US20150344549A1

US20150344549A1 - Split inteins, conjugates and uses thereof

Info

Publication number: US20150344549A1
Application number: US14/411,702
Authority: US
Inventors: Tom W. Muir; Miquel Vila-Perello; Zhihua Liu; Neel H. Shah
Original assignee: Princeton University
Current assignee: Princeton University
Priority date: 2012-06-27
Filing date: 2013-06-24
Publication date: 2015-12-03
Also published as: PT2877490T; US20200087388A1; ES2709333T3; US10526401B2; CA2877882A1; WO2014004336A3; JP2018111685A; WO2014004336A2; US20180057577A1; EP2877490A2; JP6591511B2; ES2930185T3; DK3431497T3; EP4219549A1; EP2877490B1; JP2015522020A; US12054541B2; EP3431497B1; US11161899B2; PL2877490T3

Abstract

Disclosed herein are split inteins, fused proteins of split inteins, and methods of using split inteins to efficiently purify and modify proteins of interest. Thus, provided herein are fusion proteins of a polypeptide and a split intein N-fragment, or variant thereof, as described below in greater detail. Also provided are complexes of the fusion protein and a split intein C-fragment or variant thereof as described in detail below. The complex of the fusion protein and C-fragment or variant thereof can be via a covalent interaction between the fusion protein and C-fragment or variant or via a noncovalent interaction (e.g., ionic, H-bonding, and/or van der Waals interaction). Further provided herein are split intein C-fragments or variants thereof.

Description

STATEMENT OF U.S. GOVERNMENT SUPPORT

This invention was made with U.S. government support under Grant No. GM086868 awarded by the National Institutes of Health (NIH). The U.S. government has certain rights in the invention.
This application contains, as a separate part of disclosure, a Sequence Listing in computer-readable form (filename: 47046A_SeqListing.txt; 547,838 bytes—ASCII text file—created Jun. 20, 2013) which is incorporated by reference in its entirety.

BACKGROUND

Protein splicing is a post-translational process catalyzed by a family of proteins known as inteins.(1) During this process, an intein domain catalyzes its own excision from a larger precursor protein and simultaneously ligates the two flanking polypeptide sequences (exteins) together. While most inteins catalyze splicing in cis, a small subset of these proteins exist as naturally fragmented domains that are separately expressed but rapidly associate and catalyze splicing in trans. Given their capacity to make and break polypeptide bonds (inteins can be considered protein ligases), both cis and trans-splicing inteins have found widespread use as chemical biological tools.(2)
Despite the growing use of inteins in chemical biology, their practical utility has been constrained by two common characteristics of the family, namely (i) slow kinetics and (ii) context dependent efficiency with respect to the immediate flanking extein sequences.(3,4) Recently, a split intein from the cyanobacterium Nostoc punctiforme (Npu) was shown to catalyze protein trans-splicing on the order of a minute, rather than hours like most cis- or trans-splicing inteins.(5) Furthermore, this intein was slightly more tolerant of sequence variation at the critical +2 C-extein residue than other inteins.(6)
Thus, a need exists for more robust and more efficient split inteins for use in a variety of protein purification and protein modification applications.

SUMMARY

Disclosed are split intein N- and C-fragments, variants thereof, and methods of using these split inteins in polypeptide purification and modification.
Thus, provided herein are fusion proteins of a polypeptide and a split intein N-fragment, or variant thereof, as described below in greater detail. Also provided are complexes of the fusion protein and a split intein C-fragment or variant thereof as described in detail below. The complex of the fusion protein and C-fragment or variant thereof can be via a covalent interaction between the fusion protein and C-fragment or variant or via a noncovalent interaction (e.g., ionic, H-bonding, and/or van der Waals interaction).
Further provided herein are split intein C-fragments or variants thereof. In some cases the split intein C-fragment further comprises a linker, such as a peptide linker, or other linkers as described below in detail. A specific peptide linker contemplated is -SGGC (SEQ ID NO: 705) attached to any of the split intein C-fragments described below. The linker can be tailored so as to allow for attachment of a split intein C-fragment of interest to a support, e.g., a bead, a resin, a slide, a particle.
Also provided herein are methods using the split intein N- and C-fragments, or variants thereof, as described in detail below. More particularly, provided herein are methods comprising (a) contacting (1) a fusion protein comprising a polypeptide and a split intein N-fragment, or a variant thereof, as described in detail below and (2) a split intein C-fragment or a variant thereof, as described in detail below; wherein contacting is performed under conditions that permit binding of the split intein N-fragment to the split intein C-fragment to form an intein intermediate; and (b) contacting the intein intermediate with a nucleophile to form a conjugate of the protein and the nucleophile. In various embodiments, the split intein C-fragment, or variant thereof, is bound to a support. In some embodiments, the support is a bead, a resin, a particle or a slide. It will be appreciated that selection of the N-fragment and C-fragment can be from the same wild type split intein (e.g., both from Npu, or a variant of either the N- or C-fragment as discussed in great detail below), or alternatively can be selected from different wild type split inteins or the consensus split intein sequences discussed below, as it has been discovered that the affinity of a N-fragment for a different C-fragment (e.g., Npu N-fragment or variant thereof with Ssp C-fragment or variant thereof) still maintains sufficient binding affinity for use in the disclosed methods. Moreover, such a finding allows for a single C-fragment or variant thereof bound to a support to be useful in purification and/or modification methods disclosed herein with a fusion protein wherein the N-fragment is any of the ones disclosed herein, or a variant thereof. Thus, one can select an N-fragment that has advantages for any individual polypeptide of interest, e.g., one that expresses better than others disclosed herein.
The fusion protein can be in a whole cell lysate or secreted from a cell (e.g., a mammalian cell) and in a cell supernatant. In some cases, the polypeptide of the fusion protein is an antibody, e.g., an IgG antibody. In some embodiments, the N-fragment is fused to one or both of the heavy chains of the antibody. In some embodiments, the N-fragment is fused to one or both of the light chains of the antibody. The methods disclosed herein can further comprise washing the intein intermediate (prior to contact with the nucleophile) to remove the cell lysate or cell supernatant, for example.
The methods disclosed herein can further comprise isolating the resulting conjugate of the polypeptide and nucleophile. Thus, the methods disclosed herein can be useful as an efficient purification for polypeptides prepared by recombinant protein methods.
The nucleophile can be a thiol to form a conjugate that is an α-thioester of the polypeptide. In some cases, the resulting α-thioester can be further modified by contacting with a second nucleophile, employing the well known α-thioester chemistry for protein modification. In some cases, the methods disclosed herein can provide conjugates of the polypeptide, which in some cases is an antibody (e.g., an IgG antibody), and a nucleophile (e.g., a drug, a polymer, an oligonucleotide).

BRIEF DESCRIPTIONS OF THE DRAWINGS

FIG. 1 shows trans-splicing of split DnaE inteins. (a) Scheme depicting protein trans-splicing of the KanR protein with a variable local C-extein sequence. (b) In vivo relative trans-splicing efficiencies at 30° C. with the endogenous “CFN” C-extein sequence and exogenous “CGN”, “CEN”, and “CRN” sequences. IC₅₀values (±SE, n=3-4) are normalized relative to that of intact KanR proteins with the appropriate C-extein tri-peptide.

FIG. 2 shows in vitro half-lives of trans-splicing reactions. Indicated split intein pairs fused to model exteins Ub or SUMO (Ub-Int^Nand Int^C-SUMO) were mixed at either 30° C. or 37° C., and the formation of products was monitored over time by gel electrophoresis. (a) Half-lives were extracted from the reaction progress curves fit to a standard first-order rate equation (±SE, n=3). Representative coomassie-stained SDS-PAGE gels showing (b) fast Ava splicing at 37° C. and (c) inefficient Ssp splicing at 37° C.

FIG. 3 shows sequence-activity relationships in split DnaE inteins. (a) Inteins in order of in vivo splicing activity with selected slices from the corresponding multiple sequence alignment. (b) Rendering of the Npu structure highlighting the proximity of position 120 to the terminal catalytic residues C1 and N137. (c) In vivo analysis of the C120G mutation in the Aha intein (±SD, n=3). (d) Rendering of the Npu structure highlighting key catalytic residues (sticks) and important non-catalytic positions (spheres) that modulate Ssp activity. (e) In vivo analysis of Ssp-to-Npu point mutations that improve Ssp activity (±SD, n=4). Note that all residue numberings correspond to the relevant positions on Npu as defined by the NMR structure (PDB: 2KEQ).(21)

FIG. 4 shows engineered versions of ultrafast DnaE inteins support efficient expressed protein ligation. (a) Scheme depicting the formation of the linear thioester intermediate and its use to generate a protein α-thioester for EPL. (b) Coomassie-stained SDS-PAGE gel depicting efficient MESNa thiolysis of ubiquitin from a fused AvaDnaE intein to yield the Ub-MES thioester, 4. (c) Fluorescent SDS-PAGE gels showing the formation of the Ub-CGK(Fluorescein) ligated product (6) from one-pot thiolysis and native chemical ligation reactions using the inteins indicated. (d) Reverse phase HPLC chromatographs showing pH dependence of the relative populations of precursor amide (1) and linear thioester (2).

FIG. 5 shows sequence alignments of split DnaE inteins. Numbering follows that of Npu as assigned for the NMR structure (PDB 2KEQ). Critical catalytic residues are marked with an asterisk.

FIG. 6 shows sequence logos for high- and low-activity inteins. Inteins are ranked based on in vivo activity with a “CFN” C-extein sequence. The high and low activity inteins are distinguished based on a cut-off IC₅₀value of 350 μg/mL of kanamycin, and the Aha intein is included in the high-activity set, given that the C¹²⁰G mutation dramatically restores high activity.

FIG. 7 shows purification of C-terminal α-thioesters using split-inteins. A) Scheme of the split-intein based purification of protein C-terminal α-thioesters. B) Sequence of WT Npu^Cand its mutant Npu^C-AA having a linker to immobilize it onto a solid support (underlined). For the thiolysis experiments in solution the C-terminal Cys residue was previously alkylated with iodoacetamide.

FIG. 8 shows Purification of soluble protein α-thioesters using the Npu^C-AA affinity column. A) Scheme of the purification strategy using split Npu DnaE intein. B) Purification of Ub-thioester (Ub-COSR) from cell lysates. C) Purification of MBP-thioester (MBP-COSR) from cell lysates. Both purifications were monitored by SDSPAGE analysis stained with Coomassie (top) or Western Blot using an α-His antibody.

FIG. 9 shows RP-HPLC and MS analysis of Ub and MBP α-thioesters. A) RP-HPLC (top) and MS (bottom) analysis of Ub-COSR eluted from the Npu^C-AA column. B) RP-HPLC (top) and MS (bottom) analysis of MBP-COSR eluted from the Npu^C-AA column.

FIG. 10 shows the effect of the −1 residue on the efficiency of the on-resin thiolysis. 20 different Ub-Npu^Nproteins were expressed containing each of the 20 proteinogenic amino acids at the C-terminus of Ub (−1 residue) and purified over Npu^C-AA columns. Cleavage yields from the Npu^C-AA column were estimated by gel electrophoresis and amounts of thioester versus side reactions (mainly hydrolysis) were determined by RP-HPLC and MS analysis.

FIG. 11 shows purification of H2B(1-116)-α-thioester under denaturing conditions. A) SDSPAGE analysis of the purification of H2B(1-116) α-thioester over the Npu^C-AA column in the presence of 3 M urea. RP-HPLC (B) and MS (C) analysis of E1 from panel A confirmed the presence of the desired H2B thioester.

FIG. 12 shows purification of αDEC thioesters expressed in 293T cells using the split-intein column. A) Expression levels of αDEC fused to different inteins in 293T cells. B) Purification of αDEC α-thioester through the Npu^C-AA affinity column. C) Expressed Protein Ligation (EPL) of αDEC-thioester with an N-terminal Cys containing fluorescent peptide.

FIG. 13 shows EPL directly using Int^C-column eluted thioesters. RP-HPLC (30-73% B gradient, 214 nm and 440 nm detection) and MS analysis of the reactions between the H-CGK(F1)-NH₂peptide and MBP (A) and PHPT1 (B) MES thioesters, purified from E. coli using the Int^C-column.

FIG. 14 shows a one-pot purification/ligation experiment of ubiquitin to the H-CGK(Fluorescein)-NH₂peptide (CGK(F1)). Ub-Npu^Nfrom E. coli cell lysates was bound to the Int^C-column, and after removal of contaminants through extensive washes, intein cleavage and ligation were triggered by addition of 200 mM MES and 1 mM CGK(F1) peptide. Coomassie stained SDS-PAGE analysis and in gel fluorescence of the purification/ligation (left). RP-HPLC (detection at 214 and 440 nm) and ESI-TOF MS (right) of the eluted fractions confirms the desired ligated protein was obtained in one step directly from cell lysates with a ligation yield close to 95% (quantified by RP-HPLC).

FIG. 15 shows the semi-synthesis of H2B-K120Ac under denaturing conditions. A) Coomassie stained SDS-PAGE analysis of H2B(1-116) α-thioester generation in the presence of 2 M urea (sup: cell lysate supernatant, trit: 1% triton wash of the inclusion bodies, inp: solubilized inclusion bodies used as input for the Int^C-column). E1-E6 were pooled, concentrated to 150 μM and ligated to the peptide H-CVTK(Ac)YTSAK-OH at 1 mM for 3 h at r.t. B) RP-HPLC (left) of the ligation reaction mixture and MS (right) of the ligated H2B-K120Ac product.

FIG. 16 shows the characterization of αDEC205 ligated to the H-CGK(Fluorescein)-NH2 peptide (CGK(F1)). Elution fractions from the Npu^C-column containing αDEC205-MES thioester were concentrated to 20 μM and ligated to the CGK(F1) fluorescent peptide at 1 mM for 48 h at r.t. A) ESI-TOF MS analysis of degycosylated and fully reduced HC after ligation, showing 75% of the HC are labeled. Expected mass for ligation product=50221.2 Da. Free HC=49575.0 Da. B) SEC-MALS analysis of the ligated antibody showing that it retains its tetrameric structure after thiolysis and ligation (MW=151 kDa, MW calc=148 kDa). C) Binding of αDEC205-CGK(F1) to the DEC205 receptor. Dose dependent binding of αDEC205-CGK(F1) (left) or a control α-DEC205 antibody (right) to CHO cells expressing the mouse DEC205 receptor monitored by flow cytometry using a PE labeled α-mouse IgG. Binding to control CHO/NEO cells, which don't express the receptor is shown in gray.

FIG. 17 shows purification of αDEC thioesters expressed in CHO cells using a split-intein column. Top) Coomassie stained SDSPAGE gel of the purification of αDEC-MES thioester from CHO cells using a Npu^C-column. Bottom) Western blot analysis of the same purification.

