A practical route to long-chain non-natural α,ω-diamino acids

Article abstract of DOI:10.1007/s00726-012-1349-0

An efficient method for the synthesis of long-chain α,ω-diamino acids, starting from natural α-amino acids, has been developed. The long-chain skeleton has been generated through condensation between a protected aldehyde, derived from l-aspartic acid, and an ylide obtained from an ω-hydroxy-alkyl phosphonium salt. After conversion of the ω-hydroxy group into an amine, catalytic hydrogenation produced the N,N'-protected α,ω-diamino acid. The present route to α,ω-diamino acids allows the modulation of the chain length depending on the length of the ylide used for the Wittig olefination reaction.

Full text of DOI:10.1007/s00726-012-1349-0

Amino Acids (2013) 44:435–441
DOI 10.1007/s00726-012-1349-0
ORIGINAL ARTICLE
A practical route to long-chain non-natural a,x-diamino acids
•
Giuseppina Brasile Laura Mauri
•
•
•
Sandro Sonnino Federica Compostella
Fiamma Ronchetti
Received: 10 May 2012 / Accepted: 21 June 2012 / Published online: 10 July 2012
Ó Springer-Verlag 2012
Abstract An efficient method for the synthesis of long-
chain a,x-diamino acids, starting from natural a-amino
acids, has been developed. The long-chain skeleton has
been generated through condensation between a protected
aldehyde, derived from L-aspartic acid, and an ylide
obtained from an x-hydroxy-alkyl phosphonium salt. After
conversion of the x-hydroxy group into an amine, catalytic
hydrogenation produced the N,N0-protected a,x-diamino
acid. The present route to a,x-diamino acids allows the
modulation of the chain length depending on the length of
the ylide used for the Wittig olefination reaction.
Introduction
Lipidic a-amino acids (LAAs) are of great importance for
the synthesis of drug delivery systems and bioactive lipid
mimetics (Toth 1994; Constantinou-Kokotou and Kokotos
1999). They have been conjugated to a wide variety of
different compounds in order to improve the absorption of
drugs. Due to the structural feature of the long alkyl chain,
amides or esters of saturated LAAs possess a high degree
of membrane-like character, which may facilitate their
crossing into the cell interior. Furthermore, the long alkyl
side chains may have the additional effect of protecting a
labile parent drug from enzymatic attack. LAAs have also
received considerable attention for the development of self-
adjuvanting peptide vaccines. They have been conjugated
to peptide epitopes to change their physico-chemical and
pharmacological characteristics in order to increase the
delivery of potential agents for molecular therapy (Sar-
pietro et al. 2008; Li et al. 2012). Some LAAs exhibit a
further amino group at the end of the long alkyl chain and
the effect of their long chain in increasing drug efficiency
has been widely reported. x-Amino LAA derivatives have
been used in the inhibition of metalloproteinases by
N-carboxylalkyl peptides, in which a phenethyl group had
been replaced with a long-chain x-aminoalkyl group (Esser
et al. 1995). A more selective and long-lasting response,
induced by the presence of an x-amino LAA, has been also
observed for the angiotensin converting enzyme (ACE)
inhibitor enalaprilat and some derivatives (Shirota et al.
1990; Saito et al. 1990), as well as in the case of ben-
zothiazepine- and benzoxazepine-related compounds,
where the duration of the ACE-inhibiting action is depen-
dent on the length of the chain (Itoh et al. 1986). More
recently, non-natural x-amino LAAs have been incorpo-
rated into antimicrobial peptides (AMP) with the same
Keywords Lipidic a-amino acids (LAAs) ꢀ Unnatural
fatty acids ꢀ Wittig olefination ꢀ a,x-Diamino acids
Abbreviations
TEA
Triethylamine
LiHMDS Lithium hexamethyldisilazide
TsCl
p-Toluenesulfonyl chloride
4-Dimethylamino pyridine
N,N-Diisopropylethylamine
Tetrabutylammonium iodide
Triphenylphosphine
DMAP
DIPEA
TBAI
PPh3
Boc2O
Di-tert-butyl dicarbonate
G. Brasile ꢀ F. Compostella (&) ꢀ F. Ronchetti
Dipartimento di Biotecnologie Mediche e Medicina
`
Traslazionale, Universita di Milano, Via Saldini 50,
20133 Milano, Italy
e-mail: federica.compostella@unimi.it
L. Mauri ꢀ S. Sonnino
Dipartimento di Biotecnologie Mediche e Medicina
`
Traslazionale, Universita di Milano, Via Fratelli Cervi 93,
20190 Segrate (Milano), Italy
123

436
G. Brasile et al.
rationale to impart specific physico-chemical properties to
the peptide and to develop compounds able to interact with
membranes in novel ways. The antibacterial activity of
these novel AMP peptides is affected by the length and
charge distribution of LAA, and this could have important
implications in the selectivity towards different bacterial
strains. In fact, differences in the chemical composition of
the membrane of bacteria reflect a different degree of
complementarity between the membrane and the interact-
ing AMP, which results in different activities (Russell et al.
