Synthesis of orthogonally protected l-threo-β-ethoxyasparagine

Article abstract of DOI:10.1007/s00726-009-0389-6

Orthogonally protected l-threo-β-ethoxyasparagine (Fmoc-EtOAsn(Trt)- OH, 1) was synthesized from diethyl (2S,3S)-2-azido-3-hydroxysuccinate 2 in eight steps as a building block for solid-phase peptide synthesis. The starting material is easily available in multi-gram scale from d-diethyltartrate. The transformation steps reported here are robust and scalable. Thus, a significant amount of 1 (1.8 g) was obtained in 21% overall yield. The synthesis reported is also expected to be useful for the preparation of other O-substituted l-threo-β-hydroxyasparagine derivatives.

Full text of DOI:10.1007/s00726-009-0389-6

Amino Acids (2010) 39:161–165
DOI 10.1007/s00726-009-0389-6
ORIGINAL ARTICLE
Synthesis of orthogonally protected L-threo-b-ethoxyasparagine
•
•
•
Jan Spengler Marta Pelay Judit Tulla-Puche
Fernando Albericio
Received: 31 October 2009 / Accepted: 3 November 2009 / Published online: 17 November 2009
Ó Springer-Verlag 2009
Abstract Orthogonally protected L-threo-b-ethoxyaspar-
agine (Fmoc-EtOAsn(Trt)-OH, 1) was synthesized from
diethyl (2S,3S)-2-azido-3-hydroxysuccinate 2 in eight steps
as a building block for solid-phase peptide synthesis. The
starting material is easily available in multi-gram scale from
D-diethyltartrate. The transformation steps reported here are
robust and scalable. Thus, a significant amount of 1 (1.8 g)
was obtained in 21% overall yield. The synthesis reported is
also expected to be useful for the preparation of other
O-substituted L-threo-b-hydroxyasparagine derivatives.
Abbreviations
ACN
DIEA
DMF
EDC
Acetonitrile
N,N-Diisopropylethylamine
N,N-Dimethylformamide
N-(3-Dimethylaminopropyl)-N’-
ethylcarbodiimide
EtOAc
Ethyl acetate
Fmoc-OSu N-(9-Fluorenylmethoxycarbonyloxy)
succinimide
HOAc
TLC
Acetic acid
Thin-layer chromatography
Keywords Hexafluoroacetone Á Tritylamide Á
2,4,6-Trimethoxybenzylamide Á Amino acids Á Peptide
Introduction
L-Threo-b-ethoxyasparagine is structurally closely related
to L-threo-b-hydroxyasparagine, a non-proteinogenic amino
acid present in various antimicrobial peptides. Efficient
synthetic routes to derivatives bearing common protecting
schemes for solid-phase peptide synthesis (SPPS) have
J. Spengler (&) Á M. Pelay Á J. Tulla-Puche Á F. Albericio (&)
Institute for Research in Biomedicine,
Barcelona Science Park, Baldiri Reixac 10,
08028 Barcelona, Spain
e-mail: jan.spengler@irbbarcelona.org
´
been reported (Boger et al. 2000; Guzman-Martinez and
VanNieuwenhze 2007). However, no information is
available for L-threo-b-ethoxyasparagine or its diastereo-
mers, despite this compound being a non-proteinogenic
amino acid constituent of at least one bioactive peptide
(structure not shown). We required a considerable amount
of Fmoc-EtOAsn(Trt)-OH 1 to produce this peptide by
Fmoc-SPPS. The direct introduction of the additional ethyl
ether function in Fmoc-b-OH-Asn(Trt)-OH or in earlier
stages of its synthesis calls for additional, orthogonal and
temporary carboxylic- or amide protection and this would
cause an drop in the yield of the reported multistep syn-
F. Albericio
e-mail: fernando.albericio@irbbarcelona.org;
albericio@irbbarcelona.org
J. Spengler Á M. Pelay Á J. Tulla-Puche Á F. Albericio
CIBER-BBN, Networking Centre on Bioengineering
Biomaterials and Nanomedicine,
Barcelona Science Park,
University of Barcelona,
Baldiri Reixac 10, 08028 Barcelona, Spain
F. Albericio
Department of Organic Chemistry,
University of Barcelona,
´
thesis (Guzman-Martinez and VanNieuwenhze 2007: eight
´
Martı i Franques 1-11,
08028 Barcelona, Spain
´
steps, 29% overall yield from L-threo-b-hydroxyaspartic
123

