Pilot-scale enzymatic synthesis of bioactive oligopeptides in eutectic-based media

Article abstract of DOI:10.1021/op025528y

The bioactive peptide precursors N_α-Cbz-L-Lys(N_ζ-Cbz)-Gly-OEt (1), N_α-Cbz-L-Lys(N_ζ-Cbz)-Gly-L-Asp(OAll)-OAll (2), and N_α-Cbz-L-Lys(N_ζ-Cbz)-Gly-L-Asp(OAll)-L-Glu (OAll)OEt (3) were synthesiz

Full text of DOI:10.1021/op025528y

Organic Process Research & Development 2002, 6, 684−691
Technology Reports
Pilot-Scale Enzymatic Synthesis of Bioactive Oligopeptides in Eutectic-Based
Media
,†
‡
Iqbal Gill* and Rao Valivety
BioSynTech, 512 Franklin AVenue, Nutley, New Jersey 07110, U.S.A., and Albany Molecular Research, Inc.
01 East Kensington Road, Mount Prospect, Illinois 60056, U.S.A.
6
Abstract:
condensation of oligopeptide fragments produced by chemi-
cal protocols, and the ligation/modification of proteins
generated by recombinant methods, as discussed in recent
The bioactive peptide precursors N
OEt (1), N
r
-Cbz-
-Asp(OAll)-OAll (2), and
-Asp(OAll)- -Glu(OAll)OEt (3) were
ú
L-Lys(N -Cbz)-Gly-
r
-Cbz-
ú
L-Lys(N -Cbz)-Gly-L
1
1-25
N
r
-Cbz- -Lys(N
L
ú
-Cbz)-Gly-
L
L
papers and reviews.
The advantages of using protease-
synthesized via a sequential N-to-C strategy in heterogeneous
solid-liquid mixtures of substrates using Celite-deposited
chymopapain and subtilisin as catalysts. The substrate pairs
displayed eutectic formation, and in the presence of 16-20%
w/w of water and ethanol furnished low-melting point systems
that provided substrate concentrations of 0.81-1.01 M, and
sustained protease catalysis. The di-, tri-, and tetra-peptide were
obtained in yields of 73, 74, and 67% (yield per step) respec-
tively, and product concentrations of 0.34, 0.48, and 0.45 g per
gram of reaction mixture were achieved with these systems.
The syntheses were scaled up using wiped-blade and rotating-
drum reactors, and 1, 2, and 3 were produced in yields of 71,
mediated protocols have been well recognized, with the
avoidance of racemization, dispensation with side-chain
protection, decreased requirement for N-protection, utilization
of cheap and readily available substrates, and the use of mild,
nonhazardous operating conditions being the most notable
11,12
features.
Although numerous enzymatic syntheses have
been devised for peptides, only one process, namely the
thermolysin-catalysed synthesis of Aspartame is operating
on the large scale. Critical hurdles which have blocked the
industrial transfer of lab-scale protocols is the requirement
for bulk organic solvents (which complicates reactor design
and operation and requires solvent recovery and disposal)
and low substrate concentrations which invariably lead to
low space-time yields and a lack of economic feasibility.
7
3, and 71%, at product concentrations of 0.36, 0.49, and 0.48
-
1
kg kg , on a reaction scale of 11.34, 6.30, and 5.83 kg (11.54,
.17, and 4.96 mol), respectively. The corresponding space-
6
-1
-1
time yields were 0.30, 0.64, and 0.30 kg kg d , and biocatalyst
(7) Meloen, R. H.; Puijk, W. C.; Slootstra, J. W. J. Mol. Recog. 2000, 13,
reuse provided productivities of 166, 235, and 312 kg-product
352-359.
-
1
kg-enzyme . Furthermore, a sequential one-pot, three-step
protocol was readily applied to the synthesis of the tetrapeptide
(
(
8) Yamamoto, N.; Takano, T. Nahrung 1999, 43, 159-164.
9) Silano, M.; De Vincenzi, M. Nahrung 1999, 43, 175-184.
(
(
10) Latham, P. W. Nat. Biotechnol. 1999, 17, 755-757.
11) Gill, I. S.; L o´ pez-Fandino, R.; Jorba, X.; Vulfson, E. N. Enzyme Microb.
Technol. 1996, 18, 162-183.
3
which was obtained with an overall yield of 34%, product
concentration of 0.25 kg per kg, and space-time yield of 0.07
-
1
-1
(12) Stepanov, V. M. Pure Appl. Chem. 1996, 68, 1335-1339.
(
kg kg
d .
13) Getun, I. V.; Filippova, I. Y.; Lysogorskaya, E. N.; Oksenoit, E. S. J. Mol.
Catal. B 2001, 15, 105-110.
14) Miyazawa, T.; Nakajo, S.; Nishikawa, M.; Hamahara, K.; Imagawa, K.;
Ensatsu, E.; Yanagihara, R.; Yamada, T. J. Chem. Soc., Perkin Trans. 1
(
2
001, 82-86.
Introduction
(
15) Faulconbridge, S. J.; Holt, K. E.; Lock, C. J.; Taylor, S. J. C.; Woods, M.
Org. Process Res. DeV. 2000, 4, 563-566.
(16) Khumtaveeporn, K.; Ullmann, A.; Matsumoto, K.; Davis, B. G.; Jones, J.
B. Tetrahedron: Asymmetry 2001, 12, 249-261.
17) Bordusa, F.; Dahl, C.; Jakubke, H.-D.; Burger, K.; Kokscha, B. Tetrahe-
With the increasing focus on the structures, functions, and
medical applications of bioactive oligopeptides and peptide
bioconjugates, there has been great interest in developing
(
new methods suitable for the economic and large-scale
dron: Asymmetry 1999, 10, 307-313.
synthesis of these molecules.1
-11
(18) Bj o¨ rup, P.; Torres, J. L.; Adlercreutz, P.; Clape, P. Bioorg. Med. Chem.
Lett. 1996, 6, 891-901.
Enzymatic approaches have
been applied to the total synthesis of small peptides, the
(
19) Erbeldinger, M.; Ni, X.; Halling, P. J. Biotechnol. Bioeng. 2001, 72, 69-
7
6.
*
Author for correspondence. E-mail: iqbalgill@yahoo.com
BioSynTech.
Albany Molecular Research, Inc.
(20) Wehofsky, N.; Alischb, M.; Bordusa, F. Chem. Commun. 2001, 1602-1603.
(21) G u¨ nther, R.; Bordusa, F. Chem. Eur. J. 2000, 6, 463-467.
(22) Gill, I.; L o´ pez-Fandino, R.; Vulfson, E. N. J. Am. Chem. Soc. 1995, 117,
6175-6181.
(23) Gill, I.; Vulfson, E. N. J. Am. Chem. Soc. 1993, 115, 3348-3349.
