Novel antifungal β-amino acids: Synthesis and activity against Candida albicans

Article abstract of DOI:10.1016/S0960-894X(02)00958-7

A series of novel β-amino acids has been synthesized and tested for their in vitro antifungal activity against Candida albicans. A steep SAR was observed. β-Amino acid 21 (BAY 10-8888/PLD-118) revealed the most favourable activity-tolerability profile and was selected for clinical studies as a novel antifungal for the oral treatment of yeast infections.

Full text of DOI:10.1016/S0960-894X(02)00958-7

Bioorganic & Medicinal Chemistry Letters 13 (2003) 433–436
Novel Antifungal ꢀ-Amino Acids: Synthesis and Activity Against
Candida albicans
Joachim Mittendorf,a,* Franz Kunisch,b Michael Matzke,a Hans-Christian Militzer,b
Axel Schmidtc and Wolfgang Schonfeldc,y
aMedicinal Chemistry, Business Group Pharma, Bayer AG, D-42096 Wuppertal, Germany
bCentral Research, Bayer AG, D-51368 Leverkusen, Germany
cAntiinfective Research, Business Group Pharma, Bayer AG, D-42096 Wuppertal, Germany
Received 24 September 2002; revised 28 October 2002; accepted 29 October 2002
Abstract—A series of novel b-amino acids has been synthesized and tested for their in vitro antifungal activity against Candida
albicans. A steep SAR was observed. b-Amino acid 21 (BAY 10-8888/PLD-118) revealed the most favourable activity–tolerability
profile and was selected for clinical studies as a novel antifungal for the oral treatment of yeast infections.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
Here we report the synthesis of a variety of represent-
ative cyclic b-amino acids (3–51, Table 1) and their in
vitro antifungal activity against C. albicans.
Major increases in the incidence of systemic fungal
infections caused by the yeast Candida albicans have been
observed duringthe last two decades, particularly in
immunocompromised patients.1 A critical need exists for
new antifungal agents to treat these life-threatening
infections.2
Chemistry
Several strategies were employed to synthesize the cyclic
b-amino acids listed in Table 1.5,6
The 2-aminocyclohexenecarboxylic acid 1, originally
designed as pyridoxal phosphate suicide inhibitor,
turned out to also have activity against C. albicans3
(Scheme 1). However, in toxicological studies 1 showed
a less favourable profile. Alongwith the reported anti-
fungal activity of the natural b-amino acid cispentacin
24 this prompted us to initiate a derivatization program
to identify cyclic b-amino acids with superior oral effi-
cacy and tolerability.
The synthesis of example 6 was accomplished as descri-
bed in Scheme 2 startingfrom dihydropyrane 52, which
was derived from unnatural l-glucose.7
Example 21 was prepared in a straightforward man-
ner as depicted in Scheme 3.8 In the key step a
highly enantioselective, quinine-mediated alcoholysis
of the meso-anhydride 60 provided cinnamyl ester
61 (84% yield) with eeꢀ97%. Subsequent Curtius
rearrangement and Pd-catalyzed removal of the cinna-
myl protectinggroups afforded 21 with eeꢀ99.5%. This
process was successfully used to produce 5 kgof
b-amino acid 21. The absolute configuration of 21 was
assigned by X-ray crystallography. Cispentacin 2 as
well as examples 7 and 25 were prepared in an ana-
logous fashion.8
Scheme 1. Antifungal b-amino acids.
*Correspondingauthor. Tel.:+49-202-36-4352; fax:+49-202-36-5461;
e-mail: joachim.mittendorf.jm@bayer-ag.de
yCurrent address: PLIVA d.d., Research Division, Prilaz baruna Fili-
povica 25, 10000 Zagreb, Croatia.
The trans-diastereomer 26 was obtained via isomeriza-
tion of the protected b-amino acid ester obtained from
the Curtius rearrangement of 61 usingDBU in refluxing
0960-894X/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00958-7

