Quantitative analysis of neurons with Kv3 potassium channel subunits, Kv3.1b and Kv3.2, in macaque primary visual cortex

Abstract

Voltage-gated potassium channels that are composed of Kv3 subunits exhibit distinct electrophysiological properties: activation at more depolarized potentials than other voltage-gated K+ channels and fast kinetics. These channels have been shown to contribute to the high-frequency firing of fast-spiking (FS) GABAergic interneurons in the rat and mouse brain. In the rodent neocortex there are distinct patterns of expression for the Kv3.1b and Kv3.2 channel subunits and of coexpression of these subunits with neurochemical markers, such as the calcium-binding proteins parvalbumin (PV) and calbindin D-28K (CB). The distribution of Kv3 channels and interrelationship with calcium-binding protein expression has not been investigated in primate cortex. We used immunoperoxidase and immunofluorescent labeling and stereological counting techniques to characterize the laminar and cell-type distributions of Kv3-immunoreactive (ir) neurons in macaque V1. We found that across the cortical layers approximately 25% of both Kv3.1b- and Kv3.2-ir neurons are non-GABAergic. In contrast, all Kv3-ir neurons in rodent cortex are GABAergic (Chow et al. [1999] J Neurosci. 19:9332-9345). The putatively excitatory Kv3-ir neurons were mostly located in layers 2, 3, and 4b. Further, the proportion of Kv3-ir neurons that express PV or CB also differs between macaque V1 and rodent cortex. These data indicate that, within the population of cortical neurons, a broader population of neurons, encompassing cells of a wider range of morphological classes may be capable of sustaining high-frequency firing in macaque V1.

Figure 1. Kv3 antibody controls on V1 tissue

Panels (a) and (b) show macaque V1 tissue processed to detect Kv3.1b- (a) or Kv3.2 (b) immunoreactivity by an ABC-DAB reaction. Preadsorption of the antibodies against a saturating concentration of the target antigen (see ‘Antibody controls’) abolished staining for both the Kv3.1b (c) and Kv3.2 (d) subunit proteins. Scale bar = 100μm (all panels). Panel (e) shows immunoblots on macaque V1 tissue homogenates treated with Kv3.1b (left lane) and Kv3.2 (right lane) antibodies show that each antibody labels a single broad protein band.

Figure 2. The use of epifluorescence to count certain cell populations

Kv3 immunoreactivity often appeared as a ring in a restricted plane of the tissue, illustrated here as a black ring around a grayscale neuron. While Kv3 labeling was not only observable at the equator of cells, for many cells labeling was less readily observable at other regions of the cell surface. Therefore, for counts of cells that were dually-labeled for Kv3 subunits and another protein, it was necessary to scan through the full soma of labeled neurons. Had the counts of GABA-, CB-, or PV-ir neurons, for which the immunofluorescent signal filled the volume of immunopositive somata, been conducted using single-plane images, these images would have frequently shown cells in a plane that incorrectly appeared to be Kv3-immunonegative. This counting method would have resulted in an overestimation of the number of neurons that were single-labeled for GABA-, CB-, or PV-immunoreactivity (a). Using single-plane images (i.e. confocal micrographs), we would have only correctly identified a cell as dual-labeled if the image were taken in the precise plane in which the Kv3 signal appeared (b). For this reason, when counting the GABA-, PV-, or CB-ir populations, it was necessary to examine both fluorescent channels, the tissue throughout the depth of the section. This was done using an epifluorescence microscope.

Figure 3. Laminar distribution of Kv3.1b- and Kv3.2-immunoreactivity

Photomicrographs of macaque V1, processed to show immunoreactivity for Kv3.1b (a) and Kv3.2 (b) channel subunits. Black bars between each micrograph show laminar boundaries. Kv3.1b-ir neurons (a) appear evenly distributed throughout layers 2-6. Diffuse staining of the neuropil is also present in all layers, but is stronger in layers 1-4c. Kv3.2-ir neurons (b) are also found in all layers, but appear more prominent in layers 5 and 6 in this low power micrograph as a result of fainter neuropil staining in these layers. Scale bar = 50μm (both panels).

