WO2001011950A1

WO2001011950A1 - Transgenic animals that produce altered wool

Info

Publication number: WO2001011950A1
Application number: PCT/CA2000/000945
Authority: WO
Inventors: Costas N. Karatzas; Yue-Jin Huang
Original assignee: Nexia Biotechnologies, Inc.
Priority date: 1999-08-18
Filing date: 2000-08-18
Publication date: 2001-02-22
Also published as: AU6675600A

Abstract

The invention provides transgenic animals that produce altered wool, methods of producing such animals, and wool produced by such animals.

Description

TRANSGENIC ANIMALS THAT PRODUCE ALTERED WOOL This invention relates to transgenic animals that produce altered wool.

Background of the Invention Wool is a high quality textile fiber that has great economic importance worldwide, especially in agriculturally well-developed countries like Australia, New Zealand, and Canada. The increasing demand for improved wool quality and production in sheep and goats has driven the utilization of transgenic animal technology for exploring the possibilities of varying the properties of wool in relation to its end-use as a textile fiber (Powell et ah, Reproduction, Fertility & Development 6:615-623, 1994). The promoter of a sheep keratin IF type II gene, KRT2.10 (K2.10), has been identified and used for driving the specific expression of endogenous genes, as well as foreign genes, in the wool follicles of transgenic animals, including mice and sheep (Keough et al, J. Cell Sci. 108:957-966, 1995; Bawden et al, Transgenic Res. 7:273-287, 1998; Dunn et al, J. Cell Sci. 111:3487-3498, 1998).

Summary of the Invention

The invention provides a transgenic sheep, goat, or other wool- producing mammal, such as a muse ox, alpaca, or other llama, that produces altered wool as a consequence of the expression of a transgene. The genome of the mammal contains an expressible gene encoding a silk protein, such as a spider or silkworm silk protein, and the silk gene is under the regulatory control of a promoter that directs expression of the silk-encoding gene in the wool follicles of the mammal.

The invention also includes a method of obtaining wool with altered characteristics. This method involves providing a mammal of the invention, such as a goat or sheep as described above, and sheering the mammal to obtain wool with altered characteristics.

Also included in the invention is a method of producing a transgenic mammal that produces altered wool. This method involves incorporating into the genome of a wool-producing mammal a transgene encoding a silk protein, such as a spider or silkworm silk protein. The transgene is under the regulatory control of a promoter that directs expression of the silk-encoding gene in the wool follicles of the mammal. The invention also provides a mammalian wool fiber into which there is incorporated a silk protein, such as a spider or silkworm silk protein. Other features and advantages of the invention will be apparent from the following detailed description, the drawings, and the claims.

Brief Description of the Drawings Figs. 1-9 are schematic representations of vectors that are included in the invention.

Detailed Description

Dragline silk is one of the several silks produced by spiders. It constitutes the frame and radii of the orb-web, as well as the dragline. It is able to stretch by 35% before breakage, and its tensile strength is even stronger than steel per unit weight. The size of native dragline silk protein has been observed to range from 274 kDa to 750 kDa. To date, several silk-encoding partial cDNA clones have been identified, including Spidroin 1, 2 from Nephila clavipes, the golden orb weaver found in Brazil and Southern Florida (Xu et al, Proc. Natl. Acad. Sci. U.S.A. 87:7120-7124, 1990; Hinman et al., J. Biol. Chem. 267: 19320-19324, 1992), and ADF1-4 from Araneus diadematus (Guerette et al., Science 272: 112-115, 1996). A summary of the clones isolated from Araneus, and the glands from which they are derived, is shown below, in Table 1.

Table 1 Araneus diadematus genes (libraries screened using spidroin-1, 2 from N. clavipes) Distribution-Expression of specific silk type (as determined by Northern Blot)

The sequence of ADF-1 (Araneus diadematus fibroin) is 68% poly(A)₅ (i.e., AAAAA (SEQ ID NO: l)) or (GA)₂.₇, and 32% GGYGQGY (SEQ ID NO:2); ADF-2 is 19% poly(A)₈ and 81% GGAGQGGY (SEQ ID NO:3) and GGQGGQGGYGGLGSQGA (SEQ ID NO:4); ADF-3 is 21% ASAAAAAA (SEQ ID NO:5) and 79% ((GPGQQ (SEQ ID NO:6))n, where n= 1-8, and GPGGQGPYGPG (SEQ ID NO:7); and ADF-4 is 27% SSAAAAAAAA (SEQ ID NO:8) and 73% GPGSQGPS (SEQ ID NO:9) and GPGGY (SEQ ID NO: 10). New, Improved Wool

With the goal of improving wool quality and production, we target spider dragline silk (ADF-3) gene expression directly to the wool follicle of transgenic animals under the control of the sheep K2.10 keratin gene promoter. This is described further below, in the Examples. This goal can also be accomplished by the targeted expression of other spider dragline (or other) silks, as well as non-spider silks, such as silkworm silks (e.g., a silkworm cocoon silk). Any of the silks mentioned herein, as well as other silks that are known in the art or readily obtainable, and fragments thereof, can be used in the invention.

