
1.5
1.5
—
—
—
Traces
Traces
—
—
—
CuI (0.1 equiv.), N,N-dimethylglycine (0.3 equiv.), Cs2CO3 (2.0 equiv.), dioxane, 150–200 °C, 2.5 h, microwaves (ref. 39) 1.5
CuI (0.1 equiv.), picolinic acid (0.2 equiv.), K3PO4 (2.0 equiv.), DMSO, 90 °C, 72 h (ref. 41)
CuBr (1.0 equiv.), 1,1′-azobis(cyclohexane carbonitrile) (1.0 equiv.), Cs2CO3 (2.0 equiv.), DMF, 100 °C, 4 h, microwaves
(ref. 42)
1.2
1.1
9
CuBr·Me2S (1.0 equiv.), Cs2CO3 (3.0 equiv.), pyridine, 150–200 °C, 72 h, pressure tube (ref. 38)
CuBr·Me2S (1.0 equiv.), Cs2CO3 (3.0 equiv.), pyridine, 220 °C, 3 h, microwaves
CuBr·Me2S (1.0 equiv.), Cs2CO3 (3.0 equiv.), pyridine, 110 °C, 7 d
CuBr·Me2S (1.0 equiv.), Cs2CO3 (3.0 equiv.), pyridine, 110 °C, 7 d
CuBr·Me2S (1.0 equiv.), Cs2CO3 (3.0 equiv.), pyridine, 110 °C, 7 d
1.1
1.1
1.1
1.5
2.0
2
10
11
12
13
—
12
27
41
ether 11. A predominant side reaction was debromination of acid from bromoarene 8 was unsuccessful. Diaryl ether 11 was
the aryl halide 8 at high reaction temperatures. Microwave then O-debenzylated by standard Pd-catalyzed hydrogenolysis
conditions39,40 instead of conventional heating could not giving phenol 12 in 92% yield. The debenzylation had to be
improve the reaction in terms of yield and by-product for- performed at this stage, since surprisingly the O-benzylated
mation (entry 10). To minimize debromination we decreased intermediate 11 was found to react very sluggishly in an
the temperature to 110 °C and prolonged the reaction time. attempted subsequent N-acyl Pictet–Spengler reaction with
This modification resulted in a better yield (12%; entry 11). By enol ether 13.
increasing the amount of phenol 10 from 1.1 to 1.5 equivalents
For the construction of the second tetrahydroisoquinoline
we were able to further improve the yield to 27% (entry 12). moiety our method of choice was again a TFA-mediated N-acyl
The best result was obtained when performing the reaction at Pictet–Spengler condensation, connecting intermediate 12
110 °C with 2.0 equivalents of the phenol 10 furnishing diaryl with known enol ether 1343 to give seco-bisbenzylisoquinoline
ether 11 in 41% yield (entry 13). Considering the comparatively 14 in an excellent yield of 96% (Scheme 3) and again with
low reactivity of electron-rich and sterically hindered aryl desired regioselectivity. As expected we obtained both racemic
halide 8, the obtained yield is satisfying and comparable to diastereomers in a ratio of 1 : 1 (determined by HPLC). No
the yield of 42% in Inubushi’s approach27 using related build- stereocontrol was observed, since the newly formed stereocen-
ing blocks. An attempted Chan–Lam–Evans coupling32 was ter is far away from the stereocenter in the starting material.
discarded after the synthesis of the corresponding boronic But fortunately, the obtained diastereomers of 14 were easily
Scheme 3 Route 1a: Synthesis of racemic tetrandrine (rac-1) and racemic isotetrandrine (rac-2) via seco-bisbenzylisoquinoline 14.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
separated by flash column chromatography on a preparative phenolic intermediate 17 and aryl bromide 6 (see Scheme 1)
scale. The isolated yields of the racemic diastereomers were we once again employed the established catalytic system
50% of the (R,R)/(S,S) isomers (14a) and 46% of the (R,S)/(S,R) (Scheme 4).
isomers (14b). The relative configurations of these products
This critical key step could significantly be enhanced in
were determined retrospectively after conversion into racemic this approach compared to route 1a (Scheme 2). The diaryl
tetrandrine (rac-1) and racemic isotetrandrine (rac-2), ether was now obtained in a yield of 64% (vs. 41%) in a shorter
respectively.
reaction time (3 days vs. 7 days). Also in terms of efficiency this
For the following intramolecular C–O coupling to generate step was improved, since only 1.0 equiv. of phenol instead of
the macrocycle 15 we once again conducted a screening for 2.0 equiv. was brought to reaction with 1.2 equiv. of aryl
reaction conditions. Pd2(dba)3 in combination with phosphine bromide 6. The improved yield can be explained by the fact
ligand Me4tButylXphos,36 copper(I) iodide combined with that aryl bromide 6 is less sterically hindered than the aryl
ligand N,N-dimethylglycine37 and copper(I) bromide-dimethyl- halide 8 in route 1a. Diaryl ether 18 was further debenzylated
sulfide complex38 were tested as catalytic systems on a mixture by hydrogenolysis in methanol to give phenol 19 in 94% yield
of diastereomers 14 (compare entries 2, 4 and 11 in Table 1). (Scheme 4).
All of these reactions were carried out under high dilution con-
Next seco-bisbenzylisoquinoline 21 was obtained in an
ditions (0.02 M) in order to suppress intermolecular coupling N-acyl Pictet–Spengler reaction of 19 with enol ether 20,45
reactions. Again copper(I) bromide-dimethylsulfide complex which was prepared in 80% yield by Wittig olefination of the
catalyzed the diaryl ether synthesis well, furnishing the corresponding aromatic aldehyde. To our surprise (compared
desired macrocycle 15 in 62–64% yield (Scheme 3), whereas to route 1a) the product of this Pictet–Spengler reaction follow-
the other catalytic systems failed. To our surprise, separation ing our standard protocol was not an equimolar racemic
of diastereomers of 15 via flash column chromatography could mixture of diastereomers, but the (R,R)/(S,S) isomers (21a)
not be achieved. Thus, for generation of pure racemic alkaloids were formed preferably (ratio 62 : 38, determined by HPLC) in
the cyclization had to be conducted with the previously iso- an isolated yield of 30% (Table 2, entry 1), whereas the (S,R)/
lated racemic diastereomers of the seco-intermediate 14. Both (R,S) isomers (21b) were obtained in 19% yield. As in route 1a,
isomers were cyclized in almost identical yields of 62 and 64%. the diastereomers were easily separable using flash column
To complete the total synthesis of the alkaloids, carbamates chromatography. We further isolated a by-product in 24%
15 were reduced using LiAlH4 in THF to give racemic tetran- yield, which was identified as another Pictet–Spengler product
drine (rac-1) and racemic isotetrandrine (rac-2), both in almost with the same mass as the desired product. Since the fraction
quantitative yield (98%) (Scheme 3). The analytical data of of the by-product was an inseparable mixture of diastereomers,
racemic tetrandrine (rac-1) was identical with those of an auth- exact identification was not possible, but most likely it is a
entic natural sample of tetrandrine (1), kindly donated by
Prof. P. Pachaly.
As a variation of this approach we elaborated route 1b,
Table 2 Route 1b: Conditions for synthesis of seco-bisbenzylisoquino-
line 21
which starts with the synthesis of the benzyltetrahydroisoqui-
noline unit consisting of rings A′–C′. Following the established
protocol of TFA-catalyzed N-acyl Pictet–Spengler reaction, car-
Temperature Ratio (S,S)/(R, Product/by-
bamate 10 was condensed with known enol ether 1644 to give
Entry Solvent Acid
(°C)
R) : (S,R)/(R,S) product
racemic 1-benzyltetrahydroisoquinoline 17 in 96% yield
(Scheme 4). Enol ether 16 was synthesized by Wittig olefina-
tion of commercially available 4-benzyloxybenzaldehyde in
97% yield. For the Ullmann synthesis of diaryl ether 18 from
1
2
3
4
DCM
DCM
DCM
TFE
TFA
TFA
TfOH rt
rt
−15
62 : 38
66 : 34
47 : 53
54 : 46
68 : 32
66 : 34
69 : 31
80 : 20
TFA
rt
Scheme 4 Route 1b: Synthesis of diaryl ether 19.
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
regioisomer of seco-bisbenzylisoquinoline 21, resulting from col with preference for the one or other of the diastereomers.
cyclization to the other free ortho position of precursor 19. In We presumed that altering the sequence of reactions steps,
order to decrease the amount of this by-product and to particularly utilizing an intramolecular N-acyl Pictet–Spengler
improve the diastereomeric ratio we performed the N-acyl reaction for ring closure and construction of the second stereo-
Pictet–Spengler reaction under different conditions (Table 2). center, should lead to an asymmetric induction as a result of
Conducting the reaction at lower temperature (entry 2) had the influence of the stereochemistry of the already present
only minor influence in terms of diastereomeric ratio and the asymmetric center and pre-organization of the chiral starting
formation of the by-product. Using trifluoromethanesulfonic material. This aim was achieved in routes 2a and 2b, as elabo-
acid (TfOH) instead of TFA (entry 3) increased the formation of rated in detail in the following.
(S,R)/(R,S) isomers, leading to a more balanced ratio of 47 : 53,
In route 2a, employing the established diaryl ether syn-
but did not decrease by-product formation. Following thesis protocol, phenolic intermediate 12 from route 1a was
Kayhan’s approach46 we conducted the reaction in the more coupled with 4-bromophenyl enol ether 13 to obtain 22 in 61%
polar solvent trifluoroethanol (entry 4), ending up with an yield (Scheme 6). Next we examined reaction conditions for the
improved regioselectivity of 80 : 20. A diastereomeric ratio of intramolecular N-acyl Pictet–Spengler condensation. The reac-
54 : 46 was obtained hereby for the desired regioisomer 21. tion was carried out at different temperatures (Table 3), expect-
Even though the regioselectivity was best in this entry, due to a ing that the stereochemical outcome might be controlled by
slow and incomplete conversion we decided to follow our the reaction temperature to a certain extent. We observed no
established protocol (entry 1) for the synthesis of seco-bisben- conversion below −20 °C using TFA as catalyst (entry 1). At
zylisoquinoline 21 (Scheme 5).
−20 °C, catalyzed by TFA, macrocycle 15 was formed in 13%
In the following step the macrocycle 15 was generated via yield under predominant formation of the (R,R)/(S,S) isomers
intramolecular C–O coupling of 21 (with previously separated 15a (diastereomeric ratio dr 84 : 16; entry 2), hence favouring
racemic diastereomers) using our protocol as described above,
yielding 51% of macrocyclic (R,R)/(S,S) isomers (15a) and 35%
of (S,R)/(R,S) isomers (15b). Both of these products can be
reduced to racemic tetrandrine (rac-1) and racemic isotetran-
Table 3 Conditions for cyclization via N-acyl Pictet–Spengler
condensation
drine (rac-2), respectively, in almost quantitative yields, as
shown for route 1a (Scheme 3).
Ratio (R,R)/(S,S) :
(S,R)/(R,S)
Entry
Temperature (°C)
Acid
Yield (%)
As expected and discussed above, both routes 1a and 1b
gave mixtures of diastereomers. Fortunately, chromatographic
separation of late intermediates was achieved in both routes,
so both racemic tetrandrine (rac-1) and racemic isotetrandrine
(rac-2) are available in pure form with these approaches.
Nevertheless, we intended to expand our approach to a proto-
1
2
3
4
5
6
−78 to −25
−20
−15
20
−78
40
TFA
TFA
TFA
TFA
TfOH
p-TsOH
—
—
13
28
25
9
84 : 16
87 : 13
82 : 18
78 : 22
62 : 38
>5
Scheme 5 Route 1b: Synthesis of seco-bisbenzylisoquinoline 21 and cyclization.
Scheme 6 Route 2a: Cyclization via intramolecular Pictet–Spengler condensation.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
the relative stereoconfiguration of tetrandrine (1) (S,S) and its (rac-2) (see Scheme 3). Separation of these diastereomers is
enantiomer phaeanthine (R,R). The best result was obtained at possible using preparative TLC47 or preparative HPLC, so pure
−15 °C (entry 3) in terms of diastereomeric ratio (87 : 13) and racemic alkaloids can also be obtained via routes 2a and 2b.
yield (28%). When performing the reaction at ambient temp-
In Inubushi’s synthetic approach,28 which included an
erature we obtained a comparable yield (25%) and a d.r. of intramolecular Bischler–Napieralski reaction, some stereocon-
82 : 18 (entry 4). With the stronger acid trifluoromethane- trol was also observed, but in this case the second stereocenter
sulfonic acid (TfOH) macrocycle 15 was generated even at was not built up in the course of the cyclization reaction, but
−78 °C, but in a very poor yield of 9% in a 78 : 22 ratio (entry in the following reduction of the formed 3,4-dihydroisoquino-
5). With toluenesulfonic acid (p-TsOH) no noteworthy reaction line moiety to the corresponding tetrahydroisoquinoline. The
was observed below 40 °C, and at elevated temperature only best dr obtained there was 4 : 1 (determined by NMR). In terms
inferior dr and poor yield were obtained (entry 6). Eventhough of yields our synthesis is superior to Inubushi’s approach. The
the cyclization step of route 2b suffers from low yield, the diaryl ethers in our synthesis were obtained in yields of
asymmetric induction in this step is remarkable and prompted 41–64%, whereas Inubushi yielded 42–48% (according to the
us to do further investigations.
information given in the Experimental section28).
Route 2a involved late stage construction of the tetrahydroi- Tetrahydroisoquinoline units were generated in yields of
soquinoline moiety (rings A′ + B′) forming the stereocenter at 47–96% in one single step via N-acyl Pictet–Spengler conden-
C-1′, but the same concept can also be applied to C-1 in ring sations here, whereas Inubushi’s Bischler–Napieralski protocol
B. So, in route 2b, we used the same building blocks as in with subsequent reduction of the dihydroisoquinolines furn-
route 1b, but performed the Ullmann coupling first, followed ished the corresponding tetrahydroisoquinolines in yields of
by an intramolecular Pictet–Spengler reaction. Diaryl ether 23 28–35% over three steps. Furthermore, the Bischler–
was obtained from phenolic intermediate 19 and brominated Napieralski method requires arylacetic acid precursors which
enol ether 20 employing the established diaryl ether synthesis are mostly laborious to access. The enol ethers required in our
protocol in a yield of 43%. Important for this step is the use of synthetic approach are easily obtained in a single step in excel-
freshly synthesized enol ether 20, since it tends to decompose lent yields from commercially available and cheap
quickly resulting in minor yields. In attempts to improve this benzaldehydes.
yield we once again examined Pd-catalysts (compare entry 2 in
In summary we have established a new total synthesis of
Table 1), but obtained only traces of the desired product 23. In racemic tetrandrine (rac-1) and racemic isotetrandrine (rac-2)
the subsequent intramolecular Pictet–Spengler reaction, cata- in a total of 12 steps in all routes (compared to more than 20
lyzed by TFA at −15 °C, macrocycle 15 (see Scheme 3) was steps in Inubushi’s total synthesis27). The overall yields of both
obtained in 79% yield, but in strong contrast to route 2a, here alkaloids taken together when combining the building blocks
the isotetrandrine-type (S,R)/(R,S) isomers 15b were formed 6, 7, 10 and 13 are 10.5% in route 1a and 3.0% in route 2a. In
predominantly in a ratio of 71 : 29 (determined by HPLC) route 1b the overall yield amounts 12.4% and in route 2b
(Scheme 7). In contrast to route 1b, where regioisomeric pro- 19.2% after combination of the building blocks 6, 10, 16 and
ducts were obtained in intermolecular Pictet–Spengler reac- 20. Compared to the overall yield of 2.1% in Inubushi’s
tions, we detected exclusively the desired cyclization products approach (as well starting from corresponding building
in this intramolecular reaction. Whereas route 2a does not blocks) our synthetic method is up to nine times more
provide an advantage over route 1a in terms of yield, the efficient. Whereas routes 1a and 1b provide almost equimolar,
access of isotetrandrine-type isomers in route 2b could be separable mixtures of both diastereomeric forms (rac-1 and
improved compared to route 2a.
rac-2), the routes involving intramolecular N-acyl Pictet–
Final step of both routes 2a and 2b was, as in routes 1a and Spengler condensations show good stereoselectivity for the
1b, the reduction of both ethyl carbamate moieties in 15 using one or other diastereomer. Since the stereoselectivity in our
LiAlH4 in THF to give N-methyl groups, ending up with mix- approach is limited, the synthesis of enantiomerically pure bis-
tures of racemic tetrandrine (rac-1) and racemic isotetrandrine benzylisoquinoline alkaloids based on our work remains part
Scheme 7 Route 2b: Alternative cyclization via intramolecular N-acyl Pictet–Spengler condensation.
