928 Biotechnol. Prog.

2003, 19, 928935

Reverse Micellar Extraction and Precipitation of Lysozyme Using

Sodium Di(2-ethylhexyl) Sulfosuccinate
Youn-Ok Shin, Martin E. Weber, and Juan H. Vera*
Department of Chemical Engineering, McGill University, 3610 University Street,
Montreal, Quebec, Canada H3A 2B2

Sodium di(2-ethylhexyl) sulfosuccinate, referred to as Aerosol-OT or AOT, was used

to remove lysozyme from an aqueous phase via reverse micellar extraction and
precipitation method. For both methods, when the surfactant was in excess, a complete
removal of lysozyme from the aqueous phase was obtained at the values of pH below
the pI of lysozyme. However, for the reverse micellar method, a solubilization limit of
lysozyme in the organic phase was observed, and a white precipitate was formed at
the aqueous-organic interface. This observation suggested using AOT directly as a
precipitating ligand. The lysozyme precipitated with AOT was fully recovered, with
its original enzymatic activity, using acetone as a recovery solvent. A mechanism is
suggested to explain the solubilization of lysozyme in an AOT reverse micellar system.
It is shown that a direct precipitation method can be used with advantage instead of
using the reverse micellar extraction method to recover lysozyme from an aqueous
phase.

Introduction
Proteins are produced using bacterial cells, mam-
malian cells, or plant extracts. Separation methods such
as crystallization, column chromatography, or electro-
phoresis have been applied to purify these proteins (1).
As an alternative, the use of polymers coupled with an
affinity ligand was suggested (2-5). Upon a change in
temperature or pH, the protein-polymer-affinity ligand
complex precipitates from the aqueous phase (6). The
disadvantages of this method are that the percent
precipitation is limited by the solubility of the polymer
in the aqueous phase and that a membrane separation
is required to separate protein from the polymer at the
end of the process (2, 6). Furthermore, only proteins that
can sustain elevated temperatures can be purified using
this method (5).
Liquid-liquid extraction of proteins using reverse
micellar phase was proposed as an alternative purifica-
tion method (7). Ionic surfactants are usually considered
as a protein denaturant (8, 9). Thus, for the protein
purification, the use of commercially available ionic
surfactants has been limited to reverse micellar extrac-
tion. Upon contact of an organic phase containing the
surfactant with an aqueous phase containing proteins
and electrolytes, the surfactant forms reverse micelles
in the organic phase. As a result of their electrostatic
interaction with the surfactant headgroup, proteins and
other charged species are solubilized inside the water
pools of the reverse micelles (10). This process is referred
to as forward extraction. Many reverse micellar extrac-
tion studies used sodium di(2-ethylhexyl) sulfosuccinate
(AOT) as an anionic surfactant (11-16) and trioctylm-
ethylammonium chloride (TOMAC) or cetyltrimethylam-
monium bromide (CTAB) as cationic surfactants (17-20).
* To whom correspondence should be addressed. Fax: (514) 398-
6678. Email: juan.vera@mcgill.ca.
10.1021/bp025789r CCC: $25.00 2003 American Chemical Society and American Institute of Chemical Engineers
Published on Web 04/12/2003
The ion distribution around the surfactant heads (16, 21,
22) and the reverse micellar size (23, 24) have been
considered to be factors that influence protein extraction.
The flexibility (25) and the surface charge (26) of protein
molecules have also been related to the solubilization
behavior.
Once extracted, the proteins were recovered into a
fresh aqueous phase with pH and salt concentration
adjustment (12). This process is referred to as back
extraction. The rate of the back extraction is much slower
than the rate of the forward extraction (11, 27, 28).
Adjustment of pH and ionic strength alone may not be
sufficient to recover the protein from the reverse micelles
(29, 30). Ethanol (31), 2-propanol (32), ethyl acetate (33),
and counter surfactants (34) have been added in attempts
to improve the recovery. The use of an acetone phase,
instead of using a fresh aqueous phase, was suggested
as an alternative to recover the proteins from the reverse
micellar phase (35, 36).
One mechanism of protein loss, often found in studies
of reverse micellar extraction processes, is through the
formation of a water-insoluble protein-surfactant com-
plex that precipitates at the interface between the
aqueous and organic phases (37, 38). The formation of
this precipitate depended largely on the mass transfer
kinetics (39) and on the saturation limit of the protein
in the reverse micellar phase (40). The protein-surfac-
tant complex which precipitated at the interface was
viewed as an undesirable product, since the protein was
thought to be denatured (41). Attempts were made to
avoid the formation of a protein-surfactant complex by
using new surfactants to form the reverse micellar phases
(38, 42).
Even though the technology was first suggested more
than 2 decades ago, reverse micellar extraction still
suffers from poor recovery efficiency and protein loss at
the aqueous-organic interface. In this work, it is dem-
onstrated that the lysozyme precipitated at the aqueous-
Biotechnol. Prog., 2003, Vol. 19, No. 3
organic interface during the reverse micellar extraction
process can be recovered by dissolving it in a polar
organic solvent such as acetone. On the basis of this
finding, a new method that uses AOT as a precipitating
ligand is proposed. The precipitation of lysozyme is
studied, and the yields are compared with the results
obtained with its reverse micellar extraction. Comparison
of solubilization of the lysozyme-surfactant complex in
reverse micelles with the direct extraction of lysozyme
into a reverse micellar phase suggests a mechanism to
explain the latter process.
