WO2015049230A1

WO2015049230A1 - An orthogonal set of tag-cleaving proteases for purification of proteins and stoichiometric protein complexes

Info

Publication number: WO2015049230A1
Application number: PCT/EP2014/070918
Authority: WO
Inventors: Dirk Goerlich; Steffen Frey
Original assignee: MAX-PLANCK-Gesellschaft zur Förderung der Wissenschaften e.V.
Priority date: 2013-10-01
Filing date: 2014-09-30
Publication date: 2015-04-09

Abstract

The present invention belongs to the field of biotechnology. More specifically, a widely applicable strategy for purification of recombinant protein complexes with defined stoichiometry is introduced. Further described is an orthogonal set of highly efficient and specific proteases that can be used for this procedure.

Description

An orthogonal set of tag-cleaving proteases for purification of proteins and

stoichiometric protein complexes

BACKGROUND OF THE INVENTION

Purified proteins have gained considerable impact in modern biomedical research. As the traditional chromatographic methods for purification of proteins from their natural source are tedious and often lead to limited yield and purity, most of the proteins are nowadays produced as "recombinant" proteins in suitable host organisms. In such systems, the DNA encoding the target protein is fused to foreign DNA elements. It can, e.g., be put under the control of a strong inducible promoter in order to over-express the target protein and hence allow for higher product yields. Most importantly, however, recombinant expression systems can be used to modify the target protein, i.e. to introduce mutations, deletions or to genetically fuse the target protein with engineered "tags" (Makrides (1996) Microbiol Rev. 60: 512-538; Uhlen et al. (1992) Curr Opin Biotechnol. 3: 363-369; Lichty et al. (2005) Protein Expr Purif. 41 : 98-105; Waugh (2005) Trends Biotechnol. 23: 316-320; Young et al. (2012) Biotechnol J. 7: 620-634). Such tags often promote protein expression and solubility. Typically, they also mediate high- affinity binding to standardized affinity matrices and therefore allow for highly efficient and streamlined purification schemes (Lichty (2005), supra; Young et al. (2012), supra; Nilsson et al. (1997) Protein Expr Purif. 11 : 1-16).

In the ideal case, such tags can be removed from the target protein during the purification process and thereby allow production of target proteins lacking any unwanted extensions at their termini. This step is often accomplished by site- specific proteases recognizing a unique, short and linear recognition motif that has been artificially introduced between the tag and the target protein. For this purpose, commercial suppliers offer various proteases, e.g. Thrombin, Factor Xa, enterokinase, or the 3C proteases from Tobacco etch virus (TEV) or human rhinovirus (Young et al. (2012), supra; Arnau et al. (2006) Protein Expr Purif 48: 1 - 13).

In practice, the application of these proteases is often hampered by inefficient cleavage, a requirement for elevated temperature during cleavage, pronounced preferences for certain amino acids in the P1 ^' position (the position after the scissile bond) or a narrow optimum for buffer and/or salt conditions. Also, most of these proteases leave unwanted residues at the N-terminus of the target protein (Arnau et al. (2006), supra). In addition, the specificity of some commonly used proteases (e.g. thrombin) is rather low, which might lead to the degradation of sensitive target proteins.

Recently, an alternative system utilizing the S. cerevisiae protease scUlpl has been introduced (Li & Hochstrasser (1999) Nature 398: 246-251 ; Butt et al. (2005) Protein Expr Purif. 43: 1-9; Malakhov et al. (2004) J Struct Funct Genomics 5: 75- 86). This protease can be used to specifically cleave off target proteins fused to the C-terminus of yeast Smt3p (scSUMO).

The S. cerevisiae SUMO protease Ulpl p (scUlpl ) cleaves SUMO-containing substrates also in its cellular context. SUMO (small ubiquitin-related modifier) can be covalently attached to numerous acceptor proteins, whereby an isopeptide bond is formed between SUMO^'s carboxy terminus and a lysine ε-amino group from the acceptor (Muller et al. (2001 ) Nat Rev Mol Cell Biol. 2: 202-210). The SUMO pathway involves two scUlpl -mediated proteolytic events: The enzyme removes a C-terminal extension from the scSUMO precursor protein, thereby creating the characteristic C-terminal Gly-Gly motif present in the mature scSUMO. In addition, scUlpl cleaves isopeptide bonds between scSUMO and acceptor proteins, and thereby reverses scSUMO modifications. Most importantly, scUlpl exhibits an extraordinary specificity as it recognizes not just a short peptide motif, but the folded SUMO domain including the C-terminal Gly-Gly motif (Mossessova (2000) Mol Cell. 5: 865-876). Except for proline, scUlpl can in principle accept any amino acid in the P1 ^' position after the scissile bond (Malakhov et al. (2004), supra). It is therefore suited to generate a wide variety of non-acetylated N-termini, and thus allows restoring the authentic N-terminus of most target proteins. The human orthologue of scUlpl (hsSENPI ) has been described previously (Gong et al. (1999) J Biol Chem. 275(5): 3355-3359).

SUMO is just one representative of a larger group of paralogous eukaryotic modifiers that also includes ubiquitin (Ub), Atg8 and NEDD8 (Yeh et al. (2000) Gene 248: 1 -14; van der Veen et al. (2012) Annu Rev Biochem 81 : 323-357). These proteins not only share a common fold and a similar conjugation mechanism, but also, they are similarly processed and deconjugated by dedicated proteases (van der Veen et al. (2012), supra). While SUMO, ubiquitin and NEDD8 possess a characteristic double-glycin (GG) motif at their mature C-termini, Atg8 proteins feature the sequence Phe-Gly (FG) at the corresponding position.

A priori, it is impossible to predict to which degree proteases involved in processing and deconjugation of the mentioned (and other similar) modifiers in general can discriminate between their cognate substrate and other related modifiers. This even more applies if the examined proteins originate from different species. Therefore, if mutually exclusive (i.e. orthogonal) specificity of a substrate/protease pair with respect to other substrate/protease pairs is required, the substrate/protease combinations in question have to be tested individually, as has been done in the examples below in all possible combinations.

Proteolytic removal of affinity tags is commonly accomplished in solution after elution from the affinity resin. While allowing free access of the protease to its substrate, this procedure has the disadvantage that the affinity tag released from the target protein has to be removed in a consecutive purification step. This generally necessitates a buffer exchange (to remove the prior used eluent) and a "reverse purification" on the same type of affinity resin. During this reverse purification step, the tag and any non-cleaved fusion protein (still containing the tag) are re-bound to the affinity resin and thus removed from the processed, tag- free target protein that now remains in the non-bound fraction.

An alternative to such post-elution removal of affinity tags is on-column cleavage. Here, the target protein is released from the affinity resin by directly treating the loaded resin with a specific tag-cleaving protease (Walker et al. (1994) Biotechnology 12: 601 -605; Dian et al. (2002) J Chromatogr B Analyt Technol Biomed Life Sci 769: 133-144).

Within its cellular context, the physiologically relevant form of a protein is often not a single polypeptide but a complex comprising two or even multiple subunits. Structural and functional characterization of such protein complexes thus critically relies on purification strategies that allow controlling the stoichiometry of subunits. Provided functional subunits can be produced in the absence of their binding partners, protein complexes can be assembled from individually pre-purified subunits (Fig. 1A). Alternatively, multiple subunits can be expressed and assembled in situ within the same host cell (Fig. 1 B, Fig. 2). In both cases, the assembled complex needs to be separated from an excess of non-assembled subunits and partially assembled sub-complexes (Fig. 1 , Fig. 2). This can be a challenging task, especially if the interaction between the two partners is regulated e.g. by binding to nucleotides or competing binding partners or if additional inactive subunits are in the mixture.

WO 2002/090495 (EP 1 392 717), US 6,872,551 , US 7,910,364, and US 7,498,165 describes a rapidly cleavable SUMO fusion protein expression system for difficult to express proteins. More specifically, SUMO or SUMO fragments are used to stabilize a poly-amino acid of interest and to enhance the solubility of the expressed fusion protein, enabling correct refolding and conferring monomeric expression without any toxic effects on the host cell.

WO 2003/057174 (EP 1 470 236) describes the use of SUMO and SUMO hydrolases/proteases in purifying polypeptides in general, but remains silent on the purification of stoichiometric protein complexes, or the use of these tools for on-column cleavage in affinity chromatography.

Gagnon et al. (2007) Methods in Enzymology 425: 263-282 discloses purification of a multidomain protein complex in which the sub domains have different affinity tags, but which are not linked via protease sites to the sub domains.

Removal of a Tag from a polypeptide by using proteases is known in the field (Arnau et al. (2006) Protein Expr Purif 48: 1 -13). However, the combined and simultaneous use of more than one Tag linked via a protease cleavage site in different sub domains of protein complexes, which protease cleavage sites can be mutually exclusive be cleaved, is technically challenging, and has, to the best knowledge of the inventors, not been accomplished or described in the art.

Hence, there is a need in the art for methods for purifying stoichiometric protein complexes, which circumvent the above-mentioned disadvantages.

SUMMARY OF THE INVENTION

The present disclosure now introduces a general straightforward strategy for purification of stoichiometric protein complexes that exploits the combined discriminative power of two or more affinity matrices and proteases (Fig. 1 B, Fig. 2). Briefly, by tagging individual subunits of a given protein complex with orthogonal affinity tags and orthogonal protease recognition sites, consecutive sequences of affinity capture and proteolytic release allow selecting for the presence of each tagged subunit individually. This strategy thus provides a streamlined purification scheme and a defined stoichiometry alongside with a product purity conforming the highest standards. Although figures 1 B and 2 only show the purification of a binary complex, protein complexes with more than two subunits can be purified in an analogous manner.

Evidently, this strategy requires multiple proteases with orthogonal (i.e. mutually exclusive) specificities. To prevent loss of the target proteins by thermal denaturation or buffer incompatibility during the protease cleavage step, such proteases should in addition be exceedingly efficient even at low temperature (preferably 0-4°C) and within a wide range of buffers.

Here, new proteases matching these criteria are characterized in detail: bdSENPI and bdNEDPI from Brachypodium distachyon (bd) and ssNEDPI from salmon (Salmo salar, ss) (table 1 ). Moreover, additional sets of orthogonal substrate/protease pairs such as scAtg8/scAtg4 and xlUb/xlUsp2 are provided, which may be advantageously applied in the new method for purification of stoichiometric protein complexes. Table 1 : Nomenclature of NEDD8- and SUMO-orthologues used in this disclosure along with their corresponding proteases.

Abbreviations: n.a.: not analyzed; sc: Saccharomyces cerevisiae; bd: Brachypodium distachyon; ss: Salmo salar.

Comparing the new proteases to both, yeast scUlpl and a stabilized variant of TEV protease, bdSENPI was found to be possibly one of the most efficient and versatile proteases characterized for tag removal so far. It even outperforms scUlpl in several aspects. In addition, it is shown that most here-characterized members of the four classes of SUMO-, NEDD8-, Atg8- and ubiquitin-specific proteases possess orthogonal specificities. The present disclosure further describes the successful application of the new proteases, as exemplified by bdSENPI and bdNEDPI , for the purification of a stoichiometric complex.

More specifically, the present disclosure is directed to a method for purifying a stoichiometric protein complex composed of at least two subunits from a mixture, said mixture comprising said protein complex and monomers of said at least two subunits, wherein said at least two subunits comprised in said mixture each comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS),

wherein the ATs of each of said at least two subunits differ from each other and allow affinity chromatography being selective for each AT, and

wherein the PRS of each of said at least two subunits is cleavable by a protease, which protease is orthogonal to the PRS of the other subunit(s),

wherein the method comprises the steps of

a) subjecting the mixture to a first affinity chromatography selective for the AT of the first of said at least two subunits, whereby (i) the protein complex binds to the affinity resin via the AT of the first subunit, and

(ii) impurities are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the first subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PRS of said first subunit, and

(iv) optionally removing the cleaved off AT of the first subunit; and

b) subjecting the eluate from step a) to a second affinity chromatography selective for the AT of the second of said at least two subunits, whereby

(i) the protein complex binds to the affinity resin via the AT of the second subunit, and

(ii) impurities are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the second subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PRS of said second subunit, and

(iv) optionally removing the cleaved off AT of the second subunit.

In addition, a first protease is provided, said protease having an amino acid sequence with at least 45% identity over the full length of SEQ ID NO: 2 (bdSENPI ),

wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 1 (bdSUMO) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 2; preferably wherein the protease comprises the amino acid sequence shown as amino acids 1 -224 in SEQ ID NO: 2 (bdSENP1₂₄8-48i ); more preferably wherein the protease consists of the amino acid sequence shown as amino acids 1 -224 in SEQ ID NO: 2 (bdSENP1_248-48i ).

Moreover, a second protease is provided, said protease having an amino acid sequence with at least 70% identity over the full length of SEQ ID NO: 11 (ssNEDPI ), wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 8 (ssNEDD8) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 1 1 ; preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 11 (ssNEDPI ); more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 1 1 (ssNEDPI ).

Further, a third protease is provided, said protease having an amino acid sequence with at least 35% identity over the full length of SEQ ID NO: 12 (bdNEDPI ), wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 9 (bdNEDD8) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 12; preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 12 (bdNEDPI ); more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 12 (bdNEDPI ).

Still another protease is provided, having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 21 (xlUsp2), wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 20 (xlUb) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 21 ; preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 21 (xlUsp2); more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 21 (xlUsp2).

A nucleic acid molecule, encoding any one of the above proteases is also contemplated.

In accordance with the above-described method, a kit of parts is provided, comprising at least two proteases selected from

(i) the first protease as described above,

(ii) the second or third protease as described above,

(iii) TEV protease as shown in SEQ ID NO: 18 or 19,

(iv) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 15 (scAtg4), or a protease derivative thereof having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 15, wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 14 (scAtg8) with at least 20% activity as compared to the parent protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 15 (scAtg4); and

(v) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 21 (xlUsp2), or a protease derivative thereof having an amino acid sequence with at least 80% identity over the full length of SEQ ID

NO: 21 , wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 20 (xlUb) with at least 20% activity as compared to the parent protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 21 (xlUsp2);

preferably comprising at least two proteases selected from (i)-(iii), more preferably comprising two proteases selected from (i) and (ii), and most preferably comprising the first protease and the third protease as described above.

Accordingly, the use of a protease as described above, or the kit of parts as described above in a method of purifying stoichiometric protein complexes comprising at least two subunits is further provided, wherein said at least two subunits comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS), and optionally a spacer between the AT and the PRS, and wherein the AT of each of said at least two subunits differs from each other so to allow specific affinity chromatography, and wherein the PRS of each of said at least two subunits is cleavable by a protease which is orthogonal to the PRS of the other subunit(s); preferably wherein the method is further defined as described above.

Finally, the use of a protease as disclosed above, or the kit of parts as disclosed above for on-column cleavage in an affinity chromatography is also described. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The inventors here propose a broadly applicable strategy for purification of protein complexes that combines multiple consecutive affinity purification steps with specific on-column cleavage. As an example, the purification of binary complexes according to this scheme is detailed Fig. 1 and Fig. 2.

More specifically, a method for purifying a stoichiometric protein complex composed of at least two subunits from a mixture is provided, said mixture comprising said protein complex and monomers of said at least two subunits, wherein said at least two subunits comprised in said mixture each comprise an N- terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS), wherein the ATs of each of said at least two subunits differ from each other and allow affinity chromatography being selective for each AT, and wherein the PRS of each of said at least two subunits is cleavable by a protease, which protease is orthogonal to the PRS of the other subunit(s), wherein the method comprises the steps of a) subjecting the mixture to a first affinity chromatography selective for the AT of the first of said at least two subunits, whereby

(i) the protein complex binds to the affinity resin via the AT of the first subunit, and

(ii) impurities (e.g. monomers of the second of said at least two subunits) are washed off the column, and

(iv) optionally removing the cleaved off AT of the first subunit; and b) subjecting the eluate from step a) to a second affinity chromatography selective for the AT of the second of said at least two subunits, whereby

(ii) impurities (e.g. monomers of the first of said at least two subunits) are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the second subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PCS of said second subunit, and

(iv) optionally removing the cleaved off AT of the second subunit.

In one embodiment, the stoichiometric protein complex is composed of at least two subunits. However, the stoichiometric protein complex may also be composed of three, four, five, six, seven, eight or nine subunits, which each differ from each other.

If the protein complex comprises a third subunit, one may incorporate a third affinity chromatography step. Such a third affinity chromatography step will further improve the purity, and it makes sure that only those complexes are purified, which contain all three subunits.

In this case, said third subunit comprised in said mixture comprises an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS), wherein the AT of said third subunit differs from the AT of the other subunits and allows affinity chromatography being selective for the AT of said third subunit, and wherein the PRS of said third subunit is cleavable by a protease, which protease is orthogonal to the PRS of the other two subunits, further comprising after step b) and prior to optional step c) an additional step b') subjecting the eluate from step b) to an affinity chromatography selective for the AT of the third subunit, whereby (i) the protein complex binds to the affinity resin via the AT of the third subunit, and

(ii) impurities (e.g. monomers) are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the third subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PRS of said third subunit, preferably wherein the protein complex is eluted by on-column cleavage, and

(iv) optionally removing the cleaved off AT of the third subunit.

Likewise, if the protein complex comprises a fourth subunit, it is advantageous to include a fourth affinity chromatography step. In this case said fourth subunit comprised in said mixture comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS), wherein the AT of said fourth subunit differs from the AT of the other subunits and allows affinity chromatography being selective for the AT of said fourth subunit, and wherein the PRS of said fourth subunit is cleavable by a protease which is orthogonal to the PRS of the other three subunits, further comprising after step b') and prior to optional step c) an additional step b") subjecting the eluate from step b') to an affinity chromatography selective for the AT of the fourth subunit, whereby

(i) the protein complex binds to the affinity resin via the AT of the fourth subunit, and

(ii) impurities (e.g. monomers) are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the fourth subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PRS of said fourth subunit, preferably wherein the protein complex is eluted by on- column cleavage, and

(iv) optionally removing the cleaved off AT of the fourth subunit.

Finally, if the protein complex comprises a fifth subunit, it is advantageous to include a fifth affinity chromatography step. In this case said fifth subunit comprised in said mixture comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS), wherein the AT of said fifth subunit differs from the AT of the other subunits and allows affinity chromatography being selective for the AT of said fifth subunit, and wherein the PRS of said fifth subunit is cleavable by a protease which is orthogonal to the PRS of the other four subunits, further comprising after step b") and prior to optional step c) an additional step b'") subjecting the eluate from step b") to an affinity chromatography selective for the AT of the fifth subunit, whereby (i) the protein complex binds to the affinity resin via the AT of the fifth subunit, and

(ii) impurities (e.g. monomers) are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the fifth subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PRS of said fifth subunit, preferably wherein the protein complex is eluted by on-column cleavage, and

(iv) optionally removing the cleaved off AT of the fifth subunit.

If deemed appropriate, the method comprises the additional step of c) removing the protease from the eluate originating from the last affinity chromatography step. For example, step c) may be an affinity chromatography, a size exclusion chromatography, or a precipitation step, as generally known in the art. However, any method suitable for removing the protease from the eluate may be applied. In a preferred embodiment, the protease from the eluate originating from the last affinity chromatography prior to step c) comprises an affinity tag, as further defined below, and step c) is an affinity chromatography step, whereby the protease binds to the affinity resin, and the protein complex is collected in the flow-through. Said affinity tag of the protease may be the same than one of the affinity tags used in the affinity chromatography steps a), b), b'), b") or b'"), but with the provisio that it differs from the affinity tag used in the directly preceding affinity chromatography step b), b'), b") or b'"). For example, the affinity tag of the final protease may be a polyHis-tag, and step c) is a Ni²⁺-chelate affinity chromatography.

