Please visit the new BioModels platform to access the latest content. This website is no longer updated and will be retired on 20 July 2020.
BioModels Database logo

BioModels Database


Fernandez et al. (2006), Integration of glutamate and dopamine signals

January 2009, model of the month by Michele Mattioni
Original model: BIOMD0000000152

The integration of different signals is one of the most complex and interesting features of the brain. In BIOMD0000000152 Fernandez et al [1] elucidate in which way DARPP-32, a protein present in the spine of the Medium Spiny Neuron, is able to integrate the different stimuli provided by the release of glutamate and dopamine.

The peculiar ability of a synapse to change its own strength according to the series of inputs that it receives from other neurons plays a key role in the learning ability of the brain. Learning, in the basal ganglia, is often connected with reward. Two kinds of responses are able to encode this information: An activity-dependent increase in synapse strength, known as Long Term Potentiation (LTP) and an activity-dependent weakening of the synapse, known as Long Term Depression (LTD).

The capacity of a medium spiny neuron to encode long term memory (LTM) is given by the ability to change its state between those two. The model shows how DARPP-32 is a key player in this delicate equilibrium between these two forms of excitability.

The reactions involving DARPP-32 are depicted in figure 1. Model A features the biochemical network and the receptors present in the nigro-striatal medium spiny neuron, while model B represents a nigro-pallidal medium spiny neuron, which differs slightly in the types of dopamine receptors present (D2 instead of D1) and the corresponding biochemical cascades.

Biological model of DARPP-32 regulation

Figure 1: Biological model of DARPP-32 regulation. Model A shows a nigro-striatal medium spiny neuron, model B a nigro-pallidal medium spiny neuron. Figure taken from [1].

Reactions implemented in the DARPP-32 model

Figure 2: Reactions implemented in the DARPP-32 model. This figure shows all reactions, including the different phosphorylation reactions affecting DARPP-32. Figure taken from [1].

The different effects of DARPP-32 correspond to different phosphorylation states of the protein. In figure 2 the authors explain the different combinations of phosphorylation states in which the protein can be found. As shown, while DARPP-32 can be phosphorylated in four different positions, the model takes only three of them into account, due to the fact that the fourth has not been found to play an important role in the phenomena investigated.

The release of Dopamine increases cAMP inside the cell and DARPP-32 is quickly phosphorylated by PKA (see figure 3). The isoform phosphorylated on threonine 34 is able to inhibit PP1; this inhibition will cause LTP, enabling the synapses to respond to a weaker stimulus.

In contrast, the release of glutamate will cause the phosphorylation of DARPP-32 on serine 137 (see figure 4). The effect of this is to keep DARPP-32 phosphorylated on threonine 34, but this effect only happens if DARPP-32 molecules phosphorylated on threonine 34 are already present in the synapse.

Effect dopamine release on DARPP-32 phosphorylation

Figure 3: Effect of dopamine release on DARPP-32. Dopamine is modelled by a pulse of cAMP. From [1].

Effect of glutamate release on DARPP-32 phosphorylation

Figure 4: Effect of glutamate release on on DARPP-32. Glutamate release is modelled by a train of calcium pulses. From [1].

The time course of DARPP-32 phosphorylation changes completely if the two stimuli occur in quick succession. For example, if the release of dopamine precedes the release of glutamate, DARPP-32 is phosphorylated at different residues than when stimulated by each stimulus alone.

An action followed by a reward is usually encoded in the medium-spiny neuron by the release of glutamate quickly followed by the release of dopamine [2].

It is possible to use this model to simulate the combined effect of glutamate release (Calcium) followed by dopamine release (cAMP) (see figure 6). DARPP-32 presents a completely different phosphorylation pattern from the ones shown above. The protein is phosphorylated on threonine 34. The persistence of this isoform is needed to activate a cascade of biochemical reactions resulting in an increase of the number of AMPA receptors in the postsynaptic density. Due to this the synapse will be able to react more strongly to subsequent stimuli, a key feature of LTP.

Effects of a dopamine release followed by a glutamate release

Figure 5: Effects of a dopamine release followed by a glutamate release. Dopamine release is simulated by a pulse of cAMP, glutamate release by a train of calcium pulses. Figure taken from [1].

Effects of a glutamate release followed by a dopamine release

Figure 6: Effects of a glutamate release followed by a dopamine release. Figure from a custom simulation ran with STEPS [3].

As pointed out by the authors the model should be extended downstream and upstream to include the effect of the other biochemical reactions that were not modelled in this version, however the model is already useful to simulate the integration between glutamate and dopamine and elucidate the critical importance of DARPP-32 as one of the regulators of the delicate equilibrium inside the spine.

Bibliographic References

  1. É. Fernandez, R. Schiappa , J. Girault, N. Le Novère. DARPP-32 Is a Robust Integrator of Dopamine and Glutamate Signals. PLoS Computational Biology 2(12): e176, 2006. [SRS@EBI]
  2. W. Schultz. Getting Formal with Dopamine and Reward. Neuron, 36(2):241-263, 2002. [SRS@EBI]
  3. E. De Schutter and S.M.M. Wils. STEPS: Modeling and simulating complex reaction-diffusion systems with Python. Frontiers in Neuroinformatics, [SRS@EBI]