- Email: [email protected]
Studies directed towards the synthesis of stevastelins—a macrolactonization approach to stevastelin B
TETRAHEDRON LETTERS Tetrahedron Letters 42 (2001) 5085–5088
Pergamon
Studies directed towards the synthesis of stevastelins—a macrolactonization approach to stevastelin B Tushar K. Chakraborty,* Subhash Ghosh and Shantanu Dutta Indian Institute of Chemical Technology, Hyderabad 500 007, India Received 20 April 2001; accepted 24 May 2001
Abstract—A synthetic approach towards the cyclic depsipeptide immunosuppressant stevastelin B, based on the stereoselective synthesis of its propionate-derived fatty acid segment 6 that was coupled with the tripeptide 7 leading to the advanced stage acyclic intermediate 5, is described. In spite of our best efforts, the crucial macrolactonization reaction was not successful. © 2001 Elsevier Science Ltd. All rights reserved.
Isolated from a culture of a Penicillium sp. NK374186, the stevastelins A (1), B (2), B3 (3) and C3 (4) represent a family of novel cyclic depsipeptide immunosuppresants that inhibit the dual specificity phosphatase, VHR.1–3 Three more stevastelin congeners, named A3, D3 and E3, were subsequently isolated from the same source.4 While the sulfated derivative stevastelin A (1) exhibits potent inhibitory activities against VHR in extracellular enzyme preparations, it has little effect in cellular preparations.5,6 The converse is true for stevMe
Me 3
Me(CH2)12 5
astelin B (2). The free threonyl hydroxyl group in 2 enables it to penetrate cells much better than its sulfated version 1. Once inside, it gets converted to an active form by intracellular sulfation or phosphorylation of its hydroxyl function. The pronounced biological activities of stevastelin molecules made us interested in undertaking their total synthesis. Herein, we report an attempt to synthesize stevastelin B following an approach that could also lead to some other members of the family. Me
O
O
OH HN O O
OR2
N H Me
Me(CH2)12 5
Me Me O
NH 2
OR
OR1
2: stevastelin B: R = H,
Me
R2
O
Me
OH HN O O O
1: stevastelin A: R1 = SO3H, R2 = Ac 1
Me 3
N H Me
Me O NH OR1
3: stevastelin B3: R1 = H, R2 = Ac 4: stevastelin C3: R1 = H, R2 = H
= Ac
Me
Me(CH2)12
O
Me
OH O HN PMB O HO2C N OBn H Me 5
Me O NH
Me Me(CH2)12
Me
Me
CO2H OAc OPMB
+ BocHN
Me H N
O N H Me
O BnO
OBn 6
* Corresponding author. Fax: +91-40-7173387, 7173757; e-mail: [email protected] 0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 0 - 4 0 3 9 ( 0 1 ) 0 0 8 9 3 - 0
7
OBn CO2Me
T. K. Chakraborty et al. / Tetrahedron Letters 42 (2001) 5085–5088
5086
