2
E. Alcalde et al. / Phytochemistry xxx (2016) xxx–xxx
gene acaA, cleaves the C25 fragment into C18 and C7 fragments
(Medina et al., 2011). This discovery has been extended to Blakeslea
(Barrero et al., 2011; Sahadevan et al., 2013), Mucor mucedo
(Sahadevan et al., 2013), and Mucor circinelloides (our unpublished
work).
The dedicated b-carotene biosynthetic pathway of Phycomyces
depends on two structural genes, carB for phytoene dehydrogenase
and carRA for phytoene synthase and lycopene cyclase (Arrach
et al., 2001; Eslava and Cerdá-Olmedo, 1974; Torres-Martínez
et al., 1980). The carA mutants are white and deficient in all carote-
nes, the carB mutants are white and rich in phytoene, and the carR
mutants are red and rich in lycopene. The carS mutants (Murillo
and Cerdá-Olmedo, 1976; Tagua et al., 2012) accumulate large con-
centrations of b-carotene, which makes them deep yellow, instead
of light yellow, as the wild type. Mutants of these genes lack apoc-
arotenoids (Tagua et al., 2012) and are unable to stimulate their
sexual partners (Murillo et al., 1978; Sutter, 1975).
cultures of the wild types; nineteen were found in mated cultures
only; and two were not found in mated cultures (Medina, 2013).
For convenience we designate the compounds detected as Nt or
At, where N stands for neutral extracts; A, for acid extracts; and
t, for the elution time in tens of seconds. Thus, the (2S, 6E, 8E)
and (2S, 6Z, 8E) isomers of cyclofarnesa-4,6,8-triene-2,10,11-triol
(10, 11) will be called N121 and N122 for short, since we did not
set up a column for them in Fig. 1.
In our liquid chromatography with trifluoroacetic acid added to
the mobile phase, the methylhexanoids eluted rapidly, around min.
12, the cyclofarnesoids, more slowly, min. 15–20, and the trispor-
oids thereafter. The known cyclofarnesoids absorb shorter wave-
lengths (kmax < 320 nm) than trisporic acids and other trisporoids.
All the cyclofarnesoids of Phycomyces were found in the neutral
extracts, but compounds (5), (7), and (8) were polar enough to be
present in the acid extracts as well.
Comparisons of the extracts of wild-type and carB-mutant cul-
tures of Phycomyces revealed an unforeseen diversity of apoc-
arotenoids (Medina, 2013; Polaino et al., 2012). We set out to
describe the cyclofarnesoids and the methylhexanoids in single
and mated cultures of wild-type and mutant strains under various
conditions. Fortunately, the carB mutants that make no b-carotene
retain the ability to bioconvert apocarotenoids.
2.3. Identification and chemical diversity of cyclofarnesoids and
methylhexanoids
The structures (Fig. 1) of the two methylhexanoids (1, 2) and
the eight cyclofarnesoids (4–11) that gave the highest UV absorp-
tion have been deduced from their 1H and 13C NMR spectra, UV
absorption spectra, and MS molecular formulas (see Table S1,
Figs. S1 and S2, and Text S1 in Supplementary Materials). The
structures of compounds (2), (6), (10), and (11) are supported by
previous syntheses (González-Delgado et al., 2013; Polaino et al.,
2010). Our results on compounds (1), (2), (6), (7), (10), and (11)
coincide with those on patrons isolated from nature in previous
work (Barrero et al., 2011; Polaino et al., 2012).
2. Results
2.1. A network of apocarotenoids
The apocarotenoids of the Mucorales have unpractical system-
atic names and they are so numerous and diverse that giving them
arbitrary trivial names would be untoward. A convenient nomen-
clature can be suggested on the basis of a scheme ordered as a net-
work (Fig. 1), in which contiguous compounds could be derived
from each other by simple metabolic steps. Additionally, this
scheme can be seen as a set of possible biosynthetic pathways.
