Alan Dove
 

Despite the obsession of the popular media and the public with the prospect of human reproductive cloning, nuclear transfer was developed to solve a very practical problem in biotechnology—the difficulty of creating transgenic farm animals. Cloning promises to transform the efficiency of transgenesis from a dismal 1–5% using pronuclear injection, the current method of choice, to 100%. It will also significantly reduce the time taken to produce transgenic flocks/herds for commercial protein production and ensure a more consistent level of expression in founder animals. Together with the application of site-specific recombination to embryonic, fetal, and adult cells cultured from donor animals, cloning will allow precise definition of the roles of gene products and pinpointing of genes from genome mapping information. However, current low efficiencies of survival and developmental defects in cloned offspring indicate that in practice, the method is still far from routine. Meanwhile, researchers are seeking a deeper understanding of the underlying phenomena that allow the technology to work in the first place, in the hopes of exploiting it more effectively.

Making a transgenic
The basic strategy for creating a transgenic animal has not changed much since the first transgenic farm animals were generated in 1985¹. Using a technique termed pronuclear injection, 200–300 copies of a gene of interest linked to a promoter (e.g., from a milk-specific protein) are injected through a fine glass needle into recently fertilized eggs, which are cultured briefly and subsequently implanted into surrogate mothers (see Fig. 1). Only 1–5% of the animals born are transgenic and only a proportion of these express the added gene at a high level.

Integration of the transgene into the host chromosome is a very rare and random event, which makes pronuclear injection tedious, costly, and inefficient. As the process is random, integration can occur anywhere within the genome, often inducing severe mutations that influence expression of the transgene. Thus, even though successful integration may occur, the animal may show mosaic expression of the transgene (i.e. it is switched on only in certain tissues). Some of these problems can be overcome using insulator sequences (e.g., matrix attachment regions or locus control regions) to protect transgenes from local chromosomal position effects or by using transgene rescue, where coinjection of a known highly expressed transgene is used to facilitate expression of a transgene of interest.

Experience has shown that genomic transgenes are expressed more consistently and at higher levels than cDNA transgenes. This is thought to be due to the presence of enhancers within genomic introns, particularly in the first few introns at the 5' end of a gene. It presents a problem to the genetic engineer because many genes are simply too long to be incorporated as a whole into cloning vectors, so genomic DNA cannot be used. To get around this, shorter cDNA sequences are often used together with defined regulatory elements (e.g. a portion of the 5' genomic sequence). Yeast and mammalian artificial chromosomes, which have a larger insert capacity, have also been suggested as a solution. Even the method used to purify transgene DNA has been shown to affect the efficiency of the process.

In addition to pronuclear microinjection, limited success has been obtained using alternative methods such as transfection of unfertilized oocytes by replication-defective retroviral vectors (see ref. 4) or sperm-mediated delivery. Retroviral vectors have a broad host range and can be used to transform a restricted organs (enabling the rapid evaluation of transgene expression in an animal before germline transmission is attempted), but they also have a limited insert size and may pose biosafety issues. Sperm-mediated transgene delivery is relatively simple, but of variable efficiency and poorly understood. While it is rarely used for large mammals, it may turn out to be the most effective method for producing transgenic chickens.

To date, sheep, cattle, and goats have been the main species of farm animal used to express heterologous proteins. However, rabbits and pigs have also been employed for selected applications because of their larger litters and shorter generation times (see "Barnyard biotech"). Indeed, Pharming B.V. (Leiden, The Netherlands) and Genzyme Transgenics (Framingham, MA) are currently producing -glucosidase in rabbits for a phase II trial in Pompe disease patients. The chicken, in contrast, remains difficult to transform, but once achieved promises to deliver high protein yields very quickly.

Heterologous gene expression has been directed to the milk and blood, and also shown to be feasible in urine and even semen. Most commercial systems favor the mammary gland as the production system and use the -lactoglobulin (sheep), -casein (goat), or -s1 casein (cattle) promoters to drive expression. The mouse whey acidic protein promoter is commonly used in pigs and rabbits.

Transgenic livestock represent very attractive systems for producing recombinant protein because of their high production efficiency, ability to produce bioactive large and complex proteins (e.g. factor VIII), small unstable peptides (e.g., calcitonin in fusion protein form), proteins that require extensive post-translational modification (e.g., antithrombin III), and proteins that are multimeric in nature (e.g., superoxide dismutase). More importantly for companies, they offer a much more cost-effective approach than mammalian cell culture for protein expression—a transgenic goat, for instance, produces protein at a unit cost of $10–25/g compared with $100–1000/g for cell culture.