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
of future projects. As numerous protocols have been published step using quantum mechanics-based methods, which reveal
for the synthesis of enantiomerically pure monomeric e.g. the energetic barriers and, thus, the reactivity of the
1-benzyltetrahydroisoquinolines,48–56 further elaboration of selected starting conformations. Subsequently combining
such building blocks using our protocols could provide enan- structural data of the first step with the corresponding ener-
tiomerically pure bisbenzylisoquinolines as well.
getics of the second step can provide a mechanistic as well as
To understand the surprisingly different diastereomeric the energetic explanation of the experimental results.
outcome of the routes 2a and 2b, we performed an in silico Nevertheless, as reaction pathway calculations are very
conformational analysis of the macrocyclic intermediate struc- demanding, they are normally performed only for the “final”
tures formed during the crucial intramolecular Pictet–Spengler reaction pathway, i.e. the one which was experimentally found
condensation of 22 (route 2a) and 23 (route 2b). The initial to perform best. Thus, the corresponding calculations are
step of these cyclizations is the reaction of the protonated enol either done to explain details of the reaction mechanism or to
ether building block with the arylethylamine-derived carba- investigate the basic underlying reaction mechanism if it is
mate, giving rise to a macrocyclic N-acyl iminium ion (and a unknown. However, they are not practically applicable (i.e. too
carbenium ion mesomeric structure) as the reactive intermedi- computationally demanding) for investigating differences in
ate, which then will build up the new stereocenter at C-1 (or the outcome of alternative reaction routes, as in this case the
C-1′) by electrophilic attack at C-8a (or C-8a′) of the aromatic reaction pathways for all investigated routes would have to be
ring. The putative cationic intermediates of routes 2a and 2b calculated. As the major goal of our computational studies was
are depicted in Fig. 3 (routes 1a and 1b, in which cyclization to provide a molecular explanation for exactly such differences,
occurs after the formation of both stereocenters, did not lead we asked ourselves if a stability analysis of the macrocyclic
to any (route 1a) or a satisfactory (route 1b) selectivity). intermediates based on conformational sampling (e.g. the first
Important for selectivity is the construction of the macrocycle step, see above) could already be helpful for such analysis. One
under formation of the second stereocenter in one reaction necessary prerequisite for this to be true would be that the
step.
results obtained from such simulations correlate well with the
To investigate all aspects of this complex reaction on a experimentally observed final diastereometric ratios. In this
quantitative level a two-step workflow would be necessary. context it needs to be stated that, as no reaction pathways are
First, the evaluation of the stability of different conformations calculated, a computational analysis of reaction-related quan-
of the macrocyclic intermediates by molecular mechanics- tities as e.g. the experimentally observed racemic nature of the
based sampling techniques to identify possible starting con- final reaction products would not be possible based on such
formations for the final reaction step. Second, the investi- studies.
gation of the reaction pathway of this following final reaction
Therefore, to investigate the above question, we practically
conducted extensive molecular dynamics simulations of both
isomers of the macrocyclic intermediates formed during the
reaction routes 2a and 2b and then compared the outcome of
our simulations to the diastereomeric ratios for tetrandrine
and isotetrandrine as obtained experimentally for those path-
ways. The applied computational setup was based on a proto-
col specifically established and afterwards applied successfully
for simulating macrocyclic compounds in the context of mole-
cular docking,57–59 where a similar analysis of conformational
ensembles of macrocyclic-ring conformations from molecular
dynamics simulations in solution was found to represent the
conformational space very well providing that the simulation
time and conditions are chosen properly (for a detailed
description of the sampling conditions see Computational
details). For this the Amber1660 software was used with
General Amber Force Field61 parameters for all intermediate
structures with point charges derived according to the general
guidelines for RESP charges (see Computational details for
more technical information). Both intermediate structures of
route 2a (2b) were simulated as carbenium ions containing a
single bond between N-2′ (N-2) and C-1′ (C-1) (highlighted
bonds in Fig. 3). This mesomeric form was chosen as it allows
for free rotation around that bond, which we rationalized to be
the most realistic representation for our purpose as the poten-
tial configuration around the second stereocenter to be built
up at C-1′ (C-1) is not biased by structural inflexibility in that
Fig. 3 Structures of the macrocyclic intermediates of the reactions 22
→ 15 (route 2a) and 23 → 15 (route 2b).
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 4 Summary table listing observed experimental and estimated
computational ratios deduced from and colour coded according to
values in Table S1‡ based on geometrical analysis of the first 150 (all)
conformations of the pre-reaction ensembles
region. This would have been the case for the putative
iminium mesomeric form, i.e. the explicit simulation of E- and
Z-isomers. Thus the structural ensembles obtained from the
long-term simulations of the four intermediate structures
(Fig. 3) contain pre-reaction conformers of all possible inter-
mediates formed previous to the cyclization step of routes 2a
and 2b. Afterwards, these ensembles were geometrically ana-
lyzed to, first, evaluate if there exists indeed a correlation
between these conformational ensembles and the experi-
mental diastereometric ratios, and second, provide a system-
specific structural explanation of the observed selectivities.
Therefore, we focused on two points during the analysis: results in a computational diastereomeric ratio of 74% : 26%
first, to evaluate how high the probability for each intermedi- tetrandrine : isotetrandrine ((S,S)/(R,R) : (S,R)/(R,S)) as listed in
ate is to form a pre-reactive structure which, from a geometri- Table 4. The same analysis performed for the intermediates of
cal point of view, allows the reaction to occur, and second, to route 2b led to 44 conformations of the C-1′R intermediate and
analyze which configuration is the most probable to be formed 30 of the C-1′S intermediate conformations showing pre-S
at the second stereocenter, depending on the conformation of orientation, whereas 28 of C-1′R and 48 of C-1′S intermediates
the sampled intermediates. Therefore, we concentrated on a can be assigned to pre-R for the second stereocenter.
detailed structural analysis of the obtained pre-reactive confor- Therefore, a computational ratio of 39% : 61% tetrandrine :
mational ensembles, focusing on the 1–2 (1′–2′) bond (high- isotetrandrine ((S,S)/(R,R) : (S,R)/(R,S)) could be obtained com-
lighted in red in Fig. 3) and the adjacent aromatic ring A (A′), putationally (Table 4).
as their spatial orientation was found to be important for the
In contrast to route 2b, where comparable proportions of
potential configuration of the second stereocenter, as dis- conformations are sampled for the pre-organized enantio-
cussed hereinafter. This analysis reveals the probability of the mers, the C-1S intermediate of route 2a shows surprisingly
macrocyclic scaffold being oriented towards
a specific high preference of adopting a conformation favouring pre-S
outcome and can therefore be linked to the likelihood of orientation at the second stereocenter – which is absent for
product formation. Accordingly, the diastereomeric ratio could the enantiomeric counterpart (C-1R oriented towards C-1′
be assessed computationally by combining all conformers pre-R). Considering the enantiomeric nature of the intermedi-
with similar sets of potential configurations previous to the ates, these results are unfortunately not as it is to be
crucial cyclization step.
expected. Although a reliable prediction of the stereochemical
For each reaction, both isomeric forms of the intermediates outcome for solely one enantiomeric intermediate in route 2a
(Fig. 3) were investigated. For this, molecular dynamics simu- with our computational analysis is therefore limited, the
lations were performed for all four intermediates under experi- average ratio of the combined data of both C-1S and C-1R
mental conditions (Table 3, entry 3) and the conformational intermediates reproduces the experimentally observed ratio
space of the structures was analyzed. The first 150 confor- very well (87% : 13% tetrandrine : isotetrandrine (S,S)/(R,
mations featuring a short distance between atoms C-1′ and R) : (S,R)/(R,S)). However, to minimize and investigate any
C-8a′ (C-1 and C-8a for route 2b) were extracted as these struc- potential impact of insufficient sampling, four additional
tures represent ensembles of possible starting structures for replicas were simulated for both intermediates of route 2a,
the cyclization step of the N-acyl Pictet–Spengler condensation. such that finally a total of 547 pre-reaction conformations
These sets of conformations were further analyzed (see compu- could be extracted from all combined trajectories and the
tational details at the end of this article) with respect to their above analysis was repeated for the larger pre-reaction ensem-
spatial orientation of C-1 (or C-1′) in relation to the position of bles (see computational details and values in brackets in
ring A (or ring A′), which initiates the nucleophilic attack Tables 4 and S1‡). The assessed computational ratio for all
under construction of the second stereocenter. Respective pre-reaction conformers of route 2a of 71% : 29% tetrandrine :
structures will be called “pre-reaction” ensembles or confor- isotetrandrine is very close to the previous one considering
mations in the further discussion.
only the first 150 conformations (74% : 26%) and shows
Table S1‡ lists the summarized results of the overall confor- similar trends for both enantiomers (Table S1‡).
mational analysis of the first 150 structures of the pre-reaction Summarizing, the overall final simulation times for route 2a
ensembles. From the simulations of intermediates of route 2a and 2b were 6000 and 4000 ns, respectively. These results
34 conformers of C-1R intermediates and 87 of the C-1S inter- show that sufficient sampling was guaranteed and it can be
mediates are likely to form S-configuration at the second ruled out that the observed preference of the C1-S intermedi-
stereocenter being built (referred as “pre-S”), whereas 24 of the ate is due to poor statistics or sampling, which is substan-
C-1R intermediates and 5 of the C-1S intermediates can be tiated by the fact that the potential energy of the C1-S inter-
expected to form a R-configured second stereocenter (referred mediates that form pre-S configuration on the second stereo-
as “pre-R”). Combining this data of both intermediates with center is on average 16.5 kJ mol−1 lower than the one of the
respect to the total amount of 150 pre-reaction conformations C1-R intermediate structures obtaining pre-R orientation.
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
The above results show that there exists a very good, quali- Fig. 4, A′- (or A-), highlighted in green) if a pre-S orientation is
tative agreement between the average distributions of the pre- observed at the second arising stereocenter and flipped back
reaction ensembles and the measured diastereomeric ratios of (A′+ or A+) if it is pre-R oriented, which for route 2a leads to
routes 2a (87 : 13) and 2b (29 : 71), therefore indicating that the the (S,S)/(R,R) isomers, i.e. to tetrandrine (Fig. 4, top), and to
conformational distribution of the cyclic intermediates of this the (S,R)/(R,S) isomers for route 2b, i.e. to isotetrandrine. In
reaction step is correlated to the stereoselectivity of the syn- fact, our qualitative analysis shows all pre-reaction confor-
thetic pathways 2a and 2b.
mations originating from both intermediates (C-1S and C-1R)
Given the good correlation between computational and of route 2a that obtain pre-R orientation at the second stereo-
experimental results, we performed a detailed structural ana- center were found in combination with ring A′ flipped to the
lysis based on geometries of the most prominent structures of back (A′+), suggesting a strong correlation between the posi-
the conformational ensembles as depicted in Fig. 4, which tion of the aromatic ring A′ and the dihedral angle of the high-
were chosen on the basis of the dihedral angle values of the lighted bond (data not shown). Together with the analysis
highlighted torsion (data not shown, red in Fig. 3, route 2a: above, the experimentally observed selectivity of route 2a can
C-1′ and N-2′; route 2b: C-1 and N-2) throughout the simu- be explained by the formation of highly stable pre-reaction
lation. For the pre-reaction ensembles of route 2a, a clear pre- conformations of the macrocycle. Such a distinct preference is
ference towards tetrandrine (rac-1) was observed (compu- not found for the pre-reaction conformations of route 2b,
tational product ratio of 74% : 26% (S,S)/(R,R) : (S,R)/(R,S)), which agrees with the experimentally observed lower
whereas the preference of specific intermediates of route 2b is selectivity.
less pronounced and in contrast to the former with inverted
Therefore, according to our in-depth computational ana-
stereochemical preference for the isotetrandrine type (rac-2) lysis, there exists a clear correlation between the conformation-
(Tables 4 and S1,‡ computational product ratio of 39% : 61% al ensembles of the cationic macrocyclic intermediates of the
(S,S)/(R,R) : (S,R)/(R,S)). Regarding the most dominant confor- cyclization step of routes 2a and 2b in solution and the experi-
mers of the ensembles, they are characterized by the methoxy mentally determined diastereomeric ratios for both reactions.
group of ring A′ (or the methoxy groups of ring A, respectively) Moreover, even the experimentally observed more pronounced
flipped to the front (with respect to the orientation chosen in diastereomeric ratio of macrocycle 15 via route 2a (intermedi-
Fig. 4 Most dominant conformations among the pre-reaction ensembles of macrocyclic cationic intermediate structures of the cyclization step of
route 2a (top) and route 2b (bottom), respectively. The structures are characterized by the torsional angle around the highlighted bond (orange)
determining the configuration at the newly formed, second stereocenter and the relative position of the aromatic ring (green) and labeled according
to the resulting product. The dotted line indicates the final bond formed in this reaction. Atom color scheme: carbon (light blue), nitrogen (dark
blue), oxygen (red) and hydrogen (white).
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
ates of 22 → 15) compared to route 2b (intermediates of 23 → presented here in hands, future syntheses of tetrandrine ana-
15) can be correlated to and thus explained by the relative logues for medicinal chemistry projects can be designed with
occurrence of the different pre-reaction conformations very high flexibility regarding substitution patterns of all four
sampled. The same holds for the preferred formation of tetran- aromatic rings. In our protocols, any of the four aromatic rings
drine via route 2a and isotetrandrine via route 2b. Our detailed in tetrandrine (1)/isotetrandrine (2) is derived from a readily
conformational analysis reveals the structural basis behind available substituted benzaldehyde. Depending on which aro-
these differences: the sampled pre-reaction ensemble of route matic ring is subject of intended systematic variations, the
2a is dominated by a very stable conformation featuring an corresponding building block can be inserted in a late stage of
optimal geometry for tetrandrine formation. This indicates the synthesis following the best fitting of one of our routes. If
that in this case the stereochemistry at the first stereocenter at stereocontrol is attempted or if all stereoisomers are welcome,
C-1 initiates a structural pre-organization of the macrocyclic the suitable route can be chosen accordingly. So, this work
ring leading to an optimal orientation for tetrandrine. For should help to extend the chemical diversity of libraries of ana-
route 2b a similarly stable structural pre-arrangement of the logues bisbenzylisoquinoline alkaloids significantly.
macrocyclic ring system cannot be seen and a broader range of
conformations is observed in the pre-reaction ensemble, a reasonable explanation for the observed stereocontrol gov-
which might explain the lower selectivity of that reaction path. erning the crucial cyclization step of route 2a. Furthermore,
Moreover, with computational approaches we could deliver
Although in the current study the computational work was for intramolecular reactions leading to macrocyclic molecules,
performed after the experiments and only for one specific case the combination of theory and experiment as applied herein
study the above results still indicate that – next to explaining demonstrates a promising new strategy for optimizing the
the experimental results a posteriori – such combined compu- experimental workflow for the investigation of different poss-
tational/experimental approaches could also be helpful to ible synthesis pathways. In future, the addition of compu-
computationally estimate which synthetic route might be most tational analysis might even enhance the experimental
promising with respect to the desired stereochemical outcome, outcome of similar modular synthesis setups, for which the
especially if used in the context of a modular synthesis underlying reaction mechanisms and crucial intermediates are
approach based on established reactions for which the mecha- known, e.g. by estimating diastereomeric ratios in advance.
nisms and intermediate structures are known in advance. Of Thus, the introduced method demonstrates a robust guideline
course, in such future studies, it would be necessary to re- towards targeted synthesis.
evaluate on a case-to-case basis which calculated quantities
correlate with the desired stereochemical outcome of the syn-
thesis, especially if different reaction types are investigated
and in the long run corresponding large-scale benchmark
studies need to be performed to establish a generally appli-
Experimental section
General methods
cable workflow. In addition, in this context it needs to be All solvents and reagents were purchased from commercial
stated that considering the low experimental selectivity and suppliers and were used without further purification, unless
the accuracy limits of the used molecular mechanics-based mentioned otherwise. Anhydrous THF was distilled from
computational methods in the current study, the qualitative sodium/benzophenone under nitrogen atmosphere. TLC was
reproduction of the diastereomeric trends for both reaction carried out on 0.2 mm silica gel polyester plates with a fluo-
routes is already a very good achievement and any quantitative rescence indicator (POLYGRAM SIL G/UV254). For preparative
comparison would go beyond the capability and accuracy of TLC 1 mm PLC Silica gel 60 F254 plates (20 × 20 cm) with con-
the applied methods. Nevertheless, even on such a qualitative centrating zone (Merck) were used. Flash chromatography was
level, similar computational workflows could be helpful in performed on 40–63 μm silica gel using the solvent systems
future studies by e.g. providing a computational pre-ranking of indicated. NMR spectra were recorded with a 400 MHz
different possible reactions. This might allow for an in silico- (400 MHz for 1H and 101 MHz for 13C), 500 MHz (500 MHz for
guided and thus more efficient experimental investigation of 1H and 126 MHz for 13C), or 600 MHz Bruker Biospin Avance
possible reaction pathways.
spectrometer (599 MHz for 1H and 151 MHz for 13C). Peak
assignments were based on 2D NMR experiments using stan-
dard pulse programs (COSY, HSQC/HMQC, DEPT, HMBC).
