Lambertellin system, the mechanism for fungal replacement of Monilinia fructigena with Lambertella corni-maris without competitive inhibition on agar media

Article abstract of DOI:10.1016/j.bmc.2014.02.051

The 'Lambertellin system' was disclosed which rationally explains the fungal replacement (mycoparasitism) of Monilinia fructigena (M. f., the host) with Lambertella corni-maris (L. corni-maris, the parasite) without competitive inhibition in the simultaneous incubations on agar media. The 'Lambertellin system' involves; (a) L. corni-maris secretes nontoxic lambertellols (1, 2) as the diffusible precursors of the authentic responsible substance 3 regardless of existence of the host M. f.; (b) In the absence of the host, the environment around the parasite is kept under neutral condition, and both 1 and 2 are readily transformed into 3; (c) Lambertellin (3) inhibits not only the host but also the parasite. The parasite degrades 3 for detoxification; and (d) Upon the host M. f. approaching closely to the area where the parasite inhabits, the environment around the parasite becomes acidic to stabilize 1 and 2, which gives them a chance to diffuse into the host area. Then these are gradually transformed into 3 to inhibit the host without damaging the parasite. This mechanism also accords with the progress of 'Natsu-Nenju' disease on apple fruits, which is known to be a mysterious phyto-disease because of two unique stages of its lifecycle, anamorphic (asexual) and teleomorphic (sexual). The 'Lambertellin system' would be categorized as a novel class of alleropathies.

Full text of DOI:10.1016/j.bmc.2014.02.051

Bioorganic & Medicinal Chemistry 22 (2014) 2489–2495
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Lambertellin system, the mechanism for fungal replacement
of Monilinia fructigena with Lambertella corni-maris without
competitive inhibition on agar media
⇑
Akane Hirose, Shinji Kudo, Takanori Murakami, Kazuaki Tanaka, Yukio Harada, Masaru Hashimoto
Faculty of Agriculture and Life Science, Hirosaki University, 3-Bunkyo-cho, Hirosaki 036-8561, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The ‘Lambertellin system’ was disclosed which rationally explains the fungal replacement (mycoparasit-
ism) of Monilinia fructigena (M. f., the host) with Lambertella corni-maris (L. corni-maris, the parasite)
without competitive inhibition in the simultaneous incubations on agar media. The ‘Lambertellin system’
involves; (a) L. corni-maris secretes nontoxic lambertellols (1, 2) as the diffusible precursors of the
authentic responsible substance 3 regardless of existence of the host M. f.; (b) In the absence of the host,
the environment around the parasite is kept under neutral condition, and both 1 and 2 are readily trans-
formed into 3; (c) Lambertellin (3) inhibits not only the host but also the parasite. The parasite degrades 3
for detoxification; and (d) Upon the host M. f. approaching closely to the area where the parasite inhabits,
the environment around the parasite becomes acidic to stabilize 1 and 2, which gives them a chance to
diffuse into the host area. Then these are gradually transformed into 3 to inhibit the host without
damaging the parasite. This mechanism also accords with the progress of ‘Natsu-Nenju’ disease on apple
fruits, which is known to be a mysterious phyto-disease because of two unique stages of its lifecycle,
anamorphic (asexual) and teleomorphic (sexual). The ‘Lambertellin system’ would be categorized as a
novel class of alleropathies.
Received 17 January 2014
Revised 22 February 2014
Accepted 24 February 2014
Available online 5 March 2014
Keywords:
Mycoparasitism
The Lambertellin system
Mechanism
Labeling experiments
Lambertellols
Ó 2014 Elsevier Ltd. All rights reserved.
