Idiospermine and idiospermamine

Of the calycanthaceous alkaloids identified to date, idiospermamine is the one that is closest biogenetically to idiospermine.

Idiospermamine can be envisaged as having been formed by coupling of a calycanthine moiety (through a C ortho to a tetrahydroquinoline N, ie C8) at the beta position of the indole ring of an N-methyltryptamine molecule followed by the formation of a pyrrolidine ring through the linking of the tetrahydroquinoline N to the alpha position of the N-methyltryptamine molecule and methylation of the N of the indole ring.

Idiospermine, on the other hand, can be envisaged as having been formed by the coupling of a calycanthine moiety (through a C para to a tetrahydroquinoline N, ie C6) at the beta position of the indole ring of an N-methyltryptamine molecule followed by aryl migration of the calycanthine moiety and concerted methylation of the N of the indole ring followed by loss of H+ to effect aromatization of the indole ring (This may be regarded as analogous to the coupling in the formation of hodgkinsine and the quadrigemines which are followed by N-methylation, aryl migration and aromatization.).

I envisage the following related alkaloid might also one day be identified:

Coupling of a calycanthine moiety (through a C ortho to a tetrahydroquinoline N, ie C8) at the beta position of the indole ring of an N-methyltryptamine molecule followed by aryl migration of the calycanthine moiety and concerted methylation of the N of the indole ring followed by loss of H+ to effect aromatization of the indole ring. This alkaloid might be termed idiospermine-B.

Idiospermine and other biogenetically related alkaloids

Alkaloids biogenetically related to idiospermine have been isolated from sources other than plants. For example, (-)-calycanthine and (+)-chimonanthine have been isolated from the Columbian poisondart frog, Phyllobates terribilis. The communesins have been isolated from a strain of Penicillum sp. marine fungus while perophoramidine has been isolated from the fungus Perophora namei (Ascidiaceae).

A list of the alkaloids biogenetically related to idiospermine include:

• Calycanthine
• Chimonanthine
• Calycanthidine (a mono methylated derivative of chimonanthine)
• Folicanthine (a di methylated derivative of chimonanthine)
• Chimonanthidine (same as calycanthidine but with the third methyl on an indole N rather than an indolamine N)
• Hodgkisine (related to chimonanthine – consisting of three N-methyl tryptamine units)
• Qudrigemine (related to chimonanthine – consisting of four N- methyl tryptamine units)
• Idiosperumline (a mono methylated derivative of hodgkinsine)
• Psychotrimine (related to chimonanthine – consisting of three N-methyl tryptamine units with an indole N bonded to a C7 position of chimonanthine)
• Psychopentamine (related to chimonanthine – consisting of five N-methyl tryptamine units with an indole N bonded to a C7 position of chimonanthine)
• Idiospermamine (consisting of three N-methyl tryptamine units with a calycanthine moiety attached to a beta position of an N-methyl tryptamine unit followed by cyclisation to form a pyrrolidine ring)
• Calycosidine (consisting of a calycanthine moiety attached to a pyrrolidinoindole moiety)
• Psycholeine (consisting of a calycanthine moiety attached to a pyrrolidinoindole moiety at each of the C8 positions)
• Psychotridine (consisting of five N-methyl tryptamine units)
• Isopsychotridine (consisting of five N-methyl tryptamine units)
• Vatine and Vatine-A (both consisting of six N-methyl tryptamine units)
• Vatamine (consisting of seven N-methyl tryptamine units)
• Vatamidine (consisting of eight N-methyl tryptamine units)
• Perophoramidine
• CPC-2

A helpful overview of the biogenesis of these alkaloids may be found in Jeremy A. May and Brian Stoltz, “The structural and synthetic implications of the biosynthesis of the calycanthaceous alkaloids, the communesins and nomofungin”, Tetrahedron, vol. 62 (2006), pp5262-5271.

The chemical constituents of Idiospermum australiense (Diels) S.T. Blake – Part 2

In the first post I discussed the alkaloids that were extracted from the plant. A subsequent post discussed the structure of a new triterpene lactone that was extracted. This post will discuss the other substances that were identified from the plant.

