Biochemical Systematics and Ecology 33 (2005) 831e839
www.elsevier.com/locate/biochemsyseco
Chemosystematic value of cyclopeptide
alkaloids from Heisteria nitida
(Olacaceae)
Hesham R. El-Seedi a,b,*, Sonny Larsson a,*,
Anders Backlund a
a
Division of Pharmacognosy, Department of Medicinal Chemistry, Biomedical Centre,
Uppsala University, Box 574, SE-751 23 Uppsala, Sweden
b
Department of Chemistry, Faculty of Science, El-Menoufia University,
Shebin El-Kom, Egypt
Received 14 May 2004; accepted 24 December 2004
Keywords: Heisteria nitida; Olacaceae; Cyclopeptide alkaloids; Chemosystematics
1. Introduction
Previously two cyclopeptide alkaloids have been isolated from Heisteria nitida
Engl. (Olacaceae), namely integerrenine 1 and the unusual oxide anorldianine 27-N
oxide 2, see Fig. 1 (El-Seedi et al., 1999). In this work we report the isolation of
anorldianine 3 (Fig. 1) from this plant, and discuss some chemosystematic
implications of these alkaloids. Anorldianine has previously only been reported from
Canthium arnoldianum Hepper [misspelled as Canthium anorldianum throughout the
reference, hence giving the alkaloid the name anorldianine], also known as Psydrax
arnoldiana Bridson (Rubiaceae). This alkaloid has a unique substructure containing
proline (Dongo et al., 1989). Cyclopeptide alkaloids have been found in several
* Corresponding authors. Tel.: C46 18 471 44 96/97; fax: C46 18 50 91 01.
E-mail addresses: hesham.el-seedi@fkog.uu.se (H.R. El-Seedi), sonny.larsson@fkog.uu.se (S. Larsson).
0305-1978/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bse.2004.12.023
832
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
Fig. 1. The chemical structures of the cyclopeptide alkaloids isolated from Heisteria nitida.
families. No extensive investigations into their physiological role seem to have been
performed, but reviews report antibacterial and antifungal activities (e.g. Gournelis
et al., 1997), and vignatic acid A is lethal to the larvae of the weevil Callosobruchus
chinensis (Sugawara et al., 1996). The structural type that contain 14 atoms in the
macrocyclic part have been found in nine families: Olacaceae, Celastraceae,
Phyllanthaceae, Pandaceae, Fabaceae, Rhamnaceae, Urticaceae, Malvaceae and
Rubiaceae (El-Seedi et al., 1999; Gournelis et al., 1997; Sugawara et al., 1996). The
variations in the macrocyclic entity can be grouped as shown in Fig. 2, and the
distribution of these groups among the plant families is given in Table 1.
2. Material and methods
2.1. Plant material
The bark of H. nitida was collected by Dr Felipe Ghia in 1992 at the Reserva
Biologica, Jatun Sacha, Provincia del Napo, Ecuador. Voucher specimens are
deposited in the Herbario Economica, Escuela Politecnica Nacional, EPN, Quito,
Ecuador (G. F. 539), and in the herbarium of the Botany Section, Museum of
Evolution at Uppsala University (UPS), Sweden.
2.2. Extraction and isolation
The bark of H. nitida was dried at 40 C in a dark ventilated hood before
grounding. The material (440 g) was extracted with light petroleum ether (40e60 )
three times with occasional stirring and filtered, followed by three extractions with
methanol during 4 days each. The extracts were evaporated to give 3.3 g and 34 g of
a gelatinous and an oily material, respectively. The methanol extract was partitioned
between ethyl acetate and water. This resulted in 6.7 g of an ethyl acetate soluble
fraction, and a water phase which was freeze-dried to give 27 g of a crude material
consisting mainly of carbohydrates.
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
833
Fig. 2. The five substructures used as patterns for cyclopeptide alkaloids with 14 atoms in the macrocyclic
part of the molecule. Displayed names are taken from groups proposed by Gournelis et al. (1997).
