Labelling studies on the biosynthesis of terpenes in Fusarium fujikuroi

Article abstract of DOI:10.1039/c3cc45982a

Synthetic [2-¹³C]mevalonolactone was fed to the gibberellin producer Fusarium fujikuroi and its incorporation into four known terpenoids was investigated by ¹³C NMR analysis of crude culture extracts. The experiments gave detailed in

Full text of DOI:10.1039/c3cc45982a

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Labelling studies on the biosynthesis of terpenes
in Fusarium fujikuroi†
Cite this: Chem. Commun., 2014,
0, 5224
5
a
a
b
a
Christian A. Citron, Nelson L. Brock, Bettina Tudzynski and Jeroen S. Dickschat*
Received 5th August 2013,
Accepted 30th August 2013
DOI: 10.1039/c3cc45982a
www.rsc.org/chemcomm
13
Synthetic [2- C]mevalonolactone was fed to the gibberellin producer
Fusarium fujikuroi and its incorporation into four known terpenoids
13
was investigated by C NMR analysis of crude culture extracts. The
experiments gave detailed insights into the mechanisms of terpene
biosynthesis by this fungus.
Fusarium fujikuroi is a known producer of the sesquiterpenoids
1
2
3
4
a-acorenol 1, koraiol 2, and cyclonerodiol 3, and of gibberellins, a
class of highly oxidised nor-diterpenoids (C , Scheme 1). Gibberellins
19
are produced by plants, fungi, and bacteria, and exhibit significant
biological activity as plant growth hormones, while the function of the
sesquiterpenoids is unknown. The complex gibberellin biosynthesis
starts with the cyclisation of geranylgeranyl diphosphate to ent-
kaurene 4 (Scheme S2 of ESI†), followed by a series of oxidation
steps with contraction of the B ring and loss of the C-20 methyl
5
group to give gibberellic acid GA
3
5. While the pathways in plants
4
and fungi proceed via different intermediates, bacterial gibberellin
biosynthesis is unexplored.
The plant pathogenic fungus F. fujikuroi produces large amounts
of 5 that causes the bakanae or foolish seedling disease in infected
rice plants. Conversion of the terpene hydrocarbon 4 into 5 involves
the activity of four monooxygenases (P450-1, P450-2, P450-3, and
6
P450-4) and a desaturase (Scheme 1). The first enzyme, P450-4,
catalyses the oxidation of 4 via ent-kaurenal to ent-kaurenoic acid 6,
followed by a series of four oxidations by P450-1, starting with
hydroxylation of the B ring at C-7 to ent-7a-hydroxykaurenoic acid 7
7
that is transformed into GA -aldehyde 8 by ring contraction.
12
Another hydroxylation in the A ring (C-3) to GA14-aldehyde 9
and oxidation of the aldehyde moiety yield GA14 10. Consecutive
4
oxidative conversions of 10 by P450-2, the GA desaturase, and
P450-3 eventually result in 5.
Scheme 1 (A) Structures of sesquiterpenoids from Fusarium fujikuroi; (B) structures
of diterpenoids 4 and 5 from F. fujikuroi and biosynthetic steps from 4 towards 5.
8
–10
a
Institut f u¨ r Organische Chemie, Hagenring 30, 38106 Braunschweig, Germany.
E-mail: j.dickschat@tu-braunschweig.de; Fax: +49 531 3915272;
Tel: +49 531 3915264
The ring contraction from 7 to 8 was investigated using semi-
synthetic stereospecifically deuterated isotopologues of 6 carrying
deuterium labels in the a- or b-positions of C-6 or C-7. Incubation
with cell free extracts of F. fujikuroi mutants demonstrated that the
oxidative ring contraction to 8 proceeds with stereospecific loss of the
6b- and 7b-hydrogens, whereas the 6a- and 7a-hydrogens were retained,
11
b
Institut f u¨ r Biologie und Biotechnologie der Pflanzen, Schlossplatz 8,
4
8143 M u¨ nster, Germany
13
†
Electronic supplementary information (ESI) available: Synthesis of [2- C]-12
and NMR spectra. See DOI: 10.1039/c3cc45982a
5
224 | Chem. Commun., 2014, 50, 5224--5226
This journal is ©The Royal Society of Chemistry 2014

View Article Online
Communication
ChemComm
as determined by GC-MS analysis of the corresponding methyl ester.
