METHOD FOR PRODUCING SELENOPROTEINS

BACKGROUND OF THE INVENTION

In 1817 the Swedish chemist Jons Jakob Berzelius discovered selenium (Se) — a basic element that in recent years has gained much interest for its importance in human health and disease. Selenium is an essential trace element, and is a constituent of the 21 ^st amino acid, selenocysteine (Sec), found in selenoproteins. Humans have about 25 selenoprotein genes, many encoding enzymes with a single selenocysteine residue in their active sites, eg. thyroid hormone deiodinases, thioredoxin reductases and glutathione peroxidases.

Selenocysteine confers unique properties to a protein that are dependent on the selenium atom and which can be specifically exploited for biotechnological applications. The chemical features of selenium make it an ideal catalyst for redox reactions or as a nucleophilic handle for site-specific targeting with electrophilic

compounds. Selenocysteine is more reactive than the more common sulfur-containing cysteine (Cys) residue and can in many cases be regarded as a "super cysteine", which may explain the high activities of selenoenzymes compared to their Sec-to-Cys mutants. Selenium is essential for humans due to the vital roles of some selenoproteins, and low nutritional selenium status can cause cardiomyopathy, cancer or male infertility. High selenium intake may protect against cancer in part due to the antioxidant roles of some selenoproteins, including thioredoxin reductases. Further information regarding selenium can be found in references 1 to 7, which are incorporated herein by reference.

Up until now, there have been two major methods described for producing selenoproteins: recombinantly in E. coli or by chemical synthesis.

Selenocysteine is in nature cotranslationally (at the ribosome) inserted into growing selenoprotein polypeptide chains at predefined UGA "stop" codons. This necessitates an expansion of the genetic code, which is accomplished by an intricate synthesis machinery interacting with a stem-loop secondary structure in the selenoprotein mRNA

— a SECIS (Sec Insertion Sequence) element. In bacteria, the SECIS element is situated next to the selenocysteine-encoding UGA codon in the mRNA, while other forms of SECIS elements are found in the 3'-UTR of mammalian selenoprotein mRNAs. These qualitative differences make direct expression of a mammalian selenoprotein in E. coli virtually impossible.

A limitation of the known technology for recombinant selenoprotein production is the necessity to incorporate an engineered SECIS element in the open reading frame, which is not a problem for the production of thiodoxin reductase or Sel-tagged proteins having a selenocysteine close to the C-termini of the proteins, encoded by the 3'-end of their corresponding open reading frames. However, if used for a protein with an internal selenocysteine residue, this pre-defines a sequence of four to six amino acids located three to four residues C-terminally of the selenocysteine residue (1 , 4, 8, 9)

(see also US Patent Application No 09/639,778 and PCT/EPOO/07216 (WO 01/12657), which are incorporated herein by reference).

A scheme of these limitations is shown in Figure 1 (taken from reference 1). The variant SECIS elements that in theory could be used for targeted selenocysteine insertion carry constraints to the possible codon usage. The loop region encompassing 17 nucleotides must be preserved and should be positioned on the 3'-

side 11 nucleotides from the UGA codon. However, the stem can be lengthened by 3 bp and still maintaining 50% efficiency of selenocysteine insertion, and the actual nucleotide bases of the stem are not of functional importance. The figure shows variants of SECIS elements, conforming to functional restraints of the SECIS element in E. coli, that are proposed to be possible to use in attempts for targeting insertion of selenocysteine to an internal part of recombinant selenoproteins, with the alternative amino acid sequences resulting from these variants shown in three-letter codes. Using a SECIS element for production of a selenoprotein thus yields restrictions in the sequence of that selenoprotein.

Chemical synthesis can involve either conversion of reactive Ser residues (eg. in serine proteases) into selenocysteine, or make use of expressed protein ligation. The latter involves ligation of a recombinantly expressed protein with a synthetic C-terminal peptide containing selenocysteine. The disadvantage is a reliance on chemical synthesis of the C-terminal domain, limiting its use to production of selenoproteins carrying short domains at the C-terminal part of the selenocysteine residue (typically about 20 amino acids or less). The other alternative, i.e. use of a reactive serine residue, is restricted to few proteins and cannot be utilized for synthesis of native selenoproteins.

