WO1994027428A1

WO1994027428A1 - Post-transcriptional gene regulation by trace minerals

Info

Publication number: WO1994027428A1
Application number: PCT/US1994/005388
Authority: WO
Inventors: Jack L. Leonard; Peter E. Newburger
Original assignee: University Of Massachusetts Medical Center
Priority date: 1993-05-24
Filing date: 1994-05-16
Publication date: 1994-12-08
Also published as: EP0871722A1; JP2902118B2; EP0871722A4; JPH08509381A

Abstract

Methods of controlling the in vivo and in vitro expression of a heterologous protein which involve providing a cell containing a first nucleic acid encoding the heterologous polypeptide wherein at least one codon of the mRNA transcribed from the first nucleic acid has been replaced by the codon UGA operably linked to a second nucleic acid capable of directing the translation of the UGA codon as selenocysteine are described.

Description

POST-TRANSCRIPT ONAL GENE REGULATION BY TRACE MINERALS Background of the Invention The invention relates to post-transcriptional control of heterologous gene expression.

One of the major goals of the biotechnology industry is to stably transfect genes encoding proteins of medical and commercial value into cellular and animal systems under conditions which allow control of the expression of the transfected gene. To date, the methods most widely used involve controlling gene expression at the transcriptional level by placing the gene of interest under the control of an inducible promoter. However, the majority of currently available inducible promoters allow the production of significant background levels of gene expression under non-inducing conditions, thus limiting the usefulness of these methods only to applications where low levels of the transfected gene product do not have a significant effect on the cell. Furthermore, most inducible promoters are capable of producing only a limited increase in gene expression (usually about 3- fold) under inducing conditions.

In addition, current methods usually require transfection of the gene of interest into cells which lack the gene product, because the presence of native gene product often results in a signal-to-noise ratio that is difficult to evaluate, and prevents the use of functional assays to examine the transfected gene product. In some instances, when cells lacking the gene product are not available, proteins derived from transfected genes can be differentially identified from native cellular proteins by the addition of small amino acid sequences, termed "epitope tags" or "peptide flags". However, this approach is often limited to applications which do not require functional protein because the added peptide sequence may interfere with a number of processes, including protein folding, and post- translational processing, which are essential for the functional activity of the protein. Summary of the Invention

We have discovered that an alteration of a polynucleotide sequence which results in the substitution of one or more amino acids in a polypeptide by the novel amino acid, selenocysteine (SeCys) , provides a novel method of controlling gene expression at the level of translation. Accordingly, in one aspect the invention features a method of controlling the production of a heterologous polypeptide in a eukaryotic cell which includes the steps of (a) providing a cell containing a first nucleic acid encoding the heterologous polypeptide wherein at least one codon of the mRNA transcribed from the first nucleic acid has been replaced by the codon UGA, operably linked to a second nucleic acid which is capable of directing the translation of the UGA codon as selenocysteine; and (b) growing the cell under conditions wherein the production of the heterologous polypeptide is controlled by the level of selenium available to the cell.

The method of the invention may be carried out in vitro in any eukaryotic cell type which is capable of being maintained in cell culture. Preferably, the cell is a eukaryotic cell such as a mammalian tissue culture cell, (e.g., COS-1, HL-60, CV-1, C-6, LLC/PK-1, 3T3L1 or CHO cells) or a yeast cell, e.g., Saccharomyces cerevisiae . In one preferred embodiment, the cells used do not contain a native protein which is substantially homologous to the recombinant polypeptide. However, in those cases wherein such cells are not available, or have a substantial disadvantage over cells which do contain an homologous native protein, the recombinant polypeptide may be distinguished from the native protein by the increased reactivity of the recombinant polypeptide to nucleophilic reagents due to the presence of the selenocysteine residue, or alternatively, by radiolabeling with the radioisotope ⁷⁵Se.

The first and second nucleic acids may be introduced into and maintained in the cell in a recombinant vector which is capable of autonomously replicating in the cell, or stably integrated into the genome of the cell according to standard techniques. The production of the heterologous polypeptide is controlled by the amount of the trace mineral, selenium, in the medium in which the cells are cultured. When it is desirable to inhibit expression of the polypeptide, the cells are maintained in a medium which is substantially deficient in available selenium, i.e., the concentration of selenium in the medium is less than 1 ng/ml and preferably less than 0.1 ng/ml. To induce expression of the heterologous polypeptide, the cell culture medium typically contains between 1 and 50 ng/ml, preferably 2 to 40 ng/ml, and most preferably 5 to 25 ng/ml.

Alternatively, the method of the invention may be carried out in vivo by stably incorporating the first and second nucleic acids into the genome of an embryonal cell derived from a non-human mammal, and obtaining transgenic progeny of the non-human mammal. "Transgenic" as used herein means a mammal which includes a DNA sequence which is inserted by artifice into a cell and becomes part of the genome of the animal which develops from that cell. Such a transgene may be partly or entirely heterologous to the transgenic animal. Any non-human mammal which may be produced by transgenic technology is included in the invention; preferred mammals include, in addition to mice, rats, cows, pigs, sheep, goats, rabbits, guinea pigs, hamsters, and horses. "Embryonal cells" as used herein include embryonic stem (ES) cell and fertilized oocytes. In the case of fertilized oocytes, the preferred method of transgene introduction is by microinjection, whereas for ES cells, the preferred method is electroporation. However, other methods including viral delivery systems such as retroviral infection, or liposomal fusion can be used. After introduction of the transgene into an embryonal cell, the cell is introduced into pseudo-pregnant females and progeny is obtained which is heterozygous for the transgene. A stable line of heterozygous animals may then be maintained by appropriate backcrossing to the original animal line, or the heterozygous progeny may be mated to obtain homozygous animals. When it is desirable to inhibit the expression of the heterologous polypeptide, the transgenic animals are maintained on a diet containing less than 0.02 mg/kg of food, and induction of expression is triggered by supplementing the diet with selenium to a concentration 0.1 mg/kg or higher.