GC-MS analysis also revealed that only the 6a-hydrogen is retained in 5.
These observations are in accord with a ring contraction mechanism
for the formation of 8 via radical intermediates as shown in Scheme 2,
12
pathway A. A problem of these experiments is the partial loss also of
deuterium labelling from 6a-H that is mechanistically not explained,
but may be a result of the C,H-acidity of 8 and the downstream
metabolites. The mechanism is in agreement with feeding experiments
14
by Birch and coworkers using C-labelled precursors in which the site
of incorporation was followed by degradation of the carbon backbone
13
of 5. On the other hand, the mechanism contradicts a previous report
showing that both C-6 hydrogens, if labelled with tritium, are retained
in the conversion of 6 to 8, one of which is lost in the chemical
14
oxidation of the aldehyde function.
These observations are only explainable via a mechanism that
(
a) does not involve oxidative hydrogen removal from C-6 and (b)
includes a rearrangement of one of the two hydrogens at C-6 by a
shift to C-7. A logical consequence of such a mechanism is the initial
formation of the intermediate alcohol 11 that may undergo a similar
ring contraction to 8 as suggested in Scheme 2, pathway A, albeit
with extrusion of C-6 instead of C-7 (Scheme 2, pathway B).
The critical point to distinguish between the two mechanisms is
the extrusion of C-6 or C-7 in the ring contraction step. This question
was addressed by Birch in a difficult experiment by chemical
degradation of 5 after incorporation of radioactive labelling from
13
acetate and mevalonolactone. The experiment was performed with
lack of knowledge about the structure of 5 that may have engendered
misinterpretations. Since the results from other labelling studies are
11,14
contradictory,
we decided to repeat Birch’s experiment using
modern NMR analytical techniques.
13
For this purpose racemic [2- C]-mevalonolactone 12 was synthe-
1
5–17
sised using a modification of a previously reported route (ESI†).
13
The synthetic [2- C]-12 was fed to the fungus F. fujikuroi and its
incorporation into C-7 of 4, further converted into 5 (Scheme 3), was
followed by C NMR analysis of crude culture extracts and comparison
13
13
18,19
with an authentic sample of 5 and C NMR literature data.
The
Scheme 2 Suggested mechanisms for the ring contraction in agreement
13
11
analysis revealed four carbon signals in the NMR spectrum (Table 1, with the observations by Birch et al. and Castellaro et al. (pathway A), and
14
by Hanson et al. (pathway B). Reactive intermediates are shown in brackets.
Fig. 1B) matching shifts of the standard (Fig. 1A). To further validate
13
these data the sample was mixed with 5 followed by C NMR
spectroscopy (Fig. S3 and S4 of ESI†). The results unequivocally
established the extrusion of C-7 in the ring contraction step of GA
13
biosynthesis, in full agreement with Birch’s earlier report and the
mechanism as shown in Scheme 2, pathway A, while the apparent
14
observation by Hanson and coworkers is hard to explain. A source of
error in the respective study may be the mentioned extremely low
incorporation rate. Alternative mechanisms to pathway A in Scheme 2,
e.g. via oxidation of radical intermediate B to a cation and subsequent
pinacol-type rearrangement, or formation of a diol followed by extrusion
13
Scheme 3 Labelling in 5 after feeding of [2- C]-12. Details of the
conversion of 12 to 4 can be followed in Scheme S2 of ESI.†
of a hydroxide to generate such a putative cationic intermediate, cannot indicating the expected incorporation into the assigned methylene
be excluded based on this or any of the previous labelling studies.
