This approach, originally pioneered by Leland Hartwell at the University of Washington was to find yeast cells that normally divide in response to a stimulus, then isolate mutants that can’t respond or whose response is defective. One guess would be that the proteins encoded by the mutant genes are necessary in some way for cell division. This may generate a lot of uninteresting mutants, such as mutants in microtubules that can’t separate their chromosomes; this is uninteresting because it doesn’t tell us anything that we didn’t already know--chromosome separation requires microtubules--but other mutants might affect proteins that nobody knew about.
So what did people learn by taking these kinds of approaches to studying the cell cycle? First of all they isolated a very large number of genes that were involved in the cell cycle, dozens in fact. Unfortunately, since they didn’t know what these genes did, they started numbering them as they were discovered....cdc1, cdc2, cdc3etc. (cdc means cell division cycle). Some people did this in budding yeast (S. cerevisiae) and some did it in fission yeast (S. pombe). They used different numbering systems; for example CDC 28 in S. cerevisiae is equivalent to cdc 2 in S. pombe. So the terminology has been inconsistent and confusing. Mostly what people have been doing since isolating all these mutants is trying to figure out what they do; that is, what is the function of the proteins that these genes encode?
Studies in S. pombe, and in Xenopus oocytes revealed some of the great complexity of the events necessary for cells to enter mitosis and successfully pass through it. Last time I spoke about MPF, which in oocytes is a dimer of cyclin B and a cyclin dependent kinase, and whose level in a cell must rise before the cell can enter mitosis. Thus one control on whether a cell enters mitosis or not is simply the amount of MPF (or more precisely the amount of cyclin B) that is present. Several cdc mutants in S. pombe have been isolated that involve either these proteins (i.e., ones similar, or homologous, to their frog counterparts) but many mutants involve other proteins that regulate the activity of MPF. That is, the activity of MPF is regulated by other processes besides simply variations in the level of cyclin B.
The fission yeast equivalent of cyclin B is called Cdc13, because this protein was encoded by the 13th cdc gene found (i.e., cdc13; the names of genes are written in lower case letters in italics; the name of the proteins they encode is written in unitalicized letters with the first letter capitalized. Thus the gene cdc13 encodes the protein Cdc13). The S. pombe equivalent of the cyclin dependent kinase from frog oocytes is called Cdc2, and together Cdc13/Cdc2 make up the yeast MPF.
Several mutants were isolated that affected yeast cell size--some caused much larger than usual cells and some caused much smaller than usual-sized cells. It turned out that these altered the length of G2. Some shortened G2 so that the cells entered mitosis prematurely before they were large enough, so the progeny got smaller and smaller over time. These were called wee1 mutants. Others had abnormally long G2 stages, which means they didn’t enter mitosis at the appropriate time, and so grew abnormally large; one such mutant was cdc25. Both Wee1 and Cdc25 turned out to be enzymes that could affect the phosphorylation of MPF, but in contrary ways with contrary effects.
The combination of Cdc13 and Cdc 2 forms an inactive form of MPF (Fig. 13-13 in Lodish et al.). Wee1 phosphorylates it on a tyrosine residue of Cdc2, tryosine #15. This "locks" MPF into an inactive conformation. This form can be phosphorylated by an enzyme called CAK which puts a phosphate on threonine #161 of Cdc2. (Clearly, both Wee1 and CAK are protein kinases). This doubly phosphorylated form of MPF is also inactive. Finally, however, the MPF can be activated by removal of the phosphate that is added by Wee1; this is carried out by Cdc25, which is a protein phosphatase. This exposes the active site of the cyclin dependent kinase which allows it to phosphorylate its own substrates. As you can infer, underproduction of Wee1 by a cell will result in MPF that is activated even without Cdc25, causing premature mitosis. A cell without Cdc25, on the other hand, will have only phosphorylated, inactive MPF, and thus never, or belatedly, enter mitosis. Moreover, I suspect that it won’t surprise you to learn that the activity of both Wee1 and Cdc25 are themselves regulated by other proteins, both in fission yeast and homologous proteins in oocytes. Phosphorylation of Wee1 inhibits its activity, while phosphorylation of Cdc25 is required for it to be active. Thus other protein kinases regulate the activity of Wee1 (a protein kinase) and Cdc25 (a protein phosphatase) which in turn regulate the activity of MPF, a protein kinase essential for initiation of mitosis.
This repeats a common theme in modern cell biology. The activity of many cellular processes is controlled by protein phosphorylation/dephosphorylation reactions. We’ve discussed several when we talked about signal transduction mechanisms, and here we see the same basic scheme shown to be essential to the process of cell division.
What are the substrates that must be phosphorylated by MPF to induce mitosis? Not all of them are known, but several have been indentified, including nuclear intermediate filament proteins called lamins, microtubule associated proteins (MAPs), myosin light chain, and the anaphase-promoting complex (APC), an ubiquinating enzyme that we discussed earlier.