2012). x-Amino LAAs have also been used for the con-
struction of capture agents for the detection of proteins, to
be used in alternative to expensive antibodies (Agnew et al.
2009). On the other side, they can also be considered as
valuable fatty acid analogues for the preparation of
sphingolipids, modified in the sphingoid backbone.
Sphingolipids are major components of eukaryotic plasma
membranes, with ceramide, the sphingoid backbone,
comprising a long-chain base attached to the fatty acid via
an amide bond. The amino groups appended to the fatty
acid alkyl chain could be used as derivatization points or
anchors for the attachment of probes, e.g. photoactivable
groups for the study of membrane microenvironments
(Aureli et al. 2010).
NHBOC
( )15
HOOC
NHBOC
1
NHBOC
CHO
( )10
Ph3P
Br
OH
BnOOC
2
3
Fig. 1 Chemical structure of the a,x-diamino-octadecanoic acid 1
and its precursors
rotary evaporation below 45 °C at approximately
20 mmHg. All reactions carried out under anhydrous
conditions were performed using oven-dried glassware
within an argon atmosphere. Dry solvents and liquid
reagents were distilled prior to use. Tetrahydrofuran (THF)
and toluene were distilled from sodium; dichlorometh-
ane (DCM) and pyridine (Py) were distilled from cal-
cium hydride; dimethylformamide (DMF), acetonitrile
(CH3CN), and methanol (MeOH) were dried on activated
˚
4 A molecular sieves; TEA was distilled from KOH and
DIPEA from calcium hydride. Optical rotations were
1
measured with a Perkin-Elmer 241 polarimeter. H NMR
A variety of methods for the asymmetric synthesis of
a-amino acids have been developed (for general reviews see
for example: Constantinou-Kokotou and Kokotos 1999;
and 13C NMR spectra were recorded at 298 K with a
Bruker FT-NMR AVANCETM DRX500 spectrometer
using a 5-mm z-PFG (pulsed field gradient) broadband
reverse probe. Chemical shifts are reported on the d (ppm)
scale and are relative to TMS as internal reference. The
signals were unambiguously assigned by 2D COSY and
HSQC experiments (standard Bruker pulse program).
Scalar coupling constants are reported in Hertz. The mass
spectrometric analyses were performed in positive or
negative electrospray mode (ESI–MS). MS spectra were
recorded on a Thermo Quest Finnigan LCQTMdeca ion trap
mass spectrometer; the mass spectrometer was equipped
with a Finnigan ESI interface. Data were processed by
Finnigan Xcalibur software system. All reactions were
monitored by TLC on silica gel 60 F-254 plates (Merck),
spots being developed with a phosphomolybdate-based
reagent, anisaldehyde or ninidrine solutions and subse-
quently heated at 110 °C. Flash column chromatography
was performed on silica gel 60 (230–400 mesh, Merck).
TsCl was crystallized from hexane.
´
Ma 2003; Najera and Sansano 2007; Soloshonok and
Sorochinsky 2010). The synthesis of x-amino LAAs, i.e.
non-natural long-chain a,x-diamino acids, are commonly
based on methods developed for the preparation of orni-
thine, lysine or higher homologues, which generally consist
in the alkylation of acetamidomalonate derivatives (Gaudry
1953; Payne and Boger 1985). Despite the wide range of
potential applications, there are very few examples of the
preparation of x-amino LAAs longer more than 10 carbon
atoms (Markidis and Kokotos 2002; Agnew et al. 2009).
Herein, we report a practical and general method to prepare
a,x-diamino long fatty acids, starting from readily avail-
able chiral building blocks, i.e. derivatized natural L-amino
acids. We illustrate this procedure by the synthesis of the
a,x-diamino-octadecanoic acid 1, which is appropriately
protected for inclusion in more complex structures through
its carboxylic group (Fig. 1).
Benzyl (2S)-2-(tert-butoxycarbonylamino)-4-oxo-
butanoate (5)
Materials and methods
All reagents were bought at highest commercial quality and
used without further purification except where noted. Air-
and moisture-sensitive liquids and solutions were trans-
ferred via oven-dried syringe or stainless steel cannula
through septa. Organic solutions were concentrated by
Freshly distilled oxalyl chloride (1.49 mL, 17.63 mmol)
was slowly added to a solution under argon of DMF
(0.47 mL, 6.03 mmol) in DCM (15 mL) at 0 °C. A white
solid was formed, which gradually dissolved to give a
123

A practical route to a,x-diamino acids
437
white opalescent suspension. After the mixture was stirred
for 1 h, the solvent was evaporated under reduced pressure
maintaining an inert atmosphere, by connecting directly the
flask under stirring to the vacuum pump (medium vacuum)
and then drying the solid by flushing a stream of nitrogen.
