Colloids and Surfaces B: Biointerfaces 158 (2017) 562568

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage: www.elsevier.com/locate/colsurfb

Protocols

Role of foam drainage in producing protein aggregates in foam

fractionation
Rui Li a,b,,1 , Yuran Zhang a,1 , Yunkang Chang a , Zhaoliang Wu b, , Yanji Wang b ,
Xiange Chen a , Tao Wang a
a
School of Biological Science, Jining Medical University, No. 669 Xueyuan Road, Donggang District, Rizhao, Shandong Province, 276800, China
b
School of Chemical Engineering and Technology, Hebei University of Technology, No. 8 Guangrong Road, Dingzi Gu, Hongqiao District, Tianjin, 300130,
China

a r t i c l e i n f o a b s t r a c t

Article history: It is essential to obtain a clear understanding of the foam-induced protein aggregation to reduce the loss
Received 23 March 2017 of protein functionality in foam fractionation. The major effort of this work is to explore the roles of foam
Received in revised form 11 July 2017 drainage in protein aggregation in the entire process of foam fractionation with bovine serum albumin
Accepted 16 July 2017
(BSA) as a model protein. The results show that enhancing foam drainage increased the desorption of
BSA molecules from the gas-liquid interface and the local concentration of desorbed molecules in foam.
Keywords:
Therefore, it intensied the aggregation of BSA in foam fractionation. Simultaneously, it also accelerated
Protein aggregation
the ow of BSA aggregates from rising foam into the residual solution along with the drained liquid.
Insoluble aggregates
Foam drainage
Because enhancing foam drainage increased the relative content of BSA molecules adsorbed at the gas-
Foam fractionation liquid interface, it also intensied the aggregation of BSA during both the defoaming process and the
BSA storage of the foamate. Furthermore, enhancing foam drainage more readily resulted in the formation of
insoluble BSA aggregates. The results are highly important for a better understanding of foam-induced
protein aggregation in foam fractionation.
2017 Elsevier B.V. All rights reserved.

1. Introduction adsorption-desorption process, so many aggregates will be pro-

Aggregation is an important issue in various pharmaceutical
protein processes, such as renaturation, purication, shipping, and
storage, because it often reduces protein efcacy and even causes
harmful immune responses [1,2]. As a promising alternative for
industrial protein separation, foam fractionation is hampered by
protein aggregation. Protein molecules suffer structural unfold-
ing that makes their hydrophobic groups exposed towards the gas
phase as they are adsorbed at the gas-liquid interface [3]. Owing
to bubble coalescence and defoaming, the adsorbed molecules
are desorbed from the interface and, then, some of them read-
ily aggregate via exposed hydrophobic groups [46]. In foam
fractionation, most of the protein molecules must tolerate the
Corresponding author.
Corresponding author at: School of Biological Science, Jining Medical Univer-
sity, No. 669 Xueyuan Road, Donggang District, Rizhao, Shandong Province, 276800,
China.
E-mail addresses: ruili061289@163.com (R. Li), zhaoliangwu hebut@163.com
(Z. Wu).
1
The authors contributed equally to the work.
duced and, hence, the separation performance will be poor [7]. At
present, many studies on foam fractionation of proteins have been
reported [8,9], but few have paid attention to protein aggregation.
The key to reducing the loss of protein efcacy in foam frac-
tionation is to have a clear understanding of foam-induced protein
aggregation. Several researchers have recently investigated the
mechanisms of protein denaturation induced by foam [3,1012].
However, their efforts were aimed at analyzing the relationship
between the activity loss and the foam-induced structural change
of a protein, but not at studying the foam-induced protein aggre-
gation. Maa and Hsu [4] have conrmed that the presence of the
air-liquid interface and shear force was able to enhance protein
aggregation. Furthermore, Wiesbauer et al. [13] reported that the
change of the air-water interfacial area played a critical role in
protein aggregation, conrming the results of Bee et al. [14]. This
conclusion is highly important for fundamental understanding of
foam-induced protein aggregation. However, researchers have not
yet considered the role of foam drainage, an intrinsic phenomenon
of foam. Hence, further research is required to fully understand
foam-induced protein aggregation.
Foam drainage is a critical step for protein enrichment in foam
fractionation [15]. As foam drainage is enhanced, the amount of

http://dx.doi.org/10.1016/j.colsurfb.2017.07.040
0927-7765/ 2017 Elsevier B.V. All rights reserved.
 R. Li et al. / Colloids and Surfaces B: Biointerfaces 158 (2017) 562568 563

