Stereoselective Michael addition of benzylamines to homochiral methylenebutanedioates

Article abstract of DOI:10.1039/a907060e

The Michael addition of benzylamine to the homochiral methylenebutanedioate 10 gave an adduct 11 in good yield with high stereoselectivity. By performing the reaction in methanol the (2R,3R) diastereoisomer 11 was obtained in 88% de, which was increased t

Full text of DOI:10.1039/a907060e

View Article Online / Journal Homepage / Table of Contents for this issue
Stereoselective Michael addition of benzylamines to homochiral
methylenebutanedioates
Alexander N. Chernaga,a Stephen G. Davies,*b Christopher N. Lewisc and Richard S. Todd*c
a Chemical Crystallography Laboratory, University of Oxford, South Parks Road, Oxford,
UK OX1 3PD
b The Dyson Perrins Laboratory, University of Oxford, South Parks Road, Oxford,
UK OX1 3QY
c British Biotech Pharmaceuticals Ltd., Watlington Road, Cowley, Oxford, UK OX4 5LY
Received (in Cambridge, UK) 1st September 1999, Accepted 5th October 1999
The Michael addition of benzylamine to the homochiral methylenebutanedioate 10 gave an adduct 11 in good yield
with high stereoselectivity. By performing the reaction in methanol the (2R,3R) diastereoisomer 11 was obtained
in 88% de, which was increased to 98% de after recrystallisation of the primary amine derivative 13. The ratio of
diastereoisomers was reversed by performing the reaction in aprotic solvents, with the (2R,3S) diastereoisomer 12
being obtained in 40% de in tetrahydrofuran. The Michael adduct 11 is formed under kinetic control. The primary
amine 15 is a key intermediate in the synthesis of novel matrix metalloproteinase inhibitors.
Introduction
Matrix metalloproteinases (MMPs)1 are a family of enzymes
which are responsible for the degradation of all the major com-
ponents of extracellular matrix. Over-activation of MMP’s has
been linked with a range of diseases such as arthritis and can-
cer; thus inhibition of these proteinases could serve as effective
treatments for such disease states.2 For example, other workers
at British Biotech Pharmaceuticals (BBP), have recently
described the substrate-based inhibitors, batimastat 13 and
marimastat 2.4
Fig. 1 Retrosynthetic analysis for α-aminomethyl substituted MMPs
(Bn = PhCH2).
substituted α,β-unsaturated esters are well documented.5a,7a,8
However, a survey of the literature revealed that although
attention has been focused on the addition of amines to
methylenebutanedioates9 there are no references to homochiral
3-substituted methylenebutanedioates. Other workers have
shown that amines can be added to an α-substituted acrylate
bearing a stereogenic centre to furnish a β-aminoester with high
(20:1) stereochemical control.10 Thus we were encouraged to
examine the Michael addition of benzylamine to the acrylate
ester 10, both as a synthetic goal and also to gain more insight
into the mechanistic aspects of the reaction.
Clearly these inhibitors possess very similar structural
features; in particular they possess an hydroxamic acid zinc
binding group and structure–activity relationship (SAR)
studies have been carried out involving substitution α to this
group. As an extension to that work we were interested in
gaining access to aminomethyl substituents α to the hydroxamic
acid, such as in 3 (Fig. 1). Given our interest in asymmetric
β-amino acid synthesis via Michael addition of nitrogen nucleo-
philes,5 a retro-synthetic analysis (Fig. 1) suggested that such
molecules could be constructed by a Michael addition of an
amine to an appropriately functionalised acrylate ester 5. This
in turn could be obtained from the acid 6,3,6 a readily available
bulk intermediate used by BBP for the synthesis of batimastat.
Our goal was thus twofold: (i) to synthesise an amino diester
such as 4 (Fig. 1) in which the ester groups are differentiated, to
allow selective manipulation, and (ii) to prepare the amino
diester 4 as a single stereoisomer with the (2R,3R) absolute
configuration, required for optimum enzyme inhibitory activity.
The Michael additions of amines to acrylate esters7 and β-
Results and discussion
Synthesis of the required methylenebutanedioate 10 from the
available acid 6 was achieved readily (Scheme 1). The acid 66
was protected as its tert-butyl ester 7, then the benzyl groups
were removed by hydrogenolysis. The resulting substituted
malonic acid 8 was subjected, without purification, to a
Mannich type reaction using formaldehyde and piperidine to
J. Chem. Soc., Perkin Trans. 1, 1999, 3603–3608
This journal is © The Royal Society of Chemistry 1999
3603

View Article Online
Table 1 Variation of the diastereomeric ratio 11:12 with reaction
temperature
Ratio (NMR)
[isolated yield, %] 11:12
T/ЊC
Ϫ20
94:6a
Ambient
65
90:10 [70]
72:28b
a Incomplete reaction after 168 h. b Products not isolated.