To the residual white solid, dissolved under an argon
atmosphere in THF (15 mL) and CH3CN (9 mL), a solu-
tion of BOC-L-Asp-OBzl 4 (1.5 g, 4.64 mmol) and pyri-
dine (0.45 mL) in THF (15 mL) was slowly added at
-30 °C, and the reaction was stirred at the same temperature
for 1 h. Then, the solution was cooled to -78 °C and
LiAlH(OtBu)3 (1 M THF sol., 5.10 mL) was added drop-
wise. After 0.5 h, the reaction was quenched by the addi-
tion of 1 N HCl solution (30 mL) followed by extraction
with EtOAc (3 9 20 mL). The combined organic layers
were washed with saturated NaHCO3 solution, dried over
Na2SO4 and evaporated under reduced pressure. The crude
was purified by flash chromatography (hexane/EtOAc 7:3,
v/v, 1 % Et3N) to yield compound 5 (0.85 g, 60 %) as a
colourless oil. [a]D20 ?15.5 (c 1, CHCl3). 1H NMR (CDCl3):
ESI–MS (positive-ion mode, m/z): 356.0 (100) [M?Na]?,
387.9 (83) [M?OMe?Na]?. Anal. calcd for C18H23NO5
(333.16): C 64.85, H 6.95, N 4.20. Found: C 64.52, H 7.24,
N 4.01.
12-Hydroxydodecyltriphenylphosphonium bromide (3)
12-Bromododecanol (1.00 g, 3.77 mmol) was dissolved in
EtOH (10 mL) and PPh3 (1.28 g, 4.90 mmol) was added.
The reaction was refluxed for 48 h, and then the solvent
was removed under reduced pressure. Purification by flash
chromatography (DCM/MeOH, 10:1, v/v) gave phospho-
nium salt 3 (1.79 g, 90 %) as a colourless oil, which
became a white solid after treatment with dry THF fol-
lowed by evaporation. Physical data of compound 3 were
in agreement with those reported in Tran-Thi and Falk
(1995).
Benzyl (2S)-2-(tert-butoxycarbonylamino)-18-
hydroxy-octadeca-4,6-dienoate (6)
d
1.45 (s, 9H, –tBu), 3.04 (dd, 1H, J2,3a = 4.5,
J3a,3b = 18.3 Hz, H-3a); 3.13 (dd, 1H, J2,3b = 4.9,
J3a,3b = 18.3 Hz, H-3b), 4.60–4.70 (m, 1H, H-2); 5.20
(s, 2H, CH2Ph), 5.37–5.51 (m, 1H, –NH), 7.30–7.45 (m, 5H,
arom), 9.74 (s, 1H, H-4) ppm. 13C NMR (CDCl3): d 28.3
(3C, t-Bu), 46.0 (C-3), 48.8 (C-2), 67.6 (CH2Ph), 80.3
(CMe3), 128.3–128.6 (5C, arom), 135.1 (arom), 155.3
(CONH), 170.9 (C-1), 199.3 (C-4) ppm. ESI–MS (positive-
ion mode, m/z): 361.9 (87) [M?OMe?Na]?, 700.7 (100)
[2(M?OMe) ?Na]?. Anal. calcd for C16H21NO5 (307.14):
C 62.53, H 6.89, N 4.56. Found: C 62.27, H 6.97, N 4.23.
To a suspension of the phosphonium salt 3 (2.61 g,
4.95 mmol) in dry THF (30 mL) at -60 °C under argon,
LiHMDS (1 M THF sol., 9.9 ml) was added dropwise. The
orange ylide was stirred for 40 min, then a solution of
aldehyde 2 (0.55 g, 1.65 mmol) in dry THF (15 mL) was
added dropwise within 30 min. The reaction was stirred for
1.5 h at -60 °C, then quenched with HCl 1 N (30 mL) and
extracted with EtOAc (3 9 70 mL). The combined organic
layers were washed with saturated NaHCO3 solution, dried
over Na2SO4 and the solvent evaporated under reduced
pressure. Purification by flash chromatography (DCM/
EtOAc, 9:1, v/v, 1 % Et3N) gave compound 6 (0.54 g,
65 %), a pale yellow oil, as a mixture of isomers. 1H NMR
(CDCl3): d 1.24–1.52 (m, 25H, 8 CH2 and tBu), 1.54–1.62
(m, 2H, 2 H-17), 2.03–2.11 (m, 0.8H, 2 H-8MIN), 2.11–2.19
(m, 1.2H, 2 H-8MAG), 2.47–2.68 (m, 2H, 2 H-3), 3.66 (t, 2H,
J17,18 = 6.6 Hz, 2 H-18); 4.38–4.50 (m, 1H, H-2),
5.00–5.27 (m, 3H, CH2Ph and NH), 5.32–5.52 (m, 1.6H, H-4
and H-7MAG), 5.58–5.67 (m, 0.4H, H-7MIN), 5.87–6.08 (m,
1.4H, H-6 and H-5MIN), 6.33–6.42 (m, 0.6H, H-5MAG),
7.32–7.42 (m, 5H, arom) ppm. ESI–MS (positive-ion mode,
m/z): 524.2 [M?Na]?. Anal. calcd for C30H47NO5 (501.35):
C 71.82, H 9.44, N 2.79. Found: C 71.59, H 9.71, N 2.58.
