AU6054890A

AU6054890A - Enhancement of musculature in animals

Info

Publication number: AU6054890A
Application number: AU60548/90A
Authority: AU
Inventors: Stephen H. Hughes; Pramod Sutrave
Original assignee: US Department of Health and Human Services; US Government
Current assignee: US Department of Commerce
Priority date: 1989-06-30
Filing date: 1990-07-02
Publication date: 1991-01-17
Anticipated expiration: 2010-07-02
Also published as: CA2059137C; WO1991000287A1; EP0483219A4; CA2059137A1; JPH10117785A; JPH04502559A; AU633606B2; JP2706697B2; EP0483219A1

Description

ENHANCEMENT OF MUSCULATURE IN ANIMALS

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates the c-ski gene. In particular, the present invention relates to DNA segments encoding chicken c-ski protein, to DNA constructs comprising the DNA segments and to cells transformed therewith. The present invention further relates to animals having increased muscle size.

Background Information

Viruses that contain the v-ski oncogene are not only capable of causing morphological transformation in vitro, but also can induce yogenic differentiation (Stavnezer et al., 1981, J. Virol. 39 , 920-934; Li et al., 1986, J. Virol. 57 , 1065-1072; Stavnezer et al., 1986, J. Virol. 57, 1073-1083; Colmenares and Stavnezer, 1989, Cell 59 , 293-303) . Viruses that carry and express c-ski cDNAs also induce foci and myogenic differentiation (Sutrave et al., 1990, Mol. Cell. Biol. 10 , 3137-3144). This suggests the pos¬ sibility that the ski oncogene is bifunctional since the two known functions of ski , trans- formation and differentiation, would appear to be contradictory properties. Using a v-εki probe, genomic clones for c-ski have been isolated and partially sequenced (Stavnezer et al., 1989, Mol. Cell. Biol. 9 , 4038-4045). Comparisons of the properties of two forms of c-ski that are related by alternative splicing, and of several v-ski and c-ski deletion mutants have shown that the portions of ski required for transformation and differentiation are quite similar. These results suggest that the ability of c-ski and v-ski to cause transformation and induce differentiation may be related aspects of a single property of ski rather than two separate functions.

Relatively little is known about the biochemical functions of the ski proteins. All of the biologically active forms of c-ski and v-ski that have been studied are localized primarily in the nucleus (Barkas et al., 1986, Virology 151 , 131-138; Sutrave et al., 1990, Mol. Cell. Biol. 10 , 3137-3144). When the c-ski proteins are overexpressed in chicken cells, different forms of c-s i differ in their subnuclear localization; however, the significance of these differences, if any, is as yet unclear. When chromatin condenses for cell division, the over-expressed c-ski proteins are associated with the condensed chromatin (Sutrave et al., 1990, Mol. Cell. Biol. 10 , 3137-3144). Biochemical studies have also shown that at least one form of c-ski can bind to DNA in the presence of other proteins (Nagase et al., 1990, Nucl. Acids Res. 18 , 337-343).

None of the available data make it possible to infer the normal function of c-ski either in terms of its role in growth and development (if any) or to have any direct insight into its mode of action. SUMMARY OF THE INVENTION

It is an object of the present invention to provide an isolated and characterized c-ski cDNA. It is another object of the present invention to provide a gene that increases the muscle size in animals.

It is another object of the present invention to provide domestic livestock with increased muscle size and decreased fat tissue. It is a further object of the present invention to provide a treatment for patients suffering from serious muscle injury or muscle degenerative diseases.

Various other objects and advantages of the present invention will be apparent from the drawings and the following description of the invention.

In one embodiment, the present invention relates to a DNA segment encoding a chicken c-εki protein or a DNA fragment complementary to said segment.

In another embodiment, the present invention relates to a DNA construct comprising .a DNA segment encoding a chicken c-ski protein and a vector. In a further embodiment, the present invention relates to a DNA construct comprising a DNA segment encoding a truncated chicken c-ski protein having the function of c-ski and a vector. The present invention also relates to host cells stably transformed with either to of the two DNA constructs described above, in a manner allowing expression of the protein encoded in the construct.

In yet another embodiment, the present invention relates to a animal having increased muscle size, all of whose cells contain a DNA - k - construct comprising a DNA segment encoding a ski protein and a vector, introduced into the animal, or an ancestor of the animal. The DNA segment may encode the entire protein or a truncated version thereof.

In further embodiment, the present invention relates to an animal having increased muscle size, all of whose cells contain a DNA construct comprising a DNA segment encoding a truncated ski protein having the function of ski and a vector, introduced into the animal, or an ancestor of the animal.

In another embodiment, the present invention relates to a method of stimulating muscle growth or preventing muscle degeneration comprising delivering a DNA construct of the present invention to the muscle under conditions such that the protein of the construct is expressed and muscle growth induced. In a further embodiment, the present invention relates to a method of treating a muscle degenerative disease comprising delivering a DNA construct of the present invention to the effected muscle under conditions such that the protein of the construct is expressed and treatment effected.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure l. The Structure of the c-ski cDNA clones. The lengths of the cDNAs are drawn to scale and the restriction sites indicated, v-ski is shown for comparison. The dotted boxes in v-ski represent the gag region of the gag-ski fusion in the acutely transforming virus SKV. A, B, C and D represent the regions used for generating single- strand probes for si nuclease protection analysis. The numbers next to the arrows correspond to the exon number.

Figure 2. Complete coding sequence and the potential coding region of a cDNA of. the FB29 type.

This assumes that the 5' end sequences of the FB29 are similar to the 5' ends of the FB28 and CEL clones. t indicates the site where FB28 and CEL diverge. The 25 bases found only in the CEL clones are not shown. I indicates the boundaries of v-ski . The exon boundaries are numbered and the alternately spliced exons are boxed. The single base and a ino acid change between c-ski and v-ski is also boxed with a dashed line. The translation termination codon is boxed in thick lines. The potential polyadenylation signals are densely underlined. The AT rich region containing ATTTA sequences that might be involved in mRNA stability is underlined with dashed lines.