Scheme 1 Reagents: a, isobutene, H2SO4, CH2Cl2, 97%; b, H2, 10%
Pd–C, EtOH; c, HCHO, piperidine, 86% from 7; d, K2CO3, MeI,
Me2CO, 96%.
furnish the acrylic acid 9 in 84% yield from 6. The potassium
salt of 9 was then methylated in 96% yield using iodomethane,
providing the required acrylate 10.
Addition of a methanolic solution of benzylamine to 10 gave
a 90:10 mixture of the Michael adducts 11 and 12 after 18 h at
ambient temperature (Scheme 2). Isolation of the individual
Fig. 2 X-Ray crystal structure of 1-(tert-butyl) 4-methyl (2R,3R)-3-
(aminomethyl)-2-isobutylbutanedioate, (1S)-10-camphorsulfonic acid
salt 16.
Scheme 2 Reagents: a, BnNH2, MeOH, 70%; b, HCO2H, MeOH, 10%
Pd–C, 100%.
diastereoisomers was not possible, although chromatography
did provide enriched mixtures of each diastereomer. However,
hydrogenolysis of the mixture of benzylamines 11 and 12 gave
the formate salts 13 and 14, from which the major diastereomer
13 was obtained in greater than 98% de after a single recrystal-
lisation from ethyl acetate.
Scheme
3
Reagents: a, aqueous NaHCO3, 90%; b, (1S)-10-
camphorsulfonic acid, EtOAc.
The absolute configuration of the amine 15, derived from salt
13, and by inference the precursor Michael adduct 11, was
established from X-ray crystallographic analysis of the (1S)-10-
camphorsulfonic acid salt 16 (Scheme 3), by comparison with
the known absolute configuration of (1S)-10-camphorsulfonic
acid (Fig. 2).
In order to explain the origin of the surprisingly high stereo-
selectivity found in the formation of 11, the mechanism of the
Michael addition of amines to acrylate esters was considered.
Although studies have been performed on the addition of
amines to acrylate esters,11 the nature of the mechanism is still
not fully understood, and in particular whether the Michael
adducts are formed under kinetic or thermodynamic control.12
Before the origins of the observed diastereoselectivity could
be investigated, the question of kinetic vs. thermodynamic
control in our system required addressing. Several experiments
were devised to this end and the following results are consistent
with the Michael addition of benzylamine to the acrylate 10
proceeding under kinetic control: (i) a variable temperature
study (Table 1) showed that the diastereomeric ratio of 11:12
was temperature dependent; (ii) samples taken from a reaction
at ambient temperature after 0.5, 1, 2, 4, 8 and 168 h showed no
change in the 90:10 diastereomeric ratio 11:12, indicating that
there is no equilibration between 11 and 12; (iii) no change in
the diastereomeric ratio was observed by re-subjecting a 30:70
mixture of 11:12 (from reaction in tetrahydrofuran, see Table
2) to the reaction conditions that had previously produced a
ratio of 90:10 (methanol).
The Michael addition of benzylamine to a methylenebutane-
dioate such as 10 is believed to give the enolate 1711 which is
3604
J. Chem. Soc., Perkin Trans. 1, 1999, 3603–3608

View Article Online
Table 2 Variation of diastereomeric ratio 11:12 with solvent
Ratio (NMR)
11:12
Solventa
No solventb
MeOHb
30:70
90:10
50:50
40:60
30:70
c
CHCl3
MeCNc
THFc
a All reactions performed at ambient temperature. b Complete reaction
after 18 h. c Incomplete reaction after 168 h.
Scheme 4 Reagents: a, (R)-1-phenylethylamine, MeOH, 91%; b, (S)-1-
phenylethylamine, MeOH, 93%; c, 1  HCO2H–MeOH, 10% Pd–C,
100%; d, aqueous NaHCO3.
Michael adduct and/or change the configuration of the stereo-
genic centre via double asymmetric induction13 (Scheme 4). We
were aware that the Michael addition of (S)-1-phenyl-
ethylamine to methyl crotonate furnishes adducts with poor
diastereomeric excesses (up to 20%),5a,8a so were therefore not
surprised that there were no differences observed in the con-
figuration of the major diastereomers 19 and 21 and that of 11,
and that the diastereomeric ratios 19:20 and 21:22, were iden-
tical with the ratio 11:12. Hydrogenolysis of pure 19 gave the
primary amine 15 after basification, thus establishing its abso-
lute configuration. The 90:10 mixture of 21:22 was reacted
similarly, giving 15 as the major diastereomer.
Fig. 3 Proposed mechanism for the formation of 11 and 12 under
kinetic control.
rapidly, or simultaneously, depending on the solvent, proton-
ated to furnish the enol 18. The new stereocentre in 11 is created
as a result of the tautomerisation of the enol 18, and it is the
facial selectivity of this protonation step which gives rise to the
diastereoselectivity of the reaction. The factors which may be
responsible for the origin of the diastereoselectivity during the
tautomerisation of 18 to 11 include: (i) the geometry of the enol
18 and (ii) the nature of the enol 18 in terms of intramolecular
hydrogen-bonded species vs. solvated open-chain species.11
Additionally we were interested to see what effect changing the
solvent would have on the ratio 11:12, as this might provide
more information on the nature of protonation of the enol 18.
