Recent studies indicate that the spindle attachment point of a chromosome, called a kinetochore, contains a number of proteins including two called Bub and Mad. A chromosome that is not attached to the spindle emits the Mad protein somehow, and Mad combines with a second protein, Cdc20, to form a complex that binds to and inhibits APC. In other words, as long as there are free chromosomes in the cell APC is inhibited. Once the chromosomes all bind to the spindle, they stop releasing MAD, and eventually the MAD/Cdc20 complex dissociates from APC, allowing APC to become active and degrade the anaphase inhibitor, which is called Pds1 in S. cerevesiae and Cut2 in S. pombe. This destruction of Pds1 in turn leads to inactivation of a protein called "Sister chromatid cohesion" (Scc1) protein which then allows separation of the chromatids and thus entry into anaphase.
Another protein that seems necessary for the conclusion of mitosis is called Cdc14, and it too is regulated by APC. Cdc14 is required to activate a number of proteins that are needed to break down or inhibit the cyclin B protein, that leads to termination of mitosis. The Cdc14 protein is normally inactive because it's sequestered in thenucleolus, a small structure in the nucleus where ribosomes are made. Degradation of Pds1 by APC allows the Cdc14 to escape the nucleolus (by a mechanism not yet understood), and once freed it acts as a protein phosphatase that dephosphorylates a number of proteins including Sic1, a protein that inhibits cyclin-dependent kinases, and thus helps move the cells out of mitosis.
The point here is not to memorize the details of these processes, because they are currently undergoing revision and elaboration. Rather, you should appreciate that both entry into, passage through, and exit from mitosis is a highly regulated process, requiring the interaction of many proteins whose activity is altered by phophorylation/dephosphorylation and by protein ubiquitination and degradation by proteosomes
In addition to getting into and out of mitosis, there’s a second important transition point for cells, when they "decide" to leave G1 and enter S phase; the commitment to replicate the DNA is essentially irreversible. Once a cell passes the start of S, it proceeds through G2 and M then back into G1, unless something goes dreadfully wrong. Of course a cell can be held in G1 for a long time (or more precisely, in G0); it has the proper diploid number of chromosomes, it can carry on all its normal metabolic functions, and it can carry on its specialized differentiated functions, such as contracting muscle cells, conducting electrical signals in nerve cells, and so forth. So a cell doesn’t need to divide to function, and the decision to move from G1 to S phase isn’t taken lightly. This decision point seems to mediated by cyclins and cyclin-dependent kinases, just as does the one at the end of G2. And the place it’s been best studied is in the yeast S. cerevesiae, where the notion of a START point was first crystallized. This transition point from G1 to S is said to require an S-phase promoting factor (SPF) analogous to MPF for entry into mitosis.
First let's consider the cyclins themselves. I’ve been talking as if there’s a single cyclin molecule called cyclin B. In fact there are families of cyclins--proteins whose levels rise and fall at different times during the cell cycle, and that act by activating cyclin-dependent kinases. There appear to be two broad groups of cyclins called the "regular" cyclins (abbreviated CLN) and the b-type cyclins (CLB). The regular cyclins seem to be active mostly between M phase and S phase--i.e., during G1, while the CLBs are more active during S and G2, and of course a cyclin B is part of MPF, essential for both the beginning and end of mitosis.
In yeast there seem to be three cyclins (Cln 1,2,3) active during the G1 phase and 6 active during S (Clb 5 & 6), G2 (Clb 1, 2, 5, 6), and M phases (Clb 1-4). In both S. pombe and S. cerevisiae, there seems to be a single cyclin-dependent kinase that is activated by all these different cyclins, called Cdk1; it’s encoded bycdc2 in S. pombe, and by CDC28 in S. cerevisiae. Higher organisms, like mammals, have about 6 cyclin-dependent kinases (cdks) only four of which seem to regulate the cell cycle.
At the end of metaphase, as you know, the levels of the cyclin b (clb) proteins in the cell fall drastically, as if they’ve done their job and are being cast off by the cell. In order for the cell to go through another round of division, it must rebuild the levels of these critical cyclins. It is believed that in yeast, the cyclin 3 is essential for this rebuilding process. The cyclin 3 promotes formation of cln1/2; when these reach a critical mass, the cell can enter S phase. That is, the activity of this kinase (cyclin 3/Cdc 28 = Cdc2) is a switch and unless it’s sufficiently active, the cell can’t enter S phase (see Fig. 13-26 in Lodish et al.)
