Burkholderia cepacia lipase and activated β-lactams in β-dipeptide and β-amino amide synthesis

Article abstract of DOI:10.1016/j.tetasy.2008.07.017

The work describes fluorine-activated and N-Boc-activated β-lactams as acyl donors to N-nucleophiles in the presence of Burkholderia cepacia lipase (lipase PS-D). Fluorine activation at the β-lactam ring causes the ring to open in high enantioselectivity

Full text of DOI:10.1016/j.tetasy.2008.07.017

Tetrahedron: Asymmetry 19 (2008) 1857–1861
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Burkholderia cepacia lipase and activated b-lactams in b-dipeptide
and b-amino amide synthesis
*
Xiang-Guo Li, Maria Lähitie, Liisa T. Kanerva
Department of Pharmacology, Drug Development and Therapeutics/Laboratory of Synthetic Drug Chemistry and Department of Chemistry,
University of Turku, Lemminkaisenkatu 5C, FIN-20520 Turku, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
The work describes fluorine-activated and N-Boc-activated b-lactams as acyl donors to N-nucleophiles in
the presence of Burkholderia cepacia lipase (lipase PS-D). Fluorine activation at the b-lactam ring
causes the ring to open in high enantioselectivity and allows the preparation of b-dipeptides and b-amino
amides. In the case of N-Boc-activation, the chemical ring opening is significant. b-Dipeptide formation
can then be considerably enhanced by the presence of lipase PS-D and/or by temperature.
Ó 2008 Elsevier Ltd. All rights reserved.
Received 19 June 2008
Accepted 14 July 2008
Available online 9 August 2008
1. Introduction
do not split amide bonds. However, b-lactams lack the resonance
stabilization typical to normal peptide bonds, and this explains
b-Lactam (2-azetidinone) rings are essential parts of b-lactam
antibiotics. The ring makes the drug bactericidal due to the inacti-
vation of a bacterial transpeptidase enzyme (essential for bacterial
growth). The inactivation is caused by the N1–C2 opening of the
lactam ring when the active-site serine hydroxyl reacts with the
ring.1 In addition to this, the enantiomers of b-lactams are versatile
intermediates in various types of synthetic products, for example,
as acylating agents for coupling reactions with O- and N-nucleo-
philes.2,3 Accordingly, considerable interest has been focused on
the asymmetric synthesis and regioselective ring opening of b-lac-
tams.2–5 For the chemical ring opening with free alcohols and
why the active-site serine of lipases can accept b-lactams as acyl
donors in the formation of the so-called acyl-enzyme intermediate
(step 1, Scheme 1). This intermediate subsequently reacts with the
nucleophiles present. This ability has been widely used and thor-
oughly reviewed for the kinetic resolution of b-lactams through
hydrolysis with C. antarctica lipase
B (CAL-B) in organic
solvents.10,11
b-Amino acid residues are present in many natural compounds
and synthetic peptidomimetics in the forms of b-amino amides and
b-peptides.2,3,12,13 This motivated us to start studies of b-dipeptide
formation based on the use of ring and/or N-activated b-lactams
1–3 as acyl donors to b-amino esters (Fig. 1). Fluorine as an isostere
to hydrogen was used to activate the ring in compounds 1 and 2.
N-Activation was accomplished by Boc-protection, as Boc is a com-
mon N-protecting group in peptide chemistry. The work was
started with the lipase PS-D-catalyzed ring-opening studies using
the amonolysis and aminolysis of rac-1a as novel applications of
lipase catalysis (Scheme 1). Lipase PS-D was a natural choice as a
catalyst because it was previously applicable in the alcoholysis of
rac-1a in dry organic solvents and exposed the lactam ring to
hydrolysis much less than CAL-B.8 In dipeptide synthesis, tert-butyl
esters rather than the corresponding n-alkyl esters were used in
order to prevent the previously observed lipase PS-D-catalyzed
interesterification-type reaction.9a In mechanistic terms, the two
acyl donors (a b-lactam and a n-alkyl ester) of interesterification
form their own acyl-enzyme intermediates with the serine hydro-
xyl of the lipase (step 1, Scheme 1). The n-alcohol released from the
n-alkyl ester can then serve as a nucleophile to react with the
b-lactam-based acyl-enzyme intermediate and lead to the forma-
tion of the ester as a side product. Attention was paid to minimize
the enzymatic hydrolysis of the b-lactam ring by the water in the
a-amino esters, the work of Palomo et al. have shown the impor-
tance of using N-activated b-lactams in the presence of NaN3 or
KCN so that the ring-opening rate is promoted under neutral
conditions.2,3
Lipases are serine hydrolases and used as versatile biocatalysts
for the production of enantiopure compounds. The facts that
lipases readily work in organic solvents, are economical and cata-
lyze a wide variety of reactions other than the hydrolysis of triglyc-
erides, have made many new synthetic applications possible. We
have previously used Candida antarctica lipase A (CAL-A) for the
preparation of b-dipeptides through the N-acylation of a b-amino
ester with the 2,2,2-trifluoroethyl ester of another N-protected
b-amino acid.6 The enantioselective N1–C2 opening of a b-lactam
ring with alcohols in the presence of Burkholderia cepacia lipase
(lipase PS adsorbed on Celite7 or commercial lipase PS-D) repre-
sents another application.8,9 It is generally accepted that lipases
* Corresponding author. Tel.: +358 2 333 6773; fax: +358 2 333 7955.
