Reflex Patterns in Sympathetic Neurons: Animal Experiments

Figures 2.3 and 2.4 illustrate examples of reflex responses in sympathetic neurons regulating blood vessels in skeletal muscle (muscle vasoconstrictor neurons, MVC), skin (cutaneous vasoconstrictor neurons, CVC), possibly the nasopharyngeal mucosa (‘inspiratory’ neurons, INS) and sweat glands (sudomotor neurons, SM) in the anesthetized cat. The overall results obtained on sympathetic neurons are:

Figure 2.3 Reflex patterns in sympathetic postganglionic neurons innervating skin or skeletal muscle. Reflexes in muscle (MVC) and cutaneous (CVC) vasoconstrictor and sudomotor (SM) neurons recorded from postganglionic axons in bundles isolated from a skin nerve (medial plantar nerve) or muscle nerve (gastrocnemius–soleus nerve) of the hindlimb in anesthetized cats. A. Stimulation of the carotid chemoreceptors by a bolus injection of CO₂-enriched saline (arrow) activated the MVC neurons and inhibited the CVC neuron (recorded simultaneously). Increased activity in chemoreceptor afferents in the carotid sinus nerve (CSN) was monitored. B. Stimulation of cutaneous nociceptors by pinching the ipsilateral hindpaw (indicated by bar) also excited the MVC neurons and inhibited the CVC neuron. C. Simultaneous recordings of the activity in a single CVC neuron (small signal) and in a single SM neuron (large signal) and the skin potential (SKP) from the central paw pad. The SKP is an effector response generated by activation of sweat glands. Stimulation of Pacinian corpuscles in the hindpaws by vibration excited the SM neuron and inhibited the CVC neuron. SM activation was correlated with the changes in SKP. D. Inhibition of CVC neurons to warming of the anterior hypothalamus. Note the increase in skin temperature (SKT) on the central paw pad following depression of CVC activity.

Data for A–C from Jänig and Kümmel, unpublished; D modified from Grewe et al. 1995.

Figure 2.4 Reflexes in sympathetic preganglionic neurons generated by mechanical stimulation of the nasal mucosa. Simultaneous recording from three preganglionic axons in a strand isolated from the cervical sympathetic trunk and from a phrenic nerve in an anesthetized cat. Activity recorded from a cutaneous vasoconstrictor (CVC) neuron, an inspiratory (INS) sympathetic neuron and a muscle vasoconstrictor (MVC) neuron. Phrenic nerve activity indicates central inspiration (I), post inspiration (pI) and expiration (E). The mechanical stimulus was a shearing stimulus applied to the nasal mucosa; it probably was noxious. Note that this stimulus (1) activated the INS neuron, but only during phrenic nerve discharge (PHR; i.e., during inspiration), (2) inhibited the CVC neuron, which was otherwise only active during expiration, and (3) excited the MVC neuron, which was active in inspiration and expiration before the stimulus. The reflexes in these neurons outlasted the stimulus. The increase in blood pressure (BP) was correlated with the continuous MVC discharge during and after the stimulus.

Modified from Boczek-Funcke et al. 1992.

More than 10 different functional groups of post- and preganglionic sympathetic neurons have been identified in the lumbar sympathetic outflow to skin, skeletal muscle, and viscera, and in the thoracic sympathetic outflow to the head or neck of the cat (see Fig. 2.5 for the lumbar sympathetic outflow to skin, skeletal muscle, and pelvic organs). The same types of reflex patterns have been observed in both pre- and postganglionic neurons. Most neurons in several of these pathways (e.g., the vasoconstrictor, sudomotor, and motility-regulating pathways) have ongoing activity, whereas in four investigated pathways (e.g., the pilomotor and vasodilator pathways) the neurons are normally silent. Other functionally distinct groups of sympathetic neurons which have not been studied so far are likely to innervate other autonomic target cells (e.g., the kidney [blood vessels, juxtaglomerular cells], the urogenital tract [urinary bladder, vas deferens, erectile tissue, glandular tissue], the hindgut [non-vascular smooth musculature or glandular tissue via the enteric nervous system], the spleen [immune tissue], the heart, the brown adipose tissue [Morrison 1999, 2001], etc.).

