Identification and functional characterization of a K+ channel alpha-subunit with regulatory properties specific to brain

Abstract

The physiological diversity of K+ channels mainly depends on the expression of several genes encoding different alpha-subunits. We have cloned a new K+ channel alpha-subunit (Kv2.3r) that is unable to form functional channels on its own but that has a major regulatory function. Kv2.3r can coassemble selectively with other alpha-subunits to form functional heteromultimeric K+ channels with kinetic properties that differ from those of the parent channels. Kv2.3r is expressed exclusively in the brain, being concentrated particularly in neocortical neurons. The functional expression of this regulatory alpha-subunit represents a novel mechanism without precedents in voltage-gated channels, which might contribute to further increase the functional diversity of K+ channels necessary to specify the intrinsic electrical properties of individual neurons.

Fig. 1.

Amino acid sequence of Kv2.3r and alignment with the sequence of two other K⁺ channel α-subunits: Kv2.1 (Drk1) and Kv1.1 (Rbk1). The single letter code is used for amino acid identification. Hyphens indicate identity to the sequence at the top (Kv2.3r), anddots represent gaps introduced to maintain alignment. The numbers indicate the positions in the respective sequence. Note that the C termini of Kv2.1 and Kv1.1 are shown incomplete. Bars delineate the extent of the conserved regions, including the putative transmembrane domains (S1–S6), the pore (PORE), and theA and B boxes of the N terminal. The nucleotide sequence of Kv2.3r cDNA has been deposited in GenBank (accession number X98564).

Fig. 2.

Tissue distribution of Kv2.3r.A, Northern blot analysis indicating the selective expression of Kv2.3r in the brain. Two transcripts of ∼3.3 and 5 kb were identified in neocortex, hippocampus, and cerebellum, but no signal was detected in the other tissues studied (heart, skeletal muscle, lung, and gut). B, Parasagittal and coronal sections of the entire rat brain showing the cellular distribution of Kv2.3r mRNA. In the top panels the incubation with a Kv2.3r cRNA antisense probe shows its preferential expression in neocortex, cerebellum, hippocampus, and amygdala. The bottom panels demonstrate the lack of signal when the Kv2.3r cRNA sense probe was used. C, Microphotographs of in situ hybridization demonstrating the expression of Kv2.3r in Purkinje cells of the cerebellum (left,top), hippocampal CA4 neurons (left,bottom), and neocortical pyramidal cells (right). Magnifications are 900× (left) and 2700× (right).

Fig. 3.

Expression of Kv2.3r. A,In vitro translation of Kv2.3r cDNA in reticulocyte lysates (Promega, TNT). Protein analysis was done on a 9% SDS-polyacrylamide gel. The arrow indicates the band corresponding to the Kv2.3r protein with a molecular weight of ∼59 kDa. B, Northern blot analysis of RNA obtained from CHO cells, using the same Kv2.3r probe as in Figure 2. Lane 1, Untransfected cells; lane 2, cells transfected with Kv2.1 cDNA; lane 3, cells cotransfected with Kv2.1 plus Kv2.3r. The arrow indicates a band of ∼1.6 kb corresponding to Kv2.3r cRNA. The asteriskindicates the hybridization with cyclophilin mRNA used as a control for the loading on each lane.

Fig. 4.

Coexpression of green fluorescent protein (GFP) and K⁺ channel α-subunits.A, Recordings from a cell expressing theGFP and Kv2.1 currents. B, Recordings from a cell expressing the GFP that also was transfected with Kv2.3r cDNA. Note the absence of detectable K⁺ current. Current records obtained during depolarizing pulses to 0, +20, and+40 mV are superimposed. Patch-clamp experiments were done 24–30 hr after cotransfection of the cells with 2 μg ofGFP and 2 μg of either Kv2.1 orKv2.3r cDNAs.

Fig. 5.

