An outline of the autonomic nervous system

This chapter is relevant to Section M (i) of the 2017 CICM Primary Syllabus, which expects the exam candidates to "describe the autonomic nervous system, including anatomy, receptors, subtypes and transmitters (including their synthesis, release and fate)". Specifically, this chapter addresses the first, most vague, directive in this syllabus item, mainly because it is so elemental that it could take a well-prepared exam candidate by surprise. One can know everything about the autonomic nervous system and still fail a question that asks them to describe it in five minutes, because that would require an economy of carefully chosen words within a bare structure for which one might not be prepared. Considering that the absolute hard limit for legible handwriting seems to be about 20-30 words per minute, this would cap the description of these incredibly complex systems to something like 100-150 words each, or 300 words where the full SAQ was dedicated to one of the branches by itself.

Only one SAQ in the past papers resembled this. Question 7 from the second paper of 2017 asked the candidates to compare the sympathetic and the parasympathetic nervous systems in a "compare and contrast" table. Question 4 from the second paper of 2014 and Question 1 from the first paper of 2017 were similar, but asked specifically for the anatomy and physiology of the parasympathetic nervous system. What follows is an attempt to fashion an answer to such questions which could also cover some hypothetical future SAQs about the sympathetic nervous system, all while keeping to the aforementioned word limit as much as is possible without losing vital details.

The absolute best possible reference for this topic, and the entire autonomic nervous system topic for that matter, would have to be "The integrative action of the autonomic nervous system: neurobiology of homeostasis" by Wilfrid Jänig (2022), but it would be wasteful and ridiculous to recommend this textbook to the CICM exam candidate, as is is approximately 430 pages long. Instead one is redirected to the excellent work of John Furness (2006), highly recommended mainly because the author respects the time of the audience ("I do not burden the reader with numerous caveats, quirks of observation and minor deviations that could be enumerated"). 

Definitions and boundaries of the autonomic nervous system

"Autonomic", an adjective that suggest the nobility of independent self-governance, is a relatively flattering term to describe these networks, which had historically been referred to as "vegetative" or "animal". It was coined around 1903 by John Newport Langley, a Cambridge professor whose main contribution to the long and storied history of autonomic neurology was to use the ganglionic blocking effects of nicotine to discern the physiological and anatomical extent of pre and post-ganglionic fibres. The resulting experiments defined the distribution of the nerve fibres so much better than the old anatomists that Langely was unable to keep using their older terminology, and had to devise his own:

“... To understand views of previous authors would have been even more difficult, had I added to the old terms some new features. Therefore, I called the entire system the “autonomic nervous system.” By this, I meant a “local” autonomy. The word “autonomy” indicates, without doubts, a much greater degree of independence of the central nervous system, than it is in reality, with the possible exception of innervation of the gastrointestinal canal, but I think that for novel concepts in science it is also necessary to introduce new terms, if even these terms would not represent a precise description of the subjects”

Langley went on to apply the term "sympathetic" for the system of efferent fibres arising from the thoracic spine by borrowing the anatomical works of Jacques Bénigne Winslow, who used the ancient Galenic concept of "sympathy" to rename some of the cervical and thoracic ganglionic structures he was describing in the slightly decomposed bodies of executed prisoners.  "Sympathy", in this sense, was a phenomenon which Galen used to describe the communication and cooperation between the organs, which he believed was occurring by the exchange of "animal spirits" though nerves.  Winslow observed that cervical and thoracic ganglionic fibres descended to innervate the heart and extended to other organs, concluding that these nerves were likely responsible for the "sympathy" Galen referred to. "Para-sympathetic", para meaning "alongside", was then used by Langley in 1905 to refer to the cranial and sacral autonomic systems. For mainly Anglocentric colonial reasons this terminology has become the standard in English literature, but the reader would be surprised to learn that this is not the standard worldwide, and in Europe one may find onself among professionals discussing the vegetatives Nervensystem, systema neurovegetativo, système nerveux végétatif, вегетативна нервна система, and so forth.