Thus, as a result of low expression levels, high chemical reactivity of selenocysteine and intricate translation machineries, few selenoproteins can be obtained in large amounts, making it difficult to conduct research in this field and to commercialise selenoprotein products. If selenoproteins could be produced at higher yields, and without restrictions with regard to their amino acid sequence, they would have commercial use and value in a range of applications in pharmaceutical and biomedical research and development.

The present invention discloses a new method of producing recombinant selenoproteins without restricting the position of the selenocysteine in the selenoprotein or the sequence of the subsequent amino acids in the selenoprotein.

SUMMARY OF THE INVENTION

According to a first embodiment of the invention, there is provided a method for producing a protein containing an internal selenocysteine residue, the method comprising the steps of: a) providing a first nucleic acid sequence with a UGA codon at or near its 3'- end, the nucleic acid sequence thereby encoding an N-terminal polypeptide half of the protein; b) positioning the first nucleic acid sequence at the 5'-end of a nucleic acid sequence encoding an intein; c) positioning a second nucleic acid sequence at the 3'-end of the intein nucleic acid sequence; d) introducing a functional derivative of a selenocysteine insertion sequence (SECIS) within the intein-encoding nucleic acid sequence, the SECIS decoding the UGA codon in the first nucleic acid sequence as a selenocysteine residue while allowing maintained intein splicing function on the translated protein level; e) allowing a precursor protein to be translated, the precursor protein containing a first polypeptide with a selenocysteine at or near its most C- terminal region, followed by an intein and a second polypeptide; and f) i) allowing the precursor protein to be processed by the removal of the intein with SECIS-encoded amino acids and the formation of a peptide bond between the C-terminal end of the first polypeptide and the N- terminal end of the second polypeptide, thereby forming the protein containing an internal selenocysteine residue; or ii) allowing the precursor protein to be processed by removal of the intein with SECIS loop-encoded amino acids, but preventing the formation of a bond between the first and second polypeptides and joining the C- terminal end of the first polypeptide to an N-terminal end of a third polypeptide to form the protein containing an internal selenocysteine residue.

For example, the first and second nucleic acid sequences may together encode the protein containing the selenocysteine residue, and the first and second polypeptides may be joined after the removal of the intein.

Alternatively, the first and third nucleic acid sequences may together encode the protein containing the selenocysteine residue, and the first and third polypeptides may be joined after the removal of the intein. The C-terminal end of the first polypeptide and the N-terminal end of the third polypeptide may be chemically modified after removal of the intein to facilitate peptide ligation of the first polypeptide to the third polypeptide, such as with a phosphonothiol or derivative thereof for the first polypeptide and an azido moiety or derivative thereof for the third polypeptide. The first and third polypeptides may be joined by peptide ligation using chemical means such as Staudinger ligation.

The precursor polypeptide and/or third polypeptide may be translated in a bacterium, such as E. coil

The SECIS may be inserted in the intein-encoding nucleic acid sequence corresponding to the N-terminal region of the intein.

The protein which is formed preferably does not contain any amino acids encoded from the SECIS apart from the UGA-encoded selenocysteine residue.

The nucleotide sequence of the SECIS may be selected from SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14 or SEQ ID NO: 16 or derivatives thereof, the functional prerequisite being a maintained compatibility with Sec-insertion at the bacterial ribosome and maintained intein splicing functionality.

According to a further embodiment of the invention, there is provided a DNA sequence comprising an intein-encoding sequence including a SECIS. The DNA sequence may comprise first and second nucleic acid sequences on either side of the intein sequence, respectively, wherein the first nucleic acid sequence typically has a UGA codon at or near its 3'-end and the first and second sequence together may encode a protein containing an internal selenocysteine residue.

According to a further embodiment of the invention, there is provided a vector including the DNA sequence described above.

According to a further embodiment of the invention, there is provided a host cell including the vector described above.

According to a further embodiment of the invention, there is provided a protein including an internal selenocysteine residue produced according to the method described above. The protein may be a naturally occurring protein or an artificial protein which does not include amino acids from the SECIS. The protein may be used for treating a disease or disorder, such as hyperthyroidism, cancer, muscle wasting or muscle disease.

