The concept of affinity chromatography emerged in the 1970s and exploits the reversible, biological and selective interaction between two entities such as an immobilized adsorbent on a solid support, denominated as affinity ligand, and a target biomolecule that will be further purified [6-7, 12-14]
The solid support type most widely used in affinity chromatography is non-porous materials being example of these agarose, polymethacrylate, polycrylamide, cellulose, and silica. Recently, other materials such as membranes, monolithic and expanded-bed adsorbents have been used as matrix for affinity chromatography [12-16]. The properties required for an efficient and effective chromatography matrix includes chemical inertness and stability, mechanical stability, pore size and particle size [12-13, 15-16]. The chemical inertness of the matrix is related with the minimization of nonspecific binding of the matrix material and the target protein that has been purified [12, 16]. As the support only can interact in a minimal way with the aqueous mobile phase, the matrix support should be hydrophilic and non-charged to minimize the formation of ionic interactions [16]. Concerning the stability of the material, the matrix must be physically and chemically stable under a wide range of conditions including coupling, adsorption and elution, extreme pH values, high and low temperatures and the use of organic solvents, enzymes present in the sample, detergents and disruptive eluents conditions[12-13, 15-16]. Some of conditions are those that are employed in the sterilization in place (SIP) and cleaning in place (CIP) so that the supports can be re-used, contributing for the cost-effectiveness of the purification process. In terms of mechanical stability, the matrix should be preferably rigid solid supports so that can be able to withstand high pressures during the purification process without compressing and deforming [12, 14-16]. Other requirements of an ideal matrix are related with the pore size and particle size. The gel particles must be uniform, spherical and rigid, because the beads with these characteristics provide excellent flow through properties with minimal channelling in the column applications [12, 14-17]. Moreover, the gel particles should be smaller to allows a greater surface area of the support material and because limits mass transfer effects [12, 14]. The porosity of the matrix also contributes for good flows properties and facilitates the penetration of macromolecules [12]. The high porosity also combined with small particles size, increases the surface area which reflects usually in a high capacity of the support matrix [12-14, 16]. Moreover, the pore diameter also has a significant effect on the purification step taking into account the Renkin equation [16, 18] that estimates the effective diffusion coefficient. The pore size limits the entrance of the target protein inside the pore according with the size, molecular weight, being this denominated exclusion limit. According to this, the size of the pores should be at least five times larger than the average size of a biomolecule. An example of this can be if the size of the pores should be at least 300 AËš or greater if assuming an average protein size of 60 AËš [16, 18]. The matrix selectivity can be introduced through the functionalization of the solid support with different coupling strategies generating different affinity ligands.
The affinity ligands immobilized onto a solid support can be divided into three main groups such as biological, synthetic and structural ligands (Fig.1.1). Biological ligands comprise ligands from biological sources or from in vitro techniques. These are natural receptors which target molecules with high selectivity and affinity, such as peptides, antibodies, antigens and binding or receptor proteins [19-21], and are associated with high costs of production and purification, poor stabilization under sterilization and cleaning-in-place conditions, as well as potential leakage and end-product contamination. The most common affinity biological partners involve the immobilized bacterial immunoglobulin-binding domains such as staphylococcal protein A (spA), G and L for immunoglobulin purification [22-26]. The avidin-biotin technology also has been employed on the affinity based bioseparation [27] and the purification of glycoprotein and sugars was performed through immobilized natural lectins [28]. Heparin is also used as biological ligand for affinity purification of growth factors [29].
