Patent Grant
7348408
This is a National Stage entry under 35 U.S.C. §371 of PCT Application No. PCT/GB03/000078 filed Jan. 10, 2003, published in English.
The present invention relates to fusion proteins (fusion polypeptides), particularly for use in expression and/or purification systems.
Purified proteins are required for several applications. However, the isolation of pure proteins, in sufficient quantities, is sometimes problematic. For protein function studies, large amounts of a protein of interest (for example, a mutated protein) are often needed. Various expression systems have been used for heterologous production of proteins. Escherichia coli (E. coli) is still the most common host despite huge advances in the area of protein expression in the last ten years in other hosts. E. coli is popular because expressing proteins in the bacterium is relatively simple and a vast amount of knowledge about bacterium itself exists, and (sometimes most importantly) because of the low costs associated with production.
Proteins can be expressed in E. coli either directly or as fusions (of a “fusion partner” and a protein or polypeptide), also known as fusion proteins. The purpose of fusion partners is to provide affinity tags (e.g. Hisn tag, glutathione-S-transferase, cellulose binding domain, intein tags), to make proteins more soluble (e.g. glutathione-S-transferase), to enable formation of disulphide bonds (e.g. thioredoxin), or to export fused proteins to the periplasm where conditions for the formation of disulphide bonds are more favourable (e.g. DsbA and DsbC). Proteins used as fusion partners are normally small (less than 30 kDa).
TolA is a periplasmic protein involved in (1) maintaining the integrity of the inner membrane and (2) the uptake of colicins and bacteriophages. The first function is evidenced by the increased outer membrane instability (e.g. SDS sensitivity) of TolA mutants. This function has been shown by various authors and may depend upon the interaction with the TolB protein (Levengood-Freyermuth et al., 1993, J. Bacteriol. 175: 222-228; Wan & Baneyx, 1998, Protein Expression & Purification 14: 13-22). Wan and Banex (1998, supra) have demonstrated that co-expression of the C-terminal TolAIII domain of TolA (see below) facilitates the recovery of periplasmic recombinant proteins into the growth medium of E. coli, confirming that overproduction of the TolAIII domain disrupts the outer membrane and causes periplasmic proteins to leach into the growth medium.
The second function of TolA is based upon the use of TolA as a receptor by phage proteins (Lubkowski, J. et al., 1999, Structure With Folding & Design 7: 711-722) and colicins (Gokce, I. et al., 2000, J. Mol. Biol. 304: 621-632). This has been revealed both by the phage/colicin resistance of to1A mutants and by direct demonstration of the to1A -protein interactions by physical methods. TolA is composed of three domains. A short N-terminal domain is composed of a single transmembrane helix, which anchors TolA in the inner membrane. The second, largest domain is polar and mainly α-helical. A C-terminal domain III (TolAIII) is small and composed of 92 amino acids. Its 3D structure was recently solved in a complex with N1 domain of minor coat gene 3 protein of Ff filamentous bacteriophage (Holliger, P. et al., 1999, J. Mol. Biol. 288: 649-657). It is tightly folded into a slightly elongated protein with the aid of one disulphide bond (
Lubkowski et al. (1999; supra) disclose a fusion protein comprising residues 1-86 (the N1 domain) of the filamentous Ff bacteriophage minor coat gene 3 protein g3p towards the N-terminus and residues 295-425 (including the TolAIII domain) of TolA, a coreceptor of g3p, towards the C-terminus, and a C-terminal Ala3His6 (SEQ ID NO: 1) tail. The fusion protein was used by Lubkowski et al. to elucidate the crystal structure of a complex formed between the g3p N1 and TolAIII domains.
Various homologues of the TolA protein are known, for example from E. coli (SwissProt Acc. No. P19934), Salmonella species (for example Genbank Acc. Nos gi16764117 and gi1675986, Pectobacterium species (for example Genbank Acc. No. gi16116636) and Haemophilus species (for example Genbank Acc. No. gi2126342).
The present inventors have found that the TolAIII domain has remarkable properties which are of particular use as a fusion protein partner to achieve high levels of expression in a host cell.
