The isolation, structural determination, and total synthesis of terfestatin A, a novel auxin signaling inhibitor from Streptomyces sp.

Article abstract of DOI:10.1016/j.tetlet.2004.09.055

The structure of terfestatin A was determined to be p-terphenyl β-glucoside on the basis of spectroscopic analyses, chemical degradation, and total synthesis. Indole 3-acetic acid, a plant hormone auxin, regulates every aspects of plant growth and develop

Full text of DOI:10.1016/j.tetlet.2004.09.055

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 8359–8362
The isolation, structural determination, and total synthesis of
terfestatin A, a novel auxin signaling inhibitor
from Streptomyces sp.
Atsushi Yamazoe, Ken-ichiro Hayashi, Atsuhito Kuboki, Susumu Ohira and
Hiroshi Nozaki*
Department of Biochemistry, Okayama University of Science, 1-1 Ridai-cho, Okayama city 700-0005, Japan
Received 6 August 2004; revised 3 September 2004; accepted 8 September 2004
Available online 23 September 2004
Abstract—Indole 3-acetic acid, a plant hormone auxin, regulates every aspects of plant growth and development. In the screening
for specific inhibitors on auxin signaling using the transgenic Arabidopsis plant harboring the auxin-responsive reporter gene, ter-
festain A was found to be a novel inhibitor of auxin signaling from the culture of Streptomyces sp. F40. The structure of terfestatin
A was determined to be p-terphenyl b-glucoside on the basis of spectroscopic analyses, chemical degradation, and total synthesis.
Ó 2004 Elsevier Ltd. All rights reserved.
Plant growth and development are controlled by plant
hormones such as auxin, cytokinin, abscisic acid, ethyl-
ene, and gibberellin that are low molecular organic com-
pounds, in contrast to proteinous hormones in mammal.
Indole 3-acetic acid, auxin, regulates cell division, elon-
gation, and differentiation in every aspects of plant
growth and development.1 Auxin elicits the rapid and
dramatic change of the expression patterns of a number
of genes. Therefore, its profound and diverse effects of
auxin on plant development would result from regula-
tion of the expression pattern of a number of genes.
Recent molecular biological studies have revealed that
the expressions of AUX/IAA genes are specifically in-
duced by auxin within a minute and play essential roles
in early steps of auxin signal transduction.2,3 AUX/IAA
proteins are short-lived transcriptional repressors local-
ized in the nucleus and regulate the expression of late
auxin-responsive genes linked to cell division and elon-
gation.4 However, the molecular mechanism of regula-
tion on the early auxin-responsive gene expression is
not fully understood. In addition, auxin receptor has
not been identified although many efforts have been
made to identify the receptor for several decades. Bio-
chemical and pharmacological approaches with specific
inhibitors on signal transduction have greatly contrib-
uted to the study of signaling in cellular response. Iden-
tification of the specific inhibitors of auxin signal
transduction would also provide useful tools for the
study of auxin signaling machinery. The plant growth
regulatory activity of many natural products has been
reported. However, most of this activity has been evalu-
ated with intact plants, and therefore, the site of action
have still been unclear. To screen for specific inhibitors
on auxin signaling, we used the reporter gene assay for
target-oriented screening using the transgenic Arabidop-
sis plant, which harbors the early auxin-inducible pro-
moter and b-glucuronidase (GUS) reporter gene.5 This
transgenic reporter line expresses GUS protein in re-
sponse to auxin. The GUS induction in the root is rapid
and highly specific to auxin, and can be detected easily
by enzyme activity measurement. This reporter system
enables us to perform rapid and specific screening for
inhibitors on auxin signaling. We have reported yoko-
nolides as an auxin signaling inhibitor using this repor-
ter assay from Streptomyces sp.6,7 In the continuous
screening for specific inhibitors of auxin signal transduc-
tion from microbial metabolites using the reporter
assay, we found that Streptomyces sp. F40 produced a
new inhibitor, designated terfestatin A (1). This report
deals with the isolation, structure determination, total
synthesis, and some biological activities of 1.
