Geranyl derivatives of isoquinoline alkaloids show increased biological activities

Article abstract of DOI:10.3987/COM-10-11912

Three types of isoquinoline alkaloids were tested for antimicrobial, cytotoxic, anti-malarial, anti-oxidant, and anti-HIV activities, as well as inhibitory activity against Epstein-Barr virus early antigen (EBV-EA) activation induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) in Raji cells. N-or O-Geranylation contributed to increased potency in four types of activities except anti-HIV and anti-oxidant. Some types of alkaloids may be useful as lead compounds for developing potential chemotherapeutic agents. N,N-Geranylated salsolinol was significantly active in three different assays, antimicrobial, cytotoxic, and EBV-EA, and may be a most useful compound. The Japan Institute of Heterocyclic Chemistry.

Full text of DOI:10.3987/COM-10-11912

HETEROCYCLES, Vol. 81, No. 5, 2010
1193
HETEROCYCLES, Vol. 81, No. 5, 2010, pp. 1193 - 1229. © The Japan Institute of Heterocyclic Chemistry
Received, 19th January, 2010, Accepted, 1st March, 2010, Published online, 2nd March, 2010
DOI: 10.3987/COM-10-11912
GERANYL DERIVATIVES OF ISOQUINOLINE ALKALOIDS SHOW
INCREASED BIOLOGICAL ACTIVITIES
Yumi Nishiyama,a Kinuko Iwasa,*,a Suguru Okada,a Sousuke Takeuchi,a
Masataka Moriyasu,a Miyoko Kamigauchi,a Junko Koyama,a Atsuko
Takeuchi,a Harukuni Tokuda,b Hye-Sook Kim,c Yusuke Wataya,c Kazuyoshi
Takeda,d Yi-Nan Liu,e Pei-Chi Wu,e Kenneth F. Bastow,e Toshiyuki
Akiyama,e and Kuo-Hsiung Leee
aKobe Pharmaceutical University, 4-19-1 Motoyamakita, Higashinada-ku,
b
Kobe-shi 658-8558, Japan, Department of Biochemistry and Molecular Biology,
Kyoto Prefectural University of Medicine, Kawaramachi-dori, Kamigyo-ku,
Kyoto 602-0841, Japan, cFaculty of Pharmaceutical Science, Okayama University,
d
Tsushima, Okayama 700, Japan, Yokohama College of Pharmacy, 601
e
Matanocyo, Hodogayaku, Yokohama-shi 245-0066 Japan, Natural Products
Research Laboratories, Eshelman School of Pharmacy, University of North
Carolina, Chapel Hill, NC 27599-7360, USA
Abstract – Three types of isoquinoline alkaloids were tested for antimicrobial,
cytotoxic, anti-malarial, anti-oxidant, and anti-HIV activities, as well as
inhibitory activity against Epstein-Barr virus early antigen (EBV-EA) activation
induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) in Raji cells. N- or
O-Geranylation contributed to increased potency in four types of activities except
anti-HIV and anti-oxidant. Some types of alkaloids may be useful as lead
compounds for developing potential chemotherapeutic agents. N,N-Geranylated
salsolinol was significantly active in three different assays, antimicrobial,
cytotoxic, and EBV-EA, and may be a most useful compound.
INTRODUCTION
A number of aromatic prenyltransferases, responsible for prenyl group attachment, have only recently
been isolated and characterized.1-4 Aromatic prenyltransferases catalyze the transfer of a C5

1194
HETEROCYCLES, Vol. 81, No. 5, 2010
(dimethylallyl), C10 (geranyl) or C15 (farnesyl) prenyl group derived from the corresponding isoprenyl
diphosphate metabolites onto a variety of electron-rich aromatic acceptors. Prenyl groups appear in a
wide variety of bioactive natural products of microbial and plant origin, including amino acids, stilbenes,
alkaloids, polyketides and phenylpropanoids such as flavonoids, creating natural product hybrids with
altered or enhanced bioactivities. For example, prenylated indole alkaloids containing both aromatic and
isoprenoid moieties, are widely distributed in terrestrial and marine organisms.5 Prenylation appears in
many cases to provide a higher level of bioactivity compared to the nonprenylated precursors and
sometimes to cause biological activities distinct from those of the parent compounds.6,7 Thus,
prenylated compounds represent a new frontier for the development of novel drugs, in particular,
anti-microbial, anti-oxidant, anti-inflammatory, anti-viral, and anti-cancer agents. We have previously
tested the antimicrobial, antimalarial, cytotoxic, anti-HIV, and anti-oxidant activities and inhibitory
effects on EBV-EA induction of isoquinoline alkaloids.8,9 Some of the tested isoquinolines showed
significant activities in these assays. It was expected that addition of the prenyl group to the isoquinoline
alkaloids [simple isoquinolines, 1-benzylisoquinolines, and protoberberines] would increase their potency
compared with that of the parent compounds in some assays. In this paper, we describe the synthesis of
N- and N,N-geranyl, and O- and O,O-geranyl derivatives of isoquinoline alkaloids and compare their
activities with those of non-prenylated compounds in five assays: antimicrobial activity, antimalarial
activity, cytotoxicity evaluation,10 inhibitory effects on EBV-EA induction,11 free radical scavenging
activity, and anti-HIV activity.12
RESULTS AND DISCUSSIONS
Chemistry. Geranylation of (±)-salsolinol (1)8 with geranyl bromide and sodium hydride in
N,N-dimethylformamide has been previously described to give the N- and N,N-geranyl derivatives (2,
3).13 (±)-N-Methylsalsolinol (4) and quaternary (±)-N,N-dimethylsalsolinol (5) has been prepared from
(±)-carnegine.13 Geranylation of 5 by the same methods gave the O- and O,O-geranyl derivatives (6-8)
(Scheme 1). Geranylation of (±)-tetrahydropapaveroline (9)9 gave the N- and N,N-geranyl derivatives
(10 and 11), and (±)-N-methylpapaveroline (12) produced the O,O-geranyl derivative (13) (Scheme 2).
Geranylation of (±)-2,3,9,10-tetrademethyltetrahydropalmatine (14)9 afforded the N-geranyl derivatives
(15 and 16) and (±)-2,3,10,11-tetrademethylpseudotetrahydropalmatine (17)9 gave the N-geranyl
derivative (18) (Scheme 3). Each geranylated compound was purified by preparative HPLC of the
reaction mixtures using NH4OAc (0.05% TFA)-MeOH (0.05% TFA).

HETEROCYCLES, Vol. 81, No. 5, 2010
1195
5
8
4
6
7
HO
HO
4a
X-
H
3
6'
4'
2'
8'
N
7'
3'
1a
1
1'
5'
H
7'-Me
3'-Me
HO
HO
NaH
Cl-
NH2
(±)-2 (yield 10.3%)
geranyl bromide
DMF
7'-Me
3'-Me
X-
5
4
HO
HO
4a
6
3
1'
5'
2'
6'
(±)-1
4'
7'
8'
3'
N
1a
7
3''
8''
1
7''
8
2''
1''
4''
6''
H
5''
3''-Me
7''-Me
(±)-3 (yield 12.6%)
7'-Me
8'
7'
6'
5'
3'-Me
5
4
4'
HO
O
4a
1a
6
3
3'
X-
N
2'
1'
7
8
1
5
4
O
4a
1a
6
1'
3
HO
HO
2'
3'
Br-
X-
8'
7'-Me
NaH
geranyl bromide
N
7
N
7'
4'
HO
3'-Me
1
8
DMF
6'
5'
5'
6'
7'
3'-Me
(±)-7 (yield 7.4%)
(±)-5
(±)-6 (yield 7.7%)
4'
3'
8'
7'-Me
2'
1) CH3I
2) 47% HBr
1'
4
5
4a
O
6
3
X-
N
MeO
MeO
HO
1a
7
O
1
8
47% HBr
Cl-
Br-
1''
NH
NH
2''
HO
3"
4"
3"-Me
(±)-carnegine
(±)-4
5"
6"
7"
(±)-8 (yield 8.9%)
7"-Me
8"
Scheme 1. N-geranylation of simple isoquinolines
The HPLC of the reaction mixture in geranylation of salsolinol (1) is shown in Figure 1 (I). The peaks
a1 and b1 were determined to be (±)-N-geranylsalsolinol (2) and (±)-N,N-digeranylsalsolinol (3),
respectively (Scheme 1). Their structures, including the stereostructure, have been previously
described:13 orientation of the C (1)-methyl group is quasi-axial in 2 and 3 and that of the N-geranyl
group is quasi-equatorial in 2.

1196
HETEROCYCLES, Vol. 81, No. 5, 2010
8''
5
8
4
X
6''
HO
HO
7''
4a
6
7
4''
3
2''
3'
H
3''
5''
1''
N
7''-Me
1a
1
H
3''-Me
2'
9
HO
HO
1'
OH
Br
NH2
6'
4'
OH
NaH
geranyl bromide
5'
OH
OH
(±)-10 (yield 18.7%)
DMF
7''-Me
3''-Me
5
4
(±)-9
HO
HO
4a
X
5''
6
3
7''
3''
8''
1''
4''
6''
2''
N
3'''
8'''
1'''
1a
7
7'''
4'''
2'''
1
6'''
8
H
5'''
2'
3'''-Me
1'
9
3' OH
7'''-Me
6'
OH
4'
5'
(±)-11 (yield 9.5%)
8''
7''-Me
7''
6''
5''
3''-Me
4''
3''
HO
HO
2''
1''
Br
4
5
N
O
4a
1a
6
NaH
3
geranyl bromide
X
OH
OH
N
DMF
7
O
1
8
12
2'
1'''
9
3'
1'
OH
2'''
3'''
6'
4'''
4' OH
3'"-Me
5'
5'''
6'''
13 (yield 4.1%)
7'''
8'''
7'"-Me
Scheme 2. N-geranylation of 1-benzylisoquinolines
These stereostructures were identical with the optimized geometry of 2 and 3 calculated. The optimized
geometry and molecular orbital of 2 and 3 were calculated by the DFT (Density Function Theory) method
using the Materials Studio DMol3 package of Accelrys Inc.14,15 (Figure 1) and they may be necessary to
clarify the biological activities’ mechanisms.
HPLC of the reaction mixture in geranylation of (±)-N,N-dimethylsalsolinol (5) is shown in Figure 1.
Compounds 6-8, corresponding to the peaks a2, b2, and c2, respectively, (Figure 1, II) were separated and
purified by preparative HPLC. The molecular formula of compound 6 was determined to be C22H34NO2
by analysis of its HRSIMS ([M]+, m/z 344.2606), which indicates the presence of one geranyl group.

