EP0394363A1 - Heteropolymeric protein - Google Patents

Abstract

Described is a biologically active ungulate heteropolymeric protein composed of a plurality of subunits, the two subunits being synthesized into a single cell having an expression vector comprising heterologous DNA encoding the subunits. This protein is preferably similar to the fertility hormones of ungulates, LH and FSH.

Description

HETEROPOLYMERIC PROTEIN Background of the Invention This invention relates to the use of recombinant DNA techniques to produce heteropolymeric proteins.

Various polypeptide chains have been expressed, via recombinant DNA technology, in host cells such as bacteria, yeast, and cultured mammalian cells. Fiddes et al., Nature 281:351, 1979, and Fiddes et al., Nature 286:684, 1980, describe the cloning of, respectively, the α and β subunits of human choriogonadotropin (hCG) .

Kaname US Pat. 4,383,036 describes a process for producing hCG in which human lymphoblastoid cells are implanted into a laboratory animal, harvested from the animal, and cultured iji vitro; accumulated hCG is then harvested from the culture.

Summary of the Invention In general, the invention features, in one aspect, a biologically active ungulate heteropolymeric protein composed of a plurality of subunits, both subunits being synthesized in a single cell having an expression vector containing heterologous DNA encoding the subunits. In related aspects the vector is autonomously replicating (i.e., not integrated into the chromosome of the host cell).

By ungulate is meant an animal of the group consisting of the hoofed animals, e.g., a ruminant (including sheep, cow, giraffe, deer, and camel), swine, horse, tapir, rhinoseros, elephant and hyrax. in preferred embodiments, the protein is synthesized by a eukaryotic cell, and the protein is modified post-translationally, most preferably by glycosylation; and the protein is a secreted protein such as a hormone, most preferably a fertility hormone such as chorionic gonadotropin (CG), luteinizing hormone (LH) or follicle stimulating hormone (FSH); or the hormone thyroid stimulating hormone (TSH) . In other preferred embodiments the ungulate is chosen from a horse, pig or cow.

In another aspect, the invention features a cell, containing a first expression vector, which cell is capable of producing a biologically active ungulate heteropolymeric protein that is encoded at least in part by the vector. In preferred embodiments, a second expression vector (which may be autonomously replicating) encodes a second portion of the protein or at least two subunits of the protein are encoded by a single expression vector; the protein is CG or LH, or FSH; the vector is a replicating virus or a plasmid; the cell is a monkey or mouse cell; transcription of the different subunits is under the control of the SV40 late promoter; transcription of the α. subunit of the protein is under the control of the SV40 early promoter and transcription of the β subunit is under control of the mouse metallothionein promoter, or transcription of both subunits is under the control of the mouse metallothionein promoter; and the expression vector which includes the mouse metallothionein promoter also includes at least the 69% transforming region of the bovine papilloma virus (BPV) genome. In another aspect, the invention features an expression vector (which may be autonomously replicating) including two genes encoding two different ungulate heterologous proteins, the genes being under the control of two different promoters, most preferably a metallothionein promoter and a BPV promoter; the use of different promoters advantageously minimizes the possiblity of deleterious recombinations.

In preferred embodiments of the above aspects the vector comprises at least a part of a DNA sequence chosen from the DNA sequences shown in Figures 1, 2, 3, 8, 10, 11, 13, 17 or 22; and the fragment in the vector is chosen from those fragments deposited in cells in the N.R.R.L. with accession number B18124, B18125, B18126, B18127, B18128, B18139, B18140, B18131, or B15793.

As used herein, "subunit" refers to a portion of a protein, which portion, or homolog or analog thereof, is encoded in nature by a distinct mRNA. Thus, for example, a heavy chain and a light chain of an IgG immunoglobulin are each considered a subunit. Insulin, on the other hand, is composed of two chains which are not considered subunits, because both are, in nature, encoded by a single mRNA, and cleavage into two chains naturally occurs only after translation. The term "expression vector" refers to a vector which includes heterologous (to the vector) DNA under the control of control sequences which permit expression in a host cell. Such vectors include replicating viruses, plasmids, and phages. The term "heterologous DNA" is used to refer to DNA which does not naturally occur adjacent to any other DNA which is also being referred to.

The invention permits the production of a biologically active ungulate heteropolymeric protein from a single culture of transformed cells. We have found that when the subunits of a heteropolymeric protein are produced alone they are modified by the host cell. This modification prevents recombination of the α-s^'ubunit with the β-subunit. Thus, biosynthesis of an active gonadotropin (i.e., α-and-β-subunits linked by ionic bonds) cannot occur in genetically engineered cells unless a system is used to produce both within a single cell so that combination of the subunits occurs prior to post-translational modification. This invention describes and provides the technology required for the establishment of stable cell lines transfected with foreign DNA that encodes for the biosynthesis of balanced amounts of both α and β-subunits. These cell lines allow expression of biologically active dimeric glycoprotein hormones. The system allows production of ungulate proteins, in a single culture, which undergo, in the culture, post-translational modification, e.g. glycosylation and proteolytic processing, for activity or stability.

The use of autonomously replicating expression vectors prevents undesirable influence on the desired coding regions by control sequences in the host chromosome. Other advantages and features of the invention will be apparent from the following description of the preferred embodiments thereof, and from the claims. Description of the Preferred Embodiments We turn now to the preferred embodiments of the invention, first briefly describing the drawings thereof. Drawings

Fig. 1 is a part of the DNA sequence from a bovine α clone.

Fig. 2 is a part of the DNA sequence from a bovine βLH clone.

Fig. 3 is a part of the DNA sequence from a bovine βFSH clone.

Fig. 4 is a diagram showing the cloned regions containing bovine hormone sequences. Fig. 5 is a diagrammatic representation of CL28 and bovine cDNA inserts in CL28.

Fig. 6 is a diagrammatic representation of BPV-CL28 with bovine cDNA inserts. Fig. 7 is a diagrammatic representation of

BPV-CL28 with α and β bovine cDNA inserts.

Fig. 8 is a part of the DNA sequence from the porcine α clone.

Fig. 9 is a diagrammatic representation of the cDNA clones isolated from around the porcine βFSH region. Fig. 10 is a part of the DNA sequence from the porcine βFSH clone.

Fig. 11 is a part of the DNA sequence from the equine βLH clone. Fig. 12 is a diagrammatic representation of the construction of CLSVLPX-eLH.

Fig. 13 is a part of the DNA sequence of the equine α clone.

Fig. 14 is a diagrammatic representation of the construction of CLSVLP-eα.

Fig. 15 is a diagrammatic representation of the equine βFSH clone, showing restriction endonuclease sites.

Fig. 16 is a diagrammatic representation of the construction of CLSVLP-eFSH.

Fig. 17 is a part of the DNA sequence from the equine βFSH clone.

Fig. 18 is a graphical representation of activity of recombinant dimer LH, in samples of cell supernatants, measured by stimulation of testosterone. Fig. 19 is a graphical representation of activity of recombinant dimer FSH in samples of cell supernatants, measured by estradiol production. Fig. 20 is a graphical representation of activity of recombinant dimer FSH, measured by injection of samples, containing the hormone, into immature female mice and observing ovarian weight increase. Fig. 21 is a graphical representation of the bioassay of two samples (stars and boxes) of cell supernatants containing recombinant pFSH, and a standard reference control (circles) .

