From DNA to Diversity: The Primacy of Regulatory Evolution
In the final chapter of this book, we consider why regulatory evolution is the major creative force underlying morphological diversity across the evolutionary spectrum, from variation within species to body plans. The link between DNA sequence evolution and phenotypic diversity often involves cis-regulatory elements acting as units of evolutionary change. Here, we take an in-depth look at the function of cis-regulatory elements and molecular models and examples of their evolution. We will examine a few experimental approaches that have begun to reveal the direct connections between changes in regulatory DNA, the evolution of gene regulation, and the origins of morphological diversity.
cis-regulatory elements have properties that make their evolutionary dynamics distinct from coding sequences. The turnover of transcription factor binding sites within elements with conserved functions reveals that compensatory mutation is a critical feature of cis-regulatory DNA evolution. Both theoretical considerations and experimental evidence suggest that the modification of cis-regulatory element function can readily occur through point mutations that create or destroy transcription factor binding sites. Variation within species at such sites appears to be widespread. A growing body of evidence supports the view that regulatory evolution at the species level is sufficient to account for the larger-scale patterns of morphological evolution, including the origin of humans.
WHY IS REGULATORY EVOLUTION A PRIMARY FORCE IN MORPHOLOGICAL EVOLUTION?
Many case studies in Chapters 5, 6, and 7 illustrated that selected changes in gene regulation in one part of the developing animal,
If it were possible to take judiciously chosen structural genes and put them together in the right relationship with regulatory elements, it should be possible to make any primate, with some small variations, out of human genes . . . Likewise it should be possible to make any crustacean out of the genes of higher Crustacea.
—Emile Zuckerkandl (1976)
independent of other parts, underlie the morphological diversification of serially homologous structures, the origin of novelties, and the divergence of traits between species. Regulatory evolution is the enabling genetic mechanism for the modular organization and diversity of larger bilaterian phyla and for the emergence of new morphological characters.
The argument made for the central importance of regulatory change in morphological evolution is not a new one. Indeed, the creative potential of regulatory change and the comparatively greater constraints on protein evolution were recognized early in the history of molecular biology. Now there is considerable empirical evidence which demonstrates that changes in the regulation of genes that affect morphology are implicated much more frequently in the evolution of diversity than are new genes or functional changes in protein sequences. Furthermore, our current understanding of genetic regulatory hierarchies, networks, and circuits in development is capable of revealing why this is so.
Specifically, regulatory evolution is powerful because of the following characteristics:
1. Regulatory evolution enables pleiotrophy of toolkit genes. The same transcription factor usually controls different target genes in different tissues at different stages of development. The DNA binding activity of a given protein usually remains the same throughout development, however, it is evolution within the regulatory sequences of target genes that enables these genes to respond in a context-specific fashion.
2. Regulatory evolution enables developmental modularity. For anatomical structures to become different from serial homologs within an animal or from homologs in other animals, changes must evolve in the regulatory hierarchies that operate during the development of these structures. Due to their modular organization, changes in cis-regulatory elements allow changes in gene expression to occur in one structure independently of another (this independence is also referred to as dissociation). In metazoans, the great success (in terms of species diversity and adaptation to terrestrial, aquatic, and aerial environments) of highly modular body plans such as the arthropods, vertebrates, and annelids suggests that the modularity of developmental regulatory mechanisms at both the anatomical (for example, fields) and molecular (cis-regulatory elements) levels have facilitated this diversity.
3. Regulatory evolution is a rich and continuous source of variation. The cis-regulatory regions may occupy vast stretches of DNA and contain many independent functional elements. Changes in regulatory sequences within individual elements may subtly affect the level, timing, or spatial pattern of gene expression, perhaps very selectively in terms of the tissues and stages of development involved. Without any change in protein sequences, these quantitative, temporal, and spatial changes in the deployment of regulatory genes may affect the level, timing, and spatial expression of other developmental genes. Regulatory changes are often cryptic and have little effect on morphology, but they nevertheless create genetic variations with the potential to produce the morphological variation that is the raw material of evolution. Regulatory DNA is a rich and continuous source of potential genetic, developmental, and phenotypic variation, and thus evolutionary change.
4. Regulatory evolution creates novelty. The exploration of new morphologies is facilitated by new combinations of gene expression that can arise without changes in protein function.
THE FUNCTION AND EVOLUTION OF C/S-REGULATORY DNA
Several examples of czs-regulatory elements and czs-regulatory regions were presented in Chapter 3 in the context of understanding the genetic regulatory networks that orchestrate pattern formation. To better understand the role played by czs-regulatory DNA in evolution, it is important to appreciate several general features of czs-regulatory systems.