Chemical shifts were referenced to the residual solvent signal
Conclusion
(CDCl3: δH = 7.26 ppm, δC = 77.16 ppm; CD3OD: δH
=
By the application of Ullman-type diaryl ether synthesis and 3.31 ppm, δC = 49.0 ppm). For the characterization of rotamers
N-acyl-Pictet–Spengler reactions for the construction of 1-ben- a temperature program was employed for recording both 1D
zyltetrahydroisoquinoline units, especially when employed as and 2D spectra. Hereby chemical shifts were referenced to the
central steps in generating the second stereocenter, we could signal of tetramethylsilane in deuterated tetrachloroethane
work out a new approach to bisbenzylisoquinoline alkaloids of (Tcl2 [100 °C]: δH = 5.92 ppm, δC = 74.0 ppm). IR spectra were
the tetrandrine (1) type, with the perspective for controlling recorded using
a Jasco FT/IR-4100 (type A) instrument
the stereochemical outcome. Having the four different routes equipped with a diamond ATR unit (Jasco PRO450-S). High
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
resolution mass spectra (HR-MS) were recorded using a Jeol was added and after 5 min of pre-stirring the reaction was
Mstation 700 or JMS GCmate II Jeol instrument for electron heated to 110 °C for 2–7 days under nitrogen atmosphere. The
impact ionisation (EI). Thermo Finnigan LTQ was used for reaction mixture was concentrated in vacuo, diluted with ethyl
electrospray ionisation (ESI). HPLC purity was determined acetate and filtered over a small plug of silica gel in order to
using InfinityLab Poroshell 120EC-C18 and 120CN-C18 remove the catalyst and the excess base, followed by washing
columns (2.7 µm, 100 × 3.0 mm), detecting at 210 nm and with ethyl acetate. The filtrate was concentrated in vacuo to
254 nm. For quantitative analysis of compound 15 and 21 afford a brown oil as crude product which was purified by
HPLC was performed on a Agilent InfinityLab Poroshell 120 column chromatography.
EC-C18 column (2,7 µm, 100 × 3.0 mm) using following
method: eluent 50% ACN, 49.9% water and 0.1% THF; flow
(E)-3-Bromo-4,5-dimethoxy-1-(2-nitrovinyl)benzene (4)
0.8 mL min−1; temperature 50 °C; Agilent 1100/1200 Diode Synthesized from 3-bromo-4,5-dimethoxybenzaldehyde (10.5 g,
Array Detector (λ = 210 nm). Preparative HPLC was performed 42.8 mmol) according to the procedure of Maresh et al.33 The
on a Macherey-Nagel Nucleodur® 100–5 column (5 µm, 250 × crude product was purified by flash column chromatography
10 mm) using a VWR LaPrep P110 system with a UV Detector (50% dichloromethane in hexanes, Rf = 0.25) and the title com-
P311.
pound obtained as bright yellow crystalline solid (7.37 g,
25.7 mmol, 60%). Purity (HPLC) = 92% (λ = 210 nm). Mp:
153.5 °C. 1H NMR, COSY (500 MHz, CDCl3): δ [ppm] = 7.88 (d,
General procedure 1: Wittig olefination
A
suspension of (methoxymethyl)triphenylphosphonium J = 13.6 Hz, 1H, H-1′), 7.51 (d, J = 13.6 Hz, 1H, H-2′), 7.37 (d, J =
chloride (1.2 equiv.) in anhydrous THF (2 mL per mmol aro- 2.1 Hz, 1H, H-2), 6.98 (d, J = 2.0 Hz, 1H, H-6), 3.92 (s, 3H,
matic aldehyde) was cooled to 0 °C under nitrogen atmo- 4-OCH3), 3.92 (s, 3H, 5-OCH3). 13C NMR, DEPT, HMQC, HMBC
sphere. A solution of lithium diisopropylamide (1.4 equiv., 2.0 (126 MHz, CDCl3): δ [ppm] = 154.2 (C-5), 149.9 (C-4), 137.7
M solution in THF) was added dropwise and the resulting (C-1′), 137.3 (C-2′), 127.0 (C-1), 126.4 (C-2), 118.7 (C-3), 111.6
mixture stirred for 45 min. A solution of the aromatic aldehyde (C-6), 61.0 (4-OCH3), 56.4 (5-OCH3). IR (ATR): v [cm−1] = 2919,
˜
(1.0 equiv.) in anhydrous THF (2 mL per mmol) was added 1628, 1414, 1352, 1279, 1151, 1044, 962, 831. HRMS (EI): m/z
with stirring. The mixture was allowed to warm up to ambient calcd for [C10H10BrNO4]•+ 286.9788 and 288.9767, found:
temperature and stirred for 4 h. The reaction was then 286.9801 and 288.9810.
quenched with deionized water and extracted 3× with ethyl
acetate. The combined organic phases were washed with brine,
3-Bromo-N-ethoxycarbonyl-4,5-dimethoxyphenethylamine (6)
dried over anhydrous Na2SO4 and concentrated in vacuo to Nitrostyrene 4 (7.37 g, 25.7 mmol) was reduced to 3-bromo-4,5-
afford the crude product which was purified by column dimethoxyphenethylamine (5) following
chromatography.
a procedure of
Maresh et al.33 The crude product (an orange oil) was dis-
solved in dichloromethane (100 mL) under nitrogen atmo-
sphere and the solution cooled to 0 °C. Triethylamine
General procedure 2: N-acyl Pictet–Spengler condensation
A solution of carbamate (1.0 equiv.) and enol ether (1.2–1.5 (14.1 mL, 101 mmol) was added, followed by ethyl chlorofor-
equiv.) in dichloromethane (for intermolecular reactions mate (6.06 mL, 63.4 mmol) in a dropwise fashion. The mixture
10 mL per mmol carbamate) was cooled to 0 °C under nitrogen was allowed to warm to ambient temperature and stirred for
atmosphere. Trifluoroacetic acid (10 equiv.) was added drop- 14 h, then quenched with saturated NaHCO3 solution (50 mL)
wise. The reaction was allowed to warm to ambient tempera- and extracted with dichloromethane (3 × 80 mL). The com-
ture and stirred for 6–12 h. The intramolecular reactions were bined organic phases were dried over anhydrous Na2SO4 and
conducted at an educt concentration of 0.01 mM and cooled concentrated in vacuo to afford a yellow oil. Purification was
to −15 °C. After adding TFA (10 equiv.) the temperature was accomplished by flash column chromatography (dichloro-
kept at −15 °C until the reaction has completed after 12 h. A methane, Rf = 0.12) to give 6 as pale yellow oil (5.95 g,
saturated NaHCO3 solution was then added for neutralization 18.0 mmol, 70%). Purity (HPLC) = 100% (λ = 210 nm). 1H
and the aqueous phase extracted 3× with dichloromethane. NMR, COSY (400 MHz, CDCl3) δ [ppm] = 6.96 (d, J = 1.9 Hz,
The combined organic phases were dried over anhydrous 1H, H-2), 6.69–6.66 (m, 1H, H-6), 4.68 (s, 1H, NH), 4.11 (q, J =
Na2SO4 and concentrated in vacuo to afford the crude product 7.1 Hz, 2H, OCH2–CH3), 3.85 (s, 3H, 5-OCH3), 3.83 (s, 3H,
which was purified by column chromatography.
4-OCH3), 3.40 (q, J = 6.8 Hz, 2H, NCH2–CH2), 2.73 (t, J = 7.0 Hz,
2H, NCH2–CH2), 1.23 (t, J = 7.1 Hz, 3H, OCH2–CH3). 13C NMR,
DEPT, HMQC, HMBC (101 MHz, CDCl3) δ [ppm] = 156.7
General procedure 3: Ullmann-type C–O coupling reaction
According to a modified procedure of Wang et al.,38 bromoar- (CvO), 153.8 (C-5), 145.2 (C-4), 136.2 (C-1), 124.8 (C-2), 117.8
ene (1.0–1.2 equiv.), phenol (1.0–2.0 equiv.), CuBr·Me2S (1 (C-3), 112.4 (C-6), 61.0 (OCH2–CH3), 60.7 (4-OCH3), 56.2
equiv.) and Cs2CO3 (3 equiv.) were placed in a pressure tube or (5-OCH3), 42.1 (NCH2–CH2), 35.9 (NCH2–CH2), 14.8 (OCH2–
a flask closed with a screwcap with septum inlet and sealed CH3). IR (ATR): v [cm−1] = 3354, 2937, 2826, 1697, 1488, 1271,
˜
with PTFE tape. Anhydrous pyridine (for intermolecular C–O 1235, 1140, 1044, 1001, 817. HRMS (ESI): m/z calcd for
coupling reactions 10 mL per mmol bromoarene, for intra- [C13H18BrNO4 + H]+ 332.0497 and 334.0477, found: 332.0493
molecular reaction a concentration of 0.02 mM was chosen) and 334.0474.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
(E/Z)-3-(Benzyloxy)-4-methoxy-1-(2-methoxyvinyl)benzene (7)
N-Ethoxycarbonyl-4-hydroxy-3-methoxyphenethylamine (10)
3-Benzyloxy-4-methoxybenzaldehyde (6.01 g, 24.8 mmol) 3-Methoxytyramine hydrochloride (3.46 g, 17.2 mmol) was sus-
was reacted following General procedure 1 (Wittig olefina- pended in dichloromethane (100 mL) under nitrogen atmo-
tion). The reaction was completed after 4 h. Purification sphere and the mixture cooled to
0 °C. Triethylamine
was accomplished by flash column chromatography (10% (9.58 mL, 68.7 mmol) was added, followed by ethyl chlorofor-
ethyl acetate in hexanes, Rf = 0.30) and the product mate (4.11 mL, 43.0 mmol) in a dropwise fashion. The reaction
obtained as white solid (yield 5.89 g, 21.8 mmol, 88%). The was allowed to warm to ambient temperature and stirred for
major product is the E-isomer (E,Z-isomer ratio 1 : 0.82, esti- 15 h. After this time TLC analysis showed complete conversion
mated by NMR-integrals). Purity (HPLC) = 100% (λ = to a less polar product (Rf = 0.26, dichloromethane), which is
210 nm). Mp: 95.5–96.0 °C. NMR data of the major the ethoxycarbonylated phenol intermediate. For chemo-
E-isomer: 1H NMR, COSY (500 MHz, CDCl3) δ [ppm] = selective cleavage of the phenol ester, an ethanolic sodium
7.50–7.43 (m, 2H, H-6-Ph, H-2-Ph), 7.41–7.27 (m, 3H, H-5- hydroxide solution (1 M, 100 mL) was added to the reaction
Ph, H-4-Ph, H-3-Ph), 6.86 (d, J = 13.1 Hz, 1H, H-2′), and the mixture stirred for 2 h at ambient temperature. After
6.84–6.76 (m, 3H, H-2, H-5, H-6), 5.72 (d, J = 12.9 Hz, 1H, addition of hydrochloric acid (2 M, 60 mL) the organic solvents
H-1′), 5.16 (s, 2H, OCH2-Ph), 3.87 (s, 3H, 4-OCH3), 3.65 (s, were removed in vacuo and the remaining aqueous solution
3H, 2′-OCH3). 13C NMR, DEPT, HMQC, HMBC (126 MHz, was extracted with dichloromethane (3 × 100 mL). The com-
CDCl3) δ [ppm] = 148.2 (C-3), 147.9 (C-4), 147.9 (C-2′), 137.4 bined organic phases were dried over anhydrous Na2SO4 and
(C-1-Ph), 129.4 (C-1), 128.7 (C-5-Ph, C-3-Ph), 128.0 (C-4-Ph), concentrated in vacuo to afford a beige oil. Purification by
127.5 (C-6-Ph, C-2-Ph), 118.4 (C-6), 112.4 (C-5), 111.4 (C-2), flash column chromatography (30% ethyl acetate in hexanes,
104.9 (C-1′), 71.3 (OCH2–Ph), 56.6 (2′-OCH3), 56.3 (4-OCH3). Rf = 0.20) gave the title compound as a white solid (3.70 g,
IR (ATR): v [cm−1] = 3060, 2931, 1638, 1514, 1254, 1134, 15.5 mmol, 90%). Purity (HPLC) = 100% (λ = 210 nm). Mp:
˜
1010, 746, 698. HRMS (EI): m/z calcd for [C17H18O3]•+ 95.0–95.5 °C. 1H NMR, COSY (400 MHz, CDCl3) δ [ppm] = 6.84
270.1251, found: 270.1251.
(d, J = 8.2 Hz, 1H, H-5), 6.70–6.65 (m, 2H, H-6, H-2), 5.62 (s,
1H, OH), 4.69 (s, 1H, NH), 4.10 (q, J = 7.1 Hz, 2H, OCH2–CH3),
3.86 (s, 3H, OCH3), 3.39 (q, J = 6.7 Hz, 2H, NCH2–CH2), 2.73 (t,
J = 7.0 Hz, 2H, NCH2–CH2), 1.22 (t, J = 7.1 Hz, 3H, OCH2–CH3).
13C NMR, DEPT, HMQC, HMBC (101 MHz, CDCl3) δ [ppm] =
156.8 (C-3′), 146.7 (C-3), 144.4 (C-4), 130.7 (C-1), 121.5 (C-2),
114.6 (C-5), 111.4 (C-6), 60.9 (OCH2–CH3), 56.0 (OCH3), 42.4
( )-8-Bromo-N-ethoxycarbonyl-6,7-dimethoxy-1-(3′-benzyloxy-4′-
methoxybenzyl)-1,2,3,4-tetrahydroisoquinoline (8)
Carbamate 6 (5.89 g, 17.8 mmol) and enol ether 7 (5.76 g,
21.3 mmol) were condensed following General procedure 2
(for intermolecular N-acyl Pictet–Spengler reaction) to give
tetrahydroisoquinoline 8. The reaction was completed after
12 h. The crude product was purified by flash column chrom-
atography (20% acetone in hexanes, Rf = 0.17) and the title
compound obtained as a pale yellow solid (4.66 g, 8.19 mmol,
46%). Purity (HPLC) = 98% (λ = 210 nm). Mp: 40.5–42.5 °C.
1H NMR, COSY (400 MHz, Tcl2, 100 °C) δ [ppm] = 7.46–7.41
(m, 2H, H-6-Ph, H-2-Ph), 7.38–7.32 (m, 2H, H-5-Ph, H-3-Ph),
7.31–7.26 (m, 1H, H-4-Ph), 6.79 (d, J = 8.1 Hz, 1H, H-5′), 6.74
˜
(NCH2–CH2), 35.9 (NCH2–CH2), 14.8 (OCH2–CH3). IR (ATR): v
[cm−1] = 3368, 3273, 1687, 1518, 1237, 1125, 1033, 824, 781.
HRMS (ESI): m/z calcd for [C12H16NO4]− 238.1085, found:
238.1087.
