You are on page 1of 13

Journal of Chromatography A, 1244 (2012) 1–13

Contents lists available at SciVerse ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Review

Aqueous two-phase affinity partitioning systems: Current applications


and trends夽
Federico Ruiz-Ruiz, Jorge Benavides, Oscar Aguilar, Marco Rito-Palomares ∗
Centro de Biotecnología FEMSA, Departamento de Biotecnología e Ingeniería de Alimentos, Tecnológico de Monterrey, Campus Monterrey, Ave. Eugenio Garza Sada 2501 Sur,
Monterrey, NL 64849, Mexico

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

Article history: Aqueous two-phase systems (ATPS) have been studied and used for product recovery and purification
Received 15 December 2011 from diverse biological sources. ATPS are characterized by their versatility, easy scale up parameters,
Received in revised form 16 April 2012 process integration capability and relative low cost. This technique is commonly regarded as a primary
Accepted 25 April 2012
recovery stage mainly due to its low selectivity. However, the use of strategies involving the modifica-
Available online 7 May 2012
tion of ATPS with affinity ligands have resulted in significant increases in recovery yields and purification
folds of biological products. The aim of this review is to highlight current applications, trends and chal-
Keywords:
lenges regarding affinity partitioning in aqueous two-phase systems for the fractionation, recovery and
Aqueous two-phase systems
Affinity partitioning
purification of biological products.
Affinity ligand © 2012 Elsevier B.V. All rights reserved.
Aqueous two-phase affinity partitioning

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. Common chemical modification methodologies for polymer activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3. Polymer–polymer affinity aqueous two-phase systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4. Polymer–salt affinity aqueous two-phase systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
5. Affinity enhanced ATPAP systems using alternative phase components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
6. Trends and challenges of ATPAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

the total bioprocesses costs [5,6]. Additionally, most of the unit


1. Introduction
operations comprising a biotechnology process are focused on
the recovery and purification of the product of interest. In this
There is a considerable interest in the development of efficient
context, the development of strategies for the optimization and
downstream processes for the recovery and purification of biolog-
design of efficient and economic downstream processes represents
ical products. The need to establish selective and scalable methods
a research area of great interest.
for product recovery has resulted in the implementation of diverse
Integration and intensification of unit operations are two
bioengineering strategies in the field. Nowadays, improvements
common strategies used for the optimization of downstream pro-
in liquid–liquid extraction systems, filtration devices and chro-
cessing. Process integration attempts to combine two or more unit
matography strategies allow the efficient recovery of natural dyes,
operations in one, reducing the total number of stages and increas-
antibodies, phytochemicals, proteins, virus like particles and other
ing recovery yields [2,7]. On the other hand, process intensification
products [1–4]. Purification strategies account for up to 80% of
is focused on maximizing the throughput in a given unit operation
by optimization of process parameters [2,7–9].
Aqueous two phase extraction is a liquid–liquid fractionation
夽 Presented at the 16th International Conference on BioPartitioning and Purifica-
technique that has been extensively used for the recovery and
tion, Puerto Vallarta, Jalisco, Mexico, 18–22 September 2011.
∗ Corresponding author. Tel.: +52 81 832800; fax: +52 818328 4136. purification of biological products. This technique involves the con-
E-mail address: mrito@itesm.mx (M. Rito-Palomares). struction of extraction systems that can be formed by two polymers,

0021-9673/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.chroma.2012.04.077
2 F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13

a polymer and a salt, an ionic liquid and a salt, or a low molecular polymers. These factors must be taken into consideration in order
weight alcohol and a salt mixed over a limit concentration thus to optimize ATPAP partition protocols since affinity interactions of
forming two immiscible phases [10,11]. Additionally, micellar and products and affinity ligands are highly dependent of the ionic state
reverse micelle aqueous two-phase systems (ATPS) can be formed of the active groups (according to pH level) and the ionic strength of
using solutions with ionic and/or non-ionic surfactants. Partition- the system (depending on salt concentration) [3,5,15]. In addition,
ing of biomolecules in ATPS depends on the choice of constituents ligand concentration and modified polymer concentration could
of the system, the concentration of the system constituents, the also prove to increase recovery and purification yields in ATPAP sys-
volume ratio (VR ), the pH and temperature of the system [4,12]. tems. Nevertheless, appropriate levels should be studied since high
Furthermore, the physicochemical properties (molecular weight, ligand concentration could introduce shielding effects of ligands
superficial electrochemical charge, hydrophobicity, etc.) of the toward target molecules and high modified polymer concentration
solutes being partitioned on the systems also play a major role can generate steric hindrance problems between target molecules
on their partition behavior [4]. In general, ATPS have proven to be and active receptor sites. These possible scenarios emphasize the
suitable for the recovery of biological products from different ori- importance of the additional priorities that must be studied in
gins [4,12,13]. However, one major drawback concerning the use ATPAP systems, in order to design efficient and novel purification
of ATPS is their lack of specificity. Although the manipulation of the strategies.
system parameters may favor the selective partitioning of the com- In addition to the traditional and empirical character of ATPS
pound of interest towards a particular phase, specificity usually is design, new technical issues have to be considered within ATPAP.
not high enough to achieve complete separation between product At a first glance, partitioning of modified polymers with cou-
and contaminants. pled ligands or free ligands has to be addressed. First reports
Aqueous two phase affinity partitioning (ATPAP) consists of the employing ATPAP described that phase forming curves were not
chemical modification of at least one phase forming component of greatly affected when modified polymers with coupled ligands
the system by attaching an affinity ligand with specificity for the were present at low concentrations [15]. However, a recent inves-
molecule of interest [14]. Therefore, selective interaction between tigation indicates that by increasing concentrations of a modified
specific ligands in one of the polymeric phases of the system and the polymer with a coupled ligand shifts the hydrophobic nature of the
protein, cellular fragment or biomolecule of interest followed by polymer, due to the nature of the ligands. This shift may affect not
their partition towards the modified polymer rich phase is achieved only the partition behavior of the solutes within the system but also
[14]. In addition, affinity partitioning in liquid–liquid extraction the phase formation behavior [18]. Another characteristic that may
systems can also be achieved by the addition of free ligands (bio- be affected for the use of modified polymers is the ionic nature of
logical or chemical molecules) in solution to a traditional ATPS in the polymer–ligand complex [18]. Since affinity coupling involves
order to induce a partition shift of the desired target molecule very specific molecular interactions between a target molecule and
without the chemical modification of the phase forming compo- a specific ligand, the development of an ATPAP strategy must con-
nent [3]. ATPAP was discussed by Flanagan and Barondes [15] in sider the biological and chemical nature of the proposed affinity
1975 when the selective partition of S-23 myleoma proteins in a ligand and target molecule. For example, affinity partitioning of
modified dextran–polyethylene oxide (PEO) system was studied. biological membranes has been achieved with ATPAP strategies in
The findings stated that the inclusion of polymer derivatives or which wheat germ agglutinin (WGA) has been coupled to phase
modified components to ATPS poses an advantage in downstream forming polymers. Due to the interaction between WGA with spe-
processing of high value products. In addition, the authors con- cific N-acetylglucosamine and sialic acid residues in the membrane
cluded that even though a marked increase in partition of the target surface, partitioning was favored to a specific system phase [17]. In
products using liquid–liquid affinity partitioning was observed, the this tenor, the chemical nature of both target products and ligands
use of well-established chromatographic techniques may be con- must be understood as well as the degree of saturation (DS) of the
sidered, if needed, for the subsequent purification of the product polymer (defined as the ratio of polymer/ligand molecules) in order
of interest. ATPAP provides a high compatibility system between to maximize system efficiency [17]. In addition, polymer molecular
bioaffinity ligands and specific molecules, which can be manipu- weight (MW) has an impact in product partitioning since volume
lated in order to develop more predictable purification strategies. displacements are seen with high MW polymers [2,4,15]. In the
Therefore, it is possible to anticipate the partition coefficients for case of myeloma protein purification [15], the authors stated that
a defined experimental system [16]. At preparative scale, ATPAP a 3-fold increase in polymer MW could enhance the Kp in 3.5-fold.
systems also present important advantages for the development of Also important is the hydrophobic character of the target molecule
novel downstream strategies such as: high ligand concentration per and affinity ligand, since increases in polymer MW tend to enable
volume of polymer solution, which translates to high concentration hydrophobic environments [4], and if this characteristic is not well
of bound target biomolecule in one of the system phases (process addressed the ATPAP strategy could be inefficient. These character-
intensification) and, fast equilibrium achieving [17]. In addition, istics must be considered when developing ATPAP models. Previous
since ATPAP combines the bioselectivity of an affinity technique reports have described ATPAP characteristics and particularly have
with the robust conditions of traditional liquid–liquid extraction focused on the process integration capabilities of these systems.
systems, it is well suited for the specific recovery and purification Early researches addressed the recovery and purification of macro-
of diverse biological molecules. molecules like proteins. Then, the selective potential of ATPAP
In order to determine the potential of traditional ATPS for systems was employed for separation of complex particles such
the recovery and purification of biological molecules, experiments as cellular fragments and cells. In recent years there has been an
characterizing the effect of system parameters such as polymer important growth of ATPAP based strategies, with new affinity lig-
molecular weight, tie line length (TLL), pH and VR are usually ands that allow selective purification of high value molecules like
evaluated [2,4]. In ATPAP, additional system parameters must monoclonal antibodies or nucleic acids. In general, the full poten-
be considered in order to understand the partitioning behavior tial of this technique for the recovery and purification of biological
of biomolecules. Based on the literature we have identified four products requires further attention.
additional parameters usually studied on ATPAP: (1) optimum This report focuses in presenting a practical overview to high-
pH level for product partitioning, (2) ionic strength of the sys- light the importance of ATPAP for the recovery and purification
tem in polymer–salt or polymer–polymer systems, (3) affinity of biological products. Common chemical modification method-
ligand nature and concentration and, (4) concentration of modified ologies for polymer activation are presented. A critical analysis
F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13 3

