Hofmeister effects on activity and stability of
alkaline phosphatase
Zhen Yang‡*, Xiu-Ju Liu‡, Chao Chen‡, Peter J. Halling§*
‡
College of life sciences, Shenzhen University, Shenzhen, Guangdong, China 518060
§
WestCHEM, Department of Pure & Applied Chemistry, University of Strathclyde,
Glasgow G1 1XL, United Kingdom
*Corresponding authors:
Zhen Yang, Ph.D., Professor
Address: College of life sciences, Shenzhen University, Shenzhen 518060,
China 518060
Tel.: +86 755 2653 4152
Fax: +86 755 2653 4277
E-mail: zyang@szu.edu.cn
Peter J. Halling, Ph.D., Professor
Address: WestCHEM, Department of Pure & Applied Chemistry, University of
Strathclyde, Glasgow G1 1XL, United Kingdom
Tel.: +44 (0)141 548 2683
Fax: +44 (0)141 548 4822
E-mail: p.j.halling@strath.ac.uk
Preprint of article published in Biochimica Et Biophysica Acta 1804: 821-828.
10.1016/j.bbapap.2009.12.005:
Running title: Hofmeister effects on alkaline phosphatase
Keywords:
Alkaline phosphatase, activity and stability, Hofmeister series, kosmotropicity,
chaotropicity, Jones-Dole viscosity B coefficient
1
Summary
We have studied the effects on alkaline phosphatase of adding high
concentrations (normally 1.0 M) of simple salts. It is necessary to allow for significant
effects of salts on the extinction coefficient of the reaction product, and on the
apparent pH of the buffer. Both activity and stability of the enzyme correlate well with
the Hofmeister series in terms of the salt’s kosmotropic/chaotropic properties, which
are assessed by the Jones-Dole viscosity B coefficients (B+ for cations and B for
anions). The catalytic activity or Vmax/Km of the enzyme showed a bell-shaped
relationship with the (BB+) values of the salts present, being optimal with salts
(such as NaCl, KCl, and KNO3) where the anion and cation have similar
kosmotropic/chaotropic properties. This effect is believed to be enzyme-specific and
relates to the impact of both cations and anions on the enzyme’s surface pH, active
site, and catalytic mechanism. Anions play a more predominant role than cations in
affecting enzyme stability. The rate of irreversible thermal inactivation is strongly
reduced by addition of kosmotropic anions like SO42- (half life increased from 8 to
580 min at 60oC). This effect is general and the mechanism probably involves the
ability of the ions to affect the water solvation layer around the enzyme molecule and
to interact with both the surface and internal structure of the enzyme.
2
Introduction
The specific effect of inorganic salts and ions on proteins was realized as early as
more than a century ago, when a series of cations and anions was proposed in order of
their ability to precipitate hen egg white, by Hofmeister [1]. Numerous experiments
later have revealed that the influence of ions on protein stability follows the
Hofmeister series [2,3]. There have also been examples showing that enzymatic
activity seemed to follow the Hofmeister series as well, with the earliest one being
recognized more than 40 years ago [4]. Although the Hofmeister effect has been
shown to have a universal utility not only in biochemistry, but also in areas such as
physical, colloid, polymer, and surface chemistry, the mechanism involved is still not
clearly known [5].
There has been an intensive effort devoted to explaining Hofmeister effects at the
molecular level, both through theoretical and molecular dynamics simulations and
experimental approaches [6-10]. Chemically speaking, an ion may affect the enzyme
performance by playing the role of a substrate, a cofactor, or an inhibitor of the
enzyme. But the general ranking of the Hofmeister series may be better understood by
considering the ion’s ability to alter the bulk water structure, to affect the
protein-water interaction, and to interact directly with the enzyme molecules [3]. At
low salt concentrations (up to ~0.01 M), ions affect enzyme performance
predominantly via electrostatic interactions. The Hofmeister ion effects become
important when the electrostatic forces are screened by higher salt concentrations and
the ionic dispersion forces turn out to dominate. There have been a large number of
papers dealing with Hofmeister effects on proteins and enzymes, discussed in more
detail in recent reviews [2,3,5,6,8,9]. The Hofmeister series has been associated with
the kosmotropic/chaotropic properties of the ions (Figure 1), which are characterized
by the Jones-Dole viscosity B coefficients [11]. These are ion-specific coefficients in a
fit of viscosity to the concentration of salt solutions, with the sign giving the direction
of viscosity change. With one of several plausible assumptions, the B coefficients may
then be split into contributions from individual ions. Characterizing the ion-solvent
interactions, the viscosity B coefficient is interpreted as a measure of the
3
structure-making and structure-breaking capacity of an ion in solution, thus directly
correlating to the ion’s kosmotropic/chaotropic properties. The viscosity B coefficient
is normally positive for a kosmotrope and negative for a chaotrope. While
Hofmeister’s original studies [1] dealt mainly with anion effects, a number of more
recent studies have argued that the order of cation effects on proteins is reversed, with
kosmotropic cations actually being destabilizing or deactivating [12-14]. Proteins are
usually found to be stabilized by a kosmotropic anion and a chaotropic cation and
destabilized by a chaotropic anion and a kosmotropic cation [2], and this phenomenon
can be well illustrated by a combination of ionic hydration and interaction with
surface groups on the protein [3]. However, the specific ion effect on enzyme activity
does not seem to be so straightforward, and some disobeying examples have already
been reported [15-17].
Stabilizing proteins
Destabilizing proteins
Anions: SO42-  HPO42-  CH3COO-  F-  Cl-  Br-  NO3-  I-  ClO4-  SCN-
Kosmotropes
Chaotropes
Cations: Mg2+  Li+  Na+  K+  NH4+  (CH3)4N+
Destabilizing proteins
Stabilizing proteins
Figure 1. Ranking of anions and cations according to their kosmotropic/chaotropic
properties
There is also a growing interest in the preparative use of enzymes in concentrated
salt solutions, where Hofmeister effects are likely to be important. These can be found
when enzymes are used in mainly solid systems where salts are used as reactants, and
perhaps also neutralizing agents [18]. Interestingly, a number of normal enzymes are
known to be activated in concentrated salt solutions [19,20]. There is also much recent
4
work on enzyme action in ionic liquids (reviews, including [3,21,22]). This often
includes studies on what are commonly called mixtures of ionic liquids and water. But
except at low water contents, these are essentially equivalent to concentrated salt
solutions in water. There is no obvious reason why the physical state of the pure salt
(solid or liquid) should affect the nature of these mixtures with water.