protein molecules adsorbed at the gas-liquid interface becomes Table 1

Specic composition of the feed solution of BSA.
much higher than that of protein molecules in the interstitial liq-
uid between bubbles. The increase in the relative content of the Monomer Dimer Trimer
adsorbed protein molecules in the foam may then intensify protein Molecular weight (kDa) 66.4 132.8 0.2 201.0 5.2
aggregation due to their unfolded structures [6]. Foam drainage Relative content (%) 91.5 0.7a 8.0 0.1b 0.5 0.1c
also enhances bubble coarsening and coalescence to decrease a, b and c were used to characterize the different levels in relative contents of
the gas-liquid interfacial area [16]. With the increase in bubble monomer, dimer and trimer.
size, the protein surface excess and the proteinprotein interac-
tions at the gas-liquid interface may increase to enhance protein
a steady state. The foamate and the residual solution were then
aggregation [10,14]. Thus, foam drainage has signicant effects on
sampled to obtain the required data.
foam-induced protein aggregation.
To better understand protein aggregation in foam fractionation,
2.3. Measurements of the adsorption of BSA at the gas-liquid
the role of foam drainage in the foam-induced protein aggregation
interface and bubble radius in the rising foam
was an object of the present study. Bovine serum albumin (BSA) is
a typical globular protein that has many applications in drug deliv-
The adsorption of BSA at the gas-liquid interface was character-
ery, biochemical studies, and cell culturing [1719]. Thus, BSA was
ized by three parameters: surface excess ( BSA ), and the relative
chosen as a model process. In this work, the different levels of foam
content () and mass ux (Qm ) of BSA molecules adsorbed at the
drainage were obtained by changing the foam height in the foam
gas-liquid interface in a rising foam.  BSA and Qm were calculated
fractionation column. The effects of foam drainage on the follow-
by Eqs. (1) and (2), respectively, and were measured using the
ing four areas of interest were then studied: (1) adsorption of BSA
methods detailed by Li et al. [6];  is dened as Eq. (3):
at the gas-liquid interface and bubble size in the rising foam, (2)
production rates and partition coefcients of BSA aggregates, (3) (Cf Cb ) r32 Qf
BSA = (1)
mass uxes and relative contents in the foamate of BSA aggregates, 3Qg
and (4) variations of the relative contents in the foamate of BSA
Qm = Qf (Cf Cb ), (2)
aggregates with time.
3Qg
r32 BSA Cb
= =1 , (3)
Cf Qf Cf
2. Materials and methods
where Qg and Qf are volumetric air ow rate and volumetric
2.1. Reagents foamate ow rate, respectively, and Cf and Cb are the total BSA con-
centrations in the foamate and in the entrained liquid in the rising
Bovine serum albumin (BSA, purity > 99%) was purchased from foam, respectively. Cb is considered to equal the total BSA concen-
Tianjin Unite Stars Biotechnology Co. Ltd., China. It was dissolved tration in the bulk solution in the foam fractionation column. r32 is
in a concentration of 0.30 g/L in a disodium hydrogen phosphate the Sauter mean bubble radius and is calculated with Eq. (4) [6]:
(0.165 M)citric acid (0.018 M) buffer solution with pH 7.0. The spe-

n
cic composition of the BSA solution is presented in Table 1. Tween di
3
20 (analytical grade) was purchased from Tianjin Yingdaxigui 1 i=1
Co. Ltd., China. Ultrapure water (electrical resistance 18.25 M) r32 = , (4)
2 n
prepared in a UPR-II-10T water purication system (Chengdu Ultra- di
2
pure Technology Co. Ltd., China) was used in all the experiments.
i=1

