|[Frontiers in Bioscience 1, e34-41, July 1, 1996]|
PIG alpha1,3GALACTOSYLTRANSFERASE: A MAJOR TARGET FOR GENETIC MANIPULATION IN XENOTRANSPLANTATION
Karen Strahan, Andrew Preece and Kenth Gustafsson
Division of Cell & Molecular Biology, Institute of Child Health, University of London, 30 Guilford St., London WC1N 1EH, UK
Received 06/21/96; Accepted 07/02/96; On-line 07/08/96
As alluded to above, the importance of the Gala1,3Gal epitope in the pig-to-man xeno-transplantation context has been well established. Despite the possibility that inhibition or even total eradication of the epitope in pigs may not be the only action needed, it would most certainly go a long way towards overcoming the HAR. To this end, attempts to inhibit or silence the gene are currently being carried out in several laboratories.
The most straightforward and seemingly logical approach for eradicating the action of the pig a1,3GT would be to knock-out the gene. A gene knock-out (43) involves the cloning of a gene and subsequently making a knock-out construct containing an insertion of an antibiotic resistance gene (e.g. neomycin). This insertion serves a dual purpose, in that it interrupts the targeted gene as well as provides a tool for selection. In addition, a second selection mechanism involving the Herpes Simplex Virus-thymidine kinase (HSV-TK) gene flanking the gene of interest is commonly used to provide selective killing of constructs that have not undergone gene targeting by homologous recombination. The construct is subsequently transfected into embryonic stem cells in culture, and the two forms of selection are applied. This is followed by analysis of surviving colonies of cells. ES cells displaying the desired knock-out genotype are subsequently introduced into pseudo-pregnant recipient females by either blastocyst injection or morula aggregation. This type of 'knock-out' approach has become almost a routine procedure in at least one mouse strain (i.e. 129 strain). Recently, the corresponding GGTA1 gene in mouse was 'knocked out' by Thall et al. (44). Interestingly, the resulting Gala1,3Gal-negative mice appear to secrete NAb against the epitope re-enacting the event in our primate ancestors. These mice also appear healthy and fertile indicating that a similar approach in pigs would not be harmful to the animals. However, despite intensive efforts, ES cells from pigs are not yet available and consequently this approach is not likely to be used in the immediate future.
An elegant strategy makes use of the fact that the action of a1,2fucosyltransferase (a1,2FT) adds a fucosyl residue in the position used as the substrate for addition of the terminal galactosyl residue in pigs (see Fig. 2). This strategy is being pursued by Sandrin and colleagues. They recently showed that transfection of a pig kidney epithelial cell line with human a1,2FT (45), the H-transferase synthesizing the substrate for human ABO-transferases, successfully competes with the pig a1,3GT (10). This competition leads to a substantial decrease in human NAb binding as well as in cytotoxic sensitivity of the cells. It should be possible to apply this strategy systemically, by making transgenic pigs expressing the human a1,2FT gene on pig EC. However, it is still unclear if the level of reduction achievable in this setting is enough to avoid the HAR.
By using oligonucleotide antisense technology (46, 47), we have attempted to down-regulate the expression of the Gala1,3Gal epitope. Oligonucleotide antisense DNA consists of short stretches of DNA complementary to a given mRNA molecule. Our experimental system consisted of a cultured pig EC line (PIEC) that was subjected to differing amounts of several antisense oligo-nucleotides corresponding mainly to the region for initiation of translation in the pig a1,3GT mRNA. One such oligonucleotide, when added to the medium every other day for 7 days at a concentration of 20µM, was found to inhibit expression of the epitope by approximately 40-50% (13). Since the binding of affinity-purified Ab paralleled the binding of epitope-specific lectin, these experiments also confirmed that the Gala1,3Gal epitope is the major target for human anti-pig NAb. The data also excluded the possibility that human anti-Gala1,3Gal NAb can cross-react with other pig molecules. One difficulty with such a system is that the turn-over rate of such epitopes on several glycoproteins and glycolipids may be considerable. In addition, it is hard to see how it would be possible to use such oligonucleotides in vivo, for at least two reasons. Firstly, administration of oligonucleotides would be a problem in vivo as well as ex vivo in an organ, and secondly, the effect of the oligonucleotides would be merely confined to one or two days.
A current strategy under development involves, by stable integration, the delivery of ribozymes to target cells. Ribozymes were found more than ten years ago to down-regulate in several species the expression of specific mRNA's by digesting their target in a specific location (48, 49). The hammerhead ribozyme consists of a loop structure interrupting two flanking antisense sequences carrying specificity for the target (50, 51). A number of studies have shown that it is possible to design ribozyme constructs for a desired target specificity in mammalian cells. Design and synthesis of oligonucleotides is followed by their ligation into a eukaryotic expression vector (52-55). This construct is subsequently transfected into the cells of interest leading to expression of a ribozyme RNA molecule. It has been previously shown that ribozymes expressed in a transgenic mouse were able to drastically downregulate expression of b2-microglobulin (52) and pancreatic b-cell glucokinase (55). Adoption of a similar strategy in pig may allow development of constructs that lead to ribozyme RNA that is capable of down-regulating the a1,3GT mRNA.