( )-8-(4-(2-((Ethoxycarbonyl)amino)ethyl)-2-methoxyphenoxy)-
N-ethoxycarbonyl-6,7-dimethoxy-1-(3′-benzyloxy-4′-
methoxybenzyl)-1,2,3,4-tetrahydroisoquinoline (11)
(d, J = 2.0 Hz, 1H, H-2′), 6.70 (dd, J = 8.1, 2.0 Hz, 1H, H-6′), Tetrahydroisoquinoline 8 (3.00 g, 5.26 mmol) and phenol 10
6.60 (s, 1H, H-5), 5.58–5.48 (m, 1H, H-1), 5.02 (s, 2H, OCH2- (2.52 g, 10.5 mmol) were coupled following General Procedure
Ph), 4.03–3.84 (m, 3H, H-3, OCH2–CH3), 3.83 (s, 3H, 7-OCH3), 3 (intermolecular Ullmann-type C–O coupling reaction). The
3.82 (s, 3H, 6-OCH3), 3.81 (s, 3H, 4′-OCH3), 3.43 (ddd, J = 13.2, reaction was completed after 7 days. Purification was accom-
8.8, 5.5 Hz, 1H, H-3), 3.15 (dd, J = 14.3, 3.9 Hz, 1H, H-α), plished by flash column chromatography (10% ethyl acetate in
2.84–2.70 (m, 2H, H-4, H-α), 2.53 (d, J = 13.9 Hz, 1H, H-4), dichloromethane, Rf = 0.26) and the product obtained as a
1.07 (br s, 3H, OCH2–CH3). 13C NMR, DEPT, HMQC, HMBC beige solid (1.57 g, 2.16 mmol, 41%). Purity (HPLC) = 97% (λ =
(101 MHz, Tcl2, 100 °C) δ [ppm] = 155.5 (CvO), 152.3 (C-6), 210 nm). Mp: 58.5–60 °C. 1H NMR, COSY (400 MHz, Tcl2,
149.3 (C-4′), 148.8 (C-3′), 145.7 (C-7), 137.9 (C-1-Ph), 131.7 100 °C) δ [ppm] = 7.41–7.37 (m, 2H, H-6-Ph, H-2-Ph), 7.36–7.30
(C-4a), 131.6 (C-1′), 129.9 (C-8a), 128.4 (C-3-Ph, C-5-Ph), 127.7 (m, 2H, H-5-Ph, H-3-Ph), 7.30–7.25 (m, 1H, H-4-Ph), 6.79 (d, J =
(C-4-Ph), 127.5 (C-2-Ph, C-6-Ph), 122.9 (C-6′), 118.3 (C-8), 2.0 Hz, 1H, H-2″), 6.73 (d, J = 8.1 Hz, 1H, H-5′), 6.66 (d, J = 2.0
117.5 (C-2′), 113.7 (C-5′), 112.9 (C-5), 72.0 (OCH2–Ph), 61.2 Hz, 1H, H-2′), 6.62–6.59 (m, 1H, H-6′), 6.59–6.57 (m, 1H, H-6″),
(OCH2–CH3), 60.5 (7-OCH3), 56.8 (4′-OCH3 or 6-OCH3), 56.5 6.55 (d, J = 8.2 Hz, 1H, H-5″), 6.50 (s, 1H, H-5), 5.42–5.32 (m,
(4′-OCH3 or 6-OCH3), 56.1 (C-1), 39.1 (C-α), 38.0 (C-3), 28.1 1H, H-1), 4.92 (s, 2H, OCH2–Ph), 4.56 (s, 1H, NH), 4.07 (q, J =
(C-4), 14.5 (OCH2–CH3). IR (ATR): v [cm−1] = 2934, 2836, 1691, 7.1 Hz, 2H, ‘OCH2–CH3), 4.03–3.92 (m, 1H, H-3), 3.89 (s, 3H,
˜
1596, 1513, 1425, 1318, 1246, 1137, 1101, 1024, 741, 697. 3″-OCH3), 3.81 (s, 3H, 6-OCH3), 3.91–3.69 (m, 2H, OCH2–CH3),
HRMS (ESI): m/z calcd for [C29H32BrNO6 + H]+ 570.1486 and 3.78 (s, 3H, 4′-OCH3), 3.60 (s, 3H, 7-OCH3), 3.36 (q, J = 6.9 Hz,
572.1465, found: 570.1489 and 572.1471.
2H, NCH2–CH2), 3.33–3.25 (m, 1H, H-3), 3.13 (dd, J = 14.0, 3.8
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Hz, 1H, H-α), 2.80 (dd, J = 14.1, 9.0 Hz, 2H, H-4, H-α), 2.72 (t, J (E/Z)-4-Bromo-1-(2-methoxyvinyl)benzene (13)
= 7.1 Hz, 2H, NCH2–CH2), 2.50 (d, J = 16.4 Hz, 1H, H-4), 1.20
Prepared from 4-bromobenzaldehyde (2.00 g, 10.8 mmol) fol-
(t, J = 7.1 Hz, 3H, ‘OCH2–CH3), 0.93 (br s, 3H, OCH2–CH3). 13C
NMR, DEPT, HMQC, HMBC (101 MHz, Tcl2, 100 °C) δ [ppm] =
156.5 (‘CvO), 155.4 (CvO), 152.6 (C-6), 149.5 (C-3″), 149.0
(C-4′), 148.7 (C-3′), 146.9 (C-4″), 145.4 (C-8), 140.6 (C-7), 137.9
(C-1-Ph), 133.3 (C-1″), 132.3 (C-1′), 130.3 (C-4a), 128.4 (C-5-Ph,
C-3-Ph, 127.7 (C-4-Ph), 127.6 (C-6-Ph, C-2-Ph), 124.2 (C-8a),
122.9 (C-6′), 121.0 (C-6″), 117.4 (C-2′), 115.6 (C-5″), 114.7 (C-2″),
113.6 (C-5′), 109.6 (C-5), 71.8 (OCH2–Ph), 60.9 (OCH2–CH3),
60.7 (‘OCH2–CH3), 60.7 (7-OCH3), 56.9 (4′-OCH3 or 3″-OCH3),
56.8 (4′-OCH3 or 3″-OCH3), 56.4 (6-OCH3), 52.1 (C-1), 42.3
(NCH2–CH2), 39.9 (C-α), 37.7 (C-3), 35.9 (NCH2–CH2), 28.1
lowing General procedure 1 (Wittig olefination). Purification
was accomplished by flash column chromatography (2.5%
ethyl acetate in hexanes, Rf = 0.23/0.30) to give the title com-
pound as a colourless oil (2.06 g, 9.72 mmol, 90%, E,Z-isomer
ratio 1 : 0.85, estimated by NMR integrals). Purity (HPLC) =
1
76% (λ = 210 nm). H NMR, COSY (400 MHz, CDCl3) E-isomer:
δ [ppm] = 7.40–7.35 (m, 2H, H-3, H-5), 7.11–7.07 (m, 2H, H-2,
H-6), 7.03 (d, J = 13.0 Hz, 1H, H-2′), 5.74 (d, J = 13.0 Hz, 1H,
H-1′), 3.68 (s, 3H, OCH3). 13C NMR, DEPT, HMQC, HMBC
(101 MHz, CDCl3) E-isomer: δ [ppm] = 149.5 (C-2′), 135.5 (C-1),
131.3 (C-3, C-5), 126.8 (C-2, C-6), 119.1 (C-4), 104.2 (C-1′), 56.7
(C-4), 14.6 (‘OCH2–CH3), 14.4 (OCH2–CH3). IR (ATR): v [cm−1] =
˜
(OCH3). IR (ATR): v [cm−1] = 2929, 2832, 1589, 1488, 1405,
˜
2936, 2832, 1690, 1606, 1509, 1424, 1330, 1255, 1212, 1154,
1131, 1108, 1069, 1023, 758. HRMS (ESI): m/z calcd for
[C41H48N2O10 + H]+ 729.3382, found: 729.3385.
1120, 1069, 1010, 816. HRMS (EI): m/z calcd for [C9H9BrO]•+
211.9831 and 213.9811, found: 211.9827 and 213.9806.
( )-8-(4-(2-((Ethoxycarbonyl)amino)ethyl)-2-methoxyphenoxy)-N- ( )-8-((1-(4-Bromobenzyl)-N-(ethoxycarbonyl)-6-methoxy-1,2,3,4-
ethoxycarbonyl-6,7-dimethoxy-1-(3′-hydroxy-4′-methoxybenzyl)-
1,2,3,4-tetrahydroisoquinoline (12)
tetrahydroisoquinolin-7-yl)oxy)-N-ethoxycarbonyl-6,7-dimethoxy-
1-(3′-hydroxy-4′-methoxybenzyl)-1,2,3,4-tetrahydroisoquinoline
(14) (separable mixture of racemic diastereomers)
To a solution of diaryl ether 11 (1.00 g, 1.37 mmol) in metha-
nol (50 mL) palladium (10% on carbon, 100 mg, Pd Carbamate 12 (600 mg, 0.939 mmol) and enol ether 13
0.094 mmol) was added. The reaction was stirred for 12 h at (240 mg, 1.13 mmol) were condensed following General pro-
ambient temperature under hydrogen atmosphere (1 atm). The cedure 2 (intermolecular N-acyl Pictet–Spengler reaction). The
catalyst was removed by passing the mixture through a small reaction was completed after 12 h. For separation of the
plug of Celite, followed by washing with methanol (50 mL). racemic diastereomers of 14 15% ethyl acetate in dichloro-
The filtrate was concentrated in vacuo to give a white solid methane (Rf = 0.21 for the (1R,1′R)/(1S,1′S) isomers (14a) and
(0.805 g, 1.26 mmol, 92%) as desired product with no side pro- 0.30 for the (1R,1′S)/(1S,1′R) isomers (14b)) was used.
ducts. The isolated product was used without purification for
(1R,1′R)/(1S,1′S) isomers 14a: Yield: (385 mg, 0.471 mmol,
the following Pictet–Spengler cyclization. Purity (HPLC) = 99% 50%). Purity (HPLC) = 92% (λ = 210 nm). Mp: 73.0–75.0 °C. 1H
(λ = 210 nm). Mp: 61.5–62.0 °C. 1H NMR, COSY (400 MHz, NMR, COSY (400 MHz, Tcl2, 100 °C) δ [ppm] = δ 7.23–7.18 (m,
Tcl2, 100 °C) δ [ppm] = 6.81 (d, J = 2.0 Hz, 1H, H-2″), 6.66 (d, J 2H, H-13′, H-11′), 6.86–6.80 (m, 2H, H-14′, H-10′), 6.70 (s, 1H,
= 8.2 Hz, 1H, H-5′), 6.61 (d, J = 2.1 Hz, 1H, H-2′), 6.57 (dd, J = H-5′), 6.67 (d, J = 8.1 Hz, 1H, H-13), 6.63 (d, J = 2.0 Hz, 1H,
8.2, 2.0 Hz, 1H, H-6″), 6.52 (d, J = 8.2 Hz, 1H, H-5″), 6.52–6.48 H-10), 6.55 (s, 1H, H-5), 6.50 (dd, J = 8.2, 2.1 Hz, 1H, H-14),
(m, 2H, H-5, H-6′), 5.42 (s, 1H, OH), 5.36 (dd, J = 9.4 Hz, 1H, 6.22 (s, 1H, H-8′), 5.35 (s, 2H, H-1, OH), 5.05 (t, J = 6.3 Hz, 1H,
H-1), 4.64 (s, 1H, NH), 4.07 (q, J = 7.1 Hz, 2H, ‘OCH2–CH3), H-1′), 4.13–4.04 (m, 1H, H-3), 4.01 (q, J = 7.1 Hz, 2H, ‘OCH2–
4.08–3.94 (m, 1H, H-3), 3.92 (s, 3H, 3″-OCH3), 3.82 (s, 3H, CH3), 3.97–3.89 (m, 1H, H-3′), 3.91 (s, 3H, 6′-OCH3), 3.85 (s,
6-OCH3), 3.79 (s, 3H, 4′-OCH3), 3.87–3.66 (m, 2H, OCH2–CH3), 3H, 6-OCH3), 3.87–3.66 (m, 2H, OCH2–CH3), 3.80 (s, 3H,
3.61 (s, 3H, 7-OCH3), 3.37 (q, J = 7.0 Hz, 2H, NCH2–CH2), 12-OCH3), 3.61 (s, 3H, 7-OCH3), 3.40–3.30 (m, 1H, H-3),
3.39–3.28 (m, 1H, H-3), 3.10 (dd, J = 14.0, 3.8 Hz, 1H, H-α), 3.23–3.12 (m, 1H, H-3′), 2.98 (dd, J = 14.0, 3.6 Hz, 1H, H-α),
2.89–2.78 (m, 1H, H-4), 2.79–2.70 (m, 1H, H-α), 2.73 (t, J = 7.0 2.92–2.81 (m, 1H, H-4), 2.86 (dd, J = 6.4, 3.5 Hz, 2H, H-α′),
Hz, 2H, NCH2–CH2), 2.59 (d, J = 16.4 Hz, 1H, H-4), 1.20 (t, J = 2.81–2.73 (m, 1H, H-4′), 2.70 (dd, J = 13.9, 10.0 Hz, 1H, H-α),
7.1 Hz, 3H, ‘OCH2–CH3), 0.92 (br s, 3H, OCH2–CH3). 13C NMR, 2.62 (d, J = 16.3 Hz, 1H, H-4), 2.53 (dt, J = 16.0, 4.6 Hz, 1H,
DEPT, HMQC, HMBC (101 MHz, Tcl2, 100 °C) δ [ppm] = 156.5 H-4′), 1.14 (t, J = 7.1 Hz, 3H, ‘OCH2–CH3), 0.92 (br s, 3H,
(‘CvO), 155.4 (CvO), 152.6 (C-6), 149.5 (C-3″), 146.7 (C-4″), OCH2–CH3). 13C NMR, DEPT, HMQC, HMBC (101 MHz, Tcl2,
145.5 (C-8), 145.4 (C-4′), 145.3 (C-3′), 140.7 (C-7), 133.2 (C-1″), 100 °C) δ [ppm] = 155.4 (CvO), (‘CvO), 152.5 (C-6), 148.4
132.6 (C-1′), 130.1 (C-4a), 124.4 (C-8a), 121.0 (C-6′), 120.8 (C-6′), 146.2 (C-7′), 145.5 (C-11), 145.4 (C-12), 145.0 (C-8), 140.6
(C-6″), 116.1 (C-2′), 115.4 (C-5″), 114.4 (C-2″), 111.1 (C-5′), 109.8 (C-7), 137.4 (C-9′), 132.6 (C-9), 131.2 (C-14′, C-10′), 131.1 (C-13′,
(C-5), 60.9 (OCH2–CH3), 60.7 (7-OCH3, ‘OCH2–CH3, 56.7 (3″- C-11′), 130.0 (C-4a), 128.8 (C-8a′), 128.3 (C-4a′), 124.3 (C-8a),
OCH3), 56.4 (6-OCH3 or 4′-OCH3), 56.3 (6-OCH3 or 4′-OCH3), 121.0 (C-14), 120.2 (C-12′), 116.0 (C-10), 114.0 (C-8′), 113.7
52.0 (C-1), 42.3 (NCH2–CH2), 39.8 (C-α), 37.5 (C-3), 35.9 (C-5′), 111.0 (C-13), 110.1 (C-5), 61.2 (‘OCH2–CH3), 60.9 (OCH2–
(NCH2–CH2), 28.0 (C-4), 14.6 (‘OCH2–CH3), 14.2 (OCH2–CH3). CH3), 60.7 (7-OCH3), 56.6 (6′-OCH3), 56.5 (6-OCH3), 56.4
IR (ATR): v [cm−1] = 3354, 2939, 1686, 1594, 1508, 1419, 1331, (12-OCH3), 55.6 (C-1′), 52.0 (C-1), 42.6 (C-α′), 39.8 (C-α), 38.9
˜
1244, 1211, 1129, 1109, 1023, 960, 761. HRMS (ESI): m/z calcd (C-3′), 37.3 (C-3), 28.2 (C-4′), 28.0 (C-4), 14.6 (‘OCH2–CH3), 14.2
for [C34H41N2O10]− 637.2767, found: 637.2785.
(OCH2–CH3). IR (ATR): v [cm−1] = 3527, 2923, 2853, 1686, 1607,
˜
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
1508, 1426, 1330, 1237, 1196, 1120, 1068, 1024, 804, 760. flash column chromatography (35% acetone in hexanes, Rf =
HRMS (ESI): m/z calcd for [C42H47BrN2O10 + H]+ 819.2487 and 0.29). Yield obtained from enol ether 22: 13.5 mg,
821.2466, found: 819.2479 and 821.2460.
0.0183 mmol, 28%; yield obtained from enol ether 23:
(1R,1′S)/(1S,1′R) isomers 14b: Yield: (354 mg, 0.433 mmol, 38.0 mg, 0.0514 mmol, 79%.