Spider Silk Genes

Any of a variety of procedures that are well known in the art can be utilized to clone spider biofilament-encoding genes. For example, using the known nucleic acid sequences from Nephila and Araneus silks, one skilled in the art can routinely adapt such methods to obtain a desired silk-encoding gene. The full length clones of Nephila and Araneus silk genes can be likewise isolated.

One such method for obtaining a biofilament-encoding gene sequence is to use an oligonucleotide probe generated by the Nephila clavipes spidroin 1 gene sequence (Arcidiacono et al, Appl. Microbiol. Biotechnol. 49:31-38, 1998) to screen an arachnid or insect cDNA or genomic DNA library for sequences that hybridize to the probe. Hybridization techniques are well known to those of skill in this art, and are described, for example, in Ausubel et al. , Current Protocols in Molecular Biology, John Wiley & Sons, New York, NY, 1994, and Sambrook et al, Molecular Cloning: A Laboratory Manual, (2^nd ed.), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989. cDNA or genomic DNA library preparation is also well known in the art. A large number of prepared nucleic acid molecule libraries from a variety of species are also commercially available. The oligonucleotide probes are readily designed using the sequences described herein and standard techniques. The sequences of the oligonucleotide probes can be based upon the sequences of either strand of DNA encoding, for example, the spidroin 1 gene product. Exemplary oligonucleotide probes are degenerate probes (e.g., a mixture of all possible coding sequences for the N. clavipes spidroin 1 protein).

Silks or fibroin monomers are produced within specialized glands in the abdomen of the spider and are retained in a lumen at a concentration of 20- 30%. This material polymerizes when tension is applied in the presence of drying. As the silk is very thin, the drying process is rapid. Silks are composed of alternating crystalline and amorphous regions. The dominant crystals are β-pleated sheet crystals that are arranged parallel to the long axis of the silk fiber. In Bombyx mori (silkworms), the crystal-forming domain is a six peptide (GAGAGS (SEQ ID NO: 11)) domain, which alternates with amorphous domains.

The amino acids G, A, and S make up greater than 85% of silkworms' cocoon silk fibroin. Silk's alternating Ala or Ser and Gly residues extend to opposite sides of a given β-sheet so that the Ala side chains extending from one β-sheet efficiently nestle between those of the neighboring sheet, and likewise for the Gly side chains. Gly side chains from neighboring β-sheets are in contact, as are those of Ala and Ser. The inter-sheet spacings, consequently, have the alternating values 3.5 angstroms and 5.7 angstroms, as determined from X-ray diffraction studies of Nephila dragline silk. This spacing is identical to that of synthetic poly-L-alanine peptides in their β-sheet conformation. X-ray diffraction studies of Araneuus diadematus dragline silk indicates a larger inter-sheet spacing of about 7.5 angstroms (Gosline et ah, Biomimetics: Design and Processing of Materials, p. 237-261, eds. Sarikaya and Aksay, Amer. Inst. of Physics, 1995).

In Nephila, the major ampullae gland (MA gland) produces the dragline silk. The C-terminal portion of the Nephila MA-gland produced genes cloned so far (spidroin 1 and spidroin 2) have the following sequences: a 34 amino acid long repeat motif (forming both the amorphous domain and crystal-forming domain); and a 47 amino acid long consensus sequence (Xu et al, Proc. Natl. Acad. Sci. U.S.A. 87:7120-7124, 1990; Beckwitt et al, J. Biol. Chem. 269:6661-6663, 1994). The 34 amino acid long repeat motif (forming both the amorphous domain and crystal-forming domain) of Nephila spidroin 1 has the following sequence:

AGQ GGY GGL GSQ GAG RGG LGG QGA GAA AAA AAG G (SEQ ID NO: 12). Note that the poly-alanine region (AAAAAAA (SEQ ID NO: 13)) at the C-terminal end of this sequence, as well as numerous glycine blocks forming amorphous domains (two glycine residues separated by 3 amino acids; e.g., GGLGG).

The 47 amino acid long consensus sequence of Nephila spidroin 2 has the following sequence: CPG GYG PGQ QCP GGY GPG QQC PGG YGP GQQ GPS GPG S AA AAA AAA AA (SEQ ID NO: 14).

