Robust and straightforward chemo-enzymatic enantiopure dipeptide syntheses and diketopiperazines thereof

Article abstract of DOI:10.1016/j.tetasy.2014.04.005

We have explored the scope of the synthetic route towards d-phenylglycyl diketopiperazines, involving a penicillin acylase catalysed formation of d-phenylglycyl dipeptides of l-amino acids with functional groups in the side chain. The synthesis of dipeptides from serine, threonine, glutamic acid, glutamine and methionine was successful. In contrast, aspartic acid, asparagine and cysteine only afforded trace amounts of dipeptides while no dipeptide was detected with arginine, lysine and tyrosine. Isolated dipeptide yields varied from 10% to 76%. The dipeptides were successfully converted into their corresponding enantiopure diketopiperazines by chemical esterification and cyclization under alkaline conditions, in 35-43% yield. In the case of glutamic acid, the procedure yielded the diketopiperazine with an esterified side chain. Remarkably with glutamine, the amide function in the side chain was transformed into an ester moiety, resulting in the same diketopiperazine as with glutamic acid.

Full text of DOI:10.1016/j.tetasy.2014.04.005

Tetrahedron: Asymmetry 25 (2014) 825–832
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Robust and straightforward chemo-enzymatic enantiopure dipeptide
syntheses and diketopiperazines thereof
⇑
Pedro C. Pereira, Isabel W. C. E. Arends, Roger A. Sheldon
Section of Biocatalysis and Organic Chemistry, Department of Biotechnology, Delft University of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
We have explored the scope of the synthetic route towards
penicillin acylase catalysed formation of D-phenylglycyl dipeptides of L-amino acids with functional
D
-phenylglycyl diketopiperazines, involving a
Received 6 March 2014
Accepted 3 April 2014
Available online 29 May 2014
groups in the side chain. The synthesis of dipeptides from serine, threonine, glutamic acid, glutamine
and methionine was successful. In contrast, aspartic acid, asparagine and cysteine only afforded trace
amounts of dipeptides while no dipeptide was detected with arginine, lysine and tyrosine. Isolated
dipeptide yields varied from 10% to 76%. The dipeptides were successfully converted into their
corresponding enantiopure diketopiperazines by chemical esterification and cyclization under alkaline
conditions, in 35–43% yield. In the case of glutamic acid, the procedure yielded the diketopiperazine with
an esterified side chain. Remarkably with glutamine, the amide function in the side chain was
transformed into an ester moiety, resulting in the same diketopiperazine as with glutamic acid.
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
conditions. Unsurprisingly, steps towards the environmentally
benign syntheses of peptides have been reported. Penicillin G
Diketopiperazines are cyclic dipeptides and constitute an
important family of biologically active compounds that have
attracted growing interest in several areas of research.1,2 They
are regarded as useful synthons for active pharmaceutical ingredi-
ents while histidine containing diketopiperazines are studied in
the development of therapeutic agents.3–6 Moreover, these
dipeptides have been shown to possess a wide range of biological
properties (antitumor, antiviral, antifungal and antibacterial), and
have been linked with prebiotic activity while displaying enantio-
selective catalytic activity (Fig. 1).7–12
The goal of mimicking enzymatic activity with amino acids and
peptides is not recent and important progress has been made in the
development of asymmetric reactions with amino acid and
peptide-based catalysts.12–14 Another relevant example of both
the growing interest in diketopiperazines as well as in developing
biomimetic peptides is the peptide-catalysed synthesis of diketopi-
perazines and dipeptides. This makes diketopiperazines interesting
for research, as well as showing the need for the development of
new synthetic methods and strategies to enable access to a larger
variety of such compounds.15,16
acylase, for example, is an efficient biocatalyst used in the synthe-
sis of b-lactam antibiotics and has also been shown to be effective
in the synthesis of chiral dipeptides.17–19
Chemo-enzymatic routes to diketopiperazines containing the D-
phenylglycine moiety have been described, either starting from
enantiopure natural amino acids, or from inexpensive racemic
mixtures in the case of unnatural amino acids.20,21 In the latter case
it was shown that when using a racemate, there was no interfer-
ence with either the formation or isolation of the desired pure
compound. In addition, expanding the scope of the procedure
resulted in a green, efficient and inexpensive way of obtaining
diketopiperazines containing unnatural amino acids with known
anti-viral activity on a preparative scale.20,22
The coupling of D-phenylglycine amide with the L-enantiomers
of glycine, alanine, valine, leucine, isoleucine, phenylalanine, tryp-
tophan and histidine has already been successfully performed.20
The route using penicillin G acylase provides a green and straight-
forward approach to the synthesis of dipeptides comprising
D-phenylglycine as well as their cyclic diketopiperazine deriva-
tives. The yields described are within the range of 34–70%, which
combined with room temperature, normal pressure and an aque-
ous system, make it a powerful tool in the synthesis of enantiopure
dipeptides. In order to obtain enantiopure diketopiperazines, the
linear dipeptide undergoes esterification by treatment with thionyl
chloride in methanol, followed by alkaline cyclization at room
The use of biocatalysts in organic synthesis is widespread,
mainly due to their main features: high selectivity and mild
⇑
Corresponding author. Tel.: +31 15 7600306.