Fig. 22 is a part of the DNA sequence of a cDNA clone of porcine βLH. Structure

The cloning vectors of the invention have the general structure recited in the Summary of the Invention, above. Preferred vectors have the structures shown in the Figures, and are described in more detail below. In each example the vectors were used to express the α and β subunits of each of various hormones encoded by ungulate DNAs. Three ungulate examples are presented: horse, cow and pig. It is understood that these examples are not limiting to the invention; those skilled in the art can use the methods described to isolate and express equivalent genes from other ungulates.

Each example describes the methods used to isolate the appropriate RNA, prepare a cDNA bank and then identify the desired clones.

Most of the techniques used herein are described in detail in Maniatis et al. (1982) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory), hereby incorporated by reference.

Bovine Hormones Example 1: Bovine LH and FSH (bLH, bFSH)

Bovine cDNA was synthesized from bovine pituitary mRNA by reverse transcription of poly A RNA, treatment with NaOH to hydrolyze the RNA template, and second strand synthesis using reverse transcriptase. Following treatment with nuclease SI and the EcoRI methylase, double stranded cDNA was ligated to synthetic EcoRI linkers, digested with EcoRI and ligated to the arms of the bacteriophage vector, λgtll. The recombinant phage were packaged in vitro by standard procedures and used to infect host E. coli cells to produce plaques. Clones were isolated by screening this cDNA library with appropriate probes.

The bovine α subunit cDNA was identified by hybridization of between 150,000 and 200,000 phage plaques with the 730bp EcoRI fragment containing sequences from the bovine α coding region (obtained from R. Mauer, University of Iowa), which had been labelled by nick translation to a specific activity

Q greater than 1 X 10 cpm per microgram.

The plaques to be screened were transferred to nitrocellulose filters, treated with NaOH, neutralized, washed with 2 X SSC, baked in a vacuum oven at 80°C for two hours, prehybridized and hybridized to the labelled, denatured α probe overnight. The filters were then washed twice with 1 X SSC and four times with 0.25 X SSC at 65°C before being dried and autoradiographed at -70°C for identification of positively-hybridizing plaques.

The complete nucleotide sequence of the coding region of the α cDNA was determined by use of both the chemical cleavage method of Maxam and Gilbert and the chain termination method of Sanger et al. The sequence of the coding region, including that coding for the 24 amino acid signal peptide, is given in Figure 1. In similar fashion, the cDNA for βLH is isolated and sequenced. A total of four clones were characterized, only one of which (LH7) contained the entire coding region at the 5' end of the clone. This clone, however, differed from the other three clones in that the triplet coding for amino acid 92 of the mature bLH β polypeptide was TCC (=Ser) rather than CCC (=Pro) . Since LH's from other species contain Pro at this position, we spliced the 5' end of LH7 to the 3' region of the clone LH8 (one of the other four clones) in order to reconstitute a complete coding region with a proline codon corresponding to position 92 of the mature protein subunit. The splicing was carried out at a

PvuII site located 14 base pairs 5' to the TCC codon in LH7. The nucleotide sequence of the bLH β subunit clone used in the expression system (see below) is shown in Figure 2. Identical procedures, to those described for bLH, were followed to identify bFSH clones, which were isolated by use of a porcine FSH probe (from NRRL B15793, see below). Because of abnormalities in each of the bFSH clones isolated (small deletions and a change in the region of the ATG initiator codon) , it was necessary to splice together fragments to obtain a clone suitable for expression of the protein subunit. The DNA sequence of the bFSH β subunit are given in Figure 3. Our first approach to genetically engineering the bovine gonadotropin cDNAs for expression was to construct separate expression plasmids for the α subunit cDNA and each of the β subunits cDNAs. BamHI linkers were ligated to flanking regions of each clone as shown in Figure 4. These fragments were then inserted into the Bglll site of the plasmid, CL28, shown in Figure 5, so as to place each cDNA (shown as an open rectangle) under control of the mouse metallothionein promoter (shown in a triangle). The plasmid CL28 is shown before and after insertion of a bovine gonadotropin cDNA clone. Referring to upper part of Figure 5, CL28 contains, in addition to a bacterial replication origin and a drug-reisitance marker (not shown), the mouse metallothionein gene (zig zag line), which has a unique cloning site (Bglll) located just downstream of the metallothionein promoter (triangle) . Referring to the lower figure, when a cDNA is inserted into the BgJ.II site, a plasmid such as the one shown is obtained. The inserted cDNA, when it is positioned in the orientation given by the arrow above the box, is placed under control of the murine metallothionien promoter, which allows its expression in mammalian cells. Following these constructions, the entire Bovine Papilloma Virus (BPV) genome was inserted as a 7.9 kbp BamHI - Sail fragment into each plasmid (see Figure 6). The BamHI and Sail sites are used for the insertion of the Bovine Papilloma Virus (BPV, large open rectangle) genome as shown in Figure 6.

These plasmids allow expression of bovine gonadotropin cDNA in mouse C127 fibroblast cells. The BPV portion allows the plasmid to replicate in C127 cells and frees them from contact inhibition of growth so that permanent lines expressing the product encoded by the cDNA may be obtained.

Another set of constructions, shown in Figure 7, was carried out for the expresion of bLH. In addition to the features described above, this plasmid contains a second transcriptional unit in which the α cDNA has been placed between the SV40 early promoter (open triangle) and a strong transcriptional termination signal also derived from SV40. A single such plasmid will determine the production of both subunits, α plus bLH β, or α plus bFSH β. The subunit cDNA is inserted into a site between the SV40 viral early promoter and the strong transcriptional termination region also derived from the SV40 virus. The entire unit is inserted as a BamHI fragment as shown in Figure 7.

Porcine Hormones Example 2: Porcine LH, FSH ( LH, pFSH, )

The cDNA for the common α subunit of the porcine gonadotropins was isolated from a cDNA library constructed in the E. coli plasmid, pBR322, by G/C tailing into the Pstl site in the β-lactamase gene.

Pituitaries from freshly slaughtered swine were quick-frozen in liquid nitrogen and stored at -70°C until use. Total RNA was prepared from the frozen glands, after pulverization, by extraction in (1:1) phenol: lOOmM Na-acetate (pH 5.5) at 65°C. Following separation of the phases by low-speed centrifugation, the organic phase was re-extracted with fresh acetate buffer as above and the resulting aqueous phase was pooled with the first. RNA was precipitated by the addition of 2.5 volumes of ethanol at -20°C. Redissolved total RNA was separated into poly A- and poly A -enriched fractions by chromatography on oligo (dT)-cellulose. The poly A⁺-enriched material was precipitated with ethanol, washed with 70% ethanol and dried under vacuum.

First strand cDNA synthesis was carried out by standard protocols using the AMV reverse transcriptase (60 units) to reverse transcribe five micrograms (μg) of RNA in a reaction primed by two μg of oligo (dT) in a^' final volume of 20 microliters at 42°C. At the end of the incubation period, 0.01M Tris-HCl (pH 7.5) containing 0.001M EDTA (TE) was added and the mixture was extracted with an equal volume of phenol. The organic phase was re-extracted with TE and extracted once with chloroform. Three microliters of 0.25% acrylamide were added as carrier and nucleic acid was precipitated by the addition of 600 microliters of ethanol and 150 microliters of 4M (NH.)--acetate and left overnight at -20°C.