HETEROCYCLES, Vol. 81, No. 5, 2010
1197
7'-Me
3'-Me
4
5
HO
HO
X
HO
HO
3
2
4a
6
1'
3'
5'
Br
7'
NH
13a
N
8'
2'
9
1a
4'
6'
8
1
H
NaH
8a
OH
OH
OH
OH
13
geranyl bromide
12a
DMF
10
12
11
(±)-15 (yield 2.0%)
(±)-14
7'-Me
3'-Me
4
5
X
HO
HO
3
4a
6
1'
5'
3'
7'
13a
N
8'
2'
1a
2
4'
6'
8
1
H
8a
13
OH
OH
9
12a
10
12
11
(±)-16 (yield 9.5%)
7'-Me
3'-Me
4
5
X
HO
HO
HO
HO
3
2
4a
6
1'
5'
3'
Br
7'
13a
N
8'
NH
2'
1a
4'
6'
8
1
H
8a
NaH
geranyl bromide
13
9
12a
10
DMF
12
OH
OH
11
(±)-17
OH
OH
(±)-18 (yield 21.9%)
Scheme 3. N-geranylation of tetrahydroprotoberberines
1
The H NMR spectrum displays two aromatic proton singlets at 6.68 and 6.74, a methine proton at
(q, J = 6.5 Hz), two methylene groups [3.77 (1H), 3.55 (1H), and 3.08 (2H)], two N-methyl
proton signals [3.17 (3H) and 3.14 (3H)], and a methyl proton signal [1.69 (3H, d, J = 6.5 Hz)],
characteristic of the isoquinoline moiety, and three methyl proton signals at 1.75 (3H, s), 1.65 (3H, s),
and1.59 (3H, s), two olefinic protons at 5.48 (1H, t, J = 6.5 Hz) and 5.09 (1H, t, J = 6.5 Hz), and three
methylene groups [4.62 (2H, d, J = 6.5), 2.12 (2H, m), and 2.09 (2H, m)], assignable to the geranyl
group. The olefinic protons at 5.48 and 5.09 show COSY coupling to the methylene protons at 4.62
and 2.12, respectively, which led to the assignment of H-2’ (5.48) and H-6’ (5.09), respectively. The
methylene protons at 4.62 and 2.12 were assigned to H2-1’ and H2-5’, respectively.

1198
HETEROCYCLES, Vol. 81, No. 5, 2010
a1 (2)
b1 (3)
c2 (8)
I
a3 (10)
II
III
b3 (11)
b2 (7)
a2 (6)
min
min
30
10
20
25
15
35
5
min
25
10
15
20
10
20
30
15
25
Figure 1. Analytical scale HPLC chromatograms of geranylated products of I: salsolinol (1), II: N, N-dimethylsalsolinol (5), and III: tetrahydropapaveroline (9)
Column: Cosmosil 5C18-ARII (6.0 x 150 mm); Eluent: A:0.1M NH4OAc (0.05% TFA), B:MeOH (0.05% TFA), I and II [ A/B: 80/20
(initial) 0/100 (30 min)]; III [ A/B: 75/25 (initial) 50/50 (10 min) 0/100 (30 min)]; Flow rate: 1 mL/min; Detector: UV(280 nm)
Optimized geometry of N-geranylsalsolinol (2) Optimized geometry of N,N-digeranylsalsolinol (3)
E=-986.0139433 Ha (BLYP/DNP)
E=-1376.8807811 Ha (BLYP/DNP)
The NOE correlations between an aromatic proton at 6.74 and a proton at 4.57 (H-1) indicated that
this aromatic proton was assigned as H-8. The aromatic proton at 6.68, which was correlated with
the methylene protons at 3.08 (H2-4) was assigned as H-5. NOE correlations were also observed
between the aromatic proton at 6.74 (H-8) and methylene protons at 4.62 (H-1’), indicating that a
geranyl group was attached to oxygen at C-7. Based on the above evidence, compound 6 was indicated
to be 7-O-geranyl-N, N-dimethylsalsolinol (Scheme 1). Moreover, the 13C NMR spectrum of 6 supports
this assignment, with signals corresponding to 22 carbons, including two protonated olefinic carbons
(124.90; 120.94), two protonated aromatic carbons (116.01; 113.25), six quaternary carbons
(148.79147.83; 142.30; 132.63; 124.39; 122.17), five methylene carbons (67.06; 56.86; 40.61;
27.41; 24.30), a methine carbon (69.51), two N-methyl carbons (51.50; 50.11) and four methyl
carbons (25.84; 18.63; 17.72; 16.69). The carbon signal at 147.83 (C-7) displays HMBC
correlations with the proton signal at 4.62 (H-1’), confirming 7-O-geranylation, which was already
suggested by the NOE correlation between H-8 and H-1’. On the basis of these data, the structure of 6

HETEROCYCLES, Vol. 81, No. 5, 2010
1199
was established to be (±)-7-O-geranyl-N,N-dimethylsalsolinol.
The molecular formula of compound 7 was determined to be C22H34NO2 by analysis of its HRSIMS
1
([M]+, m/z 344.2603), which indicated the presence of one geranyl group. The H NMR spectrum is
almost the same with that of 6. The NOE correlations between the aromatic proton at 6.66 and 4.55
(H-1) and the aromatic proton at 6.79 and the methylene protons at 3.12 (H2-4) indicate that the
aromatic protons at 6.66 and 6.79 were assigned to H-8 and H-5, respectively. NOE correlations were
also observed between the aromatic proton at 6.79 (H-5) and methylene protons at 4.63 (H-1’),
indicating that a geranyl group is attached to oxygen at C-6. Based on the above evidence, compound 7
was found to be 6-O-geranyl-N,N-dimethylsalsolinol (Scheme 1). Moreover, the 13C NMR spectrum of 7
supports this structure, with signals corresponding to 22 carbons. The carbon signal at 148.62 (C-6)
display HMBC correlations with the proton signal at 4.63 (H-1’), confirming 6-O-geranylation, which
was suggested by the NOE correlation between H-5 and H-1’. On the basis of these data, the structure of
7 was confirmed to be (±)-6-O-geranyl-N,N-dimethylsalsolinol.
The molecular formula of compound 8 was determined to be C32H50NO2 by analysis of its HRSIMS
1
([M]+, m/z 480.3858), which indicated the presence of two geranyl groups. The H NMR spectrum
displays two aromatic protons as singlets at 6.83 and 6.79, a methine proton at (q, J = 6.5 Hz),
two methylene groups [3.79 (1H), 3.57 (1H), and 3.14 (2H)], two N-methyl groups [3.17 (3H) and
3.15 (3H)], and a methyl proton signal [1.69 (3H, d, J = 6.5 Hz)], characteristic of an isoquinoline
moiety, and six methyl proton signals at 1.66 (6H, s), 1.60 (6H, s), 1.736 (3H, s), and 1.742 (3H, s), four
olefinic protons at 5.48 (2H) and 5.10 (2H), and six methylene groups [4.58 (2H), 4.59 (2H), 2.12 (4H,
m), and 2.07 (4H)], assignable to two geranyl groups. The aromatic proton at 6.83 shows NOE
correlations with both methylene protons at 3.14 (H-4) and 4.58 (or 4.59, H-1’). This aromatic proton
was assigned H-5. Therefore, the geranyl group is attached to the oxygen at C-6. The aromatic proton
at 6.79 displays NOE correlations with protons at 4.54 (H-1), 4.59 (or 4.58, H-1’’), and 1.69 (1-Me)
indicating that this aromatic proton was H-8. This indicated that the other geranyl group is attached to the
oxygen at C-7. From these data, the structure of 8 was deduced to be 6, 7-O-digeranyl-N,N-
dimethylsalsolinol (Scheme 1). The 13C NMR spectrum of 8 supports this structure.
Assignments of 1H and 13C signals of 6-8 were made by 1D and 2D (1H-1H COSY, NOESY, HSQC, and
HMBC) spectroscopic data. The geometry of the disubstituted olefinic bonds (between 2’ and 3’ or 2’’
and 3’’) was determined to be E on the basis of the NOE correlation between H-2’ and H-4’or H-2’’ and
H-4’’, respectively. The NOE correlations between H-1 and both quasi-equatorial and quasi-axial
N-methyl groups indicate the quasi-equatorial orientation of H-1, that is, quasi-axial orientation of the
methyl group at C-1. These stereostructures were identical with the optimized geometry of 6-8

1200
HETEROCYCLES, Vol. 81, No. 5, 2010
calculated.
HPLC of the reaction mixture in geranylation (Scheme 2) of (±)-tetrahydropapaveroline (9) is shown in
Figure 1 (III). The compounds corresponding to the peaks a3 and b3 (Figure 1, III) (10 and 11) were
purified by prep. HPLC of the reaction mixture. The molecular formula of compound 10 (peak a3)
was determined to be C26H34NO4 by analysis of its HRSIMS ([M + H]+, m/z 424.2494), which indicates
1
the presence of one geranyl group. The H NMR spectrum displays two aromatic proton singlets at
6.65 and 6.18, a methine proton at (1H, t, J = 6.5 Hz), two methylene groups [3.68 (1H), 3.38
(1H), and 2.97 (2H)] arisen from the isoquinoline moiety, three aromatic protons at 6.75 (1H, d, J = 8.5
Hz), 6.50 (1H, brd, J = 8.5 Hz), and 6.58 (1H, brs), methylene protons [3.04 (1H) and 3.14 (1H)]
characteristic of a benzyl group, and three methyl proton signals at 1.70 (3H, s), 1.64 (3H, s), and1.57
(3H, s), two olefinic protons at 5.26 (1H, m) and 5.12 (1H, m), and three methylene groups [3.83 (1H),
3.71 (1H), and 2.18 (4H)], assignable to a geranyl group. The NOE correlations between the
methylene protons (H-1’’) of the geranyl group and both the methine proton (H-1) of the isoquinoline
structure and the methylene protons (H-9) of the benzyl group indicate that the geranyl group was
attached to nitrogen. Based on the these evidences, compound 10 was established to be (±)-
N-geranyltetrahydropapaveroline (Scheme 2). The 13C NMR spectrum of 10 supports this structure, with
signals corresponding to 24 carbons (two carbons overlap), including two protonated olefinic carbons
(124.56; 114.15), five protonated aromatic carbons (116.12; 115.75), nine
quaternary carbons (147.30146.92; 146.04; 145.65; 133.29; 121.93; 121.87x2), six methylene
carbons (51.86; 44.80; 40.74x2; 27.14; 22.90), a methine carbon (63.01), and three methyl carbons
1
(25.94; 17.84; 16.80). Assignments of H and 13C signals of 10 were made by 1D and 2D (1H-1H
COSY, NOESY, HSQC, and HMBC) spectroscopic data. The geometry of the disubstituted olefinic bond
(between 2’’ and 3’’) was determined to be E on the basis of the NOE correlation between H-2’’ and
H-4’’. The geranyl group may be a quasi-equatorial orientation, because a methylene proton (H-1’’) of
the geranyl group displayed NOE correlations with both the methine proton (H-1) of the isoquinoline
structure and the methylene protons (H-9) of the benzyl group. Therefore, the benzyl group is
considered to be quasi-axial orientation. The upfield shift (+ 3.6 ppm) at C-3 in 10 compared with
N-methylsalsolinol (4) is interpreted as being due to larger steric compression by the quasi-axial benzyl
group at C-1 in 10 than that by the quasi-axial methyl group at C-1 in 4.
The molecular formula of compound 11 (peak b3) was determined to be C36H50NO4 by analysis of its
1
HRSIMS ([M]+, m/z 560.3746), which indicates the presence of two geranyl groups. The H NMR and
13C NMR spectra of 11 display the signals due to the 1-benzyltetrahydroisoquinoline structure and two
sets of signals due to two geranyl groups. The protons at 3.78 (H-1”) and 4.13 (H-1”’) show HMBC