Keywords: Terfestatin; Terphenyl; Auxin signaling inhibitor; AUX/
IAA; Arabidopsis.
*
The terfestatin A-producing organism identified as
Streptomyces sp. F40 was isolated from soil collected
Corresponding author. Tel.: +81 862569661; fax: +81 862569559;
e-mail: nozaki@dbc.ous.ac.jp
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.09.055

8360
A. Yamazoe et al. / Tetrahedron Letters 45 (2004) 8359–8362
Table 1. 1H and 13C NMR data of 1 (CD3OD)
at Okayama city, Japan.8 Two hundred milliliters of
seed culture of the producing strain was inoculated into
a 10-liter jar-fermenter containing 8L of the production
medium (soluble starch 1%, glucose 0.5%, phamamedia
0.4%, polypeptone 0.1%, meat extract 0.1%, yeast
extract 0.1%, and corn steep liquor 0.1%), and cultivated
at 28°C for 96h under agitation at 400rpm. The result-
ant culture was filtrated. The filtrate was adjusted to
pH7.0 and extracted with EtOAc. In each purification
step, the fractions were assayed for the inhibitory activ-
ity against auxin-induced reporter gene expression with
the Arabidopsis transgenic BA-GUS reporter line.5
The auxin-induced b-glucuronidase (GUS) expression
was detected by the histochemical staining of GUS enzy-
matic activity using 5-bromo-4-chloro-3-indyl-b-D-
glucuronic acid (X-Glc) as a substrate. In this reporter
assay, auxin-induced GUS expression was monitored
by the visible blue stain derived from X-Glc in root or
by a fluorescent GUS enzyme activity measurement.5
The EtOAc layer was concentrated in vacuo to give
0.48g of extract. The EtOAc extract (2.4g from 40L
of culture) was applied onto a silica gel column and
developed with stepwise elution (CHCl3–MeOH). The
active fractions were concentrated and further purified
by Sephadex LH-20 column chromatography eluting
with CHCl3–MeOH = 2:1 solvent. The active fraction
was chromatographed by ODS low-pressure column
(MeOH:H2O = 1:1) to give a white powder, as an active
compound (1: 13.7mg).16
No.
1H (d)
13C (d)
1
135.6
132.2
128.9
127.7
128.9
132.2
136.4
137.2
149.6
118.1
153.0
108.8
140.2
130.8
128.6
128.2
128.6
130.8
107.7
75.2
2
7.43 (d, 1H, J = 7.6Hz)
7.38 (m, 1H)
7.26 (m, 1H)
3
4
5
7.38 (m, 1H)
7.43 (d, 1H, J = 7.6Hz)
6
10
20
30
40
500
600
100
200
300
400
500
6
6.39 (s, 1H)
7.54 (d, 1H, J = 7.3Hz)
7.37 (m, 1H)
7.31 (m, 1H)
7.37 (m, 1H)
7.54 (d, 1H, J = 7.3Hz)
4.29 (d, 1H, J = 8.0Hz)
3.34 (t, 1H, J = 8.9, 8.0Hz)
3.22 (m, 1H)
1000
2000
3000
4000
5000
6000
77.6
71.0
3.22 (m, 1H)
2.99 (m, 1H)
77.8
62.4
3.49 (dd, 1H, J = 11.7, 2.6Hz)
3.41 (dd, 1H, J = 11.7, 5.1Hz)
TMS is used as internal reference.
