A protease—present in a complex called the proteosome—degrades many proteins in eukaryotes. Regulation of this process is crucial to cell growth and survival. For example, destruction of certain proteins at specific times is required for progression from one phase of the cell cycle to another. How does the proteosome select the proper substrates to degrade at any given time?
Substrate selection by the proteosome is determined by ubiquityla- tion. Proteins are modified by attachment of polyubiquitin chains and then are rapidly degraded by the proteosome. Thus, our question becomes: how are the proper proteins selected, under any given set of conditions, for ubiquitylation?
Figure 4.1 illustrates one of several ubiquitylating enzymes found in yeast and other eukaryotes. The enzymatic machinery itself is a multipro- tein complex to which substrates are recruited by so-called F-box proteins. There are many F-box proteins that can (individually) attach to the com- plex. Each F-box protein bears two domains: a common “F-box” domain that binds to the enzyme complex, and a unique domain that binds a spe- cific substrate. F-box proteins are analogous to transcriptional activators that work by recruitment. Both simultaneously bind substrate and enzyme—RNA polymerase and gene in one case, the ubiquitylating com- plex and target protein in the other.
Figure 4.1a shows an F-box protein in action. In this case, the F-box protein (called Cdc4) binds its target protein (called Sic1) when the latter is phosphorylated, a modification that occurs at a specific time in the cell cycle. (Thus, phosphorylation of one protein creates a docking site for another, a theme we encounter often in the pages ahead.) The F-box pro- tein then brings its target to the ubiquitylating machinery.
Figure 4.1b describes an experiment showing that specificity, i.e., which protein is ubiquitylated, is determined solely by the adhesive inter- action of an F-box protein with a target. In this experiment, the substrate- binding portion of the F-box protein (of Figure 4.1a) was substituted by a protein fragment that binds the mammalian protein Rb (retinoblastoma). When expressed in yeast, this mammalian protein is ordinarily stable, but when expressed along with the new F-box protein, it is rapidly ubiquity- lated and degraded. Thus, even a protein that is normally not a substrate
modified Cdc4 Cdc4 Sic1
FIGURE 4.1.Recruitment of substrate by F-box proteins to the ubiquitylating machinery. E1 is an enzyme that transfers ubiquitin to E2, and that enzyme in turn transfers ubiquitin chains to substrates recruited by F-box proteins. (a) Sic1 is an inhibitor of a cyclin-depen- dent kinase. When phosphorylated, Sic1 binds the F-box protein Cdc4. Destruction of Sic1 causes progression from one stage of the cell cycle to another. (b) The F-box protein from a is used here in a modified form: its substrate recognition region has been replaced by a short peptide (of sequence LXCXE) that binds the mammalian protein Rb. Rb is recruited by this hybrid F-box protein to the ubiquitylating machinery, ubiquitylated, and subse- quently destroyed. E1 is called a ubiquitylating activating enzyme; E2 (in this case Cdc34) is called a ubiquitin conjugating enzyme, and the remaining four proteins, including the F- box protein, are together called an E3 ubiquitin protein ligase. Ubiquitin is a 76-amino-acid protein and is highly conserved in eukaryotes. (Modified, with permission, from Zhou P. et al. [2000] Mol. Cell 6: 751–756 [copyright Elsevier Science].)
for the ubiquitylating machinery becomes one when artificially recruited by a modified F-box protein.
The experiment is analogous to various domain swap experiments we have encountered in our study of gene regulation. Swapping one DNA- binding domain for another, for example, preserves the function of a tran- scriptional activator, but changes the substrate (gene) it controls.3
SPLICING
Many genes, especially in higher eukaryotes, bear introns that are tran- scribed into RNA along with the coding sequences (the exons). These introns must be removed from the RNA to produce a functional mRNA. Just as there is an elaborate machinery for transcribing genes, so is there another machinery for splicing—indeed, some 50 proteins are dedicated to the process–and that process can be regulated.
Characteristic sequences flanking introns are recognized, weakly, by the splicing machinery. Splicing is activated by so-called SR proteins which bind to RNA at nearby “splicing enhancers.” Different splicing enhancers bind different SRs, just as different transcriptional enhancers bind different transcriptional activators. The following results show that the SR proteins work (i.e., activate splicing) by recruiting the splicing machinery to a near- by intron (Figure 4.2).
•A given splicing enhancer will work when placed downstream from any intron. This property is analogous to that of transcriptional enhancers, which work on any gene with which they are associated.
•Like typical transcriptional activators, an SR protein bears separable activating and nucleic-acid-binding domains. As for these transcrip-
FIGURE 4.2.Recruitment of the splicing machinery by SR proteins. (a) An SR protein is shown binding to a site (a splicing enhancer) in the RNA downstream from the intron to be removed and recruiting the splicing machinery to that intron. The machinery then apposes the 5´ and 3´ splice sites and removes the intron. (b) The RNA-binding domain of an SR protein has been replaced by that of MS2 (a bacterial virus protein), and the enhancer has been replaced by an MS2-binding site; recruitment of the splicing machinery then pro- ceeds as in a. (c) The Drosophila SR protein RBP1 binds to the so-called dsx splicing enhancer only in the presence of Tra1 and Tra2. Tra1 is expressed only in the presumptive female embryo, and this splicing does not occur in the male-presumptive embryo. (Modi- fied, with permission, from Graveley et al. 1999 [copyright Elsevier Science].)
dsx
splicing enhancer
tional activators, the role of the nucleic-acid-binding domain (RNA in this case) is simply to position the SR protein near the intron to be spliced, a point illustrated in the domain swap experiment of Figure 4.2b. In that experiment, the RNA-binding domain of an SR protein was replaced by a different RNA-binding domain, and the splicing enhancer was replaced by a sequence recognized by that new domain; splicing was found to proceed as usual.
•The activating regions on the SR proteins resemble their adhesive coun- terparts on eukaryotic transcriptional activators in that, as assayed using deletion derivatives, they work with an efficiency approximately pro- portional to their lengths.
•SR-activating regions expressed without RNA-binding domains do not activate splicing.
Many SRs are present ubiquitously. But some, like many transcrip- tional regulators, are active (or present) only under specified conditions, thus ensuring that the same RNA is spliced differently under different cir- cumstances. For example, the early Drosophila embryo develops into a male or a female depending on whether or not a specific intron is removed from a particular RNA molecule. This splicing requires the Tra1 protein, which helps a particular SR bind to the nearby splicing enhancer (Figure 4.2c). Presumptive females, but not males, express the Tra1 protein.