Preferably, the protein complex is eluted in step a) (iii) or step b) (iii) by on-column cleavage. More preferably both step a) (iii) and step b) (iii) are on-column cleavage steps. Likewise, if the method further comprises optional steps b') (iii), b") (iii) or b'"), any one of step b') (iii), b") (iii), and b'") (iii) may be independently an on column-cleavage step. Preferably, at least two of steps b') (iii), b") (iii), and b'") (iii) are on-column cleavage steps, and more preferably all of steps b') (iii), b") (iii), and b"') (iii) are on-column cleavage steps. On-column cleavage offers several advantages. It not only makes purifications more time-efficient by avoiding any lengthy buffer exchange and reverse chromatography steps. On-column cleavage also allows the target proteins to be specifically released from the resin under very mild conditions: As the elution buffer differs from the washing buffer only by a minute amount of protease, on-column cleavage bypasses more drastic elution conditions as high concentrations of competitor, significant alterations in the buffer composition or pH changes. Most importantly, however, on-column cleavage potentiates the efficiency of protein purifications by elegantly combining the specificities of the affinity resin and the protease: Only proteins containing the proper affinity tag and the proper protease recognition site will be bound and consecutively released from the resin. In contrast, contaminant proteins non- specifically interacting with the resin - and thus lacking the specific protease recognition site - will remain bound to the affinity resin during the elution step. This on-column cleavage procedure is exceptionally robust: From the inventors' experience, from several hundreds of target proteins, only one could so far consistently not be processed by the protease. In this special case, the known structure suggests that the residue in P1 ^'-position is already part of a large folded domain that effectively prevents proteolytic processing.

The term "stoichiometric protein complex" is intended to mean that each complex is composed of the same molar ratio of the same subunits, and that each complex has a definite identical size as defined by the number of subunits forming the complex. In very special cases one subunit A may form a complex comprising, e.g., either a subunit B or a subunit C, in which case there will be a mixture of stoichiometric protein complexes comprising subunits AB and complexes comprising subunits AC. However, a stoichiometric protein complex is to be distinguished from random protein aggregates, which are characterized by a random molar distribution, and which differ by its constituents.

The "mixture" may be any suitable starting material for the purification method, such as an aqueous buffered or non-buffered solution comprising the stoichiometric protein complex. The "mixture" may be a lysate, a supernatant, a pre-purified lysate or a pre-purified supernatant, or mixtures thereof, e.g. a mixture of lysates, a mixture of supernatants, or a mixture of a lysate and a supernatant, and the like. Accordingly, the mixture may originate from a mixture of lysates and/or supernatants and/or a pre-purified solution, each comprising at least one of the subunits; or the mixture may originate from a single lysate or supernatant or pre-purified solution comprising all subunits of the protein complex. The term "impurities" may also encompass an undesired buffered solution or a saline, undesired proteins other than the subunits of the complex, cell debris, and possibly monomers of the respective subunits and/or degradation products of said complex. Accordingly, apart from removing such monomers and/or degradation products, the method of the invention may also be used for replacing the buffered solution or saline, or for removing an undesired compound within the buffered solution or saline.

In analogy to the purification of binary complexes using two orthogonal tags and proteases, a purification scheme employing three or more orthogonal tags and proteases can be used for a straightforward purification of stoichiometric triple or higher order complexes. In general, the method allows for the purification of complexes comprising each orthogonally tagged subunit at least once. More specifically, the method is ideally suited for the purification of stoichiometric complexes if each orthogonally tagged subunit is comprised in the complex exactly once. If the protein complex is composed of two subunits, it preferably has a stoichiometry of 1 :1 . Likewise, if the protein complex is composed of 3 different subunits, it preferably has a stoichiometry of 1 :1 :1 , and if the protein complex is composed of 4 different subunits, it preferably has a stoichiometry of 1 :1 :1 :1 . Dependent on the nature of the protein complex, each of the subunits may be comprised once, twice or more often in the protein complex. For example, if the protein complex is composed of 2 different subunits, it may have a stoichiometry of 1 :1 , 1 :2, 2:1 , 2:2, 1 :3, 3:1 , 2:3, 3:2, or 3:3, etc. Although the method is illustrated in the examples using 1 :1 complexes, there is no reason to doubt that it will analogously also work with protein complexes of other stoichiometry. As the method only allows selecting for the presence of orthogonally tagged subunits, purification of such complexes with defined stoichiometry is preferably performed using orthogonal tags on otherwise identical subunits.

In principle, additional affinity chromatography steps can be done for each subunit comprised in the stoichiometric protein complex, as long as there are enough orthogonal protease / PRS systems. In this context, the term "orthogonal" is intended to mean that the protease exhibits only cleavage activity against its corresponding substrate recognition sequence, but not on the other PRS or sequences in the subunits.

Most of the commercially available preparations of site-specific proteases contain affinity tags and can therefore not be used for on-column cleavage from the respective affinity resins (see Fig. 7D and Fig. 8). It also has to be taken into account that some of these proteases may possess suboptimal features with respect to their efficiency, specificity, or special requirements concerning temperature and buffer.

The inventors have discovered and identified such orthogonal protease / PRS systems, which enable the above described method, and which are largely devoid of such drawbacks. They will therefore be of great practical use for labs routinely purifying recombinant proteins and protein complexes. Most importantly, the proposed purification schemes for single proteins and protein complexes are highly efficient and generally applicable. Due to the high efficiency of the provided proteases even at low temperatures and their tolerance towards various buffer conditions, the schemes can be adapted to the needs of the target proteins or complexes over a wide range of conditions. Accordingly, in the following several suitable orthogonal protease / PRS systems are disclosed or even provided for the first time.

In one embodiment, one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 1 (bdSUMO); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 60%, preferably at least 70%, more preferably at least 80%, even more preferably at least 85%, still more preferably at least 90%, most preferably at least 95%, and even most preferably at least 98% identity over the full length of SEQ ID NO: 1 (bdSUMO), wherein the protease shown in SEQ ID NO: 2 (bdSENPI ) is capable of cleaving said PRS derivative with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 1 , under identical conditions.

For determining or comparing cleavage activity of a given PRS/protease system, cleavage reactions are performed in LS-buffer (250mM NaCI, 40mM Tris/HCI pH7.5, 2mM MgCI₂, 250mM sucrose, 2mM DTT, 2pg/ml BSA). Generally, substrates and proteases are pre-diluted in LS-buffer to twice the aspired end- concentration. Cleavage is initiated by mixing identical volumes of substrate and protease pre-dilutions and stopped by mixing with a 9-fold excess of hot SDS sample buffer. A fraction corresponding to 2.5pg of substrate is separated by SDS- PAGE on 7-15% gradient gels. Gels are stained with Coomassie G250 and scanned. Cleavage activity can then be determined using e.g. a densitometer.

Preferably, the most efficient orthogonal protease is used in the final affinity chromatography step, in order to keep the protease "contamination" in the final product low. Since the bdSENPI represents a very efficient protease (as will be shown in more detail below and in the examples), the PRS as defined in (i) or (ii) above is preferably comprised in the "last" subunit to be selected for, e.g. if the stoichiometric protein complex comprises two subunits, said PRS is comprised in the second subunit. More preferably, the last subunit (e.g. the second subunit, in case the complex comprises two kinds of subunits) comprises a PRS consisting of SEQ ID NO: 1 (bdSUMO).

The AT of the subunit comprising the bdSUMO-PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in amino acids 1-224 of SEQ ID NO: 2 (bdSENPI ₂₄₈-₄8i ), or (ii) a protease derivative of (i) having an amino acid sequence with at least 45% identity, preferably with at least 50% identity, more preferably with at least 60% identity, even more preferably with at least 70% identity, still even more preferably with at least 80% identity, most preferably with at least 90% identity, and even most preferably with at least 95% identity, such as 98% identity over the full length of SEQ ID NO: 2,

wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 1 (bdSUMO) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease as defined in (i), under identical conditions. Preferably the AT is cleaved from the subunit using (i) the protease shown in in amino acids 1-224 of SEQ ID NO: 2 (bdSENP1_248-48i )-

In another embodiment, one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 3 (scSUMO); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 55% identity, preferably with at least 60% identity, more preferably with at least 70% identity, even more preferably with at least 80% identity, still even more preferably with at least 90% identity, most preferably with at least 95% identity, and even most preferably with at least 98% identity over the full length of SEQ ID NO: 3 (scSUMO),

wherein the protease shown in SEQ ID NO: 4 (scUlpl ) is capable of cleaving said PRS derivative with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 3, under identical conditions.

The AT of the subunit comprising the scSUMO-PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 4 (scUlpl ), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 35% identity, preferably with at least 40% identity, more preferably with at least 50% identity, even more preferably with at least 60% identity, still even more preferably with at least 70% identity, most preferably with at least 80% identity, and even most preferably with at least 90% identity, such as 95% or 98% identity over the full length of SEQ ID NO: 4,

wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 3 (scSUMO) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease as defined in (i), under identical conditions.

It is further contemplated that one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 5 (hsSUMOI a) or SEQ ID NO: 6 (hsSUMO2); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 60% identity, more preferably with at least 70% identity, even more preferably with at least 80% identity, still even more preferably with at least 90% identity, most preferably with at least 95% identity, and even most preferably with at least 98% identity over the full length of SEQ ID NO: 5 or 6,

wherein the protease shown in SEQ ID NO: 7 (hsSENPI ) is capable of cleaving said PRS derivative with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 5 or 6, respectively, under identical conditions.

The AT of the subunit comprising the hsSUMOI a- or hsSUMO2-PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 7 (hsSENPI ), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 45% identity, preferably with at least 50% identity, more preferably with at least 60% identity, even more preferably with at least 70% identity, still even more preferably with at least 80% identity, most preferably with at least 90% identity, and even most preferably with at least 95% identity, such as 98% identity over the full length of SEQ ID NO: 7, wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 5 (hsSUMOI a) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease as defined in (i), under identical conditions.

One PRS may comprise, preferably consist of

(i) an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8), SEQ ID NO: 9 (bdNEDD8), or SEQ ID NO: 10 (hsNEDD8); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 85% identity, preferably at least 90% identity, more preferably 95% identity, and most preferably 98% identity over the full length of SEQ ID NO: 9; or at least 99% over the full length of identity over the full length of SEQ ID NO: 8; or at least 99% over the full length of identity over the full length of 10;

wherein the protease shown in SEQ ID NO: 11 (ssNEDPI ), SEQ ID NO: 12 (bdNEDPI ), or SEQ ID NO: 13 (hsNEDPI ) is capable of cleaving said PRS derivative with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to when using the corresponding parent PRS with the amino acid sequence of SEQ ID NO: 8, 9 or 10, respectively, under identical conditions.

The AT of the subunit comprising the ssNEDD8-, bdNEDD8- or hsNEDD8-PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence selected from the group consisting of amino acid sequences shown in SEQ ID NO: 11 (ssNEDPI ), SEQ ID NO: 12 (bdNEDPI ), and SEQ ID NO: 13 (hsNEDPI ), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 70% identity, more preferably with at least 80% identity, even more preferably with at least 90% identity, still even more preferably with at least 95% identity, most preferably with at least 98% identity over the full length of SEQ ID NO: 11 (ssNEDPI ); or with at least 70% identity, more preferably with at least 80% identity, even more preferably with at least 90% identity, still even more preferably with at least 95% identity, most preferably with at least 98% identity over the full length of SEQ ID NO: 13 (hsNEDPI ); or with at least 35% identity, preferably with at least 40% identity, more preferably with at least 50% identity, even more preferably with at least 60% identity, still even more preferably with at least 70% identity, most preferably with at least 80% identity, and even most preferably with at least 90% identity, such as 95% or 98% identity over the full length of SEQ ID NO: 12 (bdNEDP1 ); wherein said protease derivative, cleaves the PRS according to SEQ ID NO: 9 (bdNEDD8) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease as defined in (i), under identical conditions.

As shown in the examples and as explained with regard to bdSUMO/bdSENPI system above, it is preferred that the PRS comprising an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8), SEQ ID NO: 9 (bdNEDD8) or a PRS derivative thereof as defined in (ii) is comprised in the first subunit. More preferably, the first subunit comprises a PRS comprising an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8) or SEQ ID NO: 9 (bdNEDD8), in particular wherein the first subunit comprises a PRS comprising an amino acid sequence as shown in SEQ ID NO: 9 (bdNEDD8). In a most preferred embodiment, the first subunit comprises a PRS consisting of an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8) or SEQ ID NO: 9 (bdNEDD8). It is thus particularly preferred that the first subunit comprises a PRS consisting of an amino acid sequence as shown in SEQ ID NO: 9 (bdNEDD8).

The AT of the subunit comprising one of these preferred PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence selected from the group consisting of amino acid sequences shown in SEQ ID NO: 11 (ssNEDPI ), and SEQ ID NO: 12 (bdNEDPI ), or

(ii) a protease derivative having an amino acid sequence with at least 70% identity, more preferably with at least 80% identity, even more preferably with at least 90% identity, still even more preferably with at least 95% identity, most preferably with at least 98% identity over the full length of SEQ ID NO: 11 (ssNEDPI ); or with at least 35% identity, preferably with at least 40% identity, more preferably with at least 50% identity, even more preferably with at least 60% identity, still even more preferably with at least 70% identity, most preferably with at least 80% identity, and even most preferably with at least 90% identity, such as 95% or 98% identity over the full length of SEQ ID NO: 12 (bdNEDPI );

wherein said protease derivative, cleaves the PRS according to SEQ ID NO: 9 (bdNEDD8) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease as defined in (i), under identical conditions. In a more preferred embodiment, the AT of the subunit comprising the ssNEDD8 or bdNEDD8-PRS is cleaved off using a protease comprising, preferably consisting of the amino acid sequence selected from the group consisting of amino acid sequences shown in SEQ ID NO: 11 (ssNEDPI ), and SEQ ID NO: 12 (bdNEDPI ). In a most preferred embodiment, the AT of the subunit is cleaved off using the protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 12 (bdNEDPI ).

Likewise, one PRS may comprise, preferably consist of

(i) an amino acid sequence as shown in SEQ ID NO: 20 (xlUb); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 80% identity, preferably at least 90% identity, more preferably 95% identity, and most preferably 98% identity over the full length of SEQ ID NO: 20, wherein the protease shown in SEQ ID NO: 21 (xlUsp2), is capable of cleaving said PRS derivative with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 20 under identical conditions.

The AT of the subunit comprising such an xlUb-derived PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 21 (xlUsp2), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 80% identity, preferably at least 90% identity, more preferably 95% identity, and most preferably 98% identity over the full length of SEQ ID NO: 21 , wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 20 (xlUb) with at least 20%, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease as defined in (i).

Alternatively, one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 14 (scAtg8); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 80% identity, more preferably with at least 90% identity, even more preferably with at least 95% identity, and most preferably with at least 98% identity over the full length of SEQ ID NO: 14,

wherein the protease shown in SEQ ID NO: 15 (scAtg4), is capable of cleaving said PRS derivative with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 14 under identical conditions.

The AT of the subunit comprising such a scAtg8-derived PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 15 (scAtg4), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 80% identity, more preferably with at least 90% identity, even more preferably with at least 95% identity, and most preferably with at least 98% identity over the full length of SEQ ID NO: 15, wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 14 (scAtg8) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease as defined in (i).

The Atg4/Atg8 system in general is already described in the art (Li et al. JBC (201 1 ) 286(9): 7327-7338).

One can easily envision that other well-established proteases recognizing linear peptide motifs (e.g. TEV protease) constitute further groups of proteases with orthogonal specificity (cf. example 2 below). Accordingly, one PRS may comprise, preferably consist of the TEV protease recognition site shown in SEQ ID NO: 16 and 17.

The AT of the subunit comprising such an TEV-PRS is cleaved off using a TEV protease as shown in SEQ ID NO: 18 or a derivative thereof having an amino acid sequence with at least 80% identity, more preferably with at least 90% identity, even more preferably with at least 95% identity, and most preferably with at least 98% identity over the full length of SEQ ID NO: 18, wherein said protease derivative is capable of cleaving the TEV-PRS shown in SEQ ID NO: 16 with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the protease as shown in SEQ ID NO: 18. One example of such a derivative is the protease as shown in SEQ ID NO: 19.

However other orthogonal protease systems are also likely to work, such as PreScission protease, enterokinase, Factor Xa, intein systems, and the like, if a subunit contains the respective PRS.

As used herein, an amino acid sequence is said to have "X % sequence identity with SEQ ID NO: Y" over a defined length of amino acids if the sequence in question is aligned with said SEQ ID NO: Y and the sequence identity between those to aligned sequences is at least X%. Such an alignment can be performed using for example publicly available computer homology programs such as the "BLAST" program, such as "blastp" provided at the NCBI homepage at http://www.ncbi.nlm.nih.gov/blast/blast.cgi, using the default settings provided therein. Subsequently, identical residues are determined, such as by counting by hand, and a subsequent calculation of the percentage identity (PID) by dividing the number of identities over the indicated length of SEQ ID NO: Y gives "X % sequence identity". If a particular length is not specifically indicated, the sequence identity is calculated over the entire/full length of SEQ ID NO: Y. Further methods of calculating sequence identity percentages of sets of polypeptides are known in the art.

Preferably, the nature of amino acid residue changes by which the polypeptide having at least X% identity to a reference sequence differs from said reference sequence is a semi-conservative and more preferably a conservative amino acid residue exchange.

Amino acid Conservative exchange Semi-conservative exchange

A G; S; T N; V; C

C A; V; L M; 1; F; G D E; N; Q A; S; T; K; R; H

E D; Q; N A; S; T; K; R; H

F W; Y; L; M; H 1; V; A

G A S; N T; D; E; N; Q;

H Y; F; K; R L; M A

I V; L; M; A F; Y; W ; G

K R; H D; E N Q; S; T; A

L M; 1; V; A F; Y; W ; H; C

M L; 1; V; A F; Y; W ; C;

N Q D; E S; T; A; G; K; R

P V; 1 L; A; M; W; Y; S; T; C; F

Q N D; E A; S; T; L; M; K; R

R K; H N; Q ; S T; D; E; A

S A; T; G; N D; E R K

T A; S; G; N; V D; E R K; 1

V A; L; 1 M; T ; C N

W F; Y; H L; M l; V; C

Y F; W; H L; M l; V; C

Changing from A, F, H, I, L, M, P, V, W or Y to C is semi-conservative if the new cysteine remains as a free thiol. Changing from M to E, R or K is semi- conservative if the ionic tip of the new side group can reach the protein surface while the methylene groups make hydrophobic contacts. Changing from P to one of K, R, E or D is semi-conservative, if the side group is on the surface of the protein. Furthermore, the skilled person will appreciate that glycines at sterically demanding positions should not be substituted and that P should not be introduced into parts of the protein which have an alpha-helical or a beta sheet structure. Residues critical for the structure and activity of the PCS or protease, and which may therefore not be made subject of substitutions, can be identified by alanine-scanning mutagenesis, as generally known in the art.

The subunit(s) may further comprise a spacer between the AT and the PRS, and/or between the PRS and the subunit. In a preferred embodiment, the subunit(s) further comprise a spacer between the AT and the PRS. A typical spacer should be flexible and hydrophilic, without representing a substrate for endogenous proteases or comprising a PRS as defined herein. Usually, spacers having a high content of glycine and serine (as well as threonine and asparagine) are used. However, charged residues (especially negative charged residues) are not excluded. The skilled person will recognize suitable spacers.

The affinity tag (AT) may be any affinity tag suitable in the above-described method. In other words, any affinity tag may be used as long as it enables purification by affinity chromatography and as long as it is specific and does not interact with other affinity resins used in the method. For example, the AT may be a peptide tag, a covalent tag or a protein tag. Examples of a peptide tag are an Avi-tag, a CBP (calmodulin-binding peptide)-tag, a Flag-tag, a HA-tag, a polyHis- tag, a Myc-tag, a S-tag, a SBP-tag, a Softag 1 , a Softag 3, a V5-tag, a Strep-tag or a Xpress-tag. Examples of a covalent tag are Isopeptag and Spytag. Examples for a protein tag are BCCP, GST-tag, GFP-tag, MBP-tag, NusA-tag, GFP-tag or a thioredoxin-tag. The AT may be selected from the group consisting of a polyHis- tag, ZZ-tag, FLAG-tag, HA-tag, GST-tag, GST-epitope tag, GFP-tag, thioredoxin, epitope tag of thioredoxin, Avi-tag, or another peptide tag. Preferably, the AT is selected from a polyHis-tag, ZZ tag, FLAG tag, HA tag, and GST tag; more preferably the AT is selected from a polyHis-tag and a ZZ-tag. In practice, in the first affinity chromatography step a resin that allows for a quick and highly efficient capture of target complexes is preferred. For this purpose, the inventors routinely use a Ni²⁺ chelate resin along with a polyHis-tagged first subunit. The protease used for on-column cleavage must therefore not contain a polyHis-tag (see Fig. 7D, Fig. 8). For elution at this initial step, bdNEDPI is ideally suited as the slightly higher amount of protease needed for efficient cleavage (in comparison to bdSENPI ) can be efficiently removed during the following purification step. In the second affinity purification step several well-established matrices can be used, amongst them the IgG-resin binding to ZZ-tag, or any antibody-based resin directed against peptide tags (for review see Lichty et al. (2005); Waugh et al. (2005); Young et al. (2012), and Nilsson et al. (1997), supra). Thus, in a specific embodiment, the first subunit comprises a polyHis-tag, and preferably the second subunit comprises a ZZ-tag.

