Seething coalition

Like you and me, cells live in a complex world. Even the simplest bacterium does not exist in isolation but needs to know what is happening outside. Is there food out there? A mate? A threat? Any cell - from a skin cell to an amoeba - also needs to have a sense of identity. It needs to recognise what is part of 'it' and what is not. As humans, we gain self-identity and perspective on the rest of the world through family, friends, facebook and trips to a therapist - oh - and by having a brain and memory. But how can a little cell recognise and respond to stuff? Even nestled within your body, individual cells make a lot of their own decisions. I hate to tell you this, but you're not much more than a coalition of cells. Don't imagine that your brain - much less you - actually tells them what to do. Sorry. See I knew you'd need a therapist.

One way that cells can detect what is happening outside is to hang receptors, or triggers, through their membrane (a thin, fluid fatty layer that defines the cell). Each receptor can detect a specific substance (or class thereof) and initiate a signal cascade when the substance is detected (by touching or binding to the receptor) and then the cell responds. Or so the textbooks say. But this disingenuous simplification plods over some absolutely fascinating questions; when the alarm is triggered, how is the signal transmitted and how does the cell (no brain remember) figure out how to respond, and is really everything - every possible thing that could happen to a cell - meticulously planned and each little signal pathway waiting like a row of dominoes? Of course I can't address all those questions here and will focus on a very early stage in signal perception; receptor dynamics at the membrane. There is so much stuff going on outside, why is a cell not overwhelmed with information? If a receptor is like an alarm light - how do you change the bulb (obscure Red Dwarf reference) or shut the thing off?




a stupefying textbook image of receptors and some elicitors (triggers).



To really appreciate receptors it's best to be clear exactly what a protein is. Previously I've drawn the analogy between a gene (DNA) and a simple computer program that does a task such as 'making protein A'. The DNA program is first copied to messenger RNA and then processed into a protein by ribosomes. The ribosome reads the letters of the genetic code in triplets, such as AUG-CAA-GAC. The code is non-overlapping and discrete, this means that there is a clear 'start' point and a correct reading frame. If the example above were to lose the first letter then UGC-AAG-AC would have quite a different meaning. A four letter code read in sets of three gives 64 possible codons. Twenty amino acids are commonly used to make proteins, so we have more than enough codons for each. Thus some amino acids have several possible codons whilst others have only one.
Image from Wiki commons
But what is an amino acid? These molecules all share the same core structure shown above but have 20 different side chains (R above). The side chains give different properties: Some amino acids are acidic and I think of them as 'tangy (citric acid, vinegar - these are of course not amino acids merely familiar acids), or basic (soapy), hydrophobic (water-hating or oily) or hydrophilic (water loving, wet) or small or large. So the coding part of a gene (the messenger RNA) is read to make new chain of amino acids. Proteins range in size from around 30 amino acids up to huge great mamas of 1,000 amino acids or more. Peptides are the also chains of amino acids but are smaller generally less than 30 amino acids. The distinction is somewhat fuzzy and context-dependent.
For a brief diversion, I have squandered a little time and some great computational resources to demonstrate the genetic code and the structure of a peptide. The DNA sequence 'gcgaacgatcgcgaatggtaa' encodes the peptide ANDREW (using single letters to represent amino acids) or Ala-Asn-Asp-Arg-Glu-Trp (if you prefer 3 letters for your amino acids). This peptide has the molecular formula: C33H47N11O12 and a molecular mass of 789.7928. Below are some representations of ANDREW. The image on the right is a 'wireframe' model, showing bonds between the atoms as bars, the grey balls carbon atoms, red oxygen, and blue nitrogen and white hydrogen. In the middle is the same model as 'spacefill' showing the relative sizes of the atoms and the overall 'shape' of the peptide. Finally there is a disappointingly simple ribbon representation that makes the underlying structure more obvious; here you can see that the hydrophobic W is predicted to coil around, giving this short peptide a twisted structure.

OK so lecture over - receptors are proteins that, in these examples, hang through a membrane hoping to bind to a trigger substance. They come in all shapes and sizes and generally (but not always) have three main parts; the external (signal receiver), the bit that crosses the membrane (with more hydrophobic amino acids) and the internal part (aka alarm bell). Some are quite beautiful - to my eyes at least - and below is one of my favourite external domains (a leucine rich repeat, LRR) that through a series of beta sheets (in yellow) folds into a horse-shoe-like structure. I find it incredible that such precise and intricate 3D structures arise from a delicate balance of amino acid combinations (sometimes folding is assisted by other proteins). Some receptors thread back and forth through the membrane, whilsts others (like a typical LRR recpetor) do so just once.