The polypeptide encoded by the first nucleic acid may be any desired polypeptide for which the nucleotide sequence is known. Methods of modifying the polypeptide to incorporate a selenocysteine amino acid residue are well known and described herein. Preferably, the selenocysteine residue is substituted for an amino acid at a position in the polypeptide which does not abolish the normal biological activity of the naturally occurring protein, e.g., a nonessential amino acid. Such amino acids may be identified by means well known to those skilled in the art and will usually occur at positions which are not involved in the catalytic or binding activity of the protein (as determined for example by mutational analysis) , or at positions which are considered critical for the structural integrity of the polypeptide (e.g., as predicted by computer analysis or crystallography) . Most often, the selenocysteine will be inserted at a position in the polypeptide which normally carries a cysteine residue. In preferred embodiments, the second nucleic acid includes a contiguous sequence of nucleotides capable of forming a stem-loop secondary structure in the mRNA transcribed from the second nucleic acid, wherein the stem-loop formed by the mRNA is capable of directing the translation of said UGA codon as selenocysteine. In one preferred embodiment, the second nucleic acid is derived from approximately 90 contiguous nucleotides from the 3' untranslated region of a gene encoding a naturally occurring mammalian selenoprotein. For example, the second nucleic acid comprises a nucleotide sequence substantially homologous to nucleotides 654 to 740 of the human selenoprotein, glutathione peroxidase, shown in Figure 8. In another preferred embodiment, the second nucleic acid is synthetically derived, and is capable of forming a stem-loop containing the sequence 5'-

NAAAUNNUAAAN-3' in the loop at the apex of the stem-loop, and the stem contains at least 12, and preferably at least 14, non-complementary nucleotides which form a bubble containing the sequence 5'-NUAGUN-3' symmetrically opposed on each half of the bubble; preferably, the stem- loop contains approximately 90 nucleotides, the bubble is placed approximately 17 nucleotides from the base of the stem, and the loop is placed approximately 11 nucleotides from the bubble at the apex of the stem-loop structure. Accordingly, the invention also features a single- stranded nucleic acid containing a contiguous stretch of nucleotides capable of forming a stem-loop secondary structure, wherein the loop of the stem-loop the sequence 5^,-NAAAUNNUAAAN-3^,, and the stem contains at least 12, and preferably at least 14, non-complementary nucleotides which form a bubble containing the sequence S-'NUAGUN-S' symmetrically opposed on each half of the bubble. The nucleic acid is capable of directing the translation of the codon, UGA, as selenocysteine when the nucleic acid is operably linked to an mRNA molecule which contains a UGA codon. Preferably, the nucleic acid of the stem-loop contains approximately 90 nucleotides, the bubble is placed approximately 17 nucleotides from the base of the stem and the loop is placed approximately 11 nucleotides from the bubble at the apex of the stem-loop structure. Also preferably, each half of the bubble is approximately 7 nucleotides and the loop contains approximately 12 nucleotides.

Also featured is a double-stranded nucleic acid which contains DNA encoding the single-stranded nucleic acid of the invention.

By "heterologous" nucleic acid is meant a nucleic acid which is partly or entirely foreign to the cell or animal in which it is introduced, or a nucleic acid which is homologous to an endogenous gene of the cell or animal with the exception that the heterologous protein contains selenocysteine substituted at least one amino acid.

By the term "operably linked" as used herein is meant that the contiguous stretch of nucleotides which form the stem-loop secondary structure is in sufficient proximity with the nucleic acid encoding the protein to allow translation of any UGA codon in the protein to be translated as selenocysteine. Preferably, the stem-loop is inserted in the 3' untranslated region of the mRNA molecule encoding the polypeptide; preferably within 2000 nucleotides of the UGA codon, more preferably within 400 to 1500 nucleotides, and most preferably within 500 to 1200 nucleotides. By "functionally active" is meant possessing any in vivo or in vitro activity which is characteristic of the naturally occurring protein.

"Homologous", as used herein, refers to the sequence similarity between two polypeptide molecules or two nucleic acid molecules. When a position in both of the two compared sequences is occupied by the same nucleotide base or amino acid subunit, then the molecules are homologous as that position. Thus, by "substantially homologous" is meant a nucleotide or amino acid sequence that is largely but not wholly homologous.

By "heterologous" nucleic acid is meant a nucleic acid which is partly or entirely foreign to the animal in which it is transfected, or a nucleic acid which is homologous to an endogenous gene of the transgenic animal, but which is inserted into the animal's genome at a location which differs from that of the natural gene.

Unless defined otherwise, all technical terms and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, the preferred methods and materials will now be described. All publications mentioned hereunder are incorporated by reference. In addition, the materials, methods and examples are illustrative only and not intended to be limiting.

The methods of the present invention provide several advantages over currently used methods of gene expression. First, the substitution of SeCys into a polypeptide sequence is absolutely dependent on the supply of selenium, thus allowing virtually absolute control of the amount of transfected gene product at the level of translation. Second, SeCys has all of the biological properties of the amino acid residue Cys, and thus substitution of SeCys for Cys does not result in a significant alteration in the normal biological activity of the transfected gene product. Third, a transfected gene product which contains SeCys can be readily distinguished from native cellular proteins via its heightened reactivity toward nucleophilic reagents, or by ⁷⁵Se incorporation.

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

Detailed Description The drawings will first be briefly described. Drawings Figure IA is a schematic diagram of the human cellular glutathione peroxidase cDNA constructs. The open reading frame (ORF) and 3'UTR are indicated by a wide bar; plasmid elements and 5'UTR are indicated by flanking lines. Nucleotide numbering starts at the beginning of the open reading frame; the ATG initiation codon is at nt 1-3, the TGA selenocysteine codon is at nt 142-144, and the TAG termination codon is at nt 607-609. Arrows indicate the positions of restriction endonuclease sites. Lines below the diagram represent the positions of the indicated deletions. The hatched bar below the diagram shows the position at which the epitope tagging sequence was inserted, and the region of cDNA replaced.

Figure IB is a schematic diagram of the potential secondary structure immediately downstream of the UGA₁₄₂ selenocysteine codon in the coding region of the human Gpx mRNA, and diagrams the positions of deletions ORF-D1 , ORF-D2, ORF-D3, and ORF-D4.

Figure 1C is a schematic diagram of an alternative potential secondary structure in the coding region of the human Gpx mRNA wherein the UGA₁₄₂ selenocysteine codon is within a hairpin structure. The deletion ORF-D5 is also indicated.

Figure 2A is an autoradiograph of an SDS- polyacrylamide gel of immunoprecipitated ⁷⁵Se-labelled COS-1 cell extracts after transfection with pCMV4 (lane

1), native GPx (lane 2), or deletion mutants 0RF-D1 through 0RF-D4 (lanes 3-6, respectively) .

Figure 2B is an autoradiograph of an SDS- polyacrylamide gel of immunoprecipitated ⁷⁵Se-labelled COS-1 cell extracts after transfection with pCMV4 vector

(lane 1) , native GPx (lane 2) , or deletion mutant 0RF-D5

(lane 3) .

Figure 3 is an autoradiograph of an SDS- polyacrylamide gel of immunoprecipitated ⁷⁵Se-labelled COs-1 cell extracts after transfection with pCMV4 vector

(lane 1) , epitope-tagged GPx (lane 2) , or deletion mutants UTR-D1 through UTR-D3 (lanes 3-5, respectively) . Figure 4 is an autoradiograph of an SDS- polyacryla ide gel of immunoprecipitated ⁷⁵Se-labelled COS-1 cell extracts after transfection with pCMV4 vector

(lane 1) , epitope-tagged GPx (lane 2) , deletion mutant

UTR-D4 (lane 3) , or deletion mutant UTR-D5 (lane 4) .