Of the substances that merit mention -

The bark afforded:

The coumarin scopoletin 1.0%
The flavone glycoside hesperidin 2.0%
β-sitosterol 0.01%


The seeds or cotyledons afforded:

Hesperidin 0.5%
Scopoletin 0.2%
p-hydroxybenzaldehyde 0.05%


The wood afforded:

The flavone hesperetin 0.05%
Scopoletin 0.14%
Scopoletin methyl ether 0.03%


The leaves afforded:

Hesperetin 0.05%
The flavone diosmetin 1.5 x 10-3%
The flavone acacetin 1.5 x 10-3%
A sesquiterpene hydrocarbon (C15H24) 0.38%

Because of time constraints, further investigative work on the sesquiterpene hydrocarbon was not carried out.

The Australian researchers, Brophy and Goldsack have examined the essential oils in Idiospermum australiense. Their results are as follows:

The leaf essential oil obtained from Idiospermum australiense (Diels) S. T. Blake in 0.3-0.5% yield (based on fresh weight) consisted of monoterpenes and sesquiterpenes. Of the 41 compounds detected in amounts >0.05%, the principal components were bicyclogermacrene (48%) and caryophyllene (8%). The tricyclic sesquiterpene alcohols globulol, viridiflorol and spathulenol were also present each in approximately 5% amounts – Jospeh J. Brophy and Robert J. Goldsack Flavour and Fragrance Journal, vol. 7 (1992), pp79-80).

Both bicyclogermacrene and caryophyllene are sesquiterpenes with the molecular formula C15H24.

D.A. Young and R.W. Sterner, Biochem. Syst. Ecol.vol. 9 (1981), p185 have examined the flavonoids in the leaves of Idiospermum australiense. Their results are cited by Robert Hegenauer in his “Chemotxonomie der Pflanzen: eine Übersicht über die Verbreitung und die sysstematische Bedeutung der Pflanzenstoffe” vol. VIII on p109. Hegenauer notes that Young and Sterner refer to the presence of 7-glucosides of luteolin and apigenin.

Our results revealed the presence of the 7- glycoside, hesperidin, and its aglycone hesperetin together with diosmetin and acacetin. Diosmetin is a methyl derivative of luteolin while acacetin is a methyl derivative of apigenin. Hesperetin is the flavanone from diosmetin.

Idiospermuline and idiospermamine – other tryptamine trimers from Idiospermum australiense

In an earlier post reference was made to both idiospermuline and idiospermamine.

Idiospermuline was isolated from the seeds of Idiospermum australiense by researchers from the Department of Pharmacology of the University of Sydney (Rujee K. Duke, Robin D. Allan, Graham A. R. Johnston, Kenneth N. Mewett and Ann D. Mitrovic, “Idiospermuline, a trimeric pyrrolidinoindoline alkaloid from the seed of Idiospermum australiense, Journal of Natural Products, vol. 58 (1995), pp1200-1208).

(-)-Idiospermuline was isolated form the seeds of the plant together with (+)-calycanthine and (-)-chimonanthine. Idiospermuline is a chimonanthine derivative and may be viewed as consisting of a chimonanthine moiety coupled via its C7’ carbon atom to the C3a” carbon atom of a pyrrolidinoindole moiety. In overall terms it is formally a dimethylated derivative of hodgkinsine, an alkaloid from Hodgkinsonia frutescens F. Muell (Rubiaceae) – A.A. Gorman, M. Hesse, H. Schmidt, P.G. Waser and W.H. Hopff in The Alkaloids, vol. 1, p203 (The Chemical Society: London 1971). Hogkinsine is also isolated from Psychotria sp.

The Overman group have made an enantioselective synthesis of (-)-idiospermuline (Tetrahedron, vol. 59 (2003), pp6905-6919).

Idiospermuline (C35H42N6) has only one secondary nitrogen atom and has a molecular weight of 546.

Idiospermamine B, on the other hand (C34H38N6, see an earlier post) has two secondary nitrogen atoms.

J.B. Bremner has reported two new alkaloids from Psychotria malayana Jack (Surya Hadi and John B. Bremner, Initial Studies on Alkaloids from Lombok Medicinal Plants, Molecules, vol. 6 (2001), pp117-129). From the plant Bremner and his colleague isolated hodgkinsine (C33H38N6, M.W. 518), chimonanthine and two other alkaloids with molecular weights of 186 and 574.

I wonder if it is possible that the alkaloid with M.W. of 574 isolated by Bremner is a trimethylated derivative of idiospermine (C34H40N6)?

The degradation of idiospermolide

An earlier post described the degradation of idiospermolide to 29-norcycloartanol. This post gives the experimental and spectral data for the intermediate degradation products.