of selected ATPAP systems reported in literature is summarized via the tresyl–chloride method [23]. This is possible due to the
to establish the benefits of the technique. Current challenges and increased number of active hydroxyl groups, in comparison with
potential trends of the practical application of ATPAP for the recov- PEG, present in DEX that can be chemically activated with affinity
ery of biological products are discussed. ligands.
The benzoate activation protocol exhibits interesting advan-
2. Common chemical modification methodologies for tages over the aforementioned activation techniques. In terms of
polymer activation overall processing time, it requires only 5 h to generate an active
matrix. In addition, an important advantage is that the active ben-
Affinity partitioning in ATPS provides a high-resolution tech- zoyl groups in the polymer matrix may act as affinity ligands
nique that can be coupled to traditional polymer–polymer and for molecules with tryptophan residues. This implies integration
polymer–salt systems. In addition, aqueous two-phase micellar between the polymer activation technique and the ligand coupling
systems have been modified in order to design new liquid–liquid requirements. Overall, the benzoate activation protocol seems well
extraction processes with enhanced product affinity. Fig. 1 presents suited for recovery and purification of proteins in ATPAP systems
the two most common strategies for the creation of ATPAP. It considering these advantages.
should be noted that polyethylene glycol (PEG) and/or dextran Even though the methodologies presented in Table 1 are tradi-
(DEX) have to be chemically active. Active polymers are charac- tional and represent several of the first approaches considered for
terized by a chemical reaction of an activator molecule with a polymer activation, more protocols have been developed in order to
polymer that renders increased activity to the functional groups increase affinity partitioning in liquid–liquid systems. Electrostatic
present in PEG or DEX. This methodology allows further covalent interactions have been well studied in order to develop chemical
coupling of affinity ligands. Nevertheless, as presented in Fig. 1, modifications that enhance this behavior towards target molecules.
recent researches emphasize the use of ligands free in solution Activation with carboxylic groups (in presence of diglutaric acid)
in order to avoid chemical activation of polymers. However, the and primary amino groups [5] has been described for recovery of
implementation of this strategy involves additional considerations antibodies and proteins, even though it must be emphasized that
such as partitioning behavior of the free ligand in ATPS prior to the amino groups tend to increase the pH level and thus have a lower
analysis of the potential interaction of the ligand with the target efficiency in product partitioning. Also, polymer activation with
molecule. pyrimidine groups has been described [3], emphasizing that active
A high diversity of molecules can be bound to ATPS poly- groups interact electrostatically with target molecules. Neverthe-
mer components as affinity ligands [15,17]. Several activation less, because of the nitrogen atom in the pyrimidine group, the net
techniques have been developed in order to achieve high ligand charge of the target product must be carefully monitored in order
coupling. In the particular case of PEG and DEX, the strategies focus to maximize system efficiency [3].
on the active hydroxyl groups present in the polymers. Examples Active phosphate groups can be coupled to PEG chains through
of traditional procedures for polymer activation for ATPAP are pre- an activation protocol involving phosphoryl-chloride [18]. It has
sented in Table 1. been found that increasing concentrations of PEG–PO4 in an extrac-
ATPAP purification strategies should consider the overall pro- tion system generates an increment of miscibility between system
cessing time in the selected activation technique, the chemical constituents. Therefore, higher concentrations of polymer and salts
nature of the active groups in the polymers and the post- are required to form the biphasic system, which favors the partition
activation ligand coupling to the polymer matrix. Considering the behavior of target molecules towards a particular phase. Additional
proposed protocols in Table 1, ethylenediamine and butanediol techniques and methodologies for the activation of PEG molecules
diglycidyl methodologies are characterized by multistep reactions are described in the literature [24,25].
that require 65–110 h in order to obtain an active polymer. The Since activation procedures often require long reaction times,
active hydroxyl groups in PEG present a chemical nature that allows optimization studies oriented in reducing the activation time may
the further addition of nucleophilic groups present in proteins, further increase the potential of ATPAP at lab and industrial scale.
peptides and aminoacids [15,17,19,20]. In addition, the principal Furthermore, the development of versatile modification reactions
advantage of these techniques involves the length of the spacer such as the PEG–benzoate (Table 1) or PEG–diglutaric acid will
arm or linker between the polymeric matrix and the affinity ligand pose serious advantages in chemical modification strategies, since
to be coupled. Since the linker arms with these protocols are longer the whole protocol (which includes polymer activation and ligand
(7–15 Å), the interaction probability between an affinity ligand and coupling) would be integrated into one step.
a particular molecule increases, generating a more efficient sepa-
ration process. Nevertheless, it should be noted that the extended 3. Polymer–polymer affinity aqueous two-phase systems
protocol time and relative complexity within these activation tech-
niques may limit, in some extent, their use in ATPAP. Traditional polymer–polymer ATPS formed by PEG and DEX
Periodate and tresyl-chloride protocols exhibit shorter times have been exploited to recover high value molecules that justify
in terms of polymer activation. This characteristic is the princi- the cost of the chemical forming phases, particularly dextran. In
pal advantage of the proposed methodologies, since 24 h or less addition, affinity enhanced polymer–polymer systems provide a
are required to generate an active polymer for the subsequent wide scope for the recovery of biological products. In this type of
ligand coupling stage. The active groups allow nucleophilic sub- ATPAP the polymer–ligand complex added to the system mainly
stitutions and additions, resulting in protein, carbohydrate binding partitions towards the phase rich on that specific polymer (e.g.
modules (CBM’s) and cofactors easy coupling [17,21–23]. Despite PEG–ligand molecules partition to the PEG rich phase). Examples
the proposed advantages, spacer arms between the polymer and of purified molecules in these systems include myeloma proteins
the affinity ligand are shorter when compared with the ethylene- [15], inner membranes [23], human cells [26], commercial enzymes
diamine and butanediol diglycidyl methodologies. This diminishes [21,27,28], nucleic acids [24,29] and monoclonal antibodies [3,5].
the complex flexibility, which in turn lowers the possible inter- Due to the particular characteristics of polymer–polymer systems
actions with the target molecules in the system. Nevertheless, it the studies have primarily focused on the recovery and purification
should be considered that a fractionation of the biomolecules that of high molecular weight products. Table 2 shows several exam-
need high ligand density in order to be partitioned to a polymeric ples of products that have been recovered using polymer–polymer
phase usually requires the chemical activation of the DEX phase ATPAP.
4
Table 1
Most common chemical modifications employed in ATPAP systems for biomolecule recovery and purification.

Polymer functional group Activation group Modified product Ligand requirement Ligand examples Reference

Yes – active amine groups: Dinitroflurobenzene, aminoacids [15]

F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13


Nucleophilic coupling

PEG Yes – oxirane active groups subject to Peptides, e.g., Glutathione, proteins, [19,20]
nucleophilic attack e.g., protein A, BSA

PEG Yes – active periodate groups: BSA, glutathione, carbohydrate binding [20]
nucleophilic substitutions modules

PEG No – active benzoyl groups allow Does not requiere biospecific ligand [18]
interaction with triptophan residues

PEG dextran Yes – Cl and O activated groups serve Proteins (avidin), coenzynes, cofactors [17,21–23,26,62]
as nucleophilic addition targets like ATP
F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13 5

Fig. 1. (A) ATPAP systems with affinity ligand (AL) chemically coupled to phase forming components. (A1) Polymer–salt ATPAP system with AL coupled in top phase polymer.
(A2) Polymer–polymer ATPAP system with AL coupled in bottom phase forming component. (B) ATPAP systems with AL free in solution in top or bottom phase. (B1)
Polymer–salt system with AL free in solution in top phase and polymer–polymer system (B2) with AL free in solution within bottom phase component. It should be noted
that the polymer solutions utilized can include traditional phase forming components such as PEG and dextran, but novel natural polymers can also be employed in each of
the ATPAP systems proposed.

Table 2
Selected examples of polymer–polymer aqueous two-phase affinity partitioning systems for recovery and purification of bioproducts.