Materials and Methods
Materials. Sodium di(2-ethylhexyl) sulfosuccinate
(AOT), lysozyme (chicken egg white, pI ) 11.0, MW )
14.3 kDa), and Micrococcus lysodeikticus were purchased
from Sigma (Oakville, ON). The purchased lysozyme
contained 5-10% buffer: sodium acetate/sodium chloride/
hydrochloric acid. Phosphate buffer (pH 6.2) purchased
from Anachemia (Oakville, ON) was used for enzymatic
activity measurements. Reagent grade isooctane and
decanol, HPLC grade acetonitrile and triflouric acid,
sodium chloride, and 1 N NaOH were obtained from
Fisher Scientific (Montreal, QC).
Extraction of Lysozyme into Reverse Micellar
Phase. A volume of 10 mL of an aqueous solution
containing lysozyme and 0.1 M NaCl was contacted with
an equal volume of an isooctane phase containing a
specified concentration of AOT. Without NaCl in the
aqueous phase, no reverse micelles were formed. The
mixture was vortexed for 5 s and left for phase separation
for 12 h at room temperature (21 ( 1 C). The pH of the
initial aqueous lysozyme solution, adjusted to 6.0 ( 0.5
using 1 N NaOH, was measured using an accupHast
reference/pH probe, purchased from Fisher Scientific
(Montreal, QC). Since the pH of the initial lysozyme
solution varied between 4 and 6, the concentration of
NaOH in the initial lysozyme solution was less than 0.04
mM.
Formation of Lysozyme-AOT Complex by Direct
Addition of AOT. An aqueous solution containing AOT
was added to 10 mL of an aqueous solution, which
contained from 0 to 5 g/L lysozyme. No pH adjustment
was made to the initial AOT or lysozyme-containing
aqueous solutions. The solubility of AOT in water was
reported to be 15 g/L or 33.8 mM in the temperature
range from 20 to 70 C (43). The critical micelle concen-
tration (CMC) of AOT in water is in the range from 2.5
to 4.1 mM at 25 C (44). The initial concentration of AOT
in the lysozyme solution was fixed between 0.02 and 1.5
mM, to avoid the formation of micelles. The volume of
the AOT solution added to the lysozyme solution was less
than 0.2 mL. As soon as the AOT solution was added,
the insoluble lysozyme-AOT complex formed. The mix-
ture was vortexed for 5 s and centrifuged, and then the
supernatant was removed. The lysozyme-ligand complex
was washed with water by adding 10 mL of distilled
water to the test tube containing precipitated lysozyme-
ligand complex, followed by a second centrifugation.
Lysozyme Recovery from Lysozyme-AOT Com-
plex. Lysozyme was recovered from the white precipitate
formed at the aqueous-organic interface during the
reverse micellar extraction process and from the
lysozyme-AOT complex formed by direct addition of AOT
using acetone as a recovery solvent. Ten milliliters of
acetone was added to the test tubes containing the
precipitated lysozyme-AOT complex. The solid dissolved
immediately in acetone and resulted in one clear solution.
929
A small amount of NaCl solution (normally less than 10
L of 0.1 M NaCl) was then added to the acetone phase
in order to neutralize the charges and dissociate the
lysozyme from the surfactant. The surfactant-free
lysozyme precipitated as a solid 1-2 min after the
addition of NaCl, while the surfactant remained in the
acetone phase. The recovered solid lysozyme was washed
with acetone to remove residual surfactant. The qualita-
tive and quantitative analyses of the final product were
carried out by dissolving the recovered-solid lysozyme
into a fresh aqueous phase.
Solubility Product Measurement of Lysozyme-
AOT Complex. For the measurement of the solubility
product of the lysozyme-AOT complex, the lysozyme-
AOT complex formed by direct addition of AOT was
collected by centrifugation, washed three times with
distilled water, and then dispersed in 5 mL of fresh
distilled water. The mixture was stirred for 24 h at room
temperature. An excess solid phase was visible in the
bottom of the test tube. The concentrations of lysozyme
and AOT in the aqueous phase were measured after
filtering the solution though a 0.45 m filter.
Analytical Techniques. The concentrations of ly-
sozyme and AOT in the aqueous phase were measured
using an Agilent high performance liquid chromatograph
(HPLC)-1100 with a Zorbax 300SB-C8 (4.6 mm i.d. 150
mm length) column. The mobile phase consisted of two
solvents, 5% acetonitrile and 0.1% triflouric acid in water
(solvent A), and 5% water and 0.085% triflouric acid in
acetonitrile (solvent B). The sample volume was set at 5
L, and the wavelength of the detector was set at 210
nm.
The lysozyme concentration in the organic phase was
measured on a Cary Varian 1/3 spectrophotometer at 280
nm (A280nm) and at 310 nm (A310nm). The reading measured
at 310 nm was subtracted from the reading measured at
280 nm to eliminate the noise caused by surfactant. A
linear calibration curve of lysozyme in the AOT reverse
micellar phase was obtained with standards containing
from 0 to 0.7 g/L lysozyme. The samples with higher
lysozyme concentration were diluted using an identical
AOT reverse micellar phase containing no lysozyme.
The enzymatic activity of lysozyme was measured
according to the method described by Davis et al. (45).
The concentrations of sodium and chloride ions in the
aqueous phase were measured using atomic absorption
spectroscopy and ion chromatography, respectively. All
experiments were performed using two to six replicates,
and the standard deviation of the measurements is
indicated in the figures by error bars.
Results and Discussion
Reverse Micellar Extraction of Lysozyme as a
Function of AOT Concentration. The effect of AOT
concentration on the solubilization of lysozyme into the
reverse micellar phase was examined using an aqueous
phase containing from 0 to 15 g/L (0 to 11 mM) lysozyme
and an organic phase containing from 0 to 100 mM AOT.
