Enantioselective synthesis of (2S)-2-(4-phosphonophenylmethyl)-3-aminopropanoic acid suitably protected for peptide synthesis

Article abstract of DOI:10.1016/S0040-4039(02)00004-7

Protected D-β-2-(4-phosphono) phenylalanine was prepared by a multistep synthesis, involving the diastereoselective alkylation of a chiral enolate. This new compound can be further incorporated into peptidic backbone by means of solid-phase peptide synthe

Full text of DOI:10.1016/S0040-4039(02)00004-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 1417–1419
Enantioselective synthesis of
(2S)-2-(4-phosphonophenylmethyl)-3-aminopropanoic acid suitably
protected for peptide synthesis
Wang-Qing Liu, Catherine Olszowy, Laurent Bischoff and Christiane Garbay*
Laboratoire de Pharmacochimie Mole´culaire et Structurale INSERM et CNRS, Faculte´ des Sciences Pharmaceutiques et
Biologiques, 4, avenue de l’Observatoire, 75270 Paris Cedex 06, France
Received 10 December 2001; accepted 21 December 2001
Abstract—Protected
D
-b-2-(4-phosphono) phenylalanine was prepared by a multistep synthesis, involving the diastereoselective
alkylation of a chiral enolate. This new compound can be further incorporated into peptidic backbone by means of solid-phase
peptide synthesis in order to synthesize short peptides adopting b-turn or a-helix conformations. © 2002 Elsevier Science Ltd. All
rights reserved.
Oligomers of b-amino acids have emerged recently as a
new class of peptidomimetics. These b-peptides are able
to fold into helices, b-turns or b-sheet structures,1 thus
receiving increasing attention in view of their better
resistance to in vivo degradation, and modified phar-
macological properties.
phonophenyl group is not structurally a true mimetic of
the O-phosphorylated tyrosine, this moiety was chosen
for two reasons. On the one hand, it has a higher
acidity8,9 than the phosphonomethyl group, the pKa of
the aryl phosphonic acid being comparable to that of
the naturally-occurring phosphate. On the other hand,
as far as the interactions with the phosphonic acid are
concerned, due to the higher flexibility exhibited by the
b-2 amino acid, we can expect that a b-2 4-phosphono-
phenylalanine would interact in a similar way to a
phosphotyrosine. Several mimetics of phosphotyrosine
already exist, some of them being commercially avail-
able. However, no compound corresponding to the
required b-2 analogue being described, we decided to
prepare it.
The asymmetric synthesis of b-amino acids substituted
only on the carbon adjacent to the acid group is much
less documented in the literature than that of a-amino
acids. However, interesting syntheses were developed
recently. Seebach2 and Juaristi3 have published a
methodology in which a chiral perhydropyrimidone is
alkylated at the a-carbon. Lavielle’s group4,5 also
recently showed that alkylation of Oppolzer’s sultam
could be extended to homoglycine enolates. More clas-
sical but versatile and efficient is the synthesis described
by Wyatt et al.6 giving several examples of 2-substituted
chiral 3-aminopropanoic acids. Using their method, we
focused on the preparation of the title compound 6c, in
which the stereochemistry of the chiral center is con-
trolled by means of diastereoselective alkylation of an
enolate bearing the Evans7 oxazolidinone chiral auxil-
iary. In the course of our structure–activity relationship
studies, molecular modeling showed that we needed a
b-2 amino acid analogous to a phosphorylated tyrosine,
having the unnatural (2S) configuration and being
resistant to phosphatases. Indeed, although the 4-phos-
To obtain an efficient multigram scale synthesis for the
precursor 2, we elected to make use of the classical
malonate synthesis. As depicted in Scheme 1, the prepa-
ration of 3-(4-diethylphosphonobenzyl)-propanoic acid
started with the alkylation of di-tert-butyl malonate by
4-bromobenzyl bromide. Di-tert-butyl malonate (5 g)
was deprotonated by sodium hydride or KOt-Bu and
the resulting anion alkylated with 4-bromobenzyl bro-
mide, leading to 1a. The aromatic bromide was subse-
quently displaced by a diethylphosphoryl group, the
precursor of the phosphonic acid, a non-hydrolyzable
phosphate analogue. This reaction was achieved by
reaction of diethyl phosphite, catalyzed by Pd(0), lead-
ing to di-terbutyl (4-diethylphosphono-benzyl)malonate
(1b). Subsequent TFA cleavage of the tert-butyl ester
groups, followed by concomittant decarboxylation by
* Corresponding author. Fax: (+33) 1 53 73 98 75; e-mail: garbay@
pharmacie.univ-paris5.fr
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00004-7

