Natural killer cells

The role of natural killer cells is to attack cells of the body which have become infected by viruses, or which are damaged or dysfunctional in some other way - e.g. cancer cells. The role of natural killer cells in tumour suppression is certainly very interesting, but I will try to focus on their role in defence against pathogens.

To put it very briefly, viruses invade the cells of the body and replicate within them; they hijack the resources and structures of the cell they invade. These are supposed to be used for creating things the body needs and new copies of the cell; instead, the virus repurposes them to create copies of itself. This is how viruses work - very simply. HIV is no different and we will see more of this soon. It is the role, then, of natural killer cells to identify cells which have been infected and to shut them down and kill them. This will destroy the virus and prevent it from replicating; although it will not, of course, undo the creation of any copies of the virus which have already been made and released.

In addition to this, natural killer cells are also able to release cytokines and there is emerging evidence that they are able to adjust to their environment and to "remember" pathogens which have been encountered previously and respond effectively. Thus they seem also to be a part of the adaptive immune system, as well as the innate immune system.

Precisely how natural killer cells achieve all of this is the subject of much current research. We do not yet have all the answers - although that's true of pretty much everything. In this case, though, we don't even have many of them. Nevertheless, for those who are interested, there is a very good paper by David H. Raulet called Missing self recognition and self tolerance of natural killer (NK) cells,¹ which I highly recommend and which is able to shed some light on the topic.

For our purposes, though, I think I can be a little cursory, but be sure to take note of the fact that I am simplifying things and eliding a lot of detail. NK cells possess a wide range of receptors - although not all NK cells possess the same receptors. Some of these receptors - such as NKG2D - are stimulatory receptors, while others - such as KIRs (Killer-cell immunoglobulin-like receptors) - are inhibitory receptors. Now, there does seem to be substantial variation in receptors among NK cells, so all NK cells are different and there will, therefore, be some variation in how they behave.

The general principle, though, seems to be that activation of NK cells depends on the balance between stimulatory and inhibitory ligands on a cell surface, or in the environment of a cell. So, if a given cell expresses lots of inhibitory molecules, the NK cell will obviously not be activated. If there is a good balance between stimulatory and inhibitory molecules then, again, the NK cell will not be activated. Again, if there are very few inhibitory molecules (or none), but also few stimulatory molecules (or none), still the NK cell is not likely to be activated. No, one requires the presence of plenty of stimulatory molecules - without there being enough inhibitory molecules to counter-balance them - for the NK cell to be activated. Or, from the NK cell's point of view, the stimulatory signal needs to be stronger than the inhibitory signal.

This, as I have said, is a simplification, however. Some cells do not actually express inhibitory receptors at all. These cells, I gather from Raulet (2006), are hyporesponsive (i.e. less responsive than the norm). I believe these cells can be stimulated to kill, but it is likely that either a strong signal is required, or their activation is dependent on external signalling, perhaps by IL-12 and/or other cytokines.

Now, let's look at this in a bit more detail. The inhibitory receptors generally recognise MHC class I (major histocompatibility complex class I) - something else which will soon be of great importance to us. MHC class I allows the NK cells to recognise "self" cells. So, the presence of MHC class I generally inhibits NK cell activity, which is an effective way of ensuring that NK cells do not attack the body's own cells. When cells become stressed or infected by viruses, MHC class I expression can be affected. In other words, many cells which are infected by viruses (and many cancer cells, too) do not express sufficient MHC class I (if any) to prevent NK cell activation. This is referred to as missing self recognition. NK cells identify damaged and infected cells by the lack of self molecules.

However, this on its own is not enough. NK cells generally do not attack the body's own cells when they are healthy because of the presence of MHC class I. However, absence of MHC class I will not precipitate an immediate attack. Instead, the presence of stimulatory ligands is required. These include heat shock proteins, extracellular matrix fragments, altered membrane phospholipids and other general markers of stressed, infected and cancerous cells. These are expressed on the surface of the cells, but stressed cells can also release stimulatory cytokines and NK cells can be stimulated by macrophages, too.

Seemingly, some molecules present on normal, healthy cells also act as stimulatory ligands. Thus, if MHC class I is poorly expressed on these cells they will be attacked. This is useful, because it means that some unhealthy cells can be killed merely by missing self recognition and expression of additional stimulatory ligands is not necessary, since stimulatory ligands occur naturally on that cell anyway.

So, in conclusion, this balance between inhibitory and stimulatory signalling allows for self-tolerance - the body's NK cells do not attack the body's own healthy cells. Meanwhile, infected cells can be "recognised" by NK cells by virtue of the stimulatory ligands which they produce and the absence of inhibitory ligands, particularly MHC class I (missing self recognition). Notice that it's really a combination of these two factors and, in fact, it seems that the triggering of NK cell activity is dependent on the end result of the interplay of numerous signals. Indeed it may well even be dependent on the particular NK cell in question.

To pick up on that last point, one final thing to add is that, in the case of cells which do not express inhibitory receptors, it appears that hyporesponsiveness is important for self-tolerance. In other words, if these cells were not hyporesponsive, they would regularly attack healthy self cells, as they would not be able to detect the presence of MHC class I and so be prevented from attacking. Thus it is necessary that they be very loath to attack and, in this way, they only attack cells which are strongly requiring of attack.

Now, there are a few ways in which NK cells actually go about killing abnormal cells. The primary method involves the release of granules. The NK cell will release perforin - a protein which creates a small pore in the target cell's membrane. Other molecules from the NK cell granules - generally proteins and proteases - can then enter the cell through the pore. These will then induce apoptosis in the target cell, or kill it via lysis, which we've seen before. Apoptosis, or "programmed cell death," is where a cell shuts itself down in response to certain signalling events. NK cells also release a number of cytokines themselves, most notably IFNγ and TNFα.

Finally, in addition to all the above, NK cells can perform antibody-dependent cell-mediated cytotoxicity (or ADCC). NK cells express FcγRIII (alternatively: CD16) receptors which bind to some antibodies (specifically IgG, I think). When this happens, the NK cells are activated and will induce apoptosis in the cell which has been opsonised with antibodies. This is, of course, in many ways similar to activation by stimulatory molecules and subsequent degranulation. It is important to note, however, that not only do antibodies activate NK cells, but they also seem to be able to expedite the process. Whether or not ADCC can completely bypass the usual mechanism involving other stimulatory receptors and inhibitory receptors is not clear to me. It's worth appreciating, though, another example of just how the adaptive immune system promotes and fine-tunes the innate immune system.

(a) An image of two natural killer cells attacking a cancer cell. (Image courtesy the NHS²)

(b) A schematic representation of a natural killer cell. (Image courtesy "A. Rad" (via Wikipedia))

Figure 1.57: The natural killer cell

¹Raulet, David H. 2006. "Missing self recognition and self tolerance of natural killer (NK) cells." Seminars in Immunology 18 (2006): 145-150
²http://www.nhs.uk/news/2008/05May/Pages/Antidepressantsandimmunity.aspx