Product System Modification Recovery yield Reference

IgG (human) Dextran T500/PEG 3350 PEG-benzyl 88% recovery/90% purity [3]
Human antibodies Dextran T500/PEG 3350 PEG-diglutaric acid 93% recovery yield [5]
IgG (human) Dextran T500/PEG 1450 PEG-IgG-HRPa 91% recovery/88% purity [11]
S-23 myeloma proteins Dextran T500/PEG 6000 PEG-dinitrophenyl Kp increase: 2.8–7 [15]
Thaumatin Dextran T500/PEG8000 PEG-gluthatione Kp increase: 0.27–4.6 [20]
Trypsin Dextran T500/PEG 8000 PEG-trypsin inhibitor Kp increase: 0.5–16 [20]
Biotinylated liposomes Dextran T500/PEG 3350 Dextran-avidin 95% recovery in BPb [22]
Biotinylated liposomes Dextran 2000/PEG 3350 Dextran-NeutrAvidin 94% recovery in BP [23]
Biologic membranes Dextran 40/PEG 3350 Dextran-NeutrAvidin 88% recovery in BP [23]
Red blood cells Dextran T500/PEG 6000 PEG-MPEG-IDA(II) and Dextran 95% recovery in BP [26]
Lymphoma cells Dextran T500/PEG 6000 Sephrose-IDA-Cu(II) 73% recovery in BP [26]
Plasmid DNA Dextran 40/PEG 600 GSTc + finger Kp increase: 0.005–29.42 [29]
␤-Galactosidase Dextran T500/PEG 8000 Dextran-benzoyl Kp decrease: 0.036–0.018 [65]
Lysozyme Dextran T500/PEG 8000 Dextran-benzoyl Kp increase from 0.20 to 0.85 [65]
␤-Galactosidase Dextran T500/Valeryl Dextran T500 Dextran benzoyl Kp increase from 0.046 to 0.18 [65]
Lysozyme Dextran T500/Valeryl Dextran T500 Dextran-benzoyl Kp decrease from 0.69 to 0.45 [65]
Lacrimal gland plasma membranes Dextran T500/PEG 3350 Dextran-WGAd 90% recovery in BP [66]
Escherichia coli inner membranes Dextran T40/PEG 3350 Ni-NTA resin in Dextran phase 93% recovery in BP [67]
a
HRP, Horseradish Peroxidase.
b
BP, bottom phase.
c
GST, Glutathione-S-Transferase.
d
WGA, wheat germ agglutinin.
6 F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13

Thaumatin and trypsin were purified by ATPAP systems formed consumption. Recombinant proteins tagged with a choline binding
using PEG 8000 and DEX T500 [20]. Epichlorohydrin and Epoxy- module (C-LytA) were isolated by Maestro et al. [6] using ATPAP
oxirane activation methods allowed coupling of glutathione and with free choline as a natural bioligand. The C-LytA module con-
trypsin inhibitor as biospecific ligands in PEG molecules. Thau- sists of six conserved ␤-hairpins that include four choline-binding
matin, a flavor enhancer and sweetener protein, presented a 17-fold sites [6]. Hence, the affinity of C-LytA for choline groups and analog
increase in Kp (from 0.27 to 4.6) in modified ATPS systems with compounds was used by the authors to design ATPAP strategies. A
PEG–glutathione in top phase. Trypsin, a proteolytic enzyme (EC strong interaction of the choline-binding sites present in the C-LytA
3.4.21.4), experienced a dramatic shift in partition coefficient from module with PEG was discovered. In consequence, proteins tagged
0.5 to 16 (32-fold increase), in aqueous systems with trypsin with the C-LytA module would experience an initial shift to the top
inhibitor bound to PEG. These results demonstrated that the recov- phase. On the other hand, addition of choline to the system reversed
ery and purification of the selected proteins could be dramatically the interaction between the C-LytA module and PEG, redirecting the
increased by using modified ATPS with biospecific ligands attached target molecule to the PEG-poor (bottom) phase. Green Fluorescent
to PEG molecules that partitioned to the PEG rich phase without Protein (GFP) fused to C-LytA was studied as a proof of concept. A
significant changes in the system composition [20]. In addition, the shift in the partition coefficient for the fusion protein from 186
authors established that the DS, which considers a saturation effect to 0.5 was observed, with an 80% total recovery of native GFP in
in the system when a molecule and an affinity ligand are added in the bottom phase. The main conclusion described by the authors
excess to the system, could decrease system efficiency. was the affinity partition modulation by choline residues present
Partition of membranes and negatively charged liposomes with in C-LtyA. Simplicity, rapidness, effectiveness and scalability confer
a biotin spacer arm was studied in ATPAP systems with DEX 40/DEX versatility to the proposed strategy.
2000 and PEG 3350 [23]. Isolation of inner and outer membranes Polymer based systems represent an important alternative for
represents an important area of opportunity for new drug delivery product recovery since both phases could be chemically active.
strategies [23]. Membranes and liposomes were recovered in affin- Furthermore, recovery and purity yields of protein products on
ity ATPS with deglycosylated avidin (NeutrAvidin) as a bioaffinity such systems are usually high (Table 2). Nevertheless, it is clear
ligand coupled (via tresyl chloride activation method) to dextran that process efficiency in ATPAP could be optimized. According to
molecules that partitioned to the bottom phase. Liposome parti- several researches, salt addition in polymer-based systems is capa-
tion was first studied in order to determine optimal parameters ble of increasing recovery yields, since marked shifts in ionic force
for further purification of biotinylated membrane fractions. In sys- can favor specific affinity interactions between target particles and
tems with absence of NeutrAvidin, 90% of the liposomes partitioned affinity ligands [3]. In this tenor, salt effects are not always consid-
to the PEG rich phase. Addition of NeutrAvidin–DEX induced a ered when applying ATPAP extraction methodologies even though
notable shift of biotinylated liposomes to the dextran rich phase, it has been determined that particular salts may positively enhance
from 85% in PEG rich phase (without ligand) to 90% in the dextran product partitioning [15,29]. Also, the effect of pH on partitioning
bottom phase. The authors concluded that the interaction between should be considered in order to increase ATPAP efficiency. In addi-
the biotin spacer arm and NeutrAvidin in liposomes and mem- tion, it is important to consider that even though polymer activation
branes determines the partitioning behavior of these molecules. techniques allow ligand coupling in order to favor product partition
This biological interaction represents a valuable approach for the in aqueous systems, spacer arms between the polymer and ligand
purification of biotinylated products. tend to have different lengths depending on the employed activa-
Affinity partitioning of pre-purified plasmid DNA (pDNA) using a tion procedure [20]. In general longer spacer arms favor affinity
PEGylated zinc finger-GST (Glutathione-S-Transferase) fusion pro- partitioning since the probability of the interaction between prod-
tein as affinity ligand was studied in PEG–DEX ATPAP by Barbosa uct and ligand increases [22,23]. In this tenor, exploiting the effect
et al. [29]. Purified pDNA based therapies emphasize the impor- of the spacer arm length represents an interesting option in order
tance of scalable and cost effective process development. The to increase recovery yields in ATPAP [22].
coupling of a fusion protein to a polymeric matrix reduces steric One important strategy within ATPAP is that of free ligands in
hindrance commonly present in resin-coupled proteins [29]. PEG solution. In order to avoid time-consuming activation protocols for
600–DEX 40 and PEG 1000–DEX 500 systems were used for pDNA polymers, direct addition of affinity ligands in order to enhance
isolation. Comparative experiments determined that when the product partitioning has been presented [6,29]. Despite the fact that
fusion protein was first administered in the system in its unPEGy- polymer activation is not necessary, a proper study of ligand parti-
lated form, its partition coefficient was of 37.89 and the partition of tioning must be undertaken in order to characterize its partitioning
the plasmid added to the system was calculated as 0.005. Neverthe- behavior, and thus determine its true potential in affinity extraction
less, pDNA partition coefficient increased 5900-fold (from Kp 0.005 procedures. Modification of the proposed affinity ligands could be
to 29.42) when adding the PEGylated form of the proposed affinity carried out in order to manipulate their partition behavior towards
ligand to system PEG 1000–DEX 500. The authors confirm that a either phase of the system. One of the most promising techniques
strong interaction between the PEGylated fusion protein and the that could be used to modify affinity ligands is PEGylation. PEGyla-
pDNA allows the complex to concentrate in a specific phase (PEG tion reactions have been well studied and characterized and allow
rich phase). PEG 600–DEX 40 systems presented a lower increase the coupling of a PEG moiety to a particular compound, in this
in Kp (from 0.0003 to 2.42). However, when increasing the PEGy- case an affinity ligand [29]. Because of the chemical nature of the
lated fusion ligand concentration to 293 ␮g/ml, pDNA partitioned modified ligand, it has been demonstrated that partitioning behav-
completely (100%) towards the PEG rich phase, thus representing iors toward PEG rich phases are greatly favored [29]. Considering
an efficient strategy for its recovery. The authors demonstrated the this methodology, PEGylation of proteins, antigens or dyes could
potential of affinity partitioning of pDNA in ATPAP, emphasizing the be developed in order to modify ligand behavior within ATPAP sys-
effect of polymer characteristics such as concentration and molecu- tems and thus design efficient unit operations for recovery of whole
lar weight, on ligand distribution through the system, which can be cells or stem cells, nucleic acids and cellular fragments. It must
exploited to develop new concentration and purification strategies. be emphasized that even though PEGylation has been widely used
Purification strategies using free ligands in solution have been for therapeutic applications, its relevance and implementation in
implemented in ATPAP to recover biological products. The princi- aqueous affinity partitioning has not been extensively addressed.
pal advantage of this methodology is that no polymer activation Hence, it is expected that novel applications of this technology
and ligand coupling are involved, reducing time and resource coupled to ATPAP strategies will be seen in the near future.
F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13 7