where di is the bubble diameter and n the number of bubbles,

2.2. Equipment
A schematic of the continuous foam fractionation of BSA
followed by defoaming is presented in Fig. 1. The foam fraction-
ation column was constructed with two transparent polymethyl
methacrylate tubes of 50 mm i. d. The heights of the two tubes
were Hl = 600 mm and Hf . The values of Hf were 200, 400, 600,
800, and 1000 mm. The tubes were used to obtain different foam
drainage levels. A gas distributor with an average pore diameter
of 250 20 m was installed at the bottom of the foam fraction-
ation column, through which air was pumped into the column
to generate bubbles. The BSA feed solution was injected near the
foam-liquid interface by a peristaltic pump (YW03, Changzhou
Yuanwang Fluid Co. Ltd., China). The foam breaker was made
of a transparent, 500-mm-high polymethyl methacrylate tube of
50 mm i. d. and four 450-mm-high, 15-mm-diam cylinders com-
posed of synthetic sponge. The cylinders were xed inside the tube
by stainless-steel wires. The foam fractionation experiments were
carried out at a volumetric feed ow rate of 10 mL/min, a volumetric
air ow rate of 200 mL/min and at room temperature, 25.0 1.0 C.
The foam owed out of the foam fractionation column into the foam
breaker, and then was completely collapsed to obtain the foamate.
When the concentration and composition of BSA in the foamate
became constant, the entire system was considered to have reached
n > 300. Note that the photographs for the bubble radius measure-
ment were taken near the top of the foam fractionation column.
2.4. Evaluation of production of BSA aggregates in foam
fractionation
The production rate (ragg ), distribution coefcient (Kagg ), and
mass ux (Qaggf ) and relative content ( agg ) in the foamate of BSA
aggregates were used to evaluate the production of BSA aggregates
in foam fractionation. Specically, ragg was used to characterize the
mass of BSA aggregates produced per unit of time in the entire foam
fractionation process. Kagg was used to characterize the distribution
of BSA aggregates in the foamate and in the residual solution. Qaggf
was used to characterize the mass of BSA aggregates obtained in
the foamate per unit of time.  agg was used to characterize the rel-
ative content of BSA aggregates in all the recovered BSA molecules
in the foamate. The four parameters are dened as Eqs. (5)(8),
respectively:
ragg(i) = Qf Cf (i) + Qr Cr(i) Qo Co(i) (5)
Qf Cf(i)
Kagg(i) = (6)
Qr Cr(i)
Qaggf(i) = Qf Cf(i) (7)
564 R. Li et al. / Colloids and Surfaces B: Biointerfaces 158 (2017) 562568
Fig. 1. Schematic diagram of continuous foam fractionation of BSA followed by defoaming.
Cf(i)
agg (i) =
Cf
where Q is volumetric ow rate; C is concentration; the subscripts
o, f, and r denote the feed solution, the foamate, and the residual
solution, respectively; and the subscript i denotes total aggregates,
oligomers, soluble polymers, or insoluble aggregates.
In the experiments, many different BSA aggregates were
detected and some of them could not be well isolated by size exclu-
sion chromatography due to their similar particle sizes. To facilitate
the analysis, all the BSA aggregates were divided into three types,
being oligomers, soluble polymers, and insoluble aggregates. The
specic morphologies, sizes, and molecular weights of the three
types of BSA aggregates are presented in Table 2. From the table,
it can be seen that oligomers and soluble polymers were spherical
while insoluble aggregates were irregularly shaped.
The concentration of each BSA aggregate and the total BSA
concentration were measured by Kjeldahl determination and size
exclusion chromatography, the specic methods of which were
described by Li et al. [6].
2.5. Measurement of volumetric liquid ow rate
The volumetric ow rates of the foamate and the residual solu-
tion were calculated by Eq. (9).
1 dG
Q =
 dt
where G is the weight of the liquid that is collected at an interval
of t = 5 min, and  is the density of the collected liquid, equal to
1000 kg/m3 .
2.6. Characterizations of BSA aggregates
A Zetasizer Nano ZS90 analyzer (Malvern Instruments, UK) and
a scanning electron microscope (SEM, S-4800, Seron Technologies,
Inc., Korea) were used to characterize the size and morphology of
the BSA aggregates, respectively.
2.7. Statistical analysis
(8)
Each of the experiments was repeated at least three times. An
analysis of the variance of the data was performed using IBM SPSS
Statistics 19.0 software (IBM, USA). The t test with P 0.05 was
used to determine the difference between mean values. Standard
deviation was provided for each mean value.
3. Results and discussion
3.1. Effects of foam height on volumetric foamate ow rate,
bubble radius, and adsorption of BSA at the gas-liquid interface in
the rising foam
To clearly explain the role of foam drainage in the aggregation
of BSA in foam fractionation, the effects of foam height (Hf ) on vol-
umetric foamate ow rate (Qf ), bubble radius (r32 ), and adsorption
of BSA at the gas-liquid interface were investigated. The latter was
characterized by measuring the surface excess ( BSA ), and relative
content () and mass ux (Qm ) of BSA adsorbed at the gas-liquid
interface in the rising foam. The results are presented in Table 3.
Increasing Hf enhanced foam drainage and intensied bubble
coarsening and coalescence [20]. Thus, in Table 3, Qf decreased
while r32 increased as Hf increased from 200 to 1000 mm. Both the
decrease and increase is signicant (P 0.05). The increased bubble
radius reduced the total area of the gas-liquid interface in the ris-
ing foam, so that increasing Hf decreased Qm signicantly (P 0.05).
Correspondingly, some BSA molecules had to be desorbed from the
gas-liquid interface to return the entrained liquid between bub-
(9)
bles. The BSA concentration in the entrained liquid then increased
to enlarge its difference from the BSA concentration at the gas-
liquid interface. As a result, BSA slightly increased with increasing
Hf . Although Qm decreased, the mass ux of BSA molecules in the
entrained liquid decreased to a larger extent due to the enhanced
foam drainage, and also signicantly increased with increasing
Hf (P 0.05).
3.2. Effects of foam height on production rates and distribution
coefcients of BSA aggregates
In this subsection, we report the results of our investigation of
the effects of foam height (Hf ) on production rates (ragg ) and dis-
 R. Li et al. / Colloids and Surfaces B: Biointerfaces 158 (2017) 562568 565