Receptors play so many roles in the life of a cell that it is with great restraint I have picked only two examples; cell movement (motility) and cell defence against a bacterium. Quite a disturbing piece of biology is my first example: Primordial germ cell migration. Germ cells make gametes (eggs and sperm) and bizarrely can be regarded as immortal because divisions of these cells pass through the generations, whilst the rest of the cells in your body will die with you. Gametes are formed during embryo development and the cells which are destined to become germ cells - primordial germ cells, PGCs - migrate from all over the embryo to congregate in the appropriate place. I am equally fascinated and repulsed to learn that PGCs squeeze through the developing gut to reach the gonads. Ewww. I don't know why other cells develop in the correct location and PGCs do not. Anyway, to migrate correctly the misplaced PGCs must know where they are headed and they follow a signal. More tractable cells, which are already at the place where gonads will develop, secrete a chemical called SDF1. As I'm sure you've already guessed, PGCs detect SDF1 by using a receptor.

To interpret the signal correctly the cell uses a couple on neat little tricks. Firstly the PGC must develop polarity, that is one part of the cell must become 'forwards' and secondly the cell must accurately gauge where is signal is coming from. The first part seems quite easy. SDF1 receptors are distributed equally over the surface of the membrane (well according to Minina et al ) so the cell 'runs' towards the signal. Most times the cell will not be pointing directly at the source of the signal and when the signal intensity diminishes the cell tumbles for a while to re-orient itself. One can imagine a program something like 'if concentration SDF1 greater than amount x then run forward, else tumble.' Such running and tumbling is a classic model for cell migration for many cell types any you may well have encountered it during high school biology lessions.

of course this image has beeen (gratefully, but possibly illegally) borrowed from Minina et al.

But this program has a problem, when moving towards a signal of increasing intensity as more and more parts of the cell surface receptors will recieve a signal >x. So the response needs to be attenuated as signal strength increases - or alternatively, the part of the cell surface receiving the most intense signal must be recognised. PGCs pull activated SDF1 receptors into the cell, by engulfing them in a little sack of membrane - a vesicle - and then destroy them. PGCs are polarised during a run, with the leading edge of the cell a veritable frenzy of endocytosis (internalisation vesicles), but during a tumble they lose this structured edge and it re-forms as a new direction is selected. Minina et al. showed that if you interfere with internalisation (and destruction) of SDF1 receptors (by creating a mutant receptor protein) then the PCGs can still head in the right direction but overshoot their target. It is not hard to imagine how these overactive receptors are bad news for any developing gonads.
De-sensitisation of activated receptors by internalisation (endocytosis) and degradation seems to be a widespread mechanism used in regulating responses to neurotransmitters, growth factors and many other stimuli. But there is some emerging evidence that endocytosis can be used in the opposite manner - to enhance active signalling. Here I'll switch over to plants and their response to pathogens.
A popular plant for research is Thale or mustard cress (Arabidopsis thaliana) and one of the best studied plant receptors is FLS2, a LRR that recognises part of bacterial flagellin. Many bacteria use a whip-like structure (flagella, think 'tails') to move around. Flagella are very similar between many species of bacteria - because it is a complex machine and not easy to change without breaking it - and therefore makes a good target for plants (or indeed animals) to recognise as 'non-self'. The flagellum is made up of repeating units of protein and FLS2 binds to a conserved part of this protein. FLS2 usually occurs at low concentrations in the plant plasma membrane (that outer layer; ignoring for now the plant cell wall). Proteins in a membrane are quite mobile; zipping around busily looking out for stuff to interact with. This obviously makes sense as there are very many things a cell needs to be aware of there is a limited about of time and space to make and present receptors to. When FLS2 hits a bit of a bacterial flagellin several changes occur, other proteins (from within the cell and within the membrane) also bind to FLS2 to form an active signalling complex. FLS2 also becomes less mobile (maybe to aid recruitment of these other proteins) and is then internalised.
Niko Geldner and Silke Robatzek suggest that FLS2 could actively signal from these internalised vesicles (a similar thing is known to occur for some animal receptors) and this could enhance signalling. A plant cell is very much larger than a bacterium, so rather than the PGC swimming in a diffusion of signal, the first sign of possible bacterial infection is rather subtle. Rather than destroying the active signalling complex to avoid over-stimulation it could be a smart trick to use the internal vesicle to provide a more sustained signal. Generally the signal needs to reach the nucleus (so that new defence genes can be switched on) and there is no reason why a bacterium should attack a cell conveniently near the nucleus. Internal vesicles can move around a cell, perhaps increasing response time and maybe making more room on the outer membrane for a few more FLS2 receptors to confirm the finding of the first.
So there you have it - a couple of ways cells figure out where to move to or how they (start) to respond to events like bacterial attack. Today I can personally attest to the importance of dealing appropriately with an alarm signal. I have written this post is a state of some distress because a neighbouring building has had an alarm sounding for exactly 12 painfully continuous hours (so far). The irresponsible fools have not provided the police with contact information, therefore their alarm is utterly useless because short of breaking into the building to shut the damned thing off there is nothing useful we can do. It is likely that I will be writing more later tonight; I apologise in advance if I am less good-humoured that usual.