E-mail address: r.a.sheldon@tudelft.nl (R.A. Sheldon).
http://dx.doi.org/10.1016/j.tetasy.2014.04.005
0957-4166/Ó 2014 Elsevier Ltd. All rights reserved.

826
P. C. Pereira et al. / Tetrahedron: Asymmetry 25 (2014) 825–832
H
N
H
N
O
O
H
N
NH2
O
N
H
O
N
H
NH
NH
N
1-{3-[2S,5S)-5-benzyl-3,6-dioxo
(3S,6S)-3-benzyl-6-(1H-imidazol
piperazin-2-yl]propyl}guanidine
-5-ylmethyl)piperazine-2,5-dione
catalyst, anti tumor
O
catalyst
H
N
O
H
N
O
O
N
H
O
N
H
OH
(3Z,6Z)-3-benzylidene-6-(4-methoxy
(3S,6S)-3-benzyl-6-(4-hydroxy
benzylidene)piperazine-2,5-dione
benzyl)piperazine-2,5-dione
anti tumor
PAI-1 Inhibitor
N
H
HN
N
N
O
H
N
O
O
O
N
3-(1-imidazol-5-ylmethyl)hexahydropyrrolo
3-(2-methylpropyl)hexahydropyrrolo[1,2-a]
pyrazine-1,4-dione
[1,2-a]pyrazine-1,4-dione
anti microbial
anti hyperglycemic
Figure 1. Examples of diketopiperazines and their functions.
temperature, after which the product formed precipitates from
solution.
in isolated yields is greater than that observed with amino acids
with non-functionalized side-chains.20 This suggests that the pres-
ence of a functional group in the side chain of one of the reactants
can interfere with the reaction.
We observed that an increase in the initial amino acid concen-
tration, starting from 200 mM, led to an improvement in purity
with serine, threonine, glutamine and methionine. In these cases,
To explore the full potential of the penicillin G route towards
the synthesis of enantiopure diketopiperazines, we decided to
screen all natural amino acids that were not yet reported (Fig. 2).
Herein we report the results obtained with the remaining amino
acids, which comprise the interesting remainder of the polar and
the charged amino acids. Proline was not tested, because
secondary amines are not accepted as a nucleophile by penicillin
G acylase. Bearing in mind the known broad selectivity of the
enzyme towards nucleophiles and the reactivity of tryptophan
and histidine in previous work, neither the volume nor the pres-
ence of a basic group in the residue was expected to constitute
as a hindrance for the procedure.20,23
this leads to a decreasing amount of the side product, D-phenylgly-
cine. This was confirmed through NMR analysis of the obtained
crude products. We concluded that, in these cases, hydrolysis of
phenylglycine amide to phenylglycine decreases with increasing
concentrations of the starting amino acid.
This was most notable in the case of the glutamine containing
dipeptide, as we could only obtain pure dipeptide when perform-
ing the reaction with an initial concentration of 500 mM. On the
other hand, we also observed that in the cases of
L-serine, L-threo-
2. Results and discussion
nine and -methionine, an increase in the initial concentration to
L
500 mM afforded an impure product. Further studies and optimi-
zation were not carried out.
2.1. Enzymatic coupling reactions
The amino acids aspartic acid, asparagine and cysteine were
tested without success. It was however possible to detect trace
amounts of the formed dipeptide through mass spectrometry in
The proposed penicillin G acylase coupling method was suc-
cessful in the coupling of the
L-amino acids serine, threonine,
glutamic acid, glutamine and methionine (Table 1). The variation

P. C. Pereira et al. / Tetrahedron: Asymmetry 25 (2014) 825–832
827
NH2
O
penicillin acylase
O
H2N
OH
+
pH 9,7 / RT
NH2
R
D-phenylglycine
amide
L-amino acid
NH2
O
H
N
OH
R:
O
R
-CH2OH
Ser
Thr
Asp
D-PG-L-amino acid
dipeptide
-CH2OHCH3
-CH2COO
Glu
-CH2CH2COOH
-CH2CONH2
-CH2CH2CONH2
-CH2C6H4OH
-CH2SH
i) SOCl2 / MeOH
ii) OH- / H2O / MeOH
Asn
Gln
Tyr
Cys
H
N
O
R
-CH2CH2SCH3
-(CH2)4NH2
Met
Lys
Arg
O
N
H
-(CH2)3NHCNHNH2
D
L
-phenylglycine- -amino acid
diketopiperazine
Figure 2. Proposed penicillin G acylase catalysed peptide synthesis with natural amino acids possessing chemical functionalities.
Table 1
solubility of
L-lysine and L-arginine might have made it impossible
Coupling reaction isolated yields and amino acid initial concentrations used (in the
successful cases)
to determine if there was product formation, because no precipita-
tion occurred at the dipeptide’s isoelectric point.
was obtained when lowering the pH.
D-Phenylglycine
Dipeptide
Isolated yield (%)
Initial concentration (mM)
The yields obtained varied considerably according to the side
chain of the starting amino acid, not necessarily showing a depen-
dency on the chemical function (Table 2). Previously, the nature of
the side chain was found to influence the enzyme selectivity
equally, that is, non-branched alkyl side-chains produced nucleo-
philes that were more effective.20,21,23
10
41
260
300
200
500
300
—
D-PG-
D-PG-
D-PG-
D-PG-
D-PG-
D-PG-
D-PG-
D-PG-
D-PG-
D-PG-
D-PG-
L-Ser
L-Thr
L-Glu
L-Gln
L-Met
L-Asp
L-Asn
L-Cys
L-Arg
L-Lys
L-Tyr
73
33
53
Traces
Traces
Traces
—
—
—
Table 2
—
Comparative apparent residue preference
—
—
Side-chain
Yield (%)
Side-chain
Yield (%)
10
—
—
OH
OH
42
>
>
these three cases. This indicates that the coupling reaction did
occur to some extent, although it did not make it feasible for syn-
thetic purposes without further studies.
