The design and synthesis of novel inhibitors of NADH:ubiquinone oxidoreductase

Article abstract of DOI:10.1016/j.bmcl.2003.10.022

Potent new inhibitors of NADH:ubiquinone oxidoreductase (complex I) have been designed, with the help of molecular modelling, by hybridisation of known complex I inhibitors with inhibitors of cytochrome c oxidoreductase. The most interesting compound was the chromone 7 which was a selective inhibitor of complex I (IC₅₀ 15 nM) and showed acaricidal activity against spider mites.

Full text of DOI:10.1016/j.bmcl.2003.10.022

Bioorganic & Medicinal Chemistry Letters 14 (2004) 511–514
The design and synthesis of novel inhibitors of NADH:ubiquinone
oxidoreductase
Stephen D. Lindell,* Oswald Ort, Peter Lummen and Robert Klein
Bayer CropScience GmbH, Werk Hochst, G836, D-65926, Frankfurt am Main, Germany
¨
Received 30 July 2003; revised 8 October 2003; accepted 8 October 2003
Abstract—Potent new inhibitors of NADH:ubiquinone oxidoreductase (complex I) have been designed, with the help of molecular
modelling, by hybridisation of known complex I inhibitors with inhibitors of cytochrome c oxidoreductase. The most interesting
compound was the chromone 7 which was a selective inhibitor of complex I (IC50 15 nM) and showed acaricidal activity against
spider mites.
# 2003 Elsevier Ltd. All rights reserved.
The enzyme complexes NADH:ubiquinone oxido-
reductase (E.C. 1.6.99.3; complex I) and cytochrome c
oxidoreductase (E.C. 1.10.2.2; complex III or bc1 com-
plex) are the first and third electron transfer complexes
of the mitochondrial respiratory chain. Complex I cat-
alyses the transfer of two electrons from NADH to
ubiquinone (coenzyme Q) to give the corresponding
ubihydroquinone. The bc1 complex transfers electrons
from the ubihydroquinone to cytochrome c via the
Rieske iron–sulfur protein and cytochrome c1. These
redox processes result in translocation of protons across
the inner mitochondrial membrane, thereby establishing
an electrochemical (proton) gradient which drives the
synthesis of ATP.1
inhibitors via hybridisation of the complex I inhibitors
1–3 with the bc1 inhibitors 4 and 5 (Fig. 1).
We postulated that the heterocyclic ring systems in
structures 1–5 were mimicking a quinone or hydro-
quinone nucleus and that the lipophilic side chain toge-
ther with the ring substituents were determining the
specificity for either complex I or the bc1 complex. With
the help of molecular modelling, we sought to design
new acaricidal complex I inhibitors by changing the ring
substituents and lipophilic side chains of the bc1 inhi-
biting chromones 4 and 5 to make them more closely
resemble those found in the complex I inhibiting struc-
tures 1–3. As a result of these efforts, the 2-benzylmer-
captochromones 6 and 7 were proposed as potential
synthetic targets. They both contained a lipophilic t-
butylphenyl side chain of the type found in compounds
1–3 and an unsubstituted fused phenyl ring as found in
fenazaquin (1). In addition, the relationship between the
lipophilic side chain and the heterocyclic ring carbonyl
group closely resembled that seen in pyridaben (2).
Using the published X-ray crystal structure of tebu-
fenpyrad (3)7 as a starting point, molecular modelling
studies suggested that the t-butyl carbon atoms and the
heterocyclic ring nitrogens (in 1 and 3)/carbonyl-oxy-
gens (in 2, 6 and 7) of structures 1, 2, 3, 6 and 7 could be
sensibly overlaid. Initial studies were performed using
simple stick models and this work was subsequently
semiquantified using computational methods.8 Thus,
the t-butyl tertiary-carbons and ring nitrogens/carbonyl
oxygens of structures 1, 2 and 7 were superimposed over
the equivalent centres in the X-ray crystal structure of 3.
Several commercial acaricides, including fenazaquine
(1),2,3 pyridaben (2)3,4 and tebufenpyrad (3)4 are potent
inhibitors of complex I which are believed to bind at, or
very close to, a ubiquinone binding site.1 The naturally
occurring chromone stigmatellin (4) and its synthetic
analogue 5 are potent inhibitors of the bc1 complex
which bind at a ubihydroquinone site.5,6 These two bc1
complex inhibitors share the same heterocyclic moiety
but have quite different side chains. In contrast, the
complex I inhibitors 1–3 contain quite different hetero-
cyles but all contain the same 4-t-butylphenyl moiety in
their side chain. This communication describes the
results of our efforts to design new acaricidal complex I
Keywords: Enzyme inhibitors; Acaricides; Insecticides.
* Corresponding author. E-mail: stephen.lindell@bayercropscience.
com
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.10.022