162
J. Spengler et al.
Compound synthesis and characterization
acid; Boger et al. 2000: nine steps, 11% overall yield from
methyl 4-methoxycinnamate). To circumvent this obstacle,
we developed an alternative route starting from diethyl
(2S,3S)-2-azido-3-hydroxysuccinate 2, which is available
in multi-gram scale from D-diethyltartrate in two steps and
represents a carboxy-protected L-threo-b-hydroxyaspartate
scaffold with a masked amino group (Charvillon and
Amouroux 1997; Saito et al. 1996). Therefore, we con-
sidered this compound perfectly suited for the introduction
of substituents at the b-hydroxy function. Although 2 was
obtained with ca. 20% of the undesired (2R,3S)-diastereo-
mer, this contamination was completely removed in later
stages (in course of the purification of compound 7) of the
synthesis.
Diethyl (2S,3S)-2-azido-3-ethoxysuccinate (3)
(as diastereomeric mixture)
A solution of 2 (10.02 g, 43.34 mmol) and iodoethane
(35.1 g, 18 mL, 225 mmol) in diethyl ether (200 mL) was
placed in a round bottomed flask equipped with a condenser
and stirred with a mixture of silver(I)oxide (13 g,
56.1 mmol) and sand (50 g) at 37°C for 15 h (the sand was
added to prevent that the silver(I)oxide sticks together).
When TLC analysis revealed complete conversion, the
solution was filtered and concentrated in vacuo. Product 3
was obtained after column chromatography (hexane/
EtOAc), 2:1, Rf = 0.5, KMnO4: yellow) as a yellow oil (all
fractions containing product, 10.0 g, 89%). The product was
a mixture of diastereomers and some side products were also
present. It was used for the next step without further purifi-
Materials and methods
1
cation. H NMR (CDCl3): d = 4.43 (d, J = 3.6 Hz, 1H),
4.39–4.17 (m, 14H), 3.91–3.79 (m, ca. 2H), 3.66 (m, ca. 1H),
General remarks
3.59–3.39 (m, ca. 2H), 1.16–1.40 (m, ca. 24H).
Diethyl (2S,3S)-2-azido-3-hydroxysuccinate 2 was pre-
pared as mixture of diastereomers from D-diethyltartrate,
following reported procedures (Charvillon and Amouroux
1997; Saito et al. 1996). All commercial reagents and
solvents were used as received unless otherwise indicated.
Solvents were from SDS (Peypin, France). TLC was per-
formed with TLC aluminum sheets coated with silica gel
60 F254 (Merck). Spots were observed with UV or after
staining and heating as indicated [15 g phosphomolybdic
acid in ethanol (100 mL), or with 5 g ninhydrin in acetone
(100 mL), or 6 g KMnO4, 40 g K2CO3 and 5 mL NaOH in
600 mL water]. For column chromatography, Silice 60
ACC 35–70 lm (SDS) was used.
