Fig. 2. Pharmacological profile of A-type potassium channels in DA SN neurons. (A) Whole-cell recording of a DA SN neuron. The membrane potential was stepped to –40 mV and current responses were recorded in increasing 4-aminopyridine (4-AP). (B) Mean dose-response for 4-AP inhibition: IC₅₀ = 472 µM, Hill coefficient of 1 (n = 6). Mean IC₅₀ calculated from Hill fits of individual neurons: 488 ± 131 µM (white circle, n = 6). (C) Same pulse protocol as in (A) under control and in the presence of HpTx3. (D) Mean dose-response for HpTx3 inhibition: IC₅₀ of 81.0 ± 4.0 nM (mean ± SD of fit), Hill coefficient of 1.1 (n = 6).

Fig. 4. A-type potassium channels control pacemaker frequencies in DA SN neurons. (A) Upper panels: current clamp whole-cell recording of pacemaker activity of a DA SN neuron under control, in the presence of 100 nM HpTx3, and after washout of the toxin (broken lines at –40 mV). After current clamp recording, an outside-out patch was pulled from the same neuron and the effect of 100 nM HpTx3 on the A-type current elicited by a voltage step to –40 mV from a holding potential of –100 mV was recorded (lower panel) to correlate the effect of a particular HpTx3 concentration on the frequency and A-type current in the same neuron. (B) Mean data from experiments in (A): the degree of HpTx3-induced frequency increase is tightly coupled to the degree of HpTx3-induced A-current inhibition determined in the same cells. The strong linear correlation (r = 0.96, slope = 1.78, n = 4–6) predicts a 2.7-fold (+170%) increase of pacemaker frequencies with complete inhibition of A-type channels. (C) Mean dose-response for HpTx3-induced frequency increase in DA SN neurons: EC₅₀=79.8 ± 4.3 nM (mean ±SD of fit), Hill coefficient of 2.1 (n = 4–6) is virtually identical to the IC₅₀ of HpTx3-induced A-channel inhibition (see Figure 2D).

Fig. 6. Double immunolabelling demonstrates Kv4.3 protein expression in DA SN neurons. Upper panels: overview of the distribution of Kv4.3 (green) and TH immunoreactivity (red) in the substantia nigra. Scale bars: 100 µm. Lower panels: Kv4.3 (green) and TH immunoreactivity (red) of a single DA SN neuron at high resolution. Scale bars: 10 µm.

Fig. 8. Linear correlation between A-type potassium charge and density, and number of Kv4.3 cDNA molecules. (A) Two examples of single-cell real-time fluorescent RT–PCR experiments on individual DA SN neurons. Different A-type current amplitudes with similar inactivation kinetics (corresponding to differences in A-type potassium charge transfer) are elicited by membrane depolarizations to –40 mV from a holding potential of –80 mV (left panels). The bold line indicates the threshold of amplification detection at R_f = 0.08. The DA SN neuron with a 4-fold smaller A-type potassium current displayed a C_t value of 38.2 (upper panels) that was about two cycles larger compared with that of the other DA SN neuron (C_t = 36.1). A difference of two C_t values corresponds to a 4-fold difference in Kv4.3L cDNA template concentration. (B) A-type potassium charge (left panel) and charge densities (right panel), and number of Kv4.3L cDNA molecules were codetermined in 21 DA SN neurons and show strong linear correlation (r = 0.90 and 0.91, respectively) over the full physiological range of both parameters. (C) A-type potassium charge densities and number of β-actin (left panel) and VMAT2 (right panel) cDNA molecules each were codetermined in 18 DA SN neurons and show no correlation (r = 0.14 and 0.21, respectively).

Fig. 1. Biophysical properties of native A-type potassium channels in DA SN neurons. (A) Whole-cell recording of a DA SN neuron. Membrane potential was stepped from a holding potential of –80 mV to depolarized potentials (–60 to –30 mV) in steps of 10 mV to elicit fast-inactivating potassium outward currents. (B) Frequency distribution of the dominant fast time constant of inactivation of A-type potassium currents recorded in 100 DA SN neurons at –40 mV (tau-i₁) in the standard whole-cell configuration. (C) Nucleated outside-out recording of a DA SN neuron. Membrane potential was stepped from a holding potential of –80 mV to 0 mV in 10 mV intervals to elicit fast-inactivating potassium outward currents. As is evident from the fit residuals (insert) in outside-out recordings, inactivation kinetics were well described with a single exponential decay. (D) Frequency distribution of the fast time constant of inactivation of A-type potassium currents recorded in 22 DA SN neurons at –40 mV (tau-i₁) in the nucleated outside-out configuration. (E) Mean activation kinetics of A-type potassium currents (insert) in nucleated outside-out recordings expressed as time-to-peak, show clear voltage dependence in contrast to their mean inactivation kinetics shown in (F) (n = 10). (G) Double-pulse experiments (–40 mV) with varying interpulse intervals at –80 mV revealed the kinetics of the recovery from inactivation of A-type potassium channels in DA SN neurons. Current responses to the second pulse after varying interpulse intervals between 25 and 400 ms recorded in nucleated outside-out patches are shown. (H) The mean recovery from inactivation in nucleated outside-out patches with an interpulse holding potential of –80 mV was well described by a single exponential function with a time constant of 172 ms (n = 6). The mean time constant of recovery calculated from fits of recovery in individual patches was 170 ± 17 ms (n = 6).

