Screening yeasts for the stereoselective reduction of oxoester clofibrate analogues

Article abstract of DOI:10.1016/j.tetasy.2005.02.010

Reduction of oxoesters 1b-d and 1f,g in the presence of different yeast strains (Saccharomyces cerevisiae DSM 11285, S. cerevisiae CBS 7336, Cryptococcus curvatus ATCC 20509, Candida bombicola ATCC 22214, Trigonopsis variabilis DSM 70714, Kluyveromyces ma

Full text of DOI:10.1016/j.tetasy.2005.02.010

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 1473–1477
Screening yeasts for the stereoselective reduction of
oxoester clofibrate analogues
Maria Grazia Perrone,a Ernesto Santandrea,a Antonio Scilimati,a,*
Christoph Syldatkb and Vincenzo Tortorellaa
a
`
Dipartimento Farmaco-Chimico, Universita di Bari, Via E.Orabona 4, 70125 Bari, Italy
bUniversita¨t Karlsruhe (TH), Engler-Bunte-Institut, Lehrstuhl fu¨r Technische Biologie, Engler-Bunte-Ring 1,
76131 Karlsruhe, Germany
Received 12 January 2004; accepted 1 February 2005
Available online 16 March 2005
Abstract—Reduction of oxoesters 1b–d and 1f,g in the presence of different yeast strains (Saccharomyces cerevisiae DSM 11285, S.
cerevisiae CBS 7336, Cryptococcus curvatus ATCC 20509, Candida bombicola ATCC 22214, Trigonopsis variabilis DSM 70714, Kluy-
veromyces marxianus CBS 6556) affords hydroxy esters 2b–d and 2f,g with diastereoisomeric excesses (de) up to >99%. Hydrolytic
enzyme(s) contained in the yeasts catalyzed to some extent the hydrolysis of the oxoesters to the corresponding acids, which undergo
decarboxylation followed by reduction of the carbonyl moiety.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
tiomerically pure 2-(4-chlorophenoxy)-3-hydroxybut-
anoic acid, a new clofibrate analogue.15,16
Clofibrate belongs to a class of drugs used to treat
hyperlipidemias, atherosclerosis and for the prevention
of heart failure. Fibrates are well tolerated in the clinic,
and show favourable safety profiles, even if fibrate-asso-
ciated toxicity has been reported in almost every organ
system.1 The most pronounced contraindication is the
liver and renal insufficiency.2,3
Usually, bakerÕs yeast constitutes the first choice catalyst
to accomplish carbonyl bioreductions, because it is easy
to use and does not require any specific laboratory
equipment. However, we have highlighted that other mi-
crobes, such as Kluyveromyces marxianus CBS 6556, can
also catalyze this asymmetric bioreduction, sometimes
outperforming bakerÕs yeast in terms of stereoselectivity
and activity, as in the case of the reduction of ethyl 2-(4-
chlorophenoxy)-3-oxobutanoate 1a (Scheme 1) (de
>99%, ee(2R,3S) = 97%, and de = 92%, ee(2R,3S) = 94%
for the reaction performed in the presence of bakerÕs
yeast).16 Herein, as a continuation of our previous
In our previous investigation, a number of racemic and
optically active clofibrate analogues were synthesized
and pharmacologically evaluated as peroxisome prolif-
erator-activated receptor activators4 and muscle tissue
chloride channel affecting agents5–7 for their therapeutic
activity and toxic side effects, respectively, with the aim
of discriminating the structural determinants responsi-
ble for the different activities. The configuration of clo-
fibrate analogues having a stereogenic centre also
affects their pharmacological profiles (therapeutic and
adverse side effects).8–14
∗
∗
O
R2
R1
COOEt
O
COOEt
OH
O
R2
R1
yeast
whole cells
Cl
Cl
1a-g
2a-g
Recently, bakerÕs yeast and a number of other yeasts,
valuable tools for stereoselectively reducing prochiral
ketones, have been used by us to prepare almost enan-
a
b
c
d
e
f
g
R1 CH3 C2H5 n-C3H7 i-C3H7 t-C4H9 C6H5 CH3
R2 CH3
H
H
H
H
H
H
*
Scheme 1. Bioreduction of compounds 1a–g by yeasts.