(24) Chen, S.-T.; Chen, S.-Y.; Wang, K.-T. J. Org. Chem. 1992, 57, 6960-
6965.
†
‡
(
(
(
1) Van Belkum, M. J.; Stiles, M. E. Nat. Prod. Rep. 2000, 17, 323-335.
2) Hancock, L. E.; Lehrer, R. Trends Biotechnol. 1998, 16, 82-88.
3) Shen, G. S.; Layer, R. T.; McCabe, R. T. Drug DiscoVery Today 2000, 5,
9
8-106.
(4) Van Belkum, M. J.; Stiles, M. E. Nat. Prod. Rep. 2000, 17, 323-335.
(5) Barra, D.; Simmaco, M. Trends Biotechnol. 1995, 13, 205-209.
(6) Hancock, R. E. W.; Diamond, G. Trends Microbiol. 2000, 8, 402-410.
(25) Wang, P.; Hill, T. G.; Wartchow, C. A.; Huston, M. E.; Oehler, L. M.;
Smith, M. B.; Bednarski, M. D.; Callstrom, M. R. J. Am. Chem. Soc. 1992,
114, 378-384.
6
84
•
Vol. 6, No. 5, 2002 / Organic Process Research & Development
10.1021/op025528y CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/04/2002

In the context of “Green Process” development and high-
productivity biocatalysis, two novel approaches have recently
been introduced, namely peptide synthesis in heterogeneous
solid-liquid systems comprised of mixtures of substrates
2
3,26-29
which undergo eutectic formation
and solid-to-solid
conversions in aqueous suspensions of substrates.1
9,30-38
The
former method relies upon the observation that amino acid
and peptide derivatives commonly employed as substrates
in peptide synthesis form low-melting point mixtures (com-
pletely liquid eutectic or liquid eutectic plus excess solid
substrate) when combined together in the absence of bulk
solvents, while the latter utilizes concentrated aqueous
suspensions of solid substrates in which reactions appear to
take place in a saturated aqueous phase rather than a liquid
eutectic. The selection of substrate combinations with or
without the judicious addition of small amounts (0-20%
w/w) of water or polar solvents (“adjuvants”, defined as
solvents used in quantities far below those required for the
dissolution of the solid substrates to form conventional
solutions) capable of depressing the melting points of the
mixtures leads to homogeneous liquids or heterogeneous
liquid-solid systems which readily sustain enzymatic activity
and enable the synthesis of oligopeptides. Such liquid-solid
systems provide very high substrate concentrations approach-
ing the theoretical maximum, typically being in the range
of 0.8-3.1 M, together with high synthetic rates and yields,
leading to product concentrations and productivities which
are typically 5-13-fold higher than those obtained in conven-
tional solution systems. This, together with the dispensation
with bulk solvent and the lower reaction volume provides a
facile, efficient, and scalable methodology which can be
readily applied to the multikilogram-scale synthesis of
peptides and complements the scale-up of enzymatic peptide
r ú
Figure 1. N-to-C synthesis of N -Cbz-L-Lys(N -Cbz)-Gly-L-
Asp(OAll)-L-Glu(OAll)-OEt.
ing a stepwise N-to-C protocol conducted totally in hetero-
geneous, low-solvent mixtures.
Results and Discussion
R ú
The synthesis of N -Cbz-L-Lys(N -Cbz)-Gly-L-Asp(OAll)-
L-Glu(OAll)OEt (3) was based around a convergent N-to-C
protocol devised for the solution-phase synthesis of the
octapeptide L-Lys-Gly-L-Asp(OH)-L-Glu(OH)-L-Glu(OH)-L-
2
2
2
Ser-L-Leu-L-Ala-NH (Figure 1). The strategy employs
stepwise N-to-C assembly, utilizing the selectivities of the
proteases employed rather than differential protection to
achieve the required amino acid couplings. For large-scale
production, particularly from the point of view of reaction
control, reactor volume, and productivity, it was highly
desirable to translate the solution-phase protocol to a
synthesis based upon heterogeneous, low-solvent reaction
media. For this purpose a screening program was initiated
to determine if the requisite substrate combinations formed
eutectics and whether the couplings could be efficiently
performed in such homogeneous/heterogeneous media. Chy-
mopapain and subtilisin deposited onto Celite were chosen
as the biocatalysts, Cbz N-terminal protection, together with
ethyl ester or allyl ester donors, was chosen for the synthesis
of peptides 1, 2, and 3, and the reactions were examined
under a variety of conditions (Table 1). In addition to water,
ethanol was chosen as a solvent (adjuvant) in view of its
efficiency in promoting the formation of low-melting point
3
0-38
synthesis using the solid-to-solid conversion approach.
The present work discloses the first example of the scale-
up of peptide synthesis in such low-melting point mixtures
derived from eutectic interactions and describes the develop-
ment of pilot processes for the kilogram-scale production of
the flavor-active peptide precursors N
Gly-OEt, N -Cbz-L-Lys(N -Cbz)-Gly-L-Asp(OAll)-OAll, and
-Cbz-L-Lys(N -Cbz)-Gly-L-Asp(OAll)-L-Glu(OAll)OEt us-
R ú
-Cbz-L-Lys(N -Cbz)-
R
ú
N
R
ú
(
26) Jorba, X.; Gill, I.; Vulfson, E. N. J. Agric. Food Chem. 1995, 9, 2536-
2
541.
(
27) Gill, I.; Vulfson, E. N. Trends Biotechnol. 1994, 12, 118-122.
(28) L o´ pez-Fandino, R.; Gill, I.; Vulfson, E. N. Biotechnol. Bioeng. 1994, 43,
1
016-1023.
29) L o´ pez-Fandino, R.; Gill, I.; Vulfson, E. N. Biotechnol. Bioeng. 1994, 43,
024-1030.
30) Ulijn, U. V.; Janssen, A. E. M.; Moore, B. D.; Halling, P. J. Angew. Chem.,
Int. Ed. 2001, 7, 2089-2098.
31) Ulijn, R. V.; Erbeldinger, M.; Halling, P. J. Biotechnol. Bioeng. 2000, 69,
(
(
(
(
(
(
(
(
(
(
26-29
1
multicomponent mixtures,
its low toxicity, and its ready
recovery and reuse in large-scale processes.
Neat combinations of the pure substrates in the complete
absence of water/solvent (adjuvant) provided simple hetero-
geneous systems consisting of eutectic plus excess solid
substrate, with the eutectic compositions displaying eutectic
6
34-639.
32) Erbeldinger, M.; Ni, X.; Halling, P. J. Biotechnol. Bioeng. 1999, 63, 316-
21.
33) Erbeldinger, M.; Ni, X.; Halling, P. J. Biotechnol. Bioeng. 1998, 59, 68-
2.
34) Erbeldinger, M.; Ni, X.; Halling, P. J. Enzymol. Microb. Technol. 1998,
3, 141-148.
