(Phlogiston version July 1997)


I have now had a chance to digest the "Models of immunological tolerance" debate. I have a number of points that I think are worth noting. The overview is this. Should it be possible to get a glimpse of the real raison d’être, the grand plan, of the system (morphostatic system for me and immune system for most if not all of you), then we could make some shortcut decisions about what each element has evolved to do. This summary is an attempt to encapsulate 3 articles into a brief summary. It is my opinion.

  1. "Nothing in biology makes sense except in the light of evolution" - Dobzansky.
  2. Let’s look at the sequence of "shells" that (eventually) form the mammalian immune system. The inner shells remain fully functional and continue to fulfil their original purpose. None are abandoned. They are layered around each other like the layers of a Russian doll
  3. Heat shock proteins identify then attempt to repair protein trash; ubiquitins escort terminally malformed proteins off the premises; endosomes chop them up into short peptides.
  4. Elective suicide (a forerunner of apoptosis) evolves in colonial bacteria as a mechanism to limit the spread of rampant disease through the colony.
  5. CAMs (cell adhesion molecules), SAMs (surface associated molecules) and ICJs (intercellular junctions) evolve. These give cells a sense of "belonging" (to a colony) and enable them to dock with one other. This is a discrimination of self from non-self. It remains the fundamental root of S/NS discrimination in mammals.
  6. Homeotic genes allow cells to adopt alternative (cellular) body plans. When gap junctions evolved (c700my ago), they enabled alternative, homeotic body plans to expand their sphere of influence to cover blocks of cytoplasmically interconnected cells - enlarging the body plan.
  7. A primitive metazoan will require the following of its component cells: when a cell’s well being is compromised it will be expected to detach itself from the colony. It can re-attach provided it has fixed the damage. If not, it will be expected to do the decent thing, sanitise its contents by trashing them and then be gently re-absorbed by the surrounding cells. Adjacent cells will replicate to fill any gaps. Individual cells are entirely responsible for maintaining their own health. If they can’t identify that they are malfunctioning and go on to spill their cytoplasmic contents, they can expect no mercy from the surrounding colonial (self) cells. Such sick cells are dangerous and will be aggressively attacked then hoovered up - together with any of their spilled contents.
  8. What has changed in mammals? Nothing - as far as the grand plan is concerned. But, the system has since been elaborated to enhance the intracellular destruction of pathogens, particularly by promoting apoptosis in cells that sport a caricature resembling those that spilled their contents on a previous occasion (more later).
  9. Phagocytes pay minimal attention to cells that are in junctional communication with each other. These are "stable" and "healthy" cells. Phagocytes concentrate their attention on the suspicious group that have detached themselves or were never attached in the first place (interlopers). They use a basic and fairly non-specific docking system to attach to underlying blocks of communicating cells. They (probably) feel for cells that are in electrical discontinuity. This system is not too specific, and it is clearly able to recognise allogeneic and even xenogeneic cells. Cells dedicated to being pathogens find ways of fooling this surveillance.
  10. Tnk like cells evolve as an extension of this non-specific recognition by phagocytes. Using CAMs already evolved to enhance neural interconnections (IgSFs; N-CAM-like), they learn to enhance recognition of detached cells so that this recognition becomes more specific to the individual. By hanging representative peptides off the cell surface, in association with an N-CAM-like molecule (Mhc class I like), local accumulations of these cells can sense when there is a significant deviation away from a healthy-self, Mhc+peptide signature. The pleomorphism of this identity system, and its germline receptor mechanism (probably effected by RNA splicing), can gradually expand from 2 to numerous different specificities. Possession of selfness ensures that a cell will not be attacked; its absence or loss of it ensures that it will. Tc cell activity is a functional inversion of this Tnk-like activity.
  11. Tc-like cells can now evolve. Detached cells that die by apoptosis have proven that they are able to deal with whatever caused their malfunction. Mhc+peptide hung off the cell surface gives the respective Tc cell receptor a snapshot of the stress response that is going on inside. Cells sporting a similar caricature to cells that previously died by apoptosis can be safely left to get on with it by themselves. But cells resembling those that spilled their cytoplasmic contents last time obviously cannot. On any new encounter, they need to be recognised and encouraged into an early apoptosis.
  12. The major point to grasp is that it matters not one jot whether the Mhc+peptide signature is formed dominantly from peptides derived from self or from some foreign intracellular nasty. A precursor Tc cell that is able to recognise this combinatorial epitope will classify the encounter into "tidy" (apoptotic) or "messy" (leaky) death. A previous "messy" encounter will encourage an aggressive response in the future. Paratopes interacting with Mhc+self-peptide are relatively resistent to an easy promotion into aggression. The huge volume of successful apoptosis that occurs naturally in the body mops most of these paratopes up into tolerance and there is also a tendency for younger, precursor Tc cells to be harder to enrol into aggression. Tc cells are designed to kill self cells. This is their primary role. That is why it is so easy to produce adjuvant arthritis and other experimental autoimmune disorders. These are precipitated when there is a great deal of membrane damage, cytoplasmic spillage and a prolonged stimulation of the aggressive T-cell system in the absence (or paucity) of clearly strange antigen that can focus attention on clearly disordered cells. If you can rapidly focus aggressive attention on the cells that are rupturing due (say) to viral infection, then the whole process can be brought to a rapid resolution by a selective aggression directed preferentially towards the virally infected cells. Since there will be only a few precursor Tc cells available that are specific to the local tissue (previously mopped up by apoptosis) and they are, anyway, young and resistant to recruitment into aggression, the differential clonal expansion of aggressive Tc cells strongly favours the strange antigen.
  13. Cells that die suddenly in trauma (eg, heart attacks, burns) have not been preparing for apoptosis. Cells that die of viral infections or other intracellular nasties have probably moved somewhere along the path towards apoptosis. This means that traumatically killed cells don’t spill much Il-1 whereas infected cells do. Spilt Il-1 is a strong contender as one of the danger signals. Nevertheless, Dressler’s syndrome is the occasional severe autoimmune sequel to a massive heart attack - so trauma does increase the risk of auto-immunity. Similarly, in burns and major trauma, auto-immune activity is easy to demonstrate - indeed, it is the probable cause of the phagocytic anergy that leads to multiple organ failure. Any system that allows auto-rejection of normal cells must sport a failsafe cut off device to inhibit piecemeal self destruction.
  14. Th1 cells evolved as an elaboration of the Tc cell system. Now, the cellular and cytoplasmic debris of leaking cells can be digested by APCs and representative peptides presented on the APC’s surface. On any future encounter of a similarly caricatured APC (NOT of the pathogen nor of its native antigen), the inflammatory response can be ramped up quickly - bringing in copious aggressive phagocytes and Tnk cells. This is designed to give inflammation a memory. The phagocytes still have to sort healthy self cells from the rest but now they are souped up to carry out this job up in a big way.
  15. Since extensive tissue destruction is undesirable, phagocytes need to be inhibited when the Th1 cell amplification process gets too intense. Phagocytes activity is consequently inhibited (as, eg, in a boil) and this increases the amount of debris that is left to cleaned up; the IgM free antibody system arose to act as a debris mop. So, progression to Th2 activity probably occurs when cell-mediated auto-rejection is inhibited in the interests of inhibiting self-tissue destruction. But this inhibits the clean disposal of mess and IgM compensates for this. All the others immunoglobulins have evolved as new shells to envelope IgM (eg, IgG, IgA, IgE).
  16. Thymic tolerance is probably targeted primarily at enhancing tolerance to lymphocytes (perhaps macrophages and epithelial cells too) to pre-emptively inhibit their auto-rejection. These cells are frequently exposed to mess making agents (lymphocytes are expected to migrate to lymph nodes - areas where destructive disease is likely to be common.). This is probably why the thymus turned up in the pharyngeal arch - close to the place where copious sea water (with its many mess making contents) is passed across the large convoluted surface of the gills.