The silk is secreted within individual glands; when the secretion is subjected to shear forces and mechanical extension, the poly-alanine (crystal- forming) segments undergo a helix to β-sheet transition, forming β-crystals that stabilize its structure. The glycine blocks are designated as portions forming the amorphous polypeptide chains interspersed among the crystalline regions. Biofilaments have evolved in certain insects and arachnids having very specialized anatomical adaptations and gene evolution. In spiders, the silk is produced by a series of abdominal glands. The formation and size of the crystal may depend upon, among other things, the primary amino acid sequence composition. The production and secretion of the major silks occurs in the major ampullae (MA) gland, the flagelliform (FL) gland, and the cylindrical (CY) gland, which produce dragline silk, viscid silk, and cocoon silk, respectively. It will be understood that the transgenic animals described herein, and the constructs used to generate such animals, can produce any of these biofilaments, or any variations thereof, such that a biofilament is produced having inter-β-sheet spacings of, for example, between 3.5 to 7.5 angstroms.

The MA gland produces two proteins, at a 3:2 ratio, which are rich in glycine, alanine, and proline. These proteins form the dragline silk, which is the lifeline, the scaffolding silk, and the frame for spider webs. Dragline silk is a high stiffness fiber and has properties similar to nylon. Dragline silk contains 20-30% crystal, by volume, and has the following characteristics: stiff (initial Young's modulus is 10 GPa), strong (tensile strength is 1.5 GPa), and tough (energy required to break is 150 Mjm^"3). The viscid silk produced by the FL gland forms the sticky spiral of the web. Niscid silk contains less than 5% crystal by volume, is elastomeric in its native state, and has properties similar to Lycra. It has the following characteristics: mechanically similar to lightly cross-linked rubber (e.g., spandex), low stiffness (initial Young's modulus is 3 MPa), and highly extensible.

The CY gland produces cocoon silk that is similar to the cocoon silk produced by silkworms (B. mori). Synthesis of Biofilament Genes

The invention can employ, in addition to native silk genes, synthetic genes. The cDNA sequences cloned to date share similarities in overall organization and in regions of sequence conservation. The consensus repeats are rich in glycine and glutamine, with poly (Ala) regions integrated into larger repeating units.

For illustration purposes, a candidate gene is represented below by one of the major dragline genes from N. clavipes, spidroin 1 (Arcidiacono et al, Appl. Microbiol. Biotechnol. 49:31-38, 1998). The highly repetitive nature of these genes raises issues of stability of the genes and the possibility of recombination. The repeats can be avoided based on suggestions offered, for example, by Arcidiacono et al, Appl. Microbiol. Biotechnol. 49:31-38, 1998; Fahnestock et al, Appl. Microbiol. Biotechnol. 47:23-32, 1997. Furthermore, a series of constructs can be generated using a similar strategy (Fahnestock et al, Appl. Microbiol. Biotechnol. 47:23-32, 1997), generating 4-20 (or more) consecutive repeats, and can be tested in cell lines prior to the generation of transgenic animals. Blocks of synthetic repeats are constructed so that they have different sizes and contain non-coding sequences (e.g., introns from immunoglobulin genes) to facilitate transcription of the encoded biofilaments and enhance expression. The blocks can be alternating, using a head-to-tail construct strategy (McGrath, K. P., Ph.D. Dissertation, University of Massachusetts at Amherst, 1991; Ferrari et al, U.S. Patent No. 5,243,038).

Codon selection can also be used to maximize expression, since premature termination can occur if a gene contains a greater number of codons recognized by tRNA species that are present in lower abundance in the cell (Rosenberg et al, J. Bacteriol. 175:716-722, 1993; Manley, J. Mol. Biol. 125:407-432, 1978). The genes to be expressed are designed and synthesized using codons favored in the tissue being expressed. Given the high frequency of alanine residues in biofilament proteins, it may be desirable to supplement the cell culture media of the cell line with additional Ala and/or Gly amino acids, to prevent the depletion of Ala tRNA and/or Gly tRNA pools. This eliminates the addition of these amino acids from being the rate-limiting step to generating the biofilament protein. Eukaryotic expression vectors can be generated that drive the synthesis and secretion of proteins (e.g., biofilament proteins) in the wool of an animal transgenic for a nucleic acid molecule encoding such a protein. These vectors are prepared according to standard molecular biology techniques. Methods of preparing such vectors are generally described as follows, and more specific examples are provided in the Examples, below.

The synthesized nucleic acid molecule(s) can have a sequence encoding an epitope tag that is attached for easy identification and/or purification of the encoded polypeptide. Such purification can be accomplished, for example, by affinity chromatography for the epitope tag. A site-specific proteolytic or chemical agent recognition site can be added to the sequence to facilitate removal of the epitope tag following purification of epitope-tagged polypeptides (Saito et al, J. Biochem. 101: 123-134, 1987). Preferably, the sites and site-specific proteolytic agents cleave at or near the junction of the epitope tag and the biofilament protein.

