Animal Health and Production for the 21st Century

This book places Australia's efforts in selected areas of animal biotechnology in the context of research developments occurring internationally. The topics chosen for discussion are relevant for the development of animal production industries into the next century. The book is divided into three sections: genetic technologies; immunological-based technologies; and intensive animal production.

It is intended for animal geneticists, agriculturalists and students.

• Language: English
• Publisher: CSIRO PUBLISHING
• Release date: Mar 1, 1993
• ISBN: 9780643105591

Book preview

Genetic Technologies

---

Transgenics

---

M.R. Brandon

University of Melbourne, Parkville, Victoria, Australia

In assessing the current status of transgenic farm animals it was apparent that Australia is in the forefront of research on transgenic pigs and sheep for the purpose of manipulating growth and introducing new metabolic pathways into sheep to enhance wool production.

Dr K Ward presented data on novel schemes to enhance wool production by the integration of two metabolic pathways into sheep, which together will increase the availability of cysteine and glucose to the wool follicle. Dr Ward has isolated 2 genes from E. coli that facilitate the conversion of serine to cysteine in the presence of acetyl-Co A and a sulphur source which can include hydrogen sulphide. The second biosynthetic pathway that was discussed and intended for introduction into the sheep germline is the glyoxylate cycle, designed to increase the availability of glucose to the follicle.

One area of research in transgenic farm animals that Australia has not achieved recognition in is the production of human pharmaceuticals by the use of transgenic animals. Dr J Simons from Edinburgh presented studies on the direct expression of human proteins in the mammary glands of transgenic sheep and their secretion in milk. Sequences from the sheep β-lactoglobulin gene, which function correctly in transgenic mice, have been used to construct fusion genes to induce expression of human factor IX and α-1-antitrypsin in the mammary gland. Human factor IX and α-1-antitrypsin genes have been used in a variety of fusion genes and very high levels of human α-1-antitrypsin have been achieved in mouse milk, while only a low amount has been produced in sheep milk. Strategies were outlined on how to increase the amounts of these valuable human pharmaceuticals produced by transgenic sheep.

Strategies and experimental approaches available for the establishment of transgenic animals were outlined. It is clear that while embryo microinjection is most widely used to transfer DNA into the germline, alternate methods including stem cell delivery and retroviral infection offer simple and efficient procedures for establishment of transgenic animals. The development of efficient and simple methods of transgene integration is of paramount importance when considering domestic animals, given their long gestation and sexual maturation times.

The symposium showed Australia was at the forefront of gene transfer in farm animals, however it was clearly identified by symposium participants that an increased international effort is needed to produce stable embryonic stem cell lines from domestic livestock to facilitate the production of transgenic animals.

---

1

Progress Towards the Introduction of New Metabolic Pathways to Sheep

---

K.A. Ward, C.R. Byrne, C.M. Shanahan, C.D. Nancarrow, J.D. Murray, Z. Leish, C. Townrow, N.W. Rigby, B.W. Wilson and C.L.H. Hunt.

CSIRO Division of Animal Production, PO Box 239, Blacktown, NSW 2148, Australia

It has long been the desire of those involved in animal breeding to be able to make precise changes to the genetic information of domestic animals without at the same time having to accept the major gene reassortment that inevitably takes place during each cycle of sexual reproduction. The recently developed techniques of genetic engineering show great promise as a way of achieving such a goal, because they provide a mechanism for the introduction of single, well-characterised gene sequences into animals. The technique most frequently used consists of injecting a small amount of recombinant DNA into the pronucleus of a single-cell embryo, where it becomes integrated into the embryo genome.

The newly incorporated DNA can contribute fully to the developmental program of the young embryo, and thus becomes an integral part of the genome of the animal that subsequently develops. The expression of the new gene in the adult is subject to numerous factors, such as the nature of the regulatory sequences included in the recombinant DNA gene construction, the position of integration of the DNA in the new genome and the compatibility of the host’s transcription factors. Nevertheless, there are now abundant examples of the expression and subsequent phenotypic modification of animals containing genes introduced by this transgenic animal technology.

