Our knowledge of the mechanisms that regulate transcription in higher eukaryotic cells has increased enormously during the past 2 years. Earlier studies, using a combination of in vitro mutagenesis and DNA-mediated gene transfer, identified two distinct types of cis-acting regulatory sequences: promoters, which are located close to the initiation site and act in a position-dependent fashion, and enhancers, which can be located far from the initiation site and act in a position-and

more »

... dependent fashion. Promoters can be subdivided into proximal elements, including the cap site itself and the TATA box, which is involved in fixing the site of initiation, and distal elements, which can be spread over several hundred base pairs. It is now clear that many promoters, particularly those of 'housekeeping' genes, lack TATA boxes and are instead composed of GC-rich elements that are often located within methylation-free islands (Bird 1986). Transcription controlled by this latter class of promoters often initiates at multiple sites. Many enhancer elements, for example, those of the immunoglobulin, insulin, and elastase genes, impose tissue-specific expression on adjacent promoters. These cis-acting elements operate by interacting with protein factors, many of which have now been identified by gel retardation and footprinting assays and some of which have been purified to homogeneity. In a few cases the corresponding genes have been cloned. In many of these studies well-characterized cis-acting elements of viruses, particularly of the DNA tumor viruses, have played a major role. At the recent ICRF-sponsored meeting on the Molecular Biology of SV40, Polyoma, and Adenoviruses (held in Cambridge, England, July 27-August 1, 1987) there was a major emphasis on such trans-acting factors. The results reported both identified new factors and considerably clarified the relationships amongst previously described factors. In this review we will discuss data reported at the meeting and, where appropriate, refer to recently published work. We will concentrate almost exclusively on work with DNA tumor viruses, referring to other systems only when there are direct connections to our major topic. Table 1 lists the factors of relevance, giving the DNA sequences they recognize, their synonyms, and the regulatory systems in which they are known to be involved.