Synthesis of a β-amino acid pharmacophore via a β-lactam intermediate

Article abstract of DOI:10.1021/jo048249c

(Chemical Equation Presented) A stereoselective synthesis of (R)-β-amino acid 1 via a β-lactam intermediate is discussed.

Full text of DOI:10.1021/jo048249c

Synthesis of a â-Amino Acid
Pharmacophore via a â-Lactam
Intermediate
Re´my Angelaud,* Yong-Li Zhong, Peter Maligres,
Jaemoon Lee, and David Askin
FIGURE 1. (R)-â-amino acid 1.
Merck Process Research, Merck Research Laboratories,
P.O. Box 2000, Rahway, New Jersey 07065
elaborate â-amino acids.7,8 The Friedel-Crafts acylation
of 1,4-difluorobenzene 3 in the presence of AlCl3 with
commercially available (S)-N-trifluoroacetylaspartic an-
hydride 2 unfortunately led exclusively to the undesired
R-acylated compound9 4 (Scheme 1). Contrary to what
has been observed by Griesbeck8 with toluene or o-xylene,
none of the desired â-acylated 5 compound could be
observed with 3.
remy_angelaud@merck.com
Received October 5, 2004
In an alternate approach, unsaturated esters are ideal
substrates for Michael addition of chiral amines10 and
as shown by Miller,11 for the asymmetric conjugate
addition of azide ion. Unfortunately, due to the strong
withdrawing electronic effect of the difluorophenyl ring,
the synthesis of R,â-unsaturated ester 8 was found to be
challenging. Precursor aldehyde 7 could only be made by
oxidation of alcohol 6 with Dess-Martin periodinane or
excess MnO2. Swern oxidation of 6 led to complete
decomposition. Furthermore, aldehyde 7 was very un-
stable and could not be purified, so it was used directly
in the subsequent Wittig-Horner reaction. Treatment of
7 with methyl dimethoxyphosphono acetate and t-BuOK
(DBU led to decomposition) in methanol gave some
desired ester 8 alongside with a major compound 9 where
the double bond has been isomerized R to the difluo-
rophenyl ring in a 23:77 ratio (Scheme 2). Attempts to
equilibrate this mixture with an excess of base led
exclusively to the thermodynamically more stable styrene
derivative 9. 12
Furthermore, the very elegant acyl halide-aldehyde
cyclocondensation (AAC) developed by Nelson13 could not
be tried because it requires the use of unstable alde-
hyde 7.
Thus, we turned our attention to the biomimetic
synthesis of â-lactams discovered by Miller (Scheme 3).14
Chiral â-lactams have been previously used as precursors
to chiral â-amino acids;15 however, little was known about
the direct opening of O-benzyloxy-protected lactams to
the corresponding â-amino acids.16
A stereoselective synthesis of (R)-â-amino acid 1 via a
â-lactam intermediate is discussed.
As part of our ongoing research program for the
discovery of new drug candidates, we recently required
the stereoselective synthesis of the BOC-protected key
intermediate (R)-â-amino acid 1 (Figure 1) on a kilogram
scale. The selective synthesis of â-amino acids1 has been
the subject of tremendous effort principally due to their
important biological activity2 as enzyme inhibitors or as
R-amino acid surrogates in the construction of peptides
possessing unique conformational properties3 (â-pep-
tides). Furthermore, the â-amino acid pattern can be
found in some interesting naturally occurring com-
pounds.4
Arndt-Eistert homologation,5 Curtius rearrangemen,5
and asymmetric addition to imines6 (i.e., Staudinger
reaction) were abandoned due to potential issues for
scale-up associated with the manipulation of hazardous
material. So we turned our attention to the most promis-
ing procedures in terms of cost, efficiency, and scalability.
Aspartic acid and its derivatives were first chosen as
potential starting materials as they are inexpensive and
have been successfully functionalized in the past into
(1) For reviews, see: (a) Cole, D. C. Tetrahedron 1994, 50, 9517. (b)
Cardillo, G.; Tomasini, C. Chem. Soc. Rev. 1996, 23, 117. (c) Juaristi,
E., Ed.; Enantioselective Synthesis of â-Amino Acids; Wiley-VCH: New
York, 1997. Juaristi, E.; Lopez-Ruiz, H. Curr. Med. Chem. 1999, 6,
983-1004. (d) Liu, M.; Sibi, M. P. Tetrahedron 2002, 58, 7991.