1
. Introduction
In 1990, Harada, one of the present authors, found that the fila-
In the course of the bioorganic investigations on this mycopar-
asitism, we have isolated instable metabolites lambertellols A (1)
and
B (2) as candidates responsible for this phenomenon
4
,5
mentous fungus Lambertella corni-maris (L. corni-maris) conquers
the other phytopathogenic fungus Monilinia fructigena (M. f.) with-
out remarkable competitive inhibition in the simultaneous cultiva-
(Fig. 1). Considerable amounts of these compounds were yielded
from the culture broth, when L. corni-maris (parasite) was incu-
bated in the presence of M. f. (host). Interestingly, incubation of
1
tion on a PSA (potato-sucrose-agar) medium. Although these two
fungi were smoothly intermixed on the basis of visual observation,
only L. corni-maris was isolated from any place of the medium after
several weeks. Harada also reported that the replacement also
occurred between these fungi on apple fruits. Since the symptom
on the apple fruits quite resembled ‘Natsu Nenju’ disease, he
concluded that this fungal replacement should be the identity of
the disease and that these phenomena would be an example of
mycoparasitism. ‘Natsu Nenju’ disease used to be explained by a
phytopathogen Seclerotinia phaeospora Hori, which featured unique
1O HO
1O HO
2
2
3
3
4
4
O
O
OH
O
O
O
lambertellol A (1): 3S
lambertellol C (5)
(racemate)
lambertellol B (2): 3R
O
O
HO
O
O
O
O
O
O
O
2
,3
anamorphic (asexual) and teleomorphic (sexual) stages,
but
details of this fungus have remained unknown. Harada’s study con-
sequently denied Seclerotinia phaeospora Hori. He suggested that
Hori misunderstood the symptom with the fungal replacement.
O
O
lambertellin (3)
4
(
MW. 270)
⇑
Corresponding author. Tel./fax: +81 172 39 3782.
E-mail address: hmasaru@hirosaki-u.ac.jp (M. Hashimoto).
Figure 1. Structures of lambertellols A (1), B (2), lambertellin (3), lambertellol C (4)
and putative degradation product 5.
http://dx.doi.org/10.1016/j.bmc.2014.02.051
968-0896/Ó 2014 Elsevier Ltd. All rights reserved.
0

2
490
A. Hirose et al. / Bioorg. Med. Chem. 22 (2014) 2489–2495
L. corni-maris alone provided only trace amounts of these metabo-
lites. We found that acidification of the medium to pH of around 4
in vitr o
6
1
, 2
3
dramatically increased the yield of both 1 and 2 even in the
absence of the host M. f. We also revealed that the host made its
surroundings remarkably acidic, while the parasite L. corni-maris
retro-Michael
i) Michael cyclization
ring opening
ii) oxidative aromatization
6
kept the conditions around neutral. On the bases of these findings,
O
HO
O
O
HO
we had proposed that acidic conditions stimulated their biosynthe-
1
oxidative quinone
formation
2
7
sis. Recently, we accomplished a total synthesis of 1 and 2, which
3
O
4
enabled us to utilize their enantiomers and a series of stable ana-
logs for biological experiments. These led to a different conclusion
that L. corni-maris excretes 1 and 2 as the diffusible precursors
which readily transformed into the authentic responsible lamber-
tellin (3) in a non-enzymatic manner after reaching to the area of
host M. f.
O
OH
O
O
II
I
Scheme 1. Mechanism of the transformation of 1 and 2 into 3.
However, these were insufficient to rationally explain the pro-
gress of the fungal replacement without competitive inhibition.
These are;
re-cyclization but at C2, and (iv) oxidative aromatization, as shown
in Scheme 1. This mechanism also explains the interconversion
between 1 and 2 that we observed on silica gel TLC.
5
(
1) The transformation of 1 and 2 should accumulate 3, but this
compound was always detected in a much lower level than 1
and 2.
2.1. Acceleration of the transformation from lambertellols into
lambertellin by increment of pH
(
2) The reason why acidic conditions increase the yield of 1 and
2
has remained unknown.
We first investigated the conditions which contribute to the
(
3) It was unclear why L. corni-maris needed to produce 1 and 2
as precursors of the real toxic substance 3.
transformation of 1 and 2 into 3. Production of 3 was monitored
with its characteristic UV absorption at 430 nm (e 6800). Since
we had experienced prompt equilibration between 1 and 2, only
Our continuous research revealed that the total mechanism
named ‘Lambertellin system’ can rationally explain the above
questions. The ‘Lambertellin system’ ingeniously controls the con-
centration of 3, the authentic toxic substance not only for the host
but also for the parasite. Although we had previously concluded
that the host stimulates biosynthesis of 1–3, the present study
proved that the biosynthesis of these metabolites functions con-
stantly regardless of the existence of the host.
major 2 was employed in the experiments. Formation of 3 was
slow, when 2 was dissolved in acidic aqueous solutions (200 lg/
mL, pH 4.5 and 5.5) as shown in Figure 2. However, that became
remarkable at pH 6.5 and 7.5. Increment of UV absorption at
430 nm seemed to be stopped at around A = 1.5 after 30 min in
these experiments. That was due to saturation of 3, that is, precip-
itation of 3 occurred. These indicated that increment of pH induced
withdrawal of the proton at C2 in 2 to trigger the retro-Michael
type ring opening and the following reactions, giving 3. The pK
a
value for 2-H is around 6.0 based on those results. Since the trans-
formation involves irreversible steps (the quinone formation and
the aromatization in Scheme 1), the reverse-transformation
(3?2) does not proceed.