46%). Purity (HPLC) = 96% (λ = 210 nm). Mp: 89.0–90.0 °C. 1H
(1R,1′R)/(1S,1′S) isomers 15a: Purity (HPLC) = 94% (λ =
NMR, COSY (400 MHz, Tcl2, 100 °C) δ [ppm] = 7.24–7.18 (m, 210 nm). Mp 132–134 °C. 1H NMR, COSY (400 MHz, Tcl2,
2H, H-13′, H-11′), 6.88–6.82 (m, 2H, H-14′, H-10′), 6.69 (s, 1H, 100 °C) δ [ppm] = 7.38 (d, J = 8.2 Hz, 1H, H-14′ or H-10′), 7.12
H-5′), 6.68 (d, J = 2.1 Hz, 1H, H-10), 6.67 (d, J = 8.2 Hz, 1H, (dd, J = 8.2, 2.5 Hz, 1H, H-13′ or H-11′), 6.79 (d, J = 8.1 Hz, 1H,
H-13), 6.55 (s, 1H, H-5), 6.53 (dd, J = 8.4, 1.9 Hz, 1H, H-14), H-13), 6.67 (s, 1H, H-13′ or H-11′), 6.64 (s, 1H, H-5′), 6.56 (dd, J
6.16 (s, 1H, H-8′), 5.38 (s, 1H, OH), 5.32 (dd, J = 9.7, 3.2 Hz, 1H, = 8.1, 2.0 Hz, 1H, H-14), 6.47 (d, J = 1.9 Hz, 1H, H-10), 6.32 (s,
H-1), 5.03 (t, J = 6.2 Hz, 1H, H-1′), 4.13–3.93 (m, 4H, H-3, H-3′, 1H, H-5), 6.19 (dd, J = 8.4, 2.2 Hz, 1H, H-14′ or H-10′), 6.16 (s,
‘OCH2–CH3), 3.90 (s, 3H, 6′-OCH3), 3.84 (s, 3H, 6-OCH3), 1H, H-8′), 5.26 (t, J = 6.1 Hz, 1H, H-1), 5.04 (s, 1H, H-1′),
3.86–3.74 (m, 2H, OCH2–CH3), 3.80 (s, 3H, 12-OCH3), 3.62 (s, 4.36–4.20 (m, 3H, H-3, ‘OCH2–CH3), 4.03–3.95 (m, 1H, H-3′),
3H, 7-OCH3), 3.42–3.31 (m, 1H, H-3), 3.22–3.08 (m, 2H, H-α, 3.87 (s, 3H, 12-OCH3), 3.73 (s, 3H, 6-OCH3), 3.89–3.59 (m, 2H,
H-3′), 2.98–2.87 (m, 1H, H-4), 2.85 (dd, J = 6.3, 3.4 Hz, 2H, H- OCH2–CH3), 3.53–3.37 (m, 3H, H-3, H-3′, H-α′), 3.34 (s, 3H, 6′-
α′), 2.84–2.71 (m, 2H, H-α, H-4′), 2.64 (d, J = 16.2 Hz, 1H, H-4), OCH3), 3.26 (s, 3H, 7-OCH3), 3.19–3.07 (m, 1H, H-4′), 2.94–2.75
2.53 (dt, J = 16.0, 4.4 Hz, 1H, H-4′), 1.14 (t, J = 7.1 Hz, 3H, (m, 2H, H-4, H-4′), 2.71 (d, J = 5.9 Hz, 2H, H-α), 2.69–2.59 (m,
‘OCH2–CH3), 0.95 (br s, 3H, OCH2–CH3). 13C NMR, DEPT, 2H, H-4, H-α′), 1.36 (t, J = 7.1 Hz, 3H, ‘OCH2–CH3), 0.86 (br s,
HMQC, HMBC (101 MHz, Tcl2, 100 °C) δ [ppm] = 155.4 (CvO, 3H, OCH2–CH3). 13C NMR, DEPT, HMQC, HMBC (101 MHz,
‘CvO), 152.4 (C-6), 148.3 (C-6′), 146.1 (C-7′), 145.5 (C-8), 145.4 Tcl2, 100 °C) δ [ppm] = 155.9 (‘CvO), 155.6 (CvO), 154.3
(C-12), 145.0 (C-11), 140.7 (C-7), 137.6 (C-9′), 132.6 (C-9), 131.2 (C-12′), 152.0 (C-6), 150.2 (C-11), 149.1 (C-6′), 147.9 (C-12),
(C-14′, C-10′), 131.1 (C-13′, C-11′), 130.0 (C-4a), 128.9 (C-8a′), 147.0 (C-8), 144.9 (C-7′), 138.9 (C-7), 134.6 (C-9′), 133.8 (C-9),
128.1 (C-4a′), 124.2 (C-8a), 121.1 (C-14), 120.1 (C-12′), 116.1 132.2 (C-14′ or C-10′), 130.1 (C-14′ or C-10′), 130.0 (C-4a′), 128.8
(C-10), 113.6 (C-5′, C-8′), 111.0 (C-13), 110.2 (C-5), 61.2 (‘OCH2– (C-8a′), 128.1 (C-4a), 122.9 (C-8a), 122.0 (C-13′ or C-11′), 121.9
CH3), 60.9 (OCH2–CH3), 60.7 (7-OCH3), 56.6 (6-OCH3 or 6′- (C-14), 121.5 (C-13′ or C-11′), 119.6 (C-8′), 116.8 (C-10), 114.1
OCH3), 56.5 (6-OCH3 or 6′-OCH3), 56.3 (12-OCH3), 55.7 (C-1′), (C-5′), 113.6 (C-13), 107.4 (C-5), 61.4, (‘OCH2–CH3), 60.7
52.0 (C-1), 42.7 (C-α′), 39.8 (C-α), 38.7 (C-3′), 37.5 (C-3), 28.3 (OCH2–CH3), 60.3 (7-OCH3), 57.7 (C-1′), 57.1 (12-OCH3 or 6′-
(C-4′), 28.1 (C-4), 14.6 (‘OCH2–CH3), 14.3 (OCH2–CH3). IR OCH3), 57.0 (12-OCH3 or 6′-OCH3), 56.3 (6-OCH3), 53.5 (C-1),
(ATR): v [cm−1] = 3438, 2931, 2840, 1692, 1609, 1510, 1427, 41.9 (C-3′, C-α′), 40.9 (C-α), 36.6 (C-3), 28.0 (C-4), 27.8 (C-4′),
˜
1332, 1237, 1201, 1122, 1069, 1026, 804, 760. HRMS (ESI): m/z 14.8 (‘OCH2–CH3), 14.2 (OCH2–CH3). IR (ATR): v [cm−1] = 2940,
˜
calcd for [C42H47BrN2O10 + H]+ 819.2487 and 821.2466, found: 2832, 1687, 1585, 1507, 1417, 1330, 1277, 1252, 1230, 1207,
819.2480 and 821.2461.
1123, 1097, 1024, 841, 768. HRMS (ESI): m/z calcd for
[C42H46N2O10 + H]+ 739.3225, found: 739.3223.
( )-N,N′-Bis-(ethoxycarbonyl)-nortetrandrine and isomers (15)
(1R,1′S)/(1S,1′R) isomers 15b: Purity (HPLC) = 95% (λ =
1
Method A – Preparation by cyclization via intramolecular 210 nm). Mp 142.5–144.0 °C. H NMR, COSY (400 MHz, Tcl2,
Ullmann-type C–O coupling reaction:
100 °C) δ [ppm] = 7.40–7.33 (m, 1H, H-14′ or H-10′), 7.07 (dd, J
Previously separated diastereomers of bisbenzylisoquino- = 8.2, 2.5 Hz, 1H, H-13′ or H-11′), 6.76 (d, J = 8.0 Hz, 1H, H-13),
line 14 (200 mg, 0.244 mmol of pure diastereomer) or of bis- 6.66–6.60 (m, 2H, 5′, H-13′ or H-11′), 6.57 (dd, J = 8.0, 2.0 Hz,
benzylisoquinoline 21, respectively (200 mg, 0.244 mmol of 1H, H-14), 6.42–6.36 (m, 1H, H-14′ or H-10′), 6.32 (s, 1H, H-10),
pure diastereomer) were reacted following General Procedure 3 6.28 (s, 1H, H-5), 6.17 (s, 1H, H-8′), 5.26 (d, J = 8.7 Hz, 1H,
(intramolecular C–O coupling). The reactions were completed H-1), 5.10 (dd, J = 10.5, 6.2 Hz, 1H, H-1′), 4.33–4.07 (m, 3H,
after 52 h. Purification was accomplished by flash column H-3, ‘OCH2–CH3), 3.96–3.89 (m, 1H, H-3′), 3.87 (s, 3H,
chromatography (35% acetone in hexanes, Rf = 0.29) and the 12-OCH3), 3.85–3.77 (m, 2H, OCH2–CH3), 3.73 (s, 3H, 6-OCH3),
products obtained as beige solids.
3.61 (s, 3H, 6′-OCH3), 3.57–3.42 (m, 2H, H-3′, H-α′), 3.38–3.25
Yields obtained from bisbenzylisoquinoline 14a: (1R,1′R)/ (m, 1H, H-3), 3.16 (s, 3H, 7-OCH3), 3.15–3.08 (m, 2H, H-α,
(1S,1′S) isomers: 115 mg, 0.156 mmol, 64%; (1R,1′S)/(1S,1′R) H-4′), 2.82 (dt, J = 15.8, 4.7 Hz, 1H, H-4′), 2.77–2.67 (m, 2H,
isomers: 112 mg, 0.152 mmol, 62%.
H-4, H-α′), 2.55 (dd, J = 13.9, 9.5 Hz, 1H, H-α), 2.50–2.35 (m,
Yields obtained from bisbenzylisoquinoline 21a: (1R,1′R)/ 1H, H-4), 1.33 (t, J = 7.1 Hz, 3H, ‘OCH2–CH3), 1.00 (br s, 3H,
(1S,1′S) isomers: 91.9 mg, 0.124 mmol, 51%; (1R,1′S)/(1S,1′R) OCH2–CH3). 13C NMR, HSQC, HMBC (101 MHz, Tcl2, 100 °C) δ
isomers: 63.1 mg, 0.0854 mmol, 35%.
Method B – Preparation by intramolecular N-acyl Pictet– 150.4 (C-11), 150.1 (C-6′), 147.9 (C-12), 144.4 (C-7′), 137.9 (C-7),
Spengler reaction: 134.8 (C-9′), 134.0 (C-9), 131.5 (C-14′ or C-10′), 130.4 (C-4a′),
[ppm] = 155.8 (CvO), 155.5 (‘CvO), 154.4 (C-12′), 152.5 (C-6),
Enol ether 22 or enol ether 23, respectively (50.0 mg, 130.0 (C-14′ or C-10′), 129.6 (C-4a), 128.3 (C-8a′), 122.1 (C-14,
0.0649 mmol) were reacted following General procedure 2 C-13′ or 11′), 122.0 (C-13′ or C-11′), 120.6 (C-8a), 119.6 (C-8′),
(intramolecular N-acyl Pictet–Spengler reaction). The reactions 117.1 (C-10), 113.3 (C-13), 111.9 (C-5′), 106.9 (C-5), 61.3
were completed after 12 h. Purification was accomplished by (‘OCH2–CH3), 60.9 (OCH2–CH3), 60.5 (7-OCH3), 57.1
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
(12-OCH3), 1′, 56.3 (6-OCH3), 56.1 (6′-OCH3), 54.1 (C-1), 41.7 ( )-7-(5-(2-((Ethoxycarbonyl)amino)ethyl)-2,3-dimethoxyphenoxy)-
(C-3′), 41.6 (C-α′), 39.0 (C-α), 28.2 (C-4), 28.1 (C-4′), 14.8 N-ethoxycarbonyl-6-methoxy-1-(4′-benzyloxybenzyl)-1,2,3,4-tetra-
(‘OCH2–CH3), 14.3 (OCH2–CH3). The resonances of C-3 and hydroisoquinoline (18)
C-8 could not be identified. IR (ATR): v [cm−1] = 2929, 2839,
˜
Phenol intermediate 17 (1.70 g, 3.80 mmol) and aryl bromide
6 (1.51 g, 4.56 mmol) were coupled following General pro-
cedure 3 (intermolecular C–O coupling reaction). The reaction
was completed after 3 days. Purification by flash column
1693, 1584, 1505, 1417, 1330, 1276, 1255, 1231, 1204, 1124,
1101, 1020, 842, 770. HRMS (ESI): m/z calcd for [C42H46N2O10
+ H]+ 739.3225, found: 739.3225.
chromatography (30% acetone in hexanes, Rf = 0.30) affording
(E/Z)-4-(Benzyloxy)-1-(2-methoxyvinyl)benzene (16)
the product as a beige solid (1.69 g, 2.42 mmol, 64%). Purity
(HPLC) = 100% (λ = 210 nm). Mp: 46.5–48.0 °C. 1H NMR,
COSY (400 MHz, Tcl2, 100 °C) δ [ppm] = 7.40–7.36 (m, 2H, H-2-
Ph, H-6-Ph), 7.33 (t, J = 7.3 Hz, 2H, H-3-Ph, H-5-Ph), 7.30–7.24
(m, 1H, H-4-Ph), 6.94–6.90 (m, 2H, H-2′, H-6′), 6.82–6.77 (m,
2H, H-3′, H-5′), 6.68 (s, 1H, H-5), 6.50 (s, 1H, H-8), 6.45 (d, J =
1.9 Hz, 1H, H-2″), 6.20 (d, J = 1.9 Hz, 1H, H-6″), 5.12 (t, J = 6.5
Hz, 1H, H-1), 4.98 (s, 2H, OCH2–Ph), 4.49 (s, 1H, NH), 4.04 (q, J
= 7.1 Hz, 2H, ‘OCH2–CH3), 4.03–3.92 (m, 3H, H-3, OCH2–CH3),
3.83 (s, 3H, 3″-OCH3), 3.79 (s, 3H, 4″-OCH3), 3.78 (s, 3H,
6-OCH3), 3.30 (q, J = 6.8 Hz, 2H, NCH2–CH2), 3.23 (ddd, J =
13.6, 9.7, 4.6 Hz, 1H, H-3), 2.94 (tt, J = 13.7, 7.2 Hz, 2H, H-α),
2.81 (ddd, J = 15.8, 9.7, 5.9 Hz, 1H, H-4), 2.64 (t, J = 7.1 Hz, 2H,
NCH2–CH2), 2.61–2.52 (m, 1H, H-4), 1.18 (t, J = 7.1 Hz, 3H,
‘OCH2–CH3), 1.13 (t, J = 7.0 Hz, 3H, OCH2–CH3). 13C NMR,
HSQC, HMBC (101 MHz, Tcl2, 100 °C) δ [ppm] = 157.8 (C-4′),
156.4 (‘CvO), 155.5 (CvO), 154.0 (C-3″), 151.1 (C-5″), 149.9
(C-6), 144.4 (C-7), 139.0 (C-4″), 137.5 (C-1-Ph), 134.1 (C-1″),
130.8 (C-1′), 130.6 (C-2′, C-6′), 130.5 (C-8a), 129.8 (C-4a), 128.5
(C-3-Ph, C-5-Ph), 127.8 (C-4-Ph), 127.4 (C-2-Ph, C-6-Ph), 119.0
(C-8), 115.1 (C-3′, C-5′), 114.1 (C-5), 111.5 (C-6″), 108.6 (C-2″),
70.5 (OCH2–Ph), 61.2 (OCH2–CH3), 60.9 (4″-OCH3), 60.7
(‘OCH2–CH3), 56.7 (3″-OCH3), 56.6 (6-OCH3), 55.9 (C-1), 42.2
(NCH2–CH2), 42.1 (C-α), 38.5 (C-3), 36.2 (NCH2–CH2), 28.4
Obtained from 4-benzyloxybenzaldehyde (5.00 g, 23.6 mmol)
following General Procedure 1 (Wittig olefination). Purification
by flash column chromatography (5% ethyl acetate in hexanes,
Rf = 0.32) gave the title compound as a white solid (5.50 g,
22.9 mmol, 97%, E,Z-isomer ratio 1 : 1.4, estimated by NMR-
integrals). Purity (HPLC) = 96% (λ = 210 nm). Mp: 37.5 °C.
NMR data of the major Z-isomer: 1H NMR, COSY (400 MHz,
CDCl3) δ [ppm] = 7.55–7.48 (m, 2H, H-6, H-2), 7.46–7.41 (m,
2H, H-2-Ph, H-6-Ph), 7.41–7.35 (m, 2H, H-3-Ph, H-5-Ph),
7.35–7.29 (m, 1H, H-4-Ph), 6.93–6.87 (m, 2H, H-3, H-5), 6.06
(d, J = 7.0 Hz, 1H, H-2′), 5.18 (d, J = 7.0 Hz, 1H, H-1′), 5.06 (s,
2H, OCH2–Ph), 3.76 (s, 3H, OCH3). 13C NMR, DEPT, HMQC,
HMBC (101 MHz, CDCl3): δ [ppm] = 156.9 (C-4), 146.6 (C-2′),
137.3 (C-1-Ph), 129.5 (C-6, C-2), 129.2 (C-1), 128.7 (C-3-Ph, C-5-
Ph), 128.0 (C-4-Ph), 127.6 (C-2-Ph, C-6-Ph), 114.8 (C-3, C-5),
105.3 (C-1′), 70.1 (OCH2–Ph), 60.6 (OCH3). IR (ATR): v [cm−1] =
˜
3034, 2834, 2080, 1639, 1606, 1236, 1093, 836, 739. HRMS (EI):
m/z calcd for [C16H16O2]•+ 240.1145, found: 240.1140.
( )-N-Ethoxycarbonyl-7-hydroxy-6-methoxy-1-(4′-benzyloxybenzyl)-
1,2,3,4-tetrahydroisoquinoline (17)
Carbamate 10 (0.950 g, 3.97 mmol) and enol ether 16 (1.43 g,
5.96 mmol) were condensed following General procedure 2
(intermolecular N-acyl Pictet–Spengler reaction). The reaction
was completed after 12 h. Purification by flash column chrom-
atography (5% ethyl acetate in dichloromethane, Rf = 0.29)
gave 17 as a white solid (1.71 g, 3.81 mmol, 96%). Purity
(C-4), 14.6 (‘OCH2–CH3, OCH2–CH3). IR (ATR): v [cm−1] = 2923,
˜
2849, 1686, 1583, 1508, 1424, 1231, 1102, 1012, 824, 764, 736.
HRMS (ESI): m/z calcd for [C40H46N2O9 + H]+ 699.3276, found:
699.3277.