Figure 2.5 Lumbar sympathetic systems supplying skeletal muscle and skin of the hindlimb and pelvic organs (including distal colon) in the cat. The preganglionic neurons of these systems are largely located in the lumbar segments L1 to rostral half L5 (the cat has seven lumbar segments) and project in the lumbar sympathetic trunk distal to paravertebral ganglion L5 or in the lumbar splanchnic nerves and the hypogastric nerves. The first group projects (with a few exceptions related to pelvic organs) to postganglionic neurons innervating somatic tissues, and the second group to postganglionic neurons innervating pelvic organs or colon. The second group includes other sympathetic pathways associated with the internal sexual organs, and probably other target cells in the pelvic organs which are not shown in the figure. Preganglionic neurons are associated with spinal circuits (shaded areas) and are under multiple supraspinal control via sympathetic premotor neurons (see Fig. 2.12). Sympathetic neurons with spontaneous activity in bold. CVC, cutaneous vasoconstrictor; CVD, cutaneous vasodilator; MR, motility-regulating; MVC, muscle vasoconstrictor; MVD, muscle vasodilator; PM, pilomotor; SM, sudomotor; VVC, visceral vasoconstrictor.

Modified from Jänig (2006).

Reflex patterns in sympathetic neurons: studies in humans

Activity in bundles with few or single discriminated sympathetic postganglionic axons in human skin or muscle nerves can be recorded in the awake subject with tungsten electrodes inserted manually into the nerve. Using this microneurographic recording technique, it has clearly been shown that muscle vasoconstrictor, cutaneous vasoconstrictor, and sudomotor neurons have the same distinct activity patterns as in the cat or rat, and that there is also evidence for the existence of sympathetic vasodilator neurons supplying skin or skeletal muscle in humans (Jänig & Häbler 2003, Wallin 2002). However, in humans we have no recordings from sympathetic neurons innervating viscera or head, and no recordings from preganglionic neurons (Jänig 2006).

Reflex patterns in parasympathetic neurons

Several systematic studies have been carried out on the functional properties of the parasympathetic pre- and postganglionic neurons involved in the regulation of pelvic organs, gastrointestinal tract, heart, airways, salivary glands, and inner eye muscles. There are good reasons to assume that the principle of organization into functionally discrete pathways is the same as in the sympathetic nervous system, the main difference being that some targets of the sympathetic system are widely distributed throughout the body (e.g., blood vessels, sweat glands, erector pili muscles, fat tissue), whereas the targets of parasympathetic pathways are more restricted (for details and references see Jänig 2006).

Functional implications of the neurophysiological studies of autonomic neurons in vivo

These neurophysiologic studies in animals and humans argue that the peripheral autonomic nervous system consists of many separate pathways. This separation is supported by a substantial amount of histochemical evidence demonstrating that, depending on their function, peripheral autonomic neurons may contain neuropeptides or combinations of neuropeptides in addition to the classic transmitters noradrenaline and acetylcholine. The co-localized peptides in autonomic neurons are to a certain degree correlated with the function of the autonomic neurons, although this correlation is not straightforward and absolute; although there are species differences; and although the function of almost all peptides in these neurons is unknown (Gibbins 1995, 2004).

The separation of the peripheral autonomic nervous system into many functionally discrete channels is fundamental to an understanding of how the autonomic nervous system works and why autonomic regulation of all effector systems by the brain is always precisely adapted to the behavior of the organism, depending on the environmental conditions and on the conditions within the body. The concept that the sympathetic nervous system operates in an ‘all-or-none’ fashion, without distinction between different effector organs, is misleading. The same applies to the idea of a simple functional antagonism between the sympathetic and parasympathetic nervous systems. This is corroborated by the finding that circulating noradrenaline has no detectable function under physiological conditions. Circulating adrenaline, the release of which from the adrenal medulla is regulated by a distinct sympathetic pathway, is a metabolic hormone and may have some other functions, but does not otherwise elicit autonomic effector responses under physiological conditions. Thus, these findings refute the idea that the sympathetic nervous system functions in a generalized fashion.

