Enzyme-catalyzed asymmetric Mannich reaction using acylase from Aspergillus melleus

Article abstract of DOI:10.1016/j.molcatb.2014.11.007

Acylase I from Aspergillus melleus (AMA) displayed catalytic promiscuity towards one-pot asymmetric Mannich reaction in acetonitrile for the first time. AMA showed favourable catalytic activity with good adaptability to different substrates. The activity

Full text of DOI:10.1016/j.molcatb.2014.11.007

Accepted Manuscript
Title: Enzyme-catalyzed asymmetric Mannich reaction using
acylase from Aspergillus melleus
Author: Zhi Guan Jian Song Yang Xue Da-Cheng Yang
Yan-Hong He
PII:
S1381-1177(14)00311-7
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcatb.2014.11.007
MOLCAB 3061
To appear in:
Journal of Molecular Catalysis B: Enzymatic
Received date:
Revised date:
Accepted date:
11-2-2014
4-11-2014
22-11-2014
Please cite this article as: Z. Guan, J. Song, Y. Xue, D.-C. Yang, Y.-
H. He, Enzyme-catalyzed asymmetric Mannich reaction using acylase from
Aspergillus melleus, Journal of Molecular Catalysis B: Enzymatic (2014),
http://dx.doi.org/10.1016/j.molcatb.2014.11.007
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Enzyme-catalyzed asymmetric Mannich reaction using acylase from
Aspergillus melleus
Zhi Guan*, Jian Song, Yang Xue, Da-Cheng Yang, and Yan-Hong He*
School of Chemistry and Chemical Engineering, Southwest University, Chongqing, 400715, P. R. China
Fax: +86-23-68254091; e-mail: guanzhi@swu.edu.cn (for Zhi Guan); heyh@swu.edu.cn (for Yan-Hong He).
Abstract: Acylase I from Aspergillus melleus (AMA) displayed catalytic promiscuity towards
one-pot asymmetric Mannich reaction in acetonitrile for the first time. AMA showed favorable
catalytic activity with good adaptability to different substrates. The activity and stereoselectivity
of the enzyme could be improved by adjusting solvent, pH, water content, temperature, molar ratio
of substrates and enzyme loading. The enantioselectivities up to 89% ee, diastereoselectivities up
to 90:10 dr (syn/anti) and yields up to 82% were achieved. This work offered a novel case of
enzyme catalytic promiscuity and a potential synthetic method for organic chemistry.
Keywords: enzyme catalysis; asymmetric Mannich reaction; acylase from Aspergillus melleus;
enzymatic promiscuity
1. Introduction
Enzymes as sustainable, eco-friendly and biodegradable catalysts have attracted increasing
attention from organic chemists. In recent years, more and more enzymes have been revealed to be
able to catalyze “non-native reactions”, and this phenomenon is known as enzyme promiscuity. In
1
Page 1 of 19

many cases, these activities for the promiscuous substrates are usually very low in comparison
with their native activities, and hence might have little physiological relevance [1]. However,
promiscuous activities are useful for organic synthesis. On the other hand, an attractive hypothesis
in enzyme evolution relies on the promiscuous activities that are thought to be not rare exceptions
but are rather widely spread, inherent features of enzymes, which can serve as starting points in
evolution [2-4]. Therefore, the research on enzyme promiscuity not only gain insight on the
evolution of new enzymes but also expands the application of biocatalysts in organic synthesis.
Some elegant works on enzymatic promiscuity have been reported [5-9]. Several enzymes have
been used in asymmetric syntheses, including asymmetric aldol reactions [10,11], asymmetric
Michael additions [12], and asymmetric Henry reactions [13].
The asymmetric Mannich reaction is a powerful synthetic strategy for the preparation
of chiral β-amino ketones and aldehydes with perfect atom economy, and the reaction
products are versatile intermediates in the synthesis of chiral amines [14]. In recent years,
though numerous successful organocatalysts and metal catalysts for asymmetric Mannich
reactions have been described with high efficiency and enantioselectivity [15], sustainable
catalysts are necessary to complement the traditional approaches to asymmetric catalysis.
Although some enzymes have been found to be able to catalyze Mannich reactions, but no
enantioselectivity was obtained [16-19], until we recently reported a protease-catalyzed
asymmetric Mannich reaction using protease type XIV from Streptomyces griseus, to date,
which is the only reported enzyme-catalyzed asymmetric Mannich reaction [20]. Since
enzymatic promiscuity is still at the exploratory stage, and no general methods are available to
2
Page 2 of 19