Alkaline phosphatase (E.C. 3.1.3.1) is a metal (Zn, Mg)-containing hydrolytic
enzyme catalyzing the hydrolysis and transphosphorylation of a wide variety of
phosphate monoesters, and its catalytic mechanism and molecular structure have been
detailed [23]. In this study, alkaline phosphatase from calf intestine was used to
catalyze the hydrolysis of p-nitrophenyl phosphate (pNPP) to p-nitrophenol (pNP).
The isoelectric point for this enzyme is 5.7 (Worthington Enzyme Manual, 1993
Edition, see http://www.worthington-biochem.com/index/manual.html). Both the
activity and stability of the enzyme was investigated by addition of different salts of
alkali and alkaline earth metals in aqueous solution in order to provide valuable
information regarding the mechanism involved in the specific ion effect on the
catalytic performance of enzymes.
Materials and Methods
Materials. Alkaline phosphatase from calf intestine (AP, 25 U/l) and its
substrate, p-nitrophenyl phosphate disodium hexahydrate (pNPP) were purchased
from Takara Biotechnology (Dalian) Co., Ltd. (China) and Amresco Inc. (USA),
respectively. NaBF4 and KPF6 were obtained from Sigma-Aldrich China Inc. All other
reagents used were of analytical grade from local manufacturers in China.
Preparation of the enzyme, the buffer, and buffer/salt solutions. The buffer
normally used for enzyme assays contains 1.0 M diethanolamine (DEA), 1.0 mM
MgCl2, and 1.0 M ZnCl2, being adjusted to pH 9.8 with HCl. The buffer/salt
solutions were prepared by dissolving different inorganic salts into the DEA buffer at
initial pH 9.8, with the pH being readjusted to 9.8 by addition of either HCl or NaOH.
The concentrated enzyme solution was diluted (usually 105-fold) before use, in the
diluent buffer consisting of 10 mM tris(hydroxymethyl)aminomethane (Tris), 1.0 mM
MgCl2, and 1.0 M ZnCl2, being adjusted to pH 8.0 with HCl. Portions of this diluted
5
enzyme solution were then added to the DEA buffer for activity assays. The pH
electrode was calibrated with normal dilute aqueous standards, so readings should be
considered apparent pH.
Determination of molar extinction coefficients for p-nitrophenol (pNP). A series
of pNP solutions (0-0.01 mM) were prepared in 1.0 M DEA buffer containing
different inorganic salts (up to 1.0 M) at different pH’s (pH 7-12), and their
absorbances at 405 nm were measured. The molar extinction coefficients of pNP were
then determined as slopes derived from the A405 vs. [pNP] plots.
Enzyme activity assays. Enzyme activity was determined by following the
formation of pNP spectrophotometrically at 37oC. After addition of 0.1 ml enzyme
solution (0.25 U/ml) to a cuvette containing 2.025 ml DEA buffer (1.0 M, pH 9.8) in
the presence of different inorganic salts (normally 1.0 M unless otherwise stated, pH
readjusted to 9.8) and 0.375 ml 100 mM pNPP in the DEA buffer, the solution was
immediately mixed and the increase in absorbance at 405 nm (indicating the
formation of pNP) was recorded with the Pharmacia Biotech Ultraspec 2000 UV/Vis
spectrophotometer equipped with a thermostated cell. One unit of alkaline
phosphatase was defined as the amount of the enzyme catalyzing the hydrolysis of 1
mol of pNPP to pNP per minute at 37oC. The Km and Vmax values of the enzyme were
obtained from Lineweaver-Burk plots.
Enzyme stability tests. The enzyme solution (0.5 U/ml in the diluent buffer with
addition of 1 M of different inorganic salts, final pH 8.0) was incubated in a water
bath controlled at different temperatures (50oC, 60oC, 70oC). Periodically 50 l of
each enzyme solution was taken for activity assay at 37oC as noted above. The
time-dependent loss in activity was used to calculate the half-life for each enzyme
solution. It is worth mentioning that for the fastest inactivation rate measurements
(e.g., enzyme in salt-free DEA buffer at 70oC), only a few points were measurable, so
the half-life estimate will be less accurate. The time taken for the incubation to heat to
70oC will also contribute to errors. The slower inactivation rates in the presence of
salts were more accurately measurable. Activity measurements throughout this study
were at least duplicated with an error of less than 5%.
6
Results and Discussion
1. Effect of salts on the molar extinction coefficient for pNP
The activity of alkaline phosphatase was determined by spectrophotometrically
following the formation of the product pNP based on its strong absorbance at 405 nm.
Determination of the enzyme activity depends on the accuracy of the molar extinction
coefficient of pNP. The variation of this molar extinction coefficient has to be
considered, as our experiments have revealed that the extinction coefficient is very
different depending on the type, pH, and concentration of the buffer and the salt
involved. For instance, the molar extinction coefficient of pNP in the DEA buffer was
raised from 15.0 to 18.6 mM-1cm-1 when Na2SO4 was gradually added to the buffer up
to a concentration of 1.0 M. Some other examples are shown in Figure 2.