The technique was pioneered in laboratory mice (Brinster et al. 1982; Costantini and Lacy, 1981; Stewart et al. 1982; Gordon et al. 1982; Gordon, 1981) and has achieved wide usage as a major research tool for the study of gene expression in this species (Brinster et al. 1985; Palmiter and Brinster, 1986). Following the initial successes in mice, the technology was applied to domestic animals (Hammer et al. 1985), where it was found possible to adapt the mouse procedures, albeit with somewhat lower efficiencies. This suggested that some modifications to the procedure would be required for optimum performance in other species, and as other laboratories began to repeat the techniques, it became apparent that each species presented its own set of unique problems (Pursel et al. 1989b; Murray et al. 1989; Wagner et al. 1984; Hammer et al. 1986; Ward et al. 1984a; Rexroad and Pursel, 1988; Simons et al. 1988). In the five years that have followed from the initial domestic animal successes, innovative solutions have been devised which today provide protocols for all the major domestic species, although the efficiency of integration of the foreign DNA in general remains less than that achievable in the mouse.

Sheep have proved to be one of the most difficult species to adapt to the pronuclear microinjection technique. The first transgenic sheep that was made contained a highly re-arranged piece of DNA, and the efficiency reported was 1.3%. Other laboratories also reported low efficiencies of integration (Ward et al. 1986; Simons et al. 1988), and it became clear that the sheep either did not integrate DNA at the same efficiency as other species, or that the species was particularly sensitive to damage during the microinjection procedure. The sheep pronuclei are frequently hidden by dense cytoplasmic granules, a problem which has also been encountered in other species such as pigs and cattle. Whereas in pigs the granules can be effectively removed from the vicinity of the pronuclei by centrifugation, however, in the case of sheep this is not very effective. The most successful approach for sheep has been to carefully synchronise the developmental stage of the embryos by controlling the ovulation of the donors, since there appears to be a narrow window during the embryo’s development where the pronuclei are more easily observable. It is also important with sheep to ensure that the optics used to visualise the pronuclei are optimised (Murray et al. 1989; Hammer et al. 1986; Rexroad et al. 1989; Simons et al. 1988). The success rates thus obtainable vary between 4–14%. This rate of success still leaves considerable room for improvement compared with some other domestic animals such as pigs, which have a success rate which varies between 10% (Pursel et al. 1989) and 35% (Vize et al. 1988). One possible route for improvement stems from a recent report that the culture of post-injected sheep embryos in a synthetic culture medium containing human serum results in a substantial increase in efficiency of production of transgenic animals (Dr. R. Seamark, unpublished).

The low efficiency associated with transgenic sheep production is, nevertheless, adequate for the experimental application of the technology. The general aim of domestic animal genetic engineering is to alter the efficiency of a selected production trait by the introduction of new or altered genetic material to the target animal. In order for this approach to be successful, two criteria apply. First, the physiology and biochemistry of the selected trait must be well understood. Second, a gene sequence must be available which can modify the trait. Because a substantial knowledge gap exists between the observed or measurable physiology of an animal and the explanation of this physiology at the molecular level, there are few production characteristics that can satisfy these criteria.