According to a further embodiment of the invention, there is provided a tag comprising a protein formed by the method described above. The tag may be for use in protein purification, labeling (such as protein labeling, lipid labeling or glycoprotein labeling), development of catalysts or use as a structural component.

According to a further embodiment of the invention, there is provided the use of a protein described above in the manufacture of a medicament for use in a method for treating a disease or disorder selected from the group consisting of hyperthyroidism, cancer, muscle wasting and muscle disease.

According to a further embodiment of the invention, there is provided a method for treating a disease or disorder selected from the group consisting of hyperthyroidism, cancer, muscle wasting and muscle disease, the method comprising administering to a patient or organism a protein described above.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 : Amino acid constraints of SECIS elements designed for targeted insertion of selenocysteine to the internal part of recombinant selenoproteins to be produced in E. coli. Figure taken from reference 1.

Figure 2: Scheme of Sectein-based selenoprotein production. The UGA encodes selenocysteine guided by the SECIS element indicated by a hair-pin structure at mRNA level, located within an intein-based sequence (here labeled Sectein) that is subsequently removed by intein splicing, thereby resulting in fusion of the two extein parts of the final selenoprotein, with no trace of the originally required SECIS loop- encoded amino acids.

Figure 3: Alignment of pGPch-SECIS variants against pGPch-1 at the position where the SECIS element is introduced at both DNA level (top panel) and protein level (bottom panel). The pGPch-SECIS-1 contains the wild type fdhψ SECIS element (underlined) and UGA is introduced at the position highlighted in a darker colour in all four constructs, respectively. The splicing positions as well as other key elements of the construct are indicated in the protein alignment.

Figure 4: Self-splicing of PRP8 mini-intein. An SDS PAGE of whole bacterial lysate of BL21(DE3) harboring pGPch-1, with or without IPTG induction, is shown. Almost all precursor protein has spontaneously spliced out the mini-intein part within the bacteria, thus joining the two exteins and resulting in a mature protein product (in this case His- tagged GST). This result and experiment is based upon descriptions in reference 20.

Figure 5: Primary sequence (bottom of left panels) and mRNA secondary structures (left and right panels) of the four SECIS variants incorporated into Sectein constructs.

Figure 6: Splicing activity test. Lane 1 : control with no IPTG added; Lane 2: pGPch-SECIS-1 ; Lane 3: pGPch-SECIS-2; Lane 4: pGPch-SECIS-3; Lane 5: pGPch- SECIS-4; Lane 6, positive control of pGPch-1.

Figure 7: Western blot (left, HisProbe), SDS PAGE (middle) and ⁷⁵ Se- radioautography (right) with crude bacterial lysate using expression from plasmids pGPch-UGA 1 , 2, ^' 3 and 4. The autoradiography has been enhanced for higher molecular weights in lanes 4-6 in the right panel in order to visualize remaining minor amounts of 75Se-labeled precursor protein.

Figure 8: Scheme of production of selenoproteins with internal selenocysteine residues, utilizing recombinant selenoprorotein production in combination with chemical synthesis and Staudinger ligation.

Figure 9: The Sel-tag and its properties - the separate applications of a Sel-tag are schematically summarized, also indicating that dithiol (solely Cys-containing) versions are not as versatile, even if they may also serve the purpose for purification. (5)

Figure 10: Nucleic acid sequence of SECIS-1 in combination with an intein-based nucleic acid sequence for selenoprotein production.

Figure 11 : Amino acid sequence of SECIS-1 and an intein-based encoded protein sequence.

Figure 12: Nucleic acid sequence of SECIS-2 in combination with an intein-based nucleic acid sequence for selenoprotein production.

Figure 13: Amino acid sequence of SECIS-2 and an intein-based encoded protein sequence..

Figure 14: Nucleic acid sequence of SECIS-3 in combination with an intein-based nucleic acid sequence for selenoprotein production.

Figure 15: Amino acid sequence of SECIS-3 and an intein-based encoded protein sequence..

Figure 16: Nucleic acid sequence of SECIS-4 in combination with an intein-based nucleic acid sequence for selenoprotein production.

Figure 17: Amino acid sequence of SECIS-4 and an intein-based encoded protein sequence.

DETAILED DESCRIPTION OF THE INVENTION

The invention describes a method for the production of recombinant selenoproteins containing an internal selenocysteine residue without the presence of amino acids derived from the SECIS element.