Ring opening of idiopsermolide with potassium t-butoxide:

After reacting idiospermolide with potassium t-butoxide in t-butanol the crude diene acid (λmax 265 nm) was methylated with diazomethane to give the diene ester which crystallized from methanol as colourless needles, m.p. 149-150° (Found: C: 79.7%; H: 9.8%. C30H46O3 requires C: 79.7%; H: 9.7%). νmax 1705, 1636, 1602, 1451, 1432, 1235, 1200, 1170, 1113, 985 cm-1; λmax 268, nm, ε 25000; mass spectrum (452, (P,22), 420 (30), 326 (12), 299 (90), 292 (70), 283 (12), 281 (14), 257 (16), 229 (18), 215 (20), 203 (28), 189 (64), 175 (80), 154 (100),149 (60), 147 (70), 145 (46), 135 (62), 133 (80)).

Hydrogenation of the diene ester:

The diene ester was hydrogenated with palladium/charcoal to give the saturated ester which crystallized from methanol as colourless prisms, m.p. 83-5-85° (Found: M.W. 456.360 ± 0.003. C30H48O3 requires M.W. 456.3603). νmax 1730, 1708, 1202 cm-1; mass spectrum 456, (P, 53), 441 (21), 425 (7), 409 (10),328 (21), 299 (66), 251 (17), 175 (28), 121 (45), 107 (52), 95 (75), 81 (52), 55 (100).

Conversion of the hydrogenated ester to 29-norcycloartanol:

The ester was reduced with lithium aluminium hydride and the crude diol tosylated with p-toluenesulphonyl chloride. The crude tosylate was reduced with lithium aluminium hydride to produce 29-cycloartanol which crystallized from methanol as colourless needles, m.p. 127-130° (lit. 128-132° G. Berti, F. Bottari, A. Marsili, I. Morelli and M. Palvani, Tetrahedron Lett., 1967, p125), νmax 3700-3300, 1200, 1120 cm-1; mass spectrum: 414 (P,5), 412 (4), 399 (20), 397 (14), 396 (18), 381 (24), 121 (46), 119 (42), 109 (52), 107 (64), 105 (56), 95 (74), 93 (56), 91 (52), 81 (76), 79 (50), 69 (76), 67 (52), 57 (84), 55 (100).

Conversion of cycloeucalenol to 29-norcycloartanol:

The acetate of cycloeucalenol was ozonlyzed to give demethyloxocycloeucalenyl acetate. This was reduced with sodium borohydride, tosylated with p-toluenesulphonyl chloride and then reduced with lithium aluminium hydride. The product was identical with the 29-norcycloartanol prepared via degradation of idiospermolide: the mixed m.p. was undepressed and the i.r. spectra were identical.

The mono-ethylenethioketal derivative of idiospermolide

The n.m.r spectrum of the mono-ethylenethioketal derivative of idiospermolide revealed that under the acidic reaction conditions the cyclcopropane ring had opened to give a compound with a 9,11 double bond. 9,19-cyclotriterpenoids are known to cleave predominantly to the Δ9(11) isomer under acid treatment (G. Ourisson, P. Crabbé and O.R. Rodig, Tetracyclic Triterpenes (Herman: Paris 1964)).

Apart from the broad doublet at δ 6.61 (J 6 Hz) for C24-H and the broad doublet at δ 4.47 (J 13 Hz) for C22-H, the characteristic features of the n.m.r. spectrum were a broad doublet at δ 5.30 (J 5Hz) for C11-H, as four proton singlet at δ 3.22 for the ethylenedithioketal protons, a singlet at δ 0.72 for C14-Me and a singlet at δ 0.66 for C13-Me. The signals for C13-Me and C14-Me were assigned by comparison with the signals for C13-Me and C14-Me in grandisolide (J.P. Kutney, D.S. Grierson, G.D. Knowles, N.D. Westcott and I.H. Rogers, Tetrahedron, 1973, vol. 29, p13); the chemical shift of C11-H at δ 5.30 is characteristic of Δ9(11) triterpenoids.