OH
OH
73
Traces
O
O
In the case of cysteine, trace amounts of phenylglycine were
found with mass spectrometry, together with data fitting cystine,
the corresponding disulfide. However, this could not be fully con-
firmed by the NMR data. The NMR profile obtained matched that
described but the chemical shifts did not.24
NH2
NH2
33
53
>
>
Traces
Traces
O
S
O
SH
Previous work indicated the affinity of
acid and -asparagine with the enzyme to be comparable, in acyl
transfer reactions where these species act as acyl acceptors.23 How-
ever, we were able to obtain a dipeptide with both -glutamic acid
and -glutamine but not with -aspartic acid and -asparagine.
In the trials with -arginine, -lysine and -tyrosine, no evidence
of the desired dipeptide was found, although the specificity of the
enzyme to -arginine was reported to be higher than the one of
-glutamic acid.23 We did detect the side product
-phenylglycine,
which indicates that there was no enzyme inactivation. The high
L-aspartic acid, L-glutamic
L
We successfully reproduced the results obtained by Khimiuk
et al.20 in the case of phenylalanine and tryptophan, where the res-
idue is bulky. This suggests that the volume of the residue should
not be an influential variable. We were, however, unable to convert
L
L
L
L
L
L
L
L-tyrosine into a dipeptide. This might have been due to tyrosine’s
low solubility.
L
The data regarding the pair aspartic or glutamic acid and the
pair asparagine or glutamine, suggest amino acids with two
L
D

828
P. C. Pereira et al. / Tetrahedron: Asymmetry 25 (2014) 825–832
Table 3
methylene groups in the residue are preferred (Table 2), whereas
the equivalent amino acid with one methylene group less is not
accepted, or only formed in very small amounts.
Esterification and condensation reactions of isolated yields
Diketopiperazine
Isolated yield (%)
The isolated yield strongly depends on the dipeptide’s solubil-
ity, not necessarily reflecting the real yield, as already observed.20
However, due to the chemical similarity between this pair, one can
consider aspartic acid and glutamic acid dipeptides to have
Cyclo (
Cyclo (
Cyclo (
Cyclo (
D-PG,
D-PG,
D-PG,
D-PG,
L
L
L
L
-Ser)
-Thr)
-Met)
-Glu/Gln-OMe)
43
48
35
68
comparable solubilities and the same applies for
-asparagine. This difference in results could then be exclusively
attributed to the enzyme’s substrate preference.
The obtained yield for -threonine was four fold higher than the
one obtained with -serine. This also suggests that the decisive
L-glutamine and
L
L
Both
L
-glutamic acid and
L
-glutamine containing dipeptide
L
esters, afforded the same product. This was confirmed by infrared
analysis and was consistently obtained in several trials (Fig. 3).
This compound was obtained pure and was positively identified
by NMR and mass spectrometry as the corresponding diketopiper-
azine methyl ester. Rather than having the acid or amide functional
factor is again not the solubility but the compatibility with the cat-
alyst, since the solubility of both amino acids is very high.
The data regarding the sulfur containing amino acids
L-methio-
nine and -cysteine, although suggestive, cannot be used to draw
L
conclusions in this matter because both compounds differ not only
in side-chain length but also in chemical functionality.
Dipeptides were obtained within a period of 2 days, either by
single or sequential precipitation, with the exception of the
group, corresponding to
compound obtained had the same ester moiety in the side chain.
Isolated yields varied between 66% and 69%.
L-glutamic acid and L-glutamine, the
With L-glutamic acid, a possible explanation for this result is
L
-serine containing dipeptide, which was obtained after 4 days. In
that the ring closes before hydrolysis of the side chain’s
carboxylate ester, which then precipitates thus preventing any
further reaction. An amine intra-molecular attack on the residue’s
ester group and not at the C terminus of the dipeptide could
occur, which would yield an eight membered ring compound.
However, we found only one of these two possible products
(Fig. 4).
The explanation is still not clear in the case of glutamine
because one must account for the loss of an amide group. One pos-
sible explanation is that the amide group is subjected to alcoholy-
sis during esterification (Fig. 5). Although amides are stable bonds,
the low pH at which esterification occurs might cause substitution
of the amide group by the ester. However, the yields of diketopi-
this, and in the -phenylglycyl- -glutamic acid cases, significant
D
L
amounts of product could still be obtained by precipitation after
50 and 10 days, respectively. These steps may be optimized by
reducing the crystallization temperature, combined with either
an earlier second isolation step or with a single isolation step, as
a first isolation will necessarily translate into a larger second crys-
tallization period. In these reactions only fresh enzyme was used.