Polymer–polymer systems offer great versatility within ATPAP The use of protein carriers in order to increase the potential
(Table 2). Phase components characteristics maximize biological of ATPAP has been also documented by Hye-Mee et al. [11]. They
interactions which can be exploited in order to develop novel reported the use of Horseradish Peroxidase (HRP) as a carrier for
downstream processing protocols. Nevertheless, the cost of the human IgG as an affinity ligand. Model mixtures with rabbit anti-
chemical forming phases in these systems should always be con- human IgG (target molecule) and goat anti-mouse IgG (impurity)
sidered, since PEG and particularly DEX may represent a significant were employed in order to establish the potential of ATPAP for the
cost at industrial scale. However, this issue can be addressed by purification of the target molecule. ATPS systems involving 10%
polymer and ligand recycling, applying back-transfer phase strate- w/w PEG 1450 and 15% w/w potassium phosphate were used. Rab-
gies in order to maximize raw material medium life. bit anti-human IgG partitioned towards the salt-rich phase when
added with other contaminants in the system without affinity lig-
4. Polymer–salt affinity aqueous two-phase systems ands. On the other hand, when human IgG–HRP was added to the
system, the complex exhibited higher affinity towards the top PEG-
Traditional polymer–salt ATPS have been used for the recov- rich phase (Kp 1.65). As a result, a recovery of 78.2% of the target
ery and purification of high value products. However, affinity molecule was achieved.
enhancement has not been extensively addressed as in the case of Purification of glucose oxidase and ␤-galactosidase from
polymer–polymer systems. A reduced number of cases are found Aspergillus niger and Kluyveromyces lactis fermentation broths was
for these systems, and applications seem to be focused in purifica- reported using modified PEG–sodium sulfate systems [33]. In gen-
tion of commercial enzymes (Table 3). One of the mayor drawbacks eral PEG–sodium sulfate systems are less used in comparison
of using salt components in affinity enhanced liquid–liquid extrac- with PEG–phosphate ATPS. In this particular case four PEGs were
tion methodologies lies within the biological interaction of ionic synthetized: PEG-trimethylamine (TMAionic interaction ligand),
groups of certain molecules towards the salt forming phase. The use PEG-benzoate (pseudo specific ligand), PEG-Coomassie brilliant
of certain salts in ATPS has been proven to increase the partition of blue G-250 (pseudo-specific ligand), and PEGpalmitate (hydropho-
target molecules below certain limits, but above critical concentra- bic ligand). PEG-Coomassie favored the partition of glucose oxidase
tions a decreased pattern of solubility for proteins and biological towards the bottom phase and the contaminants to the upper
products can be observed. Important shifts in the ionic force of phase. ␤-Galactosidase was best recovered in PEG–TMA systems
the solutions are experienced, reducing essential ionic interactions (80% recovery and 7-fold purification increase). TMA acts as a
between target molecules and proposed affinity ligands. Neverthe- positively charged ligand that repels positively induced-charged ␤-
less, applications have been reported using this type of systems for galactosidase favoring its partition towards the bottom phase. The
ATPAP [28,30,31]. authors emphasized the affinity attractions between active groups
As an example, penicillin acylase (PA) partitioning behavior in PEG and contaminant proteins, which in turn provides a new
in ATPAP polymer–salt systems has been studied [18]. PA is an strategy for purifying target molecules.
important hydrolytic enzyme that is used in the direct synthesis One of the principal advantages polymer–salt systems offer
of semi-synthetic penicillins [30,31]. Four PEG derivatives were for protein recovery is the reduced viscosity and density of the
synthesized: PEG-benzoate (PEG-bz), PEG-trimethylamine, PEG- aqueous formed phases. That will result beneficial in the poten-
palmitate and PEG-phenylacetamide (PEG-paa). PA partition in tial implementation at commercial levels [18,20]. However, further
PEG-bz systems presented a 14-fold increase in Kp (60% total recov- characterization of hydrophobic, electrostatic and biospecific inter-
ery) with 21% PEG-bz of the total PEG mass. For the particular actions is needed, since PEG molecules tend to suffer important
case of PEG-paa systems, PA presented an enrichment factor of 9.4 modifications when affinity ligands are coupled to their organic
and an 11.5-fold increase in Kp. PEG-bz and PEG-paa presented backbone. Addition of proteins or chemical compounds may
high affinity towards PA due to the fact that the interaction energy increase the hydrophobic character of the polymeric phase which
required for the coupling of PA with these components was lower could enhance partition of bioparticles with hydrophobic domains
than the interaction energy between PA and its natural substrate. [18]. This is an important strategy when partitioning of molecules
This increased the biological affinity of the target enzyme towards with known hydrophobic domains will be developed, since recov-
the activated PEG molecules present in the PEG-rich phase. The pro- ery yields could be positively improved.
posed PEG derivatives significantly increased PA partition in ATPAP For enzyme recovery, polymer–salt systems present several
systems and the primary interaction was attributed to the phenyl points to be characterized in order to design proper extraction pro-
ring and amide groups present in the proposed affinity ligands [18]. tocols. First, the use of high ligand concentration or high amounts
An additional example for the application of ATPAP using free of modified polymer molecules has demonstrated to be an effi-
ligand addition is the partition of chitinases in systems formed cient strategy to increase product recovery (thaumatin [20] and
by PEG 6000, dipotassium hydrogen orthophosphate (K2 HPO4 ) glucoamylase [21]). Nevertheless, in these cases when excessive
and free chitosan in solution as affinity ligand [28]. Chitinases are concentrations of modified PEG were added, an important shift
hydrolytic enzymes that act selectively in chitin [32]. ATPAP sys- of contaminant proteins was seen, thus reducing the overall effi-
tems were prepared with 22% PEG (w/w), 10% K2 HPO4 (w/w) and ciency of the proposed systems. Since this problem is common
0.5% (w/w) of a chitosan solution. Chitinases extracts from Neu- when designing ATPAP methodologies [15,18,34], response surface
rospora crassa, cabbage (Brassica oleracea) and puffballs (Lycoperdon analysis (RSA) or statistical methods could be employed in order
umbrinum) were loaded to independent systems. Chitinase recov- to develop prediction models and thus facilitate ATPAP design.
ered from N. crassa presented a 34-fold purification increase and Another important aspect in affinity systems design lies within
86% of total recovery in top phase. Meanwhile, extraction from optimum pH determination. Purification strategies for enzymes
cabbage and puffballs exhibited a 20-fold and 38-fold purification tend to use natural substrates (as presented by Teotia et al. [28] in
increase with 80% and 88% total activity recovery in the PEG rich chitinase recovery), in order to maximize product recovery yields.
phase, respectively. The authors suggested that due to the high In this tenor, the pH is the most important factor in the system since
concentration of chitinases, non-specific interactions between the enzyme purification must be carried out at a pH in which no enzy-
product and the affinity ligand were favored, lowering the over- matic activity will be present and thus avoid degrading the affinity
all binding capacity probably due to steric hindrance. This effect ligand [28,34,35]. Once partition has been achieved, the enzyme-
should be considered in order to develop efficient recovery and rich phase can be transferred to a fresh salt phase and elute the
purification strategies. product of interest with addition of salts. In addition, pH levels
8 F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13

Table 3
Examples of polymer–salt aqueous two phase partitioning systems for recovery and purification of biomolecules.

Product System Modification Recovery yield Reference

Thaumatin PEG 8000/phosphate PEG-gluthatione Kp increase from 0.21 to 0.71 [20]


Glucoamylase PEG 300/phosphate Starch as free ligand Kp decrease from 6.6 to 0.73 [21]
Chitinase PEG 6000/phosphate Chitosan as free ligand 90% recovery in UP after final elution [28]
Penicillin acylase PEG 4000/phosphate PEG-benzoate 14-fold increase of Kp (0.1–1.4) and 60% recovery in a single [18]
step extraction
Penicillin acylase PEG 4000/phosphate PEG-trimethylamine Kp of 2.1 with 70% recovery in single step extraction [18]
Penicillin acylase PEG 4000/phosphate PEG-palmitate Kp of 3.22 and 0.56 for contaminant proteins; 75% recovery in [18]
single step extraction
Penicillin acylase PEG 4000/phosphate PEG-phenylacetamide Kp of 2.3 and 0.1 for contaminant proteins; 69% recovery in [18]
single step extraction
Cutinase PEG 1500/phosphate Butyrate as free ligand 98% recovery in first extraction and Kp of 135 [34]
Phospholipase D PEG 6000-phosphate Alginate as free ligand 85% recovery in UP and 78-fold purification factor [35]