2
. Results and discussions
We employed two species of mycoparasites Lambertella corni-
maris (L. corni-maris) and Lambertella sp. 1346 (L. sp. 1346) in the
experiments. These had been isolated from apple fruits showing
the similar symptom. Although the former fungus leads the fungal
replacement without competitively inhibiting the host Monilinia
fructigena (M. f.) in the simultaneous cultivation on agar gel, the
These accorded with our precedent results that we detected
considerable amounts of 1 and 2 from L. corni-maris under acidic
conditions such as cultivation with acids and simultaneous cultiva-
tion with the host M. f. The host made its surroundings remarkably
acidic, while L. corni-maris kept the medium around pH 6 under
regular culture conditions. We have already disclosed that the
other parasite L. 1346 acidifies the circumstance and constantly
provides lambertellols in practical yields. Above results revealed
that the acidity in the culture medium attributes to the stabiliza-
tion of 1 and 2 to play an important role in the yield of them.
latter simply results in competitive inhibition. Lambertellols (1,
5
2) were first isolated from the culture broth of the latter fungus.
6
In contrast, the former yields 1 and 2 in only trace amounts under
regular culture conditions. Interestingly, considerable amounts of 1
and 2 were yielded, when the former was cultured in the presence
of the host M. f. or cultured under acidic conditions. We have also
disclosed that M. f. makes its surroundings acidic (around pH 3.1).
In the early stage, we had excluded lambertellin (3) as the can-
didate responsible for the fungal replacement, because 3 gave only
a small inhibition zone against the Monilinia fructicola, a model
2.2. Biodegradation of 3 by the parasites
Prompt transformation of 1 and 2 into 3 suggested that incuba-
tion of L. corni-maris accumulates 3. However, 3 was yielded less
than 1/10 of 1 or 2, as far we investigated. Furthermore, some of
3 must be artificially generated from 1 and/or 2 during isolation
process by taking their lability into account. Thus, 3 should be al-
ways in trace amounts in the L. corni-maris culture medium.
Although 1 and 2 were more stable in the culture medium of L.
sp. 1346 because of acidic conditions, Figure 2 suggests the trans-
formation was not blocked completely. Accumulation of 3 should
occur also in L. sp. 1346 culture medium. But that was not ob-
served, either.
8
fungus of M. f., in the paper disk assay on agar media. However,
we understood the inhibition was due to 3 after the structure–
activity relationship studies employing a series of our synthetic
analogs.7 These studies suggested that lambertellols (1, 2) are
diffusible precursors of the authentic responsible substance 3.
Lambertellin 3 was readily crystallized even in agar gel to give rise
to a low concentration of 3, which resulted in the small inhibition
zone. Transformation of lambertellols into 3 readily proceeded
7
in vitro for example in methanol and on PSA medium. That can
be explained by the sequential reactions of (i) retro-Michael type
butenolide opening affording putative intermediate I, (ii) auto-
oxidation into more stable naphthoquinone II, (iii) Michael type
To overcome the above inconsistency, we assumed a biodegra-
dation of 3 by the responsible fungi, L. corni-maris and L. sp. 1346
themselves. In this series of experiments, we first used L. sp. 1346

A. Hirose et al. / Bioorg. Med. Chem. 22 (2014) 2489–2495
2491
The diametermajor and diameterminor were obtained by measuring
the oval shaped colonies with a ruler through the bottom of the
culture flasks. Since Lambertella species degrade 3 as described in
the preceding section, 3 was supplemented every 24 h as the
DMSO solution (run 3, run 4) to maintain concentration. Blank
DMSO was added into a set of control flasks every 24 h. Quantity
of supplemental 3 (0.6 mg for 100 mL culture medium) was settled
based on the preliminary experiments. The total amount of 3 in
test flasks decreased because of the biodegradation, and denovo 3
slightly increased in the control flasks as the culture proceeded.
However, supplement 3 made its concentrations more than 500
times of those in the control flasks throughout the experiments.