The other signals of the n.m.r. spectrum were consistent with the structure proposed and were assigned as follows:

Broad singlet, 3H, δ 1.90 C25-Me
Doublet, 3H, δ 1.15, C4-Me (J 7 Hz)
Singlet, 3H, δ 0.98, C10-Me
Doublet, 3H, δ 0.96, C20-Me (J 7 Hz)
Multiplet, 21H, the remaining protons

Studies with the shift reagent Eu(dpm)3 gave further information. The reagent apparently bonded to the lactone carbonyl. For a 5.9 x 10-5M chloroform solution of the ethylenethioketal which was 1.8 x 10-5 M with respect to Eu(dpm)3, the following shifts were observed:

Doublet, δ 7.05, C24-H (J 6 Hz)
Singlet, δ 2.75, C25-Me
Doublet, δ 1.15, C4-Me (J 7 Hz)
Doublet, δ 1.11, C20-Me (J 7 Hz)
Singlet, δ 0.98, C10-Me
Singlet, δ 0.79, C14-Me
Singlet, δ 0.67, C13-Me

Idiospermolide a new triterpene lactone

From the dried leaves of Idiospermum australiense (Diels) S.T. Blake was extracted a new triterpene lactone that was given the name idiospermolide (0.037% of dried weight).

Idiospermolide crystallized from methanol as colourless plates,m.p. 243-245°, (Found: C, 79.5%; H, 9.8%; M.W. 438.315 ± 0.003. C29H42O3 requires C, 79.4%; H, 9.6%; M.W. 438.3134. O.R.D. (c 0.132 in methanol) : [α]589 + 34.7°, [α]350 + 87.6°, [α]304 + 362°, [α]294 0°, [α]270 - 1645°, [α]260 - 1090°; νmax 1705, 1448, 1380, 1348, 1140 cm-1; u.v. spectrum transparent above 210 nm; mass spectrum: 438 (P, C29H42O3, 22), 423 (C28H39O3, 12), 339 (C24H35O, 14), 315 (3), 314 (6), 299 (C21H31O, 12), 298 (12), 243 (3), 233 97), 175 (13), 173 (13), 161 (15), 147 (21), 133 (25), 123 (21), 121 (32), 119 (31), 111 (37), 107 (42), 105 (33), 95 (68), 94 (39), 93 (57), 91 (39), 83 (37), 81 (48), 79 (41), 77 (22), 69 (26), 67 (43), 55(100), 53 (38).

Idiospermolide was shown to be a 9,19 cyclotriterpenoid with a single methyl group at C4. It shares these features with cycloeucalenol (J.S.G. Cox, F.E. King and T.J. King, J. chem.. Soc., 1956, p1384), 29-norcycloartenol (M. Devys, A. Alcaide, F. Pinte and M. Barbier, Tetrahedron Lett., 1970, p4621) and 29-norcycloartanol (G. Berti, F. Bottari, A. Marsili, I. Morelli and M. Palvani, Tetrahedron Lett., 1967, p125 and K.N.N. Ayengar and E. Rangaswami, Tetrahedron Lett., 1967, p3567).

The other structural feature is the functionalization of the side chain as an α, β – unsaturated δ–lactone. The i.r. spectrum had a broad carbonyl absorption at 1705 cm-1. Idiospermolide readily gave a dihydro derivative whose i.r. spectrum showed carbonyl absorptions at 1730 cm-1 and 1705 cm-1. It also formed a mono-ethylenethioketal whose i.r. spectrum had a carbonyl absorption at 1710 cm-1. This indicates that there are at least two carbonyl groups present, one of which is conjugated to a double bond.

Idiospermolide had a cyclopropane ring at C9 and C10, a ketone at C3, an equatorial methyl group at C4, an α, β – unsaturated δ–lactone at C22.

The n.m.r. spectrum established the presence of a cyclopropane ring by an AB quartet at δ 0.48 and δ 0.62, each of one proton (J 4.0 Hz). The other signals observed in the n.m.r. spectrum were consistent with the structure proposed and were assigned as follows:

Broad doublet, 1H, δ 6.62, C24-H (J 6Hz)
Doublet of doublets, 1 H, δ 4.51, C22-H (J 12.8 Hz, J 3.5 Hz)
Broad singlet, 3H, δ 1.92, C25-Me
Singlet, 3H, δ 1.02, C13-Me
Doublet, 3H, δ 0.99, C20-Me (J 7.0 Hz)
Doublet, 3H, δ 0.98, C4-Me (J 7.0 Hz)
Singlet, 3H, δ 0.95, C14-Me
Multiplet, 23H, δ 2.80- 0.90 the remaining protons

The assignments of C13-Me and C14-Me were based on the assignment of the methyls of cycloartane derivatives (C. Iavarone, G. Paincatelli, E. Mincione and G. Niuta, Gazz. Chim. Ital., 1970, 100, p 888).

