Search

We have used a simple binomial model of stochastic transgene inactivation at the level of the chromosome or transgene, rather than the cellular level, for the analysis of two mouse transgenic lines that show variegated patterns of expression. This predicts the percentages of cells that express one, both or neither alleles of the transgene in homozygotes from the observed percentages of cells, which express the transgene in hemizygotes. It adequately explained the relationship between the numbers of cells expressing the transgene in hemizygous and homozygous mosaic 21OH/LacZ mouse adrenals and mosaic BLG/7 mouse mammary glands. The binomial model also predicted that a small proportion of cells in mosaic mammary glands of BLG/7 homozygotes would express both BLG/7 alleles but published data indicated that all cells expressing the transgene showed monoallelic expression. Although it didn't fit all of the BLG/7 data as precisely as a more complex model, which used several ad hoc assumptions to explain these results, the simple binomial model was able to explain the relationship in observed transgene expression frequencies between hemizygous and homozygous mosaic tissues for both 21OH/LacZ and BLG/7 mice. It may prove to be a useful general model for analysing other transgenic animals showing mosaic transgene expression.

RNA processing events modulate final productivity of a given transgene. We have evaluated a series of RNA elements for their ability to enhance alpha1-antitrypsin production in mammary cells. Our results indicate the need for a case-by-case assessment of each construct design and the occurrence of gene silencing events in vivo.

We have generated transgenic mice which express an ovine beta-lactoglobulin transgene during lactation. In two transgenic lines, BLG/7 and BLG/45, beta-lactoglobulin protein levels vary between siblings, reflected at the cellular level by a mosaic transgene expression pattern in the mammary tissue that is reminiscent of position effect variegation. To investigate whether this variegating expression profile can be affected by the introduction of an identical variegating locus on the homologous chromosome, we compared the beta-lactoglobulin expression profiles in mice hemizygous or homozygous for the transgene locus. In BLG/45 mice, milk protein analysis revealed that transgene expression was effectively doubled in homozygous compared to hemizygous mice. In contrast, beta-lactoglobulin protein in hemizygous and homozygous BLG/7 mice displayed a similar range; although minimum expression levels were doubled in the homozygous population, the maximum level of expression was indistinguishable between the two populations. Fluorescent in situ hybridisation (FISH) for transgene mRNA indicated that for a given protein level, the extent of cellular expression is similar in both BLG/7 populations. In homozygous mice genomic DNA and nuclear RNA FISH demonstrated that only one of the two BLG/7 loci is active in expressing cells, while two transcription foci were present in BLG/45 homozygous mice. This mono-allelic transgene expression pattern is not inherited through the germline, as hemizygous mice bred from homozygous parents expressed at the expected hemizygous population level. We discuss these observations in the context of known epigenetic events such as imprinting and trans-inactivation.