512
S. D. Lindell et al. / Bioorg. Med. Chem. Lett. 14 (2004) 511–514
Figure 1. Inhibitors of the electron transfer complex I and the bc1 complex. The heterocyclic ring nitrogens (in 1 and 3) and carbonyl-oxygens (in 2,
6 and 7) used in the molecular modelling overlay are marked with an asterisk (*).
All structures were then conformationally minimised
within the imposed constraints that the t-butyl tertiary-
carbons and the ring nitrogens/carbonyl-oxygens,
respectively, were confined to stay within 0.15 A of each
other. The resulting overlaid structures are shown in
Figure 2 and considering the relatively few constraints
that were imposed, the degree of overlap was considered
very encouraging. In particular, all four heterocyclic
rings ended up in a very similar plane and the side
chains connecting the heterocyclic and t-butylphenyl
moieties were all well aligned. The minimised structure
for tebufenpyrad (3) remained very close to the pub-
lished X-ray crystal structure.7
Based upon the results of the molecular modelling
studies described above, the chromones 6 and 7 were
synthesised by adapting the methodology described by
Bantick and Suschitzky9 as summarised in Scheme 1.
Biochemical testing on submitochondrial membranes
prepared from house fly flight muscles10 showed that
compounds 6 and 7 inhibited the NADH-dependent
reduction of cytochrome c (IC50 500 and 8 nM, respec-
tively) but had no effect on succinate-dependent cyto-
chrome c reduction. These results indicated that the new
chromones were binding to complex I and not to the bc1
complex. Under the same assay conditions, the Stigma-
tellin analogue 511 was (as expected5) a potent inhibitor
of the bc1 complex (IC50 2 nM) but had no effect on
Figure 2. Diagram showing the superimposed structures 1, 2, 3 and 7.
Scheme 1. Reagents and conditions: (a) 3 equiv t-BuOK, 1.1 equiv CS2, toluene, rt, 4 days then AcOH to pH 5; 30–67%; (b) 1.5 equiv 4-t-
BuC6H4(CH2)nOH, 1.5 equiv EtO2CN¼NCO2Et, 1.5 equiv PPh3, THF, rt, 24 h; 30–64%; (c) from compound 6 (R=H): 1 equiv NCS, AIBN, CCl4,
Á, 5 h; 44%; (d) from compound 7 (R=Me): 3 equiv MCPBA, CH2Cl2, rt, 16 h; 92%.