E-mail address: lkanerva@utu.fi (L. T. Kanerva).
0957-4166/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2008.07.017

1858
X.-G. Li et al. / Tetrahedron: Asymmetry 19 (2008) 1857–1861
O
O
F
F
O
OH
E
O
F
F
R-NH2
OH
F
F
E
NHR
O
F
F
NH
HN
+
NH2
NH2
Step 1
E
1a
(R)-
rac-1a
Acyl-enzyme intermediate
(S)-8
i
i
Scheme 1. Mechanism for the lipase PS-D-catalyzed kinetic resolution of rac-1 with amines RNH2 (R = H, Pr, Bu or n-Bu).
Table 1
F
O
F
O
O
Lipase PS-D-catalyzed amonolysis and aminolysis of rac-1a (0.05 M) in TBME
F
Entry
Lipase PS-D
Nucleophile
(equiv)
Time
(h)
C (x)a
(%)
ee(R)-1a
(%)
ee(S)-8
(%)
(mg mLꢀ1
)
N
N
N
R1
R1
R1
1
2
3
4
5
6
7
8
30
20
30
40
40
40
30
30
NH3 (satd)
iPrNH2 (1)
iPrNH2 (1)
iPrNH2 (1)
iPrNH2 (2)
iPrNH2 (2)b
iBuNH2 (1)
n-BuNH2 (1)
7
10
7
5
5
2
7
7
50 (0)
>99
87
>99
>99
68
87
89
46
>99
>99
>99
>99
97
96
97
80
49 (31)
50 (12)
50 (12)
42 (4)
3a; R1=H
3b; R1=Boc
1
1
1a
2a
; R =H
; R =H
1b; R1=Boc
2b; R1=Boc
50 (3)
48 (—)c
37 (—)c
Figure 1. Studied racemic and enantiopure b-lactams.
a
Proportion of hydrolysis from the observed conversion.
Reaction in DIPE.
Proportion of hydrolysis not determined.
b
c
enzyme preparation and the formation of the amino acid as a side
product.
2. Results and discussion
and (S)-8 (R = H) in enantiopure forms at 50% conversion (entry
1). The hydrolysis of the lactam ring did not play a role as long
as TBME was initially saturated with ammonia. It is worth noting
that amonolysis became slow when ammonia was continuously
bubbled into the reaction mixture. Since ammonia easily evapo-
rates and special equipments were not available to control its con-
centration, we focused our attention on aminolysis studies. In the
case of aminolysis, water in the enzymatic system clearly com-
peted with the added isopropylamine when equimolar concentra-
tions of the amine to the b-lactam were used (entries 2–4).
However, both the unreacted (R)-1a and the amide (S)-8 (R = iPr)
formed were enantiopure at 50% conversion, indicating excellent
enantioselectivity for both hydrolysis and aminolysis. The propor-
tion of the hydrolysis (the value in the brackets) was clearly
reduced when the amount of lipase was increased although more
water in the enzyme preparation came to the system at the same
time (entries 2–4). This must mean that the enzyme content affects
the aminolysis (favoring aminolysis) more than the hydrolysis rate.