FIG. 18 shows purification of αDEC thioesters using an Ava^Csplit-intein column and Western blot analysis of the purification of αDEC thioesters from mammalian cell supernatants using an Ava^C-column.

FIG. 19 shows expression tests of αDEC205 antibody fused to Ava^Nsplit intein through the C-terminus of the antibody light chain and western blot analysis of CHO cell supernatants expressing the αDEC205-AvaN fusion at different timepoints.

DETAILED DESCRIPTION

Of the roughly 600 inteins currently catalogued, (7) less than 5% are split inteins, mostly from a family known as the cyanobacterial split DnaE inteins (8). Surprisingly, only six of these, including Npu, have been experimentally analyzed to any extent, (6,9,10) and only Npu and its widely-studied, low-efficiency ortholog from Synechocystis species PCC6803 (Ssp) have been rigorously characterized in vitro.(5,11)
A rapid survey of 18 split DnaE inteins was performed using an in vivo screening method to accurately compare the efficiencies of split inteins(12,13) In this assay, the two fragments of a split intein are co-expressed in E. coli as fusions to a fragmented aminoglycoside phosphotransferase (KanR) enzyme. Upon trans-splicing, the active enzyme is assembled, and the bacteria become resistant to the antibiotic kanamycin (FIG. 1 a). More active inteins confer greater kanamycin resistance and thus have a higher IC₅₀value for bacterial growth as a function of kanamycin concentration. This assay can be carried out in the background of varying local C-extein sequences without significantly perturbing the dynamic range. Since all DnaE inteins splice the same local extein sequences in their endogenous context, this screen was originally carried out in a wild-type C-extein background (CFN) within the KanR enzyme. As expected, bacteria expressing the Npu intein had a high relative IC₅₀, whereas clones expressing Ssp showed poor resistance to kanamycin. Remarkably, more than half of the DnaE inteins showed splicing efficiency comparable to Npu in vivo at 30° C. (FIG. 1 b).
To confirm that the high IC₅₀values observed in vivo reflected rapid trans-splicing, a series of kinetic studies were performed under standardized conditions in vitro. For this, individually expressed and purified split DnaE intein fragments fused to model N- and C-extein domains, ubiquitin and SUMO were made. The endogenous local extein residues were preserved as linkers between the extein domains and intein fragments to recapitulate a wild-type-like splicing context. Cognate intein fragments were mixed at 1 μM, and the formation of the Ub-SUMO spliced product at 30° C. and 37° C. was monitored by gel electrophoresis. These assays validated that the new inteins with high-activity in vivo could catalyze trans-splicing in vitro in tens of seconds, substantially faster than Ssp (FIG. 2 a). Interestingly, all of the inteins analyzed except Ssp showed increased splicing rates at 37° C. Furthermore, all of the fast-splicing inteins showed low-to-undetectable levels of side reactions (FIG. 2 b), again in contrast to Ssp (FIG. 2 c).
The tolerance of the split inteins to C-extein sequence variation was investigated. Previously, the sensitivity of DnaE inteins was noted to changes at the +2 position in the C-extein.(6,12) Thus, all the split DnaE inteins were analyzed in the presence of a +2 glycine (CGN), glutamic acid (CEN), or arginine (CRN) in the in vivo screening assay (FIG. 1 b). Like Npu and Ssp, most of the inteins showed a dramatic decrease in activity in the presence of all three +2 mutations. Of the tested amino acids, glutamic acid was tolerated best for every intein, suggesting a conserved mechanism for accommodating a negative charge at this position. To more accurately assess the magnitude of the effect of C-extein mutations on trans-splicing, the Npu, Cra(CS505), and Cwa inteins were analyzed in vitro in the presence of a +2 glycine. All three of these reactions were characterized by rapid accumulation of thioester intermediates, which slowly resolved over tens of minutes into the spliced product and the N-extein cleavage product. Consistent with previously reported observations, these data indicate that split DnaE inteins require steric bulk at the +2 position for branched intermediate resolution and efficient splicing.(12) It is noteworthy that the Cra(CS505) and Cwa inteins showed greater C-extein promiscuity in vivo, while Ssp(PCC7002) did not tolerate any of the mutations tested. This demonstrates that subtle sequence variation between split inteins can afford differential promiscuity. Thus, this property may be further optimized through directed evolution(12) or rational design.
These data indicate that the split DnaE inteins are highly divergent in activity, despite all having evolved to catalyze trans-splicing on virtually identical substrates. Interestingly, the key catalytic residues involved in splicing are conserved across the entire family (FIG. 5). Thus, residues that affect splicing activity are non-catalytic and perhaps only moderately conserved. The measurements of relative activity can facilitate the discovery of specific sequence features that differentiate high-activity inteins from inefficient ones. Indeed, sequence homology analysis indicates that inteins with high activity are more homologous to one another than they are to the low-activity inteins. One significant outlier to this observation is the intein from Aphanothece halophytica (Aha), which despite having greater than 65% sequence identity to the high-activity inteins, was inactive with the wild-type “CFN” C-extein motif in vivo. Closer inspection of a multiple sequence alignment indicated that this intein has a non-catalytic cysteine (position 120) in place of an otherwise absolutely conserved glycine (FIG. 3 a). Furthermore, this position is close to the intein active site, where an extra nucleophile may facilitate undesirable side reactions (FIG. 3 b). Gratifyingly, mutating this cysteine to glycine reinstated high activity in the Aha intein whilst the reverse mutation destroyed the splicing activity of Npu (FIG. 3 c), validating the predictive capacity of these data.
Further analysis of the split intein sequence alignment indicated that several positions have strong amino acid conservation amongst the high-activity inteins but diverge for the low-activity inteins (FIG. 3 a, 6). These may be sites where the fast inteins have retained beneficial interactions that have been lost in slow ones. To test this idea, several positions were chosen where this sequence-activity correlation was apparent and replaced the residue in Ssp with the corresponding amino acid found in the fast inteins. Consistent with this hypothesis, several point mutations increased the activity of Ssp in vivo (FIG. 3 e). While the specific roles of these residues are not explicitly clear, especially given that they lie outside of the active site (FIG. 3 d), their locations on the intein fold (14) may provide some insights into their function. For example, at position 56, an aromatic residue is preferred in the high-activity inteins. This position is adjacent to the conserved catalytic TXXH motif (positions 69-72), and an aromatic residue may facilitate packing interactions to stabilize those residues. Similarly, a glutamate is preferred at position 122, proximal to catalytic histidine 125. The glutamate at position 89 is involved in an intimate ion cluster that was previously shown to be important for stabilizing the split intein complex.(13) Interestingly, E23 is distant from the catalytic site and has no obvious structural role. This position is conceivably important for fold stability or dynamics as has previously been observed for activating point mutations in other inteins.(15,16)
The discovery of new, fast trans-splicing inteins has broad implications for protein chemistry. Indeed, the discovery of Npu fueled a resurgence in the use of split intein-based technologies.(13,17,18) While no single intein may be ideal for every protein chemistry endeavor, the availability of several new fast-splicing split inteins can provide options to enhance the efficiency of most trans-splicing applications. For example, one common problem in working with split inteins is low expression yield or poor solubility of an intein fragment fusion to a protein of interest. Indeed, the over-expression and purification efforts here show that the Ub-IntN and IntC-SUMO fusions have markedly different yields of soluble expression, depending on the intein. Thus, a short-list of highly active split inteins with varying behavior will serve as starting point for empirical optimization of a given trans-splicing application.
Furthermore, the fragments of the different fast-splicing split inteins can be mixed as non-cognate pairs and still retain highly efficient splicing activity, further expanding the options available for any trans-splicing application. For example, the N-fragment split intein of Npu or variant thereof can bind to the C-fragment of Npu or variant thereof or any of the other split intein C-fragments or variants discussed below. Similarly, the N-fragment of Ssp or variant thereof can bind to the C-fragment of Ssp or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Aha or variant thereof can bind to the C-fragment of Aha or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Aov or variant thereof can bind to the C-fragment of Aov or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Asp or variant thereof can bind to the C-fragment of Asp or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Ava or variant thereof can bind to the C-fragment of Ava or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Cra(CS5505) or variant thereof can bind to the C-fragment of Cra(CS5505) or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Csp(CCY0110) or variant thereof can bind to the C-fragment of Csp(CCY0110) or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Csp(PCC8801) or variant thereof can bind to the C-fragment of Csp(PCC8801) or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Cwa or variant thereof can bind to the C-fragment of Cwa or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Maer(NIES843) or variant thereof can bind to the C-fragment of Maer(NIES843) or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Mcht(PCC7420) or variant thereof can bind to the C-fragment of Mcht(PCC7420) or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Oli or variant thereof can bind to the C-fragment of Oli or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Sel(PC7942) or variant thereof can bind to the C-fragment of Sel(PC7942) or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Ssp(PCC7002) or variant thereof can bind to the C-fragment of Ssp(PCC7002) or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Tel or variant thereof can bind to the C-fragment of Tel or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; the N-fragment of Ter or variant thereof can bind to the C-fragment of Ter or variant thereof or any of the other split intein C-fragments or variants thereof discussed below; and the N-fragment of Tvu or variant thereof can bind to the C-fragment of Tvu or variant thereof or any of the other split intein C-fragments or variants thereof discussed below.
The most widely used intein-based technology, expressed protein ligation, exploits cis-acting inteins to generate recombinant protein α-thioester derivatives.(2) In principle, any split intein can be artificially fused and then utilized as a cis-splicing intein in this application (1 in FIG. 4 a). Ultrafast split inteins are especially attractive in this regard due to their speed and efficiency. To test this notion, artificially fused variants of Npu, Ava, and Mcht with an N-terminal ubiquitin domain were generated. To prevent splicing, premature C-terminal cleavage or undesired high levels of competing hydrolysis residues Asn137 and Cys+1 were mutated to Ala. Upon reaction with the exogenous thiol sodium 2-mercaptoethanesulfonate (MESNa), the fused DnaE inteins were rapidly cleaved to generate the ubiquitin α-thioester, 4, in a few hours (FIG. 4 b). By contrast, MESNa thiolysis of the commonly used MxeGyrA intein was not complete even after one day under identical conditions. The fused DnaE inteins were sufficiently fast to allow for a one-pot thiolysis and native chemical ligation reaction with an N-terminal cysteine-containing fluorescent peptide, 5, to give semisynthetic protein 6 (FIG. 4 c). Furthermore, these inteins could be used to efficiently generate α-thioesters of four other structurally unique proteins domains with different C-terminal amino acid residues. These results demonstrate that fused versions of split DnaE inteins will be of general utility for protein semisynthesis.
The rapid rate of thiolysis observed for the fused DnaE inteins has mechanistic implications as well as practical ones. Without wishing to be bound by theory, one possible explanation for their enhanced reactivity over the MxeGyrA intein is that these inteins drive the N-to-S acyl shift reaction more efficiently, generating a larger population of the reactive linear thioester species 2 (FIG. 4 a). This thioester intermediate is generally thought to be transiently populated in protein splicing, and to our knowledge, it has never been directly observed.(1) Surprisingly, when analyzing the ubiquitin-DnaE intein fusions by reverse phase HPLC, two major peaks and a third minor peak were often observed, all bearing the same mass. The relative abundance of these species could be modulated by unfolding the proteins or by changes in pH, and the two major species were almost equally populated from pH 4-6 (FIG. 4 d). The major peaks most likely correspond to the precursor amide, 1, and the linear thioester, 2, and the minor peak as the tetrahedral oxythiazolidine intermediate. Importantly, only a single HPLC peak was seen for the ubiquitin-MxeGyrA fusion under identical conditions. These observations, along with the enhanced thiolysis rates, strongly support the notion that these DnaE inteins have a hyper-activated N-terminal splice junction.
Splicing activities in an entire family of split inteins has been characterized. Ultrafast protein trans-splicing may be the norm, rather than the exception, in this family. Furthermore, different split inteins have varying degrees of tolerance for C-extein mutations, suggesting that traceless protein splicing may be attainable by modestly engineering any highly active intein. A thorough comparison of the activities of a small family of homologous proteins can be used to identify important non-catalytic positions that modulate activity. Finally, by artificially fusing split DnaE intein fragments, new constructs have been provided for the efficient synthesis of protein α-thioesters used in expressed protein ligation. These results will guide the development of improved protein chemistry technologies and should lay the groundwork towards a more fundamental understanding of efficient protein splicing.

Fusion Proteins of Split Intein N-Fragment

Disclosed herein are fusion proteins of a polypeptide and a split intein N-fragment. As used herein, the term “polypeptide” refers to any amino acid based polymer, interchangeable referred to as a “protein” throughout, and can include glycoproteins and lipoproteins. In some cases, the polypeptide is a polypeptide excreted from a cell (e.g., a mammalian cell). In various cases, the polypeptide is an antibody or a fragment thereof. The polypeptide can be any naturally occurring or synthetic polypeptide of interest, including polypeptides having one or more amino acid residues other than the 20 naturally occurring amino acids.
In some cases, the polypeptide has a molecular weight of 45 kDa or greater, 50 kDa or greater, 60 kDa or greater, 75 kDa or greater, 100 kDa or greater, 120 kDa or greater, or 150 kDa or greater. The polypeptide can be, e.g., an antibody or a fragment thereof. In cases of antibodies, the split intein N-fragment can be fused to one or both of the heavy chains, and/or to one or both of the light chains. In some cases, the polypeptide is a protein secreted from a cell, e.g., a mammalian cell.
The split intein N-fragment comprises a sequence as shown in FIG. 5, e.g., Npu (SEQ ID NO: 1), Ssp (SEQ ID NO: 2), Aha (SEQ ID NO: 3), Aov (SEQ ID NO: 4), Asp (SEQ ID NO: 5), Ava (SEQ ID NO: 6), Cra(CS505) (SEQ ID NO: 7), Csp(CCY0110) (SEQ ID NO: 8), Csp(PCC8801) (SEQ ID NO: 9), Cwa (SEQ ID NO: 10), Maer (NIES843) (SEQ ID NO: 11), Mcht(PCC7420) (SEQ ID NO: 12), Oli (SEQ ID NO: 13), Sel(PC7942) (SEQ ID NO: 14), Ssp(PCC7002) (SEQ ID NO: 15), Tel (SEQ ID NO: 16), Ter (SEQ ID NO: 17), Tvu (SEQ ID NO: 18), or a variant thereof. In some cases, the spilt intein N-fragment sequence comprises a sequence other than Npu (SEQ ID NO: 1) or Ssp (SEQ ID NO: 2), and in other cases, comprises a sequence other than Npu (SEQ ID NO: 1), Ssp (SEQ ID NO: 2), or Aha (SEQ ID NO: 3). In some specific cases, the split intein N-fragment sequence comprises a sequence of Ava (SEQ ID NO: 6), Cra (SEQ ID NO: 7), Csp(PCC8801) (SEQ ID NO: 9), Cwa (SEQ ID NO: 10), Mcht (PCC7420) (SEQ ID NO: 12), Oli (SEQ ID NO: 13), Ter (SEQ ID NO: 17) and Tvu (SEQ ID NO: 18). In some cases, the split intein N-fragment has a sequence comprising a consensus sequence of SEQ ID NO: 19: (CLSYDTEILTVEYGAVPIGKIVEENIECTVYSVDENGFVYTQPIAQWHDRGEQEVFE YCLEDGSTIRATKDHKFMTEDGEMLPIDEIFEQGLDLKQVKGLPD).
As used herein, a variant of a split intein N-fragment is a mutated split intein N-fragment as disclosed herein that maintains the activity of the split intein N-fragment (e.g., its ability to bind to a split intein C-fragment and/or catalyze nucleophilic attack of the polypeptide fused to it). Contemplated variants of a split intein N-fragments disclosed herein include mutation of one or more C residues, except for Cysl, to an aliphatic residue, such as an A, I, L, or F, or to a S residue. One such variant contemplated is a mutant Npu with Cys28 and Cys59 mutated to Ser, SEQ ID NO: 20 (CLSYETEILTVEYGLLPIGKIVEKRIESTVYSVDNNGNIYTQPVAQWHDRGEQEVFEY SLEDGSLIRATKDHKFMTVDGQMLPIDEIFERELDLMRVDNLPN).

