Gene targeting in somatic cells: Since the pioneering work with sheep at Roslin cloning by nuclear transfer has now been exemplified in a number of other livestock species including cattle, goats and pigs. Although there are wide variations between experiments, efficiencies appear to be improving with 10% or better survival rates of transferred blastocyst being achieved in cattle. Nevertheless, much more research still needs to be done before the procedure is routine, especially given the high perinatal mortality and the possibility of longer term deleterious consequences. A variety of different somatic cell types have been used in cloning, including fetal and adult fibroblasts, ovarian cells, muscle cells and mammary epithelia. These donor cells are usually derived directly from fetal or adult tissues and then grown as primary cell cultures in the laboratory.

The genetic modification of somatic cells prior to nuclear transfer provides a new route for modifying and, indeed, targeting the germline. In this approach the cells are transfected, selected and tested for the genetic modification prior to NT. This method was used successfully to introduce new DNA sequences into the germline. Transgenic sheep carrying a human factor IX construct by transfecting and selecting sheep primary fibroblasts prior to NT. In qualitative terms this did not achieve anything more than pro-nuclear injection: the site of insertion was random and multiple copies of the transgene were integrated. Nevertheless, it showed that primary cell lines which had been transfected, selected and expanded in culture were still viable for the NT procedure. Moreover, the fact that the modified cells are selected prior to NT means that all the animals born are transgenic (as compared to 5~1O% by conventional transgenesis) and their gender is entirely predictable. Therefore, fewer animals are used overall, even with the relatively low efficiencies of the cloning procedure.

Most NT experiments in livestock have used primary somatic cells as the nuclear donors. These cells differ, not only from mouse ES cells, but also amongst themselves, in a variety of crucial properties including the response to selection, targeting frequency and growth conditions, all of which need to be optimised to achieve targeting.  Most gene targeting strategies are based on recombination between the transfected DNA construct and the homologous sequences in the genome. This involves the pairing of these sequences followed by DNA strand breakage and strand exchange between the targeting vector and the chromosomal sequences. Targeting vectors usually comprise two regions complementary to the chromosomal sequences, separated by a region of non-homology, which is often a selectable marker.

A popular choice of cell type for these types of experiment is fetal fibroblasts. In sheep and cattle these proliferate rapidly in vitro and can be readily transfected by electroporation or lipofection. Nevertheless, the frequency of targeting in such somatic cells is lower than in ES cells and enrichment may be required to detect and select targeting events. Many targeting vector designs are available but the so-called 'promoterless strategy' is likely to give the highest and most reliable enrichment.

Using this strategy the selectable marker should only be active when integration occurs downstream of the promoter and in-frame with an actively transcribed gene. For the strategy to work the target gene must be expressed and this requirement poses. A serious limitation if the gene is not expressed in the cell type being targeted. We now know that a range of cell types can be used for NT (mammary epithelial cells, muscle cells as well as fibroblasts) and so it-may be appropriate to choose a donor cell type that expresses the target gene.

Targeted cells for NT experiments are derived from a single selected cell, which is subsequently clonally expanded to provide cells for NT and sufficient DNA to confirm that the targeting has been successful. Given the relatively low efficiencies of targeting currently available, even with powerful targeting strategies, a relatively large number of candidate clones must be screened. This may necessitate the development of high- throughput screening protocols, for example, using a 96-well plate format whereby many hundreds of potential cell clones can be screened. For some genes it may be possible to enrich for targeted clones using immunological and flow sorting-based screens. Once potential candidates have been isolated, these need to be screened by DNA analyses which enable the targeted locus to be identified. In a somatic cell there are two copies of each autosomal gene, only one of which will be targeted. Screening involves developing DNA analyses such a Southern blotting or a polymerase chain reaction (PCR) which discriminates between the targeted and the non-targeted alleles.

Many of the approaches to somatic gene targeting have been derived from those developed using mouse ES cells. These are pluripotent stem cells which can divide many times in culture without losing their ability to contribute to the germline. This means that genetic changes can be introduced and the cells selected and expanded without the cells senescing. By contrast, the cells used for NT experiments in livestock have a finite lifespan and this has major consequences with regard to carrying out gene targeting experiments. For example, bulk populations of primary sheep fetal fibroblasts divide 80-100 times before they stop dividing and senesce.

Although this gives an indication of the longevity of the best cells, it is somewhat artificial because most primary cultures are heterogeneous and a significant proportion of the culture has a substantially shorter lifespan. As per estimate a total of ~45 population doublings are required to generate targeted cells for NT from fetal derived somatic cells using a combination of electroporation, drug selection and high- throughput screening. Although it is encouraging that populations of bovine fibroblasts retained their totipotency for nuclear transfer after more than 45 cell divisions, the senescence of primary cells reduces their overall efficiency of targeting. Importantly, this also limits the number of genetic changes that can be made to a given primary cell line.

In contrast to pro-nuclear injection, germ line modifications introduced by gene ta rgeting a re usually re ce ssive (e.g. kno ckouts) o r c o-dominant (e. g. ge ne modifications). In either case it will usually be necessary to produce the change in both alleles to see the phenotypic effect. This can be accomplished by conventional breeding, but in livestock this will take many years. It will take two generations to breed to homozygosity from a heterozygous founder animal. In cattle, this would take

3- 4 years although one generation could be eliminated if the modification was introduced into male and female cells and the cloned animals interbred.

An attractive solution to accelerating the route to homozygosity would be to target the second allele in cell culture. However, the limited proliferative capacity of the somatic donor cells currently used in NT makes targeting of the second allele a difficult proposition. One solution would be to clone by NT from the cells in which the first allele has been targeted, re-isolate cell lines from the cloned fetal material and then carry out the second targeting in these cells. In cattle, cultures have been prepared from cloned fetuses indicating this is feasible. The possibility of accruing deleterious genetic or epigenetic mutations may prove problematical and it is likely some re- isolated cell lines will not be viable for further cloning. In the long run, what is required are immortalised donor cells that retain their totipotency for NT, and in which the targeting window can be extended to carry out multiple genetic modifications, such as targeting both alleles or modifying more than one genetic locus. Although in its infancy, gene targeting in livestock will enable a range of genetic changes that will expand current applications for genetically modified animals and create wholly new opportunities.