434
J. Mittendorf et al. / Bioorg. Med. Chem. Lett. 13 (2003) 433–436
Scheme 5. (a) LDA, THF, ꢂ78 ꢁC, then EtO2C-CN, DMPU, 61%; (b)
PhCH2NH2, cat. pTsOH, CH2Cl2, 54%; (c) 60 bar H2, Pt/C, EtOH,
35 ꢁC, 76%; (d) 3 bar H2, Pd/C, 0.1 N HCl, EtOH, H2O, then (Boc)2O,
NEt3, CH2Cl2, 98%; (e) 2-nitrophenyl selenocyanate, P(nBu)3, THF,
then H2O2, 90%; (f) LiOH, H2O, THF, 98%; (g) TBDMS-OTf, 2,6-
lutidine, CH2Cl2, 52%.
Scheme 2. (a) NaOMe, MeOH, 98%; (b) TBDMSCl, NEt3, DMAP,
CH2Cl2, 84%; (c) HN3, DEAD, PPh3, THF, 64%; (d) PPh3, THF,
H2O, then (Boc)2O, NEt3, CH2Cl2, 99%; (e) TBAF, THF, 78%; (f)
cat. RuCl3, K2S2O8, KOH, H2O, CH2Cl2, then, CH2N2, Et2O, 42%;
(g) 4 N HCl, dioxane, 85%; (f) 3 N HCl, reflux, 100%.
Scheme 6. (a) Allene, CH2Cl2, hv, ꢂ70 ꢁC, 43%; (b) KOCl, KOH,
H2O, then, (Boc)2O, Na2CO3, dioxane; (c) DCC, DMAP, MeOH,
CH2Cl2, 34% (68) and 8% (69); (d) LiOH, H2O, THF; (e) 4 N HCl,
dioxane, 68%; (f) TBDMS-OTf, 2,6-lutidine, CH2Cl2, 23%.
Scheme 3. (a) EtOH, H2SO4; (b) NaOMe MeOH; (c) HCl, H2O; (d)
EtOH, H2SO4, 75%; (e) Ph3PMe+Brꢂ, KOtBu, THF, then KOH,
THF, H2O, 71%; (f) (EtCO)2O, 135 ꢁC, 75%; (g) 1.0 equiv quinine,
1.5 equiv (2E)-3-phenyl-2-propane-1-ol, toluene, ꢂ15 ꢁC, 4 h, 84%; (h)
(PhO)2PON3, NEt3, toluene, 90 ꢁC, then 3-phenyl-2-propane-1-ol, tol-
uene, reflux, 80%; (i) 0.05 mol% Pd(OAc)2, PPh3, morpholine, EtOH,
85%.
procedure described in Scheme 5, exemplified by the
synthesis of compound 34 startingfrom 62.9
Further analogues were prepared via Hofmann degra-
dation with hypochlorite as exemplified in Scheme 6 for
the synthesis of analogues 35 and 36.
The trifluoromethyl substituted derivative 42 was pre-
pared from cispentacin 2 by treatment with SF4/HF at
110 ꢁC.10 Compounds 43 and 44 were obtained starting
from diethyl 2-oxocyclopentanephosphonate11 and 2-
oxocyclopentanesulfonic acid,12 respectively, via hydro-
genation of the corresponding oximes in the presence of
PtO2 (cat.)/Ac2O.
Scheme 4. (a) TMSCl, NaI, CH3CN, then, FMOC-OSu, Na2CO3,
H2O, dioxane; chromat. separation of isomers, 15 and 22%; (b)
piperidine, 65%; (c) NMO, SeO2, CH2Cl2, MeOH, chromat. separa-
tion of isomers, 28%; (d) H2, Pd/C, EtOH, 93%; (e) MsCl, NEt3,
CH2Cl2, then DBU, THF, 93%; (f) LiOH, THF, H2O, 91%; (g) 4 N
HCl, dioxane, 100%; (h) Br3CCOONa, BnNEt3Cl, CHCl3, 74%; (i)
MeLi, Et2O, ꢂ10 ꢁC, 26%; (j) TBDMS-OTf, 2,6-lutidine, CH2Cl2,
49%; (k) Bu3SnH, hexane, 46%; (l) 4 N HCl, dioxane, 100%; (m) O3,
Me2S, MeOH, 93%; (n) NaBH4, MeOH, 80%; (o) 4 N HCl, dioxane,
97%; (p) HCl, H2O, 80 ꢁC, 83%; (q) HN3, DEAD, PPh3, THF, 81%;
(r) LiOH, H2O, THF, 96%; (s) H2, Pd/C, 0.1 N HCl, EtOH, then, 4 N
HCl, dioxane, 78%.
Other examples were synthesized as described or in
13ꢂ15
analogy to known methods,5,
cycloaddition of chlorosulfonyl isocyanate to the corre-
for example, via
5
spondingalkenes.
Results and Discussion
toluene. Several analogues were synthesized starting
from the protected b-amino acid 49 as key intermediate
(Scheme 4).
All compounds were evaluated for their inhibitory
activity against C. albicans (Table 1).16
2-Aminocyclohexenecarboxylic acid 1 and cispentacin 2
showed potent antifungal activity in this assay (IC50
0.03 and 0.13 mg/L, respectively).
Another series of analogues was synthesized via reduc-
tive amination of b-keto esters followingthe egneral

