Designing and cloning of the gene construct is the fundamental step in the generation of transgenic animals, essentially determining success or failure of all subsequent steps and whether the questions asked can in fact be answered by analyzing the transgenic animal. One helpful approach to avoid potential problems associated with the expression of the gene construct in transgenic animals is to test the construct beforehand by in vitro transfection of cells. It should be noted, however, that the in vitro tests must be regarded with caution because they will not always yield unequivocal and sufficient results for the transgenic animals.
A known fact is that the structure of DNA double helix is the same for eukaryotes and prokaryotes and that the genetic code is universal. In principle any DNA fragment, e.g., chemically synthesized DNA, cloned DNA, or fragments of chromosomes, can be micro-injected and will be integrated into the host genome with more or less the same frequency. The size of the micro-injected DNA does not appear to be subject to limitations. The fact that most of the transferred gene constructs are less than 20 kb in length is mainly due to constraints in the cloning capacities of cloning vectors currently available. The analysis of integrated gene constructs suggested that considerably longer inserts with lengths of up to 1 Mbp or longer can be integrated without any difficulty. The molecular structure of the injected DNA molecules is a significant parameter influencing the frequency of integration. The linear DNA molecules integrate approximately five-fold better than circular molecules. Under favorable conditions the use of linear DNA constructs leads to an integration frequency of approximately 25% in mouse oocytes. The use of DNA fragments with staggered ends rather than blunt-ended ones also appears to be of advantage. The constraint in random gene integration is that foreign genes are transcribed and degraded to some extent only during the first 24h following micro-injection. Transcription of micro-injected plasmids is reduced considerably after the first cell division.
In order to ensure that the transgenes are also transcribed like host own DNA it is vital to remove any prokaryotic vector sequences from the DNA constructs because they may later inhibit gene activity. It was also shown that use of transgenes is more advantageous in their original genomic form rather than using cDNA copies. The correct exon-intron structure appears to favor transcription efficiencies of transgenes. If the genomic sequences are not available or difficulties arise in preparing the gene construct, such constructs should at least contain one or shorter intron sequences. In many cases the cDNA coding for the gene to be transferred is therefore furnished at its 3' end with exon-intron sequences derived from the untranslated region of another available gene. Frequently the polyadenylation region of the gene in question or another gene is used for this purpose.
In principle a functional gene construct must carry regulatory sequences located 5' to the coding regions. Isolated DNA sequences without promoter/enhancer regions will also be integrated but they will not be expressed. Gene expression may be observed only in a rare case in which an accidental integration has taken place exactly 3´ to an endogenous promoter. Structural genes can be combined with any regulatory elements even those which are not normally associated with them. The promoter for the gene to be transferred is chosen to maximize the tissue specificity, the desired extent and the time of transgene expression. If the protein enclosed by the transgene is to be secreted, a suitable sequence encoding a signal sequence must be attached to the 5´ region of the coding sequence. Apart from regulatory and coding sequences a complete eukaryotic transcription unit must also contain sequences in its 3´region which allow correct 3´ end processing of mRNA and also influence the processing, the transcription of the transcript to the cytoplasm, and the translation of the transcript.