Recent studies Indicate that different levels of control operate within muttigene loci In addition to regulatory sequences immediately flanking the genes, there are also elements that act over long distances on more than one gene. Competition for these elements among genes can influence both the level and timing of gene expression during development. expressed properly despite the presence of enhancers. Expression levels are generally low or undetectable, and are not related to the copy-number of the transgene construct. The normal pattern of temporal and tissue-specific expression of the gene is also disturbed. Founder mice that have integrated the transgenic construct at a different chromosomal position often show different patterns of expression. This phenomenon, where the presence of other regulatory regions in the host genome at or near the site of integration of the construct influences the level and specificity of expression of the transgene, is called a position effect. These observations led to the formulation of a new functional definition, that of the locus control region (LCR). The LCR was originally identified as a control region upstream from the human [~-globin locus s (Fig. 1) and several other such elements have now been described that are important for the activation of various loci, including CD2 (Ref. 8), lymzyme 9 and the MHC (S. Carson and M. Wiles, pers. commun.). LCRs are characterized by tissue-specific, developmentally stable DNase-I-hypersensitive sites. Such sites are thought to be short stretches of DNA that are not complexed into nucleosomes. An LCR has been functionally defined as an element that confers expression upon a transgene, to a level that is independent of its site of integration in the host genome but dependent on its copy numberL Thus LCR activity is quite differently defined from enhancer activity, using very different functional assays. In particular, the fact that the DNA is not integrated into chromatin in a transient assay allows the detection of a wide variety of sequences that have enhancer activity, but not LCR activity. The converse can also be the case, for example, in the [~-globin locus the LCR hypersensitive sites HS3 and 4 (Fig. 1) do not act as enhancers in transient transfectionsm~, although they are powerful LCR elementsla-14. HS2 shows activity in both types of a~ aynaS. 16, suggesting that there can be overlap between these two functional definitions. Given this overlap, it is not surprising that LCRs also contain multiple factor-binding sites, several of which have been shown to be important for activity of the element. However, in contrast to the ease with which enhancer activity can be obtained by binding site multimerization, we have as yet been unable to reproduce LCR activity using this approach.