Corresponding author. E-mail: ascilimati@farmchim.uniba.it
0957-4166/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2005.02.010

1474
M. G. Perrone et al. / Tetrahedron: Asymmetry 16 (2005) 1473–1477
investigation aimed at preparing other optically active
clofibrate analogues,15,16 we report the results of the
reduction of the 3-oxoesters 1b–g (Scheme 1) carried
out in the presence of some microbes.
(conv. in 2b = 88%). On the contrary, no diastereoselec-
tivity was observed with both S. cerevisiae, whereas a
de = 50% was seen in the case of bakerÕs yeast. A higher
amount of the hydroxyester 2b (95%) and de = 40%
were found in the case of C. bombicola. T. variabilis gave
83% of 2b and de = 60% after 4 reaction days. It is note-
worthy that the reduction of 1b mediated by K. marxi-
anus was quite unsatisfactory, when compared to the
behaviour of the same yeast in the reduction of 1a.16
In particular, a moderate increase in the side chain
length (a methyl in 1a, an ethyl in 1b) caused a substan-
tial lowering of both 2b formation (69%) and de (54%),
together with a remarkable increase in the reaction time
(5 days vs 8 h). When K. marxianus was used to mediate
the reduction of 1c, in which the side chain is an n-pro-
pyl, 31% of 2c formed. For the reduction of 1c reaction
times were always higher (4–11 days) than those with
bakerÕs yeast (1c quantitatively converted within 1.5
days) while 2c percentages were always lower, ranging
between 0% and 69%. Comparable diastereoisomeric ex-
cesses were obtained with both strains of S. cerevisiae
(de = 60% and 66%), although they were lower than
bakerÕs yeast, which gave a de (89%) similar to that of
T. variabilis (de = 86%). A substantial de (>98%) was
obtained in the presence of C. bombicola.
2. Results and discussion
The reductive properties of a number of yeast strains
(Saccharomyces cerevisiae DSM 11285, S. cerevisiae
CBS 7336, Cryptococcus curvatus ATCC 20509, Candida
bombicola ATCC 22214, Trigonopsis variabilis DSM
70714, K. marxianus CBS 6556) in the presence of com-
pounds 1b–g have been evaluated in an attempt to devel-
op a diastereo- and enantioselective method to
transform 3-oxoesters into 3-hydroxyesters.
Compound 1a was diastereo- and enantioselectively
converted into 2a (as said in Section 1) by bakerÕs yeast
mediated reduction.15,16 The results of the reduction of
1a–g performed in the presence of the above mentioned
yeast strains are summarized in Table 1.
All the substrates were successfully reduced with the
exception of 1e. No product isolation and characteriza-
tion was possible for 2e: in all the conditions used, the
conversion was lower than 1%. Reported data (Table
1) show that each yeast behaves differently for each
substrate.
Results of the bioreduction of 1d depended on the yeast
strain used, and in any case represent an improvement
of the data obtained for the reaction performed in the
presence of bakerÕs yeast. In fact, 1d was quantitatively
converted into the product from both strains of S. cere-
visiae within 1day, while 88% of product was obtained
with C. bombicola after 5 incubation days. These results
were better than 43% conversion observed within 4 days
As far as 1b is concerned, after 5 reaction days, the con-
version into 2b mediated by both strains of S. cerevisiae
(conv. in 2b = 84%), is comparable to that obtained car-
rying out the reduction in the presence of bakerÕs yeast
Table 1. Results of the whole cell-mediated reduction of 1a–d and 1f,g
Compd Product/reaction Saccharomyces Saccharomyces Candida
Cryptococcus Trigonopsis
curvatus variabilis
ATCC 22214 ATCC 20509 DSM 70714 CBS 6556
Kluyveromyces BakerÕs yeasta
marxianus
time/de
cerevisiae
DSM 11285
cerevisiae
CBS 7336
bombicola
1ac
1b
1c
1d
1f
2a (%)b
Rt (h)
89
59
24
56
4
911 80
Quantitative
29
>99
Quantitative
8
8
8183
24
32
25
32
3
De (%)d
92
2b (%)b
84
5
84
5
95
15 h
40
2183
n.d.