Figure 5 is a schematic diagram of the potential secondary structure of the 3'UTR of human Gpx mRNA. Figure 6 is an autoradiograph of a polyacrylamide gel of the products of an RNase protection assay using a labeled riboprobe. Lane 1, undigested probe; lane 2, probe hybridized with RNA from untransfected COS-1 cells; lane 3, probe hybridized with epitope-tagged GPx COS-1 transfectants; lane 4, probe hybridized with UTR-D4 COS-1 transfectants; lane 5, probe hybridized with UTR-D5 COS-1 transfectants.

Figure 7 is an autoradiograph of an SDS- polyacrylamide gel of immunoprecipitated ³⁵S-labeled (lanes 1-4) and ⁷⁵-Se-labeled (lanes 5-8) COS-1 cells transfected with rabδb opal mutants and fusion constructs. Lanes 1 and 5, pCMV4 vector; lanes 2 and 6, rab5b(opal)GPx3'UTR; lanes 3 and 7, rabδb(opal) ; lanes 4 and 8, rab5b(wt)GPx3'UTR. Figure 8 depicts the nucleotide sequence of the human glutathione peroxidase gene including the 3^,UTR.

Figure 9 depicts the sequence and secondary structure of the "optimized" selenocysteine insertion sequence (SECIS) . Selenocysteine

A small number of eukaryotic and prokaryotic proteins, including bacterial formate dehydrogenases, the mammalian glutathione peroxidase (GPx) family (Mullenbach et al.. Nucleic Acids ites.15;5484. 1987; Chambus et al., EMBO J. 5:1221, 1986; Esworthy et al.. Arch . Biochem . Biophyε . 286:330, 1991; Takahashi et al.. Blood 68: 640, 1986) , type I iodothyronine 5'deiodinase (Berry et al. (1991) Nature 349, 438-440) , and selenoprotein P (Read et al. (1990) J. Biol . Chem . 265, 17899-17905), belong to a unique group polypeptides which contain the unusual amino acid selenocysteine. The production of selenoproteins has been reported to be strictly regulated by the level of exogenous selenium. For example. Knight et al. (J. Nutr. 117:732, 1987) reported that glutathione peroxidase activity decreased to undetectable levels in rates given a selenium deficient diet (<0.02 ppm, 0.016mg/kg). Chanoine et al. (Endocrinology 131:1787. 1992) also reported that rats receiving a selenium deficient diet for six week had a significant decrease in both type I and type II 5'-deiodinase levels (<20% normal). Speier et al. (J. Biol . Chem. 260:8951. 1985) demonstrated that, in vitro, glutathione peroxidase activity depended on the medium selenium concentration of more than 1 ng/ml, with an optimal activity observed at 5 ng/ml sodium selenate (2.6 x 10~⁸ M) , whereas cells grown in medium without Se supplementation became glutathione peroxidase deficient, with only 1-3% of the activity of Se-supplemented cells. Chada et al. (Blood 74_:2535, 1989) and Chu et al. (Nucleic Acids Res . _1_8:1531, 1990) also reported a 30 to 50 fold difference in glutathione peroxidase activity between selenium deficient and selenium replete cells.

The control of selenoprotein production by exogenous selenium is believed to occur by post- transcriptional regulation by the incorporation of selenocysteine contranslationally at a UGA codon (Bock et al. (1991) Trends Biochem . Sci . 16, 463-467), which normally acts as a translational stop codon, through the utilization of a unique selenocysteine-charged tRNA containing the appropriate UCA anticodon (Hawkes et al. (1982) Biochim . Biophys. Acta 699, 183-191; Lee et al. (1989) J^". Biol . Chem . 264, 9724-9727). Thus, regulation of selenoproteins most likely proceeds by control of the translation process at the mRNA UGA codon. Selenium incorporated into a selenocysteyl-tRNA would allow translational read through whereas, in the absence of selenium, the selenocysteine tRNA would remain unacylated and the UGA codon would then function to terminate translation.

Since the first identification of the use of the UGA codon for selenocysteine incorporation, a critical question in the interpretation of this "extended genetic code" is how the ribosomal translation assembly can discriminate the special UGA codon in the open reading frame of a selenoprotein mRNA from the termination UGA codon in other mRNA species.

In order to identify the all of the elements necessary and sufficient to signal the translation of UGA as selenocysteine, we analyzed the functional importance of sequences from both the open reading frame and the 3'untranslated region (3'UTR) of the gene encoding the human selenoprotein, glutathione peroxidase, for selenocysteine incorporation in both glutathione peroxidase and in an unrelated non-selenoprotein. Construction of GPx and rab5b subclones GPx subclone GPxR in the vector pBluescript KS (Stratagene) was used as a common template for constructing all GPx deletion subclones. It was derived by inversion of the orientation of a GPX1 cDNA (Chu et al. (1990) Nucleic Acids Res . 18, 1531-1539) in the same vector. DNA sequencing of this clone (using standard dideoxy sequencing techniques with a "Sequenase" kit [US Biochemical]) showed one additional GCG codon immediately upstream of the previously reported codon 11 (GCC) (Mullenbach et al. (1987) Nucleic Acids Res . 15, 5484; Chada et al. (1990) Genomics 6, 268-271), and a codon 92 CTG as we have reported (Chada et al. (1990) Genomics 6, 268-271) , instead of the CAG observed by Mullenbach et al. (Mullenbach et al. (1987) Nucleic Acids Res . 15, 5484) (GenBank accession numbers Y00369 and M21304) . The former insertion is a polymorphism we have observed in other normal GPX1 sequences.

Unless otherwise indicated, GPx deletion subclones were constructed by overlap extension polymerase chain reaction (PCR) according to standard methods (Ho et al. (1989) Gene 77, 51-59), using a Perkin-Elmer Cetus thermal cycler and reagents. This PCR method required two flanking primers defining the size of the final product and two mutually complementary primers directing the desired mutation in the target sequence. The sequences of the flanking primers and of one of each pair of complementary mutagenesis primers are listed in Table 1. The final PCR products were inserted back into pBluescript KS, and the sequences were confirmed by standard methods. Then, each mutant GPx sequence was subcloned into the eukaryotic expression vector pCMV4 (Andersson et al. (1989) J. Biol . Chem . 264, 8222-8229) for transfection into COS-1 cells as described below.

