Stereocontrolled formation of ketomethylene isosteres through tandem chain extension reactions

Article abstract of DOI:10.1021/jo061184o

A zinc-mediated chain extension reaction is the key step in the preparation of γ-keto amides derived from amino acids. The use of tandem reaction sequences, in which the intermediate zinc enolate is trapped with electrophilic reagents, results in the inco

Full text of DOI:10.1021/jo061184o

Stereocontrolled Formation of Ketomethylene Isosteres through
Tandem Chain Extension Reactions
Weimin Lin,† Nancy Tryder,† Fan Su,† Charles K. Zercher,*,† Jerry P. Jasinski,‡ and
Ray J. Butcher§
Department of Chemistry, UniVersity of New Hampshire, Durham, New Hampshire 03824, Department of
Chemistry, Keene State College, 229 Main Street, Keene, New Hampshire 03435, and Department of
Chemistry, Howard UniVersity, 2400 Sixth Street, Washington, D.C. 20059
ckz@cisunix.unh.edu
ReceiVed June 8, 2006
A zinc-mediated chain extension reaction is the key step in the preparation of γ-keto amides derived
from amino acids. The use of tandem reaction sequences, in which the intermediate zinc enolate is trapped
with electrophilic reagents, results in the incorporation of R-substituents, which mimic the side chains
found in natural amino acid systems. Use of the chiral amino acid L-proline as a stereo-directing element
provides a diastereoselective route to various ketomethylene isosteres.
Introduction
protease inhibitor8 further attests to the suitability of this strategy.
The synthetic methodology developed in this study for ketom-
ethylene isostere formation has potential ramifications in many
therapeutic arenas.
Isosteric replacement strategies are central to the development
of many drug candidates. Specifically, the replacement of
peptide bonds with non-hydrolyzable peptide isosteres has been
used to provide substances resistant to the metabolic proteases.
The design of a non-hydrolyzable peptide isostere that maintains
active-site recognition features constitutes a recurring theme in
protease inhibition. A variety of functional groups have been
used to mimic the peptide bond, including hydroxyethylene,1
(E)-alkenes,2 silanols,3 and dihydroxyethylene.4
Generation of peptide mimics frequently requires lengthy and
resource-intensive synthetic sequences. The desire for stereo-
controlled incorporation of the R side chain complicates the
synthetic issues even further. A host of methods5-7,9 for the
preparation of ketomethylene isosteric replacements have been
The ketomethylene (1) unit has been utilized for the inhibition
of a wide variety of enzymatic systems, including aminopep-
tidase,5 angiotensin converting enzyme (ACE),6 and others.7 The
appearance of ketomethylene isosteres in a naturally occurring
† University of New Hampshire.
‡ Keene State College.
(5) (a) Harbeson, S. L.; Rich, D. H. J. Med. Chem. 1989, 32, 1378-
1392. (b) Mansour, T. Synth. Commun. 1989, 19, 659-665.
§ Howard University.
(1) (a) Benedetti, F.; Berti, F.; Norbedo, S. J. Org. Chem. 2002, 67, 8635.
(b) Kath, J. C.; DiRico, A. P.; Gladue, R. P.; Martin, W. H.; McElroy, E.
B.; Stock, I. A.; Tylaska, L. A.; Zheng, D. Bioorg. Med. Chem. Lett. 2004,
9, 2163. (c) Tao, J.; Hoffman, R. V. J. Org. Chem. 1997, 62, 6240.
(2) (a) Wipf, P.; Henninger, T. C.; Geib, S. J. J. Org. Chem. 1998, 63,
6088. (b) Oishi, S.; Kamano, T.; Niida, A.; Odagaki, Y.; Hamanaka, N.;
Yamamoto, M.; Ajito, K.; Tamamura, H.; Otaka, A.; Fujii, N. J. Org. Chem.
2002, 67, 6162.
(3) (a) Kim, J.; Sieburth, S. M. J. Org. Chem. 2004, 69, 3008. (b) Kim,
J.; Glekas, A.; Sieburth, S. M. Bioorg. Med. Chem. Lett. 2002, 12, 3625.
(4) (a) Benedetti, F.; Magnan, M.; Miertus, S.; Norbedo, S.; Parat, D.;
Tossi, A. Bioorg. Med. Chem. Lett. 1999, 9, 3027. (b) Atsuumi, S.; Ikeura,
C.; Honma, T.; Morishima, H. Chem. Pharm. Bull. 1994, 42, 2164.
(6) Almquist, R. G.; Chao, W.-R.; Judd, A. K.; Mitoma, C.; Rossi, D.
J.; Panesovich, R. E.; Matthews, R. J. J. Med. Chem. 1988, 31, 561.
(7) (a) Angelastro, M. R.; Marquart, A. L.; Mehdi, S.; Koehl, J. R.; Vaz,
R. J.; Bey, P.; Peet, N. P. Bioorg. Med. Chem. Lett. 1999, 9, 139. (b) Ettouati,
L.; Casimir, J. R.; Raynaud, I.; Trescol-Biemont, M.-C.; Carrupt, P.-A.;
Gerlier, D.; Rabourdin-Combe, C.; Testa, B.; Paris, J. Protein Pept. Lett.
