Cns distribution of members of the two-pore-domain (KCNK) potassium channel family

Abstract

Two-pore-domain potassium (K(+)) channels are substrates for resting K(+) currents in neurons. They are major targets for endogenous modulators, as well as for clinically important compounds such as volatile anesthetics. In the current study, we report on the CNS distribution in the rat and mouse of mRNA encoding seven two-pore-domain K(+) channel family members: TASK-1 (KCNK3), TASK-2 (KCNK5), TASK-3 (KCNK9), TREK-1 (KCNK2), TREK-2 (KCNK10), TRAAK (KCNK4), and TWIK-1 (KCNK1). All of these genes were expressed in dorsal root ganglia, and for all of the genes except TASK-2, there was a differential distribution in the CNS. For TASK-1, highest mRNA accumulation was seen in the cerebellum and somatic motoneurons. TASK-3 was much more widely distributed, with robust expression in all brain regions, with particularly high expression in somatic motoneurons, cerebellar granule neurons, the locus ceruleus, and raphe nuclei and in various nuclei of the hypothalamus. TREK-1 was highest in the striatum and in parts of the cortex (layer IV) and hippocampus (CA2 pyramidal neurons). mRNA for TRAAK also was highest in the cortex, whereas expression of TREK-2 was primarily restricted to the cerebellar granule cell layer. There was widespread distribution of TWIK-1, with highest levels in the cerebellar granule cell layer, thalamic reticular nucleus, and piriform cortex. The differential expression of each of these genes likely contributes to characteristic excitability properties in distinct populations of neurons, as well as to diversity in their susceptibility to modulation.

Fig. 1.

CNS distribution of two-pore-domain channels. Coronal sections were hybridized with [³³P]-labeled cRNA probes for TASK-1, TASK-3, TREK-1, TREK-2, TRAAK, and TWIK-1 and exposed to autoradiographic film. Labeling for TASK-2 in the CNS was uniformly low and therefore is not presented here. Line drawings on the far right are adapted from Paxinos and Watson (1997; reproduced with permission) and indicate the relevant labeled areas. For abbreviations, please refer to Table 2. Scale bar, 3 mm.

Fig. 1.

CNS distribution of two-pore-domain channels. Coronal sections were hybridized with [³³P]-labeled cRNA probes for TASK-1, TASK-3, TREK-1, TREK-2, TRAAK, and TWIK-1 and exposed to autoradiographic film. Labeling for TASK-2 in the CNS was uniformly low and therefore is not presented here. Line drawings on the far right are adapted from Paxinos and Watson (1997; reproduced with permission) and indicate the relevant labeled areas. For abbreviations, please refer to Table 2. Scale bar, 3 mm.

Fig. 1.

CNS distribution of two-pore-domain channels. Coronal sections were hybridized with [³³P]-labeled cRNA probes for TASK-1, TASK-3, TREK-1, TREK-2, TRAAK, and TWIK-1 and exposed to autoradiographic film. Labeling for TASK-2 in the CNS was uniformly low and therefore is not presented here. Line drawings on the far right are adapted from Paxinos and Watson (1997; reproduced with permission) and indicate the relevant labeled areas. For abbreviations, please refer to Table 2. Scale bar, 3 mm.

Fig. 2.

Spinal cord expression of two-pore-domain K⁺ channel transcripts. Transverse sections from rat cervical spinal cord were hybridized with [³³P]-labeled cRNA probes and subsequently dipped in liquid autoradiographic emulsion. Silver grains were imaged using dark-field and bright-field microscopy. Top panels (A–D) show broad distribution of TASK-1 and TASK-3, with low-to-moderate accumulations of silver grains over many neurons. Ventral horn motoneurons were intensely labeled for both transcripts. In the alternate panels (B,D), these same motoneurons are shown at higher power using bright-field optics, with the corresponding cells indicated byarrows. Bottom panels (E–H) show dark-field images of labeling for the other transcripts in the spinal cord. Labeling for these transcripts was lower, although there were moderate levels of TREK-2 in spinal cord interneuron layers (F). Scale bar: A,C, E–H, 400 μm; B,D, 100 μm.

Fig. 3.