For the reaction in solvents other than methanol, the diastereo-
selectivity was zero, in a relatively low polarity solvent (chloro-
form), whereas a significant reversal of stereoselectivity was
observed in an ether solvent (tetrahydrofuran) (Table 2).
In order to show that the products obtained by performing
the Michael addition in aprotic solvents were derived from
kinetic control, a 90:10 mixture of diastereomers (Table 2 ) was
subjected to the Michael addition conditions in chloroform. If
the reaction is thermodynamically controlled in aprotic solvent
then the ratio of 11:12 would be expected to change to 50:50.
However, no change in the diastereomeric ratio was observed,
confirming kinetic control. Despite the above observations, the
origin of the stereocontrol of the protonation of enol 18 is still
unclear and requires further studies.
Conclusion
We have shown that high levels of stereochemical control
(90:10) are obtained from the addition of benzylamine to the
substituted methylenebutanedioate 10 in methanolic solution,
and that this ratio of diastereomers can be reversed (30:70) by
change of solvent to tetrahydrofuran. The configuration of the
newly formed stereogenic centre is controlled entirely from the
existing stereogenic centre in 10, as both (R)- and (S)-1-phenyl-
ethylamines furnish the Michael adducts 19 and 21 as the major
diastereomers with an identical absolute configuration of
the newly formed stereogenic centre and the same ratio of
diastereomers.
The Michael adduct 11 is formed under kinetic control. The
origin of the stereoselectivity requires further investigation, and
these studies are in progress. Amines 4 have been further elab-
orated to MMP inhibitors, details of which will appear in
future publications.
Experimental
The Michael additions of both (R)- and (S)-1-phenylethyl-
amines with 10 were also investigated, in order to see whether
we could enhance further the diastereomeric excess of the
Melting points (mps) were recorded on a Gallenkamp capillary
melting point apparatus and are uncorrected. Optical rotations
were measured using a Perkin-Elmer 341 polarimeter and [α]D
J. Chem. Soc., Perkin Trans. 1, 1999, 3603–3608
3605

View Article Online
values are given in units of 10Ϫ1 deg cm2 gϪ1. Microanalyses
were performed by Medac Ltd, Brunel University. IR spectra
were recorded on a Perkin-Elmer 1600 Series FTIR spectrometer
1-tert-Butyl 4-methyl (2R)-2-isobutyl-3-methylenebutanedioate
10
Potassium carbonate (43 g, 311 mmol) was added to a stirred
solution of the acid 9 (15 g, 61 mmol) in acetone (500 ml).
Stirring was continued for 0.5 h then iodomethane (7.6 ml, 122
mmol) was added and the mixture stirred for 24 h, then filtered,
the solid washed with diethyl ether (100 ml) and the filtrate
evaporated. The residue was taken up in diethyl ether, filtered
through a pad of SiO2 and evaporated to give the methyl ester
10 (15.22 g, 96%) as a pale yellow oil, [α]D20 ϩ1.1 (c 0.046 in
MeOH) (Found: C, 65.25; H, 9.3. C14H24O4 requires C, 65.6;
H, 9.27%); νmax (film)/cmϪ1 1739, 1719, 1629; δH (CDCl3; 500
1
as neat liquids or solids. The H and 13C NMR spectra were
obtained on a Bruker DPX 250 spectrometer (250 MHz and
62.6 MHz) or a Bruker AMX 500 spectrometer (500 MHz and
125.7 MHz), with tetramethylsilane as internal standard. J
values are given in Hz. High resolution mass spectra (HRMS)
were obtained by the University of Manchester Spectroscopic
Services Department.