1998, 5, 221. (c) Marshall, G. R.; Toth, M. V. U.S. Patent 320,742, 1992.
(d) Jones, D. M.; Atrash, B.; Teger-Nilsson, A.-C.; Gyzander, E.; Deinum,
J.; Szelke, M. Lett. Pept. Sci. 1995, 2, 147. (e) Jones, D. M.; Atrash, B.;
Ryder, H.; Geger-Nilsson, A.-C.; Gyzander, E.; Szelke, M. J. Enzyme Inhib.
1995, 9, 43.
(8) Umezawa, H.; Aoyagi, T.; Ohuchi, S.; Okuyama, A.; Suda, H.; Takita,
T.; Hamada, M.; Takeuchi, T. J. Antibiot. 1983, 36, 1572.
10.1021/jo061184o CCC: $33.50 © 2006 American Chemical Society
Published on Web 09/23/2006
8140
J. Org. Chem. 2006, 71, 8140-8145

Stereocontrolled Formation of Ketomethylene Isosteres
SCHEME 1
SCHEME 2
TABLE 1. Preparation of γ-Keto Amides
substrate
P
R1
R2
product
yield (%)
2a
2b
2c
2d
2e
2f
Cbz
H
H
H
H
H
Bn
i-Bu
Me
H
H
H
Me
3a
3b
3c
3d
3e
3f
43
74
65
74
83
36
54
reported. Common approaches (Scheme 1) to the ketomethylene
group include nucleophilic attack on easily epimerized amino
acid-derived aldehydes (Path A). Poor diastereoselectivity is
often observed in the formation of the γ-hydroxy acid, and
reoxidation can be troublesome. Furthermore, any R-substituent
must be incorporated through a chiral Grignard agent, which
limits side chain options to Grignard compatible functionality.
Other common methods for the generation of γ-keto ester
functionality involve enolate displacement of a bromide (or
similar leaving group) from an R-bromocarbonyl (Path B). One
such approach utilizes an electrophilic ester that is reacted with
the salt of an amino acid-derived â-keto ester. Displacement
has been reported with optically active substrates, which results
in the generation of scalemic peptide isosteres.10 The R side
chains that can be incorporated through this approach are limited
to those present in chiral R-hydroxy acids. A variation on this
theme (Path C) uses a malonate anion to effect displacement
of an amino acid-derived bromomethyl ketone. Alkylation and
decarboxylation provide access to R-substituted systems, al-
though control of stereochemistry is not possible.11
While the examples described above are representative of the
methods that have been successfully applied to the preparation
of ketomethylene-containing targets, the availability of a simple
and efficient method that will facilitate the conversion of â-keto
esters into the γ-keto ester targets could be tremendously useful.
Of additional benefit would be a method that is flexible enough
to allow the diverse incorporation of side-chain functionality.
Successful development of this flexible methodology could
contribute to a wide range of therapeutic targets.
-(CH2)5 -
Bn
H
H
Ph
H
H
2g
3g
mimics from amino acid-derived â-keto esters.12 In this ap-
proach, a suitably protected amino acid was converted to a
â-keto ester through application of the Masamune procedure.13
Preparation of the â-keto ester was followed by exposure to
ethyl(iodomethyl)zinc, which resulted in formation of the γ-keto
ester (Scheme 2). A variety of amine-protecting groups, such
as Boc, Cbz, Fmoc, and benzoyl (Bz), are tolerated with these
conditions. Furthermore, a variety of amino acid starting
materials, including lysine, aspartic acid, asparagine, and amino
acids with the hydrocarbon side chains, have been chain-
extended successfully. The efficiency of the reaction was not
dependent upon the starting amino acid, and no epimerization
of the amino acid’s stereocenters was observed in the zinc-
mediated chain extension reaction.
The facile preparation of the γ-keto ester was satisfying, since
the product contained the essential components of a dipeptide
mimic in which a ketomethylene isostere was substituted for
the peptide bond. We were interested in determining whether
amino acid-derived â-keto amides were capable of chain
extension through application of similar reaction conditions.
Furthermore, we were interested in identifying a strategy by
which to incorporate an R-substituent, which could serve to
mimic the amino acid’s side chain and could better mimic the
peptide skeleton. Studies are reported below in which amino
acid-derived â-keto amides are converted to the corresponding
γ-keto amides. Incorporation of the side chains is accomplished
through application of a tandem chain extension-aldol reaction
and tandem chain extension-homoenolate generation reaction.
We have reported that a zinc carbenoid-mediated chain
extension reaction can be used to prepare di- and tripeptide
(9) (a) Blakskjr, P.; Hoj, B.; Riber, D.; Skrydstrup, T. J. Am. Chem.
Soc. 2003, 125, 4030. (b) Hoffman, R. V.; Kim, H.-O. Tetrahedron Lett.