(2) (a) Wagner, R.; Tilley, J. J. Org. Chem. 1990, 55, 6289-6291.
(b) W. Kabawata, N.; Inamoto, Y.; Sakane, K.; Iwamoto, T.; Hashimoto,
S. J. Antibiot. 1992, 45, 513. (c) Shimamoto, K.; Shigeri, Y.; yasuda-
Kamatani, Y.; Lebrun, B.; Yumoto, N.; Nakajima, T. Bioorg. Med.
Chem. Lett. 2000, 10, 2407-2410.
(7) (a) Dexter, C. S.; Jackson, R. F. W. J. Org. Chem. 1999, 64, 7579-
7585. (b) Cundy, D. J.; Donohue, A. C.; McCarthy, T. D. J. Chem. Soc.,
Perkin Trans. 1 1999, 559-567. (c) Hunter, C.; Jackson, R. F. W.; Rami,
H. K. J. Chem. Soc., Perkin Trans. 1 2000, 219-223.
(8) Griesbeck, A. G.; Heckroth, H. Synlett 1997, 1243-1244.
(9) 1H NMR (CDCl3) for 4: 3.62 (dt, J 19.4, 3.8, 1H), 3.86 (dt, J 3.8,
19.4, 1H), 5.01 (m, 1H), 7.17 (m, 1H), 7.29 (m, 1H), 7.47 (d, J 8.1, NH),
7.59 (m, 1H).
(10) (a) Davies, S. G.; Ichihara, O.; Lenoir, I.; Walters, I. A. S. J.
Chem. Soc., Perkin Trans. 1 1994, 1411-1415. (b) Cohen, J. H.; Abdel-
Magid, A. F.; Almond, H. R.; Maryanoff, C. A. Tetrahedron Lett. 2002,
43, 1977-1981.
(3) Gellman, S. H. Acc. Chem. Res. 1998, 31, 173-180. North, M. J.
Peptide Sci. 2000, 6, 301-313.
(4) Palomo, C.; Arrieta, A.; Coss´ıo, F. P.; Aizpurua, J. M.; Mielgo,
A.; Aurrekoetxea, N. Tetrahedron Lett. 1990, 31, 6429-6432.
(5) See ref 1d and references therein for most recent applications.
(6) (a) Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M. Eur. J.
Org. Chem. 1999, 3223, 3-3235. (b) Dudding, T.; Hafez, A. M.; Taggi,
A. E.; Wagerle, T. R.; Lectka, T. Org. Lett. 2002, 4, 387-390. (c) Hafez,
A. M.; Dudding, T.; Wagerle, T. R.; Shah, M. H.; Taggi, A. E.; Lectka,
T. J. Org. Chem. 2003, 68, 5819-5825.
(11) Horstmann, T. E.; Guerin, D. J.; Miller, S. J. Angew. Chem.,
Int. Ed. 2000, 39, 3635-3638.
(12) 1H NMR (CDCl3) for 9: 3.29 (dd, J 1.4, 7.1, 2H), 3.74 (s, 3H),
6.38 (dt, J 7.1, 16, 1H), 6.59 (m, 1H), 6-86-7.04 (m, 2H), 7.15 (m,
1H).
(13) Nelson, S. G.; Spencer, K. L. Angew. Chem., Int. Ed. 2000, 39,
1323-1325.
10.1021/jo048249c CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/02/2005
J. Org. Chem. 2005, 70, 1949-1952
1949

tions14c,f,g,19 using 0.4 mol % of [(S)-BINAP]RuCl2. It is
worth noting that the rate of the reaction is considerably
increased by the addition of a catalytic amount of strong
acid (HCl) as shown previously by King.20 Under these
conditions, (S)-â-hydroxy methyl ester 13 was obtained
in quantitative yield with 91% ee.
SCHEME 1. Friedel-Crafts Acylation
Alternatively, (S)-â-hydroxy acid 14 could be obtained
by enantioselective reduction21 of â-ketoacid precursor 17
using (+)-DIP-Cl (diisopinocampheyl chloroborane) in
95% yield and 94% ee, in the presence of triethylamine
(Scheme 4). However, due to the use of a stoichiometric
amount of chiral reducer, this alternative route was not
a viable process.