Second strand synthesis was performed by the RNAase H method, which eliminates the need for treatment with SI nuclease. The RNA-cDNA hybrid formed in the first reaction (above) was incubated for 60 min at 16°C followed by 60 min at room temperature in the presence of 0.9 units of RNase H and 23 units of E. coli DNA polymerase I. Double stranded cDNA was extracted with phenol, precipitated with ethanol, washed, redissolved and tailed with dCTP by use of terminal deoxynucleotidyl transferase (40 units) in the presence of 1.4mM CoCl₂.

Tailed cDNA was annealed with pBR322 DNA, which had been tailed with dGTP at its Pstl site, and used to transform E. coli to tetracycline resistance.

Clones containing porcine α sequences were identified by hybridization to a bovine α cDNA (from pBR bov α 730 RI) , which had been radioactively labelled by nick translation. One of these clones, p α 2B, was subcloned into pUC18 for sequencing. The results (Figure 8) show it to be full length.

The cDNA for the βLH subunit was isolated from another cDNA library. This library was constructed similarly to that described above. RNA was extracted from frozen porcine pituitary tissue by the guanidinium/cesium chloride method of Maniatis et al. supra, and poly(A) RNA was isolated by two passages over oligo(dT)-cellulose. Two hybridization probes were used: a 320 bp Nael-HinfI fragment (henceforth referred to as the 3' porcine probe) from an incomplete porcine LH β- cDNA clone obtained from the library described above, using the bovine BLH cloned fragment as a probe, and the 120 bp 5' EcoRI-StuI fragment (henceforth referred to as the 5' bovine probe) isolated from bovine LH β cDNA.

First strand cDNA was synthesized from 10 ug porcine poly(A) RNA using Moloney murine leukemia virus reverse transcriptase in the presence of oligo(dT) primer. Second strand DNA was synthesized by the method of Gubler et al., Gene, 25_:263, 1983 in which RNase H was used to degrade the RNA strand and E. coli DNA polymerase I was used to sythesize the DNA replacement. 1.8 ug double stranded cDNA was obtained. The cDNA was C-tailed and annealed to Pstl digested, G-tailed pBR322 (Bethesda Research Laboratories) . Transformation of competent MC1061 cells with the annealed DNA resulted in a library of approximately 70,000 independent recombinant clones. Screening the library for porcine LH β cDNA clones was done in two steps. First the entire library was screened at a high colony per plate density (7000 cfu/100 mm dish) with the 3' porcine probe. Approximately 600 duplicate positives were obtained. Twenty positives were selected for secondary screening. This was done by scraping the positive areas into broth, diluting the clonally mixed suspensions, and plating at a low colony per plate density (250-500 cfu/100 mm dish). Two replicas were hybridized to the 3' porcine probe and to identify the most complete clones, a third replica was hybridized to the 5' bovine probe. All 20 putative positives contained clones that hybridized in duplicate to the 3' porcine probe. 13 of the 20 also hybridized to the 5' bovine probe. Plasmid DNA was isolated from each of the 13 most complete clones. Examination by restriction endonuclease analysis showed the cDNA inserts to be similar in size, ranging from 500 bp to 540 bp. All clones contained the internal PvuII site present in the original incomplete porcine LH β cDNA clone. The clone with the longest 5' PvuII fragment was selected for further subcloning and sequence analysis.

The complete sequence of the porcine LH β cDNA insert, with amino acid sequence depicted, is shown in Fig. 22. The cDNA represents 486 bp of transcribed porcine LH mRNA, but does not include the ATG initiator codon. The longest open reading frame is shown in the figure, and is 137 amino acids in length, encompassing the entire mature porcine LH β subunit and 16 amino acids of the secretory leader peptide. The DNA translation differs from the published protein sequence (Pierce et al. , Ann. Rev. Biochem. 5):465, 1981) in 6 locations:

The porcine FSH β subunit was isolated from a third cDNA library, which was constructed similarly to that described above for the isolation of the subunit, except that second strand synthesis was carried out using the Klenow fragment of the E. coli DNA polymerase in the hairpin-primed reaction, rather than by the RNase H method. Following second strand synthesis, the cDNA was treated with SI nuclease to digest the closed end of the hairpin; tailing, annealing and transformation of E. coli were performed as described above.

Screening for the pFSH β subunit was carried out by using two synthetic DNA fragments, each of which was 45 base pairs in length, corresponding to amino acids 57-71 and 74-78 of the mature protein sequence (Pierce et al. , An. Rev. Biochem. 50_.:465, 1981). The DNA sequences were chosen according to the codon usage method of Jaye et al. , Nuc. Acid Res. _11_:2325, 1983. Two clones were isolated from this screening, pF55 and pF434. As shown in Figure 9, these clones overlap. They were joined at the unique Smal/Aval site to construct a continuous coding sequence, shown in Figure 10.

Construction of CL28-αBPV

The 608 base pair cDNA was cut with Ncol at position 102 and Ddel at position 504. The ends of this fragment, which contained the entire coding region for the pre-ot protein, were filled in using the Klenow fragment of DNA polymerase I and ligated to synthetic BamHI linkers. The linkered fragment was digested with BamHI and ligated into the BamHI site of pBR322. The recombinant plasmid was identified by restriction mapping and used as a source of BamHI-ended DNA for insertion into the plasmid CL28. In brief, the inserted cDNA to be expressed is cloned into a unique Bglll site in the metallothionein gene of CL28. Plasmids containing inserts in the proper orientation are isolated and the BPV genome is then inserted as a 7.9 kb BamHI-Sall fragment, as described above. Construction of CL28-FSH-BPV

Engineering of the pFSH 55/434 clone was a bit more complex. First a Rsal-Smal fragment, extending from position 70 to position 261 was isolated and cloned into the unique Smal site of pUC18. Inserts in the correct orientation regenerate the Smal site. This plasmid was then cut with Smal and Sacl and ligated to a Smal-Sacl fragment (position 261-position 629 of the cDNA) isolated from pFSH 55/434. The recombinant plasmid was identified by restriction mapping and inserted into CL28 as a BamHI-Sau3A fragment (the BamHI site derives from the polylinker of pUC18, and the Sau3A site is located at position 605 of pFSH 55/434). BPV was then inserted into a plasmid containing the engineered pFSH cDNA in the correct orientation with respect to the metallothionein promoter to constitute the final expression vector used to transfect C127 mouse fibroblasts.

Equine Hormones Example 3: Equine LH (eLH)

The first step in the'production of equine LH β cDNA is the preparation of pituitary RNA. The pituitary gland was removed from the horse immediately after death. The pituitary (approx. 2 grams) was homogenized in a 1: 1 mixture of phenol: lOOmM Na-acetate, pH5.2 containing 0.5% SDS at 65°C. The homogenized tissue was incubated at 65°C for 20 minutes with vortexing for 10 seconds every minutes. After quick cooling on ice, the phases were separated by centrifugation. The hot phenol extraction was repeated two more times followed by two extractions with chloroform: isoamyl alcohol (24:1). The RNA was precipitated from the final aqueous phase Jby the addition of 2.5 volumes of ethanol. PolyA⁺ RNA was isolated from the total equine pituitary RNA by oligo-(dT) cellulose chromatography. The RNA was passed over the oligo-(dT) column in lOmM Tris-HCl, pH7.5, 0.5M NaCl. After several washings with this buffer solution, polyA RNA was eluted with lOmM Tris-HCl, pH7.5, ImM EDTA, 0.05% SDS. The chromatography was repeated and the final eluate precipitated with ethanol.