Figure 4. Immunoreactivity for Kv3.1b in V1

Tissue sections of macaque V1 processed for Kv3.1b-immunoreactivity show many labeled somata in all layers. The somatic labeling usually consists of faint intracellular labeling accompanied by an intense ring of immunoreactivity at the perimeter of the cell body (a, thin arrows), perhaps indicating plasmalemmal localization of the Kv3.1b protein. In many Kv3.1b-ir neurons, there is little or no labeling of the major proximal axonal or dendritic processes. Meynert cells (b), a morphologically distinct class of neurons with large somata that reside near the border between layers 5 and 6, show Kv3.1b-immunoreactivity of their somata as well as their proximal apical and basal dendrites (b, thick arrows). Similarly, neurons in layer 4b with large, horizontally-elongated cell bodies and a bipolar morphology express Kv3.1b in their somata and proximal dendrites (c,d). Neither Meynert cells nor 4b bipolars show Kv3.2-immunoreactivity (not shown). Scale bars = 20μm, all panels.

Figure 5. Non-somatic Kv3.1b-immunoreactivity

Kv3.1b-immunoreactivity is occasionally observed in neuronal processes. For example, Kv3.1b is expressed in the upper part of layer 2, in the vertically-oriented dendrites (extending perpendicular to layer 1) (thin arrows in a,b). Kv3.1b-ir is also occasionally observed in axons (c), which can be identified by the presence of varicosities (fat arrows). Scale bars: a,b = 20μm; c = 10μm.

Figure 6. Immunoreactivity for Kv3.2 in macaque V1

Tissue sections of macaque V1 processed for Kv3.2-immunoreactivity show many labeled somata (a) in all layers, although labeled neurons in layers 2 through 4 are difficult to distinguish at low power as a result of prominent neuropil-immunoreactivity (see Fig. 3). Similar to Kv3.1b, somatic labeling for Kv3.2 usually consists of an intense ring of immunoreactivity around the cell body (thin arrows), perhaps indicating plasmalemmal localization of the Kv3.2 protein. In most Kv3.2-ir neurons, there is little or no labeling of the major proximal axonal or dendritic processes, however Kv3.2 is occasionally expressed in axons. These axons are not clearly associated with nearby somata. Note the clearly distinguishable varicosities (thick arrows), which show the typical axonal “beads on a string” appearance. Scale bar, panel a = 20μm. Scale bar panel b = 10μm.

Figure 7. Number (×10³) of Kv3.1b- and Kv3.2-ir neurons per cubic millimeter of tissue in macaque V1

Means and SEM values (n=3) are given for each cortical layer. These values were obtained by the optical disector method.

Figure 8. Some Kv3.1b-ir neurons are non-GABAergic, including neurons in layer 4b

Confocal photomicrographs of tissue processed by dual immunofluorescence to show Kv3.1b-(a,b) and GABA- (c,d) immunoreactivity. The images in the left column (a,c,e), were taken from layer 4b and in the isolated ‘red’ channel (here false-colored magenta, a) several Kv3.1b-ir somata are visible, along with diffuse neuropil staining. Similar to our observations in tissue processed using the immunoperoxidase method, several of the Kv3.1b-ir neurons in layer 4b have large, elongated somata (asterisks; see Fig. 4c,d). The isolated green channel (c) shows labeling of GABA-ir somata, and some neuropil labeling. The merged image (e) shows one Kv3.1b/GABA dually-labeled neuron (thin arrow; e), whereas the two putative bipolar cells that express Kv3.1b do not express GABA (asterisks; a,e). The images in the right column (b,d,f) show a dual-immunofluorescence image from layer 4c with three Kv3.1b-ir neurons (stars; b,d,f), one of which is dual-labeled for GABA (d). Another of the Kv3.1b-ir neurons is surrounded by closely apposed varicosities which are GABA-ir (thick arrows; d,f). These varicosities may represent the puncta of a GABAergic perisomatic basket. Scale bar panel a = 20μm. Scale bar panel b = 10μm.

Figure 9. Some Kv3.2-ir neurons are non-GABAergic

This dual-immunofluorescence image from layer 4c shows labeling for Kv3.2 (a) and GABA (b). The “red” channel (colored magenta in this figure) shows somatic labeling of two Kv3.2-ir neurons and diffuse neuropil labeling. Comparison with the middle panel (b) and the merged image (c) shows that one of these Kv3.2-ir neurons (asterisk) does not express GABA. Scale bar = 10μm.

