2016_Binary Adsorption Processes of Albumin and Immunoglobulin on

Prévia do material em texto

Binary Adsorption Processes of Albumin and Immunoglobulin on
Hydrophobic Charge-Induction Resins
Qilei Zhang,† Ferdinand Schimpf,†,‡ Hui-Li Lu,† Dong-Qiang Lin,*,† and Shan-Jing Yao†
†Key Laboratory of Biomass Chemical Engineering of Ministry of Education, Department of Chemical and Biological Engineering,
Zhejiang University, Hangzhou 310027, China
‡Institute of Biochemical Engineering, Technische Universitaẗ München, 85748 Garching, Germany
ABSTRACT: Hydrophobic charge-induction chromatography with 4-
mercaptoethyl-pyridine as ligands shows promising application in
antibody purification. In this study, competitive adsorption of protein
mixtures composed with bovine serum albumin (BSA) and immunoglo-
bulin (IgG) was investigated with MEP HyperCel. Static adsorption and
dynamic binding processes were measured under different media pH and
BSA/IgG mass ratios. The results showed that MEP HyperCel had high
pH-dependent selectivity. BSA can be adsorbed quicker than IgG but part
of the adsorbed BSA would be gradually displaced by IgG as a result of
competitive adsorption. The effects of NaCl and (NH4)2SO4 on protein
mixture adsorption showed that both salts can enhance IgG selectivity on
MEP HyperCel, but the effect was different based on the combination of
electrostatic and hydrophobic interactions. Competitive adsorption
mechanism was discussed and the results obtained would be useful in the separation of albumin and immunoglobulin from
protein mixtures.
1. INTRODUCTION
Antibodies with important biological functions have shown
wide applications in diagnostic and therapeutic treatments.1
Currently, antibody drugs are one of the largest drug categories
in the pharmaceutical market and there are new antibody drugs
being approved by government bodies.2 With the development
of cell expression techniques, downstream processing such as
protein purification have become the major contributor in the
total cost of antibody production.3,4 Protein A-based affinity
chromatography is a standard technique in large-scale antibody
purification owing to its excellent selectivity.5 However, some
well-known limitations of Protein A, such as high cost, low
reusability, and ligand leaching, force the industry to explore
new resins and processes.6
Hydrophobic charge-induction chromatography (HCIC) is a
promising technique that has been successfully applied in
antibody purification.7,8 It is considered to be superficially
similar to Protein A chromatography that involves hydrophobic
interactions for adsorption and elution because of ionization of
histidine residues.9 HCIC can capture antibodies under
physiological conditions via hydrophobic interactions and
achieve effective elution through electrostatic repulsion. 4-
Mercaptoethyl-pyridine (MEP) is a typical HCIC ligand with a
pKa value of 4.8, and a resin using MEP as the ligand has been
developed which is the first commercialized HCIC resin MEP
HyperCel. It has been applied to purify monoclonal antibodies
from different resources and showed high dynamic binding
capacities.10
However, protein adsorption selectivity of MEP HyperCel
still needs to be improved as compared to Protein A affinity
chromatography. For example, albumin widely exists in blood
serum and cell culture media, and the presence of albumin can
significantly affect the adsorption process of immunoglobulin G
(IgG) with HCIC ligands.11,12
The effects of impurities or other proteins on target protein
purification have been studied with various techniques and
different resins. Cramer et al.13 investigated a selective
desorption process on ceramic hydroxyapatite for the
purification of monomeric antibody from associated aggregates
and post-Protein A impurities. A 100% yield of pure
monomeric antibody was achieved after mobile phase
optimization for selective desorption. Carta and Lewus14
developed an approximate rate equation to describe the
kinetics of multicomponent adsorption, which can be used in
the numerical simulation of adsorption systems with concen-
tration-dependent micropores. Martin et al.15 theoretically
analyzed multicomponent adsorption kinetics for protein
adsorption in porous ion exchangers, and they found the
experimental results agreed well with simulation. Confocal laser
scanning microscopy (CLSM) is a powerful technique for
visualizing protein distribution profiles in porous chromato-
graphic resins, which is useful in studying competitive
adsorption processes of different proteins in resins. Shi et
Received: December 31, 2015
Accepted: February 23, 2016
Published: February 29, 2016
Article
pubs.acs.org/jced
© 2016 American Chemical Society 1353 DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
pubs.acs.org/jced
http://dx.doi.org/10.1021/acs.jced.5b01108
Rubênia Silveira
Realce
al.16 used CLSM to study dynamic adsorption behaviors of two
intrinsic fluorescent proteins in the Q Sepharose HP resin. The
fluorescent images showed that the two proteins exhibited
distinct fading rates as compared to single component studies.