overall yield—silyl protection, LiBH4 reduction, followed by Swern oxidation. Addition of the titanium enolate derived from the acyloxazolidinethione 98 to the aldehyde 8 gave the non-Evans syn aldol product 10 as the only isolable diastereomer in 70% yield.9 The stereochemistry of the product was assigned at a later stage. Reduction of 10 with sodium borohydride in ethanol gave the diol 11 in 80% yield. Next, a p-methoxybenzylidene protection of the 1,3-diol moiety was followed by selective opening of the benzylidine ring using DIBAL-H to give the intermediate 12 in 78% yield in two steps. Swern oxidation of 12 gave an aldehyde, in 90% yield, which on treatment with CH3(CH2)12MgBr in THF furnished the expected adduct 13 in 40% yield along with the undesired isomer 14.3 Compound 13 was then converted into a known fragment 15, obtained during the degradation studies on stevastelins,3 in four steps—PMB deprotection, acetonide protection of the resulting 1,3-diol, desilylation, followed by benzoyl protection of the primary hydroxyl group. The NMR data
Retrosynthetically, stevastelin B (2) can be obtained by macrolactonization of the acyclic hydroxy acid 5 which, in turn, could be prepared by coupling the two key intermediates 6 and 7, the former with the long chain fatty acid moiety and the latter possessing the tripeptide component. The salient feature of our synthesis is the stereoselective construction of the C1–C18 fragment 6 with four consecutive stereocenters following two key steps: (i) a Ti(IV)-mediated diastereoselective nonEvans syn aldol reaction7 using a 2-oxazolidinethione based chiral auxiliary,8 which was expected to fix the C3 and C4 stereocenters; and (ii) the nucleophilic addition of a Grignard reagent derived from a long-chain halide to the C5-aldehyde to instill the requisite stereochemistry at C5. Scheme 1 outlines in detail the stereoselective synthesis of 6. The starting aldehyde 8 was prepared from methyl (S)-3-hydroxy-2-methylpropionate in three steps in 75% Bn Me OTBDPS +
OHC
TiCl4, DIPEA, Me CH2Cl2, -78 oC 70%
Me
HO
O
80%
OH
1. (COCl)2, DMSO,
CH(OMe)2 Me
CSA (cat.), CH2Cl2
OTBDPS
OTBDPS 10
1.MeO Me
S
0 oC
Me
N
O
S O 9
8
Me
N
O
NaBH4, EtOH,
Bn
Et3N, CH2Cl2, -78 oC
Me
HO
OH
2. DIBAL-H, CH2Cl2, -78 oC
11
to 0 oC
OTBDPS 90%
OPMB
2. CH3(CH2)12MgBr, 12
THF, 0 oC to rt
78% Me
Me
Me
Me
OTBDPS + CH3(CH2)12
CH3(CH2)12
OTBDPS OH OPMB Me Me 14 CH3(CH2)12 4 steps 13 O O
OH OPMB 13 (40%) 1. Ac2O, Et3N DMAP (cat.), CH2Cl2 2. TBAF, THF 85%
15 o
Me
Me
CH3(CH2)12
6
OH 70% OAc OPMB 16
TFA.H2N CH2Cl2, 0 oC
Me
Me(CH2)12
N OBn H Me 18
Me H N
O N H Me
O BnO 17
O
65%
Me
OAc O HN PMB O MeO2C
Scheme 1.
Me TFA,
7 Me
EDCI, HOBt, CH2Cl2, 0 C, then
Jones Oxidation
Me O NH OBn
5
OBn CO2Me + DIPEA
OBz
T. K. Chakraborty et al. / Tetrahedron Letters 42 (2001) 5085–5088
5087
1. TBAF, THF 2. TBSOTf, 2,6-lutidine, CH2Cl2, 0
oC
Me
CH3(CH2)12
13
OH OPMB
TBSO
3. CSA (cat.), MeOH-CH2Cl2 0
1. SO3-Py, DIPEA,
Me
Me CH3(CH2)12
Me
CO2H OPMB
TBSO
2. NaClO2, NaH2PO4,
19
oC
DMSO, CH2Cl2, 0 oC
20
2-methyl-2-butene, tBuOH, rt 95%
68% Me
Me
Me(CH2)12
Me O
Coupling with 17 TBSO O HN (same method as described PMB in Scheme 1: 6 to 18) MeO2C O N OBn H Me 21
70% LiOH, THF-MeOH-H2O 0 oC
Me
NH OBn
Me
Me(CH2)12
O
CSA (cat.), MeOH, Me 0 oC O
OH O HN PMB MeO2C O
75%
N OBn H Me 22
Me Me O
NH OBn
Macrolactonisation 5 96%
Scheme 2.