1992, 33, 3579. (c) Dominguez, M. J.; Gonzales-Muniz, R.; Garcia-Lopez,
M. T. Tetrahedron 1992, 48, 2761. (d) Szelke, M.; Jones, D. M.; Hallett,
A. Eur. Pat. Appl. EP 45,665; Chem. Abstr. 1982, 97, 39405. (e) Garcia-
Lopez, M. T.; Gonzales-Muniz, R.; Harto, J. R. Tetrahedron Lett. 1988,
19, 1577. (f) Dragovich, P. S.; Prins, T. J.; Zhou, R.; Webber, S. E.;
Marakovits, J. T.; Fuhrman, S. A.; Patick, A. K.; Matthews, D. A.; Lee, C.
A.; Ford, C. E.; Burke, B. J.; Rejto, P. A.; Hendrickson, T. F.; Tuntland,
T.; Brown, E. L.; Meador, J. W., III.; Ferre, R. A.; Harr, J. E. V.; Kosa, M.
B.; Worland, S. T. J. Med. Chem. 1999, 42, 1213. (g) Gomez-Monterrey,
I.; Gonzalez Muniz, R.; Perez-Martin, C.; Lopez de Ceballos, M.; Del Rio,
J.; Garcia-Lopez, M. T. Arch. Pharm. 1992, 325, 261.
Results and Discussion
Diketene was reacted with various N-protected and -unpro-
tected methyl esters of amino acids in the presence of sodium
bicarbonate or pyridine to generate the corresponding â-keto
amide substrates 2 (Table 1). Conditions reported earlier for
chain extension of amino acid-derived â-keto esters were utilized
for the formation of γ-keto amides 3. A variety of amino acids
were studied, although the study was limited to glycine and
(10) Hoffman, R. V.; Kim, H.-O. J. Org. Chem. 1995, 60, 5107.
(11) Garcia-Lopez, M. T.; Gonzalez-Muniz, R.; Hartoa, J. R.; Gomez-
Monterrey, I.; Perez, C.; De Ceballos, M. L.; Lopez, E.; Del Rio, J. Arch.
Pharm. 1992, 325, 3.
(12) Theberge, C. R.; Zercher, C. K. Tetrahedron 2003, 59, 1521.
(13) Brooks, D. W.; Lu, L. D.-L.; Masamune, S. Angew. Chem., Int.
Ed. Engl. 1979, 18, 72.
J. Org. Chem, Vol. 71, No. 21, 2006 8141

Lin et al.
SCHEME 3
SCHEME 4
Upon addition of catalytic TMSCl, the purified R-methylated
product 10 was isolated in 22% yield. The low yield was not
unexpected, since the yield for chain extension of this glycine-
derived substrate was only 36%.
An alternative, and potentially higher yielding, method for
the incorporation of the side chain involved a tandem chain
extension-zinc aldol reaction.16 We had reported earlier that
â-keto esters could be chain-extended and treated with various
aldehydes to provide good yields of R-substituted γ-keto esters.
In most cases, the aldol reaction was syn selective, with a typical
syn:anti ratio greater than 10:1; however, only a single â-keto
amide starting material was included in that study. The analysis
of products derived from tandem chain extension-aldol reac-
tions is complicated by hemiacetal formation of both the syn
and anti isomers. The use of derivatized chiral amino acids in
tandem chain extension-aldol reactions presents additional
complexity in the analysis due to diastereofacial-selectivity
issues and the appearance of rotameric forms. To simplify the
analysis, the reaction was first performed on a starting material
derived from an achiral amino acid. The â-keto amide 2d,
derived from dimethyl isobutyric acid, upon chain extension
and exposure to benzaldehyde, provided only chain-extended
material 3d. Once again, this result was not unexpected due to
the presence of the amide proton.
amino acids with hydrocarbon side chains. The effective chain
extension of both secondary and tertiary â-keto amides was
consistent with our previous report on chain extension of â-keto
amides.14 Poor yields of the isolated product were achieved with
unprotected glycine; however, the incorporation of a benzyl
group on the amide nitrogen provided access to modest yields
of the glycine derivative (3f). If the nitrogen was protected with
carbamate functionality (Cbz), the chain extension reaction
appeared to proceed efficiently; however, the isolated yield of
the purified γ-keto amide (3a) was diminished by the hydrolytic
instability of the imide functionality. Therefore, the simple chain
extension reactions were typically performed using secondary
amides.
Efficient mimicry of the peptide backbone must include
appropriately positioned side chains, preferably side chains with
functionality and stereochemical orientation identical to those
found in the natural amino acids. All of the substrates reported
above, as well as the amino acid-derived â-keto esters reported
in our earlier communication, lack this desired functionality at
the amide or ester R-carbon. We have reported previously two
tandem reaction strategies in which the intermediate formed in
the chain extension is functionalized at the R-carbon through
treatment with appropriate electrophilic reagents (Scheme 3).
The use of the intermediate’s nucleophilic character allows
selective functionalization at the least acidic site in the molecule.
We undertook an investigation of these tandem reaction
strategies for the functionalization of the amino acid-derived
ketomethylene skeleton.