In this description the terms "N-terminus" and "5'-end" and "C-terminus" and "3'-end" may at certain instances be used interchangeably although "N-terminus" and "C- terminus" regularly refers to regions of a protein while "5'-end" and "3'-end" refers to regions of a nucleic acid sequence.

It will be apparent to a person skilled in the art when reading this document that terms such as "polypeptide", "peptide" and "protein" as they apply to the invention are used

interchangeably herein, and refer to a sequence of amino acids. If such molecule contains selenocysteine, it may also be called a "selenoprotein".

According to the invention, a first nucleic acid sequence encoding a first part of a protein to be formed is positioned upstream of an intein, with a UGA codon (which will be translated into a selenocysteine reside) located at or near the 3'-end of the first nucleic acid sequence A SECIS sequence is then inserted into the 5'-end of the intein-encoding nucleic acid region with its design being dictated by i) maintained functionality of selenocysteine insertion at the UGA codon during bacterial translation, and ii) maintained intein splicing functionality of the mature protein with removal of restricted SECIS-encoded amino acids apart from the selenocysteine residue. A precursor protein is formed containing the first part of the protein, the intein (with the SECIS-encoded amino acids) and a second polypeptide.

In one embodiment of the invention, the intein (with SECIS-encoded amino acids) is excised from the precursor and the first and second polypeptides are joined by a peptide bond to form a protein with an internal selenocysteine residue, and the so- formed selenoprotein does not include any amino acids derived from the SECIS apart from the selenocysteine residue.

In an alternative embodiment of the invention, after excision of the intein, the first polypeptide is ligated to a third polypeptide (i.e. a polypeptide which was not part of the precursor protein) to form a protein with an internal selenocysteine residue, and the protein does not include any amino acids derived from the SECIS apart from the selenocysteine residue.

lnteins are polypeptide segments of a precursor protein that are able to excise themselves and rejoin the remaining parts of the protein (called N-extein and C-extein) with a peptide bond (Figure 3; (11)). Protein splicing is a post-translational event, releasing an internal protein fraction (intein) from the protein precursor, thereby joining two separate parts (exteins) flanking the intein. The splicing depends upon inherent features of the intein, self-catalyzing its excision and the extein ligation (10). In previously known production of selenoproteins, intein technology has been used within the so-called expressed protein ligation method, which necessitates that the selenocysteine-containing half of the selenoprotein is produced chemically and subsequently ligated to the spliced intein precursor (18, 19).

However, the present innovation describes the use of intein technology for the removal of the SECIS-encoded amino acids by extein excision, resulting in the ribosomal production of selenoproteins expressed in a host organism, for example E. coli, but without any trace of SECIS-encoded residues (and hence different amino acids) in the final product.

The present invention is further described by the following examples. Such examples, however, are not to be construed as limiting in any way either the spirit or scope of the invention.

Example 1 : Intein-quided removal of SECIS-encoded amino acids - Sectein approach

In order to construct a precursor protein, the expression of which includes both a bacterial SECIS element and an intein, herein named a "Sectein" construct, specific care must be taken to satisfy two functional criteria. The sectein should have a dual function: firstly, the mRNA must contain a functional SECIS element that directs selenocysteine insertion into the growing polypeptide during E. coli translation; and secondly, the subsequently translated Sectein precursor protein must be spliced due to maintained functional intein properties, without removal of the selenocysteine residue which is encoded by assistance of the SECIS element. This approach for recombinant protein production results in the production of a selenoprotein without any residual

SECIS-encoded amino acids, other than the selenocysteine residue. See Figure 2 for a schematic view of the invention.

Sectein construction and expression

Requirements: β Expression plasmid for intein expression; β Construction of an open reading frame encoding two halves of a selenoprotein to be expressed, positioned on either side of the intein, with the UGA for selenocysteine positioned at a site in the nucleic acid corresponding to the C-terminal end of the N-terminal half of the selenoprotein flanking the intein (Sectein); and « Introduction of a bacterial-type SECIS element within the nucleic acid sequence corresponding to the N-terminal region of the intein that decodes

the UGA in the N-terminal extein as selenocysteine without disrupting the intein self-cleavage capacity.