The enraged CICM trainee could at this stage fairly object that none of the abovementioned material is examinable, and that a supposedly exam-focused resource should surely focus on providing modern definitions which appear in contemporary textbooks and therefore in SAQ answer rubrics. This outburst would be a completely understandable reaction, and could be answered with a series of such definitions, because better late than never. Langley did explore the functional aspects of the autonomic nervous system and his major works described which organs were innervated and what those organs did in response to various dangerous drugs (adrenaline, nicotine, pilocarpine, atropine), but the main defining features of the sympathetic and parasympathetic nervous system were anatomical, i.e. based on the the origins of the nerves, and specifically on the efferent nerves. Thus:

It would be disingenuous to mislead the exam candidate into believing that this oversimplified classification is a perfect order, but that is what is necessary to conceptualise the ANS for the CICM exam. Unfortunately this binary system is responsible for a lot of confusion and misconception, and a slightly more sophisticated view would also need to incorporate some additional complexity:

• The enteric nervous system should be viewed as a separate autonomic nervous system, as it is autonomic in the sense that it is involuntary and in fact can operate in the absence of any descending control, and is functionally and histologically distinct from the two other systems,
• Ambiguous ganglia: some nervous structures cannot unambiguously be assigned to one  system or the other, eg. some of the pelvic ganglia
• The system implies directional top-down control, whereas in fact there is a lot of crosscommunication between organs which bypasses central regulatory mechanisms (especially in the enteric nervous system). 
• The classification is based on efferents, i.e. the outflow, but there are also visceral afferents which are hugely important

That last point is worth going into.  Afferent pathways dealing with autonomic functions obviously do exist (for example the baroreceptor and chemoreceptor afferents from the carotid sinus via the carotid sinus nerve), and most of these are "autonomic" in the sense that a lot of these sensory inputs integrate only vaguely or incompletely into the conscious sensorium (i.e. one is usually not directly aware of one's baroreceptors), but there is really nothing to distinguish these from other sensory nerves. To borrow a turn of phrase from Langely, "It is clear that we cannot make a like division of afferent fibres according as they run to striated muscles or to other tissue; it would lead to nothing but confusion to consider the afferent fibres of the skin as autonomic fibres and the afferent fibres of striated muscle as the only somatic afferent fibres". It would be even more difficult to classify these afferents as "sympathetic" or "parasympathetic", for example where it comes to nociception from the abdominal organs (i.e it would make no sense to define some abdominal pain as "sympathetic" and some as "parasympathetic" depending on whether the afferents travel via the sacral or thoracolumbar nerve roots)

Structural differences between the somatic and autonomic nervous systems

The major structural difference between the nervous systems is the presence of ganglia, which interrupt the neurotransmission from the CNS to the effector organs. As one can see, such ganglia are not present in the somatic nervous system, where a single heavily myelinated efferent axon stretches from the spinal cord.

In contrast, both the sympathetic and the parasympathetic nervous systems have ganglia, which contain the cell bodies of postganglionic neurons. For the thoracolumbar sympathetic fibres, these ganglia are close to the spinal cord, and for the craniosacral parasympathetic network the ganglia are closer to the target organ.

This organisation is fairly well preserved in all vertebrates and appears to have been a decision we all collectively took around the same time as we decided to have jaws and teeth (whereas cyclostomes like the hagfish have chosen a different path). Interestingly, the sacral parasympathetic circuitry seems to be a relatively recent development, most advanced among mammals (as in, amphibians and fish do not have it), whereas the cranial parasympathetic network (especially the vagus) is highly conserved. The sympathetic chain seems to become more and more disorganised the lower you climb along the branches of the evolutionary tree, and largely speculative models suggest that the earlier, more primitive forms of sympathetic signalling had relied more on the distribution of soluble mediators (similar to what happens when the adrenal medulla releases adrenaline), rather than on direct innervation of the effector organs.  In general it seems that, unless you have a centralised nervous system, the distinction between autonomic and somatic innervation becomes fairly arbitrary, as all the structures are organised into ganglia anyway,  and moreover because of the real heartbreaking difficulty in identifying voluntary behaviour which you could call "somatic" in such organisms as the Cnidaria. Still, Shimizu & Okade (2007) made some valiant efforts to relate familiar cardiovascular "rest and digest" and "fight or flight" activities to the attack and digestion behaviours of lower phyla all the way down to the Hydra,  suggesting that the autonomic nervous system massively predates such taken-for-granted structural norms as bilateral symmetry and structural concentrations of control organs. 