Benzyl (2S,4E)-2-(tert-butoxycarbonylamino)-6-oxo-
hex-4-enoate (2)
(Triphenylphosphoranylidene)acetaldehyde (1.25 g, 4.10
mmol) was added to a solution under argon of compound 5
(0.84 g, 2.73 mmol) in dry toluene (27 mL) and the
resulting orange reaction mixture was warmed to 70 °C
and stirred at this temperature for 3.5 h. After removal of
the solvent under reduced pressure, the crude was purified
by flash chromatography (petroleum ether/EtOAc, 8:2, v/v,
1 % Et3N) to afford compound 2 (0.68 g, 75 %) as a pale
1
yellow oil. [a]2D0 ?10.0 (c 1, CHCl3). H NMR (CDCl3): d
1.45 (s, 9H, tBu), 2.66–2.77 (m, 1H, H-3a), 2.83–2.96 (m,
1H, H-3b), 4.52–4.63 (m, 1H, H-2); 5.10–5.30 (m, 3H,
CH2Ph and NH), 6.09 (dd, 1H, J4,5 = 15.6, J5,6 = 7.8
(2S)-2-(tert-butoxycarbonylamino)-18-hydroxy-
octadecanoic acid (7)
Hz, H-5); 6.67 (dd, 1H, J4,5 = 15.6, J3a,4 = J3b,4
=
7.3 Hz, H-4), 7.32–7.43 (m, 5H, arom), 9.40 (d, 1H,
J5,6 = 7.8 Hz, H-6) ppm. 13C NMR (CDCl3): d 28.2 (3C,
t-Bu), 35.9 (C-3), 52.4 (C-2), 67.6 (CH2Ph), 80.4 (CMe3),
128.6–128.8 (5C, arom), 134.9 (arom), 135.6 (C-5), 151.1
(C-4), 155.0 (CONH), 171.0 (C-1), 193.2 (C-4) ppm.
To a solution of compound 6 (0.04 g, 0.08 mmol) in
CHCl3/MeOH 1:1 (3 mL) a catalytic amount of 10 % Pd/C
was added and the reaction was stirred under H2 for 24 h.
The reaction was diluted with CHCl3/MeOH and filtered
through a MILLIPORE filter. After evaporation of the
123

438
G. Brasile et al.
solvent purification by flash chromatography (DCM/MeOH
95:5, v/v) gave compound 7 (0.029 g, 87 %) as a colour-
less oil. [a]2D0 ?5.9 (c 1, CHCl3:MeOH 1:1); 1H NMR
(CDCl3/MeOD 1:1, v/v): d 1.20–1.40 (m, 26H, 13 CH2),
1.44 (s, 9H, t-Bu), 1.48–1.56 (m, 2H, 2 H-17), 1.58–1.67
(m, 1H, H-3a), 1.73–1.86 (m, 1H, H-3b), 3.55 (t, 2H,
and tBu), 1.55–1.70 (m, 2H, 2 H-17), 2.02–2.11 (m, 0.8H,
2 H-8MIN), 2.11–2.20 (m, 1.2H, 2 H-8MAG), 2.48–2.68 (m,
2H, 2 H-3), 3.28 (t, 2H, J17,18 = 6.9 Hz, 2 H-18);
4.38–4.51 (m, 1H, H-2), 4.99–5.28 (m, 3H, CH2Ph and
NH), 5.31–5.53 (m, 1.6H, H-4 and H-7MAG), 5.57–5.68 (m,
0.4H, H-7MIN), 5.86–6.09 (m, 1.4H, H-6 and H-5MIN),
6.32–6.44 (m, 0.6H, H-5MAG), 7.31–7.45 (m, 5H, arom)
ppm. ESI–MS (positive-ion mode, m/z): 549.1 [M?Na]?.