Mutated Split Intein C-Fragments and Binding to a Support

The split intein C-fragments disclosed herein are mutated from the naturally occurring sequences to mutate the N137 and C+1 residues to a residue other than Asn or Gln for N137 and a residue other than Cys for C+1 (SEQ ID NOs: 129-146). In some cases, the mutations at these two positions are to a hydrophobic residue, e.g., not containing a free SH thiol (Cys), a carboxylic acid (Asp, Glu), or a base (Arg, His, Lys) or other undesired group (e.g., Asn, Gln) on the side chain. In various cases, the two mutated aliphatic residues can be the same or different and can be A, V, I, S, M, H, L, F, Y, G, or W or can be a unnatural (e.g., not encoded by genetic code) aliphatic amino acid residue such as norleucine, 2-aminobutyric acid, nor-valine, 2-aminopentoic acid, or 2-aminohexaanoic acid (SEQ ID NOs: 219-236). Specifically contemplated are mutations where both residues are selected from A, I, V, L, Y, G and F (SEQ ID NOs: 309-326). In various cases, at least one of the two mutated residues is A.
Thus, provided herein is a mutated split intein C-fragment comprising a mutation at N137 and Cys+1 of Npu (SEQ ID NOs: 129, 147, 165, 183, 201, 219, 237, 255, 273, 291, 309, 327, 345, 363, 381 and 399), Ssp (SEQ ID NOs: 130, 148, 166, 184, 202, 220, 238, 256, 274, 292, 310, 328, 346, 364, 382 and 400), Aha (SEQ ID NOs: 131, 149, 167, 185, 203, 221, 239, 257, 275, 293, 311, 329, 347, 365, 383 and 401), Aov (SEQ ID NOs: 132, 150, 168, 186, 204, 222, 240, 258, 276, 294, 312, 330, 348, 366, 384 and 402), Asp (SEQ ID NOs: 133, 151, 169, 187, 205, 223, 241, 259, 277, 295, 313, 331, 349, 367, 385 and 403), Ava (SEQ ID NOs: 134, 152, 170, 188, 206, 224, 242, 260, 278, 296, 314, 332, 350, 368, 386 and 404), Cra(CS505) (SEQ ID NOs: 135, 153, 171, 189, 207, 225, 243, 261, 279, 297, 315, 333, 351, 369, 387 and 405), Csp (CCY0110) (SEQ ID NOs: 136, 154, 172, 190, 208, 226, 244, 262, 280, 298, 316, 334, 352, 370, 388 and 406), Csp(PCC8801) (SEQ ID NOs: 137, 155, 173, 191, 209, 227, 245, 263, 281, 299, 317, 335, 353, 371, 389 and 407), Cwa (SEQ ID NOs: 138, 156, 174, 192, 210, 228, 246, 264, 282, 300, 318, 336, 354, 372, 390 and 408), Maer(NIES843) (SEQ ID NOs: 139, 157, 175, 193, 211, 229, 247, 265, 283, 301, 319, 337, 355, 373, 391 and 409), Mcht(PCC7420) (SEQ ID NOs: 140, 158, 176, 194, 212, 230, 248, 266, 284, 302, 320, 338, 356, 374, 392 and 410), Oli (SEQ ID NOs: 141, 159, 177, 195, 213, 231, 249, 267, 285, 303, 321, 339, 357, 375, 393 and 411), Sel(PC7942) (SEQ ID NOs: 142, 160, 178, 196, 214, 232, 250, 268, 286, 304, 322, 340, 358, 376, 394 and 412), Ssp(PCC7002) (SEQ ID NOs: 143, 161, 179, 197, 215, 233, 251, 269, 287, 305, 323, 341, 359, 377, 395 and 413), Tel (SEQ ID NOs: 144, 162, 180, 198, 216, 234, 252, 270, 288, 306, 324, 342, 360, 378, 396 and 414), Ter (SEQ ID NOs: 145, 163, 181, 199, 217, 235, 253, 271, 289, 307, 325, 343, 361, 379, 397 and 415), or Tvu (SEQ ID NOs: 146, 164, 182, 200, 218, 236, 254, 272, 290, 308, 326, 344, 362, 380, 398 and 416), where the mutation at N137 and Cys+1 is a naturally occurring or unnatural hydrophobic residue. In some specific cases, at least one of the mutations is A (SEQ ID NOs: 183-200, 255-272, 327-344, and 345-416) and in more specific cases, both mutations are A (SEQ IS NOs: 399-416).
A variety of supports can be used. Generally, the solid support is a polymer or substance that allows for linkage of the split intein C-fragment, optionally via a linker. The linker can be further amino acid residues engineered to the C-terminus of the split intein C-fragment or can be other known linkers for attachment of a peptide to a support. One contemplated linker is a small peptide -SGGC (SEQ ID NO: 705), where the thiol of the C-terminal Cys can be used to attach the split intein C-fragment to the support. Thus, specifically contemplated are mutated split intein C-fragments of the Npu, Ssp, etc. sequences noted above having a -SGGC peptide linker (SEQ ID NO: 705) (e.g., specifying the residues starting at the N137 position: AAFN-SGGC) (SEQ ID NO:706). The length of a pepide linker can be modified to provide varying lengths and flexibility in any individual sitation (e.g., more than 2 Gly residues). It will also be apparent that the C-terminus residue of a peptide linker can be modified to introduce an appropriately reactive functional group to attach the split intein C-fragment to a surface of choice (e.g., Lys to react via an amine, Cys to react via a thiol, or Asp or Glu to react via a carboxylic acid). Other, unnatural amino acid residues are also contemplated for use in a peptide linker to provide other functional group moieties to allow for different attachment chemistry of the C-fragment to a support of interest (e.g., azide, alkynes, carbonyls, amino-oxy, cyano-benzothiazoles, tetrazoles, alkenes, alkyl-halides). The linker can alternatively be a polymeric linker.
Based upon an analysis of the sequences of the highly active split intein C-fragments investigated, a consensus sequence for the split intein C-fragment is derived: SEQ ID NO: 707 (VKIISRQSLGKQNVYDIGVEKDHNFLLANGLIASN), as well as a mutated version where the N137 is mutated to other than Asn or Gln (SEQ ID NO:708), or more specifically, N137 is mutant to a naturally occurring or unnaturally occurring hydrophobic residue, such as A, V, I, M, H, L, F, Y, G, S, H, or W or can be a unnatural (e.g., not encoded by genetic code) aliphatic amino acid residue such as norleucine, 2-aminobutyric acid, nor-valine, 2-aminopentanoic acid, or 2-aminohexanoic acid (SEQ ID NO: 709). Specifically contemplated mutations at N137 for the consensus sequences include A, I, V, L, Y, G, and F (SEQ ID NO: 710). Also contemplated is where N137 is mutated to A (SEQ ID NO: 711).
Further contemplated are variants of the consensus sequence having a residue at the +1 position other than Cys (SEQ ID NOs: 712, 716, 720, and 724). More specifically the +1 position can be a naturally occurring or unnaturally occurring hydrophobic residue such as A, V, I, M, H, L, F, Y, G, S, H, or W or can be a unnatural (e.g., not encoded by genetic code) aliphatic amino acid residue such as norleucine, 2-aminobutyric acid, nor-valine, 2-aminopentanoic acid, or 2-aminohexanoic acid (SEQ ID NOs: 713, 717, 721, and 725). Specifically contemplated are mutation at +1 position is selected from A, I, V, L, Y, G, and F (SEQ ID NOs: 714, 718, 722, and 726). In various cases, at least one of the mutated residues of the consensus sequence is A (SEQ ID NOs: 715, 719, and 723). In some cases, the consensus C-fragment sequence has both mutations as Ala (SEQ ID NO:727). Further contemplated is a consensus sequence comprising FN at the +2 and +3 positions (SEQ ID NO:728-743. Also contemplated is a consensus sequence comprising a peptide linker for attachment to a solid support, and one embodiment is -SGGC at positions +4-+7 (SEQ ID NO: 744-759).
The split intein C-fragment or variant thereof as disclosed herein can be attached to a solid support via a linker. In various cases, the linker is a polymer, including but not limited to a water-soluble polymer, a nucleic acid, a polypeptide, an oligosaccharide, a carbohydrate, a lipid, or combinations thereof. It is not critical what the linker's chemical structure is, since it serves primarily as a linker. The linker should be chosen so as not to interfere with the activity of the C-fragment. The linker can be made up of amino acids linked together by peptide bonds. Thus, in some embodiments, the linker comprises Y_n, wherein Y is a naturally occurring amino acid or a steroisomer thereof and “n” is any one of 1 through 20. The linker is therefore can be made up of from 1 to 20 amino acids linked by peptide bonds, wherein the amino acids are selected from the 20 naturally-occurring amino acids. In some cases, the 1 to 20 amino acids are selected from Gly, Ala, Ser, Cys. In some cases, the linker is made up of a majority of amino acids that are sterically un-hindered, such as Gly.
Non-peptide linkers are also possible. For example, alkyl linkers such as —HN—(CH₂), —CO—, wherein s=2-20 can be used. These alkyl linkers may further be substituted by any non-sterically hindering group such as lower alkyl (e.g., C₁-C₆), halogen (e.g., Cl, Br), CN, NH₂, phenyl, etc.
Another type of non-peptide linker is a polyethylene glycol group, such as: —HN—(CH₂)₂—(O-CH₂—CH₂)_n—O—CH₂—CO, wherein n is such that the overall molecular weight of the linker ranges from approximately 101 to 5000, preferably 101 to 500.
In some cases, the linker has a length of about 0-14 sub-units (e.g., amino acids).
In instances wherein the linker is a polynucleotide, the length of the linker in various embodiments is at least about 10 nucleotides, 10-30 nucleotides, or even greater than 30 nucleotides. In various aspects, the bases of the polynucleotide linker are all adenines, all thymines, all cytidines, all guanines, all uracils, or all some other modified base.
In another embodiment, a non-nucleotide linker of the invention comprises a basic nucleotide, polyether, polyamine, polyamide, peptide, carbohydrate, lipid, polyhydrocarbon, or other polymeric compounds. Specific examples include those described by Seela and Kaiser, Nucleic Acids Res. 1990, 18:6353 and Nucleic Acids Res. 1987, 15:3113; Cload and Schepartz, J. Am. Chem. Soc. 1991, 113:6324; Richardson and Schepartz, J. Am. Chem. Soc. 1991, 113:5109; Ma et al., Nucleic Acids Res. 1993, 21:2585 and Biochemistry 1993, 32:1751; Durand et al., Nucleic Acids Res. 1990, 18:6353; McCurdy et al., Nucleosides & Nucleotides 1991, 10:287; Jschke et al., Tetrahedron Lett. 1993, 34:301; Ono et al., Biochemistry 1991, 30:9914; Arnold et al., International Publication No. WO 89/02439; Usman et al., International Publication No. WO 95/06731; Dudycz et al., International Publication No. WO 95/11910 and Ferentz and Verdine, J. Am. Chem. Soc. 1991, 113:4000, the disclosures of which are all incorporated by reference herein. A “non-nucleotide” further means any group or compound that can be incorporated into a nucleic acid chain in the place of one or more nucleotide units, including either sugar and/or phosphate substitutions, and allows the remaining bases to exhibit their enzymatic activity. The group or compound can be abasic in that it does not contain a commonly recognized nucleotide base, such as adenosine, guanine, cytosine, uracil or thymine, for example at the C1 position of the sugar.
In various aspects, linkers contemplated include linear polymers (e.g., polyethylene glycol, polylysine, dextran, etc.), branched chain polymers (see, for example, U.S. Pat. No. 4,289,872 to Denkenwalter et al., issued Sep. 15, 1981; U.S. Pat. No. 5,229,490 to Tam, issued Jul. 20, 1993; WO 93/21259 by Frechet et al., published 28 Oct. 1993); lipids; cholesterol groups (such as a steroid); or carbohydrates or oligosaccharides. Other linkers include one or more water soluble polymer attachments such as polyoxyethylene glycol, or polypropylene glycol as described U.S. Pat. Nos. 4,640,835, 4,496,689, 4,301,144, 4,670,417, 4,791,192 and 4,179,337. Other useful polymers as linkers known in the art include monomethoxy-polyethylene glycol, dextran, cellulose, or other carbohydrate based polymers, poly-(N-vinyl pyrrolidone) polyethylene glycol, propylene glycol homopolymers, a polypropylene oxide/ethylene oxide co-polymer, polyoxyethylated polyols (e.g., glycerol) and polyvinyl alcohol, as well as mixtures of these polymers.
In still other aspects, oligonucleotide such as poly-A or hydrophilic or amphiphilic polymers are contemplated as linkers, including, for example, amphiphiles (including oligonucletoides).
Contemplated solid supports include resins, particles, and beads. More specific solid supports include polyhydroxy polymers, e.g. based on polysaccharides, such as agarose, dextran, cellulose, starch, pullulan, or the like, and synthetic polymers, such as polyacrylic amide, polymethacrylic amide, poly(hydroxyalkylvinyl ethers), poly(hydroxyalkylacrylates) and polymethacrylates (e.g. polyglycidylmethacrylate), polyvinyl alcohols and polymers based on styrenes and divinylbenzenes, and copolymers in which two or more of the monomers corresponding to the above-mentioned polymers are included. Specific solid supports contemplated include agarose, sepharose, cellulose, polystyrene, polyethylene glycol, derivatized agarose, acrylamide, sephadex, sepharose, polyethyleneglycol (PEG)-acrylamide, and polystyrene-PEG based supports. In some cases, the solid support can be a resin such as p-methylbenzhydrylamine (pMBHA) resin (Peptides International, Louisville, Ky.), polystyrenes (e.g., PAM-resin obtained from Bachem Inc., Peninsula Laboratories, etc.), including chloromethylpolystyrene, hydroxymethylpolystyrene and aminomethylpolystyrene, poly (dimethylacrylamide)-grafted styrene co-divinyl-benzene (e.g., POLYHIPE resin, obtained from Aminotech, Canada), polyamide resin (obtained from Peninsula Laboratories), polystyrene resin grafted with polyethylene glycol (e.g., TENTAGEL or ARGOGEL, Bayer, Tubingen, Germany) polydimethylacrylamide resin (obtained from Milligen/Biosearch, California), or Sepharose (Pharmacia, Sweden). In various embodiments, the solid support can be a magnetic bead, a glass slide, a glass bead, or a metal or inorganic particle (e.g., gold, silica, iron, or mixture thereof).