The proportion of hydrolysis (12%, entry 4), as determined by the
GC method, is in accordance with the isolated yield of (S)-8 (32%
as calculated from the racemate, the theoretical yield at 50%
conversion being 38%). The proportion of hydrolysis was further
reduced when the amount of the amine was increased from 1 to
2 equiv (entries 4 and 5). At the same time, a decrease in reactivity
was detected. The reaction in diisopropyl ether (DIPE) was most
2.1. Synthesis of compounds 1–7
Compounds rac-1a–3a and their enantiomers (ee = 99%)
were prepared by known methods from the corresponding
N-hydroxymethylated b-lactams and vinyl butanoate using chemo-
enzymatic pathways.9a,14 N-Boc protections with Boc2O in dichloro-
methane furnished (3R,4R)-2b and the enantiomers of 3b. Attempts
to prepare the enantiomers of 1b failed due to the lability caused by
the high activation stage. Enantiopure tert-butyl esters (R)-6 and
(S)-7 were obtained combining literature methods. Thus, the li-
pase-catalyzed kinetic resolution of the corresponding racemic
ethyl ester15 and b-lactam9b first afforded (R)-4 and (S)-5. The com-
pounds were hydrolyzed to the corresponding acids followed by
transformations into (R)-6 and (S)-7 in AcOtBu in the presence of
a catalytic amount of HClO4 (Scheme 2).16
2.2. Lipase PS-D-catalyzed amonolysis and aminolysis of rac-1a
The reaction of rac-1a as a novel acyl donor for ammonia and
various primary amines was first studied in the presence of lipase
PS-D in tert-butyl methyl ether (TBME) (Scheme 1). The results are
shown in Table 1. Highly enantioselective amonolysis in TBME sat-
urated with ammonia took place, giving the enantiomers (R)-1a
O
O
CO2Bu
Pr
N
H
H2N
O
O
1. AcOtBu
HClO4
(R)-4
(R)-6
HCl (18%)
reflux
or
or
O
2. K2CO3
Ph
O
NH
Ph
H2N
(S)-5
(S)-7
Scheme 2. Preparation of tert-butyl esters.

X.-G. Li et al. / Tetrahedron: Asymmetry 19 (2008) 1857–1861
1859
Table 2
effective (entry 6). The reactions of rac-1a with isobutylamine and
n-butylamine were less enantioselective (entries 7 and 8). In these
last-mentioned cases, the proportions of hydrolyses were not
quantified.
Reactions of b-lactams (0.05 M) with b-amino esters (0.1 M) in DIPE in the presence
and in the absence of lipase PS-D (40 mg mLꢀ1
)
Entry b-
Lactama
b-
Time b-
C
Yield
(%)
½ ꢁ
a 2D2
Amino (h)
Dipeptide (%)
(10ꢀ1 deg cm2 gꢀ1
)
estera
2.3. b-Dipeptide synthesis
1
(S)-1a
(R)-6
15
9
100 96
+28.8 (c 1.00,
CHCl3)
—
The next effort concerned b-dipeptide synthesis in DIPE. Enan-
tiopure b-lactams 1–3 served as acyl donors to enantiopure amino
esters 6 and 7 (Scheme 3). When enantiopure substrates are used,
the resulting dipeptide is a pure stereoisomer, as shown in Table 2.
We considered it important to look for new enzymatic methods for
the preparation of b-peptides as so far, only two methods exist: the
already mentioned CAL-A-catalyzed N-acylation of a b-amino es-
ter6 and N-acylation (and hydrolysis) with the previously found
b-peptidyl aminopeptidases from Sphingosinicella strains and
Ochrobacterum anthropi.13 The drawback of the first method is that
the size of the substituents at the acyl donor (2,2,2-trifluoroethyl
ester) is limited. This can be easily explained on the basis of the
previously published 3D-structure of CAL-A, where the acyl bind-
ing pocket is shown to be a deep and narrow tunnel.17 The benefits
of lipase catalysis, on the other hand, are the good availability of
commercial and relatively inexpensive lipase preparations, and
the fact that the normal peptide bond formed is not cleaved by
lipases.
In accordance with the excellent (S)-enantioselectivity toward
rac-1a (Table 1) and with the observed (R)-selectivity in the
N-acylation of b-amino esters,9a,15,18 (S)-1a was quantitatively
transformed into the b-dipeptide 9 (entry 1) in the presence of
(R)-6 and lipase PS-D in DIPE (Table 2). Due to the (S)-selective ring
opening, (R)-1a gave only traces of chemically formed dipeptide un-
der the same conditions (entry 2). The incorporation of (3S,4S)-2a
(the faster reacting enantiomer according to the previous methan-
olysis9a) and (R)-6 afforded a gum-like dipeptide with 51% chemical
yield (entry 3). At this point, the b-lactam was totally consumed by
the cooperative enzymatic actions of the dipeptide formation and
the hydrolysis of the lactam ring. For characterization, the formed
dipeptide was transformed into the N-Boc-protected 10. On the
other hand, the unactivated (R)-3a showed no reaction with (R)-6
under the enzymatic reaction conditions (entry 5), indicating the
need to activate b-lactams for lipase-catalyzed aminolysis. When
b-lactams were N-Boc-protected, chemical ring opening became
highly significant. Accordingly, even the less reactive enantiomers
(3R,4R)-2b and (S)-7 could be effectively transformed into the
dipeptide without lipase PS-D (entry 4). Similarly dipeptide 13
was chemically prepared from (S)-3b and (S)-7 (entry 8).
