Iromycins from Streptomyces sp. and from synthesis: New inhibitors of the mitochondrial electron transport chain

Article abstract of DOI:10.1016/j.bmc.2007.11.023

Two new α-pyridone metabolites, iromycins E and F, were isolated from cultures of strain Streptomyces sp. Dra 17, thus expanding the recently discovered iromycin family. The inhibitory potential on the mitochondrial respiratory chain was examined and reve

Full text of DOI:10.1016/j.bmc.2007.11.023

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 1738–1746
Iromycins from Streptomyces sp. and from synthesis:
New inhibitors of the mitochondrial electron transport chain
Frank Surup,a Heydar Shojaei,a Paultheo von Zezschwitz,a,*
Brigitte Kunzeb,* and Stephanie Gronda,*
aInstitute of Organic and Biomolecular Chemistry, Georg-August-Universita¨t Go¨ttingen, Tammannstrasse 2,
37077 Go¨ttingen, Germany
bResearch Group of Microbial Communication, Helmholtz Centre for Infection Research, Inhoffenstrasse 7,
38124 Braunschweig, Germany
Received 9 July 2007; revised 31 October 2007; accepted 7 November 2007
Available online 13 November 2007
Abstract—Two new a-pyridone metabolites, iromycins E and F, were isolated from cultures of strain Streptomyces sp. Dra 17, thus
expanding the recently discovered iromycin family. The inhibitory potential on the mitochondrial respiratory chain was examined
and revealed that iromycin metabolites block NADH oxidation in beef heart submitochondrial particles with different efficacy, yet
remarkably show only very low cytotoxicity. Difference spectroscopic studies indicated that iromycins inhibit the electron transport
at the site of complex I (NADH-ubiquinone oxidoreductase). Derivatives of the natural products were semisynthetically prepared
and provided detailed insights into structure–activity relationships. Drawn from these results, there are strong similarities with the
piericidins, which are among the most potent complex I inhibitors of the mitochondrial electron transport chain. Furthermore, total
synthesis afforded new analogues, and the non-natural iromycin S (IC50 = 58 ng/mL) emerged as the most active compound, thus
opening avenues of future studies with the iromycins as new valuable biochemical tools.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
of two protons per electron across the membrane in
complex I, which adds to the proton motive force used
for ATP synthesis, however, the precise mechanism is
mainly unsolved.2–4 Complex I is the largest and most
complicated enzyme in mammals, for example, bovine
heart complex I (M = approx. 1 MDa) consists of 45
subunits, seven of them are encoded by mitochondrial
DNA.5
The respiratory chain (electron transport chain) with its
particular functional complexes in the lipid bilayer
membranes is one of the most important biochemical
cascade reactions in the energy supply of all living
organisms.1 The enzymes, which mediate electron trans-
fer and ATP synthesis in a sequence of redox reactions
of the oxidative phosphorylation, are exclusively located
in mitochondria or bacterial membranes, respectively.
The impairment of this energy metabolism is associated
with dysfunction of cells or cell death.
For detailed insights in the function and dysfunction of
complex I and the affiliated physiological effects, specific
inhibitors are used as irreplaceable agents and valuable
biochemical tools. This has led to the identification of
a number of inhibitors and their use in binding studies,6
which already have extensively contributed to the ongo-
ing investigation of the properties of the enzyme and its
as yet unsolved dynamic mode of action.7 Moreover,
several modern pesticides, for example, fenpyroximate,
inhibiting complex I have been successfully brought to
market.8 Presently, inhibitors of complex I are intensely
investigated worldwide as extrogenous mediators of a
wide spectrum of neurodegenerative disorders, including
Parkinson’s disease or progressive supranuclear palsy.9
Thus, the plant-derived annonacin (1), a member of
Complex
I (NADH-ubiquinone oxidoreductase) is
responsible for the first step of the respiratory chain
and transfers electrons from NADH to ubiquinone
(coenzyme Q). This process is coupled with the transfer
Keywords: Complex I inhibitor; Mitochondrial respiratory chain;
Pyridone; Streptomyces; Total synthesis.
*
Corresponding authors. Tel.: +49 551 393095; fax: +49 551
393228; e-mail addresses: pzezsch@gwdg.de; bku@helmholtz-hzi.de;
sgrond@gwdg.de
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.11.023

F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
1739
verticipyrone (6),14 has led to attempts to classify these
compounds into three fundamental types based on their
presumed different modes of inhibition in the ubiqui-
none redox cycle and thus, different binding sites.6 How-
ever, current hypotheses of the inhibitor binding domain
suggest a common large binding pocket, which is con-
structed by multiple subunits and has several partially
overlapping binding positions for the structurally di-
verse inhibitor molecules. Yet, the understanding of
the competitive behavior of established complex I inhib-
itors is hampered by the possibility that the chemical di-
verse molecules might either compete for identical
binding sites or induce structural dynamic changes in
the enzyme which then influence the binding of a second
inhibitor.15,16
OH
OH
O
R
HO
HO
R
R
H
O
O
Annonacin (1)
H
R
OMe
OMe
N
O
O
H
H
MPTP (2)
O
O
Rotenone (3)
Recently, we reported about the isolation and total syn-
thesis of iromycins A–D, a unique family of a-pyridone
metabolites from Streptomyces sp.17,18 The striking
structural similarities and clear differences with inhibi-
tors 5 and 6 led us to evaluate their inhibitory potential
and the site of action within the mitochondrial electron
transport chain. Herein, we report on the isolation of
the new oxidized metabolites E and F, as well as the syn-
thesis of a variety of derivatives that enabled a compre-
hensive study on the structure–activity relationships
(SAR) of the iromycin family.
OH
O
NH
HO
Myxalamid A (4)
OH
4'
MeO
MeO
OH
12
6'
2'
10
N
O
1
Piericidine A1 (5)
2. Organic and biomolecular chemistry
2.1. Microbial metabolites and semisynthetic derivatives
MeO
O
The iromycins A–D (7–10) were recently isolated from
Streptomyces bottropensis sp. Go¨ Dra 17 in yields of
18 mg/L (7), 12 mg/L (8), 0.5 mg/L (9), and 0.25 mg/L
(10).17 With their fully substituted heterocyclic moiety
and the long unsaturated side chain, they show distinct
similarities with both the ubiquinones as well as verti-
cipyrone (6), that was lately isolated from the fungus
Verticillium sp. FKI-1083 as an inhibitor of the NADH
fumarate reductase and the complex I.14 Also structur-
ally related is the piericidin family with about two doz-
ens of compounds. They are produced by different
Streptomyces strains and belong to the most potent
inhibitors of complex I of the mitochondrial electron
transport chain.13,19
Verticipyrone (6)
the acetogenin family, was recently identified as the
molecular cause of neurodegenerations (atypical parkin-
sonism).10 Synthetic MPTP (1-methyl-4-phenyl-1,2,3,6-
tetrahydropyridine, 2) and the plant-derived rotenone
(3) are now regularly used in cell cultures and model
studies of Parkinson’s disease.10,11
The enormous structural diversity of known inhibitors
of complex I, which also includes microbial myxalamids
(e.g., 4)12 and piericidins (e.g., 5)13 as well as the fungal
OH
4
R1
CH3
R2
H
iromycin
A (7)
1´
7
B (8)
CH3 OH
6
9´´
2
C (9)
D (10)
H
H
H
OH
O
N
1´´
H
R1
R2
10´´
HO
4
O
OH
1'
1'
4
HO
HO
O
N
N
1''
1''
10''
10''
iromycin E (11)
iromycin F (12)

1740
F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
The iromycins 7–10 already revealed a highly interesting
biological activity as selective inhibitors of the nitric
oxide synthase (NOS) together with an apparent SAR,
since iromycin A (7) with its terminal isopropyl group
is a stronger inhibitor than the hydroxylated derivative
B (8) or C (9) that possesses just a terminal ethyl group.
