ent-Kaurene and squalene synthesis in Fusarium fujikuroi cell-free extracts

Article abstract of DOI:10.1016/S0031-9422(00)00159-X

Sterols and gibberellins are the main terpenoids in the Ascomycete Fusarium fujikuroi. Their respective precursors squalene and ent-kaur-16-ene (henceforth called kaurene) were the main terpenoids synthesised from radioactive mevalonate by extracts of F. fujikuroi in vitro. Kaurene predominated when the extracts were obtained from mycelia engaged in gibberellin production. Squalene predominated in all other cases, and particularly when the extracts were obtained from mutants with various defects in gibberellin synthesis or nitrogen-fed wild-type cultures. New protein synthesis was required to maintain the production of gibberellins in vivo and of kaurene in vitro, but not to maintain the capacity to produce squalene in vitro. Addition of a nitrogen source to cultures engaged in gibberellin production caused a large, transient increase in the mycelial concentration of L-glutamine and abolished the accumulation of gibberellins immediately and the capacity to produce kaurene later. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0031-9422(00)00159-X

Phytochemistry 54 (2000) 723±728
www.elsevier.com/locate/phytochem
ent-Kaurene and squalene synthesis in Fusarium fujikuroi cell-free
extracts
Rafael Fernandez-Martõn, Carlos Domenech 1, Enrique Cerda-Olmedo*, Javier Avalos
Â
Departamento de Genetica, Universidad de Sevilla, Apartado 1095, E-41080 Sevilla, Spain
Received 26 October 1999; received in revised form 2 May 2000
Abstract
Sterols and gibberellins are the main terpenoids in the Ascomycete Fusarium fujikuroi. Their respective precursors squalene
and ent-kaur-16-ene (henceforth called kaurene) were the main terpenoids synthesised from radioactive mevalonate by extracts
of F. fujikuroi in vitro. Kaurene predominated when the extracts were obtained from mycelia engaged in gibberellin production.
Squalene predominated in all other cases, and particularly when the extracts were obtained from mutants with various defects in
gibberellin synthesis or nitrogen-fed wild-type cultures. New protein synthesis was required to maintain the production of
gibberellins in vivo and of kaurene in vitro, but not to maintain the capacity to produce squalene in vitro. Addition of a
nitrogen source to cultures engaged in gibberellin production caused a large, transient increase in the mycelial concentration of
L-glutamine and abolished the accumulation of gibberellins immediately and the capacity to produce kaurene later. 7 2000
Elsevier Science Ltd. All rights reserved.
Keywords: Fusarium fujikuroi; Gibberella; Terpenoids; Gibberellins; Sterols; Kaurene; Squalene
1. Introduction
are physically separated from the beginning in di€erent
subcellular compartments (Domenech et al., 1996).
The various terpenoids di€er in function and timing of
synthesis. While the sterols are essential components of
cell membranes, the gibberellins are dispensable sec-
ondary metabolites secreted to the culture medium of
the fungus only after depletion of the nitrogen source;
addition of fresh nitrogen source to the cultures stops
gibberellin biosynthesis (Candau et al., 1992). The in-
tracellular signal for nitrogen inhibition is glutamine
(Munoz and Agosõn, 1993) or a metabolite thereof.
Kaurene is the ®rst tetracyclic diterpene of the gib-
berellin pathway and squalene is the ®rst triterpene
precursor of sterols. Both are produced by cell-free
extracts of F. fujikuroi incubated with R-[14C] mevalo-
nate (Avalos et al., 1988; Fernandez-Martõn et al.,
1995). To investigate the regulation of the biosyntheses
of gibberellins and sterols in F. fujikuroi we have
examined the production of kaurene and squalene in
cell-free extracts from mycelia from wild-type and
mutant strains grown under di€erent conditions.