TABLE 1. Primers used for overlap extension polymerase chain reactions for construction of GPx and rab5b subclones

Nucleotide Sequence (5'-»3'. Function

1) GGAAACAGCTATGACCAT flanking primers 2) GTAAAACGACGGCCAGTG for all GPx deletion subclones

3) AATGTGGCGTCCCTCTGAGACTACACCCAGATGAAC primers directing

4) TTACACCGCAGGGAGACTCTGATGTGGGTCTACTTG deletion in ORF-D1

5) AACGAGCTGCAGCGGCGCCTGGTGGTGCTCGGCTTC primers directing

6) TTGCTCGACGTCGCCGCGGACCACCACGAGCCGAAG deletion in ORF-D2

7) TGAGGCACCACGGTCCGGCGCCTCGGACCCCGG primers directing

8) ACTCCGTGGTGCCAGGCCGCGGAGCCTGGGGCC deletion in ORF-D3

9) CTCGGACCCCGGGGCCTGTTCCCGTGCAACCAG primers directing

10) GAGCCTGGGGCCCCGGACAAGGGCACGTTGGTC deletion in ORF-D4

11) ATCGAGAATGTGGCGTCCTGAGGCACCACGGTCCGG primers directing

12) TAGCTCTTACACCGCAGGACTCCGTGGTGCCAGGCC deletion in ORF-D5

13) ATGAGGGTGTTTCCTCCCTACGAGGGAGGAAC primers directing

14) TACTCCCACAAAGGAGGGATGCTCCCTCCTTG deletion in UTR-D4

15) ACGAGGGAGGAACACCCTTACAGAAAATACCA primers directing

16) TGCTCCCTCCTTGTGGGAATGTCTTTTATGGT deletion in UTR-D5

17) CGATAGCGCCATGTACCCATACGACGTCCCAGACTACGCTCGG primers for epitope

18) CTAGCCGAGCGTAGTCTGGGACGTCGTATGGGTACATGGCGCTAT sequence tagging

19) ATATATCGATATGACTAGCAGAAGCACAGC flanking primers

20) ATATATCCTAGGCACAGTTGCTACAACACTGGCTCTT for rab5b constructs

21) TTCCTCACCCAGTCCGTTTGACTAGATGACACAACAGTG primers directing an

22) AAGGAGTGGGTCAGGCAAACTGATCTACTGTGTTGTCAC opal mutation in rab5b

"Epitope tagging" of GPx was performed (as diagrammed in figure 1) by replacing the first 12 nucleotides (nt) of the open reading frame of GPx with a 30 nt sequence encoding an ATG start codon followed by 27 bases encoding a nine amino acid epitope of human influenza hemagglutinin protein (Chada et al. (1989) Blood 74, 2535-2541). The two oligonucleotides listed in Table 1 were annealed, then the resulting short double- stranded fragment was inserted into GPx wild-type or mutant subclones in pBluescript KS and/or pCMV4 via the Clal and Nhel restriction sites. In this process, amino acids 2 through 4 of GPx were deleted, producing an epitope-tagged "GPxEPI" possessing a net increase of six amino acid residues more than wild type GPx. Although rabbit antiserum against this epitope was available, its binding to tagged GPx molecule was much lower than that of the antisera against GPx peptide sequences, so the latter was still used to detect the tagged GPx molecule. Gpx subclones with a partial or complete deletion of the 3'UTR sequence were constructed by conventional DNA recombination techniques. In brief, the subclone UTR-D3, in which the entire GPx 3'UTR was deleted, was constructed by excision of a 250 nt Avrll-Spel fragment from the epitope-tagged GPx subclone GPxEPI in pBluescript KS, followed by religation of the remaining large fragment. Subclone UTR-D2 was constructed by excision of the Avrll-X ol fragment followed by religation of the remaining large fragment in the GPx 3'UTR sequence from GPxEPI-containing pBluexcript KS with the plasmid Xhoϊ site eliminated. The subclone UTR-D1 was obtained by inserting a GPxEPI containing fragment with a sticky Clal end and a end-filled Xhol end, excised from the construct GPxEPI in pBluescript KS, into the expression vector pCMV4 via the Clal and Smal polylinker restriction sites. The overlap extension PCT method was also used to construct mutant and fusion subclones of the rab5b gene, which encodes a member of Ras-related GTPase superfamily (Wilson et al. (1992) J. Clin . Invest . 89, 996-1005). The plasmid pMT2, carrying a 1.6 Kb rabδb cDNA clone, was obtained from D.B. Wilson (Harvard Medical School, Boston, MA). The construct rab5b(opal)GPx3'UTR contained a fusion product of the rab5b coding region with an opal (UGA) mutation at codon 63, fused with the GPx 3'UTR sequence. The oligonucleotide sequence of the flanking and mutagenesis primers are listed in Table 1. The 3'PCR flanking primer sequence resulted in the removal of the native rabSb TGA termination codon, and substitution of the last 3 codons of the GPx open reading frame, including its TAG stop codon. The resultant rabδb(opal) mutant was inserted into a pBluescript KS construct containing the entire GPx 3'UTR sequence derived from the Clal-Avrll double digestion of the native GPxR clone in pBluescript KS. The gene fusion product was then subcloned into pCMV4 as described above. The same strategy was also used to construct rab5b(WT)GPx3'UTR except, in this case, conventional PCR was applied using only the flanking primers, and the fusion product (WT, i.e. wild type without the opal mutation) was inserted into pCMV4. The construct rabδb(opal) which contains the coding region opal mutation but the native rab5b 3'UTR, was constructed by fusion of the approximately 900 nt N el-EσoRI fragment of the rabδb 3'UTR sequence with the rab5b(opal)Gpx3'UTR subclone, from which the GPx 3'UTR had been deleted as an Avrll-EcoKL fragment. The resulting rab5b(opal) sequence was then inserted into pCMV4 as above.

Transfection. labeling, and lysis of COS-1 cells COS-1 cells were transfected for transient expression of the GPx or rab5b constructs by modified calcium phosphate mediated or electroporation methods (Maniatis et al. (1990) Molecular Cloning: A Laboratory Manual Cold Spring Harbor Laboratory, Cold Spring Harbor) , and then cultured in DMEM medium supplemented with 10% fetal bovine serum, 5 ng/ml sodium selenite, 25 mM HEPES pH 7.4, and lx penicillin-streptomycin-fungizone (Gibco-BRL) . All experiments were performed 2-4 times. As a control for transfection efficiency, COS-1 cells were cotransfected with 2 μg of plasmid pXGH5 included in a human growth hormone transient expression assay system supplied by Nichols Institute. Human growth hormone secreted into the medium was detected by radiommunoassay using the Crystal Multidetector RIA System (United Technologies Packard) . For ⁷⁵Se labeling, 10 Ci of ⁷⁵Se as selenous acid diluted in nitric acid, with an original specific activity of 750-1000 Ci/g (from the University of Missouri Research Reactor Facility) , was added to the transfected cells in each plate, and the cells were incubated at 37°C for an additional 2 hours.