1,1-Dibenzyl 2-tert-butyl (2R)-4-methylpentane-1,1,2-tricarb-
oxylate 7
MHz) 6.29 (1H, s, Z-CH᎐C), 5.69 (1H, s, E-CH᎐C), 3.74 (3H, s,
᎐
᎐
Concentrated sulfuric acid (5 ml) was added to a stirred
solution of (2R)-2-[2-(benzyloxy)-1-(benzyloxycarbonyl)-2-
oxoethyl]-4-methylpentanoic acid 6 (50 g, 126 mmol) in
dichloromethane (100 ml) at Ϫ70 ЊC, in a screw-cap pressure
bottle. 2-Methylpropene was then condensed into the reaction
until the volume had approximately doubled, when the bottle
was sealed and allowed to warm to ambient temperature. After
23 h the reaction was cooled to Ϫ70 ЊC, the contents poured
slowly into well-stirred 1  aqueous sodium carbonate (250 ml)
and the mixture stirred for a further 3 h. The mixture was satur-
ated with sodium chloride to facilitate separation, separated
and the organic phase was washed with saturated aqueous
sodium chloride (100 ml) then dried (MgSO4), filtered and
evaporated to give the triester 7 (54.35 g, 97.6%) as a thick oil,
which slowly crystallised, mp 36–37 ЊC; [α]D20 ϩ5.3 (c 0.024 in
MeOH) (Found: C, 71.25; H, 7.75; C27H34O6 requires C, 71.3;
H, 7.5%); νmax (neat)/cmϪ1 1750, 1719; δH (CDCl3) 7.35–7.26
(10H, m, ArH), 5.22–5.07 (4H, m, 2 × ArCH2), 3.77 (1H, d,
J 10.2, 1-H), 3.08 (1H, ddd, J 10.2, 10.2, 4.3, 2–H), 1.66–1.44
(2H, m, 3-H and 4-H), 1.41 (9H, s, CO2But), 1.15–1.04 (1H,
m, 3-H), 0.865 (3H, d, J 6.4, CHMe), 0.835 (3H, d, J 6.4,
CHMe); δC (CDCl3) 172.6 (CO2But), 167.6 (2 × CO2Bn),
135.1 and 135.0 (2 × quaternary Ar), 128.4–128.0 (10 × Ar),
81.1 (CMe3), 67.1 and 67.0 (2 × ArCH2), 54.6 (C-2), 43.4
(C-3), 39.3 (CHCH2), 27.7 (CMe3), 25.6 (CHMe2), 23.4 and
21.0 (CHMe2).
CO2Me), 3.48 (1H, t, J 7.5, CHCO2But), 1.74–1.69 and 1.48–
1.42 (each 1H, m, CH2), 1.60–1.52 (1H, m, CHMe2), 1.40 (9H,
s, But), 0.91 and 0.88 (each 3H, d, J 6.5, CHMe2); δC (CDCl3)
172.6 and 166.9 (2 × CO), 139.1 (CH ᎐C), 125.8 (CH ᎐C), 80.5
᎐
᎐
2
2
(CMe3), 51.9 (CO2Me), 45.5 (CCO2But), 40.3 (CH2), 27.9
(CMe3), 25.9 (CHMe2), 22.4 and 22.3 (CHMe2); [Found
(CI): (M ϩ H)ϩ m/z 257.1760, C14H24O4 requires (M ϩ H)ϩ
257.1753].
(2R,3R)- and (2R,3S)-1-tert-Butyl 4-methyl 3-(benzylamino-
methyl)-2-isobutylbutanedioate 11/12
Benzylamine (10.5 ml, 97.5 mmol) was added dropwise to a
stirred solution of the diester 10 (5.00 g, 19.5 mmol) in meth-
anol (25 ml) at ambient temperature, under an argon atmos-
phere. After 18 h the solvent was evaporated and a solution of
the residue in ethyl acetate (200 ml) was washed with 1  aqueous
citric acid (100 ml; 50 ml) then water (50 ml), saturated aqueous
sodium hydrogen carbonate solution (100 ml) and saturated
aqueous sodium chloride (100 ml). The organic phase was dried
(MgSO4) and evaporated to give a 90:10 mixture of the
Michael adducts 11 and 12 (4.95 g, 70%) as a pale yellow oil.
Purification was achieved by chromatography (SiO2, eluting
with 4:1 hexane–ethyl acetate), [α]D20 Ϫ1.3 (c 0.0294 in MeOH)
for 80% de (Found: C, 69.4; H, 9.15; N, 3.85. C21H33NO4
requires C, 69.35; H, 9.0; N, 3.9%); νmax (film)/cmϪ1 3350, 1725;
δH (CDCl3) 7.33–7.19 (5H, m, ArH), 3.77–3.76 (1H, d, J 2,
ArCH2N), 3.70 (3H, s, CO2Me), 2.92–2.63 (4H, m, NCH2CH
and CHCO2But), 1.68–1.41 (3H, m, NH and CHCHMe2), 1.38
(9H, s, But), 1.05 (1H, ddd, J 13.2, 9.4, 3.8, CHCHMe2), 0.88
and 0.87 (each 3H, d, J 6.3, CHMe2); δC (CDCl3) 174.2 and
173.2 (2 × CO), 139.9 (quaternary Ar), 128.3, 128.0 and 126.9
(Ar), 80.8 (CMe3), 53.4 (ArCH2N), 51.6 and 49.4 (CCO2But
and CCO2Me), 48.5 (NCH2CH), 45.0 (CO2Me), 39.7
(CHCH2CH), 27.9 (CMe3), 26.1 (CHMe2), 23.4 and 21.3
(CHMe2) [Found (CI): (M ϩ H)ϩ m/z 364.2487, C21H33NO4
requires (M ϩ H)ϩ 364.2488].
2-[(1R)-1-(tert-Butoxycarbonyl)-3-methylbutyl]acrylic acid 9
A slurry of 10% Pd–C (9.1 g) in ethyl acetate (~50 ml) was
added to a solution of the triester 7 (90.8 g, 0.2 mol) in ethanol
(900 ml), under an atmosphere of argon. Hydrogen gas was
bubbled through the well stirred mixture for 2 h then the
reaction stirred under an hydrogen atmosphere for 16 h. The
catalyst was removed by filtration through Celite, the filtrate
transferred to a round-bottomed flask and cooled to <10 ЊC.