The â-keto amide 2f, derived from N-benzyl glycine methyl
ester, was studied in a tandem reaction sequence in which the
enolate was reacted with trimethylacetaldehyde. Chromato-
graphic purification provided a single product 11; however, the
identity of the material, in particular the stereochemical relation-
1
ships, could not be assigned with confidence. Analysis by H
NMR was complicated due to the presence of amide rotamers
and the presumed existence of both open and closed (hemiacetal)
forms of the product. Analysis via 13C NMR clearly illustrated
the existence of the acyclic form with a ketone carbonyl signal
SCHEME 5
Incorporation of an R-methyl substituent in γ-keto esters and
amides is possible through treatment of the zinc enolate
equivalent with excess carbenoid and catalytic TMSCl.15 This
method of incorporating a methyl group adjacent to the amide
carbonyl would appear to be suitable for mimicking the methyl
side chain found in alanine. The first attempt to incorporate the
R-methyl group was performed on a leucine-derived secondary
amide 2c. All efforts to effect methylation of the R-carbon
resulted in isolation of the unsubstituted γ-keto amide 3c,
presumably due to protonation of the intermediate zinc enolate
by the secondary amide’s NH. Quenching of the intermediate’s
enolate equivalent by an amide proton was reported in earlier
studies.15
An alternate achiral â-keto amide (2f), derived from N-benzyl
glycine, was selected due to the absence of the modestly acidic
NH. The substrate was rigorously dried by heating under high
vacuum prior to its addition to 4 equiv of the zinc carbenoid.
(14) Hilgenkamp, R.; Zercher, C. K. Tetrahedron 2001, 57, 8793.
(15) Hilgenkamp, R.; Zercher, C. K. Org. Lett. 2001, 3, 3037.
8142 J. Org. Chem., Vol. 71, No. 21, 2006

Stereocontrolled Formation of Ketomethylene Isosteres
SCHEME 6
SCHEME 7
SCHEME 8
above 200 ppm and the cyclic forms (hemiacetals) with
resonances ranging between 100 and 110 ppm for the acetal
carbon.
An independent synthesis of the syn isomer 11 was under-
taken to confirm the identity and stereochemistry of the product.
Benzyl acetoacetate 12 was converted to the hemiacetal 13
through the tandem chain extension-aldol protocol.16 Acid 14
was obtained via hydrogenolysis and immediately coupled with
N-benzyl glycine methyl ester via the action of EDC. Product
11 was identical to the major component in the tandem chain
extension-aldol reaction performed earlier (Scheme 5), thereby
confirming the identical diastereoselectivity of the tandem chain
extension-aldol reaction on â-keto amide 2f. When acid 14
was treated with diazomethane, the known syn-aldol product
1516 was generated, confirming the syn stereochemical assign-
ment of 11. It should be noted that syn-aldol selectivity is
required for the formation of threonine mimics.
While the tandem reaction procedures illustrated in Schemes
4 and 5 provided access to modified amino acid skeletons, the
control of absolute stereochemistry had not been explored. The
methyl ester of L-proline was reacted with diketene to provide
a â-keto amide 16. The chain extension reaction proceeded
cleanly to provide γ-keto amide 17. Exposure of 16 to the zinc
carbenoid followed by addition of benzaldehyde resulted in a
good yield of a single aldol product 18 (Scheme 6). Although
the existence of rotameric forms and the presence of both acyclic
and cyclic (hemiacetal) isomeric forms made analysis of the
crude reaction mixture difficult, generation of an X-ray quality
crystal allowed the assignment of the stereochemistry of the
major reaction product. The aldol reaction was, as expected,
syn selective. The enolate facial selectivity observed in the
reaction was influenced by the L-proline, with the newly
incorporated R-stereocenter consistent with the natural stereo-
chemistry found in L-amino acids.
The aldol product 18 could be manipulated to provide access
to side chains found in the natural amino acids. Formation of
the xanthate could be accomplished through opening the
hemiacetal. Reductive cleavage of the xanthate resulted in a
phenylalanine mimic (Scheme 7).
A serine mimic 20 was formed in the reaction of 16 with
formaldehyde. Once again, the enolate facial selectivity was
outstanding. The hydroxymethyl side chain, which prefers the
acyclic form, could be converted to the xanthate and reduced
with tributyltin hydride to provide the R-methylated γ-keto
amide (alanine mimic) 21 (Scheme 8). The stereoselectivity of
this aldol reaction was established through comparison of 21
to an identical material generated through the homoenolate
strategy described below.
An alternative approach to the alanine mimic through chain
extension and treatment with catalytic TMSCl was explored.
Protonation of the intermediate homoenolate provided the
R-methylated γ-keto amide 21. Diastereoselectivity of this
1
methylation reaction was estimated as >92:8 from H NMR
integration. The relative stereochemistry was established through
X-ray crystallographic analysis of the 2,4-dinitrophenylhydra-
zone derivative 22. Once again, the L-proline stereocenter
influenced formation of natural (L)-stereochemistry at the
R-carbon. The spectral data for 21 generated via the tandem
chain extension-aldol sequence was identical to that generated
via the homoenolate.