profile enzyme catalytic promiscuity, it is still a formidable but significant task to explore enzyme
promiscuous activities as much as possible. Herein, we report another enzyme-catalyzed
asymmetric Mannich reaction using acylase I (aminoacylase) from Aspergillus melleus
(AMA). Acylase I (N-acyl-
L
-amino-acid amidohydrolase, EC 3.5.1.14) which could
-amino acids,
catalyze the hydrolysis of -acyl-amino acids to produce the corresponding
L
L
is an important enzyme involving in amino acid metabolism in organism. The
aminoacylase is one of the top 10 enzymes used in biotechnology [21], which provides the
potential possibility for the application of this enzymatic asymmetric Mannich reaction in
industry.
2. Experimental
2.1 Materials and analytical methods
Flash column chromatography was carried out using 200-300 mesh silica gel at
increased pressure. The NMR spectra were recorded on a Bruker AMX-300MHz
spectrometer. Chemical shifts (δ) were expressed in ppm with TMS as internal standard,
and coupling constants (J) were reported in Hz. The enantiomeric excess (ee) of the
Mannich products were determined by chiral HPLC analysis performed using Chiralpak
AD-H, AS-H and Chiralcel OD-H, OJ-H columns. Relative and absolute configurations of
1
the products were determined by comparison with the known H NMR, 13C NMR and
chiral HPLC analysis from references.
Acylase I from Aspergillus melleus (AMA) (Sigma-Aldrich catalogue no. 01818,
3
Page 3 of 19

Lot#1348941V, 0.58 U/mg. EC 3.5.1.14, powder. 1 U corresponds to the amount of
enzyme which hydrolyzes 1 μmol N-acetyl-L-methionine per minute at pH 8.0 and 37°C)
was purchased from Sigma-Aldrich.
2.2 General procedure for the AMA-catalyzed asymmetric Mannich reaction
AMA (50 mg, 29.0 U) was added to a round-bottom flask containing aromatic
aldehyde (0.50 mmol), anarylamine (0.55 mmol), ketone (5.0 mmol), MeCN (0.85 mL)
and buffer (NaH2PO4-Na2HPO4, 0.2 M, pH 8.1, 0.15 mL). The resultant mixture was
stirred at 30˚C for the specified reaction time and monitored by TLC on Haiyang GF 254
silica gel plates. The reaction was terminated by filtering the enzyme. The filter cake was
washed with ethyl acetate (10 mL). Then, the filtrate was concentrated in vacuo. The
residue was purified by flash column chromatography (ethyl acetate/petroleum ether) to
give the product.
3. Results and Discussion
Initial studies were undertaken using cyclohexanone, 4-nitrobenzaldehyde and aniline as a
model reaction. It is widely proved that enzymatic activities and specificities in non-aqueous
media are markedly affected by the solvents [22]. Thus, we first explored the influence of organic
solvents on this biotransformation. The results indicated that solvents had obvious influence on the
AMA-catalyzed Mannich reaction (Table 1). Among the tested solvents, the best
enantioselectivity of 66% ee for syn-isomer was obtained in MeCN (Table 1, entry 1). On the
contrary, no enantioselectivity was observed in n-butyl acetate (Table 1, entry 17). The highest
4
Page 4 of 19