The benefits of cloning
While current transgenic technology is attractive, there is significant room for improvement to increase product development times, lower costs, and, for biopharmaceuticals, to shorten time to clinical trials. Using current methods, valuable time and resources are wasted producing nontransgenic or silenced transgenic offspring because of the inefficiency of the pronuclear microinjection method. If the transgenic animals produced turn out to be male, another generation of offspring must be produced (with an equal chance of generating males) to generate transgenic females that express the protein in their milk. What's more, those animals that are transgenic generally carry the recombinant construct at a single chromosome insertion site and thus are hemizygous for the transgene. This means that only half of their offspring (on average) will inherit the transgene, thus doubling the number of animals needed to produce a transgenic flock/herd. In principle, this can be addressed by testing progeny for the transgene using fluorescent in situ hybridization or the polymerase chain reaction (PCR). But even if the transgene is detected, there is huge variability in the levels of transgene expression among different founder animals.

Cloning offers several benefits to the animal researcher (for an outline of the process, see Fig. 1). First and foremost, for obvious reasons, there is a 100% chance that the cloned offspring of a transgenic animal will be transgenic. Thus, female transgenic donor cells can be used to produce a founder herd of lactating production animals in a single step. Once an optimal transgenic line has been identified that expresses the protein at the desired level, a breeder can clone a number of animals from the transgenic donor in a single cycle (instead of waiting for the animal to mature and reproduce). By cutting out a generation, significant savings in time can be accomplished—13–15 months for goats and two years for cattle.

The availability of donor cells isolated from transgenic animals that can be propagated and maintained in culture also offers advantages: First, it allows the production of many identical offspring over an extended period, as cultured cells can be frozen and stored indefinitely. Indeed, companies such as Genetics Savings & Clone (College Station, TX) are already offering to freeze donor cells for use in future cloning experiments. Stable and reliable ES cell lines have not as yet been developed for any farm animal, although ES-like cells from cattle and pigs have been used to generate chimeric/transgenic animals.

Second, and more importantly, the ability to use gene targeting to modify and select populations of cells of specific genotypes and phenotypes before embryo reconstruction will greatly aid in the production of livestock strains with desirable agronomic traits. It will allow not only the deletion or substitution of specific genes, but also the introduction of the single base changes in the genetic code that are typical of many human genetic diseases.

Before any of this is possible, however, nuclear transfer methods will need to be improved to overcome technical problems such as high rates of pre-, peri- and postnatal deaths and developmental defects among cloned offspring—a phenomenon that was first noted in cattle in the form of large calf syndrome.

Robert Wall, a researcher at the US Department of Agriculture (Beltsville, MD), emphasizes the need for further research to determine the stage(s) of the cell cycle that are optimal for the donor and recipient cells to develop into a normal animal. While for the most part gene expression is reprogrammed in nuclear transfer embryos, all structural changes may not be corrected, as has been demonstrated by the length of the telomeres in sheep resulting from nuclear transfer. There is an urgent need for better characterization of the role of genomic reprogramming, chromosomal reorganization, somatic mutation, and imprinting on the efficiency of the cloning process.

Transplants with a side of bacon
Ironically, one of the most speculative applications envisioned for animal cloning is the one that has received the most publicity. In theory, the ability to clone and genetically modify pigs, whose organs are similar in size to human organs, could reduce or eliminate the demand for human organ donors (see Table 1). In practice, explains Wall, "There are so many well-defined hurdles, it is hard to imagine that is going to happen any time soon."

The most immediate barrier to xenotransplantation of pig organs is hyperacute rejection. Pig tissues display a carbohydrate epitope that reacts with about 1% of human immunoglobulins, triggering a drastic immune response that destroys the transplanted organ. A targeted deletion of the gene encoding -1,3 galactosyl transferase (GT), the enzyme which produces the epitope, should significantly reduce hyperacute rejection, and several groups are hoping to use somatic cell cloning to delete this gene.

Though pig cloning has now been proven feasible (see refs 2,3 and p. 1055 ), the process remains extremely inefficient—only 1–2% of the cloned embryos manage to develop to term. Even if the gene can be deleted in pigs, "the [carbohydrate] structure may provide some essential biological function in pigs and thus destroying the -1,3 GT enzyme could be deleterious to the animals," notes Irina Polajaeva, a researcher at PPL Therapeutics (Blacksburg, VA) and an author on one of the recent pig cloning reports³.