According to the present invention, there is provided a fusion polypeptide for expression in a host cell comprising a TolAIII domain or a functional homologue, fragment, or derivative thereof and a non-TolA polypeptide, wherein the TolAIII domain or functional homologue, fragment, or derivative thereof is located towards the N-terminus of the fusion polypeptide and the non-TolA polypeptide is located towards the C-terminus of the fusion polypeptide.
As used herein, the terms “polypeptide” and “protein” are synonymous and refer to a sequence of two or more linked amino acid residues.
The TolAIII domain, when located towards the N-terminus of a fusion polypeptide, has been shown by the present inventors to facilitate higher than expected levels of the TolAIII fusion polypeptide expression in a host cell. The TolAIII domain fusions will be useful, for example, for obtaining purified protein and polypeptide partners and/or for studying the properties of these partners.
The fusion polypeptide may further comprise a signal peptide. This will allow the fusion polypeptide to be targeted to a specific intra- or extra-cellular location. The signal peptide may be located at or near the N-terminus of the fusion polypeptide. The signal peptide may be cleaved from the fusion polypeptide during the targetting process.
If the fusion polypeptide has the basic structure: N terminus-TolAIII-Protein partner-C terminus, it may be expected that it will be expressed in high yields in the cytoplasm. If, however, the fusion polypeptide has the basic structure: N terminus-Signal peptide-TolAIII-Protein partner-C terminus, the signal peptide may be used to target the construct to a non-cytoplasmic location. For example, in E. coli expression systems the ribose-binding-protein signal peptide (for example, the E. coli ribose-binding-protein signal peptide [SEQ ED NO: 2]) may be used to target a fusion protein to the periplasm. Signal peptides which may be suitable for use in the present invention conform to a set of general rules which are described in Von Heijne, G. 1985, J. Mol. Biol. 184 (1): 99-105.
The TolAIII domain or functional homologue, fragment, or derivative thereof may be codon-optimised for expression in the host cell.
The fusion polypeptide may further comprise a linker between the TolAIII domain or functional homologue, fragment, or derivative thereof and the non-TolA polypeptide. The linker may provide a physical separation between the TolAIII domain or functional homologue, fragment, or derivative thereof and the non-TolA polypeptide or may be functional. The linker may comprise at least one cleavage site for an endopeptidase. For example, the cleavage site may comprise the amino acid sequence DDDDK (SEQ ID NO: 3; for enterokinase) and/or LVPR (SEQ ID NO: 4; for thrombin) and/or IEGR (SEQ ID NO: 5; for factor Xa).
In one embodiment, the fusion polypeptide according to invention may further comprise an affinity purification tag. The affinity purification tag may be located at or near the N-terminus of the fusion polypeptide. For example, the affinity purification tag is an N-terminal Hisn tag, with n=4, 5, 6, 7, 8, 9 or 10 (SEQ ID NOs: 6-12, respectively; preferably n=6 [SEQ ID NO: 8]), optionally with the Hisn tag linked to the fusion polypeptide by one or more Ser residues (preferably two). The affinity purification tag will provide one means for immobilising the fusion polypeptide, for example as a step in purification.
In one embodiment, the fusion polypeptide comprises a signal peptide at the N-terminus and an affinity purification tag near the N-terminus. If the signal peptide is cleaved from the fusion polypeptide during targeting, then the affinity purification tag may be located at or nearer to the new N-terminus of the fusion protein.
Preferably, the TolAIII domain consists of amino acid residues 329-421 (SEQ ID NO: 13) of Escherichia coli TolA (SwissProt Acc. No. P19934).
The host cell may be bacterial (for example, Escherichia coli).
The non-TolA polypeptide of the fusion polypeptide may be human BCL-XL (SEQ ID NO: 62, SWISSPROT Accession No. B47537). The fusion polypeptide with human BCL-XL (SEQ ID NO: 62) may comprise the amino acid sequence of SEQ ID NO: 14 or SEQ ID NO: 15. As shown in Example 2 below, large amounts of BCL-XL (SEQ ID NO: 62, an important protein in apoptosis and cancer research) can be generated by expression as a TolAIII fusion polypeptide.
Further provided according to the present invention is a DNA molecule encoding the fusion polypeptide as defined above. The mRNA properties of the DNA molecule when transcribed may be optimised for expression in the host cell.
