AU700224B2 - Alpha-lactalbumin gene constructs - Google Patents

Description

WO 96/02640 PCT/GB95/01651 1 1 "Alpha-Lactalbumin Gene Constructs" 2 3 The present invention is concerned with recombinant 4 gene constructs for expressing the protein alactalbumin, or functional equivalents or parts 6 thereof.

7 8 Human milk has been shown to be superior over other 9 milk types, notably cow, sheep, camel and goat milk, for human infant nutrition. However, many mothers find 11 breast feeding difficult or inconvenient. Moreover, in 12 countries where infant food supplements are in great 13 demand, it would be highly desirable to be able to 14 supply a milk product with the nutritional benefits of human milk.

16 17 One of the major differences of human milk over milk 18 from other mammals (for example cows or sheep) is the 19 presence of a-lactalbumin as the major whey protein.

Whilst a-lactalbumin is present in other milk types, 21 the concentration is relatively low and instead the 22 major whey protein is 3-lactoglobulin. The level of a- 23 lactalbumin varies from species to species, with human 24 milk containing about 2.5 mg/ml, cow milk 0.5-1.0 mg/ml and mouse milk 0.1-0.8 mg/ml.

SUBSTITUTE SHEET (RULE 26) WO 96/02640 PCINGD5101651 2 1 The gene sequences encoding for the bovine a- 2 lactalbumin and for the human a-lactalbumin proteins 3 have been elucidated and the sequence information 4 published by Vilotte et al, in Biochemie J6: 609-620 (1987) and by Hall et al, in Biochem J 242: 735-742 6 (1987), respectively.

7 8 The present invention seeks to utilise genetic 9 engineering techniques to provide a recombinant gene construct capable of producing an a-lactalbumin 11 concentration of greater than 1.0 mg/ml, for example 12 1.2 mg/ml or above, in milk when expressed in mammalian 13 cells. Generally said construct is adapted to be 14 expressed in non-human animal, especially bovine, cells.

16 17 In one aspect, the present invention provides a 18 recombinant expression system adapted to express a- 19 lactalbumin, or a functional equivalent or part thereof in cells of a non-human, preferably bovine, animal.

21 Preferably, the recombinant expression system of the 22 present invention is adapted to express the human a- 23 lactalbumin protein, or a functional equivalent or part 24 thereof.

26 The term "expression system" is used herein to refer to 27 a genetic sequence which includes a protein-encoding 28 region and is operably linked to all of the genetic 29 signals necessary to achieve expression of the protein encoding region. Optionally, the expression system may 31 also include a regulatory element, such as a promoter 32 or enhancer, to increase transcription and/or 33 translation of the protein-encoding region, or to 34 provide control over expression. The regulatory element may be located upstream or downstream of the 36 protein-encoding region, or may be located at an intron -3- (non-coding portion) interrupting the protein encoding region. Alternatively it is also possible for the sequence of the protein-encoding region itself to comprise a regulatory ability.

Preferably, the present invention provides a recombinant DNA construct comprising, in the 5' to 3' direction and operatively linked: the 5' flanking sequence of S ID NO 21; a DNA sequence encoding a-lactalbumin or a functional equivalent thereof; and the 3' flanking sequence of any one of SEQ ID NO'S 16 to the DNA construct being adapted to express the a-lactalbumin or functional *quivalent thereof in a non-human transgenic animal.

in a further aspect the present invention provides a recombinant DNA construct comprising, in the 5' to 3' direction and operatively linked: at least about 1.8 kb of 5'-flanking sequence from the human a-lactalbumin gene including the a-lactalbumin promoter; a DNA sequence encoding a-lactalbumin or a functional equivalent thereof; at least about 3 kb of 3'-flanking sequence from the humin a lactalbumin gene the DNA construct being adapted to express the a-lactalbumin or furntional equivalent thereof in a non-human transgenic animal.

i -3/1 The recombinant DNA construct described above preferably comprising about 3.7 kb of 5' flanking sequence and about 13 kb of 3' flanking sequence.

The term "functional equivalent" refers to any derivative which is functionally substantially similar to the reference sequence or protein. In particular the term "functional equivalent" includes derivatives in which nucleotide base(s) and/or amino acid(s) have been added, deleted or replaced without a significantly adverse effect on biological function, especially biological function in milk production.

Genetic engineering has been recognised as a powerful technique not only in research but also for commercial purposes. Thus, by using genetic engineering techniques (see Maniatis et al Molecular Cloning, a Laboratory Manual Cold Spring, Harbor Laboratory, Cold Spring Harbor, New York 1982 and "Principle of Genetic Engineering", Old and Primrose, 5th edition, 1994, both incorporated herein by reference) exogenous genetic material can be transferred to a host cell ~15 and the protein or polypeptide encoded by the exogenous genetic material may i: be replicated by and/or expressed within the host. For the purposes of simplicity genetic engineering is normally carried out with prokaryotic micro-organisms, for example bacteria such as E. coli, as host. However, use has also been made of eukaryotic organisms, in particular yeasts or algae, and in certain applications 20 eukaryotic cell cultures may also be used.

e• Genetic alterations to mammalian species by micro-injection of genes into the pro-nuclei of single-cell embryos has been described by Brinster et al, in Cell WO 96102640 PCT/GB95/01651 4 1 27: 223-231, 1981. Here the foreign genetic material 2 is introduced into the fertilised egg of an animal 3 which then proceeds to develop into an embryo in the 4 normal manner having been transplanted into a foster mother. Truly transgenic animals contain copies of the 6 exogenous DNA in each cell.

7 8 Where the injected genetic material is successfully 9 incorporated into the host chromosome the animal is termed "transgenic" and the transgene is inherited in 11 the normal Mendelian manner. However, only a low 12 proportion of gene transfer operations are successful, 13 especially for large domestic animals such as pigs, 14 sheep and cattle. To date it has not been possible to control the location at which the transgene integrates 16 into the host chromosome for such animals.

17 18 For the purpose of the present invention it may, in 19 certain circumstances, be sufficient simply to produce a "mosaic" donor animal. In this situation the' 21 transgene is incorporated into the chromosome copies of 22 only certain body organs. Mosaic animals are generally 23 produced by introducing exogenous DNA into an embryo at 24 a later developmental stage.

26 One of the most promising application of transgenesis 27 in livestock aims to utilise the mammary gland as a 28 "bioreactor" to produce recombinant proteins of 29 pharmaceutical or nutritional interest in milk. The mammary gland is an attractive organ in which to 31 express heterologous proteins due to its capacity to 32 produce large quantities of protein in an exocrine 33 manner. Recombinant DNA techniques may be used to 34 alter the protein composition of cow's milk used for human or animal consumption. For example, the 36 expression of human milk proteins in cow's milk could i -s WO 96/02640 PCT/GB95/01651 1 improve its nutritional value in infant formula 2 applications (Strijker et al, in Harnessing 3 Biotechnology for the 21st Century, ed Ladisch and 4 Boser, American Chemical Society, Pages 38-21 (1992)).

Both applications would benefit from ability to 6 increase production capacity inexpensively by 7 multiplying producer animals witii conventional and 8 advanced breeding techniques.

9 The first step in developing a transgene to be 11 expressed in the mammary gland is to clone the gene for 12 the protein of interest. To direct expression into 13 milk, the promoter gene for a major milk protein 14 expressed in milk is employed. Milk protein genes are tightly regulated and are not expressed in tissues 16 other than the mammary gland, a characteristic that 17 minimises the possibility of negative effects on the 18 animal from inappropriate expression in other tissues.

19 Among the regulatory genes used to express heterologous proteins in the mammary gland are alpha-SI-casein 21 (Strijker et., 1992, supra), beta-lactoglobulin (Wright 22 et al., Bio/Technology a:831-834 (1991)) whey acidic 23 protein (Ebert and Schindler, Transgenic Farm animals: 24 Progress Report (1993)) and beta-casein (Ebert and Schindler, 1993, supra).

26 27 Newly-made gene constructs are jrmally tested in 28 transgenic mice before adopting them for use in cattle.

29 Milk obtained from the transgenic mice is assayed for quantity of recombinant protein. If enough milk is 31 available, the protein may be isolated to examine its 32 structural characteristics and bioactivity. The 33 selection of a particular construct for use in cattle 34 will depend primarily on consideration of both' expression level and authenticity of the resultant 36 recombinant protein.

WO 96/02640 PCT/GB95/01651 6 1 Transgenesis in cattle is normally initiated by 2 microinjecting several hundred copies of gene construct 3 into one of the two pronuclei in a zygote. Zygotes may 4 be obtained in vivo from the oviducts (Roschlau et al., Arch Tierz. Berlin 31:3-8 (1988); Roschlau et al., in 6 J. Reprod. Fertil. (suppl 38), Cell Biology of 7 Mammalian Egg Manipulation, ed Greve et al (1989); Hill 8 et al., Theriogenology 37:222 (1992); Bowen et al. Biol 9 Reprod. 50:664-448 (1994)) or by in vitro fertilisation of in vitro matured oocytes (Krimpenforth et al., 11 Biotechnology 9:844-847 (1991); Hill et al., 1992, 12 supra; Bowen et al., 1993 supra). Bovine zygotes must 13 be centrifuged at 15,000 x g for several minutes to 14 displace opaque lipid in order to visualise the pronuclei with phase contrast, Nomarski or Hoffman 16 interference contrast optics. 2-4 pl of buffer 17 containing several hundred copies of DNA construct are 18 injected into a pronucleus. Transgenes are thought to 19 integrate into random breaks in chromosomal DNA that result from mechanical disruption during the 21 microinjection process. Ideally, the transgenes 22 integrate at the zygote stage prior to DNA replication 23 to ensure that every cell in the adult contains the 24 transgene. In general, several "copies" of the transgene, linked together linearly, integrate in a 26 single site on a single chromosome. The site of 27 integration is random. Integration probably occurs 28 after the first round of DNA replication, and perhaps 29 as late as the 2- or 4- cell stage (Wall and Seidel, 1992), resulting in animals that are mosaic with 31 respect to the transgene. Indeed, up to 30% of animals 32 in which transgenes are detected in somatic tissues do 33 not transmit the transgenes to their offspring (or 34 transmit to less than the expected 36 After microinjection, embryos are either transferred UIn PCT/GB9501651 WO 96/02640 7 directly into the oviducts of recipients or cultured for a few days and transferred to the uterus of recipient cattle. Confirmation of transgene integration is obtained by Southern blot analysis of tissues sampled from the calf after birth. Transgene expression is measured by assaying for the gene product in appropriate tissues, or in this case milk. Embryo survival after microinjection, transgene integration frequency, frequency of expression and expression level, and frequency of germline transmission vary according to quantity and quality of DNA construct injected, strain of mice used (Brinster et al., Proc.