Piperidine (25 ml) was added in several portions, the reaction
stirred for a further 10 min then an aqueous solution of
formaldehyde (37% wt; 80 ml) added dropwise over 10 min. The
cooling bath was removed and the reaction stirred for 16 h then
the solvents were evaporated to leave a thick oil, which was
dissolved in ethyl acetate (500 ml) and washed with hydro-
chloric acid (1 ; 250 ml), then saturated aqueous sodium chlor-
ide (100 ml). The solution was dried (MgSO4) and evaporated
to leave a gum which was subjected to column chromatography
(SiO2, eluting with 2:3 ethyl acetate–hexane) to give the acid 9
(41.69 g, 86%) as a gum; [α]D20 ϩ2.1 (c 0.012 in MeOH) (Found: C,
64.4; H, 9.3. C13H22O4 requires C, 64.4; H, 9.15%); νmax (film)/
cmϪ1 3506–2613 (br.), 1731, 1696, 1625; δH (CDCl3; 500 MHz)
Minor diastereoisomer (by NMR difference): δH (CDCl3)
3.68 (s, CO2Me), 1.41 (s, But); δC 80.5, 53.6, 47.9, 47.7, 44.3,
38.6, 23.5, 21.4.
1-tert-Butyl 4-methyl (2R,3R)-3-(aminomethyl)-2-isobutyl-
butanedioate, formate salt 13
A mixture of the Michael adducts 11/12 (1.00 g, 2.75 mmol)
and 10% Pd–C (0.5 g) in a 1  solution of formic acid in meth-
anol (60 ml) was stirred for 45 min, filtered and the catalyst
washed with methanol. The filtrate was diluted with toluene (50
ml) and evaporated under reduced pressure to leave a gum
which was triturated with ethyl acetate to furnish a white solid,
0.90 g (100%), as a 9:1 mixture of diastereomers. Recrystallisa-
tion of the solid (0.80 g) from ethyl acetate gave the major
diastereomer 13 (0.584 g, 74%; >98% de), mp 116 ЊC; [α]D20 ϩ7.6
(c 0.015 in MeOH) (Found: C, 56.0; H, 9.2; N, 4.3.
C15H29NO6ؒ0.1H2O requires C, 56.1; H, 9.2; N, 4.4%); νmax
(neat)/cmϪ1 1732, 1716, 1557; δH (MeOD) 8.50 (1H, br s,
HCO2Ϫ), 3.78 (3H, s, CO2Me), 3.27–2.88 (4H, m, NCH2CH
6.45 (1H, s, Z-CH᎐C), 5.82 (1H, s, E-CH᎐C), 3.49 (1H, t, J 7.5,
᎐
᎐
CHCO2But), 1.77–1.71 (1H, m, CH2), 1.57 (1H, m, CHMe2),
1.52–1.44 (1H, m, CH2), 1.42 (9H, s, But), 0.93 (3H, d, J 6.5,
CHMe), 0.90 (3H, d, J 6.5, CHMe); δC (CDCl3) 172.6 and 171.9
(2 × CO), 138.7 (CH ᎐C), 128.2 (CH ᎐C), 80.8 (CMe ), 45.2
᎐
᎐
2
2
3
(CCO2But), 40.4 (CH2), 27.8 (CMe3), 25.9 and 22.3 (CHMe2)
[Found (CI): (MϩNH4)ϩ m/z 260.1860, C13H22O4 requires
(MϩNH4) 260.1862].
3606
J. Chem. Soc., Perkin Trans. 1, 1999, 3603–3608

View Article Online
and CHCO2But), 1.74–1.51 (2H, m, CHCHMe2), 1.44 (9H, s,
But), 1.26–1.13 (1H, m, CHCHMe2), 0.94 and 0.925 (each 3H,
d, J 6.5, CHMe2); δC (MeOD) 173.6 and 173.5 (2 × CO), 83.0
(CMe3), 53.0 and 46.5 (CCO2But and CCO2Me), 39.5 and 38.8
(2 × CH2), 28.2 (CMe3), 27.3 (CHMe2), 23.1 and 22.1 (CHMe2)
[Found (CI): (M ϩ H)ϩ m/z 274.2023, C15H29NO6 requires
(M ϩ H)ϩ 274.2018].