Also provided is an expression vector comprising the DNA molecule as defined above for expression of the fusion polypeptide of the invention. The expression vector may have an inducible promoter (for example, the IPTG-inducible T7 promotor) which drives expression of the fusion polypeptide. The expression vector may also have an antibiotic resistance marker (for example, the bla gene, which confers resistance to ampicillin and chloramphenicol).
In another aspect of the invention there is provided a cloning vector for producing the expression vector as defined above, comprising DNA encoding the TolAIII domain or a functional homologue, fragment, or derivative thereof upstream or downstream from a cloning site which allows in-frame insertion of DNA encoding a non-TolA polypeptide. The cloning vector may further comprise DNA encoding at least one cleavage site (for example, the amino acid sequence DDDDK [SEQ ID NO: 3] and/or LVPR [SEQ ID NO: 4] and/or IEGR [SEQ ID NO: 5]) for an endopeptidase, the cleavage site located between the DNA encoding the TolAIII domain or a functional homologue, fragment, or derivative thereof and the cloning site. The cloning site may comprise at least one restriction endonuclease (for example, BamHI and/or KpnI) target sequence. The cloning vector may further comprise DNA encoding an affinity purification tag as defined above. The cloning vector may further comprise an inducible promoter (for example, the IPTG-inducible T7 promotor) and/or DNA encoding an antibiotic resistance marker (for example, the bla gene, which confers resistance to ampicillin and chloramphenicol).
For example, the cloning vector may have the structure of pTolE, pTolT or pTolX (as shown in
Also provided is the use of the TolAIII domain or functional homologue, fragment, or derivative thereof for production of a fusion polypeptide as defined above.
Further provided is the use of the TolAIII domain or functional homologue, fragment, or derivative thereof for production of the DNA molecule as defined above.
Yet further provided is the use of the TolAIII domain or functional homologue, fragment, or derivative thereof for production of an expression vector as defined above.
Also provided is the use of the TolAIII domain or functional homologue, fragment, or derivative thereof for production of a cloning vector as defined above.
In one aspect there is provided a host cell containing the DNA as defined above and/or the expression vector as defined above and/or the cloning vector as defined above.
In another aspect there is provided the use of the fusion polypeptide as defined above for immobilisation of the non-TolA polypeptide, comprising the step of:
It is known that TolAIII interacts specifically with several naturally occurring proteins such as colicins, phage proteins and other Tolproteins. This range of existing binding partners makes the over expression of TolAIII fusion proteins of particular utility since these proteins may be used in purification or immobilisation technologies. The TolAIII domain therefore not only drives high expression of the fusion polypeptide but also provides an affinity tag for purification, immobilisation or analysis of the fusion polypeptide. The TolAIII binding proteins (or binding polypeptide domains thereof) could be used to provide binding sites for the TolAIII fusions (as in
Alternatively, the fusion polypeptide comprising an affinity tag as defined above may be used for immobilisation of the non-TolA polypeptide, comprising the step of:
Also provided is the use of the fusion polypeptide as defined above for purification and isolation of the non-TolA polypeptide, comprising the steps of:
In an alternative embodiment, the fusion polypeptide comprising an affinity tag may be used for purification and isolation of the non-TolA polypeptide, comprising the steps of:
The fusion polypeptide as disclosed herein may be used for studying interaction properties of the non-TolA polypeptide or the fusion polypeptide, for example self-interaction, interaction with another molecule, or interaction with a physical stimulus.
Also provided is a method for high expression of a polypeptide as a fusion polypeptide in a host cell, comprising the step of expressing the polypeptide as a fusion polypeptide as defined above in a host cell. Levels of expression of a polypeptide as a fusion protein defined herein will be high relative to levels of expression of a polypeptide not linked to the TolAIII domain.
The invention will be further described with reference to the accompanying figures. Of the figures:
In our laboratory we first prepared fusion proteins between domain III of periplasmic TolA protein (TolAIII) and T domain of colicin N. Huge amounts of fusion protein was isolated when TolAIII was at the N-terminus and T-domain at the C-terminus. On the other hand, when the colicin N domain was the N-terminal partner no expression of fusion protein was obtained.