Construction of pGPch-SECIS and pGPch-UGA variants

cDNA of P. chrysogenum PRP8 mini-intein contained in plasmid pGPch-1 (20), obtained from Dr Stefanie Poggeler, was used. Using pGPch-1 as a template, a minimal SECIS element (loop region of the wild type fdhF SECIS element, or its variants) was introduced into the plasmid as indicated in Figure 3 by PCR using the primers listed below. The PCR (Finnzyme Phusion HF PCR master mix) was performed on pGPch-1 , resulting in a completely linearized plasmid containing the mutations. This PCR product was subsequently purified by Gel extraction (Qiagen Gel Extraction kit), digested with Eco31 l (Fermentas), ligated with T4 ligase (Fermentas), and transformed into competent E.coli BL21 (DE3) cell. Mutations were confirmed by DNA sequencing (MWG biotech). SDS PAGE using the bacterial lysates of the four variants showed that pGPch-3 and pGPch-4 retained the splicing activity (see below).

Forward primers (mutations bolded, restriction site underlined): pGPch-SECIS-s1 : 5'- GGAATGGTCTCTGTCTGGTTGCAGGTCTGCACCTCTTGCGATGCGATGGAACCG (SEQ ID NO: 1 ) pGPch-SECIS-s2: 5'- GGAATGGTCTCTGTCTGGCTAAAGGTCTTTGCCTCTTGCGATGCGATGGAACCG (SEQ ID NO: 2) pGPch-SECIS-s3: 5'-

GGAATGGTCTCTGTCTGGCTAAGGGTACTTGCCTCTTGCGATGCGATGGAACCG (SEQ ID NO: 3) pGPch-SECIS-s4: 5'-

GGAATGGTCTCTGTCTGGCTAAGGGTACTCGCCTCTTGCGATGCGATGGAACCG (SEQ ID NO: 4)

Reverse Primer (same for all four PCRs, restriction underlined) pGPch-SECIS-as: 5'-GGAATGGTCTCGAGACACGCTTTCTCCCAGAAGG (SEQ ID NO: 5)

A UGA codon at 11 nucleotides upstream of the SECIS element was then introduced using the same strategy as described above, using the primers listed below. After confirmation of the mutation by DNA sequencing, these pGPch-UGA plasmid variants were transformed into BL21 (DE3) harboring pSUABC (1).

Forward primers (pGPch-SECIS-1 as template, TGA bolded) pGPTGA-s1 : δ'-GCAATGGTCTCCTTCTGGTGAAAAGCGTGTCTGGTTGCAG (SEQ ID NO: 6)

Forward primers (pGPch-SECIS-2 as template) pGPTGA-s2: δ'-GCAATGGTCTCCTTCTGGTGAAAAGCGTGTCTGGCTAAAG (SEQ ID NO: 7)

Forward primers (pGPch-SECIS-3 or 4 as template) pGPTGA-s3/4: S'-GCAATGGTCTCCTTCTGGTGAAAAGCGTGTCTGGCTAAGG (SEQ ID NO: 8)

Reverse primer (same for all four PCRs) pGPTGA-as: δ'-GCAATGGTCTCCAGAAGGATCCACGCGGAACC (SEQ ID NO: 9)

Selenium-75 incorporation

The BL21 (DE3) hosting pGPch-UGA and pSUABC was cultured in 2 ml LB medium from single colonies. After 3 hours of shaking at 37 ⁰ C, L-cysteine (100 μg/ml), selenite (5 μM) and 1 μCi [ ⁷⁵ Se]-selenite (Research Reactor Center, Missouri University, Columbia) was added into the culture, shaking was continued for another hour and then IPTG (1 mM) was added. The culture was incubated at room temperature with shaking overnight. SDS PAGE of the overnight culture was performed as described by the manufacturer (Invitrogen). The gel was stained and dried (Invitrogen gel drying system). The dried gel was stabilized in a pre-cleaned Phosphor screen (Molecular dynamics) or equivalent and exposed overnight and examined by a Phosphor Imager (Molecular dynamics) or equivalent.

Western blotting

All constructs contained a C-terminal His-tag which could be used for detection using Supersignal West HisProbe kit (Pierce). The western blotting image was developed by ChemiDoc XRS Gel Documentation system (Bio-Rad).

