Analogue modulation of back-propagating action potentials enables dendritic hybrid signalling

Abstract

We report that back-propagating action potentials (bAPs) are not simply digital feedback signals in dendrites but also carry analogue information about the overall state of neurons. Analogue information about the somatic membrane potential within a physiological range (from -78 to -64 mV) is retained by bAPs of dentate gyrus granule cells as different repolarization speeds in proximal dendrites and as different peak amplitudes in distal regions. These location-dependent waveform changes are reflected by local calcium influx, leading to proximal enhancement and distal attenuation during somatic hyperpolarization. The functional link between these retention and readout mechanisms of the analogue content of bAPs critically depends on high-voltage-activated, inactivating calcium channels. The hybrid bAP and calcium mechanisms report the phase of physiological somatic voltage fluctuations and modulate long-term synaptic plasticity in distal dendrites. Thus, bAPs are hybrid signals that relay somatic analogue information, which is detected by the dendrites in a location-dependent manner.

Figure 1. Location-dependent effects of somatic membrane potential on bAP waveforms.

(a) Superimposed IR-DIC and confocal z stack images of a GC, which was simultaneously recorded at a somatic and a dendritic site. Spikes were evoked with short current injections at the soma and the membrane potential was also set at this location with steady-state current injections to slightly depolarized (red, −64.6±0.6 mV, n=29) or hyperpolarized (blue, −77.2±0.6 mV) potentials. The second pipette was used for monitoring the undisturbed voltage either at a dendrite or at the soma. (b) Propagation of somatically set membrane potential into dendrites (left panel). Connected symbols indicate individual experiments. Arrows show the average vector of the membrane potential difference. The right panel shows the distance dependence of the absolute bAP peak voltage (see also ref. 26). (c) Average bAP waveforms at various monitored locations. Representative somatic traces are shown on the left. Spikes were aligned to their maximum rate of rise (vertical grey lines). Black traces below show the voltage difference between the bAPs evoked from hyperpolarized and depolarized potentials. (d) The same average bAPs as in c on different timescale. Numbers show the included experiment numbers. (e) Effects of hyperpolarization on the peak voltage and relative repolarization of bAPs. Grey symbols indicate individual experiments, whose running average is shown in blue (calculated from 9 neighboring data points). To quantify the differences during the repolarization relative to the loss of the peak voltage, the voltage difference between the two bAPs at 0.2–1.2 ms after the peak was compared with the voltage difference at the peak.

Figure 2. Location-dependent bidirectional effects of somatic membrane potential on bAP-evoked dendritic calcium signals.

(a) Representative experiment showing the three imaged locations (24, 67 and 168 μm) of a GC filled with calcium-sensitive Fluo-5F (183 μM) and Alexa Fluor 594 (15 μM) dyes loaded by somatic patch pipette, which was also used for setting the somatic membrane potential. Imaging traces (ΔF/F₀) from the Fluo-5F channel were alternately recorded at the two membrane potentials (−65.1 and −77.6 mV). Three-dimensional z stack Alexa Fluor 594 imaging was performed after the end of the calcium measurement and short lower intensity imaging in the red channel was used to set and maintain the line scan positions. (b) Average calcium imaging trace pairs (ΔF/F₀) along the dendrites at two somatic membrane potentials (−64.0±0.3 and −77.6±0.3 mV). Note that perisomatic imaging (<20 μm) was avoided to prevent photo-damage and that in the most distal regions bAPs do not evoke detectable calcium signals in GCs (Supplementary Fig. 3; see also ref. 26). (c) Light blue symbols mark the effect of hyperpolarization on local calcium signals along the dendritic distance in individual experiments (n=54 spots from n=25 cells) and dark blue symbols show the running average (13 individual point were used for each average data points).

Figure 3. The lower peak of the distal bAPs during hyperpolarization leads to smaller calcium influx.

(a) Comparison of simulated composite calcium currents (HN:HI, for details see Supplementary Fig. 7 and Supplementary Table 1) with recorded HVA calcium currents (in the presence of NNC55-0396) evoked by proximal and distal bAP waveforms as voltage commands. (b) Simulated calcium influxes after shifting the entire voltage command traces in 2 mV steps in a representative experiment (originally recorded 116 μm from the soma). (c) The heat map summarizes the changes of the simulated calcium influx upon offsetting the command voltage in 2 mV steps. Each column corresponds to one previously recorded bAP (six examples are shown below), which were sorted along the x axis according to the location of their recording. Zero voltage on the colour plot corresponds to the original recorded potential (depolarized, membrane potential: −64.6±0.6 mV). Notice the sensitivity of the small distal bAPs to small voltage shifts.