2
3
(R)-1a
(3S,4S)-
2a
(3R,4R)-
2b
(R)-6
(R)-6
24
18
—
7b
—
10c
100 51
+4.2 (c 0.75,
CHCl3)c
4
(S)-7
24
11d
100 97
+8.7 (c 1.00,
CHCl3)
—
+33.0 (c 1.05,
CHCl3)
+32.4 (c 1.05,
CHCl3)
5
6
(R)-3a
(R)-3b
(R)-6
(R)-6
66
144
—
0
50
—
47
12d
7
8
(R)-3b
(S)-3b
(R)-6
(S)-7
48
12e
13d
93
69
82
65
144
ꢀ15.0 (c 1.00,
CHCl3)
a
b
c
Ee 99% or higher.
Peptide formation corresponds to 7% conversion with and without lipase PS-D.
For the N-Boc protected 10.
d
e
No enzyme added.
Temperature 57 °C; enzyme content 60 mg mLꢀ1
.
110
100
90
80
R
70
60
50
40
30
20
10
0
-10
-20
0
20
40
60
80
100
120
140
160
Time (h)
Figure 2. Progression curves for the disappearance of (R)-3b in its reaction with
(R)-6 in DIPE in the absence [(s) at 23 °C, (h) at 47 °C and ( ) at 57 °C] and in the
) at 57 °C,
D
presence [(d) at 23 °C and (j) at 47 °C, 40 mg mLꢀ1; (r) at 47 °C and (
N
60 mg mLꢀ1] of lipase PS-D; (ꢂ) stability of (R)-3b in DIPE in the presence of lipase
PS-D (40 mg mLꢀ1) at 23 °C.
As the N-Boc-protection exposes the b-lactam ring to chemical
ring opening, it became necessary to study the enzymatic versus
chemical b-dipeptide formation in more detail. Accordingly, the
reaction of (R)-3b with (R)-6 was studied in the presence (filled
signs) and in the absence (open signs) of lipase PS-D in DIPE at dif-
ferent temperatures (Fig. 2). The progression curves for the disap-
pearance of (R)-3b are shown in Figure 2. Increasing the
temperature from 23 °C to 57 °C had a significant effect on the
chemical reactivities; the b-lactam (D) completely reacted to form
12 in five days at 57 °C. In accordance with the curve (s), the pre-
parative scale formation of 12 reached only 50% conversion after
six days at room temperature (Table 2, entry 6). Since the enzymatic
R2
O
R3
R4
O
R2 R3
H
N
H
N
lipase PS-D
DIPE
N
+
O
R1
R1
H2N
O
O
R4
O
(R)-6; R4=Me
(S)-7; R4=Bn
(S)-1a
(3S,4S)-2a, (3R,4R)-2b
9-13
β-dipeptide
3b
(S)- or (R)-
3b
Scheme 3. Lipase PS-D-catalyzed synthesis of b-dipeptides in DIPE.

1860
X.-G. Li et al. / Tetrahedron: Asymmetry 19 (2008) 1857–1861
reaction is always accompanied by extensive chemical reaction,
temperature effects on enzymatic reactivities stay largely unsolved
in the present work. However, it is clear that the presence of lipase
PS-D greatly favors the conversion attained after a certain time.
Thus, 92% of (R)-3b (N) had reacted after two days in the presence
of the enzyme (60 mg mLꢀ1) at 57 °C, while the same conversion
was attained chemically in three days in the absence of lipase PS-
gress of the reaction was followed by taking samples from the
reaction mixture at intervals and analyzing them by chiral HPLC
on a CHIRACEL-OD column (0.46 ꢂ 25 cm) in the case of b-dipep-
tides and chiral GC on a Chrompack CP-Chirasil-DEX CB column
in the case of amonolysis and aminolysis of rac-1.
4.2. Preparation of (3R,4R)-2b and the enantiomers of 3b
D (D). This is in accordance with the preparative scale synthesis
of 12 in the presence of lipase PS-D (Table 2, entry 7). As shown
for the reaction at 47 °C, increase in the enzyme content from none
(h) through 40 mg mLꢀ1 (j) to 60 mg mLꢀ1 (ꢀ) also gave reactivity
enhancements. Thus, the reaction without the enzyme consumed
70% of (R)-3b in four days, while only 48 and 40 h were required
to reach the same conversion in the presence of 40 and 60 mg mLꢀ1
of the enzyme, respectively. When (R)-3b (ꢂ) together with lipase
PS-D (40 mg mLꢀ1) was incubated in DIPE without an added nucle-
ophile, the b-lactam concentration was unchanged, indicating sta-
bility of the lactam ring against hydrolysis by the water in the
enzyme preparation. KCN and NaN3 were previously observed to
enhance the ring opening of b-lactams under neutral conditions.2,3
It occurred to us that lipase PS-D catalysis in the place of KCN or
NaN3 also turned the ring opening more effective.
