Highly efficient chiral-pool synthesis of (2S,4R)-4-hydroxyornithine

Article abstract of DOI:10.1021/ol016445p

(Figure Presented) A concise synthesis of the amino acid (2S,4R)-4-hydroxyornithine is described. Starting from diprotected L-aspartic acid, the scaffold of the target compound is constructed in a three-step approach: an efficient α-nitroketone formation

Full text of DOI:10.1021/ol016445p

ORGANIC
LETTERS
2001
Vol. 3, No. 20
3153-3155
Highly Efficient Chiral-Pool Synthesis of
(2S,4R)-4-Hydroxyornithine
,†
Joachim Rudolph,* Frithjof Hannig, Heidi Theis, and Ralf Wischnat
Bayer AG, Central Research, Life Science, Building Q18,
D-51368 LeVerkusen, Germany
joachim.rudolph.b@bayer.com
Received July 17, 2001
ABSTRACT
A concise synthesis of the amino acid (2S,4R)-4-hydroxyornithine is described. Starting from diprotected L-aspartic acid, the scaffold of the
target compound is constructed in a three-step approach: an efficient r-nitroketone formation through acylation of nitromethane is followed
by a diastereoselective reduction of the resulting ketone. In the last step, the nitro group is reduced to furnish the (2S,4R)-4-hydroxyornithine
scaffold. This new approach to the title compound offers advantages to the synthetic pathways previously described.
(2S,4R)-4-Hydroxyornithine (1) is a nonproteinogenic amino
acid that is widely found in nature. For example, it is a
not proceed stereoselectively,6-10 two recent selective syn-
thetic pathways to prepare the (2S,4R)-4-hydroxyornithine
structure have been described. Schmidt et al. synthesized the
scaffold of 1 starting from (R)-isopropylideneglyceraldehyde.
This sequence involves about 10 steps and proceeds non-
diastereoselectively, i.e., an asymmetric catalytic hydrogena-
tion step is needed for stereoselective introduction of the
second stereogenic center.11
component of marine organism1 and plants,2 as well as a
constituent of a number of peptide natural products, such as
the antifungal lipopeptides echinocandin and pneumocandin,3
the K 582 type antibiotics,4 and the biphenomycin antibiot-
ics.5 The interest in (2S,4R)-4-hydroxyornithine (1) is
documented in a variety of synthetic approaches previously
described in the literature. Apart from procedures that do
A very concise approach using L-serine as starting material
is described by Jackson. The key step is a palladium-
catalyzed coupling of a serine-derived zinc compound with
an acid chloride to form ketone 2, which is later reduced in
†New address: Department of Chemistry Research, Bayer Corporation,
Pharmaceutical Division, 400 Morgan Lane, West Haven, CT 06516.
E-mail: joachim.rudolph.b@bayer.com.
(1) (a) Fujita, Y. Bull Chem. Soc. Jpn. 1959, 32, 439-442. (b) Makisumi,
S. J. Biochem. (Tokyo) 1961, 49, 284-291.
(2) (a) Bell, E. A.; Tirimanna, A. S. L. Nature (London) 1963, 197, 901-
902. (b) Bell, E. A.; Tirimanna, A. S. L. Biochem. J. 1964, 91, 356-358.
(3) Denning, D. W. J. Antimicrob. Chemother. 1997, 40, 611-614.
(4) Kondo, S.; Meguriya, N.; Mogi, H.; Aota, T.; Miura, K.; Fujii, T.;
Hayashi, I.; Makino, K.; Yamamoto, M.; Nakajima, N. J. Antibiot. 1980,
33, 533-542.
(5) Ezaki, M.; Iwami, M.; Yamashita, M.; Hashimoto, S.; Komori, T.;
Umehara, K.; Mine, Y.; Kohsaka, M.; Aoki, H.; Imanaka, I. J. Antibiot.
1985, 38, 1453-1461.
a diastereoselective fashion. The elegance of this approach
(6) Hammarsten, E. Compt. Rend. TraV. Lab. Carlsberg, II 1916, 223.
(7) Talbot, G.; Gaudry, R.; Berlinguet, L. Can. J. Chem. 1956, 34, 911-
914.
(8) Mizusaki, K.; Makisumi, S. Bull. Chem. Soc. Jpn. 1981, 54, 470-
472.
(9) Girard, A.; Greck, C.; Geneˆt, J. P. Tetrahedron Lett. 1998, 39, 4259-
4260.
(10) (a) Mues, H.; Dissertation, Universita¨t Heidelberg, 2000. (b) Mues,
H.; Kazmaier, U. Synthesis 2001, 487-498.
(11) Schmidt, U.; Meyer, R.; Leitenberger, V.; Sta¨bler, F.; Lieberknecht,
A. Synthesis 1991, 409-413.
10.1021/ol016445p CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/05/2001