Autonomic ganglia

In peripheral autonomic ganglia preganglionic axons diverge to and converge on postganglionic neurons. Both divergence and convergence occur within, but not between, functionally distinct autonomic pathways. A major function of peripheral autonomic ganglia is to distribute the centrally generated signals by connecting each preganglionic neuron with several postganglionic neurons. In this way a small number of preganglionic neurons connect with a large number of postganglionic neurons. The extent of divergence varies significantly, the ratio of pre- to postganglionic axons being, in pathways such as in the ciliary ganglion to the iris and ciliary body, as low as 1:4, and in others, such as in the superior cervical ganglion with many vasoconstrictor neurons, as high as 1:150. A limited divergence and much divergence are not characteristics of the parasympathetic or sympathetic systems, respectively. Probably, by analogy with somatic motor units, limited divergence is common in pathways to small targets with discrete functions (e.g., in autonomic pathways to the inner muscles of the eye, or the parasympathetic cardiomotor pathway to cardiac pacemaker cells), whereas widespread divergence is a feature of pathways to anatomically extensive effectors that act more or less simultaneously (e.g., of cutaneous, muscle, or visceral vasoconstrictor pathways) (McLachlan 1995, Jänig 2006).

Sympathetic paravertebral ganglia

Most postganglionic neurons in sympathetic paravertebral ganglia (in the sympathetic chains) innervate autonomic targets in somatic tissues; some innervate viscera (e.g., the heart) or targets of the head (e.g., the intracranial blood vessels, salivary and nasopharyngeal glands, intraocular smooth muscles). Within these ganglia, ganglionic neurons have uniform properties. Each convergent cholinergic preganglionic axon produces excitatory postsynaptic potentials by activating cholinergic nicotinic receptor channels. The amplitude of the potentials ranges from a few millivolts (subthreshold weak synaptic inputs) to 10–40 mV (suprathreshold strong synaptic inputs). In most cases, one or a few strong preganglionic synaptic inputs are always suprathreshold, such as the endplate potential at the skeletal neuromuscular junction. They have a high safety factor and always initiate an action potential. Thus, these ganglion cells relay the incoming CNS-derived signals in only a few of their converging preganglionic inputs (McLachlan et al. 1997, 1998). The function of weak preganglionic synaptic inputs generating small subthreshold potentials in paravertebral ganglia is not clear (Fig. 2.6A).

Figure 2.6 Signal transmission in peripheral autonomic pathways: Transmission through autonomic ganglia. A. Relay functions of autonomic ganglion cells. Postganglionic neuron with one or a few strong (S; suprathreshold) preganglionic synaptic inputs (generating postsynaptic potentials of 10–30 mV) and several weak (W; subthreshold) synaptic inputs (generating synaptic potentials of a few mV). This connectivity occurs in almost all neurons in paravertebral sympathetic ganglia, in some neurons of prevertebral ganglia, and in parasympathetic ganglia. These connections function mainly to transmit the activity from pre- to postganglionic neurons (McLachlan et al. 1997, 1998). B. Relay and integrative functions of autonomic ganglion cells in prevertebral ganglia. Postganglionic neuron with weak synaptic inputs from both preganglionic neurons and interneurons of the enteric nervous system (ENS; intestinofugal neurons), and also from collaterals of spinal visceral afferents. The first two are nicotinic cholinergic (N). The afferent collaterals are peptidergic and use substance P (SP) as transmitter. These postganglionic neurons innervate neurons of the enteric nervous system and possibly other target organs in the viscera (but not blood vessels). They fire only after integration of several subthreshold cholinergic synaptic inputs and/or a slow afferent-induced depolarization.

Modified from Jänig & McLachlan (2002).