of 15 were identical with those reported for the degraded product.3 The remaining part of the synthesis was then carried out by acetate protection of the secondary alcohol in 13 followed by silyl deprotection to give the primary alcohol 16 in 85% yield from 13. Finally, Jones oxidation of 16 furnished the desired fatty acid segment 6 in 70% yield. The peptide segment 7 was synthesized from commercially available protected amino acids by standard solution phase peptide synthesis conditions using 1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide hydrochloride (EDCI) and 1-hydroxybenzotriazole (HOBt) as coupling agents and dry CH2Cl2 as solvent. Both the halves of the molecule were now ready to be coupled. The tripeptide 7 was accordingly treated with 50% TFA in CH2Cl2 at 0°C to effect Boc deprotection, leading to the free-NH2 containing tripeptide 17, which was coupled with the acid 6 following the peptide coupling method described above to obtain the coupled product 18 in 65% yield. At this stage, we thought that both the ester and the acetate groups would be hydrolyzed upon treatment with base. Unfortunately, the proton NMR spectrum of the hydrolyzed product revealed that only the ester function had been hydrolyzed, while the acetate remained intact. Attempts to deprotect the acetate first, before the hydrolysis of the ester function, using various methods did not succeed. To overcome this setback, it was decided to use silyl protection for the C5–OH, instead of an acetate group. This alternative approach is shown in Scheme 2. Desilylation of 13 and then di-TBS protection was followed by selective primary TBS-deprotection to give the compound 19 in 68% yield. A two-step oxidation of 19 gave
the C5-silyloxy acid 20 in 95% yield. The acid 20 and the peptide segment 17 were then coupled to give compound 21 in 70% yield. TBS-deprotection of 21 gave the intermediate 22 that was subjected to ester hydrolysis to furnish the desired hydroxy acid 5. The stage was now set to carry out the crucial macrolactonization reaction, but, unfortunately, even after trying several methods, for example the Yamaguchi method,10 dipyridyl disulfide,11 DCC-DMAP (cat.),12 etc. we were unable to prepare the desired cyclised product from 5. Whether the above failure should be attributed to an inherent structural feature of the cyclization intermediate 5, which possibly could not attain the required conformation for ester bond formation, or to steric hindrance due to the presence of long chain, is not yet clear. Efforts are now going on to form the ester bond first and then carry out a macrolactamization reaction in order to complete the targeted total synthesis of stevastelin B. These studies will be reported in due course. Acknowledgements The authors wish to thank Drs. A. C. Kunwar and M. Vairamani for NMR and mass spectroscopic assistance, respectively, and the CSIR, New Delhi, for research fellowships to S.G. and S.D. References 1. Morino, T.; Masuda, A.; Yamada, M.; Nishimoto, M.; Nishikiori, T.; Saito, S.; Shimada, N. J. Antibiot. 1994, 47, 1341–1343.
5088
2.
3. 4. 5. 6. 7.
T. K. Chakraborty et al. / Tetrahedron Letters 42 (2001) 5085–5088
Morino, T.; Shimada, K.; Masuda, A.; Yamashita, N.; Nishimoto, M.; Nishikiori, T.; Saito, S. J. Antibiot. 1996, 49, 564–568. Shimada, K.; Morino, T.; Masuda, A.; Sato, M.; Kitagawa, M.; Saito, S. J. Antibiot. 1996, 49, 569–574. Morino, T.; Shimada, K.; Masuda, A.; Nishimoto, M.; Saito, S. J. Antibiot. 1996, 49, 1049–1051. Hamaguchi, T.; Masuda, A.; Morino, T.; Osada, H. Chem. Biol. 1997, 4, 279–286. Burke, Jr., T. R.; Zhang, Z.-Y. Biopolymers 1998, 47, 225–241. Crimmins, M. T.; King, B. W.; Tabet, E. A. J. Am. Chem. Soc. 1997, 119, 7883–7884.
.
8. Delaunay, D.; Toupet, L.; Le Corre, M. J. Org. Chem. 1995, 60, 6604–6607. 9. All new compounds were characterized by IR, NMR and mass spectroscopic studies. Yields refer to chromatographically and spectroscopically homogeneous materials. 10. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull. Chem. Soc. Jpn. 1979, 52, 1989– 1993. 11. Corey, E. J.; Nicolaou, K. C. J. Am. Chem. Soc. 1974, 96, 5614–5616. 12. Keck, G. E.; Boden, E. P.; Wiley, M. R. J. Org. Chem. 1989, 54, 896–906.