For ³⁵S labeling, the transfected cells in each plate were first incubated for 30 min in methionine- and glutamine-free DMEM medium (Gibco) , supplemented with 10% dialyzed calf serum, lx glutamine (Gibco) , and 25 mM HEPES. Then 250 Ci of Express ³⁵S protein labeling mix (NEN DuPont) , with a specific activity of 1140 Ci/mmole for methionine, was added to the plate, and the cells were incubated at 37°C for an additional 2 hours. After ⁷⁵Se or ³⁵S labeling, 5 or 1 μl (respectively) of diisopropylfluorophosphate was added to ice-cooled labeling mixture in the COS-1 cell plates. After 5 minutes, the mixture was aspirated and 1.5 ml of COS cell lysis buffer (50 mM HEPES pH 7.8, 1% triton X- 100, 10 mM EDTA, 1 mM phenylmethylsulfonyl chloride) was added to each plate. After shaking at 4°C for 20 min, the lysed cell suspension was transferred to a microfuge tube, and subjected to centrifugation at 14,000 x g for 10 min to remove cell debris. Sodium-dodecyl sulfate (SDS) was added to the supernatant to a final concentration of 0.5%, heated in boiling water for 5 minutes, and then cooled on ice. Immunoprecipitation and protein electrophoresis

Immunoprecipitation utilized two rabbit antisera raised (by Berkeley Antibody Co., Richmond, CA) against synthetic peptide sequences from the GPx polypeptide chain, one from residues 26 to 46, and the other from residue 174 to residue 192. Fifteen μl of each antiserum, plus 20 μ.1 of protein A-Sepharose CL-4B beads (Sigma) were added to each lysate, and the mixture was incubated at 4°C overnight with constant tumbling. The beads were subsequently pellet, washed twice with washing buffer (50 mM HEPES pH 7.8, 150 mM NaCl, 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS) and once with 50 mM HEPES at pH 7.8, mixed with 30 μl SDS-gel loading buffer (50 mM Tris-HCl pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol) , heated in boiling water for 3 minutes, and then pelleted in a microfuge. The supernatant was then collected for SDS-polyacrylamide gel electrophoresis (SDS-PAGE) . For COS-1 cells transfected with rab5b constructs, the procedure was the same as above except for the use of 0.2% SDS for cell lysis and the addition of 8 μl of affinity-purified rabbit antibody (obtained from D.B. Wilson) , raised against a synthetic peptide from the hypervariable domain of rabδb.

Protein electrophoresis was performed by standard techniques (Maniatis et al. (1990) Molecular Cloning: A Laboratory Manual Cold Spring Harbor Laboratory, Cold Spring Harbor) on 12% SDS-polyacrylamide gels. RNase protection assay

Total cell RNA was isolated by the guanidine-HCl method (Ginsburg et al. (1985) Science 228, 1401-1406). Riboprobes were generated from the T7 promoter by use of an RNA transcription kit (Stratagene) to synthesize a 224 nt ³²P-labelled RNA transcript complementary to a 179 nt segment starting at the Clal site of the 5'-untranslated region of the GpxEPI transcript. The template was a Clal fragment of a construct formed by recircularization of an end-filled Spel-RsrII large fragment of GpxEPI. RNase protection assays of hybridization mixtures of 3 μg total cell RNA, 10 μg yeast tRNA, and 6 μl of the riboprobe (400,000 TCA-precipitable cpm/μ-1) were performed by standard techniques (Maniatis et al. (1990) Molecular Cloning: A Laboratory Manual Cold Spring Harbor Laboratory, Cold Spring Harbor) .

Role of the nucleotide sequence around the UGA codon

We first explored the possibility that nucleotide sequences within the open reading frame of the GPx mRNA might serve as a signal for selenocysteine insertion. Sequence analysis of GPx MRNA has revealed no conserved sequences which are common to both prokaryotic and eukaryotic selenoprotein mRNAs, but has predicted two possible loop structures around the UGA₁₄₂ codon. (UGA_{1 2} refers to the codon starting at nucleotide 142 of the cDNA sequence of GPx; all nucleotide numbering for GPx starts at the first base of the open reading frame, as indicated in figure IA) . One putative stem-loop structure, immediately downstream of the UGA₁₄₂, creates a stem-loop structure (shown in figure 1, panel B) similar to that found in the mRNA of the E coli formate dehydrogenases and related prokaryotic selenoenzyme genes (Zinoni et al. (1990) Proc. Natl . Acad. Sci . USA 87, 4660-4664) . Another, which incorporates the UGA_{1 2} codon at the tip of the "hairpin" (shown in figure 1C) , is conserved among several mammalian GPx mRNAs, as well as E. coli formate dehydrogenase mRNA sequences (Chada et al. (1988) in Oxy-Radicals in Molecular Biology and Pathology (Cerutti, p., Fridovich, I., and McCord, J. , eds) pp. 273-288, Alan R. Liss Inc., New York). To test the role of each of these potential secondary structures in the direction of selenocysteine incorporation, we constructed a series of five sequential deletions in the GPx open reading frame, designated ORF-Dl through ORF-D5 which are shown on figure 1.

The first four deletion subclones are located within the putative stem-loop region immediately downstream of the UGA₁₄₂ codon. ORF-Dl lacks a sequence from codon 49 through codon 53; 0RF-D2 lack codons 65 through 69; ORF-D3 lacks codons 54 through 63; and ORF-D4 lacks codons 71 through 74. The sequences deleted from ORF-Dl, ORF-D2, and ORF-D3 correspond, respectively to the 5' part of the stem, the 3' part of the stem, and most (29 of 31 nt) of the loop of the putative stem-loop structure (Zinoni et al. (1990) Proσ. Natl . Acad. Sci . USA 87, 4660-4664) in the open reading frame region of GPx mRNA. ORF-D4 represents a 12 nt sequence immediately downstream of the putative stem-loop structure which corresponds to a sequence which has been reported to be important to selenocysteine translation in E. coli formate dehydrogenases (Zinoni et al. (1990) Proc. Natl . Acad . Sci . USA 87, 4660-4664). GPx subclone ORF-D5 contains a deletion of codon 47, located immediately upstream of the UGA_{1 2} codon of the GPx mRNA, which forms part of the stem of the alternative, putative hairpin loop structure (Chada et al. (1988) in Oxy-Radicals in Molecular Biology and Pathology (Cerutti, p., Fridovich, I., and McCord, J. , eds) pp. 273-288, Alan R. Liss Inc., New York) . These deletion subclones, carried by the eukaryotic expression vector pCMV4, were individually transfected into COS-1 cells and GPx expression was detected by ⁷⁵Se-labeling, immunoprecipitation, SDS- polyacrylamide gel electrophoresis, and autoradiography. As shown in figure 2, panel A, COS-1 cells transfected by the vector alone (lane 1) demonstrate a low background level of ⁷⁵Se-containing polypeptide (most likely the native monkey cellular GPx) with a 23 kD size similar to that of human GPx. Transient expression of the native human GPx cDNA and of deletions ORF-Dl through 0RF-D4 (lane 2 and lanes 3-6, respectively) all show high levels of ⁷⁵Se incorporation into Gpx protein. These deletions appeared to exhibit a slight, but not substantial, decrease in GPx expression. Repeated experiments (including the creation of identical deletions in the epitope-tagged construct) also showed slightly diminished expression. Similarly, as shown in figure 2B deletion 0RF-D5 produces little or no diminution of selenocysteine insertion into GPx. Thus, while the putative loop structures in the open reading frame of the GPx mRNA may slightly modulate GPx expression, neither is absolutely necessary for translation of the UGA₁₄₂ codon as selenocysteine in human GPx. Role of the 3'UTR