EI-Sayed and Chase17 also used CLSM to study competitive
adsorption of α-lactalbumin and β-lactoglobulin on SP
Sepharose FF, and the results showed that the protein
distribution in resins were different between a single-
component system and a two-component system, and β-
lactoglobulin was displaced by α-lactalbumin despite the lower
affinity of α-lactalbumin under the experimental conditions.
One of the problems of CLSM is it usually requires protein
labeling with a fluorescent probe, which may affect the affinity
of proteins on adsorbents.18
To make HCIC applicable for the separation of a variety of
protein candidates, traditional adsorption isotherms and kinetic
studies are still needed to help understand adsorption
processes, and more data should be collected and analyzed to
understand adsorption processes and mechanisms. In this
study, Human IgG and bovine serum albumin (BSA) were used
as the model target and impurity, and dynamic binding
processes of IgG/BSA mixtures on MEP HyperCel were
investigated under different media pH and protein mixture
mass ratios. The effect of salts on the competitive adsorption
process was also studied and the mechanism of competitive
adsorption was discussed.
2. MATERIALS AND METHODS
2.1. Materials. MEP HyperCel was purchased from Pall
Life Sciences (NY, USA). Bovine serum albumin (BSA) was
bought from Sigma (Milwaukee, USA). Human IgG was
obtained from Wako Pure Chemical Industries (Wako, Japan).
All other chemicals were of analytical grade.
2.2. Adsorption Isotherms. Static adsorption of IgG and
BSA on MEP HyperCel was studied by batch adsorption
equilibrium experiments. Citrate phosphate buffers with pH of
5−8 were used as the experimental media. For the experiments,
0.8 mL of protein solutions (single-component, or protein
mixtures with various BSA/IgG mass ratios) with protein
concentration ranging from 2−40 mg/mL was mixed with
about 40 mg of drained MEP HyperCel resin. The mixture was
then shaken in a thermoshaker at 25 °C for 3 h to achieve
adsorption equilibrium. The samples were then filtered with
0.22 μm membrane filtration and analyzed by high performance
liquid chromatography (HPLC). The amount of adsorbed
protein was calculated, and Langmuir adsorption eq (eq 1) was
used for data analysis.
* =
*
+ *
Q
Q C
K C
m
d (1)
where Q* is the equilibrium adsorption capacity (mg/g resin),
C* is the equilibrium protein concentration in the liquid (mg/
mL), Qm is the saturated adsorption capacity (mg/g resin), and
Kd is the apparent dissociation constant (mg/mL).
2.3. Adsorption Kinetics. Adsorption kinetic studies were
performed under similar conditions and parameters as those of
static adsorption. A series of 40 mg of resin and 0.8 mL of
protein solutions were mixed and shaken in a thermoshaker at
25 °C. After predefined time intervals the mixtures were filtered
to obtain liquid phases. These liquid samples were analyzed by
HPLC to acquire protein concentrations. The time course of
proteinconcentration in the liquid phase was determined.
2.4. Protein Concentration Analysis. HPLC analysis was
used to measure protein concentrations in liquids. It was
performed with an LC-3000 HPLC system (Beijing Chuang-
xintongheng Science & Technology Co., Ltd., Beijing, China)
using a TSK G3000SWXL column (7.8 mm × 30.0 mm, Tosoh
Bioscience, Tokyo, Japan). The mobile phase was 0.1 M
Na2SO4 in 0.1 M phosphate solution (pH 6.7). A sample
injection volume of 20 μL with a mobile phase flow rate of 0.5
mL/min was applied, and the detection wavelength was 280
nm. All samples were filtered through a 0.22 μm microporous
membrane in advance.
3. RESULTS AND DISCUSSION
3.1. Static Adsorption. MEP HyperCel is a typical HCIC
resin and the protein adsorption is usually pH-dependent.
Figure 1 shows the saturated adsorption capacities (Qm) of
individual BSA and IgG with MEP HyperCel under a pH of
5.0−8.0. The results show that IgG had the Qm above 160 mg/
mL when pH was around 6−8, while BSA showed a maximum
Qm of 65 mg/mL at pH 5, and the adsorption capacity
decreased dramatically with the increase of pH. BSA has a
molecular weight of ∼66 kDa, while the molecular weight of
IgG is ∼150 kDa. These values indicate that a similar mole of
individual proteins was adsorbed on the resin under their own
optimized pH conditions, which may be related to the ligand
density and distribution on resins. These results are consistent
with the research by Tong et al.19 who studied similar
adsorption processes in a column separation investigation.