So the first thing that has to happen for the cell to get out of G1 is that it has to accumulate a bunch of the necessary cyclins. But apparently that alone isn’t enough to send the cell into the next phase of the cell cycle, because there are inhibitors that prevent activation of the cyclin/cdk complex whose activity is necessary to move the cell from G1 into S phase. So before the cell can move ahead, the inhibitor must be removed. Removal of this inhibitor activates Cln 1/2 which turns on the production of Clbs 1 and 2, two cyclins that are required for cells to pass through G2. Another set of proteins Clb 5 and 6 are present but not active during G1, and they must be activated in order for DNA synthesis to begin. How does all this happen?
Using the yeast example, cyclin 3/cdc2 kinase seems to be essential for transcription of cyclin 1/2. The enzyme that degrades these cyclins seems to be constantly active, so the only way for them to be built up is for the rate of synthesis to exceed the rate of degradation. The activity of the cyclin 1/2/cdc 28 complex is tightly regulated. For example yeast have two mating types. If cells of both mating types are present they secrete hormones called alpha factors or a factors that bind to each other’s specific receptors (alpha factors bind to alpha receptors on a cells and vice versa). Binding of the receptor causes the target cell to cease dividing and to arrest at START. Cells at this stage can then fuse to make diploid yeast. How does the mating factor cause arrest of the cell cycle? By keeping the cln1/2-cdc28 complex in an inactive state. Binding of the alpha factor to its receptor activates a complex cell signaling pathway, involving a MAP kinase cascade that eventually phosphorylates a protein called FAR 1. FAR 1 can bind cyclin1/2/cdc28 and inhibit its activity, probably by phosphorylating the kinase subunit. In addition, another protein, called Sic 1, can bind to the Clb 5 and 6 proteins and keep them inactive. So the cell is stuck in G1 at Start.
How does a yeast cell get past Start? Well, the same basic process that sends cells from metaphase into anaphase works here as well; a protein degradation pathway is activated. A crucial step in this pathway is phosphorylation of a protein called cdc 34, which is an enzyme that attaches ubiquitin to other proteins. Phosphorylation of the cdc 34 protein makes an active enzyme that attaches ubiquitin to both FAR 1 and Sic 1 (Fig. 13-25), which causes their rapid destruction. Destruction of Sic 1 allows the cell to begin DNA synthesis and activates Clb5 and 6, and destruction of FAR-1 activates cln1/2-cdc2, which can turn on clb 1 and 2 as well as other proteins necessary to send the cell into S phase.
The point here is that movement of cells past Start seems to depend both on the accumulation of the appropriate kinase and on the removal of particular inhibitors. The kinases then became active and presumably can phosphorylate gene regulatory factors necessary to turn on the genes needed for DNA syntheses. The activity of these kinases is regulated by inhibitors that are themselves presumably regulated by factors such as cell size, mating factors, etc. At some point the things that activate these inhibitors are eliminated, the inhibitors are destroyed by an ubiquitin dependent pathway, and the cell can move ahead. In other words not only are the levels of the cyclins cyclic, but the levels of cyclin inhibitors are cyclic as well.
You can see there’s lots to do, if you’re interested. How many regulatory factors are there? What turns them on and off? How does the signal get from the mating factor to the Far 1 protein? What genes get turned on by the cyclins and how do they function to advance the cell through its cycle? And so forth. The activity of cyclins appear to be affected not only by their concentration in the cell, or by the inhibitory factors, but also by whether or not they are phosphorylated, and even where they are phosphorylated--e.g., on tyrosine or serine residues. Sorting out the details of this will also take a while.
Well, a great deal has been learned and is being learned from studies of yeast mutants, but people are anthropocentric--we’re more interested in ourselves than any other organism. So scientists have also tried to sort out the cell cycle in mammalian--if not human--cells. While much less is known in mammalian cells, it’s probably fair to say that the current take home lesson is that mammalian cells are just like yeast cells only more so.
Cells in a multicellular organism are faced with a somewhat different set of problems than are yeast cells or fertilized egg cells. Yeast or egg cells are best off if they can reproduce fairly rapidly but in a multicellular organism, some cells produce quickly, some slowly, and the overall health of the organism requires that no one cell type predominate. So cells must not only respond to internal signals but to signals that impinge on them from their environment--contact with other cells, growth factors, etc. This requirement for exquisite fine tuning of cell division is probably why the mammalian cell cycle--while similar in most respects to what we know about yeast--is more complex and probably more highly regulated.
As Fig. 13-29 shows, there appear to be not only a bunch of cyclins in mammalian cells, but also several cyclin dependent kinases--called either cdc or cdk in this figure. This cartoon masks some of the complexity of the process, because some of these cyclins are actually classes of cyclins. Cyclin B, for example, is actually a family of closely related proteins. So while the basic idea of what propels the cell cycle is the same as in yeast , the details clearly vary.