It is at the same time pointless (from an exam perspective) and essential (from the point of view of completing the chapter with all the necessary comparisons) to discuss the enteric nervous system alongside the somatic and autonomic. Anyway it would seem silly to retreat from a digression on the ENS, having just finished a paragraph about the neurology of jellyfish. The author seeks refuge in the thought that nobody's time is wasted, as the time-poor exam candidate would have abandoned this chapter long ago (somewhere around the phrase "système nerveux végétatif") and whoever has read as far down as this will have the stamina for more ridiculous asides. So: the enteric nervous system is only loosely affiliated with the others, in the same sense as a completely separate sovereign nation might be: it receives inputs from the somatic and autonomic system, and sends afferent responses in various ways (including soluble neurohormonal ones), but is organisationally completely distinct and independent. It consists of local ganglionic structures directly embedded into the gut wall, some of which have distinct pre- and post-ganglionic fibres and some which are aganglionic (i.e. interact with their effector organ directly). The interested reader is redirected to this short review by Hansen (2002), mainly to help them develop an appreciation for how functionally and morphologically complicated this system is. To remain on brand with a foray into comparative biology, it would surprise nobody that this nervous system was the earliest to arise (Furness & Stebbing, 2018), as basically anything with a gut needs to coordinate some kind of peristaltic activity. 

Structural and functional differences and similarities between the sympathetic and parasympathetic nervous systems

Without repeating a lot of material from the chapters that deal specifically with each system, it is possible to summarise this in a few short paragraphs.  In short:

Different CNS origins: the preganglionic cell bodies of the ANS lay in different locations, but are somewhat analogous. For example, sympathetic fibres arise from the intermediolateral nucleus of the spinal cord, which is a long sausage of grey matter stretching from T1 to L3. This thing could be regarded as analogous to the grey matter nuclei of the brainstem, which contain the cell bodies of the parasympathetic preganglionic neurons. In the sacrum, there is also an intermediolateral nucleus, but the efferent fibres arising from this at the levels of S2-S4 are parasympathetic. 

Different peripheral organisation. The sympathetic nervous system radiates widely from its origin. The sympathetic fibres from nerve roots of T2 down to L2 synapse with their ganglia close to the spinal cord and then send postganglionic fibres to everywhere else. Though they arise only from the thoracolumbar nerve roots, these fibres ascend and descend in both directions and rejoin the spinal nerve roots via the grey rami communicantes, so that each anterior nerve root gets a bit of sympathetic innervation. For example, the superior cervical ganglion distributes fibres to the anterior rami of C1 to C4. A single preganglionic fibre innervates a large number of postganglionic neurons, and this amplifies and distributes the signal, such that the activation of one preganglionic neuron leads to the activation of a wide swath of innervated sites (the fibre ratio is said to be 1:20, i.e. one preganglionic axon to twenty postganglionic axons. In contrast, the parasympathetic nervous system only does the head, neck, and abdominal organs. In fact most of the parasympathetic innervation of things not inside the head comes from the vagus nerve (75% is the widely quoted percentage for how many of the total parasympathetic fibres are carried by the vagus)

Differently positioned ganglia. Parasympathetic ganglia are near, on, or in, the effector organs, whereas sympathetic ganglia are collected in neat paravertebral chains. Some believe that this distribution underlies a more effective method of functional classification for the autonomic nervous system. Consider: the sympathetic division affects wide swaths of tissue and often discharges en masse to produce systemic effects; so surely it would make the greatest sense to centralise the ganglia and fan out their numerous projections to radiate among the organs and tissues? In contrast, parasympathetic effects are often very local (i.e. specific to one particular sphincter somewhere), so it would make sense to have the bodies of regulatory neurons closer to the site of the action. Wang et al (1995) explore this position and, on balance, conclude that it is probably an oversimplification, but then one of their co-authors went on to repeat it in their chapter from Fundamental Neuroscience (2013), which means it must be a useful misunderstanding, worth propagating.