35) Eichhorn, U.; Bommarius, A. S.; Drauz, K.; Jakubke, H. D. J. Pept. Sci.
997, 3, 245-251.
36) Kuhl, P.; Eichorn, U.; Jakubke, H. D. Biotechnol. Bioeng. 1995, 45, 276-
78.
3
7
e
melting points (T ’s) of 58, 16, and 18 °C for the respective
2
substrate combinations of 1, 2, and 3. It should be noted
from Table 1 that the volume fractions of the liquid phases
were lower than expected for full eutectic formation, wherein
one substrate goes completely into the liquid phase. This is
because the substrate mixtures were held at 70 °C for only
a short period of time (5 min) to avoid decomposition of
1
2
37) Kuhl, P.; Eichorn, U.; Jakubke, H. D. In Biocatalysis in Non ConVentional
Media; Elsevier Science, Amsterdam, 1992; pp. 513-518.
38) Kasche, V.; Galunsky, B. Biotechnol. Bioeng. 1995, 45, 261-267.
Vol. 6, No. 5, 2002 / Organic Process Research & Development
•
685

Table 1. Protease-mediated synthesis of peptides 1, 2, and 3 in heterogeneous solid-liquid mixtures
adjuvanta
phase fractionc
phase changesf
substrates
mol ratio)
enzyme)
yield
(%)
(40 h)
(
% v/v, t ) 0)
extent (% v/v), time (h)
b
concnd
e
g
(
H
2
O
EtOH
T
e
initial rate
[prod.]
-
1
-1) liquefaction solidification (g g-1
(
(% w/w) (% w/w) (°C) liquid solid
(M) (mol kg
h
)
dipeptide (1):
-Cbz-L-Lys-OEt
0
2
4
6
4
4
4
4
6
6
0
0
0
0
4
58
36
36
36
23
23
23
23
23
23
0
24
27
32
37
41
48
53
60
69
100
76
73
68
63
59
52
47
40
31
1.20
1.18
1.15
1.13
1.10
1.06
1.01
0.96
0.98
0.94
3.6
19.3
37.2
51.7
39.8
38.0
33.2
30.8
45.3
41.1
8, 9.0 h
37, 7.5 h
44, 7.0 h
51, 5.0 h
61, 4.0 h
84, 3.5 h
94, 4.5 h
90, 4.0 h
93, 3.0 h
90, 2.5 h
98, 14.5 h
83, 19.0 h
87, 20.5 h
78, 22.0 h
63, 18.0 h
57, 21.5 h
51, 23.0 h
53, 23.0 h
45, 26.0 h
41, 28.5 h
0.04
0.18
0.30
0.28
0.32
0.34
0.37
0.34
0.31
0.32
7
31
56
50
58
64
73
70
64
68
N
R
+
Gly-OEt‚HCl
+
DIPEA
8
(
1:1.5:1.5)
12
16
12
16
(chymopapain)
tripeptide (2):
1
0
2
4
6
4
4
4
4
6
6
0
1
2
4
2
2
2
2
4
4
0
0
0
0
4
16
8, 11
8, 11
8, 11
2, 7
2, 7
2
2
2
2
18
12, 17
12, 17
12, 17
5, 14
5
5
5
5
5
26
28
33
37
48
57
68
73
70
77
33
36
39
43
57
63
68
70
73
77
74
72
67
63
52
43
32
27
30
23
67
64
61
57
43
37
32
30
27
23
1.22
1.20
1.17
1.15
1.12
1.07
1.03
0.98
1.00
0.95
0.99
0.98
0.96
0.95
0.93
0.91
0.89
0.85
0.87
0.77
9.8
42.6
64.1
83.0
67.6
62.8
61.8
58.3
74.1
65.3
9.4
38.1
60.4
79.2
82.5
83.0
78.6
73.1
79.0
75.5
29, 7.5 h
50, 7.0 h
57, 5.5 h
63, 6.0 h
75, 5.0 h
81, 6.0 h
89, 5.0 h
100, 4.5 h
98, 5.5 h
90, 5.0 h
44, 8.5 h
49, 7.0 h
53, 7.0 h
61, 6.0 h
73, 7.0 h
80, 6.5 h
86, 6.5 h
86, 6.0 h
95, 5.0 h
98, 5.5 h
82, 9.0 h
0.13
0.32
0.41
0.41
0.43
0.46
0.47
0.48
0.43
0.42
0.19
0.30
0.34
0.35
0.41
0.42
0.42
0.42
0.45
0.39
16
40
52
53
57
65
68
74
64
67
23
37
43
46
55
58
62
67
65
61
73, 11.0 h
70, 11.5 h
65, 13.0 h
69, 10.5 h
63, 13.5 h
75, 13.5 h
78, 14.5 h
70, 12.5 h
72, 13.0 h
67, 12.5 h
61, 15.0 h
63, 16.5 h
70, 14.0 h
60, 12.0 h
65, 16.0 h
69, 15.5 h
76, 14.0 h
71, 16.0 h
75, 13.5 h
+
L-Asp(OAll)-OAll
(
1: 1.5)
(chymopapain)
8
12
16
12
16
0
0
0
0
4
6
8
tetrapeptide (3):
2
+
L-Glu(OAll)-OAll
(
1:1.5)
(subtilisin)
12
8
16
a
Ethanol and water were used as solvents (adjuvants). b Eutectic temperature, determined by DSC. A second figure indicates the observation of two low-melting
c
temperature peaks in the DSC thermogram, the second probably due to hydrate formation. The volume fractions of the solid and liquid phases, determined by
centrifugation analysis at 40 °C. It should be noted that to avoid decomposition of the acyl acceptors, the substrate mixtures were held at 70 °C for only a short period
of time (5 min), which was insufficient for the formation of equilibrium eutectics, and thus the measured volume fractions of the liquid phases were lower than
expected. d The concentration of the acyl donor (limiting component) in the reaction mixture at t ) 0. The initial rate of product synthesis, determined at 10 min, and
e
f
quoted as mol per kg of Celite-deposited biocatalyst per h. The extent and time of the maximum (net) liquefaction observed during the reaction are given. Similarly,
g
the degree of solidification observed at the end of the reaction (40 h), and the time at which solidification began are also given. Product concentration attained at the
end of the reaction. Reactions were conducted on a 10-15-g scale in open beakers (150 mL) incubated at 40 °C, 150 rpm. Substrates and solvents (adjuvants) were
mixed at room temperature in a sealed vial, heated at ca. 5 °C min- to 70 °C, held at this temperature for 5 min, cooled at ca. 5 °C min to 40 °C, and maintained
at this temperature for 10 min prior to addition of the biocatalyst. Chymopapain and subtilisin deposited onto Celite (at 10% w/w) were used as catalysts at 10% w/w
of the reaction mixtures. DTT (0.1% w/w) was added as activator when chymopapain was used as a catalyst. Weighed samples were taken at 10 min, then at 0.5 h
intervals for centrifugation and HPLC analysis.