69
88
Rt (days)
De (%)d
4
5
5
None
None
60
54
11
50
2c (%)b
60
11
60
6125
11
0
11
69
4
3Q1uantitative
Rt (days)
De (%)d
11
86
1.5
66
>98
None
None
89
2d (%)b
Quantitative
115
94
Quantitative
16
91>99
88
10
4
76
50
43
Rt (days)
De (%)d
6
None
None
None
70
Quantitative
2f (%)b
57
8
96
8
0
7
135
4
0
7
Rt (days)
De (%)d
11
None
3
>99
>99
>99
None
54
None
1g
2g (%)b
Quantitative
Quantitative
Quantitative Quantitative Quantitative
93
6
Quantitative
Rt (days)
De (%)d
4
4
4
3
10 h
4
4
6
10 h
3
2
2
52
a See Ref. 15.
b Percentage (%) of 2a–d and 2f,g formed (GC–MS quantitative analysis).
c See Refs. 15,16.
d syn > anti pair (1H NMR data).

M. G. Perrone et al. / Tetrahedron: Asymmetry 16 (2005) 1473–1477
1475
in the presence of bakerÕs yeast. Diastereoisomeric ex-
cesses were very high, ranging from 91% to >99%, com-
pared to the de = 70% for bakerÕs yeast. No de was
observed in the transformation of 1d into 2d performed
in the presence of Cr. curvatus, T. variabilis and
K. marxianus.
The reduction of 1c took longer reaction times (up to
11 days, Tables 1and 2 ), this allowed the formation
of significantly higher amounts of 3c and 4c, which
ranged from 25% to 69%. Within 7 days, 3c and 4c
were quantitatively formed in the presence of Cr. curv-
atus. No 3d and 4d formation was observed by reduc-
ing 1d with both strains of S. cerevisiae. It is
noteworthy that in all biotransformations, these two
strains gave the lowest amounts of the secondary prod-
ucts (632%).
Diastereoisomeric excesses were still very high when 1f
was reduced in the presence of both strains of S. cerevi-
siae (de >99%). Although, after 8 reaction days, conver-
sion decreased by using S. cerevisiae DSM 11285 (57%;
quantitative conversion in 3 days for bakerÕs yeast), it
was again very high (96%), when S. cerevisiae CBS
7336 was employed.
With 1f, very high amounts of 3f and 4f were obtained
using Cr. curvatus (97%, within 4 days), C. bombicola
(62%, within 7 days) and T. variabilis (60%, within
7 days), while with K. marxianus, 36% of these products
were found.
The reduction of 1g always proceeded with complete
conversion into 2g, as with bakerÕs yeast. Reaction times
were usually higher (4–6 days compared to 2 days of
bakerÕs yeast). Using C. bombicola and T. variabilis,
reaction was even faster (10 h). The observed diastereo-
selectivity was comparable to that of bakerÕs yeast (2%),
with the exception of K. marxianus, which reduced 1g to
2g with a significantly higher de (52%).
On the other hand, compound 1g showed a striking sta-
bility towards the enzymatic hydrolysis: only the ex-
pected product 2g was always obtained. The presence
of a methyl on C2 in 1g probably does not allow a suit-
able interaction with the hydrolase active site.
In summary, among all the yeasts screened, Cr. curvatus
expresses the highest level of hydrolytic enzyme(s)
responsible for the formation of 3b–d and 3f. Bioreduc-
tions mediated by T. variabilis are not completely satis-
factory even though product formation was acceptable
with conversions ranging between 35% (1f) and quanti-
tative (1g). Stereoselection was often low to fair (de = 3–
60%), except for compound 1c, for which the highest
diastereoisomeric excess was observed (86%). Further-
more, similarly to C. bombicola and Cr. curvatus, T.
variabilis displays remarkable hydrolytic activity to-
wards all the reduced substrates. Both strains of S. cere-
visiae had the same behaviour towards all of the
compounds screened, showing a remarkable reducing
activity and high conversions. Furthermore, bakerÕs
yeast still constitutes a good biocatalyst alternative, at
least in the case of compounds 1b, 1c and 1f. According
It is noteworthy that in performing these bioreductions,
other products 3b–d, 3f, 4b–d and 4f, (Scheme 2) formed,
probably due to the simultaneous presence in the reac-
tion medium of hydrolase(s) responsible for the hydroly-
sis of the ester function, followed by spontaneous
decarboxylation of the intermediate 3-oxoacids,17
affording ketones 3b–d and 3f, which were quickly biore-
duced to 4b–d and 4f (Scheme 2). Compounds 3b–d, 3f,
4b–d and 4f did not form in blank experiments.17
All microbes, with the exception of Cr. curvatus, showed
after incubation times up to 5 days with 1b, a moderate
percentage of 3b and 4b, ranging between 5% (C. bombi-
cola) and 26% (K. marxianus). Cr. curvatus rapidly con-
verted almost all (79%) the substrate into 3b and 4b
(Table 2).
