Expression of transgenes within a single generation by direct DNA injection into vertebrate embryos has been plagued by inefficient and nonuniform gene expression. We report a novel strategy for efficient and stable expression of transgenes driven by both ubiquitous and tissue-specific promoters by direct DNA injection into developing Xenopus laevis embryos. This strategy involves flanking expression cassettes of interest with inverted terminal repeat sequences (ITRs) from adeno-associated virus. Our results suggest that the ITR strategy may be generally applicable to other systems, such as zebra fish and embryonic stem cells, and may enable tissue-specific expression of transgenes in problematic contexts.

XCsx2, a homeobox-containing gene, is expressed in cardiac muscle during Xenopus development, while the XMEF2A gene is expressed in both cardiac and skeletal muscle. Microinjection of either XCsx2 or XMEF2A mRNA into single blastomeres of two-cell stage Xenopus embryos induced precocious expression of the myosin heavy-chain alpha (XMHCα) gene at the neural plate stage (stage 14). Co-injection of both XCsx2 and XMEF2A mRNAs induced still earlier expression at the late gastrula stage (stage 12). These changes were evident in whole embryos but not in animal pole explants from injected embryos. Overexpression of XCsx2 or XMEF2A also caused an enlarged heart and abnormalities of notochord and tail in Xenopus embryos. These findings suggest that both XCsx2 and XMEF2A transcription factors have an important role in regulating the expression of the XMHCα gene and the morphogenesis of heart tissue in Xenopus development.

Previous papers have reported that DNAs exogenously injected into Xenopus laevis fertilized eggs are expressed only at and after the midblastula transition (MBT). We have injected fertilized eggs of Xenopus laevis with circular plasmids that contained bacterial chloramphenicol acetyltransferase (CAT) genes connected to the promoter of viral genes (pSV2CAT and pAd12.E1aCAT) or the Xenopus cardiac actin gene (actin-CAT fusion gene), and examined whether these DNAs are expressed during the stage before the MBT. We found that expression of CAT enzyme can be detected before the MBT when CAT genes connected to viral promoters were injected. The activity was low during the cleavage stage on a per-embryo basis; however, the time course of the accumulation of the CAT enzyme activity roughly paralleled the increase in cell number. Therefore, CAT enzyme activity per cell was constant during cleavage and did not change dramatically before and after the MBT stage. CAT mRNA level, detected by CAT antisense RNA, was roughly proportional to the levels of CAT enzyme. Therefore, we assume that the observed enzyme activity reflects the transcriptional activity of injected CAT genes before and after the MBT. When the actin-CAT fusion gene was injected, however, no enzyme activity was detected until embryos had reached the neurula stage, a stage when endogenous actin genes are first activated. On the basis of these results, we conclude that the concept of an initial transcriptional activation of exogenous genes at amphibian MBT has to be changed. We propose that the expression of polymerase-II-transcribed genes is regulated by the nature of the promoters connected to the genes rather than by changes associated with MBT.

No abstract