Diethyl (2S,3S)-2-amino-3-ethoxysuccinate (4)
A solution of 3 (as diastereomeric mixture) (10.0 g,
38.6 mmol) in EtOAc (150 mL) was stirred with Pd–C
catalyst (5% Pd, 0.5 g) under an atmosphere of H2 (air
pressure) for 15 h. When TLC (hexane/EtOAc, 2:1,
Rf = 0.5, KMnO4: starting material yellow) revealed
complete conversion, it was filtered and concentrated in
vacuo. The diastereomers were partially separated by col-
umn chromatography (dichloromethane (DCM)/methanol
(MeOH)/HOAc, 10:1:0.5, Rf = 0.3 (product), Rf = 0.26
(other diastereoisomer), ninhydrin: red brown). The frac-
tions with product were sampled and concentrated in
vacuo. HOAc was removed by co-evaporation with toluene
to give 7.85 g, 87% of 4 as a yellow oil with ca. 10% of the
(2R,3S)-diastereomer (NMR). 1H NMR (CDCl3): d = 4.33
(d, J = 3.0 Hz, 1H), 4.27–4.18 (m, 4H), 3.89 (d,
J = 3.0 Hz, 1H), 3.76 (m, 1H), 3.40 (m, 1H), 1.29 (m, 6H),
1.17 (m, 3H) ppm; 13C NMR (CDCl3): d = 170.3, 79.1,
67.0, 61.5, 61.3, 56.8, 14.8, 14.1, 14.1 ppm; [a]2D5 -33
(c 2.0, CHCl3, diastereomeric mixture); IR (film): 2,980,
1,751, 1,221, 1,164, 1,104 cm-1; HRMS (ES?) calc. for
C10H19NO5 m/z (M ? Na)? 256.1161, found 256.1154.
Diethyl (2R,3S)-2-amino-3-ethoxysuccinate: 1H NMR
(CDCl3): d = 4.14–4.28 (5H), 3.98 (d, J = 3.4 Hz, 1H),
3.80 (m, 1H), 3.49 (m, 1H), 1.32–1.19 (9H).
1H-NMR spectra were acquired at 400.125 MHz, 13C at
100.625 MHz with TMS as internal reference, and 19F at
376.494 MHz) with a Varian Mercury 400 (High-field
NMR Unit, Barcelona Science Park). The following
abbreviations are used to indicate multiplicity: s, singlet; d,
doublet; dd, double doublet; t, triplet; dt, double triplet; m,
multiplet, br s, broad signal. Chemical shifts (d) are
expressed in parts per million downfield from tetrameth-
ylsilyl chloride. IR spectra were recorded with a Thermo
Nicolet spectrometer and with the Omnic 6.0 program from
the Thermo Nicolet Corporation. Optical rotation values
were measured with a Perkin-Elmer 241 spectrometer
(concentrations are expressed in g 100 mL-1). Melting
points were determined in a Bu¨chi Melting Point B-540
apparatus and are uncorrected. High-resolution mass
spectroscopy (HRMS) was performed with the electrospray
(ion spray) (ESI–MS) method with a HRMS LC/MSD-TOF
(2006) (Agilent Technologies) in positive or negative mode
(as indicated).