4-Dimethylaminopyridine (DMAP, 2 mg, 0.02 mmol) was added
to a solution of (3R,4R)-2a (28 mg, 0.17 mmol, ee = 99%) and Boc2O
(74 mg, 0.34 mmol) in dry dichloromethane (1 mL). After 55 min,
the solvent was evaporated off, and the residue was purified on sil-
ica gel eluting with ethyl acetate in petroleum ether (1:30, v/v) to
afford (3R,4R)-2b as a solid product (36 mg, 0.14 mmol, mp 96–
97 °C, ½a 2D2
ꢁ
¼ ꢀ76:0 (c 0.50, CHCl3)) in 80% yield; HRMS: M+ found
(M+ calculated for C14H16FNO3) 265.112500 (265.111422); MS: m/z
(relative intensity) 265 (0.02), 192 (15), 164 (3), 122 (100), 96 (3),
77 (3); 1H NMR (500 MHz, CDCl3) d 1.38 (s, 9H), 5.00 (dd,
J = 13.0 Hz, 2.0 Hz, 1H), 5.25 (dd, J = 53.3 Hz, 1.8 Hz, 1H), 7.30–
7.44 (m, 5H).
(R)-3b ½a 2D2
ꢁ
¼ þ114:0 (c 0.48, CHCl3) was prepared from (R)-3a
(ee = 99%) in 97% yield and (S)-3b
½ ꢁ
a 2D2
¼ ꢀ115:0 (c 0.48,
CHCl3) from (S)-3a (ee = 99%) in 86% yield as described above;
HRMS: M+ found (M+ calculated for C14H17NO3) 247.121500
(247.120844); MS: m/z (relative intensity) 247 (3), 192 (3), 174
(5), 132 (32), 104 (100), 77 (10); 1H NMR (500 MHz, CDCl3) d
1.38 (s, 9H), 2.92 (dd, J = 16.0 Hz, 3.0 Hz, 1H), 3.43 (dd, J =
16.0 Hz, 6.0 Hz, 1H), 4.92 (dd, J = 6.0, 3.5 Hz, 1H), 7.27–7.40 (m,
5H).
3. Conclusion
We have shown that fluorine activation facilitates the enantio-
selective lipase-catalyzed b-dipeptide and b-amino amide
formation in novel reactions where b-lactams serve as acyl donors
to N-nucleophiles in lipase PS-D-catalyzed transformations
under non-aqueous conditions. As a consequence, a novel lipase-
catalyzed b-dipeptide formation is possible, provided that ring-
activation, N-activation or both are used. Accordingly, the capacity
to open the ring progressively decreases, while going from difluo-
rinated 1a through monofluorinated 2a and finally to unreactive
3a. Interestingly, N-Boc activation exposes the b-lactam ring to
strong chemical ring opening with N-nucleophiles, and the process
can be considerably enhanced by the presence of lipase PS-D and/
or by increasing the temperature.
4.3. Preparation of (R)-6 and (S)-7
A solution of (R)-4 (1.11 g, 4.85 mmol, ee >99%) in aqueous HCl
(18%) was refluxed for 16 h before the solvent was evaporated off,
affording the crude acid which was used without further purifica-
tion. Next, HClO4 (70%, 7.28 mmol, 631 lL) was added to the sus-
pension of the acid in AcOtBu (12 mL). The reaction mixture was
stirred at room temperature for 24 h, yielding (R)-6 (270 mg,
1.70 mmol, ee > 99%, ½a D22
¼ ꢀ24.8 (c 1.00, CHCl3)) in 35% yield
ꢁ
over the two steps. (R)-6 was further purified on silica gel eluting
with methanol (5% v/v) in dichloromethane as an eluent. 1H NMR
(500 MHz, CDCl3) d 1.11 (d, J = 6.5 Hz, 3H), 1.46 (s, 9H), 1.68 (br
s, 2H), 2.21 (dd, J = 15.5, 8.0 Hz, 1H), 2.33 (dd, J = 15.5, 4.5 Hz,
1H), 3.34 (m, 1H); 13C NMR (126 MHz, CDCl3) d 23.38, 28.13,
44.19, 45.48, 80.50, 171.82.