Different postganglionic neurotransmission. Parasympathetic postganglionic neurons release acetylcholine, and use muscarinic receptors, whereas sympathetic neurons release noradrenaline (and adrenaline, in the case of the adrenal medulla). Again this seems to be something we metazoans did around the same time as we decided to develop bilateral symmetry, as more complex nervous systems demanded the recruitment of all kinds of molecules for use as neurotransmitters, whereas previously they might have been used by endocrine secretory cells to communicate systemically, without the use of synapses. Acetylcholine transmission appears to have been truly ancient (even C.elegans expresses the correct genes to make a receptor), whereas noradrenergic transmission is about as recent as chordates. For example, the basic chordate Amphioxus encodes a single catecholamine receptor that seems designed to bind dopamine and appears to be the progenitor of all other catecholamine receptors. How one nervous system ended up using one neurotransmitter and not the other is not something anyone has published anything about.

Different distribution and innervation. The sympathetic nervous system is far-reaching and covers the vast majority of the body, in particular covering the entire circulatory system. In contrast, the parasympathetic nervous system mainly supplies the head and viscera. For example, the innervation of blood vessels with parasympathetic fibres is minimal. As the result, the effects of ganglionic blockade (i.e. blocking nicotinic receptors) tends to have predominantly sympatholytic  manifestations, as the parasympathetic effects are lost and overshadowed.

Different targets: the two systems rarely both innervate the same organ or tissue; most of the time either one or the other. Dual innervation is the exception, eg. the heart, detrusor muscle, blood vessels of the penis and vulva, etc.

Different (but not necessarily opposed) effects on the same organ: though many textbooks tend to present the ANS as two systems acting in opposition, this is not entirely true, and in fact under most circumstances each organ or tissue is innervated by one or the other predominantly. The impression that they produce opposite effects when they innervate the same organ seems to come mainly from the example of the myocardium. Still, there is a pervasive misconception that the sympathetic nervous system mostly coordinates systemic "fight or flight" activities, and that this is opposed and dampened by local parasympathetic "rest and digest" responses. CICM, in their answer to Question 7 from the second paper of 2017, tsked disapprovingly when "no candidate observed that the SNS is a diffuse physiological accelerator and that the PNS acts as a local brake".

Same postganglionic secondary messenger systems. Both muscarinic and adrenergic receptors are G-protein coupled, i.e. metabotropic. Their downstream effects are not necessarily the same, of course, as one can see in this excellent table from Frazer (1997):

Same preganglionic neurotransmission. Both the sympathetic and parasympathetic preganglionic synapses release acetylcholine, and use nicotinic receptors. These are ligand-gated cation channels, which means a fairly rapid transmission of signal (i.e. the postsynaptic membrane depolarises). This is what one should usually expect from nerve-to-nerve transmission, where velocity is important, and one cannot afford to wait for cAMP or somesuch. 

Same nerve fibre types.  Both sympathetic and parasympathetic preganglionic fibres are lightly and incompletely myelinated B-type fibres.  They are at the bottom of the myelinated fibre speed rankings, with a small (<3 μm) diameter and a transmission velocity of around 3-5 metres per second. For both sympathetic and parasympathetic nervous systems, postganglionic fibres are unmyelinated C fibres (thinnest, and featuring the slowest conduction).

Same direction of change (excitatory). Both sympathetic and parasympathetic systems generally stimulate the organ they innervate, i.e. the result of increased activity is rarely the relaxation of a muscle or the decrease in the secretion of a gland.