Figure 3. Diagrammatic illustration of the alternate mRNAs generated for c-εki locus as deduced from the cDNA sequence analysis. The exons are not drawn to scale. The c-εki mRNAs are shown in relation to v-ski . The dark areas are noncoding regions while the open boxes are the protein coding regions of the cDNA. The dotted boxes on both ends of v-ski are the gag regions of the gag-ski fusion in transforming ski viruses. The relative positions of the putative translational initiation codon and the translation termination codons are also shown. F gure 4. SI nuc ease protect on analys s of total RNA. Fiqure 4A shows the- uniformly labeled single-strand probes used for hybridization are shown schematically below each picture, the thick lines represent the cDNA sequence while the thin lines represent M13 sequence. The overall length of probes and expected lengths of protected fragments are also shown. The RNA hybridization is indicated at the top of each lane. The numbers (8, 10, 12, 15 or

17) above the lane indicates the age of the embryo from which the RNA was prepared. Figure 4B showsProbe A (see Figure 1) contains a Kpnl-Hindlll fragment of the FB27/29 type. This probe produced a fragment of 645 base pairs (bp) and two smaller fragments of 262 and 272 bp, shown by arrows. Figure 4C showsProbe B (see Figure 1) contains a Kpnl-Hindlll fragment of the FB28 type. This probe produced a fragment of 534 bp as shown by the arrows. Smaller fragments were not detected. Figure 4D showsprobe C (see Figure 1) contains a 497-bp Hindlll fragment of the FB27 type linked to M13 sequences. The probe yielded a 497-bp fragment and two smaller fragments of 243/254. Only the 479-bp fragment is marked by an arrow. Figure 4E shows Probe D (see Figure 1) is 1116 bp in length containing a Hindlll fragment of FB28/29 type. Probe D produced a 799-bp fragment which is marked by an arrow.

Figure 5. Sequence homology between c-ski and the pl9 region of gag from avian leukosis virus. The c-ski sequences are from positions 218 to 242 and the pl9 sequence of gag region are from positions 633 to 658. The homologous regions are boxed.

SUBSTITUTE SHEET Figure 6. The c-sJci expression cassette.

The Pvul to Nrul segment shown in the drawing was isolated by gel electrophoresis following double digestion of the plasmid. The linear DNA was used to create the transgenic mice by microinjection of fertilized eggs.

Figure 7. A transgenic mouse that expresses c-ski and a normal litter mate.

The c-ski transgene appears to segregate normally in crosses. The photograph shows a heterozygous mouse that displays the muscular phenotype (foreground) and a DNA-negative litter mate. •Double blind DNA analyses confirmed that the muscular phenotype segregates with the transgene.

Figure 8. Northern transfer analysis of transgene expression.

Panel A shows the analysis of RNA isolated from various tissues of a mouse of the line TG 8566. The upper part of the panel shows an autoradiogram after hybridization with chicken c-εki . The expected position of migration of the c-εki message appropriately transcribed from the transgene is 2.5 kb. The position corresponding to 2.5 kb is marked (εki) . The lower panel shows an autoradiogram from the same filter following hybridization to a chicken 9-actin cDNA. The β- actin cDNA will hybridize not only with 0-actin mRNA but also with other actin messages. The expected position of migration of both 0-actin and α-actin mRNAs are indicated on the right of the panel.

Panel B. The autoradiograms shown in panel B are similar to those shown in panel A except that the RNAs derive from three other transgenic lines. The lines used to prepare the RNAs are indicated - fi ¬ at the top of the figure. The filters shown in panel B were done at the same time; those in panel A were done on a different day.

Figure 9. RNase protection of RNA from the transgene.

The top of the figure shows an autoradiogram of the gel. The first lane contains the antisense RNA probe, without RNase digestion. The next two lanes show the results of digestion following hybridization of the probe either without added RNA or with tRNA. The next eight lanes show the results of hybridization to RNA isolated either from the hearts (heart) or the skeletal muscle (SK muscle) of the four transgenic lines. The next lanes show the results of hybridizing the probe to RNA from skeletal muscle of a mouse that does not carry the transgene (control SK muscle) . The last lane contains molecular weight markers. Below the autoradiogram is a diagram that shows a drawing of the MSV LTR c-εki expression cassette in relation to the antisense RNA probe. The T7 transcript begins in the middle of the c-εki coding region and goes entirely through the MSV LTR into adjacent sequences that derive from pBR322 (marked pBR) . If the transcripts deriving from the transgene initiate appropriately, then a fragment of 984 bases should be protected.

Figure 10. chicken c-εki protein expression in transgenic mice. Extracts were made from the liver (Liv) or skeletal muscle (Sk.M) of control mice (control) or ^"mice carrying and expressing the c-εki transgene. The positions of migration of radioactively labeled molecular weight standards are shown to the left of the figure. _ Q _

Figure 11. Cross sections made precisely through the middle of the plantaris muscle: (a) from control mouse and (b) from a mouse of line TG 8566. Both illustrations are at the same magnification, the size marker is 200 μ. (c)

Higher power illustration from the plantaris of control and (d) from the plantaris of an affected mouse. Size markers in (c) and (d) are 100 μ.

Figure 12. Distribution of fiber diameters in selected muscles from normal and transgenic Mice Panel A. The diaphragm appears normal in transgenic mice that express c-εki . A diaphragm from a mouse that has the muscular phenotype (TG 8566) and a diaphragm from a normal control mouse were sectioned and the number of individual muscle fibers of a given cross-sectional area tallied. Panel B. The anterior tibial muscle is grossly enlarged in mice from the line TG 8566. Transverse sections were prepared from both a transgenic mouse and a control mouse. The number of fibers of each given cross-sectional area were tallied. This muscle is composed of two distinct types of fibers, some of which are smaller, others larger, than the fibers found in the controls (see also Figure 11) .

Figure 13. Transgene expression in specific muscles from TG 8566. 20 μg of total RNA was fractionated by electrophoresis and transferred to nitrocellulose membranes. The transfers were first probed with chicken c-εki , then the filters were stripped and reprobed with a chicken _/9-actin cDNA (Cleveland et al., 1980, Cell 20 , 95-105). The RNA was isolated from the diaphragm, the soleus, or from bulk skeletal muscle (sk muscle) .

Figure 14. Immunofluorescence staining of sections made through the middle of the Rhomboideus capitis muscle of an affected mouse, (a) staining with monoclonal antibody NOQ7 5 4D, specific for slow MHC. Slow fibers are not hyper- trophied. (b) staining with monoclonal antibody SC 711 specific for Ila MHC. Ila fibers are not hypertrophied. (c) is with monoclonal antibody

2G3 which reacts with all fast MHC isoforms. All hypertrophied fibers stain with this antibody, (d) is with m/a BF-F3 specific for lib MHC. Many, but not all, hypertrophied fibers stain. Magnification x 230.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to a DNA segment encoding all, or a unique portion, of a chicken c-ski protein. The DNA segment may encode one of several chicken c-ski proteins, for example, FB29, FB28 and FB27. A "unique portion" as used herein is defined as consisting of at least five (or six) amino acids or correspondingly, at least 15 (or 18) nucleotides. The invention also relates to DNA constructs containing such DNA segments and to cells transformed therewith.

