Transient voltage-activated K+ currents in central antennal lobe neurons: cell type-specific functional properties

Abstract

In this study we analyzed transient voltage-activated K⁺ currents (I_A) of projection neurons and local interneurons in the antennal lobe of the cockroach Periplaneta americana The antennal lobe is the first synaptic processing station for olfactory information in insects. Local interneurons are crucial for computing olfactory information and form local synaptic connections exclusively in the antennal lobe, whereas a primary task of the projection neurons is the transfer of preprocessed olfactory information from the antennal lobe to higher order centers in the protocerebrum. The different physiological tasks of these neurons require specialized physiological and morphological neuronal phenotypes. We asked if and how the different physiological phenotypes are reflected in the functional properties of I_A, which is crucial for shaping intrinsic electrophysiological properties of neurons. Whole cell patch-clamp recordings from adult male P. americana showed that all their central antennal lobe neurons can generate I_A The current exhibited marked cell type-specific differences in voltage dependence of steady-state activation and inactivation, and differences in inactivation kinetics during sustained depolarization. Pharmacological experiments revealed that I_A in all neuron types was partially blocked by α-dendrotoxin and phrixotoxin-2, which are considered blockers with specificity for Shaker- and Shal-type channels, respectively. These findings suggest that I_A in each cell type is a mixed current generated by channels of both families. The functional role of I_A was analyzed in experiments under current clamp, in which portions of I_A were blocked by α-dendrotoxin or phrixotoxin-2. These experiments showed that I_A contributes significantly to the intrinsic electrophysiological properties, such as the action potential waveform and membrane excitability.NEW & NOTEWORTHY In the insect olfactory system, projection neurons and local interneurons have task-specific electrophysiological and morphological phenotypes. Voltage-activated potassium channels play a crucial role in shaping functional properties of these neurons. This study revealed marked cell type-specific differences in the biophysical properties of transient voltage-activated potassium currents in central antennal lobe neurons.

Keywords: Periplaneta americana; glomerulus; local interneuron; olfaction; projection neuron.

Fig. 1.

Example current traces of the transient potassium current (I_A) in different types of antennal lobe neurons. A: I_A was elicited by voltage steps from −100 to 70 mV. The current traces demonstrate the marked differences in inactivation kinetics of I_A between neuron types. B: time constants for inactivation of I_A. Note that in some neuron types, I_A inactivated with two time constants (τ₁ and τ₂). C: times to half-maximal inactivation of I_A. In B and C, n values are given in parentheses.

Fig. 2.

I_A in uniglomerular projection neurons (uPNs). A and B: example current traces for steady-state activation (A) and inactivation (B) of I_A. The holding potential was −60 mV. For steady-state activation, I_A was elicited by 500-ms depolarizing steps from −80 to 70 mV in 10-mV increments after a 500-ms prepulse to −100 mV. For steady-state inactivation, currents were elicited by 500-ms test pulses to 40 mV that were preceded by 500-ms pulses between −100 and 40 mV in 10-mV increments. C: mean G-V curves for steady-state activation (circles) and inactivation (triangles). Conductances were calculated with the assumption of a potassium equilibrium potential (E_K) of −98.5 mV. Values are expressed as a fraction of the calculated maximal conductance (G/G_max). The data were fit with first-order Boltzmann relations (Eq. 1; gray curve represents the mean; black curves indicate the 95% confidence interval). The numerical parameters of the G-V relations are given in Table 1. E_m, equilibrium potential. D: morphology of a uPN revealed by labeling with biocytin-streptavidin via the patch pipette. The neuron innervated a single glomerulus (shown at higher magnification in the inset) and sent a single axon along the medial antennal lobe tract to the mushroom body’s calyces and the lateral horn. Scale bars: D, 50 µm; D, inset, 50 µm. CA, calyx; GL, glomerulus; LH, lateral horn; mALT, medial antennal lobe tract; na, n-anterior; l, lateral; m, medial; np, n-posterior.

Fig. 4.

I_A in nonspiking type IIa (A1–A4) and type IIb LNs (B1–B4). A1, A2, B1, and B2: example traces for steady-state activation (A1, type IIa LN; B1, type IIb LN) and inactivation (A2, type IIa LN; B2, type IIb LN). A3 and B3: mean G-V curves for steady-state activation (circles) and inactivation (triangles) (A3, type IIa LN; B3, type IIb LN). The data were fit with first-order Boltzmann relations (Eq. 1; gray curves represent the mean; black curves indicate the 95% confidence interval). The numerical parameters of the G-V relations are given in Table 1. A4 and B4: morphology of a type IIa LN (A4) and type IIb LN (B4). Top images are overviews showing that both neuron types innervated all glomeruli. Bottom images at higher magnification show the innervation pattern in the glomeruli. The type IIa LN had a homogeneous innervation pattern, whereas the type IIb LN innervated all glomeruli only in part (zonal innervation). Scale bars: A4 and B4, top, 100 μm; A4 and B4, bottom, 20 µm. For details and abbreviations see Fig. 2 legend.