(2S,3S)-2-amino-3-ethoxysuccinic acid (5)
Diethyl (2S,3S)-2-amino-3-ethoxysuccinate 4 (7.85 g,
33.65 mmol) was refluxed with stirring with 5 N aqueous
123

Orthogonally protected L-threo-b-ethoxyasparagine
163
HCl (100 mL) for 5 h. When TLC (DCM/MeOH/HOAc,
10:1:0.5, Rf = 0.3 (starting material), ninhydrin: red
brown) revealed complete conversion, the solution was
concentrated at 70°C. The residue was dissolved in THF
(30 mL) and propylene oxide (10 mL) was added drop
wise. The supernatant was removed by suction and the
white precipitate was dried under vacuo to give 5.01 g
oxalyl chloride was removed in vacuo and the residue
was co-evaporated with dry toluene (20 mL). The acid
chloride was directly converted without further purifica-
tion [1H NMR (CDCl3): d = 4.52 (d, J = 2.0 Hz, 1H),
4.42 (m, 2H), 3.88 (m, 1H), 3.59 (m, 1H), 3.39 (br. d,
J = 7.5 Hz, 1H), 1.22 (t, J = 6.7 Hz, 3H) ppm; 19F
NMR (CDCl3): d = -79.07 (q, J = 8.4 Hz, 3F), -80.78
(q, J = 8.5 Hz, 3F) ppm; diastereomer 2: -79.87
(q, J = 8.5 Hz, approx. 0.3F), -80.53 (q, J = 8.5 Hz,
approx. 0.3F) ppm.]. The acid chloride of 6 dissolved in
dry toluene (10 mL) was dropped into a solution of tri-
tylamine (3.82 g, 14.72 mmol) in dry toluene (30 mL) at
0°C. After 1 h of stirring at 0°C, the suspension was fil-
tered and the precipitate was washed with toluene
(4 9 10 mL). The unified solutions were concentrated
under vacuo and the residue was purified by column
chromatography (hexane/EtOAc, 2:1, Rf = 0.55, UV
254 nm: blue). The undesired diastereomer was com-
pletely separated and eluted in a mixture with tritylamine
(Rf = 0.4, UV 254 nm: blue). The fractions with product
were pooled and concentrated in vacuo, then dissolved in
ACN/H2O and lyophilized to give 2.36 g (59%) of dia-
stereomerically pure product 7. Mp 49–50°C; 1H NMR
(CDCl3): d = 8.08 (br. 1H), 7.22–7.33 (18H), 4.48 (dd,
J = 1.8, 7.8 Hz, 1H), 4.23 (d, J = 2.1 Hz, 1H), 3.78 (m,
2H), 2.76 (d, J = 7.9 Hz, 1H), 1.26 (t, J = 7.0 Hz, 3H)
ppm; 19F NMR (CDCl3): d = -80.16 (q, J = 8.6 Hz,
3F), -80.79 (q, J = 8.5 Hz, 3F) ppm; 13C NMR (CDCl3):
d = 169.2, 168.0, 144.0, 128.4, 128.1, 127.3, 121.0 (q,
J = 287.4 Hz), 120.0 (q, J = 284.6 Hz), 89.5 (m), 78.7,
70.4, 69.4, 57.9, 15.2 ppm; IR (KBr): 3,396, 1,836, 1,689,
1,493, 1,239, 1,198, 974, 702 cm-1; [a]D25 -32 (c 1.6,
CHCl3); HRMS (ES ?) calc. for C28H24F6N2O4 m/z
(M ? H)? 567.1719, found 567.1712.
1
(84%) of 5. Decomp. at 210°C. The H and 13C NMR
spectra showed approx. 10% contamination with the
1
(2R,3S)-diastereomer. H NMR (DMSO-d6): d = 3.92 (d,
J = 9.2 Hz, 1H), 3.79 (m, 2H), 3.72 (d, J = 9.2 Hz, 1H),
3.42 (m, 1H), 1.13 (m, 3H) ppm; 13C NMR (DMSO-d6):
d = 170.9, 168.4, 75.2, 66.4, 53.4, 14.9 ppm; (diastereo-
mer 2R,3S: d = 170.1, 167.9, 79.7, 65.5, 53.8, 14.7 ppm);
IR (KBr): 3,431, 3,112, 1,695, 1,466, 1,138, 1,100 cm-1;
[a]2D5 -56 (c 1.5, DMSO, diastereomeric mixture); HRMS
(ES-) calc. for C6H11NO5 m/z (M-H)- 176.0559, found
176.0562.
