Questions: Excitatory-Inhibitory Balance in Neural Circuits

Question 1

A researcher chronically blocks all GABA receptors in a cortical circuit. What is the most likely outcome?

• A. The circuit becomes more efficient at processing information because inhibition was wasting energy
• B. Firing rates stabilize at a new lower baseline as neurons adapt to the absence of inhibition
• C. The circuit enters runaway excitatory activity, producing seizure-like events
• D. Homeostatic plasticity immediately compensates by reducing excitatory synapse density to pre-block levels

Answer: (C)

Removing inhibition destroys the thermostat — there is nothing to counteract ongoing glutamatergic excitation, so neurons fire more, which excites neighbors, which excites more neurons: a seizure. Option D is tempting because homeostatic plasticity does exist, but it operates on timescales of hours to days, not instantly. Seizures emerge from acute E/I imbalance, before compensation can occur.

Question 2

Postmortem studies find that patients with schizophrenia have reduced density of parvalbumin-positive interneurons in prefrontal cortex. What does the E/I framework predict about gamma-band oscillations in these patients?

• A. Gamma oscillations should increase in amplitude because reduced inhibition allows more excitatory activity
• B. Gamma oscillations should be disrupted because fast-spiking interneurons provide the rhythmic inhibitory timing that generates them
• C. Gamma oscillations should be unaffected because they are generated by excitatory neurons, not inhibitory ones
• D. Gamma oscillations should slow to alpha frequency as inhibition decreases

Answer: (B)

Parvalbumin-positive interneurons are fast-spiking and provide precisely timed inhibitory feedback that periodically resets excitatory activity — the mechanism that generates gamma oscillations (30–80 Hz). Losing these interneurons doesn't just reduce inhibition; it desynchronizes the circuit. Option A reflects the misconception that less inhibition simply means more excitation; in reality, the rhythmic structure of network activity depends on timed inhibitory pacing.

Question 3

Homeostatic plasticity operates on slower timescales than Hebbian plasticity.

• T. True
• F. False

Answer: True

Hebbian plasticity (LTP/LTD) modifies synaptic strength over minutes following coincident activity. Homeostatic plasticity — including synaptic scaling — operates over hours to days, adjusting the overall excitability of a neuron in response to chronic over- or underactivity. This difference in timescale is functionally important: Hebbian mechanisms encode specific learning, while homeostatic mechanisms maintain the network within a stable operating range.

Question 4

The E/I balance is a fixed property of mature neural circuits, set during development and unalterable in adulthood.

• T. True
• F. False

Answer: False

E/I balance is continuously and dynamically regulated throughout life via homeostatic plasticity. When a network is chronically overactive, mechanisms such as synaptic scaling down of excitatory receptors and upregulation of inhibitory synapses restore balance. Conversely, chronic underactivity triggers compensatory scaling up of excitation. This ongoing regulation is why E/I disruptions can emerge — and potentially be treated — in adult organisms.

Question 5

Why does dysfunction of parvalbumin-positive inhibitory interneurons produce more than just a reduction in overall inhibition in cortical circuits?

Model answer: Parvalbumin interneurons are fast-spiking cells that provide precisely timed inhibitory feedback to populations of excitatory neurons. This timing generates the rhythmic synchronization (gamma oscillations) underlying coherent cortical processing. When these interneurons are lost, the circuit doesn't just become 'less inhibited' — it loses the rhythmic pacemaking that coordinates activity across neurons, desynchronizing the circuit entirely and disrupting functions like working memory and perceptual binding.

This illustrates a core principle of E/I balance: inhibition is not merely a brake on excitation but a sculptor of temporal structure in neural networks. A circuit with random inhibition and a circuit with phasic, timed inhibition have vastly different computational properties, even if their average inhibitory tone is similar.