For elution at this second step it is preferred to use a protease featuring the highest possible specific activity as any added protease either has to be removed in an additional step or will remain in the final protein preparation as a contaminant. Therefore, it is recommended to use bdSENPI at this step. Further, a set of bdSENPI variants harboring different affinity tags are contemplated that can be used for efficient removal of the protease after on-column cleavage.

Therefore, in a specific embodiment, the first subunit comprises a NEDD8-PRS or NEDD8-PRS derivative as defined above, preferably the bdNEDD8-PRS, and the second subunit comprises a SUMO-PRS or SUMO-PRS derivative as defined above, preferably the bdSUMO-PRS. In a very preferred embodiment for the purification of a complex comprising two subunits, the following setup is chosen: polyHis-bdNEDD8-subunit1 and ZZ-bdSUMO-subunit2. The AT is then cleaved off using the corresponding protease, as defined for the PRS/protease systems above. The proteases itself as described in the following can be of great benefit for purifying stoichiometric protein complexes. Accordingly, provided is a protease having an amino acid sequence with at least 45% identity, preferably with at least 50% identity, more preferably with at least 60% identity, even more preferably with at least 70% identity, still even more preferably with at least 80% identity, most preferably with at least 90% identity, and even most preferably with at least 95% identity, such as 98% identity over the full length of SEQ ID NO: 2 (bdSENPI ), wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 1 (bdSUMO) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 2. Preferably the protease comprises the amino acid sequence shown as amino acids 1-224 in SEQ ID NO: 2 (bdSENP1₂₄e- 481 ). More preferably, the protease consists of the amino acid sequence shown as amino acids 1-224 in SEQ ID NO: 2 (bdSENP1₂₄8-48i )-

Also provided is a protease having an amino acid sequence with at least 70% identity, preferably with at least 80% identity, more preferably with at least 90% identity, most preferably with at least 95% identity, and even most preferably with at least 98% identity over the full length of SEQ ID NO: 11 (ssNEDPI ), wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 8 (ssNEDD8) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% as compared to the parent protease with the amino acid sequence of SEQ ID NO: 11. In a preferred embodiment, the protease comprises the amino acid sequence as shown in SEQ ID NO: 11 (ssNEDPI ). In a more preferred embodiment, the protease consists of the amino acid sequence as shown in SEQ ID NO: 11 (ssNEDPI ).

Further provided is a protease having an amino acid sequence with at least 35% identity, preferably with at least 40% identity, more preferably with at least 50% identity, even more preferably with at least 60% identity, still even more preferably with at least 70% identity, most preferably with at least 80% identity, and even most preferably with at least 90% identity, such as 95% or 98% identity over the full length of SEQ ID NO: 12 (bdNEDPI ), wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 9 (bdNEDD8) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100% as compared to the parent protease with the amino acid sequence of SEQ ID NO: 12, In a preferred embodiment, the protease comprises the amino acid sequence as shown in SEQ ID NO: 12 (bdNEDPI ). In a more preferred embodiment, the protease consists of the amino acid sequence as shown in SEQ ID NO: 12 (bdNEDPI ).

Still another protease is provided, having an amino acid sequence with at least 80% identity, preferably with at least 85% identity, more preferably with at least 90% identity, most preferably with at least 95% identity, and even most preferably with at least 98% identity over the full length of SEQ ID NO: 21 (xlUsp2), wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 20 (xlUb) with at least 20% activity, preferably at least 30% activity, more preferably at least 40% activity, even more preferably at least 50% activity, still more preferably at least 60% activity, still even more preferably at least 70% activity, most preferably at least 80% activity, even most preferably at least 90% activity such as more than 100%, as compared to the parent protease with the amino acid sequence of SEQ ID NO: 21. Preferably the protease comprises the amino acid sequence as shown in SEQ ID NO: 21 (xlUsp2); more preferably the protease consists of the amino acid sequence as shown in SEQ ID NO: 21 (xlUsp2).

Any of the proteases described above may further comprise an affinity tag, in particular if said protease is used as the "final" protease in the above-described method. The affinity tag of the protease may be chosen among those described above. Preferably, the affinity tag is a polyHis-tag.

In addition, nucleic acid molecules are contemplated, which encode one of the proteases described and disclosed herein.

The main focus of this study has been the identification, recombinant expression and characterization of protease/substrate pairs that have potential as general tools for purification of recombinant proteins and protein complexes. With the exception of TEV protease, these proteases allow for an efficient substrate cleavage at 0°C. Importantly, six of the seven proteases fall into five groups with orthogonal substrate specificity: (i) scUlpl and bdSENPI , (ii) bdNEDPI (iii) scAtg4, (iv) xlUsp2 and (v) TEV protease. The natural substrate of one protease group (as defined above) will therefore not be efficiently recognized by a protease from another group. The ssNEDPI enzyme is special as it is strictly orthogonal to groups (i), (iii) and (v), but shows some degree of cross-reactivity on an ubiquitin- containing substrate (see Fig. 5).

Species specificity and sequence conservation

The orthologous NEDP1/NEDD8 pairs from Brachypodium and salmon behave similar in most assays using the standard P1 ^'-Ala substrates. This is surprising, especially when considering the moderate degree of conservation between the corresponding NEDP1 enzymes (see Fig. 3). According to the available structure of the human NEDD8-NEDP1 complex (Shen et al. (2005) EMBO J 24: 1341 - 1351 ), the significant differences seen with respect to their P1 ^' preferences (Fig. 12) can most probably be attributed to significant exchanges in protease residues contacting the substrate C-terminal of the scissile bond. The pronounced promiscuity of NEDP1 enzymes towards orthologous substrates can easily be explained by the striking conservation between NEDD8 proteins: From a total of only 12 amino acid exchanges between salmon and Brachypodium, only 5 are non-conservative (Fig. 3). The two exchanges present within the putative interface with the proteases do not seem to crucially influence the recognition by the protease. The species promiscuity of NEDP1 enzymes has interesting practical implications: As a given NEDD8 substrate can be cleaved by both, bdNEDPI and ssNEDPI , the protease used for cleavage can be chosen freely. While bdNEDPI might be the best alternative under standard conditions, ssNEDPI is remarkably insensitive towards high salt or a suboptimal residue in the substrate^'s P1 ^'- position. The salmon enzyme might thus be the protease of choice when cutting suboptimal substrates or cleaving at special buffer conditions.

Compared to NEDD8, the SUMO orthologues analyzed herein show a low degree of sequence conservation (Fig. 3). Moreover, while the yeast SUMO (Smt3p) has a high similarity to the human SUMO1 isoform (hsSUMOI ), the bdSUMO is more related to hsSUMO2. Similar to their substrates, also the SUMO proteases from yeast, Brachypodium and human show a low degree of sequence conservation. Structural alignments (including structure predictions for the Brachypodium enzyme) (Armougom (2006) Nucleic Acids Res. 34: W604-8), however, indicate that all these enzymes adopt a similar three-dimensional structure. According to published structure of the yeast scSUMO*scUlp1 complex (Mossessova et al. (2000), supra), the substrate^»enzyme interfaces of the respective yeast and Brachypodium complexes differ in a significant number of residues that may easily account for the differences regarding cleavage efficiency (Fig. 10) and salt- or P1 ^'- sensitivity (Fig. 1 1 , Fig. 12) that could be detected in our assays. As another consequence of these exchanges, the two enzymes cleave their natural substrates better than substrates containing orthologous SUMO variants (Fig. 6): While the yeast enzyme scUlpl significantly cleaves bdSUMO-containing substrates, only «5% of the scSUMO substrate is cleaved by the Brachypodium enzyme at conditions required for =95% cleavage of its own bdSUMO substrate. In kinetic assays, bdSENPI cleaves the corresponding Brachypodium substrate even >150- fold more efficiently than the substrate containing scSUMO (Fig. 6).

Protease/substrate ratio

A variety of commercial vectors include the TEV protease recognition site ("TEV site") e.g. after the GST tag. TEV protease is thus often considered as the first choice for removing affinity tags from target proteins. While comparing the catalytic properties of a stabilized variant of TEV protease to proteases of the SENP1 and NEDP1 enzyme families, it turned out that TEV protease has major limitations that should be considered in practice.

First, the effective turnover rate of TEV protease is poor. Even at 25°C and at high substrate concentrations, each molecule of TEV protease can cleave only -150 substrate molecules per hour (Fig. 14). In addition, because of the high K_M of the reaction (50-90μΜ) (Kapust et al. (2002) Biochem Biophys Res Commun. 294: 949-955; Kapust et al. (2001 ) Protein Eng. 14: 993-1000; Parks et al. (1995) Virology 210: 194-201 )), this turnover rate can only be reached at exceedingly high substrate concentration (>100-200μΜ). At lower substrate concentrations, the number of substrate molecules cleaved per protease drops significantly. Consequently, regardless of the concentration of substrate to be cleaved, roughly the same amount of protease is required. In practice, these properties have two major consequences. First, a complete cleavage by TEV protease is hard to achieve and generally requires long incubation times at elevated temperature (generally 16-30°C, as recommended by the commercial suppliers) or high enzyme concentrations. Second, any cleavage product will be contaminated with a rather high fraction of protease unless the substrate can be supplied in unreasonably high concentrations (>200μΜ). For applications in an analytical or semi-preparative scale, the potential of TEV protease is therefore limited.

In comparison, the new proteases characterized here are highly efficient tag- removing enzymes. Remarkably, when using these enzymes, the substrate/protease ratio required for efficient cleavage remains rather constant even at low substrate concentrations. Therefore, especially bdSENPI , bdNEDPI and ssNEDPI are ideally suited for driving tag removal to completion. Importantly, when using these enzymes, the amount of protease used for cleavage can be lowered according to the substrate concentration. As a rule of thumb, at 0°C one molecule of bdSENPI will cleave roughly 2-4 substrate molecules per second, i.e. in practice, a 5.000-15.000-fold molar excess of substrate can easily be cleaved within one hour at 0°C. The NEDD8-specific enzymes have an approximately 10- fold lower turnover rate. Nevertheless, the two NEDP1 proteases can still digest an up to 1000-fold excess of substrate within one hour at 0°C.

At such enzyme/substrate ratio, the remaining "contaminant" protease that is used for cleavage can be neglected for the most common laboratory purposes. If desired, the protease concentration used for cleavage can, however, be further drastically decreased if the cleavage reaction is performed at higher temperature or for a longer time. This is easily possible as the characterized SUMO-and NEDD8-specific proteases remain fully active even after over-night incubation at 37°C or 20°C, respectively. A complete removal of the protease is possible using a protease variant harboring an engineered affinity tag. Together, these measures should allow for the removal of even trace amounts of protease.

Salt tolerance

Importantly, the most active enzyme provided herein, bdSENPI , even outperforms its yeast orthologue in several aspects: At standard conditions (see e.g. Fig. 4 and Fig. 10A), bdSENPI has a 2-3-fold higher specific activity as compared to scUlpl . In addition, bdSENPI can efficiently cleave substrates in a wide range of salt conditions while the yeast counterpart significant loses activity at NaCI concentrations above 250mM (Fig. 1 1 ). This finding contrasts the relatively mild salt sensitivity (30% remaining activity at 1 M NaCI) reported for scUlpl in the literature (Malakhov et al. (2004), supra). In this earlier report, however, the protease concentration used was significantly (presumably =10-fold) higher than needed for complete substrate cleavage under low-salt conditions. Consequently, a salt-induced decrease in protease activity by ~ 90% would have escaped detection completely.

Similar to bdSENPI , also the two NEDP1 enzymes show an excellent tolerance to high salt conditions. These enzymes can therefore conveniently be used as tag- removing proteases in a variety of different buffers.

On-column cleavage

Most critical for practical applications, all new proteases characterized here are active at 0°C (Fig. 4) and can be used for on-column cleavage (Fig. 7). Compared to traditional removal of affinity tags in solution, on-column cleavage allows for the mild elution of highly pure proteins without the need for time-consuming buffer exchange or "reverse chromatography" steps. The examples herein show that, irrespective of the protease used, efficient on-column cleavage requires a protease not directly interacting with the affinity resin. This conclusion is most probably generally true and needs to be considered for the intended experimental design. For example, as most commercial protease preparations contain polyHis- tags, such proteases cannot be used for on-column cleavage from Ni²⁺ chelate resins.

For purification of proteins using a single affinity chromatography step, the inventors routinely elute the target proteins directly from Ni²⁺ chelate columns using 30nM (untagged) bdSENPI within one hour at 4°C. In the vast majority of cases, an efficient release of the target protein is observed, generally yielding target protein concentrations between 100 and 300μΜ (up to 120 mg/ml), probably mostly limited by the binding capacity of the resin.

Since the above-described method comprises at least two affinity chromatography steps, there is further contemplated a kit of parts, comprising at least two proteases selected from

(i) a bdSENPI -derived protease, e.g. the bdSENPI protease, as defined above,

(ii) a ssNEDPI -derived protease and bdNEDPI derived protease, e.g. the ssNEDPI protease or bdNEDPI protease, as defined above

(iii) a TEV-derived protease, e.g. the protease shown in SEQ ID NO: 18 or SEQ ID NO: 19, as defined above,

(iv) an Atg4-derived protease, e.g. the scAtg4 protease, as defined above, and

(v) an Usp2-derived protease, e.g. the xlUsp2 protease, as defined above. However, the kit may also comprise more than two proteases, such as three, four or five proteases, which are orthogonal to each other. For example, the kit may comprise at least two, such as three proteases selected from (i)-(iii). More preferably the kit comprises two proteases selected from (i)-(iii). Even more preferably, the kit comprises two proteases selected from (i) and (ii). In a most preferred embodiment, the kit comprises a bdNEDPI -derived protease and a bdSENPI -derived protease, as defined above, e.g. the bdNEDPI and bdSENPI protease disclosed herein. Advantageously at least one of the proteases comprised in the kit comprises an affinity tag, as further described above.

The orthogonal proteases disclosed herein as well as the kit comprising these orthogonal proteases can be advantageously used in a method of purifying stoichiometric protein complexes comprising at least two subunits, wherein said at least two subunits comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS), and optionally a spacer between the AT and the PRS, and wherein the AT of each of said at least two subunits differs from each other so to allow specific affinity chromatography, and wherein the PRS of each of said at least two subunits is cleavable by a protease which is orthogonal to the PRS of the other subunit(s). In a preferred embodiment the method is further defined as described above. In particular, the orthogonal proteases disclosed herein as well as the kit comprising these orthogonal proteases can be advantageously used for on-column cleavage in an affinity chromatography.

In summary, a generally applicable method for purifying stoichiometric protein complexes is provided. It is parallelizable and therefore suitable for automation. The method requires a system of orthogonal proteases, which are also provided herein, which orthogonal proteases are capable of specifically cleaving affinity tags. The use of said orthogonal proteases allows an almost free choice of the N- terminus of any subunit or target protein following the PRS, including the authentic N-terminus. In particular the newly provided orthogonal proteases are capable of specifically cleaving in a wide range of buffer conditions, and can be suitably used for on-column cleavage.

For example, bdSENPI is highly specific, exhibits an extraordinary activity, even at 0 °C, which is higher than the specific activity of, e.g., scUlpl or TEV protease, has low P1 '-sensitivity, and shows a high salt tolerance. It demonstrated a higher species-specificity and merely moderate cross-reactivity with scUlpl . The NEDD8- specific protease from Brachypodium distachyon (bdNEDPI ) has not been annotated or predicted as a protein, in particular not as a protease, and the NEDD8-specific protease from Salmo salar (ssNEDPI ) has not been biochemically characterized yet. Like bdSENPI both proteases are highly specific, show high specific activity even at 0-4°C, and exhibit a high salt tolerance and a low P1 '-sensitivity.

In addition, the invention describes a quintary orthogonal protease system comprising bdSENP1/scUlp1 -proteases, the bdNEDPI -protease, TEV protease, the scAtg4-protease and the xlUsp2-protease.

In the following, the present invention is illustrated by figures and examples, which are not intended to limit the scope of the present invention. All references cited herein are explicitly incorporated by reference.

DESCRIPTION OF THE FIGURES

Figure 1 : Comparison of methodologies used to purify stoichiometric binary protein complexes.

(A) A binary complex (subunits T1 and T2) is pre-formed from purified individual components. Further chromatographic steps are required to remove surplus single subunits and binding-incompetent subunits.

(B) Purification of a stoichiometric binary complex via two consecutive affinity chromatography steps involving on-column cleavage. This strategy is further detailed in Figure 2. Figure 2: Purification of a stoichiometric binary complex using two consecutive affinity purification steps with on-column cleavage.

(A) Description of components used in the schemes (B)-(l). After co-expression of the two complex components T1 and T2 within one cell (B), the binary complex comprising the two subunits T1 and T2 is separated from host proteins and single subunits T1 and T2 by two consecutive affinity chromatography steps using orthogonal tags and protease recognition sites on each of the proteins (C)-(l). The binary complex (and surplus subunit T1 ) is bound to affinity resin 1 via the affinity tag on subunit T1 (C). After washing, specifically bound proteins are released from the resin using a site-specific protease recognizing the protease recognition site (D, E). All contaminant proteins not containing the proper protease recognition site will remain bound to the resin. In the second affinity chromatographic step (F-H), the binary complex Τ1 ·Τ2 is separated from surplus subunit T1 by binding to affinity resin 2 specifically recognizing the tag fused to component T2 and similarly cleaved off with a component T2-specific protease. The protease can be removed via an adequate affinity resin (I). Protein complexes with more than two subunits can be purified in analogously using an appropriate number of orthogonal affinity matrices and orthogonal protease systems.

Figure 3: Structure-based sequence alignment of SUMO- and NEDD8- orthologues and SUMO/NEDD8-specific proteases with their human orthologues. Relevant protein sequences were assembled from available EST and genomic sequence and aligned based on the results obtained form the Expresso server (see Example 1 ). Residue conservation at each position was classified as similar (°) or identical (·). Amino acids near the interface to the respective binding partner were highlighted in grey boxes. Residues directly involved in peptide bond hydrolysis are marked in bold. No structures were available for bdSUMObdSENPI , bdNEDD8^»bdNEDP1 and ssNEDD8^»ssNEDP1.

Figure 4: Activity of tag-cleaving proteases.

A: General design of protease substrates. All substrates contain an N-terminal polyHis-tag (Hisi₄ or Hisi₀), a protease recognition site (box left of the scissile bond) and the target protein MBP (maltose-binding protein; MBP). To ensure equivalent cleavage conditions, in SUMO-, NEDD8-, scAtg8 and xlUb-containing substrates the scissile bond is followed by identical sequences. B: Protease titration. Protease substrates (100μΜ) sketched in (A) were incubated for 1 h at 0°C (left) or at 25°C (right) in the presence of the corresponding proteases. Proteases were titrated down from 10μΜ to 1 nM. Reactions were stopped by dilution in hot SDS sample buffer. Cleavage products were separated by SDS-PAGE and stained with Coomassie G250. Shown are the non-cut (full length) proteins (fl) and the larger cleavage products (lcp). Bands of the molecular weight marker (M_r) correspond to 40kD, 50kD (more intense band) and 60kD (not always visible).

Figure 5: Cross-reactivity between various substrate/protease systems.