The gibberellins (Crozier, 1983; MacMillan, 1997)
are a group of tetracyclic terpenoids present in low
concentrations in plants, where they regulate di€erent
steps of growth and development. The Ascomycete
Fusarium fujikuroi (O'Donnell et al., 1998), previously
called Fusarium moniliforme and Gibberella fujikuroi, is
the industrial source for the various applications of
gibberellins in agriculture and brewing.
Both F. fujikuroi and plants make not only gibberel-
lins, but also other terpenoids, such as sterols and
carotenoids. In F. fujikuroi these terpenoids derive
from hydroxymethylglutaryl CoA (Fig. 1); although
gibberellins and sterols share all the early intermedi-
ates, up to farnesyl pyrophosphate, the two pathways
* Corresponding author. Tel.: +34-954624107; fax: +34-
954624107.
1 Present address: Departamento de Biologõa Molecular, Universi-
dad Nacional de Rõo Cuarto, Rõo Cuarto, Argentina.
0031-9422/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved.
PII: S0031-9422(00)00159-X

 Â
R. Fernandez-Martõn et al. / Phytochemistry 54 (2000) 723±728
724
HMG-CoA
lene and kaurene were the ®nal in vitro products
0H
-
-
CoA
0 0 C
0 0 C
0
because their conversion to later metabolites requires
NADPH and other cofactors (Ono and Imai, 1985;
Ashman et al., 1990) that were absent in our in vitro
tests. Low-nitrogen cultures were producing gibberel-
lins at the time of extraction, while high-nitrogen cul-
tures were not, as expected from the regulation of
gibberellin production by nitrogen (Candau et al.,
1992; Avalos et al., 1999). The nature of the nitrogen
source was not critical; it was NH4NO3 in the exper-
iments in Fig. 2, but similar results were obtained
when various amino acids (alanine, glycine, leucine,
glutamine, or glutamate) were used instead.
(C6)
mevalonate
0H
0H
(C5) isopentenyl PP
dimethylallyl PP (C5)
0-PP
0-PP
geranyl PP (C10)
farnesyl PP (C15)
0-PP
IPP
IPP
0-PP
geranylgeranyl PP (C20)
0-PP
squalene (C30)
kaurene
(C20)
Carotenoids
Kaurenolides
The gib mutants were isolated because they produce
no gibberellins, or smaller amounts than the wild type
(Candau et al., 1991). Extracts from low-nitrogen cul-
tures of the gib mutants produced squalene, while
0H
0
C0
H0
C00H
H0
ergosterol
gibberellic acid
Sterols
Gibberellins
90
80
Fig. 1. Main steps in the biosyntheses of gibberellins and sterols in
F. fujikuroi. Dotted arrows represent several reactions.
no ammonium addition
70
2. Results and discussion
60
50
Cell-free extracts of wild-type mycelia grown for 6
days in low-nitrogen medium converted mevalonate to
kaurene; squalene was produced when the cultures had
been grown in high-nitrogen medium (Fig. 2). Squa-
ammonium addition
40
0
8
80
4
Wild
type
Hydrophobic terpenoids
0
60
80
60
SG127
40
Kaurene
40
SG138
20
20
SG136
Squalene
Wild
type
SG121
0
0
1
2
3
0
SG139
60
Time after addition of ammonium (days)
0
20
40
80
Fig. 3. Gibberellin production and in vitro terpenoid biosynthesis
after addition of nitrogen source NH4Cl (®nal concentration 100
mM) was added (closed symbols) or not (open symbols) to wild-type
cultures grown for 8 days in low-nitrogen medium. Upper panel,
concentration of gibberellins in the medium. Central panel, total in-
corporation of radioactivity from mevalonate into hydrophobic ter-
penoids. Lower panel, incorporation into kaurene (circles) and
squalene (squares), as a fraction of total radioactivity in hydrophobic
terpenoids.
Squalene (%)
Fig. 2. Biosynthesis of terpenoids in cell-free extracts. Radioactivity
in kaurene or squalene as a fraction of total radioactivity in hydro-
phobic terpenoids (2.1 kBq per experiment in the average) after the
incubation of cell-free extracts from wild type and gib mutant myce-
lia with [2-14C]mevalonate. Cultures were grown for 6 days in low-
nitrogen medium (open symbols) or high-nitrogen medium (closed
symbols).