yield of 81% was gained under the solvent-free conditions (Table 1, entry 9) while the lowest
yield of 37% was obtained in isopropyl ether (Table 1, entry 12). The reaction in phosphate buffer
(0.2 M, pH 7.0) or water gave the product in moderate yields with low selectivity (Table 1, entries
15 and 16). To obtain the best enantioselectivity, we chose MeCN as the optimal solvent for the
following study.
To verify the specific catalytic effect of AMA on the Mannich reaction, some control
experiments were designed and performed (Table 1, entries 18-24). In the absence of the enzyme,
the product was obtained in a yield of 28% with 41:59 dr (syn:anti) and 0% ee (Table 1, entry 18),
indicating that the substrates of the model Mannich reaction had certain reactivity. However, in
comparison with the reaction in the presence of AMA (50 mg, 29.0 U), which gave the product in
a yield of 77% with 66:34 dr (syn:anti) and 66% ee (syn) (Table 1, entry 1) , it was obvious that
the addition of AMA not only could improve the reaction yield and diastereoselectivity, but also
could generate the enantioselectivity. In other words, the preparation of AMA indeed catalyzed the
model Mannich reaction in an asymmetric manner. Next, non-enzyme proteins chicken egg
albumin and bovine serum albumin were used to catalyze the reaction, and the products were
obtained in 23% yield with 0% ee and 21% yield with 7% ee, respectively (Table 1, entries 19 and
20), suggesting that only amino acid residues on the surface of the protein hardly catalyzed the
model Mannich reaction. In addition, AMA is well characterized as a Zn(II)-containing enzyme
[23-25], however the exact structure of the active site is not known. The mechanism for its natural
reaction is simply proposed on the basis of the mechanisms of Zn(II)-containing
carboxypeptidases [26]. To verify if this Zn(II)-containing enzyme, AMA, was responsible for the
observed reaction results, we attempted to take away some Zn(II) from AMA, and to test the
5
Page 5 of 19

catalytic ability of the enzyme that lost some Zn(II) in the model Mannich reaction. Thus,
phenanthroline and EDTA tetra sodium were employed as complexants to pretreat AMA
separately, and the complexant-treated AMA was then used to catalyze the model Mannich
reaction. The yields and selectivity of the reaction with complexant-pretreated AMA were reduced
(The yield was reduced from 77% to 62% and 60%, and ee was reduced from 66% to 38% and
34%, respectively) (Table 1, entries 1, 21 and 22), suggesting that a metal-containing enzyme was
related to the observed Mannich reaction. In addition, the urea pretreated AMA was also used to
catalyze the model reaction, giving the product in a yield of 57% with 62:38 dr and 21% ee (Table
1, entry 23). As a comparison, the reaction with AMA under the same reaction conditions was
performed which gave a good yield of 73% with 74:26 dr and 73% ee (Table 1, entry 24). The
results suggested that the catalytic ability and selectivity of AMA were reduced after urea
treatment. Thus, it could be inferred that enzyme was responsible for this Mannich reaction.
Table 1. Influence of solvents on the AMA-catalyzed Mannich reaction and control
experimentsa
O
CHO
NO2
NH2
O
HN
AMA
+
+
solvent / H2O, 30 oC
NO2
% ee (syn)c
Entry
Solvent
Yield (%)b
dr (syn:anti)c
1
2
3
4
5
6
7
8
MeCN
THF
77
65
63
66
46
64
60
43
66:34
64:36
64:36
55:45
68:32
57:43
55:45
60:40
66
61
58
56
53
51
50
49
EtOH
Isopropanol
DMF
CH2Cl2
MTBE
DMSO
6
Page 6 of 19

9
Solvent-free
MeOH
81
58
65
37
64
64
62
68
68
28
23
21
62
60
57
73
55:45
59:41
50:50
50:50
42:58
41:59
57:43
42:58
42:58
41:59
46:54
58:42
60:40
56:44
62:38
74:26
49
46
45
42
25
22
22
19
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Dioxane
Isopropyl ether
Cyclohexane
Ethyl acetate
Phosphate buffer (pH 7.0)
H2O
n-Butyl acetate
MeCN (no enzyme)
0
MeCN (chicken egg albumin)d
MeCN (bovine serum albumin)e
MeCN (AMA pretreated with phenanthroline)f
MeCN (AMA pretreated with EDTA tetra sodium)g
MeCN (AMA pretreated with urea)h, i
MeCNh
0
7
38
34
21
73
[a] Unless otherwise noted, the reaction conditions: a mixture of 4-nitrobenzaldehyde (0.50 mmol), aniline (0.55
mmol), cyclohexanone (5.0 mmol), deionized water (0.10 mL), solvent (0.90 mL), and AMA (50 mg, 29.0 U) was
stirred at 30 ˚C for 96 h. [b] Isolated yield after silica gel chromatography. [c] Determined by chiral HPLC analysis.
[d] Chicken egg albumin (50 mg) instead of AMA was used. [e] Bovine serum albumin (50 mg) instesd of AMA
was used. [f] A mixture of AMA (50 mg, 29.0 U) in 0.25 mM phenanthroline solution (1.0 mL) was stirred at 30˚C
for 24 h, and then water was removed under reduced pressure before use. [g] A mixture of AMA (50 mg, 29.0 U
)
in 0.74 mM EDTA tetra sodium (1.0 mL) was stirred at 30˚C for 24 h, and then water was removed under reduced
pressure before use. [h] Reaction conditions: the same as described in note [a] except deionized water (0.3 mL),
solvent (2.7 mL) was used. [i] A mixture of AMA (50 mg, 29.0 U) in 8 M urea solution (1.0 mL) was stirred at
30˚C for 24 h, and then water was removed by lyophilization before use.
Enzymes in non-aqueous solvents require water to provide them with sufficient
conformational flexibility needed for catalysis [27]. Thus, it is significant to confirm the optimal
percentage of water in the reaction system. As can be seen from Table 2, although the yield of the
7
Page 7 of 19