1
-1
the acyl acceptors and since eutectic formation in these
systems is under strong kinetic control, equilibration and
complete formation of eutectics were not achieved. This
resulted in the presence of both substrates in the solid phase
and lower than expected liquid volume fractions. However,
eutectic formation between the substrate pairs was indepen-
dently confirmed by DSC and microscopic analysis.
of product per kg of biocatalyst per h) to 37.2, 83.0, and
-1
-1
79.2 mol kg
h
respectively, and elevated the correspond-
ing product yields to 56, 53, and 46%, whilst incurring 8,
11, and 14% hydrolysis of acyl donors. This reflected direct
activation of the enzymes by water, combined with the
depression of the melting temperature of the reaction mixture
and accompanying increase in liquid-phase fraction (X
liquid-phase mobility.
L
) and
Higher water contents of 6, 8, and
6%, although increasing the synthesis rates to 51.7, 92.3,
26-29
Preliminary experiments indicated that chymopapain and
subtilisin stably tolerated temperatures of 35-45 °C when
the water content was 0-10%, and that the reaction rates
were optimal in this temperature range. Thus, the screening
was conducted at 40 °C, and at this temperature the liquid-
phase fractions of the mixtures amounted to 0, 26, and 33%
respectively, and the addition of proteases resulted in the
production of the di- tri-, and tetrapeptides with low yields
of 7, 16, and 23%, respectively, and no detectable production
of hydrolysis products. The inclusion of 4, 6, and 4% w/w
water in the respective reaction mixtures increased the rates
-
1
-1
and 91.5 mol kg
h
respectively, resulted in 11, 23, and
28% decreases in product yields due to the excessive
hydrolysis of the acyl donors.
In view of the above results, the co-addition of ethanol
e
and water, with the aim of reducing T further, and increasing
X
L
and mobility while suppressing acyl donor hydrolysis was
2
6-29
investigated.
Indeed, the inclusion of 12-16% w/w of
ethanol together with 2-4% w/w of water gave the best
results, furnishing 1, 2, and 3 in optimal yields of 73, 74,
and 67% respectively, and with 4, 6, and 9% of hydrolysis.
-1
-1
of peptide synthesis from 3.6, 9.8, and 9.4 mol kg
h
(mol
686
•
Vol. 6, No. 5, 2002 / Organic Process Research & Development

Figure 2. Schematics of the wiped-blade and rotating-drum reactors used for scale-up.
This compares with the yields of 82, 84, and 76% obtained
for the synthesis of the respective peptides in homogeneous
solution.22 This was accomplished with some reduction in
the kinetics, with overall rate enhancements of 8.2-, 4.9-,
and 7.4-fold observed as compared to the solvent (adjuvant)-
free reactions. It should be noted that the course of peptide
formation was observed to occur with a decrease in the
amount of solid phase, in other words dissolution of
suspended substrates and liquefaction of the reaction mixture,
followed by the gradual precipitation of crystalline product
and resulting partial solidification.26-29
process to the progress of the reactions was indicated with
the reduction in yields (ca. 13-26, 9-16, and 11-21%
lowering in yields of 1, 2, and 3, respectively) resulting from
the use of closed reaction vessels or the application of
reduced pressure. These deleterious effects presumably arose
from the build-up of alcohols adversely affecting biocatalyst
function in the former case, and the reduced liquefaction and
premature solidification of the reaction mixtures in the latter.
It is important to note a critical appeal of eutectic-based
media, namely the high substrate concentrations and cor-
respondingly elevated product concentrations gained (Table
1). Thus, under the optimal conditions wherein 14-20% w/w
of solvent (adjuvant) were employed, substrate (acyl donor)
concentrations of 1.01, 0.98, and 0.81 M were achieved, and
provided 0.37, 0.48, and 0.45 g of product per gram of
reaction mixture. These productivities are ca. 9-25 times
higher than those attainable when the syntheses are per-
formed in homogeneous (saturated) solutions of the substrates
The addition of water, and especially water plus ethanol,
was observed to increase the extent of liquefaction observed
during the progress of the reactions (Table 1). Thus, solvent
(adjuvant)-free syntheses of 1, 2, and 3, resulted in ca. 8,
2
9
9, and 44% (net) liquefaction of the reaction mixtures, while
4, 100, and 95% liquefaction was attained under optimal
conditions with ethanol-water. In addition the solvents
adjuvants) also generally increased the time period during
2
2
(
in ethanol-water mixtures.
which liquefaction was observed and delayed the onset of
solidification, especially in the case of 1. It can be postulated
that the speed and extent of liquefaction and subsequent
precipitation should control the rate of synthesis and ac-
cumulation of and final yield of product. Thus, reduced
liquefaction or rapid and extensive solidification or both were
often found to result in low product yields, presumably due
to only partial dissolution of substrates due to a low X in
L
the early stages, and premature termination of the coupling
reaction due to a sharp increase in viscosity and reduction
Once optimal conditions had been determined for the
syntheses, the reactions were scaled up to the multikilogram
scale (Table 2). Scale-up studies were implemented in three
types of reactor: (i) shaken beakers, where the reaction was
conducted in an open beaker placed in an orbital shaker; (ii)
wiped-blade reactors where the reaction mixture was actively
mixed with a turbine fitted with flexible blades capable of
efficiently blending the slurry (Figure 2a); (iii) rotating-drum
reactors, where the reaction mixture was mixed by placement
in baffled drums rotating at a slight inclination to the
horizontal (Figure 2b). Rather poor scale-up was observed
for the first configuration, with both synthesis rates and final
yields dropping significantly above the 50-g scale, probably
due to poor mixing of the high viscosity mixtures at this
L
in X in the final stages. It should be noted that the
implementation of reactions in open vessels resulted in the