however is hampered by the low yield (<30%) of the C-C
Dess-Martin reagent. We then performed the Henry reaction
using nitromethane and 5% sodium ethoxide in ethanol which
proved to be the best conditions with respect to substrate
conversion. However, selectivity of this reaction was found
to be unfavorable, namely, the undesired (2S,4S)-diastere-
omer (threo) (7b) was formed in a 3:2 ratio with regard to
the desired (2S,4R)-diastereomer (erythro) (7a) as determined
coupling step.12
Overall, access to (2S,4R)-4-hydroxyornithine (1) remained
problematic since either complicated purification procedures,
numerous steps, or unsatisfactory reactions were involved.
As we were interested in a more efficient access to
(2S,4R)-4-hydroxyornithine (1), we envisioned a strategy
starting from L-aspartic acid and adding a nucleophilic C-N
unit to the carboxylic acid function of its side chain.
We began our investigations with a nitroaldol reaction
approach13 using the semialdehyde of diprotected L-aspartic
acid14 and nitromethane, as shown in Scheme 1. As starting
1
by H NMR analysis (stereochemistry assigned via X-ray
analysis of the isolated threo-isomer 7b). Following this
approach, we were able to isolate only 15% of the nitro
alcohol 7a. Attempts to improve the selectivity were unsuc-
cessful.17
Because of the unfavorable diastereoselectivity and yield
of the nitroaldol reaction, we turned to a different strategy
for C-C-coupling which would involve the generation of
an R-nitroketone. It is known that R-nitroketones can be
formed by reaction of acylimidazoles with deprotonated
nitromethane.18,19 On the basis of this method, we developed
a versatile procedure which proved to be superior for our
purposes compared to the original one (Scheme 2): (S)-N-
Scheme 1. Synthetic Pathway to (2S,4R)-4-Hydroxyornithine
via Henry Reactiona
Scheme 2. Synthesis of the Hydroxyornithine Derivative 10
via Acylation of Nitromethane as the Key Stepa
a (i) NMM, ClCOOEt, THF; NaBH4, H2O; (ii) polymer-bound
bromite(I) complex 6, catalytic TEMPO; (iii) CH3NO2, NaOEt
(5%), EtOH.
a (i) CDI (1.05 equiv), THF, rt; CH3NO2 (10 equiv), t-BuOK
(1.1 equiv), rt; (ii) L-Selectride, THF, -78 °C; (iii) catalytic Pd/C,
NH4+HCOO-, -10 °C; (iv) Z-OSu, DIPEA, DMF, rt.
material for the synthesis of the aspartic acid semialdehyde,
we chose commercially available (S)-N-Boc-aspartic acid
tert-butyl ester 3. The tert-butyl ester group of γ-hydroxy
amino acids is known to be stable toward lactonization which
we preferred to avoid. (S)-N-Boc aspartic acid tert-butyl ester
3 was reduced to the corresponding homoserine derivative
4,15 which was subsequently oxidized to the semialdehyde
5. The method of choice for oxidizing the alcohol function
of 4 was found to be the polymer-bound bromite(I) complex
6 as recently described by Kirschning et al.16 This method
gave very high yields and was found to be superior to the
Boc-aspartic acid tert-butyl ester 3 was activated with
carbonyl diimidazole and transformed to the nitroketone 9
by treatment with excess nitromethane in the presence of
t-BuOK in a very clean reaction (96% yield). It was found
to be particularly important to use an excess of nitromethane
(10-fold) and an equimolar amount of t-BuOK. After the
successful implementation of the scaffold of hydroxy-
ornithine, we wanted to first reduce the keto group of 9 to
an alcohol function in a diastereoselective manner and
(12) (a) Jackson, R. F. W.; Rettie, A. B.; Wood, A.; Wythes, M. J. J.
Chem. Soc., Perkin Trans. 1 1994, 1719-1726. (b) Jackson, R. F. W.;
Wishart, N.; Wood, A.; James, K.; Wythes, M. J. J. Org. Chem. 1992, 57,
3397-3404. (c) Jackson, R. F. W.; Wood, A.; Wythes, M. J. Synlett 1990,
735-736.
(13) (a) Luzzio, F. A. Tetrahedron 2001, 57, 915-945. (b) Rosini, G.
In ComprehensiVe Organic Synthesis; Trost, B. M., Ed.; Pergamon: Oxford,
1991; Vol. 2, p 321.
(14) For a comprehensive overview on different strategies for the
synthesis of aspartaldehyde derivatives, see: Meffre, P. Amino Acids 1999,
16, 251-272.
(15) Valerio, R. M.; Alewood, P. F.; Johns, R. B. Synthesis 1988, 786-
789.
(17) Earlier studies also indicated that control of stereochemistry is
difficult for Henry reactions of nitroalkanes with aldehydes bearing a chiral
center in the â-position. See: (a) Moorman, A. R.; Martin, T.; Borchardt,
R. T. Carbohydr. Res. 1983, 113, 233-239. (b) Nakata, T.; Komatsu, T.;
Nagasawa, K.; Yamada, H.; Takahashi, T. Tetrahedron Lett. 1994, 35,
8225-8228.
(18) Baker, D. C.; Putt, S. R. Synthesis 1978, 478-479. According to
the authors, the high preference of C-acylation over O-acylation for the
ambidently nucleophilic methanenitronate anion is assigned to the unique
reactivity of the acylimidazole which behaves differently from other
activated acyl species.
(16) Sourkouni-Agirusi, G.; Kirschning, A. Org. Lett. 2000, 2, 3781-
3784.
(19) For a recent application, see: Yuasa, Y.; Yuasa, Y.; Tsuruta, H.
Synth. Commun. 1998, 28, 395-401.
3154
Org. Lett., Vol. 3, No. 20, 2001

