[24], as well as adlumidine (40) [15]and (+)- O-methylflavinan- was completely inhibited by 32, strongly inhibited by 36, and
dine (41) [25], which were isolated from Formosan plants or significantly inhibited by 30. The other members of this class
were semi-synthetic derivatives.
were inactive. Two interesting SAR observations were made.
Benzylisoquinoline alkaloid 28, which has four methoxy groups
These alkaloids were studied for their effects on the aggregation at C-6, 7, 3¢, 4¢, showed strong inhibition of platelet aggregation;
of washed rabbit platelets as induced by adenosine 5¢-diphos- however, modification to the N-methylbenzyltetrahydroisoqui-
phate (ADP, 20 ꢁM), arachidonic acid (AA, 100 ꢁM), collagen noline 36 reduced the antiplatelet effects. Also saturation of the
(
10 ꢁg/mL), and platelet-activating factor (PAF, 2 ng/mL).
benzylisoquinoline B ring sharply reduced the antiplatelet activ-
ities.
As shown in Table 1, (±)-discretamine (6) showed complete inhi-
bition of AA-induced platelet aggregation, while the other proto- The proaporphines 37± 39 as well as the two miscellaneous alka-
berberines 1±5 were inactive. The protoberberine 6 possesses a loids 40 and 41 did not exhibit any significant activity.
completely different structural skeleton from acetylsalicylic acid
(
ASA), but it exhibits ASA-like activity to prevent AA-induced Among all structural types investigated, some compounds, e.g.,
platelet aggregation [26]. Thus, the precise mechanism of action 23 and 28, exhibited potent antiplatelet activities against aggre-
still needs to be explored. The phenolic moieties may play an im- gation induced by all four activators, while others, such as 18 and
portant role in activity, because when the C-3 and C-10 hydroxy 32, demonstrated highly selective inhibition toward specific tar-
groups of 6 were changed to two methoxy groups in 5, the anti- gets. The mechanism(s) of antiplatelet actions of those potent
platelet effects were eliminated.
compounds is(are) not clear at this time. As also observed in the
previous results [4], [5], in general, a tiny change in the structure
Protopine 7, which contains a methylenedioxy group at C-9 and of different sub-types of isoquinolines will cause significant
C-10, completely inhibited collagen- and PAF-induced platelet changes in anti-platelet aggregation activity.
aggregation, and significantly prevented that by induced AA.
Modification of the methylenedioxy functionality to two meth-
oxy groups as in 9 totally abolished the antiplatelet activities. Materials and Methods
Thus, we suggest that the C-9,10 methylenedioxy group is critical
to the action of protopines.
Compounds: Alkaloids 1±41 were isolated or semi-synthesized
from various Formosan plants in our past investigation. All refer-
As shown in the assay results for pavine alkaloids 10±17, only 13 ences were cited. The preparation of compounds 34 and 35 has
exhibited significant inhibition of platelet aggregation induced been detailed in the Supporting Information.
by collagen. If the hydroxy group at C-8 of 13 was converted to a
methoxy group as in 14, no antiplatelet effects were seen.
Assay methods for platelet aggregation: The assay protocol is de-
tailed in the Supporting Information.
1
240
We also assayed the unusual spirobenzylisoquinolines 18 ± 21.
Interestingly, compound 18, which possesses an exo-methylene Data analysis: The experimental results are expressed as means
group at C-13, demonstrated excellent activity against platelet S.E. and accompanied by the number of observations. A one-way
aggregation induced by PAF. Converting the methylene group of analysis of variance (ANOVA) was used for multiple compari-
18 to the methyl group of 19 eliminated the antiplatelet activity. sons, and if there was significant variation between treatment
groups, then the mean values for inhibitors were compared
The three stilbene alkaloids 22±24 were effective against plate- with those for control by Student¢s t test, and P values of less
let aggregation induced by ADP, AA, collagen, and PAF. Among than 0.05 were considered to be statistically significant.
them, the most potent compound 23 completely inhibited plate-
let aggregation induced by all four activators. By changing the
stilbene-type base to the methyl iodide salt (22) or converting Acknowledgements
the C-5,4¢-dimethoxy groups to C-5,4¢-diethoxy groups (24), the
antiplatelet effects were reduced.
This investigation was supported by a grant from the National
Science Council of the Republic of China awarded to Y.C. Wu.
Among the series of bisbenzylisoquinoline alkaloids 25±27,
compound 25 completely inhibited platelet aggregations in-
duced by AA, collagen and PAF, and significantly inhibited that References
by ADP. On the other hand, 26 completely inhibited platelet ag-
gregation induced by only AA and collagen, while 27 partially in-
1
Wu YC, Liou YF, Lu ST, Chen CH, Chang JJ, Lee KH. Cytotoxicity of iso-
quinoline alkaloids and their N-oxides. Planta Med 1989; 55: 163±5
Tzeng CC, Wu YC, Su TL, Watanabe KA, Lu ST, Chou TC. Inhibitory ef-
fects of isoquinoline-type alkaloids on leukemic cell growth and mac-
romolecule biosynthesis. Kaohsiung J Med Sci 1990; 6: 58±65
Tsai IL, Liou YF, Lu ST. Screening of isoquinoline alkaloids and their de-
rivatives for antibacterial and antifungal activities. Kaohsiung J Med
Sci 1989; 5: 132±45
hibited collagen-, AA- and PAF-induced platelet aggregation.
2
3
4
Furthermore, in the class of the benzylisoquinoline alkaloids
2
8±36, compound 28 showed a wide range of antiplatelet aggre-
gation effects. Another active member of this class is 29, which
showed strong inhibition of both ADP- and AA-induced aggrega-
tion, and significant inhibition of collagen-induced platelet ag-
gregation. In addition, platelet aggregation induced by collagen
Chen KS, Ko FN, Teng CM, Wu YC. Antiplatelet and vasorelaxing ac-
tions of some aporphinoids. Planta Med 1996; 62: 133±6
Letter¼ Planta Med 2006; 72: 1238±1241