Natl. Acad Sci. USA 82:4438-4442 (1985)) and skill and technique of the operator performing microinjection.

This basic approach has been routinely applied to produce transgenic sheep (Wright et al., 1991, supra), goats (Ebert and Schindler, 1993, supra) pigs (Rexroad and Purcel, Proc llth Intl. Congr. Anim. Reprod. A.I.

Dublin 5:29-35 (1988)) and cattle (Krimpenfort et al., 1991, supra; Hill et al., 1992, supra; Bowen et al., 1994, supra).

Reference is also made to WO-A-88/01648 (of Immunex Corporation), to WO-A-8G/00239 and to WO-A-90/05188 (both of Pharmaceutical Proteins Limited) for describing suitable techniques and methodologies for production of recombinant gene constructs, production of transgenic animals incorporating such constructs and expression of the protein encoded in the mammary gland of the lactating adult female mammal. The disclosures of these references and the references recited above are incorporated herein by reference.

Reference is further made to Stacey et al in molecular and Cellular Biology 14(12: 1009-1016 (February 1994) which describes a knock-out experiment in which the 23 24 26 27 28 29 31 32 33 34 36 WO 96/02640 PCT/GB95/01651 8 1 mouse a-lactalbumin gene is replaced with a human a- 2 lactalbumin gene. This paper (incorporated herein by 3 reference) does not however report expression of the a- 4 lactalbumin protein.

6 In one embodiment the present invention provides an 7 expression system which has been produced by techniques 8 other than by knock-out of the gene naturally present.

9 In a further aspect, the present invention provides a 11 transgenic mammalian animal, said animal having cells 12 incorporating a recombinant expression system adapted 13 to express a-lactalbumin (preferably human a- 14 lactalbumin) or a functional equivalent or part thereof. Generally the recombinant expression system 16 will be integrated into the genome of the transgenic 17 animal and will thus be heritable so that the offspring 18 of such a transgenic animal may themselves contain the 19 transgene and thus also be covered by the present invention. Suitable transgenic animals include (but 21 are not limited to) sheep, pigs, cattle and goats.

22 Cattle are especially preferred.

23 24 Additionally, the present invention comprises a vector containing such a recombinant expression system and 26 host cells transformed with such a recombinant 27 expression system (optionally in the form of a vector).

28 29 In a yet further aspect the present invention provides a-lactalbumin produced by expression of a recombinant 31 expression system of the present invention, desirably 32 such a-lactalbumin produced in a transgenic mammal.

33 The a-lactalbumin gene is naturally activated in the 34 mammary glands of the lactating female mammal. Thus the protein expressed by the recombinant expression 36 system of the present invention would be produced at WO 96/02640 PCr/GB9/01651 9 1 such a time and would be excreted as a milk component.

2 It may also be possible for the protein of interest to 3 be produced by inducing lactation through hormonal or 4 other treatment. Processed milk products comprising such a-lactalbumin are also covered by the przesnt 6 invention.

7 8 In one preferred embodiment, the recombinant expression 9 system comprises a construct designated pHA1, pHA2, pBBHA, pOBHA, pBAHA, pBova-A or pBova-B. The 11 constructs pHA1, pHA2, pBBHA, pOBHA and pBAHA express 12 human a-lactalbumin and are thus preferred, 13 particularly pHA2. The construct pHA2 was deposited on 14 15 February 1995 at NCIMB under Accession No NCIMB 40709.

16 17 Likewise transgenic mammals comprising the specific 18 constructs listed above are preferred.

19 It has further been found th-t, in addition to 21 increased concentrations of '-lactalbumin per unit 22 volume of milk achieved by the present invention, where 23 a human a-lactalbumin gene is present the volume of 24 milk produced increases also. This finding is totally unexpected and for this reason constructs containing 26 the human a-lactalbumin gene (or functional equivalents 27 or parts thereof) and transgenic animals (especially 28 cattle) are preferred embodiments of the invention.

29 Whilst we do not wish to be bound by theoretical 31 considerations, it is further believed that the 32 promoter region of the human a-lactalbumin gene is only 33 partially responsible for the enhanced natural 34 expression of a-lactalbumin by humans. It is believed that enhanced expression may be obtained by including 36 within the recombinant expression system of the present 37 invention the 3' sequence flanking the protein-encoding 38 region and/or the 5' sequence flanking the protein- SUBSTITUTE SHEET (RULE 26)' WO 96/02640 PCT/GB95/01651 1 encoding region itself.

2 3 The flanking sequences 3' and 5' to the protein- 4 encoding region of the human a-lac gene have been sequenced for the first time. Partial sequences 6 (nucleotides 1-264, 1331-2131, 2259-2496, 2519-2680 and 7 3481-3952) of the 3' flanking region are presented in 8 SEQ ID Nos. 16 to 20 whilst the full sequence of the 9 flanking region is presented in SEQ ID No. 21. In experiments it has been observed that inclusion of 11 either or both of these sequences give a surprisingly 12 marked increase in expression levels of the a- 13 lactalbumin protein. This increase in expression may 14 be observed when the protein-encoding region is nonhuman a-lactalbumin as well as human a-lactalbumin.

16 17 Both the sequences of SEQ ID Nos. 16-20 and 21 are now 18 believed to contribute towards the higher levels of 19 expression of a-lactalbumin in human milk, and therefore comprise a further aspect of the present 21 invention.

22 23 In a further aspect, the present invention provides a 24 polynuclectide having a sequence substantially as set out in any one of SEQ ID Nos. 16-20 or SEQ ID No. 21 or 26 a portion or functional equivalent thereof.

27 28 The polynucleotides may be in any form (for example DNA 29 or RNA, double or single stranded), but generally double stranded DNA is the most convenient. Likewise 31 the polynucleotides according to the present invention 32 may be present as part of a recombinant genetic 33 construct, which itself may be included in a vector 34 (for example an expression vector) or may be incorporated into a chromosome of a transgenic animal.

36 Any vectors or transgenic animals comprising a WO 96/02640 PCT/GB95/01651 11 1 polynucleotide as described above form a further aspect 2 of the present invention.

3 4 Viewed from a yet further aspect the prese..t invention provides a recombinant expression system (preferably 6 adapted to express a-lactalbumin (preferably human a- 7 lactalbumin) or a portion or functional equivalent 8 thereof), said recombinant expression system comprising 9 a polynucleotide selected from the polynucleotide located between the EcoRI and XhoI restriction sites of 11 the wild-type a-lactalbumin gene and the polynucleotide 12 located between the BamHI and EcoRI restriction sites 13 of the wild-type human a-lactalbumin gene, or a portion 14 or functional equivalents thereof.

16 In one preferred embodiment, the recombinant expression 17 sequence of the present invention comprises both 18 polynucleotides as defined above, portions and 19 functional equivalents thereof.

21 The invention also encompasses vectors containing the 22 recombinant expression systems defined above and cells 23 transformed with such vectors. Further, the present 24 invention comprises transgenic animals wherein the transgene contains the recombinant expression system.

26 27 Figure 1 as discussed in Example 1 and shows the 28 sequence of bovine a-lactalbumin PCR primers.

29 Figure 2 is discussed in Ex.mple 1 anc 4 shows the 31 position of bovine a-lactalbumin PCR primers and 32 products.

33 34 Figure 3 is discussed in Example 2 and shows a restriction map of two overlapping genomic A clones for 36 the human a-lactalbumin gene (pHA-2 and pHA-1).

I

WO 96/02640 PCT/GB95/01651 12 1 Figure 4 is discussed in Example 3 and shows a 2 restriction map of three overlapping genomic A clones 3 for the bovine beta-lactaglobulin gene.

4 Figure 5 is discussed in Example 4 and shows SDS-PAGE 6 analysis of skimmed milk from bovine a-lactalbumin 7 transgenic mice run against non transgenic mouse milk.

8 9 Figure 6 is discussed in Example 5 and shows human alactalbumin transgene constructs.

11 12 Figure 7 is discussed in Example 5 and shows SDS-PAGE 13 analysis of skimmed milk from human a-lactalbumin 14 transgenic mice run against non transgenic mouse milk.

16 Figure 8 is discussed in Example 5 and shows a Western 17 analysis of the milk from human a-lactalbumin 18 transgenic mice run against human a-lactalbumin 19 standard.

21 Figure 9 shows the PCR cloning strategy for transgene 22 constructs PKU1 to PKU4 as discussed in Example 6.

23 24 Figure 10 gives the sequences of PKU primers 1 to 10 as discussed in Example 6.

26 27 Figure 11 shows the structure of null and humanised a- 28 lactalbumin alleles.

29 Figure 12 is a Northern analysis of total RNA from a- 31 lactalbumin-deficient lactating mammmary glands.

32 33 Figure 13 illustrates a Western analysis of a- 34 lactalbumin from targeted mouse strains.

36 Figure 14 is a histological analysis of wild type and WO 96/02640 PCGB95/01651 13 1 a-lac" lactating mammary glands.

2 3 Figure 15A shows an RNase protection assay used to 4 distinguish human replacement and mouse a-lactalbumin mRNA and Figure 15B shows an RNase protection assay of 6 mouse and human replacement a-lactalbumin mRNA.

7 8 Figure 16 gives the quantification of a-lactalbumin by 9 hydrophobic interaction chromatography.

11 SEQ ID Nos. 16 20 give parts of the sequence from the 12 BamHI site to the vector restriction sites (including 13 EcoRI sites) 3' of the protein-encoding region of the 14 endogenous human a-lactalbumin gene, as set out below: 16 SEQ ID No. 16 Nucleotides 1 to 264 (inclusive) 17 SEQ ID No. 17 Nucleotides 1331 to 2131 (inclusive) 18 SEQ ID No. 18 Nucleotides 2259 to 2496 (inclusive) 19 SEQ ID No. 19 Nucleotides 2519 to 2680 (inclusive) SEQ ID No. 20 Nucleotides 3481 to 3952 (inclusive) 21 22 SEQ ID No. 21 gives the sequence from the EcoRI 23 restriction site to the Xhol restriction site which are 24 5' to the protein-encoding region of the endogenous human a-lactalbumin gene.

26 27 In mdre detail, in Figure 11 the upper portion shows 28 the wild type murine a-lactalbumin locus. The position 29 and direction of the transcribed region is indicated by the arrow. The translational stop site and RNA 31 polyadenylation sites are also indicated. The middle 32 portion shows the structure of the null allele. The 33 striped bar indictes the HPRT selectable cassette. The 34 lower portion shows the structure of the human replacement allele. The checkered bar shows the human 36 a-lactalbumin fragment. The transcription initiation, WO 96/02640 PCr/GB9/01651 14 1 translational stop and polyadenylation sites are shown.