The present invention relates to DNA segments that encode the amino acid sequence of exon 6 given in Figure 1 or the amino acid sequence of exon 7 given in Figure 1. The present invention also relates to DNA segments that in addition to exon 7 further comprise at least four exons selected from the group consisting of: exon 1, exon 2, exon 3, exon 4, exon 5 or exon 6, given in Figure 1. Examples of such DNA segments include FB29, FB28, and FB27.

DNA segments to which the invention relates also include those encoding substantially the same proteins as those encoded in the exons of Figure l which includes, for example, allelic forms of the Figure l amino acid sequences. The invention also relates to DNA fragments complementary to such sequences. A unique portion of the DNA segment or the complementary fragment thereof of the present invention can be used as probes for detecting the presence of its complementary strand in a DNA or RNA sample.

The present invention further relates to DNA constructs and to host cells transformed therewith. In one embodiment, the DNA constructs of the present invention comprise a DNA segment encoding a c-ski protein of the present invention and a vector, for example, pMEX neo . In another embodiment, the DNA constructs comprise a DNA segment encoding a truncated c-ski protein having the function of c-ski (such as, for example, ΔFB29) and a vector (for example, pMEX neo) . The DNA construct is suitable for transforming host cells. The host cells can be procaryotic or, preferably, eucaryotic (such as, mammalian). The present invention further relates to animals, such as, for example, domestic livestock, having increased muscle size. The development of such strains of domestic livestock with increased muscles continues to be a major goal of conventional breeding schemes. (Domestic livestock as used herein refers to animals bred for their meat, such as, for example, pigs, chickens, turkeys, ducks, sheep, cows and fish, particularly, trout and catfish) .

Introduction of various genes into the germ lines of mice and of some types of domestic livestock is relatively routine to those skilled in the art. The present inventors have produced mice having increased muscle size by introducing a DNA construct comprising ΔFB29 and the pMEX neo vector into fertilized eggs. Resulting founder mice and their offspring have the DNA construct in all their cells, somatic and germ.

Introduction of DNA constructs encoding a ski protein (such as, for example a c-ski protein) , into fertilized eggs of animals (such as, by microinjection) results in strains of animals having increased muscle development and decreased fatty tissue. As one skilled in the art will appreciate, the animals with increased muscle size of the present invention can also be produced using DNA encoding a ski protein from various species (chicken being just one such example) . Furthermore, animals of the present invention can be produced using a DNA construct encoding protiens related to ski , such as, for example, εno gene.

The DNA segment ΔFB29 generated ^"by a frameshift mutation results in a truncated protein. However, as one skilled in the art will appreciate, the transgenic animals of the present invention can also be generated by DNA constructs containing DNA segments encoding a full length εki protein, a portion of a ski protein, such as, one or two exons or a biological active deletion derivative, such as, for example, v-εki , which represents a truncated c-ski fused to a viral protein. Further, it is also recognized that the selective expression of the protein in muscle tissue may result from DNA constructs created in vectors other than pMEX neo .

The present invention also relates to a method of stimulating muscle growth and preventing muscle degeneration in an animal, such as for example, a human. A possible treatment for injuries resulting in loss of muscle tissue and neurological injuries resulting in degeneration of the muscle would be to stimulate muscle growth. In the case of lose of muscle this would involve stimulating regrowth of the tissue. Whereas in the case of neurological injuries, the muscle growth would need to be rendered independent of the missing nerve stimulus. According to the present invention, muscle growth could be stimulated by delivering a DNA construct encoding a ski protein to the muscle tissue under conditions such that the protein encoded in the construct is expressed. The construct can be targeted and delivered to the muscle using standard methods known to those skilled in the art.

The present invention further relates to a method of treating a muscle degenerative disease such as, for example, muscular dystrophy and amyotrophic lateral sclerosis (also known as Lou Gehrig disease) . Treatment would comprise delivering a DNA construct of the present invention to the effected muscle under conditions such that the protein encoded in the construct is expressed and treatment effected.

Examples

Screening of cDNA Library Two chicken cDNA libraries were screened with a v-ski probe using standard protocols (Maniatis et al., 1982, Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY) . One library was made from poly A^* mRNA isolated form the body wall of 10-day embryos and the other from mRNA isolated from AEV- transformed chicken erythroblasts. Four distinct c-ski cDNAs were isolated; three were from the body wall library (these cloned were designated FB27, FB28 and FB29) and one cDNA clone from the erythroblast library (designated CEL) . These cDNAs included sequences extending both 5' and 3' of the portion of ski present in the virus. The cDNAs demonstrate that v-ski derives from a single cellular gene and suggest that multiple c-ski mRNAs, encoding distinct ski proteins, are produced from the c-ski locus by alternate splicing (Leff et al., 1986, Ann. Rev. Biochem. 55 , 1091-1117) , adding to a growing list of oncogenes known to produce multiple mRNAs in this fashion (Ben-Neviah et al., 1986, Cell 44 , 577-

586; Levy et al., 1987, Mol. Cell. Biol. 7, 4142- 4145; Martinez et al., 1987, Science 237 , 411-415; McGrath et al., 1983, Nature 304 , 501-506).

Structure of the cDNA clones The structure of all of the c-ski cDNA clones that have been characterized by DNA sequencing (Sanger et al., 1977, Proc. Natl. Acad. Sci. USA 74, 5463-5467) and their relationships to v-ski are presented in Figure 1. Only FB28 and CEL have sequences 5¹ of the v- ski sequences; both FB27 and FB29 are truncated at the 5' end. Assuming that the missing sequences at the 5¹ end of FB29 are similar to those in FB28 and CEL, a composite nucleotide sequence and the deduced amino acid sequence for a cDNA of the FB29 type is shown in Figure 2. The first ATG with a substantial downstream open reading frame is located at nucleotide position 168. Upstream of this ATG no reading frames are open, suggesting that these sequences represent the 5' untranslated region. Based on the cDNA sequence analysis and comparisons with the positions of the splice donor and acceptor sites known from the genomic sequence

(Stavnezer et al., 1989, Mol. Cell. Biol. 9 , 4038- 4045) , exon boundaries have been derived (see

Figure 3) . As seen in the figure, c-ski sequences are distributed over seven exons. FB29 contains all seven exons; FB28 and FB27 lack exon 2 and exon 6, respectively. This differential splicing affects the protein coding potential of the three cDNAs.