Synthetic affinity ligands have been developed in an attempt to overcome these disadvantages of natural ligands. They tend to combine molecular recognition features with high resistance to chemical and biological degradation, high scalability, as well as low costs and low toxicity [7, 19, 21]. Synthetic ligands have been well established over many decades and one of the major class are related with the biomimetic ligands or de novo designed ligands. These last ligands are tailor-made molecules that mimic the natural biological recognition between a target protein and a natural ligand [20]. The major concern about these biomimetic ligands is regarding the toxicity and biocompatibility of these ligands in biopharmaceutical industry for the protein purification [20]. The research strategy followed for the development of these ligands involves ligand design using in silico molecular modelling tools as a first step [20, 30], followed by the on-bead combinatorial synthesis and high-throughput screening (HTS) of ligand libraries (Fig. 1). The combinatorial chemistry can be used either in solid or solution phase. Both of the synthesis allows to increase the diversity of the compounds and also present other advantages as the cost effectiveness and the less time consuming [31]. Different scaffolds were already used for the development of synthetic ligands. The triazine scaffold is a well-established technique by the Lowe and co-workers. This scaffold was in the basis of an alternative of protein A by generating an affinity ligand that mimics protein A for the purification of immunoglobulins G (IgG) [32]. Also other bacterial immunoglobulin-binding domains such as protein G and L with the same purification purposes [33-34]. Beyond immunoglobulins purification, a wide range of affinity ligands were developed such as an artificial lectin and receptor for purification of glycoproteins [35-36], as well as other protein targets such as prion proteins [37], human recombinant factor VII, [38] and elastases [39]. This strategy was also considered for other authors for the purification of proteins as alkaline phosphatase [40] and human tissue plasminogen activator [41]. More recently, a different scaffold based on the Ugi reaction have been employed [42]. The Ugi reaction is a multicomponent reaction that involves four main compounds as an aldehyde, an amine, a carboxylic acid and an isocyanine. This Ugi reaction scaffold applied to affinity chromatography was firstly studied by Lowe et al, indicating the potential of this MCR on this field. So far, this strategy has been used mainly for the purification of immunoglobulins [42-44]. Moreover, the identification of novel affinity ligands can be also possible through phage display [45]. This technology has been developed by Smith et al by which the peptides or proteins were expressed on the surface of the phage particles. For this, DNA sequence is inserted onto the genome of a phagemid vector along with the C-terminus of the page minor coat protein, gene III, through recombinant DNA technology [45-46]. Then the cloned phagemid vector is transformed in Escherichia coli cells, and then infected with helper phage for the production of combinatorial libraries of filamentous phage with peptides displayed at surface [45, 47]. The size of the initial combinatorial library should be up to 109 clones so that could be further used in the panning for the selection of putative binders [45, 47]. The panning can comprise several cycles of binding, elution and amplification, and the increase of enrichment of phage particles over several rounds of panning can imply the existence of a progress affinity and selectivity between the putative binders displayed on phage particles and the specific target [45]. The most used phagemid vectors used are the filamentous bacteriophage and its Ff class such as M13, f1 and fd [46] and a wide range of combinatorial libraries of antibodies fragments [47-48], protein domains [49-50] and peptides [46, 51-52] have been constructed. The applications of phage display on affinity chromatography based protein purification have been described in literature [45, 53-55] and comprise the construction of combinatorial libraries based on peptides ligands for factor VIII purification [56]. Moreover, small protein domains (affibodies) based on α-helical bacterial receptor domain Z derived from spA (immunoglobulin G-binding domain), have been selected through phage display with a micromolar dissociation affinity constant for specific targets, such as Taq DNA polymerase, human insulin and apolipoprotein A-1 [49]. Additionally, the construction of combinatorial libraries based on these molecules affibodies have been extended for other targets, among them human immunoglobulin A (IgA) [57], human epidermal growth factor receptor [58] and human amyloid beta peptides implicated on Alzeihmer disease [59]. Although these molecules present highly potential to be employed as affinity ligands for purification of proteins based on affinity chromatography, these molecules have been exploited mainly for therapeutic, diagnostic and imaging applications [60]. However, Z domain is still on the basis of development of affinity ligands, where protein engineering of this domain has been performed to improve IgG purification [61]. Despite of phage display is a versatile technique that provides enormous diversity on potential binding partners at reduced costs, this also presents some limitations such as limited folding properties, stability and high product yields of the displayed binding partners [62]. Different display techniques can also be used, among them ribosome and yeast display. The ribosome display technology can overcome some disadvantages of phage display such as the dependency of the bacterial transformation efficiency to generate diversity on the combinatorial libraries and also improve the stability and folding properties of the displayed targets [62-63]. The yeast display also offers benefits regarding the proper folding of proteins as these hosts present similar post-translational machinery than mammalian cells required for the correct folding of eukaryotic proteins [64].
Another class of affinity ligands is related with structural ligands, presenting these ligands limited selectivity/affinity with production at affordable prices. These ligands comprise ion-exchange [65-67], hydrophobic [68], metal chelate [69] , covalent and thiophilic ligands [13, 20, 70]
The biological and structural ligands have been also used in purification based on affinity technologies that involves the appliance of a selective binding partner designated as affinity tag. The affinity tags displays different size range from a single amino acid to entire proteins, and can be genetically fusion to N- or C-terminal of the target biomolecule [71-76]
Afterwards, the affinity tags will recognize the affinity ligand immobilized onto the matrix of the chromatographic column, facilitating the purification process. Despite of the affinity tags being highly efficient tools for the purification of recombinant proteins through chromatography based processes, the tags can display other applications on enhancement of protein solubility and stability [77], increase of the expression levels [75, 77] and allows labelling for cellular localization and imaging studies [73]. The use of peptides as affinity tags can be more advantageous rather than protein tags because due to their smaller size, the peptides tag can contribute for a less burden for the host during the fusion protein production and might interfere less with the tertiary structure and biological activity of the target fused protein [72, 78-79]. Therefore, the removal of the peptide tags might not be extremely necessary, decreasing thus the overall costs of the purification process based on affinity peptide tags [72, 78]. Regarding therapeutic proteins and due to strict regulatory demands on integrity and biological activity, the presence of the affinity tag can compromise the protein properties and requires the removal of the tag [72-73, 75-76, 80], which will be further discussed.