O
R1
COOEt
O
O
COOH
O
O
O
yeast
not catalyzed
yeast
*
Cl
Cl
R1
Cl
R1
O
Cl
R1
OH
4b-d, 4f
1b-d, 1f
3b-d, 3f
Scheme 2. Hydrolysis of 1b–d and 1f, decarboxylation of the produced acids to 3b–d and 3f reduced, then, to 4b–d and 4f.
Table 2. Reaction time (Rt) and percentage (%) of 3b–d, 3f and 4b–d, 4f formed (together with 2b–d, 2f) from 1b–d, 1f incubated in the presence of
different yeasts
Yeast
3b + 4
Rta (days)
b
+ 4
3c
c
+ 4
3d
d
+ 4f
3f
b
b
%
b
%
b
%
Rta (days)
Rta (days)
Rta (days)
%
Saccharomyces cerevisiaec
Cryptococcus curvatus
Candida bombicola
5
16
79
5
11
7
32
100
66
24
29
0
7
4
7
7
7
0
5 h
1
90
12
24
45
97
62
60
36
11
11
7
96
Trigonopsis variabilis
Kluyveromyces marxianus
4
17
26
25
69
125
101
5
a Reaction time (Rt) are those reported in Table 1.
b Percentage (%) of formation of 3 and 4 was determined in the reaction crude by GC–MS quantitative analysis chromatogram (see Section 4).
c Same results were obtained from both strains of Saccharomyces cerevisiae.

1476
M. G. Perrone et al. / Tetrahedron: Asymmetry 16 (2005) 1473–1477
Table 3. Best yeast-mediated reduction conditions for 1a–d and 1f,g [reaction time (Rt), diastereoisomeric excess (De) values and percentage (%) of
2a–d and 2f,g formed]
Compd
Yeast
Rt (h)
Product
%
De (%)
1a
1b
1c
1d
1f
Kluyveromyces marxianus
BakerÕs yeast
BakerÕs yeast
Saccharomyces cerevisiae DSM
8
120
40
2a
2b
2c
2d
2f
Quantitative
88
>99
50
Quantitative
Quantitative
Quantitative
Quantitative
89
94
24
72
BakerÕs yeast
Candida bombicola
>99
4
1g
10
2g
to the reported data, the best yeast-mediated reduction
conditions for each of the substrates 1b–d and 1f,g are
reported in Table 3.
4.2. Culture medium
The microorganisms were cultivated under aerobic con-
ditions in a medium containing 0.3% yeast extract, 0.3%
malt extract, 0.5% peptone, 1% glucose. Agar-agar (2%)
was added to the same medium for cells preservation on
agar slants.
3. Conclusion
In conclusion, even though the synthesized substrates
1a–d and 1f,g have a very high degree of structural sim-
ilarities, a specific yeast for the reduction of each of
them was identified, enabling us to obtain 2a–d and
2f,g with high conversion (88% to quantitative) and
low to very good diastereoselectivities (4% to >99%).
This justifies the continuous effort aimed at finding
new microorganisms for the preparation of valuable
enantiomerically pure compounds for fine chemicals
and pharmaceuticals.
4.2.1. Bioreduction of 1b–f using yeasts: general proce-
dure. Cells preserved on agar slants at 4 ꢁC were used
to inoculate 250 mL flasks containing 100 mL of the cul-
ture medium. The flasks were incubated aerobically at
30 ꢁC on an orbital shaker and stirred at 250 rpm.
Flasks (250 mL) containing 100 mL of the culture med-
ium were then inoculated with 5 mL of the 24-h-old sus-
pension and incubated in the same conditions for 24 h.
Flasks (1L) containing 400 mL of the culture medium
were then inoculated with 5 mL of the latter suspension
and incubated for 24 h. Ethyl 3-oxoesters 1b–g (200 mg)
dissolved in 1mL of ethanol were added after all cul-
tures have the optical density reported in Table 4. Reac-
tion was followed by GC analysis and stopped at the
times indicated in Tables 1and 2 . The content of the
flask was then centrifuged and the supernatant extracted
with ethyl acetate. All the reactions were repeated at
least twice without any noticeable bias in the results.
