Table 2 Effect of PMSF titration on subtilisin Carlsberg activity in
hexane
propyl ester shows that methanol cannot just be a better
nucleophile since then only the rate of methyl ester production
would be affected.¶ Rather the results suggest that solvation of
the protein by methanol stimulates an increase in the intrinsic
catalytic activity of both subtilisin Carlsberg and
a-chymotrypsin under anhydrous conditions.∑ Since these
serine proteases have very different secondary and tertiary
structures similar effects might be expected with other related
enzymes.
Initial ratea/
nmol min21 mg21
Alcohol
No PMSFb
PMSFb
Inhibition (%)
None
Propanol
Methanol
22.5
11.5
42
21.3
10.5
4.5
5.4
13
89
To eliminate the possibility that trace impurities (e.g. drying
agent) or changes in particle size were responsible for the
activation, a lyophilised powder sample of subtilisin Carlsberg
was equilibrated with methanol through the vapour phase in the
presence of molecular sieves. Gravimetric measurements
showed that this procedure resulted in adsorption of low levels
of the solvent onto the protein. Following resuspension in
hexane, enzyme activity was assayed using APEE and 0.5 m
propanol. Again a transesterification rate enhancement was
observed compared to an untreated sample and with no lag
period.** However, here the main reaction product was found
to be the propyl ester. This suggests that the low levels of
methanol adsorbed onto the enzyme were sufficient to activate
it but the resultant concentration of methanol in the reaction
mixture was not high enough for it to compete as a nucleophile.
Importantly equilibration of subtilisin with ethanol or propanol
vapour under identical experimental conditions gave absolutely
no rate enhancement nor did the addition of ethanol to the
reaction mixture. Thus, the effect of methanol is very specific
and the activation mechanism cannot be simply ascribed to a
hydrogen bonding phenomena.
Further study of the rates in Table 1 reveal an interesting
effect when methanol and water were added to the enzyme
together. With dry subtilisin and 0.5 m methanol the trans-
esterification rate is 24.5 nmol min21 mg21 while with enzyme
equilibrated to aw = 0.43 and using 0.5 m propanol the rate is
1.6 nmol min21 mg21. If we now compare the rate obtained at
aw = 0.43 and 0.5 m methanol, 102.2 nmol min21 mg21, we see
that it is closer to the product than the sum of the individual
rates. This observation is not consistent with a single activation
mechanism since then the contributions would be expected to be
approximately additive. Instead it points to a cooperative effect
whereby methanol and water activate the enzyme via two or
more different processes. As a consequence the ubiquitous
‘molecular lubrication model’ will need to be revised and we are
currently working on this problem.
a
Reaction conditions: 10 mm APEE, 0.5 m methanol and 1 mg ml21
enzyme in hexane. b The enzyme was shaken with or without 10 mm PMSF
and 0.5 m alcohol in hexane for 3 h, washed with hexane to remove unbound
PMSF and dried over molecular sieves prior to rate analysis.
methanol activates anhydrous serine proteases via a solvation
process.
Footnotes
* E-mail: b.d.moore@strath.ac.uk
† The variation in amount of enzyme bound water as a function of water
activity in nonpolar solvents has been measured previously (ref. 8).
‡ Residual water levels in the dried solvents were measured via Karl–
Fischer titration and found to be hexane < 0.005%, methanol 0.009%,
propanol 0.01%.
§ The nucleophilic specificity of methanol relative to propanol has been
shown to be 4.61 for this subtilisin catalysed reaction in hexane at
aw = 0.59.9
¶ The argument that the methyl ester product may be more reactive is not
valid as the acyl intermediate will be identical whichever ester reacts with
the enzyme and hence back reaction of the methyl ester cannot lead to an
overall increase in rate.
∑
2H NMR difference measurements in hexane showed 50 deuterated
methanol molecules bound per subtilisin molecule at a methanol concentra-
tion of 3 mm.
** Transesterification rates were; untreated enzyme, 1 nmol min21 mg21
,
enzyme equilibrated with propanol,
equilibrated with methanol, 6.8 nmol min21 mg21
1 , enzyme
nmol min21 mg21
.
†† PMSF is inert towards reaction with dry alcohols over the period
employed (ref. 10).
‡‡ The exact conditions used for the assay will not affect the findings since
it is the remaining activity following titration which is being measured. The
assay used was convenient because of the known high activity of dry
subtilisin in the presence of methanol.
§§ The increased activity of the control sample in the presence of methanol
is reproducible and under study.
Finally we sought to show that methanol could activate a
serine protease without simultaneously acting as a substrate.
Hence, a series of active site titrations were carried out on
subtilisin in hexane using the irreversible inhibitor PMSF.
Anhydrous enzyme was equilibrated for 3 h in hexane in the
presence or absence of either methanol or propanol and with and
without PMSF.†† Since in the inhibition process, no deacyla-
tion occurs, methanol will not be directly involved in the
reaction scheme. Following equilibration the enzyme samples
were repeatedly resuspended in pure dry hexane (which
removed the PMSF if present) and then dried. The resultant
enzymic activities were measured by monitoring transester-
ification of APEE with methanol in hexane‡‡ and are shown in
Table 2. It can be seen that enzyme equilibrated with PMSF in
either hexane or hexane–propanol gave very similar activities to
the control samples showing only a small percentage of active
sites had been titrated. This is consistent with the low activity of
the enzyme under these condition (see Table 1). In comparison
the presence of methanol during the PMSF equilibration
resulted in significant inhibition of the enzyme preparation.§§
This is exactly what would be expected from the transesterifica-
tion measurements and provides conclusive evidence that
References
1 Enzymatic Reactions in Organic Media, ed. A. P. M. Koskinen and
A. M. Klibanov, Blackie Academic and Professional, Glasgow, 1996
and references cited therein.
2 A. Zaks and A. M. Klibanov, J. Biol. Chem., 1988, 263, 8017.
3 M. Reslow, P. Adlercreutz and B. Mattiasson, Biocatalysis, 1992, 6,
307.
¨
4 O. Almarsson and A. M. Klibanov, Biotechnol. Bioeng., 1996, 49, 87.
5 P. J. Halling, Biochim. Biophys. Acta, 1990, 1040, 225.
6 M. Dolman, P. J. Halling, B. D. Moore and S. Waldron, Biopolymers,
1997, 41, 313–321.
7 M. C. Parker, B. D. Moore and A. J. Blacker, Biocatalysis, 1994, 10,
269.
8 M. C. Parker, B. D. Moore and A. J. Blacker, Biotechnol. Bioeng., 1995,
46, 452.
9 A. K. Chaudhary, S. V. Kamat, E. J. Beckman, D. Nurok, R. M. Kleyle,
P. Hajdu and A. J. Russell, J. Am. Chem. Soc., 1996, 118, 12 891.
10 G. T. Jame, Anal. Biochem., 1978, 86, 574.
Received in Liverpool, UK, 18th February 1997; Com.
7/01162H
932
Chem. Commun., 1997