Moreover, it was shown that iromycin B (8) is the prod-
uct of a P450-dependent oxidative degradation of iro-
mycin A (7) and thus, iromycin D (10) presumably
stems from a similar oxidation of iromycin C (9).17
These findings prompted a more detailed search for
other oxidized metabolites. Cultivations of the producer
strain Streptomyces sp. Dra 17 in 10 and 50 L fermenters
under the previously reported conditions led to the iso-
lation of two further members of the iromycin family,
which were stained violet with anisaldehyde reagent.17
Iromycin E (11) was obtained from the culture filtrate
by solid phase extraction (XAD-2, MeOH), silica gel
chromatography, gel chromatography (Sephadex
LH20), and reverse phase HPLC in an amount of
0.25 mg/L. Iromycin F (12) was separated from the frac-
tion mainly containing iromycin A (7) by reverse phase
HPLC (0.38 mg/L). Structure elucidation was per-
formed using mass spectrometry and NMR methods,
and data were compared to that of iromycin A.
100-H to 3.10/3.20 ppm and of 7-H to 1.47 ppm. Striking
differences in the carbon NMR spectrum are signals at
dC = 178.1 ppm, as well as at 199.6 and 79.4 ppm indi-
cating a keto group and a hydroxy group, respectively.
The constitution of 12 was fully resolved from 1D and
2D NMR experiments which give evidence for the
known C3- and C10-side chains, and a new ring structure
with an additional hydroxy group at C-3 (see Supple-
mentary Material). Multiple HMBC correlations of 7-
H, 10-H, 100-H, and 1000-H unambigously determined
the structure of the new member of the iromycin family,
the 3H-pyridin-4-one analogue 12. Thus, oxygenation of
iromycin A (7) not only occurs at the isopropyl moiety
of the side chain to furnish the known iromycin B (8),
but also modifies the cyclic moiety of the iromycins to
generate the new iromycins E (11) and F (12).
The non-natural derivatives iromycin AH (13), AM
(14), BM (15), AA (16), and BA (17) were prepared from
the microbial metabolites iromycin A (7) and B (8)
which can be isolated from Streptomyces sp. Dra 17 in
reasonable amounts.17 General protocols for the hydro-
genation with H2 over Pd/C, the methylation with diazo-
methane, and the acetylation with acetic anhydride were
applied.17 Thus, another five iromycin analogues with
chemical changes in the side chain as well as in the het-
erocyclic core were obtained as pure compounds and
used for structural characterization and biological
profiling.
Pure iromycin E (11) was obtained as a colorless oil and
showed low chemical stability even when stored at
ꢀ20 ꢁC. The molecular formula, C19H29NO3, was de-
duced from the ESI-MS spectrum which showed ions
at m/z = 302 ([M+H–H2O]+) and m/z = 320 ([M+H]+).
It was confirmed by HR-ESI-MS (m/z = 320.22202
[M+H]+, D 0.1 ppm) and points to an oxygenated iro-
OH
1
mycin derivative. The H NMR spectrum of 11 shows
iromycin AH (13)
O
O
N
H
a signal pattern similar to that of iromycin A (7) with
only few alterations for the unsaturated side chain, for
example, with 100-H shifted upfield to 2.98 ppm. Strong
chemical shift differences were observed for the hetero-
cyclic ring system. From the 13C NMR, all signals for
the isoprenoid-like side chain (C-6, C-100to C1000) were
readily identified, however, low signal intensities of the
other carbons suggest a dynamic equilibrium, presum-
ably a keto-enol tautomerism, of the molecule in
CD3OD solution. 2D NMR experiments allowed for
the assignment of the 1H and 13C NMR data of the oxy-
genated pyridone moiety (see Supplementary Material).
Observed key HMBC correlations are 100-H (dH 2.98)
and 10-H (dH 1.59) with C-6 (dC 177.6), and 7-H (dH
1.94) with C-3 (dC 106.1) and C-4 (dC 165.7) connecting
the side chains with the heterocycle, and correspond
with the assigned structure 11 for the 5H-pyridin-2-one
metabolite iromycin E.
OR'
iromycin
R
H
R'
AM (14)
BM (15)
AA (16)
BA (17)
Me
OH Me
Ac
OH Ac
N
H
H
R
OH
iromycin
X
M (18)
MO (19)
NH
O
O
X
OH
iromycin
X
R
S (20)
NH Me
SO (21)
AO (22)
O
O
Me
H
O
X
R
The ESI-MS of iromycin F (12), also obtained as a col-
orless oil, showed an [M+H]+ ion at m/z = 320 and a
[2M+Na]+ ion at m/z = 661. The molecular formula,
C19H29NO3, was determined by HR-ESI-MS
(m/z = 320.22227 [M+H]+, D 0.8 ppm). The signal pat-
tern of the NMR spectra of 12 is very similar to both
11 and 7 and suggested a constitutional isomer of 11.
Compared to iromycin A (7), the proton NMR spec-
trum shows significant high field shifts of the signals of
2.2. Total synthesis
The structural novelty and promising biological activity
of the iromycins made their total synthesis an attractive
aim. Therefore, a highly convergent access to these com-

F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
1741
pounds was developed, that consists of a coupling of the
heterocyclic moiety as a pyrone derivative with the side
chains as alkenyl alanates on a late stage of the synthesis
and already led to the formation of iromycin A (7) in
18% overall yield and non-natural iromycin M (18) in
14% overall yield.18 As the NOS inhibition potential of
the iromycins increases with the steric bulk of the end
cap of the side chain—which could be the consequence
of a higher lipophilicity—the synthesis of the respective
tert -butyl derivative, iromycin S (20), was envisaged
and thus of a derivative, which should also resist the
degradation by side chain hydroxylation.
along this sequence offers the additional advantage of
providing the respective O-analogues by simple
debenzoylation of the pyrone intermediates instead
of their treatment with ammonia. Thus, the pyrone
21 (iromycin SO) was obtained by saponification of
compound 33, and the iromycins MO and AO (19,
22) were formed accordingly.
