Serotonin Inhibition of Claustrum Projection Neurons: Ionic Mechanism, Receptor Subtypes and Consequences for Claustrum Computation

Abstract

The claustrum is a small but densely interconnected brain structure that is innervated by axons containing serotonin (5-HT), a neuromodulator that has been implicated in control of sleep and in the actions of psychedelic drugs. However, little is known about how 5-HT influences the claustrum. We have combined whole-cell patch-clamp measurements of ionic currents, flash photolysis, and receptor pharmacology to characterize the 5-HT responses of individual claustral projection neurons (PNs) in mouse brain slices. Serotonin application elicited a long-lasting outward current in claustral PNs. This current was due to an increase in membrane permeability to K⁺ ions and was mediated mainly by the type 1A 5-HT receptor (5-HTR-1A). The 5-HT-induced K⁺ current hyperpolarized, and thereby inhibited, the PNs by reducing action potential firing. Focal uncaging of 5-HT revealed that inhibitory 5-HTR-1As were located at both the soma and dendrites of PNs. We conclude that 5-HT creates a net inhibition in the claustrum, an action that should decrease claustrum sensitivity to excitatory input from other brain areas and thereby contribute to 5-HT action in the brain.

Keywords: claustrum; neuromodulation; serotonin.

Figure 1

Identifying claustral projection neurons via their intrinsic electrical properties. (A) A whole-cell patch-clamp recording from a single claustral neuron in a coronal brain slice. The neuron was filled with neurobiotin, which was labeled by streptavidin (red). The claustrum (CLA) core is identifiable as a PV-rich (cyan) elliptical structure located between the insula and the striatum. The striatum also highly expresses PV but is separated from the cortex by the external capsule. (B) Representative recordings of AP firing (upper traces) evoked by depolarizing currents (lower traces) in PN subtypes PN1 to PN5.

Figure 2

A K⁺ conductance increase mediated the 5-HT response of claustrum PNs. (A) Representative current induced in a claustrum PN in response to local application of 5-HT (100 µM; at arrowhead). The time course of current decay was fitted with a single-exponential function (red). (B) Currents induced by 5-HT in a claustral PN held at different membrane potentials. In this cell, the 5-HT-induced current reversed its polarity at −100 mV. (C) Currents induced by 5-HT in a claustral PN held at various membrane potentials and bathed in a high external K⁺ solution. In this cell, the 5-HT-induced current reversed its polarity between −60 and −70 mV. (D) Relationships between membrane potential and 5-HT-induced currents in normal (black; n = 10) and high external K⁺ (purple n = 6) solutions. The 5-HT responses were sensitive to the electrochemical gradient of K⁺ and reversed at −95 mV and −63 mV, respectively. Points indicate mean values, and error bars show SEM. (E) Normalized chord conductances, calculated from Equation (2), of 5-HT responses measured in normal (black) and high external K⁺ solution (purple). Points indicate mean values.

Figure 3

5-HT induced K⁺ currents in all claustral PN subtypes. (A) Representative traces of 5-HT-induced outward currents in the five PN subtypes. The black arrowhead indicates the timing of pressure application of 5-HT (100 µM). (B) Response rates of the five PN subtypes. The horizontal line represents the mean response rate (67.2%) for all PNs. Asterisks indicate that PN1 cells have a significantly lower rate of responding to 5-HT (p = 0.008; see Supplementary Table S1). (C) The peak amplitude of the 5-HT-induced outward current was not significantly different across the 5 PN subtypes (refer to Supplementary Table S1 for statistical analysis). Points show individual measurements, bars indicate mean values, and error bars show SEM. ** p ≤ 0.01.

Figure 4

5-HT responses were generated by multiple types of 5-HTRs. (A) Representative responses to 5-HT (100 µM) before (Control, black) and after application of the indicated 5-HTR antagonists: 5-HTR-1A antagonist WAY100635 (1 µM, orange), 5-HTR2A antagonist MDL11939 (5 µM, red), and 5-HTR-2C antagonist RS102221 (5 µM, blue). The amplitudes of control responses are normalized to 100% to illustrate the magnitude of the block by each type of antagonist. (B) Decrease in the 5-HT-induced response of PNs, measured as response charge, following application of each 5-HTR antagonist. Points show individual measurements, bars indicate mean values, and error bars show SEM. Asterisks indicate statistically significant differences; refer to Supplementary Table S1 for statistical analyses. * p ≤ 0.05, **** p ≤ 0.0001.