22
Na2 SO4
 NaAc
r NaBF4
20
-1
(mM cm )
 KPF6
NaCl
 KNO3
 Control
18
-1
Molar extinction coefficient of pNP
24
16
14
12
7
8
9
10
11
12
pH
Figure 2. Variation of the molar extinction coefficient of pNP upon addition of 1.0 M
salts to the 1.0 M DEA buffer at different apparent pH’s
7
2. Effect of salts on buffer pH
Introduction of inorganic salts into the DEA buffer (1.0 M, initial pH 9.8) may
significantly alter the pH of the buffer (Figure 3). For instance, the pH was increased
(e.g., to pH 10.26 in the presence of 1 M Na2SO4) or decreased (e.g., to pH 9.75 in the
presence of 1 M NaNO3) with an increasing salt concentration. The pH shift caused
by salt addition has been disregarded in some of the studies involving enzymes in the
presence of salts, but has to be seriously considered, as this can trigger a change in the
charge distribution of the enzyme molecule and in turn a change in its catalytic
performance [3]. Therefore, in order to avoid the interference of buffer pH change
caused by addition of salts, for the later enzymatic reactions the salt-containing DEA
buffers were all readjusted to apparent pH 9.8 prior to use.
11.0
 Na2 HPO 4
r K oxalate
Na citrate
 Na2 SO4
NaAc
p KCl
 NaNO3
Buffer pH
10.5
10.0
 CaCl2
9.5
9.0
0.0
0.5
1.0
1.5
2.0
2.5
[Salt] (M)
Figure 3. Variation of the apparent pH of the DEA buffer (1.0 M, pH 9.8) upon
addition of salts
A few of the salts in Figure 3 may affect the buffer pH simply by acting as acids
or bases. Hence the anion in Na2HPO4 is sufficiently basic even at pH 9.8 to make the
solution more alkaline, while the cation in CaCl2 may acidify by a hydrolysis reaction
8
Ca2+ + H2O  Ca(OH)+ + H+. But most of the salts studied are either from strong
acids and bases, or contain anions that will have almost no tendency to act as bases at
pH 9.8.
Our pH data suggest that the buffer pH was higher in the presence of a salt
containing a more kosmotropic anion. Indeed, Salis et al. [24] have observed a much
higher pH than its original pH 7.0 in both 5 and 200 mM Tris buffer solutions when
Na2SO4 (final concentrations 0.5 and 1.0 M) was added, as compared to the pH values
measured in the presence of two other more chaotropic salts, NaCl and NaSCN.
Bauduin et al. [25] have also found that in a citrate buffer (25 mM, pH 5.0), the
measured pH was decreased by addition of four sodium salts; but comparatively, the
pH drop was less in the presence of Na2SO4, followed by NaCl, NaNO3, and NaBr,
which is again consistent with the Hofmeister series. A similar phenomenon was also
observed with phosphate buffer [26]: addition of four sodium salts (1.0 M) to the
buffer (5 mM, initial pH 7.0) decreased the pH in the order of NaCl (pH 6.35) > NaBr
(pH 6.33) > NaNO3 (pH 6.29) > NaClO4 (pH 6.06), with the anions matching the
decreasing order of the viscosity B coefficients.
2.1. Mechanism of effects on apparent pH
These results have therefore confirmed that the pH of an aqueous buffer solution
can be altered by addition of salts and that the effect of salts on the buffer pH is
ion-specific, following the Hofmeister series. When only electrostatic interactions are
taken into consideration, effects will depend solely on ionic strength, and not on the
identity of the ions. Bauduin et al. [25] have developed an extended Debye-Hückel
model to calculate the pH dependence of a citrate buffer on the total ionic strength,
and found that the calculated pH values were in accord with the data measured
experimentally. However, the ion-specific effect on the buffer pH cannot be explained
based only on simple electrostatic interactions.
There are in fact two possible explanations for the effects of salts on measured pH:
(1). The salts affect solvation (activity coefficients) of the buffer species, such that pH
is changed. This could be seen as a shift in pKa defined in terms of the concentrations
of buffer species. (2). The salts may directly affect the response of the pH electrode.
9
Salis et al. [24] considered that the salt effects were at least partly due to changes in
the “surface pH” around the glass electrode, rather than the bulk pH value. In contrast,
Bauduin et al. [25] argued that the pH electrode did report correctly, because the
values for the 1:1 electrolytes were in accord with those calculated by an extended
Debye-Hückel model for the dependence of pH of a citrate buffer on the total ionic
strength.
Distinguishing the above two effects is not as easy as might appear, involving
among other things the definition of pH in a solution that is no longer dilute aqueous.
(Since it is a function of the activity of a single ion, the definition of pH is much more
complicated than often recognized [27]). Moreover, both the electrostatic and
dispersion forces between salt and buffer ions might conspire together to offer an
explanation for this ion specificity.
3. Effect of salts on enzyme activity
The enzyme presented significantly different activities in the presence of different
inorganic salts. For most of the salts tested, their addition in the DEA buffer resulted
in a gradual reduction in the enzyme activity. Examination of the enzyme’s kinetic
parameters (Table 1) reveals that this activity reduction is more related to an increase
in Km and less to the change in Vmax. When no additional salts were added to the DEA
buffer, the Km and Vmax values for the enzyme were determined to be 0.69 mM and
20.4 M/min, respectively. When 0.8 M of different salts (those used in Table 1) were
present in the DEA buffer, the range of values found were 1.37-7.32 mM for Km and
11.7-29.4 M/min for Vmax, resulting in a significant decrease in the Vmax/Km ratio.
This is confirmed by measuring both these parameters in the presence of different
concentrations of Na2SO4: Increasing the salt concentration gradually to 1.0 M
induced a significant rise in the Km value to 25.1 mM and a mild increase in the Vmax
to 25.0 M/min. One way that Km can be influenced by the salts is via effects on
substrate solvation in the liquid phase – if the salts interact favorably with the
substrate, its tendency to bind with the enzyme will reduce and Km will rise. Of course
both Km and Vmax can also be affected by interactions of both cations and anions of the
salts with the enzyme molecule, especially its active site, as will be described later.