Table 1. Inhibitory effect of iromycins (7–22) on NADH oxidation in
bovine heart submitochondrial particles (SMP) in comparison with
piericidin A1 (5)
Compound
IC50 (ng/mL)
IC50 (lM)
The ring fragment 28 was obtained by propylation,
acylation, and cyclization of the b-ketoester 23 and
subsequent protection of the hydroxy moiety and
functionalization of the 6-methyl group in 24
(Scheme 1).18 The side chain of iromycin S (20)
was prepared from bromide 29, which, in turn, is
readily obtainable from pivaldehyde.20 A copper-cata-
lyzed alkynylation furnished enyne 30 which was
Iromycin A (7)
Iromycin B (8)
Iromycin C (9)
140
5700
1020
>8097
>8097
>8097
680
0.461
17.8
3.52
Iromycin D (10)
Iromycin E (11)
Iromycin F (12)
Iromycin AH (13)
Iromycin AM (14)
Iromycin BM (15)
Iromycin AA (16)
Iromycin BA (17)
Iromycin M (18)
Iromycin MO (19)
Iromycin S (20)
Iromycin SO (21)
Iromycin AO (22)
Piericidin A1 (5)
>26.5
>25.3
>25.3
2.21
890
2.80
21
desilylated and then carboaluminated with AlMe3
>8097
>8097
>8097
295
>24.3
>23.4
>22.4
1.01
to form alkenylalane 32. Treatment of this compound
with nBuLi yielded the corresponding lithium trial-
kylalkenyl alanate which smoothly underwent a cross
coupling reaction with the ring fragment 28 to fur-
nish the benzoylated pyrone analogue 33 of iromycin
S. This compound was then reacted with liquid
ammonia in an autoclave at 70 ꢁC which led to both,
formation of the respective pyridone and debenzoyla-
tion of the hydroxy moiety to yield iromycin S (20)
in 14% overall yield. The formation of the iromycins
2500
58
8.55
0.183
15.1
4800
>8097
2.1
>26.6
0.005
The rate of NADH oxidation without inhibitor was 1.1 0.46
lmol mgꢀ1 minꢀ1
.
OBzl
OBzl
OR
O
O
f
a c
e
OEt
X
O
O
O
O
CHO
O
O
26
23
27 : X = OH
28 : X = Br
24: R = H
g
d
25: R = Bzl
h
j
Br
R
Me2Al
32
29
30 : R = Si(CH3)3
31 : R = H
i
OR
OH
l
k
O
O
O
N
H
33: R = Bzl
21: R = H
20
m
Scheme 1. Total synthesis of iromycins S (20) and SO (21). Reagents and conditions: (a) LDA (2.2 equiv), THF, 0 ꢁC, 0.5 h, then nPrI, 0 ꢁC, 2.5 h;
(b) LDA (2.6 equiv), THF, 0 ꢁC, 1 h, then N-acetylimidazole, ꢀ78 ꢁC, 2 h; (c) DBU, benzene, D, 6 h, 36% from 23; (d) BzlCl, pyridine, rt, 44 h, 89%;
(e) SeO2, dioxane, 130 ꢁC, 16 h, 99%; (f) NaBH4, EtOH, 0 ꢁC ! rt, 4 h, 91%; (g) PBr3, dioxane, rt, 15 h, 99%; (h) Trimethylsilylacetylene, nBuMgBr,
THF, 0 ꢁC ! rt, 2 h, then CuCN, 29, ꢀ5 ꢁC ! rt, 16 h, 90%; (i) NaOH, MeOH/H2O, rt, 4 h, 82%; (j) Cp2ZrCl2, AlMe3, ClCH2CH2Cl, 0 ꢁC ! rt,
2 h, 89%; (k) nBuLi, THF, 0 ꢁC, 0.5 h, then 28, 0 ꢁC, 2.5 h, 79%; (l) NH3(l), 70 ꢁC, 48 h, 63%; (m) NaOH, Dioxan/H2O, rt, 16 h.

1742
F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
a+a3
100
b
c+c1
75
50
25
0
Iromycin A
0.001
0.01
0.1
1
10
Piericidin A1
620 640
Inhibitor concentration [µM]
520
540
560
580
600
Figure 1. Dose-dependant inhibition of NADH-oxidation in bovine
heart submitochondrial particles by iromycin A (7, •) in comparison
with piericidin A1 (5, m). Values represent means SEM of 2–4
independent experiments, respectively. The rate of NADH oxidation
Wavelength (nm)
Figure 2. The effect of iromycin A (7) in comparison to piericidin A1
(5) on the reduction of cytochromes by NADH. — Difference
spectrum (reduced minus oxidized) of SMP reduced with NADH
without inhibitor --- in the presence of 30 lg/mL iromycin A, and ꢁ ꢁ ꢁ
in the presence of 15 lg/mL piericidin A1.
without inhibitor was 1.2 0.2 lmol mgꢀ1 minꢀ1
.
3. Effects on the mitochondrial electron transport and
biological profile
for example, NADH, fully oxidized cytochromes in
front of the block become reduced, while those behind
it remain oxidized. The difference spectrum of NADH-
reduced minus air-oxidized SMP without inhibitor
showed the characteristic absorption maxima for the
different cytochromes (Fig. 2). Treatment of SMP with
either iromycin A (7) or piericidin A1 (5) almost com-
pletely inhibited reduction of cytochrome aa3 (alpha
band at 606 nm), cytochrome b (alpha band at
563 nm), and the cytochromes c + c1 (alpha band at
553 nm) by NADH. These results indicate that the
inhibitory effect of iromycin A (7), like that of piericidin
A1 (5), is mediated via interaction with complex I
(NADH-ubiquinone oxidoreductase) of the eukaryotic
respiratory chain. However, a more exact determination
of the mode of action of iromycins requires further
investigation.
The iromycins and their pyrone analogues 7–22 were
investigated for their potential to act as electron trans-
port inhibitors by testing their inhibitory efficacy on
NADH oxidation in beef heart submitochondrial parti-
cles (SMP) (Table 1 and Fig. 1). Piericidin A1 (5) was in-
cluded as a control. As anticipated from the structural
analogies, some iromycins indeed significantly blocked
the mitochondrial electron transport, in particular the
natural iromycins 7, 9, the semisynthetic 13, 14 and syn-
thetic agents 18, 20.
As shown in Figure 1, iromycin A (7) inhibited NADH
oxidation half-maximally (IC50) at a concentration of
140 ng/mL (0.461 lM). The IC50 value for piericidin
A1 (5) in the same test system was 2.1 ng/ml
(0.005 lM), which is comparable with the values given
in the literature.13 Metabolite 7 with its isopropyl group
as end cap is sevenfold more potent than
9
No activity was observed for iromycins A (7) and B (8)
in a test panel for herbicide, insecticide, and fungicide
activity. To evaluate the cytotoxic potential, compounds
7 and 8 were tested on cancer cell lines HM02 (gastric
adenocarcinoma), HepG2 (hepatocellular carcinoma),
and MCF7 (breast adenocarcinoma) corresponding
NCI-directives, but showed no activity (IC50 > 10 lg/
mL (> 30 lM)). Moreover, 7 is known to be non-toxic
upon intraperitoneal injection in mice (100 mg/kg).22
In addition, we checked the cytotoxicity of 5, 7, 8, 18,
and 20 in parallel, using L929 mouse fibroblasts,23 and
clearly obtained high cytotoxic activity for piericidin
A1 (5, IC50 0.18 ng/mL, 0.43 nM), but no or very low ef-
fect for the tested iromycins. The IC50 for iromycins A
(7), B (8), and M (18), respectively, was at about
20 lg/mL (aprox. 66 lM) and those of iromycin S (20)
was at about 3 lg/mL (10 lM). All results demonstrate
that even the most potent compound of the iromycin
family shows only marginal cytotoxicity in comparison
to piericidin A1 (5).