2 Restriction enzyme sites shown are: HindIII BamHI 3 XbaI 4 In Figure 12, the two autoradiographs shown are repeat 6 hybridisations of the same membrane filter using a 7 human a-lactalbumin probe followed by a rat f-casein 8 probe. The probes used are indicated under each 9 autoradiograph. The source of RNA in each lane is indicated above the lane markers.

11 12 In Figure 13 Lane A contains purified human a- 13 lactalbumin. Lanes B-F show samples of milk from 14 targeted mice, genotypes are indicated above the lane markers. Lanes G and H are a shorter exposure of Lanes 16 C and D.

17 18 The light micrographs shown in Figure 14 are 19 haemtoxylin/eosin stained sections of mammary tissue (original magnification 100x). The genotypes of each 21 gland are indicated.

22 23 In Figure 15A, the 3' junction between mouse and human 24 DNA in the a-lach allele lies between the translational stop site and the polyadenylation signal. Human a- 26 lactalbumin mRNA contains 120bp of mouse sequences in 27 the 3' untranslated end. Human replacement and mouse l ac-lactalbumin mRNA were detected by hybridisation with 29 a mouse RNA probe and distinguished by the size of RNA fragments protected from ribonuclease digestion. Human 31 sequences are indicated by the chequered bar and mouse 32 sequences by the shaded bar. Restriction enzyme sites 33 shown are: HindIII Bal XbaI 34 In Figure 15B the autoradiograph shown is of a 36 polyacrylamide urea thin layer gel. The source of RNA WO 96/02640 PCT/GB95/01651 1 is indicated above the lane markers. Lane A shows a 2 wild-type RNA hybridised to the mouse RNA proe 3 undigested with ribonuclease. Lanes D to J show RNA 4 samples from a-lacm/a-lach heterozygotes, the numbers indicate individual mice and are the source of the 6 quantitative estimates given in Figure 15. The 7 predicted size of protected fragments are indicated.

8 9 The upper portion of Figure 16 shows phenyl-Sepharose elution profiles of three milk samples. 1, a-lach/-lach 11 homozygote (mouse 2, a-lacm/a-lach heterozygote 12 (mouse 3, a-lacm/a-lacm wild type. The lower 13 portion shows a standard curve of known quantities of 14 human a-lactalbumin plotted against integrated peak 13 area.

16 17 The present invention will now be further described 18 with reference to the following, non-limiting, 19 examples.

23 Example 1 Clonina of Bovine a-lactalbumin gene 22 23 There are three known variants of bovine a-lactalbumin, 24 of which the B form is the most common. The A variant from Bos (Bos) nomadicus f.d. indicus differs from the 26 B variant at residue 10: Glu in A is substituted for 27 Arg in B. The sequence difference for the C variant 28 from Bos (Bibos) javanicus has not been established 29 (McKenzie White, Advances in Protein Chemistry 41, 173-315 (1991). The bovine a-lactalbur 'n gene 31 (encoding the B form) was cloned from ge, mic DNA using 32 the PCR primers indicated in Figure 1. The primers 33 have been given the following sequence ID Nos: 34 Ba-2 SEQ ID No 1 Ba-7 SEQ ID No 2 36 Ba-8 SEQ ID No 3 WO96/02640 PCT/GB95/01651 16 1 Ba-9 SEQ ID No 4 2 The source of DNA in all the PCR reactions was blood 3 from a Holstein-Friesian cow.

4 The length of the amplified promoter region using 6 primer Ba-9 in combination with primer Ba-8 is 0.72kb.

7 This BamHI/EcoRI fragment was cloned into Bluescript 8 (pBA-P0.7).

9 The length of the amplified promoter region using 11 primer Ba-7 in combination with primer Ba-8 is 2.05kb.

12 This BamHI/EcoRI fragment was cloned into Bluescript 13 (pBA-P2).

14 The entire bovine a-lactalbumin gene including 0.72kb 16 of 5' and 0.3kb of 3' flanking region was amplified 17 using primer Ba-9 in combination with primer Ba-2.

18 These primers include BamHI restriction enzyme 19 recognition sites, which allowed direct subcloning of the amplified 3kb fragment into the BamHI site of 21 pUC18, giving rise to construct pBova-A (see Figure 2).

22 23 Ligation of the BamHI/EcoRV fragment from clone pBA-P2 24 to the EcoRV/BamHI fragment of pBOVA-a gave rise to construct pBOVA-b (see Figure 2).

26 27 Since TAQ polymerase lacks proof-reading activity, it 28 was essential to ensure that the amplified bovine a- 29 lactalbumin DNA was identical to the published bovine a-lactalbumin gene. Sequence analysis was carried out 31 across all the exons and the two promoter fragments.

32 Comparison of the bovine a-lactalbumin exons with those 33 published by Vilotte showed 3 changes: 34 Exon I at +759 C to A. 5' non-coding region; 36 (ii) Exon I at +792 CTA to CTG. Both code for Leucine WO 96/02640 PCTGB9501651 17 1 (iii) Exon II at +1231 GCG to ACG. Alanine to 2 Threonine 3 This is indicative of the more common form of the 4 protein.

6 Although misreading of sequence during the PCR 7 amplification cannot be ruled out, the above mismatches 8 were probably due to the difference in the source of 9 bovine DNA.

11 Example 2 Cloning of Human a-lactalbumin gene 12 13 The DNA sequences of human a-lactalbumin has been 14 published (Hall et al, Biochem. 242 735-742 (1987)). Using the human sequence, PCR primers were 16 designed to clone two small fragments from human 17 genomic DNA, one at the 5' end of the gene and the 18 other at the 3' end. These were subcloned into the 19 pUC18 vector and used as probes to screen a commercial (Stratagene) A genomic library. Two recombinant 21 bacteriophages, 4a and 5b.l, which contained the a-, 22 lactalbumin gene, were isolated by established methods 23 (Sambrook et al, Molecular Cloning 2nd ed., Cold Spring 24 Harbor Laboratory (1989)). Restriction mapping demonstrated that both clones contained the complete 26 coding sequence for the human a-lac gene but differed 27 in the amount of 5' and 3' sequences present (Figure 28 Sequence analysis of exons from clone 5b.l and 29 exons and 5' flanking region of clone 4a showed that these were identical to the published sequence.

31 32 Portions of the 3' sequence are given in SEQ ID Nos. 16 33 to 20 and the 5' sequence is given as SEQ ID No. 21.

34 Example 3 Cloning of bovine beta lactoglobulin gene 36 (bBLG)

II:

WO 96/02640 PCT/GB95/01651 18 1 The DNA sequence of bovine BLG (bBLG) has been 2 published (Jamieson et al; Gene, 61; 85-90, (1987); 3 Wagner, unpublished, EMBL Data Library: BTBLACEX 4 (1991)). Using the bovine sequence, PCR primers were designed to clone a fragment from the 5' portion of the 6 bovine BLG gene. This was subcloned into the pUC18 7 vector and used as probes to screen a commercial bovine 8 (Stratagene) A genomic library. Three genomic A clones 9 were isolated and characterised by restriction enzyme analysis (see Fig. Two of the clones (BB13, BB17) 11 contain the complete bBLG coding region plus various 12 amounts of flanking regions, while clone BB25 lacks the 13 coding region and consists entirely of 5' flanking 14 region. Sequence analysis showed that the end of this clone lies 12 bp upstream of the ATG translation start 16 site. Sal I fragments containing the entire insert of 17 clone BB13 and BB17 were subcloned into pUC18, as well 18 as EcoR I fragments from clone BB25 (the latte- was 19 cloned into pBluescript (Figure 4).

21 Example 4 Assembly and expression of bovine a- 22 lactalbumin constructs 23 24 Transgene constructs (Fia. 2) 26 pBova-A consists of the bovine a-lactalbumin coding 27 regi6n, =0.72kb of 5' flank and 0.3kb of 3' flank, 28 cloned as a 3kb BamHI fragment into Bluescript vector.

29 pBova-B consists of 3 fragments: 31 1. The 1.47kb BamHI to EcoRV fragment from clone pBA- 32 P2.

33 2. The 2.78kb EcoRV to BamHI fragment from clone 34 pBova-A.

3. The cloning vector Bluescript digested BamHI.

36 i WO 96/02640 PCT/GB95/01651 19 1 Bovine a-lactalbumin expression in transcenic mice 2 3 The two constructs pBova-A and pBova-B (Figure 2) were 4 injected into mouse embryos and gave rise to transgenic animals. Milk analysis by SDS-PAGE gel stained with 6 Coomasie blue (referred to as "Coomassie gels") and 7 comparison to standard amounts of a-lactalbumin showed 8 expression levels of bovine a-lactalbumin varied from 9 non detectable for pBova-A and up to =0.5-1mg/ml for pBova-B (see Figure 5 and Table 1).

i WO 96/02640 PCT/GB95/01651 Table 1 Bovine ci-lactaibumin expression in transgenic mice Mouse Coomassie 244.12 BOVA-a 244.14 BOVA-a 244.15 BOVA-a 245.23 BOVA-b 245.8 BOVA-b 245.4 BOVA-b 245.7 BOVA-b 245.21 BOVA-b 245.13 BOVA-b 249.13 BOVA-b 246.15 BOVA-b 249.18 BQVA-b 249.23.1 BOVA-b 249.23.5 BOVA-b 249.25.3 BOVA-b 249.25.7 BOVA-b 249.30.3 BOVA-b 249.30.4 BOVA-b 249.33.2 BOVA-b 249.33.3 BOVA-b 4- ==0.5-1mg/mi =1-2mg/mi Table-1 shows the relative levels of bovine a- 1actd1buxnin in transgenic mouse milk as estimated by comparison to protein standards on Coomassie gels.

WO 96/02640 PCT/GB95/01651 21 1 Example 5 Assembly and expression of human a- 2 lactalbumin constructs 3 4 a-lactalbumin is the major whey protein in humans, beta-lactoglobulin the major whey protein in sheep and 6 cow. The level of a-lactalbumin expression varies from 7 species to species, human milk contains about 8 cow milk 0.5-1.0mg/ml, and mouse milk 0.1-0.8mg/ml. To 9 define sequences which allow maximum expression of the human a-lactalbumin gene several different constructs 11 were designed. These contain a) different amounts of 12 5' and 3' flanking regions derived from the human a- 13 lactalbumin locus, b) 5' flanking regions derived from 14 the bovine a-lactalbumin locus, or c) 5' flanking regions derived from the bovine or ovine beta- 16 lactaglobulin gene. The ovine beta-lactoglobulin gene 17 promoter has been successfully used to allow high 18 expression (>10mg/ml) of heterologous genes in mouse 19 milk.

21 Transaene constructs (Figure 6) 22 23 pHA-1 consists of the human a-lactalbumin coding 24 region, =1.8kb of 5' flank and 3kb of 3' flank derived from A clone 5b.l cloned as a 7kb EcoRI/SalI fragment 26 into pucl8.