A: 100μΜ of indicated substrates (100μΜ) were incubated with 10μΜ of indicated proteases and protease fragments for 3h at 25°C. The protease concentrations used are thus up to 10.000-fold higher than the concentrations required for efficient cleavage of their own substrates. Even at such drastic conditions, the SUMO-proteases, bdNEDPI , scAtg4, xlUsp2, and TEV protease represent five orthogonal groups of proteases. The ssNEDPI enzyme, however, shows some proteolytic activity on the xlUb-MBP substrate after 3h incubation and is therefore not strictly orthogonal to xlUspl under these conditions. Numbers in brackets refer to the amino acid numbers of full-length bdSUMO or full-length bdSENPI , respectively.

B: 100μΜ of the indicated substrates were incubated at 0°C for one hour with 1 μΜ of indicated proteases. Under these conditions, both SUMO proteases cleave substrates containing SUMOs of their own species more efficiently as compared to substrates containing an orthologous SUMO. The NEDP1 enzymes do not show any significant species preference.

Figure 6: The SUMO-specific proteases show a clear species preference for their respective SUMO substrates, but are not fully orthogonal to each other. 100μΜ of indicated substrates were cleaved at various conditions with either scUlpl or bdSENPI . The grey bars in the upper left and lower right panels in each of A and B mark lanes with efficient digestion of cognate protease/substrate pairs; the bars in the lower left and upper right panels in each of A and B, highlight lanes showing efficient digestion of substrates by the orthologous protease.

A: One hour incubation at 0°C with varying concentrations of protease. A «40-fold higher concentration of bdSENPI is needed for efficient cleavage of scSUMO- MBP as compared to scUlpl . In contrast, efficient cleavage of bdSUMO-MBP requires «10-fold higher concentration of scUlpl as compared to bdSENPI .

B: Time course at 0°C with fixed concentration (300nM) of protease. bdSENPI needs >150-times longer than scUlpl for >95% cleavage of the orthologous yeast substrate. In contrast, scUlpl needs only =15-times longer than bdSENPI to cleave >95% of the orthologous Brachypodium distachyon substrate.

Figure 7: bdSENPI , bdNEDPI , scAtg4 and xlUsp2 can be used for on-column cleavage.

A: Schematic representation of substrates used for on-column cleavage experiments using bdSENPI and bdNEDPI .

B, C: A Ni²⁺ chelate resin was pre-loaded with similar amounts of His-i -bdSUMO- GFP and His ₄-bdNEDD8-mCherry. 50μΙ aliquots were treated with indicated concentrations bdSENPI (B) or bdNEDPI (C) for 1 hour at 0°C. Control incubations were performed with buffer or with buffer containing 400mM imidazole. Resins and eluates were photographed upon illumination at 366nm. GFP and mCherry in the eluate fractions were quantified via their absorption at 488nm and 585nm, respectively. Numbers below the eluate fractions show the quantification results. Efficient on-column cleavage (>95% elution) occurred with 20nM bdSENPI and 300nM bdNEDPI , respectively. The cleavage was specific as even at a >30-fold higher protease concentration, no significant elution of the nonspecific target protein was evident.

D: 1 μΜ of either Hisi₄-tagged or non-tagged bdNEDPI protease was incubated for 1 h at 25°C with a Ni²⁺ chelate resin pre-loaded with His ₄-bdSUMO-GFP and His-i -bdNEDD8-mCherry. Efficient release of mCherry (the NEDP1 -specific target protein) from the resin was only evident when using a polyHis-tag-free protease (upper panel). In parallel, the activity of the protease preparations used in (A) was assayed using a standard solution assay (lower panel, compare to Fig. 4). Cleavage of the soluble substrate (bdNEDD8-MBP) was analyzed by SDS-PAGE after incubation for 1 h at 0°C. Shown are the non-cleaved (fl) protein and the larger cleavage product (lcp). On the soluble substrate, both the polyHis-tagged and the non-tagged proteases were equally active.

E: Schematic representation of substrates used for on-column cleavage experiments using scAtg4 and xlUsp2. F, G: Analogous to (B) and (C), on-column cleavage using scAtg4 or xlUsp2 was analyzed using a Ni²⁺ chelate resin pre-loaded with Hisi₄-scAtg8-mCherry and Hisi₄-xlUb-GFP. Specific substrate release was observed after 1 h at 0°C using 6μΜ xlUsp2 and 10μΜ scAtg4, respectively.

Figure 8: On-column cleavage using polyHis-tagged and non-tagged TEV protease. A Ni²⁺ chelate resin was separately loaded with His-n-TEV-GFP (A) or His ₀-ZZ-TEV-GFP (B). Similarly, IgG sepharose was loaded with His ₀-ZZ-TEV- GFP (C). 50μΙ aliquots of loaded resins were treated with indicated concentrations of polyHis-tagged or non-tagged TEV protease for 1 hour at 25°C. Control incubations were performed with buffer or 500 mM imidazole. Resins and eluates were photographed upon illumination at 366 nm. GFP in the eluate fractions was quantified via its absorbance at 488 nm (numbers are given below the eluate fractions). Release of the fluorescent target proteins from the affinity matrices was less efficient when TEV protease directly interacted with the affinity resin: PolyHis- tagged TEV protease matched the efficiency of the non-tagged enzyme in cleaving soluble or IgG Sepharose-immobilized substrates (C), but performed far worse in releasing substrates from a Ni²⁺ chelate resin (A, B). Combined, these experiments indicate that TEV protease can in principle be used for on-column cleavage of substrates. For efficient elution within one hour, however, TEV protease needs to be applied in high concentrations (3-1 ΟμΜ) even at 25°C. Preparative-scale purifications using TEV protease are therefore expensive and lead to significant protease contaminations of the final product.

Figure 9: Purification of a tag-free binary complex with 1 :1 stoichiometry.

His₁₄-bdNEDD8-MBP and ZZ-bdSUMO-Darpin were co-expressed in E. coli (lane1 -4). In the first affinity purification step (lanes 5-8), the cleared lysate (lanes 4 and 5) containing the binary complex along with an excess Hisi₄-bdNEDD8-MBP and binding-incompetent ZZ-bdSUMO-Darpin was applied to a Ni²⁺ chelate resin. The binary ZZ-bdSUMO-Darpin*His-|₄-bdNEDD8-MBP complex and the excess Hisi₄-bdNEDD8-MBP efficiently bound to the resin. Most lysate proteins and binding-incompetent Darpin (*) remained in the flow-through fraction (lane 6). Treatment for 1 h at 0°C with 0.5μΜ of bdNEDPI efficiently released the ZZ- bdSUMO-Darpin'MBP complex along with free MBP (lane 7); the His_i4-bdNEDD8- tag and most contaminant lysate proteins remained bound to the resin (lane 8). In the second step (lanes 9-12), the ZZ-bdSUMO-Darpin^eMBP complex was bound to a ZZ-specific affinity resin, thereby removing all MBP not complexed to MBP as well as the bdNEDPI enzyme. Upon incubation with 30nM of bdSENPI , the tag- free binary Darpin^»MBP complex was released from the resin within one hour at 0°C (lane 1 1 ). Samples corresponding to 50mOD of cells (cells or lysates) or 1/1000 of the total purification (fractions during purification) were analyzed by SDS-PAGE and Coomassie staining. The stoichiometric nature of the complex could be verified by gel filtration (not shown).

Figure 10: Cleavage kinetics, temperature dependence and temperature stability of tag-cleaving proteases.

A: Cleavage kinetics. Protease substrates were incubated with the indicated concentrations of the corresponding proteases at 0°C. Samples were taken at various time points.

B: Temperature dependence. Proteases were allowed to cleave their respective substrates for 1 h at various temperatures. Note that in comparison to A, the protease concentration was significantly reduced. bdSENPI is consistently more active than scUlpl while bdNEDPI outperforms ssNEDPI at lower temperatures.

C: All proteases were incubated for 16h at indicated temperatures in the absence of oxygen. Thereafter, the remaining activity was assayed by treating 100μΜ of substrate with indicated concentrations of the corresponding protease for h at

0°C (SENP1 and NEDP1 enzymes, scAtg4 and xlUsp2) or 25°C (TEV protease).

Figure 11 : Salt sensitivity of tag-cleaving proteases.

A: Substrates were incubated for one hour at 0°C with the corresponding proteases in buffer containing the indicated concentrations of NaCI. Strikingly, scUlpl and scAtg4 show a pronounced sensitivity to high NaCI concentrations. B, C: scSUMO- (B) or bdSUMO- (C) containing substrates were incubated at 0°C in the presence of 250mM (upper panels) or 1 M NaCI (lower panels) with 300nM of their corresponding protease. Samples were taken after various time points and analyzed by SDS-PAGE. Grey bars mark lanes with efficient digestion of cognate protease/substrate pairs. Note that scUlpl needs =50-times longer to digest 95% of its cognate substrate in the presence of 1 M NaCI as compared to 250mM NaCI. In contrast, in the presence of 1 M NaCI, the activity of bdSENPI is only decreased by a factor of «3. Figure 12: Preference of SENP1/Ulp1 -proteases, NEDP1 proteases and TEV protease for residues in the P1 ^' position.

A: Schematic representation of protease substrates with different P1 ^' residues. Substrates follow the general outline shown in Fig. 2A. To analyze the sensitivity for non-preferred amino acid residues C-terminal of the scissile bond (P1 ^' position), this position was mutated to methionine (Met), tyrosine (Tyr), arginine (Arg), glutamic acid (Glu), or proline (Pro). For TEV substrates, the respective residues were inserted before the original glycine residue.

B: scUlpl and bdSENPI substrates with different residues in the P1 ^' position were incubated for 1 hour with various concentrations of their dedicated proteases at 0°C. To facilitate a direct comparison between orthologous SUMO-specific proteases, identical concentrations of the respective proteases were used. Gray bars mark lanes with efficient digestion of cognate protease/substrate pairs.

C: Similar to (B), bdNEDPI- and ssNEDPI -specific substrates with altered ΡΓ- positions were incubated for 1 hour with the indicated concentrations of their cognate proteases at 0°C. Note that in comparison to (B) the protease concentrations used are 10-fold higher. The salmon NEDP1 enzyme shows a remarkably low P1 ^' sensitivity, but is unable to cleave a P1 ^'-Pro substrate.

D: Various TEV substrates were incubated for 1 hour with the indicated concentrations of TEV protease. To allow for efficient cleavage, the cleavage was performed at 25°C. Note that, in comparison to the proteases in (A), TEV protease was used at a 100-fold higher concentration.

Figure 13: Preference of scAtg4 and xlUsp2 for residues in the P1 ^' position.

A: Schematic representation of the protease substrates used in (B) and (C).

B, scAtg4-specific substrates sketched in (A) were incubated for 1 hour at 0°C with various concentrations of scAtg4. The scAtg4 protease shows only a mild sensitivity for residues in the P1 ' position. At 1 μΜ concentration, the protease can cleave an 80-100-fold excess of most substrates. Similar to other proteases, scAtg4 does not accept substrates harboring a proline in the P1 ' position.

C: Indicated substrates were incubated with various concentrations of xlUsp2 for h at 0°C. The xlUsp2 protease cleaves all tested substrates (except the ΡΓ-Pro substrate) with virtually identical efficiency. Remarkably, xlUsp2 even shows significant cleavage activity on the P1 ^'-Pro substrate. Figure 14: Digestion efficiency at different substrate concentrations. In two different setups, indicated concentrations of substrate were incubated with their corresponding protease. A fraction of each reaction corresponding to 1 .2 pg (-20 pmol) of substrate protein was analyzed by SDS-PAGE and Coomassie staining. Due to the variable substrate concentration, the absolute volume of the digestion reaction analyzed by SDS-PAGE was reciprocally proportional to the substrate concentration in the digestion reaction.

A: The concentrations of both substrate and protease were titrated at constant protease:substrate ratio. Note that the cleavage efficiency remains similar for SUMO- and NEDD8-, scAtg8- and xlUb- specific proteases also at low protein concentrations, while TEV protease has problems cleaving at low substrate concentrations.

B: The substrate concentration was varied from 300μΜ to 3μΜ while keeping the protease concentration constant. In this setup, the fraction of substrate cleaved by the TEV protease remains rather constant. For all other proteases assayed, the relative amount of cleaved substrate increases at lower substrate concentrations (i.e. at higher protease:substrate ratio).

Figure 15: Truncation analysis of the bdSUMO/bdSENPI -, bdNEDD8/bdNEDP1 -, and scAtg8/scAtg4 substrate/protease pairs.

A, General layout of the constructs used in B-D. Truncations at the indicated positions were introduced by site-directed mutagenesis and verified by DNA sequencing.

B-D, Analysis of in-vivo cleavage efficiency of substrate/protease pairs harboring defined truncations. Indicated constructs were expressed in E.coli for 1 .5h at 37°C. Cells were lysed in hot SDS sample buffer. Similar amounts of total protein were analyzed by SDS-PAGE and Coomassie staining. Separate panels show cleavage assays using N-terminally truncated bdSUMO (B, left panel), N-terminally truncated bdSENPI (B, middle panel), C-terminally truncated bdSENPI (B, right panel), N-terminally truncated bdNEDD8 (C, left panel), N-terminal truncated bdNEDPI (C, middle panel), C-terminally truncated bdNEDPI (C, right panel), N- terminally truncated scAtg8 (D, left panel), N-terminally truncated scAtg4 (D, middle panel), and C-terminally truncated scAtg4 (D, right panel), respectively. ln-vivo cleavage of a given fusion protein is expected if both the protease recognition site and the protease moiety are functional.

Figure 16: Detailed cleavage analysis of selected bdSUMO and bdSENPI truncations. 100μΜ of indicated substrates were incubated with various concentrations of indicated bdSENPI fragments for 1 h at 0°C. The reaction was stopped by dilution in hot SDS sample buffer. Cleavage products were analyzed by SDS-PAGE and Coomassie staining. Shown are the full-length substrate protein (fl) and the larger cleavage product (lcp). The general substrate design followed the scheme depicted in Figure 4A. Numbers in brackets refer to the amino acid numbers of full-length bdSUMO.

Figure 17: Activity of full-length TEV(SH) protease and a TEV(SH) variant lacking the six C-terminal amino acids.

100μΜ of a TEV substrate (ZZ-TEV-MBP) was incubated with the indicated concentrations of either full length TEV(SH) protease (top) or a TEV(SH) variant lacking the six C-terminal amino acids (TEV(SH)AC6; bottom) for one hour at 0°C. The reaction was stopped by dilution in hot SDS sample buffer. Cleavage products were separated by SDS-PAGE and stained with Coomassie G250. Shown are the non-cut (full length) proteins (fl) and the larger cleavage products (lcp). The C- terminal truncation has no influence on the catalytic activity.

DESCRIPTION OF THE SEQUENCES

SEQ ID NO: 1 (bdSUMO amino acids 21-97)

HINLKVKGQDGNEVFFRIKRSTQLKKLMNAYCDRQSVDMTAIAFLFDGRRLRAEQ TPDELEMEDGDEIDAMLHQTGG SEQ ID NO: 2 (bdSENPI amino acids 248-491 )

PFVPLTDEDEDNVRHALGGRKRSETLSVHEASNIVITREILQCLNDKEWLNDEVIN LYLELLKERELREPNKFLKCHFFNTFFYKKLINGGYDYKSVRRWTTKRKLGYNLID CDKIFVPIHKDVHWCLAVINIKEKKFQYLDSLGYMDMKALRILAKYLVDEVKDKSG KQIDVHAWKQEGVQNLPLQENGWDCGMFMLKYIDFYSRDMELVFGQKHMSYF RRRTAKEILDLKAG

SEQ ID NO: 3 (scSUMO/Smt3p amino acids 23-98) HINLKVSDGSSEIFFKIKKTTPLRRLMEAFAKRQGKEMDSLRFLYDGIRIQADQTP EDLDMEDNDIIEAHREQIGG

SEQ ID No: 4 (scUlpl amino acids 403-621 )

LVPELNEKDDDQVQKALASRENTQLMNRDNIEITVRDFKTLAPRRWLNDTIIEFFM KYIEKSTPNTVAFNSFFYTNLSERGYQGVRRWMKRKKTQIDKLDKIFTPINLNQS HWALGIIDLKKKTIGYVDSLSNGPNAMSFAILTDLQKYVMEESKHTIGEDFDLIHLD CPQQPNGYDCGIYVCMNTLYGSADAPLDFDYKDAIRMRRFIAHLILTDALK

SEQ ID NO: 5 (hsSUMOI a; Homo sapiens SUMOI a, amino acids 21 -97)

YIKLKVIGQDSSEIHFKVKMTTHLKKLKESYCQRQGVPMNSLRFLFEGQRIADNH TPKELGMEEEDVIEVYQEQTGG

SEQ ID NO: 6 (hsSUMO2; Homo sapiens SUMO2, amino acids 17-93)

HINLKVAGQDGSVVQFKIKRHTPLSKLMKAYCERQGLSMRQIRFRFDGQPINETD TPAQLEMEDEDTIDVFQQQTGG

SEQ ID NO: 7 (hsSENP1 ; Homo sapiens SENP1 , amino acids 419-644)

EFPEITEEMEKEIKNVFRNGNQDEVLSEAFRLTITRKDIQTLNHLNWLNDEIINFYM

NMLMERSKEKGLPSVHAFNTFFFTKLKTAGYQAVKRWTKKVDVFSVDILLVPIHL

GVHWCLAWDFRKKNITYYDSMGGINNEACRILLQYLKQESIDKKRKEFDTNGW

QLFSKKSQEIPQQMNGSDCGMFACKYADCITKDRPINFTQQHMPYFRKRMVWEI

LHRKLL

SEQ ID NO: 8 (ssNEDD8; Salmo salar NEDD8)

MLIKVKTLTGKEIEIDIEPTDKVERIKERVEEKEGIPPQQQRLIYSGKQMNDEKTAA DYKIQGGSVLHLVLALRGG

SEQ ID NO: 9 (bdNEDD8; Brachypodium distachyon NEDD8)

MIKVKTLTGKEIEIDIEPTDTIDRIKERVEEKEGIPPVQQRLIYAGKQLADDKTAKDY NIEGGSVLHLVLALRGG

SEQ ID NO: 10 (hsNEDD8; Homo sapiens NEDD8)

MLIKVKTLTGKEIEIDIEPTDKVERIKERVEEKEGIPPQQQRLIYSGKQMNDEKTAA DYKILGGSVLHLVLALRGG

SEQ ID NO: 1 1 (ssNEDPI ; Salmo salar NEDP1 ) MDPVVLSYQDSLLRRSDVALLEGPHWLNDQVIGFAFEYFAAELFKGLGEAAIFISP EVTQFIKCAACPEDLALFLEPLGLASRRWVFLAVNDNSIQTAGGSHWSLLLFLRD SGHFAHYDSQSGGNSLHARRIATKLEPFLGSGRKVPFVEEPCPLQQNSYDCGM YVICNAEALCERARVEGSPRLPVQTITPAYITQKRLEWCRLIQRLDRD

SEQ ID NO: 12 (bdNEDPI ; Brachypodium distachyon NEDP1 )

MDERVLSYGDVVLLRSDLAILRGPHFLNDRIIAFYLAHLSASFHGDGDLLLLPPSIP

YLLSNLPDPESVAEPLCLASRRLVLLPVNDNPDASVANGGSHWTLLVLDAATTDP

QAPRFVHHDSLRGSANAAAARRLARALTAGGAPLRFVEAPTPTQRNGHDCGVY

VLAVARAICGWWRSSRRRENQQGGGGDWFATMMEEVDAESVGAMRAELLQLI

HRLIQDKEQEEEKKSKAGVEDTCGQ

SEQ ID NO: 13 (hsNEDPI ; Homo sapiens NEDP1 )

MDPVVLSYMDSLLRQSDVSLLDPPSWLNDHIIGFAFEYFANSQFHDCSDHVSFIS PEVTQFIKCTSNPAEIAMFLEPLDLPNKRVVFLAINDNSNQAAGGTHWSLLVYLQ DKNSFFHYDSHSRSNSVHAKQVAEKLEAFLGRKGDKLAFVEEKAPAQQNSYDC GMYVICNTEALCQNFFRQQTESLLQLLTPAYITKKRGEWKDLITTLAKK

SEQ ID NO: 14 (scAtg8; Saccharomyces cerevisiae autophagy-related protein 8)