reaction was not obviously affected by the water content, the stereoselectivity of AMA was
dramatically influenced by the water contents. In the absence of water, a low selectivity of 11% ee
was received (Table 2, entry 1). However, at 15% water content [water / (water + MeCN), v/v] a
good selectivity of 73% ee was obtained (Table 2, entry 4). Further increase of water content led
to a decrease of the selectivity (Table 2, entries 5-8). The results indicated that appropriate amount
of water was essential for AMA to display its selectivity. Therefore, we chose 15% water content
for the following study.
Table 2. Influence of water content on the AMA-catalyzed Mannich reactiona
Entry
Water content (%)
Yield (%)b
dr (syn:anti)c
% ee (syn)c
1
2
3
4
5
6
7
8
0
77
78
77
72
72
73
73
75
46:54
49:51
66:34
76:24
66:34
68:32
64:36
57:43
11
25
66
73
67
67
61
51
5
10
15
20
25
30
35
[a] Reaction conditions: a mixture of 4-nitrobenzaldehyde (0.50 mmol), aniline (0.55 mmol), cyclohexanone (5.0
mmol), deionized water [0-35%, H2O / (H2O + MeCN), v/v, H2O + MeCN = 1.0 mL], and AMA (50 mg, 29.0 U
)
was stirred at 30 ˚C for 96 h. [b] Isolated yield after silica gel chromatography. [c] Determined by chiral HPLC
analysis.
The proportion of the substrates also affects the reaction. Thus, in order to further improve
the selectivity and yield of AMA-catalyzed Mannich reaction, the molar ratios of substrates were
investigated (Table 3). When the molar ratio of 5:1 (cyclohexanone to 4-nitrobenzaldehyde) was
used, the reaction gave the highest yield of 89% but with a low ee of 64% (Table 3, entry 2). The
8
Page 8 of 19

best ee value of 77% with a lower yield of 71% was obtained using the molar ratio of 10:1 (Table
3, entry 3). Further increasing the molar ratio of cyclohexanone to 4-nitrobenzaldehyde would
decrease both the yield and selectivity. To obtain the better selectivity, we chose the molar ratio of
ketone to aldehyde 10:1 as an optimal condition.
Table 3. Effect of the molar ratio of substrates on the AMA-catalyzed Mannich reactiona
Entry
1
Molar ratiob
1:1
Yield (%)c
61
dr (syn:anti)d
% ee (syn)d
51:49
40
2
3
4
5
6
5:1
89
71
63
61
41
66:34
81:19
78:22
76:24
65:35
64
77
72
68
58
10:1
15:1
20:1
25:1
[a] Conditions: a mixture of 4-nitrobenzaldehyde (0.5 mmol), aniline (0.55 mmol), cyclohexanone (0.5-12.5 mmol),
deionized water (0.15 mL), MeCN (0.85 mL), and AMA (50 mg, 29.0 U) was stirred at 30 ˚C for 96 h. [b]
Cyclohexanone to 4-nitrobenzaldehyde. [c] Isolated yield after silica gel chromatography. [d] Determined by chiral
HPLC analysis.
Effect of the enzyme loading on the AMA-catalyzed Mannich reaction was investigated
(Table 4). When 5 mg (2.9 U) of AMA was used, the model reaction only gave the product in a
low yield of 30% with 58:42 dr and 37% ee (Table 4, entry 1). Increasing the enzyme loading
from 5 mg (2.9 U) to 50 mg (29.0 U) led to an obvious increase in both of the yield and selectivity
(Table 4, entries 1-6). Further increasing the enzyme loading from 50 mg (29.0 U) to 125 mg
(72.5 U) hardly improved the results. Thus, 50 mg (29.0 U) of AMA was chosen as the optimal
enzyme loading for the reaction (0.50 mmol scale in 1.0 mL of MeCN/H2O).
9
Page 9 of 19