The salt concentration in the lysoyzme solution was 0.1
M NaCl. Figure 1 shows the concentration of lysozyme
in the reverse micellar phase as a function of the initial
lysozyme concentration in the aqueous phase. As the
surfactant concentration in the organic phase increased,
the amount of lysozyme solubilized into the reverse
micellar phase increased. For the case of 0 mM AOT
reverse micellar phase, no lysozyme was solubilized in
the organic phase, confirming that lysozyme was in-
soluble in isooctane. At 20 mM AOT, the maximum
930
Figure 1. Solubilization of lysozyme into AOT reverse micellar
phases; initial aqueous phase, 0.1 M NaCl, pH ) 6.0 ( 0.5
adjusted with NaOH; initial volumes of aqueous and organic
phases ) 10 mL.
concentration of lysozyme in the reverse micellar phase
was 0.09 mM. The samples prepared with an aqueous
phase containing from 0.1 to 0.15 mM lysozyme showed
a lysozyme concentration of 0.09 mM in the organic
phase, indicating that this is the saturation limit of
lysozyme in the 20 mM AOT reverse micellar phase. With
an initial lysozyme concentration of 0.18 mM, the reverse
micellar phase contained about 0.07 mM lysozyme. On
the other hand, the equilibrium lysozyme concentration
in the aqueous phase after contact with the reverse
micellar phase was below the detection limit of HPLC
for all samples, indicating essentially complete removal
of lysoyzme from the aqueous phase. The lysozyme
removed from the aqueous phase but not solubilized into
the reverse micellar phase formed a white precipitate at
the aqueous-organic interface.
An increase in AOT concentration in the organic phase
further increased the solubility limit of lysoyzme in the
AOT reverse micellar phase. At 30 mM AOT, the extrac-
tion of lysozyme into the reverse micellar phase was
complete up to an initial lysozyme concentration of 0.38
mM. The samples prepared with an aqueous phase
containing 0.5 mM lysozyme resulted in a sudden de-
crease in the lysozyme concentration in the 30 mM AOT
reverse micellar phase. Similarly, a solubilization limit
of lysozyme in the 70 and 100 mM AOT reverse micellar
phase was 0.75 and 0.83 mM, respectively. When the
initial lysozyme concentration in the aqueous phase was
greater than the saturation limit value, a noticeable
decrease in the lysozyme concentration in the reverse
micellar phase was obtained, and a white precipitate at
the aqueous-organic interface was observed. A similar
phenomenon was observed in a previous study using an
AOT (37) or a dioctyldimethylammonium chloride reverse
micellar phase (40).
Solubilization Mechanism of Lysozyme into a
Reverse Micellar Phase. On the basis of the results
in Figure 1, a mechanism of lysozyme extraction into a
reverse micellar phase is proposed:
Biotechnol. Prog., 2003, Vol. 19, No. 3
Figure 2. Initial pH of the aqueous solution as a function of
lysozyme concentration at room temperature.
(i) Upon contact, lysozyme forms a complex with the
ligand as a result of electrostatic interactions.
(ii) The lysozyme-AOT complex has a low solubility
in water and in isooctane, since no solubilization of
lysozyme was observed in the pure isooctane phase, as
shown in Figure 1.
(iii) The complex solubilizes in a reverse micellar phase
if present; otherwise, the complex remains as solid.
(iv) The solubilization of the lysozyme-AOT complex
takes place until the saturation limit of lysozyme in the
reverse micellar phase is reached. The excess lysozyme-
AOT complex remains at the aqueous-organic interface.
To verify this hypothesis, two-step experiments were
performed. First the lysozyme was directly precipitated
from the aqueous phase using AOT as a precipitation
ligand. This precipitate was then solubilized into reverse
micellar phases containing different concentrations of
AOT.
Formation of Lysozyme-AOT Complex by Direct
Addition of AOT. The precipitation of lysozyme was
performed using an initial lysozyme aqueous solution
containing 0.15-5 g/L (0-0.36 mM) lysozyme. No pH and
salt concentration adjustment was made to the lysozyme
solution. Figure 2 shows the pH of the initial lysozyme
aqueous solution. The pH of the lysozyme solution,
containing the buffer coming with the commercial
lysozyme, decreased from 5.8 to 4.2 as the lysozyme
concentration increased from 0.01 to 0.36 mM. Two sets
of experiments were performed, and the difference in the
pH between the replicates was about (0.1.
An aqueous phase containing AOT was then pipetted
into the initial lysozyme aqueous solution to form a
lysozyme-AOT precipitate. Figure 3 shows the percent
removal of lysozyme from the aqueous solution as a
function of the molar ratio, R, between AOT and lysozyme
initially added to the solution. Increasing the molar ratio
increased the percent removal until a molar ratio of about
10, at which 100% percent removal of lysozyme was
obtained. When R was greater than 10, the excess AOT
remained in the aqueous phase. When R was less than
10, the concentration of AOT remaining in the solution
was below the detection limit of HPLC, indicating that
Biotechnol. Prog., 2003, Vol. 19, No. 3 931

Figure 4. Equilibrium pH of the aqueous phase upon removal

of lysozyme as an insoluble lysozyme-AOT complex: no salt
Figure 3. Percent of lysozyme precipitated with AOT: no salt added, no pH adjustment.
added, no pH adjustment.

all of the ligand added to the solution formed lysozyme-

ligand complex and precipitated. The slope of the line in
Figure 3 is 10, indicating that 1 mol of precipitated
lysozyme was complexed with 10 mol of AOT. A similar
precipitating behavior of cytochrome c was observed
where the cytochrome c complexed with AOT at a fixed
molar ratio of 11 and precipitated from the aqueous phase
(41).
Figure 4 shows the equilibrium pH after the lysozyme-
ligand complex was formed. The pH of the solution
increased as more lysozyme was removed from the
aqueous phase. When lysozyme was removed completely
(R g 10), the excess ligand remained in the solution, and
the pH was approximately 6.
