transcriptional regulation - senecio7の日記

Weirauch MT & Hughes TR 2010 Conserved expression without conserved regulatory sequence: the more things change, the more they stay the same. Trends Genet 26:66-74.

cis- and trans-regulatory changes can contribute to sequence divergence without dramatically altering gene expression outputs
heterologous DNA often functions similarly in organisms that share little regulatory sequence similarities
indicating that trans-regulatory mechanisms tend to diverge more slowly and can accommodate a variety of cis-regulatory configurations

this capacity to 'tinker' with regulatory DNA probably relates to the complexity, robustness and evolvability of regulatory systems
but cause-and-effect relationships among evolutionary processes and properties of regulatory systems remain a topic of debate

currently, we are largely unable to recognize functionally similar regulatory DNA

among the hundreds of thousands of conserved elements in mammals – many of which are believed to be enhancers – most bear little or no sequence similarity to each other
indicating that regulatory regions are virtually all unique
expression output can be conserved despite divergence and dissimilarity in regulatory regions

what is the benefit, if any, of a system that constantly changes over evolutionary time?

there are two general types of mechanisms by which a regulatory region can change sequence without affecting the transcriptional regulation of the gene under its control
first, the trans-acting factors can change with the regulatory DNA sequence compensating for this change
second, the cis-regulatory sequences can rearrange into a different but functionally equivalent configuration
the first mechanism is easier to understand and is exemplified by several recently described concrete cases in yeast
the second mechanism seems more prevalent, but specific details are harder to pinpoint

a handful of studies have documented examples of the evolution of TF binding affinity
the changes in binding affinity presumably require changes in the cis sequences that the TF binds
on the whole, regulatory proteins are among the most slowly evolving of all protein classes
the amino acid sequences of DNA binding domains are usually highly conserved
most TF sequence preferences are thought to be largely unchanged over large evolutionary distances

presumably, TF binding affinities tend to be conserved because changes will impact all genes under the TF's control

the number of TF binding sites in vivo is large
many binding sequences are "functional" in binding proteins
even though many of them might not function in gene regulation in any particular situation
CREB (cAMP response element binding protein) binds ~4 000 human promoters in vivo
but only a small proportion are induced by cAMP in any cell type

the species-specific binding events are, for the most part, recapitulated when a human chromosome is placed in mouse
indicating that the trans-regulatory apparatus is largely conserved, and that the differing sequences of the chromosome determine the differing arrangement of proteins
the conservation of expression patterns is tolerant not only of the gain and loss of binding sequences but also the resulting rearrangement of protein binding events

the question still remains as to why cis-regulatory regions would be so prone to change in the first place, relative to coding sequences, given that the end output of gene transcription often remains so similar
providing evolutionary plasticity, while minimizing the risk of failure, might be an advantage
a system that inherently allows minor variation – for example, the addition (or removal) of a sequence that drives expression in a given tissue, without otherwise altering the regulatory properties of the gene – might have long-term advantages
in this sense, cis-regulatory turnover and shuffling might be a byproduct of organizational schemes that ensure consistent function while facilitating variation and neofunctionalization

tolerance to changes in trans might also contribute to regulatory evolution by broadening the array of cis-regulatory sequences that produce an appropriate transcriptional output
a mutation that strengthens an interaction between a TF and its cofactor can compensate for a mutation to the cofactor-DNA interaction, and so can promote cis sequence turnover and increase the possibility of interaction with a new cofactor

the redundancy offered by such a system plays a role in avoiding the deleterious effects of uncontrollable mutations
altogether, it is easy to postulate that existing regulatory schemes, and their capacity to tinker, might contribute to redundancy, robustness, modularity, complexity and evolvability
all concepts now broadly associated with regulatory network properties and hypothesized to be a product of evolutionary processes, or at least favorable for both survival and adaptation

the capacity of gene regulatory systems to tinker could be a byproduct of the fact that complex regulatory architectures are needed to successfully create an animal
the drive for complex regulatory architectures and/or the capacity to tinker might also provide an explanation for the fact that more complicated organisms tend to have a larger number of TFs, but that these TFs tend to have less sequence specificity (and, often, more promiscuous binding in vivo)
the promiscuity of eukaryotic TFs is likely to constitute one of many eukaryotic evolutionary novelties, which might enable more evolvable gene regulation and thereby be essential for the evolution of a variety of structures