As a result of these problems, Alexion Pharmaceuticals (New Haven, CT) is focusing on generating transgenic pigs using conventional technologies like pronuclear injection. According to Steven Squinto, Alexion's chief technology officer, the company is working on adding genes that will effectively shield transplanted pig organs from the human immune system. Though generating transgenic pigs this way is far easier than cloning them, Squinto concedes that xenotransplantation of pig organs into patients is still "pretty far out there" in the future. Baxter's (Deerfield, IL) subsidiary Nextran (Princeton, NJ) and Imutran (Cambridge, UK) are also developing transgenic swine expressing complement regulatory proteins for use in xenotransplantation.

In addition to the possibility of rejection, the potential for cross-species infection remains a major concern. Recent work has shown that porcine endogenous retroviruses (PERVs), which are integrated in the genomes of all pigs, can replicate in immunocompromised mice transplanted with pig cells (see p.1032). "We worry about recombination events that could occur in the context of the xenotransplanted cells in a human taking immunosuppressives," says Thomas Okarma, president of Geron (Menlo Park, CA). PERVs have not been shown to cause disease, but the idea of a porcine retrovirus integrating into a patient's genome—and possibly being passed on to the next generation—has made researchers, public health experts, and policymakers wary. In response, BioTransplant (Charlestown, MA) recently announced that it had bred miniature swine that do not produce PERV capable of replicating in human cells, although they have not tested pig cells in living animals.

Okarma argues that PERVs and other infectious agents are a major barrier to safe xenotransplantation, but he stresses that these concerns were not the primary reason Geron recently stopped funding work on pig cloning at the Roslin Institute (Edinburg, Scotland). Instead, Geron, which owns several key patents in the field (see "Bringing the barnyard into court") plans to license out its animal cloning technology while pursuing a different approach to the problem of organ transplantation: making it unnecessary. "We're interested in learning how to create embryonic stem cells from a patient's own skin biopsy," says Okarma.

In theory, pluripotent human ES cells, which Geron has succeeded in culturing, can be used to generate any type of cell in the body. If a somatic cell can be used to create a line of cloned human embryonic stem cells from each patient, the genetically identical cells could be introduced into patients without the danger of rejection. Injecting healthy cells into a damaged organ would be "a much less costly and challenging procedure than an entire organ transplant," according to Okarma, who asserts that repairing damaged organs could ultimately render whole-organ transplantation obsolete.

Squinto counters that even for transplanting tissues rather than whole organs, animals may be a better source than human ES cells. He explains that pigs naturally grow the fully-differentiated cell types that would be needed to treat diseases ranging from liver failure to Parkinson's disease, whereas "there is probably a pretty significant learning curve that one must go through to provide the necessary cues to the [ES] cell to drive it towards a specific lineage" in culture.

Down on the farm
Cloned pigs may not be employed in the operating theatre anytime soon, but it seems reasonable to expect that they will find their way to the barnyard more easily—along with the other livestock targets of recent cloning efforts. Randy Prather, a researcher at the University of Missouri (Columbia, MO), who has been working on cloning pigs, gives an example of the potential for genetically modified lifestock: "Myostatin is exciting—here's a natural mutation in cattle that results in extensive muscle growth. What happens if you knock out that gene in pigs? Do you get the same extensive muscle growth, and will it be cost-effective?"

The last question is the key in agriculture, where profit margins are much smaller than in medical biotechnology. "You're not going to be able to sell [services] for thousands of dollars" in the agricultural market, explains Robert Bremel, managing director of Gala Design (Sauk City, WI). Bremel says that cloning will have to become much more efficient before it is used for livestock. Even a less expensive technique, like the non-replicating retroviral vector Gala is using to introduce genes into cattle, will have to offer significant benefits before farmers are willing to pay for it.

Michael Bishop, vice president for research at Infigen (DeForest, WI), believes "cloned animals for breeding purposes that have been genetically modified for some disease resistance...may be the first product" to reach the agricultural marketplace. Clifton Baile, CEO of Prolinea (Athens, GA) agrees that disease resistance is high on the list, but expects "that the initial application will be in the area of improved meat quality and quantity in the hog industry." Since feed is the most significant cost in pig farming, creating a pig that produces more meat for the same amount of feed could be very profitable.

In addition to being economically worthwhile, a trait that reduces antibiotic use in farm animals or improves the quality of meat on supermarket shelves is good for public relations, a consideration that is not lost on researchers in a field that has been a lightning rod for criticism.

