To get some understanding of how the molecular biochemistry of the nAChR connects with the physiological effects of smoking, it is necessary to say a bit about the nAChRs that are found in the brain. I warn you now that a vague understanding is all we shall reach, since the details, surprisingly, are unknown (unless the tobacco companies know but aren't telling - scarcely creditable, I'm sure you'll agree).
Experimental work on the nAChR has concentrated on the heteropentameric type found in the nerve-muscle synapse and in the Torpedo ray electric organ, and this is the kind represented in the schematic figures on the work and introduction pages. However, as I said, nAChRs are also found in the brain (neuronal nAChRs), where they constitute less than 1% of the total number of ion channels, being heavily outnumbered by GABAA, glutamate and glycine receptors. Difficulties of working with live nerve tissue, and of separating out the effects of the nAChRs from the other channels, have meant that experimental data about them is scanty. It is known from genetic techniques that there are many different kinds; at least 8 different alpha-like and 3 different beta-like subunits, found to differing degrees in different parts of the brain (the concentration of nAChRs is particularly high in the hippocampus). But it is not certain how they combine in vivo; all that is known is that an all-alpha pentameter can form functional channels in vitro (it is such an all-alpha pentameter that we have used in our modelling studies, for simplicity). Even the role of neuronal nAChRs is uncertain. Compare this with the case of the muscle nAChR. The pentameric (alpha)2(beta)(gamma)(delta) composition of the channel is well established, and the part that it plays in the complex chain of events by which a nerve impulse eventually causes the contraction of a muscle is quite well understood.
The following tentative hypothesis about the function of the neuronal nAChR seems to be the likeliest at present. nAChRs do not mediate fast responses in the brain as they do in muscle. The fast responses, which are believed to constitute the physical substance of our thoughts, are carried by other channels, especially the aforementioned GABAA, glutamate and glycine receptors. The neuronal nAChRs have instead a modulatory function; they are concentrated on the end-plate bulb of neuron-neuron synapses, exposed not to the narrow (and thus fast-acting) synapses but to the intracellular medium, which is a bath of various neurotransmitters, among them acetylcholine. Contrast this to the case in muscle, where they would be found in superficially similar synapses but on the other side and facing the synaptic cleft . The level of acetylcholine that the neuronal nAChRs respond to, being in the intercellular medium, is a 'global' quantity for the entire brain (or at least for a substantial volume of it, containing thousands or millions of neurons) and is thus slowly-fluctuating: the timescale of variation will usually be minutes rather than the milliseconds of a fast-synaptic response. Hence the 'modulatory' behaviour. They respond to an increase in the level of acetylcholine by opening, allowing sodium, potassium and calcium to pass. It seems to be the calcium (flowing into the end-plate bulb) that is significant here, because it is a crucial trigger for neurotransmitter release. Consequently, there is an increase in the amount of neurotransmitter (GABAA, norepinephrine, dopamine or serotinin) released across the fast synapse when a nerve impulse arrives, and an increase in the size of the response (whether it be inhibitory or excitatory) of the channel at the other side of the synapse. An increased level of acetylcholine thus 'potentiates' the nervous system: it makes it do the same as before, only more so.
So what (to get to the point at last) about smoking? We have said that nicotine provokes the channel-opening response of nAChRs, just like acetylcholine, and hence will have the same 'potentiating' effect on the nervous system. Though the details of how this translates into mental state are clearly complex, it seems reasonable that the result should be that having a cigarette should 'wake you up', make you feel sharper and 'ready for action' (Though there will be a concomitant relaxing effect too, caused by carbon monoxide in the smoke reducing the amount of oxygen reaching the brain and depressing its activity. Biology is never simple...). The results of experments on humans and animals show exactly this. Vigilance and memory are both appreciably improved. Equally, going without a cigarette might make you feel dull, especially since the brain is likely to react to regular smoking by reducing its natural output of acetylcholine. This does not, of course, constitute a full explanation of addiction to nicotine, which probably has other more important causes, both physiological and psychological.
The amount of nicotine in cigarettes is small and its effect is noticable mainly on the neuronal nAChRs. However, it does affect muscle nAChRs too; indeed, in any more than trace quantities it is extremely poisonous, causing paralysis and death (paralysis of respiratory and cardiac muscle can seriously damage your health). This is why the tobacco plant produces nicotine; it is a poison, protecting the plant from being eaten by insects.
Lest it be thought that I have it in for smokers (anti-social, self-destructive and malodorous though they may be), I shall finish by commenting on the recent suggestion of the relation of nicotine to Alzheimer's disease. Alzheimer's disease, which affects about 5% of people over 65, and 10% of those over 80, is a neurodegenerative disease which produces first memory loss and finally dementia and senility. The brains of affected patients contain 'tangles' of filamentous protein and larger 'plaques' centred on cores of misfolded membrane proteins (the 'prions' that are thought to cause BSE are similar misfolded proteins). Exactly what causes the onset of Alzheimer's disease is still unknown, but recent research found a statistical correlation between smoking and a lower incidence of the disease. One must of course beware of arguing on the basis of statistical evidence of this type that smoking protects from Alzheimer's; to assume a causal link from this kind of correlation is a classic non sequitur (one hopes, to chose the obvious example in this case, that the researchers corrected their data for the reduced likelyhood of smokers reaching old age at all). Nevertheless, given that nicotine can affect neurons via their nicotinic receptors, a causal link cannot be ruled out, and while the exact mechanism is unclear there is reasonable circumstantial evidence: experiments show that nicotine, by stimulating nAChRs, improves memory, and memory loss is a classic symptom of Alzheimer's; moreover, nAChRs are especially numerious in the hippocampus, which is known from many experiments to be vital in the formation of new memories. At the very least, the finding surely merits further research, and indeed the Medical Research Council has recently been heavily criticised for accepting money from British American Tobacco to do just that. Perhaps in a few years time we will all start wearing nicorette patches, or at least patches containing other, less harmful nAChR agonists, and (with a stretch of the imagination) one could even see Woody Allen's vision coming true, of hypochondriacs waking up from cryogenic storage in 2250 be told by smiling doctors that smoking is 'one of the healthiest things we know.'