To test the role of the 3'UTR in selenocysteine insertion, we constructed GPx subclones containing deletions of various lengths in that region as described above. Epitope tags (Chada et al. (1989) Blood 74, 2535- 2541) were incorporated into these subclones in order to improve the resolution of the transiently expressed human GPx products from the COS-1 background. As diagrammed in figure IA, we replaced the first four codons of GPx with a 30 nt sequence encoding an ATG start codon and a 9 amino acid epitope of the human influenza hemagglutinin protein (Chada et al. (1989) Blood 74, 2535-2541). The unambiguous discrimination of the transiently expressed, epitope-tagged GPx was possible because the tagged GPx migrated slowly enough on SDS-PAGE gels that its band resolved at a position detectably higher than that of the untagged GPx. This difference of mobility permitted assessment of transient expression of transfected constructs without the need for the substantial overexpression necessary for evaluation of the coding region deletion constructs described above. The epitope sequence was also inserted into the wild type GPx subclone GPxR to yield a new GPx subclone GPxEPI, which served as a positive control for the transient expression of the GPx 3'UTR deletion constructs. These deletions are also indicated in figure IA.

The effects of three large deletions of the 3'UTR on [⁷⁵Se]-selenocysteine incorporation into GPx in transfected COS-1 cells are shown in figure 3. Lane 1 demonstrates the background GPx signal in cells transfected with vector alone. The slightly larger epitope-tagged GPx is expressed by the GPxEPI construct with its 3'UTR intact (lane 2) and is easily distinguished from the endogenous COS-1 background. Deletion of the distal 100 nt of the 3'UTR (UTR-D1, lane 3) did not diminish expression of the transfected GPx.

However, deletion of the proximal 129 nt (construct UTR- D2, lane 4) or the entire 3'UTR (construct UTR-D3, lane 5) completely eliminated detectable ⁷⁵Se incorporation into GPx. The distal and entire 3'UTR deletions did not result in a GPx mRNA without a 3'UTR, since the 3'UTR sequence of the human growth hormone gene is built into the pCMV4 vector, so as to fuse to the inserted sequence (if not separated by a transcription termination site, as in the other GPx constructs) . Computer analysis of either the entire GPx mRNA or its 3'UTR sequence using the FOLD program of the University of Wisconsin Computer Group software (Devereux et al. (1984) Nucleic Acids Res . 12, 387-395) revealed a potential secondary structure consisting of a long stem with two small loops (Figure 4) , similar to that found in the 3'UTR of rat and human 5'deiodinase and rat Gpx genes (Berry et al. (1991) Nature 353, 273-276). Moreover, two 4-nt sequences within the loop (UAAA in the first and UGAU in the second; indicated in figure 4) were identical to those at the same positions within the reported "selenocysteine insertion sequence" motif of the 5'deiodinase gene (Berry et al. (1991) Nature 353, 273- 276) . In order to test whether these two short sequences were necessary for selenocysteine insertion into GPx, we constructed two small deletion mutants, UTR-D4 and UTR-D5 (diagrammed in figure 1) , that specifically eliminated each of these sequences. The results shown in figure 5 demonstrate that either of these short deletions (lanes 3 and 4) completely abolishes detectable selenocysteine incorporation into epitope-tagged GPx in transfected COS- 1 cells.

In order to rule out the possibility that the deletion mutations affect the level of GPx transcripts, we measured levels of GPx mRNA in transfected COS-1 cells by RNase protection assays. As shown in figure 6, untransfected COS-1 cells (lane 2) contained no detectable mRNA by a riboprobe specific for the epitope- tagged GPx transcript, and COS-1 cells transfected with epitope-tagged wild-type GPx (lane 3) contain the same amount of transcript as those transfected with the UTR-D4 and UTR-D5 deletions (lanes 4 and 5) . Deletions in the open reading frame (ORF-Dl, -D2, and -D3) also produced no detectable change in GPx transcript levels (data not shown) . Transfection efficiency, assayed by cotransfection with a vector encoding human growth hormone, was also similar from group to group in these experiments (data not shown) . The GPx 3'UTR is sufficient for selenocysteine insertion Having demonstrated that sequences in the 3'UTR of GPx mRNA are necessary for translational insertion of selenocysteine, we next investigated whether this 3'UTR would be sufficient to direct the same process at a UGA codon in an unrelated coding sequence. The chosen target gene, rab5b, encodes a 25 kD GTP-binding protein which is a member of Ras-related GTPase superfamily (Wilson et al. (1992) J. Clin . Invest . 89, 996-1005). This gene was used for three constructs: rab5b(opal. had a codon 63 UGU (cysteine) modified to a UGA (opal) mutant, with the native rabδb 3'UTR; gene fusion construct rab5b (opa1)Gpx3'UTR consisted of the rabδb(opal) coding sequence fused to a 3' portion of GPx cDNA incorporating the last three codons of the GPx coding region, including its stop codon (UAG) , and the entire GPx 3'UTR; and rab5b(wt)Gpx3'UTR was also a rab5b-GPx fusion product but carried the wild type codon 63 rather than the opal mutation. The fusion constructs placed the UGU (cysteine) or UGA (potential selenocysteine) codon the same number of nt upstream from the GPx 3'UTR as in native GPx transcripts.

Figure 7 presents the results of a representative transient expression experiment of these constructs in COS-1 cells. The expression of rabδb was detected by an affinity-purified rabbit antibody against a synthetic peptide sequence, following either ³⁵S (lanes 1-4) or ⁷⁵Se (lanes 5-8) radioisotope labeling. COS-1 cells transfected with the vector alone (lanes 1 and 5) showed no detectable immunoreactive protein at the appropriate 25 kD molecular mass for rabδb. All three constructs directed the synthesis of an ³⁵S-labeled polypeptide of approximately 25 kD, at detectable but widely differing levels, but only rabδb(opal)GPx3'UTR, the fusion product of the rab5b with the opal mutation coupled to the GPx 3'UTR, incorporated ⁷⁵Se (lane 6). The rabδb(opal) transfectants expressed a very low level of ³⁵S-labelled protein, probably reflecting the existence of an alternative opal nonsense suppression mechanism (Hatfield, D. (1985) Trends Biochem . Sci 10, 201-204) in the COS-1 cells. No ⁷⁵Se was detectable even after very long exposures, indicating that no selenocysteine insertion occurred at the UGA codon in the presence of the rabδb 3'UTR. No truncated polypeptide was detectable on ³⁵S-labelled immunoprecipitates, suggesting that the short polypeptide product was unstable or not immunoreactive with the antiserum. Transfection with the rab5b(wt)GPx3'UTR fusion construct resulted in expression of immunoreactive protein without any detectable ⁷⁵Se incorporation, as expected for the wild type rabδb coding region. For these experiments, transfection efficiency was again confirmed by cotransfection with a vector encoding human growth hormone and measurement of secreted growth hormone.