Transmission of signals at the autonomic neuroeffector junctions

In peripheral tissues the transmission of activity in autonomic nerve terminals onto autonomic effector cells is complex and depends on the release of several different compounds and on the presence and distribution of the receptors in the effector membranes for these compounds. Anatomical investigations of neuroeffector junctions at arterioles, veins, pacemaker cells of the heart, cells of the vas deferens, the iris dilator myoepithelium, and the longitudinal muscle of the gastrointestinal tract have demonstrated that varicosities of autonomic nerve fibers which are not surrounded by Schwann cells form close synaptic contacts with the effector cells (Hirst et al. 1992, 1996). These structures are the morphological substrate for the transmission of the centrally generated signals by the postganglionic neurons to the effector cells (Fig. 2.7). The close synaptic contacts occur on <1% of the surface of effector cells. Neuroeffector transmission on arterioles, small arteries, the heart, the vas deferens, the iris dilator myoepithelium, and the longitudinal musculature of the gastrointestinal tract has been studied in in-vitro preparations using neurophysiological and other methods. These investigations show that signal transmission from postganglionic axons to autonomic effector cells occurs via close neuroeffector junctions and postjunctional receptors, but not – or only to a limited degree (as far as vascular target cells are concerned) – via extrajunctional receptors. This specific method of neuroeffector transmission is the basis upon which signals generated in specific central circuits (see below) are channeled to the appropriate autonomic target cells.

Figure 2.7 Simplified scheme of neuroeffector transmission to autonomic target cells (arterioles, heart, non-vascular smooth muscle cells, secretory cells). Subjunctional (subsynaptic) receptors mediate the effect of transmitter released by the nerve terminals during excitation under physiological conditions. The cell surface at which this nerve–effector communication occurs is ≤1% of the total cell surface. The subsynaptic receptors are ligand-coupled or second messenger-coupled to cellular effectors (e.g., ionic channels). Extrasynaptic receptors for the transmitter are either different from the subsynaptic ones and/or are coupled by different intracellular second-messenger pathways to the cellular effectors. The function of the extrajunctionally located receptors is unclear for most innervated effector cells. Small vesicles containing the transmitter are located close to the synaptic cleft. Large vesicles are also present in many varicosities (not shown). In these vesicles neuropeptides are co-localized with ‘classic’ transmitter. Large vesicles are not located close to the synaptic cleft. The physiological role of the neuropeptides is in most cases unclear.

Modified from Jänig (2006).

Chemical transmission at these neuroeffector junctions is based on the release of one of the ‘conventional’ transmitters, acetylcholine or noradrenaline. Most sympathetic postganglionic axons release noradre-naline, but some release acetylcholine (e.g., sudomotor and muscle vasodilator axons in some species). Most (but not all) nerve-mediated effects can be antagonized by blockade of adrenoceptors or muscarinic acetylcholine receptors. All parasympathetic neurons are cholinergic, i.e., they release acetylcholine on stimulation. However, not all effects of stimulating parasympathetic nerves are blocked by muscarinic antagonists. This clearly implies that other transmitters and/or other receptors may be involved. It is now clear that several chemical substances, often contained within individual autonomic neurons, can be released during excitation and can have multiple actions on effector tissues (Furness et al. 1989, Morris & Gibbins 1992). The compounds which may be involved are nitric oxide (NO), adenosine triphosphate (ATP), or a neuropeptide (e.g. vasoactive intestinal peptide [VIP], neuropeptide Y [NPY], galanin [GAL] and other peptides). Immunohistochemistry has revealed the presence of many peptides, although only a few of these have been demonstrated to modify the function after release from nerve terminals in vivo (e.g., NPY or VIP) (Gibbins 1995, 2004, Gibbins et al. 2003).