evaporative loss of ca. 43-61% of the added solvents during
the course of the reactions. The importance of the evaporation
Vol. 6, No. 5, 2002 / Organic Process Research & Development
•
687

Table 2. Scale-up of protease-mediated synthesis of peptides 1, 2, and 3 in heterogeneous solid-liquid mixtures
activityd
adjuvanta
2
(% w/w)
recovered
b
c
H
O:EtOH
reaction
scale (kg)
initial rate
[prod.] yield (%) biocatalyst
-
1
-1) (g g
-1
product
mol
reactor type
(mol kg
h
)
(40 h)
(%, 40 h)
1
4:12
0.01
0.01
0.05
0.11
0.51
0.17
0.55
1.22
2.93
5.17
11.54
0.54
1.52
3.11
5.08
0.01
0.06
0.13
0.47
0.15
0.49
0.89
6.17
0.52
2.00
50 mL vial, orbital shaker, 200 rpm
0.2-L beaker, orbital shaker, 200 rpm
0.5-L beaker, orbital shaker, 150 rpm
2-L beaker, orbital shaker, 150 rpm
0.5-L wiped-blade reactor, 150 rpm
2-L wiped-blade reactor, 150 rpm
10-L wiped-blade reactor, 150 rpm
10-L wiped-blade reactor, 150 rpm
20-L wiped-blade reactor, 150 rpm
50-L wiped-blade reactor, 150 rpm
4-L inclined rotating-drum, 150 rpm
12-L inclined rotating-drum, 150 rpm
12-L inclined rotating-drum, 150 rpm
46-L inclined rotating-drum, 120 rpm
50 mL vial, orbital shaker, 200 rpm
0.2-L beaker, orbital shaker, 200 rpm
0.5-L beaker, orbital shaker, 150 rpm
2-L beaker, orbital shaker, 150 rpm
0.5-L wiped-blade reactor, 150 rpm
2-L wiped-blade reactor, 150 rpm
10-L wiped-blade reactor, 150 rpm
20-L wiped-blade reactor, 150 rpm
4-L inclined rotating-drum, 150 rpm
12-L inclined rotating-drum, 150 rpm
31.8
30.6
28.8
26.1
33.5
34.7
36.0
36.5
34.8
33.5
33.0
31.6
30.2
28.1
59.0
56.1
53.8
51.9
55.6
56.0
55.7
55.3
54.4
52.2
77.5
74.6
72.0
69.8
78.3
79.2
77.4
75.9
76.0
74.3
0.35
69
65
64
62
75
73
77
78
74
71
69
72
70
66
71
72
65
61
73
76
75
73
70
67
68
63
63
58
71
74
72
71
69
72
70
73
69
74
68
70
67
69
74
73
76
77
74
68
69
64
66
67
62
69
66
61
62
63
88
87
83
81
86
85
84
82
84
87
0
0
0
0
0
1
2
5
.05
.11
.50
.17
.54
.21
.90
.12
0.33
0.32
0.31
0.38
0.37
0.39
0.39
0.38
0.36
0.35
0.37
0.36
0.34
0.46
0.47
0.42
0.40
0.47
0.49
0.49
0.48
0.45
0.43
0.46
0.42
0.42
0.39
0.48
0.50
0.49
0.48
0.46
0.48
1
1.43
0
1
3
5
.53
.35
.08
.03
2
4:16
0.01
0
0
0
0
0
1
6
0
2
.06
.13
.48
.15
.50
.93
.30
.53
.04
3
2:12
0.01
0.008 50 mL vial, orbital shaker, 200 rpm
0
0
0
0
0
1
5
0
2
.05
.10
.51
.10
.60
.29
.83
.56
.15
0.04
0.09
0.43
0.09
0.51
1.10
4.96
0.48
1.83
0.2-L beaker, orbital shaker, 200 rpm
0.5-L beaker, orbital shaker, 150 rpm
2-L beaker, orbital shaker, 150 rpm
0.5-L wiped-blade reactor, 150 rpm
2-L wiped-blade reactor, 150 rpm
5-L wiped-blade reactor, 150 rpm
20-L wiped-blade reactor, 150 rpm
4-L inclined rotating-drum, 150 rpm
12-L inclined rotating-drum, 150 rpm
a
Ethanol and water were used as solvents (adjuvants). b Initial rate of product formation, determined at 10 min, and quoted as mol per kg of Celite-deposited
biocatalyst per h. c Product concentration attained at the end of the reaction. Activity of the biocatalyst relative to that of the fresh material, after extraction of the
reaction mixture with ethanol upon completion of the reaction. The substrates and solvents (adjuvants) were mixed at room temperature in a closed glass or stainless
d
-1
-1
steel vessel, heated at ca. 5-10 °C min to 70 °C, held at this temperature for 5 min, cooled at ca. 5-10 °C min to 40 °C, and maintained at this temperature for
0-20 min prior to addition of the biocatalyst. Chymopapain and subtilisin deposited onto Celite (at 10% w/w) were used as catalysts at 10% w/w of the reaction
mixtures. DTT (0.1% w/w) was added as activator when chymopapain was used as a catalyst. Weighed samples were quenched with methanol and analyzed by HPLC.
1
scale and the formation of a dry skin of substrate at the
surface of the reaction mixtures. Rotating-drum reactors fared
better, although inadequate mixing and the formation of
agglomerates at reaction scales above ca. 1 kg significantly
reduced kinetics and yields, probably due to excessive
evaporation of solvents. In contrast, the wiped-blade reactors
performed much better and provided efficient mixing and
good dispersions throughout the course of reactions, and
syntheses were readily scaled up to 5-11 kg without
significant deterioration in rates or yields. Thus, at reaction
scales of 11.43, 6.30, and 5.83 kg (11.54, 6.17, and 4.96
mol scale respectively), 1, 2, and 3 were synthesized with
yields of 71, 73, and 71%, and with product concentrations
was also evaluated (Table 2). The shaken reactor and
rotating-drum configurations produced little discernible me-
chanical damage to the catalysts, while the wiped-blade
format did result in significant abrasion of catalyst particles
and generated 6-13% of fines. However, good activity
retentions were observed, with the Celite-deposited catalysts
from the syntheses of 1, 2, and 3 being recovered with
remaining activities of 67-77, 61-69, and 81-88%, re-
spectively, after extraction of the reaction mixtures with
ethanol. Indeed, the biocatalysts were reused 6-9 times
before the activities dropped below the economically ac-
ceptable level of 30%, and this reusability furnished net
productivities of 166, 235, and 312 kg of product per kg of
enzyme, respectively.
-1
of 0.36, 0.48, and 0.48 kg kg . These correspond to space-
-
1
-1
time yields of 0.30, 0.64, and 0.30 kg kg d , well above
To evaluate the flexibility of the methodology with regard
to multistep protocols, we attempted to carry out the N-to-C
synthesis of 3 in one-pot and three steps, using the sequential
addition of the requisite substrates and enzymes. Indeed,
providing that water and ethanol were added to replace the
-
1
-1
the acceptable margin of 0.06 kg kg
d
that was required
for industrial scale production of the peptides. An important
concern for scale-up economics, namely physical deteriora-
tion of the biocatalysts and deterioration in catalyst activity
688
•
Vol. 6, No. 5, 2002 / Organic Process Research & Development

solvents (adjuvants) lost due to evaporation and re-liquefy
the mixtures upon the completion of each step, the coupling
reactions could be readily integrated in sequence without
undue loss of yield. Thus, 3 was synthesized with a yield of
using broad-band proton decoupling and a solvent field lock.