Differential splicing of exon 2 deletes 37 amino acids without affecting the coding potential of the open reading frame downstream. Differential splicing of exon 6, however, affects the coding potential of exon 7. If exon 5 is spliced to exon 7 (seen in FB27 cDNA) , a translation termination codon is generated at the splice junction and exon 7 becomes a noncoding exon. However, if exon 5 is spliced to exon 6 and exon 6 to exon 7, as in FB28/29, then the open reading frame continues in exon 7 for 417 nucleotides encoding an additional 129 amino acids. Assuming that the missing 5' ends of the FB29 and FB27 mRNAs contain sequences identical to those present in CEL and FB28, then translation of mRNAs corresponding to the three different types of cDNAs would lead to the generation of three proteins, one containing 750 amino acids (from

FB29) , the second 713 amino acids (from FB28) , and a third protein containing 510 amino acids (from FB27) .

The CEL clone is missing 3' sequences. cDNAs that derive from the body wall (FB) library have long 3' untranslated regions that contain a 95- base pair (bp) AT-rich region from nucleotide 2803 to 2898. Within this region there are two copies of a sequence ATTTA that has been implicated in mRNA destabilization in a variety of transiently induced mRNAs including c-rayc, interferon, c-jun , and c-fos (Meijlink et al., 1985, Proc. Natl.

Acad. Sci. USA 82 , 4987-4991; Ryder et al., 1985, Proc. Natl. Acad. Sci. USA 85 , 1487-1491; Shaw et al., 1986, Cell 46 , 659-667). Addition or deletion of these sequences have also been shown to affect the transformation potential of c-fos

(Meijlink et al., 1985, Proc. Natl. Acad. Sci. USA 82 , 4987-4991).

The c-ski cDNAs contain two potential poly(A) signals (AATAAA) located at positions 3348 and 4167. Although all three clones isolated from the FB library end at the same position, none has a poly(A) tail; therefore, it is likely that the 3' ends of the c-ski mRNAs are not contained in these clones. The 4.2-kb cDNAs that were isolated and characterized are smaller than the 5.7 -8.0 and 10.0-kb mRNAs detected by Northern transfer analysis (Li et al., 1986, J. Virol. 57 , 1065- 1072) . This discrepancy has not been explained, however, it is suggested that the clones isolated so far, which lack poly(A) tails, also lack sequences from the 3' ends of the mRNAs.

The 5' End of c-ski mRNA(s)

Sequence comparisons at the 5' end reveals that both FB28 and CEL are colinear up to a position 89 bp upstream of the putative translational initiation ATG codon. FB28 has an additional 76-bp while CEL has 25 bp that are different from those found in FB28. This region where the two sequences diverge has been compared with sequences from genomic clones (Stavnezer et al., 1989, Mol. Cell. Biol. 9 , 4038-4045) and the upstream sequences in the FB28 are colinear with the genomic sequences.

Examination of the sequences in the genomic DNA at the point of divergence of CEL and FB28 does not reveal consensus sequences for donor or acceptor splice sites. In order to confirm the authenticity of the clones, SI nuclease protection analysis was carried out. The results demonstrated that the sequences present in FB28 are expressed as mRNA in normal embryos. Similar SI analyses have provided no evidence for the presence in mRNA of the 25 bp at the extreme 5¹ end of the CEL clone. It is suggested that the first 25 bases of the CEL clone are the result of a cloning artefact, and that the sequences found in the FB28 clone are expressed in c-ski mRNA(s) . In an attempt to determine how much of the 5' untranslated region of the c-εki mRNA(s) are contained in FB28, primer extension analysis was carried out. Copying the 5' segment present in

FB28 should give an extension product of about 280 bases. However, a primer extension product of 220 bases was seen. SI analyses data have shown that this segment is expressed in RNA. It is possible that the observed primer extension product is the result of premature termination; however, it is also possible that there are multiple 5' ends for the c-ski mRNAs.

Organization of the Internal Exons Sequence comparisons in the central portion of FB27, FB28 and FB29 and CEL cDNAs reveal that FB28 lacks a small region of ill bp (exon 2) from 1079- 1191 that would eliminate 37 amino acids. The genomic sequence (Stavnezer et al., 1989, Mol. Cell. Biol. 9 , 4038-4045) in the corresponding region reveals the existence of consensus splice donor and acceptor sequences at the boundaries of the deletion suggesting that the cDNA derives from a differentially spliced mRNA.

In order to confirm the existence of mRNAs of the FB28 and of the FB27/29 type, SI analysis was carried out on total RNA. Total RNA was isolated from 8, 10, 12, 15 and 17-day-old chicken embryos using standard protocols (Chirgwin et al., 1979, Biochemistry 18 , 5294-5299). Approximately 20 to 30 μg of total RNA was used for nuclear SI analysis using standard procedures (Maniatis et al., 1982, Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY) . A uniformly labeled single-stranded probe spanning the region between the Kpnl and Hindlll sites of FB29/27 or FB28 (probes A and B of Figure 1) was hybridized with total cellular RNA. As shown in Figure 4B, hybridization to mRNA and subsequent SI digestion of a probe derived from FB29 produced protected fragments of 645 bp indicating hybridization to mRNA of the FB27/29 type and the 262/272 bp fragments expected if the probe hybridized with mRNA of FB28 type. FB28 (probe B) protected a fragment of 534 bp (see Figure 4C) . Smaller fragments from the hybridization of the FB28 probe to mRNAs of the FB27/29 type were not observed; however, SI does not always cleave a DNA probe efficiently opposite a looped out region of RNA. These experiments indicate that mRNA corresponding to both the FB27/29 type, containing the second exon, and FB28 type, missing the second exon, are expressed in normal cells.

Organization of the 3' Coding Exons

Sequence comparison of the three fibroblast

SUBSTITUTESHEET co a segment (exon 6) that is absent in FB27. Examination of the genomic DNA at a position corresponding to the position where the cDNAs diverge reveals a consensus splice donor

(Stavnezer et al., 1989, Mol. Cell. Biol. 9 , 4038- 4045) . The two cDNAs appear to derive from splicing events that either include or exclude exon 6. Downstream of this alternate exon the 3¹ portion of all three cDNAs are identical.