The HR-FABMS of 1 gave a molecular formula of
C24H24O8 (m/z [M+Na]+ calcd for C24H24O8Na,
463.1368. Found, 463.1360). IR spectrum showed the
1
existence of hydroxy group (3393cmꢀ1). The H NMR
spectrum (Table 1) showed two pairs of monosubsti-
tuted benzene protons at d 7.54 (2H), 7.37 (2H), 7.31
(1H), 7.43 (2H), 7.38 (2H), 7.26 (1H), a singlet benzene
proton at d 6.39 (1H), five oxymethine protons of sugar
including an anomeric proton (d 4.29, 3.34, 3.22, 3.22,
and 2.99, each 1H) and oxymethylene protons of a sugar
at d 3.49 and 3.41. The 13C NMR (Table 1) and DEPT
spectra indicated 20 signals including 8 signals from two
pairs of monosubstituted benzene rings, 6 signals from a
pentasubstituted benzene, and 6 signals from a pyrano-
hexose. 1H–1H-COSY, HMQC spectral analyses also
confirmed the presence of two monosubstituted benzene
rings (A and C rings), a pentasubstituted benzene ring
(B ring) and pyranohexose. H-2 at the benzene A ring
and H-200 at C ring displayed HMBC correlations to
C-40 and C-10 in B ring, respectively. These NMR evi-
dences together with the molecular formula suggest that
1 is terphenyl glycoside having two phenolic hydroxy
groups at B ring. HMBC cross-peak (H-60 to C-100)
and NOESY cross-peak (H-60 to H-200) indicated that
benzene C ring is attached to o-position (C-10) of C-60.
Pyronohexose moiety was elucidated to be b-glucose
by NOESY correlations (Fig. 1) and coupling constant
(H-1000, J = 8.0Hz). We could not obtain any clear struc-
tural information from the spectroscopic analyses.
Figure 1. Structure of terfestatin A (1), and HMBC and NOESY
correlations.
2 was treated with acetic anhydride in pyridine at room
temperature for 24h to give triacetate 4.16 We deduced
the p-terphenyl skeleton to be the most plausible struc-
ture by detailed spectral analyses. Therefore, 4 was syn-
thesized by Thiel acetylation from diphenyl
benzoquinone 3.9 The synthesized 4 was completely
identical to triacetylated aglycone from 1. From these
results, the structure of aglycone was determined to be
20,30,50-trihydroxy-p-terphenyl.
Benzene proton at H-60 displayed HMBC correlation
with C-20 (d137.2) and C-50 (d 153.0). The up-field shift
of the C-20 carbon signal was attributed to hydroxyl
group at o-position. The anomeric proton (H-1000)
showed an HMBC cross-peak with C-20. These evi-
dences strongly support that glucose binds to C-20 posi-
tion at B ring.
To determine the structure of aglycone, 1 was hydro-
lyzed by 3N-HCl in MeOH at reflux for 3h to give
the aglycone 2 and D-glucose (Scheme 1). Compound
To confirm the structure of 1, we attempted the total
synthesis of 1 (Scheme 1). 30,50-Dibenzyl protected com-

A. Yamazoe et al. / Tetrahedron Letters 45 (2004) 8359–8362
8361
Scheme 1. Reagents and conditions: (a) 3N-HCl, MeOH, reflux, 3h; (b) Ac2O, pyridine, rt, 24h; (c) HClO4, Ac2O, 55°C to rt, 4h, 80%; (d) KOH,
MeOH, 0°C, 30min, 92%; (e) BnBr, NaH, TBAI, THF, 24h, 87%; (f) 3N-HCl, MeOH, reflux, 2h, 84%; (g) Cs2CO3, AcCl, CH3CN, ꢀ40°C, 1h,
73%; (h) BnBr, NaH, TBAI, THF, rt, 24h, 92%; (i) KOH, MeOH, 50°C, 2h, 90%; (j) Cs2CO3, acetobromo-a-D-glucose, rt, 24h, 97%; (k) H2, Pd–C,
EtOAc, rt, 24h, 92%; (l) KOH, MeOH, rt, 1h, 92%.
