Serotonin enhances excitability and gamma frequency temporal integration in mouse prefrontal fast-spiking interneurons

  1. Jegath C Athilingam
  2. Roy Ben-Shalom
  3. Caroline M Keeshen
  4. Vikaas S Sohal  Is a corresponding author
  5. Kevin J Bender  Is a corresponding author
  1. University of California, San Francisco, United States
8 figures, 1 table and 1 additional file

Figures

Figure 1 with 3 supplements
Serotonin alters intrinsic properties to increase FSI excitability.

(A) Experimental design: we recorded from fast-spiking interneurons labeled in a PV-Cre:: Ai14 in mPFC (top). Images of a recorded neuron in DIC and showing tdTomato expression (bottom). (B) Example …

https://doi.org/10.7554/eLife.31991.002
Figure 1—figure supplement 1
Modulation of FSI intrinsic properties by 5HT.

(A) Membrane potential (left axis) and input resistance (right axis) over time during application of 5HT. (B–C) Membrane potential (B) and input resistance (C before (averaged −5 min to 0 min) and …

https://doi.org/10.7554/eLife.31991.003
Figure 1—figure supplement 2
Dose response for 5HT.

(A) Change in membrane potential with various doses of 5HT. 0 mV indicates no change. (B) Percentage change in input resistance with various doses of 5HT. 100% indicates no change.

https://doi.org/10.7554/eLife.31991.004
Figure 1—figure supplement 3
5HT does not change membrane potential or input resistance of SOM interneurons.

(A) Experimental design: We recorded from somatostatin (SOM)-expressing interneurons labeled in a SOM-Cre:: Ai14 in mPFC. Example somatostatin (SOM)-expressing interneuron responses to …

https://doi.org/10.7554/eLife.31991.005
5HT reduces conductance through inward rectifying potassium channels.

(A–B) Top: Current recorded during a voltage ramp (3 s) from −150 mV to −50 mV before (black) and after 5HT (blue) using KGluconate in the internal solution (A) and with pre-application of the 5HT2A …

https://doi.org/10.7554/eLife.31991.006
Figure 3 with 1 supplement
Decreasing dendritic K + conductance elicits change in tau of synaptic responses in a compartmental model.

(A) Morphology of FSI model. Black circle represents location of synapse. (B) Change in membrane potential (left axis) and input resistance (right axis) in response to reducing conductance of K + cha…

https://doi.org/10.7554/eLife.31991.007
Figure 3—figure supplement 1
5HT-induced dendritic depolarization reduces synaptic driving force in compartmental model.

(A) Experimental design: Synapse was stimulated on dendrite and response was recorded at site of synapse in either voltage clamp or current clamp. (B) EPSP traces recorded in current clamp at …

https://doi.org/10.7554/eLife.31991.008
Figure 4 with 1 supplement
Local 5HT iontophoresis at FSI dendrites increases FSI firing.

(A,C) Experimental design: Neurons were patched and filled with Alexa-488. 5HT was applied locally to the dendrite using iontophoresis (50 ms), while FSIs were injected with a small amount of …

https://doi.org/10.7554/eLife.31991.009
Figure 4—figure supplement 1
Dendritic 5HT iontophoresis depolarizes neuron sufficiently to induce observed change in firing rate.

(A) Average membrane potential recorded at the soma during dendritic iontophoresis with spikes removed via median filtering. Blue line indicates time of iontophoresis. (B) Change in membrane …

https://doi.org/10.7554/eLife.31991.010
Figure 5 with 1 supplement
5HT promotes integration of synaptic inputs in a frequency-specific manner.

(A) Experimental design: Slices were bathed in a caged glutamate compound (MNI-Glutamate 2.5 mM) that is only biochemically active with photolysis. Glutamate was uncaged at five locations (1 µm …

https://doi.org/10.7554/eLife.31991.011
Figure 5—figure supplement 1
Broad 5HT iontophoresis over the slice produces typical 5HT effects.

(A) Experimental design: FSIs were patch clamped and an iontophoretic pipette containing 5HT (200 mM, pH = 4.5) or vehicle control solution (aCSF, pH = 4.5) was hovered just above the slice. (B–D) …

https://doi.org/10.7554/eLife.31991.012
Figure 6 with 2 supplements
Modeling indicates that changing tau and reducing K + conductance can modulate temporal summation.