27 28 pHA-2 consists of the human a-lactalbumin coding 29 region, =3.7kb of 5' flank and =13kb of 3' flank derived from A clone 4a cloned as a =19kb SalI fragment 31 into puclS.

32 33 pOBHA (ovine beta-lactaglobulin, human a-lactalbumin) 34 was constructed from 4 DNA fragments: 1. a 4.2kb SalI/EcoRV fragment containing the ovine 36 beta-lactoglobulin promoter (see WO-A-90/05188); WO 96/02640 PCT/GB95/01651 22 1 2. a 74bp synthetic oligonucleotide corresponding to 2 a 8bp BclI linker and bases 15-77 of the human a- 3 lactalbumin sequence used as a blunt/BglI 4 fragment; 3. a 6.2kb BglI/PstI human c-lactalbumin fragment 6 derived from A clone 4a comprising a region 7 between a BglI site at base 77 and a XhoI site in 8 the 3' flank; 9 4. pSL1180 (Pharmacia) cut with PstI and Sall.

11 pBBHA (bovine beta-lactoglobulin, human a-lactalbumin) 12 was constructed from 4 DNA fragments: 13 i. a 3.0kb EcoRI fragment containing the bovine beta- 14 lactoglobulin promoter derived from clone BB25-3 and used as a EcoRI/EcoRV fragment; 16 2. a 74bp synthetic oligonucleotide corresponding to 17 a 8bp BclI linker and bases 15-77 of the human a- 18 lactalbumin sequence used as a blunt/BglI 19 fragment; 3. a 6.2kb BglI/PstI human a-lactalbumin fragment 21 derived from A clone 4a comprising a region 22 between a Bgll site at base 77 and a XhoI site in 23 the 3' flank; 24 4. Bluescript vector cut with EcoRI and PstI.

26 pBAHA (bovine a-lactalbumin, human a-lactalbumin) was 27 constructed from 4 DNA fragments: 28 1. a 0.72kb BamHI to StuI fragment containing the 29 bovine a-lactalbumin promoter derived from clone pBA-P0.7; 31 2. a 62bp synthetic oligonucleotide corresponding to 32 bases 15-77 of the human a-lactalbumin sequence 33 used as a blunt/BglI fragment; 34 3. a 6.2kb BglI/PstI human a-lactalbumin fragment derived from A clone 4a comprising a region 36 between a BglI site at base 77 and a Xhol site in I i WO 96/02640 PCT/GB95/01651 23 1 the 3' flank; 2 4. Bluescript vector cut with BamHI and PstI.

3 4 Human a-lactalbumin expression in transqenic mice 6 5 constructs were injected into mouse embryos and gave 7 rise to transgenic animals. All constructs expressed 8 human a-lactalbumin in the milk of mice. pHA-1 and 9 pHA-2, which contain the human a-lactalbumin gene and various amounts of flanking regions expressed between 1 11 to =18mg/ml (213.5 pHA-2) in the majority of animals.

12 pOBHA and pBBHA containing the human a-lactalbumin gene 13 driven by either the ovine or bovine BLG promoter had 14 slightly lower levels of expression. pBAHA containing the human a-lactalbumin gene driven by the 0.72kb 16 bovine a-lactalbumin promoter had expression levels 17 similar to pHA-1 or pHA-2 but a lower percentage of 18 transgenic animals expressed detectable levels of 19 protein. This finding is surprising as the same bovine promoter sequence driving the bovine a-lactalbumin gene 21 gave very poor results (see Example 4 and Vilotte et 22 al; FEBS, Vol. 297, 1.2. 13-18 (1992)).

23 24 Table 2 gives a summary of the relative amount of the transgenic protein. Skimmed milk from these animals 26 was analysed by SDS-PAGE stained with Coomasie blue, 27 isoelectric focusing, Western blots visualised with a 28 commercial anti-human a-lactalbumin antibody (Sigma) 29 and chromatofocusing. The results from these analyses showed that the transgenic protein was of the correct 31 size, pi and antigenicity when compared to a human a- 32 lactalbumin standard (Sigma).

WO 96/02640 PCr/GB95/01651 24 Table 2 Human rz-1actalbumiii expression in transgenic mice mouse 205.19 204 1C 204.7 230 .15.3 230.15.5 230. 15.6 230.21.5 230.21.1 pRAl pHAl pHAI, pHAl pHAl pHAl pHAl pHAl C.Y~uma ssi~e Western n++ n.d.

n.d.

n.d.

n.d.

211.18 211.17 211.16 212. 11 213.5 212.13 212 .19 213.4 212.7 232.10 233.1 231.4 232.9 231.9 232.5 231.3 232.6 237.6 235 .15 235 .19 236.6 234.1 234.4 234. N4 pHA2 pHA2 pHA2 pHA2 pHA2 pHA2 pHA2 pHA2 pHA2

BBRA

BBHA

BBHA

BBHA

BBHA

BBHA

BBHA

BBHA

BBHA

OBHA

OBHA

OBHA

OBHA'

OBHA

OBHA

n.+d n. d.

n. d.

n. d.

n. d.

n. d.

n. d.

n.d.

n. d.

n d.

n. d.

239.14 BAHA 239.4 BAHA 240.7 BAHA 239.3 BAHA 239.6 BAHA 239.12 BAHA 243.1 BAHA 242.9 BAHA 241.16 BAHA 234.14 BAHA 243.13 BAHA 243.10 BAHA 243.4 BAHA

F++

4.

WO 96/02640 PCT/GB95/01651 1 2 Table 2 shows the relative levels of human a- 3 lactalbumin in transgenic mouse milk as estimated by 4 comparison to protein standards on Coomassie gels and Western Blots.

6 7 0.5 mg/ml 8 9 =1-2 mg/ml =2-3 mg/ml 11 >5 mg/ml 1' n.d. not determined .3 14 The results from several mice are shown in Figs. 7 and 8. Fig. 7 shows an SDS-PAGE analysis of skimmed 16 transgenic mouse milk run against a non-transgenic 17 control mouse milk. Fig. 8 shows a Western blot of 18 human a-lactalbumin transgenic milks run against a 19 human a-lactalbumin standard.

21 Example 6 Exoression of Mutaqenised Bovine a- 22 Lactalbumin under the control of the Human a- 23 Lactalbumin Promoter in vivo 24 Expression of the human a-lactalbumin transgene is 26 considerably higher than that of the native bovine a- 27 lactalbumin transgene, reflecting the difference in 28 expression levels of the endogenous bovine and human 29 genes. As this might be caused by differences in the 5' control region, the 5' region of the bovine a- 31 lactalbumin transcriptional start site was substituted 32 with sequences from the human a-lactalbumin gene.

33 34 Two constructs were made, namely PKU-5 and PKU-1H, 31 which incorporate the amino acid substitutions shown in 36 Table 3.

37 WO 96/02640 PCT/GB95/01651 The following SEQ primers used.

PKU-1 PKU-2 PKU-2L PKU-3 PKU-4 PKU-6 PKU-7 PKU-8 PKU-9 ID Nos. have been assigned to the PCR

SEQ

SEQ

SEQ

SEQ

SEQ

SEQ

SEQ

SEQ

SEQ

SEQ

SEQ

ID No.

ID No.

ID No.

ID No.

ID No.

ID No.

ID No.

ID No.

ID No.

ID No.

ID 16 17 18 In a first cloning step three fragments were subcloned 19 into the EcoRI/BamHI site of pUC18: 21 the EcoRI to PvuI fragment derived by PCR 22 amplification using PKU-primer 7 in combination 23 with 8 (6ee Figure 24 the PvuI to BsaBI fragment derived by PCR 26 amplification using PKU-primer 9 in combination 27 with 10 (see Figure 10); and 28 29 The BsaBI to HindIII fragment derived from pBA.

33 31 The final construct included 6 DNA fragments: 32 33 the 3.7kb SalI to KpnI fragment containing the 34 human a-lactalbumin promoter derived from A clone 4a (Figure 3); WO 96/02640 PcrIGB95101651 27 1 the 152bp synthetic oligonucleotide containing 2 human a-lactalbumin sequences from the KpnI site 3 to the AUG and bovine a-lactalbumin sequences from 4 the AUG to the HapI site; 6 the 1.25kb HpaI to HindIII fragment from the first 7 subcloning step; 8 9 the 0.95 kb HindIII to BglII fragment derived from pBA; 11 12 the 3.7kb BamHI to XhoI fragment from the 3' flank 13 of the human a-lactalbumin gene derived from A 14 clone 4a (Figure 3) used as a BamHI fragment; and 16 a Bluescript KS- vector cut with SalI and BamHI.

17 18 PKU-1H was constructed in the same way as PKU-5 with 19 the exception of fragment which was derived from PKU-1.

21 22 PKU-1 was constructed from six DNA fragments (see 23 Figure 9): 24 a 2.04kb SstI to Hpal fragment derived from 26 pBOVA-6; 27 28 a 0.46kb HpaI to PvuI fragment derived from PCR 29 product A (PKU-primer pair 1 and 2; see Figure 31 32 a 0.60kb Pvul to BsaBI fragment derived from PCR 33 product B (primer pair 3 and 4; see Figure 34 a 0.22kb BsaBI to HindIII fragment derived from 36 pBOVA-6; WO 96/02640 PCT/GB95/01651 28 1 a 0.95kb HindIII to BglII fragment derived from 2 pBOVA-6; 3 4 the vector pSL1180 digested with SstI and BglII.

6 Table 3 7 8 Amino Acid Substitutions present in Transgene Constructs 11 Substitutions Human promoter Plasmid 12 Human 3 flank 13 pos'n 9 30 53 14 Tyr, Tyr, Tyr, Tyr pPKU-1H Ser, Tyr, Leu, Leu 16 17 Expression in transgenic mice 18 19 The two constructs PKU-1H and PKU-5 have been injected into mouse embryos. So far transgenic animals were 21 derived for the PKU-5 construct. These animals are set 22 up for breeding to allow milk analysis.

23 24 Example 7 Effect on Lactation by disruption of a- Lactalbumin deficiency and insertion of human a- 26 Lactalbumin aene replacement in mice 27 28 Materials and Methods 29 Mouse lines 31 Mice carrying the null a-lactalbumin allele and the 32 humanised a-lactalbumin replacement allele were derived 33 by breeding chimeras produced from the targeted 34 embryonic stem cell clones M2 and F6 respectively against Balb/c mates, as described previously 36 (Fitzgerald et al J. Biol. Chem 245:2103-2108). During WO 96/02640 PCrGB95/01651 29 1 the breeding of these strains, a-lactalbumin genotypes 2 were determined by Southern analysis of genomic DNA 3 prepared from tail biopsies.