10
Table 1. Effect of salt addition on the kinetic parameters and optimal pHa of the
enzyme and the enzyme activity obtained at optimal pH
Salts
Bb
B+b
BB+
Km
Vmax
Optimal
Reaction rate
(mM)
(μM/min)
pH
obtained
at
optimal
pH
(μM/min)
Controlc
0.69
20.4
10.00
21.7  1.1
KPF6
-0.210
-0.009
-0.201
2.73
22.4
9.25
17.6  0.9
NaBF4
-0.093
0.085
-0.178
4.40
16.4
9.00
19.5  1.0
NaNO3
-0.043
0.085
-0.128
3.43
21.5
9.25
17.9  0.9
NaN3
-0.018
0.085
-0.103
1.37
11.7
NaCl
-0.005
0.085
-0.090
4.42
28.5
9.50
21.8  1.1
KNO3
-0.043
-0.009
-0.034
1.91
29.4
9.50
25.3  1.3
KCl
-0.005
-0.009
0.004
2.52
23.5
9.50
21.6  1.1
Na2SO4
0.206
0.085
0.121
6.87
22.4
9.00
14.7  0.7
NaAc
0.246
0.085
0.161
2.54
18.3
10.00
15.7  0.8
Na citrate
0.333
0.085
0.248
7.32
19.7
a
The activity of the enzyme was determined in the 1.0 M DEA buffer in the absence
and presence of different salts, with pH varying in the range between 8.0 and
11.0.
b
The viscosity B coefficients for all anions (B) and cations (B+) used in this table and
throughout the paper, except for citrate (whose B was obtained from [40]), were
obtained from [11].
c
1.0 M DEA buffer solution without addition of any extra salts.
11
Initial reaction rate (M/min)
30
KNO3
25
KCl
NaCl
20
NaNO3
Na2 SO4
15
NaAc
NaN3
10
Na citrate
CaCl2
5
0
-0.4
-0.2
0
0.2
0.4
B  B+
Figure 4. Relationship between the initial rate of the enzyme-catalyzed reaction and
the (BB+) value of the salt added (0.8 M in 1.0 M DEA buffer, pH 9.8)
The enzyme activity did not show an obvious correlation with either the viscosity
B coefficient of the anion of the salt (B) or that of the cation (B+). But as shown in
Figure 4, a bell-shaped relationship was obvious between the initial reaction rate and
the difference between B and B+, i.e., BB+, of the salt added, when the salt
concentration was 0.8 M in the reaction medium; the maximal activity was obtained
in the presence of KNO3. The dependence on ion properties was confirmed by
plotting Vmax/Km of the enzyme vs. (BB+) of the salt added (Figure 5), although no
obvious correlation was observed between the individual Km and Vmax values and the
kosmotropicity/chaotropicity of the salt. The Vmax/Km data also did not correlate well
with either B or B+ alone. These results suggest that (BB+) may be a good indicator
for quantifying the effect of inorganic salt and that alkaline phosphatase may exhibit
its optimal activity in the presence of a salt whose ions have similar
12
kosmotropic/chaotropic properties. Indeed, the enzyme activity was maximal with
salts such as NaCl, KNO3, and KCl which have a (BB+) value of -0.090, -0.034, and
+0.004, respectively. It should be noted that some of the salts tested are not 1:1 ion
combinations. Hence at the equal concentrations used they will give higher ionic
strengths. However there is no indication that they are outliers on the plots, consistent
with the view that specific ion effects, rather than simple electrostatics, dominate in
these concentrated solutions.
0.016
KNO3
-1
Vmax/Km (min )
0.012
KCl
NaN3
KPF6
0.008
NaAc
NaCl
NaNO3
0.004
NaBF4
Na2 SO4
Na citrate
0
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
B  B+
Figure 5. Correlation between Vmax/Km of the enzyme and the (BB+) value of the
salt added (1.0 M in 1.0 M DEA buffer, pH 9.8)
The pH dependence of the enzyme in the presence of different salts has also been
examined. Salt addition did trigger a significant change in the optimal pH of the
enzyme (Table 1). The correlation between the optimal pH of the enzyme and the
(BB+) value of the salt added seemed to be either linear (with Na2SO4 as an outlier)
or bell-shaped (with NaAc as an outlier). The real situation has to be further verified.
The activity of the enzyme obtained at its optimal pH showed an obvious bell-shaped
13
relationship with the (BB+) of the salt added (which can be observed from a plot
using the data given in Table 1), supporting the result obtained in the above Figures 4
and 5. However, the differences in activity were less than those obtained at constant
apparent pH 9.8, showing that alteration of the pH optimum explained part of this
effect.
3.1. Mechanism of effects on enzyme activity
Although the fact that enzymatic activity varied upon addition of salts has been
recognized for more than 40 years [4], the mechanism involved in this ion specificity
is still not clearly known. There are already some examples showing that the
salt-induced activity change does not follow the Hofmeister series [15-17]. Even for
those enzymes following the Hofmeister series, some are inhibited by all the salts
tested [4,13,16], while others get activated [15,17,28]. These complexities may be
attributed to the interactions of the ions with the enzyme molecules leading to a
change in the surface pH, net charge, active site, and catalytic mechanism of the
enzyme [3]. In the case of alkaline phosphatase, all these aspects have to be
considered in order to understand the impact of salts on its activity.
In the accepted catalytic mechanism of enzymatic hydrolysis of a phosphate
monoester (ROP) by alkaline phosphatase from E. coli [23,29], the enzyme goes
through
a
sequential
transformation
of
E
(native
enzyme)

EROP
(enzyme-substrate complex)  E-P (covalent enzyme-phosphate intermediate) 
EPi (non-covalent enzyme-phosphate complex)  E (native enzyme). The key
components involved in each active site include three metal ions (Zn12+, Zn22+, and
Mg2+), a serine residue (Ser102), and an arginine residue (Arg166). Formation of the
enzyme-substrate complex (E-ROP) involves coordination of the ester oxygen atom
of the substrate to Zn1 and additional interactions between the non-bridging oxygen
atoms of the substrate with Zn2 and the guanidinium group of Arg166. During
formation
of
the
covalent
enzyme-phosphate
intermediate
(E-P),
the
Mg2+-coordinated hydroxide ion acts as a general base to deprotonate Ser102 for
nucleophilic attack on the phosphorus atom, leading to the release of RO. The
Zn12+-coordinated OH then attacks the E-P intermediate to form the non-covalent
14
EPi complex, while in the meantime the water molecule coordinated to Mg2+ acts as a
general acid to donate a proton to the hydroxyl group of Ser102, leading to the
dephosphorylation and release of the serine residue. Finally, the EPi complex is
dissociated to release the phosphate and to give the free enzyme. Being controlled by
the Zn12+-coordinated hydroxide nucleophile, the rate-limiting step in the mechanism
is switched from the dephosphorylation of the covalent E-P intermediate under acidic
conditions to the dissociation of the non-covalent EPi complex at a higher pH [30,31].