Table 2
Morphologies, sizes and molecular weights of oligomers, soluble polymers and insoluble aggregates.

Aggregates Oligomers Soluble polymers Insoluble aggregates

Morphology

Size (nm) 1790 90160 >400

Molecular weight (kDa) 132.8664 6641062.4

Table 3
Effects of foam height (Hf ) on volumetric foamate owrate (Qf ), BSA surface excess ( BSA ), bubble radius (r32 ), and the relative content () and the mass ux (Qm ) of BSA
adsorbed at the gas-liquid interface in foam.

Hf (mm) Qf (mL/min)  BSA (kg/m2 ) r32 (mm)  (%) Qm (kg/h)

200 5.6 0.6a 1.13 0.12 106a 0.38 0.04a 51.6 5.2a 10.8 1.1 105a
400 2.3 0.3b 1.25 0.13 106a 0.51 0.05b 68.8 6.9b 9.1 0.9 105a
600 1.3 0.1c 1.43 0.15 106b 0.64 0.06c 77.2 7.3c 8.1 0.8 105b
800 0.6 0.1d 1.45 0.15 106b 0.85 0.08d 84.9 8.2c 6.2 0.7 105c
1000 0.3 0.1e 1.46 0.15 106b 1.12 0.11e 88.9 9.0c 4.7 0.5 105d

a, b, c, d, e were used to characterize the different levels of each parameter.

Fig. 2. Effects of foam height on production rates (A) and partition coefcients (B) of total aggregates, oligomers, soluble polymers and insoluble aggregates.