Here we describe cloning of pTolvectors that use TolAIII as a fusion partner at the N-terminal part of expressed fusion protein. We show that levels of expression of various fusion proteins are around 20% of total bacterial proteins and we were able to purify 50-90 mg of fusions per 1 of bacterial broth. We prepared different components of colicin N by the use of this system.
In Example 1, several proteins were expressed using the system. These were different parts and domains of colicin N (TolA binding box (peptide of amino acids 40-76), deletion mutant of T-domain (Δ10) and R domain), representing prokaryotic proteins. Human phospholipase A2, pore-forming protein from sea anemone equinatoxin II, nucleotide binding domain 1 (NBD1) of human cystic fibrosis transmembrane conductance regulator (CFTR) and human mitochondrial pyruvate dehydrogenase kinase 2 (PDK2) were examples of eukaryotic proteins. Transmembrane proteins were represented by BcrC, a component of bacitracin resistance system from Bacillus licheniformnis, and transmembrane domain 1 (TM1) of human CFTR. The expression of BCL-X, an important protein in apoptosis and cancer research, as a TolAIII fusion polypeptide is shown in Example 2.
For Example 1, in all cases except for two membrane proteins the yields of fusion protein were higher than the individual proteins. The expression of small peptides and soluble proteins was consistently good. More difficult targets were also chosen. The membrane proteins did not express at all. The human PLA, PDK2 and equinatoxin expressed well but as in the case of the individual proteins much ends up as insoluble fraction. PLA has many SS bonds and PDK has consistently resisted soluble expression in other systems. The TolAIII was not able to overcome the insoluble behaviour of these fusion partners but their recovery from inclusion bodies is still possible. In Example 2, large amounts of BCL-XL were expressed.
Materials and Methods
Cloning of pTol Vectors:
The original vector used in cloning was a derivative of pET3c (Novagen) termed pET8c. The pET8c vector was constructed by adding to the pET3c vector nucleotides encoding methionine followed by six histidine and two serine residues downstream of the cloning site (Politou, A. S. et al., 1994, Biochemistry 33(15): 4730-4737). The pET8c vector was used for an expression of fusion between domain III of TolA (amino acids 329-421; SEQ ID NO: 13) protein and T domain of colicin N. It is T7 based expression vector with bla gene, providing ampicillin selection. The fusion protein contains a methionine followed by six histidines and two serines at the N-terminal part. This linker enables easy purification using Ni-chelate affinity chromatography. The fusion partners were linked together via BamHI site. The C-terminal end of the fusion was cloned via MluI site. The T-domain gene was removed from the vector by restricting it with BamHI and MluI. An adaptor sequence was then ligated into the vector. It was composed in such a way that it removed the BamHI site within the flexible linker, but introduced a new BamHI site just after the cleavage sequence for endopeptidases (
The linker between TolAIII and fused partner is, therefore, composed of flexible part (Gly-Gly-Gly-Ser; SEQ ID NO: 18) and cleavage sequence for endopeptidases (enterokinase, factor Xa or thrombin) (
Newly cloned vectors were named pTolE, pTolX, pTolT and they comprise cleavage sequences for enterokinase, factor Xa, and thrombin, respectively. Fusion partners used to test the system were cloned into the pTol vectors via BamHI and MluI sites. If the nucleic acid sequence coding for a particular protein contained internal BamHI site, a KpnI site was used instead. Nine different proteins were used to test the system (Table 1). Coding sequences were amplified by PCR. Reaction mixture contained (in 100 μl total volume): 10 μl of 10× reaction buffer supplied by the producer, 2 μl of 100 mM MgSO4, 4 μl of dNTP mix (200 μM final concentration), 100 pmol of each oligonucleotide, approximately 20 ng of target DNA and 1 Unit of Vent DNA polymerase (New England BioLabs). Target DNA was obtained either from DNA cloned into plasmids (e.g. colicin sequences were from the plasmid pCHAP4 [Pugsley, A. P., 1984, Mol. Microbiol. 1: 317-325], equinatoxin sequences were from an equinatoxin-containing plasmid described in Anderluh G. et al., 1996, Biochem. Biophys. Res. Commun. 220: 437-42, and BcrC sequences were from an BcrC-containing plasmid described in Podlesek, Z. et al., 1995, Mol. Microbiol. 16: 969-976) or via direct PCR or RT-PCR from the host organism. The resulting DNA was sequenced after cloning into pTol to ensure that it corresponded to precisely to the section of the published sequence shown in the table. Typically the following cycles were used: 10 min at 97° C.; 30 cycles, each composed of 2 min denaturation at 97° C., 1 min of annealing at 58° C., 1 min of extension at 72° C.; 7 min at 72° C. and soak at 10° C. PCR fragments were purified using commercial kits (Qiagen) and restricted by an appropriate restriction endonucleases. Restricted fragments were cloned into pre-cleaved pTol vector. The correct nucleotide sequence of the fusion protein was verified by sequencing.
