2
J. Mukherjee, M. N. Gupta / Tetrahedron Letters xxx (2015) xxx–xxx
alcoholysis by subtilisin have been estimated many years back.37
Dordick’s group has used nucleophiles sucrose and thymidine for
bioimprinting subtilisin successfully.27,38 We have tried bioim-
printing with short chain alcohols by simply using these to precip-
itate subtilisin. Table 1 shows the initial rates of transesterification
Table 2
Transesterification of N-acetyl-L-phenylalanine ethyl ester with n-butanol in anhy-
drous n-hexane using the EPROS of subtilisin Carlsberg (SC)
Precipitating
agent
Initial rates of transesterification with n-butanol
(nmol minÀ1 mgÀ1
)
n-Propanol
n-Butanol
tert-Butanol
Ethanol
Methanol
Acetone
61
75
45
36
32
28
12
of N-acetyl-L-phenylalanine ethyl ester with n-propanol catalysed
by subtilisin bioimprinted by n-propanol, n-butanol, tert-butanol,
ethanol, methanol, acetone and acetonitrile. As different organic
solvents precipitate enzymes to different extents,39 the initial rates
by definition are normalized with respect to the amount of subtil-
isin precipitated. Both n-propanol and n-butanol were found to be
equally good bioimprinting agents. This is understandable as n-
propanol and n-butanol are structurally not very different. On
the other hand shorter chain alcohols ethanol and methanol were
less successful in bioimprinting for the reaction. The much lower
rates observed when the enzyme was precipitated in acetone and
acetonitrile indicated that the results with alcohols were not
artifacts.
Table 2 shows the initial rates of transesterification of the
amino acid ester with n-butanol with a similar set of bioimprinted
EPROS. Significantly, n-butanol was the best bioimprinting agent.
These initial rates were about half of what were observed with
the first transesterification reaction. Hence it may be concluded
that subtilisin seems to prefer n-propanol as a nucleophile as com-
pared to n-butanol. So, bioimprinting had a greater impact and
bioimprinting with either n-butanol or n-propanol made a differ-
ence during transesterification with n-butanol.
The particle size of the precipitate is likely to play an important
role because of the mass transfer constraints.40 Light scattering
measurements of the subtilisin precipitated/bioimprinted with
various organic solvents (Fig. 1) showed that all the precipitates
had particle size in a narrow range of 350–480 nm diameters.
The precipitates obtained with propanol/butanol consisted parti-
cles of higher sizes than those with methanol/ethanol. This clearly
established that the effect on initial rates (Tables 1 and 2) were due
to bioimprinting and not mass transfer constraints.
The use of CD (with a spinning cell accessory) has been
described recently to look at the closeness of the structure of
enzymes in insoluble form to the native structure.32 It was shown
that the closeness to the native structure correlated well with the
activities observed with various enzyme preparations in organic
media.32 Figure 2 shows the far UV CD spectra of the various sub-
tilisin EPROS. The corrected CD spectra show that the order of the
molar ellipticity was native subtilisin in buffer > n-propanol bioim-
printed subtilisin > n-butanol bioimprinted subtilisin > methanol
bioimprinted subtilisin P tert-butanol bioimprinted subtil-
isin P ethanol bioimprinted subtilisin > acetonitrile precipitates
of subtilisin. The CD spectra of the EPROS obtained with acetone
could not be recorded as acetone absorbs strongly in the range of
190–280 nm. The EPROS are reported to show a great shift in the
Acetonitrile
SC (1 mg) was dissolved in 50 mM Tris–HCl buffer, pH 7.8. EPROS was prepared and
the transesterification reaction was set up as mentioned in the experimental sec-
tion. The experiment was done in duplicate and the error between each set was
within 3%.
secondary structure as compared to that of native subtilisin in
aqueous buffers.32 However, among the EPROS prepared with the
various organic solvents, the ones with n-propanol and n-butanol
change signs of the CD spectra closest to the native spectra
(Fig. 2) as compared to the other spectra which change signs much
farther away. This indicates that the secondary structure of the
EPROS with n-propanol and n-butanol are better as compared to
the other EPROS. This is also in line with the activity data (Table 1).
The next obvious step was to try simultaneous/dual bioimprint-
ing with a substrate analog, for example, an amide (used by Russell
and Klibanov25) and the alcohols. Table 3 shows the effect of
bioimprinting in various cases on the initial rates of transesterifica-
tion of N-acetyl-L-phenylalanine ethyl ester with n-propanol. On
the basis of the results presented in Tables 1 and 2, we can assume
that precipitation of subtilisin with acetone did not lead to any
bioimprinting. So, just the enzyme precipitated with acetone
resulted in much higher rates than those reported for bioimprinted
subtilisin when the bioimprinting with ester analog was carried
out by freeze drying.25 Bioimprinting with N-acetyl-
L-tyrosinamide
gave a further 1.4Â enhancement in the initial rates. With both
ethanol and propanol, dual bioimprinting resulted in a more effi-
cient biocatalyst preparation. Again, n-propanol, the substrate in
the chosen transesterification reaction, was the best bioimprinting
molecule along with N-acetyl-L-tyrosinamide. It was interesting to
check whether the better initial rates also translated into the better
conversions. Figure 3 shows the use of biocatalyst preparations
with dual bioimprinting with the right alcohol and the amide
resulted in the highest conversion. The unimprinted subtilisin
(precipitated with acetone) was the least efficient biocatalyst.
While the present work was in progress, we came across a
report by Yan et al.,41 in which a crude preparation of lipase was
bioimprinted by drying its mixture with oleic acid and methanol
in aqueous organic co-solvent by freeze drying. In order to improve
the catalytic efficiency, these authors tried to synergize the
approach with immobilization, lecithin coating and salt activa-
tion.42 The best increase in catalytic efficiency reported was
19-fold over the unimprinted enzyme. We show that dual bioim-
printing was successful even in the case of a protease, subtilisin.
Table 1
Transesterification of N-acetyl-L-phenylalanine ethyl ester with n-propanol in anhy-
drous n-hexane using the EPROS of subtilisin Carlsberg (SC)
Precipitating
agent
Initial rates of transesterification with n-propanol
(nmol minÀ1 mgÀ1
)
Conclusions
n-Propanol
n-Butanol
tert-Butanol
Ethanol
Methanol
Acetone
129
130
100
96
95
72
The removal of bulk water (‘drying’) results in rigidification of
the conformation of the protein molecule. Use of this dried prepa-
ration in nearly anhydrous organic solvents ensures that this rigid
conformation is preserved.15,42 Hence, any induced conformation
by a ligand in aqueous buffer is preserved during ‘drying’. This is
the basis of bioimprinting proteins.25 The results described in the
current work exploit the earlier observations that precipitation
with organic solvents can be used for drying enzymes and in fact
generally gives more active enzyme preparations for use in organic
Acetonitrile
25
SC (1 mg) was dissolved in 50 mM Tris–HCl buffer, pH 7.8. EPROS was prepared and
the transesterification reaction was set up as mentioned in the experimental sec-
tion. The experiment was done in duplicate and the error between each set was
within 3%.