pound 9 was designed as the key intermediate to intro-
duce b-D-glucoside at C-20 position in our route and 3
was selected as a starting material. Triacetate 4 was
selectively hydrolyzed with KOH in MeOH at 0°C for
30min to yield diacetate 5. This diacetate 5 was treated
with benzyl bromide, sodium hydride, and tetra-n-butyl-
ammonium iodide in THF at room temperature for 24h
to give 6. Diol 7 was obtained by the hydrolysis of 6 with
3N-HCl in MeOH at reflux for 2h. For selective protec-
tion on C-20 position, 7 was mixed with cesium carbon-
ate and acetyl chloride in acetonitrile at ꢀ40°C for an
hour to afford monoacetate 8.16 The introduction of
benzyl group to C-30 was carried out with the treatment
of 8 with benzyl bromide, sodium hydride, and tetra-n-
butylammonium iodide in THF at room temperature
for 24h. The key dibenzyl protected intermediate 916
was obtained by the hydrolysis of 8 with potassium
hydroxide in MeOH. The structure of 9 was confirmed
by 2D NMR spectra including NOESY spectrum.
the hydrogenolysis (H2, Pd–C) and hydrolysis of acetate
(KOH in MeOH) to yield 1. Overall yield of 1 was 27%
from starting material 3. Physico-chemical, spectral, and
biological properties of synthetic 1 completely agreed
with those of 1 from natural origin.
There are few reports on the isolation of 20,30,50-trisub-
stituted p-terphenyl compounds from natural sources.
Terphenyllins11 and terprenins12 from Aspergillus candi-
dus, terferol13 from Streptomyces showdoensis, and
butlerins14 from Parmelia butleri have been reported.
These p-terphenyl derivatives have O-alkyl or acyl
groups at benzene B-ring. However, there are no reports
on p-terphenyl glycoside. Thus, terfestatin A is the first
p-terphenyl glycoside.
The inhibitory activities of 1 against auxin-induced gene
expression were assayed by the transgenic Arabidpsis
BA-GUS reporter line carrying the GUS gene under the
control of auxin-responsive promoter.5 Compound 1
(20lM) inhibited auxin-induced reporter gene expression
by histochemical staining of GUS enzymatic activity. In
addition, the quantitative GUS activity measurement by
the fluorometric substrate, 4-methyl-umbelliferyl-b-D-
glucuronide, showed that IC50 of 1 against auxin-induced
GUS gene expression in BA-GUS line was 10lM. On the
other hand, GUS enzymatic activity was not affected by 1
at 50lM. In contrast to auxin-induced gene expression, 1
showed no inhibitory effects at 50lM on estrogen-in-
duced reporter gene expression in pER 8::GFP.15 Arabid-
opsis transgenic plant, suggesting 1 is a specific inhibitor
of auxin signaling leading to auxin-responsive gene
expression, but not an inhibitor of the general translation
and transcription process. The detailed biological activi-
ties of 1 will be published elsewhere.
We performed several glycosylation methods of phe-
nolic compounds to introduce a b-glucoside at C-20
position of 9. First, 9 was treated with diisopropylethyl-
amine or Lewis acid and acetobromo-a-D-glucose, but
no reaction proceeded. We next carried out the glycosy-
lation with acetobromo-a-D-glucose as glycosyl donor
and tributyltin methoxide and tin(IV) chloride as cata-
lyst.10 This reaction afforded 10 in low yield (16%).
Therefore, we developed a new glycosylation reaction
to the phenolic group using cesium carbonate and aceto-
bromo-a-D-glucose. Phenol 9 was treated with an equi-
molecular amount of cesium carbonate and threefold
excess of acetobromo-a-D-glucose in acetonitrile at
room temperature for 24h to give 1016 in very high yield
(97%). The benzyl ether 10 was further deprotected by

8362
A. Yamazoe et al. / Tetrahedron Letters 45 (2004) 8359–8362
[a]D = (c 0.1, MeOH) ꢀ37.2°. Compound 4: Yellow
Acknowledgements
powder, IR (KBr): m 3063, 1765, 1628, 1371, 1192, 1045,
916, 757, 702cmꢀ1, UV (MeOH): kmax (nm) (loge) 251.2
We thank Dr. Yutaka Oono (Japan Atomic Energy
Institute), Dr. Athanasios Theologis (Plant Gene
Expression Center, USDA), Dr. Nam-Hai Chua
(Rockefeller University) and Dr. A. M. Jones (Univer-
sity of North Carolina) for providing transgenic repor-
ter lines and useful suggestions.