The response of autonomic target tissues to nerve-released noradrenaline or acetylcholine or other compounds usually only follows repetitive activation of many axons. High-frequency stimuli, particularly in bursts, may produce effector responses owing to the concomitant release of a neuropeptide. Alternatively, when the effects of nerve activity are not completely blocked by an adrenoceptor or muscarinic antagonist at a concentration that entirely abolishes the response to exogenous transmitter, it may not necessarily be the case that a transmitter other than acetylcholine or noradrenaline is involved (as the subjunctional receptors may be different from the extrajunctional ones; and as the concentration of transmitter released into the junctional cleft may reach millimolar concentrations). Although the effects of exogenously applied substances, which have putative transmitter function, on cellular functions are known for many tissues, the consequences of activation of postjunctional receptors by neurally released transmitters have rarely been investigated. When they have, the mechanisms of neuroeffector transmission have been found to be diverse, involving a range of cellular events (Jänig & McLachlan 2002). One important concept that has emerged is that the cellular mechanisms utilized by an endogenously released transmitter are often not the same as when this transmitter substance or its analogue is applied exogenously. The endogenously released transmitter acts primarily or exclusively on the subjunctional receptors, whereas the exogenously applied transmitter acts on the extrasynaptically located receptors (Fig. 2.7, Hirst et al. 1996).

Segmental organization of the spinal afferent and sympathetic innervation

The basis of the spatial organization of the viscerosomatic and somatovisceral reflexes, including referred pain in skin and deep somatic tissues, is the segmental (spinal) organization of the afferent and sympathetic innervation of the three body domains. This is particularly visible at the trunk of the body. Functional morphological studies on rats using tracers such as horseradish peroxidase applied either to a dorsal cutaneous nerve or an internal intercostal nerve have shown that about 90% of all sympathetic postganglionic neurons that project to the skin of the trunk through the dorsal or ventral ramus of a spinal nerve have their cell bodies in the paravertebral ganglia of the corresponding segment and of the adjacent caudal segment, the remaining postganglionic cell bodies being located in the two rostral paravertebral ganglia. For example, the cell bodies of the postganglionic neurons innervating dermatome C and projecting through segmental nerve C in Figure 2.9 are located in paravertebral ganglia 3 and 4 (90%) and 1 and 2 (10%). The cell bodies of the primary afferent neurons that innervate dermatome C and project through the segmental spinal nerve C are located in spinal ganglion 3 (Baron et al. 1995). All preganglionic neurons of a spinal segment project through the ventral root and the corresponding white ramus of this segment. In the spinal cord the preganglionic neurons are arranged in rostrocaudal columns and in clusters in the intermediate zone. There is some topical organization with respect to the type of target (Baron et al. 1985, Jänig 2006). Preganglionic axons preferentially form synapses with postganglionic neurons that innervate effector cells in the corresponding dermatome, myotomes, or sclerotome. The degree of branching of individual preganglionic axons is unknown. However, individual preganglionic neurons do not seem to project up and down into the sympathetic chain, and do not project into functionally different nerves (e.g., both into a splanchnic nerve innervating viscera and into the sympathetic chain to ganglia projecting to somatic tissues).

Figure 2.9 Segmental organization of postganglionic sympathetic neurons and primary afferent neurons to the skin of the trunk. A. Dorsal root ganglia (sensory) and paravertebral ganglia (sympathetic) are indicated by vertebral bars. The height of the bars indicates the relative numbers of neurons that project through the corresponding segmental nerve to the dermatome. Neighboring dermatomes overlap. Cell bodies of all afferent neurons are located in one dorsal root ganglion (e.g., for dermatome C in dorsal root ganglion 3). The cell bodies of the postganglionic neurons are located to 90% in the corresponding or the next caudal paravertebral (sympathetic) ganglion (e.g., for dermatome C in ganglions 3 and 4), and the remaining postganglionic neurons are located in the next two rostral paravertebral ganglia (for dermatome C in 1 and 2). This means that most postganglionic neurons in one paravertebral ganglion project in two overlapping dermatomes (e.g. 2 in B and A). According to data using the tracer horseradish peroxidase to locate cell bodies obtained in the rat from Baron et al. (1995).