Though cloning will probably prove useful in developing new traits, it is unlikely that it will ever be competitive with breeding for routine animal production: breeding a replacement heifer on a dairy farm, for example, is a simple process that currently costs $100 to $200. Breeding may also be safer, at least until cloning is better understood, as Wall points out: "Does the process of nuclear transfer create genetic diseases? Are those imprinted genes that presumably have to get un-imprinted during nuclear transfer a potential locus for a genetic mishap? Do some of the steps in the process...serve as mutagens? I for one don't know."

In addition to the possibility of introducing mutations, the prospect of producing large herds of genetically identical animals does not sit well with many agricultural researchers and farmers, who fear that inbreeding between clones would bring out recessive genetic diseases. Jeffrey Turner, president and CEO of Nexia Biotechnologies (Montreal, Quebec) and a former goat farmer, says "we applaud the use of nuclear transfer for introducing genes, but after that we have concerns for the genetic diversity issues."

Though scientists are already discussing the genetic health of cloned herds, it is clear that such herds are still a distant speck on the horizon. "You have to ask yourself how many transgenic animals have been produced by nuclear transfer. I think the number approaches a dozen, and I'd dare say not a single one of those animals were produced to try to do anything other than prove that it could be done," says Wall.

Bulletproof goats
In contrast to the narrow profit margins for traditional agricultural products, pharmaceutically and industrially useful proteins can often be sold at a huge markup, especially if they can be purified easily in large quantities. Since transgenic mammals made by pronuclear injection can express proteins in their milk, work is already well underway to use them as expression systems for protein therapeutics. "Biopharmaceutical production of proteins for the human healthcare field will be the first products to enter the market place" says Infigen's Bishop.

Chicken eggs, cow's milk, blood, and even urine may ultimately be among the starting materials for protein production, but so far one animal system seems to dominate the field. The first cash cow of livestock biotechnology will probably be a goat. The quick breeding cycle of goats has made them an easy system for genetic modification by pronuclear injection, and phase III clinical trials are already underway for blood clotting factors and other therapeutics produced in goat milk.

Goats have also been cloned by somatic cell nuclear transfer, raising the possibility of site-specific addition or deletion of genes. Though this would undoubtedly be important for agricultural applications like disease resistance or improved meat quality, it does not appear to be necessary for protein production. According to Turner at Nexia, goats with randomly integrated transgenes make effective protein factories, and "there appears to be no correlation between what chromosome [a gene is] on and whether you produce the protein well."

Regardless of the technique used to produce the animals, they offer significant advantages over traditional protein expression systems. Besides the low production cost, researchers are now discovering that where some types of bacterial fermentations or cell culture failed, udders [groan -ed] may succeed.

Spider silk proteins, for example, have been the targets of a number of protein expression efforts, because they are stronger than any man-made material. Unfortunately, the high shearing forces in conventional fermentations cause the spidroin protein to aggregate, making it useless for manufacturing fibers. In contrast, says Turner, "milking a goat is very low-shear." Within the next month, Nexia expects to have usable spider silk proteins from the milk of a strain of fast-maturing Breed Early Lactate Early (BELE) transgenic goats the company has developed. Filaments produced on an experimental scale from a tissue culture system have already demonstrated the feasibility of the spinning process.

As with agricultural animal cloning, the first industrial products from genetically modified animals will be directly beneficial to consumers. Nexia is already working with the US Army to develop bulletproof vests and surgical suture material from spider silk, and the manufacturing processes for these products should be more environmentally benign than the production of conventional plastics.

Given the experiences of researchers and companies developing GM plants for agriculture, it is no surprise that animal cloners are paying attention to the public perception of their products. Nearly every expert contacted for this article commented on the importance of developing technologies that will benefit average citizens.

So far, at least, progress in animal cloning seems to fit well with this goal. While the ethical and public health concerns of human ES cells and xenotransplantation have drawn the most attention, therapeutic proteins, bioplastics, and better quality meat will be the first products derived from this technology. Even the kids should be happy about that.

1. Hammer, R.E. et al. Production of transgenic rabbits, sheep, and pigs by microinjection . Nature 315, 680 ( 1985) MEDLINE
2. Onishi, A. et al. Pig cloning by microinjection of fetal fibroblast nuclei Science 289, 1188–1190 ( 2000) MEDLINE
3. Polejaeva, I.A. et al. Cloned pigs produced by nuclear transfer from adult somatic cells . Nature 407, 86–90 (2000). MEDLINE
4. Wells, K. et al. Transgene vectors go retro. Nat. Biotechnol. 17, 25–26 (1999). MEDLINE