Confirmation of some of these assignments were obtained by double irradiation experiments. Thus, irradiation of the olefinic proton, C24-H, at δ 6.62 resulted in the collapse of the signal from the C25-Me at δ 1.92 to a doublet (J 2.0 Hz).

In an n.O.e. experiment, irradiation of the C25-Me signal at δ 1.92 resulted in an 11% increase of the intensity of the signal from the C24-H at δ 6.62, thus confirming their structural relationship.

Studies with the shift reagent Eu(fod)3 gave further information. The reagent apparently bonded to the C3=O rather than to the lactone ring oxygens. For a 1.1 x10-4M chloroform solution of idiospermolide which was 2.5 x 10-5M with respect to Eu(fod)3, the following shifts were observed:

Multiplet, δ 4.8 – δ 4.15, protons on C2 and C4
Doublet, δ 2.71, C4-Me (J 7.0 Hz)
Singlet, δ 2.50, C25-Me
Doublet, δ 1.24, C20-Me
Singlet, δ 1.24, C13-Me
Singlet, δ1.16, C14-Me

The mass spectrum of an ethylenedithioketal derivative of idiospermolide had its base peak at m/e 131. This gave good evidence for the presence of a 3-oxo function (compare the behaviour of 5α-androstan-3-one ethylene ketal – H. Budzikiewicz, C. Djerassi and D.H. Williams, Structure Elucidation of Natural Products by Mass Spectrometry, vol. 2 (Holden Day: San Francisco 1964)).

The o.r.d. curve of idiospermolide showed a position Cotton Effect (a = +1907°). This is analogous to lophenone and lophanone which both have a single methyl group at C4 which is in the equatorial or thermodynamically favoured conformation and which both show positive Cotton Effects: a = + 1700° and a = + 1300° respectively. This supports the conclusion that there is only one methyl group at C4 which is in the equatorial conformation.

Sodium borohydride reduction of idiospermolide gave a mixture of epimeric alcohols from which only the major product could be isolated. It had the hydroxyl group in the 3β position. This is in agreement with the observation that cycloeucalenone, lophenone and cirtrostanone, each of which has a single equatorial methyl group at C4, give 3β-alcohols on reduction with lithium aluminum hydride (G. Ourisson, P Crabbé and O.R. Rodig, Tetracyclic Triterpenes (Hermann: Paris 1964)).

The n.m.r. spectrum of this alcohol as well as Eu(dpm)3 shift studies were consistent with the proposed structure for idiospermolide.

It was decided to degrade idiospermolide and correlate it with a known substance. The reaction of idiospermolide with potassium t-butoxide in refluxing t-butanol resulted in ring opening of the δ-lactone function to give a diene acid. This was characterized as both the methyl ester and the saturated methyl ester. The saturated ester was reduced with lithium aluminium hydride to the diol and the primary alcohol selectively tosylated. Reduction with lithium aluminium hydride then gave 29-norcylcoartnaol. The latter was found to be identical with a sample prepared by the ozonolysis of the acetate of cylcoeucalenol followed by sodium borohydride reduction, tosylation and reduction with lithium aluminium hydride.

With respect to the absolute configuration at C22, the proton on C22 has vicinal coupling constants of 12.8 Hz and 3.5 Hz (cf parasorbic acid which has vicinal coupling constants of 10.3 Hz and 5.4 Hz).

If this proton were pseudo equatorial, the coupling constants would have been nearly equal. Hence the large constituent at C22 is pseudo equatorial. Now the sign of the ene-lactone n →π* Cotton Effect near 250 nm in the C.D. curve of α,β-unsaturated δ-lactones have been shown to give the absolute configuration at the carbon atom adjacent to the ring oxygen (G. Snatzke, Angew. Chem. Int. Ed., 1968, 7, p14; A.F. Beecham, Tetrahedron, 1972, 28, p5543). Since the C.D. curve of idiospermolide has Δε253 = -2.2, the chirality in the pentenolide ring is opposite to that in parasorbic acid and the absolute configuration at C22 is S. This absolute configuration at C22 of idiospermolide is opposite to that found in the withanolides.

The spectral data for idiospermine

Since this work was carried out in the early seventies the technology used reflects that era. The nmr spectrum was recorded on a Varian XL-100 spectrometer.