Effect of Salt on the Formation of Lysozyme-AOT
Complex. The effects of salt on the precipitation of
lysozyme were studied using an aqueous phase contain-
ing 0.1-1 M NaCl and lysozyme. Figure 5 shows the
percent precipitation of lysozyme as a function of the
molar ratio when NaCl was present in the initial lysozyme
solution. At 0.1 M NaCl, the results were the same as
those depicted in Figure 3. At higher salt concentrations,
the amount of lysozyme complexed with AOT decreased.
At 0.3 M NaCl, only 60% of the lysozyme was precipitated
at R ) 10, and a value of R ) 20 was required for a 95% Figure 5. Effect of salt on the percent lysozyme precipitated
precipitation. A salt concentration of 1 M decreased the with AOT.
percent precipitation to 30-35% with little effect of the
molar ratio. Increasing the salt concentration increased phase containing 0.1 M NaCl and contacted with an equal
the concentration of chloride counterions competing with volume of an isooctane phase containing 5-100 mM AOT.
AOT to bind with the positively charged lysozyme. In The mixture was stirred vigorously for 72 h to ensure
addition, as the salt concentration increased, the dis- the complete solubilization of lysozyme-AOT complex.
sociation of the AOT anion from its counterion (sodium) After a phase separation of 12 h, the organic and aqueous
decreased, and complex formation between lysozyme and phase samples were collected for lysozyme analysis.
AOT decreased. The decreasing behavior of the percent Figure 6 shows the concentration of lysozyme achieved
precipitation of cytochrome c was also observed as the in an AOT reverse micellar phase using lysozyme-ligand
result of an increase in the salt concentration in the complex, compared with the reverse micellar extraction
aqueous phase (41). result from Figure 1. The error bars indicate the 95%
Solubilization of Lysozyme-AOT Complex into An confidence interval. As shown in Figure 6, for both
AOT Reverse Micellar Phase. The dried lysozyme- experiments, the concentration of lysozyme increased in
ligand complex, formed without addition of salt to the the reverse micellar phase as the concentration of AOT
initial lysozyme aqueous phase, was added to an aqueous increased. For the reverse micellar extraction, duplicate
932
Figure 6. Concentration of lysozyme in AOT reverse micellar
phase initially introduced as insoluble lysozyme-AOT complex
and extracted from an aqueous phase using a reverse micellar
phase.
results showed less than 1% standard deviation, and the
error bars are not visible in the figure, indicating the
excellent reproducibility of data. For the solubilization
of the lysozyme-AOT complex into the reverse micellar
phase, the 95% confidence interval was more noticeable.
This is believed to be due to the vigorous stirring and
the time taken to solubilize the dry lysozyme-ligand
complex into a reverse micellar phase, which could cause
physical changes to the lysozyme. Nevertheless, the
concentration of lysozyme in the reverse micellar phase
obtained by solubilizing the lysozyme-AOT complex
showed the same trend as the reverse micellar extraction
experiment. On the other hand, the lysozyme concentra-
tion in the aqueous phase was below the detection limit
of the HPLC, confirming the insolubility of lysozyme-
AOT complex in the aqueous phase. On the basis of this
result, we believe that the mechanism for the reverse
micellar extraction of lysozyme consists of two steps: first
the formation of the water insoluble lysozyme-AOT
complex that precipitates out of the aqueous phase and
then the solubilization of this lysozyme-AOT complex
in the reverse micellar phase. Once the reverse micellar
phase reaches the saturation limit, as shown in Figure
1, the lysozyme-AOT complex remains as a precipitate
at the interface.
A small-angle X-ray light scattering study of cyto-
chrome c in an AOT reverse micellar phase showed that
the protein is located at the interface of the AOT reverse
micelles (41). It is believed that the presence of both
hydrophobic and hydrophilic domains is necessary for
solubilization of the protein-AOT complex. In water or
a pure organic phase, no solubilization of lysozyme-AOT
complex was observed. In the presence of reverse mi-
celles, which contain both hydrophobic and hydrophilic
domains, the lysozyme-AOT complex may be solubilized
in the isooctane phase. However, the shape of the reverse
micelles and the location of proteins in a reverse micellar
phase are not fully understood.
Recovery of Lysozyme from Lysozyme-AOT Com-
plex. The recovery of lysozyme from precipitate was
Biotechnol. Prog., 2003, Vol. 19, No. 3
carried out using acetone as a recovery solvent. First we
tested the purity and the activity of lysozyme recovered
from the white precipitate formed at the aqueous-
organic interface during the reverse micellar extraction
process. Upon the addition of acetone, the white lysozyme-
AOT precipitate was dissolved within 1-2 min. Ligand-
free lysozyme precipitated as a solid from this solution,
while the AOT remained in the acetone phase. We then
tested the recovery of lysozyme from the lysozyme-AOT
precipitate formed by adding AOT directly as a precipi-
tating ligand. In both cases, the results were similar. The
percent recovery of lysozyme was (70 ( 18)%. Some loss
of lysozyme occurred during washing. The final product
was analyzed for its enzymatic activity by dissolving the
washed ligand-free lysozyme precipitate in distilled
water. The activity of lysozyme solubilized in the final
aqueous phase was between 45,000 and 51,000 units/mg
protein. The activity of the original lysozyme solution,
before the contact with AOT, was 48,000 ( 3,000 units/
mg protein. The use of acetone did not affect the lysozyme
activity, and the recovered lysozyme showed the same
activity as the initial lysozyme. Clearly, the formation
of the lysozyme-ligand complex and the recovery of the
lysozyme using acetone did not denature the lysozyme.
