Source: (Lander, Linton et al. 2001).

Source: (Lander, Linton et al. 2001).

Like the codon system for amino acid specification, the recognition sequences are degenerate or redundant; that is, the RE sequences for a given TF can vary. An example is given in Table 4-5 for the "paired" domain of Pax6, a gene involved in the development of eyes and other structures. These enhancer sequences may not have equal binding efficiency, but all are recognized by a given TF protein. The nominal or "consensus" binding sequence for the (vertebrate) Pax6 double-bHTH paired domain is shown on the rightmost column ("N" means any nucleotide). Pax6 also has a homeodomain (a separate bHTH domain) and binds TAAT(T/C)(A/C/G/T)(A/G)ATTA (simplified somewhat from Callaerts et al. 1997). See Table 7-5 for a general description of the different TF classes that bind these domains.

Unlike the codon system, however, in which redundancy is amino acid specific, variation in recognition sequences can lead to binding by different TFs. Variation in the copy number or location of REs can also contribute to variation in the strength of gene expression and hence have qualitative or quantitative effects. This kind of variation is shown in Figure 4-8 for REs for Pax6 among various lens protein genes. Figure 4-9 shows variation in RE location among related genes used in rhombomere (hindbrain) segmentation.

The location of REs also varies among genes and among species for the corresponding gene and can be on either side of the gene, within the gene, near the gene, or tens of kilobases away.

This variation is clearly constrained by selection because there are so many similarities, but the pattern also shows the tolerance that evolves. In life, the functional meaning of REs is determined not just by their sequence but by their context-specific combinations. For example, a 30-base pair region called DC5 is needed to express a lens protein in the eye in chick embryos; this works if particular Sox and Pax class transcription factors jointly bind the 5' and 3' region of this enhancer (Kamachi et al. 2001).

Because RE sequences are short, they can be erased or generated relatively easily by chance in any DNA sequence. That is, any mutation has a substantial probability of generating a new RE, turning an existing RE into a different one, or altering the binding efficiency of an existing one. Mutations can also erase a binding site. Over reasonably short periods of evolutionary time, it appears that the blinking on

Was this article helpful?

0 0

Post a comment