The biological ligands involved on molecular recognition with the respective partner affinity tag can comprise peptides, protein and carbohydrates. The most common biological ligands used are based on matrices with the immobilized streptadivin and antibody proteins. Regarding the IgG affinity chromatography, the first affinity tag developed was based on intrinsic selectivity and affinity between the bacterial immunoglobulin-binding domain staphylococcal protein A and the Fc region of the mammalian IgG [81]. Usually SpA is well-known to be used as an immobilized biological adsorbent for the purification of immunoglobulins; however the application of this protein on affinity chromatography was extended to their use as an affinity tag and therefore, SpA fusion constructs were developed and demonstrated with the fusion of SpA to alkaline phosphatase with the further expression in bacterial cells such as E. coli and Staphylococcus aureus [82]. Afterwards, the purification of the target protein was conducted as one-single step of IgG based affinity chromatography with protein recovery at acidic pH [82]. According to Nilsson et al [81], the SpA presents five homologous domains such as E, D, A, B and C, where the IgG presents a significant dissociation constant for domain B. Considering this, a mutated version of B domain was developed and denominated as Z domain to improve the resistance of undesirable cleavage of the purified fusion protein when using chemical tag removal strategy [83]. The bacterial immunoglobulin-binding domain staphylococcal protein G (SpG) also have been exploited as a fusion partner due to their bifunctional behaviour as these protein presents different domains that can present affinity for both IgG and serum albumin (HA) [84-85]. The SpG is composed by four different regions (A, B, C and D), being the regions C and G responsible for the binding selectivity of IgG [84-85]. Therefore, the SpG have been used as a fusion partner for the purification of recombinant proteins through human serum albumin (HSA) and IgG affinity chromatography [72, 74, 84-85].
Other affinity tags that recognized antibody based matrices are the FLAG, c-myc, T7, hemaglutinin antigen (HA) and Sogtags. A common feature between of all these tags is the fact that these peptide tags are epitopes with strong affinity for the respective antibodies, thus presenting antigenic properties [72, 76].
Most of these affinity pairs present a limited utility on purification processes due to the high costs of the adsorbents that are based on monoclonal antibodies. Moreover, these epitope peptides can also compromise the proper production of the target protein in terms of functionality [107-108]. Within all these antigenic peptides, the most widely used peptide is the FLAG affinity tag that have been employed on the purification of several recombinant proteins such as immunoglobulins, cytokines, gene-regulatory proteins [109]. The FLAG peptide is a hydrophilic peptide with five amino acid sequence DYKDDDDK with high affinity for the monoclonal antibodies M1 and M2, being the binding calcium dependent for M1 [95, 109-111]. Therefore, the elution of recombinant proteins by using M1 can be carried out by using a mild conditions with the presence of metal chelates such as EDTA, where in case of M2 is at lower pH [95, 109-111]. The most disadvantages of this system are extended to all purification processes based on biological ligands such as ligand instability and therefore leakage [109]. An attractive feature of this affinity tag is related with the tag removal, because the sequence DDDDK of the FLAG tag can be recognized by the enterokinase, an endopeptidase used on enzymatic tag removal strategies without the need of insertion of an additional sequence for further tag removal [109, 111]. Moreover, the use of this tag allows the release of the target protein without additional amino acids [109, 111].
The c-myc is an product of a proto-oncogene, and the derived epitope from this product was found to present high affinity for the monoclonal antibody 9E10 [96]. Then, this affinity pair has been mostly used as a tool for the detection of recombinant proteins through immnunoblotting assays rather than for purification processes [78, 96, 112]. Also the affinity tags T7 and hemaglutinin antigen (HA) are also purification affinity tags, being mostly used on immune-detection assays [76-77], where the T7-tag corresponds to 11 amino acid sequence being a leader peptide of phage T7 with affinity for the anti-T7 monoclonal antibody [108, 113] and the HA tag is a peptide epitope of the influenza virus hemagglutinin [114] that is recognized by the monoclonal antibody 12 CA5 [98, 115]. The epitope tags denominated as Softags are employed on immunoaffinity chromatography by using polyols responsive monoclonal antibodies as biological adsorbents [99, 116]. According to their designation, these tags allows a soft elution by using the polyols, low molecular weight compound, as a eluents agent [99, 116], being a great advantage despite of the high affinity. There are three Softags, where the Softag 1 corresponds to a thirteen amino acid sequence near the C-terminal of the β' subunit of E.coli RNA polymerase [102]; Softag 2 is a repeat heptapeptide found on C-terminal of RNA polymerase I [100]; and Softag3 is an epitope near the N-terminal of human transcription factor [99].
