Comparative study of conformational effects on stereoselective lipase catalysed acetylation of sec hydroxy groups in diastereomeric 14-membered lactones and their acyclic analogs

Article abstract of DOI:10.1016/S0040-4039(00)01634-8

Stereoselective acetylation of sec hydroxy groups in a diastereomeric (1:1) mixture of macrocyclic lactones 14, 15, catalysed by seven microbial lipases in n-heptane afforded the C(7)-O-acetyl derivative 17 with up to 99% diastereomeric excess (d.e.), whereas acetylation of their acyclic analogs 10, 11 afforded C(5)-O-acetyl derivative 13 with up to 93% d.e. Six lipases, from Pseudomonas cepacia (PCL), Pseudomonas fluorescens (PFL), Geotrichum candidum (GCL), Candida cylindracea (CCL) and Pseudomonas cepacia immobilised on ceramics (PS-C) predominantly acetylate cyclic diastereomer (3S,7S)-15 to (3S,7S)-17; only Candida antarctica-B (CAL-B) lipase predominantly acetylates (3S,7R)-14 to (3S,7R)-16. With acyclic analogs 10, 11 the same set of lipases exhibited different diastereoselective bias; CAL-B, GCL and CCL predominantly acylate (1S,5S)-11 to (1S,5S)-13, whereas PCL, PS-C, PFL and PSL acylate predominantly (1S,5R)-10 to (1S,5R)-12. Only GCL exhibited higher stereoselectivity for an acyclic pair of stereoisomers with higher conformational flexibility, over cyclic diastereomeric substrates with a conformationally restricted macrocyclic ring. The preference of PCL for macrocyclic substrates is particularly interesting, in view of the recently suggested binding mode of a series of acyclic sec alcohols in the extended conformation. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)01634-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 9135–9138
Comparative study of conformational effects on
stereoselective lipase catalysed acetylation of sec hydroxy
groups in diastereomeric 14-membered lactones and their
acyclic analogs
&
Edina Ljubovic´ and Vitomir Sunjic´*
-
Ruder Bosˇkovic´ Institute, PO Box 180, 10002 Zagreb, Croatia
Received 5 June 2000; revised 15 September 2000; accepted 20 September 2000
Abstract
Stereoselective acetylation of sec hydroxy groups in a diastereomeric (1:1) mixture of macrocyclic
lactones 14, 15, catalysed by seven microbial lipases in n-heptane afforded the C(7)-O-acetyl derivative 17
with up to 99% diastereomeric excess (d.e.), whereas acetylation of their acyclic analogs 10, 11 afforded
C(5)-O-acetyl derivative 13 with up to 93% d.e. Six lipases, from Pseudomonas cepacia (PCL), Pseu-
domonas fluorescens (PFL), Geotrichum candidum (GCL), Candida cylindracea (CCL) and Pseudomonas
cepacia immobilised on ceramics (PS-C) predominantly acetylate cyclic diastereomer (3S,7S)-15 to
(3S,7S)-17; only Candida antarctica-B (CAL-B) lipase predominantly acetylates (3S,7R)-14 to (3S,7R)-16.
With acyclic analogs 10, 11 the same set of lipases exhibited different diastereoselective bias; CAL-B, GCL
and CCL predominantly acylate (1S,5S)-11 to (1S,5S)-13, whereas PCL, PS-C, PFL and PSL acylate
predominantly (1S,5R)-10 to (1S,5R)-12. Only GCL exhibited higher stereoselectivity for an acyclic pair
of stereoisomers with higher conformational flexibility, over cyclic diastereomeric substrates with a
conformationally restricted macrocyclic ring. The preference of PCL for macrocyclic substrates is
particularly interesting, in view of the recently suggested binding mode of a series of acyclic sec alcohols
in the extended conformation. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: lipase catalysed acetylation; acyclic versus macrocyclic sec alcohols; conformational effect on stereoselectiv-
ity.
Recently we have reported a study of the lipase catalysed stereoselective acetylation of the
diastereomeric mixture 14, 15 and hydrolysis of the corresponding acetates 16, 17.1,2 Prompted
by the unexpectedly high stereoselectivity in both directions, in view of the distance of the
perturbing groups from the stereogenic centre C(7), we have proposed that absolute conforma-
tion around a stereogenic centre determines stereoselectivity in these substrates, rather than
‘large’ and ‘medium’ groups, as defined by the known model of Kazlauskas.3–5
* Corresponding author. Fax: +38 51 46 80 195; e-mail: sunjic@rudjer.irb.hr
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)01634-8