This suggests that alkaline phosphatase is enormously sensitive to the electrostatic
interactions in the active site, which can be regulated by the presence of different salts
and ions. Due to their extraordinarily high positive charge density, the three metal ions
are also important in stabilizing the transition state by neutralizing the high negative
charge density on the substrate.
E. coli AP is believed to be a prototype for all alkaline phosphatases [29],
therefore the above mechanism also applies to the calf intestine alkaline phosphatase
used in our study. As the isoelectric point for this enzyme is 5.7, it should be
negatively charged when exposed to a pH 9.8 DEA buffer.
Because both Zn2+ and Mg2+ are strongly kosmotropic cations (with a viscosity B
coefficient of 0.369 and 0.385, respectively [11]), they have a high tendency to bind
with strongly kosmotropic anions, according to “the law of matching water affinity”
[12]. Although these two metal ions incorporated in the enzyme’s active site are
coordinated, strongly kosmotropic anions such as acetate and citrate may be powerful
in competing with their coordinating ligands. This will significantly affect the
functioning of the three metal ions at the active site, so as to diminish the enzyme
activity. Particularly, strong interactions between these metal ions and the
kosmotropic anions may result in a reduction in the nucleophilicity of the
Zn12+-coordinated hydroxide ion (which is critical in controlling the rate-determining
step), hence minimizing its nucleophilic attack on the covalent E-P intermediate and
slowing down the formation of the non-covalent EPi complex.
A chaotropic anion, on the other hand, has a high polarizability and hence has a
strong tendency to bind to the protein-H2O interface and to interact with the
15
chaotropic cationic moieties (such as amino groups) on the enzyme surface [3]. This
will cause more H+ ions to be accumulated on the surface of the enzyme, leading to a
reduction in the surface pH. Boström et al. [32] have developed a modified
ion-specific double-layer model to demonstrate that the surface pH of proteins is
dependent on the salt concentration and on the ionic species following the Hofmeister
series. A low surface pH would force the active site Ser102 of alkaline phosphatase to
be protonated, thereby weakening its nucleophilic attacking power and in turn
deactivating the enzyme.
Cations can also play an important role in affecting the enzyme activity. Based on
“the law of matching water affinities”, a kosmotropic cation tends to bind strongly to
the kosmotropic carboxylic groups of Glu and Asp located on the enzyme surface,
especially those coordinating to Zn2+ and Mg2+ at the active site. This will modify the
coordinating function of the metal ions and hence modulate the enzyme activity. A
chaotropic cation, on the other hand, can bind to the enzyme surface, due to its high
polarizability, to neutralize the net negative charge of the enzyme, or ion-pair with its
chaotropic counter-anion in the solution to lessen the deactivating effect caused by the
latter.
The above considerations can act together to account for the bell-shaped
relationship between the activity of alkaline phosphatase and the (BB+) value of the
salt added. A similar bell-shaped relationship was also reported between the activity of
NADH oxidase and the anion position in the Hofmeister series [16], indicating that
the enzyme activity is modulated by both kosmotropic and chaotropic anions via
different mechanisms. Their study suggests that both the high rigidity of the active site
caused by kosmotropic anions, and the high flexibility induced by the chaotropic ones,
have a decelerating effect on the enzyme activity. The same research group has
conducted a further study of cation effect on the same enzyme [17] to reinforce the
importance of the accessibility and flexibility of the enzyme’s active site in regulating
the enzyme activity, through the perturbation of the balance between the open and
closed conformations of the enzyme’s active site. This could also be partly responsible
for the results obtained in our experiments. Indeed, a higher flexibility of cytochrome
16
c in the presence of chaotropic anions has been confirmed by a denaturation study of
the protein, as monitored by differential scanning calorimetry and circular dichroism
[33].
3.2. Mechanism of shift in the enzyme’s apparent pH optimum
The salt-induced shift in optimal pH can also be understood by considering the
effect of salts on the surface pH of the enzyme. Charges on and near the protein can
be seen as attracting or repelling H+ ions, and thus establishing a “surface pH”
different from the bulk. (An alternative picture is that charges favor or oppose
protonation of ionizable protein groups, altering pKa’s expressed in terms of the bulk
pH). If the surface charge is negative, the attraction of H+ (lower surface pH) will
mean that a given protonation state of the enzyme will now be found at a higher bulk
pH. Hence we expect an increase in the pH optimum (in terms of bulk pH). Similarly,
a more positive surface charge will be associated with a higher surface pH and a lower
pH optimum for the enzyme. Thus the pH optimum always changes in the opposite
direction to the surface pH. Two contributions have been considered in the modified
ion-specific double-layer model developed by Boström et al. [32]: a) the general
electrostatic effect, determined by ionic strength, reducing the difference between the
surface and bulk pH; and b) a specific effect of highly polarizable chaotropic anions,
which are attracted to the surface, with in consequence H+ accumulation and surface
pH reduction. This suggests that while the specific ion effect (mainly caused by
chaotropic anions) always acts to reduce the surface pH, hence raising the pH
optimum, the general electrostatic effect acts in opposite directions, depending on the
sign of protein net charge. Boström et al. [32] applied this model to lysozyme at pH
4.3, where the protein has a net positive charge, the theoretical surface pH (pH 5.0)
was indeed higher than the bulk (in the absence of salt), and salt addition induced a
gradual reduction in the surface pH, with a greater effect the lower the B value.