Other type of affinity tags is those that recognize streptadivin binding domains. These affinity tags were developed based on the natural binding pair avidin-biotin that displays a high selective and affinity constant of 1015 M-1 [27]. The first affinity tag developed was Strep-tag, a nine amino acid peptide that was selected from a random peptide library generated towards affinity for streptadivin [89], being fused to the C-terminal of the target protein. According to structural studies, this affinity tag recognizes and binds to the same pocket of biotin, the natural partner of streptadivin [90]. The main advantages of this affinity tag are related to the resistance to proteolysis in vivo, no interference with expression in E.coli and the elution conditions employed after binding on streptadivin based affinity column are mild as this step is carried through competitive elution with a biotin or analogue compound such as iminibiotin [89]. However, a limitation of the Strep-tag was their restricted fusion to the C-terminal of the target protein and due to this, an improved version of Strep-tag was developed and denominated as Strep-tag II with equilibrium dissociation constant of 37 µM [90]. Also, at the same time, an engineered streptadividin chromatographic support (Strep-Tactin) was developed to improve the binding capacity towards the Strep-tag II, through the generation of a genetic library based on streptadivin with random mutagenesis on the amino acids 44-47 positions that corresponds to flexible loop region near to the binding site [117]. Although, the affinity pair Strep-tag II - Strep tactin have been extensively used on a variety of applications [117], continuously studies on development of most robust affinity tags with high affinity for streptadivin matrices have been conducted. This affinity tags comprises strepdavidin-binding peptide (SBP) ans Nano tag. The SPB tag comprises a 38 amino acid sequence with an equilibrium dissociation constant of 2.5 nM and this peptide sequence was found from the selection of a peptide library for the immobilized streptadivin [92], presenting a great advantage over the Strep-tag II. Nano-tag displays even more robust properties that the tags already described in literature to be used on streptadividin based affinity chromatography as this tag combines the small size of Strep-tag II and also a namolar affinity constant of 4 nM [91]. This tag is a 15 amino acid peptide was developed to bind specifically to streptadivin. Therefore, a synthetic library based on the heart fatty acid-binding protein (FABP) was created and the selection was performed through ribosome display against immobilized streptadivin [118]. Other affinity tag based on Strep-tag was developed, however this affinity designated as AviD-tag displays affinity for neutravidin, a neutral form of avidin rather for streptadivin [94]. The AviD-tag is composed by a 6-amino acid cyclic peptide that was selected through phage display technique with a dissociation constant of 12 µM for both Neutravidin and avidin [94].
Besides the affinity tags used on the immunoaffinity chromatography, and the affinity tags based on streptadividin binding proteins, there are other affinity tags that are also recognized by peptides and protein based ligands. These affinity tags comprise calmodulin binding peptide (CBP), S-peptide tag, glutathione-S-transferase (GST) and maltose binding protein (MBP). The CBP is a 26 amino acid peptide sequence derived from the carboxyl-terminal of rabbit skeletal muscle myosin light chain kinase [87, 119]. The CMP affinity tag displays a high dissociation affinity constant of 10-9 for the calmodulin, revealing a strong binding interaction but calcium dependent [87, 119-120]. The main advantages related with this affinity pair are regarding to the properties of the calmodulin affinity resins, as this peptide is small (17KDa), stable and allows the production of calmodulin affinity resins at affordable prices. Moreover, this affinity tag already has been fused on N-terminal of a wide range of recombinant proteins that were further produced with high levels of expression. Moreover, the elution can be carried out at mild conditions with calcium chelating agents such as EDTA and EGTA [87, 119-120]. The enzymatic cleavage of the ribonuclease A by the protease subtilisin leads to two products: the S peptide tag and S-protein [121]. These two fragments are on the basis of purification of recombinant proteins through affinity chromatography, where the S-peptide, a fifteen amino acid sequence, is used as an affinity tag that will further recognize the S-protein immobilized on resin [88]. The major advantages are related with the small size of the affinity tag and this affinity pair displays a nanomolar dissociation affinity constant [88]. Despite the principal aim of the use of affinity tags is for purification and detection purposes, the affinity tags can also enhance solubility of the fusion partner. The solubility of a recombinant protein when produced in bacterial hosts such as E.coli can be compromised, leading to the formation of protein aggregates [122-123]. These aggregates, denominated as inclusion bodies (IBs), are unfolded or partially folded proteins [124]. Therefore, in order to address these challenges, the target proteins can be fused to solubility affinity tags to improve their solubility during protein expression, even if the mechanism is not fully understood [75-77].