Anal. calcd for C30H46N4O4 (526.35): C 68.41, H 8.80, N
10.64. Found: C 68.11, H 9.03, N 10.33.
J17,18 = 6.8 Hz, 2 H-18), 4.05–4.14 (m, 1H, H-2) ppm. 13
C
NMR (CDCl3/MeOD 1:1, v/v): d 25.4, 25.7, 27.9,
28.4–32.4 (12 C), 54.0 (C-2), 62.0 (C-18), 79.5 (CMe3),
156.3 (CONH), 176.2 (C-1) ppm. ESI–MS (negative-ion
mode, m/z): 414.2 [M-H]-. Anal. calcd for C23H45NO5
(415.33): C 66.47, H 10.91, N 3.37. Found: C 66.12, H
11.23, N 3.09.
Benzyl (2S)-2,18-di-(tert-butoxycarbonylamino)-
octadeca-4,6-dienoate (10)
Benzyl (2S)-2-(tert-butoxycarbonylamino)-18-(p-
toluensulfonyloxy)-octadeca-4,6-dienoate (8)
Compound 9 (0.27 g, 0.51 mmol) and PPh3 (0.32 g,
1.23 mmol) were dissolved in THF (10 mL) and, after 1 h,
water (1 mL) was added. The reaction was stirred over-
night, then diluted with water (15 mL) and extracted with
EtOAc (3 9 10 mL). The combined organic layers were
dried over Na2SO4 and evaporated under reduced pressure.
To a solution of the residue in dry DCM (10 mL), TEA
(0.28 mL, 2.04 mmol) was added, and, after 10 min, a
solution of BOC2O (0.28 g, 1.28 mmol) in DCM (2 ml).
The reaction was stirred at 40 °C for 6 h, then quenched
with MeOH, diluted with water and extracted with EtOAc
(3 9 10 mL). The combined organic layers were dried
over Na2SO4 and evaporated under reduced pressure. The
crude was purified by flash chromatography (hex-
ane:EtOAc, 85:15, v/v, 1 % Et3N) to yield compound 10
To a solution under argon of compound 6 (0.50 g,
1.0 mmol) in dry DCM (10 mL) DIPEA (0.70 mL,
2.0 mmol), TsCl (0.38 g, 2.0 mmol) and DMAP (0.12 g,
1.0 mmol) were added at 0 °C. The reaction was sponta-
neously warmed to room temperature and after 5 h,
quenched by addition of methanol. The solvent was
evaporated under reduced pressure and the crude purified
by flash chromatography (hexane/EtOAc, 85:15, v/v, 1 %
Et3N) to afford compound 8 (0.41 g, 63 %), an oil, as a
mixture of isomers. 1H NMR (CDCl3): d 1.12–1.53
(m, 25H, 8 CH2 and tBu), 1.58–1.70 (m, 2H, 2 H-17),
2.02–2.10 (m, 0.8H, 2 H-8MIN), 2.11–2.18 (m, 1.2H, 2
H-8MAG), 2.47 (s, 3H, CH3Ph), 2.50–2.68 (m, 2H, 2 H-3),
4.04 (t, 2H, J17,18 = 6.5 Hz, 2 H-18); 4.38–4.49 (m, 1H,
H-2), 5.00–5.10 (m, 1H, NH), 5.11–5.28 (m, 2H, CH2Ph),
5.32–5.51 (m, 1.6H, H-4 and H-7MAG), 5.58–5.67 (m, 0.4H,
H-7MIN), 5.87–6.09 (m, 1.4H, H-6 and H-5MIN), 6.32–6.42
(m, 0.6H, H-5MAG), 7.33–7.42 (m, 7H, arom), 7.78–7.84
(m, 2H, arom) ppm. ESI–MS (positive-ion mode, m/z):
678.1 [M?Na]?. Anal. calcd for C37H53NO7S (655.35):
C 67. 76, H 8.14, N 2.14, S 4.89. Found: C 67. 47, H 8.39,
N 2.01, S 4.59.
1
(0.20 g, 65 %), an oil, as a mixture of isomers. H NMR
(CDCl3): d 1.19–1.57 (m, 36H, 9 CH2 and 2 tBu),
2.01–2.10 (m, 0.8H, 2 H-8MIN), 2.10–2.19 (m, 1.2H, 2
H-8MAG), 2.46–2.69 (m, 2H, 2 H-3), 3.03–3.19 (m, 2H, 2
H-18), 4.36–4.60 (m, 2H, H-2 and NH), 4.99–5.28 (m, 3H,
CH2Ph and NH), 5.31–5.51 (m, 1.6H, H-4 and H-7MAG),
5.56–5.67 (m, 0.4H, H-7MIN), 5.86–6.08 (m, 1.4H, H-6 and
H-5MIN), 6.32–6.43 (m, 0.6H, H-5MAG), 7.31–7.44 (m, 5H,
arom) ppm. ESI–MS (positive-ion mode, m/z): 623.2
[M?Na]? (78), 1222.7 [2M?Na]? (100). Anal. calcd for
C35H56N2O6 (600.41): C 69.97, H 9.39, N 4.66. Found: C
69.62, H 9.73, N 4.45.