fixed positional and thermal parameters (only the atoms
attached to the nitrogen were refined isotropically). Chebyshev14
weighting scheme with parameters 6.39, 2.43 and 2.93 was
applied. Corrections for Lorentz and polarisation effects as
well as empirical absorption correction based on azimuthal
scan data15 were applied. In the final stage of refinement the
data were corrected for the effect of isotropic extinction. Flack
test16 was applied for the absolute configuration determinations
(enantiopole parameter was refined to 0.004 using 4529 reflec-
tions with the non-averaged Friedel equivalents). Final R and
RЈ values are 0.035 and 0.039. Maximum and minimum peaks
1-tert-Butyl 4-methyl (2R,3R)-3-(aminomethyl)-2-isobutyl-
butanedioate 15
The formate salt 13 (114 mg, 0.357 mmol) was partitioned
between ethyl acetate (10 ml) and saturated aqueous sodium
hydrogen carbonate (2 ml). The organic layer was separated,
washed with water (1 ml) and dried (MgSO4). Evaporation of
the solvent gave the free amine 15 (85 mg, 90%) as an oil, [α]D20
ϩ7.7 (c 0.011 in MeOH) (Found: C, 61.2; H, 9.95; N, 5.00.
C14H27NO4 requires C, 61.5; H, 9.95; N, 5.15%); νmax (film)/cmϪ1
3620, 3394, 3330, 1747, 1700; δH (CDCl3) 3.75 (3H, s, CO2Me),
3.00–2.60 (4H, m, CH2N, CHCO2But and CHCO2Me), 1.71–
1.49 (2H, m, CHMe2 and CHH), 1.45 (9H, s, But), 1.41 (2H, br
s, NH2), 1.05 (1H, ddd, J 12.9, 9.5, 3.5, CHH), 0.90 and 0.885
(each 3H, d, J 6.5, CHMe2); δC (CDCl3) 174.0 and 173.3
(2 × CO), 81.0 (CMe3), 51.9 and 51.7 (CCO2But and CCO2Me),
44.5 (CO2Me), 42.2 (CH2N), 39.8 (CH2CHMe2), 27.9 (But),
26.2 (CHMe2), 23.4 and 21.3 (CHMe2) [Found (CI): (M ϩ H)ϩ
m/z 274.2020, C14H28NO4 requires (M ϩ H)ϩ 274.2018].
in the final difference synthesis are 0.18 and Ϫ0.30 e ÅϪ3
.
All crystallographic calculations were carried out using the
CRYSTALS17 program package on a Micro VAX 3800
computer. Neutral atom scattering factors were taken from the
usual sources.18†
(2R,3R)- and (2R,3S)-1-tert-Butyl 4-methyl 2-isobutyl-3-
({[(1R)-1-phenylethyl]amino}methyl)butanedioate 19/20
(R)-1-Phenylethylamine (646 µl, 5 mmol) was added to a stirred
solution of the methylenebutanedioate 10 (256 mg, 1 mmol) in
methanol (1.3 ml), under an argon atmosphere. After 120 h the
solvent was evaporated and a solution of the residue in ethyl
acetate (10 ml) was washed with aqueous citric acid (1 ; 5 ml;
2 ml) then water (5 ml), saturated aqueous sodium hydrogen
carbonate (5 ml) and saturated aqueous sodium chloride (5 ml).
The organic layer was dried (MgSO4) and evaporated to give a
90:10 mixture of the Michael adducts 19/20 (344 mg, 91%) as a
viscous oil. Chromatography (SiO2, eluting with 4:1 hexane–
ethyl acetate) furnished 19 in >99% de, [α]D20 ϩ34.2 (c 0.0152 in
MeOH) (Found: C, 69.9; H, 9.1; N, 3.7. C22H35NO4 requires C,
70.0; H, 9.3; N, 3.7%); νmax (film)/cmϪ1 3350, 1732; δH (CDCl3)
7.3–7.17 (5H, m, ArH), 3.70 (1H, q, J 6.6, PhCHMe), 3.70 (3H,
s, CO2Me), 2.81–2.45 (4H, m, NCH2, CHCO2But and
CHCO2Me), 1.61–1.40 (3H, m, CHMe2, CHHCHMe2, NH),
1.34 (9H, s, But), 1.30 (3H, d, J 6.6, NCHMe), 1.08–0.97 (1H,
m, CHHCHMe2), 0.85 (6H, d, J 6.4, CHMe2); δC (CDCl3) 174.3
and 173.3 (2 × CO), 145.6 (quaternary Ar), 128.4, 121.9 and
126.6 (Ar), 80.7 (CMe3), 58.3 and 49.3 (CCO2But and
CCO2Me), 51.6 (PhCN), 47.3 (NCH2), 45.0 (CO2Me), 39.6
(CH2CHMe2), 27.9 (CMe3), 26.1, 24.4, 23.4 and 21.5 (CHMe2
and PhCHMe); [Found (CI): (M ϩ H)ϩ m/z 378.2637,
C22H36NO4 requires (M ϩ H)ϩ 378.2644].