The GST tag presents both properties such as a solubility enhancer and allows protein purification through the glutathione immobilized on affinity matrices [86]. The GST protein is a monomeric protein (26 KDa) originated from Schistosoma japonicum and belongs to a family of enzymes that catalyze the reaction between nucleophile reduced glutathione and electrophilic compounds [125-126]. Smith et al. have constructed plasmid expression denominated as pGEX vectors for the production of fusion proteins in E.coli host by using GST tag fused on N-terminal of the target protein, being this expression vector widely used for the production recombinant proteins [127-128]. This affinity pair presents several advantages on their use such as the fact the tag can protect and stabilize the target protein. Regarding the ligand point of view, the glutathione resins are cost effectiveness and the use of excess reduced glutathione as a competitive agent becomes the elution conditions milder [126]. However, the major drawback during the elution is that if the reducing conditions are not guarantee, the fusion protein can undergo oxidative aggregation due to the existence of four cystein residues exposed at surface of the subunit GST tag [129]. The other main limitation of this affinity tag is that the solubility of the fusion proteins expressed can be compromised by the total molecular weight of the fusion protein, in other words, fusion proteins with higher molecular weights than 100 KDa can lead to partially or completely insoluble proteins [128].
The other class of biological ligands that recognize affinity ligands are the carbohydrate based ligands. These ligands comprise cross-linked amylose, cellulose, and chitin among others. The greatest advantages of these affinity ligands is that these carbohydrates displays high affinity for their respective affinity tags, the affinity matrices are inexpensive and prepared in a simple manner, which contributes for the ease of scale-up purification process.
The MBP is an affinity tag that besides facilitating the purification of recombinant proteins through molecular recognition to carbohydrate amylose, also can enhance the solubility of the fusion partner. MBP is a 42 KDa periplasmic protein that is engaged on the maltose transport system of E. coli, where its principal role is related with the transport of maltose and maltodextrins across the bacterial cytoplasmic membrane [130-131]. The MPB tag fused on N-terminal target proteins can be purified in one-single step of chromatography by using a cross-linked amylose affinity resin [132]. The MBP tag presents a high affinity for the amylose, a maltose analogue, with a similar dissociation affinity constant (35 x 107 M) to maltose, the elution competitive agent [130, 133]. This presents one of the great advantages of this affinity pair because the elution of the target fusion protein can be carried out under mild conditions with 10 mM of maltose [132]. Other advantage is related with the affinity tag, because this tag does not present any cystein residue and therefore do not interferes with presence of disulfide bonds of the target protein [75, 132]. Previous studies also have demonstrated that MBP is most effective solubilizing agent when compared to other solubility affinity tags such as GST and thioredoxin (Trx),that will be described afterwards, being able to as a molecular chaperone for the fusion partner [134]. Also for this affinity tag, a plasmid vector was constructed for the expression of recombinant proteins fused on C-terminal of the GST affinity tag, denominated as pMAL [135-136]. This vector also contains a sequence coding for an endopeptidase that will be further used after the purification of recombinant proteins using for the tag cleavage. Several vectors have been developed for the production of fusion proteins with different recognition sites for Factor Xa and enterokinase [135-136].
The cellulose and chitin binding proteins present high affinity for the polysaccharides cellulose and chitin respectively. The cellulose binding proteins can be found in a widely range of carbohydrolases and can vary on their size from small (33-36 aa) to large domains with more than 180 amino acids [104]. However, as these domains are employed as an affinity tags, the smaller domains are most suitable for these applications to benefit of the smaller affinity tag advantages. These domains display considerable affinity dissociation in the low micromolar range for cellulosic supports [104]. These supports are commercially available in a wide range of supports among them fibers, beads, membranes, hydrogels with a diverse extent of porosity, polymerization and functionalization. Other great advantages between the common polysaccharides that contribute for the effectivess of this adsorbents on affinity based purification are the fact these supports are non-toxic, inert, stable with exceptional physical properties and presents non-specific binding adsorption [104, 106, 137]. The chitin binding domains present also a great impact on the purification of recombinant proteins by using this type of domain as an affinity tag for further purification of the fusion protein on a chitin based matrix [105]. These domains present a high affinity for the chitin, the second most abundant polymer after cellulose, being part of the enzymes involved on the biodegradation of chitin and chitosan [137]. Also starch binding domains already have been employed as an affinity tags, where these domains were fused to N-terminal of β-galactosidase and then purified by using starch granules [106]
Regarding the remaining classes of affinity ligands, structural and synthetic ligands, the majority of the ligands that have been used as binding partners of other affinity tags are the structural ligands being mainly cation and anion-exchange, metal chelate and hydrophobic ligands. So far, the synthetic ligands based on different chemistries have been only employed on the purification of a specific biomolecule such as immunoglobulins. The most widely structural ligand used on the purification is the metal chelate ligands and the His-tag as the respective binding partner. This type of chromatography is designated as immobilized metal affinity chromatography (IMAC).