4 RNA analysis 6 7 Total RNA was prepared by the method of Auffray and 8 Rougeon (Eur. J. Biochem 10f:303-14) from abdominal 9 mammary glands of female mice 506 days postpartum.

Northern analysis was carried out according to standard 11 procedures (Sambrook et al, Molecular cloning). Probes 12 used for hybridisation were: a 3.5kb BamHl fragment 13 containing the complete mouse a-lactalbumin gene; and a 14 l.lkb rat f-casein cDNA (Blackburn et al Nucl. Acids Res 10:2295-2307).

16 17 RNAse protection analysis 18 19 32 P-CTP radiolabelled antisense RNA was transcribed by T7 RNA polymerase (Promega) from a 455 bp HindIII-BalI 21 mouse a-lactalbumin fragment (Figure 15A) cloned in 22 Bluescript KS. The conditions for the transcription 23 :eaction, solution hybridisation and RNAse digestion 24 were as recommended by Promega. Protected fragments were separated by polyacrylamide gel electrophoresis 26 and visualised by autoradiography.

27 28 Milk composition and yield analysis 29 Milk samples were collected between days 3-7 of 31 lactation under Hypnorm (Roche)/Hypnovel (Janssen) 32 anaesthesia. 150mU of oxytocin (Intervet) was 33 administered by intraperitoneal injection and milk 34 expelled by gentle massage. Milk fat content was measured as described by Fleet and Linzell Phy iol 36 175:15). Defatted milk was assayed for protein

I

PCr/GB95/01651 WO 96/02640 1 (Bradford Anilyt. Biochem 22:248-54) and lactose was 2 measured enzymatically by sequential incubation with 3 galactosidase, (Boehringer) glucose oxidase and 4 peroxidase (Sigma) by a method adapted from that of Bergmeyer and Bernt (Methods in Enzyme Analysis 3:1205- 6 1212).

7 8 Milk yield was estimated using a titrated water 9 technique described by Knight et al., (Comp. Biochem.

Physiol 84A:127-133) in mice suckling young over a 48 11 hour period between days 3 and 6 of lactation.

12 13 Milk a-lactalbumin analysis and quantification 14 Milk samples were analysed on 16% of SDS-PAGE gels 16 (Novex) and western blotted onto Immobilon P membrane.

17 a-lactalbumin was detected by absorption with rabbit 18 anti-human a-lactalbumin antiserum (Dako), followed by 19 goat anti-rabbit IgG peroxidase antibody conjugate and visualised with an enhanced chemiluminescence system 21 (Amersham).

22 23 a-lactalbumin in milk samples was quantified by a 24 modification of the method of Lindahl et al., (Analyt.

Biochem 140:394-402) for calcium dependant purification 26 of a-lactalbumin by phenyl-Sepharose chromatography.

27 Milk'samples were diluted 1:10 with 27% w/v ammonium 28 sulphate solution, incubated at room temperature for 29 minutes and centrifuged. Supernatant was mixed with an equal volume of 100mM Tris/Cl, pH 7.5, 70mM EDTA and 31 loaded onto a column (2004l packed volume) of phenyl- 32 Sepharose (Pharmacia) pre-equilibrated with 33 Tris/Cl, pH 7.5, ImM EDTA. The column was washed with 34 the same buffer and a-lactalbumin eluted with Tris/Cl, pH 7.5, 1mM CaCl2. The optical absorbance at 36 280nm of the column was monitored and integrated peak

L--

pCTIGB950165 WO 96/02640 31 1 areas corresponding to the a-lactalbumin fraction 2 computed. A standard curve was constructed using known 3 quantities of purified human a-lactalbumin from 0 to 4 2.46 mg/ml (Figure 16).

6 Histology 7 8 Pups were removed for two hours from lactating mothers 9 on the sixth day postpartum, mothers sacrificed and thoracic mammary glands were dissected, preserved in 11 neutral buffered formalin, paraffin embedded and 12 stained with haematoxylin/eosin by standard methods.

13 14 Results 16 Mouse a-lactalbumin gene deletion 17 18 A line of mice in which a 2.7kb fragment covering the 19 complete mouse a-lactalbumin coding region and a 0.57kb of promoter has been deleted and replaced with a 2.7kb 21 fragment containing a hypoxanthine phosphoribosyl- 22 transferase (HPRT) selectable marker gene was 23 established as described in Stacey et al, 1994 supra 24 (see Figure 11). Animals carrying this allele are designated a-lac-. The wild type mouse a-lactalbumin 26 allele is designated a-lac m 27 28 Northern analysis of RNA from mammary glands taken on 29 the fifth day of lactation showed that a-lactalbumin mRNA was absent in a-lac-/a-lac- homozygotes (see 31 Figure 12) confirming that the targeted a-lactalbumin 32' gene has been removed and that no other source of a- 33 lactalbumin mRNA exists. Hybridisation of the same RNA 34 samples with a f-casein RNA in all samples (see Figure 12).

36 i I i WO 96/02640 P~IGD95/01651 32 1 a-Lactalbumin deficiency has no apparent effect in mice 2 other than during lactation. a-lac-/a-lac- homozygotes 3 and a-lacm/a-lac-heterozygotes of both sexes are normal 4 in appearance, behaviour and fertility. However, alac-/a-lac- homozygous females cannot rear litters 6 successfully. Their pups fail to thrive and die within 7 the first 5-10 days of life. Offspring of homozygous 8 a-lac-/a-lac- females do survive normally when 9 transferred to wild type foster mothers. Conversely, offspring from wild type mothers transferred to 11 homozygous a-lac-/a-lac- mothers are not sustained.

12 Table 4 shows that pups raised by a-lac-/a-lac- mothers 13 are approximately half the weight of those raised by a- 14 lac m /a-lac m wild type mice. Estimates of milk yield are consistent with this, a-lacm/a-lac- heterozygotes 16 produce similar quantities of milk as wild type, but 17 the yield of a-lac-/a-lac- homozygotes was severely 18 reduced (Table 4).

Table 4 Milk composition, pup weight, mammary tissue weight and milk yield in targeted mouse lines Genotype ct-1acm/ctlac' ca-1acm/ct1ac c-lac -Ic-lac acl1acm/ca.ac h a-lach/a-1aCh Fat ~28.23 1.65(7) 29.6 1.3(6) 45.25 25.25 1.36(7) 21.2 0.23(4)* Protein 87.52 5.82(7) 95.81 164.63 94.51 5.97(7) 77.07 1.05(4) (mg/mi) Lactose 62.44 9.27(7) 42.7 0.7 42.40 ±1.93(7) 56.85 3.8(4) Single 2.82 0.25(8) 3.14 1.52 2.9 ±0.15(8) 3.4 ±0.75(4)

PUP

weight (g) Mammary 0.34 0.06(7) 0.4 0.35 ±0.05(8) 0.31 ±0.04(6) 0.51 ±0.09(4) tissue weight per pup (g) Milk 7.51 0.44(4) 6.7 ±0.38(6) 1.37 n.t. 9.94 ±0.65(5)* yield (giday) Statistical analysis by unpaired t-test, p<0.05; ***p<0.O0l; Values are means SE.

Figures in brackets indicate the number of mothers analysed.

not tested C..3 pc/GR9501651 UWI\ OrInAn 34 1 Milk was obtained from each genotype by manual milking 2 and the composition of key components analysed. Milk 3 from a-lac'/a-lac-heterozygotes was indistinguishable in 4 appearance from wild type milk and showed fat and protein contents similar to wild type (Table While 6 lactose concentration appeared to be slightly reduced 7 in a-lacm/a-lac- heterozygotes, statistical analysis 8 showed that the difference was not significant. In 9 contrast, milk from a-lac-/a-lac- homozygotes was viscous, difficult to express from the teats and was 11 markedly different in composition to wild type. Fat 12 content was -60% greater than wild type, protein content 13 was -88% greater, and lactose was effectively absent.

14 The apparent 0.7mM lactose detected in a-lac-/a-lacfemales represents milk glucose content, since the 16 lactose assay used involved the enzymatic conversion of 17 lactose to glucose. Direct assay of glucose in wild 18 type milk indicated a concentration of 1.8mM.

19 Western analysis of milk,protein failed to detect a- 21 lactalbumin in milk from a-lac-/a-lac- homozygotes (see 22 Figure 13, Lane This was confirmed by phenyl- 23 Sepharose chromatography, a technique used to 24 specifically identify a-lactalbumin which has been adapted to obtain quantitative estimates of milk a- 26 lactalbumin content (Table 5; see also Figure 16).

27 When applied to milk from a-lac-/a-lac-homozygotes no 28 a-lactalbumin was detected. In contrast, a-lactalbumin 29 concentration in a-lacm/a-lac- heterozygote milk was estimated as 0.043mg/ml, approximately half that of 31 wild type (Table -C iI PCTIGB9S/01651 lirrr rr A~r JA VVuJ YWOAU 35 1 Table 2 Milk a-lactalbumin content.

3 4 Source a-lactalbumin (mq/ml) Human 2.9 0.1(2) 6 a-lacl'/c-lac m mice 0.09 i 0.005(6) 7 a-lac-/a-lac-mice 0 (3) 8 a-lac m /a-lac-mice 0.043 0.004(5) 9 a-lac"'/a-lach mice 0.65 0.07(4) a-lach/a-lach mice 1.38 i 0.12(5) 11 12 a-Lactalbumin content of milk samples were estimated by 13 phenyl-Sepharose chromatography.

14 Values are means SE.

16 17 Figures in brackets indicate the number of mothers 18 analysed.

19 a-Lactalbumin deficiency has no apparent effect on 21 mammary gland development. Table 4 shows that total 22 mammary tissue weights of wild type, heterozygous a- 23 lac'/a-lac- and homozygous a-lac-/a-lac- lactating 24 mothers were not significantly different. Light microscopic analysis of mammary glands (Figure 14) 26 revealed that heterozygous and homozygous a-lac-/a-lac- 27 glands were histologically normal. However, the 28 alveoli and ducts of homozygous glands were distended 29 and clogged with material rich in lipid droplets.

31 Replacement of mouse a-lactalbumin by human a- 32 lactalbumin 33 34 We have generated mice carrying the numan a-lactalbumin gene at the mouse a-lactalbumin locus. The 2.7kb mouse 36 a-lactalbumin fragment deleted at the a-lac- null 37 allele was replaced by a 2.97kb fragment containing the 38 complete human a-lactalbumin coding region and i Ml?/t A PT/t-QRO/ft1l 1l 1VVU YUIUhW 36 1 flanking sequences. The human fragment stretches from 2 0.77kb upstream of the human transcription initiation 3 site to an EcoRI site 136bp 3' of the human 4 translational stop site. Junctions with mouse sequences were made at a BamHI site 0.57kb upstream of 6 the mouse transcription initiation site and at an Xbal 7 site 147bp 3' of the mouse translational stop site (see 8 Stacey et al, 1994, supra; see also Figure 11). Here 9 we describe our analysis of animals carrying this allele, designated a-lach.