Table 4. Effect of the enzyme loading on the AMA-catalyzed Mannich reactiona
Entry
Enzyme loading
5 mg (2.9 U)
Yield (%)b
dr (syn:anti)c
% ee (syn)c
1
2
3
4
5
6
7
8
9
30
42
58
60
62
70
72
71
72
58:42
37
50
68
66
66
73
72
73
72
10 mg (5.8 U)
15 mg (8.7 U)
20 mg (11.6 U)
25 mg (14.5 U)
50 mg (29.0 U)
75 mg (43.5 U)
100 mg (58.0 U)
125 mg (72.5 U)
60:40
67:33
64:36
66:34
77:23
76:24
79:21
77:23
[a] Reaction conditions: a mixture of 4-nitrobenzaldehyde (0.50 mmol), aniline (0.55 mmol), cyclohexanone (5.0
mmol), deionized water (0.15 mL), MeCN (0.85 mL), and AMA (5-125 mg, 2.9-72.5 U) was stirred at 30 ˚C for 96
h. [b] Isolated yield after silica gel chromatography. [c] Determined by chiral HPLC analysis.
The pH of reaction medium plays a significant role in maintaining the stability and catalytic
activity of enzymes. We explored effects of pH on the AMA-catalyzed Mannich reaction (Table
5). Phosphate buffer (0.2 M, NaH2PO4-Na2HPO4, pH 5.5-9.1) was used to replace the optimized
water content in the reaction system [buffer / (buffer + MeCN) = 15%, v/v]. Blank experiments
(without enzyme) were also conducted to investigate the influence of different pH buffers on the
reaction. It can be seen that both acidic and basic buffers could catalyze the Mannich reaction, but
only showed very low or no diastereoselectivity without ee (Table 5, entries 1-6). Comparison
between the blank and the reaction in the presence of the enzyme indicated that although buffer
itself had effect on the reaction, the addition of pH 7.0-9.1 buffers could improve the
stereoselectivity of this enzymatic reaction obviously [from 77:23 dr and 73 % ee (Table 4, entry
6) to no less than 84:16 dr and 82% ee (Table 5, entries 3-6)]. The best selectivity (84% ee and
88:12 dr) was obtained in the presence of pH 8.1 buffer (Table 5, entry 5). Thus,
10
Page 10 of 19

MeCN-phosphate buffer (pH 8.1) (buffer / (buffer + MeCN) = 15%, v/v] was selected as the
optimal reaction medium for the AMA-catalyzed Mannich reaction.
Table 5. Effect of pH on the AMA-catalyzed Mannich reactiona
AMA catalysis
dr (syn:anti)c
61:39
Blank (without enzyme)
Entry
Buffer pH
Yield (%)b
% ee (syn)c
Yield (%)b
dr (syn:anti)c
% eec
1
2
3
4
5
6
5.5
6.5
7.0
7.5
8.1
9.1
87
80
67
74
77
82
57
74
82
83
84
82
52
49
32
36
39
47
58:42
0
0
0
0
0
0
77:23
53:47
85:15
49:51
86:14
50:50
88:12
47:53
84:16
49:51
[a] Reaction conditions: a mixture of 4-nitrobenzaldehyde (0.50 mmol), aniline (0.55 mmol), cyclohexanone (5.0
mmol), phosphate buffer (0.2 M, NaH2PO4-Na2HPO4, pH 5.5-9.1, 0.15 mL) , MeCN (0.85 mL), and AMA (50 mg,
29.0 U) was stirred at 30 ˚C for 96 h. [b] Isolated yield after silica gel chromatography. [c] Determined by chiral
HPLC analysis.
All enzymes need a favorable temperature to work properly. Thus, a temperature screening
was performed (Table 6). The yield and selectivity of AMA-catalyzed Mannich reaction could be
influenced obviously by temperature. The best yield of 77 % with the best ee of 84% was achieved
at 30 °C (Table 6, entry 4). The reaction at higher temperature gave lower yield with slightly
lower selectivity, probably due to the partial denaturation of the enzyme at higher temperature.
Therefore, we chose 30 oC as an optimal temperature for the reaction.
Table 6. Influence of temperature on the AMA-catalyzed Mannich reactiona
Entry
T (oC)
Yield (%)b
dr (syn:anti)c
% ee (syn)c
1
2
3
4
5
15
20
25
30
35
67
67
73
77
70
45:55
51:49
73:27
88:12
87:13
36
45
71
84
83
11
Page 11 of 19