have a direct effect in biological interactions between natural sub- are considered high added value molecules [4] could be recovered
strates and affinity ligands. Electrostatic bonding between target within this affinity approach. This represents an important research
molecules and proposed ligands heavily depends on pH level since area for polymer–salt ATPAP systems, since a low number of stud-
active groups in each molecule possess different pKa or isoelec- ies with non-chromatographic techniques have been implemented
tric point (pI) values. Considering that every molecule has different for AFPs and phycobiliproteins.
active groups, the pH level will pose an important priority since
it will be the moderator of electrostatic interactions between tar- 5. Affinity enhanced ATPAP systems using alternative
get products and affinity ligands. Nevertheless, it must be also phase components
considered that pH variations around the pI enhance hydrophobic
interactions while decreasing electrostatic forces. Affinity enhanced liquid–liquid extraction systems have been
An important option that has been developed in polymer–salt developed with phase components different to PEG, DEX or tradi-
systems consists in the coupling of molecules carriers to affin- tional salts. The use of surfactants and natural polymers has been
ity ligands in order to maximize product recovery in specific previously reported because of their versatility in terms of ATPAP
phases [11]. However, the partition of the carrier–ligand com- systems development. In addition, natural components exhibit
plex must be first characterized in order to propose a possible important advantages over traditional polymers since they are
extraction procedure, which could suppose a disadvantage of considered environmental friendly reagents. Purification of com-
employing this type of selective strategy. As can be concluded mercial enzymes [42], fusion peptides [43] and cellular membranes
when comparing data from Tables 2 and 3, recovery yields for [44] has been reported (Table 4) using alternative materials in
proteins are clearly lower in polymer–salt systems (particularly ATPAP systems.
those systems in which polymer components have been chemi- Aqueous two phase micellar systems have proven to work as an
cally modified to enhance partition). These partitioning behaviors alternative to traditional ATPS. These systems are formed when cer-
could be principally attributed to salt interference and ionic force tain surfactants (ionic, zwitterionic and non ionic) are mixed with
shifts. Salt-forming components interfere with the ionic nature of an aqueous or organic phase above a critical micelle concentration
active groups in affinity ligands and biological molecules, as a result (CMC) [45]. When aqueous phases are employed, the hydropho-
the implementation of ATPAP methodologies in polymer–salt sys- bic tails of the surfactant molecules tend to aggregate in a circular
tems is more limited since reduced affinity interactions could manner, forming structures known as micelles. On the other hand,
be maximized within these extraction systems. In polymer–salt when organic components are used, the hydrophilic heads form
systems, the direct competition between affinity ligands, their aggregates with the hydrophobic tails positioned in the exterior in
binding capacity and cross-reactivity with molecules present in a structure know as reverse micelle. Surfactants used for micelle
the environment need to be properly characterized in order to formation can be modified with affinity ligands in order to increase
avoid undesired interactions. Since a direct economic comparison partition of interest molecules [46]. Therefore, micellar ATPAP pose
between polymer–polymer systems and polymers–salt systems an interesting alternative to traditional ATPS.
will favor the latter, the recovery of proteins using polymer–salt CBM9-Green Fluorescent Protein (CBM9-GFP) was partitioned
ATPAP results interesting despite the aforementioned limitation. in a micellar system with n-decyl-␤-d-glucopyranoside (C10 G1 ) as
An important area of interest in which polymer–salt systems phase former surfactant and affinity ligand [43]. Since C10 G1 acts as
could prove to enhance product recovery is that of anti-freeze phase forming component and ligand, the proposed binary system
proteins (AFPs). These molecules play a vital role in certain (water/surfactant) significantly simplifies the recovery and purifi-
organisms in order to allow them to survive at subzero temper- cation of the product of interest. Affinity partitioning of CBM9-GFP
atures since they decrease the freeze point of the body fluids was studied first with the presence of glucose, in order to demon-
non-colligatively [36–38]. Nevertheless, actual downstream pro- strate that this carbohydrate acted as a potential inhibitor in the
cessing strategies seem to focus on chromatographic techniques system. At a 55 mM glucose concentration, the Kp for CBM9-
based on metallic affinity chromatography [39], size exclusion GFP was 0.473 ± 0.005, while in systems without glucose a Kp of
[40] and hydrophobic interaction chromatography [41]. In addi- 3.1 ± 0.2 was achieved. The observed partition coefficients show
tion, non-chromatographic techniques have been poorly addressed the strong interaction between the CBM tag and the surfactant. The
and ATPAP protocols could be designed in order to purify these use of alternative polymers to PEG and DEX is an important area
bioproducts. Since these proteins exhibit affinity to ice molecules of opportunity for ATPAP systems. Different polymers have been
through hydrogen bonds, Van der Waals and hydrophobic interac- proposed to replace PEG in ATPS extraction systems [15,34,42,47]
tions, ATPAP strategies could be developed using PEG derivatives in order to construct more economic strategies. Several research
as affinity ligands such as PEG-benzoate, PEG trimethylamine, PEG groups have proposed alternative polymers that may act as replace-
palmitate or PEG phenylacetamide. In this same tenor, molecules ments of the traditional polymeric components. The development
such as phycobiliproteins (B-Phycoerythrin, C-Phycocyanin) which of thermoseparable copolymers made of ethylene oxide and
F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13 9

Table 4
Examples of aqueous two phase affinity partitioning systems using non-traditional polymers and phase forming components.

Product System Modification Recovery yield Reference

S-23 Myeloma Proteins PEO 6000/Dextran T500 Dinitrophenyl-PEO 2.5-fold increase in Kp (2.8–7) with [15]
presence of modified PEO
␤-Mannanase Guar Gum-PES 200 Galactomannan as FLa Kp increase from 1.37 to 7.06 [42]
CBM-9-GFP n-Decyl-␤-d- C10 G1 (simultaneous phase 6-fold increase in Kp from 0.4 to 3.3 [43]
glucopyranoside/water former and affinity ligand)
Membrane proteins Tween 20/Dextran 500 Dextran-cobalt affinity resin 75% recovery in detergent phase [44]
Glucose-6-phosphate dehydrogenase PEG 6000/PES 100 Cibacron Blue F3GA (FL) ATPS with dye induced an increase in [60]
Kp from 0.73 to 1.59
Cutinase with peptide fusions BREOX PA G (EOPO) 1000-Reppal Triton X-100 and C12 E5 as FL’s Kp increase from 0.89 to 10.35 of [61]
PES 200 recombinant cutinase with fusion tags
Cellulase PEO-dextran T500 Maleic acid-PEO 2-Fold increase in Kp for cellulose with [68]
40% modification rate at PEO phase
a
FL, free ligand.

propylene oxide (EOPO polymers) constitutes an interesting alter- exploit the affinity of the protein binding module to the galac-
native for ATPAP systems. In addition, EOPO components are tomannan polymer in the aqueous extraction system. The natural
commonly mixed with natural starch polymers to form biphasic polymers hydroxypropyl starch PES 200 (Reppal) and Guar gum-
extraction systems [34]. It has been determined that affinity ligands galactomannan (9% (w/w) and 3.6% (w/w) respectively) were used
can be attached to either phase forming molecules (EOPO or natu- as phase forming components. This system allowed a top phase
ral polymer phase), thus generating an ATPAP system. Also, natural (guar gum-galactomannan) recovery yield of 90% with a Kp of 7.06
based components tend to be compatible with a wide variety of for ␤-mannanase. With the obtained results, it can be noted that
natural substrates, which increases the possibilities of developing a diverse CBM could be employed for protein purification. Hence,
novel ATPS based procedure [42]. Despite these advantages, the use a general extraction procedure for proteins could be developed if
of alternative phase components is not extensively characterized. each target molecule is fused with a specific binding module. This
In this context, the phase formation diagrams or binodal curves would represent an important advantage in the development of
must be constructed in order to validate the experimental designs novel downstream processing strategies.
and further purification protocols. Affinity partitioning of human immunoglobulin G (IgG) from
Cutinase selective partitioning was studied in modified ATPS Chinese hamster ovary cells has been also studied on ATPAP
conformed by polyalkylene glycol 50 A 1000 as top phase form- systems constructed with non-conventional constituents [49].
ing polymer and hydroxypropyl starch (BREOX-PES 200) as bottom Purification was achieved by employing a system with 8% w/w
phase polymer [47]. A dual affinity method was applied in which copolymer of ethylene oxide/propylene oxide (UCON), 6% w/w of
fusion tags (WP = Trp-Pro)2 , (WP)4 and Thr-Gly-Gly-Ser-Gly-Gly- DEX 500 and triethylene glycol–diglutaric acid as affinity ligand
(WP)4 were attached to the enzyme and served as receptors for (TEG–COOH). Addition of TEG–COOH generated an improvement
specific detergents. It was first determined that fusion tags did not on both extraction yield and purity of human IgG in the top
affect significantly cutinase partition in the aqueous systems. Con- phase. The significant increase in log Kp (from −0.2 to 3.5) of IgG
sidering this, addition of non-ionic detergents to the systems in was attributed to electrostatic interactions between the positively
order to increase the fusion tag affinity was investigated. Tween charged IgG and the negative carboxylic groups of TEG-COOH [49].
20, Tween 80 and Triton X-100 were added (1% w/w) obtaining Kp According to surface response models generated by the authors, the
fold increases of 1.84, 1.55 and 6.4, respectively. However, addi- optimal ATPS for the purification of IgG was determined to contain
tion of a 2% w/w solution of pentaethylene glycol monolauryl ether 10% w/w UCON, 5.5% w/w dextran and 20% w/w TEG–COOH. In this
(C12 E5 ) to the Breox-PES 200 systems allowed a 6.6-fold increase particular case, the use of optimization methodologies generates
in Kp and 5.8-fold increase in purity compared to systems without valuable information for downstream processing strategies.
affinity ligand. A tag effect (defined as the ratio between partition One of the current trends within natural occurring polymers
coefficients for the fusion protein and the native type protein) of 14 is their employment as phase forming components and affinity
for the TGGSGG-(WP)4 tag was observed, indicating that this deter- ligands in a simultaneous manner [42]. According to liquid affin-
gent was more efficient than the other studied to increase selective ity partitioning theory, a complex number of variables have to
partition of cutinase. The authors concluded that depending on be determined in order to design an ATPAP extraction procedure,
the fusion tag employed the interaction between a molecule and one of the most important being the presence of a specific affinity
a matrix can be increased. Furthermore, the addition of non-ionic ligand. If ligands are to be coupled or free in solution, their inclu-
detergents can modify selectively the partition of fusion proteins sion to the system requires a characterization procedure in order
in order to increase their partition to one of the phases in an ATPS. to determine optimum concentrations of phase forming compo-
In addition, some characteristics of the employed polymers should nents, ligands and sample load. In this tenor, if a proposed natural
be emphasized. Even though PES 200 can be found commercially polymer can also serve as a specific ligand for product recovery,
at lower economical costs than dextran, the relative time stock this could reduce ATPAP complexity since no additional charac-
solutions are functional may be limited due to its biodegradability terization should be made, and the system could be considered a
potential [48]. Therefore, adequate storage and processing strate- traditional ATPS although the presence of this affinity component
gies should be used in order to avoid unwanted degradation of the would enhance product recovery.
polymer. Purification of ␤-mannanase in a Guar-gum galactomannan/PES
Carbohydrate binding modules (CBM) are natural tags present starch system was possible since the upper phase of the system
in enzymes that facilitate substrate targeting and coupling [42]. was the natural substrate for the target enzyme [42]. In addition
Antov et al. [42] employed the renewable polysaccharide guar to the above mentioned strategy, the authors also concluded that
gum galactomannan and starch in a novel ATPAP system to purify the pH of the system had a significant effect on the enzyme since
␤-mannanase from Cellulomonas fimi. This enzyme has a CBM the product should be in a non-catalytic form. Even though target
with high specificity towards mannans. The approach was to products can be recovered and further eluted in a fresh salt phase
10 F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13

