Full text of DOI:10.1016/0031-9422(71)85024-0

1 9 7 1 Vol.
1 5 5 9 t o 1 5 6 7 .
Press. Printed in England.
CONSTITUENTS OF SURIANA
A TRITERP ENE DIOL OF NOVEL STRUCTURE AND A NEW
F LAVONOL GLYCOSIDE
R. E. MITCHELL and T. A.
Department of Chemistry, University of California, Los Angeles, California 90024, U.S.A.
October 1970)
Abstract-Suriana maritima L., often classified in Simaroubaceae, but regarded also as the monotypic
member of Surianaceae, is devoid of the
lactones that characterize Simaroubaceae. It contains a
novel triterpenoid diol, surianol, the structure of which is described. In addition to this,
and rhamnetin-3-rutinoside, the latter a new flavonol glycoside, were also found as constituents of the plant.
INTRODUCTION
L., a pantropical shrub or small tree common to coral coasts of
Indian Ocean islands and to the coast of Florida, has had an uncertain taxonomic history.
Although the most recent classifications placed it in a monotypic family, Surianaceae, allied
to Stylobasiae and Simaroubaceae,
after a consideration of earlier dis-
cussions, accepts the classification in Simaroubaceae.
In view of the large body of evidence that simaroubaceous genera are uniformly charac-
terized by their content of bitter lactones allied to such now well known compounds as
quassin, chaparrin, glaucarubol and many
it appeared likely that an investigation
of its chemistry might provide information to support one or the other of the conflicting
taxonomic opinions.
RESULTS AND DISCUSSION
Suriana maritima, collected in
methanol.’ Concentration of the initial extract and dilution with water resulted in the
deposition of a generous quantity of the dry wt. of plant) of a yellow crystalline
material, readily recognized as flavonoid in character. Several of this
material led to its separation into two pure (by TLC) components. One of these was identi-
fied as (I), the other as (rhamnetin 3-rutinoside) (II). Acid hydrolysis
was dried and powdered and extracted with
of I and II led to the formation, respectively, of quercetin and rhamnetin, characterized as
the acetates. Both I and II gave rhamnose and glucose upon hydrolysis, and methylation
* Contribution No. 2732 from the Department of Chemistry, University of California, Los Angeles,
California 90024, U.S.A.
We are grateful to Dr. William
for his generosity in collecting the plant material used in this work.
A.
The Evolution and Classification of Flowering Plants, Houghton-Mifflin, Boston (1968).
J. C. W ILLIS , A Dictionary of the Flowering Plants and Ferns (7th edition rbeyviHse.dK.
Cambridge University Press, Cambridge (1966).
H. P.
For examples,
Neither the ground plant material nor the extract was bitter, a fact which has been noted by others and
used in arguments to support early classifications. It is to be noted, however, that
(Simaroubaceae), recently shown to contain glaucarubol
MAN, Phytochem. 8, 1565
Flora
J. POLONSKY , Les
Ser. Z 6, 193 (1962).
Amers des
Med.,
107 (1966).
L. B. DE
and T. A.
lacks the bitterness common to other simaroubaceous species.
1559

R. E. MITCHELL and T. A.
1560
and hydrolysis of I and II gave the same compound,
The
trimethylsilyl derivative of I gave an NMR spectrum identical with that described for
The trimethylsilyl derivative of II gave an NMR spectrum essentially identical in all details
with that of I (and except for the appearance of a three-proton singlet at for the
7-methoxyl group. These data establish the identity of the new flavonol as rhamnetin
3-rutinoside (II).
0-glucose-rhqmnose
R=H
Concentration of the residual methanolic solution and extraction with light petroleum
afforded an extract which contained the bulk of the total extractives; it was not bitter.
Repartition of the petroleum-extractives between pentane and methanol-water (1: 1) gave a
pentane solution which was closely examined.