The concept of IMAC was developed by Porath and co-workers in 1975, where this type of chromatography has exploited the affinity between the proteins and heavy metal ions (Zn2+, Cu2+, Ni2+), in particularly that the zinc and copper metal could strongly adsorb polypeptides containing histidine and cysteine residues in aqueous solutions [141]. Based on this principle, afterwards Porath et al have used iminodiacetic acid (IDA) to bind metal ions bind to agarose to further chelate with proteins though histidine and cystein residues [146]. After this, Hochuli et al. has developed a new metal chelate adsorbent that uses nitrilotriacetic acid, a quadridentate chelating adsorbent, as an alternative to IDA that was found to bind stronger to Ni2+ and Cu2+ and also presents high stability [142]. Moreover, the adsorbent Ni2+-NTA also presents affinity for adjacent histidine residues in proteins or peptides [142]. Then, this technology have been reported firstly for the production and purification of recombinant protein, where the hexapepidr based on histidine was genetically fused to N-terminal of mouse dihydrofolate reductase protein, produced in E. coli and then the fusion protein was further purified by IMAC using a Ni2+-NTA adsorbent through the affinity for histidine based peptide. Subsequently, the fusion protein was eluted with pH gradient (pH 8-5) and then the hexapeptide was removed by carboxypeptidase A [142]. However, the major drawback on this pioneer purification study was the employment of denaturing conditions, and in order to overcome this, Janknecht et al. have improved the purification procedure by conducting the binding under native conditions and the elution with low concentrations of imidazole, as competitive agent, or EDTA [147-148]. Currently, this hexapeptide containing histidines is one of the most widely used affinity tag on the purification of recombinant proteins and has been comprehensively reviewed on the literature [69, 149-152]. Besides the advantages associated to structural ligands metal chelate, the compatibility of His tag with denaturing agents can be a benefit for the recombinant proteins produced as IBs, as they required further steps of solubilization with chaotropic agents [142]. The main disadvantages of the adsorbents is related with metal leaching [153], that contribute for an end-product contamination. Moreover, in case of using this technology for the production and purification of therapeutic proteins, the metal toxicity can compromise the purification process requiring additional steps of purification, contributing for associated higher costs. Other disadvantages that were found when using His-tag are related with the evidences that the imidazoyl side chain groups of histidine can interfere with protein expression and folding and also mask the binding site of the protein for the respective affinity ligand [79].
Despite of protein purification, the applications of this affinity tag have been extended to among them matrix-assisted refolding [154], detection and immobilization, protein microarrays [155-157]. The His-tag technology also has been on the basis of development of other affinity tags. The HAT tag is one of these examples and is a polyhistidine peptide sequence that can be fused on both ends of the target protein. The purification can be carried out under same conditions as His-tag, however the adsorbent used is Co2+ metal ions immobilized on carboxymethylaspartate (CMA), presenting highest capacity and high flow resistance [143]. A novel metal chelating agents based on 1,4,7-triazacyclononane (TACN) complexed with Ni2+ have been developed to be used IMAC application and intends to overcome the limitations of the technique such as leaching, stability and reduction of non-specific interactions [144] . The respective binding partner corresponds to random heptapeptides sequences that display multiple histidine, tryptophan, and/or tyrosine residues selected from a phage display to present affinity for the Ni2+- TACN based ligands [144]. The selected peptide tag to be used for the purification of recombinant proteins presents the sequence HHHNSWD and was fused to N-terminal of Green Fluorescent Protein (GFP), produced in E.coli and then purified under physiological conditions with high salt concentration: the subsequent recovery was done with imidazole due to the presence of three histidine residues [144]. This approach becomes an alternative platform for downstream IMAC.