groups (Scheme 4 and Fig. S5–S7 of ESI†). Due to the conformational
13
Subsequently, the incorporation of [2- C]-12 into the sesquiter- flexibility incorporation into one of the stereochemically distinct
penes 1, 2, and 3 was investigated. The volatiles 1 and 2 were trapped geminal methyl groups could not be distinguished. A microreaction
20
on charcoal filters from agar plate cultures of two overexpression with the headspace extract was carried out that chemically converted 1
2
13
mutants (ESI†), followed by extraction with CDCl
3
1
and direct
C
into the conformationally fixed a-cedrene 14 by treatment with 88%
2
22
13
NMR analysis of the crude headspace extracts. For 1 strongly formic acid for 10 min at room temperature. Analysis by C NMR
elevated C signals were detected at 30.2, 27.94, and 27.89 ppm, revealed enhanced signals at 119.2, 36.1, and 27.6 ppm (Scheme 4 and
13
This journal is ©The Royal Society of Chemistry 2014
Chem. Commun., 2014, 50, 5224--5226 | 5225

View Article Online
ChemComm
Table 1 Comparison of C-NMR data of authentic 5 to C-labelled 5
Communication
13
13
13
obtained after feeding of [2- C]-12 to F. fujikuroi
a
b
a
b
Carbon no.
d [ppm]
Carbon no.
d [ppm]
C-19
C-7
C-16
C-2
178.1
173.2
157.8
133.3
131.5
106.3
90.5
C-9
C-5
C-6
C-8
C-14
C-15
C-12
C-11
C-18
52.1
50.8
50.5
49.4
44.3
42.7
38.8
16.6
14.5
C-1
C-17
C-10
C-13
C-3
76.6
68.4
C-4
53.1
a
b 13
Carbon numbering according to Scheme 1.
C NMR shifts of 5 are
2
13
referenced to [ H
feeding experiment with [2- C]-12 are given in bold.
6
]DMSO. C-NMR signals of labelled carbons in the
13
13
Scheme 4 Detected C labelling patterns in sesquiterpenes from Fusarium
13
13
fujikuroi after feeding of [2- C]-12. Asterisks indicate C-labelled carbons.
Notes and references
1
N. L. Brock, B. Tudzynski and J. S. Dickschat, ChemBioChem, 2011,
2, 2667.
N. L. Brock, K. Huss, B. Tudzynski and J. S. Dickschat, ChemBioChem,
013, 14, 311.
1
2
2
3
4
D. E. Cane, R. Iyengar and M.-S. Shiao, J. Am. Chem. Soc., 1981, 103, 914.
C. B ¨o mke and B. Tudzynski, Phytochemistry, 2009, 70, 1876.
1
3
2
13
5 F. McCapra, A. I. Scott, G. A. Sim and D. W. Young, Proc. Chem. Soc.,
Fig. 1 (A) C NMR spectrum of 5 in [ H
6
]DMSO; (B) C NMR spectrum of a
13
1962, 185.
crude culture extract from F. fujikuroi after feeding of [2- C]-12. Asterisks
indicate carbon signals originating from 5 after incorporation of [2- C]-12.
6
7
8
9
M. C. Rojas, P. Hedden, P. Gaskin and B. Tudzynski, Proc. Natl. Acad.
Sci. U. S. A., 2001, 98, 5838.