FAB-MS spectra were recorded on a Kratos MS9 instrument
using a xenon gun operating at 7 kV. Samples were dispersed
in glycerol and applied to a copper probe, and spectra were
recorded in positive ionization mode using polyglycerol ions
as reference. Optical rotations were measured on an AA-
100 polarimeter at 589.3 nm, 20 °C, using 0.25- or 0.5-cm
path length cells, with methanol as solvent, and are given in
3
4% (isolated yield 30%), a product concentration of 0.25
-
1
-1 -1
kg kg , and a space-time yield of 0.07 kg kg d , at a
net) reaction scale of 5.98 kg (4.83 mol) in a 20-L wiped-
(
blade reactor. The corresponding coupling yields of 82, 66,
and 63% for the individual steps, and the overall performance
compares very well with the net yield of 36-38% obtained
when the reactions are carried out separately in heterogeneous
media.
-
1
2
-1
units of 10 deg cm g . Melting points were determined
on a Buchi 540 melting point apparatus using sealed
capillaries, and are uncorrected.
The formation of eutectics and the determine determina-
tion of eutectic melting points was established by DSC. This
was performed with a Perkin-Elmer DSC-7 instrument with
a TAC controller and with nitrogen purging. Substrate
mixtures (ca. 10-20 mg) were sealed under nitrogen in
Conclusions
Protease-mediated peptide synthesis in heterogeneous
solid-liquid mixtures composed of eutectics plus excess
solid substrate together with small amounts of ethanol and
water can be readily scaled up in wiped-blade and rotating-
drum reactors to produce kilogram amounts of di-, tri-, and
tetrapeptides in high yields and productivities. Although these
reactor configurations are not appropriate for multiton
batches, the high productivity of the reaction systems should
-
1
5
°
0-100-µL aluminium pans, heated at 5 °C min to 70
-
1
C, held at this temperature for 5 min, cooled at 5 °C min
to -20 °C, and then equilibrated at this temperature for
24-48 h. The frozen samples were then analysed at a
-
1
ramp of 1 °C min and melting and freezing peaks
determined. Calibration was performed with hexane, distilled
water, octan-1-ol, and acetamide. In addition the eutectic
temperatures were also determined from contact slides (ca.
100-300 µm thickness, ca. 100-500 µm contact zone) by
polarizing light microscopy analysis with a Zeiss Axiophot
photomicroscope fitted with a heating stage and sealed
chamber with nitrogen purging.
3
enable the use of relatively compact reactors (a 1 m reactor
with a 30% v/v working volume could deliver ca. 110-150
kg of product per cycle) which should be effective for
performing the syntheses on the hundreds-of-kilogram scale,
which is sufficient for the commercial production of such
oligopeptides. The results suggest that the methodology
provides facile, highly efficient, and economic protocols
suitable for the industrial production of short peptides. We
are further exploring the synthesis of peptides containing
multifunctional amino acids in heterogeneous media and are
studying the development of reactor configurations suitable
for performing the syntheses on the hundreds-of-kilogram
scale.
The volume fractions of the solid and liquid phases of
heterogeneous mixtures were determined by centrifugation
analysis. The substrates (ca. 5-10 g) were mixed together,
solvent (adjuvant) was added as required, and the mixture
was heated in a centrifugal microfilter tube (10 or 20 mL,
-
1
0.5 µm nylon filter) at 5 °C min to 70 °C in a heating
-
1
block, cooled at 5 °C min to 40 °C, and maintained at this
temperature for 10 min. A weighed amount of biocatalyst
was added as required, and the sample was incubated for
the required time and then centrifuged (5000g, 20 min). The
expelled liquid phase was collected and weighed, and its
density and composition were determined by microdensity
bottle and HPLC methods. The retained solid phase was
separately analysed by HPLC. These results were used to
calculate the fractions of the solid and liquid phases.
Preparative Liquid Chromatography. Preparative col-
umn chromatography was performed using a jacketed 12 cm
Experimental Section
Chemicals and Enzymes. Subtilisin, chymopapain, dithio-
threitol (DTT), N
L-Asp(OAll) ‚TosOH, L-Glu(OAll)
HCl, L-LeuOEt‚HCl, and L-AlaNH
R
-Cbz-L-Lys(N
ú
-Cbz)OH, GlyOEt‚HCl,
‚TosOH, L-Ser(OH)OEt‚
‚HBr were purchased
2
2
2
from Sigma Chemical Co. Celite, thionyl chloride, and all
solvents were obtained from Aldrich Chemical Co.
Analytical Methods. Reactions were monitored by
reverse phase HPLC using an Agilent 1100 LC system
equipped with a diode array detector and HP Chemstation
data collection and processing software. Samples were
dissolved/quenched in pure methanol, centrifuged (10000g,
×
120 cm Buchi glass chromatography column connected
to Buchi 688 LC or Ismatec Gear pumps and a Buchi UV-
vis detector fitted with a preparative flow cell. The column
was slurry packed with 40-63 µm Vydac 218TP, thermo-
stated at 40 °C, and step-eluted with mixtures of pure
methanol and 9:1 water-methanol.
1
0
0 min), then analyzed on a Hichrom RPB column, 5 µm,
.46 cm × 15 cm, maintained at 45 °C, and eluted at 1 mL
-1
min with 9:1 water-methanol (A) and methanol (B), each
containing 0.05% v/v of phosphoric acid, with detection at
Preparation of Celite-Deposited Enzymes. The pro-
teases were deposited onto Celite at a loading of 10% w/w
as follows: An ice-cold solution of the protease (4.5 L,
protease dissolved at 40 g L in 100 mM potassium
phosphate buffer, preadjusted to pH 9, containing 10 mM
each of calcium and magnesium acetates and 5 mM EDTA)
was added slowly to Celite powder (5 kg) in an inclined
2
20 and 257 nm. The following elution conditions were
used: 90:10 (A:B) to 40:60 (A:B) over 20 min; 70:30
A:B) to 10:90 (A:B) over 20 min, with a 5 min hold at
-
1
(
1
0:90 (A:B).
1H NMR spectra were recorded on a JEOL EX270 FT
6
spectrometer at 270 MHz, 40 °C, with d -DMSO as solvent,
Vol. 6, No. 5, 2002 / Organic Process Research & Development
•
689

rotating-drum mixer (20 L, operated at ca. 70-100 rpm) and
the wet adsorbate thoroughly blended for 10 min. The wet
powder was then dried at room temperature in a Buchi
fluidized bed dryer until the moisture content was below 2%
w/w, and the obtained Celite-deposited biocatalyst was stored
at 5 °C until used. Immobilization efficiencies (100 ×
fraction of applied protein immobilized × relative specific
activity of immobilized protein in an aqueous activity assay)
of 93 and 89% were obtained for subtilisin and chymopapain,
respectively.
