Temporal disparity of action potentials triggered in axon initial segments and distal axons in the neocortex

Abstract

Neural population activity determines the timing of synaptic inputs, which arrive to dendrites, cell bodies, and axon initial segments (AISs) of cortical neurons. Action potential initiation in the AIS (AIS-APs) is driven by input integration, and the phase preference of AIS-APs during network oscillations is characteristic to cell classes. Distal regions of cortical axons do not receive synaptic inputs, yet experimental induction protocols can trigger retroaxonal action potentials (RA-APs) in axons distal from the soma. We report spontaneously occurring RA-APs in human and rodent cortical interneurons that appear uncorrelated to inputs and population activity. Network-linked triggering of AIS-APs versus input-independent timing of RA-APs of the same interneurons results in disparate temporal contribution of a single cell to in vivo network operation through perisomatic and distal axonal firing.

Fig. 1.. Multiple RAF patterns in the human and rat neocortex.

(A) Human layer 1 interneurons show persistent RAF (middle) and sporadic RAF (bottom) after repetitive suprathreshold somatic current injections (top). (B) Cumulative probability of persistent and sporadic RAF frequency. (C) AIS-APs and RA-APs are easily distinguishable. Left: Superimposed AIS-APs (black) and RA-APs (red) are aligned to AP threshold, respectively. Middle: Phase plot (middle) of AIS-APs (black) and RA-APs (red).​ Right: Threshold potential and slope of membrane potential 0 to 1 ms before threshold distinguish AIS-APs and RA-APs. (D) Proportion of human layer 1 interneurons with and without RAF detected. ​ (E) Individual human interneurons showed persistent only, sporadic only, or both RAF patterns. (F) Firing pattern (left) and anatomical reconstruction (middle) of a layer 1 NGFC showing RAF in the human neocortex (black, soma and dendrites; red, axon). Right: RAF was more prevalent in cells showing NGFC axonal morphology (NGFC, n = 21; non-NGFC, n = 9). (G) Same as in (F) but in rat neocortex (NGFC, n = 22; non-NGFC, n = 10). (H to K) Simultaneous somatic and axonal bleb recordings in the same human NGFC. (H) Responses to a somatic current injection (bottom) detected on the soma (black) and the axon (magenta). An AP (red) was identified as a RA-AP. (I) Top: Somatic detection (black 1) usually preceded axonal detection (magenta) of the same APs. Bottom: Some APs were also observed with the axonally placed electrode first (magenta) and with the somatic electrode second (red 2). Note the different somatic threshold potentials for soma first versus second APs. Top and bottom panels show consecutive APs in response to the same somatic current injection. (J) Somatic threshold potentials (top) and somatic voltage derivative curves (bottom) separate superimposed AIS-APs (black) and RA-APs (red). (K) Soma to bleb latency, somatic threshold potential, and pre-AP ramp slopes before the AP differentiates AIS-APs and RA-APs.

Fig. 2.. Axonal HCN channels and local increase of potassium ion concentration promote RAF.

(A and B) Persistent RAF detected in the human neocortex under control conditions (A) was suppressed by application of the HCN channel blocker ZD7288 30 μM (B). (C) The duration of RAF (paired-sample t test, P = 0.0081) and the number of RA-APs (paired-sample t test, P = 0.00706) decreased in the presence of ZD7288 relative to control conditions. (D) Somatically recorded response of a layer 1 rosehip interneuron to a hyperpolarizing current step with components of the sag fraction indicated. (E) Somatic sag fraction was higher in layer 1 interneurons in human compared to the rat (Mann-Whitney test, P < 0.001). In both species, cells showing RAF had smaller sag fraction compared to cells where RAF could not be induced (human: Mann-Whitney test, P < 0.001; rat: Mann-Whitney test, P < 0.001). (F to H) Simultaneous somatic and axonal bleb recordings on the same layer 1 interneurons. (F) Experimental design and somatic firing pattern of a human NGFC. (G) Representative somatic (black) and axonal (magenta) responses to somatic and axonal hyperpolarization in the same human NGFC. (H) Compared to the soma, the axonal bleb showed higher sag fraction in human and rat cells (two-sample t test, P < 0.001). (I) Local application of ACSF containing elevated potassium concentration (5 mM KCl) evoked sporadic RAF in somatically recorded human layer 1 interneurons. (J) Hyperpolarized threshold potentials and somatic voltage derivative curves identify RA-APs in 5 mM KCl. (K and L) Same as (I) and (J), but in the rat. *P < 0.01, ***P < 0.001.