Even just before boosting, concentrations of 3 in test flasks were
2
5–200 times larger than those in the controls.
As shown in Figure 4, L. corni-maris grew well without supple-
Figure 2. Production of lambertellin (3) from lambertellol B (2) monitored with UV
30 nm.
mental 3. Addition of blank DMSO (100
4 h) did not affect the colony growth (run 2). On the other hand,
supplementation of 3 significantly inhibited the growth (run 3, run
). Interestingly, L. corni-maris resumed growing, when the media
lL/100 mL medium, every
4
2
4
were refreshed with the regular one after 48 h (run 4). We also per-
formed similar experiments under acidic conditions. However,
acidification led to scattering results probably due to partial colony
collapses. Experiments with L. sp. 1346 suggested that this fungus
was found to be less susceptible for 3 to lead the results less
remarkable. However, the experiments at higher concentrations
(
5.0 mg in 100 mL medium) exhibited obvious inhibition. The
results are shown in Supplemental data. These findings indicated
that 3 is toxic not only against the host M. f. but also the Lamber-
tella spp. even those are the responsible fungi. Thus, we concluded
that Lambertella spp. need to degrade toxic 3 to keep its concentra-
tion at low level.
Figure 3. Biodegradation of 3 (initially 25 lg/mL) by Lambertella spp.
2
.4. Constant production of 3 by the parasite L. cornimaris
because this fungus makes the surroundings acidic to minimize
denovo 3. The remaining 3 was quantified with UV at 430 nm.
Although 3 is a chemically inert molecule, the amount of 3 dramat-
As described, the mycoparasitism between the parasites Lam-
bertella spp. and the host M. f. involves sequential processes of;
i) biosynthesis of precursors 1 and 2, (ii) transformation into the
authentic active substance 3, and (iii) degradative detoxification
of 3 by the producer Lambertella spp.
We next investigated whether L. corni-maris operates the pro-
cesses constantly (constant production hypothesis) or only tempo-
rally when the host M. f. co-inhabits (temporal production
hypothesis). We have experienced remarkable increments of 1
(
ically decreased after 24 h, as shown in Figure 3, when 3 (25 lg/
mL) was incubated with L. sp. 1346 at room temperature. The
amount of 3 was constant within experimental error, when that
was incubated with an autoclaved L. sp. 1346 culture medium. A
similar tendency was observed by employing L. corni-maris. The
rate of decrement depended on the amount and the conditions of
the fungi, and it was technically difficult to control them com-
pletely. Thus, these experiments were insufficient for detailed
quantitative discussions. But, the decrement by L. corni-maris
was slower than that by L. sp. 1346 with reproducibility. We
assume that denovo 3 in the case of L. corni-maris might be one
of the reasons. As described, these experiments revealed that L.
sp. 1346 and L. corni-maris biologically degrade 3. As a result of
biodegradation, 3 becomes nearly invisible in the culture broths.
2
0
5
1
2
.3. Toxicity of lambertellin against the producer L. corni-maris
10
5
The next query arose as to why the parasites needed to decom-
pose 3 although that was synthesized by themselves. Our previous
studies disclosed that 3 is the authentic toxic substance for the
host M. f. We assumed that 3 is also toxic for the parasites Lamber-
7
0
tella spp.
0
12
24
36
48
60
72
84
With the above assumption, we investigated the growth rate of L.
corni-maris in the presence of 3. Technically, it was difficult to esti-
mate the growth curves of filamentous fungus Lambertella spp. by
way of optical manners. Thus we investigated that by estimating
run 1
run 2
run 3
run 4
regular culture conditions (2 flasks)
blanc DMSO (100 µL) was added in every 24 hr (2 flasks)
3 (0.6 mg in 100 µL DMSO) was added in every 24 hr, (3 flasks)
3 (0.6 mg in 100 µL DMSO) was added at 0, 24, and 36 hr, then
replaced with regular fresh media after 48 hr, (3 flasks)
the relative colony volumes by the following approximation;
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3
relativecolonyvolume¼f ðrel:diametermajorÞꢀðrel:diameterminorÞg
Figure 4. Influence of external lambertellin (3) to L. corni-maris.