(2S)-2-Ethoxy-2-[(4S)-5-oxo-2,2-bis-trifluoromethyl-
oxazolidin-4-yl]-acetic acid (6)
A solution of 5 (3.73 g, 21.05 mmol) in DMF (20 mL)
was stirred under an atmosphere of hexafluoroacetone
(anhydrous) for 22 h (uptake of 8.8 g). DMF was
removed at 1 mm Hg at 55°C and the residue was purified
by column chromatography (hexane/EtOAc/HOAc,
1:1:0.05, Rf = 0.3, phosphomolybdic acid: blue). The
fractions with product were pooled and concentrated in
vacuo. HOAc was removed by co-evaporation with toluene
1
to give 5.65 g (83%) of yellow oil. The H and 13C NMR
spectra showed ca. 10% contamination with the other
diastereomer. 1H NMR (acetone-d6): d = 5.09 (d, J =
6.8 Hz, 1H), 4.66 (dd, J = 2.4, 7.0 Hz, 1H), 4.32 (d,
J = 2.6 Hz, 1H), 3.85 (m, 1H), 3.52 (m, 1H), 1.15
(t, J = 7.0 Hz, 3H) ppm; 19F NMR (acetone-d6): d =
-80.14 (q, J = 9.0 Hz, 3F), -80.88 (q, J = 9.2 Hz, 3F)
ppm; [other diastereomer: -80.31 (m, approx. 0.3F),
-80.84 (m, approx. 0.3F) ppm]; 13C NMR (CDCl3):
d = 174.3, 169.2, 120.99 (q, J = 287.5 Hz), 119.95 (q,
J = 285.7 Hz), 89.4 (m), 75.2, 63.3, 58.1, 14.6 ppm
(additional signals of the other diastereomer: 173.6,
57.5 ppm); [a]2D5 -49 (c 2.5, CHCl3, diastereomeric mix-
ture); IR (film): 3,345, 1,831, 1,734, 1,234 cm-1. HRMS
(ES-) calc. for C9H9F6NO5 m/z (M-H)- 325.0307, found
325.0309.
(2S,3S)-2-Amino-3-ethoxy-N-trityl-succinamic
acid methyl ester hydrochloride (8)
A solution of dry HCl in dioxane (4 M, 60 mL) was added
to a solution of 7 (2.98 g, 4.72 mmol) in MeOH (60 mL) at
0°C. This mixture was stirred (closed flask) for 15 h. When
HPLC revealed complete conversion, the volatiles were
removed in vacuo and the residue was treated with diethyl
ether (30 mL). The solid precipitate was dried in vacuo to
give 2.1 g of 8 as hygroscopic solid (95%). 1H NMR
(DMSO-d6): d = 8.99 (s, 1H), 7.18–7.33 (18H), 4.49 (d,
J = 5.9 Hz, 1H), 4.36 (d, J = 5.9 Hz, 1H), 3.70 (s, 3H),
3.51 (m, 2H), 1.08 (t, J = 7.0 Hz, 3H) ppm; 13C NMR
(DMSO-d6): d = 171.9, 167.9, 144.2, 128.3, 127.6, 126.7,
78.4, 69.5, 66.3, 54.3, 52.4, 14.9 ppm; IR (KBr): 3,396,
1,751, 1,685, 1,491, 1,097, 700 cm-1; [a]D25 -11 (c 1.8,
DMSO); HRMS (ES?) calc. for C26H28N2O4 m/z
(M ? H)? 433.2127, found 433.2119.
(2S)-2-Ethoxy-2-[(4S)-5-oxo-2,2-bis-trifluoromethyl-
oxazolidin-4-yl]-N-trityl-acetamide (7)
Oxazolidinone 6 (2.28 g, 7.01 mmol) was dissolved in
oxalyl chloride (10 mL) and five drops of DMF were
added. This mixture was stirred for 30 min. The excess of
123

164
J. Spengler et al.
(2S,3S)-2-(9H-Fluoren-9-yl)methoxycarbonylamino-3-
ethoxy-N-trityl-succinamic acid (1)
diastereomeric mixture. After Pd–C-catalyzed hydrogena-
tion of the azido function to diethyl 2-amino-3-ethox-
ysuccinate 4, the ethyl ester was hydrolyzed in boiling
aqueous HCl and the 2-amino-3-ethoxysuccinic acid 5 was
precipitated from propylene oxide/THF.