 Â
R. Fernandez-Martõn et al. / Phytochemistry 54 (2000) 723±728
725
those of the wild type did not (Fig. 2). The in vitro
squalene and kaurene productions correlated nega-
tively with each other. The more extreme phenotype
was that of strain SG139, a mutant that does not ac-
cumulate gibberellins or kaurenolides in vivo. Low-
nitrogen-culture extracts of this mutant behaved
exactly as high-nitrogen-culture extracts of the wild
type. Very similar results were obtained with strains
SG121 and SG136. Less extreme was the in vitro phe-
notype of strain SG138, practically devoid of gibberel-
lins because of its inability to carry out various
oxidative reactions in the pathway (Barrero et al.,
1999). The mutant that was closest to the wild type
was strain SG127, which has a partial block of the gib-
berellin pathway.
Squalene appears to be synthesised `by default' in
various mycelial extracts, both wild type and mutant,
unless the original mycelia were engaged in gibberellin
production; in this case squalene was displaced by
kaurene as the main in vitro product of mevalonate.
The capacity to synthesize kaurene in vitro was lost
gradually after addition of fresh nitrogen source to the
cultures (Fig. 3). Squalene synthesis in vitro increased
steadily from initially undetectable levels and mirrored
the decrease in kaurene synthesis. These changes
started a few hours after the addition of the nitrogen
source and continued for several days, until squalene
became the predominant in vitro terpenoid product.
The accumulation of gibberellins ceased immediately
after the addition of fresh nitrogen source to cultures
that were synthesising them actively (Fig. 3), as
expected (Candau et al., 1992). The nitrogen source
had both immediate and delayed inhibitory e€ects.
The addition of ammonium provoked a transient
increase in the mycelial concentration of L-glutamine
(Fig. 4), which, at its peak, constituted about one
tenth of the mycelial dry mass. Lesser variations were
found in the concentrations of L-alanine and L-gluta-
mic acid. The concentrations of other amino acids
remained approximately constant.
Our results support the belief that the intracellular
nitrogen signal in the regulation of gibberellin biosyn-
thesis is glutamine (Munoz et al., 1993) or one of its
metabolites. An additional e€ect through the depletion
of 2-oxoglutarate is not excluded.
Mycelia grown in high-nitrogen medium started gib-
berellin production a few hours after they were washed
and transferred to a glucose solution (Fig. 5). Gibber-
ellins accumulated at an approximately constant rate
for several days. At the same time the mycelia took a
reddish tint owing to the accumulation of bikaverin, a
polyketide compound (Balan et al., 1970). The in vitro
production of squalene, found in extracts of mycelia
40
30
20
10
0
8
4
Hydrophobic terpenoids
0
80
60
120
Kaurene
100
40
20
Glutamine
80
60
Squalene
40
0
Alanine
20
0
1
2
3
4
5
Glutamate
Time after transfer to glucose solution (days)
0
0
4
8
24
Fig. 5. Gibberellin production and in-vitro terpenoid biosynthesis
after nitrogen deprivation. Mycelia grown for 6 days in high-nitrogen
medium were transferred to a glucose solution. Upper panel, concen-
tration of gibberellins in the solution. Central panel, total incorpor-
ation of radioactivity from [2-14C]mevalonate into hydrophobic
terpenoids. Lower panel, incorporation into kaurene (circles) and
squalene (squares), as a fraction of total radioactivity into hydro-
phobic terpenoids.
Time after addition of ammonium (hours)
Fig. 4. Intracellular concentrations of free amino acids after addition
of nitrogen source. NH4Cl (®nal concentration 100 mM) was added
(closed symbols) or not (open symbols and dashed line) to wild-type
cultures grown for 8 days in low-nitrogen medium. Triangles, L-glu-
tamine; squares, L-alanine; circles, L-glutamic acid.