Benzyl (2S)-18-azido-2-(tert-butoxycarbonylamino)-
octadeca-4,6-dienoate (9)
(2S)-2,18-di-(tert-butoxycarbonylamino)-octadecanoic
acid (1)
Tosylate 8 (0.39 g, 0.59 mmol), NaN3 (0.12 g, 1.77 mmol)
and TBAI (0.015 g, 0.04 mmol) were dissolved in dry
DMF (12 mL) and the reaction was stirred at 60 °C for 2 h.
The mixture was then diluted with EtOAc (20 mL) and
washed with water (20 mL). The organic layer was
extracted with EtOAc (2 9 20 mL) and the combined
organics dried over Na2SO4 and evaporated. Purification by
flash chromatography (hexane/EtOAc, 9:1, v/v, 1 % Et3N)
gave compound 9 (0.29 g, 92 %), an oil, as a mixture of
To a solution of compound 10 (0.18 g, 0.30 mmol) in
MeOH (10 mL) a catalytic amount of 10 % Pd/C was
added. The reaction was stirred under hydrogen atmo-
sphere for 15 h, and then diluted with MeOH and filtered
through a MILLIPORE filter, washing the filter with
CHCl3/MeOH. After evaporation of the solvent, the crude
was purified by flash chromatography (DCM:MeOH 95:5,
v/v) to give compound 1 (0.13 g, 85 %) as a colourless oil.
1
isomers. H NMR (CDCl3): d 1.21–1.52 (m, 25H, 8 CH2
123

A practical route to a,x-diamino acids
439
[a]2D0 ?4.6 (c 0.5, CHCl3/MeOH 1:1). H NMR (CDCl3/
MeOD 1:1, v/v): d 1.14–1.53 (m, 46H, 2 tBu and 14 CH2),
1.53–1.66 (m, 1H, H-3a), 1.72–1.86 (m, 1H, H-3b), 3.04 (t,
2H, J17,18 = 7.1 Hz, H-18), 3.92–4.05 (m, 1H, H-2) ppm.
13C NMR (CDCl3:MeOD 1:1, v/v): d 25.5, 26.4, 27.9 (3C,
tBu), 29.2–29.7 (15C), 32.7 (C-3), 40.3 (C-18), 55.3 (C-2),
78.8 and 79.2 (2 CMe3), 156.4 and 157.0 (2 CONH), 179.0
(C-1) ppm. ESI–MS (negative-ion mode, m/z): 513.3
[M-H]- (100); 1027 [2 M-H]- (30). Anal. calcd for
C28H54N2O6 (514.40): C 65.33, H 10.57, N 5.44. Found: C
65.20, H 10.89, N 5.22.
acid, and the phosphonium salt 3, which can efficiently be
obtained from 12-bromo-dodecanol (Fig. 1).
1
Protected aspartic acid 4 was the starting material of our
synthetic strategy (Scheme 1). A benzyl ester protection of
the carboxylic group was decided, as it can be easily
removed at the end of the synthesis by catalytic hydrog-
enolysis with the concomitant reduction of the double
bonds formed during the Wittig reactions. Compound 4
was subjected to a regioselective reduction mediated by
N,N-dimethylchloromethylenammonium chloride and lith-
ium tri(tert-butoxy)aluminium hydride, according to a
synthetic protocol already reported for a corresponding
glutamic derivative (Bycroft et al. 2003), which gave
aldehyde 5 in good yield after flash chromatography. A C-2
homologation was then performed at this stage, by reacting
compound 5 with (triphenylphosphoranylidene)-acetalde-
hyde at high temperature to give compound 2 in high yield.
Compound 2, which is a homoglutamic acid d-aldehyde,
has the advantage to be already installed with the terminal
aldehyde necessary for the main Wittig olefination for the
construction of the long alkyl skeleton. The reaction of
aldehyde 2 with the ylide generated from phosphonium salt
of 12-bromododecanol was thus tested. The yield of the
reaction is highly dependent on the temperature and the
base used for the generation of the ylide. Also literature
data on similar substrates indicate that there are not general
rules, and the reaction conditions, i.e. the base to promote
the generation of the ylide as well the temperature for the
deprotonation or the addition of the aldehyde, are finely
tuned on the particular reaction partners. We obtained our
best results by performing both the formation of the ylide
and the addition of the aldehyde at -60 °C, using lithium
hexamethyldisilazide (LiHMDS) as the base. The quality
of the hygroscopic phosphonium salt 3 also resulted very
important. This was obtained from 12-bromo-dodecanol
using standard protocols (Lei and Atkinson 2000), and,
after chromatographic purification, treatment with anhy-
drous tetrahydrofuran makes it a white solid. Finally, after
optimization of the reaction conditions alkenol 6 was
obtained in satisfactory yield as a mixture of isomers. This
is not a problem for our purposes, since we planned to
remove the double bond at the end of the synthesis together
with the carboxylic protecting group. Nevertheless, the
identity of the condensation product was unequivocally
confirmed after hydrogenation of compound 6, under the
catalysis of palladium, which gave the fully characterized
compound 7. The x-hydroxy long a-amino acid 7 is
another interesting unnatural LAA derivative, which can be
used for the construction of modified peptides.