tribution coefcients (Kagg ) of total aggregates, oligomers, soluble
polymers, and insoluble aggregates. In this work, the total aggre-
gates encompassed the sum of the oligomers, soluble polymers,
and insoluble aggregates. Our aim is to provide a clear understand-
ing of how foam drainage affects the aggregation of BSA in foam
fractionation. The results are presented in Fig. 2.
According to the work of Wiesbauer et al. [13], the pro-
duction rate of protein aggregates was essentially determined
by the change in the size of the gas-liquid interfacial area. In
foam fractionation, at each Hf , the total gas-liquid interfacial area
initially generated by the gas distributor was the same, and,
nally, all the generated interfacial area decreased to zero in the
defoaming process. Then, the change of the gas-liquid interfa-
cial area was the same, so that ragg(totalaggregates) should exhibit
no signicant changes with increasing Hf . In fact, Fig. 2A shows
that ragg(totalaggregates) signicantly increased from (1.4 0.1) 105
to (3.8 0.4) 105 kg/h as Hf increased from 200 to 800 mm
(P 0.05). Specically, the great increase in ragg(totalaggregates) was
mainly attributed to the signicant increases in ragg(solublepolymers)
and ragg(insolubleaggregates) (P 0.05). The change of ragg(oligomers)
was insignicant (P > 0.05), so its contribution to the increase of
ragg(totalaggregates) was slight. The results indicate that the enhanced
foam drainage intensied the aggregation of BSA in foam fraction-
ation, particularly the formation of soluble polymers and insoluble
aggregates. From Fig. 2B, the partition coefcients of all the BSA
aggregates signicantly decreased to lower than 1 with increas-
ing Hf (P 0.05). It is suggested that enhancing the foam drainage
intensied the ow of the BSA aggregates into the residual solution
with the drained liquid.
According to the work of Bee et al. [14], the decrease in the
gas-liquid interfacial area could cause the aggregation of protein
molecules at the gas-liquid interface and the aggregates were read-
ily desorbed from the interface. From Table 3, the higher level of
foam drainage corresponded to the larger decrease in the gas-liquid
interfacial area, so more BSA aggregates, particularly soluble poly-
mers and insoluble aggregates, could be formed. Furthermore, the
decrease in Qm should correspond to the desorption of BSA aggre-
gates from the gas-liquid interface. In addition, increasing the BSA
concentration in an aqueous solution often enhanced the aggrega-
tion of BSA [21]. By this analogy, increasing the BSA surface excess
could more readily intensify the protein aggregation at the gas-
liquid interface, because the adsorbed protein molecules possessed
more unfolded structures than those in the bulk solution. Thus, the
increased BSA due to foam drainage also enhanced the aggrega-
tion of BSA at the gas-liquid interface. The enhanced foam drainage
also improved the aggregation of BSA in defoaming, the reasons
for which are explained in the following subsection. As a result,
ragg(totalaggregates) increased with increasing Hf .
A higher level of foam drainage corresponded to a larger
decrease in the gas-liquid interfacial area and a larger number of
BSA aggregates formed in the rising foam. These aggregates were
readily desorbed from the gas-liquid interface and then owed into
566 R. Li et al. / Colloids and Surfaces B: Biointerfaces 158 (2017) 562568
Fig. 3. Effects of foam height on relative contents of total aggregates in the foa-
mate before and after the foam breaker. The foamate before the foam breaker was
obtained by collapsing the foam via stirring with a glass rod and the addition of
Tween 20 and had a molar ratio of Tween 20 and BSA of 20:1.
the bulk solution along with the drained liquid. Thus, more BSA
aggregates tended to be detected in the residual solution as foam
drainage was enhanced. As a result, the partition coefcients of all
the BSA aggregates decreased with increasing Hf .
3.3. Effects of foam height on relative contents of BSA aggregates
in the foamate
To obtain the protein-enriched liquid from the collected foam,
defoaming is an indispensable unit operation following foam frac-
tionation [22]. However, in defoaming, the sharp decrease in the
gas-liquid interfacial area certainly enhances protein aggregation
[13]. Thus, to determine if foam drainage affected the aggregation
of BSA in defoaming, we compared the relative content of total BSA
aggregates ( agg(totalaggregates) ) in the foamates taken just after the
foam fractionation (and before the foam breaker) with that after
the foam breaker. Note that when obtaining the foamate before
the foam breaker, we collapsed the foam just out of the column