1
(4.57), mp 189.7°C, H NMR (500MHz, CDCl3): d 7.47
(d, 2H, J = 5.0Hz), 7.33–7.42 (m, 6H), 7.31 (d, 2H,
J = 5.0Hz), 7.13 (s, 1H), 2.09 (s, 3H), 1.96 (s, 6H), 13C
NMR (500MHz, CDCl3): d 168.8, 167.9, 167.8, 146.0,
141.5, 138.3, 136.0, 135.9, 131.8, 129.5 (2C), 129.4, 128.7
(2C), 128.4 (2C), 128.0 (2C), 127.8, 121.9, 100.5, 20.4, 20.3,
20.1, HR-EIMS: m/z [M]+; calcd for C24H20O6, 404.1260.
Found, 404.1287. Compound 8: white powder, IR (KBr):
m 3424, 3059, 1742, 1566, 1205 1063, 745, 697cmꢀ1, UV
References and notes
1
(MeOH): kmax (nm) (loge) 258.0 (4.64), mp 148.0°C, H
NMR (500MHz, CDCl3): d 7.34–7.51 (m, 9H), 7.23–7.29
(m, 4H), 7.16–7.19 (m, 2H), 6.64 (s, 1H), 5.21 (s, 1H, OH),
5.02 (s, 2H), 2.12 (s, 3H), 13C NMR (500MHz, CDCl3): d
169.4, 153.7, 145.7, 137.5, 136.9, 135.2, 131.8, 130.8 (2C),
129.1, 129.0 (2C), 128.7 (2C), 128.3 (2C), 128.3 (2C),
128.3, 127.7, 127.6, 126.7 (2C), 118.2, 106.3, 70.6, 20.5,
HR-EIMS: m/z [M]+; calcd for C27H22O4, 410.1518.
Found, 410.1505. Compound 9: white powder, IR
(KBr): m 3498, 3061, 1600, 1381, 1227, 1098, 754,
696cmꢀ1, UV (MeOH): kmax (nm) (loge) 260.8 (4.61),
mp 117.3°C, 1H NMR (500MHz, CDCl3): d 7.60–7.63 (m,
4H), 7.38–7.49 (m, 6H), 7.32–7.35 (m, 1H), 7.19–7.30 (m,
7H), 7.06–7.08 (m, 2H), 6.87 (s, 1H), 5.76 (s, 1H, OH),
4.94 (s, 2H), 4.42 (s, 2H), 13C NMR (500MHz, CDCl3): d
149.3, 144.2, 141.2, 137.6, 137.3, 136.3, 133.5, 130.9 (2C),
129.0 (2C), 128.6 (2C), 128.5 (2C), 128.5, 128.3 (2C), 128.3
(2C), 128.0 (2C), 127.5, 127.4, 127.3, 127.0 (2C), 126.8,
124.6, 111.9, 75.4, 71.8, HR-EIMS: m/z [M]+; calcd for
C32H26O3, 458.1882. Found, 458.1881. Compound 10:
white powder. IR (KBr): m 3503, 3288, 3057, 1746, 1609,
1405, 1229, 1043, 755, 698cmꢀ1, UV(MeOH): kmax (nm)
(loge) 260.4 (4.63), mp 208.6°C, 1H NMR (500MHz,
CDCl3): d 7.51–7.53 (m, 2H), 7.35–7.44 (m, 8H), 7.19–7.29
(m, 8H), 6.94–6.96 (m, 2H), 6.80 (s, 1H), 5.15 (d, 1H,
J = 5.0Hz), 4.97–5.06 (m, 5H), 4.86 (d, 1H, J = 10Hz),
4.58 (d, 1H, J = 10Hz), 3.96 (dd, 1H, J = 12.3, 3.8Hz),
3.75 (dd, 1H, J = 12.3, 2.1Hz), 3.37–3.40 (m, 1H), 1.95 (s,