Hence the general electrostatic effect acts to reduce the surface pH back closer to the
bulk value, while the chaotrope specific effect also acts in the same direction. In our
case of alkaline phosphatase at pH 9.8, the protein is negatively charged, so the
surface pH will start lower than the bulk. Addition of salts results in two effects: The
17
non-specific electrostatic effect will be to raise the surface pH, back closer to the bulk,
thereby leading to a general reduction in pH optimum (in terms of bulk pH), which is
exactly what is observed (Table 1); whereas the specific ion effect may tend to lower
the surface pH due to the easy binding of the chaotropic anions on the protein surface
and hence accumulation of H+, thus tending to cause an increase in the optimal pH.
These two opposite effects might work together to at least partly account for the weak
correlation of pH optima with B values of the salts, as observed in Table 1. Evidently
more factors are at work here than these two simple effects, and the enzyme activity is
controlled not only by the surface pH but also by other factors involving a
complicated mechanism.
4. Effect of salts on enzyme stability
The impact of salt addition on the thermal stability of the enzyme was also
investigated. The loss of enzyme activity at different temperatures (50oC, 60oC, 70oC)
as a function of incubation time was detected. Our thermostability data seemed to fit
the first-order enzyme deactivation model, suggesting that this kinetics applied to its
thermal inactivation process in the DEA buffer with or without addition of the salts.
The half lives for the enzyme incubated in salt-free buffer at 50oC, 60oC, and 70oC
were determined to be 5.3 hr, 7.9 min, and 1.7 min, respectively. As shown in Figure 6,
all the salts tested showed the ability to stabilize the enzyme when incubated at 70oC,
whereas at the two lower incubation temperatures, there were a few salts (such as
NaNO3, KPF6, NaBF4, NaN3 at 50oC and NaNO3, KPF6, NaBF4 at 60oC) which
caused the enzyme to have a lower half life. Thus more salts exhibited a stabilizing
effect on the enzyme at higher temperatures. It may be that a general electrostatic
effect contributes here. The plots in Figure 6 also illustrate that the half life of the
enzyme is gradually increased with an increase in either the B only (Figure 6A) or
the (BB+) value (Figure 6B) of the salt added, more abruptly when the (BB+)
value becomes positive. This may suggest that the anions play a more dominant role
in affecting the enzyme stability. In fact, all the above destabilizing salts are those
with very negative (BB+) values.
18
4.0
50oC
3.5
˜50oC
3.5
60oC
3.0
Ú60o
3.0
p70oC
2.5
C
2.0
p70o
1.5
C
log (t1/2) (min)
log (t1/2) (min)
4.0
2.5
2.0
1.5
1.0
1.0
0.5
0.5
0.0
0.0
-0.3
-0.2
-0.1
0
B
0.1
0.2
0.3
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
B  B+
Figure 6. Relationship between half lives of the enzyme incubated in the DEA buffer
(1.0 M, pH 9.8) containing 1.0 M different salts at 50oC, 60oC, and 70oC
and the B (A) and (BB+) values (B) of the salts. The salts tested were:
NaCl, KCl, NaNO3, KNO3, NaAc, Na2SO4, Na2MoO4, KPF6, NaBF4, NaN3.
The three horizontal lines indicate the logarithms of the half lives obtained
by the enzyme in the DEA buffer alone, at 50oC, 60oC, and 70oC,
respectively.
The fact that higher enzyme stability was induced by salts with more kosmotropic
anions (e.g., the half life of the enzyme was increased from 8 to 580 min at 60oC in
the presence of 1.0 M Na2SO4) can be further supported by examining the enzyme
activity at high reaction temperatures. The enzyme was found to show a true optimum
temperature, with linear initial progress maintained significantly even above the
optimum. Such behavior is much more common than previously thought [34]. Figure
7 shows the dependence of the reaction rate on the reaction temperature. Addition of
NaCl, NaNO3, or KCl did not alter the optimal reaction temperature of the enzyme
(55oC), but a slight right shift was observed in the presence of NaAc, while sodium
citrate caused an even more obvious increase to 60oC. This gives us a hint that the
enzyme possessed a higher optimal reaction temperature in the presence of a salt with
a more kosmotropic anion. When exposed to temperatures higher than the optimal one,
the enzyme seemed to retain its activity following the order of Na citrate > NaAc >
NaCl ~ KCl > NaNO3. This salt sequence agreed well with the varying order of
19
(BB+) of the salts, or more specifically, with the varying order of the B values of
the involved anions, since all the salts (but KCl) used in this experiment were sodium
salts. The results obtained in this experiment regarding both the shift of optimal
reaction temperature and the retained enzyme activity under high reaction
temperatures suggest that addition of a salt with a higher B value (such as Na citrate
and NaAc) would improve the enzyme’s tolerance against higher temperatures. It is
worth mentioning that the activity of the enzyme obtained at its optimal reaction
temperature also showed a weak bell-shaped relationship with the (BB+) of the salt
added. This further supports our previous comments that alkaline phosphatase
exhibited an optimal activity in the presence of a salt (such as KCl in this experiment)
with both its cation and anion possessing equal water affinity.
40
30
Control
NaNO3
r NaCl
 KCl
 NaAc
 Na citrate
20
10
0
30
40
50
60
70
80
Reaction temperature (oC)
Figure 7. Effect of reaction temperature on enzyme activity in the presence of
different salts (0.8 M in 1.0 M DEA buffer, pH 9.8).