TASK-1 and TASK-3 show high levels of mRNA accumulation in the locus ceruleus; TWIK-1 is highly expressed by adjacent mesencephalic trigeminal neurons. Panels show dark-field images of coronal sections through the dorsolateral pons corresponding to autoradiograms in Figure 1D. Arrowheads in each panel mark the approximate boundary of the locus ceruleus, which had high levels of labeling for TASK-1 (A) and was even more densely labeled for TASK-3 (B). Large mesencephalic trigeminal neurons are located lateral to the locus ceruleus; these neurons showed high levels of TWIK-1 (arrow in C). Scale bar, 300 μm.

Fig. 4.

Expression of two-pore-domain channels in the thalamus and striatum. Shown are low-power dark-field images of emulsion-dipped sections through the telencephalon, corresponding to the film autoradiograms in Figure 1J. The line drawing (D) indicates the relevant fiber tracts and brain nuclei (for abbreviations, see Table 2). In the caudate putamen, there was uniformly high signal for TREK-1 in cells densely distributed throughout the nucleus (C); these are likely GABAergic output neurons. In contrast, TASK-3 labeled a set of large, sparsely located cells, likely cholinergic interneurons, one of which is marked with an arrow in the dark-field image (B) and in the corresponding high-power bright-field micrograph in H. Also note the intense signal for TWIK-1 in the thalamic reticular nucleus and in the choroid plexus (G). Scale bar: A–G, 500 μm; H, 25 μm.

Fig. 5.

Expression of two-pore-domain channels in different regions of the hippocampus. Low-power dark-field images of emulsion-dipped coronal sections through the hippocampus.A–F show labeling for the six channel mRNAs through the rostral/dorsal hippocampus; this view corresponds to the film autoradiograms seen in Figure 1I. Note the differential distribution through the various pyramidal cell populations (CA1, CA2, andCA3), the approximate boundaries of which are marked byarrowheads. Also, there was TASK-3 labeling in scattered cells of the nonpyramidal layers (arrow inC); these are likely hippocampal interneurons.G and H illustrate differential TASK-3 expression in the ventral/caudal hippocampus compared with the rostral/dorsal expression illustrated above. G shows TASK-3 expression in the ventral/caudal dentate gyrus and corresponds to the ventral portion of the film autoradiogram in Figure1F. Note that there was particularly intense labeling of dentate gyrus granule cells in the ventral hippocampus but very little expression by these cells in the dorsal hippocampus (C). H shows TASK-3 expression in the ventral/caudal portion of CA3 and is taken from a section corresponding to the film autoradiogram seen in Figure1G. The CA3 pyramidal cell layer is marked bycurved arrows and traverses the entire panel. Note that there is much more TASK-3 expression in cells in thebottom (i.e., ventral) part of this panel when compared with cells in the top part of the panel and when compared with TASK-3 labeling in CA3 of the dorsal hippocampus (C). Scale bar, 500 μm.

Fig. 6.

Laminar distribution of two-pore-domain K⁺ channel transcripts in the neocortex. Shown are dark-field micrographs of the somatosensory cortex. The relevant cell body layers (i.e., layer II through layer VI) are indicated on theright, as is the underlying fiber tract [external capsule (ec)]. TASK-3, TRAAK, and TWIK-1 could be seen at appreciable levels in all layers, whereas TREK-1 was highly enriched in layer IV. TASK-1 was expressed at higher levels in a scattering of cells, mostly in layers IV and VI. One of these layer VI cells is marked by an arrowhead and is shown using high-power bright-field optics in the inset on theleft. Scale bar: 250 μm; inset, 50 μm.

Fig. 7.

Two-pore-domain channel expression in dorsal root ganglia. As shown by dark-field microscopy, all seven transcripts were expressed in dorsal root ganglia, and especially high densities of silver grains were seen for TASK-1 (A), TASK-2 (B), and TRAAK (G). TASK-3 (C) was expressed at high levels but by a much more limited proportion of the cells. Labeling for TWIK-1 was for the most part restricted to larger neurons; it is shown at low power inD, with a higher-power view of the indicated neuron (arrows) shown in H. Scale bar:A–G, 400 μm; H, 50 μm.