Mono- and dicyclic unsaturated triterpenoid hydrocarbons in ...

890 R. de Mesmay et al. / Organic Geochemistry 39 (2008) 879–893
(f1, f2 and h), ii) masokocenes (i.e. dicyclic botryococcenes:
c, d, e and j) and iii) partially reduced botryococcenes
(a, b, g1, g2 and i, Fig. 9). It is noteworthy that all the
monocyclic botryococcenes and their derivatives have a
cyclohexenyl moiety in the left hand side of the molecule.
No monocyclic compound with one ring on the right side
of the molecule has been characterized in Lake Masoko
sediments, whereas the occurrence of dicyclic compounds
indicates that both sides of the molecule can be cyclized.
Except for a, all the structures in Lake Masoko are cyclic,
whereas botryococcenes found in sediments or pure strain
cultures are mostly acyclic (Metzger and Largeau, 1999). B.
braunii microalgae in Lake Masoko 32,000 years ago produced
nearly exclusively cyclic botryococcenes, which is
quite intriguing. The virtual lack in this sample of acyclic
botryococcenes, likely to be the precursors of cyclic botryococcenes,
suggests an efficient mechanism of cyclisation.
The absence of compound k and of the monocyclic C 35 botryococcene,
probable precursors for masokocenes c and e,
respectively, means that the second cyclisation is also efficient.
The occurrence of f1 and f2 suggests that their cyclisation
to j is probably less efficient due to the steric effect
of the methyl and ethyl groups on the C-21/C-22 double
bond. The stronger steric effect in h may prevent further
cyclisation to C37 masokocene.
In all partially reduced botryococcenes in the sample (i.e.
a, b, g1, g2 and i), the most easily reducible double bond (C-
26/C-27) is left unchanged compared with the corresponding
botryococcene. Moreover, only one peak for each compound
a and b was detected from GC analysis (Fig. 2A), suggesting
the formation of only one diastereoisomer for a and b via biotic
reduction of parent botryococcenes. However, in the light
of the recent work of Hebting et al. (2006) on the preservation
pathway of sedimentary organic carbon, an abiotic process
cannot be entirely excluded. For the other partially reduced
botryococcenes g1 and g2, we attribute the occurrence of
two stereoisomers to the Z and E stereochemistry of the C-
21/C-22 double bond rather than to hypothetical diastereisomers
that would be formed by an abiotic process (Hebting
et al., 2006). We then assess that all partially reduced compounds
in the ca. 32,000 year old sediments from Lake Masoko
arise from a biotic process. Huang and Murray (1995)
and Huang et al. (1996) reported similar observations on
some reduced botryococcenes found in sediment from
Sacred Lake (Kenya), and suggested that they could originate
either from a variant population of B. braunii race B or from a
microbial reduction. Moreover, the possibility of a microbial
reduction during early diagenesis was recently proposed to
explain the occurrence of partially reduced cyclic and acyclic
botryococcenes in soils of the Everglades wetlands (Gao et al.,
2007). In the present case, it is noteworthy that the only acyclic
structure in this sample is the partially reduced C34 botryococcene
a. This could suggest an in vivo competition
between reduction and cyclisation during biosynthesis. Cyclisation
cannot occur after reduction of double bonds C-1/C-2,
C-6/C-24, C-17/C-29 or C-21/C-30. Partial reduction of l to a
prevents cyclisation occurring in the left hand moiety. Furthermore,
increasing steric hindrance due to the successive
effects of the methylation of the same isoprenoid unit results
in a strong decrease in the proportion of dicyclic masokocenes
(from 100% of C 35,downto14%ofC 36 and no C 37).
4. Conclusions
Biomarkers specific for the alga B. braunii race B are the
main constituents of the hydrocarbon fraction extracted
from a ca. 32,000 year old sediment interval from Lake Masoko,
Tanzania. Thanks to GC-MS and NMR and chemical
degradation, ten new cyclic botryococcenes and partially
reduced derivatives were identified. Three C34 to C36 dicyclobotryococcenes,
named masokocenes, were characterized,
along with seven C34 to C37 monocyclic compounds.
The structures indicate that the monocyclic botryococcenes
(CnH2n-10) are likely intermediates in the biosynthesis of the
dicyclic analogues, while the partially reduced botryococcenes
(C nH 2n-2, C nH 2n-6 and C nH 2n-8) are likely end products.
The study widely extends the number of molecular structures
within the botryococcene family.
From a biogeographical point of view, the study also reinforces
the idea that botryococcene-producing B. braunii
would be a rather common colonizer of crater (Huang
et al., 1999; Zhang et al., 2007) and maar (Fuhrmann et al.,
2003) lakes, just like reservoirs (Wake and Hillen, 1981),
dams (Metzger et al., 1985a; David et al., 1988; Metzger
et al., 1988; Jaffé et al., 1995) and also water tanks (Wolf
et al., 1985; Okada et al., 1995), under almost all latitudes
and from the sea level up to alpine zones. Known as a freshwater
alga, race B of B. braunii has been also reported to be
present in some marine environments (e.g. Grice et al.,
1998; Smittenberg et al., 2005). Physical and chemical conditions
favouring its growth in some lakes, leading sometimes
to enduring blooms (e.g. Wake and Hillen, 1981; Metzger
et al., 1985a; Townsend, 2001), are still poorly understood.
Although the preference of B. braunii race B for acidic waters
is often noted, the pH does not seem to be a critical parameter
since this microalga has been found in environments with
pH up to 8.6. Besides, it would appear from the literature (e.g.
Swale, 1968; Wake and Hillen, 1980, 1981; Metzger et al.,
1985a; Huang et al., 1999; Reynolds, 2000; Townsend,
2001; Zhang et al., 2007), that it generally grows in rather
small oligotrophic lakes (ca 1–2 km 2 or less) with a small
catchment area. However, the alga can be also present in
some great lakes like Michigan (Wolf and Cox, 1981). Studies
are currently in progress to determine the physicochemical
and environmental factors that could be at the origin of the
variation of the distribution and abundance of botryococcenes
observed in the sediments of Lake Masoko.
Acknowledgments
M. Delalande, D. Williamson and L. Bergonzini are
thanked for helpful comments and discussions. We also
thank Yongsong Huang and Hans-Peter Nytoft reviews and
constructive comments. The work was supported by the
Centre National de la Recherche Scientifique (CNRS) through
the CLEHA research programme from ECLIPSE-INSU and by
the Institute of Resource Assessment (IRA) at University of
Dar es Salaam. We are grateful to C. Fosse (ENSCP, Paris)
for exact mass determination and to M.-N. Rager (ENSCP,
Paris) for NMR spectral measurements. This paper is contribution
2 of the Rungwe Environmental Science Observatory
Network (RESON) and contribution 07.50 of UMR 5125 PEPS.