One possible target for genetic manipulation, however, is the process of wool growth, the physiology of which has been the subject of considerable research. As a result, the nutritional requirements for the process are at least partially understood. It is apparent from the work of Reis and co-workers (Reis, 1979; Reis and Schinkel, 1963; Reis et al. 1973; Broad et al. 1970) that one of the limiting factors is the supply of the amino acid cysteine to the wool follicle. In addition, much of the energy supply to the wool follicle is provided by glucose (Chapman and Ward, 1979; Fraser et al. 1972), a substrate which is less readily available in ruminant animals than in monogastrics. These requirements have provided the stimulus for our research which is directed towards the integration of two metabolic pathways into sheep, which together are predicted to increase the availability of cysteine and glucose to the wool follicle. The two pathways involved are the cysteine biosynthetic pathway and the glyoxylate cycle, as shown in Figure 1. Each pathway requires the presence of two new enzymes, the coding sequences for which have been lost to the mammalian genomes during evolution. Our broad approach is to utilise the functional coding sequences from the genome of the bacteria Eschericia coli. By introducing the enzymes one at a time into transgenic animals, it will be possible to follow the metabolic effects of the genetic alterations, some of which may be unpredictable since they involve important aspects of mammalian Intermediary metabolism.

The purpose of the cysteine biosynthetic pathway is to increase the availability of cysteine to the follicle. The high cysteine content of the wool keratin proteins (Fraser et al. 1972) is consistent with findings that its supply plays an important role in fibre growth. When the amino acid is infused either abomasally or intravenously into sheep, thus increasing its availability to the follicle, wool growth is increased by 30% to 60% (Reis et al. 1973; Reis, 1967). This consists of an increase both in length growth rate and fibre diameter. Thus it follows that wool growth can be improved if a way can be found to increase the amount of cysteine that is absorbed from the diet of sheep. However, the ruminant digestion of the species makes this difficult, because if an excess of cysteine is simply added to the animal’s diet, it is degraded by the rumen bacteria. A number of different approaches have been taken over the years all aimed at increasing the amount of circulating cysteine in sheep by preventing its ruminal degradation. Thus, protein supplements have been treated with formaldehyde, methionine and methionine analogues have been encapsulated in rumen-resistant matrices, or proteins that are known to be resistant to degradation have been included in the diet (Ferguson, 1975; Kempton, 1979). While all of these have shown some effect on wool growth, none have been cost-effective in the normal farm grazing situation in Australia.

As shown in Figure 1, a new approach that has been made possible by the advent of the new genetic engineering techniques is to provide the sheep with the necessary biochemical pathway to synthesise cysteine from H2S. The pathway is catalysed by two enzymes, serine transacetylase (SAT) (EC.2.3.1.30) and O-acetylserine sulfhydrylase (OAS) (EC.4.2.99.8), as shown in Figure 1. Sheep do not possess the genetic information for either of these enzymes, and therefore can only synthesise cysteine from other sulphur-containing amino acids such as methionine, itself an essential amino acid. We have previously proposed (Ward et al. 1984b) to isolate the appropriate genetic information from the bacterium Eschericia coli, place it under eukaryotic promoter control and transfer it to sheep by pronuclear microinjection. Our aim is to construct two genes which will be transcribed and translated in the ruminal and intestinal epithelia of the sheep. We predict that the presence of SAT and OAS in these tissues will result in the biosynthesis of cysteine, utilising as substrates the endogenous serine present as a non-essential amino acid, and H2S which is produced in the rumen and intestine.

The second biosynthetic pathway we intend introducing is the glyoxylate cycle, designed to increase the availability of glucose to the follicle. Sheep, like all ruminants, have no dietary supply of glucose because it is metabolised by microbes in the rumen. Glucose is subsequently synthesised from gluconeogenic precursors such as volatile fatty acids and amino acids. The volatile fatty acids, produced in the rumen, are a complex mixture, but one of the major components is acetate, which is not able to give rise to glucose. This results from the fact that acetate must enter the tricarboxylic acid cycle, where two decarboxylation steps ensure that no net carbon utilisation can take place. While acetate can therefore be a major source of energy supply for those tissues which have a highly aerobic metabolism, it is largely lost to the glucose-demanding wool follicle. While amino acids such as cysteine limit wool production, circulating glucose appears to be adequate for follicle needs, but as this limitation is removed by, for example, the efficient operation of the cysteine biosynthetic pathway in sheep, energy supply is likely to become the rate-limiting component. In addition, a better supply of glucose from volatile acids would have a sparing effect on the overall supply of amino acids, which may increase the general metabolic efficiency of the species.