Chromatography of the pentane solution over silica gel yielded, first,
in later fractions, surianol, (III), m.p. (high resolution mass spectrum;
molecular ion at m/e 428). Surianol readily formed a diacetate (IV), the analysis and mass
spectrum of which agreed with the constitution That surianol was a triterpenoid
was indicated by the mass and NMR spectra. The latter showed two tertiary methyl groups
and
and the geminal
doublet
were seen at
groupings appearing as a triplet = 9 Hz) at
the addition of the latter signal had the appearance of a doublet of triplets at S
with
grouping of an isopropyl group, which appeared as a 6-proton
=
Hz) at
ppm. Two hydroxyl protons (lost after exchange with
the two non-equivalent carbinol protons of the secondary alcohol
and a broad multiplet at 3.43. After
and
=
9 and 4 Hz. Two one-proton doublets at
and
ppm = 4 Hz)
were indicative of a cyclopropyl grouping bearing two geminal protons only.’ Two
proton (broadened) singlets at
group.
and
indicated the presence of a terminal methylene
The mass spectrum showed a fragment ion at
loss of the side-chain, which can be formulated as the 24-methylenecholestane side-chain. In
accord with this, the dihydrodiacetate showed a fragment ion at The placing of the
which can be accounted for by
side-chain methylene group at C-24125 is supported by the appearance in the mass spectra of
III and its derivatives of a fragment M+-83, ascribed to the ion resulting from the loss of the
* The constants observed for this compound were in satisfactory agreement with those of
but
the mass spectrum disclosed a small peak at m/e 416
amount of a dihydro compound, probably stigmastanol.
=
indicating the presence of a small
T. J . MABRY, J . KAGAN and H.
(15 September 1964).
Analysis of
G.
Univ. of Texas
and H. PI O T R O W S K A, Bull.
No. 6418
G.
B. MAC C H I A, A.
France 2359 (1964).

Constituents of
maritima
1 5 6 1
allylic radical
from the side chain by cleavage at the C-22/23 bond. In agreement with
this view, the dihydro compound showed no corresponding ion.
345
The placing of the cyclopropane ring at the
is, of course, favored on bio-
genetic grounds, for numerous triterpenes containing this structure are known. Support for
this assignment is found in the fragmentation patterns of I and its derivatives. Surianol, and
derivatives bearing the methylene group at C-24, show fragment ions at m/e 300 and m/e
175 (loss of side chain from m/e 300).
(e.g. VI) characteristically show
a fragment a (Scheme I) in their mass spectra
The formation of fragments b (from III)
and c (from the dihydro derivatives) are consistent with this mode of fragmentation. More-
over, the formation of b and c make it necessary to conclude that the two hydroxyl groups
are found in the A ring, and that no carbon atoms (e.g.
are attached to ring A. Thus,
ion fragment b contains all of the methyl groups present in the parent compound.
In summary, all of the above evidence leads to the proposal of structure III for surianol.
The placing of the methyl groups at C-13 and C-14 is on biogenetic grounds, but it will be
seen that no other reasonable accommodation can be made for other skeletal dispositions
of the methyl groups in the fragment ions b and c.
Additional evidence for the structure of ring A was found in the following way. Surianol
formed an acetonide, showing the vicinal disposition of the hydroxyl groups. The i.r.
spectrum of III in dilute carbon tetrachloride solution showed free and intramolecularly
bonded hydroxyl stretching vibrations at 3610 and 3570 cm-‘, values typical of a-glycols.
Oxidation of III with chromic acid (e.g. Jones’ reagent) gave varying results, depending upon
the conditions used. One product was a dicarboxylic acid, formulated as VII.
