Info

(^Transcription factors^)

Epidermal growth factor \ // receptor, ErB2, 3, 4

Figure 7-3. Two views of the Ras regulatory pathway that responds to growth-factor signal. (A) Genetic view; (B) view from molecular cell biology. Ovals indicate individual factors whose specific names are unimportant here. Grayed ovals indicate factors classified as "effectors;" white ovals as "intracellular mediators" of the signal. Redrawn from (Downward 2001) with permission; see original for details and terms, text for explanation. Original figure copyright Nature Publishing Group 2001.

Figure 7-3. Two views of the Ras regulatory pathway that responds to growth-factor signal. (A) Genetic view; (B) view from molecular cell biology. Ovals indicate individual factors whose specific names are unimportant here. Grayed ovals indicate factors classified as "effectors;" white ovals as "intracellular mediators" of the signal. Redrawn from (Downward 2001) with permission; see original for details and terms, text for explanation. Original figure copyright Nature Publishing Group 2001.

enumeration for understanding, underestimating phenogenetic equivalence and its importance or prevalence, and viewing genotype-phenotype relationships as being more prescriptive and unitary than they are in nature. In this sense, this new work resembles the stage of biology in the 19th century, when naturalists traveled around the world collecting the beetle specimens that were then dispersed throughout the museums of Europe. This led to an accumulation of species lists that we tend today to view as merely descriptive natural history. Until we have a better ability to synthesize and digest the information from gene expression studies, the difference between then and now may be less than we think.

One problem is that manipulation of gene expression in the laboratory is almost always artificial and can even make systems seem misleadingly complex (or blind investigators to alternative pathways) (Downward 2001; Niehrs and Meinhardt 2002). This is illustrated by comparing Figures 7-3A and 7-3B, and that does not even consider variation within species. Sequence variation in all the elements, from TF protein sequences and regulation to the REs of regulated genes and their surrounding sequence and chromatin structure, can lead to quantitative variation in traits, while experimental studies in development generally focus on the invariant aspects of mechanisms (at least, as seen in model animal strains). Experiments cannot in a practical sense test all combinations of variants, and strain-by-strain variation has raised caution flags that should not be ignored.

Most pathway diagrams seen in the literature are partial, even when they appear as 7-3B, in part because they do not include all the other pathways required to activate the components that are shown. More thorough diagrams can be intimidating in complexity (Davidson et al. 2002; Davidson et al. 2002) even if the pathways shown are an exhaustive list and even if they are all correct and always operative. Individual studies report experiments on pieces of networks, often those known by experience with natural or randomly engineered mutation, but then specifically followed up experimentally. This can sometimes give the impression that the tested genes are the only genes of importance, but this can be highly misleading (expression profiling shows this as well).

Reports of only subsets of genes are understandable for practical reason, but often they consist of a mixed bag of genes, including receptors, TFs, some second messengers and the like. One can view these as akin to throwing darts randomly at a complex diagram, and then studying the result as if those were the only important genes. These facts should be kept in mind in reading this book. Many examples will follow in which we present these kinds of partial sets of genes, and in discussion of other phenomena we simply omit pathway details that do not apply to the logic of the system we are discussing.

Considering the complex nature of such cascades one can see the many points where mutation can introduce variation. Nonetheless, despite all these caveats, there is considerable and tractable order. Some broad generalizations have already emerged, and we can consider some of them and what they mean.

A Basic Functional Toolkit

The logic of cis-regulation, messenger-transduction, modular genome organization, and combinatorial expression provides a powerful new way to understand how new function arises and complex organisms are assembled: a small number of components can generate a huge number of combinations, and a new combination can in principle add novelty without destroying existing traits in an organism, for example, by only affecting specific enhancer locations, not a TF itself.

0 0

Post a comment