For comparison, the recovery of lysozyme from AOT
reverse micellar phase using acetone showed 75% recov-
ery (30). This again suggests that the mechanism of
reverse micellar extraction and direct precipitation of
lysozyme are similar.
On the basis of these observations and on the results
shown in Figure 6, we conclude that the reverse micellar
extraction of lysozyme using AOT surfactant can be
simplified by using a direct precipitation method. This
direct method eliminates the use of an isooctane organic
phase to form reverse micellar phase. The amount of AOT
required per mole of purified lysozyme was 1 order of
magnitude less using a direct precipitation method. For
the reverse micellar extraction, about 100 mol of AOT
were required to extract 1 mol of lysozyme into the
reverse micellar phase. In comparison, for the direct
precipitation method, only 10 mol of AOT were sufficient
to extract 1 mol of lysozyme from the aqueous phase.
As a practical application, the use of AOT as a
precipitating ligand was studied to test the selectivity of
the method. Lysozyme was successfully purified from hen
egg white, with 100% purity without any denaturing (46).
The potential use of this direct precipitation method for
other proteins using AOT is presently under study.
Modeling of Lysozyme-AOT Complex Formation.
For modeling, the term ligand is used here for AOT
because it was used as a ligand to form insoluble
compound rather than as a surface-active agent. The ter
crystallized lysozyme indicates the bulk lysozyme pur-
chased from the supplier (Sigma, ON). Lysozyme-ligand
complex indicates the white insoluble complex formed
upon the addition of AOT to the initial lysozyme aqueous
solution, and ligand-free lysozyme indicates the AOT-free
solid lysozyme precipitated with acetone.
The model proposed here is based on the following
assumptions:
(1) The commercially available crystallized lysozyme
contains hydrochloric acid (chlorohydrate compounds).
Thus, the positively charged lysozyme has chloride ions
as counterions in an aqueous phase.
(2) Lysozyme forms an insoluble complex when associ-
ated with a well-defined number, Z, of anionic ligand
residues. Lysozyme complexed with less than Z number
of the ligands is soluble in the aqueous phase.
Biotechnol. Prog., 2003, Vol. 19, No. 3
(3) No inorganic salts are removed from the aqueous
phase with the insoluble lysozyme-ligand complex.
(4) With hydrogen ion rather than sodium as its
couterion, AOT is insoluble in water.
(5) Sodium acetate and sodium chloride are fully
ionized in lysozyme solution.
The first assumption is based on a study of the
variation of the pH of the solution as lysozyme is
dissolved in water as shown in Figure 2. The nature of
the crystallized lysozyme depends on the type of acid
solution used in the crystallization process (47). It was
assumed that hydrochloric acid was used to precipitate
the commercial lysozyme.
Assumption (2) is based on the results shown in Figure
3. When the lysozyme was in excess, all of the added AOT
formed an insoluble complex with the lysozyme, and the
concentration of surfactant in the aqueous phase was
negligible. When AOT was in excess, all lysozyme was
removed from the aqueous phase. Thus, the insoluble
complex forms only when a mole of lysozyme binds with
Z moles of ligand.
Assumption (3) is based on the results of the sodium
and chloride ion balances before and after the formation
of the insoluble complex. When measured with atomic
absorption spectroscopy and ion chromatography, the
change in the sodium and chloride ion concentrations,
before and after the formation of lysozyme-ligand in-
soluble complex, was undetectable within experimental
error.
Assumption (4) is suggested by the pH change in the
aqueous phase upon the addition of AOT solution, as
shown in Figure 4.
Surface Charge of Lysozyme. Lysozyme contains 14
acidic (aspartic, tyrosine, and glutamic) amino acid
residues and 19 basic (arginine, histidine, and lysine)
amino acid residues (48). Barlow and Thornton (49)
reported the ionic groups of lysozyme as 11 acidic groups
excluding Tyrosine, and 18 basic groups excluding His-
tidine. The dissociation constants of most of these ion-
isable groups are reported in the literature (48, 50, 51).
One carboxylic group is in the interior of the protein
molecule, and it does not ionize in solution (50). Kuramit-
su and Hamaguchi (50) suggested the following equation
for the net surface charge, Z, of chicken egg white
lysozyme:
32 KiR
Z ) 19 - +
(1)
i)1 H + KiR
where Ki is the dissociation constant of hydrogen from
group i, and R is a parameter describing the electrostatic
interactions between a protein molecule containing a net
surface charge of Z and a proton (50, 52). The equation
considers 32 ionic groups, the 33 reported by Sakakibara
and Hamaguchi (48) less one interior carboxylic group
(50). The net surface charge of a protein is an average
over a large number of individual forms of molecule
constantly giving up and taking on protons, and its actual
net charge at any moment can be greater or less than
the average value of Z (52).
The net surface charge of lysozyme calculated from eq
1 was between +9 and +11 in the pH range between 5.8
and 4.2, shown in Figure 1. For simplicity, the net surface
charge of lysozyme was considered to be constant at +10
in the model.
The crystallized lysozyme obtained from the manufac-
turer contained about 5-10 wt % sodium acetate and
sodium chloride (Sigma, ON). The initial lysozyme was
933
assumed to contain 5% NaCl and 5% C3H2ONa. The
equilibrium pH of an aqueous solution containing an
initial concentration of lysozyme, C0Lys, can be calculated
from a charge balance:
C0LysZ + CH+ + CNa+ ) COH- + CCl- + CC2H3O2- (2)
where CC2H3O2- indicates the acetate ion concentration.