2
492
A. Hirose et al. / Bioorg. Med. Chem. 22 (2014) 2489–2495
and 2, when L. corni-maris was simultaneously cultured with the
host M. f. This might support the temporal operation hypothesis,
that is, the host might stimulate their productions. However, that
can be also explained by the constant operation hypothesis by sup-
posing that the rate of the process (iii) is larger than that of both
processes (i) and (ii).
with a large one due to 3. A detailed analysis revealed that a signal
appearing at 13.5 min reflected the isotopologue distribution pat-
1
3
tern of the C labeled 3 as shown in Figure 5, although there
was some dilution with non-labeled denovo 3. The experiment
with non-labeled 3 gave the signal with a simple isotopologue pro-
file. Thus, we assigned that the GC signal at 13.5 min was one of
biodegradation products derived from 3. The mass number m/z
270 suggested that this compound would be o- or p-quinone 4,
an oxidized product form of 3. The SIC at m/z 270 gave another sig-
nal at 15.4 min. However that did not correspond to the biodegra-
dation products, because the isotopologue distribution was
identical between those obtained after incubations with the la-
beled and non-labeled 3.
We studied this subject with mass spectral analyses by utilizing
1
3
C isotope labeled 3. We have already developed a method prepar-
1
3
ing C labeled 1 and 2 with a high incorporation level (>40% aver-
9
age incorporation). The labeled samples thus prepared afforded
characteristic isotopologue distribution patterns of molecular
(
and fragment) ions in the mass spectra. We expected that biodeg-
radation products should preserve the isotopologue distribution,
when the parasite was cultured with 1 C labeled 3 in such a high
incorporation level. We planned to explore the biodegradation
substances with a guidance of the isotopologue distribution pat-
3
Next, the GC signal at 13.5 min was explored in the culture
broth of L. corni-maris without supplemental 3. This parasite gives
3 only in trace amounts under the condition, the signal of 3 was,
however, the biggest GC signal in the TIC. Although the target sig-
nal was hard to recognize in the TIC, the SIC at m/z = 270 showed
that clearly. Lambertellols A and B (1, 2) also existed in the broth,
but these were decomposed into 3 in the GC instrument. Thus,
13
terns after incubation with C labeled 3.
The C labeled 3 was successfully prepared from 13C labeled 1
13
and 2 on silica gel, and was added into the culture broth of L. sp.
1
346. This fungus was chosen to minimize dilution with denovo
1
0
non-labeled 3. After incubation for 40 min, the GC total ion chro-
those were not seen in the GCMS. These results indicated that
matogram (TIC) of the broth provided several small signals along
3 is produced constantly, even it is almost invisible. Metabolites
Figure 5. TIC and SIC (m/z = 270) chromatograms of L. sp. 1346 culture medium after treatment of labeled 3, the molecular ions (t
corni-maris prepared by regular cultivation.
R
= 13.4, 14.9, and 15.4 min), and those of L.

A. Hirose et al. / Bioorg. Med. Chem. 22 (2014) 2489–2495
2493
(
b) In the absence of the host, the environment around the par-
asite L. corni-maris is kept under neutral condition, and both
1
and 2 are readily transformed into 3.
(
c) Lambertellin (3) inhibits not only the host but also the par-
asite. The parasite decomposes 3 for detoxification.
d) Upon the host M. f. approaching closely to the area where
the parasite inhabits, the environment around the parasite
becomes acidic to stabilize 1 and 2, which gives them a
chance to diffuse to the host area. Then these are gradually
transformed into 3 to inhibit the host.
(
The ‘Lambertellin system’ makes 1–3 nearly invisible in the reg-
ular L. corni-maris culture broth, while that works constantly. In
the case, the host is far enough from the parasite, 3 is kept at a
low concentration around host because of its less diffusible prop-
erty as well as prompt degradative detoxification by the parasite
itself. Thus, the host can grow freely. In personification, the host
cannot be aware of the parasite as shown in Figure 7. When the
host approaches closely, the environment around the parasite be-
comes acidic to make 1 and 2 stable, which gives them a chance
to diffuse into the host area. After reaching, these are gradually
decomposed into 3 to inhibit the host. Since the diffusion rate of
the acidic area is nearly equal to that of the host growth on agar
6
gel, we do not observe competitive inhibition in the simultaneous
Figure 6. Increment of lambertellol C (4) by incubation with external 3 for 40 min
and its mass spectra under labeled and non-labeled conditions.
cultivations. The mechanism as described explains the fungal
replacement on agar media in all respects. The ‘Lambertellin sys-
tem’ enables to keep the authentically toxic 3 at a low concentra-
tion, which minimizes damage of the parasite.