6H), 1.93 (s, 3H), 1.85 (s, 3H), 13C NMR (500MHz,
CDCl3): d 170.5, 170.2, 169.3, 169.1, 152.8, 149.7, 140.3,
138.2, 137.0, 136.8, 136.0, 133.5, 130.8 (2C), 129.5 (2C),
128.3 (2C), 128.1 (2C), 128.1 (2C), 128.0 (2C), 127.8, 127.7
(2C), 127.5, 127.3, 127.1, 126.7 (2C), 126.1, 111.0, 100.2,
75.0, 73.1, 71.8, 71.5, 70.7, 67.9, 61.2, 20.7, 20.6, 20.5, 20.5,
HR-FABMS: m/z [M+Na]+; calcd for C46H44O12Na,
811.2731. Found, 811.2720.
1. Walker, L.; Estelle, M. Curr. Opin. Plant Biol. 1998, 1,
434–439.
2. Abel, S.; Nguyen, M. D.; Theologis, A. J. Mol. Biol. 1995,
251, 533–549.
3. Liscum, E.; Reed, J. W. Plant Mol. Biol. 2002, 49, 387–
400.
4. Abel, S.; Oeller, P. W.; Theologis, A. Proc. Natl. Acad.
Sci. U.S.A. 1994, 91, 326–330.
5. Oono, Y.; Chen, Q. G.; Overvoorde, P. J.; Kohler, C.;
Theologis, A. Plant Cell. 1998, 10, 1649–1662.
6. Hayashi, K.; Ogino, K.; Oono, Y.; Uchimiya, H.; Nozaki,
H. J. Antibiot. 2001, 54, 573–581.
7. Hayashi, K.; Jones, A. M.; Ogino, K.; Yamazoe, A.;
Oono, Y.; Inoguchi, M.; Kondo, H.; Nozaki, H. J. Biol.
Chem. 2003, 278, 23797–23806.
8. The active strain was identified by the analysis of 16S
ribosomal DNA sequences carried out by NCIMB Japan.
9. Cain, B. F. J. Chem. Soc. 1961, 936–940.
10. Clerici, F.; Gelmi, M. L.; Mottadelli, S. J. Chem. Soc.,
Perkin Trans. 1 1994, 985–988.
11. Takahashi, C.; Yoshihira, K.; Natori, S.; Umeda, M.
Chem. Pharm. Bull. 1976, 24, 613–620.
12. Kamigauchi, T.; Sakazaki, R.; Nagashima, K.; Kawam-
ura, Y.; Yasuda, Y.; Matsushima, K.; Tani, H.; Takaha-
shi, Y.; Ishii, K.; Suzuki, R.; Koizumi, K.; Nakai, H.;
Ikenishi, Y.; Terui, K. J. Antibiot. 1998, 51, 445–450.
13. Nakagawa, F.; Enokita, R.; Naito, A.; Iijima, Y.;
Yamazaki, M. J. Antibiot. 1984, 37, 6–9.
14. Elix, J. A.; Gaul, K. L.; Hockless, D. C. R.; Wardlaw,
J. H. Aust. J. Chem. 1995, 48, 1049–1053.
15. Zuo, J.; Niu, Q. W.; Chua, N. H. Plant J. 2000, 24, 265–
273.
16. Spectral data of 1, 4, 8, 9, 10. Compound 1: IR (KBr): m
3393, 1625, 1420, 1071, 757, 701cmꢀ1, UV (MeOH) kmax
(nm) (loge) 224.0 (5.00), 264.0 (4.83), mp 134.7°C,