1
(HPLC) = 99% (λ = 210 nm). Mp: 45.5–48.0 °C. H NMR, COSY
( )-7-(5-(2-((Ethoxycarbonyl)amino)ethyl)-2,3-dimethoxyphenoxy)-
N-ethoxycarbonyl-6-methoxy-1-(4′-hydroxybenzyl)-1,2,3,4-tetra-
hydroisoquinoline (19)
(400 MHz, Tcl2, 100 °C) δ [ppm] = 7.41–7.37 (m, 2H, H-2-Ph,
H-6-Ph), 7.37–7.31 (m, 2H, H-3-Ph, H-5-Ph), 7.30–7.25 (m, 1H,
H-4-Ph), 7.00–6.94 (m, 2H, H-2′, H-6′), 6.88–6.83 (m, 2H, H-3′,
H-5′), 6.58–6.52 (m, 2H, H-8, H-5), 5.33 (s, 1H, OH), 5.14 (t, J = To a solution of diaryl ether 18 (1.69 g, 2.42 mmol) in metha-
6.7 Hz, 1H, H-1), 5.02 (s, 2H, OCH2–Ph), 4.04–3.88 (m, 3H, H-3, nol (50 mL) palladium (10% on carbon, 169 mg, Pd
OCH2–CH3), 3.84 (s, 3H, 6-OCH3), 3.22 (ddd, J = 13.7, 9.9, 4.6 1.59 mmol) was added. The mixture was stirred for 12 h at
Hz, 1H, H-3), 2.97 (d, J = 6.6 Hz, 2H, H-α), 2.76 (ddd, J = 15.9, ambient temperature under hydrogen atmosphere (1 atm). The
9.9, 5.9 Hz, 1H, H-4), 2.51 (dt, J = 15.8, 4.4 Hz, 1H, H-4), 1.10 catalyst was removed by filtration through a small plug of
(t, J = 7.1 Hz, 3H, OCH2–CH3). 13C NMR, HSQC, HMBC Celite, followed by washing with methanol (50 mL). The filtrate
(101 MHz, Tcl2, 100 °C) δ [ppm] = 157.8 (C-4′), 155.5 (C-7), was concentrated in vacuo. Purification by flash column
145.7 (C-6), 144.1 (CvO), 137.6 (C-1-Ph), 131.1 (C-1′), 130.6 chromatography (35% acetone in hexanes, Rf = 0.22) gave the
(C-2′, C-6′), 130.0 (C-8a), 128.5 (C-3-Ph, C-5-Ph), 127.8 (C-4-Ph), title compound as a white solid (1.38 g, 2.27 mmol, 94%).
127.4 (C-2-Ph, C-6-Ph), 126.1 (C-4a), 115.2 (C-3′, C-5′), 113.4 Purity (HPLC) = 100% (λ = 210 nm). Mp: 63.0–65.5 °C. 1H
(C-8), 111.3 (C-5), 70.5 (OCH2–Ph), 61.1 (OCH2–CH3), 56.3 NMR, COSY (400 MHz, Tcl2, 100 °C) δ [ppm] = 6.85–6.80 (m,
(6-OCH3), 56.1 (C-1), 42.1 (C-α), 38.5 (C-3), 28.2 (C-4), 14.6 2H, H-2′, H-6′), 6.68 (s, 1H, H-5), 6.66–6.62 (m, 2H, H-3′, H-5′),
(OCH2–CH3). IR (ATR): v [cm−1] = 3032, 2929, 2360, 1683, 1509, 6.45 (d, J = 1.9 Hz, 1H, H-2″), 6.30 (s, 1H, H-8), 6.19 (d, J = 1.9
˜
1428, 1232, 1201, 1098, 1014, 696. HRMS (ESI): m/z calcd for Hz, 1H, H-6″), 5.49 (s, 1H, OH), 5.04 (t, J = 6.6 Hz, 1H, H-1),
[C27H29NO5 + H]+ 448.2118, found: 448.2120.
4.61 (t, J = 5.7 Hz, 1H, NH), 4.08 (q, J = 7.1 Hz, 4H, OCH2–CH3,
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
‘OCH2–CH3), 3.95–3.85 (m, 1H, H-3), 3.84 (s, 3H, 3″-OCH3), Hz, 1H, H-1), 5.02 (t, J = 6.5 Hz, 1H, H-1′), 4.86 (s, 1H, OH),
3.78 (s, 3H, 6-OCH3), 3.76 (s, 3H, 4″-OCH3), 3.36–3.27 (m, 3H, 4.14–4.04 (m, 1H, H-3), 4.00 (q, J = 7.3 Hz, 2H, ‘OCH2–CH3),
H-3, NCH2–CH2), 2.99 (dd, J = 13.5, 5.9 Hz, 1H, H-α), 2.86–2.76 3.91 (s, 3H, 6′-OCH3), 3.96–3.89 (m, 1H, H-3′), 3.85 (s, 3H,
(m, 2H, H-4, H-α), 2.68–2.61 (m, 3H, H-4, NCH2–CH2), 1.23–1.17 6-OCH3), 3.87–3.71 (m, 2H, OCH2–CH3), 3.79 (s, 3H, 12-OCH3),
(m, 6H, OCH2–CH3, ‘OCH2–CH3). 13C NMR, HSQC, HMBC 3.63 (s, 3H, 7-OCH3), 3.37–3.28 (m, 1H, H-3), 3.18 (ddd, J = 13.4,
(101 MHz, Tcl2, 100 °C) δ [ppm] = 156.8 (‘CvO), 155.5 (CvO), 9.5, 4.4 Hz, 1H, H-3′), 3.05 (dd, J = 14.1, 3.7 Hz, 1H, H-α), 2.88
154.7 (C-4′), 154.0 (C-3″), 150.8 (C-5″), 149.7 (C-6), 144.7 (C-7), (d, J = 6.4 Hz, 1H, H-4), 2.84 (t, J = 6.4 Hz, 2H, H-α′), 2.75 (td, J =
139.2 (C-4″), 133.9 (C-1″), 130.7 (C-2′, C-6′), 130.1 (C-4a, C-1′), 16.2, 14.4, 9.9 Hz, 2H, H-α, H-4′), 2.61 (d, J = 6.3 Hz, 1H, H-4),
129.5 (C-8a), 118.6 (C-8), 115.3 (C-3′, C-5′), 114.0 (C-5), 112.2 2.55 (td, J = 11.3, 5.5 Hz, 1H, H-4′), 1.14 (t, J = 7.0 Hz, 3H,
(C-6″), 108.9 (C-2″), 61.2 (‘OCH2–CH3 or OCH2–CH3), 61.0 ‘OCH2CH3), 0.89 (br s, 3H, OCH2CH3). 13C NMR, HSQC, HMBC
(‘OCH2–CH3 or OCH2–CH3), 60.8 (4″-OCH3), 56.7 (3″-OCH3), (101 MHz, Tcl2, 100 °C) δ [ppm] = 155.5 (‘CvO), 155.4 (CvO),
56.6 (6-OCH3), 56.3 (C-1), 42.2 (C-α, NCH2–CH2), 39.1 (C-3), 36.2 154.8 (C-12), 154.3 (C-12′), 152.6 (C-6), 148.2 (C-6′), 146.0 (C-7′),
(NCH2–CH2), 28.4 (C-4), 14.7 (OCH2CH3 or ‘OCH2CH3), 14.6 145.0 (C-8), 140.8 (C-7), 134.2 (C-10), 133.2 (C-9), 130.5 (C-14′,
(OCH2CH3 or ‘OCH2CH3). IR (ATR): v [cm−1] = 2927, 2833, 2362, C-10′), 130.4 (C-9′), 130.0 (C-8a), 129.4 (C-14), 129.2 (C-8a′),
˜
1689, 1508, 1425, 1230, 1100, 1007, 773. HRMS (ESI): m/z calcd 128.5 (C-4a′), 123.9 (C-4a), 115.3 (C-13′, C-11′), 114.1 (C-8′), 113.5
for [C33H40N2O9 + H]+ 609.2807, found: 609.2810.
(C-5′), 112.5 (C-13), 111.7 (C-11), 110.1 (C-5), 61.2 (‘OCH2–CH3),
61.1 (OCH2–CH3), 60.7 (7-OCH3), 56.7 (6-, 6′- or 12-OCH3), 56.6
(6-, 6′- or 12-OCH3), 56.6 (6-, 6′- or 12-OCH3), 55.9 (C-1′), 52.0
(E/Z)-3-Bromo-4-methoxy-1-(2-methoxyvinyl)benzene (20)
Obtained from 3-bromo-4-methoxybenzaldehyde (0.500 g, (C-1), 42.1 (C-α′), 39.2 (C-α), 38.7 (C-3′), 37.4 (C-3), 28.3 (C-4′),
−1
˜
2.33 mmol) following General Procedure 1 (Wittig olefination). 28.0 (C-4), 14.6 (‘OCH2CH3), 14.2 (OCH2CH3). IR (ATR): v [cm
]
Purification by flash column chromatography (5% ethyl = 2918, 2850, 2366, 1668, 1612, 1497, 1427, 1331, 1255, 1120,
acetate in hexanes, Rf = 0.30) affording the product as a colour- 1020, 887, 742. HRMS (ESI): m/z calcd for [C42H47BrN2O10 + H]+
less oil (0.455 g, 1.87 mmol, 80%, E,Z-isomer ratio 1 : 0.91 esti- 819.2492 and 821.2472, found: 819.2500 and 821.2490.
mated by NMR-integrals). Purity (HPLC) = 92% (λ = 210 nm).
(1R,1′S)/(1S,1′R) isomers 21b: Yield: (70.5 mg, 0.0856 mmol,
NMR data of the major E-isomer: 1H NMR, COSY (400 MHz, 26%. Purity (HPLC) = 67% (λ = 210 nm). Mp: 71.5–73.5 °C. H
CDCl3) δ [ppm] = 7.46–7.41 (m, 1H, H-2), 7.11 (dd, J = 8.5, 2.2 NMR, COSY (400 MHz, Tcl2, 100 °C) δ [ppm] = 7.24 (d, J = 2.1
Hz, 1H, H-6), 6.93 (d, J = 12.9 Hz, 1H, H-2′), 6.82 (d, J = 5.1 Hz, Hz, 1H, H-10), 6.94 (dd, J = 8.3, 2.1 Hz, 1H, H-14), 6.77 (d, J = 8.4
1H, H-5), 5.71 (d, J = 12.9 Hz, 1H, H-1′), 3.87 (s, 3H, 4-OCH3), Hz, 2H, H-14′, H-10′), 6.74 (d, J = 8.3 Hz, 1H, H-13), 6.70 (s, 1H,
3.67 (s, 3H, 2′-OCH3). 13C NMR, DEPT, HMQC, HMBC H-5′), 6.56 (d, J = 8.3 Hz, 2H, H-13′, H-11′), 6.51 (s, 1H, H-5), 6.11
(101 MHz, CDCl3) δ [ppm] = 154.1 (C-4), 148.6 (C-2′), 130.7 (s, 1H, H-8′), 5.27 (d, J = 8.3 Hz, 1H, H-1), 5.10 (s, 1H, OH), 4.99
(C-1), 129.9 (C-2), 125.3 (C-6), 112.3 (C-5), 111.5 (C-3), 103.6 (t, J = 6.5 Hz, 1H, H-1′), 4.01 (q, J = 7.1 Hz, 2H, ‘OCH2–CH3),
1
(C-1′), 56.7 (2′-OCH3), 56.5 (4-OCH3). IR (ATR): v [cm−1] = 2937, 4.07–3.89 (m, 2H, H-3, H-3′), 3.90 (s, 3H, 6′-OCH3), 3.86–3.78 (m,
˜
2837, 2074, 1640, 1496, 1253, 1095, 1053, 816. HRMS (EI): m/z 2H, OCH2–CH3), 3.83 (s, 3H, 6-OCH3), 3.80 (s, 3H, 12-OCH3),
calcd for [C10H11BrO2]•+ 241.9937 and 243.9916, found: 3.59 (s, 3H, 7-OCH3), 3.40–3.31 (m, 1H, H-3′), 3.25 (d, J = 15.6
241.9938 and 243.9917.
Hz, 1H, H-3), 3.17 (dd, J = 14.1, 3.9 Hz, 1H, H-α), 2.92–2.73 (m,
5H, H-4, H-α, H-4′, H-α′), 2.62–2.52 (m, 2H, H-4, H-4′), 1.15 (t, J =
7.1 Hz, 3H, ‘OCH2CH3), 0.97 (br s, 3H, OCH2CH3). 13C NMR,
HSQC, HMBC (101 MHz, Tcl2, 100 °C) δ [ppm] = 155.6 (vO),
155.5 (‘CvO), 154.8 (C-12), 154.4 (C-12′), 152.6 (C-6), 148.2 (C-6′),
145.8 (C-7′), 145.2 (C-8), 140.7 (C-7), 134.3 (C-10), 133.2 (C-9),
( )-8-((1-(4-Hydroxybenzyl)-N-(ethoxycarbonyl)-6-methoxy-1,2,3,4-
tetrahydroisoquinolin-7-yl)oxy)-N-ethoxycarbonyl-6,7-dimethoxy-
1-(3′-bromo-4′-methoxybenzyl)-1,2,3,4-tetrahydroisoquinoline
(21) (separable mixture of racemic diastereomers)
Intermediate 19 (200 mg, 0.329 mmol) and enol ether 20 130.6 (C-14′, C-10′), 130.3 (C-9′), 130.1 (C-4a), 129.5 (C-14), 129.2
(96.0 mg, 0.394 mmol) were condensed following General (C-8a′), 128.4 (C-4a′), 123.5 (C-8a), 115.3 (C-13′, C-11′), 114.4
Procedure 2 (intermolecular N-acyl Pictet–Spengler reaction) to (C-8′), 113.4 (C-5′), 112.5 (C-13), 111.7 (C-11), 109.8 (C-5), 61.2
give a racemic mixture of diastereomers of tetrahydroisoquino- (‘OCH2–CH3, OCH2–CH3), 60.7 (7-OCH3), 56.7 (6-, 6′- or
line 21 in a ratio of 62 : 38 (R,R)/(S,S) : (S,R)/(R,S). The reaction 12-OCH3), 56.6 (6-, 6′- or 12-OCH3), 56.6 (6-, 6′- or 12-OCH3), 56.1
was completed after 12 h. Purification and separation of the (C-1′), 52.1 (C-1), 42.2 (C-α′), 39.4 (C-α) 38.8 (C-3), 37.9 (C-3′), 28.3
˜
diastereomers was accomplished by flash column chromato- (C-4′), 28.1 (C-4), 14.6 (‘OCH2CH3), 14.4 (OCH2CH3). IR (ATR): v
graphy (20% ethyl acetate in dichloromethane, Rf = 0.17 ((1R,1′ [cm−1] = 2945, 2344, 1686, 1613, 1497, 1431, 1331, 1255, 1120,
S)/(1S,1′R) isomers (21b) and 0.13 (1R,1′R)/(1S,1′S) isomers 1019, 949, 889, 743. HRMS (ESI): m/z calcd for [C42H47BrN2O10
(21a)) to give both diasereomers as white solids.
+
H]+ 819.2492 and 821.2472, found: 819.2502 and 821.2493.
