Fig 1.
Cultured hippocampal neurons exhibit diverse physiological properties.
(A-C) Voltage responses of 3 types of neurons evoked by current step stimulation. Blue traces show the first spike responses just above rheobase levels (also indicated by blue symbols in the I-O curves). Spike count vs. current relationship for the regular, delayed firing and stuttering neurons are shown in D, F and H, respectively. The first spike latency as a function of the injected current is plotted in E, G, I. The magenta lines are fitted Belehradek functions. (J-L) Cells assigned to one of the three phenotypes are visualized in scatter plots of various pairs of extracted physiological parameters. (J) The input membrane resistance plotted against the resting membrane potential; (K) first spike latency at 1.3-times of the rheobase current level plotted against the input resistance; (L) total (cumulative) spike count vs. rheobase. Colors indicate the subjective neuronal phenotypes, gray: regular firing, blue: delayed firing, red: stuttering type cells.
Fig 2.
Firing responses of the hippocampal neurons indicate different degree of excitability under current step stimulation vs. simulated synaptic bombardment.
Voltage traces of a delayed firing type cell and the corresponding input-output relationship are shown under current step protocol in (A) and (B), respectively. (C) Firing response and (D) the spike count vs. AMPA-conductance relationship obtained from the same delayed type neuron under dynamic clamp protocol. Corresponding panels in (E-H) demonstrate the firing responses of a stuttering type cell under identical stimulus conditions. Note that the I-O relationships are very different between the two neuron types (B vs. F and D vs. H). (I) The threshold AMPA-conductance is plotted against the rheobase for all the recorded neurons (n = 380). (J) Total (cumulative) spike count from the dynamic clamp experiments is plotted against the total static spike count (n = 399). Colors indicate the three cell types.
Fig 3.
The model neurons reproduce the firing responses and the physiological diversity of the biological neurons observed in the experiments.
(A, B, C) Voltage responses of the regular, delayed and stuttering type models driven by current step stimulation. Blue traces show the first spike responses just above rheobase levels. (D, F, H) Spike count vs. current plots and (E, G, I) first spike latency plotted as the function of the injected current are shown for the regular, delayed and stuttering type model neurons. Note that the stuttering model exhibits a sudden jump in spike count at the transition to repetitive firing (+195 pA, H). (J) Input membrane resistance vs. resting membrane potential of all model instances. (K) The threshold AMPA-conductance is plotted against the rheobase of the model neurons. (L) Cumulative spike counts from synaptic vs. current step responses are scatter plotted. Data from 600 model instances are shown.
Fig 4.
Pharmacological blocking of the Kir- and D-currents produce differential effects in delayed vs. stuttering type neurons.
The voltage responses of a delayed firing type neuron are shown before (A) and after the application of BaCl2 (B). (C) Input-output functions obtained from the current step experiments indicate a moderate increase of the firing under BaCl2. (D) I-O relationship obtained from the dynamic clamp experiments. (E and F) Total spike counts in control and BaCl2-treated cells are shown for the current step vs. DCl experiments, respectively. (G) The average relative change in total spike counts in the current step (IV) and dynamic clamp (DCl) experiments. Panels (H-N) show the same for the stuttering type neurons under the application of 4-AP. The firing of the demonstrated neuron is markedly increased by 4-AP application when the current step stimulation is used (J) but increased to a less degree when the simulated synaptic inputs are used (K). The pooled data (L-M) reveal that spike counts under current step stimulation increase far more than firing under simulated synaptic inputs. (N) Average relative change of total spike counts following 4-AP application in the current step (IV) and dynamic clamp (DCl) experiments.
Fig 5.
Removal of two K-currents from the model neurons produce differential effects on their firing output.
(A) First spike latency vs. input resistance plot for the delayed firing type neurons. Non-filled symbols represent the data after the removal of the Kir-current. (B) Scatter plot of cumulative spike counts from the same set of model neurons (synaptic vs. current step responses). (C) Pooled cumulative spike counts shift in a differential manner when static vs. dynamic inputs are used. D-F: Same plots demonstrating the effects of D-current removal from the stuttering type model neurons.
Fig 6.
Manipulations of the voltage dependence and kinetics of the K-currents exert profound effects on the firing responses of the model neurons.
(A) The half-voltage of the steady-state activation curve (Vm,1/2) of the Kir-current is shifted from -91 to -75 mV and the cumulative spike counts of 25 model instances are calculated. Black arrows indicate the model responses shown in A1 and A2. (B) The peak activation time constant of the Kir-current (τm,max) is shifted from 10 to 130 ms and the cumulative spike counts of the corresponding model responses are plotted. (C) Similar manipulation is performed on the steady-state activation midpoint of the D-current of the stuttering type model (25 instances). (D) Peak activation time constant of the D-current is shifted from 2 to 18 ms. Here, models featuring the ‘fast’ D-current exhibit high static and low dynamic excitability, while those with the ‘slow’ D-current exhibit low static spike counts and high dynamic spike counts.