Figure 5

5-HT decreased the AP firing of claustral PNs. (A) Representative traces of AP firing (top) in a claustral PN in response to depolarizing current pulses (bottom). 5-HT (100 µM) was applied 5 s before the second depolarization (arrowhead) and hyperpolarized the membrane potential of the cell. (B) AP firing elicited by current pulses (bottom) before (Control; top) and after (center) 5-HT application. (C) Relationship between the magnitude of depolarizing current pulses and frequency of resulting APs in control conditions (black) and after application of 5-HT (green). Points indicate mean values and error bars show ±1 SEM.

Figure 6

The temporal dynamics of AP firing reflects the time course of 5-HT-induced K⁺ conductance. (A) An example of claustral PN responses to pairs of depolarizing current ramps (bottom); 5-HT (100 µM) was applied at different times (arrowheads) prior to the second current ramp. (B) Relationship between the mean time course of 5-HT-induced K⁺ conductance (purple), measured in voltage-clamp conditions, and mean reduction in AP firing (pink) produced by application of 5-HT at variable times. Points indicate mean values, and error bars show ±1 SEM. (C) Correlation between the 5-HT-induced K⁺ conductance and reduction in AP firing measured in individual neurons. (D) Relationship between the mean time course of the 5-HT-induced K⁺ conductance (purple) and mean reduction in the AP current threshold (CT) produced by application of 5-HT at indicated times. Points represent mean values, and error bars show ±1 SEM. (E) Correlation between the 5-HT-induced K⁺ conductance and the reduction in CT (green) measured in individual neurons.

Figure 7

The subtractive action of 5-HT on neuronal arithmetic. (A) The input–output (I-O) relationship of a neuron could be modulated either linearly or non-linearly, changing processing of the input, the output or both. (B) I-O relationships of claustral PNs before (Control; black) or after (maroon) application of 5-HT (100 µM). Points indicate mean values and error bars show ±1 SEM. (C–F) I-O curve parameters obtained from Boltzmann function fits to the curves shown in (B): (C) Input₅₀, (D) slope, (E) half-maximal AP frequency, and (F) maximum AP frequency. Points show individual measurements, bars indicate mean values, and error bars show SEM. Asterisks indicate a statistically significant difference (p = 0.0016); refer to Supplementary Table S1 for statistical analyses.

Figure 8

The spatial resolution of 5-HT uncaging. (A) Outward currents evoked by photolyzing caged 5-HT (10 µM) over a 100 µm-by-100 µm area at different locations, indicated by squares numbered 1–3. (B,C) Determination of the distance-dependence of uncaging, calculated from both the peak amplitude (B) and the charge (C) of the 5-HT-induced outward currents. Distance was calculated according to the distance to the nearest neighboring process of a neuron. Points indicate mean values and error bars show ±1 SEM. Lines represent fits of exponential functions to the data; the length constants of the exponential fits are also indicated.

Figure 9

5-HT uncaging revealed the distribution of 5-HTRs in claustral PN neuronal compartments. (A) Outward currents evoked by photolyzing caged 5-HT (10 µM) over nearly the whole neuron (upper large square in left image), soma (upper small square in left image), 1 dendrite (lower small square in left image), or many dendrites (lower large square in left image). (B) The charge of outward currents induced by uncaging 5-HT at the whole neuron, soma, and calculated dendrite component (whole neuron–soma). (C) The charge of the 5-HT-induced outward currents for the calculated dendrite component and actual responses measured after uncaging 5-HT over a large area that included many dendrites. Points in (B,C) show individual measurements, bars indicate mean values, and error bars show SEM. Asterisks indicate statistically significant differences; refer to Supplementary Table S1 for statistical analyses. * p ≤ 0.05, *** p ≤ 0.001, **** p ≤ 0.0001.

Figure 10

The localization of 5-HTR-1A on claustral PNs. (A) Outward currents evoked by photolyzing caged 5-HT (10 µM) over the whole neuron (large left square), soma (small square), and dendrites (large right square). Black traces represent responses measured in normal ACSF, while red traces represent responses measured in the presence of WAY100635 (1 µM). (B) The total charge of responses to uncaging 5-HT in the indicated neuronal compartments, measured in control conditions and in the presence of WAY1000635. Points show individual measurements, bars indicate mean values, and error bars show SEM. (C) The percentage reduction in responses to uncaging 5-HT produced by WAY1000635. Bars indicate mean values, and error bars show SEM. Refer to Supplementary Table S1 for statistical analyses.