Striatal neurons get a kick out of dopamine

Dopamine (DA) is critically involved in reinforcement learning, and malfunction in dopamine signalling is associated with numerous brain disorders, including attention deficit hyperactivity disorder, Parkinson's disease, drug abuse and schizophrenia. DA neurons are spontaneously active, but diverge from the baseline activity with brief bursts or pauses when animals are presented with cues that predict future reward (punishment or omission of expected reward counts as negative reward and causes a brief pause in DA cell firing). It is often assumed that phasic variations in striatal DA concentration are integrated by postsynaptic D1‐like and D2‐like receptors (Dreyer et al. 2016). These are linked to biochemical cascades that regulate the synaptic strength of cortical inputs to the basal ganglia that underlie incentive‐based behaviour. Consequently compounds that amplify DA signalling are strongly addictive (Di Chiara et al. 2004). But are D1‐ and D2‐regulated cascades fast enough for real‐time processing of reward? Are phasic changes in DA levels sufficient to activate these cascades? What concentration of DA is required to activate or inhibit these cascades? Since the pathways in question involve complex G‐protein‐regulated biochemical cascades, the question of how temporal variations in DA levels affect postsynaptic neurons has been difficult to address experimentally.

In this issue of The Journal of Physiology, Yapo et al. (2017) provide important new information on D1‐ and D2‐regulated signals in striatal slices. Using UV‐uncaged DA they examined activation of genetically encoded biosensors for cAMP and protein kinase A (PKA). Further, they constructed mathematical models of D1 and D2 receptor‐regulated signalling to simulate conditions that resemble the expected natural reward signals. Thereby they showed that brief variations in DA signals are sufficient to elicit changes in second messenger signals and thereby possibly mediate long‐term plasticity.

Yapo et al. (2017) first showed that levels of cAMP in MSNs relate to phasic changes in dopamine: DA transients evoked a transient increase in cAMP levels in D1‐positive MSNs while in D2‐positive MSNs DA transients could inhibit cAMP evoked by an adenosine A2A receptor agonist. They then investigated intracellular signalling downstream of cAMP. Here they used AKAR3, a surrogate substrate of PKA that reports the equilibrium state of PKA and phosphatase (for example PP1) activity. Thus while in D1‐MSNs cAMP accumulation also led to increased phosphorylation status, the corresponding link between cAMP and AKAR3 signals was much weaker in D2‐MSNs. The dose–response relationship for DA activation and inhibition of MSNs revealed that D1‐MSNs and D2‐MSNs responded equally potently to transient DA stimulation, challenging the common notion that D2 receptors are more potent than D1 receptors.

These data were next used to constrain mass‐action models for D1‐ and D2‐regulated signalling (Nair et al. 2015). The models included the canonical signalling cascades from DA receptors, over G‐proteins, adenylate cyclase, cAMP, PKA, to phosphorylation of T34 on DARPP‐32. The biosensors Epac SH150 and AKAR3 were explicitly included while setting up the models, but excluded when the models were used to make predictions regarding the functional system. Importantly, the authors also included the effect of T34 via PP1 on the AKAR3 construct.

In models of D2 cells, A2A receptors and D2 receptors competed for activation of adenylate cyclase, but downstream of cAMP the models of D1 and D2 pathways were identical. So how was the experimental observation of divergent AKAR3 signals in D1 and D2 pathways handled in the model? The authors noted that the basal level of DARPP‐32 phosphorylation is higher in D2 cells than in D1 cells. This could be simulated by reducing the activity of PP2B in the D2‐regulated pathway causing increased dephosphorylation of AKAR3 by PP1. Simulations of the native D2 system showed that PKA was indeed influenced by phasic DA. The analysis by Yapo et al. (2017) shows that biosensors like AKAR3, which are subjected to multiple competing factors, should be interpreted carefully.

One of the goals of the authors was to investigate if the intracellular signalling was sensitive to fast variations in dopamine levels observed from phasic cell firing, in particular whether a brief reduction of DA from the baseline, for example evoked by negative reward, can activate postsynaptic signals (Nair et al. 2015; Dreyer et al. 2016). This situation is not easily captured by UV‐uncaging experiments. Therefore the authors used their mathematical model to simulate conditions that more closely fit the expected dopamine reward signal. They first showed that brief DA transients could evoke a temporal elevation of cAMP and PKA in D1‐regulated pathways. Conversely, in D2‐regulated pathways, the brief DA transients could transiently block A2A‐evoked cAMP and PKA. They then investigated the effect of a sudden reduction from a constant baseline. When the DA baseline was sufficiently high to inhibit A2A‐mediated cAMP, brief dips in DA, around 1–2 s, would elicit a strong response in cAMP in D2 regulated signals. Furthermore, the response of PKA scaled with the duration of the phasic dip.

The work by Yapo et al. (2017) expands our knowledge of the interplay between dynamic variations and steady‐state baseline in DA signalling and shows great promise for future experiments. For example, partial DA cell loss has been proposed to reduce the dynamics of DA signals and the resulting compensatory changes in postsynaptic signalling may be a symptom‐generating mechanism for early stages of Parkinson's disease (Dreyer, 2014). It could be interesting to see if the physiological approach by Yapo et al. (2017) can pinpoint how postsynaptic pathways adapt to long‐term changes in DA physiology mediated by lesions, drugs of abuse or long term medication.

Additional information

Competing interests

None declared.

Author contributions

Both authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.