Figure 1 Diagram of the glyoxylate cycle and the cysteine biosynthetic pathway integrated in the sheep to utilize the acetate and H2S produced by the rumen and intestine.

The glyoxylate cycle provides a metabolic shunt which prevents the loss of acetate carbon in the decarboxylation steps of the tricarboxylic acid cycle. In E. coli, it consists of two genes, aceA, encoding the enzyme isocitrate lyase (EC.4.1.3.1), and aceB, encoding malate synthase (EC.4.1.3.2). The operation of these genes to produce succinate and malate is shown in Figure 1. Our aim with these genes is two-fold. Initially, we wish to express the genes in tissues such as liver, kidney and intestine, in order to produce more glucose for general circulation. We predict that this will give rise to an overall increase in general metabolic efficiency. In addition, we intend to place the genes under the direct control of a promoter which will direct expression specifically in the wool follicle, in order to increase glucose in the follicle itself.

Manipulation of the Bacterial Genes for Eukaryotic Expression

To achieve our goal of introducing the two metabolic pathways to sheep, our initial strategy is to utilise a metallothionein promoter to regulate transcription of the gene, and the sheep growth hormone gene to provide sequences 3’ to the coding sequence which are necessary for message stability and translation. We are utilising this particular combination in order to achieve expression of the pathways in tissues such as the intestinal tract, liver and kidney. We have accumulated a substantial body of information about the expression in transgenic animals of the sheep MT-Ia/sheep GH gene MTSGH9. This gene is expressed in the intestine, liver and kidney of transgenic mice when 25 mM zinc sulphate is included in their drinking water (Shanahan et al. 1989) and the expression is dependent on the presence of the added zinc. The gene is also expressed in transgenic sheep, where some expression is obtained even in the absence of added dietary zinc, and the addition of zinc further stimulates expression ((Ward et al. 1988). Attempts to reduce or eliminate the constitutive expression of this gene in sheep are currently the subject of a major research effort within our laboratory. Nevertheless, while the promoter\gene combination is not yet completely satisfactory for our purposes, it possesses many of the features which are desirable in the regulation of the new metabolic pathways.

The modification of the various bacterial genes was similar in principle for each gene.

cysE and cysK

The cysE gene was isolated from a lambda transducing phage (Ingle and Loughlin, 1980). An EcoR1-PvuII fragment was cloned into the plasmid pBR322, giving rise to plasmid pCLH1. The approximate position of the gene was determined by assaying the ability of different portions of the pCLH1 insert to support growth of cysE- -mutants of E. coli on media lacking in cysteine. The gene was then sequenced and compared with the published sequence (Denk and Bock, 1987), with which full agreement was obtained. In preparation for subsequent manipulations, two changes were made to the sequence by means of site-directed mutagenesis. As shown in Figure 2a, upstream from the ATG initiation codon, a BamH1 site was introduced to provide a site for splicing a eukaryotic promoter, and base -3 was altered from a T to an A to provide for increased translational efficiency in eukaryotic cells (Kozak, 1986). The sequence 3’ to the gene was modified to remove a Sal1 site from the region of the stop codon. The second modification was made in order to facilitate splicing the bacterial gene into the sheep growth hormone gene.

The promoter sequence isolated from the sheep metallothionein-Ia (MT-Ia) gene (Peterson and Mercer, 1986) was spliced to the cysE coding sequence by insertion of a 860 base pair EcoR1-BamH1 fragment of plasmid pMT010 (Dr. J. Mercer, University of Melbourne) into the modified cysE gene. This modification results in a gene whose transcription is predicted to start at the transcription start site of the MT-Ia promoter, and to contain within the 5’ untranslated portion of the transcript, 24 bases of MT-Ia sequence and 816 bases of the bacterial cysE gene.