After solvent exchange from methanol to THF, the
crude solution of 13 from hydrogenation was saponified
with LiOH at room temperature and the resulting
lithium carboxylate salt of 14 was extracted into water.
Following partial reacidification to pH ) 5, 14 was
allowed to react with O-benzyl hydroxylamine using EDC
as a coupling agent. Chiral O-benzyloxy amide 15 directly
crystallized out from the aqueous solution after stirring
at room temperature for a couple of hours in an excellent
overall yield of 93% for the last four steps, starting from
Meldrum derivative 11. This sequence was run on a 4 kg
scale without any isolation of intermediates (Scheme 4).
SCHEME 2. Wittig-Horner Reaction
Ring closure of 15 to â-lactam 16 via Mitsunobu
reaction was found to be more problematic. On a small
scale and using typical conditions (THF, PPh3/DIAD 1
equiv), the reaction proceeded very smoothly and gave
the expected â-lactam in 93% isolated yield with complete
inversion of configuration. Unfortunately, we were unable
to drive the reaction to completion on a larger scale (>100
g). Increasing of the amount of PPh3/DIAD or using a
more electron-rich phosphine, P(t-Bu)3, led to the forma-
tion of a large amount of elimination byproduct (styrene
derivative). Fortunately, by conducting the reaction in
toluene instead of THF, we were able to reach complete
conversion and minimize the elimination side reaction
to less than 3% yield. Furthermore, H2-DIAD and triph-
enylphosphine oxide produced during the reaction can
be easily removed by simple filtration of the toluene
solution, facilitating the â-lactam isolation. Then, solvent
exchange to MeOH and cooling to -20 °C led to the
crystallization of the chiral â-lactam 16 in 78% yield and
with >99.5% ee22 (Scheme 4).
SCHEME 3. Retrosynthetic Scheme
Precursor â-ketoester 12 was easily made using the
following sequence (Scheme 4): acylation17 of Meldrum’s
acid in the presence of 2,4,6-collidine18 with 2,5-difluo-
rophenylacetyl chloride, which was generated in situ from
the corresponding acid 10 and oxalyl chloride. Meldrum
intermediate 11 was simply isolated by crystallization
after acidification of the crude basic aqueous extracts
(NaOH) in 83% yield. Methanolysis of 11 gave the crude
â-ketoester 12 after decarboxylation, and this solution
was directly submitted to Noyori hydrogenation condi-
Finally, we were delighted to see that when treated
with MeONa23 in methanol, opening of â-lactam 16 and
subsequent catalytic hydrogenation24 (Pd/C) gave the
corresponding â-aminoester 18; saponification with LiOH
with in situ BOC trapping then revealed the desired
optically pure (R)-â-amino acid 1 in 85% overall yield
(Scheme 5).
(14) (a) Miller, M. J.; Mattingly, P. G.; Morrison, M. A.; Kerwin, J.
F. J. Am. Chem. Soc. 1980, 102, 7026-7032. (b) Miller, M. J.; Bajwa,
J. S.; Mattingly, P. G.; Peterson, K. J. Org. Chem. 1982, 47, 4928-
4933. (c) Guzzo, P. R.; Miller, M. J. J. Org. Chem. 1994, 59, 4862-
4867. (d) Durham, T. B.; Miller, M. J. Org. Lett. 2002, 4, 135-138. (e)
Bulychev. A.; Bellettini, J. R., O’Brien, M.; Crocker, P. J.; Samama,
J.-P.; Miller, M. J.; Mobashery, S. Tetrahedron 2000, 56, 5719-5728.
(f) Durham, T. B.; Miller, M. J. J. Org. Chem. 2003, 68, 27-34. (g)
Durham, T. B.; Miller, M. J. J. Org. Chem. 2003, 68, 35-42. See also:
Yang, H. W.; Romo, D. J. Org. Chem. 1999, 64, 7657-7660.
(15) Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M. Synlett
2001, 12, 1813-1826.
(19) Noyori, R.; Ohkuma, T. Angew. Chem., Int. Ed. 2001, 40, 40-
73 and references therein.