Oxidation to the dicarboxylic acid appears to proceed through the two intermediate
a-ketols as shown by TLC, using as a reference standard one of the a-ketols isolated in an
alternative oxidation in which chromic acid in acetic acid was used. The
ester
(VIII) of the acid showed two carbomethoxy groups in its NMR spectrum, and a mass
spectral molecular ion at m/e 486. A high resolution mass determination agreed closely with
the expected composition
In addition to the dicarboxylic acid, oxidation yielded
an a-ketol which was easily oxidized with bismuth trioxide in acetic acid. The mixture of
products, which could not be separated nor individually characterized, showed mass spectral
molecular ions at m/e 424 (for a diosphenol) and m/e 422, the latter being assumed to be the
(X) resulting from further oxidation of the
(IX) (diosphenol).
Support for this interpretation is found in the spectrum of the mixture, which showed
H. E.
R . B EUGELMANS a n d B. C.
Tetrahedron Letters 4341 (1966).

1.562
R. E. MITCHELL and T. A.
a
26.6
m/e 300
m/e 302
from dihydro
SCHEME 1.
at 310 nm, shifted to 352 nm after addition of alkali, a shift characteristic of
sphenol. Although the usual range of absorption for the diosphenol chromophore is
270-280 nm, the absorption of a chromophore conjugated with a cyclopropane ring often
undergoes a bathochromic shift which is normally
but which has been reported
to be as high as 40 nm in the structure
269 nm
The formation of a dehydrodiosphenol is regarded as supporting evidence for an A-ring
diol system in a skeleton in which the tertiary methyl groups are found
at the C/D ring junction. Oxidation of diols in which the hydroxyl groups were at
* When a TLC plate of the mixture was sprayed with ferric chloride solution a dark spot appeared at the
position to which the compound(s) migrated.
I. A.
Interpretation of Ultraviolet Spectra of
Pergamon Press, Oxford (1964).

Constituents of
maritima
1563
I
or
would not proceed past the
stage (short of more extensive
degradation). It is compelling on biogenetic grounds to place one of the hydroxyl groups at
thus leaving a choice between a and a The carbinol proton signals in the
NMR spectrum of III are consistent only with the where the atoms
are disposed, thus accounting for the triplet with 9 Hz coupling observed
for the C-4 proton, and the doublet (4 Hz) of triplets Hz) for the C-3 proton.
Although certain other stereochemical features of surianol are not established with
=
certainty, it seems
that the stereochemistry shown in III is correct. Surianol has
ROOC
surianol
439
euphorbol
S
CHEME II.

R. E. MITCHELL and T. A. GEI SSMAN
1564
comparable to other compounds of this class (cf. XII and XIII), but quite
+
different from those recorded for compounds of the euphol (XIV) or tirucallol series.
Although the occurrence in nature of a triterpene is not new (e.g.
the is unique in the series. It is noteworthy, in respect to the
taxonomic question to which this study was originally addressed, that surianol belongs to a
stereochemical series different from that which includes the triterpenes (euphol or tirucallol)
from which the simaroubaceous lactones are presumably derived.
EXPERIMENTAL
were determined on a
melting point apparatus and are corrected.
spectra were taken as
nujol mulls unless otherwise specified; NMR spectra were recorded in
solution unless otherwise
specified and peak positions are given in values with tetramethylsilane as internal standard; mass spectra
were obtained on AEI-MS-9 and CEC-MS-21-491 spectrometers, with direct inlet, at 70
potential, and significant ions of higher m/e values as well as ions with relative abundance of 40
the base peak are quoted with their relative abundance in parenthesis. TLC was run on
ionization
or more of
Merck
plates (5 10, or 5 20 cm when the adsorbate was particularly complex), visualization being satis-
factory after spraying with and heating in an oven at 120” for a few min; column
grams used Baker silica gel 60-200 mesh (30 g of adsorbate) packed in hexane or benzene and were
eluted with the appropriate solvent (usually increasing portions of then increasing portions of acetone
their progress always being monitored by TLC analysis of the fractions collected. Analytical samples
in
were dried at
mm or 1
maritima. 1 kg of dried powdered plants was processed as recorded in the summar-
Each extract was separately examined on TLC and since the and
mm for 16 hr immediately prior to analysis.