1418
W.-Q. Liu et al. / Tetrahedron Letters 43 (2002) 1417–1419
Scheme 1. (i) CH2(COOt-Bu)2, NaH or KOt-Bu, THF; (ii) HPO(OEt)2, Pd(PPh3)4; (iii) TFA, CH2Cl2 then toluene, reflux; (iv)
i-BuOCOCl, N-methylmorpholine, THF then (4R,5S)-4-methyl-5-phenyl-2-oxazolidinone, lithiated with n-BuLi; (v) NaHMDS,
THF −78°C then BrCH2COOt-Bu, −78°C, 2 h; (vi) TFA, CH2Cl2; (vii) Ph2PON3, Et3N, toluene rt then 1 h reflux; (viii) LiOH,
H2O2, THF–water then aqueous Na2SO3; (ix) H2, Pd/C, EtOH; (x) Fmoc-Cl, NaHCO3, dioxane–water.
refluxing for several hours in toluene, led to the
expected 3-(4-diethylphosphono)-propanoic acid (2).
The N-acyl oxazolidinone 5b underwent a smooth
saponification with lithium hydroperoxide in THF–
water, giving the free acid which was easily purified by
flash chromatography, eluting with EtOAc–cyclohexane
in the presence of 1% formic acid. The acid 6a and
recovered chiral oxazolidinone were obtained in 75 and
94% yields, respectively. Replacement of the Cbz pro-
tecting group by Fmoc, suitable for solid-phase peptide
synthesis, was achieved by hydrogenolysis under atmo-
spheric pressure and subsequent reprotection with
Fmoc-Cl in dioxane–water affording 0.97 mmol of the
final amino acid 6c. Yields: 1a 63%, 1b 81%, 2 84%, 3
53%, 4 63%, 5a quant, 5b 67%, 6a 75%, 6b quant, 6c
67%. The optical rotatory power for 6c was measured
in chloroform and methanol, giving two different val-
ues: [h]D (20°C)=−8 (c=0.5, CHCl3); [h]D (20°C)=−3
(c=0.5, MeOH).
Having this precursor in hand, we performed the N-
acylation of the lithiated Evans chiral oxazolidinone
(4R,5S)-4-methyl-5-phenyl-1,3-oxazolidin-2-one
with
the ‘mixed anhydride’ from our substituted propanoic
acid 2, via activation with isobutyl chloroformate. The
use of the ‘mixed anhydride’ was preferred to the
chloride, the acylation with the chloride giving very low
yields. The resulting compound 3 bearing the chiral
auxiliary required for diastereoselective alkylation was
obtained with a 53% yield only, presumably due to a
partial decomposition of the mixed anhydride. Taking
into account the preliminary observations of Wyatt and
coworkers,6 we incorporated an acid moiety as the
precursor of the b-amino group. This was achieved
through diastereoselective alkylation of the sodium eno-
late (NaHMDS) of 3 with tert-butyl bromoacetate at
−78°C, giving 4 with a diastereomeric excess of 80%,
the minor diastereomer being readily separated by
means of flash chromatography. The absolute configu-
ration of the chiral center was postulated by analogy
with the numerous examples of alkylation of such
chiral enolates. Deprotection of the acid (TFA–CH2Cl2)
was readily achieved but a slight racemization of 5a was
observed by 1H NMR, the undesired diastereomer
being separated by flash chromatography in the follow-
ing step. The straight transformation of the acid into a
Cbz-protected amino group was obtained by a variant
of the Curtius rearrangement. First, treatment at room
temperature with diphenylphosphoryl azide afforded
the acyl azide, the latter being subsequently decom-
posed by refluxing in toluene, giving the isocyanide
which was trapped in situ by benzyl alcohol in the
presence of triethylamine.
In conclusion, ( )-4-phosphono-b-2-phenylalanine suit-
D
ably protected for further incorporation into peptides,
was prepared in ten steps from inexpensive materials.
Using the widely used Evans oxazolidinone, the synthe-
sis proved to be efficient, providing this novel com-
pound with a good enantioselectivity. It is worth noting
that, as both enantiomers of cis-4-methyl-5-phenyl-1,3-
oxazolidin-2-one are commercially available, either of
the two enantiomers of the b-amino acid can be pre-
pared following this synthetic procedure. This com-
pound is
a
good candidate for a-helices and
b-turn-containing short peptides, bearing a 4-phos-
phono group to mimic the phosphorylated tyrosine
residue.
Structural data of compounds 1a–6a: 1a lH (DMSO)
1.28 (18H, s), 2.87 (2H, d, J=8.2 Hz), 3.52 (1H, t,
J=8.2 Hz), 7.16 (2H, d, J=7.8 Hz), 7.42 (2H, d, J=8.2