Results and discussion
The length of the alkyl skeleton of compound 1, com-
bined with the functionalization of the unnatural fatty
acid, immediately oriented the synthetic strategy towards
the use of a series of Wittig reactions for the convergent
preparation of the C-18 fatty acid, by condensations of
readily available and suitably functionalized building
blocks. In a Wittig reaction the elongation of the alkyl
chain occurs through a condensation between an aldehyde
and a phosphonium salt, via the formation of an inter-
mediate ylide, which evolves to an alkene. It is possible to
take advantage of properly derivatized L-amino acids as
the aldehyde counterpart, since they already contain the
a-aminoacid group, while, the phosphonium salt could be
derived from x-functionalized alkyl bromides. It is
unquestionable that many different disconnections are
possible for the preparation of compound 1, but our
choice was oriented towards the reduction of critical
synthetic steps and by the commercial availability of the
building blocks, easily obtainable from common suppliers.
Due to this, the use of N-protected x-amino-alkyl bromide
was discarded for two main reasons. First, due to the
availability of only short homologues those are expensive.
Second, because of problems relating to the commonly
used amino protecting groups which are incompatible
with the reaction conditions for the Wittig condensation.
In fact, the N-FMOC base labile protecting group does not
resist to the basic conditions of the Wittig reaction, while,
with N-BOC, there are problems of instability connected
with the preparation of the phosphonium salt at high
temperature, as already reported in the literature and
confirmed by our preliminary experiments (Wiktelius and
Luthman 2007). On the contrary, the use of x-bro-
moalkanol is highly encouraged for the commercial
availability of different homologues and because the
conversion of an x-hydroxy group into an amine can be
easily accomplished by well-established chemical trans-
formations. Consequently, the choice of the reaction
partners entailed into aldehyde 2, derived from L-aspartic
Compound 6 was finally converted into the x-amino
acid by replacing the hydroxyl group with an amine. This
was accomplished by initial conversion of 6 into tosylate 8,
in order to have a good leaving group which was efficiently
123

440
G. Brasile et al.
Scheme 1 i (COCl)2, DMF,
Py, MeCN, THF, -30 °C, then
LiAlH(OtBu)3, -78 °C, 60 %;
ii Ph3P=CH–CHO, toluene,
70 °C, 75 %; iii 3, LiHMDS,
THF, -60 °C, 65 %; iv H2,
Pd/C, CHCl3/MeOH, 87 %;
v TsCl, DMAP, DIPEA, DCM,
63 %; vi NaN3, TBAI, DMF,
60 °C, 92 %; vii Ph3P, THF,
H2O, then BOC2O, DCM,
40 °C, 65 %; viii H2, Pd/C,
MeOH, 85 %
NHBOC
COOH
NHBOC
CHO
NHBOC
CHO
i
ii
BnOOC
iii
BnOOC
BnOOC
HOOC
4
5
2
NHBOC
NHBOC
iv
( )15
OH
( )9
BnOOC
R
7
6. R = OH
8. R = OTs
9. R = N3
v
vi
vii
viii
1
10. R = NHBOC
Sharpless KB, Heath JR (2009) Iterative in situ click chemistry
creates antibody-like protein-capture agents. Angew Chem Int
Ed 48:4944–4948
displaced by sodium azide in dimethylformamide at high
temperature. The azido group of compound 9 was then
selectively reduced by means of a Staudinger reaction to
give the crude amine, which was in turn protected as a
BOC derivative. Finally, catalytic hydrogenation of 10
produced the saturated a,x-diamino octadecanoic acid
derivative 1, by simultaneous reduction of the double
bonds and deprotection of the carboxylic group.