The nmr spectrum of idiospermine showed several characteristic features: twelve protons in the aromatic region; three one-proton doublets in the range δ 4.6 – δ 4.2; four three-proton singlets indicative of four N-Me groups; and three protons upfield at δ 1.30.

The nmr spectrum showed the following signals:

Doublet, 1H, δ 7.57 (J 6 Hz)
Multiplet, 11H, δ 7.35 – δ 6.20
Doublet, 1H, δ 4.64 (J 4.5 Hz)
Doublet, 1H, δ 4.31 (J 4.5 Hz)
Doublet, 1H, δ 4.19 (J 4.5 Hz)
Singlet, 3H, δ 3.56
Singlet, 3H, δ2.45
Singlet, 3H, δ2.40
Singlet, 3H, δ2.34
Multiplet, 10H, δ 3.46 – δ 2.04
Doublet, 2H, δ 1.30 (J 12.5 Hz)
Broad signal, 1H, δ 1.24

When the spectrum was run at 40 degrees C the one-proton signal at δ 1.24 had shifted to δ 1.30 and when the temperature was 60 degrees C it had shifted back to δ 1.24 while a one-proton doublet (J 4.5 Hz) moved out of the aromatic region upfield to δ 5.99 and the one-proton doublet at δ 4.64 became broader. When the spectrum was run at 70 degrees C the doublet which had moved out of the aromatic region had moved to δ 5.85 and had become broad, as did the doublet originally at δ 4.64 which moved to δ 4.56. These experiments suggested the presence of three N-H groups.

Exchange with deuterium oxide resulted in the disappearance of one of the signals in the range δ 7.35 - δ 6.20, and of the signals at δ 4.64 and δ 1.24; the doublets at δ 4.64 and δ 4.31 also collapsed to singlets. This confirms the presence of three N-H groups and two NH-CH-N residues of the type found in the calycanthaceous alkaloids.

The upfield signal at δ 1.30 has the characteristic position and appearance of the signal of the C3’-Heq proton in calycanthine. Chimonanthine type alkaloids do not show upfield signals in their nmr spectra (J.B Hendrickson, R. Göschke and R. Rees, Tetrahedron, vol 20 (1964), p565; E.S. Hall, F. McCapra and A.I. Scott, Tetrahedron, vol 23 (1967), p4131). The aliphatic region of the spectrum run at 70 degrees C was more clearly resolved; integration showed the presence due to four additional protons but, nevertheless, the spectrum closely resembled the corresponding region of the spectrum of calycanthine, but was doubled up. In particular, the doubling up of the triplet of doublets, which in calycanthine arises from C3-Hax and C3’’’-Hax, can be clearly seen. This doubling up of the signals for the protons of the ethylamine bridges of calycanthine must presumably be due to an unsymmetrical influence in the molecule.

If there is, in fact, a calycanthine moiety present, then the characteristic doublet of doublets (o- and m-coupling) for the upfield aromatic protons C8-H and C8’’-H would be expected to be visible in the spectrum of idiospermine. In fact, the spectrum shows a doublet at δ 6.26 (J 8.0 Hz) superimposed on a doublet of doublets (J 8.0 Hz, J 1.2 Hz) integrating for two protons. On the reasonable assumption that there is a calycanthine moiety present, then it follows that one aromatic ring has a 6-substituent to remove meta coupling.

Accepting this hypothesis, it remains to add to the calycanthine molecule at the C6 position the elements of a unit of N-methyltryptamine and one N-methyl group. The nature of this moiety emerged from a consideration of the uv data. The uv spectrum of idiospermine was different from those of the calycanthaceous alkaloids in several respects; in particular, there was an absorption at 302 nm of greatly enhance intensity (ε 32400) compared with an absorption at 310 nm (ε 5350) in the uv spectrum of calycanthine. This can be most satisfactorily understood if idiospermine contains a 2-arylindole residue.

The structure proposed for idiospermine satisfies the requirements of the nmr spectral data as follows: there are only two NH-CH-N protons as required; the aliphatic N-H probably resonates upfield; the three-proton singlet at δ 3.56 has a chemical shift which is very close to that for protons of the methyl group in N-methylindole (δ 3.60); the downfield proton in the aromatic region (δ 7.57) is suggestive of a proton at C7 of an indole nucleus.