(IC50 = 1020 ng/mL, 3.5 lM) with its less branched side
chain. Compared to these, the 700-hydroxylated metabo-
lites 8 and 10 and the ring oxidized analogues 11 and 12
display no significant inhibitory activity. Hydrogenation
and methylation of iromycin A (analogues 13 and 14) re-
duces the potency by a factor of 5–6. In contrast, the
other semisynthetic derivatives 15–17 as well as the syn-
thetic pyrone intermediates 19, 21, and 22 are not active.
Remarkably, iromycin M (18) resembling a 500,600-dihyd-
roanalogue of iromycin C (9) is 3.5-fold more potent
than the latter. Yet, the most potent of all these congen-
ers is iromycin S (20), that carries a more bulky t-butyl
group in the side chain and is about 2.5-fold more active
in the complex I inhibition assay than the natural
metabolite 7.
The site of inhibition of iromycin A within the respira-
tory chain was investigated by means of difference spec-
troscopy. Upon reduction with physiological substrates,

F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
1743
4. Discussion of structure–activity relationships
acting as antagonists of the semiquinone intermediate
stabilized within the complex, and thus they could be
similar to rotenone and those piericidin derivatives con-
taining a hydroxy group on the ring and a non-hydrox-
ylated side chain. In contrast, natural piericidins also
show type A inhibition and act as an ubiquinone
competitor.
Isolated as a group of metabolites from Streptomyces
sp., the piericidins (e.g., 5) feature the fully substituted
4-hydroxypyridine core as a ‘cyclic head’, together with
the branched, unsaturated side chain as a ‘hydrophobic
tail’. They have been widely recognized as highly active
complex I inhibitors. They exhibit promising biological
activity in a broad range as insecticides, antimicrobial
or antifungal agents, and also as inhibitors of immune
response proteins. Additionally, the highly selective
antitumor activity and great potency of certain pierici-
dins prompted a detailed investigation of SAR.13,19,24–26
However, the different mechanisms of action have yet
not been established, and the high chemical instability
and significant cytotoxicity of the piericidins impairs their
broad use in biochemical studies.
5. Conclusion
New iromycins have been isolated and synthesized. To
determine structure–activity relationships, their efficacy
on NADH oxidation in beef heart SMP was tested in
comparison to iromycin A (7) and piericidin A1 (5).
The inhibitory potential of the iromycins is clearly mod-
ulated by the structural features discussed here and
shows analogies with the piericidins. However, some
pronounced differences between both metabolite fami-
lies were stated. It was already recognized that the cyto-
toxicity of different piericidin congeners does not
necessarily correlate with their respective complex I inhi-
bition.13 Thus, the marginal cytotoxicity of the iromyc-
ins paired with a significant complex I inhibiton
provides another hint, that additional effects than just
blocking of the electron transport chain might cause
the high toxicity of the piericidins. Moreover, a vasodi-
lating activity was noted for some piericidins,28 while the
iromycins are presumably vasoconstrictors due to their
inhibition of the NOS.17 Clearly, more detailed investi-
gations are necessary to both, broadly evaluate and then
understand the biological profile of these Streptomyces
metabolites. This includes competitive binding studies
of the iromycins and known inhibitors. With the estab-
lished total synthesis at hand, we are now working to-
wards the preparation of hybrids of the iromycins and
piericidins, and additional analogues should be obtained
from precursor-directed biosynthesis.
The iromycins possess similar structural features as the
piericidins (e.g., 5) and verticipyrone (6), in particular
a heterocyclic core structure and a side chain with a
non-conjugated double bond in b,c-position to the
ring. However, the distinct differences between the
compounds investigated in this study allow for the dis-
cussion of structure–activity relationships with the fol-
lowing points:
(a) The most active iromycin S (IC50 = 0.183 lM) and
the natural iromycin A (IC50 = 0.461 lM) are 40- to
90-fold less potent than the known piericidin 5
(IC50 = 0.005 lM). Thus, their lower potency might
either originate from the different type of heterocyclic
moiety, the longer alkyl substituent at C-5, or the short-
er isoprenoid-like side chain.25,26 (b) Non-natural 20
with its t-butyl moiety presumably contains the most
lipophilic side chain, while 8, 10, 15, and 17 are
700-hydroxylated and show no inhibition on NADH oxi-
dation in SMP (IC50 > 17 lM). In contrast, the highly
active piericidin A1 (5) carries a hydroxy group at
C-10 of the side chain, but also tolerates its removal as
long as the 4-hydroxy pyridine moiety remains un-
changed.13,24 (c) The b,c-positioned double bond is an
exceptional feature of the cyanobacterial toxic kalkipy-
rone27 and the complex I inhibitors such as 5, 6, 7,
and 20. Surprisingly, mono-unsaturated iromycin M
(18) only containing this double bond is even threefold
more active than the natural, bis-unsaturated 9. While
complete hydrogenation of the side chain in piericidins
was reported to reduce the potency by a factor of 36,19
the saturated iromycin AH (13) is still more active than
9 and only fivefold less active than its bis-unsaturated
homologue 7. (d) The a-pyrone analogues of the iromyc-
ins turned out to be inactive (IC50 > 9 lM), though the
c-pyrone family of the verticipyrones shows strong
inhibitory capacity (IC50 > 0.046 lM).14 (e) As stated
for the piericidins, the 4-hydroxy function on the hetero-
cycle is of high importance. This does not apply to addi-
tional hydroxy functions on the cyclic head such as in 11
or 12, abolishing the significant acidity of the 4-OH. In
analogy, methylated and acetylated iromycins 14, 15, 16,
and 17 show weak or no inhibitory action, respectively.
(f) With respect to the modes of action classified by
Esposti,6 the iromycins might be type B inhibitors,
Altogether, the new iromycin family sets the stage for
the identification of potent but non-cytotoxic complex
I inhibitors which would be ideal biochemical tools
not only for the investigation of the complex I enzyme,
but also for pharmacologically relevant studies on neu-
rodegenerative diseases.
6. Experimental
6.1. General experimental methods
Fermentations, extractions, and column chromatogra-
phy were performed as described elsewhere.17 For special
equipment and solvents, see Supplementary Material.