W.-Q. Liu et al. / Tetrahedron Letters 43 (2002) 1417–1419
1419
Hz); 1b lH (DMSO) 1.16 (6H, t, J=7.2 Hz), 1.26 (18H,
s), 2.99 (2H, d, J=8.0 Hz), 3.57 (1H, t, J=8.0 Hz),
3.86–4.01 (4H, m), 7.36 (2H, dd, J=7.8, 4.1 Hz), 7.42
(2H, dd, J=13.1, 8.2 Hz); 2 lH (DMSO) 1.17 (6H, t,
J=7.2 Hz), 2.53 (2H, t, J=7.5 Hz), 2.84 (2H, t, J=7.5
Hz), 3.86–4.03 (4H, m), 7.36 (2H, dd, J=7.8, 4.1 Hz),
7.57 (2H, dd, J=13.1, 7.8 Hz); 3 lH (CDCl3) 0.82 (3H,
d, J=6.6 Hz), 1.26 (6H, t, J=7.2 Hz), 2.99 (2H, t,
J=7.4 Hz), 3.12-3.27 (2H, m), 3.95–4.12 (4H, m), 4.68
(1H, quint, J=6.7 Hz), 5.58 (1H, d, J=7.2 Hz), 7.20–
7.38 (7H, m), 7.67 (2H, dd, J=13.1, 8.0 Hz)); 4 lH
(CDCl3) 0.81 (3H, d, J=6.6 Hz), 1.16–1.25 (6H, m),
1.31 (9H, s), 2.26 (1H, dd, J=16.9, 4.3 Hz), 2.65 (1H,
dd, J=12.1, 10.5 Hz), 2.74 (1H, dd, J=16.9, 10.5 Hz),
3.00 (1H, dd, J=12.1, 5.8 Hz), 3.96–4.08 (4H, m),
4.42–4.48 (1H, m), 4.54–4.62 (1H, m), 5.30 (1H, d,
J=7.2 Hz), 7.19–7.36 (7H, m), 7.69 (2H, dd, J=13.1,
8.0 Hz); 5a lH (CDCl3) 0.76 (3H, d, J=6.6 Hz),
1.17–1.25 (6H, m), 2.35 (1H, dd, J=17.5, 4.5 Hz), 2.64
(1H, dd, J=17.4, 9 Hz), 2.84 (1H, dd, J=17.4, 10.4
Hz), 3.03 (1H, dd, J=17.5, 7.1 Hz), 3.64–4.09 (4H, m),
4.42–4.61 (1H, m), 4.58 (1H, quint, J=7.2 Hz), 5.35
(1H, d, J=7.2 Hz), 7.18–7.70 (9H, m); 5b lH (CDCl3)
0.76 (3H, d, J=6.6 Hz), 1.17–1.26 (6H, m), 2.76–3.10
(2H, m), 3.48 (2H, t, J=6.8 Hz), 3.95–4.08 (4H, m),
4.26–4.32 (1H, m), 4.49 (1H, quint, J=7.0 Hz), 5.02
(2H, s), 5.28 (1H, d, J=7.2 Hz), 7.18–7.36 (7H, m) 7.66
(2H, dd, J=13.2, 7.8 Hz); 6a lH (CDCl3) 1.19 (6H, t,
J=7.1 Hz), 2.85–3.49 (5H, m), 3.92–4.14 (4H, m), 5.02
(2H, s), 5.30 (1H, br s), 7.20–7.34 (7H, m), 7.62 (2H,
dd, J=13.2, 7.9 Hz); 6b lH (D2O) 1.29 (6H, t, J=7.3
Hz), 3.03–3.29 (5H, m), 4.07–4.18 (4H, m), 7.46 (2H,
dd, J=7.8, 4.1 Hz), 7.46 (2H, dd, J=13.2, 7.8 Hz), lC
(D2O) 43.3, 58.3, 61.6, 65.0, 80, 125.6 (d, J=214 Hz),
129.8 (d, J=18 Hz), 131.5 (d, J=12 Hz), 140.2, 164.7;
6c lH (CDCl3) 1.22 (3H, t, J=6.8 Hz), 1.23 (3H, t,
J=6.8 Hz), 2.60-3.48 (5H, m), 3.88–4.07 (4H, m), 4.14
(1H, t, J=6.9 Hz), 4.30 (2H, d, J=6.9 Hz), 5.34 (1H,
br t), 7.20–7.70 (12H, m).
Acknowledgements
We are grateful to INSERM for providing a grant to C.
Olszowy and the Ligue Nationale Contre le Cancer for
financial support to the group.
References
1. Seebach, D.; Ciceri, P. E.; Overhand, M.; Jaun, B.; Rigo,
D. Helv. Chim. Acta 1996, 79, 2043–2066.
2. Seebach, D.; Boog, A.; Schweizer, W. B. Eur. J. Org.
Chem. 1999, 335–360.
3. Juaristi, E.; Quintana, D.; Balderas, M.; Garcia-Perez, E.
Tetrahedron: Asymmetry 1996, 7, 2233–2246.
4. Ponsinet, R.; Chassaing, G.; Lavielle, S. Tetrahedron:
Asymmetry 1998, 9, 865–871.
5. Ponsinet, R.; Chassaing, G.; Vaissermann, J.; Lavielle, S.
Eur. J. Org. Chem. 2000, 83–90.
6. Arvanitis, E.; Ernst, H.; Ludwig, A. A.; Robinson, A. J.;
Wyatt, P. B. J. Chem. Soc., Perkin Trans. 1 1998, 521–528.
7. Evans, D. A.; Ennis, M. D.; Mathre, D. J. J. Am. Chem.
Soc. 1982, 104, 1737–1739.
8. Domchek, S. M.; Auger, K. R.; Chetterjee, S.; Burke, T.
R., Jr.; Shoelson, S. E. Biochemistry 1992, 31, 9865–9870.
9. Liu, W.-Q.; Roques, B. P.; Garbay, C. Tetrahedron Lett.
1997, 38, 1389–1392.