Production of human cvtosolic glutathione peroxidase 4 (GPx4)

Human cytosolic glutathione peroxidase 4 (GPx4) as a Sectein-produced selenoprotein using a P. chrysogenum PRP8 mini-intein (20) was produced according to the method described above.

The mini-intein is a group of inteins lacking a homing endonuclease gene (HEG) domain while retaining splicing activity solely depending upon the intein sequence. The plasmid (pGPch-1 ) coding for this protein is in the public domain and was obtained

from Dr. Stefanie Pόggeler. The self-splicing of the mini-intein occurs simultaneously with expression of the precursor protein (Figure 4).

To produce variants of a functional Sectein, thereby containing a UGA encoding selenocysteine at or near the end of the N-terminal extein, the sequence of a functional bacterial SECIS element and the property of intein splicing of the PRP8 mini-intein, four different constructs were made to exemplify the approach of the innovation (Figure 5), using pGPch-1 as the backbone vector.

In Figure 5 GPch-SECIS-1 (SEQ ID NOs 10 and 11 ) represents a wild type fdhF SECIS element and consequently introduces four mutations compared to the scaffold sequence of the PRP8 mini-intein. pGPch-SECIS-2 (SEQ ID NOs 12 and 13) and pGPch-SECIS-3 (SEQ ID NOs 14 and 15) maintain a similar secondary structure as the wild type SECIS but have two and one mutations, respectively, which have been introduced. pGPch-SECIS-4 (SEQ ID NOs 16 and 17) introduces no mutation as the PRP8 mini-intein is maintained, yet the secondary structure is slightly different from the wild type SECIS hairpin.

When the constructs of Figure 5 were tested for intein splicing capabilities in BL21 (DE3), this showed that pGPch-SECIS-1 and pGPch-SECIS-2 had completely lost their splicing activity while pGPch-SECIS-3 maintained partial splicing activity

(Figure 6). It appeared that pGPch-SECIS-4 behaved as the wild type pGPch-1 regarding intein splicing since there was no difference between these two constructs on the protein level for the intein portion of the protein. Importantly, however, pGPch- SECIS-3 maintained splicing since this construct could also encode Sec insertion if a

UGA was positioned at the end of the N-extein.

Next, the codon positioned 11 nucleotides upstream of the SECIS element in all four constructs was changed to "UGA" (to yield plasmids pGPch-UGA-1 , 2, 3, 4). In addition, an extra pSUABC plasmid was co-transformed with these constructs, respectively, for increased capacity of cotranslational selenocysteine insertion. Using these new constructs, a selenium-75 ( ⁷⁵ Se selenite addition to growth medium) incorporation test was subsequently performed and verified by Western blot analysis for probing a His-tag present in the N-terminal end of the N-extein (Figure 7). The data showed that pGPch-UGA-1 had maintained an efficient [ ⁷⁵ Se]-selenocysteine incorporation (Lane 2) even when pSUABC was not included, but no detectable splicing occurred with this construct. However, pGPch-UGA-2 and 3 showed low [ ⁷⁵ Se]-

selenocysteine incorporation, which was also supported by the Western blot, and, importantly, pGPch-UGA-3 also maintained splicing activity although the efficiency was low (Lane 5, Western blot data). These results demonstrate the feasibility of the Sectein concept. The applicants are not aware of any prior art which describes the combination of SECIS-encoded selenocysteine insertion with intein cleavage in the same protein, resulting in bacterial production of a selenoprotein with an internal selenocysteine residue and without traces of SECIS-encoded amino acid residues other than this selenocysteine residue.

Example 2: Intein-mediated production of selenoproteins in two halves combined with Staudinger ligation

As an alternative approach to selenoprotein production using solely the intein methodology described in example 1 , recombinant selenoprotein production in bacteria can also be combined with chemical modification and a Staudinger ligation reaction.