(1R,1′R)/(1S,1′S) isomers 21a: Yield: 113 mg, 0.138 mmol,
42%. Purity (HPLC) = 75% (λ = 210 nm). Mp: 68.5–70.5 °C. H
NMR, COSY (400 MHz, Tcl2, 100 °C) δ [ppm] = 7.18 (d, J = 2.1
Hz, 1H, H-10), 6.91 (dd, J = 8.3, 2.1 Hz, 1H, H-14), 6.81–6.76
1
( )-8-(4-(2-((Ethoxycarbonyl)amino)ethyl)-2-methoxyphenoxy)-N-
eth-oxycarbonyl-6,7-dimethoxy-1-(4′-methoxy-3′-(4-(2-methoxy-
vinyl)phenoxy)benzyl)-1,2,3,4-tetrahydroisoquinoline (22)
(m, 2H, H-14′, H-10′), 6.74–6.68 (m, 2H, H-13, H-5′), 6.58–6.52 Phenol 12 (100 mg, 0.157 mmol) and enol ether 13 (36.7 mg,
(m, 3H, H-5, H-13′, H-11′), 6.21 (s, 1H, H-8′), 5.30 (d, J = 10.1 0.172 mmol) were coupled following General Procedure 3
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
(intermolecular Ullmann-type C–O coupling reaction). The 6.6 Hz, 2H, H-α), 2.81 (ddd, J = 16.0, 9.5, 6.6 Hz, 1H, H-4), 2.64
reaction was completed after 5 days. Purification by flash (t, J = 7.1 Hz, 2H, NCH2–CH2), 2.61–2.52 (m, 1H, H-4), 1.17 (t,
column chromatography (15% ethyl acetate in dichloro- J = 7.1 Hz, 3H, ‘OCH2CH3), 1.13 (t, J = 8.0 Hz, 3H, OCH2CH3).
methane, Rf = 0.26) gave 22 as a white solid (74.0 mg, 13C NMR, HSQC, HMBC (101 MHz, Tcl2, 100 °C) δ [ppm] =
0.0960 mmol, 61%). Purity (HPLC) = 100% (λ = 210 nm). Mp 156.8 (C-4′), 156.4 (‘CvO), 155.5 (C-3′′′), 154.0 (C-3″), 151.1
59.5–61.0 °C. 1H NMR, COSY (400 MHz, Tcl2, 100 °C) (C-5″), 149.9 (C-4′′′), 149.9 (C-6), 148.4 (C-2′′′′), 144.6 (C-7),
Z-isomer: δ [ppm] = 7.45–7.38 (m, 1H, H-2′′′ or H-6′′′), 139.0 (C-4″), 134.2 (C-1″), 132.2 (C-1′), 130.6 (C-2′, C-6′), 130.4
7.11–7.05 (m, 1H, H-2′′′ or H-6′′′), 6.86 (d, J = 12.9 Hz, 1H, (C-4a), 130.0 (C-8a), 129.7 (C-1′′′), 121.5 (C-6′′′), 118.9 (C-8),
H-2′′′′), 6.81 (d, J = 8.3 Hz, 1H, H-5′), 6.78 (d, J = 2.0 Hz, 1H, 118.1 (C-2′′′), 117.1 (C-3′, C-5′), 114.8 (C-5′′′), 114.1 (C-5), 111.6
H-6′), 6.77–6.71 (m, 3H, H-2″, H-3′′′, H-5′′′), 6.67 (d, J = 1.9 Hz, (C-6″), 108.7 (C-2″), 105.0 (C-1′′′′), 61.2 (OCH2–CH3), 60.9
1H, H-2′), 6.55 (dt, J = 8.3, 1.6 Hz, 1H, H-6″), 6.50 (d, J = 8.2 Hz, (4″-OCH3), 60.7 (‘OCH2–CH3), 56.8 (2′′′′-OCH3), 56.7 (4″-OCH3
1H, H-5″), 6.48 (s, 1H, H-5), 5.80 (d, J = 12.8 Hz, 1H, H-1′′′′), or 3″-OCH3), 56.7 (4″-OCH3 or 3″-OCH3), 56.6 (6-OCH3), 55.8
5.32 (d, J = 8.7 Hz, 1H, H-1), 4.58 (s, 1H, NH), 4.07 (q, J = 7.1 (C-1), 42.2 (NCH2–CH2, C-α), 38.6 (C-3), 36.2 (NCH2–CH2), 28.4
Hz, 2H, ‘OCH2–CH3), 3.97 (d, J = 9.9 Hz, 1H, H-3), 3.91–3.75 (C-4), 14.6 (OCH2CH3, ‘OCH2CH3). The resonance of CvO
(m, 2H, OCH2–CH3), 3.81 (s, 6H, 6-OCH3, 3″-OCH3), 3.72 (s, could not be identified. IR (ATR): v [cm−1] = 2930, 2851, 1692,
˜
3H, 4′-OCH3), 3.64 (s, 3H, 2′′′′-OCH3), 3.59 (s, 3H, 7-OCH3), 3.35 1584, 1505, 1424, 1229, 1085, 1028, 941, 830, 767. HRMS (ESI):
(q, J = 6.9 Hz, 2H, NCH2–CH2), 3.32–3.24 (m, 1H, H-3), 3.13 m/z calcd for [C43H50N2O11 + K]+ 809.3052, found: 809.3045.
(dd, J = 14.1, 3.9 Hz, 1H, H-α), 2.85–2.73 (m, 2H, H-4, H-α),
( )-Tetrandrine (rac-1) and ( )-isotetrandrine (rac-2)
2.71 (t, J = 7.1 Hz, 2H, NCH2–CH2), 2.51 (d, J = 16.2 Hz, 1H,
H-4), 1.20 (t, J = 7.1 Hz, 3H, ‘OCH2–CH3), 0.96 (br s, 3H, OCH2– To a suspension of LiAlH4 (19 mg, 0.49 mmol) in anhydrous
CH3). 13C NMR, HSQC, HMBC (101 MHz, Tcl2, 100 °C) THF (3 mL),
a solution of carbamate 15 (45.0 mg,
Z-isomer: δ [ppm] = 156.5 (‘CvO), 156.4 (C-4′′′), 152.6 (C-6), 0.0609 mmol of either (1R,1′S)/(1S,1′R) or (1R,1′R)/(1S,1′S) dia-
150.4 (C-4′), 149.4 (C-3″), 148.1 (C-2′′′′), 146.6 (C-4″), 145.1 stereomer) in anhydrous THF (2 mL) was added dropwise
(C-3′), 140.7 (C-7), 133.2 (C-1″), 132.6 (C-1′), 130.4 (C-1′′′), 130.2 under nitrogen atmosphere. The mixture was refluxed for 6 h.
(C-4a), 129.4 (C-2′′′ or C-6′′′), 126.2 (C-2′′′ or C-6′′′), 125.8 (C-6′), After cooling to 0 °C deionized water (5 mL) was added very
124.1 (C-8a), 122.8 (C-2′), 120.8 (C-6″), 117.3 (C-3′′′ or C-5′′′), slowly. The mixture was brought to pH 12–14 with sodium
116.6 (C-3′′′ or C-5′′′), 115.4 (C-5″), 114.3 (C-2″), 114.2 (C-5′), hydroxide solution (2 M) and extracted with ethyl acetate (4 ×
109.7 (C-5), 105.4 (C-1′′′′), 61.0 (OCH2–CH3), 60.7 (‘OCH2–CH3), 20 mL). The combined organic phases were dried over anhy-
60.4 (7-OCH3), 56.9 (2′′′′-OCH3), 56.8 (4′-OCH3), 56.6 (3″- drous Na2SO4 and concentrated in vacuo to afford a yellow oil.
OCH3), 56.4 (6-OCH3), 52.1 (C-1), 42.3 (NCH2–CH2), 39.6 (C-α), Purification was accomplished by flash column chromato-
37.7 (C-3), 35.9 (NCH2–CH2), 28.0 (C-4), 14.6 (‘OCH2–CH3), graphy (dichloromethane → 1% triethylamine and 2% metha-
14.4 (OCH2–CH3). The resonances of C-8 and CvO could not nol in dichloromethane, Rf = 0.18). Racemic tetrandrine (rac-1)
be identified. IR (ATR): v [cm−1] = 2911, 1693, 1603, 1504, was obtained from (1R,1′R)/(1S,1′S)-15 as colourless needles
˜
1423, 1329, 1262, 1213, 1152, 1125, 1108, 1026, 839, 766. (37.0 mg, 0.0594 mmol, 98%) and racemic isotetrandrine (rac-
HRMS (ESI): m/z calcd for [C43H50N2O11 + H]+ 771.3487, found: 2) from (1R,1′S)/(1S,1′R)-15 as pale yellow prisms (37.3 mg,
771.3497.
0.0599 mmol, 98%), both after recrystallization from acetone.
For the separation of the mixtures of diastereomers resulting
from intramolecular N-acyl Pictet–Spengler reactions of 22 and
23 a preparative TLC method described by Lu et al.47 was used
with some modifications (0.5% triethylamine and 7.5% metha-
( )-7-(5-(2-((Ethoxycarbonyl)amino)ethyl)-2,3-dimethoxyphenoxy)-
N-ethoxycarbonyl-6-methoxy-1-(4′-(2-methoxy-5-(2-methoxyvinyl)
phenoxy)benzyl)-1,2,3,4-tetrahydroisoquinoline (23)
Phenol intermediate 19 (200 mg, 0.329 mmol) and enol ether nol in chloroform, Rf (rac-1) = 0.32; Rf (rac-2) = 0.22).
20 (87.9 mg, 0.361 mmol) were coupled following General Alternatively, preparative HPLC was performed on a Macherey-
Procedure 3 (intermolecular Ullmann-type C–O coupling reac- Nagel Nucleodur® 100–5 column (5 µm, 250 × 10 mm), eluent
tion). The reaction was completed after 48 h. Purification by 1% methanol and 0.1% diethylamine in chloroform; flow rate
flash column chromatography (15% ethyl acetate in dichloro- 7.0 mL min−1; temperature 25 °C; UV-detection at λ = 283 nm;
methane, Rf = 0.23) affording the product as a beige solid (Rt (rac-1) = 4.11 min; Rt (rac-2) = 4.93 min).
(114 mg, 0.141 mmol, 43%). Purity (HPLC) = 85% (λ =
( )-Tetrandrine (rac-1): Purity (HPLC) = 98% (λ = 210 nm).
210 nm). Mp: 47.0–50.5 °C. 1H NMR, COSY (400 MHz, Tcl2, Mp: 217–220 °C (literature: 252–253 °C (racemic tetran-
100 °C) δ [ppm] = 6.95–6.84 (m, 4H, H-5′′′, H-6′′′, H-2′, H-6′), drine)62). 1H NMR, COSY (500 MHz, CDCl3) δ [ppm] = 7.34
6.81–6.74 (m, 4H, H-2′′′, H-3′, H-5′, H-2′′′′), 6.68 (s, 1H, H-5), (dd, J = 8.2, 2.2 Hz, 1H, H-14′ or H-10′), 7.14 (dd, J = 8.2, 2.5
6.56 (s, 1H, H-8), 6.46 (s, 1H, H-2″), 6.21 (s, 1H, H-6″), 5.71 (d, Hz, 1H, H-13′ or H-11′), 6.88 (dd, J = 8.2, 1.6 Hz, 1H, H-14),
J = 12.8 Hz, 1H, H-1′′′′), 5.14 (s, 1H, H-1), 4.51 (s, 1H, NH), 4.04 6.86 (d, J = 8.1 Hz, 1H, H-13), 6.80 (dd, J = 8.2, 2.6 Hz, 1H,
(q, J = 7.1 Hz, 2H, ‘OCH2–CH3), 4.04–3.90 (m, 3H, 3, OCH2– H-13′ or H-11′), 6.55 (d, J = 1.7 Hz, 1H, H-10), 6.51 (s, 1H, H-5′),
CH3), 3.84 (s, 3H, 3″-OCH3), 3.80 (s, 3H, 4″-OCH3), 3.78 (s, 3H, 6.30 (s, 1H, H-5), 6.30 (dd, J = 8.3, 2.2 Hz, 1H, H-14′ or H-10′),
6-OCH3), 3.73 (s, 3H, 4′′′-OCH3), 3.60 (s, 3H, 2′′′′-OCH3), 3.30 (q, 5.99 (s, 1H, H-8′), 3.93 (s, 3H, 12-OCH3), 3.87 (dd, J = 11.0, 5.6
J = 6.9 Hz, 2H, NCH2–CH2), 3.27–3.18 (m, 1H, H-3), 2.95 (d, J = Hz, 1H, H-1′), 3.75 (s, 3H, 6-OCH3), 3.73 (s, 1H, H-1), 3.55–3.48
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
(m, 1H, H-3), 3.43 (ddd, J = 12.6, 10.0, 5.9 Hz, 1H, H-3′), 3.37 mediate structure were derived as follows: the antechamber
(s, 3H, 6′-OCH3), 3.25 (dd, J = 12.5, 5.7 Hz, 1H, H-α′), 3.19 (s, module of AmberTools17 was used for calculating initial
3H, 7-OCH3), 2.99–2.89 (m, 3H, H-3, H-4, H-4′), 2.89–2.84 (m, atomic charges on the AM1 level applying the bond charge cor-
1H, H-3′), 2.80 (t, J = 11.7 Hz, 1H, H-α′), 2.75–2.72 (m, 1H, rection and a net atomic charge of 1. Using Amber16,60 20.000
H-4′), 2.70 (dd, J = 14.2, 10.2 Hz, 1H, H-α), 2.62 (s, 3H, 2′- steps of initial conjugate-gradient minimization were per-
NCH3), 2.52 (d, J = 13.7 Hz, 1H, H-α), 2.45–2.39 (m, 1H, H-4), formed for all structures, which were subsequently heated up
2.33 (s, 3H, 2-NCH3). 13C NMR, HSQC, HMBC (151 MHz, to 300 K and simulated in gas phase for 4 μs. The trajectory
CDCl3) δ [ppm] = 153.9 (C-12′), 151.5 (C-6), 149.5 (C-11), 148.7 was clustered according to the dihedral torsions of the macro-
(C-6′), 148.6 (C-8), 147.2 (C-12), 143.9 (C-7′), 138.0 (C-7), 135.4 cycle using cpptraj.65 The representative structures of the first
(C-9′), 135.1 (C-9), 132.8 (C-14′ or C-10′), 130.3 (C-14′ or C-10′), 10 clusters were optimized with the Gaussian0966 program
128.3 (C-4a), 128.2 (C-8a′), 128.1 (C-4a′), 123.1 (C-8a), 122.9 package on the HF/6-31G(d) level of theory and charges were
(C-14), 122.1 (C-13′ or 11′), 122.0 (C-13′ or 11′), 120.3 (C-8′), fit to the electrostatic potential calculated according to the
116.4 (C-10), 112.9 (C-5′), 111.7 (C-13), 105.9 (C-5), 64.1 (C-1′), Merz–Singh–Kollman scheme.67 Those 10 conformations were
61.6 (C-1), 60.4 (7-OCH3), 56.3 (12-OCH3), 56.0 (6′-OCH3), 56.0 used as input structures for a multi-configurational RESP pro-
(6-OCH3), 45.4 (C-3′), 44.3 (C-3), 42.8 (2′-NCH3), 42.5 (2-NCH3), cedure to derive the final point charges. For the solvent the
42.1 (C-α), 38.4 (C-α′), 25.4 (C-4′), 22.2 (C-4). The NMR data are minimized DCM conformation was used for a charge deri-
identical with those of an authentic sample of tetrandrine. IR vation applying single-configurational RESP.
˜
(ATR): v [cm−1] = 2940, 2838, 1579, 1506, 1446, 1410, 1355,
Conformational sampling. All Molecular Dynamics (MD)
1268, 1213, 1121, 1023, 844, 767, 744. HRMS (ESI): m/z calcd simulations in this study were carried out with Amber16.60
for [C38H42N2O6 + H]+ 623.3116, found: 623.3111.
Each intermediate structure, prepared as described in the pre-
( )-Isotetrandrine (rac-2): Purity (HPLC) = 97% (λ = 210 nm). vious paragraph, was placed inside a truncated octahedron,
Mp: 175.5–182.0 °C (literature: 166–168 °C (enantiopure isote- which was solvated with DCM molecules within a 20 Å region
trandrine28) A melting point of racemic isotetrandrine is not around the solute and a chloride ion for neutralization. A
published yet). 1H NMR, COSY (500 MHz, CDCl3) δ [ppm] = series of minimizations was performed to achieve a target
7.27 (dd, J = 8.4, 2.3 Hz, 1H, H-14′ or H-10′), 7.10 (dd, J = 8.2, density of 1.0 g cm−3. Each system was heated to a target temp-
2.5 Hz, 1H, H-13′ or H-11′), 6.83–6.77 (m, 2H, H-13, H-14), erature of 258.15 K (−15 °C) to mimic the most favorable
6.67–6.63 (m, 1H, H-13′ or H-11′), 6.53 (s, 1H, H-5′), 6.48–6.37 experimental conditions (see entry 3 in Table 3). The heating
(m, 2H, H-10, H-14′ or H-10′), 6.27 (s, 1H, H-5), 5.98 (s, 1H, was performed over 150 ps with positional constraints on all
H-8′), 3.92 (s, 3H, 12-OCH3), 3.88–3.81 (m, 2H, H-1, H-1′), 3.75 atoms with a force constant of 3.0 kcal mol−1 Å−1 while
(s, 3H, 6-OCH3), 3.61 (s, 3H, 6′-OCH3), 3.44–3.36 (m, 1H, H-3′), heating to 20 K. Afterwards the constraints were removed from
3.33–3.21 (m, 2H, H-3, H-α′), 3.13 (s, 3H, 7-OCH3), 3.03 (d, J = non-solute atoms while heating to 200 K. In the final step the
14.2 Hz, 1H, H-α), 2.96–2.86 (m, 3H, H-4, H-4′, H-α′), 2.85–2.76 unconstraint system was further heated up to 258.15 K. All
(m, 3H, H-3, H-4′, H-3′), 2.60 (d, J = 10.4 Hz, 1H, H-α), 2.57 (s, heatup steps were performed in the NVT ensemble, whereas
3H, 2′-NCH3), 2.45–2.31 (m, 1H, H-4), 2.26 (s, 3H, 2-NCH3). 13
C
for all following simulations NPT ensemble was used. For
NMR, HSQC, HMBC (151 MHz, CDCl3) δ [ppm] = 154.2 (12′), pressure regulation the Berendsen barostat68 and for tempera-
152.0 (C-6), 150.0 (C-6′), 149.7 (C-11), 148.5 (C-4a), 147.1 (C-12), ture regulation the Langevin thermostat69 with a collision fre-
143.8 (C-7′), 137.3 (C-7), 135.4 (C-9′), 132.2 (C-14′ or C-10′), quency of 4.0 ps−1 were used. Throughout the simulation,
130.3 (C-14′ or 10′), 129.0 (C-4a′), 127.9 (C-8a′), 122.9 (C-14), bonds involving hydrogen atoms were constrained using the
122.2 (C-13′ or C-11′), 121.8 (C-13′ or C-11′), 120.7 (C-8a), 120.0 SHAKE70 algorithm, electrostatic energies were calculated
(C-8′), 116.0 (C-9), 111.5 (C-13), 111.4 (C-5′), 105.7 (C-5), 64.0 using the Particle Mesh Ewald method71 and periodic bound-
(C-1′), 62.1 (C-1), 60.6 (7-OCH3), 56.2 (12-OCH3), 55.9 (6-OCH3), ary conditions were applied. All simulations were performed
55.7 (6′-OCH3), 46.3 (C-3′), 43.0 (2′-NCH3), 42.8 (2-NCH3), 38.9 with a time step of 1 fs and a 12 Å cutoff for non-bonded inter-
(C-α), 37.9 (C-α′), 25.9 (C-4′), 23.2 (C-4). The resonances of C-3 actions, employing the pmemd.cuda engine of Amber1660 on
and C-8 could not be identified. IR (ATR): v [cm−1] = 2933, graphics processing units with mixed precision mode.60
˜
2834, 1582, 1506, 1445, 1413, 1260, 1229, 1114, 1030, 973, 865,
For each intermediate two replica simulations were per-
837, 772. HRMS (ESI): m/z calcd for [C38H42N2O6 + H]+ formed and the simulations were continued until 150 confor-
623.3116, found: 623.3108. The NMR data are in accordance mations could be extracted from the combined replica trajec-
with published data.63
tories, which featured a distance smaller than 3.1 Å (3.2 Å)
between atoms C-1′ and C-8a′ (C-1 and C-8a), i.e. the two atoms
forming the final bond. It has to be mentioned that in the first
Computational details
Structure preparation. All four macrocyclic intermediate 1000 ns less than 100 suitable structures were sampled for inter-
structures as shown in Fig. 3 as well as a single molecule of di- mediates of route 2b, which is why the simulation was
chloromethane (DCM) were built and minimized in increased to 2000 ns. This way, the same amount of structures
Avogadro.64
(150) could be extracted for intermediates of route 2a and 2b.