In conclusion, we have demonstrated that zinc carbenoid-
mediated chain extension reactions can be performed on amino
acid-derived â-keto amides. The products of these reactions are
γ-keto amides, which appear suitable to serve as ketomethylene
(16) Lai, S.-J.; Zercher, C. K.; Jasinski, J. P.; Reid, S. N.; Staples, R. J.
Org. Lett. 2001, 3, 4169.
J. Org. Chem, Vol. 71, No. 21, 2006 8143

Lin et al.
3H), 3.22 (dd, 1H, J ) 10.3, 18.3 Hz), 2.92 (m, 1H), 2.79 (dd, 1H,
J ) 2.9, 18.3 Hz), 2.55 (s, 3H), 2.14 (s, 3H), 2.06 (m, 1H), 1.88-
1.73 (m, 3H); 13C NMR (125 MHz, CDCl3) δ 214.1, 207.2, 172.2,
169.8, 137.0, 128.5, 128.4, 127.4, 84.1, 59.2, 51.9, 47.0, 45.4, 43.6,
30.1, 29.2, 24.8, 19.2. Resonances observed for the minor rota-
mer: 1H NMR (500 MHz, CDCl3) δ 6.51 (d, 1H, J ) 10.0 Hz),
3.84 (d, 1H, J ) 8.6 Hz), 3.77 (s, 3H), 2.53 (s, 3H); 13C NMR
(125 MHz, CDCl3) δ 214.4, 205.7, 172.3, 169.7, 137.4, 128.5,
126.8, 85.0, 59.1, 52.3, 46.1, 45.8, 43.5, 30.6, 30.2, 22.1, 19.2. IR
(neat, cm-1): 2954-2852, 1740, 1715, 1640, 1436, 1202. HRMS
(ESI) Calcd for [M+ + H]): 424.1247; Found: 424.1255.
isosteric replacements for peptide bonds. The use of tandem
reaction processes facilitates the incorporation of side chains
into these mimics. The chain extension-aldol reaction proceeds
with threonine-like stereoselectivity. The alcohol functionality
can be deoxygenated to provide hydrocarbon side chains.
Methylation can be induced through the intermediacy of an
intermediate homoenolate, which provides direct access to
alanine analogues. Use of L-proline-derived â-keto amides offers
efficient diastereocontrolled formation of L-amino acid mimics.
Studies involving more complex starting materials are underway
in our laboratory, as well as studies that target specific proteases
through ketomethylene isostere formation.
A 10-mL oven-dried, round-bottomed flask, equipped with a
septum with a flow of nitrogen through a needle and a stir bar,
was charged with toluene (4 mL). The xanthate (0.16 g, 0.4 mmol),
tributyltin hydride (0.16 mL, 0.6 mmol), and AIBN (10 mg, 0.06
mmol) were added sequentially to the flask. The solution was heated
to 80 °C for 8 h. The mixture was concentrated in vacuo, and the
resulting residue was purified by column chromatography on silica
gel (hexane/ethyl acetate, 1:1; Rf ) 0.11) to give 88 mg of
compound 19. [R]25D +3.8° (c 0.002 g/mL, CHCl3). 1H NMR (500
MHz, CDCl3) δ 7.32-7.18 (m, 5H), 4.46 (dd, 1H, J ) 4.7, 8.9
Hz), 3.81 (m, 1H), 3.73 (s, 3H), 3.40 (m, 1H), 3.20 (m, 1H), 3.12-
3.00 (m, 2H), 2.59 (dd, 1H, J ) 8.8, 13.7 Hz), 2.38 (dd, 1H, J )
Experimental Section
2-(5-Hydroxy-5-methyl-2-phenyltetrahydrofuran-3-carbonyl)-
cyclopentanecarboxylic Acid Methyl Ester (18). A 25-mL oven-
dried, round-bottomed flask, equipped with a septum with a flow
of nitrogen through a needle and a stir bar, was charged with 10
mL of methylene chloride and diethyl zinc (1.0 M in hexanes, 2.5
mL, 2.5 mmol). The solution was cooled to 0 °C, and methylene
iodide (0.20 mL, 2.5 mmol) was added slowly by syringe. After
the solution was stirred for 10 min, compound 16 (0.22 g, 1.0 mmol,
in 2 mL of methylene chloride) was added by syringe to the
resulting white suspension. The mixture was stirred for 30 min, at
which time freshly distilled benzaldehyde (0.13 g, 1.2 mmol) was
added into the reaction mixture by syringe. After TLC analysis
(hexane/ethyl acetate, 1.5:1; Rf ) 0.20) indicated the chain extension
intermediate was consumed, the solution was quenched by cautious
addition of saturated aqueous ammonium chloride (10 mL). The
mixture was extracted with diethyl ether (2 × 15 mL). The
combined organic extracts were washed with brine (10 mL) and
dried over anhydrous sodium sulfate. The organic solution was
filtered and concentrated under reduced pressure. The residue was
chromatographed on silica (hexane/ethyl acetate, 1:1; Rf ) 0.25)
to yield 0.23 g (70%) of compound 18 as white crystals (mp 113-
3.2, 18.0 Hz), 2.18 (m, 1H), 2.06 (s, 3H), 2.06-1.92 (m, 3H); 13
C
NMR (125 MHz, CDCl3) δ 207.8, 173.4, 172.7, 138.7, 129.1, 128.4,
126.4, 58.7, 52.0, 46.8, 45.1, 40.6, 37.7, 29.9, 29.1, 24.8. IR (neat,
cm-1): 2923-2851, 1744, 1713, 1634, 1436, 1365. HRMS (ESI)
Calcd for [M+ + Na]: 340.1519; Found: 340.1517.