In this approach, intein technology is utilized for removal of SECIS-derived residues from the N-terminal half of the selenoprotein, without direct fusion with the C-terminal half of the selenoprotein as described in example 1. Instead, an N-azidoprotein derivative for the C-terminal half of the selenoprotein is used, e.g. such as that made by the subtilisilin-catalyzed formation, to thereby combine the technology for recombinant selenoprotein synthesis with a ligation of the two halves of the protein, producing a native selenoprotein having an internal selenocysteine residue with no trace of any other SECIS-associated amino acid. The steps of this synthesis (Figure 8) are described in more detail as follows:

1a) Using recombinant selenoprotein production of the N-terminal half of a selenoprotein, a construct is designed for subsequent removal of the portion of the protein C-terminal of the selenocysteine residue (containing the SECIS-encoded amino acids) using intein technology, such as that enabled by chitin binding for purification with the New England BioLab IMPACT system Qntein-Mediated Purification with an Affinity Chitin-binding Tag) (12). pTWIN vectors can be used, undergoing cleavage after a temperature or pH shift, which may be advantageous for purifying target selenoproteins, thus avoiding reducing agents.

1 b) Making a phosphonothiol derivative of the selenium-containing extein using (diphenyl-phophino)methanethiol will give high yield and avoid side reactions with other

nucleophiles (such as the selenocysteine) (13) although the selenocysteine residue may have to be shielded from other side reactions by the employment of a protective group.

2) Production of the C-terminal half of the selenoprotein in E. coli (i.e. initiating at the residue coming after the selenocysteine), using common recombinant protein production technology, is performed, preferably with a C-terminal tag for purification. Subsequently, this protein is converted to an N-terminal azidoprotein, e.g. by using the subtilisin-catalyzed method (14).

3) The two halves of the selenoprotein can thereafter be ligated using trace-less Staudinger ligation (13, 15).

4) The final synthetic selenoprotein ligation product can, for example, be analyzed with SDS-PAGE, mass spectrometry or enzymatic assays (e.g. for GPx activity when

GPx is the target protein) to reveal the extent of successful production and any presence and identity of potential side products.

The production of recombinant selenoproteins with internal selenocysteine residues (the "N-terminal half in Figure 8) is thereby made by use of bacterial-type SECIS elements, which are spliced away using the intein approach, whereupon the C-terminal half of the final selenoprotein is ligated to the N-terminal half using a Staudinger ligation. This combination of recombinant selenoprotein production with intein cleavage and Staudinger-based protein ligation has not been described in any prior art of which the applicant is aware.

High-yield site-directed introduction of selenocysteine into internal regions of proteins would be of substantial importance. First of all, many uncharacterized native selenoproteins of humans and other organisms could be produced in useful quantities for in-depth studies, such as, thioredoxin reductase, four glutathione peroxidase enzymes (classical GPxI , gastrointestinal GPx2, plasma GPx3, phospholipid hydroperoxide GPx4), iodothyronine deiodinase enzymes and selenoprotein P. Since several selenoproteins have functions of medical importance, producing large amounts of the proteins would allow them to be studied in detail. The following table provides some potential examples.

In addition to the production of native selenoproteins, the introduction of selenocysteine into proteins that normally do not contain selenocysteine would enable a range of biomedical applications due to the unique features of selenocysteine (8). Such methodologies include selenocysteine-targeted labeling, use of selenocysteine as protein-bound radical trap, helping to improve phase determination in X-ray crystallography, trapping of protein folding intermediates, development of novel selenocysteine-dependent catalysts for redox reactions, new peptide conjugation reactions, and more.

The method of the invention will also allow Sel-tag technology to be developed further. The first study of the Sel-tag was published in the launch issue of Nature Methods (6) and subsequently described in further detail in Nature Protocols (7) and in US Patent Application No. 09/639,778 and PCT/EPOO/07216 (WO 01/12657). To briefly summarize, the Sel-tag forms a redox active selenolthiol/selenenylsulfide motif. The selenolate of the reduced Sel-tag provides a "handle" which can be utilized for several applications, including residue-specific selenolate-targeted fluorescent labeling, or radiolabeling with either gamma emitters or positron emitters. It was also found that the selenolthiol motif could form the basis for one-step purification involving binding to phenyl arsine oxide sepharose. The selenenylsulfide of the Sel-tag in its oxidized state shields and thereby protects the otherwise reactive selenocysteine residue from

unspecific reactions with electrophilic compounds. Thus, the reactivity of the selenocysteine residue in a Sel-tag can be deliberately employed for targeting under controlled conditions, i.e. upon reduction with a reductant such as DTT, whereas under normal aerobic conditions, the Sel-tag exists in an inert oxidized state. Dithiol variants of the motif are not as versatile (Figure 9). The combined properties of the Sel-tag enable a range of new biotechnological applications, but current production methods necessitate that it is positioned at the very C-terminal end of a protein. Its use also has limitations because of the arsenic-based purification of a Sel-tag.