by stirring with a glass rod and the addition of Tween 20, a widely
used antifoaming agent which was also able to prevent the pro-
tein aggregation induced by the gas-liquid interface [2325]. In
the experiments, we found that  agg(totalaggregates) had no signicant
changes with increasing the molar ratio of Tween 20 and BSA in the
foamate when the molar ratio was more than 20:1. So it is at the
molar ratio of 20:1 that the values of  agg(totalaggregates) before and
after the foam breaker were compared. The results are presented
in Fig. 3.
Fig. 4. Effects of foam height on mass uxes and relative contents of total aggregates, oligomers, soluble polymers and insoluble aggregates in the foamate.
In Fig. 3,  agg(totalaggregates) before the foam breaker had a slight
decrease (P > 0.05) and that after the foam breaker signicantly
increased (P 0.05), with increasing Hf from 200 mm to 1000 mm.
More importantly, the difference in  agg(totalaggregates) before and
after the foam breaker signicantly increased (P 0.05). In the
defoaming process. the aggregation of BSA at the gas-liquid inter-
face could not be effectively prevented by Tween 20, so the
results in Fig. 3 indicate that most of BSA aggregates were formed
in their desorption from the interface. The slight decrease of
 agg(totalaggregates) before the foam breaker was because more BSA
aggregates formed in the rising foam owed into the bulk solution
with the drained liquid with increasing Hf .
Based on the above results, we further studied the effects of
foam drainage on the aggregation of BSA in the defoaming process
by analyzing the effects of foam height on the mass uxes (Qaggf )
and the relative contents ( agg ) of total BSA aggregates, oligomers,
soluble polymers, and insoluble aggregates in the foamate obtained
after the foam breaker were investigated. The results are presented
in Fig. 4.
From Fig. 4A, Qagg f (totalaggregates) exhibited no signicant
changes with increasing Hf from 200 to 1000 mm (P > 0.05). Accord-
ing to the results of Wiesbauer et al. [13], the smaller gas-liquid
interfacial area corresponded to a lower protein aggregation rate.
In this case, increasing Hf reduced the gas-liquid interfacial area
so that Qaggf(totalaggregates) should have decreased. This conjecture
was in disagreement with the current result. It is indicated that
the change of the gas-liquid interfacial area was not the only factor
that enhanced the interface-induced protein aggregation. It is well
known that enhancing foam drainage could improve the protein
concentration in the foamate [26]. Table 3 shows that the rela-
tive content of BSA molecules adsorbed at the gas-liquid interface
in the rising foam () increased as foam drainage was enhanced.
Thus, the concentration of BSA molecules adsorbed at the gas-
liquid interface was higher at a higher level of foam drainage. The
higher concentration of the adsorbed BSA molecules corresponded
to a higher possibility for BSA molecules to aggregate during the
defoaming process. The increased aggregation of BSA due to the
increased concentration of the adsorbed BSA molecules made up
for the decrease due to the reduced gas-liquid interfacial area.
As a result, Qaggf(totalaggregates) was not signicantly affected by Hf
(P > 0.05). Fig. 4A also shows that Qaggf(oligomers) exhibits a sharp
decrease with increasing Hf (P 0.05), while Qaggf(solublepolymers)
and Qaggf(insolubleaggregates) exhibit a slighter increase (P 0.05). This
indicates that enhancing foam drainage tended to induce the for-
mation of BSA aggregates of larger sizes. The result was attributed
to the fact that the enhanced foam drainage increased the content of
the adsorbed BSA molecules in the rising foam, so that their number
for forming an aggregate increased during the defoaming process.
 R. Li et al. / Colloids and Surfaces B: Biointerfaces 158 (2017) 562568
Fig. 5. Dependency of relative content of total aggregates in the foamate on relative
content of BSA molecules adsorbed at the gas-liquid interface in foam.
Although Qaggf(totalaggregates) did not signicantly change,
agg(totalaggregates) signicantly increased as Hf increased (P 0.05).
Based on the previous work of Li et al. [6],  agg(totalaggregates)
was not essentially determined by the total BSA concentration
in the foamate, but by , because the higher  corresponded
to the higher possibility for aggregation of the adsorbed BSA
molecules with unfolded structures in the defoaming process.
As a result, the enhanced foam drainage increased , so that
 agg(totalaggregates) increased with increasing Hf . For clearly elaborat-
ing how  agg(totalaggregates) depended on , their corresponding data
were analyzed using different models. The results show that the
