by
Abstract
A recent meeting on cellular asymmetry - how one end of the cell is marked as different from the other - illustrated that a cell may use a simple cue to begin the development of the asymmetrical cell. Understanding all the steps that follow this initial cue will require a great deal more work.
The question of cellular asymmetry is a simple one. How does an amorphous bag of chemicals develop a front and a back? The generation of a difference in one part of a cell (asymmetry) leading to the definition of a front and a back (polarity) happens repeatedly in development. Only when polarity occurs can the cell make an axon at one end and a dendrite at the other, or excrete digestive fluids only into the stomach and not into surrounding muscles, or split in two to produce two different cell types.
Participants at a recent two-day symposium held April 17-18 at Stanford University, "Making the Difference: Mechanisms of Cellular Asymmetry," detailed their approaches to this problem in bacteria, yeast, ciliates, fruit flies, worms, and mammalian cells. Their results - encouraging but by no means complete - suggest that the cell uses whatever is handy to get asymmetry going and may have more than one way to convert an initial mark into a permanent polarity change.
The field is clearly beyond the descriptive stage. By screening directly for perturbations in cell polarity (what Ira Herskowitz of the University of California, San Francisco, termed "the awesome power of brute force"), researchers have identified a number of proteins that are found at only one end of a cell. This impressive list includes: Ash1p (Herskowitz) and possibly Bud8p and Bud9p (John Pringle, University of North Carolina) in budding yeast; Numb, Prospero, Miranda, Staufen and Inscuteable in fly neuroblasts (Chris Doe, University of Illinois); and the Par proteins in worm embryos (Kenneth Kemphues, Cornell). Other strategies have identified asymmetric Notch in mammalian neuroblasts (Susan McConnell, Stanford), Na+/K+-ATPase in MDCK epithelial cells (W. James Nelson, Stanford), DnaB in Bacillus subtilus (Alan Grossman, Massachusetts Institute of Technology), and CtrA in Caulobacter crescentus (Lucy Shapiro, Stanford).
As researchers had hoped, proteins (and, in the case of PROSPERO and Ash1, the mRNAs) are located at one end of the cell and can often explain the divergent development of that part of the cell. But does this explain the genesis of asymmetry?
Doe thinks not. He described his discovery of Prospero and its anchoring protein, Miranda, but wanted more: "We've just pushed back the question to what localizes Miranda," he says.
Shapiro faces a similar problem with CtrA. She knows that CtrA turns on flagellum production and turns off DNA replication in Caulobacter swarmer cells. And as Caulobacter divides to form one swarmer cell and one non-motile, replication-competent stalk cell, CtrA is found only in the prospective swarmer cell. This can be explained. Proteolysis of CtrA is localized to the prospective stalk cell. How the proteolysis is localized, however, remains a mystery. "We're still peeling back the onion," says Shapiro. "But unless you have an entrée to the system, you will never get anywhere. Now we need to follow all the way back through from CtrA to the source of the asymmetry."
As the proteins pile up, the researchers have been struck by the lack of similarity between the different systems. "One of the amazing things that I find about this whole meeting and this whole field of asymmetry is that everyone is finding lots of molecules, and everyone is finding lots of different molecules," says Kemphues. "I'm waiting to see the homologies. They have to be there. Or maybe we all have different systems."
"The only nuts and bolts things that seem to be emerging is that microfilaments are needed in all cases," says Doe. "Right now that's pretty skimpy. We all scratch our heads and say it's amazing how little similarity there is right now."
There are some conceptual similarities, however, between the two most completely characterized systems: Nelson's MDCK cells and budding yeast. And Anita Sil of the University of California at San Francisco, who was one of the discoverers of Ash1p in yeast (figure 1), says that the field is not ready for detailed comparisons.  |
| Figure 1 |
In addition to defining mechanisms of localization, researchers are still struggling to define how asymmetric localization of determinants translates into asymmetric cell fate. Kemphues, for example, has wonderful localization data on his six Par proteins (figure 2), but little idea of how they actually work. Genetic evidence suggests that Par3 anchors the spindle in the anterior cell to prevent the 90-degree rotation that occurs in the posterior cell. But Kemphues has no idea how the proteins are initially localized (Par3 and Par6 at the anterior periphery of the embryo, and Par1 and Par2 at the posterior periphery), or how they effect differences in cell fate once they are there.