Research has shown that the optimized pH values for protein
adsorption are usually around their individual isoelectric points
because of strong hydrophobic interactions between proteins
and HCIC ligands.20,21 However, with the addition of
impurities with competitive adsorption phenomenon, purifica-
tion conditions may be changed in order to achieve optimum
performance, and this optimum condition may change when
different impurities exist. To better understand the competitive
adsorption behaviors for these two proteins under different
conditions, it may be useful to study how pH, mass ratio, and
salt can affect the competitive adsorption behaviors when the
two proteins are loaded together. Meanwhile, the discrepancy
of both adsorption capacity and optimum pH between IgG and
Figure 1. Saturated adsorption capacities (Qm) of individual BSA
(white) and IgG (black) on MEP HyperCel under different pH.
Journal of Chemical & Engineering Data Article
DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
1354
http://dx.doi.org/10.1021/acs.jced.5b01108
Rubênia Silveira
Realce
Rubênia Silveira
Realce
Rubênia Silveira
Realce
BSA may result in high capacity and adsorption selectivity of
IgG to BSA.
Figure 2 shows the adsorption isotherms of protein mixtures
under different mass ratios and pH. The presence of another
protein clearly affects the static adsorption processes of IgG/
BSA, and Figure 2 indicates that this competitive adsorption
behavior is pH, protein concentration, and mass-ratio depend-
ent. Figure 2 panels a and b show that BSA and IgG had similar
adsorption capacity individually at pH 5. The adsorption
capacity of both proteins declined with the addition of another
component. For protein concentrations less than ∼2 mg/mL,
the adsorption capacity of BSA under different IgG/BSA mass
ratios was comparable to that of the pure BSA profile.
Meanwhile, the adsorption capacity of IgG under same
condition showed over 25% decrease as compared to that of
pure IgG data. This may because the ligands were not fully
occupied by these proteins when the protein concentration was
less than ∼2 mg/mL, and pH 5 was a good adsorption
condition for BSA. However, with the increase of IgG
concentration, the BSA adsorption capacity decreased gradu-
ally. For example, the BSA adsorption capacity of the 1:1
(BSA/IgG) mixture was only one-third of that of pure BSA
(pH 5). With the increase of pH from 5 to 6, the adsorption of
IgG increased dramatically, and it shows similar static
adsorption from pH 6 to 8. Moreover, the increase of BSA in
the protein mixture resulted in the decrease of IgG adsorption.
However, at pH 8, the maximum adsorption capacity of BSA
was lower than 20 mg/mL in protein mixtures, and the IgG
adsorption seems not affected by the addition of BSA (Figure
2g,h).
Figure 2 indicates that these proteins may compete for
binding sites on resins, and the increase of one component can
lead to the decrease of adsorption capacity of another. It is
possible that based on the Vroman effect,22 BSA with smaller
molecular size and higher mobility and concentration would
reach and adsorb onto binding sites faster than IgG, which will
later be replaced by larger proteins with higher affinity, that is,
IgG.23 Therefore, adsorption kinetics was used to reveal the
competitive binding process.
3.2. Adsorption Kinetics. Figure 3a shows the adsorption
kinetics of BSA under conditions of BSA/IgG = 4:1 and pH 5−
8. These kinetic profiles clearly show the dynamic adsorption
process of BSA on MEP HyperCel. The results show that
Figure 2. Adsorption isotherms of binary BSA and IgG under different
pH and mass ratios: (a) BSA, pH 5; (b) IgG, pH 5; (c) BSA, pH 6;
(d) IgG, pH 6; (e) BSA, pH 7; (f) IgG, pH 7; (g) BSA, pH 8; (h) IgG,
pH 8. Mass ratios: (■) single-component; (●) BSA/IgG = 4:1; (○)
BSA/IgG = 3:1; (△) BSA/IgG = 2:1; (×) BSA/IgG = 1:1.
Figure 3. Adsorption kinetics of BSA and IgG under different pH: (a)
BSA, (b) IgG adsorption kinetic profiles; BSA/IgG = 4:1; initial
protein concentration 10 mg/mL. (■) pH 5; (□) pH 6; (△) pH 7;
(×) pH 8. (c) IgG fractions on resin: (q[IgG]/(q[IgG] + q[BSA])). Dotted
data were measured and solid lines were fitting curves following the
modified Linear Driving Force (LDF) model.