HETEROCYCLES, Vol. 81, No. 5, 2010
1201
correlations with carbons at 58.30 (C-1’’’) and 55.79 (C-1”), respectively. Furthermore, HMBC
correlations were observed from protons at 3.96 (H-1’’) and 4.13 (H-1’’’) to the carbon at 52.01
(C-3). The proton at 4.20 (H-1’’’) also shows a correlation with the carbon at 70.72 (C-1). It was
concluded that both geranyl groups are attached to nitrogen on the basis of these HMBC correlations, as
well as by NOE correlations between H-1 (4.42) and H-1’’ (3.96), H-3 (3.53) and H-1’’’ (4.20),
13
and H-3 (3.68) and H-1’’’ (4.13 and 4.20), respectively. Assignments of 1H and C signals of 11
were made by 1D and 2D (1H-1H COSY, NOESY, HSQC, and HMBC) spectroscopic data. Based on
these evidences, compound 11 was determined to be (±)-N,N-digeranyltetrahydropapaveroline (Scheme
2). The geometries of the disubstituted olefinic bonds (between 2’’ and 3’’ and between 2”’ and 3’’’)
were determined to be E on the basis of the NOE correlations between H-2’’ and H-4’’, and H-2”’ and
H-4”’, respectively. NOE correlations were observed between the proton at 4.20 (H-1’’’) and H2-3 at
3.68 and 3.53 (quasi-axial and quasi-equatorial protons), indicating that C-1’’’ of the geranyl group is a
quasi-equatorial orientation. The quasi-equatorial orientation of the proton at C-1 was suggested by the
NOE correlations between the methine proton at 4.42 (H-1) and the proton (3.96) of C-1’’ of the
quasi-axial geranyl group. Therefore, the orientation of the benzyl group is quasi-axial. This was
suggested from the fact that the upfield shift (+ 4.1 ppm) at C-3 in 11 compared with N,
N-dimethylsalsolinol (5) arises from larger steric compression by the quasi-axial benzyl group at C-1 in
11 than that by the quasi-axial methyl group at C-1 in 5.
HPLC of the reaction mixture in geranylation (Scheme 2) of N-methylpapaveroline (12) is shown in
Figure 2 (I). Some compounds (peaks a4, b4, and d4) in the reaction mixture changed during prep.
c4 (13)
b4
b
5 (16)
I
III
II
b6 (18)
a4
a6
d4
a5 (15)
min
20
10
15
min
10
5
min
25
5
15
20
25
30
10
15
20
Figure 2. Analytical scale HPLC chromatograms of products obtained by geranylation of I: N-methylpapaveroline (12),
II: 2, 3, 9, 10-tetrademethyltetrahydropalmatine (14), and III: 2, 3, 10, 11-tetrademethylpseudotetrahydropalmatine (17)
Column: Cosmosil 5C18-ARII (6.0 x 150 mm); Eluent: A:0.1M NH4OAc (0.05% TFA), B:MeOH (0.05% TFA), I [A/B: 75/25
(initial) 50/50 (10 min) 0/100 (30 min)], II and III [A/B: 80/20 (initial) 0/100 (20 min); Flow rate: 1 mL/min; Detector: UV(280 nm)

1202
HETEROCYCLES, Vol. 81, No. 5, 2010
HPLC and were not purified. Only peak c4 was purified by prep. HPLC. The molecular formula of
compound 13 (peak c4) was determined to be C37H48NO4 by analysis of its HRSIMS ([M]+, m/z
1
570.3592), which indicates the presence of two geranyl groups. The H NMR spectrum displays two
aromatic protons as doublets at 8.32 and 8.08 (J = 7.0 Hz), two aromatic protons as singlets at 7.70
and 7.60 originating in the isoquinoline moiety, three aromatic protons at 6.71 (1H, d, J = 8.5 Hz),
6.36 (1H, dd, J = 8.5, 1.5 Hz), and 6.49 (1H, d, 1.5 Hz), a methylene proton at 4.83 (2H) characteristic
of a benzyl group, six methyl proton signals at 1.83 (3H, s), 1.70 (3H, s), 1.61 (3H, s), 1.59 (3H, s),
1.58 (3H, s), and 1.55 (3H, s), four olefinic protons at 5.58 (1H), 5.39 (1H), 5.09 (1H), and 5.02 (1H),
and six methylene groups [4.90 (2H), 4.80 (2H), 2.15 (2H), 2.13 (2H), 2.05 (2H), and 1.98 (2H)],
assignable to two geranyl groups. The aromatic proton at 7.60 shows NOE correlations with both
protons at 8.08 (H-4) and 4.90 (H-1’’). Thus, the proton at 7.60 was assigned to H-5. Therefore,
the geranyl group is attached to the oxygen at C-6. The aromatic proton at .70 displays NOE
correlations with protons at  4.80 (H-1’’’), indicating that this proton is H-8. This indicates that the
other geranyl group is attached to the oxygen at C-7. Based on these evidences, compound 13 was
13
identified as O,O-digeranylpapaveroline (Scheme 2). Moreover, the C NMR spectrum of 13 shows
four protonated olefinic carbons (124.91; 124.84; 119.65; 119.32), seven protonated aromatic carbons
(108.15; 108.08), 12 quaternary carbons (
147.50146.17; 144.48; 143.93; 137.50; 132.81; 132.63; 126.47; 125.62), seven methylene
carbons (67.71; 67.33; 40.63; 40.46; 34.96; 27.30; 27.29), one N-methyl carbon (46.28), and six
methyl carbons (25.91; 25.87; 17.82; 17.81; 16.88; 16.84). The carbon signal at 158.86 (C-6)
displayed HMBC correlations with the proton signal at 4.90 (H-1’’), confirming 6-O-geranylation,
which was also suggested by the NOE correlation between H-5 and H-1’’. The carbon signal at
154.18 (C-7) displayed HMBC correlations with the proton signal at 4.80 (H-1’’’), confirming
7-O-geranylation, which was corroborated by the NOE correlation between H-8 and H-1’’’.
Assignments of 1H and 13C signals of 13 were made by 1D and 2D (1H-1H COSY, NOESY, HSQC, and
HMBC) spectroscopic data. On the basis of this evidence, the structure of 13 was established to be O,
O-digeranylpapaveroline (Scheme 2). The geometries of the disubstituted olefinic bonds (between 2’’
and 3’’ and between 2’’’ and 3’’’) were determined to be E on the basis of the NOE correlation between
H-2’’ and H-4’’, and between 2’’’ and 3’’’, respectively.
HPLC of the reaction mixture in geranylation (Scheme 3) of 2, 3, 9, 10-tetrademethyltetrahydropalmatine
(14) is shown in Figure 2 (II). The compounds corresponding to peaks a5 and b5 (Figure 2, II) (15 and
16) were purified by prep. HPLC. The molecular formula of compound 15 (peak a5) was determined to
be C27H34NO4 by analysis of its HRSIMS ([M]+, m/z 436.2486), which indicates the presence of one

HETEROCYCLES, Vol. 81, No. 5, 2010
1203
1
geranyl group. The H NMR spectrum displays two aromatic proton singlets at 6.72 and 6.66, two
aromatic proton doublets at 6.78 (1H, d, J = 8.0 Hz), 6.57 (1H, d, J = 8.0 Hz), methine and methylene
protons at (3H, brs), three methylene groups [3.81 (1H), 3.54 (1H), 3.28 (1H), 3.18 (2H), 3.12
(1H)] arising from the protoberberine moiety and three methyl proton signals at 1.65 (3H, s), 1.59 (3H,
s), and1.58 (3H, s), two olefinic protons at 5.44 (1H, t, J = 8.0 Hz) and 5.08 (1H, brs), and three
methylene groups [4.08 (2H), 2.22 (2H), and2.20 (2H)], assignable to the geranyl group. The NOE
correlations between a methylene proton (4.08, H-1’) of the geranyl group and the protons at 
(H-8 and/or H-13a) indicate that the geranyl group is attached to nitrogen. Based on these evidences,
compound 15 was determined to be N-geranyl-2, 3, 9, 10-tetrademethyltetrahydropalmatinium salt. The
13C NMR spectrum of 15 supports this structure, with signals corresponding to 26 (one carbon was
overlapped) carbons including four protonated aromatic carbons (x2), eight
quaternary carbons (143.60124.76; 121.90; 120.93; 114.80), four methylene
carbons (57.06; 51.66; 34.89; 23.75), and a methine carbon (64.34) which arise from the
tetrahydroprotoberberine structure, and two protonated olefinic carbons (124.12; 111.74), two
quaternary carbons (133.47), three methylene carbons (59.93; 41.04; 26.97), three methyl
groups () assignable to the geranyl group. Assignments of 1H and 13C signals of 15
were made by 1D and 2D (1H-1H COSY, NOESY, HSQC, and HMBC) spectroscopic data. On the basis
of these data, the structure of 15 was established to be (±)-N-geranyl-2,3,9,10-tetrademethyl-
tetrahydropalmatinium salt with the B/C-cis form as described later (Scheme 3).
The molecular formula of compound 16 (peak b5) was determined to be C27H34NO4 by analysis of its
1
HRSIMS ([M]+, m/z 436.2484), which indicates the presence of one geranyl group. The H NMR
spectrum displays two aromatic proton singlets at 6.84 and 6.71, two aromatic proton doublets at 6.85
and 6.77 (J = 8.2 Hz), a methine proton at (1H, dd, J = 12.5, 5.0 Hz), and four methylene groups
[(1H, d, J = 15.5 Hz), (1H, d, J = 15.5 Hz), 3.89 (1H), 3.83 (1H), 3.69 (1H), 3.26 (1H), 3.20
(1H), and 3.05 (1H)] which arise from the tetrahydroprotoberberine structure, and three methyl proton
signals at 1.72 (3H, s), 1.66 (3H, s), and1.33 (3H, s), two olefinic protons at 5.44 (1H, dd, J = 9.0, 5.5
Hz) and 5.14 (1H, brs), and three methylene groups [3.86 (1H, m), 3.61 (1H, m), and 2.20 (4H, m)],
assignable to the geranyl group. The NOE correlation between one of the methylene protons (,
H-1’) of the geranyl group and one of the methylene protons ( H-8) indicates that the geranyl
group is attached to nitrogen. Based on these evidences, compound 16 was postulated to be the isomer of
N-geranyl-2,3,9,10-tetrademethyltetrahydropalmatininium salt (15). Moreover, the 13C NMR spectrum
of 16 supports this structure, with signals corresponding to 27 carbons, including four protonated