However, due to the known stability of this industrial catalyst,
we think the effective re-use should be feasible.25
2.2. Esterification and condensation reactions
The subsequent steps in the chemo-enzymatic diketopiperazine
syntheses are esterification with thionyl chloride in methanol, fol-
lowed by in situ aqueous alkaline cyclization. The resulting dipep-
tide was isolated by precipitation from solution. This procedure
allowed the diketopiperazine to be obtained pure and was success-
perazine methyl ester found in both L-glutamic acid and L-gluta-
mine are similar, although one equivalent of thionyl chloride is
used over the course of the esterification of the latter and two
equivalents in the former. This would presumably lead to a lower
yield in the case of L-glutamine, thus suggesting amide alcoholysis
ful in the cases of
L
-serine,
L-threonine and L-methionine (Table 3).
was not the cause for our results.
O
O
H
H
N
N
OH
O
H2N
OH
O
H2N
O
O
H2N
HO
i) SOCl2 / MeOH
ii) H2O, MeOH / pH 8.5
i) SOCl2 / MeOH
ii) H2O, MeOH / pH 8.5
H
N
O
O
O
N
H
O
found
H
N
H
O
O
N
O
O
O
N
H
N
H
O
NH2
OH
not found
not found
Figure 3. Same final product obtained in both cases is the corresponding diketopiperazine methyl ester and not the two expected corresponding diketopiperazines.

P. C. Pereira et al. / Tetrahedron: Asymmetry 25 (2014) 825–832
829
NH2
O
H
N
O
O
NH
O
O
HN
H2O / MeOH
pH 8.5
O
O
H2O / MeOH
pH 8.5
HN
HN
O
O
O
O
O
O
not found
found
Figure 4. Amine intramolecular attack affords a cyclic dipeptide methyl ester and not eight membered ring product.
NH2
NH2
H
N
O
O
OH
O
HN
HN
H2O / MeOH
pH 8.5
SOCl2
MeOH
O
O
O
O
N
H
O
O
O
H2N
O
O
Figure 5. Putative formation of
D-PG-
L-Gln methyl ester via amide esterification.
An alternative, perhaps more likely explanation, involves a
neighbouring group participation mechanism to account for the
conversion of the amide into an ester moiety. In this hypothesis,
this transformation takes place after ring-closure and not during
the esterification (Fig. 6).
the time required for the precipitation during the ring-closure step
was faster in the case of the -glutamine containing dipeptide than
in the case of the -glutamic acid containing dipeptide, thus
L
L
suggesting a different mechanism in the two cases.
H2N
H2N
O
O
O
HO
N
- NH3
N
O
O
HN
NH
O
NH
O
O
N
H
MeOH
H
N
N
OH
O
O
O
O
N
H
O
N
H
O
O
Figure 6. Proposed mechanism for the diketopiperazine ester formation with an enol intramolecular attack.
The hydroxyl group adjacent to the imine attacks the terminal
3. Conclusion
amide, affording a six membered lactone with a loss of ammonia.
Subsequent attack of methanol on the lactone carbonyl regener-
ates the open ring configuration, affording a methyl ester as indeed
we have obtained.
Attack of a water molecule and not of a methanol molecule is
unlikely, since no evidence of the formation of the diketopipera-
zine acid was found. In accordance with our results, and as
observed in the case of the glutamic acid containing dipeptide,
the ester presumably precipitates before hydrolysis of the side
chain ester can take place. In addition to this, we observed that
The coupling procedure is applicable in the case of L-serine,
-threonine, L-methionine and L-glutamine, thus leading to a pure
product without further purification. Data suggest there is room
for optimization. In some cases it was possible to improve the
product purity, simply by altering the initial concentration of
amino acid.
L
The complete procedure is valid in the synthesis of the corre-
sponding diketopiperazines, except in the case of
L-glutamine
and -glutamic acid. In these cases, the same compound was
L

830
P. C. Pereira et al. / Tetrahedron: Asymmetry 25 (2014) 825–832
obtained pure and it was identified as the corresponding ester of
the desired products. Two mechanisms have been proposed in
order to explain this result.
d = 0.852 (d, J = 6.3 Hz, CHOHCH3); 13C NMR (75 MHz, D2O+DCl):
d = 174.8, 170.7 (C@O), d = 133.5, 132.0, 131.2, 129.5 (Ar);
d = 68.51 CH2OH, d = 59.57, 58.10 (C ), d = 20.05 (CHCH3CH2OH).
a
A green and straightforward procedure was therefore estab-
lished for the synthesis of dipeptides and diketopiperazines, con-
4.2.3. Synthesis of
At first, 14.7 g (100 mmol) of
(36.2 mmol) of -PGA were suspended in water and the pH was
adjusted to 9.7, making an overall volume of 500 mL. To this,
12.2 g of immobilized penicillin acylase was added at room tem-
D
-phenylglycyl-
L
-glutamic acid
taining both the
D
-phenylglycyl and chemically functionalized
L
-glutamic acid and 5.52 g
amino acid moieties, which are compounds of growing interest
due to their catalytic and biological activity.
D
perature. Two further portions of D-PGA (4.74 g, 31.6 mmol) were
4. Experimental
added, with a 2-hour interval each, at which the pH was adjusted
to 9.9 with the NH4OH. After 6 hours, the enzyme was filtered and
a turbid mixture was obtained. The pH was then lowered to 3.3
with approximately 30 mL of H2SO4 (6 M). The mixture was left
stirring for a period of two days, after which the formed precipitate
was filtered, washed with petroleum ether 40–60, dried and 13.2 g
was obtained. The mixture was kept stirring for an extra period of
10 days. More precipitate was formed and collected, following the
same procedure, affording 7.28 g, which corresponds to an overall
yield of 73%. The product was isolated NMR pure. Mp: 217–219 °C.
