Volume Transmission in the Brain: Beyond the Synapse

The concept of volume transmission as an important mode of chemical communication in the brain was introduced in the 1980s.^1,3,17–20 New techniques (e.g., receptor autoradiography, immunohistochemistry) that allowed detailed mapping of the locations of neurotransmitters (NTs) and their receptors led to several experimental findings that were not compatible with a purely synaptic mode of operation. Mapping studies found relative mismatches in some brain areas between the density of nerve terminals positive for a particular NT and the density of receptors for that NT. Situations in which there were very few terminals, as compared with the number of receptors, indicated that active substances must be getting to receptors in some manner other than release from an adjacent terminal. Studies demonstrating presence of NTs outside of synapses under several sets of experimental conditions (e.g., release enhanced, reuptake blocked) indicated that spillover from the synaptic cleft into the surrounding extracellular space (ECS) was one possible source.² After release, diffusion of NTs through the ECS is likely to be sufficient to activate nonsynaptic receptors for a considerable distance (Figure 1).^1–3 Mapping studies have identified axon terminals without postsynaptic specializations. Also, varicosities containing granules filled with monoamine NTs have been localized, not only in axon terminals, but also in other portions of neurons (e.g., dendrites, soma, non-terminal axon segments). These findings support the possibility of nonsynaptic release of NTs into the ECS.^1,3,20 Nonsynaptic release was also supported by the finding of measurable levels of specific NTs in the ECS in regions lacking synaptic terminals for that NT.²⁰ On the basis of animal (primarily rodent) studies, it was concluded that the modulatory NTs (e.g., acetylcholine, norepinephrine, dopamine, serotonin) act in great part by volume transmission because the majority of their NT-containing varicosities are not part of a synapse, and many of their projections do not make synaptic contact.¹ This finding is supported by a recent study in nonhuman primates, reporting that 94% of tyrosine hydroxylase-positive boutons in prefrontal cortex Area 10 had no identifiable synaptic structures, indicating primarily nonsynaptic release of dopamine.²¹ There may be species-specific differences, as one study found a much higher percentage of acetylcholine-containing varicosities associated with synapses in the human cerebral cortex (67%) than had been reported for the rat (15%).²² Mechanisms for release of NT from nonsynaptic varicosities include exocytosis (similar to synaptic release) and reverse operation of NT transporters.^1,17–20 Many other neuroactive substances (e.g., glial transmitters, peptides, nitric oxide, neurosteroids) are also released into the ECS.^1,17–19 There is growing recognition that volume transmission is an important mode of action for other NTs (e.g., glycine, glutamate, GABA), as well.^23–26 The location of a receptor, compared with the source of the neuroactive substance, is likely to be very important in determining activation.^1–3 Many receptors that are not part of a synapse are located sufficiently far from the synaptic cleft that they will be unable to detect individual release events. Rather, they will experience a summation of surrounding events. Thus, they are more likely to signal slower (tonic) changes that are occurring over a larger area.

At the local level, volume transmission occurs via movement of neuroactive substances through the interstitial (extracellular) fluid that fills the ECS.^4,27 On the microscopic scale, the physical makeup of the ECS will influence the speed and, in some cases, direction that substances and interstitial fluid move.^4,27 The ECS, which occupies about 20% of the tissue volume, has a very complex, three-dimensional (3D) structure. This lengthens the actual path (measured as tortuosity) that a molecule must traverse as it moves around and between structures (e.g., neurons, glia, blood vessels).^4,27 Recent 3D reconstructions (rat hippocampus) have allowed identification of two major structural patterns within the ECS differing in width, confirming an earlier conclusion by Van Harreveld (Figure 1).^4,5 In areas where adjacent cell membranes are parallel, the ECS is narrower, constraining movement of interstitial fluid and molecules. Areas where several membranes converge are wider, allowing faster movement.⁵ Many factors can potentially affect the speed of movement of a neuroactive substance within the parenchyma of the brain, including its concentration and molecular size, binding to receptors or transporters, uptake processes, localized gradients in electrical potential, pressure changes (e.g., due to arterial pulses), temperature gradients (e.g., between an area of brain that is active and surrounding tissue), and bulk flow.^4,17–19,27 Most signaling by volume transmission via the ECS is short-distance, particularly within gray matter. It has been suggested that the highly linear nature of white-matter tracts may provide a faster route for interstitial fluid movement than occurs in gray-matter areas.^4,14

Psychiatric medications, which must cross the blood–brain barrier, will distribute throughout the brain parenchyma and therefore are accessing sites of action by volume transmission.¹ Also, it has been proposed that the therapeutic efficacy of CNS drugs such as NT reuptake-inhibitors is mediated primarily by effects at nonsynaptic receptors.¹ One reason is that the majority of NT transporters are located extrasynaptically. Another is the drug-induced increased NT concentration in the ECS. NT receptor subtypes that are synaptic generally have a lower affinity, and therefore require a higher NT concentration for activation, than the subtypes that are nonsynaptic. After synaptic release, diffusion quickly reduces the NT concentration in the synaptic cleft below the level required to activate synaptic receptors. However, it is still sufficiently high in the ECS to activate nonsynaptic receptors (Figure 1).^1,2 Nonsynaptic receptor subtypes may also initiate very different changes from those of synaptic receptors and so may provide targets for development of new therapeutics. For example, the lower-affinity NMDA receptor subtype (GluN2A), which is primarily synaptic, has been linked to neuroprotective effects, whereas the high-affinity NMDA receptor subtype (GluN2B), which is primarily nonsynaptic, has been linked to glutamate-mediated excitotoxicity.²⁸ It has been suggested that the failure of nonselective NMDA antagonists in clinical trials is due to the simultaneous activation of both pathways.²⁸