The equine cDNA library was constructed by conventional techniques. The first strand cDNA was synthesized by reverse transcription of pituitary polyA RNA using the enzyme AMV reverse transcriptase. Second strand synthesis was carried out by sequential reactions with the Klenow fragment of E. coli DNA polymerase and AMV reverse transcriptase. The resulting hairpin loop was cleaved by SI nuclease digestion. The double-stranded cDNA was C-tailed by reaction with calf-thymus terminal deoxynucleotidyl transferase and annealed to pBR322 which had been cleaved at the Pstl site and G-tailed. These recombinant plasmids were then used to transform E. coli MC1061 cells to generate a cDNA library. Transformants were selected on the basis of tetracycline resistance. In order to identify the equine LH β clone, a 520bp fragment of the above bovine LH β cDNA clone was used as a hybridization probe. The probe, radioactively labeled with 32p by nick-translation, was used to screen the equine cDNA library by the colony hybridization technique of Grunstein et al. (Proc. Nat. Acad. Sci. U.S.A. 72:3961, 1975). Pre-hybridization was carried out at 37°C in 50% formamide, 5X Denhardt's, 5X SSC, pH7.4, 0.1% SDS, and 100 ug/ l E. coli tRNA. Hybridization was carried out at 37°C with the same solution containing denatured probe at a concentration of 5 x 10 cpm/ml. The filters were washed at 55°C with 2 x SSC, 0.01M Na₂P0₄, pH7.2, 0.1% SDS. The washed filters were exposed to Kodak film overnight at -70°C with one intensifying screen. Screening yielded one equine LH β cDNA clone

(pBR322-eLH Pst) whose sequence is shown in Figure 11. Equine α was isolated by a similar method to that described above for βLH, except that a 730 bp EcoRI fragment of bovine α cDNA was used as a probe. The DNA sequence of the α is shown in Fig. 13. Construction of CLSVLPX-eLH

CLSVLP is an expression vector constructed to make use of the SV40 late transcriptional signals and the SV40, T-antigen-dependent origin of replication. An SV40 late transcriptional cassette was constructed by joining the SV 40 fragment from Taql (0.56 m.u.) to Hind III (0.83 mu) to the SV40 poly A signal fragment which spans the region of SV 40 from 0.14 to 0.19 mu. To accomplish this, the Hind III site at 0.83 mu and the Bell site at 0.19 mu were changed to Bglll sites by filling in the overhanging basis with Klenow and ligating Bglil linkers. The two fragments were then joined at this Bglll site. This cassette, which now has Taql and BamHI, at the 5' and 3' ends respectively, was put into pML2 at the Clal and BamHI sites. pML2 is a deleted version of pBR322; the deleted sequences include from 1120 bp to 2490 bp on the pBR322 map (Lusky et al. , Nature 293: 19, 1981).

Referring to Fig. 12, pBR322-eLH Pst was digested with Pst-1 and the 610 bp fragment was isolated. This Pst-1 fragment contains the entire coding sequences of equine β LH subunit. As an intermediate step, this Pstl fragment was subcloned into the unique Pst I site in the multi-cloning region of 02757

- 18 - pUC18 to create the vector, pUC18 eLH Pst. The orientation of the cDNA fragment with respect to the pUC18 is such that the unique EcoRI site in the multi-cloning region is at the 3' end of the cDNA. pUC18 eLH Pst was digested with HindiII and EcoRI, which flank the cDNA insert and the 650bp fragment was isolated. This isolated DNA was subsequently digested with Banl, to make use of the unique Banl site lObp 5' to the ATG, filled-in with Klenow polymerase, and Sail linkers were added. A Sail digest was carried out and the mixture ligated to CLSVLP-Xho (which is simply CLSVLP with the Bglll site changed to an Xhol site by the addition of Xhol linkers). Orientation was checked and the final DNA, CLSVLPX-eLH, was purified on CsCl gradients.

Construction of plasmid CLSVLP-eα.

The Ba Hl-NcoI fragment coding for eαLH was cloned into pUClδ by standard methodology to give pUC18eα. Referring to Fig 14, pUClδe α DNA was digested with Ncol, filled in with Klenow polymerase, and BamHI linkers added. This construct, now called pUC18e α Bam, has two BamHI sites flanking the entire equine cDNA region. pUC18e α Bam was digested with BamHI to isolate the cDNA, which subsequently was cloned into the unique Bglll site of CLSVLP, and plasmids having the correct orientation (shown in Fig.

14), purified.

Example 4: Equine FSH(eFSH)

An equine FSH β clone was isolated from a θgenomic DNA library constructed in the phage EMBL3.

The first step in the production of the equine genomic library was the preparation of equine DNA. Equine DNA was prepared from buffy coat cells which were isolated from total horse blood by centrifugation. The buffy coat cells were freed of contaminating red blood cells by lysis with freshly prepared red blood cell lysis buffer (10:1, 0.144 M NH₄C1: 0.01 M NaHC0₃). The buffy coats were pelleted and resuspended in nuclei lysis buffer (0.4 M NaCl, 0.01M Tris-HCl, pH8.0, 0.002 M EDTA) . SDS was added to 0.5% and proteinase K to a final concentration of 250μg/ml. After overnight incubation at 37°C, the lysed cells were extracted two times with phenol and two times with CHC1₃. The DNA was ethanol precipitated, spooled, and resuspended at a concetration of 300μg/ml in 0.01M Tris-HCl, pH8.0, 0.001 M EDTA at 4°C.

A genomic library was prepared in λ vector EMBL3 according to the method of Frischauf et al. (J. Mol. Biol. 170:827, 1983). Briefly, total horse DNA was partially digested with Mbol to achieve a fragment length of 20kb. The digested DNA was then treated with calf intestinal alkaline phosphatase, ethanol precipitated, and resuspended in 0.001 M Tris-HCl, pH7.5, 0.0001 M EDTA at a concentration of 0.5μg/μl. EMBL3 vector DNA was digested to completion with BamHI and EcoRI■ The DNA was phenol-extracted and precipitated with 0.6 vol of isopropanol. The DNA pellet was washed two times with 0.35 M NaAcetate (pH6) :ethanol (1:25) and resuspended in 0.01 M Tris-HCl, pH7.5, 0.0001 M EDTA at a concentration of 0.5 μg/μl.

For ligation, vector and insert DNA were mixed in a 10:1 weight ratio (vector: insert) in IX ligase buffer (10 mM Tris-HCl, pH7.6, 10 mM MgCl₂, ImM DTT, 0.5 mM ATP, 0.1 mg/ml BSA) and T4 ligase was added. Ligation was carried out overnight at 15°C.

Packaging of the ligated DNA was carried out using commerical packaging extract "Packagene" (Promega Biotech) and the resulting bacteriophage plated on the

E. coli strain NM539. Bacteriophage were eluted with ice cold SM buffer and the resulting amplified library stored at 4°C over CHC1₃. Screening was carried by the method of

Gruenstein et al. , supra using a 520 bp coding region fragment of the above bovine FSH cDNA as probe. The probe was nick-translated. Pre-Hybridization was carried out at 37°C in 50% formamide, 5X Denhardt's, 5X SSC, 0.1% SDS, and lOOμg/ml E. coli tRNA.