aMr of fusion protein calculated from the sequence.
bRestriction site used for cloning at the N-terminal part of the fusion protein. In all cases C-terminal site used was MluI.
cRefSeq accession number.
dOligonucleotides to amplify the desired proteins were of the following sequences (all 5′-3′; see Table 1):
Expression of Proteins in E. coli
All proteins were expressed in an E. coli BL21(DE3)pLysE strain (from Novagen). The strain was transformed with plasmid and grown on LB plates with appropriate selection (Ampicillin, Chloramphenicol). One colony was used to inoculate 5 ml of LBAC medium (Ampicillin at 100 μg/ml, Chloramphenicol at 34 μg/ml, both from SIGMA). Bacteria were grown on rotating wheel at 37° C. After 60 min the expression of recombinant proteins was induced by an addition of 1 mM (final) IPTG and bacteria were grown for additional 4 h. Small samples (corresponding to a volume of bacteria which when resuspended in 1 ml yields A600=0.5) was analysed on SDS-PAGE. Gels were stained with Coomassie and scanned at 600 dpi using commercial scanner. The amount of expressed proteins was estimated from the gels using the program Tina 2.0. For large-scale expression, 5 ml of bacterial culture in stationary phase was used to inoculate 250 ml of LBAC medium and grown at 37° C. in orbital shaker at 180 rpm overnight. The next morning 20-25 ml of overnight culture was used to inoculate 500 ml of M9 LBAC medium. In total 3-5 l of bacterial culture were grown for a single protein. Bacteria were grown at the same conditions until A600 reached approximately 0.8. Then the production of recombinant proteins was induced by adding IPTG to final 1 mM concentration. Bacteria were grown for additional 4-5 h, centrifuged for 5 min at 5000 rpm at 4° C., and stored at −20° C.
Isolation of Proteins from Bacteria
Pelleted bacteria were resuspended (2 ml of buffer/g of cells) in 50 mM NaH2PO4, pH 8.0, 300 mM NaCl, 10 mM imidazole, 20 mM β-mercaptoethanol (buffer A), with following enzymes and inhibitors of proteases (final concentrations): DNase (10 μg/ml), RNase (20 μg/ml), lysozyme (1 mg/ml of buffer), PMSF (0.5 nM), benzamidine (1 mM). They were incubated on ice for an hour and occasionally vigorously shaken. The resuspended bacteria were sonicated for 3 min with a Branson sonicator and then centrifuged in a Beckman ultra-centrifuge at 40000 rpm, 4° C. in 45ti rotor. Supernatant was removed and placed at 4° C. Pellet was resuspended in the same buffer without enzymes and inhibitors (1 ml/g of weight) and kept on ice for 15 min. Centrifugation at the same conditions followed after additional 1 min of sonication. Supernatants from both centrifugations were merged and applied at approximately 1 ml/min to 1-3 ml of Ni-NTA resin (Qiagen) equilibrated with buffer A. Typically, column with bound protein was washed with two fractions of 3 ml of buffer A, two fractions of buffer A with 20 mM imidazole and 6-10 fractions of buffer A with 300 mM imidazole. Fractions were analysed on SDS-PAGE. Fractions of interest were pooled and dialysed three times against water (5 l) at 4° C. Purity was checked by SDS-PAGE. Proteins were stored at 4° C. in 3 mM NaN3. Protein concentration was determined by using extinction coefficients calculated from the sequence.