 Â
R. Fernandez-Martõn et al. / Phytochemistry 54 (2000) 723±728
726
that had not yet started to produce gibberellins, was
replaced by that of kaurene following nitrogen
removal (Fig. 5).
cumstances only and had to be continuously renewed.
Squalene and kaurene are produced in separate subcel-
lular compartments (Domenech et al., 1996), but dis-
ruption of these compartments made the enzymes
compete in vitro for the last common substrate, farne-
syl diphosphate. Under these conditions, conversion to
geranyl±geranyl diphosphate and kaurene predomi-
nates over conversion to squalene.
The isolation of the genes for gibberellin biosyn-
thesis in F. fujikuroi (Tudzynski et al., 1998; Tudzynski
and Holter, 1998) opens the way for a correlation
between the biosynthesis in vivo and in vitro with gene
expression. The gene cluster for gibberellin biosyn-
thesis includes two contiguous genes, one for a speci®c
geranyl±geranyl diphosphate synthase and the other
for a bifunctional copalyl diphosphate and kaurene
synthase whose transcription is enhanced by nitrogen
starvation.
Addition of cycloheximide blocked the production
of gibberellins in mycelia that had been transferred to
a glucose solution (Fig. 6). No gibberellins were pro-
duced when cycloheximide was added in the ®rst hours
after transfer, before the onset of gibberellin pro-
duction in the mycelia and of kaurene production in
their extracts. Accumulation of gibberellins stopped a
few hours after addition of cycloheximide to mycelia
actively engaged in gibberellin synthesis. This suggests
that new protein synthesis was required for the onset
and for the maintenance of gibberellin biosynthesis.
New protein synthesis was apparently not needed to
maintain the squalene-synthesising ability. Extracts
were prepared from the mycelia that had been incu-
bated in a glucose solution for 3 days (Fig. 6). When
cycloheximide had been present in the glucose solution
all the time, the extracts synthesised squalene, but not
kaurene. The later cycloheximide was added, more
kaurene and less squalene was synthesised in vitro.
The break-even point was reached when cycloheximide
was added 1.4 days after the transfer to the glucose
solution.
Enzyme stability should be advantageous in a pro-
cess that is maintained for many days in the absence
of a nitrogen source. We found, however, that mycelia
ceased to produce gibberellins a few hours after ad-
dition of cycloheximide and their extracts lost their
ability to synthesise kaurene about 2 days later.
The results, particularly those with the mutants and
with cycloheximide, indicate that squalene synthesising
enzymes were always present in the cells, while kaurene
synthesising enzymes were available under some cir-
3. Experimental
F. fujikuroi IMI58289, a gibberellin producing wild
type, was obtained from the Commonwealth Myco-
logical Institute, Kew, England. The gib mutants defec-
tive in gibberellin biosynthesis were obtained after
exposure of wild-type spores to N-methyl-N'-nitro-N-
nitrosoguanidine (Candau et al., 1991).
The fungi were grown in 500-ml Erlenmayer ¯asks
with 250 ml of minimal medium that contains D(+)-
1
glucose (80 gl 1), NH4NO3 (high-nitrogen, 4.8 g l
;
low-nitrogen, 0.48
g
l
1), KH2PO4 (5
g
l
1),
MgSO4Á7H2O (1 g l 1), and trace elements (Geissman
et al., 1966). The ¯asks were inoculated with about 106
spores and incubated in the dark at 308C in an orbital
shaker (150 rpm). Spores were collected from cultures
1
grown on sporulation agar (1 g l yeast extract, 1 g
1
1
1
D-glucose, 1 g l NH4NO3, 1 g l KH2PO4, 0.5 g
18
12
6
l
l
1
1
MgSO4Á 7H2O, 16 g l
agar) under white light
(about 4 W m 2). When required, NH4NO3 was
replaced by amino acids, maintaining the total nitro-
gen concentration. When indicated, cultures grown in
high-nitrogen
minimal
medium
were
washed
thoroughly with water and transferred to a glucose sol-
ution that contained D(+)-glucose (20 g l 1) and the
trace elements of the minimal medium.
