De‐mobilisation of NMDA receptors in midbrain dopamine neurons: a quantum of reward?

Abstract

Dopamine neurons in the midbrain ventral tegmental area (VTA) are a key component of the brain reward system. Dopamine neurons increase their action potential firing when unexpected salient events such as rewards are encountered and decrease their firing if an expected reward is unforthcoming; they encode errors in reward prediction (Schultz, 2016). The change from irregular low-frequency firing to high-frequency bursts of firing, or to no firing, reflects both their intrinsic ionic conductances and the balance of their excitatory and inhibitory synaptic inputs (Marinelli et al. 2006). This has important health implications; for example, many drugs of abuse alter either the intrinsic firing properties or the synaptic inputs of VTA dopamine neurons (Francis et al. 2019), hijacking brain reward pathways. Glutamatergic activation of NMDA receptors has long been considered a key driver of burst firing in midbrain dopamine neurons (Marinelli et al. 2006). But how exactly does this happen? In this issue of the journal Etchepare et al (2021) bring a fresh approach to this question by using single molecule tracking, whereby a quantum dot of fluorescence on individual NMDA receptors in the membrane can be microscopically visualised. This enabled them to explore for the first time the membrane dynamics of NMDA receptors in midbrain dopamine neurons. The authors first used dissociated cell cultures of tdTomato-expressing dopamine neurons from the midbrain nuclei (including VTA) of neonatal mice (Etchepare et al. 2021). Quantum dots of fluorescence were associated with an antibody to the NMDA receptor obligatory subunit, GluN1. The authors observed that individual NMDA receptors in dopamine neuron dendrites were highly mobile; over 40% moved at a rate of more than 100 nm/s (Etchepare et al. 2021). They travelled impressive dendritic distances; in one example, a single NMDA receptor moved back and forth across 6 μm of dendrite. When NMDA receptors were activated by agonist binding, this mobility was decreased. This provides a fascinating snapshot of the change in dynamics of NMDA receptors in the membrane of cultured midbrain dopamine neurons as the receptor changes to the agonist-bound state. Whether NMDA receptors show the same changes in mobility in vivo remains to be tested. The authors next explored the importance of this change in membrane mobility for spike firing (Etchepare et al. 2021). They used rat brain slices and identified VTA dopamine neurons using electrophysiological and histochemical markers. They first confirmed the well-established effect of NMDA to make firing more irregular and more frequent. Next, they explored the question of whether NMDA receptor activation per se, or their reduced mobility specifically, is important for this effect. Antibodies that cross-link NMDA receptors were used to reduce their mobility in the membrane; this was confirmed in midbrain cell cultures. What effect would this have on action potential firing? Rats were given either the cross-linking antibody or a control antibody and recordings were made in midbrain slices. VTA dopamine neurons from rats given the cross-linking antibody showed more irregular firing than those from rats given the control antibody; in other words, reducing the mobility of NMDA receptors had the same effect as NMDA receptor activation on spike firing. Importantly, the authors showed that cross-linking NMDA receptors had no effect on their surface expression, agonist binding or ionic permeability. Thus, NMDA receptors are present and functional, but their mobility is impaired and spike firing is de-regularised, suggesting that de-mobilisation is the critical event. In midbrain dopamine neurons in brain slices, regular pacemaker firing is the predominant spike pattern and burst firing is rarely seen (Marinelli et al. 2006). It is not yet known whether decreased NMDA receptor mobility in vivo would convert irregular firing into bursts of action potentials, an important change that encodes positive reward prediction errors. If it does, then the de-mobilisation of individual NMDA receptors in the dendrites of VTA dopamine neurons may represent a quantum of reward. A potentially interesting addendum is that the NMDA receptor cross-linking also affected the small conductance calcium-dependent potassium (SK) channels. SK channel activity sustains regular pacemaker firing in midbrain dopamine neurons in vitro and blocking them with apamin causes irregular firing (Marinelli et al. 2006). Following cross-linking of NMDA receptors, the authors noted a reduced effect of apamin on pacemaker firing (although this did not reach significance). This might indicate less contribution of SK channels to pacemaker firing when NMDA receptors are immobilised; however, it might also reflect the higher baseline level of irregularity seen after cross-linking NMDA receptors. No changes in SK channel expression were observed in VTA dopamine neurons in brain slices. Using cultured dopamine neurons, the authors detected an altered distribution of SK channel subunits in the dendrites following NMDA receptor cross-linking. Therefore, one possible explanation for the findings of this study is that NMDA receptor activation and the concomitant reduction in NMDA receptor mobility causes redistribution of SK channels, uncoupling them from NMDA receptor-mediated calcium influx and thus promoting irregular firing. Further work is needed to substantiate this idea. Single molecule quantum dot tracking is well established in hippocampal neurons as a means of visualising membrane receptor dynamics at high resolution (Varela et al. 2016). Hitherto it has not been applied in midbrain dopamine neurons. Given the proposed role of NMDA receptors in firing patterns relevant to reward and addiction, this is an important study. The findings open the door to further work on the role of glutamate receptor mobility in synaptic plasticity in VTA dopamine neurons, including the many examples of drug-induced synaptic potentiation.