Fractionation of Bacterial Proteins
All bacterial proteins were fractionated in order to see the amount of insoluble expressed proteins. Pelleted bacteria from 100 ml of broth were resuspended in 40 ml of 20% sucrose, 1 mM EDTA, 30 mM Tris-HCl, pH 8.0 and incubated 10 min at room temperature. They were centrifuged at 9000 g for 10 min at 4° C. Supernatant was removed and pellet was gently resuspended in 8 ml of ice-cold 5 mM MgSO4. Bacteria were gently shaken and incubated on ice for 10 min. Bacterial protoplasts were centrifuged again at the same conditions. Supernatant was removed as periplasmic fraction. Pellet was resuspended in 10 ml of 20 mM NaH2PO4, pH 8.0, with 1 mg of lysozyme and benzamidine. It was shaken vigorously and incubated on ice for 30 min, and finally, sonicated 5×30 s. Cytoplasmic proteins were removed from insoluble material by centrifugation at 35 000 g at 4° C. for 30 min. Supernatant was removed as cytoplasmic fraction and pellet was resuspended in 2 ml of 8 M urea, 10 mM Tris-HCl, pH 7.4, 0.5% Triton X-100 as insoluble fraction (membrane proteins and putative inclusion bodies).
Cleavage and Purification of TolAIII-R-Domain Colicin N Fusion
Pure R-domain of colicin N was produced using the pTol expression system. 45 mg of TolAIII-R-domain was incubated in 35 ml of cleavage mixture at 20° C. for 20 h. Cleavage mixture contains buffer as specified by producer and thrombin (Restriction grade, Novagen) at 0.1 U/mg of fused protein. Cleaved products were dialysed three times against 5 l of 40 mM Tris-HCl, pH8.4 at 4° C., each time at least 4 h. Cleaved R domain was separated from TolAIII and uncleaved fusion protein by ion-exchange chromatography on FPLC system (Pharmacia). Proteins were applied to Mono S column (Pharmacia) at 1 ml/min in 40 mM Tris-HCl, pH8.4. After unbound material was washed from the column, R-domain was eluted by applying gradient of NaCl from 0 to 500 mM in the same buffer in 30 min. Large peak at approximately 70% of NaCl (app. 350 mM) was collected and checked for purity by SDS-PAGE.
Cloning of pTol Vector
A DNA fragment encoding BCL-XL (SEQ ID NO: 62) was amplified by PCR from the plasmid pETBCLXL using the oligonucleotides SenseBCL-STU (5′-TTT TTT AGG CCT TCT CAG AGC AAC CGG GAG-3′; SEQ ID NO: 60) and Mlu-BCL-Rev (5′-TTT TAC GCG TTC ATT TCC GAC TGA AGA G-3′; SEQ ID NO: 61). BCL-XL (SEQ ID NO: 62) was introduced into pTOLT plasmid using StuI and MluI restriction sites. The final plasmid was named as a pTOLT-BCLXL (
Protein Purification
BCL-XL (SEQ ID NO: 62) protein was expressed in an E. coli BL21 DE3 (pLysE) strain. The strain was transformed with plasmid and grown on LB plates with ampicillin (200 μg/ml) and chloramphenicol (35 μg/ml) selection. 5 ml of LB medium with antibiotics was inoculated with single colony and grown overnight at 37° C. A 5 ml overnight culture was introduced into 500 ml of LB medium in 2 liter flasks containing ampicillin and chloramphenicol. Bacteria were grown until OD600: 0.8 and induced by addition of final concentration 1 mM IPTG then grown for additional 3 hours. Cells were harvested and resuspended in 20 mM phosphate, 300 mM NaCl, pH: 8.0 buffer containing RNAse, DNAse, PMSF (1 mM) and Benzamidine (1 mM). The cells were lysed by French press and the supernatant was obtained by ultra-centrifugation at 40 000 rpm for 1 h. The N-terminal 6× Histidine-tag (SEQ ID NO: 8) facilitated purification of the Tol-BCL fusion by means of Ni-NTA affinity column. The fusion protein was washed onto the column with 20 mM phosphate, 300 mM NaCl, pH: 8.0, buffer, additionally washed with the same buffer containing 50 mM imidazole and eluted in 300 mM imidazole, pH 7.0. The expression of fusion protein was analysed by SDS-PAGE (
Thrombin Cleavage of the BCL-XL Protein
20 mg of TolA-BCL fusion was incubated in 20 ml of cleavage buffer at 4° C. for 4 h. Cleavage buffer contains 50 mM Tris-HCl, 150 mM NaCl, 2.5 mM CaCl2, 5 mM DTT and Thrombin (1Unit of thrombin (Sigma)/mg of fused protein). The released protein was recovered applying overnight dialysed cleavage mixture to a Ni-NTA column. After unbound protein was washed from the column, remains of the BCL-XL (SEQ ID NO: 62) protein was washed by 2 M NaCl. All flow through and washes were collected and analysed by SDS-PAGE (