20
4.1. Mechanisms of effects on enzyme stability
There have been many reports showing that enzyme stabilization is generally
favored by the presence of kosmotropic anions [35] and chaotropic cations [14]. Our
study offers another example to support this point of view. As was discussed in [3],
the specific ion effect on enzyme stability can be attributed to the ion’s ability to
affect the hydration shell surrounding the enzyme molecule and to interact directly
with the functional groups on the surface of the enzyme molecule. Because of its
strong interaction with water, a kosmotropic anion like SO42- competes effectively for
the water molecules originally associated with the enzyme molecule, together being
excluded from the enzyme surface. The thermodynamic analysis of such ion exclusion
effects has been presented by Pegram and Record [36]. This encourages the enzyme
molecule to minimize its surface area exposed to the solvent, thereby favoring its
native compact state due to the restored driving force of the hydrophobic effect [10].
Unlike its kosmotropic partner, however, a chaotropic anion like SCN- has not only a
low water affinity but also a high polarizability, thereby preferring to bind to the
protein-water interface and to interact with the chaotropic cationic groups of the
enzyme (such as amino groups), and with the amide moieties, especially those located
in the peptide backbone mainly buried in the native state of the enzyme, according to
“the law of matching water affinity”. This may induce a conformational change to the
enzyme, shifting from its native state to its denatured one. The enzyme is therefore
easily destabilized.
In our experiments, all the stability data correlated similarly well with both the B
and the (BB+) values of the salts tested. This suggests that in terms of affecting the
enzyme stability, the anion of the salt plays a more predominant role relative to its
cation. Broering and Bommarius [37] have also found no significant cation effect on
deactivation of the monomeric red fluorescent protein. This is actually not surprising,
because cations are less polarizable [38] and hydrate less strongly [39]. However, the
stabilizing effect of a kosmotropic anion can be lessened in the presence of a
kosmotropic cation, because both of them have similar water affinity and hence a high
tendency of ion-pairing together. For the same reason, the destabilizing effect of a
21
chaotropic anion can also be weakened in the presence of a chaotropic cation.
Conclusion
The present work has demonstrated explicitly the impact of inorganic salts and
ions on both the buffer pH and the activity and stability of alkaline phosphatase. All
three aspects are controlled directly or indirectly by the Hofmeister effect. There has
been a consensus that a strong electrolyte does not affect the pH of a buffer solution,
or does, so only via electrostatic interactions. Our pH data and its associated
discussion have clearly indicated that the buffer pH can be significantly altered by
addition of salts and that this variation is ion-specific, following the Hofmeister series.
Both the enzyme activity and stability correlate well with the Hofmeister series, and
the combination of the general electrostatic interactions and the specific ionic
dispersion forces make the major contributions to this effect. The stability study of
alkaline phosphatase offers another example of more kosmotropic anions and
chaotropic cations favoring higher enzyme stability. The mechanism involved
concerns the ability of the ions to affect the water solvation layer around the enzyme
molecule and to interact with both the enzyme surface and internal structure. Anions
play a more predominant role than cations in affecting the enzyme stability. The
impact of ions on enzyme activity, on the other hand, is not so straightforward. Unlike
enzyme stability that generally follows a common trend and is dependent more
predominantly on the anion, the alteration of enzyme activity with addition of
different salts is affected by both anions and cations and is different from enzyme to
enzyme, because the mechanism is related to the impact of both these ions on the
surface pH, active site, and catalytic mechanism of each specific enzyme. All these
can be explained with the aid of the Hofmeister series. The activity of alkaline
phosphatase showed a bell-shaped relationship with the (BB+) values of the salts
present, with an optimal activity presented when in the presence of a salt, such as
KNO3, the cation and anion of which have similar kosmotropic/chaotropic properties.
22
References
[1]
F. Hofmeister, Zur lehre der wirkung der salze. Zweite mittheilung, Arch. Exp.
Pathol. Pharmakol. 24 (1888) 247260. Translated and republished by W. Kunz,
J. Henle, B.W. Ninham, Curr. Opin. Colloid Interface Sci 9 (2004) 19–37.
[2]
H. Zhao, Effect of ions and other compatible solutes on enzyme activity, and its
implication for biocatalysis using ionic liquids, J. Mol. Catal. B: Enzym. 37
(2005) 1625.
[3]
Z. Yang, Hofmeister effects: an explanation for the impact of ionic liquids on
biocatalysis, J. Biotechnol. (in press)
[4]
J.C. Warren, L. Stowring, M.F. Morales, The effect of structure-disrupting ions
on the activity of myosin and other enzymes, J. Biol. Chem. 241 (1966)
309316.
[5]
W. Kunz, P. Lo Nostro, B.W. Ninham, The present state of affairs with
Hofmeister effects, Curr. Opin. Colloid Interface Sci. 9 (2004) 118.
[6]
R.L. Baldwin, How Hofmeister ion interactions affect protein stability, Biophys.
J. 71 (1996) 20562063.
[7]
B. Hribar, N.T. Southall, V. Vlachy, K.A. Dill, How ions affect the structure of
water, J. Am. Chem. Soc. 124 (2002) 1230212311.
[8]
P. Bauduin, A. Renoncourt, D. Touraud, W. Kunz, B.W. Ninham, Hofmeister
effect on enzymatic catalysis and colloidal structures, Curr. Opin. Colloid
Interface Sci. 9 (2004) 4347.
[9]
M. Boström, D.R.M. Williams, B.W. Ninham, Why the properties of proteins in
23
salt solutions follow a Hofmeister series, Curr. Opin. Colloid Interface Sci. 9
(2004) 4852.
[10] K.D. Collins, G.W. Neilson, J.E. Enderby, Ions in water: Characterizing the
forces that control chemical processes and biological structure, Biophys. Chem.
128 (2007) 95104.
[11] H.D.W. Jenkins, Y. Marcus, Viscosity B-coefficients of ions in solution, Chem.
Rev. 95 (1995) 26952724.
[12] K.D. Collins, Ions from the Hofmeister series and osmolytes: effects on proteins
in solution and in the crystallization process, Methods 34 (2004) 300311.
[13] E.M. Bowers, L.O. Ragland, L.D. Byers, Salt effects on -glucosidase:
pH-profile narrowing, Biochim. Biophys. Acta 1774 (2007) 15001507.