To investigate whether both types of cDNAs represent normal cellular mRNA, Si analysis was carried out. Uniformly labeled single-stranded probes were generated from an 799-bp Hindlll fragment (see probe D, Figure 1) from the 3* end of FB28 and a 497-bp Hindlll fragment from the 3' end of FB27 (probe C) subcloned in M13mpl8. Single-stranded probes were prepared and were hybridized to 20 μg of total RNA prepared from whole chicken embryos of different ages. As seen in Figure 4E, hybridization of the RNA to probe D (FB28) followed by Si treatment produced a 799-bp fragment that would be expected if mRNA of FB28/29 type was present. Smaller fragments that would be generated by the FB28 probe hybridizing with mRNA of the FB27 type were not detected. SI digestion of the FB27 probe (probe C) produced a fragment of 497-bp and also two smaller fragments of 243 to 254 bp (see Figure 4D) . It is likely that these smaller fragments derive from SI cleavage of the FB27 probe at the site where the probe differs from mRNAs of the FB28/29 type. These data confirm the existence of the FB28/29 mRNA, but suggest that, if FB27 mRNA exists, it is likely to be present at a lower level than FB28/29.

Differential Splicing

SUBSTITUTE SHEET - 20 -

A schematic illustration of the differential splicing of three c-ski mRNAs as deduced from the nucleic acid sequence analysis of the cDNA clones is shown in Figure 3. SI analysis of total RNA derived from chicken embryos has confirmed the existence of two classes of mRNAs, those that do, and those that do not, contain exon 2. Using a similar protocol, the existence of mRNA that contains exon 6 was confirmed. Utilizing this technique, the existence of mRNA lacking exon 6 which would correspond to FB27 cDNA, could not be demonstrated; however, several lines of argument suggest that the FB27 cDNA is not a simple cloning artefact. The sequences absent from FB27 are bounded by apparent splice junctions (as judged by an examination of both the cDNA sequences and the available genomic DNA sequence) . Although cDNAs are occasionally obtained that contain one or more introns, presumably because a partially processed mRNA was reverse transcribed, isolating cDNA artefactually missing an exon is less likely on theoretical grounds, and seems to occur rarely, if at all, in the manufacture of a cDNA library. For these reasons, the interpretation that FB27 represents a real, if relatively rare, c-εki mRNA is currently favored.

Comparison of c-ski with v-ski

Transduction of c-oncs by retroviruses often results in truncation, deletion, or point mutation(s) in the c-αnc (Bishop, J. M., 1983, Ann Rev. Biochem. 52 , 301-354). Comparing c-ski and v-ski sequences shows that the v-ski gene is truncated at both 5• and 3' ends and represents only part of the c-ski coding region (see Figure l and 3) . v-ski sequences begin at position 244 (the putative initiator ATG is at position 168) and end at position 1541 (see Figure 3) . The biological significance of the 5¹ and 3' truncations are unknown. There is only one base change in the v-ski relative to the c-ski, at position 1284 where v-ski has a 'C and c-ski has a 'T'. This base change alters the amino acid from Trp in c-ski to Arg in v-ski . Deletion analysis of v-ski implies that this amino acid change does not play a significant role in the transforming potential of v-ski . The biological activities of the cDNAs are being evaluated by expressing the coding regions from the cDNAs using replication-competent retroviral vectors (Hughes et al., 1987, J. Virol. 61 , 3004-3012). As shown in Figure 5, 18 of 20 bp are identical between c-ski and the pl9 region of gag in the parental ALV. This region of ho ology contains the 5¹ junction between viral sequences and v-ski . With such a long stretch of homology, it is impossible to assign the 5' recombinational joint precisely. No substantial homology can be seen precisely at the 3' ski/ALV joint. However, just downstream of the 3 ' junction, there is an ALV sequence that is closely homologous to a segment of c-ski found just 3' of the v-ski/AVL junction. It is possible to invoke this region of homology in the alignment of the nucleic acids involved in the recombination event.

Transduction mechanism The exact mechanism(s) by which retroviruses acquire cellular oncogenes remains uncertain. It has been suggested that as a first step, the viral DNA is integrated next to or within a cellular oncogene (Bishop, J. M. , 1983, Ann Rev. Biochem. 52 , 301-354). The retroviral and cellular sequences can then be fused by DNA deletion events (Bishop, J.M. 1983, Ann. Rev. Biochem. 52 , 301- 354; Czernilofsky et al., 1983, Nature 301, 736- 738) or RNA read through (Herman et al. , 1987, Science 236, 845-848; Nilsen et al., 1985, Cell 41 , 1719-726); however, homology is not known to be involved in either of these processes. In the case of ski, a comparison of c-ski sequences with the pl9 coding region implies that some event in the generation of the ski viruses involved homologous recombination. However, the homologous recombination event may be secondary. If the original event were nonhomologous and generated a viral genome that contained a direct repeat, it would be expected, as has been observed for Rous sarcoma virus, that sequences between the repeats could be lost by recombination during viral passage, presumably through copy choice during reverse transcription. It is possible, however, to propose models in which the oncogene is acquired by homologous recombination between the c-onc and the replication competent viruses. Short stretches of homologies have been observed at both ends of viral oncogenes (Banner et al., 1985, Mol. Cell. Biol. 5 , 1400-1407; Van Beveren et al., 1983 Cell 32 , 1241-1255) or only at the left hand or 5' recombination joint (Besmer et al., 1986, Nature (London) 320 , 415-421; Roebroek et al., 1987, J. Virol. 61 , 2009-2016). The present data does not allow the determination of whether the 5' homologous recombination event was primary or secondary.

Transgene and Generation of the Transgenic Mice

The largest form of the isolated cDNAs, that is FB29 which contains sequences that derive from all seven coding exons of c-ski, and judged by DNA sequence, encodes a c-ski protein of 750 amino acids, was used to create the derivative ΔFB29. ΔFB29 has a fra eshift mutation at position 1475 in the fifth coding exon (one C in a run of five Cs was lost in the frameshift mutant) , and is predicted to give rise to a protein of 448 amino acids of which the first 436 are identical to the first 436 amino acids of the FB29 form of c-ski (the last 12 amino acids are past the frameshift mutation and thus differ from those of the FB29 form of c-ski) . ΔFB29 used in the generation of a the transgene is shown schematically in Figure 6.