Methods of Purifying and Modifying Polypeptides

Site-specific modification of proteins is an invaluable tool to study the molecular details of protein function (19). Moreover, its potential for the discovery and development of protein therapeutics has also been recently acknowledged (20). Several methods have been developed over the years to generate site-specifically modified proteins; one of the most widely used is Expressed Protein Ligation (EPL), which has been applied to many different proteins in a variety of studies to address fundamental questions of protein function. EPL was first described in 1998 (21), as an expansion to recombinant proteins of Native Chemical Ligation (NCL) (19,22), and it consists on the reaction between a C-terminal recombinant protein α-thioester with a synthetic peptide containing a Cys at its N-terminus through the formation of a new native peptide bond between the two fragments. The synthetic nature of the Cys containing peptide allows for the incorporation of almost any chemical modification into the protein of interest.
In order to apply EPL to any given protein the generation of protein C-terminal thioesters in good yields and high purity is an absolute requirement. A family of single turn-over enzymes, named inteins, has been used since the dawn of EPL for the generation of such thioesters. Inteins are able to catalyze protein splicing, a naturally occurring post-translational modification by which they excise themselves from the polypeptide in which they are embedded, concomitantly forming a new peptide bond between their flanking protein regions (23). Importantly, this reaction occurs via several protein α-thioesters, which can be trapped through a trans-thioesterification reaction with an exogenous thiol.
Inteins, such as GyrA or VMA, have been successfully harnessed to prepare a wide variety of protein thioesters. In order to isolate the desired protein thioesters inteins are usually fused to affinity tags such as the chitin bidning domain or the hexa-His tag. However, despite the notable success of this strategy, the reaction conditions required for efficient thiolysis (reducing agents, large concentration of thiols and long incubation times) affect the performance of such tags and subsequent additional purification steps are often required to obtain the desired pure product for ligation (24-26). Moreover, depending on the identity of the C-terminal residue of the protein of interest, significant levels of in vivo premature cleavage can occur, significantly reducing the final product yield.
An ideal system should combine the thioester formation capabilities of inteins with a built-in affinity purification strategy (fully compatible with the thiolysis reaction conditions) and reduced risk of premature cleavage. Naturally split-inteins were investigated, which can perform a reaction analogous to protein splicing but in which the intein itself is split into two different polypeptides. Each of the two intein fragments are completely inactive by themselves but have a strong affinity for each other and, upon binding, they adopt their splicing competent active conformation and are able to carry out protein trans-splicing. Recently, an artificially split version of the DnaB intein has been reported for the purification of unmodified proteins (27). Thus, a purification strategy is provided using naturally split-inteins instead and to harness them for the one pot purification and generation of recombinant protein α-thioesters (FIG. 7) directly from cell lysates.
Due to the extremely fast reaction kinetics of naturally split inteins, several mutations were introduced to allow efficient thioester formation and minimize in vivo and in vitro undesired cleavage reactions. Specifically, both the C-intein C-terminal catalytic Asn137 and the Cys+1 residues had to be mutated to Ala to prevent premature C- and N-terminal cleavage, respectively. Mutation to two sequential aliphatic residues, natural or unnatural, is also expected to yield comparable results as the AA mutation. Other mutated split intein C-fragments as described above can be used in the described purification and/or modification methods, and are specifically contemplated.
To develop a split-intein based purification and thioester formation strategy the Npu split-intein was chosen, which is one of the fastest DnaE split-inteins previously known [10]. Initially the ability of split Npu to generate protein thioesters in solution was tested by mixing the model protein ubiquitin fused to NpuN with a mutant NpuC (Asn137 and Cys+1 to Ala) in the presence (and absence) of the thiol MESNa. SDSPAGE, HPLC and MS analysis of the reactions showed the formation of the desired ubiqutin C-terminal α-thioester in a few hours. Encouraged by these results an affinity purification strategy was designed based on the covalent immobilization of the NpuC intein mutant onto a solid support. The immobilized mutated NpuC could then be used to purify NpuN tagged proteins from complex mixtures and addition of an exogenous thiol would cleave off the desired protein α-thioester, which would elute from the column in a highly purified form. Other split intein N-fragments as described above can be used in the methods disclosed herein, and are specifically contemplated.
An NpuC mutant (Asn¹³⁷and Cys⁺¹mutated to Ala, NpuC-AA (SEQ. ID NO: 777) was prepared with a Cys residue at the C-terminus of its C-extein, which was used to immobilize the peptide onto a iodoacetyl resin. With the NpuC-AA affinity resin in hand, it was shown that several protein C-terminal α-thioesters (Ubiquitin, MBP, PHPT1) could be easily produced and purified out of cell lysates (FIG. 8). HPLC and MS analysis confirmed the formation of the desired protein thioesters with very low levels of undesired hydrolysis (FIG. 9). Recovery yields varied between 75 and 95% and the NpuC-AA resin had a consistent loading capacity of 3-6 mg of protein per mL. The utility of the α-thioester derivatives of Ub, MBP, and PHPT1 obtained from the column was demonstrated by ligating each of them to an N-terminal Cys-containing fluorescent peptide (CGK(F1)) to give the corresponding semisynthetic products in excellent yield (FIG. 13). Importantly, one-pot thiolysis/ligation reactions can be carried out, which provides a site-specifically modified protein directly from cell lysates without isolating the intermediate thioester (FIG. 14).
A concern when working with split-inteins (and also inteins) is the effect of the flanking amino acid sequences on splicing and/or thiolysis activity. Although the N-terminal junction is regarded as more tolerant towards deviations from the native N-extein residues it was important to evaluate the effect that the C-terminal amino acid of the protein of interest (−1 residue according to intein numbering conventions) would have on the yields of thioester formation. A complete library of Ub-X-NpuN fusion proteins was constructed where the C-terminal Ub residue (X) was varied from its native Gly to all other 19 proteinogenic amino acids. Proteins were expressed in E. coli and cell lysates, applied to the NpuC-AA affinity resin and purified. Protein yields were estimated from the SDSPAGE analysis for each purification and hydrolysis levels from RP-HPLC and MS analysis of the elution fractions (FIG. 10). Results show similar trends to those known for non-split inteins, such as GyrA (29), and most amino acids display high yields of cleavage after overnight incubation with MESNa, the exceptions being Pro and Glu, for which recovery were 49 and 50%, respectively. As expected, the Asn α-thioester could not be isolated due to the well-known reaction of its side-chain with the adjacent α-thioester to form a succinimide.
This purification strategy was very successful for the purification of several soluble proteins under native conditions. However, protein fragments required for EPL sometimes suffer from poor solubility and high toxicity and tend to accumulate in cellular inclusion bodies during expression. The Npu split-intein has been shown to retain a significant level of activity in the presence of denaturants (28), which suggested that this strategy would be compatible under such conditions. Using the model Ub-NpuN protein fusion it was confirmed that both, binding to the NpuC-AA-resin and thioester formation, worked well in the presence of 2 and 4 M urea. Similar levels of both, binding and thiolysis, were obtained than in the absence of denaturant and same reaction conditions.
The system was next tested for the purification from inclusion bodies of a fragment of the histone H2B. Preparation of site-specifically modified histones using EPL is a topic of major interest due to their crucial role in the understanding of epigenetic regulation. However, histone fragments are remarkably poorly behaved and the preparation of their recombinant C-terminal α-thioesters particularly challenging. A H2B(1-116) fragment fused to NpuN was expressed in E. coli and the inclusion bodies extracted with 6 M urea. The H2B-NpuN fusion was subsequently diluted into a 3 M urea buffer and the corresponding C-terminal α-thioester generated concomitant with purification over the NpuC-AA affinity resin (FIG. 11). Due to its tendency of aggregation, very dilute protein solutions bound more efficiently to the resin, and also longer reaction times were required for efficient thioester generation, obviously, these are parameters that would need to be optimized on a protein to protein basis. Using these conditions, H2B(1-116)C-terminal thioester was obtained in excellent purity (>90% by RP-HPLC) and isolated yield (˜20 mg per L of culture). This represents a significant improvement over previous protocols which afford less protein (4 mg per L of culture) and require the use of multiple chromatographic purification steps including RP-HPLC. Importantly, the H2B(1-116)-MES thioester obtained from the IntC-column can be directly used in EPL reactions without further purification. Accordingly, the protein was successfully ligated to a synthetic H2B(117-125) peptide containing an acetylated Lys at position 120 to yield semi-synthetic H2B-K120Ac (FIG. 15).
The potential of this thioester formation purification strategy was demonstrated by applying it to the site-specific modification of a monoclonal antibody. Thus, specifically contemplated is a method of purifying an antibody by using a fusion protein of an antibody and a split intein N-fragment as disclosed herein and a mutated split intein C-fragment as disclosed herein.
The modification of antibodies is a field of intense research, specially focused on the development of therapeutic antibody-drug-conjugates (30). The identity of the N-intein could have a significant effect on the expression levels of its fusion to a given protein of interest. The N-fragment of several of the fastest split DnaE inteins cross-reacted with NpuC, allowing one to use any of them with the same NpuC based affinity column. Accordingly, the expression levels of a model antibody (αDEC, antibody against the DEC205 receptor) were tested and found the highest levels of expression were obtained when αDEC was fused to the AvaN intein (FIG. 12A). αDEC-AvaN fusions were transfected into 293T cells and after 4 days of culture the supernatants were collected and purified over the NpuC-AA-column. The presence of a C-terminal thioester in the purified αDEC was confirmed by reacting it with a short fluorescent peptide with an N-terminal Cys residue (FIGS. 12B and C). MS of the deglycosylated and reduced αDEC-fluorophore conjugate was used to confirm its identity and SEC-MALS demonstrate the product was monodisperse and of the expected size for an IgG antibody.
Split-inteins can be engineered for the preparation of protein α-thioesters and that the strong affinity between the two split-intein fragments provides a powerful handle for their purification. The generality of the approach is demonstrated by using it to generate highly pure thioesters of both soluble (ubiqutin, MBP, PHPT) and insoluble proteins (H2B fragment) as well as monoclonal antibodies (αDEC). Moreover, several N-inteins can be tested for optimal expression levels of the protein of interest and used with one single NpuC-column.
Thus, the split inteins disclosed herein can be used to purify and modify a polypeptide of interest. A polypeptide of interest is provided in a fusion protein with a split intein N-fragment, e.g., via well-known recombinant protein methods. The fusion protein is then contacted with a corresponding split intein C-fragment under conditions that allow binding of the N-fragment and C-fragment to form an intein intermediate. The split intein C-fragment can be bound to a support (e.g., a solid support such as a resin) or can subsequently (e.g., after binding to the split intein N-fragment to form the intein intermediate) be bound to a support. This allows for the removal via washing of components that were in the mixture due to the recombinant protein synthesis, allowing the fusion protein to be isolated from the other components. Washes can include detergents, denaturing agents and salt solutions (e.g., NaCl).
Then, the intein intermediate can be reacted with a nucleophile to release the polypeptide of interest from the bound N- and C-fragment inteins wherein the C-terminus of the polypeptide is modified by the nucleophile added. The nucleophile can be a thiol to directed yield the polypeptide as an α-thioester, which in turn can be further modified, e.g., with a different nucleophile (e.g., a drug, a polymer, another polypeptide, a oligonucleotide), or any other moiety using the well-known α-thioester chemistry for protein modification at the C-terminus. One advantage of this chemistry is that only the C-terminus is modified with a thioester for further modification, thus allowing for selective modification only at the C-terminus and not at any other acidic residue in the polypeptide.
The nucleophile that is used in the methods disclosed herein either with the intein intermediate or as a subsequent nucleophile reacting with, e.g., a α-thioester, can be any compound or material having a suitable nucleophilic moiety. For example, to form a α-thioester, a thiol moiety is contemplated as the nucleophile. In some cases, the thiol is a 1,2-aminothiol, or a 1,2-aminoselenol. An α-selenothioester can be formed by using a selenothiol (R-SeH). Alternative nucleophiles contemplated include amines (i.e. aminolysis to give amides directly), hydrazines (to give hydrazides), amino-oxy groups (to give hydroxamic acids). Additionally, the nucleophile can be a functional group within a compound of interest for conjugation to the polypeptide of interest (e.g., a drug to form a protein-drug conjugate) or could alternatively bear an additional functional group for subsequent known bioorthogonal reactions such as an azide or an alkyne (for a click chemistry reaction between the two function groups to form a triazole), a tetrazole, an α-ketoacid, an aldehyde or ketone, or a cyanobenzothiazole.
Additional aspects and details of the invention will be apparent from the following examples, which are intended to be illustrative rather than limiting.