[14] D. Constantinescu, H. Weingärtner, C. Herrmann, Protein denaturation by ionic
liquids and the Hofmeister series: A case study of aqueous solutions of
ribonuclease A, Angew. Chem. Int. Ed. 46 (2007) 88878889.
[15] M. Nagaoka, T. Shiraishi, F. Furuhata, Y. Uda, Effects of inorganic anions on the
activation of acid sialidases, Biol. Pharm. Bull. 26 (2003) 295298.
[16] G. Žoldák, M. Sprinzl, E. Sedlák, Modulation of activity of NADH oxidase from
Thermus thermophilus through change in flexibility in the enzyme active site
induced by Hofmeister series anions, Eur. J. Biochem. 271 (2004) 4857.
[17] K. Tóth, E. Sedlák, M. Sprinzl, G. Žoldák, Flexibility and enzyme activity of
NADH oxidase from Thermus thermophilus in the presence of monovalent
cations of Hofmeister series, Biochim. Biophys. Acta 1784 (2008) 789795.
24
[18] R.V. Ulijn, P.J. Halling, Solid-to-solid biocatalysis: thermodynamic feasibility
and energy efficiency, Green Chem. 6 (2004) 488496.
[19] O. Wesołowska, I. Krokoszyńska, D. Krowarsch, J. Otlewski, Enhancement of
chymotrypsin-inhibitor/substrate interactions by 3 M NaCl, Biochim. Biophys.
Acta 1545 (2001) 7885.
[20] H. Oneda, Y. Muta, K. Inouye, Substrate-dependent activation of thermolysin by
salt, Biosci. Biotechnol. Biochem. 68 (2004) 18111813.
[21] F. van Rantwijk, R.A. Sheldon, Biocatalysis in ionic liquids, Chem. Rev. 107
(2007) 27572785.
[22] C. Roosen, P. Müller, L. Greiner, Ionic liquids in biotechnology: applications
and perspectives for biotransformations, Appl. Microbiol. Biotechnol. 81 (2008)
607614.
[23] B. Stec B, K.M. Holtz, E.R. Kantrowitz, A revised mechanism for the alkaline
phosphatase reaction involving three metal ions, J. Mol. Biol. 299 (2000)
13031311.
[24] A. Salis, D. Bilaničová, B.W. Ninham, M. Monduzzi, Hofmeister effects in
enzymatic activity: weak and strong electrolyte influences on the activity of
Candida rugosa lipase, J. Phys. Chem. B 111 (2007) 11491156.
[25] P. Bauduin, F. Nohmie, D. Touraud, R. Neueder, W. Kunz, B.W. Ninham,
Hofmeister specific-ion effects on enzyme activity and buffer pH: Horseradish
peroxidase in citrate buffer, J. Mol. Liquids 123 (2006) 1419.
[26] M.C. Pinna, A. Salis, M. Monduzzi, B.W. Ninham, Hofmeister series: the
25
hydrolytic activity of Aspergillus niger lipase depends on specific anion effects,
J. Phys. Chem. B. Lett. 109 (2005) 54065408.
[27] H.
Galster,
pH
Measurement:
Fundamentals,
Methods,
Applications,
Instrumentation, VCH Publishers, New York, 1991.
[28] I.E. Gouvea, W.A.S. Judice, M.H.S. Cezari, M.A. Juliano, T. Juhάsz, Z. Szeltner,
L. Polgάr, L. Juliano, Kosmotropic salt activation and substrate specificity of
poliovirus protease 3C, Biochemistry 45 (2006) 1208312089.
[29] K.M. Holtz, E.R. Kantrowitz, The mechanism of the alkaline phosphatase
reaction: insights from NMR, crystallography and site-specific mutagenesis,
FEBS Lett. 462 (1999) 711.
[30] W.E. Hull, S.E. Halford, H. Gutfreund, B.D. Sykes,
31
P nuclear magnetic
resonance study of alkaline phosphatase: the role of inorganic phosphate in
limiting the enzyme turnover rate at alkaline pH, Biochemistry 15 (1976)
15471561.
[31] P. Gettins, J.E. Coleman,
31
P nuclear magnetic resonance of phosphoenzyme
intermediates of alkaline phosphatase. J Biol Chem. 258 (1983) 408416.
[32] M. Boström, D.R.M. Williams, B.W. Ninham, Specific ion effects: Why the
properties of lysozyme in salt solutions follow a Hofmeister series. Biophys. J.
85 (2003) 686694.
[33] N. Tomáčková, R. Varhač, G. Žoldák, D. Sedláková, E. Sedlá, Conformational
stability and dynamics of cytochrome c affect its alkaline isomerization, J. Biol.
Inorg. Chem. 12 (2007) 257266.
26
[34] R.M. Daniel, M.J. Danson, R. Eisenthal, The temperature optima of enzymes: a
new perspective on an old phenomenon, Trends Biochem Sci. 26 (2001)
223225.
[35] J.M. Broering, A.S. Bommarius, Evaluation of Hofmeister effects on the kinetic
stability of proteins, J. Phys. Chem. B 109 (2005) 2061220619.
[36] L.M. Pegram, M.T. Record, Jr., Thermodynamic origin of Hofmeister ion effects,
J. Phys. Chem. B 112 (2008) 94289436.
[37] J.M. Broering, A.S. Bommarius, Cation and strong co-solute effects on protein
kinetic stability, Biochem Soc Trans. 35 (2007) 16021605.
[38] A. Grossfield, P. Ren, J.W. Ponder, Ion solvation thermodynamics from
simulation with a polarizable force field, J. Am. Chem. Soc. 125 (2003)
1567115682.
[39] J.E. Combariza, N.R. Kestner, J. Jortner, Energy-structure relationships for
microscopic solvation of anions in water clusters, J. Chem. Phys. 100 (1994)
28512864.
[40] H. Zhao, Z. Song, Nuclear magnetic relaxation of water in ionic-liquid solutions:
determining the kosmotropicity of ionic liquids and its relationship with the
enzyme enantioselectivity. J. Chem. Technol. Biotechnol. 82 (2007): 304–312.
27