Journal of Chemical & Engineering Data Article
DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
1355
http://dx.doi.org/10.1021/acs.jced.5b01108
adsorption was the main process in the first few minutes for all
pH conditions studied, and desorption gradually became the
dominant process later on and last until the end of the
experiments. Moreover, the maximum capacity appeared earlier
with the increase of pH. For example, the curve of pH 8 started
to decline after only 15 s, and there was no BSA detectable in
the end of the experiment.
The adsorption of protein in resin is a complex mass
transport process and it can be affected by physical and
chemical properties of resins, media, and proteins. A series of
models have been proposed to study and predict protein
adsorption kinetics in resins, such as pore diffusion and solid
diffusion models, but for general curve fitting applications these
models may show similar results.24 The Linear Driving Force
(LDF) model is a simple empirical model (eq 2)
∂
∂
= −
q t
t
k q q t
( )
( ( ))max (2)
which can be represented by the following equation by solving
mass transfer differential equation:
= −q t q kt( ) (1 exp( ))max (3)
where q(t) is the adsorption capacity at time t, qmax is the
maximum adsorption capacity, and k is a constant. Good
agreement between the LDF model and more rigorous
diffusion models can be achieved when k = 15D/r2 (D is
diffusivity and r is the radius of resins). This empirical model
can greatly simplify the mathematical complexity and also is
adequate for many practical applications.24
To use eq 3 to predict BSA adsorption behaviors, two
adjustments are needed. (1) A time correction term tc is needed
to compensate the time required for sample preparation before
measurements; (2) the desorption process is also need to be
included for BSA. The modified equation is as follows
= − ′ − − ′q t q k t q k t( ) (1 exp( ) (1 exp( )ads ads des des (4)
where qads is the adsorption capacity, qdes is the desorption
capacity, kads is the adsorption parameter, and kdes is the
desorption parameter. t′ = t − tc is the adjusted time. tc is a
displacement factor which practically indicates the timerequired for the sample preparation and filtration. For
experimental data obtained at longer times, this problem can
be neglected. Meanwhile, a desorption term is applied because
of the competitive adsorption between these two proteins. It
was found that the amount of BSA adsorbed on the resin
decreased after certain adsorption times. Therefore, the classical
adsorption model is not suitable for fitting such data. Moreover,
this phenomenon is likely indicating the existence of dynamic
adsorption−desorption processes during the competitive
adsorption.
Figure 3b shows the adsorption kinetics of IgG, which has a
gradual increase in the profiles, and the results show that pH
did not have a significant effect on the adsorption of IgG.
Equation 4 can also be used to fit the adsorption kinetics of
IgG; however, as IgG did not show an obvious desorption
process, the desorption part of eq 4 was neglected. The fitting
curves in Figure 3a,b show that this modified LDF model can
well fit the experimental data. The related fitting parameters are
listed in Table 1.
By comparing BSA and IgG adsorption kinetics shown in
Figure 3a,b, the IgG fractions on resin (q[IgG]/(q[IgG] + q[BSA])),
which indicates the changes of IgG percentage on the resin, can
be calculated as a function of time. Figure 3c indicates that pH
is an important factor determining IgG fraction on resin, and it
may take over 15 min for IgG to reach its maximum adsorption
on the resin. A column separation study is a useful verification
approach to further confirm the data profiles shown in this
study.19 The separation study indicates that pH 7−8 combined
with small flow rates or sufficient column length is a good
separation condition to purify IgG from BSA/IgG mixtures.
The adsorption kinetics was also studied under pH 5 where
BSA and IgG had similar static adsorption capacities, with
different mass−ratio protein mixtures. Similar to the results
shown in Figure 3a, those in Figure 4a show that BSA can be
quickly adsorbed under all mass ratio conditions, and within
less than 5 min the maximum capacity was reached. The curves
then show a decline of BSA adsorption capacity, which implies
a competitive adsorption between BSA and IgG. Figure 4b
shows that the IgG adsorption process was relatively slow
compared to that of BSA, and the adsorption capacity is related
to the mass percentage of IgG in the protein mixtures. The
results of “IgG fraction on resin” in the figure give direct
information about the final proportion of the two proteins
bound on the resin. Figure 4c indicates that the maximum
fraction of IgG adsorbed on the resin was less than 0.7 under
pH 5. Therefore, Figure 3c and Figure 4c indicate that pH is an
important factor that can dramatically affect the mass fraction of
IgG adsorbed on the resin during competition adsorption.