S. D. Lindell et al. / Bioorg. Med. Chem. Lett. 14 (2004) 511–514
513
Scheme 2. Reagents and conditions: (a) 1 equiv K2CO3, 1.1 equiv MeI, acetone, rt, 20 h; 75%; (b) 1 equiv MCPBA, CH2Cl2, 0 ꢁC, 16 h; 67%; (c)
1.1 equiv 4-t-BuC6H4(CH2)nNH2, MeCN, 50 ꢁC, 24 h; 57–72%.
Table 1. Inhibition of complex 1 (membrane bound) and acaricidal
activity of the chromones 6–13
Compd
R
X
n
Complex I
IC50 (nM)10
TETRUR activity
@ 300 ppm (%)13
6
7
8
9
10
11
12
13
H
Me
Cl
S
S
S
1
1
1
1
1
1
2
2
500
8
6
>200
160
20
10
200
0
100
50
0
0
0
Et
S
Me
Me
Me
Me
SO2
NH
S
Figure 3. Inhibition of purified NADH:Q oxidoreductase and inhibi-
tion of specific radioligand binding to purified complex I from Musca
domestica as a function of chromone 7 concentration. Flight muscle
complex I (25 mg protein per mL) was purified as described; inhibition
of enzyme activity (*); inhibition of specific radioligand binding
(*).10
80
0
NH
(compound 11) gave a reasonable inhibitor (IC50
20 nM) which was, however, biologically inactive. The
effect of elongating the connecting chain between the
chromone and t-butylphenyl moieties was also investi-
gated. Inhibitory potency and significant biological
activity were maintained in the sulfur analogue 12 but
the NH analogue 13 was inactive. Overall, there is a
reasonably good correlation between in vitro and in
vivo activity, whereby compounds with IC50 values of
ꢀ10 nM were biologically active.
complex I activity. Further testing showed that com-
pound 7 inhibited purified soluble house fly complex I10
with an apparent IC50 of 15 nM (Fig. 3), whereas com-
pound 6 was much less active (IC50=118 nM). The
enzymatic results were further substantiated by equili-
brium binding data:10 Compound 7 inhibited the
specific binding of the radiolabelled complex I inhibitor
3H-AE F11920912 in a concentration-dependent manner
(Fig. 3) with an IC50 of 38 nM and an apparent Ki of
25 nM. Taken together, the biochemical results indicate
that the chromone 7 occupies the same, or a closely
overlapping, binding site to the known complex I inhi-
bitors 1–3 and 3H-AE F119209. Biological testing of
compound 7 showed that it possessed acaricidal activity
against spider mites (LD50 ca. 50 ppm) whereas chro-
mones 5 and 6 were inactive.13
In conclusion, we have succeeded in designing and syn-
thesising potent new inhibitors of complex 1, starting
out from known bc1 inhibitors.14 The most interesting
compound, chromone 7, is a selective inhibitor of com-
plex I (IC50 15 nM) and exhibits acaricidal activity
against spider mites.
Acknowledgements
In order to probe the structure–activity requirements of
this new lead, a number of other chromone analogues of
compound 7 were prepared according to Schemes 1 and
29 and tested for biochemical10 and biological13 activity.
The results (Table 1) showed that the 3-chloro com-
pound 8 maintained biochemical and some biological
activity, whereas the 3-ethyl substituted compound 9
was inactive. Oxidation of the sulfur (compound 10) led
to a 20-fold drop in inhibitory potency and loss of bio-
logical activity. Replacement of the sulfur by NH
We wish to thank Dr. P. West for useful discussions,
Mr. T. Goody and Mr. F. Fenkl for technical assistance
and Dr. U. Sanft for the biological testing results.
References and notes
1. Hollingworth, R. M.; Ahammadsahib, K. I. Rev. Pestic.
Toxicol. 1995, 3, 277. Brandt, U. Biofactors 1999, 9, 95.