4.1. Materials and methods
L
-Amino acids and D-phenylglycine amide were obtained from
commercial suppliers. Immobilized penicillin acylase from Esche-
richia coli, Assemblase 7500Ò, was kindly donated by DSM Anti-
Infectives (Delft, The Netherlands). Mass spectrometry analysis
gave coherent results, Mw or Mw+1, according to the technique
used. The melting points were measured with a Büchi 510 and
are not corrected. Optical rotations were measured with either a
Perkin Elmer 241 Polarimeter or a Perkin Elmer 343 Polarimeter
and concentrations were converted to the standard value of c 1.
[
a
]
24 = À87.5 (c 1, 2.5 M HCl). 1H NMR (300 MHz, D2O+DCl):
D
d = 7.487 (m, aromatic protons), d = 5.173 (s, CHNH2), d = 4.485
(dd, J1 = 3.9 Hz, J2 = 9.6 Hz, CHCH2CH2COOH), d = 2.15, 1.81 (mm,
CHCH2CH2COOH); 13C NMR (75 MHz, D2O+DCl): d = 178.1, 176.0,
170.3 (C@O), d = 133.4, 132.0, 131.3, 129.5 (Ar); d = 58.14, 53.44
4.2. Synthesis of D-phenylglycyl-L-amino acid dipeptides
(C ), d = 30.99, 26.91 (CHCH2CH2COOH).
a
4.2.1. Synthesis of
At first, 6.30 g (60.0 mmol) of
-PGA were suspended in water and the pH was adjusted to 9.70
with NH4OH 25%, making an overall volume of 230 mL. Next,
6.01 g of immobilized penicillin acylase was added at room tem-
D
-phenylglycyl-
L-serine
4.2.4. Synthesis of
At first, 15.1 g (104 mmol) of
(34.7 mmol) of -PGA were suspended in 200 mL of water and
the pH was adjusted to 9.7 with NH4OH 25%. To this, 4.12 g of
immobilized penicillin G acylase was added at room temperature.
D
-phenylglycyl-
L
-glutamine
L-serine and 2.25 g (15.0 mmol) of
L
-glutamine and 5.21 g
D
D
perature. Three further portions of D-PGA (2.25 g, 15.0 mmol) were
Two further portions of D-PGA (5.20 g, 34.7 mmol) were added,
added, with a 1.5-hour interval each, at which the pH was read-
justed to 9.8 with the same base. After 6 hours, the enzyme was fil-
tered and a turbid mixture was obtained. The pH was adjusted to
5.6 with H2SO4 (6 M). The mixture was left stirring for 4 days, after
which precipitation occurred. The solid formed was filtered,
washed with petroleum ether 40–60, dried, and 200 mg was
obtained. The mixture was kept stirring for a period of 50 days
and more precipitate was formed. This was collected, following
the same procedure, after which 1.17 g was obtained in the second
fraction, affording a total yield of 10%. The product was isolated
with a 2-hour interval each, at which the pH was readjusted to
9.7 with the same base. After 6 hours of reaction time, the enzyme
was filtered and the pH was then lowered to 5 with H2SO4 (6 M);
precipitation was immediate. The mixture was left stirring over-
night, after which the precipitate was filtered, washed with petro-
leum ether 40–60, dried and 8.59 g (29.7%) was obtained. A second
precipitate fraction was filtered from the mother liquor the day
after and a third fraction 2 days after, affording 840 mg and
1.68 g, respectively. The first fraction was NMR pure while the
second had minor impurities. Mp: 238–240 °C. [
a
]
24 = À89.5
D
pure. Mp: 224–226 °C (d). [
a
]
D
20 = À68.9 (c 1, 2.5 M HCl). 1H
(c 1, 2.5 M HCl). 1H NMR (300 MHz, D2O+DCl): d = 7.376 (m, aro-
matic protons), d = 5.061 (s, CHNH2), d = 4.287 (dd, J1 = 4.5 Hz,
J2 = 9.3 Hz, CHCH2CH2CONH2), d = 2.08, 1.944 (mm, CHCH2CH2-
COOH); 13C NMR (75 MHz, D2O+DCl): d = 177.6, 174.5 , 168.9
NMR (300 MHz, D2O+DCl): d = 7.49–7.51 (m, aromatic protons),
d = 5.22 (s, CHNH2), d = 4.54 (t, J = 4.5 Hz, CHCH2OH), d = 3.861,
d = 3.86, 3.73 (m, m, J1 = 3.9 Hz, J2 = 11.7 Hz, CHCH2OH); 13C NMR
(75 MHz, D2O+DCl): d = 174.5, 170.3 (C@O), d = 133.3, 132.0,
(C@O), d = 132.0, 130.7, 129.9, 128.2 (Ar), d = 56.74, 52.47 (C ),
a
131.3, 129.6 (Ar); d = 62.26 CH2OH, d = 58.08, 56.55 (C ).
a
d = 30.89, 26.1 (CHCH2CH2CONH2).