MKSTFKSEYPFEKRKAESERIADRFKNRIPVICEKAEKSDIPEIDKRKYLVPADLTV GQFVYVIRKRIMLPPEKAIFIFVNDTLPPTAALMSAIYQEHKDKDGFLYVTYSGENT FG

SEQ ID NO: 15 (scAtg4; Saccharomyces cerevisiae autophagy-related protein 4)

MQRWLQLWKMDLVQKVSHGVFEGSSEEPAALMNHDYIVLGEVYPERDEESGA

EQCEQDCRYRGEAVSDGFLSSLFGREISSYTKEFLLDVQSRVNFTYRTRFVPIAR

APDGPSPLSLNLLVRTNPISTIEDYIANPDCFNTDIGWGCMIRTGQSLLGNALQILH

LGRDFRVNGNESLERESKFVNWFNDTPEAPFSLHNFVSAGTELSDKRPGEWFG

PAATARSIQSLIYGFPECGIDDCIVSVSSGDIYENEVEKVFAENPNSRILFLLGVKL

GINAVNESYRESICGILSSTQSVGIAGGRPSSSLYFFGYQGNEFLHFDPHIPQPAV

EDSFVESCHTSKFGKLQLSEMDPSMLIGILIKGEKDWQQWKLEVAESAIINVLAKR

MDDFDVSCSMDDVESVSSNSMKKDASNNENLGVLEGDYVDIGAIFPHTTNTEDV

DEYDCFQDIHCKKQKIVVMGNTHTVNANLTDYEVEGVLVEKETVGIHSPIDEKC SEQ ID NO: 16 (P1 ^'-P2^' spacer sequence in TEV protease recognition site- containing substrates is underlined; cf. Fig. 4A)

ENLYFQGT

SEQ ID NO: 17 (Spacer sequence in TEV protease recognition site-containing substrates used for P1 ^'-sensitivity assays is underlined; cf. Fig. 12)

ENLYFQXGT, wherein X = none, Met, Tyr, Arg, Glu or Pro

SEQ ID NO: 18 (TEV protease; Tobacco etch virus Nla protease)

GESLFKGPRDYNPISSTICHLTNESDGHTTSLYGIGFGPFIITNKHLFRRNNGTLLV

QSLHGVFKVKNTTTLQQHLIDGRDMIIIRMPKDFPPFPQKLKFREPQREERICLVT

TNFQTKSMSSMVSDTSCTFPSSDGIFWKHWIQTKDGQCGSPLVSTRDGFIVGIH

SASNFTNTNNYFTSVPKNFMELLTNQEAQQWVSGWRLNADSVLWGGHKVFMS

KPEEPFQPVKEATQLMNELVYSQ

SEQ ID NO: 19 (TEV(SH)AC6)

ESLFKGPRDYNPISSSICHLTNESDGHTTSLYGIGFGPFIITNKHLFRRNNGTLLVQ

SLHGVFKVKDTTTLQQHLVDGRDMIIIRMPKDFPPFPQKLKFREPQREERICLVTT

NFQTKSMSSMVSDTSCTFPSSDGIFWKHWIQTKDGQCGSPLVSTRDGFIVGIHS

ASNFTNTNNYFTSVPKNFMELLTNQEAQQWVSGWRLNADSVLWGGHKVFMNK

PEEPFQPVKEATQLMN

SEQ ID NO: 20 (xlUb; Xenopus laevis ubiquitin)

MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQLEDGRTLS DYNIQKESTLHLVLRLRGG

SEQ ID NO: 21 (xlUsp2; Xenopus laevis ubiquitin-specific processing protease 2)

MRSHTLRIHGMGAGREHQIPGTVILSSIMDFILHRAKSSKHVQGLVGLRNLGNTC

FMNSILQCLSNTKDLRDYCQQNSYRRDLSSKKCNTAIMEEFARLLQAIWTSSANE

VVSPSEFKTQIQRYAPRFMGYNQQDAQEFLRFLLDGLHNEVNRVTVKPRPSSQD

LDHMPDSEKGKKMWKRYLEREDSRIVELFVGQLKSSLTCTDCGYCSTVFDPFW

DLSLPIAKKSASEVSLVDCMRLFTKEDVLDGDEKPTCCRCKARRRCTKKFTIQRF

PKILVLHLKRFSEGRIRSGKLSTFVNFPLKDLDLREFSSESNPHATYNLYAVSNHS

GTTMGGHYTAYCKNPSNGEWYTFNDSRVTAMSSSQVKSSDAYVLFYELSGPSS

RM SEQ ID NO: 22 (His₁₄-bdSUMO-MBP)

MSKHHHHSGHHHTGHHHHSGSHHHSGSAAGGEEDKKPAGGEGGGAHINLKVK

GQDGNEVFFRIKRSTQLKKLMNAYCDRQSVDMTAIAFLFDGRRLRAEQTPDELE

MEDGDEIDAMLHQTGGAGTKTEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKV

TVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLY

PFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSAL

MFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNK

HMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSK

PFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELA

KDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALKDAQT

NGTGC.

SEQ ID NO: 23 (His₁₀-ZZ-TEV-MBP)

MHHHHHHHHHHGSNKFNKEQQNAFYEILHLPNLNEEQRNAFIQSLKDDPSQSA

NLLAEAKKLNDAQAPKVAMNKFNKEQQNAFYEILHLPNLNEEQRNAFIQSLKDDP

SQSANLLAEAKKLNDAQAPKVAMSGENLYFQGTKTEEGKLVIWINGDKGYNGLA

EVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLL

AEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIP

ALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAG

AKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNY

GVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDK

PLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAAS

GRQTVDEALKDAQTNGTGC.

SEQ ID NO: 24 (His₁₄-bdNEDD8-MBP)

MSKHHHHSGHHHTGHHHHSGSHHHSGTMIKVKTLTGKEIEIDIEPTDTIDRIKERV

EEKEGIPPVQQRLIYAGKQLADDKTAKDYNIEGGSVLHLVLALRGGAGTKTEEGK

LVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIF

WAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIY

NKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYE

NGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTIN

GPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLE

NYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQM

SAFWYAVRTAVINAASGRQTVDEALKDAQTNGTGC. SEQ ID NO: 25 (His₁₄-ssNEDD8-MBP)

MSKHHHHSGHHHTGHHHHSGSHHHSGMLIKVKTLTGKEIEIDIEPTDKVERIKER

VEEKEGIPPQQQRLIYSGKQMNDEKTAADYKIQGGSVLHLVLALRGGAGTKTEE

GKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDII

FWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLI

YNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKY

ENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTI

NGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFL

ENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQ

MSAFWYAVRTAVINAASGRQTVDEALKDAQTNGTGC.

SEQ ID NO: 26 (His₁₄-scAtg8-MBP)

MSKHHHHSGHHHTGHHHHSGSHHHTGGSSGSESSKSTFKSEYPFEKRKAESE

RIADRFKNRIPVICEKAEKSDIPEIDKRKYLVPADLTVGQFVYVIRKRIMLPPEKAIFI

FVNDTLPPTAALMSAIYQEHKDKDGFLYVTYSGENTFGAGTKTEEGKLVIWINGD

KGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFG

GYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPN

PPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIK

DVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWS

NIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEG

LEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVR

TAVINAASGRQTVDEALKDAQTNGTGC.

SEQ ID NO: 27 (His₁₄-xlUb-MBP)

MSKHHHHSGHHHTGHHHHSGSHHHTGGSSGSESSMQIFVKTLTGKTITLEVEP

SDTIENVKAKIQDKEGIPPDQQRLIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRG

GAGTKTEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVA

ATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAY

PIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAA

DGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFN

KGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPN

KELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKG

E I M PN I PQMSAFWYAVRTAVI N AASG RQTVDEALKDAQTNGTG C, SEQ ID NO: 28 (Hisi₄-scSUMO-MBP)

MSKHHHHSGHHHTGHHHHSGSHHHTGSDSEVNQEAKPEVKPEVKPETHINLKV

SDGSSEIFFKIKKTTPLRRLMEAFAKRQGKEMDSLRFLYDGIRIQADQTPEDLDM

EDNDIIEAHREQIGGAGTKTEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVE HPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFT WDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMF NLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHM NADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPF VGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAK DPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALKDAQTN GTGC.

Percent identity tables

The present invention is further described by the following embodiments. A method for purifying a stoichiometric protein complex composed of at least two subunits from a mixture,

said mixture comprising said protein complex and monomers of said at least two subunits,

wherein said at least two subunits comprised in said mixture each comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS),

wherein the PRS of each of said at least two subunits is cleavable by a protease, which protease is orthogonal to the PRS of the other subunit(s), wherein the method comprises the steps of

a) subjecting the mixture to a first affinity chromatography selective for the AT of the first of said at least two subunits, whereby

(ii) impurities are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the second subunit is cleaved off, or the protein complex is eluted by on- column cleavage, using said orthogonal protease which is specific for the PRS of said second subunit, and

(iv) optionally removing the cleaved off AT of the second subunit.

The method of embodiment 1 , wherein in step a) (iii) and/or step b) (iii) the protein complex is eluted by on-column cleavage.

The method of embodiment 1 or 2, wherein the method further comprises the step of c) removing the protease from the eluate originating from the last affinity chromatography.

The method of any one of embodiments 1 -3, wherein the protein complex comprises a third subunit,

wherein said third subunit comprised in said mixture comprises an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS),

wherein the AT of said third subunit differs from the AT of the other subunits and allows affinity chromatography being selective for the AT of said third subunit, and

wherein the PRS of said third subunit is cleavable by a protease, which protease is orthogonal to the PRS of the other subunits,

further comprising after step b) and prior to optional step c) an additional step b') subjecting the eluate from step b) to an affinity chromatography selective for the AT of the third subunit, whereby

(i) the protein complex binds to the affinity resin via the AT of the third subunit, and

(ii) impurities are washed off the column, and

(iv) optionally removing the cleaved off AT of the third subunit.

The method of embodiment 4, wherein the protein complex comprises a fourth subunit,

wherein said fourth subunit comprised in said mixture comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS),

wherein the AT of said fourth subunit differs from the AT of the other subunits and allows affinity chromatography being selective for the AT of said fourth subunit, and

wherein the PRS of said fourth subunit is cleavable by a protease which is orthogonal to the PRS of the other subunits, further comprising after step b') and prior to optional step c) an additional step

b") subjecting the eluate from step b') to an affinity chromatography selective for the AT of the fourth subunit, whereby

(ii) impurities are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the fourth subunit is cleaved off, or the protein complex is eluted by on- column cleavage, using said orthogonal protease which is specific for the PRS of said fourth subunit, preferably wherein the protein complex is eluted by on-column cleavage, and

(iv) optionally removing the cleaved off AT of the fourth subunit.

The method of embodiment 5, wherein the protein complex comprises a fifth subunit,

wherein said fifth subunit comprised in said mixture comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS),

wherein the AT of said fifth subunit differs from the AT of the other subunits and allows affinity chromatography being selective for the AT of said fifth subunit, and

wherein the PRS of said fifth subunit is cleavable by a protease which is orthogonal to the PRS of the other subunits,

further comprising after step b") and prior to optional step c) an additional step

b'") subjecting the eluate from step b") to an affinity chromatography selective for the AT of the fifth subunit, whereby

(ii) impurities are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the fifth subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PRS of said fifth subunit, preferably wherein the protein complex is eluted by on-column cleavage, and optionally removing the cleaved off AT of the fifth subunit.

The method of any one of embodiments 1 to 6, wherein one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 1 (bdSUMO); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 60% identity over the full length of SEQ ID NO: 1 (bdSUMO),

wherein the protease shown in SEQ ID NO: 2 (bdSENPI ) is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 1 , under identical conditions;

preferably wherein the PRS as defined in (i) or (ii) is comprised in the second subunit;

more preferably, wherein the second subunit comprises a PRS comprising SEQ ID NO: 1 (bdSUMO);

most preferably wherein the second subunit comprises a PRS consisting of SEQ ID NO: 1 (bdSUMO).

(i) an amino acid sequence as shown in SEQ ID NO: 3 (scSUMO); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 55% identity over the full length of SEQ ID NO: 3 (scSUMO),

wherein the protease shown in SEQ ID NO: 4 (scUlpl ) is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 3, under identical conditions.

(i) an amino acid sequence as shown in SEQ ID NO: 5 (hsSUMOIa) or SEQ ID NO: 6 (hsSUMO2); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 60% identity over the full length of SEQ ID NO: 5 or 6,

wherein the protease shown in SEQ ID NO: 7 (hsSENPI ) is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 5 or 6, respectively, under identical conditions. The method of any one of embodiments 1 to 9, wherein one PRS comprises, preferably consists of

(ii) a PRS derivative of (i) with an amino acid sequence having at least 85% identity over the full length of SEQ ID NO: 9 or at least 99% over the full length of identity over the full length of SEQ ID NO: 8 or 10, wherein the protease shown in SEQ ID NO: 11 (ssNEDPI ), SEQ ID NO: 12 (bdNEDPI ), or SEQ ID NO: 13 (hsNEDPI ) is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 8, 9 or 10, respectively, under identical conditions;

preferably wherein the PRS comprising an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8), SEQ ID NO: 9 (bdNEDD8) or a PRS derivative thereof as defined in (ii) is comprised in the first subunit;

more preferably wherein the first subunit comprises a PRS comprising an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8) or SEQ ID NO: 9 (bdNEDD8), in particular wherein the first subunit comprises a PRS comprising an amino acid sequence as shown in SEQ ID NO: 9 (bdNEDD8); most preferably wherein the first subunit comprises a PRS consisting of an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8) or SEQ ID NO: 9 (bdNEDD8), in particular wherein the first subunit comprises a PRS consisting of an amino acid sequence as shown in SEQ ID NO: 9 (bdNEDD8).

The method of any one of embodiments 1 to 10, wherein at least one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 14 (scAtg8); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 80% identity over the full length of SEQ ID NO: 14,

wherein the protease shown in SEQ ID NO: 15 (scAtg4), is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 14 under identical conditions.

The method of any one of embodiments 1 to 11 , wherein at least one PRS comprises, preferably consists of the TEV protease recognition site shown in SEQ ID NO: 16. 13. The method of any one of embodiments 1 to 12, wherein at least one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 20 (xlUb); or

(ii) a PRS derivative of (i) with an amino acid sequence having at least 80% identity over the full length of SEQ ID NO: 20,

wherein the protease shown in SEQ ID NO: 21 (xlUsp2), is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 20 under identical conditions.

14. The method of embodiment 7, wherein the subunit is eluted from the column using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in amino acids 1 -224 of SEQ ID NO: 2 (bdSENP1_{2 8-} 48i ), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 45% identity over the full length of SEQ ID NO: 2,

wherein said protease derivative is capable of cleaving the PRS according to ID NO: 1 (bdSUMO) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions;

preferably wherein the subunit is eluted from the column using (i) the protease shown in in amino acids 1 -224 of SEQ ID NO: 2 (bdSENP1_248-48i )-

15. The method of embodiment 8, wherein the subunit is eluted from the column using

(ii) a protease derivative of (i) having an amino acid sequence with at least 35% identity over the full length of SEQ ID NO: 4,

wherein said protease derivative is capable of cleaving the PRS according to ID NO: 3 (scSUMO) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions.

16. The method of embodiment 9, wherein the subunit is eluted from the column using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 7 (hsSENPI ), or (ii) a protease derivative of (i) having an amino acid sequence with at least 45% identity over the full length of SEQ ID NO: 7,

wherein said protease derivative is capable of cleaving the PRS according to ID NO: 5 (hsSUMOI a) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions.

The method of embodiment 10, wherein the subunit is eluted from the column using

(ii) a protease derivative of (i) having an amino acid sequence with at least 70% identity over the full length of SEQ ID NO: 11 (ssNEDPI ), or SEQ ID NO: 13 (hsNEDPI ), or with at least 35% identity over the full length of SEQ ID NO: 12 (bdNEDPI ),

wherein said protease derivative, cleaves the PRS according to SEQ ID NO: 9 (bdNEDD8) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions;

preferably wherein the subunit is eluted from the column using

(ii) a protease derivative having an amino acid sequence with at least 70% identity over the full length of SEQ ID NO: 11 (ssNEDPI ) or with at least 35% identity over the full length of SEQ ID NO: 12 (bdNEDPI ), wherein said protease derivative, cleaves the PRS according to ID NO: 9 (bdNEDD8) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions;

more preferably wherein the subunit is eluted from the column using a protease comprising, preferably consisting of the amino acid sequence selected from the group consisting of amino acid sequences shown in SEQ ID NO: 11 (ssNEDPI ), and SEQ ID NO: 12 (bdNEDPI ); and most preferably wherein the subunit is eluted from the column using the protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 12 (bdNEDPI ).

The method of embodiment 11 , wherein the subunit is eluted from the column using

(ii) a protease derivative of (i) having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 15, wherein said protease derivative is capable of cleaving the PRS according to ID

NO: 14 (scAtg8) with at least 20% activity as compared to the parent protease as defined in (i).

The method of embodiment 12, wherein the subunit is eluted from the column using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 18 or 19 (TEV protease), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 18 or 19, wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 16 (TEV) with at least 20% activity as compared to the parent protease as defined in (i).

The method of embodiment 13, wherein the subunit is eluted from the column using

(ii) a protease derivative of (i) having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 21 , wherein said protease derivative is capable of cleaving the PRS according to ID NO: 20 (xlUb) with at least 20% activity as compared to the parent protease as defined in (i).

The method of any one of embodiments 1-3 and 7-20, wherein the protein complex is composed of 2 different subunits, preferably with a stoichiometry of 1 :1.

The method of any one of embodiments 1-4 and 7-20, wherein the protein complex is composed of 3 different subunits, preferably with a stoichiometry of 1 :1 :1.

The method of any one of embodiments 1-20, wherein the protein complex is composed of 4 different subunits, preferably with a stoichiometry of 1 :1 :1 :1. 24. The method of any one of embodiments 1 -23, wherein the mixture originates from a mixture of lysates and/or supernatants and/or a pre-purified solution, each comprising at least one of the subunits.

25. The method of any one of embodiments 1 -23, wherein the mixture originates from a single lysate or supernatant or a pre-purified solution comprising all subunits of the protein complex.

26. The method of any one of embodiments 1 -25, wherein the first subunit comprises a poly-His tag, and preferably wherein the second subunit comprises a ZZ tag.

27. The method of any one of embodiments 1 -26, wherein the first subunit comprises the PRS as defined in embodiment 10, and wherein the second subunit comprises the PRS as defined in embodiment 7.

28. The method of embodiment 27, wherein the elution in step a) is carried out using the protease as defined in embodiment 17, and wherein the elution in step b) is carried out using the protease as defined in embodiment 14.

29. The method of embodiment 27 or 28, wherein the first subunit comprises a poly-His tag, and wherein the second subunit comprises a ZZ tag.

30. The method of any one of embodiments 1 -29, wherein step c) is an affinity chromatography, a size exclusion chromatography, or a precipitation step. 31. The method of any one of embodiments 1 -30, wherein the protease used for cleaving an affinity tag from a subunit comprises an affinity tag which is the same than one of the affinity tags used in the affinity chromatography steps a), b), b'), b") or b'"), with the provisio that the affinity tag differs from the affinity tag used in the directly preceding affinity chromatography step.

32. The method of any one of embodiments 1 -31 , wherein the protease from the eluate originating from the last affinity chromatography prior to step c) comprises an affinity tag, preferably as defined in embodiment 26, and wherein step c) is an affinity chromatography step, whereby the protease binds to the affinity resin, and the protein complex is collected in the flow- through.

33. The method of any one of embodiments 1 -32, wherein the subunit(s) further comprise a spacer between the AT and the PRS, and/or between the PRS and the subunit; preferably wherein the subunit(s) further comprise a spacer between the AT and the PRS.

34. A protease having an amino acid sequence with at least 45% identity over the full length of SEQ ID NO: 2 (bdSENPI ), wherein said protease is capable of cleaving the PRS according to ID NO: 1 (bdSUMO) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 2;

preferably wherein the protease comprises the amino acid sequence shown as amino acids 1 -224 in SEQ ID NO: 2 (bdSENP1₂₄8-48i );

more preferably wherein the protease consists of the amino acid sequence shown as amino acids 1 -224 in SEQ ID NO: 2 (bdSENP1_248-48i ).