3.3. Salt Addition. NaCl and (NH4)2SO4 are two widely
used salts for protein separation and purification applications.
These salts can precipitate proteins under certain concentration
conditions. Meanwhile, the addition of these salts may affect
the competitive adsorption process between BSA and IgG.
Therefore, the precipitation of the two proteins was studied
under experimental conditions in advance of adsorption studies,
and no precipitation was found under the salt and protein
concentrations studied. Figure 5 and Figure 6 show the
adsorption isotherms of BSA and IgG with the addition of
varied amounts of NaCl or (NH4)2SO4. The results show that
the BSA adsorption capacity decreased dramatically when NaCl
concentration increased (Figure 5a), while the adsorption
capacity of IgG gradually increased (Figure 5b). The adsorption
of BSA was almost not detectable with the addition of 0.75 or
1.0 M NaCl. Therefore, the IgG fraction on the resin can reach
close to 1 when 1.0 M NaCl is added, which is significantly
higher than that without NaCl under same conditions (∼0.4,
see Figure 4c).
However, Figure 6 shows slightly different results when
(NH4)2SO4 was added. Although IgG showed the same trend
with the increase of (NH4)2SO4 concentration, the BSA
adsorption capacity showed a “U” shape change. The BSA
adsorption capacity decreased after adding 0.25 M (NH4)2SO4;
however, it increased with the further addition of (NH4)2SO4.
IgG has a much larger molecular weight than BSA. Research
has shown that electrostatic interactions play a dominant role in
Table 1. Kinetic Fitting Parameters of BSA and IgG at
Different pH
BSA IgG
pH qads kads qdes kdes qads kads
5 85.51 0.85 54.62 0.12 22.70 0.11
6 66.30 0.31 57.68 0.06 30.96 0.11
7 28.23 0.13 26.85 0.13 28.11 0.17
8 31.22 0.07 30.12 0.13 25.84 0.27
Journal of Chemical & Engineering Data Article
DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
1356
http://dx.doi.org/10.1021/acs.jced.5b01108
the BSA adsorption processes on HCIC resins. Meanwhile,
hydrophobic interactions are the main adsorption mechanism
for IgG binding onto MEP HyperCel.10,19,25 NaCl is a neutral
salt which can behave as an electrostatic interaction screener
but may not change the hydrophobic property of proteins
within certain concentrations. Therefore, the addition of NaCl
impedes the electrostatic interactions between BSA and the
ligands, which results in the decrease of BSA adsorption
capacity. However, the interaction between IgG and ligands was
not much affected. The increase of adsorption capacity shown
in Figure 5b was probably due to desorption of BSA that left
more binding sites available for IgG, which is also a result of
competitive adsorption.
Figure 4. Adsorption kinetics of BSA and IgG under different BSA/
IgG mass ratios: (a) BSA, (b) IgG adsorption kinetic profiles; pH 5,
initial protein concentration 10 mg/mL. (○) single-component; (■)
BSA/IgG = 4:1; (□) BSA/IgG = 3:1; (△) BSA/IgG = 2:1; (×) BSA/
IgG = 1:1. (c) IgG fractions on resin: (q[IgG]/(q[IgG] + q[BSA])). Dotted
data were measured and solid lines were fitting curves following the
modified Linear Driving Force (LDF) model.
Figure 5. Adsorption isotherms of BSA and IgG under different NaCl
concentrations at pH 5. (a) BSA, (b) IgG; BSA/IgG = 4:1. (○) 0 M;
(■) 0.25 M; (□) 0.5 M; (Δ) 0.75 M; (×) 1.0 M. Dotted data were
measured and solid lines were fitting curves following the Langmuir
model.
Figure 6. Adsorption isotherms of BSA and IgG under different
(NH4)2SO4 concentrations at pH 5. (a) BSA, (b) IgG; BSA/IgG = 4:1.
(○) 0 M; (■) 0.25 M; (□) 0.5 M; (△) 0.75 M; (×) 1.0 M. Dotted
data were measured and solid lines were fitting curves following the
Langmuir model.