4.2.2. Synthesis of
At first, 7.16 g (60.4 mmol) of
(15.0 mmol) of -PGA were suspended in water and the pH was
D
-phenylglycyl-
L
-threonine
4.2.5. Synthesis of
At first, 8.98 g (60.2 mmol) of
(15 mmol) of -PGA were suspended in water and the pH was
D
-phenylglycyl-
L
-methionine
L-threonine and 2.25 g
L-methionine and 2.25 g
D
D
adjusted to 9.80 with NH4OH 25%, making an overall volume of
200 mL. Next, 6.27 g of immobilized penicillin acylase was added
adjusted to 9.80 with NH4OH 25%, making an overall volume of
200 mL. To this, 6.05 g of immobilized penicillin G acylase was
at room temperature. Three further portions of
D-PGA (2.25 g,
added at room temperature. Three further portions of
D-PGA
15 mmol) were added, with a 1.5-hour interval each, at which
the pH was readjusted to 9.8 with the same base. After 6 hours,
the enzyme was filtered and a turbid mixture was obtained. The
pH was then lowered to 6.4 with H2SO4 (6 M). The mixture was left
stirring overnight, after which precipitation occurred. The solid
was filtered, washed with petroleum ether 40–60 and dried. The
yield was 41% (6.25 g) and the product was isolated NMR pure.
(2.25 g, 15 mmol) were added, with a 1.5-hour interval each, at
which point the pH was readjusted to 9.8 with the same base. After
6 hours, the enzyme was filtered and a turbid mixture was
obtained. The pH was then lowered to 5.8 with H2SO4 (6 M). The
mixture was left stirring overnight, after which precipitation
occurred and the formed solid was filtered, washed with petro-
leum ether 40–60 and dried under vacuum and P2O5 for a period
of 9 days, affording 9.05 g of product corresponding to a yield of
53%. The product was isolated NMR pure. Mp: 235–237 °C.
Mp: 250–253 °C (d). [
a]
24 = À27.8 (c 1, 2.5 M HCl). 1H NMR
D
(300 MHz, D2O+DCl): d = 7.54 (m, aromatic protons), d = 5.26 (s,
CHNH2), d = 4.515 (CHCHOHCH3), d = 4.312 (m, CHCHOHCH3),
[a]
24 = À97.9 (c 1, 2.5 M HCl). 1H NMR (300 MHz, D2O+DCl):
D

P. C. Pereira et al. / Tetrahedron: Asymmetry 25 (2014) 825–832
831
d = 7.485 (m, aromatic protons), d = 5.157 (s, CHNH2), d = 4.6 (m,
4.2.9. Attempted synthesis of
diketopiperazine
At first, 1.05 g (3.75 mmol) of
D
-phenylglycyl-
L
-glutamic acid
J1 = 3.9 Hz, J2 = 10.2 Hz, CHCH2CH2SCH3), d = 2.246, 2.061 (mm,
CHCH2CH2SCH3), d = 1.856 (mm, CHCH2CH2SCH3); 13C NMR
(75 MHz, D2O+DCl): d = 176.4, 170.3 (C@O), d = 133.5, 132.0,
D-phenylglycyl- -glutamic acid
L
was suspended in 20 mL of dry methanol. A white suspension
131.3, 129.5 (Ar), d = 58.14, 52.86 (C ), d = 30.97, 30.71 (CHCH2CH2-
was obtained which was cooled to 0–5 °C. To this suspension
a
SCH3), d = 15.48 (CHCH2CH2SCH3).
570 lL of SOCl2 (7.86 mmol, 2.1 equiv) was added dropwise. The
mixture was allowed to reflux for 4 hours and after cooling to room
temperature, 40 mL of water was added. The pH was adjusted to
8.7 with a saturated solution of sodium hydroxide. The mixture
was left stirring and after 7 days, the formed precipitate was fil-
tered. This was washed with petroleum ether 40–60 and dried.
The product was isolated and 680 mg was obtained, which
corresponds to a yield of 66%. The product was identified as the
4.2.6. Synthesis of
At first, 1.00 g (4.20 mmol) of
pended in 44 mL of dry methanol. A white suspension was obtained
which was cooled. To this suspension 370 L of SOCl2 (5.09 mmol,
1.21 equiv) was added dropwise. The mixture was allowed to reflux
for 4 hours and after cooling to room temperature, 168 mL of water
was added. The pH was adjusted to 8.6 with a saturated solution of
sodium hydroxide. The mixture was left stirring and after 7 days,
the formed precipitate was filtered. This was washed with petro-
leum ether 40–60 and dried. The product was isolated and
400 mg was obtained, which corresponds to a yield of 43%. Mp:
D
-phenylglycyl-
L
-serine diketopiperazine
D
-phenylglycyl-
L-serine was sus-
l
corresponding methyl ester. Mp: 236–238 °C.
[a]
24 = À43.8
D
(c 1, DMSO). 1H NMR (300 MHz, DMSO-d6): d = 8.64 (d, J = 2.4 Hz,
CHNHCO), d = 8.26 (s, CHNHCO), d = 7.35 (m, aromatic protons),
d = 4.93 (d, J = 2.7 Hz, CHPh), d = 4.06 (t, CHCH2CH2CO2CH3),
d = 3.32 (s, CH2CH2CO2CH3), d = 2.905, d = 2.905 (m, m, CH2CH2CO2-
CH3); 13C NMR (75 MHz, DMSO-d6): d = 173.6, 168.2, 167.5 (C@O),
d = 139.4, 129.2, 128.6, 127.7 (Ar), d = 59.40 (CHCO2CH3), d = 53.3
(CHPh), d = 52.1 (CHCH2CH2CO2CH3), d = 29.5 (CHCH2CH2CO2CH3),
d = 27.9 (CHCH2CH2CO2CH3).