6
7
8
40
45
50
64
65
60
85:15
84:16
84:16
83
82
79
[a] Reaction conditions: a mixture of 4-nitrobenzaldehyde (0.5 mmol), aniline (0.55 mmol), cyclohexanone (5.0
mmol), buffer (0.2 M, NaH2PO4-Na2HPO4, pH 8.1, 0.15 mL), MeCN (0.85 mL), and AMA (50 mg, 29.0 U) was
stirred at 15-50 ˚C for 96 h. [b] Isolated yield after silica gel chromatography. [c] Determined by chiral HPLC
analysis.
The time course of the AMA-catalyzed Mannich reaction at 30, 40 and 50 oC was
investigated (Table 7). Generally, both stereoselectivity and yield increased as the reaction time
went on at early stage and finally kept basically unchangeable. Specifically, the influence of
reaction time was more remarkable on dr and ee values than on yield. The best yield and
selectivity (up to 80 % yield, 88:12 dr and 84 % ee) were obtained at 30 oC. The lower yield and
selectivity were received at 40 and 50 oC, probably due to the partly denaturation of the enzyme at
higher temperatures.
Table 7. The time course of the AMA-catalyzed Mannich reaction at different temperaturesa
Entry Time (h)
30 oC
drc
40 oC
drc
50 oC
drc
Yield
(%)b
56
% eed
Yield
(%)b
49
% eed
Yield
(%)b
47
% eed
1
2
3
4
5
6
7
8
12
24
46:54
57:43
65:35
81:19
85:15
88:12
88:12
87:13
39
53
64
75
80
84
84
84
60:40
54
63
75
81
82
83
--
62:38
60
67
69
78
77
79
--
58
60
65:35
83:17
86:14
84:16
85:15
--
50
67:33
86:14
83:17
85:15
84:16
--
36
60
63
52
48
62
68
54
72
65
65
61
96
77
64
60
120
144
79
--
--
80
--
--
--
--
--
--
[a] Reaction conditions: a mixture of 4-nitrobenzaldehyde (0.5 mmol), aniline (0.55 mmol), cyclohexanone (5.0
mmol), buffer (0.2 M, NaH2PO4-Na2HPO4, pH 8.1, 0.15 mL), MeCN (0.85 mL), and AMA (50 mg, 29.0 U) was
12
Page 12 of 19

o
stirred at 30, 40 or 50 C for specified time. [b] Isolated product. [c] dr of syn/anti, determined by chiral HPLC
analysis. [d] ee values of syn-isomer, determined by chiral HPLC.
Finally, with the optimized conditions in hand, we further investigated the substrate scope
and the generality of the AMA-catalyzed Mannich reaction. Some other aromatic aldehydes,
ketones and aromatic amines were tested (Table 8). Generally, aromatic aldehydes with
electron-withdrawing substituents showed better stereoselectivity than those with
electron-donating or neutral substituents (Table 8, entries 1-9). Besides aniline, various substituted
arylamines were also used as substrates (Table 8, entries 10-15). It was interesting that
heterocyclic ketone tetrahydrothiopyran-4-one was also tolerant as a substrate to the
transformation (Table 8, entries 16 and 17). Among these examples, the best enantioselectivity of
89% ee and the best diastereoselectivity of 90:10 dr (syn:anti) was obtained for the reaction of
3-bromobenzenamine, 4-nitrobenzaldehyde and cyclohexanone (Table 8, entry 10). In addition,
acetone as an acyclic ketone was also used to react with 4-methoxybenzenamine and
4-nitrobenzaldehyde, but the product was only obtained in 22 % yield without enantioselectivity
after 168 h. Notably, AMA showed different degrees of enantioselectivity for syn-isomers, but low
or no enantioselectivity for anti-isomers.
Table 8. Scope of the AMA-catalyzed Mannich reactionsa
NH2
R2
O
O
O
X
HN
AMA
+
+
H
R1
R1
MeCN / buffer, 30 oC
R2
X
1
2
3
4
Entry
1
X
R1
CH2 4-NO2C6H4
R2
Product Time (h) Yield (%)b dr (syn:anti)c % ee (syn)c
4a
168 80 88:12 84
H
13
Page 13 of 19

