Protein Expression and PuriWcation 48 (2006) 1���13 www.elsevier.com/locate/yprep 1046-5928/$ - see front matter �� 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2005.12.002 Review Current strategies for the use of aYnity tags and tag removal for the puriWcation of recombinant proteins Jos�� Arnau ��, Conni Lauritzen, Gitte E. Petersen, John Pedersen Unizyme Laboratories A/S, Dr. Neergaards vej 17, DK-2970 H��rsholm, Denmark Available online 28 December 2005 Abstract AYnity tags are highly eYcient tools for protein puriWcation. They allow the puriWcation of virtually any protein without any prior knowledge of its biochemical properties. The use of aYnity tags has therefore become widespread in several areas of research e.g., high throughput expression studies aimed at Wnding a biological function to large numbers of yet uncharacterized proteins. In some cases, the presence of the aYnity tag in the recombinant protein is unwanted or may represent a disadvantage for the projected application of the protein, like for clinical use. Therefore, an increasing number of approaches are available at present that are designed for the removal of the aYnity tag from the recombinant protein. Most of these methods employ recombinant endoproteases that recognize a speciWc sequence. These process enzymes can subsequently be removed from the process by aYnity puriWcation, since they also include a tag. Here, a survey of the most common aYnity tags and the current methods for tag removal is presented, with special emphasis on the removal of N-terminal histidine tags using TAGZyme, a system based on exopeptidase cleavage. In the quest to reduce the signiWcant costs associated with protein puriWcation at large scale, relevant aspects involved in the development of downstream processes for phar- maceutical protein production that incorporate a tag removal step are also discussed. A comparison of the yield of standard vs. aYnity puriWcation together with an example of tag removal using TAGZyme is also included. �� 2005 Elsevier Inc. All rights reserved. Keywords: AYnity puriWcation AYnity tag His-tag IMAC Recombinant protein Protein puriWcation Tag removal Endoproteases Exopeptidases Aminopeptidases Therapeutic protein puriWcation With the modern advances in genomics, proteomics and bioinformatics, the number of proteins being produced using recombinant techniques is exponentially increasing. High throughput screening approaches are being per- formed to rapidly identify proteins with a potential applica- tion as therapeutic, diagnostic or industrial enzymes [1]. For this purpose, diVerent expression hosts (e.g., Esche- richia coli, Saccharomyces cerevisiae, Pichia pastoris, insect and mammalian cell lines) have been developed to express heterologous proteins [2���7]. Additionally, genomic approaches are being pursued to solve the structure of numerous proteins [8,9]. The above-mentioned screening approaches would not be feasible if speciWc puriWcation procedures were to be developed for each individual protein. Here, the use of aYnity tags enables diVerent proteins to be puriWed using a common method as opposed to highly customized procedures used in conventional chromatographic puriWcation. When designing a downstream processing strategy for a protein, the inclusion of an aYnity tag might be attrac- tive for a number of additional reasons. In many cases, the protein candidate may exist as a version that includes an aYnity tag from its early research stages where no bio- chemical characterization or functional assay is yet avail- able. For structural studies, more than 60% of the proteins produced include a polyhistidine tag (his-tag, [10]). Additionally, the fact that aYnity puriWcation nor- mally results in high yields���often over 90%���makes this * Corresponding author. E-mail address: ja@unizyme.dk (J. Arnau).