Hybridization was carried out at 37°C with the same solution containing denatured probe at a concentration

5 of 5 X 10 cpm/ml. Filters were washed at 45°C in

2xSSC, 0.01 M sodium phosphate, 0.1% SDS and exposed to Kodak XAR film for two days at -70° with an intensifying screen.

Screening yielded one recombinant bacteriophage containing the full-length equine FSH coding region. The restriction map of the 4.4 kb BamHI fragment of the remitting clone is shown in Fig 15. The coding regions are shown below the map as dark boxes, which are connected by a broken line representing an intron. Construction of plasmid CLSVLP-βFSH Referring to Fig. 16, pUC18 eFSHBam contains the eFSH genomic DNA clone as a 4. Kb BamHI fragment. Partial sequencing revealed that the translational initiation codon (ATG) was about 25bp downstream from a Stul site (Fig. 17). Restriction mapping showed at least one more Stul site about 2450bp downstream from this first site. Partial sequencing suggested that the entire coding region for the β FSH subunit was present between the two Stul sites. pUC18 eFSHBam was digested with Stul (which leaves blunt ends) and a 2450bp fragment was isolated on agarose gels. CLSVLP DNA was digested with Bglll, filled in using Klenow polymerase, and ligated to the isolated Stul fragment. The correct orientation CLSVLP-βFSH was selected and DNA prepared. Expression and Use The above described vectors are useful for the expression of hormone subunits in appropriate cells. The incorporation of virus-containing vectors into eukaryotic cells for the production of a heteropolymeric protein is generally accomplished as follows. First, if the viral DNA and homopolymeric protein-encoding DNA are incorporated into a plasmid, which is maintained, in, say, E. coli, the plasmid sequences (e.g., the pBR322 sequences) are removed and the resulting DNA is ligated to form circular DNA including the viral region and the heteropolymeric protein-encoding sequence or sequences. This circular DNA generally does not contain all of the genetic information needed to produce a replicating virus, the other necessary sequences (e.g. those encoding coat protein) having been replaced by the heteropolymeric protein-encoding sequence or sequences. The circular DNA, minus the plasmid DNA, must be close enough in size to the naturally occurring viral DNA from which it is derived to permit the DNA to enter and replicate in appropriate host mammalian cells. The circular DNA is used to transfect host cells in order to produce virus stock for later infections.. Since some of the DNA necessary to produce virus is missing, the transfection must occur in conjunction with helper virus DNA encoding enough of the missing function to produce replicating virus.

Transfected host cells are grown and incubated until lysed by replicating virus. The resulting replicating virus stock, including helper virus, is then used to infect host cells for production of the heteropolymeric protein. Virus stock is maintained, since it generally needs to be reused to infect fresh batches of host cells, as each culture of infected, protein-producing host cells generally is eventually lysed by the virus.

The specific recombinant DNA sequences described above are used to transfect, and then infect, host cells, as follows. Again these examples are not meant to be limiting, those skilled in the art will readily see that other methods can be used for protein expression. Example 5: Bovine hormones

In order to derive cell lines producing bLH, the α-containing plasmid and the bLH β-containing plasmid were mixed together and introduced into mouse fibroblasts. Similarly, for the expression of bFSH, α-containing plasmid DNA was mixed with the bFSH β-containing plasmid and introduced into the mouse cells by the Ca-phosphate transfection procedure, described above. The mouse C127 fibroblasts were transfected with plasmids as described above and transformed cell lines derived from these transfections. Eighty-six cell lines were screened for bLH expression. Seven of these were selected for further characterization. For bFSH expression, eight of ninety-four lines were further characterized. Assays carried out both ^n vitro and in vivo demonstrated the biological activities of both bLH (Figure 18) and bFSH (Figures 19, 20).

Referring to Fig. 18, the in vitro bioassay of recombinant bLH is shown. This assay measures the ability of LH to stimulate testosterone secretion from an isolated preparation of steroidogenic cells (Leydig cells) prepared from rodent testes. Testes from mature rats are collected and the tunica albuginea is removed. The decapsulated testes are mechanically disrupted and the seminiferous tubules separated from the Leydig cells by filtration through nylon mesh. The Leydig cells, which pass through the mesh, are centrifuged at low speed and resuspended in culture medium for use. Each determination is carried out using a tube containing one million viable cells. Reference standard LH or different amounts of unknown are added to each tube. Following incubation at 37°C for three hours, the cells are removed by centrifugation and the supernatants are assayed for testosterone by a specific radioimmunoassay. The standard curve is shown in open circles, while assays of tissue culture supernatants from lines expressing bLH from the cloned cDNAs are given in closed circles. The concentraton of bLH can be calculated from noting the number of microliters required to give a response equivalent to the known number of miu of the LH reference preparation. Results of in vivo bioassays are given in Table

1.

Table 1

SAMPLE Number VOLUME INJECTED (mL) UTERINE WEIGHT (

Control 3 10 14 . 6 + 1 . 6 bLH Medium 3 2 16 . 7 + 2 . 9 3 5 21 . 2 + 7 . 6 3 10 31 . 8 + 9 . 3 Immature female mice (19-22 days) were injected i.p. with either control tissue culture medium or medium removed from cultures of a bLH-producing line as shown in the table. The indicated volumes were injected over a 72 h period in a series of six injections. At the end of the experiment, the mice were sacrificed and their uteri were carefully dissected out, blotted dry and weighted on an analytical balance.

Referring to Fig. 19, the in vitro bioassay of recombinant bFSH is shown. Suspensions of Sertoli cells were prepared from rat seminiferous tubules by treatment with collagenase and mechanical dispersion. The cells are pelleted by centrifugation and plated in microtiter wells at a density of 1 X 10 5 viable cells per well for 24 h. The culture medium is then removed and replaced with fresh medium containing androstenedione plus either a reference FSH preparation or sample of the cell supernatants to be assayed from bFSH production. Following a 24 h incubation period, the medium in each well is assayed from estradiol by radioimmunoassay. The presence of biologically active FSH stimulates the production of the enzyme aromatase by the Sertoli cells. Aromatase, which is induced by FSH in a dose-dependent manner, converts the androstenedione to estradiol. The standard FSH preparation generated the curve shown in open circles, while tissue culture supernatant from a bFSH-producing line gave the results shown in closed circles. The assay demonstrates in vitro bioactivity from the recombinant bFSH. In vivo bioassay of recombinant bFSH was also performed (Figure 20). Increased ovarian weight results from the action of FSH in stimulating follicular growth and development. Since the recombinant bFSH is free of LH activity, there is no need to employ CG before the FSH-containing sample (Steelman and Pohley assay). The curve shown demonstrates the biological activity jin vivo of the recombinant material. Example 6: Porcine hormones To produce cells which synthesize and secrete pFSH, C127 mouse fibroblasts were transfected with a mixture of two plasmids: the α cDNA in the metallothionein-BPV plasmid and the pFSH cDNA in this same type of construction, as described above. Fifty micrograms of each plasmid was added to 0.5 mL of a 250mM CaCl_ solution containing 10 mg of salmon sperm DNA as carrier. The mixture was bubbled into 0.5 mL of 280 mM NaCl, 50mM HEPES and 1.5mM Na-phosphate. The calcium phosphate precipitate was allowed to form for 30 minutes at room temperature.