1-3 are in trace amounts as a result of dynamic equilibrium among
the processes (i), (ii), and (iii). Thus, the temporal production
hypothesis was experimentally denied.
2.6. Discussions about the fungal replacement on apple fruit
In early stage, we performed the experiments with UV detected
HPLC to observe a remarkable increment of lambertellol C (5)
1
1
L. corni-maris conquers M. f. also on the apple fruits in the fields.
upon the addition of 3 into the L. sp. 1346 culture medium, as
Harada, one of the present authors, reported that this would be the
shown in Figure 6. This peak was confirmed with the retention
time as well as the H NMR spectrum after isolation. These results
real progress of ‘Natsu-Nenju’ disease which had been reported by
1
1–3
Hori in 1912.
Finally, we discuss this phenomenon based on the
might imply a biological transformation of 3 into 5. However, that
was the reverse conversion of that we observed in vitro as shown
in Scheme 1, and hard to understand based on chemistry, because
that transformation involves both enthalpically and entropically
disfavored reductive dearomatization step. This transformation
results we obtained. The skin of apple fruits must have tolerance
against pathogenic microorganisms. Thus, Lambertella spp. cannot
get into healthy apple fruits. Since the host M. f. is also known as
the pathogen of Monilia disease, a serious one in apple production
in the early spring, M. f. might be somehow more infective to the
fruits and it can inroad solely while the detailed infection route
has remained unknown. Propagation of M. f. in the apple fruits
reduces the defense system of the skin, which allows intrusion
by the parasite. Then the intruder Lambertella spp. exterminates
the former occupant M. f. by using 3.
Mycoparasitisms are usually observed in soil. It is not neces-
sary for Lambertella spp. to utilize chemicals of the host M. f.
because of the phenomenon in nutritious fruits. We assume that
co-inhabitation with M. f. is inconvenient for Lambertella spp. due
to some reason. So, they need to wipe out the former occupant
M. f. Both Monilinia and Lambertella species belong to sclerotinia-
ceae family, that is, they are genetically close to each other. Due
to their resembling biological system, Lambertella spp., probably,
could not obtain substances giving damage specifically to M. f.
Lambertella spp. overcame this inconvenience by acquiring such
ingenious ‘Lambertellin system’ in their revolution. This system
enables Lambertella spp. to inhibit the host M. f. with minimum
damage for themselves.
(
1
3 ? 5) was denied experimentally because incubation of L. sp.
346 with C labeled 3 did not affect the isotopologue distribution
1
3
of 5 in the EI mass spectra. The mass spectruma of 5 was obtained
by the direct ionization method after isolation. The GCMS condi-
tions totally decomposed 5 to give 3. Although HPLC provides the
signal of 5 without remarkable decomposition, unfortunately the
LC-ESI and LC-APCI methods were less sensitive for 1–3 and 5, as
far we investigated.
Formation of 4 can be explained by an oxidation of 1 and/or 2. We
presumed that the addition of toxic 3 stimulated secretions of oxida-
tive enzymes in order to detoxify it, and one of those gave rise to the
oxidation of denovo 1 and/or 2 to provide 5. This accorded with the
putative structure 4, because that is also an oxidant form of 3.
1
2
2
.5. The ‘Lambertellin system’
Comprehensive analyses of these results revealed a total mech-
anism of this fungal replacement on the agar media without com-
petitive inhibition as described below. This phenomenon cannot be
explained simply by secretions of antibiotics Hence we named the
3. Conclusion
‘
‘Lambertellin system’’ which is comprised of;
As described, we succeeded in disclosing the ‘Lambertellin
system’ which rationally explains the fungal replacement of M. f.
with L. corni-maris on agar media. This mechanism also accords
with the fungal replacement on apple fruits. ‘Lambertellin system’
(
a) L. corni-maris secretes nontoxic lambertellols (1, 2) as the
diffusible precursors of the authentic responsible substance
3
regardless of existence of the host M. f.

2
494
A. Hirose et al. / Bioorg. Med. Chem. 22 (2014) 2489–2495
Figure 7. The lambertellin system disclosed in the present study.
would be categorized as a novel class of alleropathies. Although
productions of 1 and 2 should be coded in gene, the authentic
responsible 3 is not controlled by the genetic system. The acidity
around the parasite governs its production. Thus, only bioorganic
investigations could dissolve the mechanism in this particular
phenomenon.