(20) King, S. A.; Thompson, A. S.; King, A. O.; Verhoeven, T. R. J.
Org. Chem. 1992, 55, 6689-6691.
(21) Wang, Z.; Zhao, C.; Pierce, M. E.; Fortunak, J. M. Tetrahe-
dron: Asymmetry 1999, 10, 225-228.
(22) The ee is increased from 92 to >99.5% during crystallization
in MeOH.
(16) Known examples required two- to four-step deprotection/
protection for opening. See ref 14g.
(23) Similarly, treatment of â-lactam 16 with LiOH in THF/water
give the corresponding â-amino acid.
(17) Capozzi, G.; Roelens, S.; Talami, S. J. Org. Chem. 1993, 58,
7932-7936.
(24) When the hydrogenation is performed on the â-lactam 16, the
NO-Bn bond is cleaved instead of the N-OBn bond, giving the
N-hydroxyl â-lactam.
(18) Use of triethylamine or pyridine led to the formation of bis-
acylated byproduct up to 15%.
1950 J. Org. Chem., Vol. 70, No. 5, 2005

SCHEME 4. Synthesis of â-Lactam Intermediate 16
89-91 °C. 1H NMR (CDCl3): 1.77 (s, 6H), 4.49 (s, 2H), 6.95-
7.07 (m, 3H). 13C NMR: 26.9, 35.1, 91.9, 105.4, 115.8 dd, 116.4
dd, 118.0 dd, 122.6 dd, 156.8 d, 159.2 d, 160.3, 170.5, 193.0. IR:
1204, 1425, 1498, 1579, 1736, 2848, 2922. Anal. Calcd for
C14H12F2O5: C, 56.38; H, 4.06; F, 12.74. Found: C, 56.13; H,
3.66; F, 13.04.
SCHEME 5. Opening of â-Lactam
Methyl 4-(2,5-Difluorophenyl)-3-oxobutanoate 12. A so-
lution of Meldrum’s derivative 11 (4.3 kg) was refluxed in
anhydrous methanol (20 L) for 2.5 h. CO2 gas evolved during
the reflux period. A pure sample was obtained through MTBE/
hexane crystallization. Mp: 38 °C. 1H NMR (CDCl3): 3.50 (s,
2H), 3.69 (s, 3H), 3.83 (s, 2H), 6.86-6.94 (m, 2H), 6.94-7.00 (m,
1H). Enol form: 3.48 (s), 3.67 (s), 12.04 (s). 13C NMR: 42.7, 48.3,
52.3, 115.5 dd, 116.3 dd, 118.1 dd, 122.3 dd, 172.8, 198.3. IR:
1014, 1497, 1716, 2956. Anal. Calcd for C11H10F2O3: C, 57.90;
H, 4.42; F, 16.65. Found: C, 57.69; H, 4.11; F, 16.56.
In conclusion, we have successfully developed a scaled
process of chiral amino acid 1. This process can be run
economically on a kilogram scale. The sequence involves
a four-step “through process” where isolation procedures
are kept to a minimum.
N-Benzyloxy-4-(2,5-difluorophenyl)-3(S)-hydroxybutan-
amide 15. The preceding solution was divided into two batches
(approximately 12 L each) and degassed for 15 min with
bubbling N2, and then 70 mL of 2 N HCl followed by (S)-BINAP-
RuCl2 (22.3 g, 0.4 mol %) were added to each batch, which was
then submitted to hydrogenation (H2, 150 psi) at 60 °C for 4-5
h. Both batches were combined for the next step. Chiral HPLC
indicated 91% ee for â-hydroxy ethyl ester 13. 1H NMR
(CDCl3): 2.46 (dd, J 16.7, 9, 1H), 2.56 (dd, J 16., 3.2, 1H), 2.75
(br, OH), 3.71 (s, 3H), 4.29 (m, 1H), 6.90 (m, 1H), 6.96-7.02 (m,
2H).
The preceding methanol solution of â-hydroxyester was
solvent switched to THF (17 L final volume) using 35 L of THF
with a minimal residual volume of 5 L (20-25 °C, 28 mmHg).