Twenty-four hours prior to transfection, 5 X ₅ 10 mouse C127 fibroblasts were transferred to a 100 mm culture dish. Immediately before adding the exogenous DNA, the cells were fed with fresh medium (Dulbecco's Modified Medium, 10% fetal bovine serum).

One mL of the calcium phosphate precipitate was added to each dish (10 mL) , and the cells were incubated at 37°C for 8 hours. The medium was then aspirated and replaced with 5 mL of 20% glycerol in phosphate-buffered saline, pH 7.0 (PBS) for two minutes at room temperature. The cells were washed once with PBS, fed with 10 mL of medium and incubated at 37°C for 24 hours. At this time, the medium was changed, and subsequent refeeding was carried out every three days. Expression of pFSH was determined by ij vitro assay. Cell supernatants, potentially containing FSH, were collected and concentrated by dialysis against polyethylene glycol. These concentrated samples were then assayed for jLn vitro activity as follows: Immature female rats (21 days old) were injected with 5 IU of PMSG (pregnant mouse's serum gonadotropin) and killed 48 hours later. Their ovaries were removed and placed in ice-cold Medium 199. Granulosa cells were expressed from the ovarian follicles into McCoy's 5A medium with 4mM glutamine and antibiotics. Approximately 0.25 X 10 cells per well were inoculated into microtiter dishes in a volume of 1 mL and samples to be assayed (the above cell supernatants) and standards were added in a volume of 0.3 mL.

The microtiter dishes were placed in a water-saturated incubator at 37°C in an atmosphere of 95% air and 5% CO,. After 48 hours, supernatants were removed, centrifuged at 100 X g for 20 min and samples were taken for determination of progesterone by a highly sensitive and specific radioimmunoassay. The results (Fig. 21) show a dose-dependent increase in the amount of progesterone found in the medium, which parallels the standard curve for FSH. Example 7: Equine Hormones

COS-7 cells (Gluzman, Cell 23:175, 1981) were split (on the day prior to the transfection) into 100mm dishes, at 10 cells per dish. 25ug of equine α DNA (CLSVLP-e α) plus 25ug of the appropriate equine β subunit DNA (CLSVLP-eFSH or CLSVLP-eLH) were mixed together with 5-10ug of pJL DNA (which expresses SV40 large-T and has its own SV40 origin of replication, Dunn et al. , Cancer cells 3:227, Cold Spring Harbor, 1985) with 0.5ml of 0.28M NaCl, 0.05M HEPES, and 0.0015M Sodium Phosphate, pH7. 0.5ml of a mixture of 0.25M CaCl₂ and 0.01M HEPES, pH 7.1 was added dropwise, and a precipitate was allowed to form, undisturbed, for 40 minutes. The precipitate, as a suspension, was added to the cells and allowed to incubate for at least 5 hours at 37°C in a C0₂ incubator. The precipitate and media were removed from the cells and 5ml of a solution of 15% ₅ glycerol in PBS was added to each plate and allowed to incubate for 2 minutes at room temperature. This solution was then removed, the cells were washed twice with 10ml of PBS, and 5ml of DMEM+10% fetal bovine serum was added to each plate. Samples were collected after ₁₀ 48 hours incubation and assayed for FSH and LH activities.

Culture medium was collected from the COS cells three days after transfection. Medium was centrifuged at 1000 x g for ten minutes to remove particulate -, -■ matter. Supernatants were concentrated with the

Centricon Micro Concentration apparatus purchased from Amicon. Two milliliters of clarified culture medium was added to the sample reservoir. The reservoir was attached to the membrane retentate cup (mw cutoff 24,000 ₀ daltons), and the entire apparatus centrifuged at 4000 X g for 20 minutes. Solutes with a mw less than 24,000 daltons passed through the membrane while proteins with a mw greater than 24,000 were retained (recovery of proteins with mw greater than 24,000 is typically ₅ greater than 95%). All samples were prepared aseptically to allow for analysis by in vitro bioassay as well as radioimmunoassay. eLH bioactivity was analyzed with the mouse

Leydig cell-testosterone production assay (described 0 above). Leydig cells were diluted to 2.5 X 10 viable cells/ml with Medium 199 containing penicillin (lOOU/ml), streptomycin (100 ug/ml) and 0.1% bovine serum albumin. One milliliter of the Leydig cell suspension was added to each 12 X 75 mm test tube that contained increasing concentrations of a purified LH standard (range from O-lOng/tube) or increasing volumes of the concentrated medium samples. All tubes were incubated in a shaking water bath (at 37°C) for three hours. At the end of this incubation, cells were pelleted by centrifugation and the resulting supernatant was analyzed for testosterone content with a specific testosterone radioimmunoassay. The purified LH preparation stimulated the secretion of testosterone in a dose related fashion to a point where maximal stimulation of the Leydig cells was achieved. eLH bioactivity was present within the concentrated medium range from the equivalent of 1-6 ng/ml of the bovine LH standard. Concentrated medium samples were prepared as described above from cells transfected with the equine α- and equine FSH β-containing plasmids. Medium samples were analyzed for FSH activity with a dimer-specific bovine FSH radioimmunoassay. This antibody was obtained through the generosity of Dr.

James Dias of the Albany Medical College and is directed against the β portion of the FSH dimer. Two of the media samples contained measurable amounts of recombinant equine FSH activity. The two concentrated medium samples were also assessed for FSH jin vitro bioactivity using the granulosa cell-progesterone assay described above. Briefly, granulosa cells were washed several times in Ham F-10 medium containing penicillin, streptomycin and 0.1% bovine serum albumin. Cells were plated at a density of 0.25 X 10 viable cells/ml/culture well in Falcon 24-well culture dishes. To each culture was added a dose of a purified FSH standard or increasing volumes of concentrated culture medium. After 3 days incubation, medium samples were collected, centrifuged at 1000 X g for 15 minutes to remove debris and analyzed for progesterone content with a specific radioimmunoassay. Increasing doses of FSH cause an increase release of progesterone from the cultured cells into the medium. The concentrated culture samples exhibited FSH activity equivalent to approximately 10-40 ng/ml of the bovine FSH standard employed in the assay. Having demonstrated how to produce cell lines which express recombinant LH or FSH, we now demonstrate how production and purification of these recombinant proteins can be achieved on a commercial scale. Once again the following examples are not meant to be limiting to the invention. Example 8: Production of recombinant hormones

Recombinant bovine LH was produced in microcarrier spinner cultures from the Cbl 19b cell

₂ line. Cell stocks were maintained in 850 cm roller bottles with 50 mL of DNA containing 10% fetal bovine serum (FBS) and were routinely screened for bacterial, fungal and mycoplasma contamination.