The ethyl acetate fraction (6.5 g) was subjected to SEPARO column chromatography on 20 g of silica gel 60, using gradient elution with hexane:dichloromethane
and ethyl acetate:methanol as previously described (El-Seedi et al., 2003). The eluted
fractions were evaluated by TLC on silica gel using dichloromethane:methanol (98:2)
as eluent, giving 16 main fractions. Fraction 7 (76 mg) showed presence of alkaloids
as detected using Dragendorff’s reagent. This fraction was extracted with 0.1 M
hydrochloric acid, and subsequently made alkaline with solid sodium bicarbonate
before re-extracted with chloroform. Further purification over a Sephadex LH-20
834
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
Table 1
The distribution of structural types for cyclopeptide alkaloids with 14 atoms in the macrocyclic
substructure, with the number of investigated genera and species in each family
Ordera
Santalales
Celastrales
Malpighiales
Fabales
Rosales
Malvales
Gentianales
a
b
Familya
Olacaceae
Celastraceae
Pandaceae
Phyllanthaceae
Fabaceae
Rhamnaceae
Urticaceae
Malvaceae
Rubiaceae
Genera:species
1:1
1:1
1:1
2:2
1:1
13:31
1:1
2:5
3:4
Structural typeb
1
2
3
4
5
C
C
ÿ
C
ÿ
C
C
C
C
ÿ
ÿ
C
ÿ
C
C
ÿ
ÿ
ÿ
C
ÿ
ÿ
ÿ
ÿ
ÿ
ÿ
ÿ
C
ÿ
ÿ
ÿ
ÿ
ÿ
C
ÿ
ÿ
ÿ
ÿ
ÿ
ÿ
ÿ
ÿ
C
ÿ
ÿ
ÿ
Order and family according to APGII (APG, 2003).
The signs denote presence (C) and no reported structure (ÿ), respectively.
column using dichloromethane as eluent afforded anorldianine (10 mg) (Dongo
et al., 1989), reported for the second time from natural origin.
2.3. Structural elucidation
Our spectral data are deviating from that previously reported (Dongo et al.,
1989), and accordingly extensive analysis with NMR and MS has been performed.
The relative intensities of the m/z fragments from the electron impact mass
spectrometry are given in Table 2.
Dongo et al. (1989) assigned the structure based only on 1H, 13C and HeH COSY
NMR spectral analysis, which could explain the difference in chemical shifts. The
complete assignment of 1H NMR is presented together with the 13C NMR data in
Table 2
The relative intensities of the m/z fragments of anorldianine electron impact mass spectroscopy experiment
m/z
Relative intensity
485
440
439
328
287
194
189
166
135
114
97
70
8
38
75
23
11
30
100
22
25
32
65
98
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
835
Table 3
1
H and 13C NMR assignments of anorldianine, recorded at 400 MHz and 100.6 MHz, respectively
Pos.
13
C shift dCa
1
3
4
5
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
28, 29
30
31
32
157.1 (s)
83.6 (d)
54.1 (d)
171.3 (s)
62.8 (d)
167.1 (s)
e
126.7 (d)
116.5 (d)
131.6 (s)
130.4 (d)
120.6 (d)
120.9 (d)
132.1 (d)
28.9 (d)
15.6 (q)
20.4 (q)
e
172.0 (s)
68.2 (d)
36.2 (t)
26.4 (d)
23.4 (q)
22.9 (q)
42.2 (q)
27.9 (t)
23.6 (t)
46.7 (t)
1
H shift dH (multi. JHeH, Hz)b
e
5.01
4.75
e
4.23
e
5.00
6.64
6.40
e
6.96
7.07
7.29
7.10
1.37
1.24
0.99
8.45
e
2.50
1.49
1.65
0.94
0.92
2.12
1.46
1.91
3.86
(dd, J3,17 Z 1.5; J3,4 Z 6.9)
(dd, J4,3 Z 6.9; J4,20 Z 9.8)
(dd, J7,23 Z 6.9 and 9.8)
(d, J9,10 Z 9.7)
(dd, J10,11 Z 7.3; J10,9 Z 9.7)
(d, J11,10 Z 7.3)
(dd, J13,16 Z 1.5;
(dd, J14,15 Z 1.5;
(dd, J15,14 Z 1.5;
(dd, J16,13 Z 1.5;
(m)
(d, J18,17 Z 6.7)
(d, J19,17 Z 6.7)
(d, J20,4 Z 9.8)
J13,14 Z 7.3)
J14,13 Z 7.3)