A Sel-tag for recombinant proteins has now been produced which mimics the naturally occurring C-terminal motif of mammalian thioredoxin reductase (-Gly-Cys-Sec-Gly- COOH). It was found that this motif could be coupled to virtually any recombinant protein expressed in E coli and it is expected that it could be used for novel biochemical and biomedical applications.

While the invention has been described in detail with respect to specific embodiments thereof, it will be appreciated by those skilled in the art that various alterations, modifications and other changes may be made to the invention without departing from the spirit and scope of the present invention. It is therefore intended that the claims cover or encompass all such modifications, alterations and/or changes.

References

1. Arner, E. S. J., Sarioglu, H., Lottspeich, F., Holmgren, A., and Bock, A. (1999) J. MoI. Biol. 292,

1003-16.

2. Rengby, O., Johansson, L., Carlson, L. A., Serini, E., Vlamis-Gardikas, A., Karsnas, P., and Arner, E. S. J. (2004) Appl. Environ. Microbiol. 70, 5159-67.

3. Jiang, Z., Arner, E. S., Mu, Y., Johansson, L, Shi, J., Zhao, S., Liu, S., Wang, R., Zhang, T., Yan, G., Liu, J., Shen, J., and Luo, G. (2004) Biochem Biophys Res Commun 321 , 94-101.

4. Arner, E. S. J. (2002) Methods Enzymol. 347, 226-35.

5. Cheng, Q., Johansson, L., Thorell, J.-O., Fredriksson, A., Samen, E., Stone-EIander, S., and Arner, E. S. J. (2006) ChemBioChem 7, 1976-81.

6. Johansson, L., Chen, C, Thorell, J.-O., Fredriksson, A., Stone-EIander, S., Gafvelin, G., and Arner, E. S. J. (2004) Nat Methods 1 , 61-66.

7. Cheng, Q., Stone-EIander, S., and Arner, E. S. J. (2006) Nat Protocols 1 , 604-13.

8. Johansson, L., Gafvelin, G., and Arner, E. S. J. (2005) Biochim Biophys Acta 1726, 1-13.

9. Johansson, L., and Arner, E. S. J. (2006) in "Selenium: its molecular biology and role in human health" (Hatfield, D. L., and Gladyshev, V. N., Eds.), pp. Chapter 20, Springer Science + Business Media, Norwell.

10. Xu, M. Q., and Evans, T. C, Jr. (2001) in "Methods", Vol. 24, pp. 257-77. 11. Xu, M. Q., and Evans, T. C, Jr. (2005) in "Curr Opin Biotechnol", Vol. 16, pp. 440-6.

12. Evans, T. C, Jr., Benner, J., and Xu, M. Q. (1999) in "J Biol Chem", Vol. 274, pp. 18359-63.

13. Soellner, M. B., Nilsson, B. L., and Raines, R. T. (2006) J Am Chem Soc 128, 8820-8.

14. Liu, L., Hong, Z. Y., and Wong, C. H. (2006) Chembiochem 7, 429-32.

15. Nilsson, B. L., Soellner, M. B., and Raines, R. T. (2005) Annu Rev Biophys Biomol Struct 34, 91- 118.

16. Rengby, O., and Arner, E. S. J. (2007) Appl Environ Microbiol 73, 432-41.

17. Johansson, L., Svensson, L., Bergstrδm, U., Jacobsson-Ekman, G., Arner, E. S. J., van Hage- Hamsten, M., Bucht, A., and Gafvelin, G. (2005) FEBS J 272, 3449-60.

18 Hondal RJ, Nilsson BL, Raines RT.(2001)J Am Chem Soc. 30;123(21 ):5140-1. 19. Hondal RJ.(2005) Protein Pept Lett.12(8):757-64. 20. Elleuche S, et al. BBRC 2008 366:239