Fig. 3.. Different temporal domains for AIS-APs and RA-APs in layer 1 interneurons in vitro.

(A to G) Spontaneous AIS-APs and RA-APs detected in active human slices. (A) Schematic experimental design. Elevated network activity was induced with bath application of 2 μM carbachol and 10 μM SCH23390. (B) AIS-APs (black) and RA-AP (red) occur at depolarized and hyperpolarized states of membrane potential fluctuations in a human layer 1 interneuron, respectively. Dashed lines separate up, transition, and down states. (C) Probability of subthreshold membrane potential fluctuations corresponding to cellular up, down, and transition states in the human interneuron shown in (B). (D) Ratio of RA-AP and AIS-AP in different cellular states. (E) AIS-APs (black) and RA-APs (red) were identified according to hyperpolarized threshold potentials (left), different somatic voltage derivative curves (middle), and distinct pre-AP ramp slopes (right). (F) SD of membrane potential 0 to 50 ms before APs. (Mann-Whitney test, P < 0.001). (G) Proportion of layer 1 interneurons showing spontaneous RA-APs in active human slices. (H to M) Detection and timing of AIS-APs and RA-APs in oscillating rat slices. (H) Schematic experimental design. Oscillation was induced with bath application of 2 μM carbachol and 10 μM SCH23390. (I) Simultaneous whole-cell recording from a layer 1 interneuron (top) and LFP recording in layer 5 (middle with Hilbert transformation shown below). A spontaneous RA-AP (blue) occurs during a hyperpolarized state, and AIS-APs are synchronized to LFP deflections. (J) AIS-APs (black) and RA-APs (blue) were identified according to hyperpolarized threshold potentials (left), different somatic voltage derivative curves (middle), and distinct pre-AP ramp slopes (right). (K) SD of membrane potential 0 to 50 ms before APs (Mann-Whitney test, P < 0.001). (L) Proportion of layer 1 interneurons showing spontaneous RA-APs in active rat slices. (M) Circular plot of AP probability relative to LFP phase (population data: n = 22 cells, n = 76 RA-APs, and n = 534 AIS-APs; nonuniformity Rayleigh test: RA-APs, P = 0.071; AIS-APs, P < 0.001). ***P < 0.001.

Fig. 4.. Layer 1 interneurons fire AIS-APs and RA-APs in vivo.

(A) Schematic of simultaneous layer 1 whole-cell recording and layer 5 LFP recordings in mice. (B) Proportion of layer 1 interneurons showing spontaneous RA-APs in the mouse neocortex in vivo. (C) Confocal image of a whole-cell recorded layer 1 interneuron that generated RA-APs. (D) Simultaneous whole-cell recording from a layer 1 interneuron (top) and LFP recording in layer 5 (middle with Hilbert transformation shown below). A spontaneous RA-AP (green) occurs during a hyperpolarized state, and AIS-APs are fired from depolarized up states. (E) AIS-APs (black) and RA-APs (green) were identified according to hyperpolarized threshold potentials (left), different somatic voltage derivative curves (middle), and distinct pre-AP ramp slopes (right). (F) SD of membrane potential 0 to 50 ms before APs (Mann-Whitney test, P < 0.001). (G to I) Proportion (G), number (H), and frequency (I) of AIS-APs and RA-APs detected in individual layer 1 interneurons. (J) Circular plot of AP probability related to LFP phase (AIS-APs, black; RA-APs, green; nonuniformity Rayleigh test for population data: RA-APs, P = 0.15; AIS-APs, P < 0.001). ***P < 0.001.