J15,16 Z 8.8)
J16,15 Z 8.8)
(dd, J22,23 Z 5.8 and 8.3)
(m), 1.31 (m)
(m)
(d, J25,24 Z 6.4)
(d, J26,24 Z 6.4)
(s)
(m), 2.32 (m)
(m), 1.69 (m)
(m), 3.36 (m)
HSQC-DEPT
via 2,3JCeH
HMBC correlations
qCa
CH
CH
CaO
CH
CaO
NH
CHa
CHa
qCa
CHa
CHa
CHa
CHa
CH
CH3
CH3
NH
CaO
CH
CH2
CH
CH3
CH3
N(CH3)2
CH2
CH2
CH2
H-3, H-14, H-15
H-17, H-18, H-19
H-3, H-17
H-3, H-4, H-7
H-30, H-31
H-7, NH-9, H-30
e
NH-9, H-11
NH-9, H-10
H-10, H-11, H-13, H-16
H-11, H-14
H-13
H-14, H-16
H-11, H-15
H-3, H-18, H-19
H-3, H-19
H-3, H-18
e
e
H-28, H-29
H-25, H-26
H-25, H-26
H-23, H-26
H-23, H-25
e
H-7, H-32
H-7, H-30
H-30, H-31
The sample was dissolved in CDCl3 with TMS as internal standard.
a
Assignments were based on HETCOR, HSQC-DEPT and HMBC experiments.
b
Assignments were based on 1He1H COSY, NOESY and TOCSY experiments.
Table 3. The most important NOESY correlations of anorldianine are shown in
Fig. 3 and in Fig. 4 the most important HMBC correlations are presented.
2.4. Confirmation of proline
As a method of confirming the structure, the presence of proline after acid
hydrolysis was investigated. Anorldianine (500 mg) was hydrolysed with 6 M
hydrochloric acid (300 ml at 100 C during 21 h) in a sealed glass tube. After
evaporation, esterification was performed with 300 ml of 1 M hydrochloric acid in
methanol (at 100 C for 40 min). The methyl esters was N-acetylated by adding
methanol:acetic anhydride (4:1) in 300 ml of water at room temperature and
a reaction time of 60 min. The product was dissolved in methanol and analysed by
GCeMS using an HP-5, 25 m fused silica WCOT column with a temperature
836
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
Fig. 3. The most important NOESY correlations of anorldianine.
program of 140 C for 3 min and 230 C for 6 min. The MS of the peak at retention
time 8.2 min was identical to an authentic sample of N-acetylproline methyl ester.
3. Chemosystematic significance
The systematic placement of Santalales has not been possible to deduce. In the
past it has been associated with rosid taxa, of which many today are placed among
asterids, e.g. the Apiales (Backlund and Bremer, 1997), and Icacinaceae within
Aquifoliales (Kårehed, 2001). A recent phylogenetic study utilizing the sequences of
the 18 S, rbcL, atpB and 26 S genes, however, show weak support for a placement of
Santalales as sister group to the asterids (Soltis et al., 2003). Plotting the five
structural subgroups on the proposed ordinal relationships of the core eudicots
(APG, 2003), as in Fig. 5, raise interesting implications. The type 3 cyclopeptides
have a seemingly restricted distribution, including only Santalales and Gentianales,
and could hence be interpreted as supporting an asterid placement. This connection
is strengthen if taking into account that Heisteria olivae Steyerm., have been shown
to contain scopolamine (Cairo Valera et al., 1977), usually associated with the asterid
family Solanaceae, and the hypothesis merits even further investigation.
Fig. 4. The most important HMBC correlations of anorldianine.
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
837
Fig. 5. The ordinal classification of APG II (APG, 2003) with the presence of cyclopeptide alkaloids of
substructures 1e5 given after the name.