11 12 Deletion of the murine a-lactalbumin gene established 13 that a-lactalbumin deficiency blocks lactose synthesis 14 and severely disrupts milk production. We have used the a-lach allele to test the ability of human a- 16 lactalbumin to restore milk production in the absence 17 of mouse a-lactalbumin. a-lacm/a-lach heterozygous and 18 a-lach/a-lach homozygous mice were normal in appearance, 19 fertility and behaviour.

21 In contrast to a-lac-/a-lac- mice, a-lach/o-lach 22 homozygous mothers produce apparently normal milk and 23 rear offspring successfully. Table 4 shows that pups 24 raised by i-lacm/a-lac h heterozygous and a-lach/a-lach homozygous females are similar in weight to those of 26 wild type mothers. This is supported by our 27 observation that these animals raised successive 28 litters of pups entirely normally. These data 29 constitute clear evidence that the human gene can functionally replace the mouse gene. Analysis of milk 31 composition (Table 4) shows that lactose concentration 32 is similar in all genotypes. Although both protein and 33 fat concentrations seem reduced in a-lach/a-lach 34 homozygous animals, only the fat reduction was judged statistically significant by unpaired t-test. These 36 animals show an increase in milk volume over wild type 37 (Table 4).

38 .VY1 WV% 04I/0MAn PCTrrrB95/1651 1 Relative quantification of human and mouse a- 2 lactalbumin RNA.

3 4 Human milk contains considerably more a-lactalbumin (2.5mg/ml) than murine milk (0.lmg/ml). We wished to 6 determine whether the human a-lactalbumin fragment 7 retained a high level of expression when placed at the 8 mouse locus, or assumed a lower level more 9 characteristic of the mouse gene. a-lacm/a-lach heterozygous mice provided an ideal means of addressing 11 this question, as the expression of the human gene 12 could be directly compared with its mouse counterpart 13 in the same animal.

14 Figure 15A shows the strategy used to compare levels of 16 mouse and human a-lactalbumin mRNA. Because the 17 junction between human and mouse a-lactalbumin 18 sequences lies upstream of the polyadenylation site, a- 19 lach mRNA contains a "tag" of 120 bases of untranslated mouse sequences at the 3' end. A uniformly 21 radiolabelled mouse RNA probe was used in a 22 ribonuclease protection assay to detect and distinguish 23 human and mouse a-lactalbumin mRNA in the same RNA 24 sample. The relative abundance of each mRNA was calculated from the amount of label in fragments 26 protected by human and mouse mRNAs.

27 28 A ribonuclease protection assay was performed and the 29 results are shown in Figure 15B. Lane A shows the undigested 455 base probe and Lane K shows that yeast 31 tRNA did not protect any fragments. Wild type mouse 32 RNA protected a fragment consistent with the predicted 33 305 base RNA from endogenous mouse a-lactalbumin RNA 34 (see Lane Homozygous a-lach/a-lach gland RNA protected a smaller band consistent with the predicted 36 120 base RNA protected by human a-lactalbumin mRNA (see 37 Lane Lanes D-J show results consistent with a 38 series of heterozygous a-lacm/a-lach animals were e wn oI/0)nin SCIrr/o95/01651 1 obtained (see Lanes D t, Small and large protected 2 fragments in each sample indicate the presence of both 3 human and mouse a-lactalbumin mRNA. Protected 4 fragments were excised from the gel, radioisotope content measured, adjusted for the size difference and 6 the ratio of human to mouse a-lactalbumin mRNA 7 estimated. Table 6 shows the amount of radioisotope 8 present in the 305 base and 120 base fragments excised 9 from Lanes D to J of the gel shown in Figure 15B, and the calculated ratio of human to mouse a-lactalbumin 11 mRNA in each a-lacm/a-lach heterozygote. It is apparent 12 that, although there was variation between individual 13 animals, human a-lactalbumin mRNA was significantly 14 more abundant than mouse mRNA. Averaging the seven alacm/a-lach heterozygotes gives a value of 16 greater expression for human a-lactalbumin mRNA.

M

WO 96/02640 PCT/GB9S/016S1 1 Table 6 2 3 Relative quantification of human and mouse a- 4 lactalbumin mRNA in a-lacm/a-lac" mamindry glands Lane Mouse# 9 D 120 base fragmenta 5957 5770 4825 305 base fragmenta 1000 547 810 human/mouse RNA ratiob 15:1 26:1 15:1 6018 5206 5481 26858 1077 1452 1117 3561 14:1 9:1 12:1 19:1 Lane designations indicate the source of protected fragments and correspond to those shown in Figure a. numbers are expressed in counts per minute b. Ratio between c.p.m. of 120 base fragment multiplied by 2.54 (to adjust for size difference) and c.p.m. of 305 base fragment.

IL 1/1 PrT/fl5/1651 i tf n lnE jall VW0UIVAV"U 40 Human a-lactalbumin protein expression 2 3 A Wstern analysis of a-lactalbumin .n targeted mouse 4 lines was conducted. Human a-lactalbumin can be distinguished from mouse a-lactalbumin by its faster 6 electrophoretic mobility (see Lanes A, A prominent 7 lower band in a-lach/a-lach homozygotes and a-lach/a-lacm 8 heterozygotes was observed (see Lanes C, D, G, and 9 corresponds to the position of the human a-lactalbumin standard and was only observed in mice which express 11 human a-lactalbumin generated either by gene targeting 12 or by pronuclear microinjection (data not shown). This 13 identity was confirmed by phenyl-Sepharose 14 chromatography (.See Figure 16) and analysis of ?eptides released by cyanogen bromide cleavage (data not shown).

16 The band with slower mobility, similar to mouse a- 17 lactalbumin, is also a human a-lactalbumin gene product 18 the nature of which is unknown. This species varied in 19 intensity with a-lach gene dosage (see Lanes G, H) and was also present in milk from human a-lactalbumin 21 transgenic mice generated by pronuclear microinjection 22 (data not shown).

23 24 The a-lactalbumin content of milk samples was quantified by phenyl-Sepharose chromatography. Figure 26 16 shows superimposed absorbance profiles of column 27 eluates of three illustrative milk samples including 28 the a-lacm/a-lach heterozygote and a-lach/a-lach 29 homozygote shown in Figure 13. The peaks corresponding to eluted a-lactalbumin are marked. a-Lactalbumin 31 contents were estimated by comparing the integrated 32 peak areas with the human a-lactalbumin standard curve 33 shown. The relationship between integrated peak area 34 and a-lactalbumin quantity was linear and highly reproducible. a-Lactalbumin content for the samples 36 shown in Figure 16 were estimated as follows: a-lacoa- 37 lac" wild-type 0.mg/ml; a-lacm/a-lach heterozygote #76 38 0.45mg/ml; a-lach/a-lach homozygote #22 0.9mg/ml. Table I I WII 60 trc\A PVrCT/GBni01651 UJIv (btanien T41 1 5 shows the concentration of a-lactalbumin in milk 2 samples from targeted mouse lines and lactating women.

3 It is clear that the concentration of a-lactalbumin in 4 milk is directly related to gene dosage, eg a-lacm/alac- heterozygotes shown an a-lactalbumin concentration 6 half that of wild type. Given that the volumes of milk 7 produced by these mice are similar (Table the 8 concentration of a-lactalbumin provides a reasonable 9 indication of the quantity synthesised. The relative proportions of human and mouse a-lactalbumin components 11 in a-lac'/a-lach heterozygote milk were estimated by 12 assuming that a-lactalbumin expression from a single 13 mouse allele was 0.043mg/ml and the rest represented 14 human a-lactalbumin. This is consistent with the amounts of a-lactalbumin expressed by a-lacm/a-lac- 16 heterozygotes and wild type mice. Therefore, a-lacm/a- 17 lac-heterozygotes were estimated as expressing 18 0.61mg/ml human and 0.043mg/ml mouse a-lactalbumin.

19 Thus, human a-lactalbumin is approximately 14-fold more abundant than mouse a-lactalbumin in a-lacm/a-lach 21 heterozygote milk. This is remarkably consistent with 22 the relative proportions of mRNA.

23 24 Example 8 Enhanced expression of a heterologous gene.

26 These data confirm that the upstream promoter region 27 (AUG to about -3.7 kb) which is included in the pHA-2 28 construct enhances expression of a heterologous gene.

29 Table 7 shows the results of milk analysis from pHA-2 transgenic founder females. Out of 10 females, 6 31 animals expressed high levels of human a-lac. 3 32 animals failed to express detectable levels of human a- 33 lac (less than 0.2 mg/ml in this assay), all 3 also 34 failed to transmit the transgene. We can neither be certain whether they were low expressors or not 36 transgenic.

37 rw t4lanCIA14MI VD PJAFYIUIO'U 42 Z I7I')U Tale8 Constructs PKU-0 to BALT-B all contain the bovine a-lac promoter (about 2kb). Constructs i'KU-1H to PKU-16 all contain the human cz-lac promoter (3.7kb).

Using the human cx-lac promoter increased the expression of the transgene to almost 100%.

These data show that the use of the human ca-lac promoter achieves a higher level of expression than the use of the bovine promoter, and induces expression in more animals than the bovine promoter.

TABLE 7 CONSTRUCT MOUSE SEX C mg/mi TRANS. MI FREQ.

MILK

___ANALYS -pHA2 210-17 14- 1/21=5% 211-12 F <1 14/54 211-16 F 10 5 3/4 *F >>10 ND 0/8 211-24, 27, 14 29, 31, 36, 37, 42, 46, 47, 54 F 10 ND 0/5 8/77=10% F <10 1 0/4 212/13 F >1 1-5 2/4 212-19 F >10 ND 0/8 212-36, 44, M4 46 213-4 F <10 5 0/2 2/14=14% F 110 2/8 OVERALL MIF 25/166=15% ND Not detectable TBA To be analysed TBO To bred on C Copy number TRANS. Transmission MI FREQ. Integration frequency WO 96/02640 WO 9602640PCrIGJ39SI01651 43 TABLE 8 Constr. Transgenics Expressers max. exeprq PKU-0 6F/5M 3/5 500 PKU-1 18F/23M 5/18 200 PKU-2 8F/714 3/8 400* PI(U-3 6F/5M 2/6 100* PKTJ-4 7F/3M 5/18 300* BALT-A 34F/24M n.t. n.t.

BALT-B 10F/18M n.t. n.t.

PKU-1H 6F/5M 4/5 100 13F/14M 13/13 800 P1(1-6 3F/414 2/2 100 PKU-7 4F/11M 1/1 PKU1-16 13F/12M n.t. n.t.