Pergamon
Studies directed towards the synthesis of stevastelins—a macrolactonization approach to stevastelin B Tushar K. Chakraborty,* Subhash Ghosh and Shantanu Dutta Indian Institute of Chemical Technology, Hyderabad 500 007, India Received 20 April 2001; accepted 24 May 2001
Abstract—A synthetic approach towards the cyclic depsipeptide immunosuppressant stevastelin B, based on the stereoselective synthesis of its propionate-derived fatty acid segment 6 that was coupled with the tripeptide 7 leading to the advanced stage acyclic intermediate 5, is described. In spite of our best efforts, the crucial macrolactonization reaction was not successful. © 2001 Elsevier Science Ltd. All rights reserved.
Isolated from a culture of a Penicillium sp. NK374186, the stevastelins A (1), B (2), B3 (3) and C3 (4) represent a family of novel cyclic depsipeptide immunosuppresants that inhibit the dual specificity phosphatase, VHR.1–3 Three more stevastelin congeners, named A3, D3 and E3, were subsequently isolated from the same source.4 While the sulfated derivative stevastelin A (1) exhibits potent inhibitory activities against VHR in extracellular enzyme preparations, it has little effect in cellular preparations.5,6 The converse is true for stevMe
Me 3
Me(CH2)12 5
astelin B (2). The free threonyl hydroxyl group in 2 enables it to penetrate cells much better than its sulfated version 1. Once inside, it gets converted to an active form by intracellular sulfation or phosphorylation of its hydroxyl function. The pronounced biological activities of stevastelin molecules made us interested in undertaking their total synthesis. Herein, we report an attempt to synthesize stevastelin B following an approach that could also lead to some other members of the family. Me
O
O
OH HN O O
OR2
N H Me
Me(CH2)12 5
Me Me O
NH 2
OR
OR1
2: stevastelin B: R = H,
Me
R2
O
Me
OH HN O O O
1: stevastelin A: R1 = SO3H, R2 = Ac 1
Me 3
N H Me
Me O NH OR1
3: stevastelin B3: R1 = H, R2 = Ac 4: stevastelin C3: R1 = H, R2 = H
= Ac
Me
Me(CH2)12
O
Me
OH O HN PMB O HO2C N OBn H Me 5
Me O NH
Me Me(CH2)12
Me
Me
CO2H OAc OPMB
+ BocHN
Me H N
O N H Me
O BnO
OBn 6
* Corresponding author. Fax: +91-40-7173387, 7173757; e-mail: [email protected] 0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 0 - 4 0 3 9 ( 0 1 ) 0 0 8 9 3 - 0
7
OBn CO2Me
T. K. Chakraborty et al. / Tetrahedron Letters 42 (2001) 5085–5088
5086
overall yield—silyl protection, LiBH4 reduction, followed by Swern oxidation. Addition of the titanium enolate derived from the acyloxazolidinethione 98 to the aldehyde 8 gave the non-Evans syn aldol product 10 as the only isolable diastereomer in 70% yield.9 The stereochemistry of the product was assigned at a later stage. Reduction of 10 with sodium borohydride in ethanol gave the diol 11 in 80% yield. Next, a p-methoxybenzylidene protection of the 1,3-diol moiety was