Date of Completion


Embargo Period



Neuroscience, Adenosine, Neocortex, Synaptic Plasticity, Plasticity, Neuromodulation, Homeostasis

Major Advisor

Maxim Volgushev

Associate Advisor

John Salamone

Associate Advisor

Merce Correa

Field of Study



Doctor of Philosophy

Open Access

Campus Access


Changes in the strength of synaptic connections are widely believed to be the neural basis for learning and memory. Associative homosynaptic plasticity dictates that inputs which contribute to the firing of a postsynaptic neuron (i.e., inputs which initiate postsynaptic potentials in close preceding temporal proximity to postsynaptic firing) should be potentiated, while inputs whose activity does not help drive firing should be depressed. This classic Hebbian learning rule suffers from a major caveat. Under these rules, potentiated inputs will have the tendency to continue potentiating, and depressed synapses will continue to depress. These ‘runaway synaptic dynamics’ must be constrained by additional plasticity rules to maintain a balance of learning and synaptic homeostasis. Heterosynaptic plasticity, or changes in the synaptic strength of connections which were not active during an induction protocol, is a co-occurring process and has been shown in models to serve this homeostatic role. In heterosynaptic plasticity, the direction of plastic changes is correlated with the input’s initial paired-pulse ratio (PPR; itself inversely related to the release probability at the presynaptic terminal). Weak inputs get stronger, strong inputs tend to get weaker, and thus heterosynaptic plasticity is weight-dependent.

As a metabolite of ATP, adenosine might be the most widespread neuromodulator in the brain. Neurons and astrocytes release adenosine and ATP in an activity-dependent manner. Consequently, adenosine is implemented in a multitude of functions associated with physiological and pathological alterations of brain activity. In the present work, we induce both homosynaptic and heterosynaptic plasticity on the background of differing degrees of adenosine receptor activation. We demonstrate that fluctuations in adenosine concentration alter the strength of weight-dependent plasticity, and by consequence alter the ability of heterosynaptic plasticity to constrain runaway dynamics. Blockade of endogenous adenosine tone acting on A1 receptors abolished the weight-dependence of plasticity. Simulations replicating this result found synaptic weight distributions were prone to runaway dynamics. In contrast, increasing adenosine tone strengthened the weight-dependence of plasticity. Simulations modeling this result demonstrated strong homeostatic regulation of synaptic weights. Thus fluctuating adenosine concentrations shift regimes of plasticity toward those dominated by associative learning rules or those dominated by homeostatic learning rules.