A variety of chemical cleavage agents and their recognition sites (in single letter code) are known in the art, and include the following: hydroxylamine (N or G); formic acid (D or P); cyanogen bromide (M); and acetic acid (D or P). For example, a cyanogen bromide (CNBr) can be used to cleave a Met (M) residue introduced between the epitope tag and the biofilament protein. Alternatively, natural or synthetic proteases can be used. Examples of these (and their recognition sites) include enterokinase (DDDK (SEQ ID NO: 15)); Factor Xa (IEGR (SEQ ID NO: 16)); chymotrypsin (W, Y, and F); renin (YIHPFHLL (SEQ ID NO: 17)); trypsin (R and K); and thrombin (RGPR (SEQ ID NO: 18)). For example, the epitope tag can be attached to the biofilament via a thrombin-recognition site. Following affinity purification of epitope tag-containing proteins, since biofilaments are generally resistant to proteolysis, the epitope tag can be easily removed upon proteolytic cleavage with thrombin.

Expression Vectors

A typical expression cassette used in the invention consists of elements necessary for proper transcription, translation, and secretion in the desired eukaryotic cell (i.e., a promoter, a signal sequence for secretion, intron sequences, and a polyadenylation signal). Many wool follicle-specific promoters can be used with their signal sequences or with the silk and/or fibroin gene signal sequence. In the former case, the biofilament-encoding nucleic acid molecule should not contain its own translation initiation codon but, rather, should be in-frame with the 3' end of the signal sequence. The 3' end of the biofilament-encoding nucleic acid molecule can contain its own polyadenylation signal, or can contain the 3' untranslated sequence normally found on the gene used for the promoter and/or signal sequence. For example, a biofilament-encoding nucleic acid molecule can be placed in an expression vector cassette with a promoter, signal sequence, and 3' untranslated sequence that are all from the same gene. The eukaryotic expression constructs to be used can include one or more of the following basic components. 1. Promoter or transcriptional initiation regulatory region

These sequences can be heterologous to the cell to be modified and can include synthetic or natural viral sequences (e.g., human cytomegalovirus immediate early promoter (CMN); simian virus 40 early promoter (SN40); Rous sarcoma virus (RSN); or adenovirus major late promoter) that confer a strong level of transcription to the nucleic acid molecule to which they are operably linked. The promoter can also be modified by the deletion of non- important sequences and/or addition of sequences, such as enhancers (e.g., an enhancer element of CMN, SN40, or RSN) or tandem repeats of such sequences. The addition of strong enhancer elements can increase transcription by, for example, 10-100 fold. Expression from the above-noted viral promoters is constitutive (i.e., expression occurs in the absence of an apparent external stimulus).

2. Intron

Genes containing an intron (i.e., genomic clones or cDΝA or synthetic genes into which introns have been inserted) are expressed at higher levels than intron-less genes. Hence, inclusion of an intron placed between the transcription initiation site and the translational start codon, 3' to the translational stop codon, or within the coding region of the biofilament- encoding gene can result in a higher level of expression.

The intron sequence includes a 5' splice site (donor site) and a 3' splice site (acceptor site). Sequences of at least 100 basepairs can be found between these two sites. These intronic sequences can be derived from the promoter being used or from the native gene (Ichimura et al., 3. Mol. Evol. 35: 123-130, 1992), and can be positioned 5' to the coding sequence of the silk or fibroin gene. Since the highly repetitive nature of the construct raises concerns over the stability of the gene and the possibility of recombination due to the repetitive sequences, they can be disrupted by inserting the introns of the gene from which the promoter is used. The introns can be positioned, for example, in a manner similar to those present in the fibroin gene of Bombyx mori (Tsujimoto et al, Cell 18:591-600, 1979; Tsujimoto et al, Cell 16:425-436, 1979). This strategy allows for increased levels of expression, in addition to increased stability of the gene.

3. Signal (leader) sequences

Each expression vector can contain a signal sequence that directs the expressed gene product to be secreted from the follicle cells. Such a signal sequence is present in any gene that is secreted naturally. A signal sequence from a relative fibroin gene (e.g., B. mori heavy and light fibroin gene, P25, and sspl60), from a gene specific to the tissue of expression, or a general signal sequence (e.g., from human alkaline phosphatase, mellitin, or CD33 signal peptides) can be used. Further, the signal sequences for secretion can be interchanged between mammalian and insect genes. For example, signal sequences from genes of the follicle or the sequences from native silk (from arachnids) or fibroin (from insects) genes can be used.

4. Termination region

The transcription termination region of the nucleic acid constructs can include the 3'-end and polyadenylation signal from which the 5'-promoter region is derived. Alternatively, the 3'-end of the nucleic acid construct can contain other transcription termination and polyadenylation signals that are known to regulate post-transcriptional mRNA stability, such as those derived from bovine growth hormone, β-globin genes, or the SN40 early region. 5. Other features of the expression vectors

The expression vectors designed for gene transfer can also contain an origin of replication for propagation in E. coli, an SN40 origin of replication, an ampicillin resistance gene, a neomycin resistance gene for selection in eukaryotic cells, or genes (e.g., dihydrofolate reductase gene) that amplify the dominant selectable marker and the gene of interest. In addition, the expression vectors can contain appropriate flanking sequences at their 5' and 3' ends that allow for enhanced integration rates in the transduced cells (ITR sequences; Lebkowski et al, Mol. Cell. Biol. 8:3988-3996, 1988). Further, prolonged expression of the silk or fibroin protein in vitro can be achieved by the use of sequences (e.g., EBΝA-1 and oriP from the Epstein- Barr virus) that allow for autonomous replication. (The transduced, circular nucleic acid replicates as a plasmid in mammalian cells.)