The sheep growth hormone (GH) gene was fused 3’ to the stop codon of the cysE gene to provide appropriate polyadenylation signals and an intron-splicing structure which has been postulated as necessary for good expression of cDNA’s and bacterial sequences in eukaryotic cells (Palmiter et al. 1987). The GH gene was introduced by a series of standard cloning steps, resulting in the plasmid pMTCE11. During the cloning manipulations, the GH ATG initiation codon was removed, thus reducing the likelihood of a growth hormone translation product being formed from the gene transcript. The predicted RNA transcript from this fusion gene is about 2 kb, consisting of 24 bp of MT-Ia gene, 816 bp of cysE gene and about 1.1 bp of GH gene, including the poly A tract. The only translation product predicted is the bacterial SAT enzyme.

The cysK gene was isolated from a plasmid clone pAB101 (Boronat et al. 1984) (generously donated by Dr. Jones-Mortimer, Cambridge), and the cysK gene identified by sequencing (Byrne et al. 1988a). It was then modified upstream from the ATG in the same way as the cysE gene, but no changes were necessary to the sequence downstream from the stop codon. The same EcoR1-BamH1 DNA fragment of the MT-Ia gene was fused 5’ to the cysK coding sequence, and the sheep GH gene was joined to the 3’ portion of the gene, giving rise to plasmid pMTCK11 (Figure 2b). The predicted transcript from this fusion gene is about 2 kb, similar in size to that of the MTCE11 fusion gene. The predicted translation product is the bacterial OAS enzyme.

Figure 2 The plasmids pMTCE11 and pMTCK11 were constructed by mutating the bacterial cysE and cysK genes immediately upstream from the ATG Initiation codon, and the sheep MT-Ia promoter sequence was Inserted as shown. The sheep growth hormone gene was joined to the 3’ end of the bacterial coding sequence immediately downstream from the stop codon. The vectors used for the initial cloning are shown in brackets. MT = sheep metallothionein-Ia promoter. GH = sheep growth hormone gene: the direction of transcription of the bacterial genes is shown by arrows.

Figure 3 The plasmids pMTaceA1 and pMTaceB1 were constructed by mutating the bacterial aceA and aceB genes immediately upstream from the ATG initiation codon, and inserting the sheep MT-Ia and GH gene sequences essentially as described in Figure 2. In this case, however, 41 bp of the first exon of the sheep GH gene was placed between the MT-Ia promoter and the bacterial coding sequence in order to extend the amount of 5’-untranslated DNA sequence in the fusion gene.

aceA and aceB

The glyoxylate operon was cloned from the E. coli chromosome and the aceA, aceB and portion of the aceK genes sequenced (Byrne et al. 1988b). Both genes were then modified by procedures essentially similar to those used for the cys genes, utilising the EcoR1-BamH1 MT-Ia promoter fragment and the sheep GH gene. In this case, however, 41 bp of the first exon of the sheep growth hormone gene were placed between the MT-Ia promoter and the bacterial coding sequences. This was done to increase the length of the 5’-untranslated sequence of the fusion genes. The resulting plasmids were named pMTaceA1 and pMTaceB1 (Figure 3a,b). The sizes of the predicted transcripts are 2.4 kb for pMTaceA1 and 2.7 for pMTaceB1, based on a 1.1 kb GH gene transcript and 1.3 kb (aceA) or 1.6 kb (aceB) bacterial gene transcripts.

