So I heard about this PNAS paper all summer long but never got around to reading it until yesterday.
Neutrophils are white blood cells, whose job it is to chase around invaders that enter your bodily fluids. They can sense foreign invaders by sensing chemical traces. Thus the chemical traces are said to be chemoatractants for the neutrophils. A couple of weeks ago I showed you a movie of a migrating neutrophil chasing a bacterium. Here it is again:
It turns out that if you uniformly stimulate migration by giving the cells a uniform dose of chemoatractant, cells start to migrate in seemingly random directions. What Henry Bourne's lab did was to record the cell's movements to see if there was some hidden pattern to this stimulus, and what they found was truly bizarre.
First I have to explain to you what I mean by cell polarity.
All cells have some degree of polarity, that is to say that they have a front and a back. For epithelial cells, there is an up and a down - one side, called the basal-lateral side, which faces the inside of your body where all the blood and lymph flows, and another side, called the apical side, which faces the outside world, like the inside of the gut or the entrance to some duct. Other cells display different forms of cell polarity. Neurons have dendrites, which receive information, and axons, which send information. Migrating cells have a leading edge that pulls the cell forward and a tail which it retracts as it crawls.
In most migrating cells you can draw a straight line from the leading edge (top of the image on the right) to the tail and find that the line cross first through the lamella, which is filled with actin and generates most of the locomotive portion of the cell, then trough the golgi which directs the flow of secreted proteins and lies next to the centrosome (the red dots in the image on the right) from which all the microtubules (green in the image) emanate from. After the centrosome comes the nucleus (blue in the image) and finally the line would reach the cell tail. For neutrophils the leading edge and lamella is often reffered to as the pseudopod and the tail is called the uropdo. Now I must say that in certain cells the order is reversed in that the nucleus comes before the golgi and centrosome, but this order is highly dependent on the cell type.
When one of these crawling cells wants to move, it has to first position the centrosome, golgi and nucleus along the axis of migration. This repositioning of the major organelles is called polarization. This process is highly dependent on the microtubule cytoskeleton and on two specific polarization signaling pathways both triggered by the protein Cdc42. This very important signalling protein has two functions
1) It positions the centrosome by activating the Par6-Par3-aPKC pathway that results in an activation of the motor dynein. Activated dynein at the periphery pulls on microtubules which are connected to the centrosome and thus act to reposition the centrosome either at the centroid or towards the leading edge.
2) It activates actin retrograde movement - which consists of a stream of actin that moves away from the leading edge. This actin flow pushes the nucleus towards the back of the cell and helps the cell generate locomotive force.
OK so getting back to the PNAS paper ...
Before flooding the neutrophils with a uniform gradient of chemoatractant, the Bourne group drew an imaginary line through the cell's axis of polarity using the centrosome and the nucleus. 
Then they added the chemoatractant and recorded the cells' movements. In an incredible surprise ALL THE CELLS TURNED LEFT! Now note that they turned left with respect to their axis of polarity. So if the cell was facing north, after being stimulated it crawled west, if the cell was facing south, it crawled east. If the Par3-Par6-aPKC polarity pathway was disrupted, by depolymerizing the microtubules or by inhibiting aPKC or dynein, then the leftward bias disappeared. In other words the cells migrated randomly. But if the researchers introduced active GSK3-beta into the cells, another signaling molecule that is involved in multiple polarity pathways, the critters migrated TOWARDS THE RIGHT!
So what does this mean? At the very least, it would say that cells can tell the difference between left and right. Now if one thinks a bit about the implications of this result, one would realize that to properly analyze what is left/right, the cells would have to integrate information from the other two axises, namely the front/back and up/down axises. To tell the front from the back they clearly must compute the centrosome/nucleus axis. To tell the up from the down the cells must analyze which side faces the coverslip and which side of the cell faces the medium (indicating that integrin signalling must be involved). But then the cells would need some chiral structure that could compute the result of the third axis (left/right) ... now every protein in the cell is chiral, but the largest known cellular structure that preserves chirality on a large scale is the centrosome itself. Interestingly, ablation of the centrosome abolishes the leftward migration. Now since this procedure would also abolish the front/back axis you could argue that the effect was indirect.
All in all, of of the most bizarre findings of the last year.
- Log in to post comments
That's a strange story. Why was it published in PNAS?
Two reasons:
1) the result is so weird
2) unfortunately the numbers (i.e. how many cells turned left vs right) are on the low side - but the percentages are drastic and thus seem statistically relevant (as in 90% turned left)
My feeling is that if this result is true, many other groups will repeat these findings. Of course, that's the way science works ...
Thanks for the paper and explanation Alex.
Here's my question - if these cells are getting direction cues by being attached to a single surface, what happens if they were to attach "above" as well as below, or better yet, attach in all 3 dimensions? Would they lose the left bias? Would attaching "above" also introduce a "left" turn signal, (actually a right turn signal if looking from below), that would cancel out the other "left" signal, so that the cells would turn neither left nor right? It's hard to guess exactly what the physiological relevant situation is for this cell line - what types of surfaces to neutrophils need to attach to in the body - none? 1? 2? many? in 2D? 3D? It would be interesting to know whether polarized epithelial cell also exhibit such behavior, and perhaps easier to understand the physiological relevance if they do...
Very interesting...
The whole question of a cell crawling in a 3d matrix is a good one. I guess in that situation, only one dimension would be defined (front/back) so chirality would be impossible, just as ethane does not have any chiral carbons even if each carbon has the ability to become chiral if one of it's hydrogens could be changed.
The important result is that if migrating cells are placed in the appropriate environment where chirality could be expressed, they do display some chiral properties. The idea that crawling cells can display chirality is such a challenge to most people's assumptions that it is striking.
I have to say that cells that rotate their primary cilia are chiral as well. The epithelial layer defines the z (or apical/basal-lateral) axis, and the clockwise movement specifies the x and y axis. This chirality of ciliar rotation moves growth and signaling factors around our developing bodies and ends up specifying the chirality inherent in our body plan. Thus our hearts are on the left and our appendix(ces?) are on the right. Come to think of it, the axis of ciliary rotation also depends on the centrosome, so in this way these two forms of cellular chirality that I've described are related.
Hmmm Interesting. I hadn't thought about cilia in that way before.
I agree with you that these observations are quite revealing about how cells work, whether they exhibit chirality in their real environments or not...but it's a good point that body plan chirality does indeed suggest there might be some relevance in vivo.
I must admit that I haven't read the paper you are discussing but the cilia rotation argument caught my attention. Are these migrating cells ciliated or you are talking about primary cilia?
The primary cilia - this is a well know phenomena (ciliary rotation causes the left/right axis specification in vertebrates). I must added that most cells have a primary cilia, whether they be epithelial cells (that have microvilli), endothelial cells or migrating cells.