(S-23 myeloma proteins [15] or ␤-Mannanase [42]) product stabil- bioprocesses [54]. Nevertheless, as presented in Table 5, actual
ity could be affected if an adequate system pH is not used, favoring purity standards for stem cells have not been well addressed with
the partition of contaminant molecules towards the target phase. liquid–liquid fractionation techniques. The most recent applica-
A disadvantage of natural polymers is their high viscosities. tion of ATPAP for CD34+ stem cells recovery presented yields of
Despite the fact that natural polymers (Reppal PES starch or Guar 42% in purity. This purity is significantly lower than that obtained
Gum Galactomannans), are considered environmental-friendly, with traditional cell sorting methodologies [53]. In this case, the
high viscosities must be reduced in order to employ them as phase strategy presented by the authors involved a monoclonal antibody
forming chemicals. Pretreatments based on heat and partial enzy- against the CD34 surface antigen, which favored partition of tar-
matic hydrolysis have been proposed in order to fulfill this objective get cells to the PEG rich phase of the constructed systems. Even
[42]. Nevertheless the cost-effective approach within the natural though a strong immunoaffinity interaction was established, a cen-
system would be hampered. As with polymer–polymer systems, trifugation step was employed for target cell concentration in the
the use of modified natural polymers in high concentrations may interphase of the system, which also concentrated an important
enhance protein recovery since a biocompatible environment will number of contaminant cells. In this tenor, affinity strategies based
be available for the product of interest [15,42]. However, it should on target product concentration in the top or bottom phase should
also be taken into account that once the stoichiometric limits be established, in order to avoid working with interphase steps,
for ligand concentration and polymer addition are reached, sys- which tend to increase the presence of contaminant molecules. In
tem saturation is achieved. In terms of affinity systems composed addition, it should be considered that the scalability, integration
with alternative phase components and the present discussion and and intensification potential of ATPAP for stem cell recovery are
examples described within this section and Table 4, protein recov- comparative advantages against traditional cell sorting method-
ery is well suited with this technology. Starch based polymers and ologies (FACS and MACS), which cannot be employed at industrial
other plant polysaccharides (e.g. guar gum) allow protein recovery levels [53]. Although ATPAP methodologies for stem cell purifica-
once they have been pretreated. Furthermore, their natural prece- tion have not addressed purity standards, it must be emphasized
dence gives them added value since actual global trends emphasize that overall recovery yields with this technique are superior to cell
the importance of working with natural resources [34,42,48]. Natu- sorting methodologies. As reported by Sousa et al. [54], recover-
ral polymers are environmental-friendly components with biologic ies of 95–96% of CD34+ stem cells have been achieved by liquid
characteristics suitable for their application in downstream pro- affinity partitioning and just 85% recovery has been reported with
cessing operations. However, the production of such materials in MACS procedures. In this context, improved ATPAP will result in an
some cases needs to be standardized, which represents an impor- increase in purity for stem cells. Consequently, this extraction tech-
tant issue to be addressed for the future application of such phase nology could prove to be one of the most important at industrial
forming chemicals. Also important is the chemical stability of the level, since a critical advantage relies on its potential for bioprocess
polymer solutions, since several natural polymers cannot be pre- scale-up.
pared and kept in stock for extended time. In terms of enzyme recovery, penicillin acylase was selected
since recovery through ATPAP was lower when compared to
metallic affinity chromatography (IMAC). It was determined that
6. Trends and challenges of ATPAP recovery and purity could be maximized with pre-established chro-
matographic methods and by optimization of the affinity resins in
One of the mayor trends in downstream processing involves order to increase superficial area and decrease pressure drops, and
process integration and intensification. In this context, the imple- thus confer an increased scalability potential [52]. However, it is
mentation of ATPAP systems could provide an integrative approach important to consider that the recovery of the enzyme with ATPAP
in process design. Biological affinity strategies can be used, as seen, offers advantages of simplicity, scale up and potential of process
to purify commercial enzymes [5,6,21], membranes [23], plasmidic intensification [18].
DNA [24], and recombinant proteins [6] achieving high recovery As it has been discussed, recovery yields and purification
yields. One of the mayor advantages of working with ATPAP sys- achieved using ATPAP and chromatographic methodologies are
tems is their relative ease of set up and construction, minimizing similar. Even though ATPAP has not been extensively characterized,
chromatography based strategies. its scale up potential and intensification capabilities could posi-
Nevertheless, the well-established chromatographic techniques tion this alternative technique as a valuable approach in industrial
applied for the selective recovery and purification of biological applications. Furthermore, the current existing process challenge
products could possibly hinder ATPAP application at industrial lev- that the industry is facing will most likely require the use of com-
els. The main comparison between these techniques involves their bined chromatography and non-chromatography approaches. For
efficiency in terms of yield and purity for a specific product, and the the particular case of penicillin acylase recovery, liquid affinity
economic aspects of each one. In terms of product yield and purity, partitioning can provide similar results when compared to stan-
a vast number of publications can be found in terms of chromatog- dard affinity chromatographic-based protocols (see Tables 2 and 3).
raphy or affinity chromatography for product purification. Nevertheless, it must be considered that the effect of several param-
Table 5 presents a selective comparison of representative eters must be characterized in order to obtain optimal results.
biomolecules studied with ATPAP and chromatographic tech- One of the principal strategies to increase soluble protein recovery
niques. In terms of efficiency, it can be seen that recovery yields and purity relies on the strict control of electrostatic interactions
for human IgG [3,50,51] and CD34+ stem cells [53,54] are very between amino acid residues and active sites present in affin-
similar since they tend to be above 90%. Even though the proto- ity ligands. This can be manipulated by studying the optimum
cols for antibody purification with affinity chromatography have pH levels for protein stability and activity within an extraction
been well studied in the past decades, stem cell recovery through system, since different behaviors are expected for each protein
cell sorting methodologies will be insufficient in order to meet [20,28]. An additional strategy to increase product partitioning
the need for future stem cell therapies due to their low scalabil- is the implementation of affinity ligands that exhibit hydropho-
ity potential [53]. Since recovery yields are comparable with those bic domains that will enhance recovery of molecules with similar
obtained with MACS techniques, authors have emphasized the rela- domains or structures [18]. It should be noted that this effect can
tive simplicity of ATPAP for recovery of stem cells and thus, provide be maximized when protein extraction is made near the isoelec-
a more economical and scalable technology for future stem cell tric point, since product solubility will be low and thus facilitate
F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13 11

Table 5
ATPAP and chromatographic techniques applied in bioproduct recovery.

Product Application Method Conditions Yield Reference

Human IgG Therapeutic treatments ATPAP Dextran T500-PEG 3350 (PEG benzyl) 91% recovery and > 90% purity [3]
Protein A-Affinity Product elution with 0.1 M glycine HCl 95% recovery and > 93% purity [51]
chromatography pH 3 and column regeneration 50 mM
NaOh
Penicillin Synthesis of ATPAP PEG-phosphate (PEG trimethylamine) >70% recovery [18]
Acylase semi-synthetic IMAC chromatography Protein capture with pH 8.5 and 0.5 M 92% recovery and >90% purity [52]
penicillins NaCl, elution with 1 M NH4 Cl buffer
with 10 mM sodium phosphate
CD34+ Therapeutic treatment ATPAP PEG 8000-dextran T500 (with anti 95.5% recovery with 42% purity [54]
stem cells of selected dystrophies CD34+ mAb)
MACS device Immunomagnetic cell sorting from a 85% recovery and >95% purity [53]
50% SSEA-1+ mESC cell mixture