Fig. 3.

I_A in spiking type I local interneurons (LNs). A and B: example current traces for steady-state activation (A) and inactivation (B). C: mean G-V curves for steady-state activation (circles) and inactivation (triangles). The data were fit with first-order Boltzmann relations (Eq. 1; gray curve represents the mean; black curves indicate the 95% confidence interval). The numerical parameters of the G-V relations are given in Table 1. D: morphology of a type I LN. Top, overview showing the neuron innervated many, but not all, glomeruli and gave rise to the Y-shaped tract (arrow) typical for type I LNs. Bottom, higher magnification of a different focal plane showing single glomeruli to demonstrate that the density of neurites varied between glomeruli (1, 2, and 3). Scale bars: D, top, 50 μm; D, bottom, 25 μm. For details and abbreviations see Fig. 2 legend.

Fig. 5.

Concentration-dependent block of I_A by 4-aminopyridine (4-AP) in uPNs, type I LNs, type IIa LNs, and type IIb LNs. The concentration-response relations were similar in all 4 neuron types. Curves are sigmoidal fits (Eq. 2) with the following EC₅₀ values: uPNs, 1.7 mM (1.2–2.5 mM); type I LNs, 1.5 mM (1.1–2.0 mM); type IIa LNs, 1.2 mM (0.8–1.8 mM); and type IIb LNs, 1.1 mM (0.8–1.7 mM); n values are given in parentheses.

Fig. 6.

α-Dendrotoxin (DTX; red) and phrixotoxin-2 (PaTX2; blue) block portions of I_A in uPNs (A), type I LNs (B), type IIa LNs (C), and type IIb LNs (D). A–D: conductance density-voltage relations of I_A before and during the application of DTX (A1–D1) and PaTX2 (A2–D2). Boxplots show the I_A amplitudes for a voltage pulse to 40 mV. The example current traces show I_A before and during the application of toxins elicited by test pulses to 40 mV. Boxplots were generated using Tukey’s method. *P ≤ 0.05; **P ≤ 0.01; paired t-test; n values are given in parentheses.

Fig. 7.

α-Dendrotoxin (DTX; red)- and phrixotoxin-2 (PaTX2; blue)-sensitive components of I_A. A–D: mean G-V curves of the DTX- and PaTX2-sensitve currents in uPNs (A), type I LNs (B), type IIa LNs (C), and type IIb LNs (D). The voltage dependence for activation, measured as the voltage for half-maximal activation (V_0.5,act), is different between the two current components in uPNs, type I LNs, and type II LNs (uPNs: DTX, −1.8 ± 1.3 mV; PaTX2, −5.5 ± 1.3 mV, P < 0.05; type I LNs: DTX, 8.7 ± 1.5 mV; PaTX2, −0.8 ± 1.4 mV, P < 0.0001; type IIa LNs: DTX, 12.5 ± 1.7; PaTX2: −2.5 ± 2.6, P < 0.0001; type IIb LNs: DTX, 0.7 ± 1.55 mV; PaTX2, 3.2 ± 2.9; P > 0.05; F-tests); n values are given in parentheses.

Fig. 8.

Physiological role of I_A in uPNs and type I LNs. Summary and quantification of the α-dendrotoxin (DTX)and phrixotoxin-2 (PaTX2) effects on electrophysiological properties in uPNs (A and B, respectively) and type I LNs (C and D, respectively). A1–D1: example action potential waveforms. A2–D2: repolarization rate. A3–D3: width of half-maximal amplitude (spike width). A4–D4: rate of amplitude of afterhyperpolarization. A5–D5: latency to first action potential during a depolarizing voltage step. A6–D6: action potential threshold (spike threshold). A7–D7: number of action potentials during 500-ms depolarizing current pulses (frequency). Inset in C1 shows the type I LN action potential at lower magnification to capture the long-lasting afterhyperpolarization of this neuron type. Boxplots were generated using Tukey’s method. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; paired t-test; n values are given in parentheses. Ctrl, control.

Fig. 9.

Physiological role of I_A in nonspiking type II LNs. Example recording of a type IIa LN that was depolarized by a 2-s current pulse of 660 pA before and during the application of phrixotoxin-2 (PaTX2; 670 nM).