Both one-liter (culture volume) and eight-liter

Bellco glass spinner vessels were ^'used for production runs. Five grams (dry weight) of Cytodex III microcarrier beads were used in the one-liter vessels, while 40 g were used in the eight-liter cultures. The microcarriers were swollen, washed, autoclaved and washed with growth medium following standard procedures specified by the manufacturer. Cultures were initiated by adding freshly trypsinized cells to the prepared microcarriers, to a final concentration of 0.8 to 1.3 x

10 cell per L. Three days after inoculation, the medium was changed by allowing the microcarriers to settle, withdrawing 80% of the conditioned medium and adding fresh growth medium. Five days after inoculation, the cells were put into production by removing the spent medium, washing the microcarriers once with one half the culture volume of DME (Delbecco's Modified Eagles medium) and adding the production medium (CEM 2000, Cellular Enhancement Medium, Scott Labs, West Warwick, RI) . Thereafter, cultures were harvested every two days by allowing the microcarriers to settle, withdrawing 80% of the conditioned medium and adding fresh CEM 2000 (Cellular Enhancement Medium, Scott

Laboratory Labs, West Warwick, RI). Harvests from day nine onwards were considered to be serum-free and were pooled for purification. Thimerosal was added to all harvested medium at a final concentration of 0.01%, to limit microbial growth.

Yields in one-liter spinners were between 1.3 and 2.0 mg/L per harvest period (48 hours). Approximately 80% of this production occurred on the first day after a medium change. Total production might be increased by changing the medium daily. Production in the eight-liter spinners was from 0.8 to 1.3 mg/L per harvest. These lower levels resulted from lower cell concentrations (1 X 10 cells/mL compared with 1.5 - 2 g X 10 cells/mL in one-liter spinners). Production was continued for 43 days in the one-liter spinners and for

31 and 37 days in the two eight-liter spinners used. A total of 177 mg of crude material in serum-free medium was produced for purification.

Methods employed for the production of recombinant bovine FSH were the same as those described above for LH production. The serum-free growth medium used was CEM + insulin.

Yields in the one-liter spinner cultures were between 1.9 and 5.4 mg/L/harvest. Two one-liter cultures produced 72 mg of crude material for purification, and two more were later set up to produce an additional 109 mg. Cultures were maintained for as long as 38 days. Example 9: Purification of recombinant bovine LH

Conditioned media was made to 0.01% Tween 80 and clarified with a 0.5 micron Pall Profile filter cartridge at a flow rate of 1 1/min. Clarified media was concentrated and flow dialyzed using a Pellican

2 (Millipore) tangential flow cassette with 5 ft. of surface area. The membranes were prepared according to manufacturer specifications. Media was concentrated approximately 40 fold and the final buffer composition was 50 mM MES (2-[N-Morpholino] ethanesulfonic acid, pH 6.0), 20 mM NaCl, and 0.01% Tween 80.

The concentrate was loaded on to 2.5 x 17.5 cm column of Trisacryl M-SP (LKB) which was equilibrated with 50 mM MES, pH6.0, and 20 mM NaCl. The column was washed with two column volumes of this buffer and batch eluted with 0.1M sodium phosphate, pH 9.0, and 0.15 M NaCl. Fractions were assayed by RIA (Radioimmunoassay) and pooled. All procedures were carried out at 4°C.

Reversed phase-HPLC was performed at room temperature using a Waters model 6000A solvent delivery system and model 660 gradient programmer. A

Micro-Bondapak phenyl column (Waters, 4.8 x 300 mm) was equilibrated with 10 mM sodium phosphate pH 7.2 at a flow rate of 1 ml/min. Pooled, S-Sepharose fractions were loaded directly on to the column. After the absorbance at 280 nm returned to baseline, a 30-min. linear gradient from 0 to 70% ethanol in 10 mM phosphate buffer pH 7.2 was applied. One ml fractions were assay by RIA, pooled, and the ethanol removed under vacuum in a Speed Vac Concentrator (Savant) . 90/02757

- 32 - Pooled fractions were lyophylized and dissolved in about 1/5 volume of 10 mM phosphate, pH 7.2, 0.5M NaCl. HPLC molecular exclusion chromatography was performed with a GF-250 column (DuPont) equilibrated 5with the same buffer. A maximum of 0.5 ml of sample was loaded. The flow rate was 0.5 ml/min. and the absorbance monitored at 280 nm. Highly purified dimeric hormone was recovered in 2 to 3 0.5 ml fractions at a retention volume of approximately 10 ml. lOExample 10: Purification of recombinant bovine FSH r-bFSH was purified by a three-step procedure employing anion exchange, dye-affinity and hydrophobic interaction chromotography (HIC) . The entire process was done at 4°C with the exception of the HPLC step 15which was performed at room temperature.

23 L of serum-free, conditioned media was first clarified using a 0.5μm Pall Profile filter cartridge (P/N: MCY 1001/1005). 0.01% Tween 80 was added to minimize loss by non-specific adsorption. The flow rate 2cwas 2.5 L/minutes. There was no loss of activity as measured by RIA.

Clarified media was concentrated and dialyzed with an Amicon spiral tangential flow cartridge (SlOYlO) having 10 ft. 2 of surface area. The cartridge was

23?repared and used according to the manufacturer' s specifications. The flow rate was 1.5-2 L/min. and the outlet pressure was 25-30 PSI. The media was concentrated approximately 30 x then flow dialyzed against lOmM phosphate buffer, pH 5.0 and 0.01% Tween

3®0. Dialysis was continued until the conductivity matched that of the starting buffer. This procedure required 6 L of buffer. The final volume of the concentrated and dialyzed pool was approximately 1L. Anion exchange chromatography was performed on a 500ml DEAE Trisacryl LS (LKB) column (9.0 x 8.0 cm). The column was equilibrated with lOmM phosphate, pH 6.0, containing 0.01% Tween 80 and flowed at a rate of 42 cm/hr. After loading, the column was washed with three column volumes of buffer then eluted with a 5 L linear gradient from 0 to 0.3M NaCl in lOmM phosphate. 45 ml fractions were collected, scanned for A_₈₀ and assayed by RIA.

The pH of the pooled activity was adjusted to 7.5 with 2M NaOH and applied directly onto a 120ml Blue Trisacryl M (LKB) column at a rate of 37 cm/hour. The column was previously equilibrated in lOmM phosphate, pH 7.5, lOOmM NaCl and 0.01% Tween 80 and, following the load, washed with three column volumes of the same buffer. The activity was eluted with a 1.2 L linear gradient from 0.1 to 5M NaCl in lOmM phosphate, pH 7.5, 0.01% Tween. Fractions were again scanned for _28Q and assayed by RIA. Pooled activity off the Blue-Trisacryl column, in approximately 2M NaCl, was loaded directly onto a Synchropak Pentyl HPLC column (4.1 x 250mm) (SynChrom). The column was equilibrated with 0.8M (NH. SO. in lOmM phosphate buffer, pH 7.0 at a flow rate of 1.0 ml/min. When the A_28Q returned to baseline, a linear, decreasing salt gradient from 0.8 to OM (NH.)₂SO. in lOmM phosphate, pH 7.0 was applied over 30 minutes. Activity was assayed by RIA, and appropriate fractions collected.