The study of the mammalian cell cycle isn’t easy to do in intact animals, so most work is done in tissue culture, where some cell types can be grown and maintained indefinitely. While this is obviously a somewhat artificial situation, it has had some advantages. First of all, one can control the environment in which the cells grow, adding and subtracting substances at will, to learn what is necessary for cells to grow, divide, stay alive or die. One consequence of this is that it was clear fairly early on that cells would not grow in a rich medium in culture unless the culture medium was supplemented with serum, the liquid portion of blood after clotting. It was subsequently learned that serum supplied rare proteins necessary for cells to divide.
These proteins were called growth factors as we have discussed previously--i.e., these are factors (proteins) necessary for the growth of cells in culture, and many of them have been discovered over the last 25 years or so. This result--that cells can’t divide unless they’re given a signal to do so from outside--suggests that unlike yeast cells, which will divide unless they are inhibited, mammalian cells require a positive signal in order to divide. We already encountered a number of these growth factors when we talked about signal transduction. So we now know in a general way how these factors act. They bind to receptors on the cell surface; the receptors are often tyrosine kinases that in turn activate a series of internal protein kinases--especially the mitogen-activated protein kinases--which presumably activate genes necessary for the cells to move from G1 to S phase. That is, there’s an internal cascade of enzymatically catalyzed events, mostly involving the phosphorylation of proteins, that activates the transcription of genes necessary for cell division. (Recall the mitogens are substances that cause mitosis, so it’s pretty clear that MAP kinases are involved in the cell division process.)
Most of what we know about control of the mammalian cell division cycle is phenomenology--that is, that one can describe what happens when you do some sort of manipulation and that tells you something about the details of the process. By withdrawing serum from cells in culture you cause their arrest at G1/S checkpoint. If you add back serum it takes the cells about 8 hours to get back into the cell cycle. That is, they are out of the cell cycle and in G0 if you withdraw serum. To get back in the cell cycle they must actively do something that probably involves making and assembling the cyclin/cdk complex that’s necessary to move beyond G1. This kind of experiment suggests that the cyclin/cdk complex is pretty fragile--that is, it’s easily broken down and hard to reassemble. In addition, experiments in culture show that the level of growth factors in the medium is limiting for cell division--i.e., different cells with receptors for the growth factor compete for whatever factor is available, and if they do not receive an adequate supply of it, they do not divide. That suggests one way to regulate cell division and organ size in vivo. Why--because if the amount of growth factor provided by the cells from which it comes is constant, then cells can divide as long as each cell receives more than the threshold amount of growth factor. But the dividing cells must share a fixed amount of growth factor, so as the number of cells increases, each individual cell receives a declining amount of growth factor until the amount falls below the threshold needed to induce cell division.
One of the reasons that people are so interested in the regulation of mammalian cell division is that it’s widely believed that cancer is a malfunction of the normal controls on cell division. We’ll talk about this in more detail beginning next week, but I want briefly mention the kinds of proteins that are defective in cancer cells. One kind we’ve already discussed--the so-called oncogenes that encode proteins in cell signaling pathways. It’s thought that mutations in these proteins cause cancer by permanently activating some signaling protein that normally is tightly regulated; that is, a protein that normally exists in both active and inactive states is mutated into the permanently active conformation.. One example of this is the Ras oncogene, which can mutate so that it does not hydrolyze GTP, and thus constantly activates the MAP kinase pathway. Many other proteins in this pathway have been found to be defective in cancer cells.
However, as you might guess from our discussion of yeast and Far1, there are also proteins in mammalian cells that are inhibitors of cell division. If these are mutated so that they become inactive, then a brake on division is removed and the cell can enter the cell cycle. Such a protein was found to be mutated in people who developed retinoblastoma, a tumor of the eye. This disease has a hereditary form that runs in families, and so it was possible to use this clue and the techniques of molecular genetics to identify the gene and its protein product. The Rb gene has been called a tumor suppressor, because in its normal, active form it prevents cell division. It does this by binding to a regulatory protein that is necessary for transcription of some of the genes required for cell division (see Fig. 13-31). When growth factors are present, the Rb protein is phosphorylated, by a cyclin/CDK complex, and the phosphorylated form is inactive. Thus it unbinds from the transcription factor, allow transcription to proceed, and producing the proteins necessary for cell division. So in normal cells the Rb protein and probably several others like it, are part of the mechanism that allows the cell to divide when appropriate growth factors are present, but that prevent division otherwise. When the Rb gene is mutated, it apparently can no longer bind to and inhibit the regulatory factor, even when it is not phosphorylated, so the transcription of cell division genes proceeds unimpeded, and the cell divides constantly, the mark of cancer.
No comments:
Post a Comment