The mass spectrum of idiospermine did not show the characteristic fragmentations of the calycanthaceous alkaloids (eg the loss of C3H8N from the ethylamine bridge of calycanthine). The base peak was at P-43, corresponding to the loss of C2H5N. This can be interpreted in terms of the presence of an N-methyltryptamine unit, giving rise to a McLafferty rearrangement.

Support for the tryptamine unit was obtained by examining the nmr spectrum with the presence of Eu(dpm)3. It was found that addition of 10-3M Eu(dpm)3 to a 4 x 10-3M chloroform solution of idiospermine caused a downfield shift of signals due to the tryptamine side chain leaving the calycanthine pattern more clearly visible.

The chemical constituents of Idiospermum australiense (Diels) S.T. Blake

This blog has been set up so that results of the study of the chemical constituents of Idiospermum australiense that we carried out at Sydney University in 1972 and 1973 may be more widely accessed. This blog is also a testimony to the pioneering work into natural product chemistry in Australia by the late Prof Ern Ritchie and the late Prof Wal Taylor who were my supervisors. Prof Taylor passed away just under a year ago and was working towards publishing the results of my study. In the event these results are not actually published in a recognized scientific journal then this blog makes available some the results that can otherwise be sourced from my thesis. It is well documented that in 1912 Diels described the tree as Calycanthus australiense Diels from a specimen collected in northern Queensland at Harvey Creek near the Russell River. Then in 1971 the plant was rediscovered in the Daintree area after a case of stock poisoning. T.S. Blake, a taxonomist with the Queensland herbarium, reclassified the plant into a new genus and family, Idiospermum australiense, Idiospermaceae.

This first post will briefly describe the results concerning the alkaloids found in the plant.
The bark, cotyledons, wood and leaves all yielded (+)-calycanthine
Bark: 0.1% calycanthine
Cotyledons: 0.05% calycanthine
Wood: 0.03% calycanthine
Leaves: 0.38% crude alkaloid fraction (calycanthine plus another alkaloid)

This result is significant in that it indicates that the plant is closely related to the calycanthaceae since according to Hegenauer, members of this family characteristically contain calycanthine type alkaloids (Hegnauer, R., Chemotaxonomie der Pflanzen, Band 3, p338 (Birkhauser Verlag: Basel 1964).

After exhaustive counter current separation of the alkaloid fraction from the leaves of the plant a new alkaloid, idiospermine (0.002%) was isolated.

Idiospermine (C34H40N6) can be viewed as a trimer of methyltryptamine plus an additional methyl group. From spectral data it was concluded that its structure is that of a calycanthine moiety linked via one of its para positions to the alpha position of methyltryptamine where the indole nitrogen has been methylated.

I moved on from chemistry in 1973 and I hope that in a future post I can obtain the assistance of a friend to post the actual chemical structures.

The uv spectrum displayed the presence of an alpha-aryl indole chromophore. Prof Taylor subsequently confirmed the structure via X-ray crystallography carried out by Prof Trevor Hambley of Sydney University.

Idiospermine is different from idiospermuline (extracted from the seeds of this plant) which is a chimonanthine analogue - Rujee K. Duke, Robin D. Allan, Graham A.R. Johnston, Kenneth N. Mewett and Ann D. Mitrovic, "Idiospermuline, a trimeric pyrrolidinoindoline alkaloid from the seed of Idiospermum australiense, Journal of Natural Products, vol 58 (no.8), 1995, pp1200-1208.

Idiospermine is, however, similar to idiospermamine B recently reported by Anthony R. Carroll, Gregory A. Feshner, Sandra Duffy and Vicky M. Avery at the 50th Annual Meeting of the American Society of Pharmocognosy held in Hawaii in June of 2009. This alkaloid was isolated from the bark of the plant and displays antimalarial, antitrypanosomal and cytoxic activities.

The structure of idiospermamine B ( C34H38N6) is that of a calycanthine moiety coupled at the beta position of methytryptamine followed by cyclisation (formation of a pyrrolidine ring). This work from researchers at Griffith University is about to be published in the very near future.

In 2001 Bremner published results into study of a Psychotria sp from Lombok, Indonesia - http://www.mdpi.com/1420-3049/6/2/117/pdf. He referred to an alkaloid with a molecular weight of 574. My guess is that this compund has the molecular formula ( C37H46N6). Bremner does not seem to have published a structure for this alkaloid but my guess is that it is a trimethylated compound from idiospermine.