4.4. Decomposition of 3 by the parasites
Five media (200 mL) in 500 mL baffled flasks were prepared;
those contained (i) culture broth that Lambertella. sp. 1346 was
cultured for 4 days (ꢀ3), (ii) the culture broth that Lambertella.
sp. 1346 was cultured for 4 days and then autoclaved (ꢀ1), and
(
(
iii) distilled water (ꢀ1). Lambertellin (3, 5.0 mg) in acetone
2.0 mL) was added into each flask. After stirring (100 rpm) at
4
4
. Experimental
room temperature for 24 h, each culture broth was filtered,
and extracted with AcOEt. Each extract was independently dried
over MgSO , filtered and then concentrated in vacuo. Then each
4
.1. General
residue was diluted again with acetone (2.0 mL) and subjected
to HPLC. Recovered 3 was quantified by comparing the peak
area at UV 430 nm with that of the standard solution (5.0 mg
of 3 in 2.0 acetone). Experiments with L. corni-maris were also
performed.
L. corni-maris, L. sp. 1346, and M. f. were provided from Yukio
Harada of Hirosaki University. Isotope labeled and non-labeled
lambertellols A (1) and B (2) were prepared with established pro-
4
,5,9
Ò
cedure.
HPLCs were performed with SunFire C18 (3.5 lm
4
.6 ꢀ 100 mm) employing a Waters 1525 Binary Pump equipped
with a Waters 996 Photodiode Array Detector. Measurements of
Ò
4
.5. Inhibition of L. corni-maris by addition of external 3
GCMS were performed with a RESTREK Rtx -1 (0.25 mm ꢀ 30 m)
using SIMADZU GCMS-QP2010 gas chromatography mass
spectrometer.
Colonies of L. corni-maris were prepared in a 200 mL Erlenmeyer
flask with established procedure but stirred at 60 rpm. Each colony
ca. 0.7 cm) was transferred into a 100 mL flask and filled with
(
4
.2. Non-labeled lambertellin (3)
potato-sucrose medium (100 mL). Totally 10 flasks were prepared.
These flasks were gently shaken (50 rpm) at 24 °C. For the first two
flasks, nothing was added (run 1). Into the second two flasks, blank
Lambertella sp. 1346 extracts (51 mg) were prepared by the
established procedure employing PS medium (1.0 L) and the
4
DMSO (100
flasks, 3 (0.6 mg in 100
Into the last three flasks, 3 (0.6 mg in 100
l
L) was added every 24 h (run 2). Into the third three
L DMSO) was added at 0, 24, 36, and 60 h.
L DMSO) was added at 0
following extraction with EtOAc. The crude extracts were loaded
l
onto a silica gel (Merck 7734) column. Slow development with
l
EtOAc/hexane (70:30) gave non-labeled lambertellin (3, 6.0 mg).
1
and 24 h. For the last three flasks, culture media in these flasks
were removed by decantation after 48 h and then filled with fresh
regular PS medium (100 mL).
The H NMR spectrum of the sample was identical to that of
authentic sample. These operations were repeated when we
needed.
For every 12 h, the major and minor diameters of colonies were
measured from the bottom of the flasks with a ruler. The relative
volumes were estimated by the following approximation;
4
.3. In vitro transformation of 2 into 3
Lambertellol B (2, 300
.5 acetate buffer solution (3.0 mL) and pH 6.5 phosphate buffer
lg) in DMSO (50 lL) was added into pH
relativecolonyvolume
5
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
solutions. After each solution was kept at room temperature for
4 h, the absorbance at UV 430 nm was measured. The similar
experiments were performed with pH 4.5 acetate buffer and pH
.5 phosphate buffer solutions. Obvious yellow colorization was
fðdiametermajorÞ ꢀ ðdiameterminorÞg
:
2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3
fðinitialdiametermajorÞ ꢀ ðinitialdiameterminorÞg
7
When considerable colony collapses were observed, the exper-
iment was stopped. The similar experiments employing L. sp. 1346
with 3 (0.6 and 5.0 mg in 100 mL media) were also performed. The
results are shown in the Supplemental data.
observed in pH 6.5 and 7.5 solution after the experiments. Decom-
position of 2 with 330 nm was also examined, but to give unclear
curves, because both 2 and 3 absorb the UV light at 330 nm.

A. Hirose et al. / Bioorg. Med. Chem. 22 (2014) 2489–2495
2495
4
.6. 13C Labeled 3
HPLC. The similar experiment was performed with non-labeled 3.