One of the early events in mitosis is the dissolution of the nuclear membrane during prophase. This appears to require phosphorylation of the lamins by MPF; you may recall that lamins are important in maintaining the structural integrity of the nuclear membrane, but when phosphorylated, the filamentous lamin structures dissociate into monomers of lamin, and this leads to vesiculation (breaking up into small vesicles) of the nuclear membrane. (These vesicles retain one kind of lamin, called lamin B, which is probably what marks them as derived from the nucleus, destined to reassemble again into nuclear membranes at the end of mitosis.) This was elegantly shown in a series of experiments in Frank McKeon’s lab in which mutant lamins, that lacked a phosphorylation site, did not dissociate during mitosis because they couldn’t be phosphorylated, so the nuclear membrane remained intact (Fig. 13-16).
A second important substrate is APC (anaphase-promoting complex). APC is phosphorylated and activated by MPF, and it in turn ubiquinates cyclin B (Cdc2), targeting it for degradation. I already mentioned that this was necessary to end mitosis. But APC has an earlier essential function in mitosis as well (Fig. 13-19). There is a protein that blocks movement of the cell from metaphase into anaphase, the transition that is characterized by the separation of sister chromatids in early anaphase. As we discussed earlier, this shouldn’t occur until all the chromosomes have aligned properly at the middle of the mitotic spindle. To prevent premature separation of the chromosomes before all proteins are aligned, there is a protein, called "anaphase inhibitor" that blocks the initiation of anaphase. This protein must be removed in order to allow anaphase to begin, and it is removed by degradation, after it has been ubiquinated by APC. The cell then enters anaphase and telophase, but doesn’t complete mitosis until cyclin B is degraded, which is also promoted by APC. Thus, initiation of mitosis is caused by buildup and activation of MPF, the cyclin B/cyclin-dependent kinase complex. This begins the process of mitosis by phosphorylating, and altering, the activity of many cellular proteins, but completion of mitosis actually requires removal, i.e. degradation, of some proteins, including cyclin B itself.
Finally, and briefly, phosphorylation of MAPs interferes with microtubule function, and halts vesicle traffic in the cells. It also causes the ER and Golgi to fragment into small pieces which are apparently randomly and passively distributed to the daughter cells. (It’s probably easier to distribute lots of small bits of ER and Golgi membrane between two cells than large bits which might end up predominantly in one cell or another.) And phosphorylation of myosin inhibits its function and prevents premature formation of the contractile ring around the center of the cell before telophase. Only when myosin is dephosphorylated at that time can it interact with actin to "pull the noose tight" around the membrane separating the two new nuclei, and thus lead to formation of two separate daughter cells.
To summarize, MPF is in some sense the primary controller of mitosis. It is necessary for cells to enter mitosis, and it must be degraded in order for cells to leave mitosis. Its activity is regulated in a large number of ways, by alterations in stability and by alterations in activity mediated by phophorylation and dephosphorylation reactions. MPF in turns regulates the activities of many cellular proteins necessary for proper functioning of the cell during mitosis, including separation of chromatids, dissolution of the nuclear membrane, movement of organelles, and contraction. The decision to enter mitosis is an irreversible one for cells, so it is perhaps no surprise that this process is subject to a tight interrelated cascade of controls. What should also not be surprising is to learn that the passage into S phase, itself an irrevocable decision is similarly tightly controlled, as shown initially by genetic studies in S. cerevesiae, which we’ll take up next time.
Now, before we go on, I’d like to step back from details a minute to indicate some general principles that seem to govern the passage of cells through their cycle.
1. What causes a cell to move from place to place in the cell cycle is the sequential activation of cyclin dependent kinases (there are only a few of these in yeast but many in mammals).
2. Activity of the cdk depends on their binding to cyclins; the concentrations of different cyclins rise and fall at different times. The resulting activation or inactivation of their partner cdk moves the cell through a new stage of the cell cycle. Presumably this means that the cdks in turn activate essential components in the cell for movement through the cell cycle. In particular the activity of a cdk/cyclin pair probably activates the next cyclin/cdk combination in the sequence.
3. Activity of various cyclin dependent kinases is highly regulated, positively and negatively.
a. Binding cyclins activates cdks, and binding cyclin kinase inhibitors (CKIs) turns them off.
b. In addition the activity of cdks and cyclins can be activated by phosphorylation or inhibited by phosphorylation.
c. Removal of a phosphate by phosphatases can reverse the effects of phosphorylation.
d. Finally, the levels of the cyclins can be altered by changes in the rates of synthesis (transcription) or degradation of these proteins.
4. While any movement through the cell cycle seems to require the cyclin/cdk complexes, there seem to be two major points at which the cell pauses in the cycle to take stock of itself.
5. One of these is as the end of G1, before the cell enters S-phase, and is called START.
6. The other one of these is at the end of G2, before the cell enters M phase. These two places in the cell cycle are sometimes called checkpoints.
7. The proteins that are responsible for moving the cell past START and into M phase are the same in yeast and different but very similar in higher organisms. Of course MPF is the factor for entry into mitosis. Mutations that block the activity of MPF or its homologs cause cells to arrest in the cell cycle at one or the other of the checkpoints.
8. The cdks can regulate not only other proteins such as transcription factors, but also their own activity. For example MPF seems to be involved in a positive feed back loop so that activation of some results in a dramatic increase in the amount of MPF present in the cell.
No comments:
Post a Comment