Organic compounds -- part of the class 532-570 series – Organic compounds – Sulfur containing
Reexamination Certificate
2000-01-27
2002-12-10
Weber, Jon P. (Department: 1651)
Organic compounds -- part of the class 532-570 series
Organic compounds
Sulfur containing
C424S070510
Reexamination Certificate
active
06492562
ABSTRACT:
TECHNICAL FIELD
The invention concerns methods for the efficient synthesis of oligoketide thioesters, including diketide and triketides, which are useful as intermediates in polyketide production and to methods to use these intermediates. The methods of synthesis are suitable for liquid phase as well as solid-phase combinatorial synthesis. The invention also includes polyketide and tailored polyketide products.
BACKGROUND OF THE INVENTION
The creation of novel macrolide polyketides has been achieved through genetic manipulation of polyketide synthases. The modular nature of the Type 1 polyketide synthases allows for domain exchange between different polyketide synthase genes, resulting in hybrid genes which produce polyketide synthases with altered properties that result, in turn, in modified macrolide structures. Thus, it is possible to control chain length, choice of chain extender unit, degree of &bgr;-carbon oxidation level, and to some degree stereochemistry. The choice of starter unit has been more difficult to control. Two complementary approaches have been described.
Dutton, et al.,
J. Antibiotics
(1991) 44:357-365 demonstrated that the avermectin polyketide synthase was somewhat flexible in choice of starter units. When denied the natural starter unit through inactivation of the branched-chain amino acid dehydrogenase, the avermectin polyketide synthase will accept a variety of &agr;-branched carboxylic acids as the starter unit. However, only about 30 acids out of nearly 800 candidate acids tried were accepted. Acids without an &agr;-branch appear to be metabolized through &bgr;-oxidation until an &agr;-branch is reached, further limiting this methodology. Marsden, et al.,
Science
(1998) 79:199-202 exchanged the native loading domain of the erythromycin PKS with that from the avermectin polyketide synthase, resulting in a hybrid PKS having the same loosened starter unit specificity as the avermectin PKS. Clearly, the native specificities of enzymatic domains will always be a limitation on the flexibility of resulting hybrid systems.
A more general method for controlling starter unit specificity has been described by Jacobsen, et al.,
Science
(1997) 277:367-369. Inactivation of the ketosynthase in module 1 (KS1) of the erythromycin PKS (DEBS) results in an enzyme (KS1°-DEBS) incapable of initiating polyketide synthesis using precursors normally available to the cell. When supplied with a suitable thioester of the diketide product of module 1 or its analogs, however, KS1°-DEBS efficiently incorporates these into full-length polyketides. Subsequent experiments have demonstrated that a very wide range of diketide analogs are accepted by KS1°-DEBS, making this a very general method for production of analogs of the polyketide precursor of erythromycin, 6-deoxyerythronolide B (6-dEB), with variations at the positions controlled by the starter unit. Further, this method allows for production of 6-dEB analogs altered at the 12-position; this is equivalent to altering the substrate specificity of the module 1 acyltransferase (AT1) which transfers the first extender unit. While this has been accomplished through the above described domain exchange experiments as well, the “diketide method” allows for introduction of 12-position substituents which are not available from nature. Furthermore, triketide analogs are accepted, opening the 10- and 11-positions of 6-dEB for modification. The 6-dEB analogs obtained can be further converted into analogs of erythromycin by feeding to a suitable converter strain, such as a strain of
Saccharopolyspora erythraea
containing a non-functional erythromycin PKS. The resulting erythromycins have altered side-chains at the 13-position as well as other optional modifications, and show altered biological activity. These erythromycin analogs can also be produced by introducing the KS1°-mutation into an erythromycin-producing strain of
Saccharopolyspora erythraea
, then supplying the mutant strain with suitable diketide or triketide thioesters as described above.
Implementation of this method requires the availability of the N-acylcysteamine oligoketide thioesters. Synthetic methods available in the art for these thioesters do not lend themselves to efficient, economical synthesis, or to the systematic production of variants. The diketide and triketides also typically contain chiral centers requiring the control of absolute or relative stereochemistry.