The construction of the ski portion of the transgene is already described (Sutrave and Hughes, 1989, Mol. Cell. Biol. 9 , 4046-4051; Sutrave et al., 1990, Mol. Cell. Biol. 10 , 3137- 3144) . Briefly, a truncated chicken c-ski cDNA called ΔFB29 had been previously cloned into the adaptor plasmid Clal2Nco. The ΔFB29 segment was released from the adaptor plasmid by Clal digestion and the 5' overhangs filled in using the Klenow fragment of E. coli DNA polymerase I and all four dNTPs. This blunt-ended fragment was ligated to the pMEX neo vector which have been digested with EcoRl restriction enzyme and blunt- ended with the Klenow fragment. Clones were selected that had inserts in the correct orientation and were digested with both Pvul and Nrul restriction endonucleases. These enzymes release a segment that contains the ΔFB29 cDNA flanked by an MSV LTR and the SV40 polyA signal (see Figure 6) . This fragment was gel purified and used to inject fertilized mouse eggs [Hogan et al., 1986, Manipulating the mouse embryo. A Laboratory Manual. (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory) ] .

The ΔFB29 clone was placed in the pMEX expression plasmid in such an orientation that the truncated c-ski cDNA between an MSV LTR and an SV40 polyadenylation site (see Figure 6) . The plasmid was digested with Pvul and Nrul to release the expression cassette. The expression cassette was purified by gel electrophoresis and introduced into fertilized mouse eggs by microinjection. Forty-four founder mice were obtained after two independent injections. The mice were identified by dot blot analysis of DNA isolated from tail clips. This analysis was confirmed by Southern transfer. Three of the 44 founder mice showed a distinct muscle phenotype (TG 8566, TG 8821, and TG 8562) . These three founders and a single mouse that contained unrearranged copy of the complete transgene but did not show any phenotype (TG 8542) were used to generate lines. Southern transfer analysis of DNA from TG 8566, TG 8821, TG 8562, and TG 8542 suggests that the site of integration of the transgene in each line is different and the copy number varies from approximately 5-35 copies per genome.

DNA positive mice from the three lines (TG 8566, TG 8821 and TG 8562) had a similar distinct appearance resulting from abnormal muscle growth. Although the three lines of mice carry an oncogene, none of the lines appears to have an increased incidence of tumors. This result is not totally unexpected, since the v-ski virus is not tumorigenic in chickens unless the birds are injected with infected cells (E. Stavnezer, 1988, in The Oncogene Handbook, E. P. Reddy, A. Skalka, and T. Curran, eds. (Amsterdam: Elsevier Science Publishing Co.), pp. 393-401). The three strains of mice do not express high levels of the trans- gene except in skeletal muscle. This is consistent with the interpretation that c-ski affects skeletal muscle cells directly and not as a secondary consequence of altered motor neuron function. The majority of the skeletal muscles are involved. Mice with this phenotype can be readily identified by looking for enlarged limb and jaw muscles (Figure 7) . Since the phenotype was obtained with three separate founders, the most reasonable explanation is that the phenotype was caused by the chicken c-ski transgene. Therefore, all four lines of mice were examined, the three with the muscular phenotype and the one line that did not have an observable phenotype, for the expression of the transgene.

The expression cassette was also introduced into fertilized pig eggs using standard procedures known to those in the art. Founder pigs show the expected muscle specific selection of expression of the gene and accordingly, it is expected that the pigs will have the same muscle phenotype as that seen in the founder mice.

DNA/RNA Analysis

Total cellular DNA and RNA were isolated by standard procedures. For RNA isolation, tissues were frozen in liquid nitrogen immediately following dissection and homogenized in RNAzol (Cinna Biotex) and processed according to the manufacturer's recommendation. For Northern transfer analysis, approximately 20 μg of total RNA from different tissues was fractioned by electrophoresis on 1.5% agarose gels containing 2.2 M formaldehyde. The RNA was transferred to nitrocellulose membranes and probed either with a nick-translated chicken ski cDNA or a chicken β- actin cDNA. The coding region of the 3-actin cDNA cross reacts with the messages for the other actions and can be used to validate the quantity and quality of RNA from most tissues. Total RNA from spleen, lung, brain, kidney, liver, stomach, heart, and leg (skeletal) muscle was isolated as described and the results are shown in Figure 8. All three lines with the phenotype (TG 8566, TG 8821 and TG 8562) expressed a 2.5-kb chicken c-ski specific transcript at high levels in skeletal muscle; however, some lines of mice showed low levels of chicken c-ski RNA in other tissues. The TG 8562 line has RNA from the transgene in the heart, although at a lower level than in skeletal muscle. Histopathology of hearts from TG 8562 mice showed that there is no significant effect on this tissue. Line TG 8542, which does not show any phenotype, had a much lower level of RNA from the transgene in muscle than did the lines that showed the phenotype. The observation that expression was restricted to muscle was unexpected since the MSV LTR has been shown to express in a variety of tissues when linked to other genes (Khillan et al., 1987, Genes Dev. 1, 1327-1335) .

To determine whether the transcript was initiated at the proper site in tissues expressing the ski transgene, RNase protection assays were carried out as described by Melton (Melton et al., 1984, Nucl. Acids Res. 12 , 7035-7036) . A 1.8-kb Pvυl to Bgll fragment was subcloned in Bluescript KS vector and used to generate radioactively labeled RNA from the T7 promoter. Approximately 10 μg of total RNA was hybridized with 5 x 10⁵ cpm of probe. The hybridizations were carried out overnight at 50°C in 80% formamide and lx buffer (5x hybridization buffer is 0.2 M Pipes, pH 6.4, 2 M sodium chloride, 5 mM EDTA) . After hybridization, the samples were diluted in ribonuclease digestion buffer (10 mM Tris CI, pH 7.5, 0.3 M sodium chloride, 5 mM EDTA) and treated with RNase TI at a concentration of l u/μl for 60 min at 30°C. The RNase digestions were stopped by adding 10 μl of 20% SDS and 4 μl of Proteinase K (stock 10 mg/ml) and incubating at 37°C for 15 min. The digested samples were extracted with phenol chloroform (1:1 mixture) and ethanol precipitated with carrier tRNA. The pellet was rinsed once with 70% ethanol, dried and dissolved in formamide containing bromophenol blue and xylene cyanol dyes. The samples were denatured at 100^βC and separated on 6% polyacrylamide gels containing 7.5 M urea.