Then, a solution of LiOH‚H2O (2370 g, 56.48 mol) in water (12
L) was added at 20-25 °C. The solution was stirred for 30 min,
MTBE (8 L) was added, and the layers were separated. The
aqueous layer was extracted with MTBE (2 × 8 L), and then
the combined organic layers were back-extracted with water (8
L). The combined aqueous layers containing â-hydroxy acid 14
were used directly for the next step. 1H NMR (CDCl3): 2.51 (dd,
J 16.7, 8.7, 1H), 2.59 (dd, J 16.7, 3.5, 1H) 2.85 (m, 2H), 4.31 (m,
1H), 6.88 (m, 1H), 6.90-7.02 (m, 2H).
Experimental Section
5-[1-Hydroxy-2-(2,5-difluorophenyl)ethylidene]-2,2-di-
methyl-1,3-dioxane-4,6-dione 11. Oxalyl chloride (2.544 kg,
1.719 L, 20.049 mol) was slowly added to a slurry of 2,5-
difluorophenylacetic acid 10 (3 kg, 17.429 mol, MW 172.13) and
DMF (9 mL) in dichloromethane (11 L) at 25-30 °C over 1.5 h.
CO, CO2, and HCl gas evolved through the addition. The mixture
was stirred for 2 h at rt until the end of gas evolution. CH2Cl2
and excess (COCl)2 were distilled off (20-25 °C, 28 mmHg). CH2-
Cl2 (2 L) was added to the residue, and the resulting solution
was then slowly added to a mixture of Meldrum’s acid (2.637
kg, 18.30 mol) and collidine (4.224 kg, 4.6 L, 34.857 mol) in
dichloromethane (18 L) at -5 °C over 2 h. The resulting pale
yellow solution was stirred for 1 h at 0 °C, and then10.4 L of 6
M HCl was charged at 0 °C over 5 mn. The layers were
separated, and the organic layer was extracted with 69 L of 1
M NaOH and then with 34 L of 1 M NaOH. The combined basic
aqueous layers were acidified with 37% HCl (approximately 8.5
L) until pH ) 1. The solid was filtered off and washed with water
(30-50 L) to remove the yellow color and then dried over N2
stream for 5 days to yield 4.3 kg (83%) of white solid 11. Mp:
J. Org. Chem, Vol. 70, No. 5, 2005 1951

To the preceding aqueous solution was added HCl (3.5 L, 42
mol, 37% aq) and then O-benzylhydroxyamine hydrochloride
(2705 g, 16.94 mol) followed by slow addition of EDC‚HCl (4060
g, 21.18 mol) over 15 min at 20-30 °C. The resulting mixture
was stirred for 1 h at 20-22 °C while the hydroxamate
precipitated. The light pink solid was filtered off and washed
with water (3 × 12 L) and with heptane (3 × 12 L) and then
dried over N2 stream for 3 days (yield: 4.27 kg, 93% last four
steps). Chiral HPLC indicated 92% ee. [R]589 ) +12.96 (c ) 1
CHCl3). Mp: 70-72 °C. 1H NMR major rotamer (273 K,
CDCl3): 2.15 (dd, J 9.2, 15.1, 1H), 2.26 (dd, J 2.5, 15.1, 1H),
2.75 (m, 2H), 3.70 (br, OH), 4.21 (m, 1H), 4.87 (dd, J 11.6, 15.3,
2H), 6.87-7.01 (m, 3H), 7.35 (m, 5H), 8.9 (br, NH). IR: 1495,
1653, 2847, 3214. Anal. Calcd for C17H17F2NO3: C, 63.54; H,
5.33; F, 16.65; N, 4.36. Found: C, 63.59; H, 5.04; F, 12.01; N,
4.32.
N-Benzyloxy-4(R)-[1-methyl-(2.5-difluorobenzyl)]-2-
oxoazetidine 16. Triphenylphosphine (3638 g, 13.87 mol) was
slowly added to a solution of diisopropyl azadicarboxylate (2804
g, 2730 mL, 13.87 mol) in toluene (43 L) at such a rate that the
temperature did not rise above 25 °C over 30 min. Then,
â-hydroxamate 15 (4051 g, 12.61 mol) was added portionwise
at 20-30 °C over 30 mn. During the addition of hydroxamate,
a precipitate appeared (mixture of triphenylphosphine oxide and
reduced DIAD) and the color turned black. After 2 h, the solids
were filtered off and washed with 5 L of toluene, and the
combined filtrates were solvent switched to methanol (final
volume 20 L) using 35 L of methanol (20-25 °C, 28 mmHg).