(A) Experimental design: Single EPSPs were modeled using a double exponential with two different decay constants (taubaseline = 15 ms, tau5HT = 23 ms) to match the change in tau observed with 5HT …

https://doi.org/10.7554/eLife.31991.013
Figure 6—figure supplement 1
Model is robust to changes in synaptic parameters.

(A) Experimental design: Five synapses were placed either on the same dendrite (1 µm apart, arrowhead, black), separate dendrites (blue), or soma (green). The synapses were stimulated at varying …

https://doi.org/10.7554/eLife.31991.014
Figure 6—figure supplement 2
Background synaptic noise does not change summation enhancement.

(A) Experimental design: Synaptic noise was included at various levels in a subset of dendrites of the compartmental model and synapses in model (Figure 6C) were stimulated at variable ISIs. Example …

https://doi.org/10.7554/eLife.31991.015
Figure 7 with 2 supplements
Mimicking 5HT effects elicits preferential firing to gamma frequency inputs in FSIs.

(A) Experimental design: The Gq-coupled Designer Receptor Exclusively Activated by Designer Drugs (DREADD) was expressed specifically in FSIs using a Cre-dependent virus injected into PV-Cre mice. …

https://doi.org/10.7554/eLife.31991.016
Figure 7—figure supplement 1
CNO has no effect on FSIs not expressing DREADD.

(A–B) Change in membrane potential (A) and input resistance (B) over time with application of CNO in FSIs that either express the Gq-coupled DREADD, hM3D(Gq)-mCherry (blue) or a control fluorophore, …

https://doi.org/10.7554/eLife.31991.017
Figure 7—figure supplement 2
5HT2A agonist increases probability of FSI firing in response to gamma frequency inputs.

(A) Experimental design: A stimulating electrode was placed in the tissue within 100 µm of the recorded FSI and a 2 s train of randomly distributed stimulating current pulses (200 µs) with varied …

https://doi.org/10.7554/eLife.31991.018
Mimicking 5HT effects in FSIs produces gamma frequency events in downstream pyramidal neurons.

(A) Experimental design: The Gq-DREADD was expressed specifically in FSIs using a cre-dependent virus injected into PV-Cre mice. Prefrontal slices were bathed in carbachol (2 µM) to induce …

https://doi.org/10.7554/eLife.31991.019

Tables

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
genetic reagent (M. musculus)PV-CreJackson LaboratoryStock#:017320
genetic reagent (M. musculus)Ai14Jackson LaboratoryStock#:007914
genetic reagent (M. musculus)SERT-CreJackson LaboratoryStock#:014554
transfected construct (virus)AAV5-Ef1-DIO-ChR2-eYFPUNC Vector CoreAAV5-Ef1a-DIO-hChR2(H134R)-EYFP-WPRE-pA
transfected construct (virus)AAV5-CaMKII-ChR2-eYFPUNC Vector CoreAAV5-CaMKIIa-hChR2(H134R)-EYFP
transfected construct (virus)AAV-DJ-Ef1a-DIO-hM3D(Gq)-mCherryStanford Vector CoreGVVC-AAV-130
transfected construct (virus)AAV-DJ-Ef1a-mCherryStanford Vector CoreGVVC-AAV-14
chemical compound, drugDL-AP5TocrisCatalog#:3693
chemical compound, drugCNQXTocrisCatalog#:1045
chemical compound, drugGabazineTocrisCatalog#:1262
chemical compound, drug5HTTocrisCatalog#:3457
chemical compound, drugMDL100907TocrisCatalog#:4173
chemical compound, drugα−methyl−5ΗTTocrisCatalog#:0557
chemical compound, drugCarbacholTocrisCatalog#:2810
chemical compound, drugMNI-GlutamateTocrisCatalog#:1490
antibody (rabbit)Rabbit anti-5HTImmunostarCatalog#:200801:500
antibody (mouse)mouse anti-GFPInvitrogenCatalog#:A111201:500
antibody (goat)Alexa 405 goat anti-rabbitInvitrogenCatalog#:A315561:250
antibody (goat)Alexa 488 goat anti-mouseInvitrogenCatalog#:A110291:250

Additional files

Download links