HaPKU-2 n.t. n.t. ii.t.

n.t. =not tested estimate WO 96/02640 PCT/GB95/01651 44 SEQUENCE LISTING GENERAL INFORMATION:

APPLICANT:

NAME: PPL THERAPEUTICS (SCOTLAND) LIMITED STREET: ROSLIN CITY: EDINBURGH COUNTRY: UNITED KINGDOM POSTAL CODE (ZIP): EH25 9PP (ii) TITLE OF INVENTION: Alpha-lactalbumin Gene Constructs (iii) NUMBER OF SEQUENCES: 21 (iv) COMPUTER READABLE FORM: MEDIUM TYPE: Floppy disk COMPUTER: IBM PC compatible OPERATING SYSTEM: PC-DOS/MS-DOS SOFTWARE: Pacentln Release Version #1.30 (EPO) INFORMATION FOR SEQ ID NO: 1: SEQUENCE CHARACTERISTICS: LENGTH: 27 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1: GCGGATCCAC AACTGAAGTG ACTTAGC 27 INFORMATION FOR SEQ ID NO: 2: SEQUENCE CHARACTERISTICS: LENGTH: 35 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 2: SUBSTITUTE SHEET (RULE 26)' WO 96/02640 PCT/GB95/1651 GATGGATCCT GGGTGGTCAT TGAAAGGACT GATCC INFORMATION FOR SEQ ID NO: 3: SEQUENCE CHARACTERISTICS: LENGTH: 43 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 3: GCAGGCGAAT TCCTCAAGAT TCTGAAATGG GGTCACCACA CTG 43 INFORMATION FOR SEQ ID NO: 4: SEQUENCE CHARACTERISTICS: LENGTH: 33 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 4: GAGGATCCAA TGTGGTATCT GGCTATTTAG. TGG 33 INFORMATION FOR SEQ ID NO: SEQUENCE CHARACTERISTICS: LENGTH: 46 base pairs TYPE: nucleic acid STRANDEDNESS: single (D)-TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: GCTGAATTCG TTAACAAAAT GTGAGGTGTA TCGGGAGCTG AAAGAC 46 INFORMATION FOR SEQ ID NO: 6: SUBSTITUTE SHEET (RULE 26) WO 96/02640 PCr/GB95/01651 46 SEQUENCE -CHARACTERISTICS: LENGTH: 58 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID No: 6: GCGGATCCGA TCGCTTGTGT GTCATAACCA CTGGTATGGT ACGCGCTACA GACCCCTG 58 INFORMATION FOR SEQ ID NO: 7: SEQUENCE CHARACTERISTICS: LENGTH: 58 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cONA (xi) SEQUENCE DESCRIPTION: SEQ ID No: 7: GCGGATCCGA TCGCTTGTGT GTCATAACCA CTGCTATGGA GCGC3GTACA GACCCCTG 58 INFORMATION FOR SEQ ID NO: 8: SEQUENCE CHARACTERISTICS: LENGTH: 69 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (Ui) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID INOr: GCGGATCCGA TCGTACAAAA CAATGACAGC ACAGAATATG GACTCTACCA CAT.A-ATAAT AAATTTGG 69 INFORMATION FOR SEQ ID NO: -9: SEQUENCE CHARACTERISTICS: SUBSTITUTE SHEET (RULE 26) WO 96/02640 PCTIGB95101651 47 LENGTH: 34 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 9: GCTCTAGATC ATCATCCAGG TACTCTGGCA GGAG 34 INFORMATION FOR SEQ ID NO: SEQUENCE CHARACTERISTICS: LENGTH: 24 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: GCTGAAGCTT CACTTACTTC ACTC 24 INFORMATION FOR SEQ ID NO: 11: SEQUENCE CHARACTERISTICS: LENGTH: 65 base pairs TYPE: nucleic STRANDEDNESS: s; TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 11: GCGGATCCAA AGACAGCAGG TGTTCACCGT CGACGACGCC TACGTAACTT CTCACAGAGC CACTG INFORMATION FOR SEQ ID NO: 12: SEQUENCE CHARACTERISTICS: LENGTH: 46 base pairs SUBSTITUTE SHEET (RULE 26) WO 96/02640 Per/GB95/01651 48 TYPE: nucleic acid STRLANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 12: GCTGAATTCG TTAACAAAAT GTGAGGTGAG CCGGGAGCTG AAAGAC 46 INFORMTION FOR SEQ ID NO: 13: SEQUENCE CHARACTERISTICS: LENGTH: 54 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 13: GCGGATCCGA TCGCTTGTGT GTCATAACCA CTGGTATGAT ACGCGGTACA GACC 54 INFORMATION FOR SEQ ID NO: 14: SEQUENCE CHARACTERL7STICS: LENGTH: 69 base pairs TYPE:. nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTI ON: SEQ ID NO: 14: GCGGATCCGA TCGTACAAAA CAAT-GACAGC ACAGAATATG GACTCCTCCA GATAAATAAT AAAATTTGG 69 INFORMATION FOR SEQ ID NO: SEQUENCE CHARACTERISTICS- LENGTH: 34 base pairs TYPE: nucleic acid SUB8STITUTE SHEET (RULE 26) WO 96/02640 WO 9602640PCT/GB95101651 49 STRANDEDNESSi single TOPOLOGY: linear (ii) MOLECULE TYPE: cPNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: CCTCTAGATC ATCATCCAGC AGCTCTGGCA GGAG 34 INFORMATION FOR SEQ ID NO: 16; SEQUENCE CHAR.ACTERISTICS: LENGTH: 264 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 16: GGATCCAAAG TTGGCTAAAC ACTGGCCGGG TGCAGTGCTT CCACCTGTA.; TTCCAGCACT TTGGAkAGGCT GAGCGTGCGCA GATTGCTTG?. GGTCAGG.A.T TTGAC ACCAG CTTGGCTAAC 120 AGCAANACCC TGTCTCTACC AAAPAGTACAA AAATTATCTG GGTGTGGTGG CAGGCGCCTG 180 TAATCCCAGC TACTCGGGAG GCTGAGGCAG AAGAATTGTT TGAACCTGGG AGGCAGAGGT 240 TGTAGTGAGC TGAGATCGCT CATT 264 INFORMATION FOR SEQ ID NO: 17: SEQUENCE CHARACTERISTICS: LENGTH: 803 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 17: TCTTTTTCAA TTATTCATTT GTTACAGTGG GTTATGATAC AAATGTTTAT AGATGCCTAC SUBSTITUTE SHEET (RULE 26) WO 96/02640 PCT/GB95/01651 TCTGTACTAG TACTACAG AG CAGTTTTTCT GTGTTTATAT TCAGTTCAkAT TGTAOTGTOT

TGAGTTGTAT

GTACAAGCA-A

TCTCTTGTTC

GATTGGAGTT

ACCTCCACGT

AGGCGCCTGC

TGTTGCTCAG

CAGOTGTGAG

TCCACTGAAT

TATCCAGTGA

CTCCTTTTTT

WWTAATC CAT

CACTCACCAA

CTAGTCTGTA

GCCCTTGTTG

CTTGGGTTAA

CACCATO CCC

GCTGGTCTTG

CCACAGTGCC

TAAGAATTCT

GGTATGGCTG

ACAAATAAAA

GTATTAAATC AAATAALACtA

AGGCATTTGG

CTTATTTTCC

CCCAGGCTGG

AG CGTTTCT C

GGCTAATTTT

AAAkCTCCCCA

TGGCCTGTAC

TTTGAGAAAA

GATCAGCTTC

GGTCTGCATT

CCACACTAC

AGTO CAG TG

CTGCCTCAGC

TGTATTTTTA

CCTCGGCCCT

ATTTTTTAAA

TGAATCAATA

ATGACATACA

ACAAAATGCC

TGGAATTCTC

TTATGTATAT

CACGATTCTT

CTCCTGACTA

OTAGAGATGG

TCCCAAATGC

TTTCAATGTC

AATCT AT ACA

ATGTTCTTTG

AGO CAAACTC

ATATTTTTGA

GGCTCACGAG

CTGGATTAC

GCTTTCACCA

GCTGGGATTA

TAATATGGTG

CTG CCTCCTT 120 180 240 300.

360 420 480 540 600 660 720 780 TGCCAGTAGT TCTCCTCCTC ATTGTATATG TTGAAGGOT ACAACTTOAT OTTTGTTATA TOTATACACT TAAATGTCAC CAC INFORMATION FOR SEQ ID NO: 18: SEQUENCE CHARACTERISTICS: LENGTH: 238 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 18: TTTGOTAOA ACCAGACAGT AAACTTGCTG TTCTCTTCCT GAGATCTTTT OTTOAGATOC TGAATAGGAG GCAGCATGGC AG'CTGAGCTA TCTGTTCTG C TTCTCTACC TCTGTCTCTT TCCCTTAOOC CTAAAATGA.A GCTCTAAOCC AAOCAAAOGT CGAAGTCAT CCAGACTAAT TGGGAAGCGG OTAOOCTCCA GGGAGTGGCT CTCAOAGAGC AG.ACCATTTA CTGAGCTC INFORMATION FOR SEQ ID NO: 19: SEQUENCE CHARACTERISTICS: LENGTH: 162 base pairs TYPE: nuclei4c acid 803 120 180 238 SUBSTITUTE SHEET (RULE 26) 74,m WO 96/02640 PCTIGD95/01651 STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA (xi) SEQUENC. DESCRIPTION: SEQ ID NO: 19: AATACAGAGT TTTGTTCTCT ACTCTTATCC TGCTTI±CTCC TCCCTCCTAC TTTTCCCTGA CACCTATCTT GTTGTGAAOA CAGGAATTG C ATT? GATAAA ATCAAATCTT TTTTATTTTT TTTTGAGATG GAATCTTOCT CTGTTTCCAG GCTGOAGTGC TO INFORMATION FOR SEQ ID NO: SEQUENCE CHARACTERISTICS: LENGTH: 472 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA SEQUENCE DESCRIPTION: SEQ ID NO: ATCTGGTCAG CAGTGAAGCT CAGTGTACAC ATTCATTCCT T CACCAAGTGG TTATTGAGGA TATGCTGTTT GCTAGGTACT A TATTTAGAGA TGGGGTCTCA CAATGTTGCC CAGTCTACAG G

CCTTCACTG

CTTTACTTA

ACAGTGGCT

GTGAG CACAO CACACTACAG CCTCAAACTC CTGA TCTCGAGTAG CTGGGACTAC AGGGATGTGC CACC TGCTACTTGA AGGATGAAGA TAGGAGGAGA CACT TTTTTTTTTT TTGACAGAGT TTTGCTCTGT TGCC ACCTCCAGGT CAAGCAATTC TCCTGCTCAG CCTC INFORMATION FOR SEQ ID NO: 21: SEQUENCE CHARACTERISTICS: LENGTH: 2119 base pairs TYPE: nucleic acid STRANDEDNESS: double TOPOLOGY: linear .GTTCAA OAGATCCTCC

CTTOATTTGT

TTTATTTGTT

ATTCACAGGT

TGCCTCAGTC

CTACTTTAGC

TCTTTTTTTT

TOCAACCTCC.