After 28 h, a semisolid mixture was obtained, and the yield
of dipeptide had reached 76%. Ethanol (ca. 4 L) was added,
and the suspension was stirred for 20 min and filtered, and
the filtrate containing the product was collected and rotary-
evaporated to dryness at room temperature. The recovered
biocatalyst retained 73% of its initial activity. The sticky
solid (ca. 2.3 kg) was dissolved in methanol (ca. 2.4 L) and
purified by column chromatography (step elution with 7:3,
then 5:5, then 4:6 water-methanol) to give the dipeptide as
a white solid: 1.031 kg; 69% yield; 98-99% pure (by
2
0
1
R ú
Preparation of Substrates. N -Cbz-L-Lys(N -Cbz)OEt
HPLC); mp 90-92 °C; [R]D -12.3° (c 0.10, MeOH); H
NMR (DMSO-d , 40 °C): d 1.18 (t, 3H, J ) 7.2 Hz), 1.25-
.47 (m, 4H), 1.57 (m, 1H), 1.65 (m, 1H), 2.97 (m, 2H),
.74 (dd, 1H, J ) 6.0 Hz, 16.7 Hz), 3.88 (dd, 1H, J ) 5.7
was prepared from the corresponding acid using the thionyl
chloride method.39 L-Asp(OAll)OAll and L-Glu(OAll)OAll
were prepared from the corresponding tosylate salts by
neutralizing ice-cold concentrated aqueous solutions with
aqueous sodium hydroxide, followed by extraction with ethyl
acetate or ether and rotary evaporation. The obtained free
bases were stored at -20 °C until use.
Reactors for Scale-Up of Heterogeneous Reactions.
Two types of reactor configurations were employed for scale-
up trials, as illustrated in Figure 2: (a) wiped-blade reactors,
consisting of jacketed, open-top cylindrical glass vessels with
volumes of 0.5, 1, 2, 5, 10, 20, and 50 L, with aspect ratios
of ca. 1.5:1-2.2:1 (height: diameter). Each reactor was fitted
with an overhead stirrer carrying a turbine consisting of six
6
1
3
Hz, 16.9 Hz), 4.02 (dt, 1H, J ) 7.0 Hz, 7.2 Hz), 4.08 (q,
2
7
H, J ) 6.9 Hz), 4.98 (s, 2H), 5.03 (s, 2H), 7.15 (m, 1H),
.26 (m, 1H), 7.31-7.46 (m, 10H), 8.26 (dd, 1H, J ) 5.6
Hz, 5.8 Hz); FAB-MS (M + H) calculated for C26
00.2396, observed 500.2438.
Preparation of N -Cbz-L-Lys(N
l)OAll (2). N -Cbz-L-Lys(N -Cbz)-GlyOEt (1, 936 g, 1.87
34 3 7
H N O
5
r
ú
-Cbz)-Gly-L-Asp(OAl-
R
ú
mol) was mixed with L-Asp(OAll)OAll (600 g, 2.80 mol),
water (81 g), ethanol (310 g), and DTT (5 g) in an open
1
0-L wiped-blade reactor, and the slurry was heated with
stirring (200 rpm) to 50 °C to form a viscous suspension.
The final water and ethanol contents were ca. 4.1% w/w and
40° inclined polyethylene blades, each with a width ca. 0.2
×
reactor vessel height and a thickness of 1-6 mm. Three
1
5.4% w/w, respectively. The mixture was allowed to cool
of the blades were of length ca. 0.25 × vessel diameter, while
the remaining blades were of length ca. 0.52 × vessel
diameter, and these blades wiped the reactor walls. (b)
inclined-rotating-drum reactors, consisting of cylindrical
polypropylene or stainless steel vessels with nominal volumes
of 4, 12, and 46 L, with aspect ratios of ca. 2.1:1-2.5:1
to 40 °C, Celite-deposited chymopapain (200 g) was added,
and stirring was continued (ca. 190 rpm) at 40 °C for 18 h.
Tripeptide formation was accompanied by near complete
liquefaction of the reaction mixture, with a homogeneous
liquid being formed at ca. 3-7 h, followed by an increase
in viscosity and gradual precipitation of the tripeptide
thereafter. After 18 h, a viscous semisolid mixture was
obtained, and the yield of tripeptide had reached 71%.
Ethanol (ca. 1.7 L) was added, the suspension was stirred
for 20 min and filtered, and the filtrate containing the product
was collected and rotary evaporated at room temperature to
a pale yellow syrup (ca. 1.2 kg). The recovered biocatalyst
retained 69% of its initial activity. The crude product was
dissolved in methanol (ca. 1.5 L) and purified by column
chromatography (step elution with 8:2, then 6:4, then 5:5,
then 4:6 water-methanol) to furnish the tripeptide as a white
solid: 0.799 kg; 64% yield; >99% pure (by HPLC); mp
(height: diameter). The reactors were fitted with six or eight
stainless steel or polyethylene baffles (triangular cross
section) of length ca. 0.7 × reactor length and of width ca.
0
.15 × reactor diameter, and were inclined at ca. 10-20°
(
to the horizontal) during operation.
Preparation of N -Cbz-L-Lys(N
Cbz-L-Lys(N -Cbz)OEt (1.30 kg, 3 mol) was mixed with
r
ú R
-Cbz)-GlyOEt (1). N -
ú
water (121 g) and ethanol (353 g) in an open 10-L wiped-
blade reactor, and the slurry was heated with stirring (200
rpm) to 70 °C to form a viscous liquid. This was cooled to
4
0 °C, and to this were added GlyOEt‚HCl (558 g, 4 mol),
diisopropylethylamine (505 g, 3.92 mol, 98 mol %), DTT
10 g) and a 1:1 mix of finely powdered potassium carbonate
and potassium hydrogen carbonate (50 g), and stirring (ca.
60 rpm) continued at 40 °C to furnish a semiliquid
2
0
1
1
(
(
15-116 °C; [R]D -7.6° (c 0.10, MeOH); H NMR
DMSO-d , 40 °C): 1.26-1.43 (m, 4H), 1.53 (m, 1H), 1.60
m, 1H), 2.75 (dd, 1H, J ) 6.3 Hz, 16.0 Hz), 2.88 (dd, 1H,
(
6
1
J ) 6.7 Hz, 16.1 Hz), 2.98 (m, 2H), 3.72 (dd, 1H, J ) 6.1
Hz, 15.7 Hz), 3.80 (dd, 1H, J ) 5.5 Hz, 15.3 Hz), 4.00 (dt,
heterogeneous multicomponent mixture. The final water and
ethanol contents were ca. 4.2 and 12.1% w/w, respectively.
Celite-deposited chymopapain (500 g) was added to the
mixture, and stirring continued (200 rpm) at 40 °C for 28 h.
The formation of dipeptide was accompanied by partial
liquefaction of the reaction mixture, with the mixture
transforming to a mobile liquid at 4-7 h, and this was
followed by an increase in viscosity as the reaction pro-
gressed further and the dipeptide partially crystallized out.