To clone and propagate large segments of DNA, cosmids and similar vectors, such as fosmids and artificial chromosomes (such as bacterial artificial chromosomes (B ACS)) are the vectors of choice. Although plasmid vectors can, in theory, carry large inserts, the resulting recombinants transform Escherichia coli very inefficiently. Cosmids have the capacity to propagate large pieces of foreign DNA (Royal et al, Nature 279: 125, 1979). The expression vectors used for the generation of transgenic animals can be linearized by restriction endonuclease digestion prior to transformation of a cell. In a variation of this method, only a digestion fragment that includes the coding, 5'-end regulatory sequences (e.g., the promoter), and 3'-end regulatory sequences (e.g., the 3' untranslated region) from a gene will be used to transform cells. A cell transformed with such a fragment will not, consequently, contain any sequences that are necessary solely for plasmid propagation in bacteria. (E.g., the cell will not contain the E. coli origin of replication, or a nucleic acid molecule encoding an antibiotic-resistance protein (e.g., an ampicillin-resistance protein) that is useful for selecting prokaryotic cells.)

In another variation of this method, the digestion fragment used to transform a cell will include the coding' region, the 5' and 3' regulatory sequences, and a nucleic acid molecule (including a promoter and 3' untranslated region) encoding a protein capable of conferring resistance to an antibiotic useful for selecting eukaryotic cells (e.g., neomycin or puromycin). The biofilament gene of interest can be modified in its 5' untranslated region (UTR), its 3' UTR, or its N-terminus coding region, to preferentially improve expression. Alternatively, sequences within the coding sequence of the biofilament encoding-nucleic acid molecule can be deleted or mutated to increase secretion or avoid retention of the gene product within the cell due to, for example, the presence of endoplasmic reticulum (ER) retention signals or other sorting inhibitory signals. Further, the transgenic construct can contain sequences that possess chromatin opening domain activity, such that they confer reproducible activation of tissue-specific expression of a linked transgene (Ellis et al, PCT Application No. WO 95/33841; Chung et al, PCT Application No. WO 96/04390).

Example 1. Targeting spider silk gene expression to the wool follicle of transgenic animals under the control of the sheep K2.10 gene promoter Construction of the ADF-3 expression cassette with the sheep keratin K2.10 gene promoter in pBluescript is shown in Fig. 1. With the goal of improving wool quality and production, we elected to target Biosteel-3™ (ADF-3) expression directly into the wool follicle of transgenic animals, under the control of the sheep K2.10 keratin gene promoter. A 2.7 kilobase fragment of the K2.10 gene promoter is obtained by PCR amplification of the sheep K2.10 gene-containing plasmid pK2.10 (Dr. B. Powell, Department of Biochemistry, University of Adelaide, Australia) with a 5' sense primer containing a Swal site and a 3' antisense primer containing an Agel site. The PCR product is digested with Swal and Agel. The WAP promoter sequence of pM3, an expression vector (courtesy of Dr. Zhou, Nexia Biotechnologies Inc.) that harbors 2 copies of the chicken β-globin sequences and that served as an insulator, together with a 1.7 kilobase ADF-3 cDNA fragment, are replaced at the Swal and Agel sites with the Swal- Agel digested K2.10 promoter PCR product to form pK210M3wap. PCR is performed using the same pK210 as a template with a 5' sense primer containing a Pmel site and a 3' antisense primer containing a Notl site and a distal Avrll site. The amplified 3 kilobase-long fragment, including a few 3' exons, introns, and partial 3' untranslated region of the K2.10 gene to provide exon/intron splicing sites and a polyadenylation signal, is digested with Pmel and Avrll. The WAP 3' end fragment in the pK210M3wap is replaced at the Pmel and Avrll sites with the Pmel- Avrll digested K2.10 3'end PCR product to generate pK210M3.

Generation of transgenic animals with the transgenes

Since there are no hair (wool) cortical cell lines, the expression cassettes are tested directly by transgenic techniques. The generation of transgenic mice prior to the generation of transgenic goats is preferable, not only because of the greater time expenditure (i.e., longer gestation period) required to generate a transgenic goat as compared to a transgenic mouse, but also because of the higher expenses in maintaining and housing the animals. Notl digestion of the expression vector pK210M3 releases the K210-ADF3 transgene for microinjection. DNA is purified and injected by established standard techniques. Transgenic goats are generated by injecting the same transgene, using standard techniques.