Aureli M, Prioni S, Mauri L, Loberto N, Casellato R, Ciampa MG,
Chigorno V, Prinetti A, Sonnino S (2010) Photoactivable
sphingosine as a tool to study membrane microenvironments in
cultured cells. J Lip Res 51:798–808
Bycroft BW, Chhabra SR, Kellam B, Smith P (2003) Convenient
synthesies of (3S,5S)-carbapenam-3-carboxylates and their bio-
synthetic relevance. Tetrahedron Lett 44:973–976
Constantinou-Kokotou V, Kokotos G (1999) Synthesis of optically
active lipidic a-amino acids and lipidic 2-amino alcohols. Amino
Acids 16:273–285
Conclusions
Esser CK, Kopka IE, Durette PL, Harrison LK, Niedxwiecki LM,
Izquierdo-Martin M, Stein RL, Hagmann WK (1995) Inhibition
of matrix metalloproteinases by N-carboxyalkyl peptides con-
We have described a practical methodology for the syn-
thesis of a,x-diamino acids, which has been applied to the
preparation of an octadecanoic derivative. The preparation
of compound 1 provides a general route to a,x-diamino
analogues of long fatty acids, and resulted in the longest
compound of the family ever synthesized. Furthermore,
this approach permits the modulation of the alkyl chain
length depending on the chain length of the ylide used for
the Wittig olefination reaction. In order to allow the
insertion in a peptide sequence, it would be also possible to
differentiate the two amino groups at the end of the syn-
thesis, by protecting the amine obtained after Staudinger
reduction with a different protecting group. During the
synthetic sequence another class of unnatural LAAs, i.e.
a-amino-x-hydroxy long fatty acids, is recovered. The
synthetic protocol herein reported is of general applica-
bility and is a useful sequence to obtain LAA derivatives
for the synthesis of drug delivery system or for the study of
biological membranes.
0
taining extended alkyl residues at P1 . Bioorg Med Chem Lett
5:539–542
Gaudry R (1953) A convenient synthesis of DL-ornithine. Can J Chem
31:1060–1063
Itoh K, Kori M, Inada Y, Nishikawa K, Kawamatsu Y, Sugihara H
(1986) Synthesis and angiotensin converting enzyme-inhibitory
activity of 1,5-benzothiazepine and 1,5-benzoxazepine deriva-
tives. Chem Pharm Bull 34:2078–2089
Lei H, Atkinson J (2000) Synthesis of phytyl- and chroman-
derivatized photoaffinity labels based on a-tocopherol. J Org
Chem 65:2560–2567
Li C, Shen J, Wei XL, Xie C, Lu WY (2012) Targeted delivery of a
novel palmitylated D-peptide for antiglioblastoma molecular
therapy. J Drug Target 20:264–271
Ma J-A (2003) Recent developments in the catalytic asymmetric
synthesis of a- and b-amino acids. Angew Chem Int Ed 42:
4290–4299
Markidis T, Kokotos G (2002) Synthesis of enantiopure x-function-
alized C15 a-amino carboxylates. J Org Chem 67:1685–1688
´
Najera C, Sansano JM (2007) Catalytic asymmetric synthesis of a-
amino acids. Chem Rev 107:4584–4671
Payne LS, Boger J (1985) Synthesis of a,x-diamino acids: an efficient
preparation of DL-homolysine and protected derivatives. Synth
Commun 15:1277–1290
Acknowledgments We gratefully acknowledge the University of
Milan for financial support.
Russell AL, Williams BC, Spuches A, Klapper D, Srouji AH, Hicks
RP (2012) The effect of the length and flexibility of the side
chain of basic amino acids on the binding of antimicrobial
peptides to zwitterionic and anionic membrane model systems.
Bioorg Med Chem 20:1723–1739
References
Saito S, Matsui S, Watanabe M, Waga T, Kajiwara Y, Shirota M, Iijima
M, Kitabatake K (1990) Synthesis and angiotensin converting
enzyme inhibitory activity of N-carboxymethyldipeptides with an
omega-aminoalkyl group. Arzneimittelforschung 40:257–260
Agnew HD, Rohde RD, Millward SW, Nag A, Yeo WS, Hein JH,
Pitram SM, Tariq AA, Burns VM, Krom RJ, Fokin VV,
123

A practical route to a,x-diamino acids
441
Sarpietro MG, Micieli D, Pignatello R, Liang MT, Toth I, Castelli F
(2008) Effect of variation in the chain length and number in
modulating the interaction of an immunogenic lipopeptide with
biomembrane models. Thermochim Acta 471:14–19
Shirota M, Kajiwara Y, Iijima M, Kitabatake K (1990) Pharmaco-
logical properties of the new non-sulfhydryl angiotensin con-
verting enzyme inhibitor N-[8-amino-1(S)-carboxyoctyl]-L-
alanyl-L-proline. Arzneimittelforschung 40:515–519
Toth I (1994) A novel chemical approach to drug delivery: lipidic
amino acid conjugates. J Drug Target 2:217–239
Tran-Thi NH, Falk H (1995) An efficient synthesis of the plant
growth hormone 1-triacontanol. Monatsh Chem 126:565–568
Wiktelius D, Luthman K (2007) Taking control of P1, P10 and double
bond stereochemistry in the synthesis of Phe–Phe (E)-alkene
amide isostere dipeptidomimetics. Org Biomol Chem 5:603–605
Soloshonok VA, Sorochinsky AE (2010) Practical methods for the
synthesis of symmetrically a,a-disubstituted a-amino acids.
Synthesis 2319–2344
123