As demonstrated above, the transformed cell lines of the invention are used to produce biologically active heteropolymeric proteins. Further, the vectors of the invention which contain enough of one or more of the above cDNA sequences to encode a biologically active or i munologically cross-reactive hormone are useful for production of such hormones, or for production of antibodies to the subunits. By biologically active, we mean that the hormone produced from the heterologous DNA exhibits biological activity in the ungulate of the same type as the naturally occurring hormone. Generally about 80% of the amino acid sequence of the naturally occurring hormone is sufficient. By immunologically cross-reactive, we mean that the hormone undergoes an immunological reaction with monoclonal or polyclonal antibodies that also react with the natural hormone. The immunologically cross-reactive hormones are particularly useful for immunodiagnostics, by well known techniques. For example, antibodies to specific subunits are useful for detecting the presence of a specific hormone; for example, in the urine of a female ungulate to determine the turning of the estrous cycle. Deposits

The following vectors have been deposited in the NRRL (Illinois): pUC18 containing the eα 480 BamHI fragment

(IG 5004), NRRL B18127; pUC18 containing the eLHβ 610 bp Pstl fragment

(IG 5005), NRRL B18128; pUC18 containing the e FSHβ 4.4 kb BamHI fragment (IG 5006), NRRL B18139; pBR322 containing the bov α 730 bp EcoRI fragment (IG 5001), NRRL B18124; pBR322 containing the bov LHβ 540 bp EcoRI fragment (IG 5002), NRRL B18125; pBR322 containing the bov FSHβ 540 bp EcoRI

BamHI fragment (IG 5003), NRRL B18126; pBR322 containing the pα 410 bp BamHI fragment (IG 5007), NRRL B18140; pBR322 pLHβ 540 bp (Pstl fragment (IG 5008), NRRL B18131; and pBR322 containing the pFSHβ fragment, NRRL

B15793 Deposits B18124, B18126, B18127, B18128 and

B18131 were deposited in accordance with the Budapest treaty on October 17, 1986; deposits B18139 and B18140 were deposited on November 19, 1986; and deposit B15793 was deposited on . Applicants acknowledge their responsibility to replace these cultures should they die before the end of the term of a patent issued hereon, 5 years after the last request for a culture, or 30 years, whichever is the longer, and their responsibility to notify the depository of the issuance of such a patent. The deposits will be made available to the public under the terms of the Budapest treaty.

Other Embodiments Other embodiments are within the following claims. For example:

Other heteropolymeric proteins, e.g., thyroid stimulating hormone, can be produced, as can ungulate animal immunoglobulins or immune response antigens. Other host cells, vectors, promoters, transforming sequences, and viruses can also be employed. The host cell employed generally is dependent on the vector being used. For example, when the vector is replicating virus or non-replicating viral DNA, the host cells are cells capable of being infected or transfected, respectively, by those vectors; e.g., SV40-containing vectors require monkey host cells, preferably CV-1 cells. Where the cloning vector is a plasmid having procaryotic control sequences, prokaryotic host cells, e.g., E. coli, are used. Where the cloning vector is a plasmid having eukaryotic control sequences, appropriate eukaryotic host cells, e.g., mouse C127 cells, are used.

Claims

1. A biologically active heterodimeric ungulate fertility hormone selected from LH, and FSH, said hormone comprising an α subunit and a β subunit, both said subunits being synthesized in a single cell comprising an expression vector comprising heterologous DNA encoding one said subunit.

2. The hormone of claim 1 wherein said ungulate is chosen from a horse, pig or cow.

3. A cell comprising a first expression vector, said cell being capable of producing a biologically active ungulate fertility hormone selected from LH, and FSH, said hormone being encoded at least in part by said first expression vector.

4. The cells of claim 3 wherein said ungulate is chosen from a horse, pig or cow.

5.- The cell of claim 3, said first expression vector encoding a first subunit of said hormone.

6. The cell of claim 3, further comprising a second expression vector encoding the second subunit of said hormone.

7. The cell of claim 6, a second subunit of said hormone being encoded by said second vector.

8. The cell of claim 6, both subunits of said hormone being encoded by said first expression vector.

9. The cell of claim 3, said first vector being a plasmid.

10. The cell of claim 3, said cell being a mammalian cell.

11. The cell of claim 3, the α and β subunits of said heterodimeric hormone being encoded by said first expression vector.

12. The cell of claim 11, transcription of said α and β subunits of said heterodimeric hormone being under the control of the SV40 late promoter.

13N The cell of claim 11, transcription of the cc subunit of said heterodimeric hormone being under the control of the SV40 early promoter, and transcription of the β subunit of said heterodimeric hormone being under the control of the mouse metallothionein promoter.

14. The cell of claim 3, said first vector comprising at least the 69% transforming region of the bovine papilloma virus genome.

15. The cell of claim 3, the α subunit of said heterodimeric hormone being encoded by said first expression vector and the β subunit of said heterodimeric hormone being encoded by a second expression vector.

16. The cell of claim 14, transcription of the α and β subunits of said heterodimeric hormone being under the control of the SV40 late promoter.

17. The cell of claim 8, said cell being a monkey cell.

18. The cell of claim 14, transcription of the α and β subunits of heterodimeric hormone being under the control of the mouse metallothionein promoter.

19. The cell of claim 18, said first and said second vectors each comprising at least the 69% transforming region of the bovine papilloma virus genome.

20. The cell of claim 9, said cell being a mouse cell.

21. An expression vector encoding the two different subunits of an ungulate fertility hormone selected from LH and FSH.

22. The vector of claim 21 wherein said ungulate is chosen from a horse, pig or cow.

23. The expression vector of claim 21, said expression vector comprising a plasmid.

24. The expression vector of claim 21, said expression vector comprising a replicating virus.

25. Biologically active hormone produced by the- cell of claim 3.

26. A method for producing a biologically active ungulate fertility hormone selected from LH, and FSH comprising culturing host cells comprising a first expression vector encoding at least a portion of said hormone.

27. The method of claim 26 wherein said ungulate is chosen from a horse, cow or pig.

28. The method of claim 26, the first subunit of said hormone being encoded by said first expression vector and a second subunit of said hormone being encoded by a second expression vector comprised in said host cell.

29. The method of claim 26, both subunits of said hormone being encoded by said first expression vector.

30. The hormone of claim 1, said expression vector being autonomously replicating.

31. The cell of claim 3, said expression vector being autonomously replicating.

32. The vector of claim 21, said vector being autonomously replicating.

33. The method of claim 26, wherein said expression is autonomously replicating.

34. The hormone of claim 1, wherein said vector comprises enough of one of the DNA sequences shown in Figures l, 2, 3, 8, 10, 11, 12, 17, or 22 to encode a biologically active or immunologically cross-reactive c hormone.

35. The cell of claim 3, wherein said vector comprises enough of one of the DNA sequences shown in Figures 1, 2, 3, 8, 10, 11, 13, 17 or 22 to encode a biologically active or immunologically cross-reactive hormone.

36. The vector of claim 21, wherein said vector comprises enough of one of the DNA sequences shown in Figures 1, 2, 3, 8, 10, 11, 13, 17 or 22 to encode a biologically active or immunologically cross-reactive hormone.

37. The method of claim 26, wherein said vector comprises enough of one of the DNA sequences shown in Figures 1, 2, 3, 8, 10, 11, 13, 17 or 22 to encode a biologically active or immunologically cross-reactive hormone.

38. A DNA fragment comprising at least 80% of the ungulate cDNA, chosen from the ungulate cDNA present in the vectors in cells deposited in the N.R.R.L. with deposit number B18124, B18125, B18126, B18127, B18128, B18139, B18140, B18131, or B15793.

39. cDNA encoding at least 80% of the α or β subunit of porcine LH or FSH.

40. cDNA encoding at least 80% of the α or β subunit of bovine LH or FSH.

41. cDNA encoding at least 80% of the α or β subunit of equine LH or FSH.