The value of COH- is obtained using the ion product for
water, Kw, taken as 10-14 (53), and the dissociation
constant of acidic acid, Ka, taken as 1.8 10-5 (54):
Kw ) CH+COH- (3)
CH+CC2H3O2-
Ka ) (4)
CHC2H3O2
Equations 1 and 2 were solved for CH+ at different
concentrations of lysozyme, and these values were used
to calculate the pH values. These results are shown in
Figure 2. If the presence of HCl in lysozyme is ignored,
the equations fail to reproduce the variation pH shown
in the figure.
Formation of Lysozyme-Ligand Complex. At a
pH less than its pI, lysozyme has an overall positive
surface charge (Z > 0). Assuming Cl- as counterion, the
dissociation constant of lysozyme can be written as
CLys+10(CCl-)10
KLysCl10 ) (5)
CLysCl10
By the common ion effect, the solubility of AOT is
limited in an acidic solution or in a salt solution, which
is written as
KLH ) CH+CL- (6)
KLNa ) CNa+CL- (7)
where L- is the negatively charged group of AOT, or
ligand. The dissociated lysozyme reacts with L- forming
an insoluble lysozyme-ligand complex. The solubility
product of this complex, KLysL10, is
KLysL10 ) (CLys+10)(CL-)10 (8)
The solubility product of AOT, KLNa ) 1.1 10-3 (M2),
was obtained from the data reported by Caryl (43). The
solubility product of the lysozyme-ligand complex, KLysL10,
was determined experimentally to be 4 10-55 (M11). The
values of KLysCl10 and KLH were adjusted to fit the data.
Values of KLysCl10 ) 2.5 10-23 (M10), and KLH ) 9.2
10-9 (M2) gave a good fit to the data depicted in the
figures.
The values of KLysCl10 and KLysL10 are very small because
the concentration of one of the ions is raised to the power
10 in eqs 5 and 8. These values indicate that the lysozyme
is associated with chloride in solution unless the ligand,
AOT, is present. The solubility product of lysozyme-AOT
is in the order of 10-55 indicating that the complex is
virtually insoluble in water and precipitates upon forma-
tion.
The moles of lysozyme-ligand complex formed in the
aqueous phase, nLysL10, can then be calculated using the
mole balances on the lysozyme and the ligand written
as
934
nLysL10
C0Lys ) CLys+10 + CLysCl10 + (9)
V antibodies by nonstoichiometric polyelectrolyte complexes.
nLysL10
C0L ) CL- + 10 (10)
V Biochemistry 1997, 62, 5571-577.
where nLysL10 refers to the moles of the insoluble lysozyme-
ligand complex formed in the solution, V refers to the
total volume of the mixture, and CLysCl10 indicates the
concentration of lysozyme that is associated with the
chloride ions and not available to react with the ligand.
Finally, the percent removal of lysozyme as precipitated
complex is calculated as
nLysL10
% removal ) 100 (11)
C0LysV
Results of the Model. Figure 3 shows the percent
removal of lysozyme as the insoluble complex with AOT
for different mole ratios of AOT to lysozyme. Assuming
Z ) 10 and using the experimentally measured KLysL10
value, the model calculation of the percent mass removal
agrees well with the experimental results.
Figure 4 compares the model results with measure
values of the equilibrium pH of the aqueous solution after
the precipitation of the lysozyme-ligand complex. The
value of KLH was adjusted to fit these data.
Figure 5 shows the effect of sodium chloride concentra-
tion on the formation of the lysozyme-ligand complex
using the adjusted value of KLysCl10. The model agrees well
with the experimental results at salt concentrations up
to 0.3 M. For 1 M NaCl, the model predicts zero percent
removal because essentially all of the lysozyme and the
ligand remain associated with their counterions. The
experimental results show about 30-35% removal of
lysozyme, which is almost independent of the molar ratio,
R. This phenomenon is believed to be due to the salt
precipitation of lysozyme (55, 56).
Conclusions
A new method for lysozyme purification is proposed.
The method uses the commercially available anionic
surfactant, AOT, to precipitate lysozyme from an aqueous
solution and uses acetone to dissociate the complex and
recover the lysozyme. The direct precipitation method
requires significantly less AOT than the reverse micellar
extraction method to extract lysozyme from an aqueous
phase. This precipitation method has advantages over
the fractional sedimentation method, which requires long
processing time (many hours) and the gradient gels
operating under protein denaturing conditions (57). The
selectivity and the application of this method to the
precipitation of other proteins using different surfactants
are currently under investigation using an industrial
fermentation broth. Results presented in this work also
suggest that the solubilization of lysozyme into an AOT
reverse micellar phase consists of two steps: formation
of lysozyme-AOT complex to precipitate from the aque-
ous phase and the solubilization of lysozyme-AOT
complex into the reverse micellar phase.
Acknowledgment
The authors are grateful to the Natural Sciences and
Engineering Research Council of Canada for financial
support.
References and Notes
(1) Ladisch, M. In Bioseparations Engineering: Principles,
Practice, and Economics; Wiley-Interscience: New York,
2001.
Biotechnol. Prog., 2003, Vol. 19, No. 3
(2) Dainiak, M. B.; Izumrudov, V. A.; Muronetz, V. I.; Galaev,
I. Y.; Mattiasson, B. Affinity precipitation of monoclonal
Bioseparation 1999, 7, 213-240.
(3) Galaev, I. Y.; Mattiasson, B. New methods for affinity
purification of proteins. Affinity precipitation: a review.
(4) Lali, A.; Balan, S.; John, R.; DSouza, F. Carboxymethyl
cellulose as a new heterobifunctional ligand carrier for affinity
precipitation of proteins. Bioseparation 1999, 7, 195-205.
(5) Mattiasson, B.; Kumar, A.; Galaev, I. Y. Affinity precipita-
tion of proteins: design criteria for an efficient polymer. J.
Mol. Recognit. 1998, 11, 211-216.