4. Experimental
Reaction progress was monitored by TLC and GC
analysis. Thin-layer chromatography (TLC) was per-
formed on silica gel sheets with a fluorescent indicator
(Statocrom SIF, 60 F254 Merck); TLC spots were ob-
served under ultraviolet light or visualized with I2
vapour. Column chromatography was performed using
silica gel Merck 60 (0.063–0.200 lm). GC analyses were
performed by using a HP5 column (methyl silicone
gum; 30 m · 0.25 mm · 250 lm film thickness) on a
Hewlett Packard 5890 model, Series II. GC–MS ana-
lyses were performed on a Hewlett Packard 6890-
5793MSD.
Table 4. Culture medium optical density at which 1b–g have been
added
Yeast
Optical densitya
Saccharomyces cerevisiae CBS 7336
Saccharomyces cerevisiae DSM 11285
Trigonopsis variabilis DSM 70714
Cryptococcus curvatus ATCC 20509
Candida bombicola ATCC 22214
Kluyveromyces marxianus CBS 6556
4.9
4.5
4.4
8.2
8.8
7.2
Compounds 1b–g were prepared as reported.16 Spectro-
scopic data for 2b–g,18 3b–d, 3f and 4b–d and 4f19–23 are
consistent with those ones reported. H NMR spectra
1
a Measured at 620 nm.
were recorded in CDCl3 on
300 MHz spectrometer.
a Varian Mercury
Optical density was measured at 620 nm by a Perkin–
Elmer UV/Vis Lambda Bio 20 spectrophotometer.
4.2.2. Blank experiment.
A
1L flask containing
400 mL of the culture medium was stirred at 30 ꢁC on
an orbital shaker at 250 rpm. Ethyl 3-oxoesters
(200 mg) 1b–f dissolved in 1mL of ethanol were added.
The reaction was monitored by GC analysis and
stopped after 13 days. The content of the flask was ex-
tracted with ethyl acetate and analyzed.
4.1. Microorganism sources
The bakerÕs yeast used to reduce 3-oxoesters 1a–g was
supplied by Lievitalia. K. marxianus CBS 6556, S. cere-
visiae DSM 11285, S. cerevisiae CBS 7336, Cr. curvatus
ATCC 20509, C. bombicola ATCC 22214 and T. varia-
bilis DSM 70714 were obtained from public type culture
collections (ATCC, CBS, DSM).
4.2.3. Analytics. The reaction mixture (0.5 mL) was ex-
tracted by adding ethyl acetate (1mL). The sample was
centrifuged at 10,000g for 5 min in order to separate the

M. G. Perrone et al. / Tetrahedron: Asymmetry 16 (2005) 1473–1477
1477
10. Esbenshade, T. A.; Kamanna, V. S.; Newman, H. A. I.;
Tortorella, V.; Witiak, D. T.; Feller, D. R. Biochem.
Pharmacol. 1990, 40, 1263–1274.
two phases. The supernatant was dried by anhydrous
Na2SO4 and then analyzed by GC and/or TLC.
11. Inomata, N.; Yoshida, H.; Aoki, Y.; Tsunoda, M.;
Yamamoto, M. Tohoku J. Exp. Med. 1991, 165, 171–
182.
Acknowledgements
12. Ciolek, E.; Dauca, M. Biol. Cell. 1991, 71, 313–320.
13. Ferorelli, S.; Franchini, C.; Loiodice, F.; Perrone, M. G.;
Scilimati, A.; Sinicropi, M. S.; Tortorella, P. Tetrahedron:
Asymmetry 2001, 12, 853–862, and references cited
therein.
14. Perrone, M. G.; Santandrea, E.; Di Nunno, L.; Scilimati,
A.; Tortorella, V.; Capitelli, F.; Bertolasi, V. Tetrahedron:
Asymmetry, 2005, 16, 783–792.