1-tert-Butyl 4-methyl (2R,3R)-3-(aminomethyl)-2-isobutyl-
butanedioate, (1S)-10-camphorsulfonate salt 16
(1S)-10-Camphorsulfonic acid (168 mg, 0.722 mmol) was
added to a solution of the free amine 15 (197 mg, 0.722 mmol)
in ethyl acetate (6 ml) and the solvent allowed to evaporate
slowly over several days, affording white crystals of the (1S)-10-
camphorsulfonate salt 16, mp 107–109 ЊC; [α]D20 ϩ22.7 (c 0.0196
in MeOH) (Found: C, 56.9; H, 8.5; N, 2.8. C24H43NO8S requires
C, 57.0; H, 8.6; N, 2.8%); νmax (neat)/cmϪ1 3107, 1736, 1725,
1208, 1149; δH (CDCl3) (ammonium fragment) 7.74 (3H, br s,
NH3ϩ), 3.78 (3H, s, CO2Me), 3.35–3.11 and 3.02–2.94 (4H, m,
NCH2, CHCO2But and CHCO2Me), 1.77–1.60 (2H, m, CHMe2
and CHHCHMe2), 1.41 (9H, s, But), 1.36–1.24 (1H, m, CHH-
CHMe2), 0.93–0.89 (6H, m, CHMe2); [(1S)-10-camphor-
sulfonate fragment] 3.24 (1H, d, J 14.6), 2.77 (1H, d, J 14.6),
2.60–2.48 (1H, m), 2.35–2.24 (1H, m), 2.04–1.97 (2H, m), 1.90
(1H, d, J 18.2), 1.77–1.60 (1H, m), 1.41 (1H, obscured m), 1.05
(3H, s), 0.82 (3H, s); δC (CDCl3) 217.5, 172.8, 172.1, 81.5, 58.3,
52.5, 48.0, 47.3, 44.2, 43.9, 42.8, 42.6, 37.9, 37.5, 27.9, 26.9,
25.9, 24.5, 22.6, 22.0, 19.8, 19.7.
Minor diastereoisomer (by NMR difference): δH (CDCl3)
3.695 (CO2Me), 1.43 (But).
1-tert-Butyl 4-methyl (2R,3R)-3-(aminomethyl)-2-
isobutylbutanedioate 15
The Michael adduct 19 was subjected to the hydrogenolysis
conditions already described for 11/12, giving the primary
amine 15 (68%) after basification.
Crystal data for 16
C24H43NO8S, M = 505.67, monoclinic, space group P21,
a = 15.064 (2), b = 6.245 (1), c = 16.603 (2) Å, β = 115.99 (1)Њ,
V = 1403.9 D3 (by the least squares refinement of the setting
angles for 24 automatically centered reflections), Z = 2,
Dc = 1.20 g cmϪ3, F(000) = 548, µ = 13.5 cmϪ1. Crystal dimen-
sions 0.25 × 0.31 × 0.53 mm.
(2R,3R)- and (2R,3S)-1-tert-Butyl 4-methyl 2-isobutyl-3-
({[(1S)-1-phenylethyl]amino}methyl)butanedioate 21/22
The Michael adducts 21/22 were obtained in an analogous
fashion to 19/20 (350 mg, 93%; 88:12 mixture of diastereoi-
somers after chromatography), [α]D20 Ϫ37.6 (c 0.0205 in MeOH)
for an 88:12 mixture (Found: C, 69.7; H, 9.3; N, 3.6.
C22H35NO4 requires C, 70.0; H, 9.3; N, 3.7%); νmax (film)/cmϪ1
3350, 1731; δH (CDCl3) 7.34–7.17 (5H, m, ArH), 3.73 (1H, q,
J 6.6, PhCHN), 3.71 (3H, s, CO2Me), 2.82–2.48 (4H, m,
NCH2CH and CHCO2But), 1.65–1.37 (3H, m, CHMe2, CHH-
CHMe2 and NH), 1.32 (9H, s, But), 1.29 (3H, d, J 6.6,
NCHMe), 1.01 (1H, m, CHHCHMe2), 0.85 (6H, d, J 6.5,
CHMe2); δC (CDCl3) 174.2 and 173.2 (2 × CO), 145.3 (quater-
nary Ar), 128.4, 126.8, 126.4 (Ar), 80.8 (CMe3), 57.7 and 49.1
Data collection and processing. Enraf-Nonius CAD4 dif-
fractometer, graphite monochromated Cu-Kα radiation (λ =
1.54180 Å) ω-2θ scan mode with the
(0.81 ϩ 0.15tanθ)Њ; 4865 reflections measured (2 > θ > 70Њ, 0, h,
k, l), 2929 unique (merging R = 0.041), giving 2819 with
I > 3σ(I).
ω scan width
Structure analysis and refinement. Direct methods. Full-
matrix least-squares refinement with all non-hydrogen atoms in
anisotropic approximation (319 variables, observations/
variables = 8.8). All hydrogen atoms were located in the differ-
ence-Fourier maps and included in the final refinement with
† CCDC reference 207/375.