The solution was cooled to -30 °C for 30 min while the â-lactam
crystallized out. The slurry was then filtered, and the white
crystals were washed with cold methanol/water 9:1 (4 × 8 L,
-20 C). Yield: 3 kg, 78%. Chiral HPLC indicated 99.7% ee. [R]589
) +9.24 (c ) 1, CHCl3). Mp: 80 °C. 1H NMR (CDCl3): 2.38 (dd,
J 13.8, 2.4, 1H), 2.64-2.72 (m, 2H), 2.92 (ddd, J 14.2, 5.1, 0.9,
1H), 3.68 (m, 1H), 4.92 (d, J 14.8, 1H), 4.94 (d, J 14.8, 1H), 6.81
(m, 1H), 6.88-7.01 (m, 2H), 7.38 (m, 5H). 13C NMR: 31.8, 37.9,
57.2, 78.3, 115.1 dd, 116.5 dd, 117.6 dd, 125.2 dd, 128.7, 129.1,
129.3, 135.4, 156.7 d, 159.1 d, 163.9. IR: 728, 1043, 1496, 1774.
Anal. Calcd for C17H15F2NO2: C, 67.32; H, 4.98; F, 12.53; N, 4.62.
Found: C, 67.22; H, 4.80; F, 12.66; N, 4.68.
N-BOC-4-(2,5-difluorophenyl)-3(R)-aminobutanoic Acid
1. To a solution of â-lactam 16 (0.52 g, 1.72 mmol) in MeOH (10
mL) was added MeONa (0.5 M in MeOH, 5.2 mL) at rt. After 10
min, 10% Pd/C (0.5 g) was added, and the mixture was submitted
to hydrogenation (rt, H2 40 psi) for 1 h. Catalyst was filtered off
and the solvent evaporated to give crude free amine 18 which
was used directly due to instability. 1H NMR (CDCl3): 1.45 (br,
2NH), 2.35 (dd, J 15.9, 8.8, 1H), 2.51 (dd, J 15.9, 4.1, 1H), 2.68
(dd, J 13.5, 7.6, 1H), 2.78 (dd, J 13.5, 5.5, 1H), 3.50 (m, 1H),
3.70 (s, 3H), 6.88-6.96 (m, 2H), 6.97-7.03 (m, 1H).
Crude amine 18 was dissolved in THF (5 mL), and triethy-
lamine (0.36 mL, 2.6 mmol) was added followed by BOC2O (0.45
g, 2.0 mmol). After 10 min at rt, LiOH‚H2O (0.14 g, 3.5 mmol)
was added and the resulting mixture was stirred overnight at
rt. The solution was extracted twice with MTBE (2 × 20 mL),
and after phase separation, the aqueous layer was reacidified
with satd NaH2PO4 and back-extracted with MTBE (20 mL) to
yield after crystallization (MTBE/hexane) 0.46 g (85% overall
yield from 16) of 1. [R]589 ) +27.0 (c ) 0.38 CHCl3). Mp: 103-
105 °C. 1H NMR (CDCl3) mixture of rotamers: 1.38 (s, 9H),
2.55-2.67 (m, 2H), 2.93 (m, 2H), 4.2 (m, 1H), 5.11 (br, NH),
6.88-7.02 (m, 3H), 10.2 (br, OH). 13C NMR: 28.3, 33.6, 37.9,
17.7, 114.8, dd, 116.3 dd, 117.9 dd, 126.5 dd, 125.2, 156.7 d, 159.2
d, 176.3. IR: 813.5, 1053.6, 1165, 1496, 1522, 1686, 2912, 2939,
3359. HRMS: calcd for C15H19F2NO4 315.1282, found 315.1281.
Acknowledgment. We gratefully thank Mr. R.
Reamer for the NMR expertise and Dr. F. Wang for
analytical support.
JO048249C
1952 J. Org. Chem., Vol. 70, No. 5, 2005