Extraction
ized form of Scheme
methanol extracts appeared the same they were combined.
was obtained as yellow needles after recrystallization of the crude yellow solid
TLC
several times from methanol (filtered hot); it was less soluble in methanol than
after development with
10); red color in
methylsilyl derivative) 7.45 (H-6’ doublet of doublets, = 2, 8 Hz),
(H-2’, d, = 2 Hz),
(H-5’, d,
8 Hz), 6.48 (H-8, d,
427 (rhamnose H-l,
(shoulder)
=
Hz), 6.23 (H-6, d,
and
=
Hz), 5.82 (glucose H-l, broadened d, = 6 Hz),
carbohydrates protons);
(log 257
265
297 (3.92) and 361 (4.27) nm;
238
282 (3.87) and 307 (3.91) nm.
Hydrolysis of this compound (250 mg) proceeded by refluxing for 2
10 aq.
in pyridine (2 ml) and acetylated with acetic anhydride
The reaction was poured into iced 2N HCI and the product extracted with
extract was washed portions of 2N HCI, then saturated solution, and after drying and filtration
was evaporated, affording colorless needles (1.50 mg) after recrystallization of the residue from ethanol,
in methanol (10 ml) containing
(10 ml). Cooling and filtration afforded a yellow crystalline solid which was dried, dissolved
ml) by standing at room temperature for 24 hr.
the combined
m.p.
(lit.”
with spectral properties
NMR and mass spectra) consistent for rhamnetin tetraacetate
192-193’). Anal. Found: C, 59.58; H, 444;
for
C, 59.50; H,
aq.
Hydrolysis
(5 ml) gave
295298”
was
of this compound (100 mg) by refluxing it for 3 hr in methanol (7.5 ml) containing
a yellow crystalline product which was filtered from the cooled reaction mixture; 58 mg,
(dec.) (lit.” for rhamnetin 294-296”). The aqueous solution remaining after hydrolysis of the
neutralized
solution), filtered, then studied on thin layer chromatograms of silica gel G (Merck)
whereby two components were observed, these corresponding in color and in
when co-chromatographed with each of these and compared with
to glucose and rhamnose
mixtures.
Methylation of
mg) was accomplished by stirring it with anhydrous
(150 mg) in refluxing
acetone (20 ml) containing an excess of
and refluxed for 1
ml) for 16 hr. The reaction mixture was then evaporated
with
aq.
(10 ml). A cream-colored solid was filtered from the cooled solution,
dried, then twice recrystallized from ethanol, affording yellow needles (39 mg),
and mixed m.p. with
authentic
prepared in the same way from
spectrum identical with that of authentic material
(I) was obtained from the mother-liquor remaining after the initial recrystallization of the crude
yellow solid. Several successive concentrations of, and recrystallization from, this mother-liquor, led to a
* The chromatograms were developed with n-propyl
with an anisaldehyde spray and heated to show green and blue-black spots, on a pink background, for
nose and glucose which had values of and0.48, respectively, when chromatographed as a mixture.
See E. STAHL, Thin-layer Chromatography, Springer Verlag, Berlin (1965).
acetate-water
sprayed
C.
T. A.
J. C. KNIGHT and D. I. WILKINSON, Am. Chem.
85, 835 (1963).
The Chemistry of Flaconoid Compounds, Pergamon Press, Oxford (1962).

Constituents of
1567
and 41 (70);
(d,
=
Hz),
and 4.62.