00 120 240 300 360 420 472

ACACAT

C.TTATT

AGO CTG

COAOTA

GGCTTAGGCT

TTATTTGATT

OAOTGCTCAC

OTCOOACCAA

SUBSTITUTE SHEET (RULE 26)t WO 96/02640 W096/2640PCr1GB95101651 (il) MOLECULE TYPE: CONA (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 21:

:GRATTCCCP

i rGGCTCCP

CATCCCTTCI

GGTTATCCAT

CATOTAGGAG

GATAAATGTT

TTGGGAGGGC

GCCAACATGC

CGTGCCTGTA

GCAGAGGTTG

ACTTGGTCAA

GAGATGAGGA

TGACATGAAG

TACTTGCTCA

TCCAATCTTT

CAAGTTATTG

CGGTAGGACT

TTTTTTTTTT

GTGCAGTGGA

TO CCTCAGC C

TTTTGTAG?.G

TGATCC?.CCA

GCCTAGATGC

GATCCAAGTT

TTCCTOTCOT GTACCCTTGC AGTGCCTCTO GGTGGAATOC GGPAGAATGG

CTTCTGTTGT

OOCACACCAA

GGTACAGAAG

ATA.ACGATAT

CCCTGAAACA

CGAGGCCAOG

AGA.AACTCCG

GTCCCAGCTA

CAGTGAGCCA

AAAAAAAA

CAAAOAAGAC

CTTCATGCCA

GGGGCACTGA

CCACTGGCTT

GTCTTAGATT

GGTTAGGGT

TTTTTTTTA

GTGAACATGG

CCTCAAGTAG

ATGGGATTTT

GACTCGGCCT C TTTCATACAG G ATATCGGATT G

GTTTCTGAAC

GATCCTCCAC

ACAGGATTTA

GGAAGACCTT

TAAGAAACAG

CAGGCAAATT

TCTCTACTAA

CTCGGGAGGC

ATGTATCTCT

ATTCCCTTCA

ACTO AG AGG0A

CAGTATGTAA

CO CAGCCGCT

GCCTGAGCTC

AAA TA CATA.A

TOAGOCAGGA

AGATCGCOCC ACTGCATTCC AAAAkAAAAAA GAkATCGGTOG O CAAATTAAA

CCTTATAGAG

GOTCCTTCCC

TATGTAA;TGT

GGGGAATCTG

PGATAGGGTC

CTCACTGCAG

CrGGGACTAC kCCATGTTGC

CCAAAATGC

;CTTTTCAAT

TTGTAGTCT

AAAAGGAAGA

CATGAAGGAG

GO AG CTATTC

ATTCCAGACA

ACTTTCTGTG

CT CAATG0C CA

ATAATAGCTC

TCACTTTOTC

CCTCGACCTC

AGOCACATGT

CCAGGCTGOT

CGGGATTACA

TATGCATTTT

A.CGTTCCCAT

TGCTATCAGA

CTCATGCCAC

CTTTTCCCTC;

GTCTTAAATA

CCTGTCTGTA

AGkAG TTTGA ATTAA CCCGOG

GAATCACTTG

AGCCTGGGCA

AGA.AGAAGAA

CTAAGAGCTA

A.GAACTAGTA

TAAGCTTGTT

GCCAACTCTG

GTGTAGTATT T STAGOATAGC TI rcCCAGGATG G TOTGCTCAA G :ACC.;TOCCC A TCOAGCTCC T OTOTGAGCC A

ACTTTCTGCT

TTCATATACT

ACTCTGAATA

GATTOOTTOG

ATCTAGCACT

GACCAGCCTG

CATGGTAACA

AG CCTGGGAO

ACAGAGTGAG

ATC AGO TTTA

CTTOTCACCA

1'C CT C.ZACT C

:AGCCTTAAG

G.GTTGTCTA

~GGTTATTTA

'AG ATTCTTT

ATGGATGGA

TGTTCCTCC

GCTnA'TTTT 000 CTCAAG

CTGTGCCTG

120 180 240 300 360 420 480 540 600 6 6 U 720 780 840 900 960 1020 1080 1140 1200 1260 1320 ,1380 1440 CCTTAAGTAO GOAAGTCTTAA ATTCTATTCC TATTTCTGAO SUBSTITUTE SHEET (RULE 26) WO 96/02640 PcTIGB95/01651

CCTTCACTCA

TCCTTGGACA

CTCCACATTC

ATTTTCTGAT

CAGACACACT

GGGAAGCAGG

GATACTGGGA

CGTTTCTGTC

CCAGAG CCC C

TTCAGGTTCT

TGAGCTACCA

AGTGCCAGCT

TGCATGTCTC

TCCCTCACTT

GCTTCCCAGA

ATCA;TGGTTT

GAGGGA. ;G TTGG CATGAC TG AGGCCTTT C

TGGGGGTAGC

T-ATTA.-AGAA-

CTGATCCTGG

TAGGGGGGAA

CTTATATTGC

ACCAACCCTA

GAACTCZTTTC

:LAAAGTAGGG

CAGTCTCTCT

TGCATU.-ATA

CAAAITG AC

CTAATTCTCC

GACTGTCCCA

GO CGGPAC CT CCCCATC CCC

CAAGAAAC.AA

TCCCCAGACA

TGAATTATGG

TCATTCTCTZ'

TAAATPLAATG

TTCTTTCTCC

CCCTTCTTAC

TCTGATGACA

CCGTP.TAGPA

TTCTTTGTTC

AGOG CTAAAC

ACAATACCTC

PL;GG AAGCTO C'CTAGATGTe

AAACTOACTC

CTCTOTTCCT

AT0GCCTCAT TACAPC CCC CT

CCTTTATTGT

CTCAAGTAAC

.kAAGCCAA.AT

CTATGGACTA

GCAGGCTCAG

GG CTTGGTA ATG CTTCCAT CC TO C CATC 1500 1560 1620 1680.

1740 1800 1860 1920 1980 2040 2100 2i19

CT'CTTCCC'TC

GACATAGATC

CCZLTCCTGC

CT !.AGGAG CZAGCC1AATC ACAA.aTC7G !GTG'.CCC! GCTGCCGAA.- SUBSTITUTE SHEET (RULE 26) WO 96/02640 PCT/GB95/01651 54 INDICATIONS RELATING TO A DEPOSITED MICROORGANISM (PCT Rule l3bis) A. The indications made below relate to the microorganism referr ed to in the description on page 9 lIine 13-15 B. IDENTIFICATION OF DEPOSIT Further deposits are identified on an additional sheet E Name of depositary institu~tion NATIONAL COLLECTIONS OF INDUSTRIAL AND MARINE BACTERIA LTD Address of depositary institution (inc~luding postal code and country) 23 ST MAClIAR DRIVE ABERDEEN AB2 iRY

SCOTLAND

-UNITED KINGDOM C. ADDITONAL INDICATIONS (leave blank if not applicable) This information is continued on an additional sheet E ESCHERICHIA COLI D115 OLF1 (K12) CONTAINING PLASMID pJIA-2 D. DESIGNATED STATES FOR WHICH INDICATIONS ARE MADE (if the iddoarc nfor all dignaed Satr) ALL STATES E. SMIA3-ATE FURNISHING OF INDICATIONS (leave blank if not appflcable) The indications listed below will be submitted to the InternationalI Bureau later (spcci)y the gencral nature of Iheindications 'Acccsion Number ofDeposi:) ACCESSION NO 40709 DATE OF DEPOSIT 15 FEBRUARY 1995 CONSTRUCT DEPOSITED pHA-2 For receiving Office use only For International Bureau use only Ibis sheet was received with Uh. international application El This sheet was received by the International Bureau on: ?1 SEPTEMBER 1995 Authorized officer D. J. MACKERNESS (MRSI Form PCr/RO/134 (July 1992) IAuthorized officer

Claims (9)

1. A recombinant DNA construct comprising, in the 5' to 3' direction and operatively linked: the 5' flanking sequence of SEQ ID NO 21; a DNA sequence encoding a-lactalbumin or a functional equivalent thereof; and the 3' flanking sequence of any one of SEQ ID NO'S 16 to the DNA construct being adapted to express the a-lactalbumin or functional equivalent thereof in a non-human transgenic animal.

2. A recombinant DNA construct comprising, in the 5' to 3' direction and Soperatively linked: at least about 1.8 kb of 5'-flanking sequence from the human a- 20 lactalbumin gene including the a-lacalbumin promoter; a DNA sequence encoding a-lactalbumin or a functional equivalent thereof; 25 at least about 3 kb of 3'-flanking sequence from the human a- lactalbumin gene the DNA construct being adapted to express the a-lactalbumin or functional equivalent thereof in a non-human transgenic animal. 4 i 0 56

3. A recombinant DNA construct as claimed in claim 1 or 2 comprising about 3,7 kb of 5' flanking sequence and about 13 kb of 3' flanking sequence. 4, A recombinant DNA construct as claimed in claim 1, 2 or 3, wherein the a- lactalbumin is human a-lactalbumin. A recombinant DNA construct as claimed in any one of claims 1 to 4, wherein the functional equivalent of a-lactalbumin is an a-lactalbumin nutein incorporating amino acid substitutions as compared to wild type a- lactalbum in.

6. The recombinant DNA construct pHA-I as described herein.

7. The recombinant DNA construct pHA-2 as described herein.

8. A vector containing a recombinant DNA construct as claimed in any one of S. claims 1 to 7. 9 20 9. A host cell containing a vector as claimed in claim 8. 9 A non-human transgenic animal having a recombinant DNA construct as claimed in any one of claims 1 to.7 integrated into its genome. 9 25 11. A non-human transgenic animal as claimed in claim 10 which is capable of 9 transmitting the construct to its progeny. 9

12. A non-human transgenic animal as claimed in claim 10 or 11 which is a lactating female. 57

13. Transgcnic cattlc as clainicd fil claim' 10, 11 or 12.

14. A nictliod of producing milk having an enhiancced contcnt of c-lactalbunlin, thc mcthod comprising extractinig milk from a lactating f1cmalce ton-1humnan transgcilic animial as claiti~cd in claim 12. Milk produl~ccdl by a incthod as claimcd in claim 14. 16, ot-factal1unin cxtractcd from milk as clainmcd in claim Lee,.