4. Experimental
4.1. Materials and methods
TBME (water content 25 ppm according to Karl Fischer titra-
tion) and DIPE (water content 28 ppm) were of the highest analyt-
ical grade and were stored over molecular sieves (4 Å). Lipase PS-D
from B. cepacia was purchased from Amano Europe, England. Pre-
parative chromatographic separations were performed by column
(S)-7 ½a 2D2
¼ þ3:1 (c 1.00, CHCl3) was prepared from (S)-5
ꢁ
(ee = 99%) in 61% yield with mp 30–31 °C; MS: m/z (relative inten-
sity) 236 (M++1, 0.22), 191 (8), 144 (29), 120 (37), 88 (100); 1H
NMR (500 MHz, CDCl3) d 1.46 (s, 9H), 1.93 (br s, 2H), 2.28 (dd,
J = 16.0, 8.5 Hz, 1H), 2.43 (dd, J = 16.0, 4.0 Hz, 1H), 2.64 (dd,
J = 13.0, 8.0 Hz, 1H), 2.77 (dd, J = 13.0, 5.5 Hz, 1H), 3.45 (m, 1H),
7.20–7.33 (m, 5H); 13C NMR (126 MHz, CDCl3) d 28.15, 42.67,
43.52, 49.77, 80.79, 126.52, 128.56, 129.35, 138.54, 171.79.
chromatography on Merck Kieselgel 60 (0.063–0.200 lm). TLC
was carried out with Merck Kieselgel 60F254 sheets. Unless stated
otherwise, all enzymatic reactions were performed at room tem-
perature (23 °C). Melting points were measured on a Sanyo instru-
ment at a heating rate of 2 °C. Optical rotations were determined
with a Perkin–Elmer polarimeter, and [
a]D values are given in units
4.4. Preparative-scale kinetic resolution of rac-1a with
isopropylamine
of 10ꢀ1 deg cm2 gꢀ1
.
The NMR spectra were recorded on a Bruker 500 spectrometer
with tetramethylsilane (TMS) as an internal standard for 1H and
13C NMRs, and with 2,2,2-trifluoroethanol as an internal standard
for the 19F NMR. 1H–1H COSY, 1H–13C HQSC, and 1H–13C HMBC
spectra were used for the assignment of the chemical shifts when
Compound rac-1a (200 mg, 1.09 mmol) and lipase PS-D
(880 mg) were added to the solution of iPrNH2 (64 mg, 1.09 mmol)
in TBME (22 mL). The reaction was stopped by filtering off the en-
zyme at 50% conversion after 6 h. The solvent was evaporated off
and the residue was purified on silica gel eluting with 40% ethyl
acetate in petroleum ether to afford (R)-1a [93 mg, yield 46%,
ee(R)-1a >99%] and (S)-8 (R = iPr) as a solid product [83 mg, 32%,
necessary. Mass spectra were taken on
spectrometer.
a VG 7070E mass
In a typical small-scale experiment, lipase PS-D was added to
one of the b-lactams (0.05 M) in a TBME or DIPE in the presence
of ammonia, an amine, or a b-amino ester (0.05–0.1 M). The pro-
ee(S)-8 >99%, ½a 2D2
ꢁ
¼ þ4:2 (c 1.00, CHCl3)], mp 91–92 °C; HRMS: M+
found (M+ calculated for C12H16F2N2O) 242.122700 (242.123070);

X.-G. Li et al. / Tetrahedron: Asymmetry 19 (2008) 1857–1861
1861
MS: m/z (relative intensity) 242 (0.24), 222 (15), 164 (5), 146 (6),
106 (100), 79 (13), 43 (6); 1H NMR (500 MHz, CDCl3) d 0.95 (d,
3H, J = 6.5 Hz, CH3), 1.08 (d, 3H, J = 6.5 Hz, CH3), 2.00 (br s, 2H,
NH2), 3.97 (m, 1H, CH (CH3)2), 4.59 (t, 1H, J = 13.5 Hz, CHNH2),
6.04 (br s, 1H, NHCO), 7.31–7.38 (m, 5 arom. H); 13C NMR
(126 MHz, CDCl3) d 22.07 (CH3), 22.19 (CH3), 41.62 (CH(CH3)2),
57.87 (t, J = 23.9 Hz, CHNH2), 116.35 (t, J = 258.3 Hz, CF2), 128.03,
128.50, 128.54, 136.26 (arom. C), 162.61 (t, J = 28.9 Hz, CO); 19F
128.15, 128.40, 128.55, 129.17, 136.10, 137.35, 154.67, 165.77 (d,
J = 18.9 Hz), 170.55.