263–265 °C.
[
a
]
24 = À59.9 (c 1, DMSO). 1H NMR (300 MHz,
D
DMSO-d6): d = 8.47 (s, CHNHCO), d = 7.99 (s, CHNHCO), d = 7.35
(m, aromatic protons), d = 5.08 (t, J = 4.9 Hz, CHCH2OH), d = 4.90
(d, J = 1.2 Hz, CHPh), d = 3.90 (s, CHCH2OH), d = 3.81 (dd, CHCH2OH),
d = 3.63 (dd, CHCH2OH); 13C NMR (75 MHz, DMSO-d6): d = 167.6,
167.5 (C@O), d = 140.0, 129.0, 128.5, 128.2 (Ar), d = 63.27 (CH2OH),
d = 59.45 (CHPh), d = 57.62 (CHCH2OH).
4.2.10. Attempted synthesis of
diketopiperazine
D-phenylglycyl-L-glutamine
At first, 1.02 g (3.66 mmol) of
suspended in 20 mL of dry methanol. A white suspension was
obtained which was cooled to 0–5 °C. To this suspension 320
D-phenylglycyl-L-glutamine was
4.2.7. Synthesis of
1.00 g (3.97 mmol) of
pended in 20 mL of dry methanol. A white suspension was
obtained which was cooled. To this suspension 350 L of SOCl2
D
-phenylglycyl-
L
-threonine diketopiperazine
D-phenylglycyl-
L
-threonine was sus-
lL
of SOCl2 (4.41 mmol, 1.2 equiv) was added dropwise. The mixture
was allowed to reflux for 4 hours and after cooling to room temper-
ature, 40 mL of water was added. The pH was adjusted to 8.6 with a
saturated solution of sodium hydroxide. The mixture was left stir-
ring and after 7 days, the formed precipitate was filtered. This was
washed with petroleum ether 40–60 and dried. The product was
isolated and 700 mg was obtained, which corresponds to a yield
of 69%. This was identified as the same methyl ester that was
obtained from the phenyl glycine glutamic acid dipeptide (see
l
(4.82 mmol, 1.2 equiv) was added drop-wise. The mixture was
allowed to reflux for 4 hours and after cooling to room tempera-
ture, 40 mL of water was added. The pH was adjusted to 8.6 with
a saturated solution of sodium hydroxide. The mixture was left
stirring and after 7 days, the formed precipitate was filtered. This
was washed with petroleum ether 40–60 and dried. The product
was isolated and 450 mg was obtained, which corresponds to a
yield of 43%. Mp: 274–276 °C. [
a]
24 = À71.4 (c 1, DMSO). 1H
D
above). Mp: 236–238 °C. [
a
]
D
24 = À44.9 (c 1, DMSO). 1H NMR
NMR (300 MHz, DMSO-d6): d = 8.39 (s, CHNHCO), d = 8.13 (d,
J = 1.6 Hz, CHNHCO), d = 7.33 (m, aromatic protons), d = 5.10 (d,
J = 5.6 Hz, CHCHOHCH3), d = 4.94 (s, CHPh), d = 4.12 (t, CHCHOH
CH2), d = 3.64 (s, CHCH2OH), d = 1.14 (d, J = 6.4 Hz, CHCHOHCH3);
13C NMR (75 MHz, DMSO-d6): d = 167.8, 167.7 (C@O), d = 138.9,
128.0, 127.9, 127.6 (Ar), d = 67.95 (CHOH), d = 60.56 (CHPh),
d = 58.28 (CHCHOHCH3), d = 19.79 (CHCH2OHCH3).
(300 MHz, DMSO-d6): d = 8.63 (d, J = 2.7 Hz, CHNHCO), d = 8.24 (s,
CHNHCO), d = 7.35 (m, aromatic protons), d = 4.92 (d, J = 2.7 Hz,
CHPh), d = 4.05 (t, CHCH2CH2CO2CH3), d = 3.60 (s, CH2CH2CO2CH3),
d = 2.45, d = 2.03 (m, m, CH2CH2CO2CH3); 13C NMR (75 MHz,
DMSO-d6): d = 173.6, 168.2, 167.5 (C@O), d = 139.3, 129.2, 128.6,
127.7 (Ar), d = 59.4 (CO2CH3), d = 53.3 (CHPh), d = 52.1 (CHCH2CH2-
CO2CH3), d = 32.21 (CHCH2CH2CO2CH3), d = 29.5 (CHCH2CH2CO2-
CH3), d = 27.9 (CHCH2CH2CO2CH3).
4.2.8. Synthesis of
At first, 1.04 g (3.69 mmol) of
suspended in 15 mL of dry methanol. A white suspension was
obtained which was cooled. To this suspension 325 L of SOCl2
D-phenylglycyl-L-methionine diketopiperazine
D
-phenylglycyl-
L-methionine was
References
l
1. Ressurreição, A. S. M.; Delatouche, R.; Gennare, C.; Piarulli, U. Eur. J. Org. Chem.
2011, 217–228.