Figure 4. Faster repolarization promotes larger proximal dendritic calcium signals.

(a) The acceleration of the repolarization was mimicked in conductance clamp configuration using an I_A-like conductance (Supplementary Fig. 9) without affecting the AP peak and membrane potential. Traces show APs with (green, average of 11 cells) and without (grey) the I_A-like conductance and their voltage difference (black). The APs with accelerated repolarization resulted in larger proximal dendritic calcium signals, unlike the APs affected by a passive conductance. (b) Exploration of the necessary activation and inactivation time constants that allows for membrane potential-dependent calcium influx using originally R- (HI) and N-type (HN) calcium current parameters. For these calcium current simulations, we employed an AP waveform pairs (24 μm), which were recorded at two membrane potentials (Fig. 1). These voltage commands were modified by offsetting the preceding membrane potentials to −80 mV (traces at the top) to measure the contribution of the bAP shape changes to differential calcium influx in isolation. Thus, the colour maps show the differences in the calcium influx evoked by APs with faster and slower repolarization, which derived from APs from hyperpolarized and depolarized membrane potentials. The necessity of fast inactivation time constant is demonstrated in the two example trace pairs below. The graph below is the horizontal cross-sections of the colour graphs showing the calcium influx differences at the standard activation kinetics (tau at 0 mV: 1.137 and 0.825 ms). Arrows indicate the registered inactivation time constant. (c) Ensemble calcium currents in GCs contain fast-inactivating components. Traces of calcium currents together with the fast (τ_f) and slow (τ_s) components of double exponential fits are shown from a single nucleated patch (upper trace with black fit, R²=0.916) and the average (light grey area is the s.e.m., individually fitted data, relative weights of τ_f and τ_s were 52.8±8.9% and 47.2±8.9%, respectively).

Figure 5. Hyperpolarization-induced enhancement of calcium influx requires inactivating HVA calcium currents.

(a) Average traces and changes in charge (0.35–1.35 ms after the peak of bAP) of the calcium currents in nucleated patches evoked by previously recorded proximal bAP pairs (24 μm) at two membrane potentials, in control conditions and in the presence of Cav3 blocker, NNC55-0396 (10 μM), or in the presence of Cav2.3 and Cav3 blocker, Ni²⁺ (500 μM). Crosses indicate the changes predicted by simulations (see b). (b) Simulated HVA calcium currents with inactivating and non-inactivating HVA or only non-inactivating HVA components corresponding to pharmacologically isolated currents in NNC55-0396 or Ni²⁺, respectively. (c) Average calcium imaging traces at two membrane potentials and the effect of somatic hyperpolarization in control conditions and during the blockade of Cav3 (10 μM NNC55-0396), or Cav2.3 and Cav3 channels (50 μM Ni²⁺, the lower concentration was needed to avoid interference with imaging). Numbers indicate numbers of experiments.

Figure 6. Physiological impacts of the analogue content of the hybrid dendritic bAPs.

(a) Average calcium signals evoked by single APs, which were preceded by five sine waves in the theta range (5.2 Hz). (b) Changes of the bAP-evoked calcium signal amplitudes during the theta cycle relative to rest at proximal (light blue) and distal (green) regions. For better visualization, the first and last data points in the theta cycle are shown after and before the actual data (open symbols). See also Supplementary Fig. 4. (c) Average traces (n=8 and 8 cells) before and after pairing of glutamate uncaging-evoked distal EPSPs (150–200 μm) with postsynaptic APs (300 pairing at 1 Hz, timing ±4 ms; Supplementary Fig. 13). The pairing was made either at depolarized (red, −62.5±0.5 mV) or hyperpolarized (blue, −81.4±0.5 mV) membrane potentials. The graph shows the differential long-term changes of EPSP amplitudes depending on the pairing membrane potential (n=8 and 8 cells). (d) Same experimental arrangement as in c except that calcium channels were blocked with Ni²⁺, nifedipine, NNC55-0396 and ω-conotoxin (n=8 and 8 cells at −62.3±0.2 and −81.1±0.4 mV, respectively). (e) Summary of the long-term changes in control conditions, with inhibited calcium channels and using a pairing protocol that lacks postsynaptic firing (n=8, 8, 8, 8, 6 and 6, respectively, at hyper- and depolarized pairing protocols). Bars show the average relative effects of pairing at the two membrane potentials in the three conditions (right axis, P=0.013, P=0.99, P=0.45, respectively, t-test, averages of the 30–50 min period after pairing) and circles indicate the absolute EPSP amplitudes before and after pairing (left axis).