Thus far it has not been shown that Heisteria is parasitic, a feature otherwise
common in Santalales. The possibility that presence of alkaloids due to such a lifestyle
cannot be disregarded. Accumulation of host species secondary metabolites have been
demonstrated for the mistletoes Amyema sp. [sic!] (Loranthaceae) (Boonsong and
Wright, 1961), and Viscum cruciatum Sieber (Martı́n Cordero et al., 1993), as well as for
Osyris alba L. (Santalaceae) (Woldemichael and Wink, 2002).
838
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
The cyclopeptides, with a reduced macrocycle corresponding to our substructure
type 2, have a restricted distribution within the rosid clade. Biosynthetic rationale
implies that these reduced macrocycles can be thought of as precursors of the more
abundant styryl-type macrocycles. The presence of a retained carboxylic function in
the vignatic acids, placed in type 2, of Fabales (Sugawara et al., 1996), further
suggest a biosynthetic series from the carboxylic acid via the saturated carbon chain
to the styryl-function.
4. Conclusion
The pattern of cyclopeptide alkaloids with macrocycles of 14 atoms suggests
a closer relationship between Santalales and asterids than hitherto thought. The
scarce data on their biosynthesis (Baig et al., 1993) may also be complemented from
a selection of study organism on phylogenetic grounds. The low amounts of
cyclopeptide alkaloids in plants, usually in the range of 0.0002e1% (Gournelis et al.,
1997) imply that their presence in many cases may have been overlooked. The
disjunct distribution could of course also be explained by differential gene
expression, gene loss or simple convergence in response to similar ecological factors
(e.g. discussed in Wink, 2003).
A phylogenetic approach to investigate this suggests that the chemosystematic
value of cyclopeptide alkaloids would increase with a greater sampling of asterid
taxa. This while the biosynthetical investigations of these alkaloids may gain more
from a greater sampling of rosid taxa, in particular, outside the already well studied
Rhamnaceae.
Acknowledgements
We are very grateful to The Swedish Institute for a fellowship to H. R. E., and
generous financial support from the International Foundation of Science (Grant in
Aid F/3334-1).
References
APG [Angiosperm Phylogeny Group], 2003. Bot. J. Linn. Soc. 141, 399.
Backlund, A., Bremer, B., 1997. Plant Syst. Evol. 207, 225.
Baig, M.A., Banthorpe, D.V., Coleman, A.A., Tampion, M.D., Tampion, J., White, J.J., 1993.
Phytochemistry 34, 171.
Boonsong, C., Wright, S.E., 1961. Aust. J. Chem. 14, 449.
Cairo Valera, G., de Budowski, J., Delle Monache, F., Marini-Bettòlo, G.B., 1977. Atti Accad. Naz.
Lincei 62, 363.
Dongo, E., Ayafor, J.F., Sondengam, B.L., Connolly, J.D., 1989. J. Nat. Prod. 52, 840.
El-Seedi, H.R., Gohil, S., Perera, P., Torssell, K.B.G., Bohlin, L., 1999. Phytochemistry 52, 1739.
El-Seedi, H.R., Ringbom, T., Torssell, K., Bohlin, L., 2003. Chem. Pharm. Bull. 51, 1439.
H.R. El-Seedi et al. / Biochemical Systematics and Ecology 33 (2005) 831e839
839
Gournelis, D.C., Laskaris, G.G., Verpoorte, R., 1997. Nat. Prod. Rep. 14, 75.
Kårehed, J., 2001. Am. J. Bot. 88, 2259.
Martı́n Cordero, C., Gil Serrano, A.M., Ayuso Gonzalez, M.J., 1993. J. Chem. Ecol. 19, 2389.
Soltis, D.E., Senters, A.E., Zanis, M.J., Kim, S., Thompson, J.D., Soltis, P.S., Ronse De Craene, L.P.,
Endress, P.K., Farris, J.S., 2003. Am. J. Bot. 90, 461.
Sugawara, F., Ishimoto, M., Le-Van, N., Koshino, H., Uzawa, J., Yoshida, S., Kitamura, K., 1996. J.
Agric. Food Chem. 44, 3360.
Woldemichael, G.M., Wink, M., 2002. Biochem. Syst. Ecol. 30, 139.
Wink, M., 2003. Phytochemistry 64, 3.