DNA analysis Genomic DNA of transgenic animals is prepared and analyzed by

PCR and Southern blot. In situ hybridization is performed with the ADF-3 specific probe in the hair (wool) follicles of the transgenic animals.

Electron microscopic analysis of fiber structure of the transgenic animals

Preparation of hair (wool) fibers for transmission EM is carried out as described by G. E. Rogers and colleagues (Bawden et al, Transgenic Res. 7:273-287, 1998).

Hair (wool) measurement

The physical characteristics of the hair (wool) of the transgenic animals are measured using standard methods (Damak et al., Biotechnology (N.Y.) 14: 185-188, 1996).

Example 2. Integrating Biosteel™ (ADF-3) with the keratin IF K2.10 for expression of the fusion protein in the wool follicle of transgenic animals under the control of the sheep K2.10 gene promoter

This example provides an approach to facilitate the association of silk proteins with the high sulfur and disulfide bond-containing keratin intermediate filament proteins. In particular, the silk proteins, which can be, for example, spider silk or silkworm silk proteins (see above), are produced as fusion proteins that include silk and keratin sequences. The sequences of these two components of the fusion proteins can be arranged in any order. For example, the silk protein sequences can be positioned at the amino terminal end of the keratin, at the carboxyl terminal end of the keratin, or at both ends of the keratin. A specific, non-limiting example of a fusion protein that includes keratin and silk sequences is as follows.

N-terminal fusion of ADF-3 with K2.10 in pK210M3 (Fig. 2)

A linker containing a 5 Agel site, an ATG start codon, the N-terminal end domain of K2.10 (Steinert, J. Invest. Dermatol. 100:729-734, 1993), and the ADF-3 cDNA 8901-9061 basepair sequence, including a 3' distal Ball site (for the N-terminal sequence of ADF-3 without the signal peptide) in pM3 (courtesy of Dr. Zhou, Nexia Biotechnologies, Inc.) is ligated at the Ball site with the ADF-3 fragment (above, Example 1) excised from pM3 with Ball-Scil digestion. The ligated fragment is connected at the Scil site with a Sci-Pmel digested PCR product amplified from K2.10 cDNA (Dr. B. Powell, Department of Biochemistry, University of Adelaide, Australia) with a 5' sense primer containing a Scil site, the 3' end of ADF-3 open reading frame without the stop codon, and the sequence from the N-terminal variable domain of K2.10, and a 3' antisense primer containing a distal Pmel site and the sequence of the 3' end of K2.10 including its stop codon. The ADF-3 fragment in pK210M3 (see Example 1, above) is replaced at the Agel and Pmel sites with the newly ligated fragment (fused ADF3-K210) digested with Agel-Pmel to form pK210-N-ADF3.

Internal fusion of ADF-3 into K2.10 in pK210M3 (Fig. 3)

The following steps are used to generate the vector carrying the internal fusion of ADF-3 with K2.10: 1. An Agel-Hindlll digested PCR product amplified from the K2.10 cDNA with a 5' sense primer containing an Agel site and the 5' end sequence of K2.10, and a 3' antisense primer containing a Hindlll site and the sequence from the L2 domain of K2.10 (Steinert, J. Invest. Dermatol. 100:729-734, 1993), are ligated at the Hindlll site with a linker containing a 5' Hindlll site, and the ADF-3 cDNA 8901-9061 basepair sequence including a 3' distal Ball site (for the N-terminal sequence of ADF-3 without the signal peptide) in pM3.

2. The ligated fragment is connected at the Ball site with the ADF-3 fragment excised from pM3 with Ball-Scil digestion to form the K2.10-linker-ADF3 fragment.

3. PCR is performed using the K2.10 cDNA as a template with a 5' sense primer containing a Scil site, the 3' end sequence of the ADF-3 open reading frame without the stop codon, and the sequence from the L2 domain of K2.10, a 3' antisense primer containing a distal Pmel site and the 3' end sequence of the K2.10 cDNA open reading frame including the stop codon. The fragment is digested with Scil and Pmel. 4. The Scil-Pmel digested PCR product is ligated at the Scil site with the Agel-Scil digested K2.10-linker-ADF3 fragment. The sheep K2.10 gene promoter is replaced at the Agel and Pmel sites with the newly ligated and Agel-Pmel digested fragment (fused K2.10-ADF3) to form the final product, pK210-I-ADF3.

C-terminal fusion of ADF-3 into K2.10 in pK210M3 (Fig. 4)

The following steps are used to generate a vector that harbors the C-terminal fusion of ADF-3 with K2.10:

1. An Agel-Hindlll digested PCR product amplified from the K2.10 cDNA with a 5' sense primer containing a distal Agel site and the 5' end sequence of K2.10, and a 3' antisense primer containing a distal Hindlll site and the sequence from the 3' end of K2.10 cDNA open reading frame without the stop codon, are ligated at the Hindlll site with the linker containing a 5' Hindlll site, and the ADF-3 cDNA 8901-9061 basepair sequence including a 3' Ball site in pM3.