Parameterization. The General Amber Force Field61 para- These sets of structures represent possible starting confor-
meters were used for all molecules. Charges for each inter- mations for the final reaction and were thus named “pre-reac-
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
tion ensemble” and analyzed further regarding their orientation
around the second stereo center. This analysis revealed a struc-
tural preference of the C1-S intermediate of route 2a favouring
tetrandrine formation compared to the enantiomeric counter-
part (C1-R intermediate) that lacks a conformational preference
(see results section). Thus, to rule out possible sampling issues,
four additional replica trajectories for both intermediates of
route 2a were simulated for additional 1000 ns each and a
larger set of pre-reaction conformers could be extracted from all
combined trajectories, see next paragraph.
6 V. F. King, M. L. Garcia, D. Himmel, J. P. Reuben,
Y. K. Lam, J. X. Pan, G. Q. Han and G. J. Kaczorowski,
J. Biol. Chem., 1988, 263, 2238–2244.
7 S. Patel, Sci. Signaling, 2015, 8, re7.
8 Y. Sakurai, A. A. Kolokoltsov, C.-C. Chen, M. W. Tidwell,
W. E. Bauta, N. Klugbauer, C. Grimm, C. Wahl-Schott,
M. Biel and R. A. Davey, Science, 2015, 347, 995–998.
9 O. N. P. Nguyen, C. Grimm, L. S. Schneider, Y.-K. Chao,
C. Atzberger, K. Bartel, A. Watermann, M. Ulrich, D. Mayr,
C. Wahl-Schott, M. Biel and A. M. Vollmar, Cancer Res.,
2017, 77, 1427–1438.
Dihedral analysis. For all structures of the pre-reaction
ensembles the relative pre-reaction orientation (pre-S or pre-R) 10 D. Falzarano and H. Feldmann, Science, 2015, 347, 947–948.
at the second stereocenter was analyzed manually based on 11 M. Shamma, The isoquinoline alkaloids: chemistry and
the relative position of the hydrogen atom. Table S1‡ lists the
pharmacology, Academic Press, New York, 1972, ch. 5,
amount of structures within the first 150 pre-reaction confor-
vol. 25.
mations sampled for route 2a and 2b, respectively, that orig- 12 D. Longley and P. Johnston, J. Pathol., 2005, 205, 275–292.
inate from either intermediate of each route and are oriented 13 P. Joshi, R. A. Vishwakarma and S. B. Bharate, Eur. J. Med.
towards either pre-S or pre-R at the second stereocenter. With
Chem., 2017, 138, 273–292.
4 additional replicas sampled for route 2a the pre-reaction 14 S. Dewanjee, T. Dua, N. Bhattacharjee, A. Das,
ensembles increased to a total amount of 547 conformations,
for which the same analysis was repeated corresponding to the
M. Gangopadhyay, R. Khanra, S. Joardar, M. Riaz, V. Feo
and M. Zia-Ul-Haq, Molecules, 2017, 22, 871.
numbers in brackets in Table S1.‡ Furthermore, for the pre- 15 R. J. Kelly, R. W. Robey, C. C. Chen, D. Draper,
reaction ensemble the position of ring A′ (A) relative to the
second stereocenter was determined visually and defined in a
binary fashion as + (corresponding to the methoxy group(s)
V. Luchenko, D. Barnett, R. K. Oldham, Z. Caluag,
A. R. Frye, S. M. Steinberg, T. Fojo and S. E. Bates,
Oncologist, 2012, 17, 512–e523.
flipped backward in Fig. 4) and – (corresponding to the 16 C.-Z. Wu, L. Lai, X. Hu, R.-R. Lei and Y.-F. Yang, J. Asian
methoxy group(s) flipped forward in Fig. 4). Nat. Prod. Res., 2013, 15, 993–1002.
Potential energies. The average potential energies for the 17 Q.-H. Yang, C.-S. Jiang, T. Jin, J.-F. Xu, T.-L. Qu, Y.-W. Guo
intermediate structures of routes 2a were calculated with our
in-house program DynaDock72 by averaging over all pre-reac-
tion conformers.
and Z.-B. Zhao, J. Asian Nat. Prod. Res., 2018, 1–11, DOI:
10.1080/10286020.2018.1478817.
18 Z. Cao, M. Wright, J. Cheng, X. Huang, L. Liu, L. Wu and
P. Yang, Oncol. Rep., 2014, 32, 1211–1217.
19 J. Lan, N. Wang, L. Huang, Y. Liu, X. Ma, H. Lou, C. Chen,
Y. Feng and W. Pan, Eur. J. Med. Chem., 2017, 127, 554–566.
20 J. Lan, L. Huang, H. Lou, C. Chen, T. Liu, S. Hu, Y. Yao,
J. Song, J. Luo, Y. Liu, B. Xia, L. Xia, X. Zeng, Y. Ben-David
and W. Pan, Eur. J. Med. Chem., 2018, 143, 1968–1980.
21 J. Song, J. Lan, C. Chen, S. Hu, J. Song, W. Liu, X. Zeng,
H. Lou, Y. Ben-David and W. Pan, MedChemComm, 2018, 9,
1131–1141.
22 V. Calvillo-Páez, R. R. Sotelo-Mundo, M. Leyva-Peralta,
J. C. Gálvez-Ruiz, D. Corona-Martínez, R. Moreno-Corral,
R. Escobar-Picos, H. Höpfl, O. Juárez-Sánchez and
K. O. Lara, Chem.-Biol. Interact., 2018, 286, 34–44.
23 Y. Tian and J. Zheng, in Advances in Molecular Toxicology,
ed. J. C. Fishbein and J. M. Heilman, Academic Press,
Cambridge, San Diego, Oxford, London, 2017, ch. 6, vol.
11, pp. 241–272.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
Financial support was provided for F. B. by the Deutsche
Forschungsgemeinschaft (project BR1034/7-1) and for
M. M. and I. A. by the Deutsche Forschungsgemeinschaft
(SFB749/C08 and SFB1035/A10, CIPSM).
References
1 N. Bhagya and K. R. Chandrashekar, Phytochemistry, 2016,
125, 5–13.
2 N. Bhagya and K. R. Chandrashekar, Biomed.
Pharmacother., 2018, 97, 624–632.
24 H. Jin, L. Li, D. Zhong, J. Liu, X. Chen and J. Zheng, Chem.
Res. Toxicol., 2011, 24, 2142–2152.
25 Y. Tian, S. Shen, Y. Jiang, Q. Shen, S. Zeng and J. Zheng,
Arch. Toxicol., 2016, 90, 1737–1748.
3 H. Kondo and K. Yano, Yakugaku Zasshi, 1928, 48, 107–123.
4 K. K. Chen and A. L. Chen, J. Biol. Chem., 1935, 109, 681– 26 L. Tainlin, H. Tingyi, Z. Changqi, Y. Peipei and Z. Qiong,
685. Ecotoxicol. Environ. Saf., 1982, 6, 528–534.
5 C.-Y. Kwan and F. Achike, Acta Pharmacol. Sin., 2016, 23, 27 Y. Inubushi, Y. Masaki, S. Matsumoto and F. Takami,
1057–1068.
Tetrahedron Lett., 1968, 9, 3399–3402.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Organic & Biomolecular Chemistry
Paper
28 Y. Inubushi, Y. Masaki, S. Matsumoto and F. Takami, 55 L. Evanno, J. Ormala and P. M. Pihko, Eur. J. Chem., 2009,
J. Chem. Soc. C, 1969, 1547–1556. 15, 12963–12967.
29 J. Qingping, Y. Bin, G. Zhenzhen, D. Yanqun, C. Yu, J. Yan, 56 N. Blank and T. Opatz, J. Org. Chem., 2011, 76, 9777–9784.
Z. Mingqiang and Y. Weicheng, CN109942593, 2019.
30 D. L. Comins, P. M. Thakker, M. F. Baevsky and
M. M. Badawi, Tetrahedron, 1997, 53, 16327–16340.
31 J. F. Hartwig, Angew. Chem., Int. Ed., 1998, 37, 2046–2067.
32 D. M. T. Chan, K. L. Monaco, R.-P. Wang and
M. P. Winters, Tetrahedron Lett., 1998, 39, 2933–2936.
57 I. Ugur, M. Schroft, A. Marion, M. Glaser and I. Antes,
J. Mol. Model., 2019, 25, 197.
58 S. Wang, A. H. Crevenna, I. Ugur, A. Marion, I. Antes,
U. Kazmaier, M. Hoyer, D. C. Lamb, F. Gegenfurtner,
Z. Kliesmete, C. Ziegenhain, W. Enard, A. Vollmar and
S. Zahler, Sci. Rep., 2019, 9, 9731.
33 J. J. Maresh, A. A. Ralko, T. E. Speltz, J. L. Burke, 59 B. Eyermann, M. Meixner, H. Brötz-Oesterhelt, I. Antes and
C. M. Murphy, Z. Gaskell, J. K. Girel, E. Terranova, S. A. Sieber, ChemBioChem, 2020, 21, 235–240.
C. Richtscheidt and M. Krzeszowiec, Synlett, 2014, 2891– 60 D. A. Case, R. M. Betz, D. S. Cerutti, T. E. Cheatham, III,
2894.
T. A. Darden, R. E. Duke, T. J. Giese, H. Gohlke,
A. W. Goetz, N. Homeyer, S. Izadi, P. Janowski, J. Kaus,
A. Kovalenko, T. S. Lee, S. LeGrand, P. Li, C. Lin, T. Luchko,
R. Luo, B. Madej, D. Mermelstein, K. M. Merz, G. Monard,
H. Nguyen, H. T. Nguyen, I. Omelyan, A. Onufriev,
D. R. Roe, A. Roitberg, C. Sagui, C. L. Simmerling,
W. M. Botello-Smith, J. Swails, R. C. Walker, J. Wang,
R. M. Wolf, X. Wu, L. Xiao and P. A. Kollman, AMBER 2016,
University of California, San Francisco, 2016.
34 S. G. Davies, E. C. Goddard, P. M. Roberts, A. J. Russell,
A. D. Smith, J. E. Thomson and J. M. Withey, Tetrahedron:
Asymmetry, 2016, 27, 274–284.
35 A. Yokoyama, T. Ohwada and K. Shudo, J. Org. Chem., 1999,
64, 611–617.
36 C. H. Burgos, T. E. Barder, X. Huang and S. L. Buchwald,
Angew. Chem., Int. Ed., 2006, 45, 4321–4326.
37 D. Ma and Q. Cai, Org. Lett., 2003, 5, 3799–3802.
38 J. Wang and G. Evano, Org. Lett., 2016, 18, 3542–3545.
39 J. Shi, J. Wu, C. Cui and W.-M. Dai, J. Org. Chem., 2016, 81,
10392–10403.
40 L. Shen, C. J. Simmons and D. Sun, Tetrahedron Lett., 2012,
53, 4173–4178.
41 D. Maiti and S. L. Buchwald, J. Org. Chem., 2010, 75, 1791–
1794.
42 L. Navarro and M. D. Pujol, Tetrahedron Lett., 2015, 56,
1812–1815.
61 J. Wang, R. M. Wolf, J. W. Caldwell, P. A. Kollman and
D. A. Case, J. Comput. Chem., 2004, 25, 1157–1174.
62 S. M. Kupchan, N. Yokoyama and B. S. Thyagarajan,
J. Pharm. Sci., 1961, 50, 164–167.
63 H. Guinaudeau, A. J. Freyer and M. Shamma, Nat. Prod.
Rep., 1986, 3, 477–488.
64 M. D. Hanwell, D. E. Curtis, D. C. Lonie,
T. Vandermeersch, E. Zurek and G. R. Hutchison,
J. Cheminf., 2012, 4, 17.
43 S. Kang, R. J. Ono and C. W. Bielawski, J. Polym. Sci., Part A: 65 D. R. Roe and T. E. Cheatham, J. Chem. Theory Comput.,
Polym. Chem., 2013, 51, 3810–3817. 2013, 9, 3084–3095.
44 S. J. Gharpure, A. M. Sathiyanarayanan and 66 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
P. Jonnalagadda, Tetrahedron Lett., 2008, 49, 2974–
2978.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers,
K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox,
Gaussian 09, Revision E.01, Gaussian, Inc., Wallingford, CT,
2009.
45 C. G. Caldwell, Y.-C. Chiang, C. Dorn, P. Finke, J. Hale,
M. Maccoss, S. Mills and A. Robichaud, WO98/17660, 1998.
46 J. Kayhan, M. J. Wanner, S. Ingemann, J. H. van
Maarseveen and H. Hiemstra, Eur. J. Org. Chem., 2016,
3705–3708.
47 S.-T. Lu, I.-L. Tsai and S.-P. Leou, Phytochemistry, 1989, 28,
615–620.
48 E. Mons, M. J. Wanner, S. Ingemann, J. H. van Maarseveen
and H. Hiemstra, J. Org. Chem., 2014, 79, 7380–7390.
49 D. L. Comins and M. M. Badawi, Tetrahedron Lett., 1991,
32, 2995–2996.
50 C. Gremmen, M. J. Wanner and G.-J. Koomen, Tetrahedron
Lett., 2001, 42, 8885–8888.
51 M. Chrzanowska, A. Grajewska and M. D. Rozwadowska,
Chem. Rev., 2016, 116, 12369–12465.
52 R. E. Gawley, G. Hart, M. Goicoechea-Pappas and
A. L. Smith, J. Org. Chem., 1986, 51, 3076–3078.
53 N. Uematsu, A. Fujii, S. Hashiguchi, T. Ikariya and
R. Noyori, J. Am. Chem. Soc., 1996, 118, 4916–4917.
54 B. Horst, M. J. Wanner, S. I. Jørgensen, H. Hiemstra and
J. H. van Maarseveen, J. Org. Chem., 2018, 83, 15110–15117.
67 U. C. Singh and P. A. Kollman, J. Comput. Chem., 1984, 5,
129–145.
68 H. J. C. Berendsen, J. P. M. Postma, W. F. v. Gunsteren,
A. DiNola and J. R. Haak, J. Chem. Phys., 1984, 81, 3684–
3690.
This journal is © The Royal Society of Chemistry 2020
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
69 W. F. van Gunsteren and H. J. C. Berendsen, Angew. Chem., 71 T. Darden, D. York and L. Pedersen, J. Chem. Phys., 1993,
Int. Ed. Engl., 1990, 29, 992–1023. 98, 10089–10092.
70 J.-P. Ryckaert, G. Ciccotti and H. J. C. Berendsen, J. Comput. 72 I. Antes, Proteins: Struct., Funct., Bioinf., 2010, 78, 1084–
Phys., 1977, 23, 327–341.
1104.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2020