I work on T cells. These are a type of white blood cell, or leukocyte (leuko for white, cyte for cell). Specifically, they are immune cells: they function to fight pathogens – entities your body recognises as non-self that could cause you harm. These might be tiny viruses, bacteria, fungi or parasites; even cancer. Some of these infect cells, and all of them can cause disease. Because these things can happen anywhere in your body, you need a cell that is highly motile to identify it as a threat (or not! As we will see).

Enter T cell. T cells are the type of white blood cell that patrols your entire body and carries out checks, to ensure that you are free of infection. They are the great orchestrators of the immune response, and once they have discovered something wrong, they call on an army of other cells to eliminate the threat, and leave the rest of your cells intact. Distinguishing between self and non-self is the basic job of the T cell.

Your body is massive, though, and T cells are really small. A T cell is about 15 um3, a human body is around 664 MILLION um3. That means each T cell has to cover a lot of ground each day to try to stop you getting ill. So, T cells: highly motile.

How does T cell migration work?

Your body is laced with vessels to speed up transport of blood cells to certain areas. T cells flow through the blood, detect chemical signals towards potential sites of infection then adhere to the cylindrical blood vessel walls. Once they are out of the main thrust of the blood flow, they start crawling along these walls towards the target site. The T cell then squeezes through gaps in the wall, into the tissues and identifies/deals with the threat. Central to this process are ‘integrins’ and ‘actin’ and this sequence of events is what I’m interested in.

Each T cell (each cell, indeed) has a cytoskeleton made of actin. Literally a cellular skeleton made up of this actin protein. Unlike your skeleton, this skeleton is dynamic. It is an intricately regulated web of fibres that allows the T cell to shift and change its shape. If you looked at just the bottom of a T cell as it moves, you would see this web of fibres undulating and flowing like a tank track, from the front of the cell towards the back. This constant flow of actin is what acts as the motor.

With this actin motor comes integrin. This is a hook protein, thousands of examples of which protrude from the bottom of the cell. With the flow of the actin cytoskeleton happening constantly, it is these hooks that grip onto the blood vessels, link to the flowing actin, and cause forward movement. Think of it like a car, the engine is constantly turning over (the flow of the actin cytoskeleton), but it is only when the clutch is engaged that the wheels turn round and the car moves forward (integrin engagement). This is called, rather prettily, the ‘molecular clutch’.

So, very simply, that’s how T cells move. But to understand it more deeply it’s useful to look again at some sizes. If you thought a T cell was small, integrin proteins are smaller at about 5 nm3 apiece. Compared to the area of T cell that is in contact with the blood vessel (~30 000 nm2), asking a single integrin to move an entire cell would be like trying to move the titanic with a single toothpick as an oar. In order to get enough of a grip, therefore, these integrin hooks have to work together, or – and here’s where we come in – form clusters.

Protein clustering

Get enough of these hook proteins grouping together in a small area of the cell membrane, all engaging with the actin cytoskeleton as well as links or ‘ligands’ on the blood vessels and you get forward cell movement. Clusters of integrin increase the force translated from the substrate (the blood vessel wall) to the cytoskeleton (the constantly flowing cell motor).

Play with these clusters even a tiny bit by increasing their size, decreasing their density, location or dynamics and you get huge changes in the way the T cells move. This then affects, of course, their ability to correctly traffic around the body, get to those sites of infections and interact with many other cells in specific ways.

The next idea I want to get across is that it isn’t just integrin that clusters. In fact, you would be hard pressed to find a protein that doesn’t rely on some sort of nanoscale location based functioning.

One example is “Lck”: an effector protein linked to integrin signalling. It is a great example of a protein that is always ‘on’. So it isn’t just the activation status of the protein that matters (whether it is on or off, or in an intermediate state) it is also the location in the cell that matters, and not only this but also the nanoscale clustering status of the protein.

So we are now beginning to appreciate that it is the nanoscale changes within microscale cellular compartments or locations that are at the root of cellular function. But if they are so small, how do we image them?

I use super resolution microscopy to investigate protein clusters

A couple of years ago a group of techniques together called ‘super resolution microscopy’ earned Eric Betzig, William Moerner and Stefan Hell the Nobel prize in chemistry. These allow us to image single molecules within clusters of proteins, using more or less traditional fluorescent labelling (tagging something you want to look at with something that lights up), at 10 to 20 nm resolution.

With techniques like STORM or PALM, it is now possible to detect nanospatial changes as described above, in specific protein populations. In combination with our new clustering algorithm that makes use of Bayesian statistics to reduce human decision making, we can reliably quantify them, too.

With this I hope to characterise T cell movement in terms of the nanoclustering of integrins and engagement of actin as they flow, crawl and otherwise manoeuvre through different regions of the body, interact with other cells and lastly when they go wrong and start attacking your own tissue (autoimmune disease) – all by mimicking these physiological environments in the lab.


  • Super resolution microscopy
  • An easy-going guide to Bayesian statistics and cluster analysis
  • What happens to integrin clustering and cell migration when your T cells turn against you

Written by Michael Shannon