Gibberellin concentrations in the media were esti-
mated ¯uorometrically (Candau et al., 1991) using gib-
berellic acid (Sigma, St. Louis, MO, USA) as a
reference and were expressed as the mean and its stan-
dard error in two to six independent experiments, with
two or more separate determinations in each. Amino
acid concentrations were determined by reverse-phase
HPLC (Farfan et al., 1996). Proline cysteine, and
tryptophan could not be measured by this method.
Cell-free extracts were obtained by lyophilising
0
0
1
2
3
Time after transfer to glucose solution (days)
Fig. 6. Inhibition of gibberellin production by cycloheximide. Myce-
lia grown for 6 days in high-nitrogen medium were transferred to a
glucose solution and cycloheximide was added at di€erent times.
Open symbols and dashed lines, concentration of gibberellins in the
solution after cycloheximide addition at the time indicated by the
arrows. Closed symbols and continuous lines, controls without cyclo-
heximide.

 Â
R. Fernandez-Martõn et al. / Phytochemistry 54 (2000) 723±728
727
mycelial samples and passing them immediately
through a sieve (pores 0.5 mm in diameter). The pow-
der was mixed with 8 vol extraction bu€er (0.4 M
with a liquid scintillation counter (Wallac 1409, Turku,
Finland) with 4 ml Biogreen 1 (Sharlau SA, Barcelona,
Spain). The results are the mean and s.e. of the ratio
of radioactivity in kaurene and squalene to that in
hydrophobic terpenoids in at least two independent ex-
periments, with two separate determinations in each.
Cycloheximide (actidione, Serva Feinbiochemica,
1
Tris±HCl pH 8, 5 mM dithiothreitol, 200 ml l gly-
cerol) and centrifuged at 104 Â g; 0.2 ml supernatant
was mixed with 0.3 ml extraction bu€er without gly-
cerol containing (®nal concentrations) 18 mM (18.5
kBq) RS-[2±14C]mevalonate, 10 mM ATP, 2 mM
NAD, 2 mM NADP, 6 mM MnCl2, 4 mM MgCl2 and
10 mM KF and incubated at 358C for 2 h. The
required sodium mevalonate was obtained by drying
the lactone (Amersham International, Amersham, UK)
and dissolving it in 10 ml 10 mM NaOH. After stop-
ping the reaction with 1 ml of MeOH, the mixture was
extracted three times with petrol (bp 40±608C). An ali-
quot was radioassayed to determine labelled hydro-
phobic terpenoids ꢀ4:322:4 kBq, mean and s.e. in 136
tests.) The rest of the extract was dried with N2 and
developed with petrol on Silica Gel G TLC. Kaurene
and squalene, the main products (Fig. 7), were scraped
o€ and radioassayed. Radioactivity was determined
Heidelberg) was used at a ®nal concentration of 50 mg
1
l
(Candau, 1991).
which totally inhibits the growth of F. fujikuroi
Acknowledgements
We thank C. Vallejo for their able collaboration and
acknowledge the ®nancial support of the European
Union (FAIR CT96-1633), the Spanish Government
(Direccion General de Ensenanza Superior PB96-1336
and Programa de colaboracion con Iberoamerica), and
Junta de Andalucõa (Group CVI0119).
References
Ashman, P.J., Mackenzie, A., Bramley, P.M., 1990. Characterization
of ent-kaurene oxidase activity from Gibberella fujikuroi. Biochim.
Biophys. Acta 1036, 151±157.
Avalos, J., Fernandez-Martõn, R., Prado, M.M., Cerda-Olmedo, E.,
1999. Gibberellin biosynthesis in Gibberella. J. Bot. Gall. 146, 55±
65.
Avalos, J., Mackenzie, A., Nelki, D.S., Bramley, P.M., 1988.