A solution of KOH (1.22 g, 21.85 mmol) in H2O (20 mL)
was added to a solution of 8 (1.89 g, 4.37 mmol) in dioxane
(40 mL) at 0°C. After 1 h of stirring, TLC revealed com-
plete conversion. Glacial HOAc (1.25 mL) was added and
the solution was cooled to 0°C. After addition of Fmoc-OSu
(2.21 g, 6.55 mmol), DIEA was dropped into the milky
suspension until pH 9 was reached (ca. 2 mL). After 1 h of
stirring, complete conversion was observed by TLC. The
dioxane was removed under vacuo at 30°C, and 1 N
aqueous HCl (100 mL) was added. The product was
extracted with ethyl acetate (3 9 50 mL). The organic layer
was dried over MgSO4, filtered and concentrated in vacuo.
The residue was purified by column chromatography
(hexane/EtOAc, 3:2) until the excess of Fmoc-OSu had
eluted, and the product was then eluted with hexane/
EtOAc/HOAc, 1:1:0.05, Rf = 0.2, UV 254 nm: blue). The
fractions with product were pooled and concentrated in
vacuo. HOAc was removed by co-evaporation with toluene
to give 1.98 g (70%) of 1 as a foamy white solid. Mp 113–
The differentiation between the two carboxyl functions
of 5 was accomplished by the use of the bidentate pro-
tecting/activating reagent hexafluoroacetone (HFA). HFA
undergoes a cyclocondensation with the a-amino carbox-
ylic acid unit to give 2,2-bis(trifluoromethyl)-1,3-oxazoli-
din-4-ones. This reaction proceeds site selectively in the
presence of further carboxylic acid residues in the molecule
(Spengler et al. 2006). Thus, 5 was reacted with HFA in
DMF to give the oxazolidinone 6 in 83% yield (Scheme 1).
For the transformation of the unreacted carboxylic acid
of 6 into a suitable protected amide group, we reasoned that
it would be much easier to obtain the 2,4,6-trimethox-
ybenzyl-protected amide than the trityl-protected amide,
because 2,4,6-trimethoxybenzylamine is less sterically
hindered and more nucleophilic than tritylamine. The
2,4,6-trimethoxybenzyl-group is reported to provide effi-
cient amide protection orthogonally to the Fmoc-group and
can be removed with acids (Weygand et al. 1968). Indeed,
the amide formation was found to be possible by activation
of the terminal carboxylic group of 6 with carbodiimide
(EDC) and subsequent reaction with 2,4,6-trimethoxyben-
zylamine. Amide 9 was isolated in 46% yield (Scheme 2).
These conditions failed completely when tritylamine
was tried to react. Although 2,2-bis(trifluoromethyl)-1,
1
116°C; H NMR (DMSO-d6): d = 8.40 (s, 1H), 7.99 (d,
J = 9.3 Hz, 1H), 7.86–7.91 (3H), 7.81 (d, J = 7.5 Hz,
1H), 7.37–7.44 (2H), 7.33–7.14 (20H), 4.39 (m, 2H), 4.34
(d, J = 2.9 Hz, 1H), 4.19 (m, 1H), 4.07 (m, 1H), 3.72 (m,
1H), 3.57 (m, 1H), 1.17 (t, J = 7.0 Hz, 3H) ppm; 13C NMR
(DMSO-d6): d = 171.3, 167.9, 156.3, 144.3, 143.7, 143.5,
140.6, 140.8, 128.4, 127.6, 127.6, 127.5, 127.0, 126.9,
126.5, 125.8, 125.4, 120.0, 119.9, 80.1, 69.3, 67.1, 66.2,
55.9, 46.5, 15.0 (some signals are doubled) ppm; IR (KBr):
3,397, 1,726, 1,696, 1,492, 700 cm-1; [a]D25 13 (c 1.75,
DMSO); HPLC (70–100% ACN, 8 min): tr 4.97 min
(98%); HRMS (ES?) calc. for C40H36N2O6 m/z (M ? H)?