2. The ligated fragment is connected at the Ball site with the ADF-3 fragment excised from pM3 with Ball-Scil digestion to form

K2.10-linker-ADF3 fragment.

3. The K2.10-linker-ADF3 fragment is ligated at the Scil site with another small linker containing a 5' Scil site, the sequence of the 3' end of K2.10 cDNA with the stop codon, and a 3' Pmel site. 4. The ADF-3 fragment in pK210M3 is replaced at the Agel and

Pmel sites with the newly ligated and Agel-Pmel digested fragment (fused K2.10-ADF3) to form the final product, pK210-C-ADF3.

Generation of transgenic animals with the transgenes

Not I digestion of the expression vectors pK210-N-ADF3, pK210-1- ADF3 , and pK210-C- ADF3 releases the fused K210- ADF3 transgenes for microinjection. DNA is purified and injected by established, standard techniques. Transgenic goats are generated by injecting the same transgenes.

DNA analysis Genomic DNA of transgenic animals is prepared and analyzed by

PCR and Southern blot. In situ hybridization is performed with the ADF-3 specific probe in the hair (wool) follicles of the transgenic animals. Electron microscopic analysis of fiber structure of the transgenic animals

Hair (wool) measurement

The physical characteristics of the hair (wool) of the transgenic animals are measured with the standard methods (Damak et al., Biotechnology (N.Y.) 14: 185-188, 1996).

Example 3. Targeting silkworm fibroin heavy and light chains into the wool follicle of transgenic animals under the control of the sheep K2.10 gene promoter

Fibroins of the silkworm, Bombyx mori, are secreted into the lumen of posterior silk gland (PSG) from the surrounding PSG cells as a protein complex consisting mainly of a heavy (H) chain of 350 kDa and a light (L) chain of 25 kDa, which are connected by disulfide bonds (Takei et al, J. Cell. Biol. 105:175-180, 1987). The H chain is a fibrous protein and is characteristically rich in glycine, alanine, and serine. It is a giant molecule (4,700 amino acids) comprising a "crystalline" portion of about two-thirds and an "amorphous" region of about one-third. The crystalline portion contains about 50 repeats of a polypeptide of 59 amino acids, the sequence of which is known (Heslot, Biochimie 80:19-31, 1998). The L chain is nonfibrous and comprises relatively high amounts of leucine, isoleucine, valine, and acidic amino acids. It has been strongly suggested that the H-L chain combination in a one to one molar ratio is important for efficient secretion of silk fibroin. The mechanical properties of silk fibroin, for example, elasticity and toughness, would make it feasible to express the fibroin H and L chains in the wool follicle of transgenic animals under the control of the sheep K2.10 keratin gene promoter for the aim of improving wool quality and production.

Construction of the fibroin H chain expression cassette with the sheep keratin K2.10 gene promoter in pBluescript (Fig. 5)

A synthetic monomer fibroin H chain DNA encoding the known polypeptide of 59 amino acids (Heslot, Biochimie 80: 19-31, 1998) with 5' Nhel (G/CTAGC) and 3' Spel (A/CTAGT) termini is obtained and inserted into plasmid pQE9 (Qiagen). Synthetic inserts are released from the plasmid by digestion with Nhel and Spel. These inserts are self-ligated to form linear multimers of the target DNA. The ligated fragments are re-digested with Nhel and Spel, which cut all inverted repeats, but not the "head-to-tail" ligated multimers as in this case any internal restriction sites are already destroyed. Two more synthetic linkers, one containing the mouse Ig kappa signal sequence with 5' Agel and 3' Nhel termini and another containing the fibroin H chain DNA encoding the non-repetitive C terminal amino acids with 5' Spel and 3' Pmel termini, are ligated to the Nhel and Spel sites of the multimers, respectively. The joined fragments are digested with Agel and Pmel, and ligated to the Agel-Pmel digested pK210M3 (see Example 1, above) to generate ρK210FBH.

Construction of the fibroin L chain expression cassette with the sheep keratin K2.10 gene promoter in pBluescript (Fig. 6)

PCR is performed using pFLlδ DNA that harbors the full-length fibroin L chain cDNA of Bombyx mori (Yamakuchi et al., J. Mol. Biol. 210: 127-139, 1989) as a template with a 5' sense primer: (5'-GCGCAGACCGGTGCGGCCGCTCTAGACTCG-3' (SEQ ID NO: 19)) containing an Agel site and a 3' antisense primer: (5'-GCGCAGGTTTAAACGTCGACGCCCCATCCTCAC-3' (SEQ ID NO: 20)) containing a Pmel site. The PCR product is digested with Agel and Pmel. The fibroin H chain fragment in pK210FBH is replaced at Agel and Pmel sites with the Agel-Pmel digested fibroin L chain PCR product to generate pK210FBL.

