H. Zhang et al. / Tetrahedron 67 (2011) 2591e2595
2593
data for 1 revealed key correlations that defined the complete
relative configuration. More specifically, ROESY correlations (Fig. 1)
HRESI(ꢀ)MS analysis of fascioquinol F (8) revealed a pseudo-mo-
lecular ion (MꢀH)ꢀ consistent with a molecular formula C26H36O2
established that H-5, H-9 and H-14 were positioned one (
while H3-26, H3-23, H3-24 and H2-15 were positioned on the op-
posite ( ) face of the fused tricyclic system. Thus the complete
a
) face,
(Dmmu ꢀ0.2), while the NMR (CDCl3) data (Supplementary data,
Table S8) disclosed resonances (dH 5.58, d, 9.8 Hz, H-10; 6.25, d,
9.8 Hz, H-20; 1.35, s, H3-170) suggestive of a chromene moiety. The
planar structure assigned to 8 was confirmed by spectroscopic com-
parison to the closely related marine brown alga metabolite dictyo-
chromenol8 and the sponge metabolite 2-(hexaprenylmethyl)-2-
methylchromenol.9 As the latter two natural products exhibited small
b
relative stereostructure for fascioquinol A (1) could be assigned as
indicated.
During HPLC fractionation of 1 it proved advantageous to em-
ploy an H2O/MeCN gradient, with a constant 0.01% TFA modifier to
enhance peak shape and improve resolution. The benefits of a TFA
modifier notwithstanding, during concentration in vacuo (freeze
drying) samples of 1 underwent acid mediated degradation to yield
3e5 (structure assignments detailed below). The proposition that
the cyclic ethers 3e5 were handling artifacts was further supported
by acid hydrolysis studies (see Experimental section), and the fact
that formation could be avoided by either neutralizing HPLC frac-
tions with sodium bicarbonate prior to concentration in vacuo, or
by avoiding HPLC in favour of gel chromatography (Sephadex LH-
20) in neutral solvent (MeOH).
but positive [
the near zero [
a
]D measurements (þ4andþ3.6(c1, CHCl3), respectively),
a]
D
22 þ 0.06 (c 0.58, CHCl3) value for 8 was taken as evi-
dence of a near racemic mixture. Literature reports of 8 are limited to
a single occurrence in a 1989 patent as a synthetic anti-ulcer agent,10
with the current report being the first account of 8 as a natural product.
In an attempt to explain the mechanism for the acid mediated
transformations noted above we propose (Fig. 2) that following
hydrolysis of the sulfate 1 to the phenol 2, cyclization of 2 to 3e5
proceeds via initial protonation of D12,13 to yield a C-13 tertiary
carbocation. This intermediate carbocation then undergoes either
(i) pro R facial SN1 addition by the adjacent 17-OH to yield 4, or (ii)
pro S facial SN1 addition to yield 5, or (iii) stereocontrolled intra-
HRESI(ꢀ)MS analysis of fascioquinol B (2) revealed a pseudo-
molecular ion (MꢀH)ꢀ attributed to a molecular formula (C26H38O2,
D
mmu ꢀ0.5) consistent with a desulfated analogue of 1. Supportive
molecular 1,2-hydride transfer (across the
tertiary carbocation which in turn undergoes SN1 addition by the
adjacent 17-OH from the less sterically hindered face. This latter
stereocontrolled 1,2-hydride transfer and facial attack suggest
a face) to yield a C-14
of such a structure assignment, the 1H NMR (methanol-d4) data for
2 (Supplementary data, Table S2) was almost identical to that for 1,
with the significant exception being an w0.5 ppm shielding of
resonances for H-19 and H-21, which now flanked a phenol rather
than a sulfate functionality. Likewise, diagnostic 13C NMR chemical
shift differences between 1 and 2 (C-17 DdC ꢀ4.4, C-19 DdC ꢀ6.9, C-
20 DdCþ4.6, C-21 DdC ꢀ6.9) could be attributed to hydrolysis of the
C-20 sulfate moiety to a phenol.
a
a
assignment of the C-13/C-14 relative configuration to fascioquinol C
(3) as indicated. Furthermore, given the proposed mechanistic
relationship to 5, a sponge metabolite of known absolute configu-
ration,5 we propose that 1e4 belong to the same antipodal series.
The natural product meroterpenes 1 and 6e8 were assessed in
cytotoxicity and cell proliferation assays against human gastric (AGS)
HRESI(ꢁ)MS analysis of the acid degradation products 3e5
revealed pseudo-molecular ions consistent with molecular for-
mulae (C26H38O2,
D
mmu ꢀ0.6, ꢀ1.1 and 0.6, respectively) isomeric
with 2. Detailed analysis of the NMR (methanol-d4) data for 3e5
(Supplementary data, Tables S3eS5) revealed a loss of the D12,13
moiety with concomitant ring closure to yield cyclic ethers. For
example, analysis of the NMR data for fascioquinol C (3) revealed
diagnostic resonances for a secondary aliphatic H3-25 methyl
(dH 0.68, d, 6.5 Hz), a quaternary oxygenated C-14 (dC 96.2), and an
isolated H2-15 benzylic spin system (dH 2.73, d, 16.6 Hz, Hb-15; 3.19,
d, 16.6 Hz, Ha-15), with strong ROESY correlations observed be-
tween Hb-15 and H3-25, and between Ha-15 and H3-24dconsistent
with the five member cyclic ether as indicated (with the C-13 and
C-14 relative stereochemistry assigned on mechanistic grounds as
indicated below). Similar analysis of the NMR data for 4 and 5
confirmed retention of H-14 (4 dH 1.26, d, 8.2 Hz; 5 dH 1.61, m), and
the appearance of quaternary oxygenated C-13 (4 dC 75.3; 5 dC 76.8)
and tertiary aliphatic H3-25 methyl (4 dH 1.11, s; 5 dH 1.15, s) reso-
nances, consistent with the C-13 epimeric six member cyclic ethers
as indicated. On reviewing the scientific literature 5 proved to be
identical in all respects, including absolute configuration, with the
known marine sponge metabolite strongylophorine-22 first
reported in 2003 by Miyaoka et al. from an Okinawan specimen of
Petrosia (Strongylophora) corticata.5 Based on this comparison, fas-
cioquinol D (4) was assigned as the C-13 epimer as indicated.
HRESI(ꢀ)MS analysis of fascioquinol E (6) revealed a highest
mass ion isomeric with 1, whereas the NMR (methanol-d4) data
(Supplementary data, Table S6a) disclosed a simplified E,E,E-ger-
anylgeranyl side chain. Consistent with this interpretation, acid
mediated hydrolysis of 6 yielded the known sponge metabolite
geranylgeranyl 1,4-hydroquinone (7).6 Comparison of NMR shifts
between 6 and 7 (as discussed above for 1 and 2) permitted structure
assignment of the meroterpene sulfate fascioquinol E (6) as in-
dicated. Literature reports on 6 are limited to a single occurrence in
a 1989 patent as a synthetic reversetranscriptase inhibitor,7 with the
current report being the first account of 6 as a natural product.
Fig. 2. Biosynthetic and acid mediated relationships of 1e8.