Journal of Chemical & Engineering Data Article
DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
1357
http://dx.doi.org/10.1021/acs.jced.5b01108
(NH4)2SO4 is a kosmotropic salt which can improve
hydrophobic interactions between proteins and ligands, and it
has been widely used in protein precipitation and hydrophobic
interaction chromatography. Therefore, the addition of
(NH4)2SO4 could facilitate the binding of IgG onto HCIC
ligands. This is probably one reason for the increase of IgG
adsorption as shown in Figure 6b. The adsorption of BSA on
the ligands was likely due to the results of both electrostatic and
hydrophobic interactions depending on (NH4)2SO4 concen-
tration. When (NH4)2SO4 concentration was relatively low
(∼0.25 M), electrostatic interactions still dominated, and
(NH4)2SO4 has similar effect as NaCl, which leads to the
decrease of BSA adsorption. With the increase of (NH4)2SO4
concentration, hydrophobic interactions gradually turn into
action, which helps the adsorption of BSA on the resin.
However, the maximum adsorption capacity of BSA after
adding (NH4)2SO4 was lower than that without (NH4)2SO4. A
detailed mechanism may need further studies using other
techniques, such as Surface Plasmon Resonance. In summary,
the addition of salts can improve the purification of IgG from
BSA/IgG mixtures.
Figure 7 shows that the adsorption kinetics of BSA and IgG
with the addition of salts had similar profiles of thatwithout
salts (Figures 3 and 4). BSA showed maximum adsorption in
less than 5 min, and then the desorption process overtook the
adsorption process (Figure 7a). The addition of (NH4)2SO4 led
to the decline of BSA adsorption, although 0.25 M (NH4)2SO4
showed lower adsorption capacity than that of 1.0 M. Figure 7c
shows that the addition of salts resulted in an improvement of
IgG fraction on resin under the experimental conditions, and
the results showed that the addition of 0.25 M NaCl or
(NH4)2SO4 had similar enhancing effects, which is better than
1.0 M (NH4)2SO4 or no salt conditions.
3.4. Discussion. Competitive adsorption in protein
purification has been studied using different proteins mixtures
in the literature,26,27 and techniques such as confocal laser
scanning microscopy are used as a powerful imaging tool to
qualitatively study competitive adsorption processes.28 Mean-
while, competitive adsorption or displacement of proteins on
different material surfaces has been widely investigated since
the 1960s, which is now commonly referred to as the Vroman
effect.29 This effect has been discussed on solid surfaces with
different charge properties30 and different protein pairs,23,31 and
the mechanism has been discussed on a molecular level.32
However, this phenomenon is still not well understood, and
there are typically three possible mechanisms to explain the
competition process.29
Figure 8a shows a desorption−adsorption mode, in which
BSA desorbs first and then IgG adsorbs onto the ligand. In the
competitive binding process (Figure 8b), BSA is displaced by
neighboring IgG attached on ligands. Figure 8c shows the third
mode where a transient complex between BSA and IgG is first
formed and then BSA gradually moves away from the ligand. In
the end the ligands were occupied by IgG. It is difficult to
confirm which mode actually happened in this study. The
competitive exchange mode can be easily achieved on material
surfaces, but it may require high ligand density of resins in
chromatography. The transient complex formation process may
need a longer period for IgG to be absorbed onto the ligands.
However, the kinetic profiles of BSA adsorption show that it
can usually happened within 5 min. Therefore, the desorption−
adsorption mode may be the most possible mechanism of the
BSA/IgG mixture adsorption processes.
4. CONCLUSIONS
A commercial HCIC resin MEP HyperCel was used to
investigate competitive adsorption processes of BSA and IgG in
order to better understand the adsorption behaviors and
optimize IgG separation efficiency from protein mixtures with
HCIC resins. The adsorption isotherms and dynamic binding
processes were determined under different media pH and mass
ratio of BSA/IgG mixtures. The results showed that pH was an
important factor on determining the protein adsorption
capacity of the resin, and the increase of one component in
the mixture would result in the decrease of adsorption capacity
of another component, as an indication of competitive
adsorption. Kinetic results showed that BSA can be adsorbed
on the ligands quicker than IgG but part of the adsorbed BSA
Figure 7. Adsorption kinetics of BSA and IgG under different salt
concentrations at pH 5. (a) BSA, (b) IgG; BSA/IgG = 4:1; (○) 0 M;
(■) 0.25 M (NH4)2SO4; (×) 1 M (NH4)2SO4; (▽) 0.25 M NaCl. (c)
IgG fractions on resin: (q[IgG]/(q[IgG] + q[BSA])). Dotted data were
measured and solid lines were fitting curves following the modified
Linear Driving Force (LDF) model.