Uniformly labeled antisense RNA was generated by T7 RNA polymerase from a fragment that spans the MSV LTR and c-ski (see Figure 9) . Using RNA from the three positive transgenic lines, a protected fragment of approximately 980 bases was seen, which is the expected size if the transcript is initiated at the authentic initiation site within the MSV LTR (Figure 9) . This analysis also gives a more quantitative estimate of the level of transgene RNA in the heart and skeletal muscle of both the phenotypically positive and the pheno- typically negative lines of mice. Figure 9 shows that the level of transgene RNA in the heart of TG 8821 is much lower (estimated at perhaps 1/10- 1/20) than the level found in the skeletal muscle. In addition, this analysis shows that the phenotypically negative line, TG 8542, has a low but detectable level of transgene RNA in skeletal muscle. These data suggest not only that the muscular phenotype is associated with the expression of the chicken c-εki transgene, but also that a minimum threshold level of c-εki RNA must be reached to produce the muscular phenotype. In fact, these data suggest that the minimal threshold level to see an effect of the transgene is high, probably several thousand-fold the levels of endogenous c-ski expression in the chicken tissues examined (Sutrave and Hughes, 1989, Mol. Cell. Biol. 9 , 4046-4051).

Protein Analysis

The underlying assumption is not that the expression of the c-ski RNA gives rise directly tt the muscular phenotype, but rather that the phenotype results from the presence of the c-ski protein. Accordingly, the c-ski protein was looked at in Western transfer assays. Although rabbit antisera have been prepared that specifically recognize the 50-kd form of c-ski (Sutrave et al., 1990, Mol. Cell. Biol. 10 , 3137- 3144) , these antisera do not work well in Western transfer assays. Mouse monoclonal antibodies that recognize c-ski and that work well in Western transfer assays have been developed, however, the use of these reagents presents a technical problem. These monoclonals were not available in sufficient quantity to permit direct labeling. Indirect labeling procedures using, for example, labeled rabbit anti mouse detect not only the anti ski monoclonal but also the endogenous mouse heavy chain, which comigrates with the 50-kd form of c- ski made from the transgene. To avoid this problem, extracts of muscle and control tissue (liver) from normal controls and from the transgenic mice were prepared. The endogenous mouse antibodies were removed from these extracts by precipitation with rabbit anti mouse antibody as described below.

For detection of εki protein in tissues from the transgenic and control mice, 1-5 mg of tissue was homogenized in 1 ml of RIPA buffer, 20 mM Tris CI pH 7.5, 150 mM NaCl, 0.5% SDS, 0.5% NP40, 0.5% - 2^Q -

sodium deoxycholate, 1 mM EDTA, 1 mM PMSF, and 35 μ/ml of aproteinin. The homogenate was clarified by centrifugation at 10,000 rpm for 10 min. Mouse IgG was removed from 100 μl of the supernatant by incubation with 10 μl of 1 mg/ml rabbit anti mouse IgG (in PBS) for 2 hr on ice. The complex was removed by adding 100 μl of 40% protein A sepharose beads in RIPA buffer. The resulting supernatant was collected and 20 μl was fractionated on 10% SDS polyacrylamide gels. The proteins were transferred to nitrocellulose membranes overnight in buffer containing 0.125 M Tris CI, 0.092 M Glycine and 20% Methanol, pH 8.3. The filters were blocked with 4% dry nonfat milk in TBS buffer (0.5 M Tris CI, pH 7.4 and .2 M sodium chloride) for 2 hr at room temperature and incubated with a mixture of three anti-ski monoclonal antibodies at a dilution of 1:3000 for 2 hr at room temperature and were then washed 3 with TBS. Secondary incubations with rabbit anti mouse IgG were done for 2 hr at room temperature (1:2000 dilution from a l mg/ml stock). The filter was washed as described above and finally incubated with 5 μCi of ^,J5I protein A (Amersham, sp. act. 30 mCi/mg) for 2 hr at room temperature. The filter was washed 3x with TBS and exposed to XAR Kodak film at -70°C for 6 days.

Only affected tissues (skeletal muscle) from the transgenic animals contain the 50-kd c-ski protein (Figure 10) . These data also suggest that there may be differences in the level of the c- εki protein in the muscles of the three positive lines; however, the complexities of the manipulations in this experiment make quantitative interpretation a questionable proposition.

Histology For histology, selected muscles from the line TG 8566 were isolated so that they remained attached at their origin and insertion and they were then fixed in 2% formaldehyde, 2% gluteraldehyde. Fixed muscle were transected precisely through the middle of the muscle belly and embedded in JB4 plastic (Polysciences, PA) . For im unocytochemistry, tissues were snap frozen in isopentane cooled in liquid nitrogen. The procedure for immunochemical staining was as outlined in Narusawa et al. [(1987), J. Cell. Biol. 104 , 447-459].

The results showed that not only is expression of the transgene and its effect restricted to muscle, the expression and the effects appear, at least in the line TG 8566, to be confined to certain muscles, and apparently to specific fiber types. However, the affected fibers are not all of the same type. The myocardium was normal and there were no abnormalities of visceral smooth muscle in these animals. Figures Ila and lib compare cross sections made precisely through the middle of the plantaris muscle from mature male controls and transgenic mice. Cross sectional area of the control is 2.7 μ' and that of the TG 8566 mouse is 9.4 μ², more than twice the control value. This massive growth is generalized. Comparable increases in cross section were found in almost all axial and appendicular muscles throughout male and female mice of line TG 8566. Only three muscles were found that appear to be normal: the tongue, the diaphragm, and the soleus; these are the same size in transgenic as in control muscles (see Figure 12) . RNA was isolated from the diaphragm and soleus muscles of TG 8566 mice. Northern transfer analysis shows that the level of chicken c-ski RNA is much lower in these two phenotypically normal muscles than in the affected muscles from the same line (Figure 13) .

The most obvious additional gross morphologic abnormality is that transgenic animals were almost totally devoid of fat whereas control animals contained substantial amounts of subcutaneous and intraperitoneal fat. For this reason, there is little difference in weights between control and TG 8566 mice. There are also skeletal abnormalities; the tibia of transgenic animals is normal in size but was bowed cranially, apparently as an adaptation to accommodate the more than two¬ fold increase in size of the anterior tibia and extensor digitorum muscles.