Transcription and Translation of the Fusion Genes in Cell Culture

The four plasmids, pMTCE11, pMTCK11, pMTaceA1 and pMTaceB1 were used to transfect mouse L-cells by the calcium phosphate method (Wigler et al. 1979), using neomycin resistance encoded on the co-transfected plasmid pSV2neo (Southern and Berg, 1982) as a selectable marker for transfected cells. Transcription of the genes was measured by Northern blot analysis both in transient and stably transformed cells. In order to determine whether the introduced genes retained heavy metal inducibility, the cultures were grown in the presence and absence of 40 uM zinc. The results obtained for pMTCK11 are shown in Figure 4, and for pMTaceA1 and pMtaceB1 in Figure 5. All four genes are transcribed in mouse L-cells, giving transcripts of the sizes predicted above. In addition, pMTCE11 and pMTCK11 each gave a smaller transcript of 1.6 kb. It is not clear what this corresponds to, but it may reflect alternate splicing of the primary RNA transcript. When RNA was extracted from cells grown in the presence of zinc, a significant increase was found in the intensity of the detected bands. This is consistent with the MTIa promoter retaining its heavy metal inducibility in the fusion genes.

Figure 4 Northern blot analysis of total RNA extracted from mouse L-cells transformed with plasmid pMTCK11. The probe used to detect specific transcripts consisted of an anti-sense RNA to the cysK coding sequence. RNA from E. coli strain C600 is also shown as a control. Zinc induction of cells is indicated by (+ Zn).

---

TABLE 1 Specific activity of serine transacetylase and O-acetylserine sulfhydrylase in transformed cultures of mouse L-cells

Cells were harvested, homogenised in extraction buffer (50 mM Tris-CI, pH 7.5) and centrifuged for 20 min. at 10 000 × g. The supernatant solution was used for the assay of SAT or OAS activity. All values have been corrected for low non-specific activity in cells transformed with vector alone. E. coli activity was measured in strain AL509 (SAT) and strain C600 (OAS).

TABLE 2  Specific activity of isocitrate lyase and malate synthase in transformed mouse L-cells

The mRNA transcripts detected in cell culture are translated into active enzyme. Table 1 shows the measured levels of the cysteine biosynthetic enzymes SAT and OAS. Thus, SAT activity of 225 nmoles O-acetylserine formed / mg protein / 30 min was measured in cells transformed with the pMTCE11 plasmid, and OAS activity of 1500 nmoles cysteine formed / mg protein / 30 min was measured in cells transformed with the plasmid pMTCK11. The activity of both enzymes was increased in the cell cultures when they were grown in the presence of zinc, which further supports the zinc inducibility of the fusion genes. Consistent with our previous findings with the MT-Ia promoter in cell culture (Ward et al. 1988; Shanahan et al. 1989), however, substantial basal activity is also detected in the absence of added zinc in the culture medium.

Table 2 shows the measured levels of the two glyoxylate cycle enzymes. Isocitrate lyase activity of 3.9 nmoles of isocitrate utilised/min/mg protein was measured in cells transformed with pMTaceA1, and malate synthase activity of 0.28 nCi14C-acetyl CoA Incorporated/20 min/mg protein detected in cells transformed with pMTaceB1.

Cell lines containing complete pathways (cysteine or glyoxylate) were established by introducing the appropriate plasmids as a DNA mixture during the cell transformation. The cell line containing the cysteine pathway produced RNA and enzyme activities similar to those of cell lines containing only one of the genes (Table 1). In this cell line, cysteine biosynthesis from serine and H2S was demonstrated by incubating radioactively-labelled serine with cell extracts and observing the appearance of radioactive cysteine. This was detected, but the amount produced was much less than that predicted from the levels of the two individual enzymes. Similarly, the cell line containing the glyoxylate cycle produced activities similar to those containing only a single fusion gene (Table 2). Despite the presence of the two enzymes in this cell line, however, we have not been able to demonstrate in cell extracts the production of malate from isocitrate, which suggests that at least one of the enzyme levels is too low for significant operation of the pathway.