The importance of the distance between the UGA codon and the selenocysteine-insertion sequence remains unkown. Our target for induced selenocysteine incorporation, rab5b, is similar in amino acid number to GPx, and the codon mutated to UGA is similar distance (550 nt) from the 3'UTR as the UGA₁₄₂ in GPx. However, in 5'deiodinase the span is approximately 1200 nt, so the precise distance between these elements necessary for selenocysteine incorporation is probably not critical.

These data demonstrate that small segments of the 3'UTR of the human GPx gene, specifically the conserved AAA and UGAU sequences within the potential stem-loop structure, and not the potential stem-loop or hairpin structures in the coding region, are essential for selenocysteine translation in human GPx. Moreover, these data also demonstrate that the GPx 3'UTR alone is sufficient to signal the translation, as selenocysteine, of an opal mutation (UGA) in the open reading frame of an unrelated non-selenoprotein, rab5b. Synthesis of "Optimized" Selenocysteine Insertion Sequence Examination of the genes encoding various known mammalian selenoproteins indicates that the 3'UTRs have little primary sequence similarity, but have similar potential stem-loop structures (Hill et al., J. Biol . Chem . 266:10050. 1991; Zinoni et al. Proc. Natl . Acad. Sci . 82:4660, 1990; Ho et al. Nucleic Acids Res . 16:5207. 1988; Berry et al., Nature 353:273. 1991). This lack of homology between the 3'UTRs of these genes combined with our analyses of the 3'UTR of human glutathione peroxidase, which have demonstrated that two 3 to 4 nucleotide stretches of the putative stem-loop structure are essential for selenocysteine incorporation, have allowed us to design a synthetic nucleotide sequence which is capable of forming a stem-loop structure which contains the essential elements and which is orientation independent. This "optimized" synthetic sequence contains a stem-loop containing a "bubble" 17 nucleotides from base of the stem-loop, followed by an additional 11 nucleotide stem with a 12 nucleotide loop, or balloon, at the top of the structure. To render the "optimized stem loop orientation nonspecific, the essential 3-4 nucleotide targeting elements are positioned in mirror image on the appropriate bubble and balloon regions of the artificial stem loop. The structure of this optimized element is shown in Fig. 9. In this figure, MRS, denotes a multiple restriction site for ease of insertion of the element into any appropriate cloning vector, and N indicates any nucleotide: N_x denotes a stretch of two or more nucleotides of any sequence; N:N denotes complementary base pairs. A nucleotide sequence containing the elements of this optimized stem-loop may be constructed by standard techniques known to those skilled in the art of molecular biology. For example, we have synthesized a 92-mer oligonucleotide comprising the loop-bubble-balloon using an Applied Biosystems DNA synthesizer. After gel purification, this single-stranded oligonucleotide served as the template for PCR amplification using 22-mer sense and antisense PCR primers containing 12 and/or 6 nucleotide overlapping sequences: 1) template oligonucleotide:

5'GTCGACCCTAGGAGAAATTTAGATATGTTATGAGCCCTCT CAGTAAAGTTCACACTGAGAGGGTTAGTGTGAACATATCTAA ATTTCTAGACCC-3' 2) PCR primers: 5'TTTTTTTTTTGTCGAGACCCTAGG-3'

5'TATATATATATCCCCCGGGTCT-3' The PCR reaction was carried out using 0.1 μg/μl template oligonucleotide, 50 pmole/μl of each PCR primer according to standard methods for 10 cycles of 95°C for 1 min, 50°C for 1 min, and 70°C for 1 min.

The double-stranded PCR product was then ligated into the pCRII vector (InVitrogen, San Diego, CA) using the TA cloning system (InVitrogen) and transformed into INVαF' cells. The sequence of the construct was confirmed by nucleotide sequencing using fmol PCR sequencing from Promega (Madison, WI) . Construction of Recombinant Selenocysteine containing Polypeptides.

Any desired polypeptide for which the DNA sequence is known may be used in the method of the invention by substitution of codon encoding any amino acid which is not essential for the natural activity of the polypeptide. The approach to the preparation of these "TGA" mutants may be generally accomplished by site- directed or oligonucleotide based mutagenesis techniques, e.g., using commercially available kits (Promega) . For example, the cDNA encoding the human thyroid hormone receptor-01 was cloned into the multiple cloning site of the vector p-alter (Promega) and the first cysteine codon was mutated to TGA by oligonucleotide based mutagenesis. The cDNA was then rescued by standard laboratory procedure, and the mutation was confirmed by nucleotide sequencing. Expression of Selenopolvpeptides Polypeptides according to the invention may be produced by the expression from a recombinant nucleic acid having a sequence encoding the polypeptide linked to a recombinant nucleic acid containing the stem-loop structure required for translation of selenocysteine, using any appropriate expression system: e.g., transformation of a suitable eukaryotic host cell with the recombinant nucleic acid in a suitable expression vehicle such as those described above. Those skilled in the field of molecular biology will understand that any of a wide variety of expression systems may be used to provide a selenocysteine containing recombinant protein of the invention. The precise host cell used is not critical to the invention and includes Saccharomyces cerevisiae or mammalian cells (e.g., COS-1, HL-60, CV-1, LLC/PK-1, C-6, 3T3L1, and CHO cells). Such cells are available from a wide range of sources (e.g., the .American Type Culture collection, Rockland, MD) . The method of transformation or transfection, and the choice of expression vehicle will depend on the nature of the polypeptide to be expressed and the host system selected. Transformation and transfection methods are described, e.g., in Ausebel et al. (Current Protocols in Molecular Biology, John Wiley & Sons, New York, 1989) ; expression vehicles may be chosen from those well-known in the art, e.g., in Cloning Vectors: A Laboratory Manual (P.H. Pouwels et al., 1985, Suppl. 1987).

For example, the cDNA encoding a desired polypeptide is inserted into the eukaryotic expression vectors pcDNAl/neo and pRC/CMV (InVitrogen) which are especially preferred as parent vectors for the selenocysteine expression system in an orientation designed to allow expression.

Alternatively, selenocysteine containing polypeptides according to the invention may be produced by a stably-transfected mammalian cell line. A number of vectors suitable for stable transfection of mammalian cells are available to the public, e.g., see Pouwels et al., supra; methods for constructing such cell lines are also publicly available, e.g., in Ausebel et al., supra . Once the desired selenopolypeptide is stably transfected into a host system, the production of the polypeptide may be controlled by the content of the selenium in the medium. Selenium deficient cell culture systems have been described (Speirer et al., supra; Chada et al., supra) . Normally, the production of a selenocysteine protein will be inhibited at selenium concentrations below 0.1 ng per ml medium and induced at concentrations above 1 ng per ml. The optimal induction of selenopolypeptide production occurs at approximately 5 to 25 ng per ml medium with concentrations above 50 to 100 ng/ml being cytotoxic, depending on the cell type used.