The work was carried out under the framework of the
National Project ÔProgettazione, Sintesi e Valutazione
Biologica di Nuovi Farmaci CardiovascolariÕ supported
`
by the Ministero dellÕIstruzione, dellÕUniversita e della
Ricerca (MIUR, Rome). Thanks are also due to the Uni-
versity of Bari and CNR-Istituto di Chimica dei Comp-
osti OrganoMetallici-ICCOM/Sezione di Bari (Italy).
15. Perrone, M. G.; Santandrea, E.; Scilimati, A.; Tortorella,
V.; Capitelli, F.; Bertolasi, V. Tetrahedron: Asymmetry
2004, 15, 3501–3510.
References
16. Perrone, M. G.; Santandrea, E.; Scilimati, A.; Syldatk, C.;
Tortorella, V.; Capitelli, F.; Bertolasi, V. Tetrahedron:
Asymmetry 2004, 15, 3511–3517.
1. Sgro, C.; Escousse, A. Therapie 1991, 46, 351–354.
2. Gariot, P.; Barrat, E.; Mejean, L.; Pointel, J.; Drouin, P.;
Debry, G. Arch. Toxicol. 1983, 53, 151–163.
3. Angelin, B.; Einarsson, K.; Leijd, B. Eur. J. Clin. Invest.
1984, 14, 73–78.
4. Rangwala, S. M.; OÕBrien, M. L.; Tortorella, V.; Longo,
A.; Loiodice, F.; Noonan, D. J.; Feller, D. R. Chirality
1997, 9, 37–47.
5. Ferorelli, S.; Loiodice, F.; Tortorella, V.; Conte-Camer-
ino, D.; De Luca, A. Farmaco 2001, 56, 239–246.
6. Carbonara, G.; Fracchiolla, G.; Loiodice, F.; Tortorella,
P.; Conte-Camerino, D.; De Luca, A.; Liantonio, A.
Farmaco 2001, 56, 749–754.
7. Liantonio, A.; Accardi, A.; Carbonara, G.; Fracchiolla,
G.; Loiodice, F.; Tortorella, P.; Traverso, S.; Guida, P.;
Pierno, S.; De Luca, A.; Conte-Camerino, D.; Push, M.
Mol. Pharmacol. 2002, 62, 265–272.
8. Feller, D. R.; Kamanna, V. S.; Newman, H. A. I.;
Romstedt, K. J.; Witiak, D. T.; Bettoni, G.; Bryant, S. H.;
Conte-Camerino, D.; Loiodice, F.; Tortorella, V. J. Med.
Chem. 1987, 30, 1265–1267.
17. Hydrolysis of 3-oxoesters 1a–d and 1f by LiOH/THF–
H2O affords 3-oxoacids, which quickly and spontaneously
undergo decarboxylation (as many carboxylic acids bear-
ing at the a-position a strong electron withdrawing group
behave) to the corresponding ketones 3a–d and 3f.
Conversely, the hydrolysis of the 3-hydroxyesters 2a–d
and 2f into the corresponding 3-hydroxyacids by KOH/
THF–H2O was not followed by their decarboxylation, and
it was possible to isolate the stable 3-hydroxyacids.
18. Perrone, M. G.; Santandrea, E.; DellÕUomo, N.; Gian-
nessi, F.; Milazzo, F. M.; Sciarroni, A. F.; Scilimati, A.;
Tortorella, V. Eur. J. Med. Chem. 2005, 40(2), 143–154.
19. Eberle, M.; Stierli, D.; Mueller, U. PCT WO 2004000797,
Chem. Abstr. 140:59414.
20. Ogawa, K.; Terada, T.; Muranaka, Y.; Hamakawa, T.;
Hashimoto, S.; Fujii, S. Chem. Pharm. Bull. 1986, 34,
1118–1127.
21. Imai, T.; Takao, H.; Yamaguchi, K.; Uchida, T.; Goto,
T.; Umetsu, N. Pestic. Sci. 1992, 34, 119–125.
22. Luerssen, K.; Brandes, W.; Haenssler, G.; Holmwood, G.;
Reinecke, P. DE3325761, 1985.
9. Bettoni, G.; Loiodice, F.; Tortorella, V.; Conte-Camerino,
D.; Mambrini, M.; Ferrannini, E.; Bryant, S. H. J. Med.
Chem. 1987, 30, 1267–1270.
23. Boeckmann, K.; Regel, E.; Buechel, K. H.; Luerssen, K.;
Konze, J.; Brandes, W. CA1257263, 1989.