Results
Expression of TolAIII Protein in E. coli
In Example 1, the third domain of TolAIII with tags (
Expression of other Proteins in E. coli as Fusions with TolAIII
Ten proteins were tested in order to check the suitability of pTol expression system for expression and preparation of other proteins (see Example 1, Table 1, and Example 2). These were different parts and domains of colicin N (TolA binding box (peptide of amino acids 40-76), deletion mutant of T-domain (Δ10) and R domain), representing prokaryotic proteins. Human phospholipase A.sub.2, pore-forming protein from sea anemone equinatoxin II, nucleotide binding domain 1 (NBD1) of human cystic fibrosis transmembrane conductance regulator (CFTR), human mitochondrial pyruvate dehydrogenase kinase 2 (PDK2) and BCL-XL (SEQ ID NO: 62) were examples of eukaryotic proteins. Transmembrane proteins were represented by BcrC, a component of bacitracin resistance system from B. licheniformis, and transmembrane domain 1 (TM1) of human CFTR. Proteins chosen represent variations in size (app. 4.4 of colicin 40-76 kDa vs. 44 kDa of PDK2), genetic code (prokaryotic vs. eukaryotic proteins), protein location (soluble vs. membrane), and disulphide content (PLA2, 7 disulphides vs. equinatoxin, none). Fusion proteins were expressed at high proportion in E. coli using pTol system (
Isolation and Cleavage of Fusion Proteins
In Example 1, expressed fusions were isolated from the cytoplasm by simple extraction into buffered solution, which was applied onto Ni-NTA column. By this single step proteins were already more than 95% pure (
15N TolE-Tdomain 40-76
aProteins are named after plasmid used for expression of fusion protein.
Pure R-domain was prepared from TolT-Rdomain fusion by cleavage with thrombin and separation of cleavage products by ion-exchange chromatography. The results of such purification scheme are presented on
We show in Example 2 that BCL-XL (SEQ ID NO: 62), an important protein in apoptosis and cancer research, can be expressed in large quantities as a fusion with TolAIII (see
The first table lists values computed assuming ALL Cys residues appear as half cystines, whereas the second table assumes that NONE do.
The TolAIII domain was cleaved from the TolA-BCL fusion using thrombin and the BCL partner purified on a Ni-NTA column (
The first table lists values computed assuming ALL Cys residues appear as half cystines, whereas the second table assumes that NONE do.
Discussion
TolAIII is expressed in huge quantities in soluble form in bacterial cytoplasm. Among the reasons for high expression of proteins in E. coli are most commonly cited appropriate codon usage, stability of mRNA transcript, size, content of disulphide bonds, and non-toxicity to the cell. TolAIII is small protein, with only one disulphide bond. It is very stable and monomeric in solution even at concentrations as high as 30 mg/ml (data from analytical ultracentrifugation and gel filtration, not shown). The small size and tendency not to aggregate are certainly important in tolerance of heterologous material in the cytoplasm of bacteria A further advantage of TolAIII gene is, that it is bacterial protein and as such it possesses only 5 codons (4.7% of 106 amino acids excluding protease cleavage site) rarely transcribed in E. coli genome. They are scattered along the sequence. An improvement of its expression could be achieved by engineering of the conformation of its mRNA transcript. It was shown that, for a high yield of transcribed RNA, sometimes the conformation of RNA should be such, that the ribosome binding site and start codon should be exposed and not involved in base pairing. In the case of TolAIII mRNA both are involved in building short stems and not always completely exposed (analysis of transcribed RNAs of 60-120 nucleotides (step of 10 nt) by Mfold on http://bioinfo.math.rpi.edu/˜zukerm/). High expression of TolAIII protein in the T7 based vector and the high yields of pure product are comparable or even better than published and existing systems for production of fusion proteins in E. coli.