J. Chem. Soc., Perkin Trans. 1, 1999, 3603–3608
3607

View Article Online
(CCO2But and CCO2Me), 51.6 (PhCN), 47.1 (NCH2), 45.1
(CO2Me), 39.8 (CH2CHMe2), 27.9 (CMe3), 26.1, 24.7, 23.5 and
21.4 (CHMe2 and PhCHMe) [Found (CI): (M ϩ H)ϩ m/z
378.2643, C22H36NO4 requires (M ϩ H)ϩ 378.2644].
Minor diastereoisomer (by NMR difference): δH (CDCl3)
3.67 (CO2Me), 1.37 (But); δC (CDCl3) 80.6, 58.4, 48.2, 46.2,
44.3, 38.4, 26.2, 24.5, 21.5.
6 R. Hirayama, M. Yamamoto, T. Tsukida, K. Matsuo, Y. Obato,
F. Sakamoto and S. Ikeda, Bioorg. Med. Chem., 1997, 5, 765.
7 (a) For a review, see D. C. Cole, Tetrahedron, 1994, 50, 9517; (b)
S. Torii, T. Inokuchi and M. Kubota, J. Org. Chem., 1985, 50, 4157;
(c) D. H. R. Barton, A. Fekih and X. Lusinchi, Bull. Soc. Chim. Fr.,
1988, 4, 681; (d) S. E. Drewes, O. L. Njamela, N. D. Emslie,
N. Ramesar and J. S. Field, Synth. Commun., 1993, 23, 2807;
(e) G. Jenner, Tetrahedron Lett., 1995, 36, 233; ( f ) S. Kwiatkowski,
A. Jeganathan, T. Tobin and D. S. Watt, Synthesis, 1989, 946 (and
references therein).
8 (a) E. Juaristi, J. Escalante, B. Lamatsch and D. Seebach, J. Org.
Chem., 1992, 57, 2396; (b) J. M. Hawkins and G. C. Fu, J. Org.
Chem., 1986, 51, 2820.
9 (a) E. S. Stratford and R. W. Curly, J. Med. Chem., 1983, 26, 1463;
(b) A. Zilka, E. S. Rachman and J. Rivlin, J. Org. Chem., 1961, 26,
376.
10 P. Perlmutter and M. Tabone, Tetrahedron Lett., 1988, 29, 949.
11 (a) S. Patai and Z. Rappoport, in The Chemistry of Alkenes,
ed. S. Patai, Interscience, New York, 1964, ch. 8; (b) H. E.
Zimmerman, J. Org. Chem., 1955, 20, 549; (c) S. I. Suminov and A.
N. Kost, Russ. Chem. Rev., 1969, 38, 884.
12 For a discussion of kinetic vs. thermodynamic control, see Klumpp,
in Reactivity in Organic Synthesis, Wiley, New York, 1982, pp. 36.
13 S. Masamune, W. Choy, J. S. Peterson and L. R. Sita, Angew. Chem.,
Int. Ed. Engl., 1985, 24, 1.
14 J. S. Rollet, Computing Methods in Crystallography, Pergamon Press,
Oxford, 1965.
15 A. C. T. North, D. C. Philips and F. S. Mathews, Acta Crystallogr.,
Sect. A, 1968, 24, 351.
1-tert-Butyl 4-methyl (2R,3R)-3-(aminomethyl)-2-
isobutylbutanedioate 15
The 88:12 mixture of 21 and 22 was subjected to the hydro-
genolysis conditions already described for 11/12, giving an 88:12
mixture (93%), with the primary amine 15 as the major
diastereoisomer.
Acknowledgements
We would like to thank the DTI, EPSRC and British Biotech
Pharmaceuticals for funding R. S. T. under the Asymmetric
Synthesis LINK Scheme and the latter for allowing R. S. T. to
pursue these studies.
References
1 R. P. Beckett, Exp. Opin. Ther. Patents, 1996, 6, 1305.
2 E. Baramova and J.-M. Foidart, Cell Biol. Int., 1995, 19, 239.
3 C. Campion, A. H. Davidson, J. P. Dickens and M. J. Crimmin,
WO 90/05719, 1990; Chem. Abstr., 1990, 113, 212677.
4 J. P. Dickens, M. J. Crimmin and R. P. Beckett, WO 94/02447, 1994;
Chem. Abstr., 1994, 121, 180211.
5 (a) S. G. Davies and O. Ichihara, Tetrahedron: Asymmetry, 1991, 2,
183; (b) S. G. Davies and I. A. S. Walters, J. Chem. Soc., Perkin
Trans. 1, 1994, 1129; (c) S. G. Davies, O. Ichihara and I. A. S.
Walters, J. Chem. Soc., Perkin Trans. 1, 1994, 1141.
16 H. D. Flack, Acta Crystallogr., Sect. A, 1983, 39, 876.
17 D. J. Watkin, J. R. Carruthers and P. W. Betteridge, CRYSTALS user
guide, Chemical Crystallography Laboratory, University of Oxford,
1985.
18 International Tables for Crystallography, Kynoch Press,
Birmingham.
Paper 9/07060E
3608
J. Chem. Soc., Perkin Trans. 1, 1999, 3603–3608