A protease having an amino acid sequence with at least 70% identity over the full length of SEQ ID NO: 11 (ssNEDPI ), wherein said protease is capable of cleaving the PRS according to ID NO: 8 (ssNEDD8) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 1 ;

preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 11 (ssNEDPI );

more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 11 (ssNEDPI ).

A protease having an amino acid sequence with at least 35% identity over the full length of SEQ ID NO: 12 (bdNEDPI ),

wherein said protease is capable of cleaving the PRS according to ID NO: 9 (bdNEDD8) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 12;

preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 12 (bdNEDPI );

more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 12 (bdNEDPI ).

A protease having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 21 (xlUsp2),

wherein said protease is capable of cleaving the PRS according to SEQ ID NO: 20 (xlUb) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 21 ;

preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 21 (xlUsp2);

more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 21 (xlUsp2). The protease of any one of embodiments 33 to 37, wherein the protease further comprises an affinity tag, preferably an affinity tag as defined in embodiment 26.

A kit of parts, comprising at least two proteases selected from

(i) the protease according to embodiment 34,

(ii) the protease according to embodiment 35 or 36,

(iii) the protease as defined in embodiment 19,

(iv) the protease as defined in embodiment 18, and

(v) the protease according to embodiment 37,

preferably comprising at least two proteases selected from (i)-(iii),

more preferably comprising two proteases selected from (i) and (ii), and most preferably comprising the protease according to embodiment 34 and the protease according to embodiment 36.

The kit of parts of embodiment 39, wherein at least one of the proteases further comprises an affinity tag, preferably an affinity tag as defined in embodiment 26.

Use of a protease as defined in any one of embodiments 34-38, or the kit of parts as defined in embodiment 39 or 40 in a method of purifying stoichiometric protein complexes comprising at least two subunits,

wherein said at least two subunits comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS), and optionally a spacer between the AT and the PRS, and

wherein the AT of each of said at least two subunits differs from each other so to allow specific affinity chromatography, and

wherein the PRS of each of said at least two subunits is cleavable by a protease which is orthogonal to the PRS of the other subunit(s);

preferably wherein the method is further defined as in any one of embodiments 1 -30.

Use of a protease as defined in any one of embodiments 34-38, or the kit of parts as defined in embodiment 39 or 40 for on-column cleavage in an affinity chromatography.

A nucleic acid molecule, encoding the protease according to any one of embodiments 34-38. EXAMPLES

Example 1

Identification of SUMO/SENP1 and NEDD8/NEDP1 substrate/protease pairs from Brachypodium distachyon and Salmo salar.

In order to develop a convenient purification procedure for binary and higher order protein complexes, the present inventors searched for highly specific and efficient proteases with orthogonal specificity to the Saccharomyces cerevisiae (sc) scSUMO/scUlpl system. To this end, they followed two alternative approaches: First, assuming that a large evolutionary distance would be sufficient to generate an orthogonal system, the present inventors looked for clearly identifiable SUMO and scUlpl orthologues in organisms that diverged from S. cerevisiae early in evolution. In an alternative approach, the present inventors looked for paralogous substrate/protease pairs participating in parallel protein modification pathways within the same cell, which might thus be optimized by natural evolution for mutually exclusive specificities. Nevertheless, given the possibility that such parallel protease pathways might still have overlapping specificities, it could not a priori be assumed that such paralogous substrate/protease pairs would be orthogonal.

Iterative BLAST searches were performed on existing EST and genomic databases using the known human and yeast SUMO and SENP1 orthologues Smt3p and Ulpl p as a template. Full-length sequences of the primary hits were assembled from multiple overlapping clones and used as input sequences for further BLAST searches. Similarly, NEDD8- and NEDP1 orthologues were identified.

Using iterative homology searches and by assembling multiple cDNA as well as genomic sequences, several so-far non-annotated putative orthologues of SENP1- like and NEDP1 -like proteases along with their probable substrates were identified (not shown). As the inventors aimed at proteases efficiently cleaving at 0°C, the search was further focused on organisms tolerating a low temperature. From the remaining hits, the putative SENP1 and NEDP1 proteases from the grass Brachypodium distachyon (purple false brome) and the putative NEDP1 orthologue from Salmo salar (salmon) along with their putative substrates were cloned. Figure 3 shows a structure-based sequence alignment of the respective protein sequence in comparison to their putative human orthologues. Folded protein domains were subjected to structure-based sequence alignments using Expresso (Armougom et al. (2006), supra) and the following templates: hsSUMOI : 2G4D_B (Xu et al. Biochem J. 398: 345-352); hsSUM02/bdSUMO: 2CKH_B (Shen et al. (2006) Biochem J. 397: 297-288); scSUMO: 1 EUV_B (Mossessova et al. (2000), supra); hsSENP1/bdSENP1 : 2CKH_A (Shen et al. (2006), supra); scUlpl : 1 EUV_A (Mossessova et al. (2000), supra); all NEDD8 orthologues: 2BKR_B (Shen et al. (2005), supra); all NEDP1 orthologues: 2BKR A (Shen et al. (2005), supra). Interacting residues (< 4.6A center-center distance) were computed using MacPyMol (DeLano Scientific). To annotate SUMO-interacting surface of hsSENPI , the structure of its complex with hsSUM02 (accession number 2CKH; Shen et al. (2006), supra) was analyzed.

The catalytic domains of the Brachypodium and Salmon proteases were over- expressed in soluble form in E. coli and purified on a Ni²⁺ chelate resin using an engineered polyHis-tag. The catalytic domain of the reference protease scUlpl (Malakhov et al. (2004), supra) and a stabilized version of TEV protease (TEV(SH); van den Berg et al. (2006) J Biotechnol. 121 : 291 -298) lacking the C-terminal autoinhibitory peptide (Nunn and Djordijevic, 2005, supra) were purified accordingly. All proteases could be purified in large amounts and were highly active. The activity of the C-terminally truncated TEV(SH) variant used in all further experiments was indistinguishable from the full-length parent TEV protease (Fig. 17).

As further examples of proteases involved in processing of SUMO-like modifiers, the inventors expressed and purified the S. cerevisiae Atg8-specific protease Atg4p (scAtg4; Kirisako et al. (2000) J Cell Biol. 151 (2): 263-276) and the catalytic domain of the Xenopus laevis (xl) ubiquitin-specific protease Usp2 (xlUsp2; cf. SEQ ID NO: 21 ). A preferred truncated fragment of xlUsp2 is xlUsp2_43-3₈3, which was used in the examples herein. If required, untagged proteases were obtained by proteolytically removing the polyHis-tag. Design and expression of protease substrates for cleavage assays in solution

It was chosen to directly compare the efficiency of substrate processing by the identified proteases with cleavage by the established alternatives scUlpl and TEV protease. Therefore, the design of SUMO- and NEDD8-containing substrates followed a common scheme that ensured an identical sequence context (and therefore a similar accessibility) of the scissile bond (Fig. 4A): The common Gly- Gly-motif preceding the scissile bond was followed by the tri-peptide Ala-Gly-Thr (AGT) and the target protein (maltose binding protein, MBP). Within the TEV protease substrate, the sequence after the scissile bond was Gly-Thr (GT), in agreement with the natural and preferred TEV recognition sequence (Kapust et al. (2002), supra; Kostallas et al. (201 1 ) PLoS One 6: e16136). Here, a ZZ-tag of 14kDa (Nilsson et al. (1987) Protein Eng. 1 : 107-1 13) was fused to the N-terminus of the protease recognition site in order to allow for an easy electrophoretic discrimination between full length and cleaved substrate. All substrate proteins were expressed in E. coli from appropriate expression vectors and purified via an engineered N-terminal polyHis-tag using Ni²⁺ chelate chromatography with imidazole elution. All proteins were quantified via their absorption at 280nm and computed extinction coefficients. Accuracy of quantification and purity of the proteins were analyzed by SDS-PAGE.

All new proteases are highly efficient tools for tag removal

The inventors chose to directly compare the properties of all six proteases recognizing substrates with ubiquitin-like fold (scUlpl , bdSENPI , bdNEDPI , ssNEDPI , scAtg4, xlUsp2) to the established TEV protease in a defined in-vitro system. Throughout this study, the inventors used a variant of TEV protease based on the solubility-enhanced and autocleavage-resistant TEV(SH) variant (Berg and Berglund, 2006) lacking the C-terminal autoinhibitory peptide (Nunn and Djordijevic, 2005), that displayed a catalytic activity indistinguishable from the parent enzyme (Fig. 17). The amino acid sequence of this protease is detailed in SEQ ID NO: 19.

If not stated otherwise, all in-vitro cleavage reactions were performed in LS-buffer (250mM NaCI, 40mM Tris/HCI pH7.5, 2mM MgCI₂, 250mM sucrose, 2mM DTT, 2pg/ml BSA). Generally, substrates and proteases were pre-diluted in LS-buffer to twice the aspired end-concentration. Cleavage was initiated by mixing identical volumes of substrate and protease pre-dilutions. After incubation, the reactions were stopped by dilution in hot SDS sample buffer. A defined fraction (generally corresponding to 2.5pg of substrate) was separated by SDS-PAGE on 7-15% gradient gels. Gels were stained with Coomassie G250 and scanned.

To initially compare the activity of the proteases, a fixed concentration (100μΜ) of each substrate protein was incubated with a wide concentration range of the corresponding protease for one hour at 4°C or 25°C (Fig. 4B). These initial tests allowed to roughly estimate that already «15-50nM of scUlpl or bdSENPI efficiently (>95%) cleave the corresponding substrate proteins within one hour at 0°C. At 25°C, no more than 10nM of either protease was required. For a similarly efficient cleavage of their substrates, the two NEDP1 enzymes and scAtg4 required ^¾10- to 15-fold and xlUsp2 ~30- to 50-fold higher concentrations at both temperatures. Strikingly - but consistent with the low specific activities of commercial "TEV protease" (Sigma-Aldrich #T4455) and "TurboTEV" (Nacalai USA #NU0102) preparations - even at 10μΜ concentration, TEV protease was unable to cleave more than =40% of its substrate protein within one hour at 0°C. At 25°C, =2-3μΜ of TEV protease was required for efficient cleavage. Example 2

SUMO-specific proteases, bdNEDPI, scAtg4, xlUsp2 and TEV protease represent five orthogonal groups of proteases.

As the present inventors aimed at finding orthogonal substrate/protease systems, it was crucial to test for cross-reactivity between the proteases. This parameter was analyzed by incubating each substrate for three hours with a high concentration (10μΜ) of each protease at 25°C (Fig. 5A). A truncated version of bdSUMO lacking the first 20 amino acids (bdSUMO(21-97)) and a further N-terminally shortened version of bdSENPI (bdSENPI (248-481 )) were also included in the analysis (see Example 5 below). Strikingly, even under these drastic conditions (representing an up to 10.000-fold over-digestion), the SUMO proteases (i), the bdNEDPI enzyme (ii), the scAtg4 protease (iii), xlUsp2 (iv), and TEV protease (v) separated into five groups of proteases (i-v) that did not cleave substrates of the respective other groups. These five classes can therefore indeed be regarded as truly orthogonal. An interesting case is ssNEDPL This protease is truly orthogonal to scUlpl , bdSENPI , scAtg4 and TEV protease. When used at high concentrations it can, however, cleave a substrate containing xlUb, although with low efficiency. The Brachypodium orthologue bdNEDPI , in contrast, does not show this cross- reactivity. This example illustrates that a priori a prediction of orthogonality based on sequence distance or evolutionary distance is virtually impossible. In addition, the differences in cross reactivity on the xlUb-containing substrate of bdNEDPI and ssNEDPI shows impressively that it is not possible to directly extrapolate from the specificity found for a protease from a given species to the specificity of a corresponding orthologue of another species.

The yeast and Brachypodium SUMO proteases prefer their natural SUMO substrates, but are not fully orthogonal.

Under drastic cleavage conditions, none of the SUMO- or NEDD8-specific proteases could discriminate between its natural and the corresponding orthologous substrate (Fig. 5A). Even under less drastic conditions (incubation with 1 μΜ protease for 1 hour at 0°C) the NEDP1 enzymes cleaved the two orthologous NEDD8 substrates equally well (Fig. 5B). In this assay, however, the two SUMO- specific proteases visibly preferred the substrate containing the matching SUMO variant. More specifically, when compared to cleavage of their natural substrates, >5-times (scUlpl ) or >50-times (bdSENPI ) more protease was required to efficiently cleave the respective orthologous SUMO substrate (Fig. 6A). Accordingly, the time needed for efficient cleavage of the orthologous SUMO substrates was «15- and even 150-times longer for scUlpl and bdSENPI , respectively (Fig. 6B). These results show that both SUMO proteases clearly prefer their natural substrates. Compared to scUlpl , the species specificity of bdSENPI is «10-times more pronounced.

Example 3

On-column cleavage of immobilized substrates

To address the applicability of the characterized proteases for on-column cleavage, it was tested if bdSENPI and bdNEDPI can specifically cleave off the fluorescent target proteins outlined in Figure 7A from a Ni²⁺ chelate resin loaded with His-|₄- bdSUMO-GFP and Hisi₄-bdNEDD8-mCherry. Briefly, 50μΙ of substrate-loaded affinity resin was extensively washed with LS- buffer in mini-spin columns (MoBiTec). For elution, the buffer was removed by mild centrifugation and the resins were mixed with 10ΟμΙ LS-buffer containing the indicated protease or appropriate buffer controls. After elution, material released from the resin was collected by centrifugation. The resin was washed with another 100μΙ of LS-buffer. The wash fraction was combined with the elution fraction. Fluorescent proteins were quantified via their absorption at 488 nm and 585 nm, respectively.

These experiments confirm the orthogonal specificities of these two proteases: While 20-30nM bdSENPI efficiently released its specific target (GFP) from the resin (Fig. 7B), even at much higher protease concentrations NEDD8-tagged mCherry remained firmly bound to the resin. Conversely, bdNEDPI did not recognize the bdSUMO-GFP substrate but specifically and efficiently released mCherry from the resin (Fig. 7C). Strikingly, in both cases the protease concentrations needed for efficient release of the specific substrates were similar to the concentrations needed for substrate cleavage in solution (compare Figures 7B and C to Fig. 4), although substrate immobilization is expected to limit the diffusion and accessibility of the substrate.

On-column cleavage reactions might be influenced by interactions between the protease and the affinity resin, as an immobilized protease can only access substrates in its immediate vicinity. To address this issue, the efficiency of substrate release from a Ni²⁺ chelate resin by non-tagged and polyHis-tagged bdNEDPI was compared (Fig. 7D). Strikingly, while the non-tagged protease was very effective, the polyHis-tagged version interacting with the Ni²⁺ chelate resin failed to release its specific target protein even at high enzyme concentration (1 μΜ) and at elevated temperature (Fig. 7D, upper panel). This effect was not caused by a general interference of the polyHis-tag with the protease activity, as both the tagged and the non-tagged enzymes efficiently cleaved a soluble substrate (Fig. 7D, lower panel).

Similar results were also obtained with TEV protease (Fig. 8). Due to the lower activity of TEV protease, however, efficient on-column cleavage within one hour required much higher enzyme concentrations and incubation at 25°C. Similar to bdSENPI and bdNEDPI , also scAtg4 and xlUsp2 were tested for their applicability in on-column cleavage (Fig. 7E-G). Both proteases could specifically release their substrate proteins from the resin, efficient substrate cleavage, however, required higher protease concentrations as compared to cleavage in solution (compare to Fig. 4).

Example 4

Purification of a stoichiometric protein complex using an orthogonal protease pair

The orthogonal specificity of bdSENPI and bdNEDPI and their applicability for on- column cleavage was exploited for purification of a stoichiometric protein complex. As a proof of principle, the stable and well-characterized complex between the E. coli maltose binding protein (MBP) and the MBP-specific designed ankyrin-repeat protein "off7" (anti-MBP DARPin, here called Darpin) was chosen (Binz et al. (2004) Nat Biotechnol 22: 575-582).

The two binding partners were co-expressed in E coli as His-i₄-bdNEDD8-MBP and ZZ-bdSUMO-Darpin and purified by two consecutive capture-release steps (Fig. 9, see also Fig. 1 B and Fig. 2). More specifically, E. coli strain NEB Express (New England Biolabs) harboring expression plasmids for both proteins was grown at 25°C in 200ml TB medium with appropriate antibiotics to an OD₆oo of 6. The culture was diluted in 600ml fresh medium containing antibiotics and 0.1 mM IPTG and further shaken at 18°C over night. After adding EDTA (5mM) and PMSF (1 mM) directly to the culture, cells were harvested by centrifugation, resuspended in lysis buffer (290mM NaCI, 45mM Tris/HCI pH 7.5, 4.5mM MgCI₂, 10mM DTT) supplemented with 25mM imidazole, and lysed by sonication. The cleared lysate obtained by ultracentrifugation (1 h, 200.000xg) was incubated with 10ml Ni²⁺ chelate resin for one hour at 4°C to capture Hisi₄-bdNEDD8-MBP and its binary complex with ZZ-bdSUMO-Darpin via their Hisi₄-tags. The resin was extensively washed with lysis buffer containing 15mM imidazole followed by lysis buffer containing 250mM sucrose. Bound protein complexes were specifically released by incubation with lysis buffer containing 250mM sucrose and 500nM bdNEDPI for 1 h at 4°C. 1 ml of the eluate fraction was incubated with 1 ml anti-ZZ-resin for 1 h at 4°C with and extensively washed with lysis buffer followed by buffer WB2 (100mM NaCI, 10mM Tris/HCI pH 7.5, 5mM DTT). A highly pure and stoichiometric binary complex could be eluted by incubation with 30nM bdSENPI for 1 h at 4°C in buffer WB2. Most importantly, the two target proteins were cleaved off from their tags during the purification procedure, thereby yielding a non-tagged complex. The stoichiometric nature of the complex could be verified by gel filtration.

Example 5

Detailed characterization of tag-cleaving proteases

So far, the characterization of the analyzed protease systems was mainly restricted to the essential properties needed for the purification of recombinant proteins and protein complexes. In the following paragraphs, the proteolytic properties of the enzymes will be compared in greater detail.

The Brachypodium SENP1 enzyme is the most efficient tag-cleaving protease tested.

A first rough titration (see Fig. 4) already suggested that, at 0°C, 15-50nM of the two SUMO-specific proteases is sufficient for efficient substrate cleavage, while both NEDP1 and the scAtg4 proteases need roughly 10-fold higher and xlUb even up to 50-fold higher protease concentrations. At the same temperature, efficient substrate cleavage by TEV protease required protease concentrations significantly exceeding 10μΜ. For a more detailed comparison of cleavage efficiencies, the cleavage kinetics at 0°C were analyzed for each substrate/protease pair (Fig. 10A). To facilitate a qualitative comparison, orthologous protease pairs were used at identical concentrations (30nM for SUMO-specific proteases, 300nM for NEDP1 enzymes). This setup revealed even subtle differences between similar proteases: While digestion was =90% complete after 30 min with 30nM of bdSENPI , a comparable digestion efficiency with scUlpl required more than twice as long. Similarly, at 300nM concentration, bdNEDPI cleaved its substrate «2-times faster than the corresponding salmon enzyme. In this assay, the activity of scAtg4 was comparable to bdNEDPI . The xlUsp2 enzyme required about 3-fold higher concentration to process its substrate at the same rate as ssNEDPI . At 0°C, even 20μΜ TEV protease was insufficient to process 100μΜ of substrate (i.e. a 5-fold molar excess) within 4 hours. These results classified bdSENPI as the most efficient tag-cleaving protease tested. Taking into account the different protease concentrations and cleavage kinetics, this direct comparison corroborated that, at 0°C, the two NEDP1 proteases and scAtg4 are roughly 10-15 times less active than the SUMO-specific enzymes but still outperform TEV protease by a factor of «150-300. The xlUsp2 protease as the least active of the newly characterized proteases is still >50-fold more active than TEV protease.

The SUMO-, NEDD8- and Atg8-specific proteases are highly active between 0°C and 37°C.