2
3
CH2 4-CNC6H4
H
H
4b
4c
4d
4e
4f
161
144
166
119
166
168
96
82
82
81
78
77
82
75
75
31
58
73
74
74
63
79
80
65:35
80:20
81:19
72:28
65:35
63:37
63:37
59:41
90:10
87:13
87:13
84:16
70:30
43:57
50:50
48:52
82
79
76
72
71
66
56
44
89
84
83
81
73
23
54
52
CH2
CH2
4-BrC6H4
4-ClC6H4
4
H
5
CH2 4-CF3C6H4
H
6
CH2
CH2
CH2
3-FC6H4
4-FC6H4
C6H5
H
7
H
4g
4h
4i
8
H
9
CH2 4-MeC6H4
CH2 4-NO2C6H4
CH2 4-NO2C6H4
CH2 4-NO2C6H4
CH2 4-NO2C6H4
H
192
166
189
142
120
166
168
168
168
10
11
12
13
14
15
16d
17d
3-Br
4-Cl
3-Me
4-Me
4j
4k
4l
4m
4n
4o
4p
4q
CH2 4-NO2C6H4 4-OMe
CH2
S
4-BrC6H4
4-ClC6H4
4-CF3C6H4
4-OMe
H
H
S
[a] Reaction conditions: a mixture of aromatic aldehyde (0.50 mmol), arylamine (0.55 mmol), ketone (5.0 mmol),
buffer (0.2 M, NaH2PO4-Na2HPO4, pH 8.1, 0.15 mL), MeCN (0.85 mL), and AMA (50 mg, 29.0 U) was stirred at
30 oC. [b] Isolated yield after silica gel chromatography. [c] Determined by chiral HPLC analysis. [d]
tetrahydrothiopyran-4-one (1.0 mmol).
4. Conclusion
We reported the AMA-catalyzed one-pot asymmetric Mannich reaction. Compared with
current chemical technologies, this enzymatic reaction is more environmental-friendly and
sustainable by using a biocatalyst from inexpensive regenerable resources. A wide range of
substrates could be accepted by AMA. The activity and stereoselectivity of AMA could be
improved by adjusting solvent, pH, water content, temperature, molar ratio of substrates and
enzyme loading. The enantioselectivities up to 89% ee, diastereoselectivities up to 90:10 dr and
yields up to 82% were achieved. This work not only offered a novel case of enzyme catalytic
promiscuity and a potential synthetic method for organic chemistry but also might help to better
14
Page 14 of 19

understand metabolic pathways for the biosynthesis of nitrogen-containing compounds. It
provided an insight into a new area of enzyme catalysis and a building block for further research.
Supporting Information
1
Characterization data of Mannich products, H NMR and 13C NMR spectra, and chiral HPLC
trace.
Acknowledgements
This work was financially supported by the National Natural Science Foundation of China (No.
21276211).
References
[1] O. Khersonsky, D. S. Tawfik, Annu. Rev. Biochem. 79 (2010) 471–505.
[2] A. Fernandez-Gacio, A. Codina, J. Fastrez, O. Riant, P. Soumillion, ChemBioChem, 7 (2006)
1013–1016;
[3] P. J. O’Brien, D. Herschlag, Chem. Biol. 6 (1999) R91–R105;
[4] R. J. Kazlauskas, Curr. Opin. Chem. Biol. 9 (2005) 195–201.
[5] O. P Ward, A. Singh, Curr. Opin. Biotechnol. 11 (2000) 520–526;
[6] A. Babtie, N. Tokuriki, F. Hollfelder, Curr. Opin. Chem. Biol. 14 (2010) 200–207;
[7] M. S. Humble, P. Berglund, Eur. J. Org. Chem. (2011) 3391–3401;
[8] E. Busto, V. Gotor-Fernández, V. Gotor, Chem. Soc. Rev. 39 (2010) 4504–4523;
[9] Q. Wu, B. K. Liu, X. F. Lin, Curr. Org. Chem. 14 (2010) 1966–1988.
15
Page 15 of 19