641.2652, found 641.2649.
EtO2C
N3
CO2Et
OH
EtO2C
N3
CO2Et
OEt
a
D-diethyltartrate
89 %
3
2
c
HO2C
H2N
CO2H
OEt
b
EtO2C
H2N
CO2Et
OEt
87 %
84 %
5
4
Results and discussion
O
N
O
O
NHTrt
e
CO2H
OEt
O
Highly efficient procedures for the formation of ethyl
ethers from hydroxy carboxylic acids by base-induced OH-
deprotonation and reaction with iodoethane have been
described (Aikins et al. 2005). However, the risk of epi-
merization under basic conditions must be taken into
consideration because the starting material diethyl (2S,3S)-
2-azido-3-hydroxysuccinate 2 contains significant amounts
of the (2R,3S)-diastereomer and epimerization at the (3S)-
carbon of the diastereomeric contamination would produce
the (2R,3R)-enantiomer of our target molecule. Therefore,
we used the mild, non-basic Ag2O/iodoethane protocol,
which is reported to be useful for the formation of ethyl
ethers of hydroxy carboxylic acids (Parmenon et al. 2008)
and gives 89% of diethyl 2-azido-3-ethoxysuccinate 3 as a
d
O
F3C
F3C
F3C
59 %
OEt
N
F3C
H
83 %
H
7
6
O
O
HO2C
FmocHN
NHTrt
OEt
f
g, h
70 %
MeO2C
H2N
NHTrt
95 %
OEt
1
8
a. Ag2O/EtI, Et2O, reflux. b. H2, Pd-C, ethyl acetate, r.t. c1. aqu. HCl, reflux,
c2. propylene oxide, r.t.. d. hexafluoroacetone, DMF, r.t. e1. oxalyl chloride,
r.t., e2. tritylamine, toluene, 0 ºC. f. MeOH/dry HCl, r.t. g. KOH, dioxane/H2O,
r.t. h. Fmoc-OSu, DIEA, dioxane/H2O, r.t.
Scheme 1 Synthesis of Fmoc-EtOAsn(Trt)-OH 1 from D-diethyl-
tartrate
123

Orthogonally protected L-threo-b-ethoxyasparagine
165
Fmoc-EtOAsn(Trt)-OH 1 in eight steps, which is amenable
for SPPS. The starting material is easily available in multi-
gram scale from D-diethyltartrate. Contaminations with the
undesired (2R,3S)-diastereomer which was obtained toge-
ther with the starting material could be removed in step 5.
The transformation steps reported here allow the synthesis of
the target compound in 21% overall yield. This strategy can
be used for the synthesis of other non-natural O-substituted
L-threo-b-hydroxyasparagine analogs.
Acknowledgments We are deeply indebted to Professor Klaus
Burger, who inspired the HFA chemistry, for his continuous support.
This study was partially funded by PharmaMar S.A.I., CICYT
(CTQ2006-03794), the Generalitat de Catalunya (2009SGR 1024),
the Barcelona Science Park and the Institute for Research in
Biomedicine.
Scheme 2 The Tmob-protected amide reacts as nucleophile with the
oxazolidinone
3-oxazolidin-4-ones, like 9, represent carboxy-activated
species, their direct use in SPPS is limited (Albericio et al.
2005). Therefore, we envisioned the exchange of the HFA-
protecting group for N-Fmoc protection. However, mild
hydrolysis (room temperature) of the lactone 9 produced
only traces of the desired 1-carboxy-a-amino acid, but
mainly products such as 10, originating from an intramo-
lecular cyclization reaction between the (lactone-) acti-
vated carboxylic acid and the amide-group (HPLC–MS).