The ion-exchange resins are also used as structural ligands with affinity for several tags. In this case, the affinity tags must be charged so that can interact though electrostatic interactions to the opposite charge of the matrices, allowing protein purification. The polyarginine-tag consists of six arginine residues and the proteins fused with this peptide already have been purified by cation exchange resin and was firstly reported by Sassenfeld et al with the production and purification of the human urogastrone by using two steps of chromatography that leads to a purity higher than 95% and 44% yield with the elution conducted with a NaCl linear gradient at alkaline pH [138]. The use of polyarginine-tag has also been extended on refolding studies of a target protein upon immobilization. The refolding process is required once fusion proteins are produced as inclusion bodies and a successful refolding strategy should be performed by using low amounts of protein due to the kinetic competition between the correct folding and incorrect aggregation [158]. In order to develop strategies for refolding on matrix, proteins fused to an arginine tail were denatured with a chaotropic agent such as urea and then attempts on refolding on column were performed through the binding of the affinity tag through the cation exchange resins. Upon removal of denaturant agent and the high salt concentrations, the fusion protein was able to successfully refold without aggregation [158-159]. The main limitations of this system are that charge interactions can provide low selectivity and aggregation issues and could occur do to charge-charge interactions [72]. A positively charged domain designated as Zbasic was also developed to be used as affinity tag and subsequently in cation exchange chromatography [66]. A great advantage on using cation-exchange chromatography for purification of recombinant proteins produced in E.coli is that the most proteins of this host are neutral or acidic, which can contribute for the reduction of non-specific interactions as this proteins do not bind to cationic resins at physiological and alkaline pH [65, 79]. This domain was engineered to create a highly charged Z domain, the mutated version of B domain of SpA, through the substitution of residues at specific positions by charged ones so that at physiological conditions the positively charged Zbasic tag could interact with the negatively charged matrix [65-66]. The choice of this domain to create a novel affinity tag present several reasons such as the domain is soluble, do not display disulfide bonds and present a reversible folding after the exposure to chaotropic agents [66]. This last reason was relevant as this system was then extended to matrix assisted refolding of proteins that were solubilised with chaotropic agents after being produced as inclusion bodies. This successful strategy is more advantageous over the common solubilization and refolding procedures based on dialysis, contributing for an effective, less expensive and time-consuming procedure. The major drawback of using poly-arginine tails for the purification reported in literature [79] is that the these polypeptides can compromise the target protein due to degradation and precipitation problems. In order to overcome these issues, the poly-anionic affinity tails have been used developed, and tails containing from one to eight glutamic residues were used to purify human growth hormone [160] and virus-like particles [140] produced in E.coli . In both situations, these affinity tags were found to be useful on the production of proteins as these were produced as a soluble form and the purification was also found effective. Moreover, an affinity tag also based on negatively charged residues such as aspartates was described to be a good approach and also the fusion proteins with these tails remain stable with the unaltered activity [139]. The Z domain was also the basis for the generation of the Zacidic, with the same advantages as the Zbasic, however it was engineered to have negatively charged to be used on anion-exchange chromatography [67].
The affinity tags that have been described so far are based on the purification through affinity chromatography. However, the use of the affinity tags have been extended to other techniques such as aqueous two-phase systems (ATPS), where hydrophobic tags based on tryptophan, phenylalanine and tyrosine are used [161]. Other example is related with Halo-tag, a 34 KDa protein derived from bacterial haloalkane dehalogenase that is able to yield pure proteins through the covalent binding to chloroalkane [162]. In order to combine several functionalities, a multipurpose peptide tag was developed with the sequence HYDHYD that comprises a bi-repetition of the tripeptide constituted by a histidine, tyrosine and aspartate residues [145]. Due to these functionalities, this peptide can be fused on N-or C-terminal of the target proteins enabling the purification by several techniques such as IMAC, IEX, ATPS and hydrophobic interaction chromatography (HIC). This versatile affinity tag has proven to have multifunctional purification capabilities and was used for the purification of GFP, lactate dehydrogenase and human haemoglobin in different manners [145]
Other proteins such as small ubiquitins-related modifier (SUMO) [163-164], N-utilization substance A (NusA) [165-166] and thioredoxin A (TrxA) [167] are also widely used as affinity tags; however these affinity tags are mainly used as solubility tags because their use is based on improvement of solubility of the fusion partner rather than for purification purposes [75, 77]. In order to use these affinity tags for affinity based purification, a purification tag must be added to the fusion partner so that can be used later onto affinity matrices with an immobilized receptor. The TrxA is a 12 KDa oxido-reductase small protein that presents relevant properties such as an intrinsic thermal stability, solubility and robust folding properties that when fused to a target protein contributes for high yields of soluble protein [168-169]. The SUMO is 11KDa protein used as a part of pros-translational modification in eukaryotic cells, and was found that enhances solubility and stability of the binding partner when fused on N-terminal [163-164]. A great advantage of this affinity tag is that the natural sequence of this tag is recognized by the SUMO protease (S. cerevisae Ulp1), in particular the conserved Gly-gly [164] . However, when used in eukaryotic host, the affinity tag can be cleaved due to the presence of SUMO proteases in vivo [170-171]. As the other solubility tags, the NusA is a 55 KDa transcription elongation and antitermination factor that is also known to increase the solubility of fusion proteins, however due to their large size can contribute for a high metabolic burn for the host cell [172].
