Real-time imaging of hippocampal network dynamics reveals trisynaptic induction of CA1 LTP and "circuit-level" effects of chronic stress and antidepressants

Today’s pervasive presence of stress renders stress-related
psychiatric disorders (SRPDs), a relevant global health problem.
Memory impairment is a major symptom likely mediated by the
hippocampus (HIP), a limbic brain region highly vulnerable to
stress. Recent evidence suggests that information processing
problems within specific neuronal networks might underlie SRPDs.
However, the precise functional neurocircuitry that mediates
hippocampal CA1 long-term potentiation (LTP), a putative correlate
of mammalian learning and memory, remains unknown at present.
Furthermore, valuable assays for studying stress and drug effects
on polysynaptic activity flow through the classical input/output
circuit of the HIP are missing. To engage a circuit-centered
approach, voltage-sensitive dye imaging was applied in mouse brain
slices. Single pulse entorhinal cortex (EC) to dentate gyrus (DG)
input, evoked by perforant path stimulation, entailed strong
neuronal activity in the DG, but no distinct neuronal activity in
the CA3 and CA1 subfield of the HIP. In contrast, a thetafrequency
(5 Hz) stimulus train induced waves of neuronal activity
percolating through the entire hippocampal trisynaptic circuit
(HTC-waves). Spatially restricted blocking of glutamate release at
CA3 mossy fiber synapses caused a complete disappearance of
HTC-waves, suggesting frequency facilitation at DG to CA3 synapses
the pivotal gating mechanism. In turn, non-theta frequency
stimulations (0.2/1/20 Hz) proved much less effective at generating
HTC-waves. CA1 long-term potentiation (CA1 LTP) is the best
understood form of synaptic plasticity in the brain, but
predominantly at the monosynaptic level. Here, HTC-waves comprise
high-frequency firing of CA3 pyramidal neurons (>100 Hz),
inducing NMDA receptordependent CA1 LTP within a few seconds.
Detailed examination revealed the existence of an induction
threshold for LTP. Consequently, baseline recordings with a reduced
number of HTC-waves were carried out to test the effects of memory
enhancing drugs and HPA axis hormones on hippocampal network
dynamics. Bath application of caffeine (5 mM), corticosterone (100
nM) and corticotropin-releasing hormone (5 & 50 nM) rapidly
boosted HTC-waves. Cognitive processes taking place within the HIP
are challenged by stress exposure, but whether and how chronic
stress shapes "net" neuronal activity flow through the HIP remains
elusive. The HTC-wave assay, refined for group comparisons,
revealed that chronic stress markedly lowers the strength of evoked
neuronal activity propagation through the hippocampal trisynaptic
circuit. In contrast, antidepressants (ADs) of several classes, the
mood stabilizer lithium, the anesthetic ketamine, and the
neurotrophin brainderived neurotrophic factor amplified HTC-waves.
An opposite effect was obtained with the antipsychotic haloperidol
and the anxiolytic diazepam. The tested ADs exert this effect at
low micromolar concentrations, but not at 100 nM, and nearly
always, also not at 500 nM. Furthermore, the AD fluoxetine was
found to facilitate LTP of HTC-waves. Finally, pharmacological
blockade of the tyrosine-related kinase B receptor abolished
fluoxetine effects on HTC-waves. These results highlight a
circuit-centered approach suggesting evoked synchronous theta
rhythmical firing of EC principal cells as a valuable tool to
investigate several aspects of neuronal activity flow through the
HIP. The physiological relevance is emphasized by the finding that
the resulting HTC-waves, which likely occur during EC theta
oscillations, evoke NMDA receptor-dependent CA1 LTP within a few
seconds. Furthermore, HTC-waves allow to integrate molecular,
cellular and structural adaptations in the HIP, pointing to a
monoaminergic neurotransmission-independent, "circuit-level"
mechanism of ADs, to balance the detrimental effects of chronic
stress on HIP-dependent cognitive abilities.