Insertion of the Cysteine and Glyoxylate Cycle Genes into Mice and Sheep

The inserts from plasmids pMTCE11, pMTCK11, pMTAceA1 and pMTaceB1 have been inserted into transgenic mice by pronuclear microinjection. Two mice contain the cysE fusion gene, 4 mice contain the cysK fusion gene, 3 mice contain the aceA fusion gene and 2 mice contain the aceB fusion gene. Southern blot analysis of three of the cys-gene containing mice is shown in Figure 6a. Digestion of the transgenic DNA with the enzyme Bam H1 produces a band of 4 kb, identical in size with that of the DNA originally microinjected into the mouse embryos. This is to be expected because the integration of foreign DNA into transgenic mice normally occurs as a tandem array of the foreign gene, the size of the array varying between animals, and, since there exists a single recognition sequence for Bam H1 in the fusion gene, a fragment the size of the original DNA is excised from the array. As shown in Figure 6a, the number of copies integrated varies between the various mice, reflecting the different sizes of the arrays. Each of the fusion genes contains a single Bam H1 site situated between the MT-Ia promoter and the bacterial coding sequence (Figures 2 and 3), and therefore digestion with this enzyme would be expected to excise from the tandem array a band corresponding in size to that of the gene injected into the animals. Since each of the mice contain such a band, this indicates that most of the genes in the transgenic animals have retained their sequence integrity, and are organised largely head-to-tail.

Figure 5 Northern blot analysis of total RNA extracted from mouse L-cells transformed with plasmids pMTaceA1 and pMTaceB1, either singly or as a mixture of both plasmids. The probes used to detect specific transcripts consisted of DNA fragments specific for the coding sequences of either the aceA or aceB genes. L = untransformed cells. A = transformed with pMTaceA1. B = transformed with pMTaceB1. AB = transformed with a mixture of pMTaceA1 and pMTaceB1.

Figure 6 Southern blot analysis of (A) three transgenic mice and (B) one transgenic sheep. CK47 contains the fusion gene insert of plasmid pMTCK11, while CE12 and CE32 contain the insert from pMTCE11. Sheep CK47 contains the DNA from the insert of plasmid pMTCK12, which is essentially the same as pMTCK11, but with 41 bp of the first exon of the GH gene placed between the MT-Ia promoter and the bacterial coding sequence. DNA was extracted from tail tissue and digested with the restriction enzyme BamH1, or in the case of sheep DNA, with EcoRI (a) or BamH1 (b). The probes used consisted of DNA encoding the cysE or cysK. genes. A tandem array of the insert arranged head-to-tail produces a DNA fragment of size 4 kb for each gene. Head-to-head or tail-to-tail arrangements produce bands of larger sizes.

Expression of the various fusion genes in the transgenic progeny was examined both by Northern blot analysis of RNA extracted from a variety of tissues, and by enzyme analysis of tissue extracts. As shown in Figure 7, the only gene to give rise to detectable levels of expression in animals was MTAceA1. In the transgenic mouse line established with this gene, a low-level transcript was present in RNA extracted from liver, but not in testis or brain. The other ace-containing gene, MTaceB1 and the two cys-containing genes, MTCE11 and MTCK11 did not give rise to RNA transcripts in any of the animals examined.

Figure 7 Northern blot analysis of total RNA extracted from various tissues of a transgenic mouse containing the gene MTaceA1. The probe used was a DNA a DNA fragment specific for the coding sequence of the bacterial aceA gene. NT = untransformed mouse L-cells. AB, A = cells transformed with a mixture of pMTaceA1 and pMTaceB1, or with pMTaceA1 only. For mouse tissues, Te = testis, Br = brain, and Liv = liver.