The temperature dependence of protease activity was analyzed by incubating a fixed concentration of substrate proteins at various temperatures with a limiting amount of the respective proteases (Fig. 10B). As expected, the cleavage efficiency increased from 0°C to 37°C for all substrate/protease pairs. Also in this assay, bdSENPI performed better than its yeast orthologue and consistently showed a more efficient cleavage of its substrate at all temperatures. In a direct comparison of the two NEDP1 enzymes, the Brachypodium enzyme was more active than its salmon counterpart - at least between 0°C and 25°C and showed a similar temperature dependence as scAtg4. The activity of Usp2 greatly improved at 37°C. Although also TEV protease was more active at higher temperatures, it was in this assay at all temperatures at least 10-fold less efficient than any of the other proteases tested. Thus, while all SUMO- and NEDD8-specific proteases tested can be used for efficient tag removal at 0°C, TEV protease needs higher temperatures, more enzyme and/or significantly longer incubation times for similar results.

Long-term stability of tag-cleaving proteases

If desired, increasing the reaction temperature or the incubation times will generally reduce the amount of protease necessary for tag removal. This, however, requires the protease to be stable under these conditions. To exemplarily test for long-term stability, the activities of all five proteases were assayed after over-night incubation at 0, 20, 37 or 50°C in the absence of oxygen (Fig. 10C). While the two SUMO proteases and xlUsp2 sustained temperatures up to 37°C for 16h without visibly loosing activity, the enzymes were completely inactive after over-night incubation at 50°C. The NEDP1 enzymes and scAtg4 remained fully active after over-night incubation at 20°C. After over-night incubation at 37°C, scAtg4 retained «50% of its activity while both NEDP1 enzymes lost most of their activity. TEV protease was stable after over-night incubation at 37°C but significantly lost activity at 50°C. These results indicate that all tested enzymes are fully compatible with over-night incubations at room temperature.

The yeast scUlpl and scAtg4 enzymes are sensitive to high-salt conditions. In order to test for salt tolerance, protease activity was assayed in the presence of various NaCI concentrations ranging from 200mM to 1 M (Fig. 1 1 A). In this assay, TEV protease and ssNEDPI showed the highest salt tolerance and were largely insensitive to salt concentrations up to 1 M. A moderate activity loss of «30% and ^«50% with increasing salt was observed for bdNEDPI and xlUsp2, respectively. The bdSENPI enzyme efficiently cleaved its substrate in a wide salt range between 200 and 750mM NaCI, while a slightly decreased activity was noticeable at 1 M NaCI. In contrast, the yeast enzymes scUlpl and scAtg4 showed a striking salt sensitivity and significantly lost activity between 200mM and 1 M NaCI. A detailed analysis of the cleavage kinetics of scUlpl and bdSENPI indicated that scUlpl cleaves its substrate «50 times slower in the presence of 1 M NaCI as compared to 250mM NaCI (Fig. 1 1 B). In comparison, the salt-induced kinetic inhibition was only »3-fold for bdSENPI (Fig. 1 1 C).

Overall, at all salt concentrations, bdSENPI showed the highest cleavage efficiency of all proteases tested. Strikingly, the advantage of bdSENPI over its yeast orthologue, which can already be seen at moderate ionic strength, is even further increased at higher salt.

P1 ^'-sensitivity

A number of proteases show sensitivity towards the residue following the scissile bond within the substrate (ΡΓ position; see e.g. Arnau et al. (2006) and Kapust et al. (2002), supra). In such cases, the residues C-terminally flanking the scissile bond may be regarded as a part of the recognition sequence, i.e. such proteases cleave within their recognition sequences.

In order to analyze the sensitivities of the analyzed SUMO-, NEDD8-, scAtg8- and xlUb-specific proteases towards various amino acids in the P1 ^' position, the standard substrates (P1 '-Ala; Fig. 4A) were compared with substrates containing the non-preferred P1 ' residues Met, Tyr, Glu, Arg or Pro (Fig. 12, Fig. 13). Methionine was chosen as the most common case if the authentic N-terminus of a target protein needs to be restored. Tyrosine represents a non-related bulky hydrophobic residue while glutamic acid and arginine are examples for charged residues. Last, the performance of all five proteases on substrates containing proline in the P1 ^' position, a residue known to cause problems for most proteases, was also tested.

As expected, most SUMO- and NEDD8-containing substrates with altered residues in the P1 ^' position required higher protease concentrations for efficient cleavage than the corresponding P1 ^'-Ala constructs (Fig. 12 B, C). These effects were generally moderate (=3-10-fold) for P1 ^'-Met, -Tyr and -Arg substrates. An up to 30- fold higher protease concentration was, however, required for P1 ^'-Glu substrates. A remarkable exception is represented by ssNEDPI , which cleaved all tested substrates (except for the P1 ^'-Pro substrate) with comparable efficiency (Fig. 12C, right panel).

In contrast to published data (Kapust et al. (2002), supra), substrate cleavage by TEV protease was virtually unaffected by the P1 ^'-Met or P1 ^'-Tyr mutations (Fig. 12D). The P1 ^'-Arg substrate, however, required a ~3-fold elevated protease concentration. The P1 ^'-Glu mutant could only be cleaved by an equimolar concentration of TEV protease within one hour at 25°C.

A low sensitivity for residues in the P1 ' position was observed for scAtg4 (Fig. 13B). At 25°C, 1 μΜ protease was, however, generally sufficient to process an 80- 100-fold excess of all tested P1 ^'-variants (except for P1 ^'-Pro).

With regard to the P1 ^'-position, xlUsp2 was remarkably promiscuous and processed its P1 ^'-Ala, -Met, -Tyr, -Arg and -Glu substrates with virtually identical efficiency (Fig. 13C). As the only protease tested here, xlUsp2 could even process a P1 ^'-Pro substrate - although with significantly reduced efficiency.

SUMO-, NEDD8-, Atg8- and ubiquitin-specific proteases show high turnover rates also at low substrate concentrations.

In practice, the amount of protease needed for efficient substrate cleavage strongly depends on the substrate concentration. At high, "saturating" substrate concentrations, the rate of substrate conversion is limited only by the maximal turnover rate of the enzyme (substrate conversions per enzyme per time unit). At lower substrate concentrations, in contrast, the effective turnover is limited by the availability of the substrate. In other words, the number of processed substrate molecules per molecule of enzyme drops when lowering the substrate concentration. A measure for the transition between these two regimes is the Michaelis-Menten constant (KM). The turnover rate reaches half of its maximum when the substrate concentration equals K_M. The two kinetic parameters (maximal turnover rate and K_M) are characteristic for each enzyme/substrate pair and can be used to describe and predict the performance of an enzyme at different substrate concentrations.

Here, the effect of substrate concentration on the protease activity was analyzed in two slightly different setups (Fig. 14). First, the concentrations of both, substrate and protease were reduced proportionally in order to maintain a constant substrate/protease ratio (Fig. 14A). In a second assay, the substrate concentration was titrated down over two orders of magnitude while keeping the absolute protease concentration constant (Fig. 14B). For these experiments, the standard substrates (see Fig. 4A) were used.

These assays showed that the two SUMO-specific proteases can cleave a >10.000-fold excess of substrate within one hour at 0°C. Remarkably, the number of substrate molecules cleaved per molecule of protease remained largely constant when the concentrations of both substrate and protease were reduced proportionally (Fig. 14A). These results indicated that the K_M for the reaction is low, probably in the lower one-digit micromolar range. In line with this interpretation, the digestion efficiency drastically increased if the concentration of substrate was reduced at a constant protease concentration (Fig. 14B). The two NEDD8-specific proteases as well as scAtg4 and xlUsp2 showed a similar general behavior with low K_M, although the maximal turnover rate of these enzymes was significantly lower. The substrate excess successfully cleaved within one hour at 0°C was therefore limited to =500-1000-fold (NEDP1 enzymes), «400-fold (scAtg4) and « 00-fold (xlUsp2), respectively.

According to these results, all analyzed SUMO-, NEDD8-, scAtg8- and xlUb- specific proteases cleave highly efficient also at low substrate concentrations. At substrate concentrations typical for preparative applications (>10μΜ substrate) these enzymes can therefore operate near their maximal turnover rates. Similar titration experiments performed with TEV protease and a TEV protease substrate showed that, here, even at 25°C and at an exceedingly high substrate concentration (300μΜ), the number of substrate molecules cleaved per molecule of TEV protease within one hour was limited to « 100-150. Moreover, the substrate turnover per protease is further reduced at low substrate concentration: When titrating down the substrate at constant protease concentration the fraction of cleaved substrate increased only marginally (Fig. 14B). Along the same lines, reducing the concentration of both the substrate and the protease significantly impaired cleavage (Fig. 14A). These results are consistent with the rather high KM of the reaction that is reported in the literature (50-90μΜ) (Kapust et al. (2002); Kapust et al. (2001 ); Parks et al. (1995), supra). Thus, complete substrate cleavage by TEV protease generally requires a high protease/substrate ratio. At low substrate concentration, the required ratio is even higher.

Analysis of truncated bdSUMO/bdSENPI , bdNEDD8/bdNEDP1 and scAtg8/scAtg4 substrate/protease pairs

In order to narrow down the minimal domains required for proteolytic cleavage, a detailed truncation analysis was performed for selected substrate/protease pairs (Fig. 15). Fusion proteins consisting of a N-terminal maltose-binding protein (MBP), a protease recognition site (PRS; here bdSUMO, bdNEDD8 or scAtg8) and the respective protease (bdSENPI , bdNEDPI , or scAtg4) harboring truncations at defined positions (Fig. 15A) were expressed in E. coli. In this assay, an in vivo cleavage of the fusion protein after the PRS as analyzed by SDS-PAGE of whole cell lysates in SDS sample buffer is indicative for a decent functionality of both the protease recognition site and or the respective protease.

According to these experiments it can be expected that bdSUMO₈3-97 and bdSENPI 288-477 are sufficient for a basal activity of the bdSUMO/bdSENPI system. Likewise, bdNEDD8_4-75, bdNEDPI -13-219, scAtg8₂9-n6, and scAtg4₉i_-388, are required for a basal activity in the bdNEDD8/bdNEDP1 or scAtg8/scAtg4 systems, respectively. Proper cleavage and stability of the proteins can be expected when using larger fragments. According to Fig. 15, proper cleavage and stability in E.coli can be expected for bdSUMO₂i-97, bdSENPI _248-48i , bdNEDD8_2-75, bdNEDPI _{4-2 9}, scAtg8_{9- 6}, and scAtg4 _-494. To also detect subtle differences in cleavage efficiency, selected substrates (according to Fig. 4A) harboring bdSUMO truncations and selected bdSENPI truncations were analyzed in detail in an in vitro assay (Fig. 16). While bdSENPI 242-481 and bdSENPI 248-481 showed virtually identical activity, deletion of five more amino acids from the N-terminus of bdSENPI lead to a significant decrease in proteolytic activity (compare left and middle column to the right column of panels). Likewise, bdSUMO₂-97 and bdSUMO₂i-97 could be cleaved with identical efficiency while bdSUMO_23-9₇ is cleaved with reduced efficiency and cannot be cleaved to completion (compare upper two rows to lower row of panels). In summary, a preferred minimal bdSUMO/bdSENPI system is represented by bdSUMO₂i-97 and bdSENPI _{2 8}- ₈i -

LIST OF REFERENCES

WO 2002/090495 (EP 1 392 717)

WO 2003/057174 (EP 1 470 236)

WO 2008/083271

US 7,498,165

US 7,910,364

US 6,872,551

Makrides SC (1996) Strategies for achieving high-level expression of genes in Escherichia coli. Microbiol Rev 60: 512-538.

Uhlen M, Forsberg G, Moks T, Hartmanis M, Nilsson B (1992) Fusion proteins in biotechnology. Curr Opin Biotechnol 3: 363-369.

Lichty JJ, Malecki JL, Agnew HD, Michelson-Horowitz DJ, Tan S (2005) Comparison of affinity tags for protein purification. Protein Expr Purif 41 : 98-105.

Waugh DS (2005) Making the most of affinity tags. Trends Biotechnol 23: 316-320.

Young CL, Britton ZT, Robinson AS (2012) Recombinant protein expression and purification: a comprehensive review of affinity tags and microbial applications.

Biotechnol J 7: 620-634.

Nilsson J, Stahl S, Lundeberg J, Uhlen M, Nygren PA (1997) Affinity fusion strategies for detection, purification, and immobilization of recombinant proteins.

Protein Expr Purif 1 1 : 1 -16. Arnau J, Lauritzen C, Petersen GE, Pedersen J (2006) Current strategies for the use of affinity tags and tag removal for the purification of recombinant proteins. Protein Expr Purif 48: 1 -13.

Butt TR, Edavettal SC, Hall JP, Mattern MR (2005) SUMO fusion technology for difficult-to-express proteins. Protein Expr Purif 43: 1 -9.

Malakhov MP, Mattern MR, Malakhova OA, Drinker M, Weeks SD et al. (2004) SUMO fusions and SUMO-specific protease for efficient expression and purification of proteins. J Struct Funct Genomics 5: 75-86.

Muller S, Hoege C, Pyrowolakis G, Jentsch S (2001 ) SUMO, ubiquitin's mysterious cousin. Nat Rev Mol Cell Biol 2: 202-210.

Mossessova E, Lima CD (2000) Ulp1 -SUMO crystal structure and genetic analysis reveal conserved interactions and a regulatory element essential for ceil growth in yeast. Mol Cell 5: 865-876.

van der Veen AG, Ploegh HL (2012) Ubiquitin-like proteins. Annu Rev Biochem 81 : 323-357.

Walker PA, Leong LE, Ng PW, Tan SH, Waller S et al. (1994) Efficient and rapid affinity purification of proteins using recombinant fusion proteases. Biotechnology (N Y) 12: 601 -605.

Dian C, Eshaghi S, Urbig T, McSweeney S, Heijbel A et al. (2002) Strategies for the purification and on-column cleavage of glutathione-S-transferase fusion target proteins. J Chromatogr B Analyt Technol Biomed Life Sci 769: 133-144.

van den Berg S, Lofdahl PA, Hard T, Berglund H (2006) Improved solubility of TEV protease by directed evolution. J Biotechnol 121 : 291 -298.

Kapust RB, Tozser J, Copeland TD, Waugh DS (2002) The P1 " specificity of tobacco etch virus protease. Biochem Biophys Res Commun 294: 949-955.

Kostallas G, Lofdahl PA, Samuelson P (201 1 ) Substrate profiling of tobacco etch virus protease using a novel fluorescence-assisted whole-cell assay. PLoS One 6: Θ16136.

Nilsson B, Moks T, Jansson B, Abrahmsen L, Elmblad A et al. (1987) A synthetic IgG-binding domain based on staphylococcal protein A. Protein Eng 1 : 107-1 13. Binz HK, Amstutz P, Kohl A, Stumpp MT, Briand C et al. (2004) High-affinity binders selected from designed ankyrin repeat protein libraries. Nat Biotechnol 22: 575-582. Kapust RB, Tozser J, Fox JD, Anderson DE, Cherry S et al. (2001 ) Tobacco etch virus protease: mechanism of autolysis and rational design of stable mutants with wild-type catalytic proficiency. Protein Eng 14: 993-1000.

Parks TD, Howard ED, Wolpert TJ, Arp DJ, Dougherty WG (1995) Expression and purification of a recombinant tobacco etch virus Nla proteinase: biochemical analyses of the full-length and a naturally occurring truncated proteinase form. Virology 210: 194-201.

Shen LN, Liu H, Dong C, Xirodimas D, Naismith JH et al. (2005) Structural basis of NEDD8 ubiquitin discrimination by the deNEDDylating enzyme NEDP1. EMBO J 24: 1341-1351.

Armougom F, Moretti S, Poirot O, Audic S, Dumas P et al. (2006) Expresso: automatic incorporation of structural information in multiple sequence alignments using 3D-Coffee. Nucleic Acids Res 34: W604-8.

Xu Z, Chau SF, Lam KH, Chan HY, Ng TB et al. (2006) Crystal structure of the SENP1 mutant C603S-SUMO complex reveals the hydrolytic mechanism of SUMO-specific protease. Biochem J 398: 345-352.

Shen LN, Dong C, Liu H, Naismith JH, Hay RT (2006) The structure of SENP1- SUMO-2 complex suggests a structural basis for discrimination between SUMO paralogues during processing. Biochem J 397: 279-288.

Li & Hochstrasser (1999) A new protease required for cell-cycle iirogression in yeast. Nature 398: 246-251.

Gong L, Nillas S, Maul GG, Yeh ETH (2000) Differential Regulation of Sentrinized Proteins by a Novel Sentrin-specific Protease. J. Biol. Chem. 275(5): 3355-3359. Yeh ETH, Gong L, Kamitani T (2000) Ubiquitin-like proteins: new wines in new bottles. Gene 248: 1-14.

Nunn & Djordijevic (2005) Crystal structure of tobacco etch virus protease shows the protein C terminus bound within the active site. J Mol Biol. 350(1 ): 145-155. Kirisako et al. (2000) The reversible modification regulates the membrane-binding state of Apg8/Aut7 essential for autophagy and the cytoplasm to vacuole targeting pathway. J Cell Biol. 151 (2): 263-276.

Li, et al. (2011 ) Kinetics comparisons of mammalian Atg4 homologues indicate selective preferences toward diverse Atg8 substrates. J Biol Chem. 286(9):7327- Gagnon et al. (2007) In Vitro Reconstitution and Affinity Purification of Catalytically Active Archaeal Box C/D sRNP Complexes. Methods in Enzymology 425: 263- 282.

Sequence Database entry XP_003567671 .1 .

Sequence Database entry XP_003567672.1 .

Sequence Database entry ACX1 1 191.1 (sequence 36453 from US 7,569,389). Sequence Database entry C0H840 (UniProt).

Sequence Database entry I1 IVA9 (UniProt).

Sequence Database entry B4F6P0 (UniProt).

Claims

A method for purifying a stoichiometric protein complex composed of at least two subunits from a mixture,

(ii) impurities are washed off the column, and

(iv) optionally removing the cleaved off AT of the second subunit.

The method of claim 1 , wherein in step a) (iii) and/or step b) (iii) the protein complex is eluted by on-column cleavage.

The method of claim 1 or 2, wherein the method further comprises the step of c) removing the protease from the eluate originating from the last affinity chromatography.

The method of any one of claims 1-3, wherein the protein complex comprises a third subunit,

wherein said third subunit comprised in said mixture comprises an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (P S),

(ii) impurities are washed off the column, and

(iv) optionally removing the cleaved off AT of the third subunit.

The method of claim 4, wherein the protein complex comprises a fourth subunit, wherein said fourth subunit comprised in said mixture comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS),

wherein the AT of said fourth subunit differs from the AT of the other subunits and allows affinity chromatography being selective for the AT of said fourth subunit, and wherein the PRS of said fourth subunit is cleavable by a protease which is orthogonal to the PRS of the other subunits,

further comprising after step b') and prior to optional step c) an additional step b") subjecting the eluate from step b') to an affinity chromatography selective for the AT of the fourth subunit, whereby

(ii) impurities are washed off the column, and

(iii) the protein complex is eluted from the column and the AT of the fourth subunit is cleaved off, or the protein complex is eluted by on-column cleavage, using said orthogonal protease which is specific for the PRS of said fourth subunit, preferably wherein the protein complex is eluted by on-column cleavage, and

(iv) optionally removing the cleaved off AT of the fourth subunit.

6. The method of claim 5, wherein the protein complex comprises a fifth subunit, wherein said fifth subunit comprised in said mixture comprise an N-terminal affinity tag (AT) separated from the subunit by a protease recognition site (PRS),

further comprising after step b") and prior to optional step c) an additional step b'") subjecting the eluate from step b") to an affinity chromatography selective for the AT of the fifth subunit, whereby

(ii) impurities are washed off the column, and

optionally removing the cleaved off AT of the fifth subunit.

The method of any one of claims 1 to 6, wherein one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 1 (bdSUMO); or

most preferably wherein the second subunit comprises a PRS consisting of SEQ ID NO: 1 (bdSUMO); and.

wherein the AT of the subunit comprising said PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in amino acids 1-224 of SEQ ID NO: 2 (bdSENP1₂₄₈-48i), or

preferably wherein the subunit is eluted from the column using (i) the protease shown in in amino acids 1 -224 of SEQ ID NO: 2 (bdSENP1₂₄₈-48i ).