1-(2-Hydroxymethyl-4-oxopentanoyl)-pyrrolidine-2-carboxy-
lic Acid Methyl Ester (20). A 100-mL oven-dried, round-bottomed
flask, equipped with a septum with a flow of nitrogen through a
needle and a stir bar, was charged with 35 mL of methylene chloride
and diethyl zinc (1.0 M in hexanes, 15.0 mL, 15.0 mmol). The
solution was cooled to 0 °C, and methylene iodide (1.20 mL, 15.0
mmol) was added slowly by syringe. After the solution was stirred
for 10 min, compound 16 (0.64 g, 3.0 mmol, in 3 mL of methylene
chloride) was added by syringe to the resulting white suspension.
The mixture was stirred for 30 min. Paraformaldehyde (0.5 g) was
placed in a dry one-necked round-bottomed flask, which was capped
with a rubber septum. One end of a wide-bore cannula was inserted
through the septum in the flask that contained paraformaldehyde,
and the other end of the cannula was inserted through the septum
in the flask that contained the zinc reagent. The paraformaldehyde
flask was heated with a heat gun to induce formation of formal-
dehyde, which was bubbled into the zinc-carbenoid solution for
10 min. (Caution: Polymerization of the gaseous formaldehyde in
the cannula, which can cause blockage of the cannula, must be
avoided. Use of a tygon tube with barrels of disposable syringes at
either end provides a suitable cannula for this purpose) After TLC
analysis (hexane/ethyl acetate, 1.5:1; Rf ) 0.20) indicated the chain
extension intermediate was consumed, the solution was quenched
by cautious addition of saturated aqueous ammonium chloride (15
mL). The solution was extracted with ethyl acetate (3 × 20 mL).
The combined organic layers were washed with brine (20 mL) and
dried over anhydrous sodium sulfate. The resulting liquid was
filtered and concentrated under reduced pressure. The residue was
chromatographed on silica (hexane/ethyl acetate, 1:1; Rf ) 0.20)
to yield 0.57 g (74%) of compound 20 as a yellow oil with some
115 °C) in which one major hemiacetal form was present. [R]25
D
1
-11.6° (c 0.011 g/mL, CHCl3). H NMR (500 MHz, CDCl3) δ
7.41-7.26 (m, 5H), 6.65 (s, 1H,), 5.18 (d, 1H, J ) 5.0 Hz), 4.56
(dd, 1H, J ) 3.5, 8.6 Hz), 3.80 (s, 3H), 3.38-3.23 (m, 3H), 2.31-
1.84 (m, 6H), 1.69 (s, 3H); 13C NMR (125 MHz, CDCl3) δ 174.8,
171.9, 142.0, 128.5, 127.8, 125.4, 105.3, 83.2, 58.8, 52.2, 51.0,
47.6, 41.6, 28.9, 24.9, 24.3. The presence of the open chain form
as a minor constituent is confirmed by the 13C resonance of 207.7.
IR (neat, cm-1): 3400 (b, OH), 2956, 1744, 1640, 1448.
1-(2-Benzyl-4-oxopentanoyl)-pyrrolidine-2-carboxylic Acid
Methyl Ester (19). A 10-mL oven-dried, round-bottomed flask,
equipped with a septum with a flow of nitrogen through a needle
and a stir bar, was charged with sodium hydride (24 mg, 0.6 mmol,
60% in the mineral oil). Hexanes (2 mL) were added to the flask
to wash the sodium hydride. The hexane was removed by syringe,
and then tetrahydrofuran (4 mL) was added by syringe. The solution
was cooled to 0 °C, and carbon disulfide (76 mg, 1.0 mmol) was
added by syringe. After the solution was stirred for 5 min,
compound 18 (0.10 g, 0.3 mmol, in 1 mL of tetrahydrofuran) was
added by syringe, followed by methyl iodide (0.14 mg, 1.0 mmol).