Once the recombinant polypeptide is expressed, it may be isolated according to methods well known in the art and the functional activity may be determined by assays appropriate for the particular polypeptide, e.g., enzymatic activity or binding affinity. In the case where the desired selenopolypeptide is expressed in cells which contain a native protein with the same functional activities, the selenopolypeptide may be distinguished from the native protein by its higher reactivity with nucleophilic agents due to the selenocysteine moiety as described (Leonard et al., Biochim. Biophys . Acta 787:122. 1984), or alternatively by radiolabeling with ⁷⁵Se, as described herein.

Expression of Selenopolypeptides In vivo

The gene for any desired polypeptide which has been modified according to the methods described herein to encode a selenocysteine amino acid residue may be used to produce a transgenic animal wherein production of the polypeptide is controlled by the selenium content in the diet of animal. Methods for producing transgenic animals are well known (e.g., see Hogan et al., Manipulating the Mouse Embryo: A laboratory manual , CSH Press, Cold Spring Harbor, NY, 1986; Leder et al., U.S. Patent No. 4,736,866). Typically, expression of the desired selenopolypeptide in a transgenic animal will be inhibited when the animal is given a diet containing less than 0.016 mg/kg selenium, whereas high levels of the protein will be produced when the animal is given a diet containing 0.1 mg/kg or more selenium (e.g., as Na₂Se0₃, Sigma) .

Other Embodiments The methods of the invention may also be used to produce high levels of any commercially desirable selenopolypeptide. As discussed above, the presence of available selenium produces a 30 to 50 fold increase in the expression of a selenopolypeptide over the level produced under selenium deficient conditions. This level may be further increased by cotransfecting the cell with the gene encoding the selenocysteine tRNA in an expression vehicle which will allow overexpression of the tRNA under the appropriate conditions, e.g., when selenium is present. For example, this may be accomplished by putting the gene encoding the tRNA under the control of an inducible promoter and then supplying the factor required for induction of the gene at the same time, or before, the medium is supplemented with selenium.

What is claimed is:

Claims

1. A method of controlling the production of a heterologous polypeptide in a eukaryotic cell, said method comprising, providing a cell containing a first nucleic acid encoding said heterologous polypeptide wherein at least one codon of the mRNA transcribed from said first nucleic acid has been replaced by the codon UGA, and a second nucleic acid operably linked to said first sequence, said second nucleic acid being capable of directing the translation of said UGA codon as selenocysteine; and growing said cell under conditions wherein the production of said polypeptide is controlled by the level of selenium available to said cell.

2. The method of claim 1 wherein said is grown in vitro.

3. The method of claim 2 wherein said cell is a mammalian cell.

4. The method of claim 1 wherein said cell is a yeast cell.

5. The method of claim 1 wherein said first and second nucleic acids are maintained in said cell in a recombinant vector capable of autonomously replicating in said cell.

6. The method of claim 1 wherein said first and second nucleic acids are stably integrated into the genome of said cell.

7. The method of claim 1 wherein said cell is an embryonal cell derived from a non-human mammal, and said method further involves the step of obtaining transgenic progeny of said non-human mammal which contain said first and second nucleic acids stably incorporated into the genome.

8. The method of claim 1 wherein said second nucleic acid comprises a contiguous sequence of nucleotides capable of forming a stem-loop secondary structure in the mRNA transcribed from said second nucleic acid.

9. The method of claim 8 wherein said second nucleic acid is derived from approximately 90 contiguous nucleotides from the 3' untranslated region of a gene encoding a naturally occurring mammalian selenoprotein.

10. The method of claim 9 wherein said mammalian selenoprotein is human glutathione peroxidase, and said second nucleic acid comprises a nucleotide sequence substantially homologous to nucleotides 654 to 740 of Figure 8.

11. The method of claim 8 wherein said second nucleic acid is synthetically derived, and the loop of said stem-loop comprises the sequence 5'-NAAAUNNUAAAN-3' and the stem of said stem-loop comprises at least 12 non- complementary nucleotides which form a bubble containing the sequence 5'-NUAGUN-3' symmetrically opposed on each half of said bubble, wherein each N is selected independently from the group consisting of adenine, guanine, cytosine and uracil.

12. The method of claim 2 wherein said polypeptide is produced by said cell when the concentration of selenium is 1 to 25 ng per milliliter of growth medium.

13. The method of claim 7 wherein said polypeptide is produced in said transgenic progeny when the daily dietary intake of said transgenic progeny contains greater than 0.1 mg per kg of selenium.

14. The method of claim 1 wherein said cell does not comprise a native protein substantially similar to said recombinant polypeptide.

15. The method of claim 1 wherein said cell contains a native protein substantially similar to said recombinant polypeptide and said recombinant polypeptide is distinguished from said native protein by the increased reactivity of said recombinant polypeptide to nucleophilic reagents.

16. The method of claim 1 wherein said cell contains a native protein substantially similar to said recombinant polypeptide and said recombinant polypeptide is distinguished from said native protein by the ability of said recombinant polypeptide, but not said native protein, to incorporate the radioisotope ⁷⁵Se.

17. A method of producing radiolabeled recombinant polypeptide comprising, providing a first nucleic acid encoding said heterologous polypeptide wherein at least one codon of the mRNA transcribed from said first nucleic acid has been replaced by the codon UGA; providing a second nucleic acid operably linked to said first sequence, said second nucleic acid comprising a contiguous sequence of nucleotides capable of forming a stem-loop secondary structure in the mRNA transcribed from said second nucleic acid, said stem-loop being capable of directing the translation of said UGA codon as selenocysteine; introducing said first and second nucleic acids into a cell; growing said cell in the presence of ⁷⁵Se under conditions sufficient to allow the incorporation of ⁷⁵Se into said polypeptide.

18. A single-stranded nucleic acid comprising a contiguous stretch of nucleotides capable of forming a stem-loop secondary structure, wherein the loop of said stem-loop comprises the sequence 5'-NAAAUNNUAAAN-3', and the stem of said stem-loop comprises approximately 14 non-complementary nucleotides which form a bubble containing the sequence 5-'NUAGUN-3' symmetrically opposed on each half of said bubble, wherein each of said N, independently, is selected from the group consisting of adenine, guanine, cytosine and uracil, and said nucleic acid being capable of directing the translation of the codon, UGA, as selenocysteine when said nucleic acid is operably linked to an mRNA molecule comprising said codon.

19. The nucleic acid of claim 18 wherein said stem-loop contains approximately 90 nucleotides, said bubble is placed approximately 17 nucleotides from the base of said stem and said loop is placed approximately 11 nucleotides from said bubble.

20. The nucleic acid of claim 18 wherein each half of said bubble is approximately 7 nucleotides.

21. The nucleic of acid of claim 18 wherein.said loop contains approximately 12 nucleotides.

22. A double-stranded nucleic acid comprising DNA encoding the single-stranded nucleic acid of claim 18.