An aggregate of 30 years of work gave him six cloned genes that have various motifs but are short on mechanistic clues. "What I would like to have is a hypothesis of a cell biological mechanism that I could attack with experiments," he says. Lacking such a hypothesis, he says, "I am faced with working my way out step by step. We've got the Pars, now let's find what they bind to."
Figure 2
His first step is a Par3-binding protein called Pkc3. "Par3 is like Velcro - it brings other proteins together," says Kemphues. "Pkc3 says that, yes, we are dealing with a signaling complex, so it is a modest step forward."
Filling in the missing links may be easier in budding yeast, where there are good candidates for the landmarks at the two poles (distal and proximal) of the cell. These hypothetical landmarks are used to determine where the next daughter cell will emerge (the bud site). Pringle described his theory that Bud8p marks the distal pole as a bud site (as the mutant always buds at the proximal pole) whereas Bud9p marks the proximal pole. Both proteins are found at both poles when overexpressed, but Pringle suggested that timing and levels of expression may be critical to limit localization to the appropriate pole. The initial cue that localizes Bud9p may be a remnant of the previous site of cell division, whereas the cue for Bud8p may be the site of membrane delivery into the growing bud. Pringle and Herskowitz have not yet resolved whether these proteins are important for the localization of Ash1 mRNA, Herskowitz's daughter-cell determinant.
 |
| Figure 3 |
Nelson, however, is happy to look to yeast for inspiration. His latest work, described at the symposium and in the May 29 issue of Cell, involves a ~17S complex originally discovered in yeast. The yeast Sec6/8 complex is localized at the bud tip, suggesting that it restricts vesicle delivery to a single site on the plasma membrane. In MDCK cells, Nelson says, the complex moves to the site of cell contact soon after the cells stick together. And antibodies that interfere with Sec6/8 stop several proteins from being sent to this side of the cell (the basal-lateral side), although other proteins are still sent correctly to the apical side of the cell.
Although Nelson does not know what anchors Sec6/8, he described previous work that identifies E-cadherin as the initial trigger of asymmetry in these cells. Cultured MDCK cells naturally stick together because of the E-cadherin glue on their outside. Nelson discovered that, soon after two cells stick together, the basal-lateral protein Na+/K+-ATPase finds its way to the site of contact. A mesh of proteins, including E-cadherin, catenins, ankyrin, fodrin, and the ATPase, keep the ATPase fixed in place.
The cell is now asymmetric, but the asymmetry is not oriented in a useful direction. The directionality comes from extracellular matrix (ECM) secreted either by surrounding cells (in the kidney) or by the MDCK cells themselves (in culture). "The ECM tells the cell what is down and what is up," says Nelson.
Nelson tested his E-cadherin hypothesis by inducing fibroblasts - which normally put the ATPase all over their surface - to produce E-cadherin. The new cells clumped and put the ATPase only where one cell contacted another. "This is the ultimate experiment," he says. "You turn a nonpolarized cell into a polarized cell. You turn base metal into gold."
"All the cells have to do is to take a surface that is symmetrical and tap it," says Nelson. In this case that tap is the random collision and gluing of one cell to another. "As soon as the cell responds to that mark the cell will be polarized. It's a simple idea, not very controversial, and it was probably published a couple of hundred years ago, most likely in German."
If Nelson is right, all the brute-force screens will yield a simple answer to the simple question of how asymmetry is established. That is, the cell uses and builds on whatever happens to be around. It will, however, take years to fill out the entire story, detailing how the cell marks the site, reinforces the mark, and propagates the new polarity throughout the cell. As Stanford's Shapiro observes, "It will probably take the rest of my life."
William Wells, Ph.D., is a scientific journalist with Biotext, Ltd. in San Francisco.
To come.