10.1098/rsta.2003.1195Regulated recruitment and cooperativity in thedesign of biological regulatory systemsB y M a r k Pt a s h n eMemorial Sloan{Kettering Cancer Cen ter, 1275 York Avenue,Box 595, New York, NY 10021, USA ([email protected])Published online 6 May 2003What distinguishes a man from a mouse is not so much di¬erent proteins, butrather the appearance of common proteins (etc.) at di¬erent times and posi tionsin the developing organisms. Thus speci c genes are transcribed or repressed, pro-teins degraded or stabilized, RNA transcripts spliced one way or another, and soon. These are examples of `regulatory’ decisions. A rather simple mechanism|calledregulated recruitment|lies at the heart of many of these regulatory decisions.Keywords: regulated recruitment; coop erative binding; evolvability;synergy; enzyme speci¯ city; gene activation1. IntroductionA goal of this symposium was to identify simple rules, or strategies, that can beused to develop complex systems. I describe here a simple mole cular strategy thatnature has used widely in evolving biological complexity. This description is part ofan extended argument put forth in the recently published book Genes and signals(Ptashne & Gann 2002). The biological complexity I refer to has two aspects, onehaving to do with the development of a complex organism f rom a fertilized egg, andthe other with the evolution of di¬erent forms of life. A similar theme is sounded inthe two cases: limited sets of gene products (proteins, mainly and RNA molecules)are used, in di¬erent combinations, to generate diversity.For example, proteins that generate human hands also generate human feet, butthe di¬erent app endages form as a result of di¬erent patterns of appearance of theseproteins as the embryo develops. And, speaking somewhat more informally, a com-mon set of gene products can be used to gen erate a human, a mouse, and perhapseven a ®y, by appropriately regulating their appearances as each of these organismsdevelops. There is more to it than this, of course, but these statements su¯ ce topose the following problems.How is the appearance of proteins regulated during development of any one organ-ism? What kinds of molecular changes to the regulatory apparatusare required toOne contribution of 18 to a Theme `Self-organization: the quest for the origin and evolution of structure’.Phil. Trans. R. Soc. Lond. A (2003) 361, 1223{12341223c°2003 The Royal Society1224 M. PtashneFigure 1. Cooperative binding. Interactions between proteins binding to DNA (for example) aretypically weak, involving, say, 1{2 kcal of interaction energy. The resulting e® ect on binding,say 10- to 100-fold, can be physiologically important. Modi¯ed with permission from Ptashne& Gann (2002).modify those patterns as evolution proceeds? To return to our original theme, howis biological complexity generated from a limited se t of common elements?2. Protein regulationThere are many ways to regulate the appearance of a protein in a cell at a given timeand place, but for the purposes of illustration we consider just three: a gene can betranscribed (or not); a RNA molecule can be spliced, sometimes one way or another;and a protein can be destroyed (by proteolysis). In each case th e regulation can beeither/or, or it can be graded: a gene can be transcribed more or less, and so on. Mygeneral point is that a common mole cular strategy has been u sed in the evolution ofregulatory systems that e¬ect each of these three kinds of control (and many othersas well). The mechanism is called `regulated recrui tment’. To understand it we need rst to consider the ubiquitous phenomenon of cooperative binding, as illustrated in gure 1. I will then illustrate how the principle is used in transcription, discuss someof the implications, and then return, near the e nd, to RNA splicing and proteolysis.Figure 1 shows three macromolecules; we will call A and B `proteins’, and therod `DNA’, but the following characteristics apply more generally. Given speci eda¯ nities of A and B for their sites, and assumi ng that A and B are present atapproximately some speci ed conc entrations (as would be found in a cell, for exam-ple), neither A nor B is bound e¯ ciently to its site when the other protein is absent.When both proteins are present, however, b oth proteins are bound, an e¬ect calledcooperative binding of proteins to DNA. The word `cooperativity’ is laden withhistorical associations, and so we have to be clear what is required in this case:Phil. Trans. R. Soc. Lond. A (2003)Regulated recruitment and cooperativity 1225Figure 2. Gene activation e® ected by cooperative binding. The rod is a DNA fragment containingthe lacZ gene, a site that binds the activator CAP, and a more extended sequence that binds theRNA polymerase, called the promoter. The `activator’ CAP binds DNA only when complexedwith the small molecule cyclic AMP. Glucose in the medium depletes intracellular cyAMP, andso CAP only works when glucose is absent from the medium. Reproduced with permission fromPtashne & Gann (2002).the proteins must simply `touch’ one another when bound to DNA. Conformationalchanges, though they may occur, are not necessary for the e¬ect, nor is energy inthe form of adenosine triphosphate (ATP) (for example) used. Rather, one proteinhelps the other bind simply by providing binding energy in the form of an `adhesive’interaction between the proteins.To see how the reaction of gure 1 applies to gene regulation, consider gure 2,which shows a bacterial transcriptional activator in action. The gene is the famouslacZ (beta-galactosidase) gene of Escherichia coli. (For reasons that become evidentupon considering the entire argument of Ptashne & Gann (2002), analysis of geneactivation, rather than gene repression|the usual entry into the subject of generegulation|is the more revealing approach).The catabolite activator protein (CAP), binds speci cally to its site on DNA, andRNA polymerase binds to its site, called the promoter. In th e absence of CAP (andalso in the absence of another regulatory protein, the lac repressor, see below), RNApolymerase binds to the promoter, but only infrequently. CAP binds cooperativelywith the bac terial RNA polymerase as shown, and thereby `activates’ the gene.Note that the term `activation’ as used here can be misleading. N either the enzymenor the gene has in any meaningful sense been `activated’; rather, the
View Full Document
Unlocking...