subsequently reduce the nitro group to an amino function.
As is known from detailed studies on the reduction of 4-keto
δ-amino acids by Jackson,12 the R-center of such compounds
exerts moderate stereochemical control and use of the
reducing agent L-Selectride20 gives the highest levels of
stereoselectivity in favor of the erythro-diastereoisomer.
In summary, a practical and concise synthesis of the
nonproteinogenic amino acid (2S,4R)-4-hydroxyornithine has
been developed starting from commercially available dipro-
tected L-aspartic acid. The synthesis involves an efficient
homologization of the acid side chain to form an R-nitro-
ketone in the first step and subsequent generation of the
second stereogenic center through diastereoselective ketone
reduction. We find that this approach allows a very direct
access to (2S,4R)-4-hydroxyornithine derivatives that offers
some advantages over the syntheses previously described.
In addition, the described two-step approach to nitro alcohols
is a useful and general alternative to the classical Henry
reaction.
Accordingly, we found that using L-Selectride gave rise
to a 85:15 mixture in favor of the desired erythro compound
7a () (2S,4R)-diastereomer). We were able to separate the
diastereomers via silica gel chromatography. In the last step
of the sequence, the nitro group was reduced to furnish the
vicinal amino alcohol function of the tert-butyl (2S,4R)-N-
Boc-4-hydroxyornithinate 8. Surprisingly, hydrogen reduc-
tion using different catalysts under a variety of reaction
conditions (different pressures, temperatures) was unsuc-
cessful. Instead, hydrogen transfer, i.e., ammonium formate
in methanol in the presence of palladium on carbon, proved
to be the method of choice for this transformation.21 The
resulting amino alcohol 8 was found to be unstable and was
therefore further transformed to the corresponding Nδ-Z-
protected hydroxyornithine derivative 10.
Acknowledgment. We thank Jordi Benet-Buchholz for
the crystal structure determination, Benjamin Meseguer for
his support in reduction catalyst screening, and finally,
Isabelle Lefebvre for experimental advice.
Supporting Information Available: Description of all
experimental procedures and product characterization. Crystal
structure data for compound 7b. This material is available
free of charge via the Internet at http://pubs.acs.org.
(20) Wittmann, S.; Schoenecker, B. J. Prakt. Chem./Chem.-Ztg. 1996,
338, 759-762.
(21) (a) Ranu, B. C.; Sarkar, A.; Guchhait, S. K.; Ghosh, K. J. Indian
Chem. Soc. 1998, 75, 690-694. (b) Ram, S.; Ehrenkaufer, R. E. Synthesis
1988, 91-95.
OL016445P
Org. Lett., Vol. 3, No. 20, 2001
3155