1H, J ) 6.5 Hz, 7.4 Hz), 4.51-4.62 (m, 4H), 4.75 (ddd,
1H, J ) 6.1, 6.8, 8.2 Hz), 4.99 (s, 2H), 5.03 (s, 2H), 5.15-
5.30 (m, 4H), 5.82-6.01 (m, 2H), 7.17 (t, 1H, J ) 6.4 Hz),
7.25-7.40 (m, 11H), 8.15 (dd, 1H, J ) 5.7 Hz, 6.3 Hz),
8.28 (dd, 1H, J ) 5.7 Hz, 5.9 Hz); FAB-MS (M + H)
calculated for C34
Preparation of N
L-Glu(OAll)OEt (3). N
(OAll)OAll (2, 720 g, 1.08 mol) was mixed with L-Glu-
H
43
N
4
O
10 667.2979, observed 667.2983.
-Cbz-L-Lys(N -Cbz)-Gly-L-Asp(OAll)-
-Cbz-L-Lys(N -Cbz)-Gly-L-Asp-
r
ú
R
ú
(39) Bodanzsky, M. Peptide Chemistry, 2nd ed.; Springer-Verlag: Berlin, 1993.
6
90 Vol. 6, No. 5, 2002 / Organic Process Research & Development
•

(
OAll)OAll (392 g, 1.73 mol), water (9 g), and ethanol (167
form a viscous liquid. This was cooled to 40 °C, and to this
were added GlyOEt‚HCl (0.898 kg, 6.44 mol), diisopropyl-
ethylamine (0.81 kg, 6.31 mol, 98 mol %), DTT (10 g), and
a 1:1 mix of finely powdered potassium carbonate and
potassium hydrogen carbonate (70 g). Stirring (ca. 160 rpm)
continued at 40 °C to furnish a semiliquid heterogeneous
mixture. Celite-deposited chymopapain (0.8 kg) was added
to the mixture, and stirring continued (200 rpm) at 40 °C
for 28 h, when 82% of dipeptide had been formed. Water
(0.10 kg) and ethanol (0.41 kg) were added to liquefy the
mixture, followed by L-Asp(OAll)OAll (0.56 kg, 2.60 mol)
and DTT (7 g), and the mixture was stirred (ca. 170 rpm)
for 20 min to give a viscous heterogeneous mixture. A further
portion of Celite-deposited chymopapain (0.26 kg) was
added, and the mixture was stirred at 40 °C for a further 23
h, by which time the yield of tripeptide had reached 54%
(66% based upon dipeptide), and a thick crystalline suspen-
sion had formed. Once again, this was liquefied by adding
water (0.06 kg) and ethanol (0.60 kg) and stirring, followed
by the addition of L-Glu(OAll)OAll (0.78 kg, 3.44 mol).
Stirring was continued (ca. 150 rpm) for 30 min to provide
a viscous suspension. Celite-deposited subtilisin (0.35 kg)
was added, and stirring was continued (ca. 130 rpm) at 40
°C for 32 h, when the tetrapeptide yield had reached 34%
(63% based upon tripeptide) and the mixture had partially
solidified. Ethanol (ca. 7.0 L) was added, the suspension was
stirred for ca. 1 h min. It was then filtered through Celite,
and the filtrate containing the product was collected and
rotary-evaporated at room temperature to a yellow syrup (ca.
5.3 kg). This was dissolved in methanol (ca. 4.8 L) and
purified by column chromatography (as above) in six batches
to furnish the tetrapeptide as a white solid: 1.22 kg; 30%
g) in an open 5-L wiped-blade reactor, and the slurry was
heated with stirring (200 rpm) to 50 °C to form a viscous
suspension. The final water and ethanol contents were ca.
1
.9% w/w and 12.6% w/w, respectively. The mixture was
cooled to 40 °C, Celite-deposited subtilisin (200 g) was
added, and stirring was continued (ca. 140 rpm) at 40 °C
for 38 h. The reaction was accompanied by near complete
liquefaction of the reaction mixture, with minimum viscosity
occurring at ca. 6-9 h, followed by a gradual increase in
viscosity and increasing precipitation of the tetrapeptide
thereafter. After 38 h, the reaction mixture had partially
solidified, and the yield of tetrapeptide had reached 73%.
Ethanol (ca. 2.1 L) was added, the suspension was stirred
for 20 min and filtered, and the filtrate containing the product
was collected and rotary evaporated at room temperature to
a pale yellow syrup (ca. 1.0 kg). The recovered biocatalyst
retained 78% of its initial activity. The crude product was
dissolved in methanol (ca. 0.8 L) and purified by column
chromatography (step elution with 8:2, then 6:4, then 5:5,
then 4:6, then 2:8 water-methanol) to give the tetrapeptide
as a white solid: 0.551 kg; 62% yield; > 99% pure (by
20
1
HPLC); mp 112-115 °C; [R]D -16.2° (c 0.10, MeOH); H
NMR (DMSO-d , 40 °C): 1.25 (t, 3H, J ) 7.3 Hz), 1.30-
.54 (m, 4H), 1.58 (m, 1H), 1.67 (m, 1H), 1.96 (m, 1H),
6
1
2
1
.08 (m, 1H), 2.50 (dd, 2H, J ) 7.4 Hz, 7.6 Hz), 2.68 (dd,
H, J ) 8.0 Hz, 15.9 Hz), 2.89 (dd, 1H, J ) 5.7 Hz, 15.5
Hz), 3.01 (dt, 2H, J ) 6.2 Hz, 6.5 Hz), 3.75 (d, 2H, J ) 5.0
Hz), 4.05 (m, 1H), 4.11 (q, 2H, J ) 7.5 Hz), 4.30 (m, 1H),
4
Hz), 5.09 (s, 2H), 5.13 (d, 1H, J ) 12.7 Hz), 5.17-5.35 (m,
4
.58-4.63 (m, 4H), 4.75 (m, 1H), 5.07 (d, 1H, J ) 12.8
H), 5.85-6.07 (m, 2H), 7.20 (t, 1H, J ) 6.3 Hz), 7.29-
.48 (m, 11H), 8.15-8.26 (m, 3H); FAB-MS (M + H)
20
7
yield; 98-99% pure (by HPLC); mp 111-113 °C; [R]
-15.8° (c 0.10, MeOH); H NMR spectrum identical to that
D
1
calculated for C41
One-Pot Synthesis of N
Asp(OAll)-L-Glu(OAll)OEt (3). N
2.10 kg, 4.83 mol) was mixed with water (0.19 kg) and
H
54
N
5
O
13 824.3718, observed 824.3741.
-Cbz-L-Lys(N -Cbz)-Gly-L-
-Cbz-L-Lys(N -Cbz)OEt
r
ú
above.
R
ú
(
Received for review March 7, 2002.
OP025528Y
ethanol (0.56 kg) in an open 20-L wiped-blade reactor, and
the slurry was heated with stirring (200 rpm) to 70 °C to
Vol. 6, No. 5, 2002 / Organic Process Research & Development
•
691