Design of expression vectors

Expressing a transgene in an appropriate manner is a huge task. The gene construct that is quite active in transfected cells may be only weakly expressed in transgenic animals. Moreover, each line of animals shows a specific pattern of transgene expression. In addition, a number of transgenes remain silent in some cells but are expressed in others. This is a clear limitation for the use of transgenic animals. A large number of transgenic founders are attempted until they find the lines expressing the foreign gene as they expected. This approach implies wastage of time and animals but it appears sufficient to obtain 1,000-10,000 copies of recombinant protein per cell exhibiting the expected biological action. This relative but real success of the expression vectors dissuaded researchers for years from improving them.

The situation is altogether different when recombinant proteins are to be prepared in milk. Highly efficient expression vectors are often needed to obtain the expected large amounts of proteins recovery from milk. However, it must be ensured that the recombinant proteins are restricted to milk as much as possible to avoid any potential deleterious effects on host animals. A significant amount of work was done for the production of recombinant proteins in milk and a few efficient expression vectors were designed and patented. Yet basic studies on the mechanisms of gene expression and empirical observations have led to defined rules for vector con­struction and to identification of pitfalls that can be avoided. A functional transcription unit is composed of several elements that cooperate with each other. These elements and their mode of action are not yet fully known. A gene construction consists of associating more or less known elements. This may generate nonfunctional vectors for unknown reasons. The major recommendations for design­ing reliable vectors may be summarized as follows.