followed by selective opening of the benzylidine ring using DIBAL-H to give the intermediate 12 in 78% yield in two steps. Swern oxidation of 12 gave an aldehyde, in 90% yield, which on treatment with CH3(CH2)12MgBr in THF furnished the expected adduct 13 in 40% yield along with the undesired isomer 14.3 Compound 13 was then converted into a known fragment 15, obtained during the degradation studies on stevastelins,3 in four steps—PMB deprotection, acetonide protection of the resulting 1,3-diol, desilylation, followed by benzoyl protection of the primary hydroxyl group. The NMR data
Retrosynthetically, stevastelin B (2) can be obtained by macrolactonization of the acyclic hydroxy acid 5 which, in turn, could be prepared by coupling the two key intermediates 6 and 7, the former with the long chain fatty acid moiety and the latter possessing the tripeptide component. The salient feature of our synthesis is the stereoselective construction of the C1–C18 fragment 6 with four consecutive stereocenters following two key steps: (i) a Ti(IV)-mediated diastereoselective nonEvans syn aldol reaction7 using a 2-oxazolidinethione based chiral auxiliary,8 which was expected to fix the C3 and C4 stereocenters; and (ii) the nucleophilic addition of a Grignard reagent derived from a long-chain halide to the C5-aldehyde to instill the requisite stereochemistry at C5. Scheme 1 outlines in detail the stereoselective synthesis of 6. The starting aldehyde 8 was prepared from methyl (S)-3-hydroxy-2-methylpropionate in three steps in 75% Bn Me OTBDPS +
OHC
TiCl4, DIPEA, Me CH2Cl2, -78 oC 70%
Me
HO
O
80%
OH
1. (COCl)2, DMSO,
CH(OMe)2 Me
CSA (cat.), CH2Cl2
OTBDPS
OTBDPS 10
1.MeO Me
S
0 oC
Me
N
O
S O 9
8
Me
N
O
NaBH4, EtOH,
Bn
Et3N, CH2Cl2, -78 oC
Me
HO
OH
2. DIBAL-H, CH2Cl2, -78 oC
11
to 0 oC
OTBDPS 90%
OPMB
2. CH3(CH2)12MgBr, 12
THF, 0 oC to rt
78% Me
Me
Me
Me
OTBDPS + CH3(CH2)12
CH3(CH2)12
OTBDPS OH OPMB Me Me 14 CH3(CH2)12 4 steps 13 O O
OH OPMB 13 (40%) 1. Ac2O, Et3N DMAP (cat.), CH2Cl2 2. TBAF, THF 85%
15 o
Me
Me
CH3(CH2)12
6
OH 70% OAc OPMB 16
TFA.H2N CH2Cl2, 0 oC
Me
Me(CH2)12
N OBn H Me 18
Me H N
O N H Me
O BnO 17
O
65%
Me
OAc O HN PMB O MeO2C
Scheme 1.
Me TFA,
7 Me
EDCI, HOBt, CH2Cl2, 0 C, then
Jones Oxidation
Me O NH OBn
5
OBn CO2Me + DIPEA
OBz
T. K. Chakraborty et al. / Tetrahedron Letters 42 (2001) 5085–5088
5087
1. TBAF, THF 2. TBSOTf, 2,6-lutidine, CH2Cl2, 0
oC
Me
CH3(CH2)12
13
OH OPMB
TBSO
3. CSA (cat.), MeOH-CH2Cl2 0
1. SO3-Py, DIPEA,
Me
Me CH3(CH2)12
Me
CO2H OPMB
TBSO
2. NaClO2, NaH2PO4,
19
oC
DMSO, CH2Cl2, 0 oC
20
2-methyl-2-butene, tBuOH, rt 95%
68% Me
Me
Me(CH2)12
Me O
Coupling with 17 TBSO O HN (same method as described PMB in Scheme 1: 6 to 18) MeO2C O N OBn H Me 21
70% LiOH, THF-MeOH-H2O 0 oC
Me
NH OBn
Me
Me(CH2)12
O
CSA (cat.), MeOH, Me 0 oC O
OH O HN PMB MeO2C O
75%
N OBn H Me 22
Me Me O
NH OBn
Macrolactonisation 5 96%
Scheme 2.