HPLC: pump Jasco PU-1587; detector Jasco UV/VIS
1575; software Jasco-Borwin 1.50; preparative HPLC
conditions: column A: Machery& Nagel Nucleodur
100-5 C18 ec. (250 · 8 mm), flow rate: 2.5 mL minꢀ1
;
column B: Grom, Supersphere 100 C18 ec. (4 lm,
100 · 8 mm), flow rate 2.5 mL minꢀ1; column C: Grom,
Supersphere 100 C18 ec. (4 lm, 100 · 20 mm), flow rate
14 mL minꢀ1; HPLC programs: Solution A: H2O, solu-
tion B: acetonitrile, solution C: MeOH; program A:

1744
F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
40% A to 100% C in 25 min, 5 min 100% C; program B:
55% A to 100% C in 25 min, 5 min 100% C. Iromycin E
(11): isocratic 30% B, 70% A, column A. Iromycin F
(12): isocratic 60% B, 40% A, column A. Iromycin AH
(13), AM (14), BM (15), AA (16): program A, column
B. Iromycin BA (17): program B, column B. Butyl lithium
was titrated according to the method of Suffert.29 THF
and hexane were distilled from sodium benzophenone
ketyl, 1,2-dichloroethane was distilled from CaH2.
Iromycins AM (14) and AA (16),17 iromycin M (18) and
4-benzoyloxy-6-bromomethyl-3-methyl-5-propylpyran-
2-one (28),18 and (E)-1-bromo-4,4-dimethylpent-2-ene
(29)20 were prepared as described in the literature.
m 3421, 2959, 2929, 1718, 1636, 1560, 1458, 1380,
1050 cmꢀ1; UV (MeOH): kmax nm (e): 279 (278), 203
(2526); UV (MeOH+NaOH): kmax nm (e): 278 (197),
220 (749); Rf-values: 0.73 (cyclohexane/EtOAc/MeOH
5:10:2), 0.49 (MeOH/H2O 7:3); 1H NMR (300 MHz,
CD3OD): d 0.89 (m, 3H, 30-H3), 0.96 (d, J = 7.0 Hz, 6H,
800-H3, 900-H3), 1.24 (m, 2H, 20-H2), 1.94 (s, 3H, 7-H3),
1.59 (s, 3H, 1000-H3), 2.26 (m, 1H, 700-H), 1.59/2.26 (m,
2H, 10-Ha/10-Hb), 2.66 (d, J = 6.5 Hz, 2H, 400-H2), 2.98
(d, J = 7.0 Hz, 2H, 100-H2), 5.35 (m, 1H, 200-H), 5.34 (m,
1H, 500-H), 5.40 (m, J = 15.5 Hz, 1H, 600-H) ppm; 13C
NMR (75.5 MHz, CD3OD): d 8.6 (+, C-7), 14.3 (+, C-
30), 16.1 (+, C-1000), 22.9 (+, C-800, C-900), 27.8 (ꢀ, C-10),
30.4 (ꢀ, C-20), 32.3 (+, C-700), 35.1 (ꢀ, C-100), 43.8 (ꢀ, C-
400), 71.0 (Cquat, C-5), 106.1 (Cquat, C-3), 118.8 (+, C-200),
126.1 (+, C-500), 138.2 (Cquat, C-300), 140.3 (+, C-600),
165.7 (Cquat, C-4), 176.6 (Cquat, C-2), 177.6 (Cquat, C-6)
ppm.
6.2. Mitochondrial preparations
Submitochondrial particles (SMP) of bovine heart were
obtained by ultrasonic treatment of the mitochondria,
which were isolated by differential centrifugation,
following the protocol of Smith by using a blender to
homogenize the heart mince.30 The initial homogeniza-
tion buffer consisted of 250 mM sucrose, 10 mM
KH2PO4, 10 mM Tris, 2 mM EGTA, 2 mM MgCl2,
pH 7.4. Further isolation procedures were carried out
in the same medium without EGTA.31
6.5. Iromycin F (2,3-dihydroxy-6-[(E,E)-(3,7-dimethyloc-
ta-2,5-dienyl]-3-methyl-5-propyl-3H-pyridin-4-one, 12)
Total amount of isolated metabolite: 3.8 mg. C19H29NO3
(Mr = 319.45); ESI-MS: m/z = 320 [M+H]+, 342
[M+Na]+, 661 [2M+Na]+; HR-ESI-MS: Found:
320.22227, calculated for C19H30NO3 320.22202 (D
20
D
6.3. Measurements of NADH-oxidation and difference
spectra
0.8 ppm); ½aꢂ ¼ ꢀ23ꢃ (c = 1.0, MeOH); IR (KBr): m
3420, 2962, 2933, 2873, 1716, 1662, 1624, 1458, 1384,
1196, 1116, 1062 cmꢀ1; UV (MeOH): kmax nm (e): 312
(1561), 202 (3818); UV (MeOH+NaOH): kmax nm (e):
316 (1295), 206 (3056); Rf-values: 0.25 (CHCl3/MeOH
9:1), 0.74 (cyclohexane/EtOAc/MeOH 5:10:2), 0.24
(MeOH/H2O 7:3); 1H NMR (300 MHz, CD3OD): d
0.91 (t, J = 7.0 Hz, 3H, 30-H3), 0.96 (d, J = 7.0 Hz, 6H,
800-H3, 900-H3), 1.38 (m, J = 7.0 Hz, 2H, 20-H2), 1.47 (s,
3H, 7-H3), 1.71 (s, 3H, 1000-H3), 2.24 (m, 1H, 700-H),
2.10–2.40 (m, 2H, 10-H2), 2.69 (d, J = 6.5 Hz, 2H, 400-
H2), 3.10 (m, J = 16.0, 7.0 Hz, 1H, 100-Ha), 3.20 (m,
J = 16.0, 7.0 Hz, 1H, 100-Hb), 5.16 (m, J = 7.0 Hz, 1H,
200-H), 5.35 (m, J = 15.5, 6.5 Hz, 1H, 500-H), 5.40 (m,
J = 15.5, 6.5 Hz, 1H, 600-H) ppm; 13C NMR
(75.5 MHz, CD3OD): d 14.4 (+, C-30), 16.7 (+, C-1000),
23.0 (+, C-800, C-900), 23.8 (ꢀ, C-20), 27.5 (ꢀ, C-10), 29.2
(+, C-7), 31.3 (ꢀ, C-100), 32.3 (+, C-700), 43.6 (ꢀ, C-400),
79.4 (Cquat, C-3), 114.9 (Cquat, C-5), 119.2 (+, C-200),
125.6 (+, C-500), 140.1 (Cquat, C-300), 140.9 (+, C-600),
154.4 (Cquat, C-6), 178.1 (Cquat, C-2), 199.6 (Cquat, C-4)
ppm.
In both assays, SMP were suspended in 75 mM air sat-
urated sodium potassium phosphate buffer, pH 7.4, with
1 mM each of EDTA and MgCl2. The test compounds
were added in methanolic solution to the reaction cham-
ber and the methanol concentration in all tests did not
exceed 2% (v/v).
The tests were performed at 30 ꢁC and contained in the
different sample preparations 42–219 lg/mL of bovine
protein. After 4 min of preincubation at 30 ꢁC with or
without inhibitors, the reaction was started by the addi-
tion of NADH to give a final concentration of 0.16 mM.