We have employed a domain of a periplasmic bacterial protein as a fusion partner in the overexpression of various proteins of bacterial and eukaryotic origin. Some small peptides or domains could be attached to TolAIII without significantly changing its size. The same amount of expressed protein would then be expected. In fact, the yield of fusion containing colicin N 40-76 peptide was the same as for TolAIII itself. The system is suitable for the preparation of eukaryotic proteins as well. In particular, the level of expression of EqII is much more improved over the published one. Approximately 20% of total expression of the fusion contrasted with approximately 5% in the case of direct expression. The majority of EqtII expressed from the pTol system is in the insoluble fraction, but isolation of the soluble cytoplasmic fraction still resulted in a large improvement in yield over the published method. The pTol system might also be applicable for proteins expressed as inclusion bodies. For example, the amount of expressed PLA2 is similar to other expression systems, however the fusion protein can easily be isolated by Ni-NTA chromatography and then refolded and cleaved on the column matrix. An interesting observation was that the two membrane proteins studied did not express as fusion proteins with pTolA system, although the reason for this is unclear at the moment.
Three expression vectors were constructed providing three different cleavage sites for endopeptidases widely used in molecular biology, e.g. enterokinase, factor Xa and thrombin. Recognition sites for endopeptidases differ in amino acid sequence and size. These differences dramatically change properties of the small TolAIII partner in fusion proteins (Table 3). TolAT and TolAX are basic, calculated pI more than 8.5, TolAE is acid in nature, pI of 6.6. This is the result of four aspartates in the recognition sequence for enterokinase (DDDDK; SEQ ID NO: 3). The constructed vectors thus enable higher flexibility, i.e. one can easily choose appropriate vector on the basis of the properties of fused partner. In our case, R-domain of colicin N was expressed in pTolT vector since R-domain is even more basic (pI 9.7) than cleaved TolAIII. On the other hand, colicin N peptide 40-76 has almost the same pI as TolAT or TolAX. This make subsequent purification much more difficult, the peaks representing the peptide and TolAIII would then overlap in ion-exchange chromatography. Therefore, peptide was expressed in pTolE. Cleaved TolAIII was not bound to the column at chosen conditions and the difference in pI of the uncleaved fusion (pI 7.2) and peptide was large enough to get clearly resolved peaks (not shown).
aProteins are named according to the vector in which they were produced.
bCalculated from the sequence.
We could produce functional parts of the colicin N toxin by using the pTol expression system. We produced functional R-domain and 39 residue peptide composed of colicin residues 40-76. His-tagged R-domain expresses poorly and irreproducibly and the to1A fusion expressed consistently well and improved the yield by more than two fold. Peptide was produced as 15N labelled sample for NMR structure determination. Preparation of large quantities of labelled peptide sample for NMR structure analysis can be problematic and a significant financial burden to research groups. High yields and versatility of the pTol system should make preparation of short peptides and proteins much cheaper and alternative to chemical synthesis and other expression systems. The system may be particularly useful for reproducible high level expression of small (<20 kDa) soluble proteins and unstructured peptides. For example, the system might prove useful in the preparation of 15N or 13C labelled small peptides for NMR structural studies.
The expression of BCL-XL, an important protein in apoptosis and cancer research, is difficult to express at high yield since it has a hydrophobic C-terminal region which causes instability and toxicity. Thus most structural work has been carried out on truncated versions lacking this region. We were unable to express this protein in satisfactory yields for structural studies and thus used the TolAIII fusion protein system to improve our yields. We can now express large amounts of this protein as a TolAIII fusion partner (
| Number | Date | Country | Kind |
|---|---|---|---|
| 0200689.8 | Jan 2002 | GB | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/GB03/00078 | 1/10/2003 | WO | 00 | 2/14/2005 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO03/057708 | 7/17/2003 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5087564 | Mai et al. | Feb 1992 | A |
| 20020137049 | Mark et al. | Sep 2002 | A1 |
| Number | Date | Country |
|---|---|---|
| 0299810 | Jan 1989 | EP |
| WO0121817 | Mar 2001 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20050130269 A1 | Jun 2005 | US |