[10] C. Li, X.-W. Feng, N. Wang, Y.-J. Zhou, X.-Q. Yu, Green Chem. 10 (2008) 616–618;
[11] H.-H. Li, Y.-H. He, Y. Yuan, Z. Guan, Green Chem. 13 (2011) 185–189.
[12] J.-F. Cai, Z.Guan, Y.-H. He, J. Mol. Catal. B: Enzym. 68 (2011) 240–244.
[13] T. Purkarthofer, K. Gruber, M. Gruber-Khadjawi, K. Waich, W. Skranc, D. Mink, H. Griengl,
Angew. Chem, Int. Ed. 45 (2006) 3454–3456.
[14] M. Arend, B. Westermann, N. Risch, Angew. Chem. Int. Ed. 37 (1998) 1044–1070.
[15] X.-H. Cai, H. Guo, B. Xie, Eur. J. Chem. 3 (2012) 258–266.
[16] K. Li, T. He, C. Li, X.-W. Feng, N. Wang, X.-Q. Yu, Green Chem. 11 (2009) 777–779;
[17] T. He, K. Li, M.-Y. Wu, X.-W. Feng, N. Wang, H.-Y. Wang, C. Li, X.-Q. Yu, J. Mol. Catal. B:
Enzym. 67 (2010) 189–194;
[18] S.-J. Chai, Y.-F. Lai, H. Zheng, P.-F. Zhang, Helv. Chim. Acta 93 (2010) 2231–2236;
[19] M. López-Iglesias, E. Busto, V. Gotor, V. Gotor-Fernández, Adv. Synth. Catal. 353 (2011)
2345–2353.
[20] Y. Xue, L.-P. Li, Y.-H. He, Z. Guan, Sci. Rep. 2 (2012) 761. DOI: 10.1038/srep00761.
[21] Y. B. Tewari, Appl. Biochem. Biotechnol. 23 (1990) 187–203.
[22] C. R. Wescott, A. M. Klibanov, Biochim. Biophys. Acta 1206 (1994) 1–9.
[23] H. A. Lindner, V. V. Lunin, A. Alary, R. Hecker, M. Cygler, R. Ménard, J. Biol. Chem. 278
(2003) 44496–44504;
[24] H. K. Chenault, J. Dahmer, G. M. Whitesides, J. Am. Chem. Soc. 111 (1989) 6354–6364;
[25] M. W. Anders, W. Dekant, Adv. Pharm. 59 (1994) 431–448.
[26] R. B. Silverman, The Organic Chemistry of Enzyme-Catalyzed Reactions, Academic Press,
San Diego, 2000, pp. 55–60.
16
Page 16 of 19

[27] A. Zaks, A. M. Klibanov, J. Biol. Chem. 263 (1988) 8017–8021.
17
Page 17 of 19

Graphical abstract
Enzyme-catalyzed asymmetric Mannich reaction using acylase from Aspergillus melleus
Zhi Guan, Jian Song, Yang Xue, Da-Cheng Yang, and Yan-Hong He
NH2
R2
O
X
O
O
X
HN
AMA
+
+
H
R1
R1
o
MeCN / buffer, 30 C
R2
17 examples
yields up to 82%
dr up to 90:10
ee up to 89%
AMA = Acylase I from Aspergillus melleus
18
Page 18 of 19

Highlights
1. Acylase I from Aspergillus melleus (AMA) displayed catalytic promiscuity towards one-pot
asymmetric Mannich reaction for the first time.
2. AMA showed favorable catalytic activity with good adaptability to different substrates.
3. The activity and stereoselectivity of the enzyme could be improved by adjusting solvent, pH
and water content. The enantioselectivities up to 89% ee, diastereoselectivities up to 90:10
dr (syn/anti) and yields up to 82% were achieved.
4. This work offered a novel case of enzyme catalytic promiscuity and a potential synthetic
method for organic chemistry.
19
Page 19 of 19