hydrophobic interactions. In addition to the previous technical polysaccharides are readily being used for their phase forming
comparison between ATPAP and chromatographic methodologies, capabilities and their user-friendly disposability.
the bioprocessing trends and economic aspects of these two tech- The nature and origin of affinity ligands used on ATPAP has
niques should also be considered in order to establish the appealing experienced significant changes over time. First reports on ATPAP
of ATPAP for its future extensive use on the biotechnology indus- systems emphasized the specific interactions between affinity lig-
try. According to the potential trends in bioprocess engineering ands and molecules of interest. Such ligands were chemically
and monoclonal antibodies production, current advances in pro- designed and synthesized to exert specific affinity. During the last
duction systems will condition the actual downstream processing decade the transition from chemical molecules to ligands such
strategies to process higher concentrations of target molecules as enzymes, carbohydrate binding modules (CBM’s), biologic sub-
[55–57]. In this context, several promising alternatives, like aque- strates and antigens, has become an important trend in ATPAP
ous two phase affinity partitioning, could emerge in order to reduce applications. This fact marks a distinction between first generation
the number of unit operations in a process and at the same time and second generation ATPAP systems. It can be expected that a
enhance the overall productivity when coupled to a smaller affinity great variety of biomolecules will be studied within the next few
chromatographic step [55]. Rosa et al. recently published a whole years as possible bioligands.
economical comparison between aqueous partitioning and affin- In terms of free affinity ligands in solution, they provide several
ity chromatography for the selective recovery of high added value advantages over chemically coupled ligands. Polymer activation
bioproducts [55]. In the proposed investigation, a pharmaceutical techniques are time and resource consuming, while free ligands do
plant was designed and modeled with mathematical and economi- not need to be covalently coupled to the phase constituent, hence
cal software in order to predict total production cost of 1.2 tons of a time saving protocols can be developed. Nevertheless, after placing
mAb per year. Two processes were designed: (1) one with a down- an affinity ligand in the system, the molecule by itself must exhibit
stream process based on ATPS and (2) another that included affinity a strong preference for either the top or bottom phase in order to
chromatography. Total inversion costs were of $10,959,672 USD develop a purification strategy based on that behavior. In addition,
for the ATPS process against $ 12,695,523 USD for the chromato- the effect of adding a particular ligand should be well studied. Ide-
graphic approach. In terms of annual operating costs, ATPS included ally it would be expected that the system characteristics such as VR
$ 8.7 million dollars and affinity chromatography supposed $ 14.4 and composition stay well balanced and unaltered.
million dollars. The most expensive component of the operating Affinity partitioning has studied a wide scope of biomolecules
costs was the affinity resin employed in chromatographic proce- in liquid–liquid extraction systems and the primary focus of this
dures for mAb extraction ($ 16,250 USD/L of resin which constitutes technique has been non-therapeutic proteins [33,58–62]. How-
79% of the operating cost). On the other side, raw materials consti- ever, the application of ATPAP for the commercial recovery of high
tute the 58% of the total costs for the ATPS annual expenditures commercial value biologicals with therapeutic uses is expected.
(PEG and salts as phase forming components). The authors found One particular example makes reference to the recent advances
that working with low titers of mAb (1–2 g/L) decreased annual on monoclonal antibodies (MABs) therapies [3,63]. In this con-
operating costs in chromatography but increased ATPS expendi- text, the use of ATPAP systems may be an attractive alternative
tures. The opposite behavior was found when increasing the mAb to chromatography-based strategies. Another area that expects a
concentration to 4–7 g/L, since ATPS can process more sample of sudden increase in demand is that of gene therapy mediated by
the desired product without changing the system composition. It plasmidic DNA vaccination [24,29]. Based on the numerous on-
was concluded that the proposed ATPS methodology could contin- going trails for pDNA therapies novel strategies for the recovery
uously process the same amount of mAbs than a chromatographic and purification of these macromolecules are going to be needed
based process, with reduced operating costs and high titer process- in order to satisfy the increasing demands in this niche [24,29].
ing capabilities [55]. Hence, industrial processes with ATPS could The effectiveness of ATPAP has been proved on a wide variety of
represent an alternative to affinity chromatography or traditional biologicals, primarily macromolecules, cell fragments and whole
chromatography in order to reduce operating costs and facilitate cells. Fig. 2 shows the correlation among the absolute change in
process intensification. The biocompatibility and low toxicity of ln Kp and average particle diameter (nm) of different types of bio-
renewable polymers make them attractive constituents of ATPS logical products (specific characteristics of the products included in
related techniques. Currently, there is a clear interest for studying Fig. 2 are shown in Table 6) when ATPAP strategies are used instead
the use of new polymers such as natural occurring polysaccharides, of traditional ATPS. Most of the proteins that have been studied on
like starch. The use of guar gum, a non-ionic storage polysaccha- ATPAP have a molecular weight lower than 100 kDa (besides anti-
ride from guar plants, has become a primary focus of study because bodies), while nucleic acids are between 100 and 125 kDa. Small
of its neutral chemistry and biological activity when used in aque- cells and cell fragments (primarily membranes) have been effec-
ous two-phase systems [42]. Hydroxypropyl starch and guar gum tively recovered in ATPAP systems constituted by two polymeric
12 F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13

Fig. 2. Correlation of absolute change in ln Kp with the nature and average particle size of selected examples of macromolecules recovered in polymer–polymer and
polymer–salt ATPAP systems. Despite molecular nature and particle size, ATPAP methodologies can be widely adapted for recovery and purification of molecules in
bioprocesses.

Table 6
Selected examples of biomolecules partitioned in ATPAP systems according to increasing particle size.

Number System Ligand Product Size Reference

1 PEG 8000-phosphate Gluthatione Thaumatin 3.5 nm [71] [20]


2 PEG 4000–Phosphate Benzoate Penicillin Acylase 5.5 nm [18] [19]
3 PEG 10000–dextran 40 PEG–Cu-IDA (DSa ) pDNA 2686 bp [72] [24]
4 PEG 600–dextran 40 GST-zinc finger Protein (DS) pDNA 2693 bp [29] [29]
5 PEG 3350–dextran 500000 Glutaric acid (FL) IgG 28 nm [73] [3]
6 PEG 3350–dextran 500000 Diglutaric acid IgG 28 nm [73] [5]
7 PEG 3350–dextran 500000 Avidin Biotinylated Liposomes 45 nm [74] [22]
8 PEG 3350–dextran 2000 NeutrAvidin Microsomal membranes 150 nm [74] [23]
9 PEG 8000–dextran 500000 Anti-human IgG Erythrocytes 7700 nm [75] [70]
10 PEG 8000–dextran 500000 Anti-mouse IgG NS-1 Cells 13,000 nm [76] [69]

Examples presented in this table refer to the selected molecules from Fig. 2.
a
DS, double stranded.

phases. As can be observed in Fig. 2, for these selected biological the addition of a displacer may be used to promote desorption of
products the ln Kp values range from >1 to <9, demonstrating the molecule of interest from the affinity ligand. It is expected that
that regardless identity, nature and size of the biological products efficient displacement methodologies will be developed and ATPAP
a specific partitioning can be achieved using ATPAP methodologies. will gain increased popularity among downstream processing unit
It would be of great interest to study the recovery and purification operations.
potential of ATPAP for several other macromolecules of commercial
interest such as phytochemicals, pigments, hormones, antifreeze
proteins, phycobiliproteins, transcription factors, polysaccharides 7. Conclusions
and tandem/fusion proteins. In addition, recovery of bionanoparti-
cles such as dendrimers, organelles and virus-like particles would Over the last two decades, affinity partitioning in liquid–liquid
suppose an interesting field for ATPAP partitioning. In this context, extraction systems has experienced a marked growth. Since almost
research groups and biotechnological industries could apply these every biological molecule exhibits a natural affinity towards a par-
new methodologies in order to design bioprocesses that involve ticular substance, general ATPAP protocols may be developed for
less unit operations [62–67], hence increasing purity and over- a wide variety of organic particles, which could possibly trigger
all recovery yield. Furthermore, the recovery and purification of new research areas within downstream processing. One mayor
low molecular weight compounds of commercial interest, such as trend in ATPAP systems is using new raw materials as phase form-
nutraceuticals and natural pigments on ATPAP may represent an ing components, emphasizing eco-friendly polymers. In addition,
interesting new field to explore. development and implementation of new molecules as affinity
Regarding the isolation of the biomolecules of interest after ligands that have not been used in affinity partitioning will sup-
ATPAP, strategies for chemical or biological displacement of the tar- pose an important area of opportunity in the near future. Major
get molecule from the affinity ligand should be further studied and challenges in ATPAP, such as the optimization of displacement
optimized. Desorption methodologies have been poorly addressed protocols, the characterization of novel and efficient chemical mod-
or emphasized on the available scientific literature. Traditional ification methodologies, and the cost of some affinity ligands, may
methodologies including ionic force shifting, pH adjustment, and be addressed in the next few years. This could probably increase
F. Ruiz-Ruiz et al. / J. Chromatogr. A 1244 (2012) 1–13 13