2. Fischer, P. M. J. Peptide Sci. 2003, 9, 9–35.
(4.48 mmol, 1.2 equiv) was added dropwise. The mixture was
allowed to reflux for 4 hours and after cooling to room tempera-
ture, 30 mL of water was added. The pH was adjusted to 8.6 with
a saturated solution of sodium hydroxide. The mixture was left
stirring and after 10 days, the formed precipitate was filtered. This
was washed with petroleum ether 40–60 and dried. The product
was isolated and 590 mg was obtained, which corresponds to a
3. Avendaño, C.; de la Cuesta, E. Curr. Org. Synth. 2009, 6, 143–168.
4. Borthwick, A. D.; Liddle, J. Med. Res. Rev. 2011, 31, 576–604.
5. McCleland, K.; Milne, P. J.; Lucieto, F. R.; Frost, C.; Brauns, S. C.; Van De Venter,
M.; Du Plessis, J.; Dyason, K. J. Pharm. Pharmacol. 2004, 56, 1143–1153.
6. Song, M. K.; Hwang, I. K.; Rosenthal, M. J.; Harris, D. M.; Yamaguchi, D. T.; Yip,
I.; Go, V. L. W. Exp. Biol. Med. 2003, 228, 1338–1345.
7. Martins, M. B.; Carvalho, I. Tetrahedron 2007, 63, 9923–9938.
8. Carrea, G.; Colonna, S.; Kelly, D. R.; Lazcano, A.; Ottolina, G.; Roberts, S. M.
Trends Biotechnol. 2005, 23, 507–513.
9. Kowalski, J.; Lipton, M. A. Tetrahedron Lett. 1996, 37, 5839–5840.
10. Jackson, W. R.; Jayatilake, G. S.; Matthews, B. R.; Wilshire, C. Aust. J. Chem. 1988,
41, 203–213.
11. Noe, C. R.; Weigand, A.; Pirker, S. Monatsh. Chem. 1996, 127, 1081–1097.
12. Jarvo, E. R.; Miller, S. J. Tetrahedron 2002, 58, 2481–2495.
13. List, B. Tetrahedron 2002, 58, 5573–5590.
yield of 35%. Mp: 253–255 °C. [
a]
24 = À47.6 (c 1, DMSO). 1H
D
NMR (300 MHz, DMSO-d6): d = 8.62 (d, J = 2.7 Hz, CHNHCO),
d = 8.29 (s, CHNHCO), d = 7.35 (m, aromatic protons), d = 4.93 (d,
J = 2.4 Hz, CHPh), d = 4.08 (t, J = 5.0 Hz, CHCH2CH2SCH3), d = 2.905
(m, CH2CH2SCH3); d = 2.024 (m, CH2CH2SCH3) 13C NMR
(75 MHz, DMSO-d6): d = 168.3, 167.4 (C@O), d = 139.4, 129.2,
128.6, 127.7 (Ar), d = 59.40 (CHPh), d = 53.32 (CHCH2CH2SCH3),
d = 32.21 (CHCH2CH2SCH3), d = 29.14 (CHCH2CH2SCH3), d = 15.21
(CHCH2CH2SCH3).
14. Wennemers, H. Chem. Commun. 2011, 47, 12036–12041.
15. Huang, Z.-Z.; Leman, L. J.; Ghadiri, M. R. Angew. Chem., Int. Ed. 2008, 47,
1758–1761.

832
P. C. Pereira et al. / Tetrahedron: Asymmetry 25 (2014) 825–832
16. Pichowicz, M.; Simpkins, N. S.; Blake, A. J.; Wilson, C. Tetrahedron 2008, 64,
3713–3755.
17. Datta, S.; Sood, A.; Török, M. Curr. Org. Synth. 2011, 8, 262–280.
18. Kallenberg, A. I.; van Rantwijk, F.; Sheldon, R. A. Adv. Synth. Catal. 2005, 347,
905–926.
21. Pereira, P. C.; Sheldon, R. A.; Arends, I. W. C. E. Submitted for publication.
ˇ
ˇ
22. Vanzura, J.; Palát, K.; Celadník, M.; Hradec, K.; Klimes, J.; František, S.; CS Patent
210383, 1983.
ˇ
23. Didziapetris, R. J.; Švedas, V. K. Biomed. Biochim. Acta 1991, 50. 10/11, S 237–S
242.
19. van Langen, L. M.; van Rantwijk, F.; Švedas, V. K.; Sheldon, R. A. Tetrahedron:
Asymmetry 2000, 11, 1077–1083.
24. Pouchert, C. J.; Behnke, J. The Aldrich Library of 13C and 1H FT NMR Spectra 1;
Aldrich Chemical Company, 1993. p 894.
20. Khimiuk, A. Y.; Korennykh, A. V.; van Langen, L. M.; van Rantwijk, F.; Sheldon,
R. A.; Švedas, V. K. Tetrahedron: Asymmetry 2003, 14, 3123–3128.
25. Schroën, C. G. P. H.; Mohy Eldin, M. S.; Janssen, A. E. M.; Mita, G. D.; Tramper, J.
J. Mol. Catal. B: Enzym. 2001, 15, 163–172.