Fig. 3. Correlation of pacemaker frequencies and A-type potassium channel densities over the full physiological spectrum in DA SN neurons. (A) Frequency distribution of pacemaker frequencies from 146 DA SN neurons (standard whole-cell recordings). (B) Frequency distribution of A-type potassium charge densities (pC/pF) calculated from A-type potassium currents elicited at –40 mV from a holding potential of –80 mV recorded in 120 DA SN neurons (standard whole-cell). (C) Pacemaker frequencies and A-type potassium charge densities were codetermined in 16 DA SN neurons and show a strong linear correlation (r = 0.94, slope = –2.01, n = 16) over the full physiological range of both parameters. Insert: typical spontaneous pacemaker activity of a DA SN neuron (broken line at –40 mV).

Fig. 5. Fast inactivating A-type potassium channels are mediated by Kv4.3L and KChip3.1 expression in DA SN neurons. (A) PCR positive control: all A-type potassium channel candidate subunits and the two marker transcripts are detected using mouse midbrain cDNA (<10 fmol) as a PCR template. The amplified fragments had sizes (bp) predicted by their corresponding mRNA sequences. (B) Multiplex PCR expression profiling of a single dopaminergic (TH⁺, GAD₆₇^–) SN neuron displaying fast-inactivating A-type potassium current evoked by membrane depolarizations to –50 and –40 mV from a holding potential of –80 mV (upper panel). Agarose gel analysis (lower panel) shows that TH and Kv4.3L were coexpressed in this neuron. This homogeneous expression pattern was found for all analyzed DA SN neurons (n = 32). (C) Frequency distribution of the dominant fast time constant of inactivation of A-type potassium currents recorded (standard whole cell) in 32 genotyped (TH⁺, Kv4.3L⁺) DA SN neurons at –40 mV. This distribution is very similar to those of non-genotyped DA SN neurons shown in Figure 1B and D. (D–F) Same type of single-cell multiplex RT–PCR experiments as shown in (A–C) for KChip1-4 expression. (D) Positive control: all probed KChip subunits were expressed in mouse brain with KChip2.4, KChip3.1 (as well as KChip3.2) and KChip4.1 being the most prominent splice variants. (E) Agarose gel analysis shows that only TH and KChip3.1 were coexpressed in this DA SN neuron. This homogeneous expression pattern was found for all analyzed DA SN neurons (n = 12). (F) Frequency distribution of the dominant fast time constant of inactivation of A-type potassium currents recorded in 12 genotyped (TH⁺, KChip3.1⁺) SN neurons at –40 mV.

Fig. 7. Real-time fluorescent RT–PCR standard curve for single-cell Kv4.3L quantification. (A) Sensitivity and reproducibility of the real-time fluorescent RT–PCR protocol for Kv4.3L. Relative fluorescence intensities (R_f) were plotted against PCR cycle number for quadruple experiments with five different calculated numbers of Kv4.3L dsDNA template molecules (ranging from 1 to 1 000 000 molecules, as indicated). (B) Same data as in (A) are plotted on a logarithmic R_f scale, for better illustration of the exponential phase of the PCR. Slopes of the exponential amplification phases are highly reproducible and independent of template concentration. The bold line indicates the fluorescence threshold of amplification detection at R_f = 0.08, manually set within the exponential PCR phase. (C) Absolute standard curve for Kv4.3L cDNA detection derived from data shown in (A) and (B). The cycle number where the relative fluorescence crosses this threshold of amplification detection is defined as the C_t value. Mean C_t values for different numbers of Kv4.3L template molecules are plotted against their respective numbers of template molecules on a logarithmic scale. The linear regression fit (r = 0.999) defined the intercept as PCR cycle 40.47.

Fig. 9. Linear correlation between A-type potassium charge/density and number of KChip3 cDNA molecules. (A) Two examples of single-cell real-time fluorescent RT–PCR experiments on individual DA SN neurons (same type of experiment as described in caption to Figure 8). The DA SN neuron with an ∼4-fold smaller A-type potassium current displayed a C_t value of 39.2 (upper panels), which was about two cycles larger compared with that of the other neuron (C_t = 37.3). (B) A-type potassium charge and charge densities, and number of KChip3 cDNA molecules were codetermined in 13 DA SN neurons and show strong linear correlation (r = 0.90 and 0.92, respectively) over the full physiological range of both parameters.