(6) Vaidya, A. A.; Lele, B. S.; Deshmukh, M. V.; Kulkarni, M.
G. Design and evaluation of new ligands for lysozyme
recovery by affinity thermoprecipitation. Chem. Eng. Sci.
2001, 56, 5681-5692.
(7) Luisi, P. L.; Bonner, F. J.; Pellegrini, A.; Wiget, P.; Wolf, R.
Micellar solubilization of proteins in aprotic solvents and their
spectroscopic characterisation. Helv. Chim. Acta 1979, 62,
740-753.
(8) Marcozzi, G.; Domenico, C. D.; Spreti, N. Effects of surfac-
tants on the stabilization of the bovine lactoperoxidase
activity. Biotechnol. Prog. 1998, 14, 653-656.
(9) Naoe, K.; Nishino, M.; Ohsa, T.; Kawagoe, M.; Imai, M.
Protein extraction using sugar ester reverse micelles. J.
Chem. Technol. Biotechnol. 1999, 74, 221-226.
(10) Goklen, K. E.; Hatton, T. A. Protein extraction using
reverse micelles. Biotechnol. Prog. 1985, 1, 69-74.
(11) Dungan, S. R.; Bausch, T.; Hatton, T. A.; Plucinski, P.;
Nitsch, W. Interfacial transport processes in the reversed
micellar extraction of proteins. J. Colloid Interface Sci. 1991,
145, 33-50.
(12) Goklen, K. E.; Hatton, T. A. Liquid-liquid extraction of
low molecular-weight proteins by selective solubilization in
reversed micelles. Sep. Sci. Technol. 1987, 22, 831-841.
(13) Jarudilokkul, S.; Stuckey, D. C. Continuous forward and
back extraction of lysozyme from egg white using reverse
micelles. Sep. Sci. Technol. 2001, 36, 657-669.
(14) Paradkar, V. M.; Dordick, J. S. Purification of glycoproteins
by selective transport using concanavalin-mediated reverse
micellar extraction. Biotechnol. Prog. 1991, 7, 330-334.
(15) Shield, J. W.; Ferguson, H. D.; Bommarius, A. S.; Hatton,
T. A. Enzymes in reversed micelles as catalysts for organic-
phase synthesis reactions. Ind. Eng. Chem. Fundam. 1986,
25, 603-612.
(16) Wolbert, R. B. G.; Hilhorst, R.; Voskuilen, G.; Nachtegaal,
H.; Dekker, M.; Vant Riet, K.; Bijsterbosch, B. H. Protein
transfer from an aqueous phase into reversed micelles: The
effect of protein size and charge distribution. Eur. J. Biochem.
1989, 184, 627-633.
(17) Dekker: M.; Vant Riet, K.; Weijers, S. R.; Baltussen, J.
W. A.; Laane, C.; Bijsterbosch, B. H. Enzyme recovery by
liquid-liquid extraction using reversed micelles. Chem. Eng.
J. 1986, 33, B27-B33.
(18) Jolivalt, C.; Minier, M.; Renon, H. Extraction of R-chymot-
rypsin using reversed micelles. J. Colloid Interface Sci. 1990,
135, 85-96.
(19) Krei, G. A.; Hustedt, H. Extraction of enzymes by reverse
micelles. Chem. Eng. Sci. 1992, 47, 99-111.
(20) Lazarova, Z.; Tonova, K. Integrated reversed micellar
extraction and stripping of R-amylase. Biotechnol. Bioeng.
1999, 63, 583-592.
(21) Luisi, P. L.; Giomini, M.; Pileni, M. P.; Robinson, B. H.
Reverse micelles as hosts for proteins and small molecules.
Biochim. Biophys. Acta 1988, 947, 209-246.
(22) Rabie, H. R.; Suyyagh, T.; Vera, J. H. Reverse micellar
extraction of proteins using dioctyldimethylammonium chlo-
ride. Sep. Sci. Technol. 1998, 33, 241-257.
(23) Kelly, B. D.; Wang, D. I. C.; Hatton, T. A. Affinity-based
reversed micellar protein extraction: I. Principles and protein-
ligand systems. Biotechnol. Bioeng. 1993, 42, 1199-1208.
(24) Woll, J. M.; Hatton, T. A.; Yarmush, M. L. Bioaffinity
separations using reversed micellar extraction. Biotechnol.
Prog. 1989, 5, 57-62.
Biotechnol. Prog., 2003, Vol. 19, No. 3
(25) Naoe, K.; Kai, T.; Kawagoe, M.; Imai, M. Extraction of
flexibly structured protein in AOT reverse micelles: the
flexible structure of protein is the dominant factor for its
incorporation into reverse micelles. Biochem. Eng. J. 1999,
3, 79-85.
(26) Cassin, G.; Illy, S.; Pileni, M. P. Chemically modified
proteins solubilized in AOT reverse micelles: Influence of
proteins charges on intermicellar interactions. Chem. Phys.
Lett. 1994, 221, 205-212.
(27) Lye, G. J.; Asenjo, J. A.; Pyle, D. L. Protein extraction using
reverse micelles: Kinetics of protein partitioning. Chem. Eng.
Sci. 1994, 49, 3195-3204.
(28) Sun, Y.; Ichikawa, S.; Sugiura, S.; Furusaki, S. Affinity
extraction of proteins with a reversed micellar system com-
posed of cibacron blue-modified lecithin. Biotechnol. Bioeng.
1998, 58, 58-64.
(29) Rahaman, R. S.; Chee, J. Y.; Cabral, J. M. S.; Hatton, T.
A. Recovery of an extracellular alkaline protease from whole
fermentation broth using reversed micelles. Biotechnol. Prog.
1988, 4, 218-224.