ATPAP popularity in the development of novel downstream strate- [35] S. Teotia, M.N. Gupta, J. Chromatogr. A 1025 (2004) 297.
gies for recovery and purification of biological products. Based on [36] S.P. Graether, C.M. Slupskym, B.D. Sykes, Biophys. J. 84 (2003) 552.
[37] A. Wierzbicki, P. Dalal, T.E. Cheatham, J.E. Knickelbein, A.D.J. Haymet, J.D.
the advantages described herein it can be concluded that ATPAP Madura, Biophys. J. 93 (2003) 1442.
may be an attractive alternative to traditional chromatographic [38] M. Bar, Y. Celik, D. Fass, Cryst. Growth Des. 8 (2008) 2954.
techniques at industrial scale. [39] M. Bar, R. Bar-Ziv, T. Scherf, D. Fass, Protein Expr. Purif. 48 (2006) 243.
[40] Y.C. Liou, M.E. Daley, L.A. Graham, C.M. Kay, V.K. Walker, B.D. Sykes, P.L. Davies,
Protein Expr. Purif. 19 (2000) 148 (SEC).
Acknowledgments [41] Z. Li, F. Xiong, Q. Lin, M. d‘Anjou, A.J. Daugulis, D.S.C. Yang, C.L. Hew, Protein
Expr. Purif. 21 (2001) 438 (HIC).
[42] M. Antov, L. Anderson, A. Andersson, F. Tjerneld, H. Stalbrand, J. Chromatogr. A
The authors want to thank Tecnológico de Monterrey Research 1123 (2006) 53.
chair (Grant CAT161) and CONACyT (the National Council for Sci- [43] H. Lam, M. Kavoosi, C.A. Haynes, D. Wang, D. Blankschtein, Biotechnol. Bioeng.
ence and Technology in Mexico) for the scholarship #295368. 89 (2005) 381.
[44] M. Roobol-Bóza, V. Dolby, M. Doverskog, A. Barrefelt, F. Lindqvist, U.C. Opper-
man, K. Köhler Van Alstine, F. Tjerneld, J. Chromatogr. A 1043 (2004) 217.
References [45] Y.J. Nikas, C.L. Lui, T. Srivastava, N.L. Abbott, D. Blankschtein, Macromolecules
25 (1992) 4797.
[1] P. Mazzola, A. Lopes, F. Hasmann, A. Jozala, T. Penna, P. Magalhaes, C. Rangel- [46] D. Xiao-Yan, M. Yao, F. Xu-Dong, S. Yan, Biotechnol. Progr. 26 (2009) 150.
Yagui, A. Pessoa Jr., J. Chem. Technol. Biotechnol. 83 (2008) 143. [47] A. Nilsson, M. Mannesse, M. Egmond, F. Tjerneld, J. Chromatogr. A 946 (2002)
[2] J. Benavides, O. Aguilar, B. Lapizco-Encinas, M. Rito-Palomares, Chem. Eng. 6400.
Technol. 31 (2008) 838. [48] S. Sturesson, F. Tjerneld, G. Johansson, Appl. Biochem. Biotechnol. 26 (1990)
[3] A.M. Azevedo, P. Rosa, I.F. Ferreira, A. Pisco, J. de Vries, R. Korporaal, T.J. Visser, 281.
M.R. Aires-Barros, Sep. Purif. Technol. 65 (2009) 31. [49] I.F. Ferreira, A.M. Azevedo, P.A.J. Rosa, M.R. Aires-Barros, J. Chromatogr. A 1195
[4] J. Benavides, M. Rito-Palomares, J. Chem. Technol. Biotechnol. 83 (2008) 133. (2008) 94.
[5] P. Rosa, A.M. Azevedo, I.F. Ferreira, J. de Vries, R. Korporaal, H.J. Verhoef, T.J. [50] M. Arnold, H. Bittermann, B. Kalbfuss-Zimmermann, T. Neumann, K. Schmidt,
Visser, M. Aires-Barros, J. Chromatogr. A 1162 (2007) 103. R. Sekul, F. Hilbrig, H. Ludolph, R. Freitag, J. Chromatogr. A 1218 (2011) 4649.
[6] B. Maestro, I. Velasco, I. Castillejo, M. Arévalo-Rodríguez, A. Cebolla, J.M. Sanz, [51] H. Yang, P. Gurgel, R.G. Carbonell, J. Chromatogr. A 1216 (2009) 910.
J. Chromatogr. A 1208 (2008) 189. [52] L. Yung-Chuan, C. Chih-Chiang, S. Shing-Yi, J. Chromatogr. B 794 (2003) 67.
[7] J. Benavides, M. Rito-Palomares, J. Chromatogr. B 807 (2004) 33. [53] K. Schriebl, S. Lim, A. Choo, A. Tscheliessnig, A. Jungbauer, Biotechnol. J. 5 (2010)
[8] T. Hernández-Mireles, M. Rito-Palomares, J. Chem. Technol. Biotechnol. 81 50.
(2006) 989. [54] A.F. Sousa, P.Z. Andrade, R. Pirzgalska, T. Galhoz, A.M. Azevedo, C.L. da Silva,
[9] M. Rito-Palomares, A. Lyddiatt, Chem. Eng. J. 87 (2002) 313. M.R. Aires-Barros, J.M.S. Cabral, Biotechnol. Lett. 12 (2011) 2373.
[10] S.C. Silverio, P.P. Madeira, O. Rodrıguez, J.A. Teixeira, E.A. Macedo, J. Chem. Eng. [55] P.A.J. Rosa, A.M. Azevedo, S. Sommerfeld, W. Bäcker, M.R. Aires-Barros, Biotech-
Data 53 (2008) 1622. nol. Adv. 29 (2011) 559.
[11] P. Hye-Mee, L. Sang-Woo, C. Woo-Jin, K. Yoon-Mo, J. Chromatogr. B 856 (2007) [56] M. Trexler-Schmidt, S. Sze-Khoo, A.R. Cothran, B.Q. Thai, S. Sargis, B. Lebreton,
108. Biopharm. Int. (Suppl.) (2009) 8.
[12] J. Schindler, H.G. Nothwang, Proteomics 6 (2006) 5409. [57] A.A. Shukla, B. Hubbard, T. Tressel, S. Guhan, D. Low, J. Chromatogr. B 848 (2007)
[13] M. Rito-Palomares, J. Chromatogr. B 807 (2004) 3. 28.
[14] A. Cordes, J. Flossdorf, M.R. Kula, Biotechnol. Bioeng. 30 (1987) 375. [58] W.L. Hinze, E. Pramauro, Anal. Chem. 24 (1993) 133.
[15] S.D. Flanagan, S.H. Barondes, J. Biol. Chem. 250 (1975) 1484. [59] L. Yang, D. Xiao-Yan, S. Yan, Sep. Purif. Technol. 53 (2007) 289.
[16] M. Rito-Palomares, J. Microbiol. Biotechnol. 12 (2002) 535. [60] Y. Xu, M. Vitolo, C.N. Albuquerque, A. Pessoa Jr., J. Chromatogr. B 780 (2002) 54.
[17] A. Persson, B. Jergil, FASEB J. 9 (1995) 1304. [61] M. Mannesse, R. Cox, B. Koops, H. Verheij, G. De Haas, M. Egmond, H. van der
[18] M.R. Gavasane, V.G. Gaikar, Enzyme Microb. Technol. 32 (2003) 665. Hijden, J. de Vlieg, J. Biochem. 34 (1995) 141.
[19] D.M. Head, B.A. Andrews, J.A. Asenjo, Biotechnol. Tech. 3 (1989) 27. [62] X. Yan, A. Souza, M. Pontes, M. Vitolo, A. Pessoa Jr., Braz. Arch. Biol. Technol. 46
[20] B.A. Andrews, D.M. Head, P. Dunthorne, J.A. Asenjo, Biotechnol. Tech. 4 (1990) (2003) 741.
49. [63] P.A. Rosa, A.M. Azevedo, I.F. Ferreira, S. Sommerfeld, W. Bäcker, M.R. Aires-
[21] T. Gouveia, B. Kilikian, J. Chromatogr. B 743 (2000) 241. Barros, J. Chromatogr. A 1216 (2009) 8741.
[22] L. Ekblad, J. Kernbichler, B. Jergil, J. Chromatogr. A 815 (1998) 189. [64] L.O. Persson, B. Olde, J. Chromatogr. 457 (1988) 183.
[23] I.R. Barinaga-Rementeria, S. Mebrahtu, B. Jergil, J. Chromatogr. A 971 (2002) [65] M. Lu, F. Tjerneld, J. Chromatogr. A 766 (1997) 99.
117. [66] L. Ekblad, B. Jergil, J.P. Gierow, J. Chromatogr. B 743 (2000) 397.
[24] H.S.C. Barbosa, A.V. Hine, S. Brocchini, N.K.H. Slater, J.C. Marcos, J. Chromatogr. [67] H. Everberg, J. Clough, P. Henderson, B. Jergil, F. Tjerneld, I. Barinaga-
A 1217 (2010) 1429. Rementeria, J. Chromatogr. A 1118 (2006) 244.
[25] J. Harris, E. Struck, M. Case, S. Paley, J. Polym. Sci. 22 (1984) 341. [68] J.H. Kyung, T. Kajiuchi, J. Park, K. Park, H. Kim, H. Lee, Y. Kim, Appl. Chem. 7
[26] E. Laboureau, Vijayalakshmi, J. Mol. Recognit. 10 (1997) 262. (2003) 105.
[27] L. Yuan-Qing, S. Sai-Nan, C. Branford-White, Z. Li-Min, Sep. Purif. Technol. 73 [69] R. Hamamoto, M. Kamihira, S. Iijima, J. Ferment. Bioeng. 82 (1996) 73.
(2010) 665. [70] K. Sharp, M. Yalpani, S.I. Ioward, D. Brooks, Anal. Biochem. 154 (1986) 110.
[28] S. Teotia, R. Lata, M.N. Gupta, J. Chromatogr. A 1052 (2004) 85. [71] Y. Kuznetsov, J. Konnert, A. Malkin, A. McPherson, Surf. Sci. 440 (1999) 69.
[29] H. Barbosa, A.V. Hine, S. Brocchinic, N.K.H. Slater, J.C. Marcos, J. Chromatogr. A [72] M. Fitzgerald, P. Skowron, J. Van Etten, L. Smith, D. Maed, Nucleic Acids Res. 20
1206 (2008) 105. (1992) 3753.
[30] J.M.T. Hamilton-Miller, Bacteriol. Rev. 30 (1966) 761. [73] C.J. Roberts, P.M. Williams, J. Davies, A.C. Dawkes, J. Sefton, J.C. Edwards, A.G.
[31] A. Bruggink, E. Roos, E. de Vroom, Org. Process Res. Dev. 2 (1998) 343. Haymes, C. Bestwick, M.C. Davies, S.J.B. Tendler, Langmuir 11 (1995) 1822.
[32] A.V. Ilyina, V.P. Varlamov, V.E. Tikhonov, I.A. Yamskov, V.A. Davankov, Biotech- [74] K. Kawajiri, A. Ito, T. Omura, J. Biochem. 81 (1977) 779.
nol. Appl. Biochem. 19 (1994) 199. [75] J. Hainfield, T. Steck, J. Supramol. Struct. 6 (1977) 301.
[33] B.S. Dhoot, J.M. Dalal, V.G. Gaikar, Sep. Sci. Technol. 42 (2007) 1859. [76] Y. Fukushima, S. Hagiwara, Proc. Natl. Acad. Sci. U.S.A. 80 (1983) 2240.
[34] S. Fernandes, G. Johansson, R. Hatti-Kaul, Biotechnol. Bioeng. 73 (2001) 465.

You might also like