9136
To test this finding further, and also prompted by the potential biological activity of the
resorcillic esters 10 and 11, open-chain analogs of the macrocyclic lactones 14 and 15, we have
undertaken their synthesis and biocatalytic separation. They are formally derived from the
macrocyclic counterparts 14 and 15 by cleaving the C(12)ꢀC(121) bond. Compound 14 (generic
name a-zearanol) has found application as an anabolic agent in meat-production;6 its 7b-isomer
15 is eliminated from the commercial product because of its estrogenic side effects.7 The
possibility that their open-chain analogs 10, 11 retain anabolic activity whilst avoiding the
undesired side effects was an additional motive for undertaking this study. Open-chain
diastereomers 10, 11 (1:1) were prepared with the (1S) configuration, which corresponds to
biosynthetically generated (3S) configuration in 14 and 15, Scheme 1.8
Scheme 1. (a) Al(OCH(CH3)2)3/2-PrOH/D; (b) PS-C/vinyl acetate/n-hexane/30°C; (c) KOH/MeOH/rt; (d) t-
Bu(CH3)2SiCl/imidazole/CH2Cl2/0°C; (e) CH3(CH2)4Br/Mg/I2/Et2O–toluene/D; (f) 1% HCl/EtOH/rt; (g) PPh3/
DEAD/Et2O/rt; (h) 5% HCl/THF (1:2)/rt; (i) NaBH4/MeOH; (j) PS-C/vinyl acetate/n-heptane/30°C

9137
Key steps in this synthesis comprise the efficient kinetic resolution of 2 by PS-C lipase that
affords R-3 with >94% e.e., the subsequent Grignard reaction of the TBDMS-protected R-4,
and acylation of R-6 with inversion of configuration (Mitsunobu reaction) by MEM-protected
resorcillic acid 7; the overall yield from 1 to a 1:1 mixture 10, 11 was 17%. A standard sample
of 12, 13 (1:1) was prepared by exhaustive enzymatic acetylation.
The results of a parallel study of the stereoselective acetylation of 10, 11 and 14, 15 by a series
of microbial lipases in n-heptane are presented in Table 1.
Table 1
Stereoselectivity of acetylation of a 1:1 mixture of acyclic alcohols 10, 11 and macrocyclic alcohols 14, 15 by
various microbial lipases
Comp. Lipase
t (h)
T (°C)
Conv. (%) D.e. (%) alcohol D.e. (%) acetate Da
10, 11
Pseudomonas cepacia
Pseudomonas cepacia
Pseudomonas fluorescens 72
Pseudomonas species
Pseudomonas cepacia-C
Candida antarctica B
Geotrichum candidum
Candida cylindracea
Pseudomonas cepacia
Pseudomonas cepacia
48
120
+50
+30
+30
+30
+30
+30
+30
+30
+50
+30
+30
+30
+30
+30
+30
+30
42.9
33.8
41.5
46.9
43.6
36.9
35.7
49.2
27.1
33.6
45.6
45.1
40.4
50.0
3.6
24.0 11
16.8 11
21.8 11
29.3 11
5.0 11
31.8 12
32.9 12
30.8 12
33.2 12
9.5 12
10.2 13
92.5 13
46.2 13
95.4 17
95.1 17
94.7 17
89.0 17
57.8 17
42.1 16
99.9 17
98.7 17
2.4
2.3
2.3
2.6
1.3
1.3
42.7
4.1
60
64
89
38
5.4
3.6
48
3
120
6
24
5.5
48
6.0 10
51.5 10
44.6 10
35.5 14
48.1 14
79.5 14
73.2 14
39.1 14
42.1 15
3.7 14
14, 15
Pseudomonas fluorescens 24
Pseudomonas species
Pseudomonas cepacia-C
Candida antarctica B
Geotrichum candidum
Candida cylindracea
24
6
48
48
48
\200
\200
36.4
56.6 14
a D-value defines diastereoselectivity according to the Sih’s equation, developed for enantioselectivity (E-value).9
Higher stereoselectivity was observed for most lipases with the cyclic substrates 14, 15. The
exceptional selectivity of GCL is particularly interesting; it shows a higher D-value and much
higher affinity for the acyclic pair 10, 11; in spite of the high D-value for the macrocyclic pair
14, 15, calculated at <4% conversion after 48 h, they can be regarded as nonsubstrates.
Most lipases accept the macrocyclic pair of diastereomers 14, 15 better, and on average
exhibit higher diastereoselectivity (d.e. 58–99%) in favour of the (3S,7S) diastereomer, whereas
the acyclic analogs are less well accepted and the stereoselectivity is generally lower (d.e.
46–93%), Table 1.10 The stereoselectivity of PCL catalysed acetylation of both pairs of
stereoisomers is similar after 48 h at +50°C and after 120 h at +30°C, revealing that a larger
temperature effect on the conformational mobility of the more flexible open-chain analogs is not
reflected in enhanced stereoselectivity at lower temperatures.