514
S. D. Lindell et al. / Bioorg. Med. Chem. Lett. 14 (2004) 511–514
2. Hackler, R. E.; Suhr, R. G.; Sheets, J. J.; Hatton, C. J.;
10. Submitochondrial membranes and purified soluble
NADH:Q oxidoreductase from house fly (Musca domes-
tica) flight muscles were prepared according to previously
described methods (Lummen, P. Biochem. Soc. Transact.
1999, 27, 602). Activities of membrane bound redox
complexes were measured photometrically at l=550 nm
using either NADH (complex I+III) or succinate (com-
plex II+III) as electron donors and cytochrome c as
electron acceptor. Cytochrome oxidase was blocked by
4 mM possium cyanide. NADH oxidation by solubilised
complex I was determined photometrically at l=340 nm
with n-decylubiquinone (50 mM) as electron acceptor.
Equilibrium radioligand binding to purified house fly
complex I was performed using the labeled aminopyr-
imidine inhibitor 3H-AE F119209 (Kd=9 nM) as pre-
viously described.12 Kinetic parameters were calculated
with the SigmaPlot software (SPSS).
11. We wish to thank Dr. M. Pettett of Bayer Crop Science,
Lyon for having supplied us with a sample of compound 5
which had been prepared according to the method descri-
bed in ref 5.
12. Okun, J. G.; Lummen, P.; Brandt, U. J. Biol. Chem. 1999,
274, 2625.
13. Test for acaricidal activity against the Two Spotted Spider
Mite: Bean plants (Phaseolus vulgaris) which were infested
with Tetranychus urticae (TETRUR) were dipped into an
aqueous solution of the formulated (EC 10) test com-
pound at a concentration of 300 ppm. After storage in a
controlled environment cabinet at approx. 25 ꢁC for 6
days, the % mortality was determined.
14. For other approaches to the design and synthesis of
hybrid complex I inhibitors, see: Arndt, S.; Emde, U.;
Baurle, S.; Friedrich, T.; Grubert, L.; Koert, U. Chem.
Eur. J. 2001, 7, 993. Yabunaka, H.; Kenmochi, A.;
Nakatogawa, Y.; Sakamoto, K.; Miyoshi, H. Biochim.
Biophys. Acta 2002, 1556, 106.
Johnson, P. L.; Davis, L. N.; Edie, R. G.; Kaster, S. V.;
Jourdan, G. P.; Jackson, J. L.; Krumkalns, E. V. In
Advances in the Chemistry of Insect Control III; Briggs, G.
G., Ed.; RCS: Cambridge, 1994; p. 70.
3. Hollingworth, R. M.; Ahammadsahib, K. I.; Gadelhak,
G.; McLaughlin, J. L. Biochem. Soc. Trans. 1994, 22, 230.
4. Degli Eposti, M. Biochim. Biophys. Acta 1998, 1364, 222.
5. Thierbach, G.; Kunze, B.; Reichenbach, H.; Hofle, G.
Biochim. Biophys. Acta 1984, 765, 227.
6. Zang, Z.; Huang, L.; Shulmeister, V. M.; Chi, Y.-L.; Kim,
K. K.; Hung, L.-W.; Crofts, A. R.; Berry, E. A.; Kim,
S.-H. Nature 1998, 392, 677.
7. Osano, Y. T.; Okada, I.; Okui, S.; Matsuzaki, T. Anal.
Sci. 1991, 7, 181.
8. Energy calculations were performed based on the force-
field method using an augmented AMBER force field
(Weiner, S. J.; Kollman, P. A.; Case, D. A.; Singh, U. C.;
Ghio, C.; Alagona, G.; Profeta, S.; Weiner, P., J. Am.
Chem. Soc. 1984, 106, 765). Gasteiger’s method (Gastei-
ger J.; Marsili, M. Tetrahedron 1980, 36, 3219) was used
to compute approximate atomic charges for the mole-
cules. To overlay the molecules, harmonic constraints
between corresponding atoms were applied. The con-
straint energy term was Econstr=2000 (dꢂ0.15)2 kJif
d
>0.15 A and it was E=0 kJif d <0.15A where d is the
distance between the corresponding atoms. The sum of all
molecular energies and constraint energies was minimised
with respect to the geometries and spatial orientation of
the molecules. Changes in geometry were restricted to
rotations about rotatable bonds. For related work on the
superimposition of different complex I inhibitors, see:
Akagi, T.; Takahashi, Y.; Sasaki, S.-S. Quant. Struct.-Act.
Relat. 1996, 15, 290.
9. Bantick, J. R.; Suschitzky, J. L. J. Heterocyclic Chem.
1981, 18, 679.