S. Malonek, M. C. Rojas, P. Hedden, P. Gaskin, P. Hopkins and
B. Tudzynski, Appl. Environ. Microbiol., 2005, 71, 1462.
B. Tudzynski, M. C. Rojas, P. Gaskin and P. Hedden, J. Biol. Chem.,
13
Fig. S8–S10 of ESI†). The signal at 27.6 ppm is due to incorporation into
the geminal pro-S methyl group of 14, whereas only low incorporation
ca. 10% by C NMR peak integration) into the adjacent methyl group
was detected (25.6 ppm). This defined stereochemical course is in
strong favour of a concerted S 2 type reaction with major inversion of
configuration, which in turn demonstrates incorporation of labelling
mainly into the pro-R methyl group of 1. This experiment also shows
that attack of the acorenyl cation E by water must have occurred from
the Re face of the molecule (Re/Si distinguished due to labelling,
2002, 277, 21246.
13
(
B. Tudzynski, M. Mihlan, M. C. Rojas, P. Linnemannst o¨ ns, P. Gaskin
and P. Hedden, J. Biol. Chem., 2003, 278, 28635.
23
1
0 A. Bhattacharya, S. Kourmpetli, D. A. Ward, S. G. Thomas, F. Gong,
S. J. Powers, E. Carrera, B. Taylor, F. N. de Caceres Gonzalez,
B. Tudzynski, A. L. Phillips, M. R. Davey and P. Hedden, Plant Physiol.,
2012, 160, 837.
N
1
1
1 S. J. Castellaro, S. C. Dolan, P. Hedden, P. Gaskin and J. Macmillan,
Phytochemistry, 1990, 29, 1833.
2 P. Hedden, Physiol. Plant., 1991, 101, 709.
Scheme 4). In contrast, an alternative stereochemical course with 13 A. J. Birch, R. W. Richards and H. Smith, Tetrahedron, 1959, 7, 241.
1
1
4 J. R. Hanson, J. Hawker and A. F. White, J. Chem. Soc., Perkin
Trans. 1, 1972, 1892.
5 J. A. Lawson, W. T. Colwell, J. I. De Graw, R. H. Peters, R. L. Dehn
and M. Tanabe, Synthesis, 1975, 729.
elimination of water from 1 to generate a cation that is intramolecularly
reattacked by the olefinic double bond from the same side, i.e. with
retention of configuration, is difficult to understand.
1
1
6 M. Tanabe and R. H. Peters, Org. Synth., 1981, 60, 92.
7 J. S. Dickschat, C. A. Citron, N. L. Brock, R. Riclea and H. Kuhz,
Eur. J. Org. Chem., 2011, 3339.
For 2 the obtained spectrum showed enhanced signals at 46.3, 37.6,
and 31.0 ppm (Scheme 4 and Fig. S11–S13 of ESI†). This, together with
2
published NMR data of 2, allowed for conclusions on the stereoche- 18 I. Yamaguchi and N. Takahashi, J. Chem. Soc., Perkin Trans. 1, 1975, 992.
19 A. Preiss, G. Adam, D. Saman and M. Budesinsky, Magn. Reson.
Chem., 1987, 25, 239.
mical course of the FPP cyclisation that proceeds with conversion of the
terminal E methyl group of FPP into the pro-S of the two geminal
2
0 J. S. Dickschat, S. C. Wenzel, H. B. Bode, R. M u¨ ller and S. Schulz,
ChemBioChem, 2004, 5, 778.
13
methyl groups in 2. Finally, incorporation of labelling from [2- C]-12
13
21 N. L. Brock, S. R. Ravella, S. Schulz and J. S. Dickschat, Angew.
Chem., Int. Ed., 2013, 52, 2100.
2
into 3 was observed by enhanced C signals at 40.39, 40.36, and
5.7 ppm (Scheme 4 and Fig. S14 and S15 of ESI†), in agreement with
2
2 T. G. Randall and R. G. Lawton, J. Am. Chem. Soc., 1969, 91, 2127.
3
Cane’s earlier report.
23 G. D. Brown, G.-Y. Liang and L.-K. Sy, Phytochemistry, 2003, 64, 303.
5
226 | Chem. Commun., 2014, 50, 5224--5226
This journal is ©The Royal Society of Chemistry 2014