(t, = 9 Hz),
(with the appearance of a
doublet of multiplets with = 9 Hz),
Oxidation of Surianol
(i) With chromium trioxide. Surianol (III) (50 mg, 0.117
in a solution of hot acetic acid (5 ml), was
ml, excess) of in (v/v) aq.
heated on a steambath for 1 hr with a 10% (v/v) solution
acetic acid, and the reaction mixture was then evaporated. The residue was slurried with aq.
and then evaporated, the salts remaining being dried by azeotropic distillation with benzene and extracted
several times with
column x 8 cm) of silica gel
The combined extracts were evaporated and cbromatographed on a
g) packed in benzene and eluted with benzene. Thus, fractions were
obtained which showed on TLC two spots of closely similar
reacted compound III) and the product-component having the smaller
colorless crystals, by recrystallization from ethanol;
880 cm-‘; m/e 426 411 (M+-15)
301 (M+-125) (26). 175 (12). 126 95 81
was tentatively. identified one of the two expected
mixture are probably the two a-ketols. Oxidation of the ‘a-ketol’ mixture
(25 mg) in refluxing acetic acid (1 ml) for hr, evaporation of the solution, and extraction of the residue
with led to a crude product which was by column chromatography and from
(while later fractions consisted of (TLC)
was obtained in small amount, as
3400
383 (M+-43)
57 55
3055, 3020, 1715
1640 and
342
343 (M+-83)
69
43 (77) and 41 (100). This
a-ketols; the two components of the crude
20 mg) with bismuth trioxide
ethanol to give a small quantity of solid, apparently consisting of two components (TLC) as indicated by a
mass spectrum which had m/e 424 (24) 422
310 shifting to 352
(ii) With the Sarett
299 (424-125) (7) and 297 (422-125) (4). This was not further
after the addition of two drops of 2N
purified; it had
Much the same result as in (i) was obtained. Surianol (III) (40 mg) in a
solution of pyridine (2 ml) was added to a solution of Sarett reagent (43 mg, ca. 25% excess) in pyridine
(2 ml), and after standing at room temp. for 16 hr the solution was poured into
then washed with followed by saturated solution, dried
(25 ml) which was
and evaporated. The crude
product contained unreacted compound III and the ‘a-ketol’ mixture (by TLC comparison with the products
in (i)).
(iii) Jones reagent:“’ A solution of surianol (III) (25 mg) in acetone (3 ml) was stirred at room temp.
during the slow
addition of Jones reagent until the orange color of the reagent persisted and no
compound III remained (TLC). A main component of lower
compound III having on the same chromatogram) had formed, and an i.r. spectrum
of the crude product indicated carboxylic acid material. was methylated with in in an
g silica gel) packed in and eluted by
mg, m.p. 90-100” but TLC homogeneous:
after development with
bath, and the crude product purified on a column chromatogram
benzene, and
crystallized from ethanol to give
(0.005
1735 (ester
m/e 386.3703
for
57
486.3709; m/e486
43 (78) and 41 (86); (re-
471 (M+-15)
361
83
and
81
69
55
crystallization mother liquor)
(carbomethoxyl
Oxidation of dihydrosurianol (50 mg) with Jones reagent in acetone (10 ml) at ice-bath temp. was
monitored by TLC, with compound III and the a-ketol derived from III as reference standards (these having
the same R, values as the corresponding dihydro compounds). After 30 min and addition of three drops of
Jones reagent most of the diiydrosurianol bad disappeared with the appearance of two components (nearly
identical
appeared on the addition of more Jones reagent, with the concomitant appearance of a very polar component
(and no component with higher than the a-ketol as expected for a diketone) corresponding to the diacid
with TLC
the same as those of the ‘a-ketol’ mixture previously encountered; these dis-
previously encountered. This was methylated and isolated as the diester as previously; unfortunately the
quantity was insufficient for a NMR spectrum of high enough precision to provide information on the nature
of the protons a to the carbonyl groups; m.p.
2 mg colorless needles.
Acknowledgement-This study was supported by a research grant,
Service. We thank Dr. Arthur Cronquist for helpful discussion and Dr. William
material used in the work. Analyses are by Miss Heather King, U.C.L.A.
from the U.S. Public Health
for providing the plant
G. I. Poos, G. E.
K. I. M.
R. E. BEYLER and L. H.
J. Am.
75,422 (1953).
E. R. H. J ONES and B. C. L.
J.
39, (1946).