Experiment with L. corni-maris in place of L. sp. 1346 was also
performed.
Colonies of Lambertella sp.1346 (ca. 5 ꢀ 5 mm ca. 20 colonies)
were suspended with distilled water (15 mL) and the fungus was
recovered after centrifugation (4000 rpm, 10 min). The recovered
hyphae were filled again with distilled water (15 mL), suspended,
centrifuged (4000 rpm, 10 min), and then recovered. The recovered
hyphae were suspended with PS medium (200 mL) containing so-
dium [1- C]acetate (20 mg). After the resulting broth was cultured
for 2 days at 24 °C, MeOH (200 mL) was added. The resulting sus-
pension was filtered in suction and the filtrate was concentrated
in vacuo until the whole volume became 150 mL. The resulting
solution was extracted with AcOEt, washed with brine, dried over
Acknowledgment
Part of this work was supported by Grants-in-Aids for Scientific
Research (B) (40212138) of Japan Society for the Promotion of
Science. Authors appreciate Miwa Hashimoto who supported
making illustrations.
1
3
Supplementary data
MgSO
9.8 mg). The residue was loaded onto a relatively large amount
of silica gel column (20 g). Slow development with EtOAc/hexane
4
and concentrated in vacuo to give the crude residue
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2014.02.051.
(
1
(
70:30) gave 3 (5.8 mg) in almost pure state based on the H
NMR spectrum. Since, GC analysis of the sample suggested some
small impurities, a part of sample was further purified with HPLC
under the established procedure to afford pure 3 (0.7 mg). The
incorporation rate (ca. 40%) of the sample was checked with
References and notes
1
2
3
.
.
.
Harada, Y.; Sasaki, M. Rep. Tottori Mycol. Inst. 1990, 28, 275.
Hori, S. SHOKUBUTU BYOUGAI KOUWA (in Japanese) 1916, II, 97.
Hori, S. ENGEINO TOMO (in Japanese) 1912, 8, 949.
9
GCMS.
4. Murakami, T.; Takahashi, Y.; Fukushi, E.; Kawabata, J.; Hashimoto, M.; Okuno,
T.; Harada, Y. J. Am. Chem. Soc. 2004, 126, 9214.
5
6
7
.
.
.
Murakami, T.; Morikawa, Y.; Hashimoto, M.; Okuno, T.; Harada, Y. Org. Lett.
004, 6, 157.
Murakami, T.; Takada, N.; Harada, Y.; Okuno, T.; Hashimoto, M. Biosci.
Biotechnol. Biochem. 2007, 71, 1230.
Nomiya, M.; Murakami, T.; Takada, N.; Okuno, T.; Harada, Y.; Hashimoto, M. J.
Org. Chem. 2008, 73, 5039.
4
.7. Exploration of the degradation products from 3
2
The 13C labeled 3 (1.0 mg in 0.5 mL, DMSO, ca. 40% incorpora-
tion based on the isotopolgue distribution) was added into L. sp.
346 culture medium (200 mL) and the medium was incubated
1
8. Monilinia fructigena makes the spores only limited conditions. Monilinia
fructicola is isolated from peach fruit which also belongs to Rosaceae family.
9
for 40 min. Culture broth (2.0 mL) was sampled and extracted with
ethyl acetate (20 mL). After concentration, the residue was diluted
with MeOH (1.0 mL) and then subjected to GCMS (conditions:
.
Murakami, T.; Sasaki, A.; Fukushi, E.; Kawabata, J.; Hashimoto, M.; Okuno, T.
Bioorg. Med. Chem. Lett. 2005, 15, 2591.
10. Lambertellols A and B (1, 2) could be ionized without decomposition when
Ò
Rtx -1, 0.25 mm ꢀ 30 m, 100 ? 320 °C, 20 °C/min and then
these were introduced into the instrument directly (EI method).
1
1
1. Murakami, T.; Sasaki, A.; Fukushi, E.; Kawabata, J.; Hashimoto, M.; Okuno, T.
Bioorg. Med. Chem. Lett. 2005, 15, 2587.
2. Jeffries, P.; Young, T. W. K. Interfungal parasitic relationships; CABI International:
Wallingford, 1994.
3
20 °C for 10 min). The extracts were also analyzed with HPLC
(
UV 350 nm). The signal for lambertellol C (4) was confirmed with
1
EIMS (direct insertion) and H NMR spectrum after preparative