After TLC analysis (hexane/ethyl acetate, 1:1; Rf ) 0.25) indicated
the starting material 18 was consumed, the solution was quenched
with saturated aqueous ammonium chloride (3 mL). The solution
was extracted with diethyl ether (2 × 5 mL). The combined organic
layers were washed with brine (5 mL) and dried over anhydrous
sodium sulfate. The resulting liquid was filtered and concentrated
under reduced pressure. The residue was chromatographed on silica
(hexane/ethyl acetate, 1.5:1; Rf ) 0.20) to yield 96 mg (74%) of
the xanthate as a yellow oily mixture of two rotamers in a 3.3:1
ratio. [R]25D +15° (c 0.004 g/mL, CHCl3). Major rotamer: 1H NMR
(500 MHz, CDCl3) δ 7.37-7.27 (m, 5H), 6.59 (d, 1H, J ) 9.0
Hz), 4.31 (dd, 1H, J ) 4.4, 8.4 Hz), 3.77-3.64 (m, 2H), 3.64 (s,
minor hemiacetal forms present. [R]25 -25.3° (c 0.003 g/mL,
D
CHCl3). 1H NMR (500 MHz, CDCl3) δ 4.52 (dd, 1H, J ) 4.8, 8.9
Hz), 3.88-3.84 (m, 2H), 3.72 (s, 3H), 3.68-3.66 (m, 2H), 3.28
(m, 1H), 3.20 (t, 1H, J ) 7.0 Hz), 3.01 (dd, 1H, J ) 10.0, 18.4
Hz), 2.48 (dd, 1H, J ) 3.7, 18.4 Hz), 2.24 (m, 1H), 2.12 (s, 3H),
2.02-1.93 (m, 3H); 13C NMR (125 MHz, CDCl3) δ 207.0, 173.6,
173.3, 64.5, 58.8, 52.5, 47.3, 42.1, 41.2, 29.8, 29.0, 24.8. Minor
hemiacetal forms: 13C NMR (125 MHz, CDCl3) δ 175.3, 172.6,
172.0, 105.0, 104.9, 69.5, 69.4, 25.1, 24.6. IR (neat, cm-1): 3435
(b, OH), 2956-2852, 1743, 1714, 1624, 1447. HRMS (ESI) Calcd
for [M+ + Na]: 280.1155; Found: 280.1152.
8144 J. Org. Chem., Vol. 71, No. 21, 2006

Stereocontrolled Formation of Ketomethylene Isosteres
1-(2-Methyl-4-oxopentanoyl)-pyrrolidine-2-carboxylic Acid
Methyl Ester (21) (via Xanthate Formation and Reduction). A
10-mL oven-dried, round-bottomed flask, equipped with a septum
with a flow of nitrogen through a needle and a stir bar, was charged
with sodium hydride (48 mg, 1.2 mmol, 60% in the mineral oil).
Hexanes (2 mL) were added to the flask to wash the sodium
hydride. The hexane was removed by syringe, and then tetrahy-
drofuran (4 mL) was added by syringe. The solution was cooled to
0 °C, and carbon disulfide (0.15 g, 2.0 mmol) was added by syringe.
After the solution was stirred for 5 min, compound 20 (0.26 g, 1.0
mmol, in 1 mL of THF) was added by syringe, followed by methyl
iodide (0.21 g, 1.5 mmol). After TLC analysis (hexane/ethyl acetate,
1:1; Rf ) 0.20) indicated the starting material 20 was consumed,
the solution was quenched with saturated aqueous ammonium
chloride (3 mL). The solution was extracted with diethyl ether (2
× 5 mL). The combined organic layers were washed with brine (5
mL) and dried over anhydrous sodium sulfate. The resulting liquid
was filtered and concentrated under reduced pressure. The residue
was chromatographed on silica (hexane/ethyl acetate, 1.5:1; Rf )
The flask was immersed in an ice-water bath, and the solution
was cooled to 0 °C. Methylene iodide (1.1 mL, 7.5 mmol) was
added in one portion, and the solution was stirred for 10 min. â-Keto
ester 16 (0.400 g, 1.9 mmol, dissolved in 5 mL of methylene
chloride) was added to the white suspension, and the mixture was
stirred for 40 min, at which time chlorotrimethylsilane (0.4 mL,
2.7 mmol) was added. The reaction was allowed to stir for 80 min,
then quenched with the addition of 20 mL of saturated aqueous
ammonium chloride. The mixture was extracted with diethyl ether,
which was washed with brine and dried over anhydrous sodium
sulfate. The resulting liquid was filtered and concentrated on a rotary
evaporator. The residue was chromatographed on silica (hexane/
ethyl acetate, 3:1; Rf ) 0.15) to yield 0.187 g (39%) of compound
21 as a yellow oil. [R]D25 -58.4° (c 0.125, CHCl3). 1H NMR (500
MHz, CDCl3) δ 4.49-4.45 (dd, 1H, J ) 4.4, 8.1 Hz), 3.84-3.64
(m, 2H), 3.71 (s, 3H), 3.14-3.02 (m, 2H), 2.12 (s, 3H), 2.42-
1.94 (m, 4H), 1.15 (d, 3H, J ) 10 Hz); 13C NMR (125 MHz, CHCl3)
δ 208.1, 174.8, 173.2, 58.8, 52.3, 47.4, 47.0, 33.3, 30.3, 29.3, 25.1,
16.9; IR (neat, cm-1) 2957, 1746, 1645, 1434.
0.2) to yield 0.24 g (68%) of the xanthate as a yellow oil. [R]25
D
2,4-Dinitrophenylhydrazone Derivative of 21 (22). Concen-
trated sulfuric acid (0.2 mL) was added to a test tube that contained
2,4-dinitrophenylhydrazine (119 mg, 0.6 mmol). Distilled water was
added dropwise with swirling until the hydrazine reagent dissolved.