4.8. Preparation of b-dipeptide 12
Lipase PS-D (60 mg mLꢀ1) was added to a solution of (R)-3b
(31 mg, 0.13 mmol, ee = 99%) and (R)-6 (40 mg, 0.25 mmol,
ee = 99%) in dry DIPE (2.5 mL) and was stirred at 57 °C for 48 h be-
fore the work-up as above, except that ethyl acetate in petroleum
ether (1:3, v/v) was used as an eluent to afford dipeptide 12
NMR (471 MHz, CDCl3)
ꢀ117.38 (dd, J = 247.0, 11.8 Hz).
d
ꢀ116.39 (dd, J = 247.0, 14.1 Hz),
{43 mg, 0.11 mmol, mp 140–141 °C, ½a D22
¼ þ32:4 (c 1.05, CHCl3)}
ꢁ
in 82% yield; HRMS: M+ found (M+ calculated for C22H34N2O5)
406.247800 (406.246773); MS: m/z (relative intensity) 406 (8),
294 (65), 249 (36), 119 (62), 106 (74), 57 (100); 1H NMR
(500 MHz, CDCl3) d 0.97 (d, J = 6.5 Hz, 3H), 1.42 (s, 18H), 2.28 (d,
J = 5.5 Hz, 2H), 2.54 (dd, J = 14.0 Hz, 5.5 Hz, 1H), 2.68 (m, 1H),
4.19 (m, 1H), 5.01 (m, 1H), 6.03 (d, J = 8.0 Hz, 1H), 6.26 (br s, 1H),
7.20–7.35 (m, 5H).
4.5. Preparation of b-dipeptide 9
Lipase PS-D (40 mg mLꢀ1) was added to a solution of (S)-1a
(16 mg, 0.08 mmol, ee = 99%) and (R)-6 (24 mg, 0.15 mmol,
ee = 99%) in dry DIPE (1.5 mL). The reaction was stopped after
15 h by filtering off the enzyme. The solid part was washed with
dichloromethane (4 mL) and the filtrate was concentrated under
vacuum. The residue was purified on silica gel eluting with ethyl
acetate in petroleum ether (1:1, v/v) to afford dipeptide
9
4.9. Non-enzymatic preparation of b-dipeptide 13
{25 mg, 0.07 mmol, mp 100–101 °C, ½a D22
¼ þ28:8 (c 1.00, CHCl3)}
ꢁ
in 96% yield; HRMS: M+ found (M+ calculated for C17H24F2N2O3)
342.174300 (342.175499); MS: m/z (relative intensity) 342 (0.7),
285 (4), 191 (7), 106 (100); 1H NMR (500 MHz, CDCl3) d 1.13 (d,
J = 6.5 Hz, 3H, CH3CH), 1.42 (s, 9H, (CH3)3), 2.12 (dd, J = 15.8 Hz,
5.5 Hz, 1H, CH2), 2.27 (dd, J = 15.8 Hz, 4.5 Hz, 1H, CH2), 4.19 (m,
1H, CH), 4.60 (m, 1H, CH), 7.05 (d, J = 8.0 Hz, 1H, CONH), 7.32–
7.40 (m, 5H).
Solution of (S)-3b (25 mg, 0.10 mmol, ee = 99%) and (S)-7
(47 mg, 0.20 mmol, ee = 99%) in dry DIPE (2.0 mL) was stirred
for 144 h, before the work-up as above, except that ethyl acetate
in petroleum ether (1:4, v/v) was used as an eluent to afford
dipeptide 13 (31 mg, 0.07 mmol, mp 139–140 °C, ½a D22
¼ ꢀ15.0
ꢁ
(c 1.00, CHCl3)) in 65% yield; HRMS: M+ found (M+ calculated
for C28H38N2O5) 482.278300 (482.278073); MS: m/z (relative
intensity) 482 (4), 409 (13), 353 (15), 279 (20), 192 (60), 88
(100); 1H NMR (500 MHz, CDCl3) d 1.42 (s, 9H), 1.43 (s, 9H),
2.17 (dd, J = 16.0 Hz, 5.5 Hz, 1H), 2.26 (dd, J = 16.0 Hz, 5.5 Hz,
1H), 2.53 (m, 2H), 2.69 (m, 2H), 4.33 (m, 1H), 5.04 (m, 1H),
6.98–7.33 (m, 10H).
4.6. Preparation of b-dipeptide 10
Lipase PS-D (40 mg mLꢀ1) was added to the solution of (3S,4S)-
2a (30 mg, 0.18 mmol, ee = 99%) and (R)-6 (58 mg, 0.36 mmol,
ee = 99%) in dry DIPE (3.6 mL). After 18 h, the reaction was worked
up as above to afford the gum-like dipeptide 10 (30 mg,
0.09 mmol) in 51% yield. In order to facilitate the characterization,
the free amino group of 10 was protected in the quantitative yield
in the presence of Boc2O (2 equiv) and NEt3 (2.5 equiv) in dry
dichloromethane. The data for N-Boc-protected 10 are as follows.
Acknowledgment
This work was supported by the Academy of Finland (Grant No.
109743 to L.K.)