J. Mittendorf et al. / Bioorg. Med. Chem. Lett. 13 (2003) 433–436
Table 1. In vitro activity of b-amino acids against Candida albicans
435
Compd
Structure
IC50 (mg/L)
Compd
Structure
IC50 (mg/L) Compd
Structure
IC50 (mg/L)
1
0.03
18
32
35a
36a
>256
2
0.13
19a
128
>256
3a
4a
0.5
20a
128
37a
38a
>256
(3:1 m.d.)b
128
21
0.13
16
5a
128
64
22a
128
32
39
40a
41a
42a
43a
44a
45
>256
128
32
(3:1 m.d.)b
6
23a
7
64
24a
(5:1 m.d.)b,c
64
8a
32
25
26
27
28
128
128
8
1
9a
>256
>256
128
>256
4
>256
>256
2
10a
11a
32
12a
29
(3:1 m.d.)b,c
16
46a
47
32
13 (5:1 m.d.)b,c
14b (4:1 m.d.)c
30a
(2:1 m.d.)b,c
128
128
16
128
31
48
2
15a
16a
diast. Ac 8
diast. Bc 16
32a
64
128
32
49
50a
51
16
128
32
(single diast.)c
64
33a
(2:1 m.d.)b,c
17a
256
34a
aRacemic.
bm.d., mixture of diastereomers.
cConfiguration not known.