Examples

Materials

All buffering salts, isopropyl-β-D-thiogalactopyranoside (IPTG), and N,N-diisopropylethylamine (DIPEA) were purchased from Fisher Scientific (Pittsburgh, Pa.). Kanamycin sulfate (Kan), β-Mercaptoethanol (BME), DL-dithiothreitol (DTT), sodium 2-mercaptoethanesulfonate (MESNa), ethanedithiol (EDT), Coomassie brilliant blue, N,N-dimethylformamide (DMF), Tetrakis(triphenylphosphine)palladium(0) (Pd(PPh₃)₄), phenylsilane, triisopropylsilane (TIS), sodium diethyldithiocarbamate trihydrate, and 5(6)-carboxyfluorescein were purchased from Sigma-Aldrich (St. Louis, Mo.). Tris(2-carboxyethyl)phosphine hydrochloride (TCEP) was purchased from Thermo Scientific (Rockford, Ill.). Fmoc-Gly-OH, Fmoc-Lys(Alloc)-OH, and Boc-Cys(Trt)-OH were purchasd from Novabiochem (Laufelfingen, Switzerland). Piperidine was purchased from Alfa Aesar (Ward Hill, Mass.). Dichloromethane (DCM) and rink amide resin were purchased from EMD Chemicals (Billerica, Mass.). 1-Hydroxybenzotriazole hydrate (HOBt) was purchased from AnaSpec (Fremont, Ca). Trifluoroacetic acid (TFA) was purchased from Halocarbon (North Augusta, S.C.). Complete protease inhibitor tablets were purchased from Roche Diagnostics (Mannheim, Germany). Nickel-nitrilotriacetic acid (Ni-NTA) resin was from Novagen (Gibbstown, N.J.). The QuikChange XL II site directed mutagenesis kit was from Agilent (La Jolla, Calif.). DpnI and the Phusion High-Fidelity PCR kit were from New England Biolabs (Ipswich, Mass.). DNA purification kits (QIAprep spin minikit, QIAquick gel extraction kit, QIAquick PCR purification kit) were from Qiagen (Valencia, Calif.). Sub-cloning efficiency DH5 competent cells and One Shot BL21(DE3) chemically competent E. coli were purchased from Invitrogen (Carlsbad, Calif.) and used to generate “in-house” high-competency cell lines. Oligonucleotides were purchased from Integrated DNA Technologies (Coralville, Iowa). The new intein genes were generated synthetically and purchased from GENEWIZ (South Plainfield, N.J.). All plasmids used in this study were sequenced by GENEWIZ.
Criterion XT Bis-Tris gels (12%), Immun-blot PVDF membrane (0.2 μm), and Bradford reagent dye concentrate were purchased from Bio-Rad (Hercules, Calif.). 20×MES-SDS running buffer was purchased from Boston Bioproducts (Ashland, Mass.). Mouse anti-myc monoclonal antibody (α-myc) was purchased from Invitrogen (Carlsbad, Calif.). Anti-His Tag, clone HIS.H8 mouse monoclonal antibody (α-His6) was purchased from Millipore (Billerica, Mass.). Mouse HA.11 monoclonal antibody (α-HA) was purchased from Covance (Princeton, N.J.). IRDye 800CW goat anti-Mouse IgG secondary antibody (Licor mouse 800) and Licor Blocking Buffer were purchased from LI-COR Biotechnology (Lincoln, Nebr.).

Equipment

Size-exclusion chromatography was carried out on an ÄKTA FPLC system from GE Healthcare. Both preparative and analytical FPLC were carried out on a Superdex 75 10/300 or S200 10/300 column. For all runs, proteins were eluted over 1.35 column volumes of buffer (flow rate: 0.5 mL/min). Analytical RP-HPLC was performed on Hewlett-Packard 1100 and 1200 series instruments equipped with a C18 Vydac column (5 μm, 4.6×150 mm) at a flow rate of 1 mL/min. Preparative RP-HPLC was performed on a Waters prep LC system comprised of a Waters 2545 Binary Gradient Module and a Waters 2489 UV detector. Purifications were carried out on a C18 Vydac 218TP1022 column (10 μM; 22×250 mm). All runs used 0.1% TFA (trifluoroacetic acid) in water (solvent A) and 90% acetonitrile in water with 0.1% TFA (solvent B). For all runs, a two minute isocratic period in initial conditions was followed by a 30 minute linear gradient with increasing buffer B concentration. Electrospray ionization mass spectrometric analysis (ESI-MS) was performed on a Bruker Daltonics MicrOTOF-Q II mass spectrometer. In vivo intein activity assays were carried out on a VersaMax tunable microplate reader from Molecular Devices. Cells were lysed using an S-450D Branson Digital Sonifier. Western blots and coomassie-stained in vitro splicing assay gels were imaged on a LI-COR Odyssey Infrared Imager. Fluorescent fluorescein-containing gels were imaged using the GE ImageQuant LAS 4000 imager.

Compilation of the DnaE Sequence Library and Sequence Analysis

Protein sequences of the split DnaE inteins were obtained from the NEB InBase1. This list consisted of 23 entries as of May 2011. Of these entries, two were discarded from the study as they did not have a C-intein sequence: Csp(PCC7822) and Nosp(CCY9414). Two pairs of inteins had identical sequences: Nsp(PCC7120) with Asp (these are most likely the same organism with two different names) and Sel(PCC6301) with Sel(PC7942). Thus, Nsp(PCC7120) and Sel(PCC6301) were removed from the library. The Mcht(PCC7420) and Oli C-intein sequences were identical, but both inteins were kept in the library as their N-intein sequences were different. In the InBase, the Aov intein had an “X” at position 87 in place of an absolutely conserved isoleucine (I), so 187 was utilized at this position. The plasmid for the kanamycin resistance assays bearing the Csp(PCC7424) intein proved to be unstable and yielded highly variable results; thus this intein was excluded from the analyses. The final library contained 18 inteins, Table 1.

TABLE 1

DnaE Intein
Name	Genus	Species	Strain

Npu	Nostoc	punctiforme	PCC73102
Ssp	Synechocystis	species	PCC6803
Aha	Aphanothece	halophytica
Aov	Aphanizomenon	ovalisporum
Asp	Anabaena	species	PCC7120
Ava	Anabaena	variabilis	ATCC29413
Cra(CS505)	Cylindrospermopsis	raciborskii	CS-505
Csp(CCY0110)	Cyanothece	species	CCY0110
Csp(PCC8801)	Cyanothece	species	PCC8801
Cwa	Crocosphaera	watsonii	WH 8501
Maer(NIES843)	Microcystis	aeruginosa	NIES-843
Mcht(PCC7420)-2	Microcoleus	chthonoplastes	PCC7420
Oli	Oscillatoria	limnetica	Solar Lake
Sel(PC7942)	Synechococcus	elongatus	PC7942
Ssp(PCC7002)	Synechococcus	species	PCC7002
Tel	Thermosynechococcus	elongatus	BP-1
Ter-3	Trichodesmium	erythraeum	IMS101
Tvu	Thermosynechococcus	vulcanus

Given the high homology of DnaE intein sequences, the N- and C-inteins were manually aligned using the multiple alignment software Jalview2. All N-intein sequences were “left-justified” to align the first cysteine residue, and the variable N-intein tail region was not aligned. All C-intein sequences were “right-justified” to align the C-terminal asparagine. The residue numbering used in this study is based on the numbering for the NMR structure of a fused Npu intein (PDB code 2KEQ). Thus, the variable N-intein tail region after residue 102 (the last residue of NpuN) is excluded from the numbering, as is the N-terminal methionine of the C-intein. The C-intein numbering starts at 103, except for the Tel and Tvu inteins, which have a gap at this position and start at 104. For the sequence logos (FIG. 6), the N- and C-intein alignments were each separated into two alignments based on high and low activity. The high activity sequence logos were comprised of Cwa, Cra(CS505), Csp(PCC8801), Ava, Npu, Csp(CCY0110), Mcht(PCC7420), Maer(NIES843), Asp, Oli, and Aha (which was included based on the high activity of the C120G mutant). The low activity sequence logos were comprised of Aov, Ter, Ssp(PCC7002), Tvu, Tel, Ssp, and Sel(PC7942). The sequence logos were generated using WebLogo. (4) Heat maps were generated using the statistical computing and graphics program “R”.

Cloning of Plasmids for In Vivo Screening

The aminoglycoside phosphotransferase (KanR) and Npu gene fragments were cloned into a pBluescript KS (+) vector between KpnI and SacI restriction sites as previously described (36,37). This construct contained the following architecture:
[KanR promoter]-[RBS]-[myc-KanRN]-[IntN]-[iRBS]-[IntC]-[CFN-KanRC] where the KanR promoter is the constitutive promoter found in most kanamycin-resistant plasmids, RBS is a common E. coli ribosomal binding site, iRBS is an intervening ribosomal binding site preceded by a linker, myc encodes for a c-myc epitope tag (EQKLISEEDL) (SEQ ID NO: 760), KanRN and KanRC are fragments of the KanR protein, and IntN and IntC are split intein fragments. An analogous Ssp plasmid was also constructed as previously described (36,37). These plasmids are referred to as myc-KanR-NpuDnaE-Split and myc-KanR-SspDnaE-Split. To generate the screening vectors for the remaining split inteins, synthetic genes were designed and purchased from GENEWIZ containing the following architecture:

- [5′ overhang]-[IntN]-[iRBS]-[IntC]-[3′ overhang]
  where the 5′ and 3′ overhangs were the exact 39 bp found upstream of NpuN and 25 by found downstream of NpuC, respectively, in the myc-KanR-NpuDnaE-Split plasmid. For all inteins, the purchased gene sequences were codon-optimized with the default E. coli codon usage table generated based on all E. coli coding sequences in GenBank8. The synthetic genes were received in pUC57 vectors.

To clone the screening plasmids, the entire synthetic gene was amplified with Phusion High-Fidelity Polymerase using primers annealing to the 5′ and 3′ overhangs. The resulting megaprimer was inserted into the myc-KanR plasmid in place of Npu by overlap-extension PCR with Phusion polymerase (39). This resulted in 18 homologous plasmids containing identical backbones, promotors, and KanR genes, but with different codon-optimized intein genes. The plasmids are named as: myc-KanR-XyzDnaE-Split (where Xyz indicates the intein name as given in Table 1). Specific point mutations were made to various inteins using a QuikChange Site-Directed Mutagenesis kit with the standard recommended protocol.

In Vivo Screening of Relative Intein Activities

96-Well Plate Assay:
Intein activity-coupled kanamycin resistance (KanR) assays were conducted in 96-well plate format as previously described (36,37). Typically, plasmids were transformed into 15 μL of sub-cloning efficiency DH5α cells by heat shock, and the transformed cells were grown for 18 hours at 37° C. in 3 mL of Luria-Bertani (LB) media with 100 μg/mL of ampicillin (LB/amp). The over-night cultures were diluted 250-fold into LB/amp solutions containing 8 different kanamycin concentrations (150 μL per culture). The cells were grown at 30° C. on a 96-well plate, monitoring optical density (OD) at 650 nm every 5 minutes for 24 hours while shaking for one minute preceding each measurement. The endpoint of this growth curve (typically in the stationary phase) was plotted as a function of kanamycin concentration to visualize the dose-response relationship and fitted to a variable-slope dose-response equation to determine IC₅₀values.
${OD}_{Obs} = {OD}_{Min} + \frac{({OD}_{Max} - {OD}_{Min})}{1 + 10^{[(\log {IC}_{50} - \log [Kan]) \cdot HillSlope]}}$
In each regression analysis, typically three or four independent dose response curves were collectively fit to the equation above using the GraphPad Prism software. In each fit, OD_Minwas fixed to the background absorbance at 650 nm, and all other parameters were allowed to vary. The reported error bars for the IC₅₀bar graphs (FIG. 1 b) represent the standard error in the best-fit IC₅₀value from three or four collectively fit dose-response curves.

Western Blot Analysis of In Vivo Splicing:

For the western blot analyses, DH5α cells were transformed with the assay plasmids identically as for the 96-well plate setup and grown for 18 hours at 37° C. while shaking. The overnight cultures were used to inoculate 3 mL of fresh LB/amp at a 1:300 dilution, and the cells were incubated at 30° C. for 24 hours. The ODs of the 30° C. cultures were measured at 650 nm to assess relative bacterial levels, then 150 μL of each culture was transferred to an Eppendorf tube and centrifuged at 17,000 rcf for 2 minutes. The supernatant was aspirated off, and the cell pellets were resuspended/lysed in ˜200 μL of 2×SDS gel loading dye containing 4% BME (the resuspension volumes were varied slightly to normalize for differences in OD). The samples were boiled for 10 minutes, then centrifuged at 17,000 rcf for 1 minute. Each sample (5 μL) was loaded onto a 12% Bis-Tris gel and run in MES-SDS running buffer. The proteins were transferred to PVDF membrane in Towbin transfer buffer (25 mM Tris, 192 mM glycine, 15% methanol) at 100V for 90 minutes. Membranes were blocked with 4% milk in TBST, then the primary antibody (α-myc, 1:5000) and secondary antibody (Licor mouse 800, 1:15,000) were sequentially applied in 4% milk in TBST. The blots were imaged using the Licor Odyssey scanner.

Cloning of Plasmids for In Vitro Splicing Assays

Ub-IntN Plasmids:
The N-intein expression plasmids were derived from a previously described NpuN plasmid, pMR-Ub-NpuN(WT) (36,37). This plasmid encoded for the following protein sequence:

(SEQ ID NO: 761)

MHHHHHHGGMQIFVKTLIGKTITLEVESSDTIDNVKSKIQDKEGIPPDQQ

ELIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGGGGGGKFAEY CLSYET

EILTVEYGLLPIGKIVEKRIECTVYSVDNNGNIYTQPVAQWHDRGEQEVF

EYCLEDGSLIRATKDHKFMTVDGQMLPIDEIFERELDLMRVDNLPN

where the NpuN sequence is given in bold, the immediate native local extein residues are underlined, and these residues are preceded by His₆-Ub with a Gly₄linker. Significant in vivo proteolysis was previously observed during expression of this construct, so this plasmid was modified using QuikChange to remove the Gly₄sequence. The resulting plasmid, pMR-Ub-NpuN-ΔGly₄was used as the template for all other Ub-IntN plasmids and encoded for the following protein sequence:

(SEQ ID NO: 762)

MHHHHHHGGMQIFVKTLTGKTITLEVESSDTIDNVKSKIQDKEGIPPDQQ

ELIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGGKFAEY CLSYETEILT

VEYGLLPIGKIVEKRIECTVYSVDNNGNIYTQPVAQWHDRGEQEVFEYCL

EDGSLIRATKDHKFMTVDGQMLPIDEIFERELDLMRVDNLPN

All other IntN plasmids were cloned using overlap-extension PCR to generate Ub-IntN fusion genes in homologous plasmids in a traceless manner (39). Specifically, N-intein genes were amplified by Phusion polymerase from the synthetic gene plasmids using primers with overhangs that anneal to the plasmid sequences surrounding NpuN in pMR-Ub-NpuN-ΔGly₄. The resulting megaprimer was then used to insert the new N-intein gene in place of NpuN to generate a new plasmid called pMR-Ub-IntN that was identical to the NpuN plasmid except for the N-intein gene.
IntC-SUMO Plasmids:
The C-intein plasmids were all derived from a previously described NpuC plasmid, pET-NpuC(WT)-SUMO (37). This plasmid encoded for the following protein sequence:

(SEQ ID NO: 763)

MGSSHHHHHHGENLYFQ|GIKIATRKYLGKQNVYDIGVERDHNFALK

NGFIASN CFNSGLVPRGSASMSDSEVNQEAKPEVKPEVKPETHINLKVSD

GSSEIFFKIKKTTPLRRLMEAFAKRQGKEMDSLRFLYDGIRIQADQTPED

LDMEDNDIIEAHREQIGGYPYDVPDYA

where the NpuC sequence is given in bold, and the immediate native local extein residues are underlined, followed by a linker sequence and SUMO-HA. This construct is preceded by a His₆-tag and a tobacco etch virus (TEV) protease recognition sequence. The TEV protease cleavage site is indicated by “|” and leaves behind a glycine residue in place of an N-terminal IntC methionine.