This observation led us to question the suitability of 2,4,6-
trimethoxybenzyl-(Tmob)-protection for the amide func-
tion in our case because the carboxylic group has to be
activated for coupling on solid-phase reactions and the
observed intramolecular cyclization would then compete
with the reaction with the resin-bound peptide.
References
Aikins JA, Haurez M, Rizzo JR, Van Hoeck J-P, Brione W,
Kestemont J-P, Stevens C, Lemair X, Stephenson GA, Marlot
E, Forst M, Houpis IN (2005) Synthesis of a peroxime
proliferator activated receptor (PPAR) a/c agonist via stereo-
controlled Williamson ether synthesis and stereospecific SN2
reaction of S-2-chloro propionic acid with phenoxides. J Org
Chem 70:4695–4705
Albericio F, Burger K, Cupido T, Ruiz J, Spengler J (2005)
Application of hexafluoroacetone as protecting and activating
reagent in solid phase peptide and depsipeptide synthesis.
Arkivoc 2005(vi):191–199
Boger DL, Lee RJ, Bounaud P-Y, Meier P (2000) Asymmetric
synthesis of orthogonally protected L-threo-b-hydroxyaspara-
gine. J Org Chem 65:6770–6772
Charvillon FB, Amouroux R (1997) Synthesis of 3-hydroxylated
analogs of D-aspartic acid b-hydroxamate. Synthetic Commun
27:395–403
We reasoned that the steric demanding trityl group could
prevent undesired intramolecular cyclizations. Several
conditions were tested to generate the trityl-protected amide
7 (Scheme 1). Finally, the best results gave the activation of
6 as acid chloride, and subsequent reaction with 2.1 equiv.
of tritylamine. A substitution of the excess of tritylamine for
pyridine caused a significant drop in yield. The tritylamide 7
could be completely separated from the remaining amounts
of the (2R,3S)-diastereomer by chromatography on silica
and 7 was obtained diastereomerically pure in 59% yield.
Also, parts of the tritylamine used as excess were recovered.
The hydrolysis of the oxazolidinone 7 proceeded very
slowly. Therefore, the methyl ester hydrochloride 8 was
prepared as intermediate by stirring in methanol containing
dry HCl. After hydrolysis of the methyl ester and amino
group protection with Fmoc-OSu under standard condi-
tions, 1 was obtained in 21% yield over eight steps.
´
Guzman-Martinez A, VanNieuwenhze MS (2007) An operationally
simple and efficient synthesis of orthogonally protected L-threo-
b-hydroxyasparagine. Synlett 1513–1516
Parmenon C, Guillard J, Caignard D-H, Hennuyer N, Staels B,
Audinot-Bouchez V, Boutin J-A, Dacquet C, Ktorza A,
Viaud-Massuard M-C (2008) 4, 4-Dimethyl-1, 2, 3, 4-tetrahy-
droquinoline-based PPAR a/c agonists. Part I: Synthesis and
pharmacological evaluation. Bioorg Med Chem Lett 18:1617–
1622
Saito S, Komada K, Moriwake T (1996) Diethyl (2S, 3R)-2-(N-tert-
butoxycarbonyl)amino-3-hydroxysuccinate. Org Synth 73:184–
200
Spengler J, Boettcher C, Albericio F, Burger K (2006) Hexafluoro-
acetone as protecting and activating reagent: new routes to
amino, hydroxy, and mercapto acids and their application for
peptide and glyco- and depsipeptide modification. Chem Rev
106:4728–4746
Weygand F, Steglich W, Bjarnason J (1968) Easily cleavable
protective groups for acid amide groups. III. Derivatives of
asparagine and glutamine with 2, 4-dimethoxybenzyl- and 2, 4,
6-trimethoxybenzyl-protected amide groups. Chem Ber
101:3642–3648
Conclusions
Here, we describe a robust and scalable strategy for the
preparation of diastereomerically pure (2S,3S-configuration)
123