436
J. Mittendorf et al. / Bioorg. Med. Chem. Lett. 13 (2003) 433–436
A very limited study on the structure–activity-relation-
ship (SAR) of cispentacin 2 as previously described13
indicated strict structural requirements for antifungal
activity. The same trend could be observed in this more
extensive study. Whereas dehydro-cispentacin 3 showed
only slightly lower potency (IC50 0.5 mg/L), transposi-
tion or hydrogenation of the double bond in 1 resulted
in a significant loss of activity (examples 4, 5). Likewise,
insertion of heteroatoms (examples 6–10), reduction of
ringsize ( 11) or methyl substitution (13–17) had a
negative impact on antifungal activity. A variety of
open-chain analogues, as exemplified by compound 12,
were inactive.
as PLD-118 by Pliva, Croatia for the oral treatment of
yeast infections.
Acknowledgements
The authors are grateful to D. Arlt, P. Babczinski, U.
Geschke, D. Habich, W. Hartwig, J. Hotho, S. Martitz,
M. Plempel, E. Schwenner, W. Schroeck and K. Ziegel-
bauer for support and helpful discussions, P. Binger and
A. Marhold for intermediates and A. Goehrt for X-ray
analysis.
We also investigated the SAR at position 4 of the
cyclopentane ringin more detail. Amongthese deriva-
tives (18–24), only the introduction of an exo-methylene
group resulted in strong antifungal activity. Compound
21 (IC50 0.13 mg/L) was equipotent to cispentacin 2 and
selected for further derivatizations.
References and Notes
1. (a) Pfaller, M. A. J. Hosp. Infect. 1995, 30, 329. (b) Dixon,
D. M.; McNeil, M. M.; Cohen, M. L.; Gellin, B. G.; La
Montagne, J. R. Public Health Rep. 1996, 111, 226. (c)
Edmond, M. B.; Wallace, S. E.; McClish, D. K.; Pfaller, M. A.;
Jones, R. N.; Wenzel, R. P. Clin. Infect. Dis. 1999, 29, 239.
2. Hossain, M. A.; Ghannoum, M. A. Expert Opin. Investig.
Drugs 2000, 9, 1797.
3. Kunisch, F.; Babczinski, P.; Arlt, D.; Plempel, M. Ger.
Offen. DE 4028046 A1, Chem. Abstr. 1992, 117, 20486.
4. (a) Konishi, M.; Nishio, M.; Saitoh, K.; Miyaki, T.; Oki,
T.; Kawaguchi, H. J. Antibiot. 1989, 42, 1749. (b) Kawabata,
K.; Inamoto, Y.; Sakana, K.; Iwamoto, T.; Hashimoto, S. J.
Antibiot. 1990, 43, 513. (c) Iwamoto, T.; Tsujii, E.; Ezaki, M.;
Fujie, A.; Hashimoto, S.; Okuhara, M.; Kohsaka, M.; Ima-
naka, H.; Kawabata, K. J. Antibiot. 1990, 43, 1. (d) Oki, T.;
Hirano, M.; Tomatsu, K.; Numata, K.; Kamei, H. J. Antibiot.
1989, 42, 1756.
The (1R, 2S)-configuration of 21 turned out to be
essential, since stereoisomers 25 and 26 demonstrated
only weak antifungal activity. Again, a very steep SAR
was observed, resultingin significant loss of potency,
when the double bond of 21 was shifted to other posi-
tions (27–29, 34), small substituents were introduced
(30–33) or the ringsize was altered ( 35–38). Modifi-
cations of the carboxyl (39–47) and the amino sub-
stituent (48–51) indicated that both groups are crucial
for potent in vitro activity against C. albicans. Methyl
ester 45 and dipeptides such as 48, however, showed
strongin vivo antifungal activity probably due to pro-
teolytic cleavage in plasma to release 21.
5. Fulop, F. Chem. Rev. 2001, 101, 2181.
6. Mittendorf, J.; Kunisch, F.; Plempel, M. Ger. Offen. DE
4217776 A1, Chem. Abstr. 1994, 121, 83988.
7. Grynkiewicz, G. Carbohydr. Res. 1984, 128, C9.
8. Mittendorf, J. Eur. Pat. Appl. EP 805145 A1. Chem. Abstr.
1997, 127, 359100.
9. Murata, S.; Suzuki, M.; Noyori, R. Tetrahedron Lett. 1980,
21, 2527.
10. Feldhoff, R.; Haas, A.; Lieb, M. J. Fluorine Chem. 1990,
49, 225.
11. Lee, K.; Wiemer, D. F. J. Org. Chem. 1991, 56, 5556.
12. Prepared in analogy to: Djerassi, C. J. Org. Chem. 1948,
13, 848.
b-Amino acid 21 (BAY 10-8888) exhibits its antifungal
activity by a unique dual mode of action.17 First, it is
accumulated about 200-fold in yeast cells by active
transport via permeases specific for branched-chain
amino acids. Inside the cell 21 inhibits specifically iso-
leucyl-tRNA synthetase, resultingin inhibition of pro-
tein synthesis and cell growth. In contrast, active
transport and inhibition of protein synthesis of cis-
pentacin 2 appears to be mediated by the corresponding
enzymes specific for proline.17 Actingas mimetics of
small amino acids may explain the observed narrow
SAR of antifungal b-amino acids.
13. Ohki, H.; Inamoto, Y.; Kawabata, K.; Kamimura, T.;
Sakane, K. J. Antibiot. 1991, 44, 546.
14. Duncia, J. V.; Pierce, M. E.; Santella, J. B. J. Org. Chem.
1991, 56, 2395.
15. Manfre, F.; Kern, J.-M.; Biellmann, J.-F. J. Org. Chem.
1992, 57, 2060.
Amongall so far prepared b-amino acids 21 exhibited
the most favourable activity-tolerability profile and was
selected for further development. It showed high efficacy
in rat and mouse systemic candidiasis models including
azole-resistant strains18 and a favourable pharmaco-
kinetic (almost 100% oral bioavailability in rats, dogs,
rabbits; 7 h half-life in man) and safety profile.
16. In vitro testingwas performed by a microbroth dilution
technique (96-well microtitre plates) with an inoculum of 103
cfu / mL of C. albicans ATCC 36082 in YNB medium (yeast-
nitrogen base powder, 6.7 g/L, glucose 10 g/L, pH7). IC50-
values were determined after incubation for 24 h at 37 ꢁC.
Read-out was performed photo-/nephelometrically on a Spec-
tra III Elisa reader at 360 nm.
17. Ziegelbauer, K.; Babczinski, P.; Schoenfeld, W. Anti-
microb. Agents Chemother. 1998, 42, 2197.
In conclusion, an extensive chemical optimization on
b-amino acids revealed very steep SAR for antifungal
activity and led to the identification of the novel
b-amino acid 21. An efficient asymmetric synthesis
could be developed for this compound. BAY 10-8888 21
is currently beinginvestigated in phase II clinical studies
18. Schoenfeld, W.; Mittendorf, J.; Schmidt, A.; Geschke, U.
Abstracts, 41st Interscience Conference on Antimicrobial
Agents and Chemotherapie, Chicago, IL, Sept 22–25, 2001;
American Society for Microbiology: Washington, DC, 2001;
F-2144.