First, most elements are regulated by a minimum of four to six transcription factors of various structural types. Transcription factors virtually never act alone, so the output of individual czs-regulatory elements is determined by the integration of multiple diverse inputs.
Second, the spatial relationship of binding sites for transcription factors within a cis-regulatory element can be of utmost functional importance. Within the few hundred base-pairs of a typical element, the proximity of binding sites can determine whether transcription factors interact cooperatively. For example, multiple lower-affinity sites may have a greater effect on transcription through cooperative interactions than a single high-affinity site. Most metazoan transcription factors bind a family of DNA sequences with some degree of degeneracy in the sequences recognized. The "flexibility" of transcription factor binding opens up greater opportunities for cooperative interactions.
Third, repression—not activation—is generally the ground state of gene expression. That is, the chromatin that packages DNA has generally repressive effects on transcription. The ci's-regulatory elements represent sites where protein complexes are assembled that attempt to overcome this repression.
Fourth, spatial boundaries of gene expression in a cellular field are usually set by both positive and negative inputs into ci's-regulatory elements. Loss of positive inputs (in trans) or of their binding sites (in as) eliminates or contracts the spatial domains of gene expression. Conversely, loss of repressors or their sites expands spatial domains. The sites where activators and repressors bind may overlap; in those cases, competition for sites may determine gene activity. In other cases, they may not overlap and the activators and repressors modulate the output of the element through short-range (fewer than 100 bp) or long-range (more than 100 bp) effects on transcriptional functions.
Fifth, expression of terminal differentiation genes, such as those encoding structural proteins that carry out functions specific to certain cell types, are often regulated by czs-elements that respond largely or entirely to positive transcriptional activators. Hence, genes found progressively farther downstream from the regulators that establish spatial domains in the developing embryo may not require negative inputs to prevent their expression in inappropriate positions or cell types.
In summary, czs-regulatory elements are evolved devices. The number, type, and topology of transcription factor binding sites within any element is shaped by very different kinds of functional considerations than those affecting coding sequences. These distinct functional features govern the rate and means by which new elements arise and extant elements vary and evolve.
Evolutionary changes in a given gene's expression may arise through a variety of mechanisms. That is, they may arise directly from alterations in the czs-regulatory DNA of the gene or they can occur indirectly, through changes in the deployment or activity of upstream transcription factors that regulate the gene (which ultimately reflect changes in the czs-regulatory elements of genes encoding these transcription factors or their regulators). In the remainder of this chapter, we focus on the developmental and evolutionary consequences of changes in cis-regulatory elements.
The cz's-regulatory DNA function may evolve through any of several molecular mechanisms. Here, we distinguish between two major potential sources of cz's-regulatory DNA evolution:
• The de novo evolution of czs-regulatory elements through changes in nonfunctional DNA (Fig. 8.1a).
• The evolution of czs-regulatory elements from preexisting functional elements (Fig. 8.1b-d).
Within the latter category, we distinguish between three modes of sequence evolution:
• Duplications and DNA rearrangements involving existing functional elements that create additional copies of elements or new elements (Fig. 8.1b).
• Modifications of existing cz's-regulatory elements, for example, through the gain or loss of binding sites for positive and/or negative regulators (Fig. 8.1c).
• The special case of co-option of an existing element that expands its developmental function (Fig. 8.1d).
First, we examine, from a theoretical viewpoint, the factors affecting the probabilities of these different mechanisms of cz's-regulatory DNA evolution. Next, we analyze some illuminating case studies of cz's-regulatory element evolution.
De novo evolution of cis -regulatory elements from preexisting nonfunctional DNA sequences
The vast majority of animal genomes are composed of noncoding DNA. Consequently, extensive stretches of DNA sequences exist that could potentially function to regulate the expression of adjacent genes. Much of this DNA consists of various lengths of repetitive elements, some of which may have architectural functions (for example, the chromosome centromere), but most of which are believed to be nonessential and not involved in the regulation of specific genes. In addition, an abundant amount of single-copy DNA occurs between coding regions.
In principle, such unconstrained sequences could evolve into functional czs-regulatory elements. The key question is, How "difficult" is it for random DNA sequences to evolve into a new regulatory element? To answer this question, we must consider two major factors. The first factor is the probability that one or more binding sites evolve for individual specific DNA-binding proteins. The second consideration is the minimal input required for czs-element
Was this article helpful?