The rate of NADH oxidation was measured within
4 min as a decrease in optical density at 340 nm in a
UV2 Unicam UV/VIS spectrophotometer. Difference
spectra of NADH-reduced minus air-oxidized were
recorded at room temperature in a DW- 2000 UV/VIS
SLM Aminco double-beam spectrophotometer (Ameri-
can Instruments, Silver Springs, MD, USA). The SMP
were suspended to give a final concentration of
2.35 mg protein/mL. The preincubation with or without
inhibitors in the sample cuvette was performed for 2 min
and the reduction of SMP in the sample cuvette was car-
ried out by the addition of NADH (final concentration
2 mM). The bandwidth was 2 nm and the speed 1 nm/s.
6.6. Iromycin AH (6-(3,7-dimethyloctanyl)-4-hydroxy-3-
methyl-5-propyl-1H-pyridin-2-one, 13)
Iromycin A (7, 41.7 mg, 0.137 mmol) was dissolved in
CH3OH (10 mL) and Pd/C (20 mg) was added to the stir-
red solution under H2 atmosphere. After 2 h the suspen-
sion was filtered, and evaporation of the solvents gave
the crude product which was purified using preparative
HPLC (column C, program A) to yield 5.8 mg (14%) of
pure compound 13. C19H33NO2 (Mr = 307.48); ESI-MS:
m/z = 308 [M+H]+; IR (KBr): m 3421, 2963 (sh), 2873,
1654, 1560, 1547, 1375, 1167 cmꢀ1; UV (MeOH): kmax
nm (e): 209 (4192), 287 (1051); UV (MeOH+NaOH): kmax
nm (e): 204 (6564), 217 (4343), 274 (1030). Rf-values: 0.10
6.4. Iromycin E (4,5-dihydroxy-6-[(E,E)-(3,7-dimethyloc-
ta-2,5-dienyl]-3-methyl-5-propyl-5H-pyridin-2-one, 11)
Total amount of isolated metabolite: 2.5 mg. C19H29NO3
(Mr = 319.45); ESI-MS: m/z = 302 [M+H–H2O]+, 320
[M+H]+, 342 [M+Na]+, 661 [2M+Na]+; HR-ESI-MS:
Found: 320.22200, calculated for C19H30NO3 320.22202
20
D
(D 0.06 ppm); ½aꢂ ¼ ꢀ380ꢃ (c = 1.0, MeOH); IR (KBr):

F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
1745
(CHCl3/MeOH 9:1), 0.52 (cyclohexane/EtOAc/MeOH
5:10:2), 0.77 (MeOH/H2O 7:3); 1H NMR (300 MHz,
CD3OD): d 0.88 (d, J = 7.0 Hz, 6H, 800-H3, 900-H3), 0.95
(m, 3H, 1000-H3), 0.96 (m, J = 7.0 Hz, 3H, 30-H3), 1.14-
1.20 (m, 3 H, 400-Ha, 600-H2), 1.28-1.41 (m, 4 H, 200-Ha,
400-Hb, 500-H2), 1.47-1.61 (m, 5H, 20-H2, 200-Hb, 300-H, 700-
H), 1.95 (s, 3H, 7-H3), 2.42 (m, 2H, 10-H2), 2.47-2.59 (m,
2H, 100-H2) ppm; 13C NMR (75.5 MHz, CD3OD): d 8.7
(+, C-7), 14.5 (+, C-30), 19.9 (+, C-1000), 23.0 (+, C-800,
C-900), 24.4 (ꢀ, C-20), 25.9 (ꢀ, C-500), 28.0 (ꢀ, C-10), 29.0
(ꢀ, C-100), 29.2 (+, C-700), 34.2 (ꢀ, C-300), 38.1 (ꢀ, C-200),
38.1 (ꢀ, C-400), 40.4 (+, C-600), 105.8 (Cquat, C-3), 113.3
(Cquat, C-5), 144.7 (Cquat, C-6), 166.0 (Cquat, C-2), 166.3
(Cquat, C-4) ppm.
(10), 93 (10), 81 (17), 57 (100) [C4H9 þ], 41 (39); IR (film):
m 3309, 2960, 2872, 2120, 1478, 1464, 1364, 1260, 972,
1
846, 635 cmꢀ1; H NMR (250 MHz, CDCl3): d 1.01 (s,
9H), 2.08 (t, J = 2.8 Hz, 1H), 2.95 (mc, 2H), 5.31 (dt,
J = 15.6, J = 5.5 Hz, 1H), 5.68 (dt, J = 15.6,
J = 1.5 Hz, 1H) ppm; 13C NMR (62.9 MHz, CDCl3,
DEPT): d 21.6(ꢀ), 29.5(+), 32.8 (Cquat), 69.7(+), 82.3
(Cquat), 118.2(+), 143.3(+) ppm.
6.9. (E,E)-2,6,6-Trimethylhepta-1,4-dienyldimethylalane
(32)
Cp2ZrCl2 (0.274 g, 0.937 mmol) was treated at 0 ꢁC with
AlMe3 (4.68 mL, 9.4 mmol, 2.0 M in hexane, Caution:
pyrophoric!), and the solvent was removed in vacuo. 1,2-
Dichloroethane (5 mL) was added, and the solution was
stirred for 30 min at 0 ꢁC. A solution of enyne 31
(0.572 g, 4.68 mmol) in 1,2-dichloroethane (10 mL) was
added dropwise, and the mixture was stirred for 2 h at
rt. At this point, GC analysis of a hydrolyzed aliquot
showed complete consumption of the enyne. The solvent
and the excess of AlMe3 were removed in vacuo, and hex-
ane (1 mL) was added at 0 ꢁC to precipitate the zirconium
salts. The mixture was filtered through a frit and the solids
were washed with hexane (1 mL). The filtrate was concen-
trated in vacuo to furnish 0.81 g (89%) of alane 32 as a yel-
low oil whose purity was >95% as determined by GC
Iromycin AM (14), BM (15), AA (16), and BA (17) were
prepared according to the known procedures and the
experimental details of 14 and 16 are described else-
where.17 See Supplementary Material for the character-
ization data of 15 and 17.
6.7. (E)-6,6-Dimethyl-1-trimethylsilylhept-4-en-1-yne (30)
nButylmagnesium bromide (22 mL, 35 mmol, 1.6 M in
Et2O) was added dropwise at 0 ꢁC to a solution of tri-
methylsilylacetylene (5.05 mL, 3.51 g, 35.7 mmol) in
THF (40 mL), and the mixture was stirred for 2 h at
rt. The thus obtained trimethylsilylethynylmagnesium
bromide was transferred by cannula to a precooled
(ꢀ5 ꢁC) mixture of (E)-1-bromo-4,4-dimethylpent-2-
ene (29) (4.165 g, 23.52 mmol) and CuCN (0.11 g,
1.2 mmol) in THF (50 mL), and the mixture was stirred
for 16 h at rt. The reaction mixture was poured into sat-
urated NH4Cl solution and extracted with Et2O
(3 · 100 mL). The combined organic phases were dried
over Na2SO4, then filtered, and concentrated in vacuo
(100 mbar). The residue was purified by Kugelrohr dis-
tillation (70 ꢁC oven temperature, 12 mbar) to yield
4.11 g (90%) of compound 30 as a colorless liquid.