1204
HETEROCYCLES, Vol. 81, No. 5, 2010
aromatic carbons (116.15), eight quaternary carbons (
143.66122.77; 122.43; 122.06; 115.38), four methylene carbons (58.43; 57.62; 29.60; 24.48), and a
methine carbon (68.21) which arise from the tetrahydroprotoberberine structure, and two protonated
olefinic carbons (124.71; 114.42), two quaternary carbons (133.45), three methylene carbons
(46.96; 41.14; 27.10) and three methyl groups () assignable to the geranyl group.
The proton signals (H-1’) at 3.86 and 3.61 display HMBC correlations with the carbons at 57.62 (C-8)
and 58.43 (C-6), respectively, confirming N-geranylation, which was also suggested by the NOE
1
13
correlation between H-8 and H-1’. Assignments of H and C signals of 16 were made by 1D and 2D
(1H-1H COSY, NOESY, HSQC, and HMBC) spectroscopic data. On the basis of these evidences, the
structure of 16 was confirmed to be (±)-N-geranyl-2,3,9,10-tetrademethyltetrahydropalmatininium salt
with the B/C-trans form as described below (Scheme 3).
Compounds 15 and 16 are stereoisomers of the B/C ring juncture. The chemical shifts of C-6 in 15 and 16
are 51.66 and 58.43 ppm, respectively. The upfield shift of C-6 in 15 compared with 16 is interpreted as
being due to steric compression in the B/C-cis form because C (6)-H and C (13a)-C (13) bonds are 1,
3-diaxial in the B/C-cis form. The chemical shifts at C-13 in 15 and 16 are 34.89 and 29.60 ppm,
respectively. The upfield shift of C-13 in 16 compared with 15 arises from steric interaction
(1,3-diaxial) between C (13)-H and the N-(C-1’) bonds in the B/C-trans form. Compounds 15 and 16
were established to be (±)-cis- and trans-N-geranyl-2,3,9,10-tetrademethyltetrahydropalmatininium salts
(Scheme 3). The geometry of the disubstituted olefinic bond (between 2’ and 3’) was determined to be
E in both compounds on the basis of the NOE correlation between H-2’ and H-4’.
The HPLC of the reaction mixture in geranylation (Scheme 3) of 2,3,10,11-tetrademethylpseudo-
tetrahydropalmatine (17) is shown in Figure 2 (III). Compound 18, which corresponds to peak b6
(Figure 2, III), was purified by prep. HPLC, although a6 was not purified. The molecular formula of
compound 18 was determined to be C27H34NO4 by analysis of its HRSIMS ([M]+, m/z 436.2487), which
1
indicates the presence of one geranyl group. The H NMR spectrum displays four aromatic proton
singlets at 6.83, 6.81, 6.70, 6.55, a methine proton at and four methylene groups [(1H),
4.34 (1H), 3.85 (1H), 3.78 (1H), 3.63 (1H), 3.25 (1H), 3.15 (1H), and 3.04 (1H)] which arise from the
protoberberine skeleton and three methyl proton signals at 1.73 (3H, s), 1.67 (3H, s), and1.32 (3H, s),
two olefinic protons at 5.44 (1H) and 5.12 (1H), and three methylene groups [3.87 (1H), 3.64 (1H),
1
and 2.20 (4H)], assignable to the geranyl group. From a comparison of the H NMR spectrum with
that of compounds 15 and 16, compound 18 was postulated to be N-geranyl-2,3,10,11-tetrademethyl-
13
pseudotetrahydropalmatinium salt with a B/C-cis or B/C-trans junction. The C NMR spectrum of 18
displays signals corresponding to 27 carbons, including four protonated aromatic carbons

HETEROCYCLES, Vol. 81, No. 5, 2010
1205
( 114.17), eight quaternary carbons (146.63122.72;
122.53; 122.05; 118.26), four methylene carbons (61.61; 58.27; 29.83; 24.38), a methine carbon
(68.55) arising from the tetrahydroprotoberberine structure, and two protonated olefinic carbons
(124.86; 111.41), two quaternary carbons (133.27), three methylene carbons (46.69; 41.01;
27.04), and three methyl groups () assignable to the geranyl group. Compound 18
was identified as (±)-trans-N-geranyl-2,3,10,11-tetrademethylpseudotetrahydropalmatinium salt (Scheme
3) from comparison of the chemical shifts of C-6 (58.27) and C-13a (68.55) with those (51.66 and
64.34) in 15 (B/C-cis) and those (58.43 and 68.21) in 16 (B/C-trans). The geometry of the
disubstituted olefinic bond (between 2’ and 3’) was determined to be E on the basis of the NOE
correlation between H-2’ and H-4’. Assignments of 1H and 13C signals of 18 were made by 1D and 2D
(1H-1H COSY, NOESY, HSQC, and HMBC) spectroscopic data.
Antimicrobial activity.
The synthetic simple isoquinolines [salsolinol (1), N-geranylsalsolinol (2), N,N-digeranylsalsolinol (3),
N-methylsalsolinol (4), N,N-dimethylsalsolinol (5), 7-O-geranylsalsolinol (6), 6-O-geranylsalsolinol (7),
and 6,7-O-digeranylsalsolinol (8)], the synthetic 1-benzylisoquinolines [tetrahydropapaveroline (9),
N-geranyl- and N,N-digeranyltetrahydropapaveroline (10 and 11), N-methylpapaveroline (12), and O,
O-digeranyl-N-methylpapaveroline (13)], and the synthetic tetrahydroprotoberberines [2,3,9,10-
tetrademethyltetrahydropalmatine (14), cis- and trans-N-geranyl-2,3,9,10-tetrademethyltetrahydro-
palmatinium salts (15 and 16), 2,3,10,11-tetrademethylpseudotetrahydropalmatine (17), and
trans-N-geranyl-2,3,10,11-tetrademethylpseudotetrahydropalmatininium salt (18)] were tested against
Staphylococcus aureus (Gram-positive) and Escherichia coli (Gram-negative) by the liquid dilution
method. The minimum inhibitory concentrations (MIC) are presented in Table 1. The results for 1-5
have already been published.13
N-Geranylation increases the activity against both bacteria compared with salsolinol (1),
N-methylsalsolinol (4), and N,N-dimethylsalsolinol (5), which are inactive. Notably, N,N-digeranylated
salsolinol (3) displays significant activity (7.8 g/mL) against S. aureus. Thus, according to these data,
N-quaternization by N-geranylation (3), not simply N-alkylation (5), appears to be important for enhanced
antimicrobial activity.

1206
HETEROCYCLES, Vol. 81, No. 5, 2010
Table 1. Antibacterial activity of several 1soquinoline-type alkaloids 1-18
MIC (g/mL)a
Staphylococcus
aureus
Escherichia
coli
group
compound
ATCC25923
ATCC25922
1b
2b
3b
4b
5b
6
1000
1000
＞
＞
62.5
1000
≧
≧
7.8
31.3
≧
≧
1000
1000
＞
＞
simple isoquinolines
1000
1000
＞
＞
62.5
62.5
250
250
250
250
250
250
250
250
250
250
250
250
250
250
＞
＞
＞
＞
＞
＞
＞
＞
＞
＞
≧
＞
＞
7
8
250
250
125
250
125
250
250
250
250
250
9
＞
≧
≧
＞
≧
＞
≧
≧
＞
＞
10
11
12
13
14
15
16
17
18
1-benzylisoquinolines
protoberberines
Benzalkonium chlloride
Benzethonium chloride
3.9
7.8
15.7
7.8

HETEROCYCLES, Vol. 81, No. 5, 2010
1207
Enhancement of activity against S. aureus was observed in O-gerany isoquinolines (6-8). Enhanced
activity was not observed in N- or O-geranylation of 1-benzylisoquinolines and tetrahydroproto-
berberines.
Cytotoxicity evaluation.
The synthetic simple isoquinolines (1-8), 1-benzylisoquinolines (9-13), and tetrahydroprotoberberines
(14-18) were assayed for in vitro cytotoxicity against five human tumor cell lines, including lung
carcinoma (A-549), prostate carcinoma (DU145), epidermoid carcinoma of the nasopharynx (KB), a
drug-resistant KB-subline (KBvin), and human promyelocytic leukemia (HL-60). The cytotoxicity data
are given as an ED50 value for each cell line, the concentration of a compound that caused a 50%
reduction in absorbance at 562 nm relative to untreated cells using SRB10 and MTT16 and/or WST-8
assays (HL60 is a non-adherent cell line; therefore, the SRB assay could not be used with it), and are
shown in Tables 2 and 3. The results for 1-5 have already been published.13
The parent salsolinol (1), N-methylsalsolinol (4), and N,N-dimethylsalsolinol (5) showed no activity in
SRB and MTT assays, and N-geranylsalsolinol (2) showed either weak or no activity against all cell lines
in both assays. In comparison, N,N-digeranylsalsolinol (3) showed especially increased activity against
DU-145, KB, and HL-60 cell lines. Compound 3 exhibited the highest potency (1.2 g/mL) against the
HL-60 cell line, and also displayed high activity (0.77 g/mL) against this cell line in the WST-8 assay
(Table 3).17 Thus, N-geranylation, particularly digeranylation, increased cytotoxicity, while
N-methylation, either mono or di, had no effect. The increase in cytotoxicity was found in O- and
O,O-digeranyl compounds (6-8) compared with the parent N,N-dimethylsalsolinol (5) in the SRB assay,
though activity of O,O-digeranyl compounds (8) was weaker than that of N,N-digeranylsalsolinol (3).
Thus, from these data, N-quaternization by N-geranylation, not simply N-alkylation, appears to be
important for enhanced cytotoxicity.
In 1-benzyltetrahydroisoquinoline, N- and N,N-digeranylations of tetrahydropapaveroline (9) increased
remarkably the activity against three or four cell lines tested in the SRB assay. Compounds 10 and 11
showed a broad spectrum of activity. It was demonstrated that N-geranylation in the 1-benzyltetrahydro-
isoquinoline-type alkaloids contributes especially to the increase in cytotoxicity. The increase in
cytotoxicity by N,N-digeranylations (11) compared with N-geranylation (10) was not observed to be
distinct from that of the simple isoquinolines.
An increase in cytotoxicity was found in the O,O-digeranyl derivative (13) of papaveroline (12), except
for the KBvin cell line in SRB and MTT assays. Compound 13 exhibited high potency (1.49 g/mL)
against the HL-60 cell line, and also displayed high activity (1.24 g/mL) against this cell line in the
WST-8 assay.17