½
a 2D2
ꢁ
¼ þ4:2 (c 0.75, CHCl3); HRMS: M+ found (M+ calculated for
References
C22H33FN2O5) 424.236800 (424.237351); MS: m/z (relative inten-
sity) 424 (0.5), 295 (32), 206 (43), 191 (24), 150 (81), 106 (77),
57 (100); 1H NMR (500 MHz, CDCl3) d 1.09 (d, J = 7.0 Hz, 3H),
1.38 (s, 9H), 1.43 (s, 9H), 1.66 (m, 1H), 2.10 (dd, J = 15.5 Hz,
4.0 Hz, 1H), 4.09 (m, 1H), 5.22 (d, J = 51.5 Hz, 1H), 5.26 (m, 1H),
7.27–7.33 (m, 5H).
1. Patrick, G. L. An Introduction to Medicinal Chemistry, 3rd ed.; Oxford University
Press, 2005. pp 391–393.
2. Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M. In Enantioselective Synthesis
of b-Amino Acids; Juriasti, E., Soloshonok, V., Eds., 2nd ed.; Wiley-VHC: New
York, 2005; p 477.
3. Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M. Synlett 2001, 1813–1826.
4. Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M. Eur. J. Org. Chem. 1999,
3223–3235.
5. Benaglia, M.; Cinquini, M.; Cozzi, F. Eur. J. Org. Chem. 2000, 563–572.
6. Li, X.-G.; Kanerva, L. T. Org. Lett. 2006, 8, 5593–5596.
4.7. Non-enzymatic preparation of b-dipeptide 11
7. Sundholm, O.; Kanerva, L. T. J. Chem. Soc., Perkin Trans. 1 1993, 2407–2410.
8. Brieva, R.; Crich, J. Z.; Sih, C. J. J. Org. Chem. 1993, 58, 1068–1075.
9. (a) Li, X.-G.; Lähitie, M.; Päiviö, M.; Kanerva, L. T. Tetrahedron: Asymmetry 2007,
18, 1567–1573; (b) Li, X.-G.; Kanerva, L. T. Adv. Synth. Catal. 2006, 348, 197–205.
10. Forró, E.; Fülöp, F. Mini-Rev. Org. Chem. 2004, 1, 93–102.
11. Liljeblad, A.; Kanerva, L. T. Tetrahedron 2006, 62, 5831–5854.
12. Seebach, D.; Beck, A. K.; Bierbaum, D. J. Chem. Biodivers. 2004, 1, 1111–1239.
13. Heck, T.; Kohler, H.-P. E.; Limbach, M.; Flogel, O.; Seebach, D.; Geueke, B. Chem.
Biodivers. 2007, 4, 2016–2030.
14. Li, X.-G.; Kanerva, L. T. Tetrahedron: Asymmetry 2007, 18, 2468–2472.
15. Gedey, S.; Liljeblad, A.; Fülöp, F.; Kanerva, L. T. Tetrahedron: Asymmetry 1999,
10, 2573–2581.
16. Hu, J.; Miller, M. J. J. Am. Chem. Soc. 1997, 119, 3462–3468.
17. Ericsson, D. J.; Kasrayan, A.; Johansson, P.; Bergfors, T.; Sandström, A. G.;
Bäckvall, J.-E.; Mowbray, S. L. J. Mol. Biol. 2008, 376, 109–119.
18. Kanerva, L. T.; Csomós, P.; Sundholm, O.; Bernáth, G.; Fülöp, F. Tetrahedron:
Asymmetry 1996, 7, 1705–1716.
(S)-7 (38 mg, 0.16 mmol, ee = 99%) was added to a solution of
(3R,4R)-2b (21 mg, 0.08 mmol, ee = 99%) in dry DIPE (1.6 mL).
The reaction was worked up after 24 h as above except that ethyl
acetate in petroleum ether (1:8, v/v) was used as an eluent to
afford solid 11 {39 mg, 0.08 mmol, mp 136–137 °C, ½a D22
¼ þ8:7
ꢁ
(c 1.00, CHCl3)} in 97% yield; HRMS: M+ found (M+ calculated for
C28H37FN2O5) 500.270000 (500.268651); MS: m/z (relative inten-
sity) 500 (1), 409 (11), 297 (35), 191 (46), 106 (55), 88 (86), 57
(100); 1H NMR (500 MHz, CDCl3) d 1.43 (s, 18H), 1.64 (m, 1H),
2.04 (dd, J = 16.0 Hz, 4.0 Hz, 1H), 2.74 (m, 2H), 4.21 (m, 1H), 5.19
(d, J = 50.1 Hz, 1H), 5.23 (m, 1H), 7.08–7.27 (m, 10H); 13C NMR
(126 MHz, CDCl3) d 28.11, 28.32, 37.23, 39.32, 46.37, 55.98 (d,
J = 17.6 Hz), 80.15, 81.34, 92.75 (d, J = 194.9 Hz), 126.70, 127.93,