C12H22Si (Mr = 194.39); EI-MS: m/z (%)=194 (1) [M+],
179 (3) [M+–CH3], 153 (13), 137 (9) [M+–C4H9], 125
(3), 109 (3), 83 (11), 73 (37) [SiMe3þ], 57 (100)
[C4H9 þ]; IR (film): m 2960, 2177, 1464, 1363, 1250,
1
analysis of a hydrolyzed aliquot. H NMR (250 MHz,
CDCl3): d ꢀ0.79 (s, 6H), 1.01 (s, 9 H), 2.01 (s, 3H), 2.95
(d, J = 5.0 Hz, 2H), 5.24-5.60 (m, 3H) ppm.
6.10. 4-Benzoyloxy-3-methyl-5-propyl-6-[(E,E)-3,7,7-
trimethylocta-2,5-dienyl]pyran-2-one (33)
nBuLi (0.35 mL, 0.81 mmol, 2.3 M in hexane) was added
to a solution of alane 32 (2.4 mL, 1.2 mmol, 0.50 M in
hexane) in THF (1.0 mL) at 0 ꢁC, and the mixture was
stirred for 30 min. A solution of 4-benzoyloxy-6-bromo-
methyl-3-methyl-5-propylpyran-2-one (28) (150.0 mg,
0.411 mmol) in THF (2.0 mL) was slowly added at
0 ꢁC, and stirring was continued for 2.5 h. The reaction
mixture was poured into saturated NH4Cl solution
(3 mL) and extracted with EtOAc (3 · 15 mL). The com-
bined organic phases were dried over MgSO4, then fil-
tered, and concentrated in vacuo. The residue was
purified by flash column chromatography on SiO2
(10 g, hexane/EtOAc 6:1) to yield 137 mg (79%) of pyr-
971, 842, 760 cmꢀ1 1H NMR (250 MHz, CDCl3): d
;
0.19 (s, 9H), 1.00 (s, 9H), 2.95 (dd, J = 1.1, J = 6.6 Hz,
2H), 5.28 (dt, J = 15.5, J = 6.6 Hz, 1H), 5.67 (dt,
J = 15.5, J = 1.1 Hz, 1H) ppm; 13C NMR (62.9 MHz,
CDCl3, DEPT): d 0.1(+), 23.0(ꢀ), 29.5(+), 32.8 (Cquat),
86.0 (Cquat), 104.9 (Cquat), 118.4(+), 143.1(+) ppm.
one 33 (Rf = 0.45) as
a colorless oil. C27H34O4
(Mr = 422.56); ESI-MS: m/z (%)=423 (100) [M+H]+,
846 (41) [2M+H]+; HR-ESI-MS: Found: 423.25326, cal-
culated for C27H35O4 423.25299 (D 0.6 ppm); IR (film): m
2959, 1747, 1718, 1577, 1452, 1363, 1243, 1175, 1107,
1063, 1022, 973, 706 cmꢀ1; 1H NMR (250 MHz, CDCl3):
d 0.87 (t, J = 7.4 Hz, 3H), 1.00 (s, 9H), 1.47 (mc, 2H),
1.68 (s, 3H), 1.95 (s, 3H), 2.23 (mc, 2H), 2.67 (d,
J = 7.3 Hz, 2H), 3.30 (d, J = 7.5Hz, 2H), 5.20-5.31 (m,
2H), 5.48 (d, J = 15.5 Hz, 1H), 7.56 (t, J = 7.1 Hz, 2H),
7.70 (mc, 1H), 8.18 (mc, 2H) ppm; 13C NMR
(75.5 MHz, CDCl3, APT): d 10.5(+), 14.3(+), 16.5(+),
24.2(ꢀ), 28.3(ꢀ), 30.2(+), 31.1(ꢀ), 33.7(ꢀ), 43.8(ꢀ),
114.3(ꢀ), 114.4(ꢀ), 119.4(+), 123.3(+), 129.0(ꢀ),
130.3(+), 131.3(+), 135.8(+), 139.5(ꢀ), 144.7(+),
160.4(ꢀ), 161.5(ꢀ), 163.8(ꢀ), 166.3(ꢀ) ppm.
6.8. (E)-6,6-Dimethylhept-4-en-1-yne (31)
NaOH (40 mL, 40 mmol, 1.0 M in H2O) was added at rt
to a solution of silylalkyne 30 (4.54 g, 23.4 mmol) in
MeOH (80 mL), and the mixture was stirred for 4 h,
poured into saturated NH4Cl solution, and extracted
with pentane (4 · 40 mL). The combined organic phases
were dried over MgSO4, then filtered, and concentrated
by careful distillation at atmospheric pressure using a
20 cm Vigreux column. The residue was purified by Ku-
gelrohr distillation (70 ꢁC, 65 mbar) to yield 2.34 g
(82%) of enyne 31 as a colorless liquid. C9H14
(Mr = 122.21); EI-MS: m/z (%)=121 (7) [M+–H], 108

1746
F. Surup et al. / Bioorg. Med. Chem. 16 (2008) 1738–1746
6.11. Iromycin S (4-hydroxy-3-methyl-5-propyl-6-[(E,E)-
3,7,7-trimethylocta-2,5-dienyl]-1H-pyridin-2-one, 20)
References and notes
1. Nelson, D. L.; Cox, M. M. Lehninger Principles of
Biochemistry; Worth: New York, 2000, Chapter 19.
2. Yagi, T.; Matsuno-Yagi, A. Biochemistry 2003, 42, 2266–2274.
3. Hirst, J. Biochem. Soc. Trans. 2005, 33, 525–529.
4. Sherwood, S.; Hirst, J. Biochem. J. 2006, 400, 541–550.
5. Carroll, J.; Fearnley, I. M.; Skehel, J. M.; Shannon, R. J.;
Hirst, J.; Walker, J. E. J. Biol. Chem. 2006, 281, 32724–32727.
6. (a) Eposti, M. D. Biochim. Biophys. Acta 1998, 1364, 222–
235; (b) Eposti, M. D.; Ghelli, A. Biochem. Soc. Trans.
1999, 27, 606–609.
In an autoclave a mixture of pyrone 33 (61.7 mg,
0.146 mmol) and liquid NH3 (15 mL) was stirred for 48 h
at 70 ꢁC. After the mixture was cooled to rt, the NH3
was carefully evaporated and the residue was diluted with
KHSO4 solution (1.0 M, 3 mL) and extracted with Et2O
(3 · 20 mL). The combined organic phases were dried over
Na2SO4, then filtered, and concentrated in vacuo. The res-
idue was purified by flash column chromatography on SiO2
(8 g, hexane/EtOAc 2:1+1% HOAc) to yield 29.1 mg (63%)
of pyridone 20 (Rf = 0.33) as a colorless oil. C20H31NO2
(Mr = 317.47); ESI-MS: m/z (%) = 340 (10) [M+Na]+, 657
(100) [2M+Na]+; HR-ESI-MS: Found: 318.24282, calcu-
lated for C20H32NO2 318.24276 (D 0.2 ppm); IR (KBr): m
7. Yamashita, T.; Nakamaru-Ogiso, E.; Miyoshi, H.; Matsuno-
Yagi, A.; Yagi, T. J. Biol. Chem. 2007, 282, 6012–6020.