The muscles in control mice are made up of fibers with a range of cross sectional areas. The range of sizes is greatly extended in TG 8566 mice (Figures 11 and 12) . Not all fibers are affected; the hypertrophy is limited to a select population of fibers (Figure lid) . Hypertrophy of these fibers apparently accounts for the increase in muscle mass in line TG 8566. No evidence was found that the numbers of fibers is significantly increased in individual muscles. For example, the population of fibers in the control plantaris is 912 ± 111 (n=3) and in TG 8566 is 991 ± 87 (n=3). Similarly, no difference in the total number of fibers in the extensor digitorum longus muscle of TG 8566 and control mice was found.

To investigate which types of fibers are affected in the line TG 8566, three monoclonal antibodies were used that specifically recognize slow myosin heavy chains (MHC) (NOQ7 5 4D, Narusawa et al., 1987, J. cell Biol. 104 , 447-

459), all fast MHCs (2G3, Narusawa et al., 1987, J. Cell Biol. 104 , 447-459) type II a fast MHC (SC 7 11, Schiaffino et al., 1989, J. Muscle Res. and Cell Motil. 10 , 197-205) and type lib fast MHC (BF-F3, Schiaffino et al., 1989, J. Muscle Res. and Cell Motil. 10 , 197-205). Figure 14 shows immunofluorescent staining of sections from the rhomboideus capitis muscle with these three antibodies. No evidence was found that slow fibers are enlarged (Figure 14a) and the total number of slow fibers in the rhomboideus capitis from a TG 8566 mouse (120) approximates the number found in rhomboideus capitis from control mice (117) . Type Ila fibers are also not affected (Figure 14b) . Many of the type Ila fibers lie between hypertrophied fibers and are frequently distorted in shape as if compressed by the expansion of their neighbors (Figure 14b) .

The small, fast Ila fibers and all of the hypertrophied fibers stain with the monoclonal antibody 2G3 (Figure 14c) indicating that all hypertrophied fibers are fast. In the rhomboideus capitis, most, but not all, large fibers also stain with the monoclonal antibody BF-F3 that is specific for lib MHC (Figure 14d) . In the plantaris, there is more variation in reactivity and only 50% of hypertrophied fibers stain with BF-F3. These results show that the hypertrophic modification of fibers in TG 8566 mice involves at least two types of fast fibers. One of these is type lib. By exclusion, we suggest that the others are IIx fibers (Schiaffino et al., 1989, J. Muscle Res. and Cell Motil. 10 , 197-205; Termin et al. , 1989, Histochemistry 92 , 453-457; Gorza, 1990, J. Histochem. Cytochem. 38 , 257-265).

Variation in staining of hypertrophic fibers with the actomyosin ATPase histochemical reaction after acid pre-incubation, DPNH staining for mitochondrial enzyme activity and PAS staining all support the conclusion that more than one fast fiber type is affected in this line of transgenic mice. These results also support the interpretation that both lib and IIx fibers are hypertrophied.

Occasional necrotic and regenerating fibers were found in some, but not all, muscles of line TG 8566 mice. In the hind limb, these appeared to be most prevalent in the anterior tibial muscle; they were never found in the rhomboideus capitis, a superficial muscle of the neck.

All publications mentioned hereinabove are hereby incorporated in their entirety by reference.

While the foregoing invention has been described in some detail for purposes of clarity and understanding, it will be appreciated by one skilled in the art from a reading of this disclosure that various changes in form and detail can be made without departing from the true scope of the invention and appended claims.

Claims

WHAT IS CLAIMED IS:

1. A DNA segment encoding a chicken c-ski protein or a DNA fragment complementary to said DNA segment.

2. A DNA segment comprising exon 6 defined in Figure 1.

3. A DNA segment comprising exon 7 defined in Figure 1.

4. The DNA segment according to claim 3, further comprising at least four exons selected from the group consisting of: exon 1, exon 2, exon 3, exon 4, exon 5 and exon 6, defined in Figure l.

5. A DNA construct comprising: i) said DNA segment according to claim 1; and ii) a vector.

6. The construct according to claim 5 wherein said vector is pMEX neo.

7. A DNA construct comprising: i) a DNA segment encoding a truncated c-ski protein having the function of c-ski; and ii) a vector.

8. The DNA construct according to claim 7, wherein said DNA segment is ΔFB29.

9. The DNA construct according to claim 7, wherein said vector is pMEX neo.

10. A host cell stably transformed with said construct according to claim 5, in a manner allowing expression of said protein encoded in said construct.

11. A host cell stably transformed with said construct according to claim 7, in a manner allowing expression of said protein encoded in said construct.

12. The host cell according to claim 10 or claim 11 which is a mammalian cell.

13. A animal having increased muscle size, all of whose cells contain a DNA construct comprising a DNA segment encoding a ski protein and a vector, introduced into said animal, or an ancestor of said animal.

14. The animal according to claim 13 wherein said ski protein is a c-ski protein.

15. The animal according to claim 13 wherein said protein is a chicken c-ski protein.

16. The animal according to claim 13 wherein said vector is pMEX neo .

17. The animal according to claim 13 which is domestic livestock.

18. The animal according to claim 17 which is a pig.

19. The animal according to claim 13 which is a mammal.

20. The mammal according to claim 19 which is a mouse.

21. A animal having increased muscle size, all of whose cells contain a DNA construct comprising a DNA segment encoding a truncated ski protein having the function of ski and a vector, introduced into said animal, or an ancestor of said animal.

22. The animal according to claim 21 wherein said protein is a chicken c-εki protein.

23. The animal according to claim 21 wherein said DNA segment is ΔFB29.

24. The animal according to claim 21 wherein said DNA segment encodes a v-εki protein.

25. The animal according to claim 21 wherein said vector is pMEX neo.

26. The animal according to claim 21 which is domestic livestock.

27. The animal according to claim 26 which is a pig.

28. The animal according to claim 21 which is a mammal.

29. The mammal according to claim 28 which is a mouse.

30. A method of stimulating muscle growth or preventing muscle degeneration comprising delivering a DNA construct comprising a DNA segment encoding a ski protein and a vector, to said muscle under conditions such that said protein of said construct is expressed and muscle growth induced.

31. A method of treating a muscle degenerative disease comprising delivering a DNA construct comprising a DNA segment encoding a ski protein and a vector, to the effected muscle under conditions such that said protein of said construct is expressed and treatment effected.

32. The method according to claim 31 wherein said muscle degenerative disease is muscular dystrophy or a yotrophic lateral sclerosis.