(i) an amino acid sequence as shown in SEQ ID NO: 3 (scSUMO); or

wherein the protease shown in SEQ ID NO: 4 (scUlpl) is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 3, under identical conditions; and

wherein the AT of the subunit comprising said PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 4 (scUlpl ), or (ii) a protease derivative of (i) having an amino acid sequence with at least 35% identity over the full length of SEQ ID NO: 4,

wherein said protease derivative is capable of cleaving the PRS according to SEQ ID NO: 3 (scSUMO) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions.

wherein the protease shown in SEQ ID NO: 7 (hsSENPI) is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 5 or 6, respectively, under identical conditions; and

wherein the AT of the subunit comprising said PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 7 (hsSENPI), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 45% identity over the full length of SEQ ID NO: 7,

wherein said protease derivative is capable of cleaving the PRS according to ID NO: 5 (hsSUMO a) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions.

0. The method of any one of claims 1 to 9, wherein one PRS comprises, preferably consists of

(ii) a PRS derivative of (i) with an amino acid sequence having at least 85% identity over the full length of SEQ ID NO: 9 or at least 99% over the full length of identity over the full length of SEQ ID NO: 8 or 10,

wherein the protease shown in SEQ ID NO: 1 1 (ssNEDPI ), SEQ ID NO: 12 (bdNEDPI ), or SEQ ID NO: 13 (hsNEDPI ) is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 8, 9 or 10, respectively, under identical conditions;

more preferably wherein the first subunit comprises a PRS comprising an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8) or SEQ ID NO: 9 (bdNEDD8), in particular wherein the first subunit comprises a PRS comprising an amino acid sequence as shown in SEQ ID NO: 9 (bdNEDD8);

most preferably wherein the first subunit comprises a PRS consisting of an amino acid sequence as shown in SEQ ID NO: 8 (ssNEDD8) or SEQ ID NO: 9 (bdNEDD8), in particular wherein the first subunit comprises a PRS consisting of an amino acid sequence as shown in SEQ ID NO: 9 (bdNEDD8); and wherein the AT of the subunit comprising said PRS is cleaved off using

(i) a protease comprising, preferably consisting of the amino acid sequence selected from the group consisting of amino acid sequences shown in SEQ ID NO: 11 (ssNEDPI), SEQ ID NO: 12 (bdNEDPI), and SEQ ID NO: 13 (hsNEDPI), or

(ii) a protease derivative of (i) having an amino acid sequence with at least 70% identity over the full length of SEQ ID NO: 11 (ssNEDPI), or SEQ ID NO: 13 (hsNEDPI), or with at least 35% identity over the full length of SEQ ID NO: 12 (bdNEDPI),

preferably wherein the subunit is eluted from the column using

(i) a protease comprising, preferably consisting of the amino acid sequence selected from the group consisting of amino acid sequences shown in SEQ ID NO: 11 (ssNEDPI), and SEQ ID NO: 12 (bdNEDPI), or

(ii) a protease derivative having an amino acid sequence with at least 70% identity over the full length of SEQ ID NO: 11 (ssNEDPI) or with at least 35% identity over the full length of SEQ ID NO: 12 (bdNEDPI), wherein said protease derivative, cleaves the PRS according to ID NO: 9 (bdNEDD8) with at least 20% activity as compared to the parent protease as defined in (i), under identical conditions;

more preferably wherein the subunit is eluted from the column using a protease comprising, preferably consisting of the amino acid sequence selected from the group consisting of amino acid sequences shown in SEQ ID NO: 11 (ssNEDPI), and SEQ ID NO: 12 (bdNEDPI); and

most preferably wherein the subunit is eluted from the column using the protease comprising, preferably consisting of the amino acid sequence shown in SEQ ID NO: 12 (bdNEDPI).

1 . The method of any one of claims 1 to 10, wherein at least one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 14 (scAtg8); or

wherein the protease shown in SEQ ID NO: 15 (scAtg4), is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 14 under identical conditions; and

wherein the AT of the subunit comprising said PRS is cleaved off using

(ii) a protease derivative of (i) having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 15, wherein said protease derivative is capable of cleaving the PRS according to ID NO: 14 (scAtg8) with at least 20% activity as compared to the parent protease as defined in (i).

12. The method of any one of claims 1 to 11 , wherein at least one PRS comprises, preferably consists of the TEV protease recognition site shown in SEQ ID NO: 16; and wherein the AT of the subunit comprising said PRS is cleaved off using

13. The method of any one of claims 1 to 12, wherein at least one PRS comprises, preferably consists of

(i) an amino acid sequence as shown in SEQ ID NO: 20 (xlUb); or (ii) a PRS derivative of (i) with an amino acid sequence having at least 80% identity over the full length of SEQ ID NO: 20,

wherein the protease shown in SEQ ID NO: 21 (xlUsp2), is capable of cleaving said PRS derivative with at least 20% activity as compared to when using the parent PRS with the amino acid sequence of SEQ ID NO: 20 under identical conditions; and

wherein the AT of the subunit comprising said PRS is cleaved off using

14. The method of any one of claims 1 -3 and 7-13, wherein the protein complex is composed of 2 different subunits, preferably with a stoichiometry of 1 :1.

15. The method of any one of claims 1-4 and 7-13, wherein the protein complex is composed of 3 different subunits, preferably with a stoichiometry of 1 :1 :1.

16. The method of any one of claims 1-13, wherein the protein complex is composed of 4 different subunits, preferably with a stoichiometry of 1 : 1 :1 :1.

17. The method of any one of claims 1 -16, wherein the mixture originates from a mixture of lysates and/or supernatants and/or a pre-purified solution, each comprising at least one of the subunits.

18. The method of any one of claims 1-16, wherein the mixture originates from a single lysate or supernatant or a pre-purified solution comprising all subunits of the protein complex.

19. The method of any one of claims 1 -18, wherein the first subunit comprises a poly-His tag, and preferably wherein the second subunit comprises a ZZ tag.

20. The method of any one of claims 1-19, wherein the first subunit comprises the PRS as defined in claim 10, and wherein the elution in step a) is carried out using the protease as defined in claim 10; and wherein the second subunit comprises the PRS as defined in claim 7 and wherein the elution in step b) is carried out using the protease as defined in claim 7.

21 . The method of claim 20, wherein the first subunit comprises a poly-His tag, and wherein the second subunit comprises a ZZ tag.

22. The method of any one of claims 1-21 , wherein step c) is an affinity chromatography, a size exclusion chromatography, or a precipitation step.

23. The method of any one of claims 1 -22, wherein the protease used for cleaving an affinity tag from a subunit comprises an affinity tag which is the same than one of the affinity tags used in the affinity chromatography steps a), b), b'), b") or b'"), with the provisio that the affinity tag differs from the affinity tag used in the directly preceding affinity chromatography step.

24. The method of any one of claims 1-23, wherein the protease from the eluate originating from the last affinity chromatography prior to step c) comprises an affinity tag, preferably as defined in claim 19, and

wherein step c) is an affinity chromatography step, whereby the protease binds to the affinity resin, and the protein complex is collected in the flow-through.

25. The method of any one of claims 1 -24, wherein the subunit(s) further comprise a spacer between the AT and the PRS, and/or between the PRS and the subunit; preferably wherein the subunit(s) further comprise a spacer between the AT and the PRS.

26. A protease having an amino acid sequence with at least 45% identity over the full length of SEQ ID NO: 2 (bdSENPI ),

wherein said protease is capable of cleaving the PRS according to ID NO: 1 (bdSUMO) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 2;

preferably wherein the protease comprises the amino acid sequence shown as amino acids 1-224 in SEQ ID NO: 2 (bdSENP1_248-48i ).

27. The protease of claim 26, wherein the protease consists of the amino acid sequence shown as amino acids 1 -224 in SEQ ID NO: 2 (bdSENP1_248-48i ).

28. A protease having an amino acid sequence with at least 70% identity over the full length of SEQ ID NO: 1 1 (ssNEDPI), wherein said protease is capable of cleaving the PRS according to ID NO: 8 (ssNEDD8) with at least 20% activity as compared to the parent protease with the amino acid sequence of SEQ ID NO: 1 1 ;

preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 1 1 (ssNEDPI );

more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 1 1 (ssNEDPI ).

29. A protease having an amino acid sequence with at least 35% identity over the full length of SEQ ID NO: 12 (bdNEDPI),

preferably wherein the protease comprises the amino acid sequence as shown in SEQ ID NO: 12 (bdNEDPI);

30. A protease having an amino acid sequence with at least 80% identity over the full length of SEQ ID NO: 21 (xlUsp2),

more preferably wherein the protease consists of the amino acid sequence as shown in SEQ ID NO: 21 (xlUsp2).

31 . The protease of claim 30, wherein the protease has an amino acid sequence with at least 98% identity over the full length of SEQ ID NO: 21 (xlUsp2).

32. The protease of any one of claims 26 to 31 , wherein the protease further comprises an affinity tag, preferably an affinity tag as defined in claim 19.

33. A kit of parts, comprising at least two proteases selected from

(i) the protease according to claim 26 or 27,

(ii) the protease according to claim 28 or 29,

(iii) the protease as defined in claim 12, (iv) the protease as defined in claim 1 1 , and

(v) the protease according to claim 30 or 31 ,

preferably comprising at least two proteases selected from (i)-(iii),

more preferably comprising two proteases selected from (i) and (ii), and most preferably comprising the protease according to any one of claims 26-27 and the protease according to claim 29.

34. The kit of parts of claim 33, wherein at least one of the proteases further comprises an affinity tag, preferably an affinity tag as defined in claim 19.

35. Use of a protease as defined in any one of claims 26-32, or the kit of parts as defined in claim 33 or 34 in a method of purifying stoichiometric protein complexes comprising at least two subunits,

preferably wherein the method is further defined as in any one of claims 1-30.

36. Use of a protease as defined in any one of claims 26-32, or the kit of parts as defined in claim 33 or 34 for on-column cleavage in an affinity chromatography.

37. A nucleic acid molecule, encoding the protease according to any one of claims 26-32.

38. An enzyme formulation comprising a purified, recombinant polypeptide comprising an amino acid sequence with at least 45% sequence identity to SEQ ID NO: 2 (bdSENP1₂₄₈-48i ) ,

wherein said polypeptide

(i) is capable of cleaving at least 90% of a 2 000-fold molar excess of a native substrate protein shown in SEQ ID NO: 22 (His₁ -bdSUMO-MBP) at standard conditions of 1 hour incubation at 0 °C, 100 μΜ initial concentration of substrate protein in a buffer consisting of 250 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2Mg/ml BSA, and (ii) is capable of cleaving at least 90% of a 100-fold molar excess of native substrate protein variants in which only residue 123 in SEQ ID NO: 22 (the Pt ' position of His₁₄-bdSUMO-MBP) has been mutated to Met, Tyr, Arg or Glu relative to SEQ ID NO: 22 at standard conditions as defined in (i), and

(iii) is capable of cleaving at least 50% of a 2 000-fold molar excess of a native substrate protein as shown in SEQ ID NO: 22 (His₁₄-bdSUMO- MBP) within one hour at 0°C at high-salt conditions of 100 μΜ initial concentration of substrate protein in a buffer consisting of 1000 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2pg/ml BSA, and

(iv) cleaves at standard conditions as defined in (i) any of the substrates shown in SEQ ID NO: 23 (His₁₀-ZZ-TEV-MBP), SEQ ID NO: 24 (Hisi₄- bdNEDD8-MBP), SEQ ID NO: 25 (His₁ -ssNEDD8-MBP), SEQ ID NO: 26 (His₁₄-scAtg8-MBP) or SEQ ID NO: 27 (His -xlUb-MBP) at least 10 000 fold less efficiently than the substrate shown in SEQ ID NO: 22 (Hisi₄-bdSUMO-MBP), and

(v) cleaves at standard conditions as defined in (i) the substrate shown in SEQ ID NO: 28 (Hisi₄-scSUMO-MBP) at least 25-fold less efficiently than the substrate as shown in SEQ ID NO: 22 (His₁ -bdSUMO-MBP), and

(vi) if the polypeptide does not comprise a polyHis-tag, the polypeptide is capable of cleaving a substrate protein as shown in SEQ ID NO: 22 (Hisi₄-bdSUMO-MBP) immobilized on a Ni(ll) chelate resin with at least 10% efficiency as compared to the non-immobilised substrate at standard conditions as defined in (i).

An enzyme formulation comprising a purified, recombinant polypeptide comprising an amino acid sequence with at least 70% sequence identity to SEQ ID NO: 1 1 (ssNEDPI ),

wherein said polypeptide

(i) is capable of cleaving at least 90% of a 100-fold molar excess of a native Hisi₄-ssNEDD8-MBP substrate protein (SEQ ID NO: 25) at standard conditions of 1 hour incubation at 0 °C, 100 μΜ initial concentration of substrate protein in a buffer consisting of 250 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2pg/ml BSA, and

(ii) is capable of cleaving at least 90% of a 100-fold molar excess of native substrate protein variants in which only residue 103 in SEQ ID NO: 25 (the Pi' position His₁₄-ssNEDD8-MBP) has been mutated to Met, Tyr, Arg or Glu relative to SEQ ID NO: 25 at standard conditions as defined in (i), and

(iii) is capable of cleaving at least 70% of a 100-fold molar excess of a native substrate protein as shown in SEQ ID NO: 25 (His ₄-ssNEDD8-MBP) within one hour at 0°C at high-salt conditions of 100 μΜ initial concentration of substrate protein in a buffer consisting of 1000 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2pg/ml BSA, and

(iv) cleaves at standard conditions as defined in (i) any of the substrates shown in SEQ ID NO: 23 (Hisi₀-ZZ-TEV-MBP), SEQ ID NO: 22 (His₁₄- bdSUMO-MBP) or SEQ ID NO: 26 (His₁₄-scAtg8-MBP) at least 1000 fold less efficiently than the substrate shown in SEQ ID NO: 25 (Hisi - ssNEDD8-MBP), and

(v) if the polypeptide does not comprise a polyHis-tag, the polypeptide is capable of cleaving a substrate protein shown in SEQ ID NO: 25 (His-| - ssNEDD8-MBP) immobilized on a Ni(ll) chelate resin with at least 10% efficiency as compared to the non-immobilised substrate at standard conditions as defined in (i).

An enzyme formulation comprising a purified, recombinant polypeptide comprising an amino acid sequence with at least 35% sequence identity to SEQ ID NO: 12 (ssNEDPI),

wherein said polypeptide

(i) is capable of cleaving at least 90% of a 200-fold molar excess of a native substrate protein shown in SEQ ID NO: 24 (His₁ -bdNEDD8-MBP) at standard conditions of 1 hour incubation at 0 °C, 100 μΜ initial concentration of substrate protein in a buffer consisting of 250 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2pg/ml BSA, and

(ii) is capable of cleaving at least 90% of a 30-fold molar excess of native substrate protein variants in which only residue 103 in SEQ ID NO: 24 (the Pi' position of His -bdNEDD8-MBP) has been mutated to Met, Tyr, Arg or Glu relative to SEQ ID NO: 24 at standard conditions as defined in (i), and

(iii) is capable of cleaving at least 70% of a 200-fold molar excess of a native substrate protein as shown in SEQ ID NO: 24 (His₁ -bdNEDD8-MBP) within one hour at 0°C at high-salt conditions of 100 μΜ initial concentration of substrate protein in a buffer consisting of 1000 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2Mg/ml BSA, and

(iv) cleaves at standard conditions as defined in (i) any of the substrates shown in SEQ ID NO: 23 (His₁₀-ZZ-TEV-MBP), SEQ ID NO: 22 (His₁₄- bdSUMO-MBP), SEQ ID NO: 26 (His₁₄-scAtg8-MBP) or SEQ ID NO: 27 (Hisu-xlUb-MBP) at least 1000 fold less efficiently than the substrate shown in SEQ ID NO: 24 (His₁₄-bdNEDD8-MBP), and

(v) if the polypeptide does not comprise a polyHis-tag, the polypeptide is capable of cleaving a substrate protein as shown in SEQ ID NO: 24 (Hisi₄-bdNEDD8-MBP) immobilized on a Ni(II) chelate resin with at least 10% efficiency as compared to the non-immobilised substrate at standard conditions as defined in (i).

An enzyme formulation comprising a purified, recombinant polypeptide comprising an amino acid sequence with at least 80% sequence identity to SEQ ID NO: 21 (xlUsp2),

wherein said polypeptide

(i) is capable of cleaving at least 70% of a 100-fold molar excess of a native substrate protein shown in SEQ ID NO: 27 (Hisi₄-xlUb-MBP) at standard conditions of 1 hour incubation at 0 °C, 100 μΜ initial concentration of substrate protein in a buffer consisting of 250 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2pg/ml BSA, and

(ii) is capable of cleaving at least 90% of a 100-fold molar excess of native substrate protein variants in which only residue 1 1 1 in SEQ ID NO: 27 (the Ρ- position of Hisu-xlUb-MBP) has been mutated to Met, Tyr, Arg or Glu relative to SEQ ID NO: 27 at standard conditions as defined in (i), and

(iii) is capable of cleaving at least 40% of a 30-fold molar excess of a native substrate protein variant in which only residue 1 1 1 in SEQ ID NO: 27 (the P-i' position Hisu-xlUb-MBP) has been mutated to Pro relative to SEQ ID NO: 27 at conditions defined in (ii), and

(iv) is capable of cleaving at least 40% of a 100-fold molar excess of a native substrate protein shown in SEQ ID NO: 27 (HiSi -xlUb-MBP) within one hour at 0°C at high-salt conditions of 100 μΜ initial concentration of substrate protein in a buffer consisting of 1000 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2Mg/ml BSA, and (v) cleaves at standard conditions as defined in (i) any of the substrates shown in SEQ ID NO: 23 (His₁₀-ZZ-TEV-MBP), SEQ ID NO: 22 (His₁₄- bdSUMO-MBP), SEQ ID NO: 24 (Hisi₄-bdNEDD8-MBP), SEQ ID NO: 25 (Hisi4-ssNEDD8-MBP), or SEQ ID NO: 26 (His₁₄-scAtg8-MBP) at least 1000 fold less efficiently than the substrate SEQ ID NO: 27 (His₁₄-xlUb- MBP).

An enzyme formulation comprising a purified, recombinant polypeptide comprising an amino acid sequence with at least 80% sequence identity to SEQ ID NO: 15 (scAtg4),

wherein said polypeptide

(i) is capable of cleaving at least 90% of a 200-fold molar excess of a native substrate protein shown in SEQ ID NO: 26 (His ₄-scAtg8-MBP) at standard conditions of 1 hour incubation at 0 °C, 100 μΜ initial concentration of substrate protein in a buffer consisting of 250 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2pg/ml BSA, and

(ii) is capable of cleaving at least 90% of a 100-fold molar excess of native substrate protein variants in which only residue 150 in SEQ ID NO: 26 (the Pi' position His₁₄- scAtg8-MBP) has been mutated to Met, Tyr, Arg or Glu relative to SEQ ID NO: 26 at standard conditions as defined in (i), and

(iii) is capable of cleaving at least 25% of a 200-fold molar excess of a native substrate protein shown in SEQ ID NO: 26 (Hisi₄-scAtg8-MBP) within one hour at 0°C at high-salt conditions of 100 μΜ initial concentration of substrate protein in a buffer consisting of 1000 mM NaCI, 40 mM Tris/HCI pH 7.5, 2 mM MgCI₂, 250 mM sucrose, 2 mM DTT, 2pg/ml BSA, and

(iv) cleaves at standard conditions as defined in (i) any of the substrates shown in SEQ ID NO: 23 (His₁₀-ZZ-TEV-MBP), SEQ ID NO: 22 (His₁₄- bdSUMO-MBP), SEQ ID NO: 24 (His₁₄-bdNEDD8-MBP), SEQ ID NO: 25 (His₁₄-ssNEDD8-MBP), or SEQ ID NO: 27 (His₁₄-xlUb-MBP) at least 1000 fold less efficiently than the substrate shown in SEQ ID NO: 26 (His ₄-scAtg8-MBP).