Journal of Chemical & Engineering Data Article
DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
1358
http://dx.doi.org/10.1021/acs.jced.5b01108
would be gradually displaced by IgG as a result of competitive
adsorption. A modified Linear Driving Force model was used,
which showed good fitting results of the adsorption kinetic
profiles. The effect of salts on the adsorption process was also
studied, and the results showed that both NaCl and (NH4)2SO4
can enhance the adsorption selectivity of IgG on MEP
HyperCel, although the effects and mechanism of these two
salts were different based on the electrostatic and hydrophobic
interactions. In addition, the desorption−adsorption mode may
be suitable to explain the competitive adsorption process of IgG
and BSA.
■ AUTHOR INFORMATION
Corresponding Author
*E-mail: lindq@zju.edu.cn.
Funding
This work was supported by the National Natural Science
Foundation of China and the Zhejiang Provincial Natural
Science Foundation of China.
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
Mr. Ferdinand Schimpf was a master student from Technische
Universitaẗ München.
■ REFERENCES
(1) Menachem, A.; Chapman, J.; Deri, Y.; Pick, C.; Katzav, A.
Immunoglobulin-Mediated Neuro-Cognitive Impairment: New Data
and a Comprehensive Review. Clin. Rev. Allergy Immunol. 2013, 45,
248−255.
(2) Reichert, J. M. Antibodies to watch in 2014: Mid-year update.
mAbs 2014, 6, 799−802.
(3) Li, F.; Vijayasankaran, N.; Shen, A.; Kiss, R.; Amanullah, A. Cell
culture processes for monoclonal antibody production. mAbs 2010, 2,
466−479.
(4) Low, D.; O’Leary, R.; Pujar, N. S. Future of antibody purification.
J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2007, 848, 48−63.
(5) Hober, S.; Nord, K.; Linhult, M. Protein A chromatography for
antibody purification. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci.
2007, 848, 40−47.
(6) Li, R.; Dowd, V.; Stewart, D. J.; Burton, S. J.; Lowe, C. R. Design,
synthesis, and application of a protein a mimetic. Nat. Biotechnol. 1998,
16, 190−195.
(7) Bak, H.; Thomas, O. R. T. Evaluation of commercial
chromatographic adsorbents for the direct capture of polyclonal rabbit
antibodies from clarified antiserum. J. Chromatogr. B: Anal. Technol.
Biomed. Life Sci. 2007, 848, 116−130.
(8) Tong, H.-F.; Lin, D.-Q.; Gao, D.; Yuan, X.-M.; Yao, S.-J.
Caprylate as the albumin-selective modifier to improve IgG
purification with hydrophobic charge-induction chromatography. J.
Chromatogr. A 2013, 1285, 88−96.
(9) Ghose, S.; Hubbard, B.; Cramer, S. M. Protein interactions in
hydrophobic charge induction chromatography (HCIC). Biotechnol.
Prog. 2005, 21, 498−508.
(10) Arakawa, T.; Kita, Y.; Sato, H.; Ejima, D. MEP chromatography
of antibody and Fc-fusion protein using aqueous arginine solution.
Protein Expression Purif. 2009, 63, 158−163.
(11) Guerrier, L.; Girot, P.; Schwartz, W.; Boschetti, E. New method
for the selective capture of antibodies under physiolgical conditions.
Bioseparation 2000, 9, 211−221.
(12) Guerrier, L.; Flayeux, I.; Boschetti, E. A dual-mode approach to
the selective separation of antibodies and their fragments. J.
Chromatogr., Biomed. Appl. 2001, 755, 37−46.
(13) Morrison, C. J.; Gagnon, P.; Cramer, S. M. Purification of
monomeric mAb from associated aggregates using selective desorption
chromatography in hydroxyapatite systems. Biotechnol. Bioeng. 2011,
108, 813−821.
(14) Carta, G.; Lewus, R. Film Model Approximation for
Multicomponent Adsorption. Adsorption 2000, 6, 5−13.
(15) Martin, C.; Iberer, G.; Ubiera, A.; Carta, G. Two-component
protein adsorption kinetics in porous ion exchange media. J.
Chromatogr. A 2005, 1079, 105−115.
(16) Shi, Q.-H.; Shi, Z.-C.; Sun, Y. Dynamic behavior of binary
component ion-exchange displacement chromatography of proteins
visualized by confocal laser scanning microscopy. J. Chromatogr. A
2012, 1257, 48−57.