Effect of coexpression of Kv2.3r with Kv2.1 (A), Kv1.1 (B), and ShakerBΔ6–46 (ShB; C). In each case we superimposed K⁺ current traces obtained in cells transfected with a K⁺ channel α-subunit alone (Kv2.1, Kv1.1, or ShB) with the recordings obtained from cells transfected with a mixture of each type of α-subunit plus an equal amount of Kv2.3r (Kv2.1+Kv2.3r,Kv1.1+Kv2.3r, and ShB+Kv2.3r).Traces in the middle and right columns have been scaled to facilitate the comparison of activation and inactivation time courses, respectively. Note the marked and selective effect of Kv2.3r on Kv2.1 currents. In theleft and middle columns the depolarizing pulses used to open the channels were applied to 0 mV. In theright column the pulses were applied to +20 mV. In all cases the holding potential was −80 mV.

Fig. 6.

Comparison of the kinetic properties of homomultimeric Kv2.1 channels and of channels formed by the coexpression of Kv2.1 plus Kv2.3r. A, Families of macroscopic K⁺ currents obtained during depolarizing pulses to various membrane potentials (from −40 to +40 in steps of 10 mV) from a holding potential of −80 mV. B, Current–voltage curves obtained by plotting current amplitude measured at the end of each pulse (ordinate) as a function of membrane potential during the pulse (abscissa). C, Time to reach half-maximal (t_1/2) activation of Kv2.1 and Kv2.1+Kv2.3r currents at various membrane potentials. Each point represents the mean ± SD of measurements done in seven experiments.

Fig. 7.

Comparison of the kinetic properties of homomultimeric Kv2.1 channels and of channels formed by the coexpression of Kv2.1 plus Kv2.3r in high external K⁺ (70 mm). A, Families of macroscopic K⁺ currents obtained during depolarizing pulses to various membrane potentials (from −20 to +40 in steps of 20 mV) from a holding potential of −80 mV. Note the tail currents recorded on repolarization, representing the deactivation time course of the channels. B, Current–voltage curves obtained by plotting current amplitude measured at the end of each pulse (ordinate) as a function of membrane potential during the pulse (abscissa). The intersection of the curves with the x-axis indicates the reversal potential.C, Average conductance–voltage relationships. Conductance was estimated from the amplitude of tail currents recorded on repolarization to −80 mV of voltage pulses delivered to variable membrane potentials. The values in the ordinate were normalized to the amplitude of the tails at +50 mV. Eachpoint represents the mean ± SD of measurements done in five to eight experiments. Curves drawn on the data points are least-squares fits to a Boltzmann function of the form: $G=1-(G_{max}/[1+exp(V-V_{1/2})/k])$in which G_max is maximal conductance, V the membrane potential during the pulse,V_1/2 the potential at which 50% of G_max is obtained (+9 mV for Kv2.1 and +19 mV for Kv2.1+Kv2.3r currents), and k a slope factor that indicates the steepness of the curve (7 mV for Kv2.1 and 7.9 mV for Kv2.1+Kv2.3r currents).

Fig. 8.

Macroscopic K⁺ currents recorded from cells expressing either Kv2.1 channels or the Kv2.1/Kv2.3r tandem dimer. Traces are superimposed (and scaled in themiddle and right panels) to facilitate the comparison of the activation and inactivation time courses in the two types of currents. Because we wanted to stress the similarity between the Kv2.1+Kv2.3r and the Kv2.1/Kv2.3r tandem dimer currents, the Kv2.1 current traces of this figure are the same as in Figure5A. Depolarizing pulses are applied to 0 mV (left and middle panels) or +20 mV (right panels). Holding potential is −80 mV. For Kv2.1/Kv2.3r currents, t_1/2 of activation at 0 mV is 25 ± 4 msec (mean ± SD, n = 6), and inactivation time constant at +20 mV is 9500 ± 2600 msec (n = 7).

Fig. 9.

Differential effects of external Zn²⁺(1 mm) on Kv2.1 and Kv2.1/Kv2.3r dimer currents.A, Reversible inhibition of Kv2.1 currents recorded at two different membrane potentials. B, Reversible inhibition of Kv2.1/Kv2.3r dimer K⁺ currents. Currents were generated by depolarization to the indicated membrane potentials. Holding potential is −80 mV.