All other IntC plasmids were cloned using overlap-extension PCR to generate IntC-SUMO fusion genes in homologous plasmids in a traceless manner (39). Specifically, C-intein genes were amplified by Phusion polymerase from the synthetic gene plasmids using primers with overhangs that anneal to the plasmid sequences surrounding NpuC in pET-NpuC(WT)-SUMO. The resulting megaprimer was then used to insert the new C-intein gene in place of NpuC to generate a new plasmid called pET-IntC-SUMO that was identical to the NpuC plasmid except for the C-intein gene.

Purification of Proteins for In Vitro Splicing Assays

Over-Expression and Purification of Ub-IntN Constructs (Except Ub-CwaN):
E. coli BL21(DE3) cells transformed with each N-intein plasmid were grown in 1 L of LB containing 100 μg/mL of ampicillin at 37° C. until OD₆₀₀=0.6. The cells were then cooled down to 18° C., and expression was induced by addition of 0.5 mM IPTG for 16 hours at 18° C. After harvesting the cells by centrifugation (10,500 rcf, 30 min), the cell pellets were transferred to 50 mL conical tubes with 5 mL of lysis buffer (50 mM phosphate, 300 mM NaCl, 5 mM imidazole, 2 mM BME, pH 8.0) and stored at −80° C. The cell pellets were resuspended by adding an additional 15 mL of lysis buffer supplemented with Complete protease inhibitor cocktail. Cells were lysed by sonication (35% amplitude, 8×20 second pulses separated by 30 seconds on ice). The soluble fraction was recovered by centrifugation (35,000 rcf, 30 min). The soluble fraction was mixed with 2 mL of Ni-NTA resin and incubated at 4° C. for 30 minutes. After incubation, the slurry was loaded onto a fritted column. After discarding the flow-through, the column was washed with 5 column volumes (CV) of lysis buffer, 5 CV of wash buffer 1 (lysis buffer with 20 mM imidazole), and 3 CV of wash buffer 2 (lysis buffer with 50 mM imidazole). The protein was eluted with elution buffer (lysis buffer with 250 mM imidazole) in four 1.5 CV elution fractions. The wash and elution fractions were analyzed by SDS-PAGE.
After enrichment over the Ni-NTA column, the proteins were purified by gel filtration. The wash and elution fractions were all treated with 50 mM DTT for 30 minutes on ice. For well-expressing proteins, the first elution fraction was then directly injected on an S75 10/300 gel filtration column (3×1 mL injections) and eluted over 1.35 CV in freshly prepared, degassed splicing buffer (100 mM phosphates, 150 mM NaCl, 1 mM DTT, 1 mM EDTA, pH 7.2). For the more dilute, low-yielding proteins, typically the 50 mM imidazole wash fraction and the first two elution fractions were pooled and concentrated four-fold to 3 mL. Then, the concentrated protein was purified by gel filtration identically to the high-yielding constructs. FPLC fractions were analyzed by SDS-PAGE, and the purest fractions were pooled and analyzed by analytical gel filtration, analytical RP-HPLC, and mass spectrometry. The concentration of pure proteins were determined by UV A280 nm and by the Bradford assay.
Over-Expression and Purification of Ub-CwaN:
The Ub-CwaN protein did not express well in the soluble fraction, and all of the enriched protein was aggregated, as observed by gel filtration analysis. Thus, after expression, cell lysis, and fractionation, as described above, the protein was extracted from the insoluble fraction of the lysate as follows. First, the lysate pellet was resuspended in 20 mL of Triton wash buffer (lysis buffer with 0.1% Triton X-100) and incubated at room temperature for 30 minutes. The Triton wash was centrifuged at 35,000 rcf for 30 minutes, and the supernatant was discarded. Next, the pellet was resuspended in 20 mL of lysis buffer containing 6 M urea, and the mixture was incubated overnight at 4° C. The mixture was centrifuged at 35,000 rcf for 30 minutes, and then the supernatant was mixed with 2 mL of NiNTA resin. The Ni column was run identically as for the native purifications described above, except that every buffer had a background of 6 M urea. Following enrichment over a Ni-NTA column, the 50 mM imidazole wash and the first two elution fractions were pooled and diluted to 0.2 mg/mL. The diluted protein was refolded into lysis buffer (without urea) by step-wise dialysis removal of the urea at 4° C. The protein was concentrated four-fold to 3 mL and immediately purified by gel filtration as indicated for the native purifications above. The pure protein was analyzed by analytical gel filtration, analytical RP-HPLC, and mass spectrometry. Note that this construct was highly susceptible to aggregation. When re-folded at 2 mg/mL rather than 0.2 mg/mL, less than 10% of the obtained protein was monomeric, whereas more dilute refolding yielded roughly 50% monomeric protein. The obtained protein was 80% monomeric, and the monomer to aggregate ratio did not change after 24 hours of storage at 4° C. The concentration of pure protein was determined by the Bradford assay.
Over-Expression and Purification of IntC-SUMO Constructs:
E. coli BL21(DE3) cells transformed with each C-intein plasmid were grown in 1 L of LB medium containing kanamycin (50 μg/mL) at 37° C. until OD₆₀₀=0.6. Then, expression was induced by addition of 0.5 mM IPTG for 3 hours at 37° C. The cells were lysed, and the desired protein was enriched over Ni-NTA resin identically as for the natively purified Ub-IntN proteins. The AvaC-SUMO and Csp(PCC8801)C-SUMO proteins did not express well at 37° C., so the proteins were re-expressed by induction at 18° C. for 16 hours. For each protein, the 50 mM imidazole wash and the first two elution fractions were pooled and dialyzed into TEV cleavage buffer (50 mM phosphate, 300 mM NaCl, 5 mM imidazole, 0.5 mM EDTA, 0.5 mM DTT, pH 8.0) then treated with 40 μg of His-tagged TEV protease overnight at room temperature. The cleavage was confirmed by RP-HPLC/MS, after which the reaction solution was incubated with Ni-NTA resin at room temperature for 30 min. The flow-through and two 1.5 CV washes with wash buffer 1 were collected and pooled. The protein was then concentrated to 3-4 mL, injected onto the S75 10/300 gel filtration column (3×1 mL injections), and eluted over 1.35 CV in freshly prepared, degassed splicing buffer (100 mM phosphates, 150 mM NaCl, 1 mM DTT, 1 mM EDTA, pH 7.2). FPLC fractions were analyzed by SDS-PAGE, and the purest fractions were pooled and analyzed by analytical gel filtration, analytical RP-HPLC, and mass spectrometry. The concentration of pure protein was determined by UV A_{280 nm}and by the Bradford assay.
Usage and Storage of the Ub-IntN and IntC-SUMO Constructs:
All of the purified proteins were stored at 4° C. and used within two days for splicing assays with their cognate IntC-SUMO. The remaining protein (2 vol. eq.) was mixed with splicing buffer containing 60% glycerol (1 vol. eq.) to yield a 20% glycerol stock that was aliquoted and flash frozen in liquid N₂. The protein aliquots were stored at −80° C. The proteins were fully functional after thawing on ice and could be flash-frozen and re-thawed at least once without detectable loss of function.

In Vitro Splicing Assays

Kinetic Assay Procedure:
For a typical assay, individual protein stock solutions of Ub-IntN and IntC-SUMO constructs were prepared in filtered splicing buffer (100 mM phosphate, 150 mM NaCl, 1 mM DTT, 1 mM EDTA, pH 7.2) at 2× the final concentration (e.g. 2.0 μM stock solution for a 1.0 μM reaction). 1 mM TCEP was added (from a pH-neutralized 100 mM stock solution) to each protein solution, and the proteins were incubated at 30° C. or 37° C. for 5 min depending on the reaction temperature. To initiate a reaction, the N- and C-intein were mixed at equal volumes (i.e. equimolar ratios). A typical reaction volume was 300 μL and was carried out in an Eppendorf tube on a heat block. During the reaction, 20 μL aliquots of the reaction solution were removed at the desired time points and quenched in 20 μL of 2× concentrated SDS gel loading dye on ice to afford a final quenched solution with 40 mM Tris (˜pH 7.0), 10% (v/v) glycerol, 1% (w/v) SDS, 0.02% (w/v) bromophenol blue, and 2% (v/v) BME. For each reaction, an artificial zero time point was taken by mixing equivalent amounts of starting materials directly into the quencher solution. Samples were boiled for 10 minutes then centrifuged at 17,000 rcf for 1 minute. Aliquots of starting materials and time points (15 μL) were loaded onto Bis-Tris gels and run in MES-SDS running buffer. The gels were Coomassie-stained then imaged using the Licor Odyssey scanner.
Note that for the reactions with a CGN C-extein sequence, no BME was used in the quencher solution. Furthermore, before boiling the samples, each sample was treated with 1 μL of 2 N HCl. After boiling and cooling the samples, they were treated with 1 μL of 2 N NaOH. This procedure prevented undesired hydrolysis or thiolysis of the branched intermediate.
Determination of Kinetic Parameters:
To determine reaction rates, each lane of a gel was analyzed using the Licor Odyssey quantification function or ImageJ. Given the close proximity of the starting material bands, these bands were typically integrated together. To normalize for loading error, the integrated intensity of each band in a lane was expressed as a fraction intensity of the total band intensity in that lane (which remained relatively constant between lanes). These normalized intensities were plotted as a function of time, and data from three independent reactions were collectively fit to first-order rate equations using the GraphPad Prism software:
Y=S·(e ^−k ^obs ^·t)+Z For reactant depletion:
Y=Y _max·(1−e ^−k ^obs ^·t) For product formation:
Y is the fractional intensity of a species, t is time in minutes, S is a scaling factor for reactant depletion (allowed to vary), Z indicates the fraction of reactant remaining at the reaction endpoint (allowed to vary), Y_maxis a scaling factor for product formation, and k_obsis the observed first-order rate constant for the splicing reaction (allowed to vary). Half-lives were calculated from the best-fit value for the first-order rate constant:
$t_{1 / 2} = \frac{\ln 2}{k_{obs}}$
For reactions with no detectable side product formation, the rate of product (Ub-SUMO) and IntN formation were consistent with the rate of starting material depletion.
Western Blot Analysis of Reactions:
Western blots of the zero time point and reaction endpoint were carried out to confirm the identities of the observed bands. The quenched time points from the reactions described above were loaded onto 12% Bis-Tris gels (5 μL per sample, two identical gels) and run in MES-SDS running buffer. The resolved proteins were transferred from the gel onto PVDF membrane in CAPS transfer buffer (10 mM N-cyclohexyl-3-aminopropanesulfonic acid, 10% (v/v) methanol, pH 10.5) at 100 V for 60 minutes. Membranes were blocked with Licor Blocking Buffer, then the primary antibody (α-His₆, 1:3000, or α-HA, 1:25,000) was applied in Licor Blocking Buffer. The secondary antibody (Licor mouse 800, 1:15,000) was applied in 4% milk in TBST. The blots were imaged using the Licor Odyssey scanner. Blots from the 30° C. and 37° C. reactions were virtually identical.

HPLC/MS Analysis of Npu-CGN and Cra(CS505)-CGN Reactions

For the HPLC/MS analysis of Npu-CGN and Cra(CS505)-CGN, individual protein stock solutions of Ub-IntN and IntC-CGN-SUMO were prepared in filtered splicing buffer (100 mM phosphate, 150 mM NaCl, 1 mM DTT, 1 mM EDTA, pH 7.2) at 8.01 μM. 1 mM TCEP was added (from a pH-neutralized 100 mM stock solution) to each protein solution, and the proteins were incubated at 30° C. for 5 min. To initiate a reaction, the N- and C-intein were mixed at equal volumes (i.e. equimolar ratios) and incubated at 30° C. During the reaction, 90 μL aliquots of the reaction solution were removed at the desired time points and quenched in 30 μL of a quenching solution (6 M guanidine hydrochloride with 4% trifluoroacetic acid). 100 μL of each quenched time point were injected onto an analytical C18 RP-HPLC column and eluted over a 25-73% buffer B gradient in 30 minutes, preceded by a two minute isocratic phase in 25% buffer B (see Equipment section for column and running buffer specifications). At different time points, various HPLC peaks were collected and their identities were confirmed by mass spectrometry. The IntC-(Ub)SUMO species were identified by MS, verifying branched intermediate formation and depletion.

Kinetic Modeling

When comparing the Npu, Cra(CS505), and Cwa reactions in the presence of CGN, higher amounts of cleaved ubiquitin (i.e. N-extein cleavage) were obtained than splice product, despite the fact that the rate of the former was slower than the latter. This observation is inconsistent with N-extein cleavage and splice product formation only occurring from the branched intermediate, since in this scenario splicing and cleavage would be competing first-order reactions occurring from the same reactant (the branched intermediate), leading to more splice product than cleavage (the opposite of that observed). In an attempt to reconcile these observations, a series of kinetic modeling simulations were carried out. All modeling was carried out using the kinetic modeling applet from BPReid (40). Themodels have three basic assumptions about the splicing pathway:
1. The forward and reverse reactions in the first equilibrium are fast. In addition, the position of this equilibrium lies slightly towards the amide.
2. The second equilibrium is also fast and should have K_eqclose to 1, since both intermediates are cysteinyl thioesters.
3. For fast inteins, the rate of branched intermediate resolution (k₅) is on the same order of magnitude as the rates of the first two reversible steps, whereas the cleavage rates from L (k₆) and B (k₇) are relatively slow. For slow inteins, branched intermediate resolution (k₅) is also slow, on the same order of magnitude as the cleavage.
With these assumptions, six scenarios were devised that assess how the relative rates of cleavage and branched intermediate resolution and the equilibrium between the linear and branched intermediates could affect the rates and extents of formation of the cleavage and spliced products. For slow inteins, such as those bearing exogenous C-extein residues, the rate of branched intermediate resolution is similar to the rate of N-extein cleavage. Under these circumstances, three factors are important:
1. The relative rates of cleavage from L versus B (k₆vs. k₇).
2. The relative rates of branched intermediate resolution versus cleavage (k₅vs. k₆+k₇).
3. Most importantly, the rates of exchange between the linear and branched intermediates (k₃/k₄).
These analyses suggest not only that cleavage should be occurring both from the linear and branched intermediate, but also that cleavage at the linear intermediate may be favored.
Protein Thiolysis and Ligation from Fused DnaE Inteins and MxeGyrA
Solid-Phase Synthesis and Purification of H-Cys-Gly-Lys(Flourescein)-NH₂(CGK-Fluorescein):
Fmoc-based solid phase peptide synthesis (SPPS) was used to produce a peptide with the sequence H-Cys-Gly-Lys(Fluorescein)-NH₂. The peptide was synthesized on Rink amide resin at a 0.2 mmol scale as follows: 20% piperidine in DMF was used for Fmoc deprotection using a one minute equilibration of the resin followed by a 20 minute incubation. After Fmoc deprotection, amino acids were coupled using DIC/HOBt as activating agents. First, the amino acid (1.1 mmol) was dissolved in 50:50 DCM:DMF (2 mL) and was activated with DIC (1.0 mmol) and HOBt (1.2 mmol) at 0° C. for 15 minutes. The mixture was added to the N-terminally deprotected resin and coupled for 10 minutes at room temperature.
After the cysteine was coupled, the lysine side chain was deprotected by treatment with Pd(Ph₃)₄(0.1 eq.) and phenylsilane (25 eq.) in dry DCM for 30 minutes. The peptidyl resin was washed with DCM (2×5 mL) and DMF (2×5 mL) followed by two washes with 0.5% DIPEA in DMF (v/v) and two washes with 0.5% sodium diethyldithiocarbamate trihydrate in DMF (w/v) to remove any remaining traces of the Pd catalyst. 5(6)-Carboxyfluorescein was then coupled to the lysine side chain using the DIC/HOBt activation method overnight at room temperature. Finally, the peptide was cleaved off the resin using 94% TFA, 1% TIS, 2.5% EDT, and 2.5% H₂O (6.5 mL) for one hour. After cleavage, roughly half of the TFA was evaporated under a stream of nitrogen. The crude peptide was precipitated with cold ether and washed with cold ether twice. Finally, the peptide was purified by RP-HPLC on C18 prep column over a 15-80% buffer B gradient in 40 minutes. The purified peptide was analyzed by analytical RP-HPLC and ESI-MS to confirm its identity. Note that no attempt was made to separately isolate the 5-carboxyfluorescein and 6-carboxyfluorescein conjugates, thus the peptide is a mixture of these two isomers.
Cloning of Ub-Intein Fusions:
All Ub-Intein fusions were cloned into a modified pTXB1 vector from NEB containing ubiquitin in which a His₆-tag and stop codon were inserted between the MxeGyrA intein and the chitin binding domain. This resulted in a plasmid, pTXB1-Ub-MxeGyrA-ATEA-H₆that encodes for the following protein, called called Ub-MxeGyrA-ATEA-H₆:

(SEQ ID NO: 764)

MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQL

EDGRTLSDYNIQKESTLHLVLRLRGGCITGDALVALPEGESVRIADIVPG

ARPNSDNAIDLKVLDRHGNPVLADRLFHSGEHPVYTVRTVEGLRVTGTAN

HPLLCLVDVAGVPTLLWKLIDEIKPGDYAVIQRSAFSVDCAGFARGKPEF

APTTYTVGVPGLVRFLEAHHRDPDAQAIADELTDGRFYYAKVASVTDAGV

QPVYSLRVDTADHAFITNGFVSHATEAHHHHHH

in which the intein sequence for MxeGyrA (N198A) is shown in bold, preceded by ubiquitin and followed by the endogenous local C-extein sequence (underlined) and a His₆-tag.

This plasmid was modified to replace the MxeGyrA intein with a fused Npu intein. First, the myc-KanR-NpuDnaE-Split plasmid was modified by QuikChange to remove the iRBS sequence separating the NpuN and NpuC genes. The resulting plasmid, myc-KanR-NpuDnaE-Fused, was then used as a template to amplify megaprimers bearing the fused Npu intein with overhangs homologous to the sequences surrounding MxeGyrA in the modified pTXB1 vector. The Npu gene with the N137A mutation was inserted in place of MxeGyrA using overlap-extension PCR with the Phusion polymerase. (39) Importantly, this construct was modified to include the native C-extein residues of Npu (CFN) instead of those for MxeGyrA (TEA). The resulting plasmid, pTXB1-Ub-NpuDnaE-ACFN-H₆encoded for the following protein:

(SEQ ID NO: 765)

MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQL

EDGRTLSDYNIQKESTLHLVLRLRGGCLSYETEILTVEYGLLPIGKIVEK

RIECTVYSVDNNGNIYTQPVAQWHDRGEQEVFEYCLEDGSLIRATKDHKF

MTVDGQMLPIDEIFERELDLMRVDNLPNIKIATRKYLGKQNVYDIGVERD

HNFALKNGFIASACFNHHHHHH

This fusion showed substantial in vivo hydrolysis of ubiquitin when expressed in E. coli. Thus, it was further modified using QuikChange mutagenesis by mutating the +1 cysteine to alanine, generating the plasmid pTXB1-Ub-NpuDnaE-AAFN-H₆. This plasmid encoded for the following protein (Ub-NpuDnaE-AAFN-H₆) that was used for in vitro thiolysis experiments:

(SEQ ID NO: 766)

MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQL

EDGRTLSDYNIQKESTLHLVLRLRGGCLSYETEILTVEYGLLPIGKIVEK

RIECTVYSVDNNGNIYTQPVAQWHDRGEQEVFEYCLEDGSLIRATKDHKF

MTVDGQMLPIDEIFERELDLMRVDNLPNIKIATRKYLGKQNVYDIGVERD

HNFALKNGFIASA AFNHHHHHH

The pTXB1-Ub-AvaDnaE-AAFN-H₆and pTXB1-Ub-MchtDnaE-AAFN-H₆plasmids, encoding for the following protein sequences (Ub-AvaDnaE-AAFN-H₆and Ub-MchtDnaE-AAFN-H₆, respectively), were cloned analogously by modifying the pTXB1-Ub-NpuDnaE-AAFN-H₆plasmid.

Ub-AvaDnaE-AAFN-H₆

(SEQ ID NO: 767)

MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQL

EDGRTLSDYNIQKESTLHLVLRLRGGCLSYDTEVLTVEYGFVPIGEIVDK

GIECSVFSIDSNGIVYTQPIAQWHHRGKQEVFEYCLEDGSIIKATKDHKF

MTQDGKMLPIDEIFEQELDLLQVKGLPEIKIASRKFLGVENVYDIGVGRD

HNFFVKNGLIASA AFNHHHHHH

Ub-MchtDnaE-AAFN-H₆

(SEQ ID NO: 768)

MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQL

EDGRTLSDYNIQKESTLHLVLRLRGGCLSYDTQILTVEYGAVAIGEIVEK

QIECTVYSVDENGYVYTQPIAQWHNRGEQEVFEYLLEDGATIRATKDHKF

MTDEDQMLPIDQIFEQGLELKQVEVLQPVFVKIVRRQSLGVQNVYDIGVE

KDHNFCLASGEIASA AFNHHHHHH

As a control for the removal of the +1 Cys residue in the DnaE intein constructs, the +1 Thr residue was mutated from the pTXB1-Ub-MxeGyrA-ATEA-H₆plasmid by QuikChange mutagenesis to yield the plasmid pTXB1-Ub-MxeGyrA-AAEA-H₆, encoding for the protein Ub-MxeGyrA-AAEA-H₆.

(SEQ ID NO: 769)

MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQL

EDGRTLSDYNIQKESTLHLVLRLRGGCITGDALVALPEGESVRIADIVPG

ARPNSDNAIDLKVLDRHGNPVLADRLFHSGEHPVYTVRTVEGLRVTGTAN

HPLLCLVDVAGVPTLLWKLIDEIKPGDYAVIQRSAFSVDCAGFARGKPEF

APTTYTVGVPGLVRFLEAHHRDPDAQAIADELTDGRFYYAKVASVTDAGV

QPVYSLRVDTADHAFITNGFVSHA AEAHHHHHH

Cloning of Additional Fusions to Fused DnaE Inteins:
Several other proteins were fused to AvaDnaE or MchtDnaE to test the sequence dependence on thiolysis from these inteins. The proteins utilized were the N-terminal SH3 domain of human Grb2 (AAs 1-55+/− an exogenous C-terminal Gly), the SH2 domain of human Abl kinase (AAs 122-217), eGFP, and the catalytic domain of human PARP1 (AAs 657-1015). All plasmids were cloned using the aforementioned methods to yield plasmids encoding the following proteins:

SH3-AvaDnaE-AAFN-H₆

(SEQ ID NO: 770)

MEAIAKYDFKATADDELSFKRGDILKVLNEECDQNWYKAELNGKDGFIPK

NYIEMCLSYDTEVLTVEYGFVPIGEIVDKGIECSVFSIDSNGIVYTQPIA

QWHHRGKQEVFEYCLEDGSIIKATKDHKFMTQDGKMLPIDEIFEQELDLL

QVKGLPEIKIASRKFLGVENVYDIGVGRDHNFFVKNGLIASAAFNHHHHH

H

SH3-Gly-AvaDnaE-AAFN-H₆

(SEQ ID NO: 771)

MEAIAKYDFKATADDELSFKRGDILKVLNEECDQNWYKAELNGKDGFIPK

NYIEMGCLSYDTEVLTVEYGFVPIGEIVDKGIECSVFSIDSNGIVYTQPI

AQWHHRGKQEVFEYCLEDGSIIKATKDHKFMTQDGKMLPIDEIFEQELDL

LQVKGLPEIKIASRKFLGVENVYDIGVGRDHNFFVKNGLIASAAFNHHHH

HH

SH3-MchtDnaE-AAFN-H₆

(SEQ ID NO: 772)

MEAIAKYDFKATADDELSFKRGDILKVLNEECDQNWYKAELNGKDGFIPK

NYIEMCLSYDTQILTVEYGAVAIGEIVEKQIECTVYSVDENGYVYTQPIA

QWHNRGEQEVFEYLLEDGATIRATKDHKFMTDEDQMLPIDQIFEQGLELK

QVEVLQPVFVKIVRRQSLGVQNVYDIGVEKDHNFCLASGEIASAAFNHHH

HHH

SH2-AvaDnaE-AAFN-H₆

(SEQ ID NO: 773)

MLEKHSWYHGPVSRNAAEYLLSSGINGSFLVRESESSPGQRSISLRYEGR

VYHYRINTASDGKLYVSSESRFNTLAELVHHHSTVADGLITTLHYPACLS

YDTEVLTVEYGFVPIGEIVDKGIECSVFSIDSNGIVYTQPIAQWHHRGKQ

EVFEYCLEDGSIIKATKDHKFMTQDGKMLPIDEIFEQELDLLQVKGLPEI

KIASRKFLGVENVYDIGVGRDHNFFVKNGLIASAAFNHHHHHH

eGFP-AvaDnaE-AAFN-H₆

(SEQ ID NO: 774)

MVSKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICT

TGKLPVPWPTLVTTLTYGVQCFSRYPDHMKQHDFFKSAMPEGYVQERTIF

FKDDGNYKTRAEVKFEGDTLVNRIELKGIDFKEDGNILGHKLEYNYNSHN

VYIMADKQKNGIKVNFKIRHNIEDGSVQLADHYQQNTPIGDGPVLLPDNH

YLSTQSALSKDPNEKRDHMVLLEFVTAAGITLGMDELYKCLSYDTEVLTV

EYGFVPIGEIVDKGIECSVFSIDSNGIVYTQPIAQWHHRGKQEVFEYCLE

DGSIIKATKDHKFMTQDGKMLPIDEIFEQELDLLQVKGLPEIKIASRKFL

GVENVYDIGVGRDHNFFVKNGLIASAAFNHHHHHH

PARP_C-AvaDnaE-AAFN-H₆

(SEQ ID NO: 775)