2 J. Arnau et al. / Protein Expression and PuriWcation 48 (2006) 1���13 alternative economically favorable. Other economical and practical issues like the number of unit operations needed and the timesaving resulting from a reduction in chromatographic steps are also relevant aspects for con- sideration. Introducing an aYnity tag may have a positive eVect in the biochemical properties of the target protein. A lit- erature survey reveals that aYnity tags have been observed to: (i) improve protein yield [11,12], (ii) prevent proteolysis [13], (iii) facilitate protein refolding [14], (iv) protect the antigenicity of the fusion protein [15], and (v) increase solubility [16���19]. AYnity tags have also been used to increase the sensitivity of binding assays for tagged ScFv [20]. On the other hand, adding a tag has also been reported to negatively aVect the target protein resulting in e.g., (i) a change in protein conformation [21], (ii) lower protein yields [22], (iii) inhibition of enzyme activ- ity [23,24], (iv) alteration in biological activity [25], (v) undesired Xexibility in structural studies [26] and (vi) toxicity [27]. Due to the somehow unpredictable changes that adding a tag may introduce in a protein and its behavior, it is usu- ally desirable to remove the tag. This reXects on the design of the protein fusion. Importantly, removal of the tag needs to be considered when designing a process for the produc- tion of a recombinant protein that is intended for human use to enable production of a ���native��� (i.e., tagless) protein. And consequently, both the enzyme(s) used to cleave the tag and the cleaved fusion partner need to be removed from the puriWed protein. Here, a review of aYnity tags commonly used for recombinant protein production and the methods avail- able for tag removal are discussed. A comparison of puriWcation processes for a recombinant enzyme with and without aYnity tag is also presented together with an overview of a downstream process that incorporates aYnity puriWcation and tag removal. Finally, an exam- ple of process for tag removal is presented for a his-tag thioredoxin (Trx).1 An overview of aYnity tags and the design of the protein fusion Recent reports have included several overviews of the cur- rently available aYnity tags for protein production and puriW- cation [28���32]. Nevertheless, since the choice of aYnity tag and the method for tag removal are mutually dependent, an intro- duction to aYnity tags is given herein for clarity. AYnity tags can be deWned as exogenous amino acid (aa) sequences with a high aYnity for a speciWc biological or chemical ligand. A major group of aYnity tags consists of a peptide or protein that binds a small ligand linked on a solid support (e.g., his-tags bind to immobilized metals, dis- cussed below). Another group includes tags that bind to an immobilized protein partner such as an antibody or anti- body puriWcation using protein A aYnity chromatography ([29], Table 1). The protein A-based methodology, used for e.g., puriWcation of monoclonal antibodies (mAb), has been extensively reviewed elsewhere [33,34] and will not be dis- cussed here. His-tags are the most widely used aYnity tags. PuriWca- tion of his-tagged proteins is based on the use of chelated 1 Abbreviations used: Trx, thioredoxin 6��his, a stretch containing six consecutive histidine residues aa, amino acid(s) AAP, Aeromonas proteol- ytica aminopeptidase APM, aminopeptidase M B., Bacillus CPA, car- boxypeptidase A CPB, carboxypeptidase B DAPase, recombinant dipeptidyl peptidase, part of TAGZyme E., Escherichia ELP, elastin-like polypeptides FMN, Xavin mononucleotide FP, green Xuorescent protein GST, glutathione S-transferase His-tag, a polyhistidine tag IMAC, im- mobilized metal���ion aYnity chromatography mAb, monoclonal antibod- ies MBP, maltose-binding protein NMR, nuclear magnetic resonance pGAP, recombinant pyroglutamyl aminopeptidase pGAPase, an engi- neered version of recombinant pGAP used in TAGZyme PHB, poly- hydroxybutyrate Qcyclase, recombinant glutamine cyclotransferase, part of TAGZyme ScFv, single chain antibodies. Table 1 AYnity and solubility tags for recombinant proteins a Only a few relevant references are included. Tag Size (aa) Comments Referencesa His-tag 5���15 PuriWcation under native or denaturing conditions [10,38,85,109] FLAG 8 Calcium-dependent, mAb-based puriWcation [42,43,45] Streptag II 8 ModiWed streptavidin, elution with biotin analog [49,51,52,110] HA-tag 9 InXuenza virus hemagglutinin tag, Ab-based puriWcation [36] Softag1, Softag 3 13, 8 Recognized by polyol-responsive mAb [46���48] c-myc 10 mAb-based puriWcation [31] T7-tag 11���16 mAb-based puriWcation [56] S-tag 15 S-protein resin aYnity puriWcation [45] Elastin-like peptides 18���320 Protein aggregation by temperature shift, intein used to remove tag [64���66] Chitin-binding domain 52 Binds only insoluble chitin (see intein, Table 3) [68,111] Thioredoxin 109 AYnity puriWcation with modiWed resin [17,26,89,112] Xylanase 10A 163 Cellulose based capture, elution with glucose [113,114] Glutathione S-transferase 201 Glutathione or GST-Ab aYnity [9,26,87] Maltose binding protein 396 Amylose aYnity puriWcation [17,26,53,54,57,60] NusA 495 Increased solubility in E. coli. AYnity tag needed for puriWcation [19]