6. INTERACTIONS WITH OTHER TRANSMITTERS AND MODULATORS

6.1 Glycine and D-Serine

The existence of a modulatory site for glycine on the NMDA receptor was known for some years before the discovery of the various NR subunits (Johnson & Ascher, 1987). Thus, the NMDA receptor is allosterically regulated by a binding site at which glycine and other compounds such as L- or D-serine act as agonists (in the micro-molar range) in a strychnine insensitive manner (Johnson and Ascher, 1987). It appears that agonist binding at this site is a requirement for receptor activation (Kleckner & Dingledine, 1991). It is intriguing that the NR1-NR2A subunit combination has the lowest affinity for glycine (Laurie & Seeburg, 1994), and that there is a considerable amount of the NR2A subunit in the lateral thalamus. One interpretation of this data could be that in the NMDA-receptor glycine site is not so easily saturated in the thalamus, and thus is available for modulation. Indeed, in the VB complex, it has been shown that D-serine can potentiate NMDA-receptor-mediated responses in vivo (Salt, 1989), indicating that the glycine site is not saturated. Furthermore, it is now known that D-serine can be found in the vertebrate brain, probably in glial elements (Hashimoto et al., 1992; Schell et al., 1995). It is thus possible that this amino acid is the endogenous modulator at the glycine site, and that it is tonically released. This is supported by the finding that ligands acting as glycine site antagonists can reduce NMDA-receptor-mediated responses in both the LGN and the VB (Levy et al., 1990; Salt, 1987).

6.2 Amines

The diffuse innervation of thalamic nuclei by cholinergic, serotonergic and noradrenergic fibres arising from the brainstem has been extensively studied and reviewed (McCormick, 1989, 1992). In particular, the ways in which these amines can modulate the various classes of thalamic neurones and how this is important for state-dependent effects in the thalamo-cortical circuit have been determined by a series of elegant studies (reviews: McCormick, 1992; Steriade et al., 1990; Steriade 1993; Steriade et al., 1993).

Noradrenaline (acting at -adrenoceptors) and acetylcholine (acting at muscarinic receptors) can depolarise thalamic relay neurones and allow them to become more sensitive to other synaptic inputs. These effects are mediated via a pertussis toxin-insensitive G-protein and result in the reduction of a K+ conductance, termed IKleak (McCormick, 1992). Interestingly, the action of one transmitter acting via one of these intrinsic membrane conductances can occlude the action of another transmitter acting via the same conductance (McCormick, 1992; McCormick & Prince, 1987; Pape and McCormick, 1989), suggesting that simultaneous activation of all of these modulatory influences may not necessarily have a greater effect than maximal activation of one. It has been shown that activation of postsynaptic metabotropic glutamate receptors in the LGN with 1S,3R-ACPD is also exerted via this same common mechanism (McCormick & Von Krosigk, 1992).

Noradrenaline (acting at -adrenoceptors) and serotonin (via an action at 5HT receptors other than 5HT1A or 5HT2 receptors) modulate thalamic activity by a different intrinsic membrane mechanism. Both transmitters enhance Ih , a strongly voltage-dependent conductance, activated by hyperpolarisation and carried by Na+/K+. These actions effectively result in the reduction of neuronal responsiveness to hyperpolarising inputs (McCormick & Pape, 1990; Pape and McCormick, 1989). As described above for acetylcholine and noradrenaline, these responses to serotonin and noradrenaline can also occlude each other.

Interactions between serotonin and ionotropic glutamate receptor-mediated-responses have been demonstrated in the rat VB complex in vivo: serotonin was able to enhance NMDA and non-NMDA responses and in some cases sensory responses (Eaton & Salt, 1989). It is unclear as to whether this represents an interaction at the receptor level. It is more likely that this reflects an action of serotonin via Ih, dampening recurrent inhibitory input from the nrt evoked by sensory stimuli or iontophoretically applied agonists. It is particularly notable that 5HT greatly enhances sensory responses which are normally strongly controlled by sensory-evoked inhibitory input (Eaton & Salt, 1989).

6.3 Glutathione

A modulatory site on the NMDA receptor which is sensitive to sulphydryl redox agents was first described by Aizenman et al. (1989). More recently, it has been shown that the endogenous redox agent glutathione can selectively modulate the NR1-NR2A receptor-mediated responses: glutathione, in its reduced form, enhanced responses (Koehr et al., 1994). This is particularly interesting, as it has been shown that glutathione is released from brain slices in a Ca2+-dependent manner, suggesting it may have a transmitter/modulator role (Zaengerle et al., 1992). It is at present not known from which cell types glutathione is released, but reduced glutathione immunostaining has been observed in cerebellar neurones and glia (Hjelle et al., 1994). Furthermore, both the oxidised and reduced forms of glutathione have been shown to reduce responses of VB neurones to sensory stimuli and ionotropic glutamate receptor agonists (Shaw & Salt, 1995).

6.4 Nitric Oxide

The nitric oxide (NO) system has been the subject of much study since the discovery that it plays an important part in neural signalling (Garthwaite, 1993). NO is a free radical gas which can diffuse across membranes rapidly, thus acting on neural elements which are at some distance from the site of production. One of the modes of action of NO is to stimulate soluble guanylate cyclase, leading to an increase in intracellular cyclic GMP in target cells, and this can then lead to further effects, depending on the cell. NO is synthesised from L-arginine by the action of NO synthase (NOS), with the production of L-citrulline (Garthwaite, 1993). A variety of techniques have been used to demonstrate a widespread distribution for NOS within the brain, including in the thalamus and the cerebral cortex, predominantly in neurones but also in astrocytes (Bickford et al., 1993; Bredt & Snyder 1992; Valtschanoff et al., 1993; Vincent & Kimura, 1992). Some of these neurones are also immunoreactive for citrulline (Pasqualotto et al.., 1991). There is evidence to suggest that NOS in the thalamus is located in the terminals of cholinergic brainstem afferents (Bickford et al., 1993; Sugaya & McKinney, 1994; Vincent & Kimura 1992). The only anatomical evidence regarding arginine in the thalamus to date is scanty, but indicates that it is localised in glia (Aoki et al., 1991; Petrusz et al., 1995). A diversity of cellular compartments suggests not only that NO may be an active end-product but that, as intermediates may move between compartments, L-arginine and L-citrulline might have a signalling function in their own right. This is speculative, but is supported by the release of arginine upon stimulation of pathways in cerebellar slices (Hansel et al., 1992) and in the thalamus in vivo (Do et al., 1994).

Interactions of NO with excitatory amino acids have been described in addition to direct post-synaptic effects, and it has been suggested that NO may have presynaptic actions, regulating the release of transmitters. These diverse effects may be related to regional differences, but also to the diverse actions of the various components of the NO/arginine cycle described above (Garthwaite 1993). In the thalamus, it has been shown that NO donors and/or L-arginine can have a direct post-synaptic effect (presumably mediated via a cyclic GMP system) on thalamic relay neurones (Cudeiro et al. 1994; Do et al., 1994; Pape & Mager, 1992; Shaw & Salt, 1995), which can in turn affect the responses of these cells to sensory stimuli and excitatory amino acids.

NOTE Please note: this document is part of the HTML version of a paper originally published in print in Progress in Neurobiology.

contents Contents back Previous Section Forward Next Section References Reference List

Return to Tom Salt's Home Page