(30) Shin, Y.-O.; Weber, M. E.; Vera, J. H. Comparison of two
methods to recover lysozyme from reverse micellar phases.
Sep. Sci. Technol. 2003, 38, 1733-1748.
(31) Aires-Barros, M. R.; Cabral, J. M. S. Selective separation
and purification of two lipases from Chromobacterium visco-
sum using AOT reversed micelles. Biotechnol. Bioeng. 1991,
38, 1302-1307.
(32) Chen, J. P.; Jen, J. T. Extraction of concanavalin A with
affinity reversed micellar systems. Sep. Sci. Technol. 1994,
29, 1115-1132.
(33) Hagen, A. J.; Hatton, T. A.; Wang, D. I. C. Protein refolding
in reversed micelles. Biotechnol. Bioeng. 1990, 35, 955-965.
(34) Jarudilokkul, S.; Poppenborg, L. H.; Stuckey, D. C. Back-
ward extraction of reverse micellar encapsulated proteins
using a counterionic surfactant. Biotechnol. Bioeng. 1999, 62,
593-601.
(35) Goto, M.; Hashimoto, Y.; Fujita, T.; Ono, T.; Furusaki, S.
Important parameters affecting efficiency of protein refolding
by reversed micelles. Biotechnol. Prog. 2000, 16, 1079-1085.
(36) Kabanov, A. V.; Nametkin, S. N.; Evtushenko, G. N.;
Chernov, N. N.; Klyachko, N. L.; Levashov, A. V.; Martinek,
K. A new strategy for the study of oligomeric enzymes:
-glutamyltransferase in reversed micelles of surfactants in
organic solvents. Biochim. Biophys. Acta 1989, 996, 147-152.
(37) Lye, G. J.; Asenjo, J. A.; Pyle, D. L. Extraction of lysozyme
and ribonuclease-a using reverse micelles: Limits to protein
solubilization. Biotechnol. Bioeng. 1995, 47, 509-519.
(38) Jauregi, P.; Varley, J. Colloidal gas aphrons: A novel
approach to protein recovery. Biotechnol. Bioeng. 1998, 59,
471-481.
(39) Paradkar, V. M.; Dordick, J. S. Mechanism of extraction
of chymotrypsin into isooctane at very low concentrations of
Aerosol OT in the absence of reverse micelles. Biotechnol.
Bioeng. 1994, 43, 529-540.
935
(40) Shin, Y.-O.; Vera, J. H. Solubilization limit of lysozyme in
DODMAC reverse micelles. Biotechnol. Bioeng. 2002, 80,
537-543.
(41) Adachi, M.; Harada, M. Solubilization mechanism of cy-
tocrome c in sodium bis(2-ethylhexyl) sulfosuccinate water/
oil microemulsion. J. Phys. Chem. 1993, 97, 3631-3640.
(42) Ono. T.; Goto, M.; Nakashio, F.; Hatton, T. A. Extraction
behavior of hemoglobin using reversed micelles by dioleyl
phosphoric acid. Biotechnol. Prog. 1996, 12, 793-800.
(43) Caryl, C. R. Sulfosuccinic esters: Structure and wetting
power. Ind. Eng. Chem. 1941, 33, 731-737.
(44) Linfield, W. M. In Anionic Surfactants Part II; Marcel
Dekker: New York, 1976; pp 405-443.
(45) Davies, R. C.; Neuberger, A.; Wilson, B. M. The dependence
of lysozyme activity on pH and ionic strength. Biochim.
Biophys. Acta 1969, 178, 294-305.
(46) Shin, Y.-O.; Rodil, E.; Vera, J. H. Selective precipitation
of lysozyme from egg white using AOT. J. Food. Sci. 2003,
68(2), 595-599.
(47) Alderton, G.; Ward, W. H.; Fevold, H. L. Isolation of
lysozyme from egg white. J. Biol. Chem. 1945, 157, 43-58.
(48) Sakakibara, R.; Hamaguchi, K. Structure of lysozyme. J.
Biochem. 1968, 64(5), 613-618.
(49) Barlow, D. J.; Thornton, J. M. The distribution of charged
groups in proteins. Biopolymers 1986, 25, 1717-1733.
(50) Kuramitsu, S.; Hamaguchi, K. Analysis of the acid-base
titration curve of hen lysozyme. J. Biochem. 1980, 87, 1215-
1219.
(51) Tanford, C.; Roxby, R. Interpretation of protein titration
curves. Application to lysozyme. Biochemistry 1972, 11,
2192-2198.
(52) Edsall, J. T.; Wyman, J. In Biophysical Chemistry; Aca-
demic Press: New York, 1958.
(53) Smith, R. M.; Martell, A. E. In Critical Stability Constants.
Volume 4: Inorganic Complexes; Plenum Press: New York,
1976.
(54) Noggle, J. H. In Physical Chemistry; HarperCollins College
Publishers: New York, 1996; p 407.
(55) Curtis, R. A.; Prausnitz, J. M.; Blanch, H. W. Protein-
protein and protein-salt interactions in aqueous protein
solutions containing concentrated electrolytes. Biotechnol.
Bioeng. 1998, 57, 11-21.
(56) Kuehner, D. E.; Blanch, H. W.; Prausnitz, J. M. Salt-
induced protein precipitation: Phase equilibria from an
equation of state. Fluid Phase Equilib. 1996, 116, 140-147.
(57) Arvanitis, D. N.; Yang, W.; Boggs, J. M. Myelin proteolipid
protein, basic protein, the small isoform of myelin-associated
glycoprotein, and p42MAPK are associated in the Triton
X-100 extract of central nervous system myelin. J. Neurosci.
Res. 2002, 70(1), 8-23.
Accepted for publication February 28, 2003.
BP025789R