9138
The results reveal two interesting aspects of PCL and GCL stereoselectivity, however. First,
GCL although known to possess one large cavity that extends from the active site up to the
surface of the protein, substantially longer than in other lipases, and completely non-polar,12,13
accommodates acyclic substrates 10, 11 better than the macrocyclic ones, and acetylates them
with a D-value of ca. 43. This indicates that the acyclic pair 10, 11, presumably bound on the
GCL active site in an extended conformation, is accommodated better than their cyclic analogs.
Secondly, it has been argued recently that PCL exhibits higher stereoselectivity in acylation of
acyclic substrates that have substituents at the stereogenic centre, which differ significantly in
shape but not in volume,14 and it is stated that acyclic substrates do not fit the hydrophobic
binding regions of the active site of PCL in a folded conformation. Our results, however, reveal
the macrocyclic sec alcohols 14, 15 as better substrates, though they are equal to the folded
conformers of their acyclic counterparts 10, 11.
To explain the above results and to get a better insight into the origin of stereoselectivity we
are presently performing computer docking of both pairs of stereoisomers, 10, 11 and 14, 15, in
the PCL and GCL active site.
References
&
1. Gelo, M.; Sunjic´, V. Tetrahedron 1992, 48, 6511–6520.
&
2. Gelo, M.; Antolic´, S.; Kojic´-Prodic´, B.; Sunjic´, V. Tetrahedron 1994, 50, 13753–13764.
3. Kazlauskas, R. J.; Weissfloch, A. N. E.; Rappaport, A. T.; Cuccia, L. A. J. Org. Chem. 1991, 56, 2656–2665.
4. Zuegg, J.; Honig, H.; Schrag, J. D.; Cygler, M. J. Mol. Catal. B: Enzymatic 1997, 3, 83–98.
5. Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic Synthesis; Wiley-VCH: Weinheim, 1999; pp. 1–88.
6. Brown, R. G. In Anabolics in Animal Production, Symposium OIA Paris, February 15–17; Meissonier, E., Ed.;
1983; pp. 155–180.
7. Mirocha, C. J.; Pathre, S. V.; Behrens, J.; Schauerhammer, B. Appl. Environ. Microbiol. 1978, 35, 986.
1
8. All new compounds exhibited correct HRMS molecular ions, and their IR, H, and 13C NMR spectra conform
with the structures.
9. Chen, C.-S.; Fujimoto, Y.; Girdauskas, G.; Sih, C. J. Am. Chem. Soc. 1982, 104, 7294–7299.
10. We assign the (1S,5S) configuration to the more reactive diastereomer 11, based on the HPLC eluation order and
on the observed enantioselectivity bias of its conformational congener S-18, which is preferentially acylated
enantiomer of ( )-18 by GCL.11 In order to confirm this assignment, crystal structure determinations for 11 and
12 are in preparation.
&
11. Ljubovic´, E.; Sunjic´, V. Tetrahedron: Asymmetry 1997, 8, 1–4.
12. Herzber, O.; Moult, J. Proteins 1991, 11, 223–229.
13. Schrag, J. D.; Cygler, M. J. Mol. Biol. 1993, 230, 575–591.
14. Lemke, K.; Lemke, M.; Theil, F. J. Org. Chem. 1997, 62, 6268–6273.