MVNPGTKSKLPKPVQDLIKMIFDVESMKKAMVEYEIDLQKMPLGKLSKRQ

IQAAYSILSEVQQAVSQGSSDSQILDLSNRFYTLIPHDFGMKKPPLLNNA

DSVQAKAEMLDNLLDIEVAYSLLRGGSDDSSKDPIDVNYEKLKTDIKVVD

RDSEEAEIIRKYVKNTHATTHNAYDLEVIDIFKIEREGECQRYKPFKQLH

NRRLLWHGSRTTNFAGILSQGLRIAPPEAPVTGYMFGKGIYFADMVSKSA

NYCHTSQGDPIGLILLGEVALGNMYELKHASHISKLPKGKHSVKGLGKTT

PDPSANISLDGVDVPLGTGISSGVNDTSLLYNEYIVYDIAQVNLKYLLKL

KFNFKTSLWCLSYDTEVLTVEYGFVPIGEIVDKGIECSVFSIDSNGIVYT

QPIAQWHHRGKQEVFEYCLEDGSIIKATKDHKFMTQDGKMLPIDEIFEQE

LDLLQVKGLPEIKIASRKFLGVENVYDIGVGRDHNFFVKNGLIASAAFNH

HHHHH

Purification of Various Protein-Intein Fusions:
E. coli BL21(DE3) cells transformed with each Protein-Intein fusion plasmid were grown in 1 L of LB medium containing ampicillin (100 μg/mL) at 37° C. until OD₆₀₀=0.6. Then, expression was induced by addition of 0.5 mM IPTG and incubation for 3 hours at 37° C. or incubation for 16 hours at 18° C. All Ub fusions were expressed at 37° C., the eGFP fusion was expressed at 18° C., and the SH3, SH2, and PARP_Cfusions were expressed at both temperatures. After harvesting the cells by centrifugation (10,500 rcf, 30 min), the cell pellets were transferred to 50 mL conical tubes with 5 mL of lysis buffer (50 mM phosphate, 300 mM NaCl, 5 mM imidazole, No BME, pH 8.0) and stored at −80° C. The cell pellets were resuspended by adding an additional 15 mL of lysis buffer supplemented with Complete protein inhibitor cocktail. Cells were lysed by sonication (35% amplitude, 8×20 second pulses separated by 30 seconds on ice). The soluble fraction was recovered by centrifugation (35,000 rcf, 30 min). The soluble fraction was mixed with 2 mL of Ni-NTA resin and incubated at 4° C. for 30 minutes. After incubation, the slurry was loaded onto a fritted column. After discarding the flow-through, the column was washed with 5 column volumes (CV) of lysis buffer, 5 CV of wash buffer 1 (lysis buffer with 20 mM imidazole), and 3 CV of wash buffer 2 (lysis buffer with 50 mM imidazole). The protein was eluted with elution buffer (lysis buffer with 250 mM imidazole) in four 1.5 CV elution fractions. The wash and elution fractions were analyzed by SDS-PAGE with loading dye containing no thiols. The cleanest fractions were pooled and treated with 10 mM TCEP for 20 minutes on ice. Then, the solution was injected on an S75 or 5200 10/300 gel filtration column (2×1 mL injections), and eluted over 1.35 CV in thiolysis buffer (100 mM phosphates, 150 mM NaCl, 1 mM EDTA, 1 mM TCEP, pH 7.2). The FPLC fractions were analyzed by SDS-PAGE with loading dye containing no thiols, and the purest fractions were pooled and analyzed by analytical RP-HPLC and mass spectrometry. The concentration of pure protein was determined by UV A_{280 nm}.
Thiolysis of Ub-Intein Fusions and Ligation of Ubiquitin to a Small Fluorescent Peptide:
For each Ub-Intein fusion protein, four reactions were carried on a 100 μL scale at 30° C. In the first reaction to monitor background hydrolysis, the fusion protein (501.1M) was incubated in thiolysis buffer (100 mM phosphate, 150 mM NaCl, 1 mM EDTA, 1 mM TCEP, pH 7.2) supplemented with freshly added TCEP (an additional 5 mM). In the second and third reactions, the protein was incubated identically as for the first reaction, except that each reaction had either 100 mM MESNa or 1 mM CGK-Fluorescein. In the fourth reaction, both MESNa and the peptide were added. At various time points, 5 μL of reaction solution were removed and quenched in 30 μL 2×SDS loading dye containing no thiols. As time points were collected, they were stored at −20° C. until the end of the reaction. After the reaction, the 35 μL quenched time points were thawed, treated with 1 μL of a 1 M TCEP stock solution, boiled for 10 minutes, and centrifuged at 17,000 rcf for 1 minute. Time points (5 μL) were loaded onto 12% Bis-Tris gels and run in MES-SDS running buffer. The gels were first imaged on a fluorescence imager to visualize the Ub-CGK-Fluorescein ligation product. Then the gels were coomassie-stained and imaged using the Licor Odyssey scanner. In addition, the reaction endpoints were quenched by 20-fold dilution in H₂O with 0.1% TFA and injected on an analytical C18 RP-HPLC column. The mixture was separated over a 2 minute isocratic phase in 0% B followed by a 0-73% B linear gradient in 30 minutes. The major peaks were collected and analyzed by MS.
Thiolysis of SH3-, SH2-, eGFP-, and PARP_c-Intein Fusions:
Thiolysis reactions with several other proteins fused to the AvaDnaE and MchtDnaE fused inteins were carried out analogously to the ubiquitin reactions described above. In a typical reaction, carried out on a 300 μL scale at 30° C., 10 μM fusion protein was treated with 5 mM TCEP in thiolysis buffer then incubated in the presence or absence of MESNa (either 100 mM or 200 mM) added from a pH-adjusted 1 M stock solution. At various time points, aliquots (15 μL) of the reaction solution were quenched in 30 μL of 2×SDS gel loading dye containing no thiols and stored at −20° C. until the end of the reaction. After the reaction, the 45 μL quenched time points were thawed, treated with 1 μL of a 1 M TCEP stock solution, boiled for 10 minutes, and centrifuged at 17,000 rcf for 1 minute. Time points (15 μL) were loaded onto 12% Bis-Tris gels and run in MES-SDS running buffer. Then the gels were coomassie-stained and imaged using the Licor Odyssey scanner. In addition, the reaction endpoints were quenched by 4-fold dilution in H₂O with 0.1% TFA and injected on an analytical C18 RP-HPLC column. The mixture was separated over a 2 minute isocratic phase in 0% B followed by a 0-73% B linear gradient in 30 minutes. The product peaks were collected and analyzed by MS.

Observation of the Linear Thioester Intermediate in Fused DnaE Inteins

For Npu, Ava, and Mcht fusions to ubiquitin, three peaks were visible for the purified protein when directly injected onto a C18 RP-HPLC column from a neutral buffer. These peaks all had the same mass of the desired protein. When diluted 20-fold in H₂O containing 0.1% TFA (pH 2) and incubated for at least two hours at room temperature, the first two peaks merged into the third peak (FIG. 4 d). The same observation could not be made for MxeGyrA under identical conditions. To further confirm that an equilibrium between the precursor amide and linear thioester was occurring, the Ub-NpuDnaE-AAFN-H₆protein was diluted 20-fold in thiolysis buffer containing 1% SDS. Before boiling, the two major peaks were visible. After boiling for 10 minutes, when the protein was unfolded, the first major peak partially converged into the second major peak, suggesting that the latter was the amide, which should be more stable in the unfolded intein. Additional evidence that the three peaks were in equilibrium came from pH titrations. The protein was diluted 20-fold into citric acid/phosphate buffers ranging from pH 2 to pH 8, incubated at room temperature for 3-4 hours, then analyzed by HPLC over a 30-73% B gradient in 30 minutes (FIG. 4 d). The relative abundance of the three species was modulated and showed a bell-shaped pH dependence, similar to the activities of enzymes containing multiple ionizable functional groups in their active sites.
In addition to observing the desired protein mass from all three observed HPLC peaks, the presence of a −18 Da species was observed in the first two peaks. This mass change is characteristic of a dehydration reaction, and such a reaction has been previously reported by Mootz et. al. for a mutant form of the SspDnaB intein that cannot efficiently catalyze the initial N-to-S acyl shift. (41) Specifically, the tetrahedral intermediate of the forward and reverse acylation reactions can undergo acid-catalyzed dehydration to yield a thiazoline side product. For Mootz and co-workers, this species was an irreversible side-product for their mutant intein under normal reaction conditions, and it lead to low yields. In the systems herein, where the DnaE inteins can react to completion, this species is either an artifact of acidification during RP-HPLC or it is fully reversible under normal reaction conditions. It is noteworthy that the observation of the thiazoline by MS further validates the presence of detectable levels of the tetrahedral intermediate in the present reaction mixtures.
For the DnaE intein fusions to proteins other than ubiquitin, similar HPLC profiles were observed with multiple peaks at neutral pH, however the ratios of the three peaks varied depending on the sequence. Additionally, for sequences more similar to the endogenous “A-E-Y” DnaE N-extein (such as the SH3 fusions with an “I-E-M” sequence), substantial accumulation of a dehydrated product (as much as 50% by HPLC/MS) was seen, similar to that observed by Mootz et. al. for the split SspDnaB intein. (41) For these constructs, this species appears to accumulate during protein expression resulting in a mixture of “trapped” (dehydrated) and “free” (native, hydrated) fusion protein. Upon addition of MESNa at neutral pH, the “free” protein rapidly undergoes thiolysis to yield the desired product, and the “trapped” protein slowly rehydrates and is also thiolyzed to yield the same desired product. Thus, in the reaction progress curves, a “burst” phase was observed followed by a slower phase. Importantly, the accumulation of dehydrated fusion protein could be reduced by expression at lower temperatures (18° C. instead of 37° C.), and these reactions could be driven faster and closer to completion by increasing the MESNa concentration from 100 mM to 200 mM MESNa. In addition, it is noteworthy that for the SH3 thiolysis reaction, the MchtDnaE intein was substantially more efficient that the AvaDnaE intein, suggesting that different fused DnaE inteins may be preferable depending on the protein of interest.
Numerous modifications and variations in the invention as set forth in the above illustrative examples are expected to occur to those skilled in the art. Consequently only such limitations as appear in the appended claims should be placed on the invention.

REFERENCES

(1) Mills, K. V.; Perler, F. B. Protein Pept. Lett. 2005, 12, 751-5.
(2) Vila-Perelló, M.; Muir, T. W. Cell 2010, 143, 191-200.
(3) Southworth, M.; Amaya, K.; Evans, T.; Xu, M.; Perler, F. Biotechniques 1999, 27, 110-120.
(4) Amitai, G.; Callahan, B. P.; Stanger, M. J.; Belfort, G.; Belfort, M. Proc. Natl. Acad. Sci. USA 2009, 106, 11005-10.
(5) Zettler, J.; Schütz, V.; Mootz, H. D. FEBS Lett. 2009, 583, 909-14.
(6) Iwai, H.; Züger, S.; Jin, J.; Tam, P.-H. FEBS Lett. 2006, 580, 1853-8.
(7) Perler, F. B. Nucleic Acids Res. 2002, 30, 383-4.
(8) Caspi, J.; Amitai, G.; Belenkiy, O.; Pietrokovski, S. Mol. Microbiol. 2003, 50, 1569-77.
(9) Dassa, B.; Amitai, G.; Caspi, J.; Schueler-Furman, O.; Pietrokovski, S. Biochemistry 2007, 46, 322-330.
(10) Chen, L.; Zhang, Y.; Li, G.; Huang, H.; Zhou, N. Anal. Biochem. 2010, 407, 180-7.
(11) Martin, D. D.; Xu, M. Q.; Evans, T. C. Biochemistry 2001, 40, 1393-402.
(12) Lockless, S. W.; Muir, T. W. Proc. Natl. Acad. Sci. USA 2009, 106, 10999-1004.
(13) Shah, N. H.; Vila-Perelló, M.; Muir, T. W. Angew. Chem. Int. Ed. Engl. 2011, 50, 6511-5.
(14) Oeemig, J. S.; Aranko, A. S.; Djupsjöbacka, J.; Heinämäki, K.; Iwaï, H. FEBS Lett. 2009, 583, 1451-1456.
(15) Du, Z.; Liu, Y.; Ban, D.; Lopez, M. M.; Belfort, M.; Wang, C. J. Mol. Biol. 2010, 400, 755-67.
(16) Appleby-Tagoe, J. H.; Thiel, I. V.; Wang, Y.; Wang, Y.; Mootz, H. D.; Liu, X.-Q. J. Biol. Chem. 2011, 286, 34440-7.
(17) Busche, A. E. L.; Aranko, A. S.; Talebzadeh-Farooji, M.; Bernhard, F.; Dötsch, V.; Iwaï, H. Angew. Chem. Int. Ed. Engl. 2009, 48, 6128-31.
(18) Dhar, T.; Mootz, H. D. Chem. Commun. 2011, 47, 3063-5.
(19) Pellois J-P, Muir T W: Current Opinion in Chemical Biology 2006, 10:487-491.
(20) Cheriyan M, Perler F B: Adv Drug Deliv Rev 2009.
(21) Muir T W, Sondhi D, Cole P A: Proc. Natl. Acad. Sci. U.S.A 1998, 95:6705-6710.
(22) Dawson P E, Muir T W, Clark-Lewis I, Kent S B: Science 1994, 266:776-779.
(23) Evans T C J R TC, Xu M-Q: Chem Rev 2002, 102:4869-4884.
(24) Wu Y-W, Oesterlin L K, Tan K-T, Waldmann H, Alexandrov K, Goody R S: Nat Chem Biol 2010, 6:534-540.
(25) Frutos S, Goger M, Giovani B, Cowburn D, Muir T W: Nature chemical biology 2010, 6:527-533.
(26) lsen SK, Capili A D, Lu X, Tan D S, Lima C D: Nature 2010, 463:906-912.
(27) Lu W, Sun Z, Tang Y, Chen J, Tang F, Zhang J, Liu J-N: Journal of Chromatography A 2011, 1218:2553-2560.
(28) Zettler J, Schütz V, Mootz H D: FEBS Letters 2009, 583:909-914.
(29) Southworth M W, Amaya K, Evans T C, Xu M Q, Perler F B: BioTechniques 1999, 27:110-114, 116, 118-120.
(30) Carter P J: Exp Cell Res 2011, 317:1261-1269.
(31) Perler, F. B. Nucleic Acids Res. 2002, 30, 383-4.
(32) Waterhouse, A. M.; Procter, J. B.; Martin, D. M. A.; Clamp, M.; Barton, G. J. Bioinformatics 2009, 25, 1189-91.
(33) Oeemig, J. S.; Aranko, A. S.; Djupsjöbacka, J.; Heinämäki, K.; Iwaï, H. FEBS Lett. 2009, 583, 1451-1456.
(34) Crooks, G. E.; Hon, G.; Chandonia, J.-M.; Brenner, S. E. Genome Res. 2004, 14, 1188-90.
(35) R Development Core Team; R Foundation for Statistical Computing: Vienna, Austria, 2011.
(36) Lockless, S. W.; Muir, T. W. Proc. Natl. Acad. Sci. USA 2009, 106, 10999-1004.
(37) Shah, N. H.; Vila-Perelló, M.; Muir, T. W. Angew. Chem. Int. Ed. 2011, 50, 6511-5.
(38) Stothard, P. BioTechniques 2000, 28, 1102, 1104.
(39) Bryksin, A. V.; Matsumura, I. BioTechniques 2010, 48, 463-5.
(40) Reid, B. P.; BPReid: Hanover, N. H., 2009.
(41) Schwarzer, D.; Ludwig, C.; Thiel, I. V.; Mootz, H. D. Biochemistry 2012, 51, 233-42.
(42) Sun, P.; Ye, S.; Ferrandon, S.; Evans, T. C.; Xu, M.-Q.; Rao, Z. J. Mol. Biol. 2005, 353, 1093-105

Claims

1. A fusion protein comprising a split intein N-fragment and a polypeptide, wherein the split intein N-fragment comprises a sequence selected from the group consisting of SEQ ID NOs: 1-19, or a variant thereof.

2. The fusion protein of claim 1, wherein the split intein N-fragment comprises a sequence of SEQ ID NO: 19, or a variant thereof.

3. The fusion protein of claim 1, wherein the split intein N-fragment comprises a sequence selected from the group consisting of SEQ ID NO: 20-38 and 761, or a variant thereof.

4. The fusion protein of claim 1, wherein the polypeptide has a molecular weight of 40 kDa or greater.

5. The fusion protein of claim 4, wherein the polypeptide is an antibody or fragment thereof.

6. The fusion protein of claim 5 wherein the antibody is an IgG antibody or fragment thereof.

7. The fusion protein of claim 6, wherein a heavy chain of the antibody is fused to the split intein N-fragment.

8. The fusion protein of claim 7, wherein each heavy chain of the antibody is fused to the split intein N-fragment.

9. The fusion protein of claim 6, wherein a light chain of the antibody is fused to the split intein N-fragment.

10. The fusion protein of claim 1, wherein the polypeptide is secreted from a cell.

11. A complex comprising the fusion protein of claim 1 and the split intein C-fragment of claim 12.

12. A split intein C-fragment comprising a sequence selected from the group consisting of SEQ ID NOs: 57-74, or a variant thereof.

13. The split intein C-fragment of claim 12 comprising a sequence selected from the group consisting of SEQ ID NOs: 75-128, or a variant thereof.

14. The split intein C-fragment of claim 12 comprising a sequence of SEQ ID NO: 707, or a variant thereof.

15. The split intein C-fragment of claim 14 comprising a sequence selected from the group consisting of SEQ ID NOs: 708-711, or a variant thereof.

16. The split intein C-fragment of claim 12 attached to a support.

17. The split intein C-fragment of claim 16, wherein the support comprises a particle, a bead, a resin, or a slide.

18. The split intein C-fragment of claim 17, wherein the split intein C-fragment comprises a sequence selected from the group consisting of SEQ ID NOs: 417-704 and 745-759, or a variant thereof.

19. A method comprising

(a) contacting (1) a fusion protein comprising a polypeptide and a split intein N-fragment comprising a sequence selected from the group consisting of SEQ ID NO: 1-19, or a variant thereof, and (2) a split intein C-fragment comprising a sequence selected from the group consisting of SEQ ID NO: 57-74, or a variant thereof, wherein contacting is performed under conditions that permit binding of the split intein N-fragment to the split intein C-fragment to form an intein intermediate; and

(b) contacting the intein intermediate with a nucleophile to form a conjugate of the protein and the nucleophile.

20-34. (canceled)