So, in synopsis, the whole process works by a differential suicide in which irremediably dysfunctional cells trash their cytoplasm and so sanitise their contents. The fact that infection is a frequent cause of intracellular dysfunction is of no interest to the system. Cells don’t think "I am (or you are) infected". They realise "I am (or you are) irrecoverably sick". The corollary is that there is a differential nurturing of healthy self cells that are in junctional communication with their neighbours. The adaptive immune system simply remembers some caricature of the cells (or their debris) that failed to do the decent thing last time and watches out for them the next time round.

I have had some further thoughts on how the morphostatic response progresses from the innermost shell and progressively employs each successive shell until morphostasis is achieved. It will stop at whichever shell has finally been successful. The heat shock protein system may achieve morphostasis on its own, fixing the problem with minimal disruption. When the volume of malformed protein is too great the next stage is induced (eg, the dumping of representative peptides on the surface, associated with Mhc molecules). Phagocytes try to remove cells or organisms that start making a mess - and they are mostly successful. But this can fail. At this stage, the identity system they use has probably been overcome by the agent precipitating the crisis. So, Tnk cells are brought into play. These are more specific about self identity (and are dominantly directed at disconnected cells). Even so, some mess makers still fool the system and lead to cytoplasmic rupture. Now the Tc cell system is brought into play to focus (remember) attention on suitably caricatured cells. The Tc cell killing system provides a clean clearance of sick cells -while this works. But even this may be overcome, at which time we are left with a mess of spilled cell debris. This debris is easy to recognise and eliminate (by phagocytes). The Th1 system then categorises the epitopes formed on APC membranes (ClassII+peptide) on the basis of it having been previously hung out for display during an inflammatory event while cell debris and cytoplasm was being cleaned up. When Th1 cells meet a similarly caricatured phagocyte (which must have already started to clean up cell debris) then they will call up an army of phagocytes to carry out this inflammatory job more swiftly and aggressively. But this then lays the system open to the failsafe cutoff of macrophage anergy when the inflammatory destruction becomes too intense. Extensive tissue destruction must be averted by inhibiting macrophage activity. This anergy leads to the accumulation of a lot of cellular debris. And so, the need arises to mop this up with IgM. I am sure that a little thought would allow us to go on to the next rounds (the other Igs) but I’m going to stop here. The complement "shell" is discussed in my in print article.

Thus, for each insult, there is a sequential move from an inner to an outer shell until morphostasis is achieved. Now this is rather like the evolutionary steps taken in forming the mammalian immune (morphostatic) system in the first place. Remember, also, that ontogeny tends to recapitulate phylogeny. So is the gradual aging of the immune system partly the stochastic accumulation of a multitude of responses to various "mess making" insults as they progress through these shells? For some agents, the move to the outer shells may be early in life and for others it is late. Could this, with the superimposition of the gradual genetic divergence of the zygote derived colony, be found to be a source of the immune changes that occur as an animal ages?