of 15 were identical with those reported for the degraded product.3 The remaining part of the synthesis was then carried out by acetate protection of the secondary alcohol in 13 followed by silyl deprotection to give the primary alcohol 16 in 85% yield from 13. Finally, Jones oxidation of 16 furnished the desired fatty acid segment 6 in 70% yield. The peptide segment 7 was synthesized from commercially available protected amino acids by standard solution phase peptide synthesis conditions using 1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide hydrochloride (EDCI) and 1-hydroxybenzotriazole (HOBt) as coupling agents and dry CH2Cl2 as solvent. Both the halves of the molecule were now ready to be coupled. The tripeptide 7 was accordingly treated with 50% TFA in CH2Cl2 at 0°C to effect Boc deprotection, leading to the free-NH2 containing tripeptide 17, which was coupled with the acid 6 following the peptide coupling method described above to obtain the coupled product 18 in 65% yield. At this stage, we thought that both the ester and the acetate groups would be hydrolyzed upon treatment with base. Unfortunately, the proton NMR spectrum of the hydrolyzed product revealed that only the ester function had been hydrolyzed, while the acetate remained intact. Attempts to deprotect the acetate first, before the hydrolysis of the ester function, using various methods did not succeed. To overcome this setback, it was decided to use silyl protection for the C5–OH, instead of an acetate group. This alternative approach is shown in Scheme 2. Desilylation of 13 and then di-TBS protection was followed by selective primary TBS-deprotection to give the compound 19 in 68% yield. A two-step oxidation of 19 gave
the C5-silyloxy acid 20 in 95% yield. The acid 20 and the peptide segment 17 were then coupled to give compound 21 in 70% yield. TBS-deprotection of 21 gave the intermediate 22 that was subjected to ester hydrolysis to furnish the desired hydroxy acid 5. The stage was now set to carry out the crucial macrolactonization reaction, but, unfortunately, even after trying several methods, for example the Yamaguchi method,10 dipyridyl disulfide,11 DCC-DMAP (cat.),12 etc. we were unable to prepare the desired cyclised product from 5. Whether the above failure should be attributed to an inherent structural feature of the cyclization intermediate 5, which possibly could not attain the required conformation for ester bond formation, or to steric hindrance due to the presence of long chain, is not yet clear. Efforts are now going on to form the ester bond first and then carry out a macrolactamization reaction in order to complete the targeted total synthesis of stevastelin B. These studies will be reported in due course. Acknowledgements The authors wish to thank Drs. A. C. Kunwar and M. Vairamani for NMR and mass spectroscopic assistance, respectively, and the CSIR, New Delhi, for research fellowships to S.G. and S.D. References 1. Morino, T.; Masuda, A.; Yamada, M.; Nishimoto, M.; Nishikiori, T.; Saito, S.; Shimada, N. J. Antibiot. 1994, 47, 1341–1343.
5088
2.
3. 4. 5. 6. 7.
T. K. Chakraborty et al. / Tetrahedron Letters 42 (2001) 5085–5088
Morino, T.; Shimada, K.; Masuda, A.; Yamashita, N.; Nishimoto, M.; Nishikiori, T.; Saito, S. J. Antibiot. 1996, 49, 564–568. Shimada, K.; Morino, T.; Masuda, A.; Sato, M.; Kitagawa, M.; Saito, S. J. Antibiot. 1996, 49, 569–574. Morino, T.; Shimada, K.; Masuda, A.; Nishimoto, M.; Saito, S. J. Antibiot. 1996, 49, 1049–1051. Hamaguchi, T.; Masuda, A.; Morino, T.; Osada, H. Chem. Biol. 1997, 4, 279–286. Burke, Jr., T. R.; Zhang, Z.-Y. Biopolymers 1998, 47, 225–241. Crimmins, M. T.; King, B. W.; Tabet, E. A. J. Am. Chem. Soc. 1997, 119, 7883–7884.
.
8. Delaunay, D.; Toupet, L.; Le Corre, M. J. Org. Chem. 1995, 60, 6604–6607. 9. All new compounds were characterized by IR, NMR and mass spectroscopic studies. Yields refer to chromatographically and spectroscopically homogeneous materials. 10. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull. Chem. Soc. Jpn. 1979, 52, 1989– 1993. 11. Corey, E. J.; Nicolaou, K. C. J. Am. Chem. Soc. 1974, 96, 5614–5616. 12. Keck, G. E.; Boden, E. P.; Wiley, M. R. J. Org. Chem. 1989, 54, 896–906.