Srf

JEF1,

JEF1,

FIGURE 1.5 cfs-Regulatory elements of some muscle-specific genes. These genes can be considered as members of muscle gene batteries: MCK, muscle creatine kinase; MLC, myosin light chain; MHC, myosin heavy chain. They share target sites for several specific transcriptional regulators, shown in color: MEF2, red; SRF, tan; MyoD family (MDF) bHLH factor, blue. In addition each gene has other specific target sites, shown as open symbols. Data for mouse MCK are from Donoviel et al. (1996), Fabre-Suver and Hauschka (1996), and Grayson et al. (1998); for MLC, from Rosenthal et al. (1992) and Rao et al. (1996); for chicken skeletal a-actin, from MacLellan et al. (1994); for mouse smooth muscle MHC, from Zilberman et al., (1998). In the latter gene the target site sequences are highly conserved amongst rat, mouse, and rabbit smooth muscle MHC genes, and thus are liable to play significant regulatory roles; MyoD family factors are not expressed in smooth muscle (Olson, 1990) and its target sites are consequently absent from this regulatory element. [Adapted from Arnone and Davidson (1997) Development 124, 1851-1864 and The Company of Biologists, Ltd.]

FIGURE 1.5 cfs-Regulatory elements of some muscle-specific genes. These genes can be considered as members of muscle gene batteries: MCK, muscle creatine kinase; MLC, myosin light chain; MHC, myosin heavy chain. They share target sites for several specific transcriptional regulators, shown in color: MEF2, red; SRF, tan; MyoD family (MDF) bHLH factor, blue. In addition each gene has other specific target sites, shown as open symbols. Data for mouse MCK are from Donoviel et al. (1996), Fabre-Suver and Hauschka (1996), and Grayson et al. (1998); for MLC, from Rosenthal et al. (1992) and Rao et al. (1996); for chicken skeletal a-actin, from MacLellan et al. (1994); for mouse smooth muscle MHC, from Zilberman et al., (1998). In the latter gene the target site sequences are highly conserved amongst rat, mouse, and rabbit smooth muscle MHC genes, and thus are liable to play significant regulatory roles; MyoD family factors are not expressed in smooth muscle (Olson, 1990) and its target sites are consequently absent from this regulatory element. [Adapted from Arnone and Davidson (1997) Development 124, 1851-1864 and The Company of Biologists, Ltd.]

SRF (the "serum response factor"); and the muscle-specific MEF2 factor. Thus the basic principle of the gene battery: genes are indeed linked into given batteries by the presence of czs-regulatory target sites for common sets of transcriptional regulators. This principle is of large consequence, for it tells us where to look for the structural genomic features that assign given genes to given gene batteries, in both development and evolution.

Since cis-regulatory function devolves from cis-regulatory structure, one might expect there to exist structural features characteristic of elements that control terminal differentiation processes, and indeed there are. These elements need not act as regulatory pioneers, laying out new territories as do modules that execute pattern formation processes. Instead they are activated by transcription factors that have already been placed in the appropriate cells by prior specification processes. Single factors can sometimes provide the trigger needed to activate terminal differentiation genes; sometimes several different positive factors act synergistically, as in the examples shown in Fig. 1.5. When the spatial boundaries within which terminal functions are to be expressed have been set upstream, boundary-forming repressors are not required, and so some czs-regulatory elements that function in this situation are lacking any target sites for spatial repressors. Terminal czs-regulatory elements are often compact, and if a gene is to run in several different cell types, they may contain within the same module sites for diverse activators that are present in these cell types. Furthermore, terminal differentiation processes often mediate expression at very high levels, for the products of the genes they control are in some cases needed in relatively enormous quantities. The classic examples are the globin genes in mammalian erythrocytes. At certain stages these genes support > 90% of all protein synthesis in these cells (Hunt, 1974). Very high transcript levels imply that particular cis-regulatory features mediate maximal rates of initiation in the basal transcription apparatus, which of course must be instructed quantitatively as well as qualitatively (for transcription initiation rates in development see Davidson, 1986).

The special features of cz's-regulatory elements that control terminal differentiation processes remind one of the general point that the kind of role played by a given element in the regulatory network is reflected in the nature of its complement of target site sequences. Thus we return to the conclusion that the regulatory logic of the system is directly encoded in the DNA (Arnone and Davidson, 1997).

Was this article helpful?

0 0

Post a comment