Example 4. Targeting the Biosteel™ (ADF-3) expression into the wool follicle of transgenic animals under the control of the mouse UHS keratin gene promoter

It has been shown that the promoter of a mouse IFAP gene (UHS keratin gene) can drive an ovine insulin-like growth factor 1 (IGF1) expression in the skin of the transgenic sheep, and the wool production in such transgenic sheep was improved (Damak et al, Biotechnology (N.Y.) 14: 185-188, 1996). With the goal of improving wool quality and production, we elect to target Biosteel-3™ (ADF-3) gene expression directly to the wool follicle of transgenic animals under the control of the mouse UHS keratin gene promoter.

Construction of the ADF-3 expression cassette with the mouse UHS keratin gene promoter in pBluescript (Fig. 7)

A 700 basepair fragment is obtained from PCR amplification of the mouse genomic DNA with a 5' sense primer:

(5'-CGCTATGTCGACAACCATGTTGGAACATCCTGC-3' (SEQ ID NO:21)) containing a Sail site and a 3' antisense primer: (5'-GCGTATACCGGTGAGGAGTTGTGTTCACAGGAG-3' (SEQ ID

NO:22)) containing an Agel site. The amplified PCR product of the mouse UHS keratin gene promoter is digested with Sail and Agel. The sheep K2.10 promoter fragment in pK210M3 (see Example 1, above) is replaced by the digested PCR fragment of the mouse UHS keratin gene promoter at Sail and Agel sites to form pUHSM3.

Example 5. Targeting silkworm fibroin heavy and light chains into the wool follicle of transgenic animals under the control of the mouse UHS keratin gene promoter

The mechanical properties of silk fibroin, for example, elasticity and toughness, would make it feasible to express the fibroin H- and L-chains in the wool follicle of transgenic animals under the control of the mouse UHS keratin gene promoter for the aim of improving wool quality and production.

Construction of the fibroin H chain expression cassette with the mouse UHS keratin gene promoter in pBluescript (Fig. 8)

(5'-CGCTATGTCGACAACCATGTTGGAACATCCTGC-3' (SEQ ID NO:23)) containing a Sail site and a 3' antisense primer: (5'-GCGTATACCGGTGAGGAGTTGTGTTCACAGGAG-3' (SEQ ID NO:24)) containing an Agel site. The amplified PCR product of the mouse UHS keratin gene promoter is digested with Sail and Agel. The sheep K2.10 promoter fragment in pK210FBH (see Example 3, above) is replaced by the Sail- Agel digested PCR fragment of the mouse UHS gene promoter at Sail and Agel sites to form pUHSFBH. Construction of the fibroin L chain expression cassette with the mouse UHS keratin gene promoter in pBluescript (Fig. 9)

A 700 basepair fragment is obtained from PCR amplification of the mouse genomic DNA with a 5' sense primer: (5'-CGCTATGTCGACAACCATGTTGGAACATCCTGC-3' (SEQ ID NO:25)) containing a Sail site and a 3' antisense primer: (5'-GCGTATACCGGTGAGGAGTTGTGTTCACAGGAG-3' (SEQ ID NO:26)) containing an Agel site. The amplified PCR product of the mouse UHS keratin gene promoter is digested with Sail and Agel. The sheep K2.10 promoter fragment in pK210FBL (see Example 3, above) is replaced by the

Sail- Agel digested PCR fragment of the mouse UHS gene promoter at Sail and Agel sites to form pUHSFBL.

Generation of transgenic animals with the transgenes

Notl digestion of the expression vectors pUHSFBH and pUHSFBL releases UHS-FBH and UHS-FBL transgenes for microinjection, respectively. UHS-Fibroin transgenic mice are generated by co-injecting the H chain and the L chain fragments in a 1: 1 molar ratio. DNA is purified and injected by established, standard techniques. Transgenic goats are generated by injecting the same transgene.

Claims

1. A transgenic, wool-producing mammal that produces altered wool as a consequence of the expression of a transgene, wherein the genome of the mammal contains an expressible gene encoding a silk protein, the silk gene being under the regulatory control of a promoter that directs expression of the gene in wool follicles of the mammal.

2. A method of producing a transgenic wool-producing mammal that produces altered wool, the method comprising incorporating into the genome of a wool-producing mammal a gene encoding a silk protein, the gene being under the regulatory control of a promoter that directs expression of the gene in wool follicles of the mammal.

3. A mammalian wool fiber into which there is incorporated a silk protein.

4. A method of obtaining wool with altered characteristics, said method comprising: (a) providing a mammal of claim 1 , and

(b) sheering said mammal to obtain the wool.

5. The mammal of claim 1, wherein the silk protein is spider silk protein or silkworm silk protein.

6. The method of claim 2, wherein the silk protein is spider silk protein or silkworm silk protein.

7. The mammalian wool fiber of claim 1, wherein the silk protein is spider silk protein or silkworm silk protein.