1208
HETEROCYCLES, Vol. 81, No. 5, 2010
Table 2. In vitro cytotoxic activity of several isoquinoline-type alkaloids 1-18 against various human
tumor cell lines
——————————————————————————————————————
cell lineb
groupa
compound
A-549
DU-145 KB
KBvin
HL-60
——————————————————————————————————————
ED50 (g/mL)c-SRB
—————————————————————————————
1d
2d
3d
4d
5d
NAe
7.75
5.72
> 20
NA
NA
NA
NDe
ND
ND
ND
7.75
3.84
> 20
10.1
3.28
> 20
5.64
NA
> 20
I
> 20
> 20
7.12
8.90
> 20
12.4
13.9
11.5
> 20
12.8
7.70
7.60
> 20
> 20
> 20
> 20
ND
ND
ND
ND
6
7
8
9
10
11
12
13
NA
NA
NA
NA
ND
ND
ND
ND
ND
1.31
1.34
11.95
6.25
3.56
2.00
7.46
5.89
3.38
2.35
8.05
5.37
3.60
7.96
5.07
NA
II
14
15
16
17
18
6.05
7.40
3.63
NA
2.88
3.29
3.26
NA
5.40
5.87
5.70
14.6
2.90
1.45
3.19
1.33
NA
ND
ND
ND
ND
ND
III
9.68
2.31
NA
Taxolf
2.91
1.91
3.10
> 850
ND
ED50 (g/mL)c –MTTg
————————————————————————————
1d
2d
NA
NA
NA
NA
NA
I
NA
NA
NA
NA
NA

HETEROCYCLES, Vol. 81, No. 5, 2010
1209
3d
15.8
3.28
5.70
NA
1.20
II
12
NA
NA
5.26
NA
NA
NA
12.3
1.49
13
7.50
6.43
———————————————————————————————————————
aI: simple isoquinolines, II: 1-benzylisoquinolines, III: protoberberines
bA-549: lung carcinoma, DU-145: prostate carcinoma , KB: epidermoid carcinoma of the nasopharynx,
KBvin: drug-resistant, HL-60: human promyelocytic leukemia.
cCytotoxicity as ED50 for each cell line, the concentration of compound that causes a 50% reduction in
adsorbance at 562 nm relative to unreated cells using the SRB or MTT assay. Pure compound is
considered to be significantly active when its ED50 < 4.0 g/mL.
dThese data have already been prepared.13 eNA: no activity (if it does not have 50% inhibition at 20
g/mL, we suggest it has no activity); ND: not determined; fng/mL; gDifferent time of treatment (because
of long doubling time in HL-60): 24 hr for A-549, DU-145, KB, and KBvin; 72 hr for HL-60 in MTT
assay
In 2,3,9,10-tetrademethyltetrahydropalmatine (14), an increase in cytotoxicity by N-geranylation (15, 16)
of the stereostructure was not observed in almost all cell lines, while in the N-geranyl derivative (18) with
a B/C-trans junction of 2,3,10,11-tetrademethylpseudotetrahydropalmatine (17), the activity increased,
except for the KBvin cell line. Contrary to this, compounds 15 and 16 displayed higher activity than 18
against the HL-60 cell line.
Comparing the activity between the compounds and the cell lines, N- and N,N-geranyl derivatives (10 and
11) of the 1-benzyltetrahydroisoquinolines and N-geranyl compound (16, tetrahydroprotoberberine) with
a B/C-trans junction showed high potency (1.31-3.63 g/mL) in three or four cell lines. The N,N- and
N-geranylated compounds (3, simple isoquinoline and 18, tetrahydroprotoberberine) displayed high
activity (2.31-3.84 g/mL) in DU145 and KB cell lines. 2,3,9,10-Tetrademethyltetrahydroproto-
berberine (14-16) exhibited high potency (1.33-3.19 g/mL) in the KBvin cell line, independent of the
presence of the geranyl group. In both MTT and WST-8 assays, only two N,N- and O,O-geranylated
compounds (3, simple isoquinoline and 13, 1-benzylisoquinoline) displayed high activity in the HL-60
cell line. It was suggested that the structures of the test samples had a strong effect on the HL-60 cell
line.

1210
HETEROCYCLES, Vol. 81, No. 5, 2010
Table 3. In vitro cytotoxic activity of several isoquinolines-type alkaloids1-4 and 9-18 against
HL-60
group
compound
IC50 value (g/mL)-WST-8
1
7.17
7.33
0.77
5.09
6.99
4.78
7.32
3.68
1.24
＞20
8.05
12.38
6.15
16.3
2
simple isoquinolines
3
4
9
10
11
12
13
14
15
16
17
18
1-benzylisoquinolines
protoberberines
Inhibitory effects on EBV-EA induction.
The Epstein-Barr virus early antigen (EBV-EA) activation assay is considered to be an effective
indicator for the evaluation of anti-tumor-promoting activity.11 The inhibitory effects of simple
isoquinolines (1-8), 1-benzylisoquinolines (9-13), and tetrahydroprotoberberines (14-18) on EBV-EA
activation induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) in Raji cells were examined as a
primary screening of anti-tumor-promoting activity.

HETEROCYCLES, Vol. 81, No. 5, 2010
1211
Table 4. Inhibitory effects of several isoquinoline-type alkaloids 1–18 on
TPA Induced EBV-EA activation (100%)a
concentration (mol ratio/32 pmol TPA)
b
group
compound
1000
500
100
10
IC50
% to control (% viability)
1c
2c
3c
4c
5c
6
12.0 (60)
3.1 (70)
0 (70)
37.6
24.3
20.6
36.8
35.1
36.8
36.1
32.6
79.0
71.6
69.3
78.8
78.0
69.8
69.0
67.3
100
384
350
296
372
369
371
369
335
97.4
91.7
100
simple isoquinolines
9.3 (60)
8.7 (60)
8.9 (60)
8.0 (60)
5.3 (60)
100
98.7
98.0
96.5
7
8
9
13.9 (60)
11.8 (60)
8.9 (60)
15.0 (60)
9.1 (60)
37.5
36.1
35.0
39.2
31.2
76.9
75.4
74.3
80.6
72.4
100
100
100
100
100
387
380
369
410
323
10
11
12
13
1-benzylisoquinolines
14
15
16
17
18
0 (60)
0 (60)
0 (60)
0 (60)
0 (60)
25.8
23.2
23.7
23.7
23.2
76.8
74.2
74.8
71.5
71.1
96.9
94.3
94.7
94.0
93.8
321
307
308
305
303
protoberberines
Ginsenoside-Rg1 0 (80)
-Carotene 9.1 (60)
32.5
34.3
72.6
82.7
91.0
100
310
400
aValues represent percentages relative to the positive control value. TPA (32 pmol, 20 ng) = 100%.
Values in parentheses are viability percentages of Raji cell. bIC50 represents the mol ratio to TPA
that inhibits 50% of positive control (100%) activated with 32 pmol of TPA.
cThese data have been published.13

1212
HETEROCYCLES, Vol. 81, No. 5, 2010
The inhibitory effects of the test compounds on TPA-induced EBV-EA activation, their effects on the
viability of Raji cells, and the 50% inhibitory concentration (IC50) values are shown in Table 4. The
results for 1-5 have already been published.13
All simple isoquinolines (1-8) displayed stronger inhibition (IC50 296-384) than that of the reference
-carotene (IC50 400), which has been studied extensively in cancer chemoprevention using animal
models.18 The inhibitory activity was more increased by geranylation than by methylation on nitrogen
(compared with 3 and 5). Therefore, N-geranylation but not N-dimethylation appears to be important for
enhanced activity. N,N-Digeranylsalsolinol (3) displayed the strongest inhibition (IC50 296), and its
activity was higher than that of ginsenoside-Rg1, which is known as a strong anti-tumor-promoter.19
The 6,7-O-digeranylated derivative (8) had increased activity (IC50 335) compared with the parent
compound (5), though 6- and 7-O-geranylation (6 and 7) scarcely affected the activity. Thus, these
compounds, especially N,N- and O,O-digeranylated derivatives, appear to be useful leads for further
development of potential cancer chemopreventive agents.
There was little increase in inhibition by N-, and N,N-geranylation (10 and 11) of tetrahydropapaveroline
(9). Increased inhibition (IC50 323) was observed in O, O-geranylation (13) of papaveroline (12). All
tetrahydroprotoberberines (14-18) displayed strong inhibition (IC50 303-321) comparable to
ginsenoside-Rg1 (IC50 310), independent of whether the compound has the geranyl group or not. They
might be valuable antitumor promoters. In 2, 3, 9, 10-tetrademethyltetrahydropalmatine (14), there was
little increase in inhibition by N-geranylation (15, 16) independent of the stereostructure.
In simple isoquinolines and 1-benzylisoquinolines, digeranylated derivatives (3, 8, and 13) displayed
potent activity, suggesting that relative lipophilicity of the geranyl groups may contribute to the inhibitory
effect.
Free radical scavenging activity.
Several human illnesses, such as cancer, diabetes, atherosclerosis, etc., can be linked to the damaging
action of reactive free radicals.20 The ability of three types of compounds (1–18) to scavenge
1,1-diphenyl-2-picrylhydrazyl (DPPH) free radicals was examined; the results are presented in Table 5.
The results for 1-3 have already been published13. To evaluate the free radical scavenging activity of these
compounds, the concentration required to scavenge DPPH free radicals by 50% (SC50) was determined.
The antioxidant -tocopherol was used as a reference compound. Compounds 1-3 displayed similar
potency and were more active than -tocopherol. 6-O- or 7-O-Geranylation (6 or 7) of 5 reduced the
activity and 6,7-O-digeranylation (8) further reduced the activity.