To this warm solution was added 2 mL of 95% ethanol. The freshly
prepared 2,4-DNP solution was added to a solution of 21 (120 mg
in 2 mL of 95% ethanol, 0.5 mmol). The solution was allowed to
sit at room temperature for 24 h, at which time crystallization was
observed. The solution that contained the crystals was concentrated
on a rotary evaporator, and the resulting solid orange residue
dissolved in a mixture of EtOAc/CHCl3 (approximately 15:1) over
a steam bath. The clear orange solution was stored in a refrigerator
(approximately 5 °C). Formation of crystals (clear prisms, mp )
-2.0° (c 0.007 g/mL, CHCl3). 1H NMR (400 MHz, CDCl3) δ 4.78
(dd, 1H, J ) 7.3, 10.7 Hz), 4.58 (dd, 1H, J ) 7.3, 10.7 Hz), 4.46
(dd, 1H, J ) 4.6, 8.6 Hz), 3.90 (m, 1H), 3.79 (m, 1H), 3.70 (s,
3H), 3.58 (m, 1H), 3.14 (dd, 1H, J ) 10.4, 18.2 Hz), 2.60 (dd, 1H,
J ) 3.3, 18.2 Hz), 2.58 (s, 3H), 2.22 (m, 1H), 2.15 (s, 3H), 2.10-
1.96 (m, 3H); 13C NMR (125 MHz, CDCl3) δ 215.5, 206.5, 172.4,
170.5, 72.9, 59.0, 52.2, 47.3, 42.8, 38.1, 29.9, 29.2, 24.9, 19.0. IR
(neat, cm-1): 2955-2850, 1743-1640, 1640, 1214, 1068. HRMS
(ESI) Calcd for [M+ + H]: 348.0934; Found: 348.0936.
A 10-mL oven-dried, round-bottomed flask, equipped with a
septum with a flow of nitrogen through a needle and a stir bar,
was charged with toluene (4 mL). The xanthate (35 mg, 0.1 mmol),
tributyltin hydride (0.05 mL, 0.2 mmol), and AIBN (5 mg, 0.03
mmol) were added to the flask in the indicated order. Then the
solution was heated to 80 °C for 8 h. The mixture was concentrated
in vacuo, and the resulting residue was purified by column
chromatography on silica gel (hexane/ethyl acetate, 3:1; Rf ) 0.15)
25
164-165 °C) was observed within 24 h. [R]D -60.7° (c 0.173,
CHCl3). 1H NMR (400 MHz, CDCl3) δ 11.03 (s, 1H), 9.12 (s, 1H),
8.26-7.65 (m, 2H), 4.49-4.45 (dd, 1H, J ) 4.4, 8.1 Hz), 3.84-
3.64 (m, 2H), 3.71 (s, 3H), 3.14-3.02 (m, 2H), 2.12 (s, 3H), 2.42-
1.94 (m, 4H), 1.15 (d, 3H, J ) 10 Hz); 13C NMR (100 MHz, CDCl3)
δ 174.9, 173.2, 156.8, 145.3, 137.5, 130.0, 124.0, 115.8, 59.3, 52.8,
47.1, 42.4, 35.2, 29.8, 25.0, 17.2; IR (neat, cm-1) 3338, 1744, 1637,
1431. HRMS (ESI) Calcd for [M+ + H]: 422.1676; Found:
422.1670.
1
to give 17 mg (70%) of compound 21 as a yellow oil. H NMR
(400 MHz, CDCl3) δ 4.46 (dd, 1H, J ) 4.6, 8.7 Hz), 3.83-3.64
(m, 2H), 3.70 (s, 3H), 3.12-2.94 (m, 2H), 2.36 (dd, 1H, J ) 3.2,
17.3 Hz), 2.12 (s, 3H), 2.25-1.93 (m, 4H), 1.15 (d, 3H, J ) 6.9
Hz); 13C NMR (125 MHz, CDCl3) δ 207.9, 174.5, 172.9, 58.5,
52.1, 47.2, 46.8, 33.1, 30.1, 29.1, 24.9, 16.7. IR (neat, cm-1): 2957,
1746, 1645, 1434.
1-(2-Methyl-4-oxopentanoyl)-pyrrolidine-2-carboxylic Acid
Methyl Ester (21) (via Homoenolate). A 100-mL round-bottomed
flask was equipped with a stir bar and septum and was charged
with 15 mL of methylene chloride and diethyl zinc (5.0 mL of a
1.0 M solution in hexane, 5.0 mmol) under a nitrogen atmosphere.
Supporting Information Available: Experimental procedures
for the preparation of 2a-g, 3a-g, 10, 11, 16, and 17. 1H and 13
C
NMR spectra of 2a-g, 3a-g, 10, 11, 16, 17, 18, 19, 21, and 22.
Crystal structure data for 18 and 22. This material is available free
of charge via the Internet at http://pubs.acs.org.
JO061184O
J. Org. Chem, Vol. 71, No. 21, 2006 8145