Longer-distance chemical signaling also occurs.^{14,16,18,19,29,30} It was demonstrated in the 1980s that a tracer introduced into the cerebrospinal fluid (CSF) quickly distributed into the brain parenchyma, outlining the microvasculature.³¹ Transport within the CSF potentially allows neuroactive substances rapid access to areas of the brain distant from their point of entry into the CSF. CSF is continuously produced, with a rate of formation in the human sufficient to fully replace its volume several times per day.^29,30,32,33 CSF flow provides access to distant areas of the brain because it enters the perivascular spaces (PVS, Virchow-Robin space) that surround all penetrating arterioles and venules (intracerebral extrinsic microvasculature).^6–9,34 The PVS is bounded on the inner side by the basal laminae of the penetrating vessels and on the outer side by the continuation of the external glial-limiting membrane (glia limitans) on the surface of the brain (Figure 2). The PVS may provide a mechanism for coordinated changes in activity across multiple areas.^29,30,35 It has been proposed, for example, that distribution by this route of CSF containing neuroactive substances is important for onset and maintenance of specific behavioral or motivational states (e.g., fear, appetite, mood, circadian rhythms).^29,30

In recent animal studies (rodent, canine, nonhuman primate), it has been shown that even rather large-molecular-weight substances are carried into the brain by the flow of subarachnoid CSF into the PVS surrounding penetrating arterioles.^{12–16,27,36} Arterial pulsations are an important driving force, and movement of substances via the PVS route was virtually abolished by piercing the dura, indicating that hydraulic integrity is essential for maintaining this flow.^36–38 Larger-molecular-weight substances were retained within the PVS, likely because of the filtering function of the encompassing astrocytic endfeet (Figure 2). Small lipophilic tracers were also retained within the PVS and rapidly cleared.³⁷ Water and lower-molecular-weight substances in the periarterial PVS followed the basement lamina of the arterioles to the capillary bed, passing freely into the ECS along this route.^12,13 There is some controversy on the path taken thereafter. Clearance via the basement membranes in the walls of arteries to the cervical lymph nodes is one reported route.^13,38,39 Alternatively, clearance via the perivenous PVS into the subarachnoid CSF has been reported.^12,36,37 Subarachnoid CSF is removed by uptake via arachnoid villi into the venous sinuses and via the perineuronal spaces surrounding cranial nerves (e.g., olfactory, optic) to drain into the nasal and cervical lymphatics.^29,30,32,36 Although clearly present in many species, a recent study has questioned the importance of the second route in primates.¹⁴ The flow of CSF into the brain via the PVS may also provide a new route for CNS medication administration. Recent animal studies (nonhuman primate, canine, rodent) have demonstrated that substances can be delivered into the brain parenchyma after lumbar intrathecal injection (Figure 3).^14–16

Overall, this system provides a path by which substances can be both circulated throughout and removed from the brain parenchyma.^{12–16,30,36,38,39} A recent (rodent) animal study reported that the rate of flow was most rapid during sleep, slowing abruptly when animals were awakened.⁴⁰ This was a result of a reduction in the volume fraction of the ECS, which decreased from about 22%−24% (asleep or anesthetized) to about 13%−15% (awake). As would be predicted, clearance of substances was much faster during sleep, when the ECS was largest. The authors noted that these results suggest that one important function of sleep is efficient clearance of toxins and wastes from the brain by this pathway.⁴⁰ Any diminution in this flow through the ECS could impair clearance of solutes (e.g., amyloid beta), and this has been proposed as an important pathophysiological mechanism for several neurodegenerative diseases.^12,38 It may also contribute to sequelae of low-level brain injury (microinfarction, mild traumatic brain injury).^12,38 These conditions may trigger a prolonged widespread reactive gliosis and/or focal disruption of vascular perfusion. Both reactive gliosis and focal loss of vascular perfusion are associated with impaired bulk flow through the ECS, which may promote accumulation of neurotoxic substances.^12,38 Volume transmission also provides a route for toxic substances in the blood or CSF to reach the brain parenchyma. This route has been implicated in both multiple sclerosis and viral encephalitis.³⁰ It may also be the origin of the neuropsychiatric symptoms that can occur after some systemic infections.⁴¹

Recent scientific evidence is leading to a greater understanding of the complexities and intricacies of chemical communication in the brain. The importance and relevance of both fast/precise and slow/summative types of communication have yet to be fully appreciated. Advancing knowledge is gradually elucidating the roles that volume transmission plays in supporting normal function. Also, this mode of chemical communication provides new avenues with great potential for intervening in neurodegenerative conditions, after acute trauma, and in the development of novel therapeutics for chronic psychiatric conditions that, to date, have had limited available treatments.