HETEROCYCLES, Vol. 81, No. 5, 2010
1213
Table 5. Radical scavenging activity of several isoquinoline-type alkaloids 1-18
SC50 (M)a
12.0
group
compound
1b
2b
3b
4
17.0
11.5
32.1
simple isoquinolines
5
31.3
6
40.7
7
41.1
8
57.7
9
2.64
10
7.81
1-benzylisoquinolines
11
12
13
6.47
2.95
9.94
14
5.53
8.39
3.81
6.04
5.36
24.3
15
protoberberines
16
17
18
α-tocopherol
aThe compound concentration showing radical scavenging efficacy of 50% was defined as SC50.
bThese data have already been prepared.13

1214
HETEROCYCLES, Vol. 81, No. 5, 2010
All 1-benzyltetrahydroisoquinolines (9-13) are more active than -tocopherol. N-Geranylation (10, 11) of
1-benzyltetrahydroisoquinoline (9) reduced the activity. The activity also decreased by
O,O-digeranylation (13) of papaveroline (12). All tetrahydroprotoberberines (14-18) proved to be more
active than -tocopherol; increased activity by N-geranylation was small.
In all three types of isoquinoline alkaloids, the activity was reduced by N- or O-geranylation. Obviously
the phenolic hydroxyl groups on the aromatic rings strongly influenced the activity, without an affect of
the geranyl group.
Antimalarial activity.
The simple isoquinolines (1-3), 1-benzylisoquinolines (12, 13), and tetrahydroprotoberberines (14, 16-18)
were tested in vitro against human malaria parasite, Plasmodium falciparum FCR-3. The antimalarial
activity of each compound was determined as a percentage of reduction compare of control. The
compound concentration required to inhibit cell growth by 50% was expressed as EC50. To evaluate the
toxicity of the compounds for mammalian cells, the concentration causing a 50% growth reduction (IC50)
of mouse mammary FM3A cells, a model of the host, was determined. The IC50/EC50 ratios for the
compounds were calculated as an evaluation of antimalarial activity. The results are presented in Table
6. In three types of isoquinoline derivatives, N-geranylation increased the inhibitory activity; however,
the geranylated derivatives of the simple isoquinolines and 1-benzylisoquinolines showed only slight
inhibitory activity and no selectivity. Only N-geranyl derivatives (16 and 18) of the
tetrahydroprotoberberines inhibited P. falciparum with EC50 values in the order of 10-7 M. This is an
increase antimalarial activity compared with the parentcompounds (14 and 17), and their selectivity
indexes were > 15. Compounds 16 and 18 were potent antimalarial agents with higher selectivity
indexes compared with the other test compounds.
In addition, N-geranyl derivatives, 16 and 18, it is noted that the selectivity of the compounds (selectivity;
>15) was comparable to that of mefloquine (selectivity; 90). It means that these compounds have safety
for human clinical treatment of malaria if these compounds developed new antiamalrial drug.

HETEROCYCLES, Vol. 81, No. 5, 2010
1215
Table 6. In vitro antimararial activity of several isoquinoline-type alkaloids 1-3, 12-14, and 16-18
50 % inhibitory concentration (M)a
selectivity
indexb
group
compound Plasmodium falciparum Mouse mammary cells
FCR-3 EC50 (growth %) FM3A IC50 (growth %)
IC50 / EC50
1
2
3
>27 (100 %)
9.0
-
-
-
simple isoquinolines
1-benzylisoquinolines
5.3
3.4
0.08
0.45
12
13
5.1
3.6
2.9
1.5
-
-
14
16
>17 (100 %)
0.33
>8.3 (73 %)
>4.9 (69 %)
-
>15
protoberberines
17
18
9.5
8.1
-
0.53
>7.7 (69 %)
>15
Mefloquine
0.032
2.9
91
aThe 50 % inhibitory concentration was defined by comparison with drug-free controls incubated
under same condition.
bIn vitro selectivity index was estimated from the ratio ( IC50 / EC50) of the drug concentrations
necessary to inhibit the growth rate of cells to 50 % of the growth value between the malaria
parasites
and mouse mammary FM3A cells which served as a model host.

1216
HETEROCYCLES, Vol. 81, No. 5, 2010
Table 7. Anti-HIV activity of several isoquinoline-type alkaloids 1-18
EC50
IC50(g/mL)a
＞25
group
compound
(g/mL)b
1c
＞2.5
＞2.5
＞2.5
＞2.5
＞25
2c
20.23
10.68
＞25
simple isoquinoline type
3c
4
5
＞25
19.69
24.59
18.02
＞25
9
＞2.5
＞2.5
＞2.5
＞2.5
＞2.5
＞2.5
＞2.5
＞25
10
11
12
13
14
15
16
17
18
AZTd
1-benzylisoquinoline type
17.95
＞25
19.84
＞25
protoberberine type
18.13
＞25
22.54
＞25
500
0.014
aThe agent concentration that inhibited H9 cell growth by 50%. bThe agent concentration that
inhibited viral replication in H9 cell by 50%. cThese data have already been prepared.13
dAzidothymidine

HETEROCYCLES, Vol. 81, No. 5, 2010
1217
Anti-HIV activity.
The isoquinoline alkaloids (1-4 and 9-18) were tested against HIV-1 replication in H9 lymphocytes in
order to evaluate their anti-HIV activity. However, none of them displayed anti-HIV activity (Table 7).
Interestingly, results in the anti-HIV assay did not parallel those in the antimicrobial and cytotoxicity
assays.
Conclusions.
In summary, three types of isoquinoline alkaloids (1-18) were tested for antimicrobial, cytotoxic,
anti-malarial, anti-oxidant, and anti-HIV activities, as well as inhibitory activity against Epstein-Barr
virus early antigen (EBV-EA) activation induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) in Raji
cells. N- or O-Geranylation contributed to increased potency in four types of activities except anti-HIV
and anti-oxidant. N,N-Geranylation of salsolinol (simple isoquinoline, 1) strongly increased the potency
in antimicrobial activity. N,N- and N-Geranylations of tetrahydropapaveroline (1-benzylisoquinoline, 9)
strongly enhanced cytotoxic activity. N,N-Geranylation of 1 and N-geranylation of 17 (2,3,10,11-
oxygenated protoberberine) increased the same activity to a lesser extent. N,N-Geranylation of 1 and
O,O-geranylation of papaveroline (1-benzylisoquinoline, 11) strongly increased cytotoxic activity against
the HL-60 cell line. N,N-Geranylsalsolinol (3) showed potent inhibitory effects on EBV-EA induction
compared with those of the parent compound (1). O,O-Geranylation of 5 (simple isoquinoline) and
N,N-geranylation of 11 (1-benzylisoquinoline) enhanced the inhibitory activity. The protoberberines
(14-18) tested also displayed strong inhibitory activity.
N-Geranylation of 2,3,9,10- and
2,3,10,11-oxygenated protoberberines increased the antimalarial activity. Among the tested biological
activities of the isoquinolines 1-18, N,N-geranylation of salsolinol (1) strongly increased potency in three
assays, antimicrobial and cytotoxic activities and inhibitory effects on EBV-EA induction, while
N-geranylation increased the same activities to a lesser extent. However, N,N- and N-methylation did
not increase the activities in these assays. These simple N-geranylated isoquinolines also have free
radical scavenging activity. These findings indicate that the N-geranyl group plays an important role in
mediating these biological activities.
Compound 3 shows antimicrobial, cytotoxic, and inhibitory effects on EBV-EA induction, compounds 10,
11, 13, and 18 are cytotoxic, 8, 13, and 14-18 inhibit EBV-EA induction, and 16 and 18 display
antimalarial activity. In the present studies, we have identified new biologically active N- or
O-geranylated isoquinolines, which may be considered as lead structures for developing potential
chemotherapeutic agents. It was first suggested that the addition of a geranyl residue to isoquinoline
skeletons may contribute to the enhancement of the biological activities of isoquinoline alkaloids.

1218
HETEROCYCLES, Vol. 81, No. 5, 2010
EXPERIMENTAL
1
General procedures. Conventional H NMR, NOESY, COSY, HMBC, and HMQC spectra were
obtained on a Varian VXR-500 spectrometer (1H: 500 MHz) using CD3OD solvent, except where noted,
with TMS as int. standard. 13C NMR and DEPT spectra were measured on a Varian VXR-500
spectrometer (125 MHz). Mass spectra were determined on a Hitachi M-4100 instrument at 75 eV. The
secondary ion mass spectra (SIMS) were measured using glycerol as matrix.
HPLC and prep. HPLC
analyses were performed using a Hitachi M-6200 intelligent pump (1 mL/min) and Hitachi M-6250 or
Jasco PU-2089 intelligent pump (6 mL/min), respectively, and a Hitachi L-4000 UV detector (280 nm).
Cosmosil 5C -AR reversed-phase column of small (4.6 i.d. X 150 mm) and large (20 i.d. X 250 mm)
18
sizes were used for HPLC and prep. HPLC, respectively. Analyses with a Hitachi HPLC system were
made using a solvent system, (A) 0.1M NH OAc (0.05% TFA) / (B) MeOH (0.05% TFA) under the
4
following gradient conditions: A/B, initial (75/25), 10 min (50/50), 30 min (20/80) or initial (80/20), 30
min (0/100) or initial (80/20), 20 min (0/100) (flow rate 1 mL/min). Prep. HPLC analyses for
purification were performed using a solvent system, (A) H2O (0.05% TFA) / (B) MeOH (0.05% TFA)
under the following gradient conditions: A/B 50/50 to 0/100, 30 to 60 min (flow rate 6 mL/min).
()-Tetrahydropapaverine hydrochloride and azidothymidine (AZT) were purchased (Sigma).
()-Salsolinol (1),8 ()-carnegine,9 ()-N-methylpapaveroline (11),9 ()-2,3,9,10-tetrademethyl-
tetrahydropalmatine (14),9 and ()-2, 3, 10, 11-tetrademethylpseudotetrahydropalmatine (17)9 have
previously been prepared.
Optimized geometry and molecular orbital were calculated by the DFT Method (Density Function
Theory) using the Materials Studio DMol3 package of Accelrys Inc.14,15 First, optimized geometry was
obtained using the Perdew-Wong LDA functional (PWC)21 and double numerical plus d-functional
(DND) basis set. Second, the optimized geometry obtained was further calculated for molecular orbitals
using the Becke exchange plus Lee-Young-Parr correlation (BLYP)22, 23 and the double numerical plus
polarization (DNP) basis set.
Preparations of ()-N- and ()-N,N-geranylsalsolinol (2 and 3). To a stirred suspension of sodium
hydride (245 mg, 10.2 mmol) in DMF (15 mL) at room temperature under N2 ()-salsolinol
hydrochloride8 (1, 1 g, 4.65 mmol) was added by portions followed by a catalytic amount of
hydroquinone. The mixture was stirred for 30 min. Geranyl bromide (953 mg, 4.39 mmol) was added
dropwise and the mixture was stirred for 2 h. NaH (245 mg, 10.2 mmol) and geranyl bromide (953 mg,
4.39 mmol) were added again, and the mixture was stirred for 2 h. After decomposition of excess NaH
with MeOH (1 mL), the mixture was poured onto ice-water and extracted with Et2O followed by CHCl3.
Et2O and CHCl3 were separately dried and evaporated. The CHCl3 extract was subjected to prep.