A new class of affinity tags designated as elastin-like polypeptides (ELPs) raised with the advantages as these tags do not require a purification chromatographic step and the tag removal is not needed [173]. This tags encloses the fusion of the sequence VPGXG, where X is any amino acid, to a self-cleaving intein, being this useful because they can endure a reversible inverse temperature transition [173-174]. In other words, the ELPs can reversible aggregate with temperatures higher than their phase transition and high salt concentrations [173-174]. Moreover, the presence of intein allows self-cleaving as a reaction to a mild pH shift [173-175]. In order to purify recombinant proteins, the fusion proteins just need to subject to a several cycles of salt addition, heating and centrifugation until precipitation [175]. The main advantages related with this technology is the elimination of using chromatographic resins, the simple methodology and the potential for scaleable, becoming a very promising alternative [174]
TAG REMOVAL
The presence of the affinity tags fused to the purified target protein not always interferes with the structure and biological function of the target protein, unless if the affinity tag used corresponds to a large protein. Moreover, in order to apply these technologies to the purification of therapeutic proteins, there is a demand for tag cleavage to avoid immunological responses and to guarantee the authenticity of the protein structure. The removal of the affinity tag can be performed mainly by enzymatic cleavage or harsh chemical treatments. The chemical methods mostly used involve the cyanogens bromide or hydroxylamine. Although, the main advantage of these methods is the low cost, these treatments presets low specificity and can also lead to protein denaturation [72, 79]. Therefore, the enzymatic methods are preferably as they can be used under physiological conditions and presents selectivity due to the fact the endoproteases used recognizes specific amino acid sequences or motifs [72, 79-80]. The main endoproteases used on the tag removal have been extensively reviewed and comprises enterokinase [176], tobacco etch virus (TEV) [177] Factor Xa [178], thrombin [178-179] and SUMO protease [164]. An attractive feature of the endoproteases enterokinase and factor Xa is that after cleavage there is no additional amino acid residues on the protein structure, contributing for intact end product [79, 176-177]. The main limitations are related with the tag incompatibility in certain conditions (e.g. buffers, temperature) that can influence their activity and therefore become inefficient, the steric hindrance also can unable the endoproteases to access the cleavage site [80] and also the high costs associated on the use of this proteases, limiting the scalability of the purification process [180]. In order to overcome the higher costs, strategies on the reusability of these proteases without losing their efficiency have been studied, and recently Santana et al have reported new strategies for the immobilization of enterokinase on magnetic supports [181]. According to this study, the immobilized enterokinase seems to maintain the activity as this immobilized enzyme was still able to cleave a fusion protein. Moreover, this study revealed that these magnetic supports can be reusable and are stable [181]. Therefore the use of magnetic supports seems to be a promising platform for the immobilization of enzymes, could be extended to other different proteases contributing for a more effective purification process at lower costs.
Concluding Remarks
Nowadays there is still a demand to develop downstream purification process that can combine selectivity at affordable prices. The affinity chromatography is the most frequently unit for protein purification, however this technique is dependent of the adsorbents availability. Within all the affinity ligands, the biological ones are still the preferably choice as this ligands are well-established and present high selectivity. However, the synthetic ligands raised to overcome the high costs associated with these ligands and have been developed to mimic the natural binding of the biological for their respective pair. The structural ligands are more embracing as their use is dependent on the properties of the target protein, could be extended to the purification of a wide range of proteins and are inexpensive. Nevertheless, the main limitation of these ligands is the lack of specificity that could interfere with the protein purity at the end of the process.
However, the affinity ligands developed so far do not enclose a wide range of purification process, being mostly developed for therapeutic proteins such as immunoglobulins. Therefore, there is still a need to choose a general strategy that could be extended to the purification of any protein of interest with high yields of purity. Currently there is a wide range of affinity tags that can be used for the production and consequently can facilitate the purification of recombinant proteins. However, the choice of using the best affinity tag is dependent on the protein of interest that will be produced and then purified. Moreover, the choice of the affinity tag should be taken into account the respective binding partner. This is due to the choice of the respective affinity ligand can also influence the entire purification process. The biological ligands still present the majority of the ligands that are able to recognize peptide or protein affinity tags and therefore can compromise the efficiency and costs of the entire purification process. One of the mostly used affinity pair is the His-tag and the respective metal chelate supports and even if this pair seems to be applied in a wide range of applications, there is a still a lack of selectivity between the pair and also problems such as metal leaching regarding the structural ligand.
Therefore there is still a need to progress on the development of affinity pairs that can combine the robustness and cost effectivess of the affinity ligand and at the same time presents a high selectivity for the tag that should comprise the advantages of a peptide tag and also promotes the solubility and then the proper folding of the target protein.