dependency of agg(totalaggregates) on  was well tted by the Gaus-
sian curve model, as presented in Fig. 5. The gure illustrates that
 agg(totalaggregates) increased very slowly in a low- range from 0.5 to
0.7, and its increase became much larger as  increased to a value
greater than 0.7. Furthermore,  agg(totalaggregates) did not increase to
1.0 as  approached 1.0, because some adsorbed BSA molecules pos-
sibly returned into their native structures in the defoaming process,
so that they did not readily aggregate. In addition, Fig. 4B shows
Fig. 6. Effects of foam height (Hf ) on the variations of relative contents of total aggregates (A), oligomers (B), soluble polymers (C) and insoluble aggregates (D) in the foamate
with time.
567
that  agg(solublepolymers) and  agg(insolubleaggregates) also signicantly
increased with increasing Hf (P 0.05) while  agg(oligomers) exhibited
no signicant changes (P > 0.05). The results also conrmed that the
enhanced foam drainage induced the formation of BSA aggregates
with larger sizes in the defoaming process.
3.4. Effects of foam height on the variations of relative contents of
BSA aggregates in the foamate with time
In a large-scale foam fractionation process, the foamate often
had to remain in the collector for a long time to obtain a cer-
tain volume before the subsequent treatment [27]. During this
period, the desorbed protein molecules that were unable to return
their native structures could further aggregate in the foamate. We
investigated the effects of foam height (Hf ) on variations of the
relative contents of total aggregates, oligomers, soluble polymers,
and insoluble aggregates in the foamate with time at room tem-
perature (25.0 1.0 C). Our aim was to determine if the extent of
foam drainage led to further protein aggregation in the foamate.
The results are presented in Fig. 6.
From Fig. 6A,  agg(totalaggregates) at each Hf linearly increased with
time, and its increase, i.e., the slope of the tted linear, increased sig-
nicantly (P 0.05) with increasing Hf . By comparing the results in
Fig. 6BD, the signicant increase in  agg(totalaggregates) with time was
mainly attributed to the signicant increase in  agg(insolubleaggregates)
(P 0.05). The results indicate that BSA could further aggregate in
the foamate and that most of the formed aggregates were insoluble.
More importantly, the enhanced foam drainage improved the gen-
eration rate of insoluble aggregates. Furthermore, at a low level of
foam drainage (Hf = 200400 mm), the insoluble aggregates were
directly formed by the BSA monomers, while at a high level of
foam drainage (Hf = 4001000 mm) the insoluble aggregates were
formed by the soluble polymers and the BSA monomers. Thus, the
aggregation of BSA in the foamate obeyed the multistage aggre-
gation model. In addition,  agg(oligomers) exhibited no signicant
changes with time (P > 0.05).
As the unfolded BSA molecules were desorbed from the gas-
liquid interface, some of them immediately aggregated in the
defoaming process, but some of them that did not aggregate were
568 R. Li et al. / Colloids and Surfaces B: Biointerfaces 158 (2017) 562568
present in the foamate. Owing to their structural instability, these
unfolded molecules could further aggregate in the foamate. By com-
paring the data of in Table 3 with those of  agg(totalaggregates) in
Fig. 4B, the desorbed BSA molecules that did not aggregate in the
defoaming process had a high concentration and a high relative
content in the foamate when the extent of foam drainage was high.
Thus, enhancing foam drainage could intensify the aggregation of
BSA in the foamate, particularly the formation of insoluble aggre-
gates.
4. Conclusions
By decreasing the gas-liquid interfacial area, enhancing foam
drainage intensied the foam-induced aggregation of BSA and
increased the aggregation rate of BSA. Because most aggregates
were readily desorbed from the gas-liquid interface, they owed
into the bulk solution with the drained liquid and were detected
in the residual solution. As a result, enhancing foam drainage
decreased the partition coefcients of BSA aggregates. Enhancing
foam drainage increased the concentration and the relative content
of BSA molecules adsorbed at the gas-liquid interface in the rising
foam, so it intensied the aggregation of BSA during the defoaming
process and the storage of the foamate. Furthermore, enhancing
foam drainage readily resulted in the formation of insoluble BSA
aggregates.
Acknowledgements
This work was nancially supported by the Natural Science
Foundations of China (Grant Nos. 21346008 and 21236001), the
Graduate Student Innovation Foundation of Hebei, China (Grant