HETEROCYCLES, Vol. 81, No. 5, 2010
1219
HPLC [0.1M NH4OAc (0.05% TFA) / MeOH (0.05% TFA) (A/B) initial 60/40, 0/100 (30 min)] to give
()-N-geranylsalsolinol (2, 205.7 mg, yield 10.3%) and ()-N,N-geranylsalsolinol (3, 329.9 mg; yield
12.6%) as trifluoroacetate. 2: SIMS m/z [M + H]+ 316; HRMS m/z [M + H]+ 316.2257 (C20H30NO2
requires 316.2271). 3: SIMS m/z [M]+ 452; HRMS m/z [M]+ 452.3527 (C30H46NO2 requires 452.3523).
The 1H and 13C NMR spectral data of these compounds have been previously presented.13
Preparation of ()-N-methylsalsolinol (4). Carnegine-HCl (500 mg, 1.94 mmol) in 47% HBr (2 ml)
was refluxed for 2 h at 130° in an oil bath. The solvent was evaporated in vacuo, the residue in MeOH
was crystallized to give ()-N-methylsalsolinol-HBr (4, 398.5 mg, yield 64.8%). SIMS m/z [M + H]+
1
13
194; HRMS m/z [M + H]+ 194.1201 (C11H16NO2 requires 194.1176). The H and C NMR spectral
data of 4 have been previously presented.13
Preparation of ()-N,N-dimethylsalsolinol (5). A solution of carnegine-HCl (3.1 g, 12.06 mmol) and
CH3I (1 ml) in MeOH (25 ml) and (Me)2CO (25 ml) in a glass-stoppered bottle was heated for 75 min at
110 °C in a oil bath. The solvent was evaporated in vacuo, the residue was crystallized in
(Me)2CO-Et2O to give N-methylcarnegine iodide (4.75 g, yield 86.5%). The iodide (3.8 g) in 47% HBr
(10 ml) was refluxed for 1 h at 140 °C in an oil bath. The solvent was evaporated in vacuo, the residue
was crystallized in (Me)2CO-Et2O to give ()-N,N-dimethylsalsolinol bromide (5, 2.97 g, yield 98.5%).
SIMS m/z [M]+ 208; HRMS m/z [M]+ 208.1350 (C12H18NO2 requires 208.1336). The 1H and 13C NMR
spectral data of 5 have been previously presented.13
Preparations of ()-7-O-geranyl-, ()-6-O-geranyl-, and ()-6,7-O-digeranyl-N,N-dimethylsalsolinol
(6-8). To a stirred suspension of NaH (260 mg, 10.8 mmol) in DMF (15 mL) at rt under N2 ()-N,N-
dimethylsalsolinol hydrobromide (5, 1 g, 3.47 mmol) was added portionwise followed by a catalytic
amount of hydroquinone. The mixture was stirred for 30 min. Geranyl bromide (1 g, 4.61 mmol) was
added dropwise and the whole mixture was stirred for 3 h. After decomposition of excess NaH with
MeOH (1 mL), the mixture was poured onto ice-water and extracted with CHCl3. The combined
organic layer was dried and evaporated. The CHCl3 extract was subjected to prep. HPLC [0.1M
NH4OAc (0.05% TFA)/MeOH (0.05% TFA) (A/B) initial 100/0, 0/100 (540 min)] to give
()-7-O-geranyl-N,N-dimethylsalsolinol (6, 102.2 mg, yield 7.7%), ()-6-O-geranyl-N,N-
dimethylsalsolinol (7, 98.8 mg, yield 7.4%) and ()-6,7-O-digeranyl-N,N-dimethylsalsolinol (8, 154.2 mg,
yield 8.9%) as trifluoroacetates. 6: 1H NMR (500 MHz, CD3OD)  6.68 (1H, s, H-5), 6.74 (1H, s, H-8),
5.48 (1H, t, J = 6.5 Hz, H-2´), 5.09 (1H, t, J = 6.5 Hz, H-6´), 4.62 (2H, d, J = 6.5 Hz, H-1´), 4.57 (1H, q, J
= 6.5 Hz, H-1), 3.77 (1H, dt, J = 12.5, 8.0 Hz, Hax´-3), 3.55 (1H, dt, J = 12.5, 5.0 Hz, Heq´-3), 3.17 (3H,
s, ax´NMe), 3.14 (3H, s, eq´NMe), 3.08 (2H, m, H-4), 2.12 (2H, m, H-5´), 2.09 (2H, m, H-4´), 1.75 (3H, s,
3´-Me), 1.69 (3H, d, J = 6.5 Hz, Me-1), 1.65 (3H, s, 8´), 1.59 (3H, s, 7´-Me); 13C NMR (125 MHz,

1220
HETEROCYCLES, Vol. 81, No. 5, 2010
CD3OD) C-6), 147.83 (C-7), 142.30 (C-3´), 132.63 (C-7´), 124.90 (C-6´), 124.39 (C-1a),
122.17 (C-4a), 120.94 (C-2´), 116.01 (C-5), 113.25 (C-8), 69.51 (C-1), 67.06 (C-1´), 56.86 (C-3), 51.50
(eq´NMe), 50.11 (ax´NMe), 40.61 (C-4´), 27.41 (C-5´), 25.84 (C-8´), 24.30 (C-4), 18.63 (C-1-Me), 17.72
(7´-Me), 16.69 (3´-Me); SIMS m/z [M]+ 344; HRMS m/z [M + H]+ 344.2606 (C22H34NO2 requires
1
344.2588), m/z [M-C10H16]+ 208.1355 (C12H18NO2 requires 208.1338) ; 7: H NMR (500 MHz, CD3OD)
 6.79 (1H, s, H-5), 6.66 (1H, s, H-8), 5.50 (1H, m, H-2´), 5.10 (1H, m, H-6´), 4.63 (2H, d, J = 6.5 Hz,
H-1´), 4.55 (1H, q, J = 6.5 Hz, H-1), 3.78 (1H, dt, J = 12.5, 7.5 Hz, Hax´-3), 3.56 (1H, dt, J = 12.5, 6.0 Hz,
Heq´-3), 3.17 (3H, s, ax´NMe), 3.13 (3H, s, eq´NMe), 3.12 (2H, m, H-4), 2.12 (2H, m, H-5´), 2.07 (2H, m,
H-4´), 1.75 (3H, s, 3´-Me), 1.67 (3H, d, J = 6.5 Hz, Me-1), 1.65 (3H, s, Me-8´), 1.60 (3H, s, 7´-Me); 13C
NMR (125 MHz, CD3OD) C-6), 147.90 (C-7), 142.26 (C-3´), 132.63 (C-7´), 124.90 (C-6´),
124.39 (C-1a), 122.17 (C-4a), 120.94 (C-2´), 116.01 (C-5), 113.25 (C-8), 69.36 (C-1), 66.86 (C-1´), 57.19
(C-3), 51.58 (eq´NMe), 49.80 (ax´NMe), 40.62 (C-4´), 27.39 (C-5´), 25.86 (C-8´), 24.51 (C-4), 18.36
(C-1-Me), 17.74 (7´-Me), 16.67 (3´-Me); SIMS m/z [M]+ 344; HRMS m/z [M + H]+ 344.2603
1
(C22H34NO2 requires 344.2588), m/z [M-C10H16]+ 208.1317 (C12H18NO2 requires 208.1338); 8: H NMR
(500 MHz, CD3OD)  6.83 (1H, s, H-5), 6.79 (1H, s, H-8), 5.48 (2H, m, H-2´, H-2´´), 5.10 (2H, m, H-6´,
H-6´´), 4.59 (2H, m, H-1´´), 4.58 (2H, m, H-1´), 4.54 (1H, q, J = 6.5 Hz, H-1), 3.79 (1H, dt, J = 12.5, 8.0
Hz, Hax´-3), 3.57 (1H, dt, J = 12.5, 5.0 Hz, Heq´-3), 3.17 (3H, s, ax´NMe), 3.15 (3H, s, eq´NMe), 3.14
(2H, m, H-4), 2.12 (4H, m, H-5´, H-5´´), 2.07 (4H, m, H-4´, H-4´´), 1.742 (3H, s, 3´´-Me), 1.736 (3H, s,
3´-Me), 1.69 (3H, d, J = 6.5 Hz, Me-1), 1.66 (6H, s, 8´, 8´´), 1.60 (6H, s, 7´-Me, 7´´-Me); 13C NMR (125
MHz, CD3OD) C-6), 149.78 (C-7), 142.58(C-3´´), 142.56 (C-3´), 132.68 (C-7´), 125.09 (C-1a),
124.96 (C-6´), 122.33 (C-4a), 120.94 (C-2´´), 120.85 (C-2´), 115.05 (C-5), 114.41 (C-8), 69.60 (C-1),
67.38 (C-1´), 67.04 (C-1´´), 56.83 (C-3), 51.55 (eq´NMe), 50.25 (ax´NMe), 40.66 (C-4´, C-4´´), 27.48
(C-5´), 27.44 (C-5´´), 25.93 (C-8´), 24.56 (C-4), 18.63 (C-1-Me), 17.81 (7´-Me), 16.77 (3´-Me), 16.74
(3´´-Me); SIMS m/z [M]+ 480; HRMS m/z [M]+ 480.3858 (C32H50NO2 requires 480.3837), m/z
[M-C10H16]+ 344.2586 (C22H34NO2 requires 344.2588).
Preparation of ()-tetrahydropapaveroline (9). A solution of ()-tetrahydropapaverine hydrochloride
(Sigma, 2 g, 5.26 mmol) in 47% HBr (10 ml) was refluxed for 2.5 h. The solvent was evaporated in
vacuo. The crystalline product was recrystallized from MeOH to give ()-tetrahydropapaveroline
1
hydrobromide 9 (1.84 g, yield 95.1%) which was identified by comparing its H NMR and HPLC with
data of an authentic sample.
Preparations of ()-N- and ()-N,N-digeranyltetrahydropapaveroline (10 and 11). To a stirred
suspension of NaH (65 mg, 2.71 mmol) in DMF (15 mL) at rt under N2 ()-tetrahydropapaveroline
hydrobromide (9, 1 g, 2.72 mmol) was added portionwise followed by a catalytic amount of