8. Lummen, P. Biochim. Biophys. Acta 1998, 1364, 287–296.
¨
9. Dauer, W.; Kholodilov, N.; Vila, M.; Trillat, A.-C.;
Goodchild, R.; Larsen, K. E.; Staal, R.; Tieu, K.; Schmitz,
Y.; Yuan, C. A.; Rocha, M.; Jackson-Lewis, V.; Hersch,
S.; Sulzer, D.; Przedborski, S.; Burke, R.; Hen, R. Proc.
Natl. Acad. Sci. U.S.A. 2002, 99, 14524–14529.
10. Lannuzel, A.; Michel, P. P.; Ho¨glinger, G. U.; Champy,
P.; Jousset, A.; Medja, F.; Lombes, A.; Darios, F.; Gleye,
C.; Laurens, A.; Hocquemiller, R.; Hisch, E. C.; Ruberg,
M. Neuroscience 2003, 121, 287–296.
11. Perier, C.; Bove, J.; Wu, D.-C.; Dehay, B.; Choi, D.-K.;
Jackson-Lewis, V.; Rathke-Hartlieb, S.; Bouillet, P.;
Strasser, A.; Schulz, J. B.; Przedborski, S.; Vila, M. Proc.
Natl. Acad. Sci. U.S.A. 2007, 104, 8161–8166.
2958, 1627, 1456, 1362, 1218, 1163, 1113, 1029, 972 cmꢀ1
;
1H NMR (300 MHz, CD3OD): d 0.94 (t, J = 7.0 Hz,
3H), 0.99 (s, 9H), 1.47 (mc, 2H), 1.70 (s, 3 H), 1.98 (s,
3H), 2.40 (mc, 2H), 2.65 (d, J = 7.0 Hz, 2 H), 3.28 (mc,
2H), 5.15 (t, J = 7.0 Hz, 1H), 5.28 (dt, J = 15.5,
J = 7.0 Hz, 1H), 5.47 (d, J = 15.5 Hz, 1H) ppm; 13C
NMR (75.5 MHz, CD3OD, APT): d 8.7(+), 14.6(+),
16.5(+), 24.1(ꢀ), 28.0(ꢀ), 30.1(+), 30.4(ꢀ), 33.7(ꢀ),
43.8(ꢀ), 106.0(ꢀ), 113.7(ꢀ), 121.1(+), 123.4(+), 138.7(ꢀ),
143.0(ꢀ), 144.7(+), 165.96(ꢀ), 165.99(ꢀ) ppm.
12. Beyer, S.; Kunze, B.; Silakowski, B.; Muller, R. Biochim.
Biophys. Acta 1999, 1445, 185–195.
¨
6.12. Iromycin MO (19), SO (21), and AO (22)
13. Schnermann, M. J.; Romero, F. A.; Hwang, I.; Naka-
maru-Ogiso, E.; Yagi, T.; Boger, D. L. J. Am. Chem. Soc.
2006, 128, 11799–11807.
14. Ui, H.; Shiomi, K.; Suzuki, H.; Hatano, H.; Morimoto,
H.; Yamaguchi, Y.; Masuma, R.; Sunazuka, T.; Shimam-
ura, H.; Sakamoto, K.; Kita, K.; Miyoshi, H.; Tomoda,
H.; Omura, S. J. Antibiot. 2006, 59, 785–790.
The respective 4-benzoyloxy-3-methyl-6-(octa-2-enyl)-5-
propylpyran-2-one derivative (50 lmol) was dissolved in a
solution of NaOH (0.5 M in dioxane/water 1:1, 1.5 mL)
and stirred for 16 h at rt. The solution was carefully acidi-
fied with hydrochloric acid (1 N) to approx. pH 2 and ex-
tracted with EtOAc (3 · 15 mL). The extracts were dried
over Na2SO4, then filtered, and concentrated in vacuo.
The crude product was purified by column chromatogra-
phy on silica gel to furnish the respective 4-hydroxy-3-
15. Murai, M.; Ishihara, A.; Nishioka, T.; Yagi, T.; Miyoshi,
H. Biochemistry 2007, 46, 6409–6416.
16. Okun, J. G.; Lummen, P.; Brandt, U. J. Biol. Chem. 1999,
¨
274, 2625–2630.
17. Surup, F.; Wagner, O.; von Frieling, J.; Schleicher, M.;
Oess, S.; Muller, P.; Grond, S. J. Org. Chem. 2007, 72,
methyl-6-(octa-2-enyl)-5-propylpyran-2-one
See Supplementary Material for characterization data.
derivatives.
¨
5085–5090.
18. Shojaei, H.; Li-Bo¨hmer, Z.; von Zezschwitz, P. J. Org.
Chem. 2007, 72, 5091–5097.
19. Yoshida, S.; Takahashi, N. Heterocycles 1978, 10, 425–467.
20. Hoffmann, R. W.; Brinkmann, H.; Frenking, G. Chem.
Ber. 1990, 123, 2387–2394.
21. Negishi, E.-i.; Van Horn, D. E.; Yoshida, T. J. Am. Chem.
Soc. 1985, 107, 6639–6647.
Acknowledgments
This work was financially supported by the Bundes-
ministerium fur Bildung und Forschung (GenoMik+:
¨
MetabolitGenoMik) and the Deutsche Forschungs-
gemeinschaft (SFB 416). The authors are indebted to
H.J. Langer for valuable discussions and technical
assistance. We thank B. Engelhardt for her technical
assistance and J. Reinheimer/BASF AG, W. Beil/Han-
nover Medical University, and F. Sasse/Braunschweig
Helmholtz Centre for Infection Research for conduct-
ing biological profiling tests. S.G. and P.v.Z. are grate-
ful to Prof. Axel Zeeck and Prof. Armin de Meijere,
respectively, for their continuing support.
22. Sukenaga, Y.; Yamazaki, T.; Aoyama, T.; Takayasu, Y.;
Harada, T. Japan Patent JP10 237044, 1998, Chem. Abstr.
1998, 129, 244203S.
23. Mosmann, T. J. Immunol. Meth. 1983, 65, 55–63.
24. Gutman, M.; Kliatchko, S. FEBS Lett. 1976, 67, 348–353.
25. Chung, K. H.; Cho, K. Y. Z. Naturforsch. 1989, 44c, 609–
616.
26. Yoshida, S.; Nagao, Y.; Watanabe, A.; Takahashi, N.
Agric. Biol. Chem. 1980, 44, 2921–2924.
27. Graber, G. M.; Gerwick, W. H. J. Nat. Prod. 1998, 61,
677–680.
28. Kominato, K.; Watanabe, Y.; Hirano, S.-I.; Kioka, T.;
Terasawa, T.; Yoshioka, T.; Okamura, K.; Tone, H. J.
Antibiot. 1995, 48, 99–102, 103–105.
Supplementary data
29. Suffert, J. J. Org. Chem. 1989, 54, 509–510.
30. Smith, A. L. Methods Enzymol. 1967, 10, 81–86.
31. Thierbach, G.; Reichenbach, H. Biochim. Biophys. Acta
1981, 638, 282–289.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmc.
2007.11.023.