Cane, D. E., et al.,
J Am Chem Soc
(1987) 109:1255-1257 describes a three-step process to produce the N-acetylcysteamine thioester of (2S,3R)-2-methyl-3-hydroxy pentanoic acid. The method relies on the use of a chiral reagent N-propionyl-(4S)-4-isopropyl-2-oxazolidinone for control of the absolute stereochemistry of the product:
This material results from the acylation of (4S)-4-isopropyl-2-oxazolidinone with propionyl chloride, typically using a strong base such as n-butyllithium at low temperature. The Cane process is an aldol condensation of this starting material with propionaldehyde in the presence of dibutylboron triflate (Bu
2
BOTf), followed by hydrolysis of the imide (lithium hydroperoxide) and thioesterification with N-acetylcysteamine in the presence of diphenyl phosphorylazide and triethylamine. This multi-step process is inefficient, with substantial losses accompanying the hydrolysis step.
Cane, D. E., et al.,
J Antibiotics
(1995) 48:647-651 was able to improve yields using a five-step process which replaces the aldol condensation with a Claisen condensation between the lithium enolate of the propionyl oxazolidinone (N-propionyl-(4S)-4-benzyl-2-oxazolidinone was used as the stereochemistry controlling starting material in this method) and propionyl chloride followed by reduction of the resulting &bgr;-ketoester product using zinc borohydride. Protection of the &bgr;-hydroxy substituent as a tert-butyldimethylsilyl ether preceded hydrolysis of the imide, which again required lithium hydroperoxide. The protecting group gave improved yields from hydrolysis, but required an additional two steps to add and remove. This longer process also suffers from the use of zinc borohydride, which is not commercially available.
Cleavage of the N-acyloxazolidinones resulting from either aldol or Claisen condensations as described above is problematic. Various methods of cleavage are known in the art, including that of Evans, D. A., et al.,
Tetrahedron Lett
(1987) 28:6141, in which undesired reaction at the oxazolidinone carbonyl during hydrolysis is suppressed by the use of lithium hydroperoxide. This process requires the use of concentrated solutions of hydrogen peroxide, which are explosive and dangerous for large-scale processes. The N-acyloxazolidinones are unreactive towards thiols or thiolates, although some conversion to thioesters can be observed using concentrated solutions of lithium thiolates in tetrahydrofuran. The low solubility of the thiolates in tetrahydrofuran combined with epimerization of chiral diketides due to the basicity of the thiolates limits the utility ofthis method. Miyata, O., et al.,
Syn Lett
(1994) 637-638, describes conversion of N-acyloxazolidinones to S-benzylthioesters through the use of lithium benzylthiotrimethylaluminate. The production of more complex thioesters containing groups capable of binding Lewis acids like trimethylaluminum, such as those based on N-acylcysteamine, has not been reported.
The N-acetylcysteamine thioesters of larger oligoketides have also been prepared. Cane, D. E., et al.,
J Am Chem Soc
(1993) 115:522-526 synthesized the N-acetylcysteamine thioester of (4S,5R)-5-hydroxy-4-methyl-2-heptenbic acid using the stereochemically controlled aldol condensation product of N-propionyl-(4S)-4-benzyl-2-oxazolidinone as the starting material:
This imide was converted to the corresponding aldehyde, and extended at the carbonyl group by a Wittig reaction to obtain the desired triketide as the ethyl ester which was then hydrolyzed and converted to the acylcysteamine thioester in a two-step process. Yields were improved by addition of steps
Ashley Gary
Burlingame Mark A.
Chan-Kai Isaac C.
Ashley Gary
Kaster Kevin
Kosan Biosciences, Inc.
Murashige Kate H.
Weber Jon P.
LandOfFree
Racemic thioesters for production of polyketides does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Racemic thioesters for production of polyketides, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Racemic thioesters for production of polyketides will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2938448