5
52
Can. J. Chem. Vol. 80, 2002
dissolved in potassium phosphate buffer (1 mL, 5 mM,
pH 7), divided into 0.3-mg portions, and added to PEG
Fig. 1. Circular dichroism spectra, in water and 1,4-dioxane, of
CALB colyophilized with PEG. PEG–protein ratio of 250 for the
spectrum recorded in water (g) and 2.5 (a); 5 (b); 25 (c); 50 (d);
125 (e); and 250 (f) for those recorded in 1,4-dioxane.
(
2.5–125 mg) dissolved in water (0.4 mL) (PEG–protein
ratio (w/w) 8–417); final volume 0.5 mL. The samples were
then frozen and lyophilized. Each sample was prepared in
double to allow for the measurement of activity and to re-
cord the circular dichroism spectrum. Control samples were
analogously prepared, but without PEG. The protein content
was determined by Biorad Protein Assay.
2
00
210
220
230
240
250
0
a
b
-
-
-
-
2
4
6
8
Circular dichroism spectra
Far-UV circular dichroism (CD) spectra were recorded
with a Jasco 500 A spectropolarimeter over the range of
c
2
2
2
00–250 nm for the sample dissolved in water and
10–250 nm for the samples dissolved in dioxane. (Below
10 nm, the high absorbance of dioxane causes a low signal-
d
to-noise ratio that prevents the measurements.) All spectra
were recorded at 25°C. The optical path was 0.5 cm in the
case of lipases and 0.1 cm in the case of subtilisin. The
PEG–enzyme samples were added to 1 mL of 1,4-dioxane
e
g
f
(
or water), vigorously shaken, and immediately used for CD
-
10
analysis. All spectra were baseline corrected.
Measurements of catalytic activity
-12
nm
The transesterification between 1-octanol and vinyl ace-
tate to give 1-octyl acetate and acetaldehyde was used as
model reaction for the three enzymes considered. In all
cases, 1 mL of 1,4-dioxane containing 1-octanol (0.19 M)
and vinyl acetate (1.1 M) was added to the different enzyme
samples. All the reagents, enzyme samples, and 1,4-dioxane
ally caused by sizeable aggregates or very fine protein dis-
persions (17). Moreover, it is worth noting that the enzyme
only showed a CD signal in dioxane when it was previously
lyophilized from an aqueous solution containing PEG. When
the enzyme was suspended in dioxane, it did not give a CD
signal in the 210–350 nm range, even when PEG was subse-
were equilibrated against molecular sieves (a < 0.1) before
w
being used. The reaction mixtures were shaken at 200 rpm
and 25°C. At scheduled times, aliquots were withdrawn
from the reaction mixtures and the conversion determined by
GLC. (HP-1 Crosslinked Methyl Silicone Gum, 25 m,
–
1
quently added (up to 5 mg mL ) to the protein suspension.
This result, which is in agreement with previous observa-
tions on other proteins (14), demonstrates that CALB is sol-
uble in 1,4-dioxane and its secondary structure is the same
as in water. On the other hand, the spectra obtained for
CALB at PEG–protein ratios lower than 125 showed a pro-
portionally lower signal, which can be attributed to a de-
crease in the solubility of the protein (14). Analogous
behavior was observed for lipase PC and subtilisin (spectra
not shown).
0
.32 mm ID, Hewlett-Packard; conditions: oven temperature
from 35°C (initial time 10 min) to 160°C with a heating rate
of 15°C min , H as carrier gas.) The retention times for vi-
–
1
2
nyl acetate, acetic acid, 1-octanol, and 1-octyl acetate were
1
.8, 2.2, 11.5, and 13.0 min, respectively. Using these reac-
tion conditions, the catalytic activity for CALB, lipase PC,
and subtilisin was, respectively, 2.20, 0.16, and
–
1
0
.028 µmol min for the samples without PEG, whereas it
Figure 2 compares the increment of activity and solubility
of the three enzymes as a function of the PEG–protein ratio.
From zero to the highest ratio tested, not only did the activ-
ity increase 3.1-, 4.3-, and 3.5-fold for CALB (a); lipase PC
(b); and subtilisin (c), respectively, but the enzymes were
also fully solubilized. Nevertheless, it is evident that for all
the enzymes the initial increment of activity was much
steeper than that of solubility. Table 1 reports the increments
of activity and solubility for CALB, lipase PC, and subtilisin
at low PEG–protein ratios. It can be seen that the increment
of activity was for all the enzymes much higher than the in-
crement of solubility. For example, passing from 0 to a
PEG–protein ratio of 2.5 for the lipases or 8 for protease, the
increment of activity was 4.4-, 7.1-, and 4.1-fold higher than
that of solubility for CALB, lipase PC, and subtilisin, re-
spectively.
–
1
was 6.75, 0.68, and 0.098 µmol min for the samples at the
highest PEG–protein ratio tested (250 for lipases and 417 for
subtilisin).
The possibility that PEG could affect the enzymatic activ-
ity by competing as a nucleophile with octanol seems very
unlikely because, as shown in Fig. 2, the activity increased
as a function of PEG concentration or, when the plateau
value was reached, remained constant.
Re s ults a nd dis c us s ion
Figure 1 shows that CALB, at a PEG–protein ratio higher
than 125 (w/w), is fully solubilized in 1,4-dioxane, as shown
by the almost complete overlap between the spectra recorded
in the organic solvent and in water. In addition, the similar-
ity between the spectra outside the range typical for a pro-
tein, in our case from 350 to 250 nm (data not shown),
should rule out the presence of scattering phenomena usu-
These data prove that the activating effect of PEG is not
directly correlated to enzyme solubility. As a consequence,
the activity increase should be due to a lyoprotective effect
©
2002 NRC Canada
Secundo, Francesco