One of the cys-encoding genes has been inserted into a transgenic sheep, in order to determine whether the poor expression of these genes in mice was also found in sheep. The gene used was pMTCK12, a slight variant of pMTCK11 such that 41 base-pairs of the GH first exon have been interposed between the MT-Ia promoter and the bacterial cysK coding sequence. This gene was used because cell culture data indicated that this construction may be slightly better expressed in mouse L-cells. Southern blot analysis of the DNA from the transgenic sheep is shown in Figure 6b. Digestion with EcoR1, which does not possess a recognition sequence within the fusion gene, gives rise to a fragment of about 6 kb in length, hybridising to a growth hormone probe with about the same intensity as that of the endogenous growth hormone gene. Digestion with BamH1, which has a single recognition site in the fusion gene, produces a fragment of about 5 kb, slightly larger than the size of the injected DNA. We interpret these results to show that in the sheep, a single copy of the gene has been integrated into the sheep genome. While the results are consistent with the gene not having undergone any re-arrangement, we have not yet rigorously excluded this possibility. Examination of the expression of this gene in the sheep has been limited to skin and liver, both of which were obtained as biopsy samples. However, no detectable expression has been observed in Northern blots of RNA extracted from these tissues, and it appears likely that the same poor expression observed in mice will also be manifest in this sheep.

Since the genes are not re-arranged in mice and probably not in the transgenic sheep, the conclusion is inescapable that by inserting a bacterial gene between the MT-Ia promoter and the sheep growth hormone gene, the high levels of zinc-inducible expression observed with the unaltered MTGH9 gene in transgenic animals cannot be obtained with the modified sequences.

Conclusions

The modification of domestic animal biochemistry to provide for better productivity is an attractive application of the newly-developed techniques of gene manipulation, and the approach described in this paper, using bacterial coding sequences to provide the genetic information necessary for the required additional enzymes, is potentially very powerful. It is well-documented that bacterial sequences can be expressed in mammalian cells in culture (Mulligan and Berg, 1980) and several have also been expressed in transgenic animals (Palmiter et al. 1987). Furthermore, of specific relevance to the work described in this paper, it has been reported recently that the cysE gene of Salmonella typhimurium has been expressed in the skin of transgenic mice (Sivaprasad et al. 1989) and transgenic sheep (Prof. G.E. Rogers, unpublished), utilising a murine leukaemia virus promoter to initiate transcription of the gene. However, expression of the S. typhimurium cysM gene (equivalent to the E. coli cysK gene) utilising this same promoter appears to be very low or absent. Nevertheless, this work is highly encouraging and suggests that the goal of a practical application may be appreciably closer.

The problem that faces those laboratories interested in applying the techniques to the farming community, however, lies in the fact that only a very limited choice of promoter sequences is currently available that are of potential practical value. In order for the genetic modifications to be at all meaningful in a practical farm application, the expression of the genes must be tissue-specific, and preferably under the regulation of a promoter which is controllable by a compound that is compatible with farm operation. It is of considerable academic interest but little practical value to obtain expression of the cysteine biosynthetic pathway, for example, in tissues in which the essential substrate H2S is not present. Our laboratory also views regulation of the genes by an inducible promoter as an important goal. We have previously inserted a modified sheep growth hormone gene into sheep in order to alter the concentration of circulating growth hormone in these animals. The gene used was well-regulated in transgenic mice, but when inserted into transgenic sheep, was expressed constitutively. The high growth hormone levels that resulted in these animals proved to be physiologically deleterious, causing metabolic disturbances which culminated in the animals’ deaths at about 10 to 12 months of age. While the operation of newly introduced biochemical pathways into animals may not always require the same level of regulation as that which is obviously necessary for growth hormone concentration manipulation, our objective is to provide the maximum level of regulation that we can achieve over the new pathways.

From the results presented in this paper, it is clear that obtaining such expression can prove difficult. Our choice of MT-Ia promoter and growth hormone gene was made both on the basis of our existing knowledge of the high expression of this combination in transgenic animals, and the findings of others that metallothionein-growth hormone genes provide a suitable background for supporting expression of bacterial genes in transgenic mice (Palmiter et al. 1987). It is apparent that our initial gene design is not appropriate for expression in transgenic animals. There are several differences between the genes that