No.220056) and the Doctoral Scientic Research Foundation of Jin-
ing Medical University (Grant Nos. 600493001, JY2015BS12 and
JY2015BS11).
References
[1] M.E. Cromwell, E. Hilario, F. Jacobson, Protein aggregation and bioprocessing,
AAPS J. 8 (2006) E572E579.
[2] A.J. Chirino, M.L. Ary, S.A. Marshall, Minimizing the immunogenicity of
protein therapeutics, Drug Discov. Today 9 (2004) 8290.
[3] I. Barackov, A. Mause, S. Kapoor, R. Winter, G. Schembecker, B. Burghoff,
Investigation of structural changes of -casein and lysozyme at the gas-liquid
interface during foam fractionation, J. Biotechnol. 161 (2012) 138146.
[4] Y.F. Maa, C.C. Hsu, Protein denaturation by combined effect of shear and
air-liquid interface, Biotechnol. Bioeng. 54 (1997) 503512.
[5] W. Wang, Protein aggregation and its inhibition in biopharmaceutics, Int. J.
Pharm. 289 (2005) 130.
[6] R. Li, N. Fu, Z. Wu, Y. Wang, Y. Wang, Protein aggregation in foam
fractionation of bovine serum albumin: effect of protein concentration,
Biochem. Eng. J. 103 (2015) 234241.
[7] R. Li, Z. Wu, Y. Wang, L. Ding, W. Liu, Prevention of irreversible aggregation of
whey soy proteins in their foam fractionation from soy whey wastewater,
Asia Pac. J. Chem. Eng. 11 (2016) 673682.
[8] B. Burghoff, Foam fractionation applications, J. Biotechnol. 161 (2012)
126137.
[9] A.G. Wouters, I. Rombouts, N. Schoebrechts, E. Fierens, K. Brijs, C. Blecker, J.A.
Delcour, Foam fractionation as a tool to study the air-water interface
structure-function relationship of wheat gluten hydrolysates, Colloids Surf. B
151 (2017) 295303.
[10] J.R. Clarkson, Z.F. Cui, R.C. Darton, Protein denaturation in foam: I. Mechanism
study, J. Colloid Interface Sci. 215 (1999) 323332.
[11] J.R. Clarkson, Z.F. Cui, R.C. Darton, Protein denaturation in foam: II. Surface
activity and conformational change, J. Colloid Interface Sci. 215 (1999)
333338.
[12] J.R. Clarkson, Z.F. Cui, R.C. Darton, Effect of solution conditions on protein
damage in foam, Biochem. Eng. J. 4 (2000) 107114.
[13] J. Wiesbauer, R. Prassl, B. Nidetzky, Renewal of the air-water interface as a
critical system parameter of protein stability: aggregation of the human
growth hormone and its prevention by surface-active compounds, Langmuir
29 (2013) 1524015250.
[14] J.S. Bee, D.K. Schwartz, S. Trabelsi, E. Freund, J.L. Stevenson, J.F. Carpenter, T.W.
Randolph, Production of particles of therapeutic proteins at the air-water
interface during compression/dilation cycles, Soft Matter 8 (2012)
1032910335.
[15] Y. Wang, Z. Wu, R. Li, Z. Zhang, Enhancing foam drainage using inclined foam
channels of different angles for recovering the protein from whey
wastewater, Colloids Surf. A 419 (2013) 2836.
[16] S.A. Magrabi, B.Z. Dlugogorski, G.J. Jameson, Bubble size distribution and
coarsening of aqueous foams, Chem. Eng. Sci. 54 (1999) 40074022.
[17] M.M. Rahimnejad, M. Jahanshahi, G.D. Najafpour, Production of biological
nanoparticles from bovine serum albumin for drug delivery, Afr. J. Biotechnol.
5 (2006) 19181923.
[18] C.A. Kreader, Relief of amplication inhibition in PCR with bovine serum
albumin or T4 gene 32 protein, Appl. Environ. Microbiol. 62 (1996)
11021106.
[19] M.T. Kane, Variability in different lots of commercial bovine serum albumin
affects cell multiplication and hatching of rabbit blastocysts in culture, J.
Reprod. Fertile 69 (1983) 555558.
[20] P. Stevenson, Hydrodynamic theory of rising foam, Miner. Eng. 20 (2007)
282289.
[21] A. Bhattacharya, R. Prajapati, S. Chatterjee, T.K. Mukherjee,
Concentration-dependent reversible self-oligomerization of serum albumins
through intermolecular -sheet formation, Langmuir 30 (2014)
1489414904.
[22] P.R. Garrett, Defoaming Antifoams and mechanical methods, Curr. Opin.
Colloid Interface Sci. 20 (2015) 8191.
[23] D.K. Chou, R. Krishnamurthy, T.W. Randolph, J.F. Carpenter, M.C. Manning,
Effects of Tween 20 and Tween 80 on the stability of Albutropin during
agitation, J. Pharm. Sci. 94 (2005) 13681381.
[24] L. Krielgaard, L.S. Jones, T.W. Randolph, S. Frokjaer, J.M. Flink, M.C. Manning,
J.F. Carpenter, Effect of tween 20 on freeze-thawing-and agitation-induced
aggregation of recombinant human factor XIII, J. Pharm. Sci. 87 (1998)
15931603.
[25] R. Li, Z. Wu, Y. Wang, L. Ding, W. Liu, Prevention of irreversible aggregation of
whey soy proteins in their foam fractionation from soy whey wastewater,
Asia Pac. J. Chem. Eng. 11 (2016) 673682.
[26] S. Bhattacharjee, R. Kumar, K.S. Gandhi, Prediction of separation factor in
foam separation of proteins, Chem. Eng. Sci. 52 (1997) 46254636.
[27] R. Li, Z. Wu, Y. Wang, W. Liu, Pilot study of recovery of whey soy proteins from
soy whey wastewater using batch foam fractionation, J. Food Eng. 142 (2014)
201209.