Terpenoid biosynthesis in cell extracts of wild-type and mutant
strains of Gibberella fujikuroi. Biochim. Biophys. Acta. 966, 257±
265.
Balan, J., Fuska, J., Kuhr, I., Kuhrova, V., 1970. Bikaverin, an anti-
biotic from Gibberella fujikuroi, e€ective against Leishmania brasi-
liensis. Folia Microbiol. 15, 479±484.
Barrero, A.F., Oltra, J.E., Cabrera, E., Reyes, F., Alvarez, M., 1999.
Metabolism of gibberellins and ent-kaurenoids in mutants of
Gibberella fujikuroi. Phytochemistry 50, 1133±1140.
Candau, R., 1991. Produccion de giberelinas por Gibberella fujikuroi.
Ph.D. Thesis, University of Sevilla.
Candau, R., Avalos, J., Cerda-Olmedo, E., 1991. Gibberellins and
carotenoids in the wild type and mutants of Gibberella fujikuroi.
Appl. Environ. Microbiol. 57, 3378±3382.
Candau, R., Avalos, J., Cerda-Olmedo, E., 1992. Regulation of gib-
berellin biosynthesis in Gibberella fujikuroi. Plant Physiol. 100,
1184±1188.
Crozier, A. (Ed.), 1983. The Biochemistry and Physiology of
Gibberellins, vol. 11. Praeger, NY.
Domenech, C.E., Giordano, W., Avalos, J., Cerda-Olmedo, E., 1996.
Separate compartments for the production of sterols, carotenoids
and gibberellins in Gibberella fujikuroi. Eur. J. Biochem. 239,
720±725.
Farfan, M.J., Martõn-Rendon, E., Calderon, I.L., 1996. E€ect of
gene ampli®cation on threonine production by yeast. Biotechnol.
Bioeng. 49, 667±674.
Fig. 7. Terpenoids synthesised from [2-14C]mevalonate by cell-free
extracts from mycelia incubated in high-nitrogen medium for 6 days
and in a glucose solution for the time indicated. Autoradiography of
a TLC plate exposed on X OMAT S ®lm (kodak) for 14 days.
Fernandez-Martõn, R., Reyes, F., Domenech, C.E., Cabrera, E.,
Bramley, P.M., Barrero, A.F., Avalos, J., Cerda-Olmedo, E.,
1995. Gibberellin biosynthesis in gib mutants of Gibberella fuji-
kuroi. J. Biol. Chem. 270, 14970±14974.

 Â
R. Fernandez-Martõn et al. / Phytochemistry 54 (2000) 723±728
728
Geissman, T.A., Verbiscar, A.J., Phinney, B.O., Cragg, G., 1966.
Studies on the biosynthesis of gibberellins from ( )-kaurenoic
acid in cultures of Gibberella fujikuroi. Phytochemistry 5, 933±
947.
tematics and phylogeography of the Gibberella fujikuroi species
complex. Mycologia 90, 465±493.
Ono, T., Imai, Y., 1985. Squalene epoxidase from rat liver micro-
somes. Methods Enzymol. 110, 375±380.
MacMillan, J., 1997. Biosynthesis of the gibberellin plant hormones.
Nat. Prod. Rep. 14, 221±243.
Tudzynski, B., Holter, K., 1998. Gibberellin biosynthetic pathway in
Gibberella fujikuroi: evidence for a gene cluster. Fungal Genet.
Biol. 25, 157±170.
Munoz, G.A., Agosin, E., 1993. Glutamine involvement in nitrogen
control of gibberellic acid production in Gibberella fujikuroi.
Appl. Environ. Microbiol. 59, 4317±4322.
Tudzynski, B., Kawaide, H., Kamiya, Y., 1998. Gibberellin biosyn-
thesis in Gibberella fujikuroi: cloning and characterization of the
copalyl diphosphate synthase gene. Curr. Genet. 34, 234±240.
O'Donnell, K., Cigelnik, E., Nirenberg, H.I., 1998. Molecular sys-