(17) El-Sayed, M. M. H.; Chase, H. A. Confocal microscopy study of
uptake kinetics of α-lactalbumin and β-lactoglobulin onto the cation-
exchanger SP Sepharose FF. J. Sep. Sci. 2009, 32, 3246−3256.
(18) Teske, C. A.; Von Lieres, E.; Schröder, M.; Ladiwala, A.;
Cramer, S. M.; Hubbuch, J. J. Competitive adsorption of labeled and
native protein in confocal laser scanning microscopy. Biotechnol.
Bioeng. 2006, 95, 58−66.
(19) Tong, H.-F.; Lin, D.-Q.; Yuan, X.-M.; Yao, S.-J. Enhancing IgG
purification from serum albumin containing feedstock with hydro-
Figure 8. A schematic diagram illustrating three competitive adsorption processes:(a) desorption/adsorption processes; (b) competitive binding
processes; (c) displacement through the formation of transient complexes. Adapted with permission from ref 29. Copyright 2012 Elsevier.
Journal of Chemical & Engineering Data Article
DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
1359
mailto:lindq@zju.edu.cn
http://dx.doi.org/10.1021/acs.jced.5b01108
phobic charge-induction chromatography. J. Chromatogr. A 2012,
1244, 116−122.
(20) Zhao, G.; Peng, G.; Li, F.; Shi, Q.; Sun, Y. 5-Aminoindole, a new
ligand for hydrophobic charge induction chromatography. J.
Chromatogr. A 2008, 1211, 90−98.
(21) Lin, D.-Q.; Tong, H.-F.; Wang, H.-Y.; Yao, S.-J. Molecular
Insight into the Ligand−IgG Interactions for 4-Mercaptoethyl-pyridine
Based Hydrophobic Charge-Induction Chromatography. J. Phys. Chem.
B 2012, 116, 1393−1400.
(22) Vroman, L.; Adams, A.; Fischer, G.; Munoz, P. Interaction of
high molecular weight kininogen, factor XII, and fibrinogen in plasma
at interfaces. Blood 1980, 55, 156−159.
(23) Holmberg, M.; Stibius, K.; Larsen, N.; Hou, X. Competitive
protein adsorption to polymer surfaces from human serum. J. Mater.
Sci.: Mater. Med. 2008, 19, 2179−2185.
(24) Carta, G.; Jungbauer, A., Adsorption Kinetics. In Protein
Chromatography; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim,
Germany, 2010; pp 161−199.
(25) Arakawa, T.; Futatsumori-Sugai, M.; Tsumoto, K.; Kita, Y.; Sato,
H.; Ejima, D. MEP HyperCel chromatography II: Binding, washing
and elution. Protein Expression Purif. 2010, 71, 168−173.
(26) Xu, X.; Lenhoff, A. M. Binary adsorption of globular proteins on
ion-exchange media. J. Chromatogr. A 2009, 1216, 6177−6195.
(27) Traylor, S. J.; Xu, X.; Lenhoff, A. M. Shrinking-core modeling of
binary chromatographic breakthrough. J. Chromatogr. A 2011, 1218,
2222−2231.
(28) Ljunglöf, A.; Hjorth, R. Confocal microscopy as a tool for
studying protein adsorption to chromatographic matrices. J.
Chromatogr. A 1996, 743, 75−83.
(29) Hirsh, S. L.; McKenzie, D. R.; Nosworthy, N. J.; Denman, J. A.;
Sezerman, O. U.; Bilek, M. M. M. The Vroman effect: Competitive
protein exchange with dynamic multilayer protein aggregates. Colloids
Surf., B 2013, 103, 395−404.
(30) Lassen, B.; Malmsten, M. Competitive Protein Adsorption at
Plasma Polymer Surfaces. J. Colloid Interface Sci. 1997, 186, 9−16.
(31) Huetz, P.; Ball, V.; Voegel, J. C.; Schaaf, P. Exchange Kinetics
for a Heterogeneous Protein System on a Solid Surface. Langmuir
1995, 11, 3145−3152.
(32) Jung, S.-Y.; Lim, S.-M.; Albertorio, F.; Kim, G.; Gurau, M. C.;
Yang, R. D.; Holden, M. A.; Cremer, P. S. The Vroman Effect: A
Molecular Level Description of Fibrinogen Displacement. J. Am. Chem.
Soc. 2003, 125, 12782−12786.
Journal of Chemical & Engineering Data Article
DOI: 10.1021/acs.jced.5b01108
J. Chem. Eng. Data 2016, 61, 1353−1360
1360
http://dx.doi.org/10.1021/acs.jced.5b01108