Trisporic acids were named, as they were discovered, with the
letters A through E to indicate the substitutions at their C-2, C-3,
and C-13 atoms. These letters can be used to define extended sub-
families that include the neutral trisporoids. For the cyclofarne-
soids we define here subfamilies with additional capital letters
that indicate the substitutions at their C-2, C-8, and C-14 atoms.
We arrange both the trisporoids and the cyclofarnesoids in col-
umns for the subfamilies and in rows according to the substitu-
tions at C-4 and at the b carbon linked to C-1. The proposed
semisystematic names consist of the roots ‘‘trispor” or ‘‘cyclofar-
nes”, the suffixes and prefixes of the rows and the capital letters
of the columns.
4-Dihydrocyclofarnesine S (4), 4-dihydrocyclofarnesine T (5),
trans-cyclofarnesine U (8), and cyclofarnesine T (9) are natural
products that had not been described previously. The structure of
(4) was deduced from its formula, C15H24O3, its IR absorption
bands at 3409 and 1010 cmꢀ1, its UV absorption spectrum, which
pointed out to a cyclofarnesa-5,7,9-triene, and its 1H and 13C
NMR spectra. The slight differences (Table S1) in the NMR spectra
of (4) and a synthetic compound (12) (González-Delgado et al.,
2013), involve the arrangement of the OH groups on C-2 and C-4
(Fig. 2) and identify these compounds as trans- and cis-4-dihydro-
cyclofarnesine S, respectively.
4-Dihydrocyclofarnesine T (5), with the molecular formula
C15H24O4 and a UV spectrum characteristic of a conjugate triene,
is the most polar of the cyclofarnesoids and only partially extracted
in basic medium. It contains one more hydroxyl group than (4),
tentatively located on C-16 on the basis of the proposed conversion
of (4) to cis-cyclofarnesine U (7) (see subsection 2.5). trans-Cyclo-
farnesine U (8), C15H23O3, must be the trans isomer of (7)
(Polaino et al., 2012), as indicated by its polarity and its UV
absorption.
Compound (9), C15H23O4, with UV spectrum characteristic of a
conjugated trienone (Grasselli and Selva, 1970), is very similar to
cyclofarnesine S, but has an additional hydroxyl group (Supple-
mentary Text S1). Compound (9) could be either cyclofarnesine T
or cyclofarnesol S, but the last is unlikely, because it would be
the only cyclofarnesoid from Phycomyces hydroxylated at the b car-
bon linked to C-1.
The scheme in Fig. 1 generalizes the semisystematic names
already used for some trisporoids and can be widened to include
new variations to be discovered in the future. We have left cyclo-
farnesa-4,6,8-triene-2,10,11-triol (10, 11, isomers 2S, 6E, 8E and
2S, 6Z, 8E, respectively) out of the scheme, but it could be placed
in a new column in the future.
2.2. Putative apocarotenoids
We analyzed by LC/MS and LC/UV the neutral (pH 8) and acidic
(pH 2) extracts of single and mated cultures of strains NRRL1555
and A56, two largely isogenic wild types of opposite sex. Putative
apocarotenoids were immediately recognized as compounds that
were absent in similar extracts of the white carB mutant strains,
which completely lack b-carotene. There were thirty-two such
compounds in five-days cultures, disregarding those with very
low absorbance. Eleven of them were present in single and mated
2.4. Variations with age and sex
The total amounts of methylhexanoids and cyclofarnesoids
increased at least from the second to the fifth day in cultures
grown from spores (Table 1). Judging from their molecular struc-
tures, the apocarotenoids of each family are expected to have very
similar molar absorption coefficients, thus allowing the use of the
absorption in the chromatograms for quantitative comparisons.
Please cite this article in press as: Alcalde, E., et al. Cyclofarnesoids and methylhexanoids produced from b-carotene in Phycomyces blakesleeanus. Phyto-