HETEROCYCLES, Vol. 81, No. 5, 2010
1221
hydroquinone. The mixture was stirred for 30 min. Geranyl bromide (590 mg, 2.72 mmol) was added
dropwise, and the mixture was stirred for 10 h. Geranyl bromide (295 mg, 1.36 mmol) was added
dropwise and the mixture was further stirred for 12 h. After decomposition of excess NaH with MeOH
(1 mL), the mixture was poured onto ice-water and extracted with Et2O followed by CHCl3. The Et2O
and CHCl3 phases were separately dried and evaporated. The CHCl3 extract was subjected to prep.
HPLC [0.1M NH4OAc (0.05% TFA)/MeOH (0.05% TFA) (A/B) initial 80/20, 0/100 (60 min)] to give
()-N-geranyltetrahydropapaveroline (10, 274 mg, yield 18.7%) and ()-N,N-digeranyltetrahydro-
papaveroline (11, 175 mg, yield 9.5%) as trifluoroacetate. 10: 1H NMR (500 MHz, CD3OD)  6.75 (1H,
d, J = 8.5 Hz H-5´), 6.65 (1H, s, H-5), 6.58 (1H, brs, H-2´), 6.50 (1H, brd, J = 8.5 Hz, H-6´), 6.18 (1H, s,
H-8), 5.26 (1H, m, H-2´´), 5.12 (1H, m, H-6´´), 4.42 (1H, t, J = 6.5 Hz, H-1), 3.83 (1H, dd, J = 13.0, 8.5
Hz, H-1´´), 3.71 (1H, m, H-1´´), 3.68 (1H, m, H-3), 3.38 (1H, m, H-3), 3.14 (1H, m, H-9), 3.04 (1H, m,
H-9), 2.97 (2H, m, H-4), 2.18 (4H, s, H-4´´, H-5´´), 2.12 (2H, m, H-5´), 2.07 (2H, m, H-4´), 1.70 (3H, s,
Me-8´´), 1.64 (3H, s, 7´´-Me), 1.57 (3H, s, 3´´-Me); 13C NMR (125 MHz, CD3OD) 149.65 (C-3´´),
C-6), 146.92 (C-3´), 146.04 (C-4´), 145.65 (C-7), 133.29 (C-7´´), 127.63 (C-1´), 124.56 (C-6´´),
121.93 (C-4a), 121.87 (C-1ª, C-6´), 117.67 (C-2´), 116.76 (C-5´), 116.12 (C-5), 115.75 (C-8), 114.15
(C-2´´), 63.01 (C-1), 51.86 (C-1´´), 44.80 (C-3), 40.74 (C-4´´, C-9), 27.14 (C-5´´), 25.94 (C-8´´), 22.90
(C-4), 17.84 (7´´-Me), 16.80 (3´´-Me); SIMS m/z [M + H]+ 424; HRMS m/z [M + H]+ 424.2494
(C26H34NO4 requires 424.2482); 11: 1H NMR (500 MHz, CD3OD)  6.67 (1H, d, J = 8.0 Hz H-5´), 6.64
(1H, s, H-5), 6.38 (1H, brs, H-2´), 6.30 (1H, dd, J = 8.0, 2.0 Hz, H-6´), 5.89 (1H, s, H-8), 5.61 (1H, d, J =
7.0 Hz H-2´´´), 5.42 (1H, t, J = 7.0 Hz, H-2´´), 5.15 (1H, m, H-6´´´) 4.99 (1H, m, H-6´´), 4.42 (1H, dd, J
= 9.5, 3.0 Hz, H-1), 4.20 (1H, dd, J = 14.5, 7.0 Hz, H-1´´´) 4.13 (1H, dd, J = 14.5, 7.0 Hz, H-1´´´), 3.96
(1H, dd, J = 14.5, 9.0 Hz, H-1´´), 3.78 (1H, dd, J = 14.5, 6.0 Hz, H-1´´), 3.68 (1H, m, Hax´-3), 3.53 (1H,
m, Heq´-3), 3.45 (1H, dd, J = 13.5, 3.5 Hz, H-9), 2.88 (1H, dd, J = 13.5, 9.5 Hz, H-9), 2.97 (2H, m, H-4),
2.28 (4H, s, H-4´´´, H-5´´´), 2.18 (4H, s, H-4´´, H-5´´), 2.07 (2H, m, H-4´), 1.85 (3H, s, 3´´´-Me), 1.69
(3H, s, Me-8´´), 1.68 (3H, s, Me-8´´´), 1.65 (3H, s, 7´´´-Me), 1.62 (3H, s, 7´´-Me) , 1.46 (3H, s, 3´´-Me);
13C NMR (125 MHz, CD3OD) 151.18 (C-3´´), 150.09 (C-3´´´), C-6), 146.65 (C-3´), 145.82
(C-4´), 145.24 (C-7), 133.37 (C-7´´), 133.32 (C-7´´´), 127.88 (C-1´), 124.70 (C-6´´), 124.65 (C-6´´´),
121.42 (C-4a), 122.52 (C-1a), 122.25 (C-6´), 118.04 (C-2´), 116.51 (C-5´), 115.85 (C-5), 116.81 (C-8),
112.11 (C-2´´, C-2´´´), 70.72 (C-1), 58.30 (C-1´´´), 55.79 (C-1´´), 52.01 (C-3), 41.06 (C-4´´, C-4´´´),
38.32 (C-9), 27.14 (C-5´´´), 27.05 (C-5´´), 25.98 or 25.99 (C-8´´or C-8´´´), 24.0 (C-4), 17.87 or 17.89
(7´´-Me or 7´´´-Me), 17.28 (3´´´-Me), 16.90 (3´´-Me);SIMS m/z [M]+ 560; HRMS m/z [M]+ 560.3746
(C36H50NO4 requires 560.3746), m/z [M-C10H16]+ 424.2498 (C26H34NO2 requires 424.2482).
Preparation of 6,7-O,O-digeranyl-N-methylpapaveroline (13). To a stirred suspension of NaH (115

1222
HETEROCYCLES, Vol. 81, No. 5, 2010
mg, 4.79 mmol) in DMF (15 mL) at rt under N2 N-methylpapaveroline bromide9 (12, 1 g, 2.66 mmol) was
added portionwise followed by a catalytic amount of hydroquinone. The mixture was stirred for 30 min.
Geranyl bromide (240 mg, 1.11 mmol) was added dropwise and the mixture was stirred for 10 h.
Geranyl bromide (295 mg, 1.36 mmol) and NaH (70 mg, 2.92 mmol) were added dropwise and the
mixture was stirred for a further 5 h. The reaction mixture was treated and separated by prep. HPLC as
described in preparation of 10 and 11 to give 6,7-O,O-digeranyl-N-methylpapaveroline (13, 31 mg; yield
4.1%) as trifluoroacetate. 13: 1H NMR (500 MHz, CD3OD)  8.32 (1H, d, J = 7.0 Hz, H-3), 8.08 (1H, d,
J = 7.0 Hz, H-4), 7.70 (1H, s, H-8), 7.60 (1H, s, H-5), 6.71 (1H, d, J = 8.5 Hz, H-5´), 6.49 (1H, d, J = 1.5
Hz, H-2´), 6.36 (1H, dd, J = 8.5, 1.5 Hz, H-6´), 5.58 (1H, t, J = 6.5 Hz, H-2´´), 5.39 (1H, t, J = 6.5 Hz,
H-2´´´), 5.09 (1H, m, H-6´´), 5.02 (1H, m, H-6´´´), 4.90 (2H, d, J = 6.5 Hz, H-1´´), 4.83 (2H, s, H-9), 4.80
(2H, d, J = 6.5 Hz, H-1´´´), 4.30 (3H, s, N-Me), 2.15 (H-5´´), 2.13 (4H, s, H-4´´), 2.05 (2H, m, H-5´´´),
1.98 (2H, m, H-4´´´), 1.83 (3H, s, 3´´-Me), 1.70 (3H, s, 3´´´-Me), 1.61 (3H, s, 8´´), 1.59 (3H, s, 8´´´), 1.58
(3H, s, 7´´-Me), 1.55 (3H, s, 7´´´-Me); 13C NMR (125 MHz, CD3OD) C-6),
C-1154.18 (C-7), 147.50 (C-3´), 146.17 (C-4´), 144.48 (C-3´´), 143.93 (C-3´´´), 137.50 (C-4a),
136.61 (C-3), 132.81 (C-7´´), 132.63 (C-7´´´), 126.47 (C-1´), 125.62 (C-1a), 124.91 or 124.84 (C-6´´or
C-6´´´), 123.58 (C-4), 120.39 (C-6´), 119.65 (C-2´´´), 119.32 (C-2´´), 117.14 (C-5´), 116.13 (C-2´),
108.15 (C-8), 108.08 (C-5), 67.71 (C-1´´), 67.33 (C-1´´´), 46.28 (N-Me), 40.63 (C-4´´) , 40.46 (C-4´´´),
34.96 (C-9), 27.30 or 27.29 (C-5´´or C-5´´´), 25.91 (C-8´´), 25.87 (C-8´´´), 17.82 or 17.81 (7´´ or
7´´´-Me), 16.88 or 16.84 (3´´ or 3´´´-Me); SIMS m/z [M]+ 570; HRMS m/z [M]+ 570.3592 (C37H48NO4
requires 570.3583).
Preparations of ()-(cis) and ()-(trans)-N-geranyl-2,3,9,10-tetrademethyltetrahydropalmatinium
salts (15 and 16). To a stirred suspension of NaH (126 mg, 5.25 mmol) in DMF (15 mL) at room
temperature under N2 ()-2,3,9,10-tetrademethyltetrahydropalmatine hydrobromide9 (14, 1 g, 2.63 mmol)
was added by portion followed by a catalytic amount of hydroquinone. The mixture was stirred for 30
min. Geranyl bromide (571 mg, 2.38 mmol) was added dropwise and the mixture was stirred for 2.5 h.
Geranyl bromide (220 mg, 1.01 mmol) and NaH (50 mg, 2.08 mmol) were added again and the mixture
was further stirred for 2.5 h. After decomposition of excess NaH with MeOH (1 mL), the mixture was
poured onto ice-water and extracted with Et2O followed by CHCl3. The Et2O and CHCl3 phases were
separately dried and evaporated. The CHCl3 extract was subjected to preparative HPLC [0.1M NH4OAc
(0.05% TFA)/MeOH (0.05% TFA) (A/B) initial 60/40, 0/100 (30 min)] to give ()-(cis)-N-geranyl-
2,3,9,10-tetrademethyltetrahydropalmatinium salt (15, 29.5 mg, yield 2.0%) and ()-(trans)-N-geranyl-
2,3,9,10-tetrademethyltetrahydropalmatinium salt (16, 137.6 mg, yield 9.5%) as trifluoroacetates. 15:
1H NMR (500 MHz, CD3OD)  6.78 (1H, d, J = 8.0 Hz, H-11), 6.72 (1H, s, H-4), 6.66 (1H, s, H-1), 6.57