Figure 10. Most PV-ir neurons express Kv3.1b

This dual immunofluorescence image taken from layer 2/3 of V1 shows that Kv3.1b (a) and PV (b) are frequently co-expressed in the same neurons. All of the somata shown in this image are immunoreactive for both Kv3.1b and PV (i.e. all somata appear in both channels; a,b) and look white in the merged image (c). This image also shows that the subcellular localization of the two antigens differs. Kv3.1b labeling (a) is most intense at the somatic periphery, and usually includes only very proximal axonal or dendritic processes. PV labeling (b), conversely, appears cytosolic and often includes dendrites and axons as well as the cell somata. Scale bar = 20μm.

Figure 11. Kv3.1b and PV do not always co-label somata in macaque V1

The top row (a,d,g) shows a dual immunofluorescence image taken from layer 4c of V1. The isolated “red” channel (a) shows two Kv3.1b-ir somata, as well as neuropil labeling. The green channel (d) shows PV-immunoreactivity in one soma as well as in the neuropil. The merged image (g) shows that one of the Kv3.1b-ir neurons is dually-labeled, while one neuron (asterisk; a,g) expresses Kv3.1b but not PV. In the middle row (b,e,h), a dual immunofluorescence image (also from layer 4c) shows a PV-ir neuron (asterisk; e,h) that does not express the Kv3.1b channel subunit. The merged (h) and separate (b,e) images show one neuron dually-labeled for Kv3.1b (b) and PV (e), and one neuron that expresses PV (asterisk; e,h). The bottom row (c,f,i) shows labeling for Kv3.1b (c) and PV (f). In the “red” (c) and merged (i) channels, Kv3.1b-ir puncta can be seen (thin and fat arrows; c,i). The thin arrows indicate puncta which appear to be clustered close to a soma that is neither Kv3.1b- (c) nor PV-ir (f). These clustered puncta are suggestive of a Kv3.1b-ir perisomatic axonal basket. The puncta indicated by the fat arrow (c,i) are more difficult to clearly interpret as axonal because the postsynaptic neuron (asterisk; c,f,i) is immunoreactive for Kv3.1b. These puncta could equally well represent axonal varicosities or clusters of Kv3.1b subunits on the postsynaptic membrane. Scale bars = 20μm, all panels.

Figure 12. Meynert cells express Kv3.1b but not PV

In this dual immunofluorescent image of neurons in layer 6 the “red” channel (a) shows Kv3.1b-immunoreactivity in two layer 6 neurons, including a Meynert cell (asterisk). The green channel (b) shows one neuron is immunopositive for PV. The merged image (c) shows that the Meynert cell (asterisk) expresses Kv3.1b but not PV. Scale bars = 20μm, all panels.

Figure 13. Some Kv3.2-ir neurons do not express PV

In this dual immunofluorescence image of neurons in layers 2/3 the “red” channel (a) shows three Kv3.2-ir somata. The green channel (b) shows two PV-ir neurons. The merged image (c) shows that both of these PV-ir neurons are dually-labeled for Kv3.2 and PV (asterisks), and the other neuron expressed Kv3.2 but not PV (arrow). Scale bar = 20μm.

Figure 14. Some Kv3.1b-ir neurons in layers 1–4 express CB

This dual immunofluorescence image of neurons in layer 2/3 shows two Kv3.1b-ir somata (a) one of which is also CB-ir (b,c; asterisk). Scale bar = 20μm.

Figure 15. Some Kv3.1b-ir neurons do not express GAD67 mRNA

These panels show a region from layer 3 of tissue dually labelled for Kv3.1b protein (by immunocytochemistry) and for GAD67 mRNA (by in situ hybridization). In panel (a) a number of Kv3.1b-ir neurons can be seen (some of which have been marked with asterisks for comparison with other panels). In the panel (b) the hybridization signal for the same region of tissue shows that while most of these neurons also express GAD67, there is one neuron which does not (arrow). Given that it has been shown that GAD67 is probably expressed in all inhibitory neurons in macaque V1 (Hendrickson et al., 1994), and given that glycine is not a primary neurotransmitter in this region, the GAD67-negative cell is likely to be excitatory. The DAPI stain in the panel (c) confirms that these immunoreactive cells are all neurons. In the three-channel merged image (d) the in situ signal has been false-colored green and inverted. Scale bar = 20μm.

Figure 16. CaMKIIα gene expression by a Kv3.1b-ir neuron

These micrographs show a neuron from layer 2 which is immunoreactive for Kv3.1b protein (asterisk; a). This same neuron also expresses mRNA for CaMKIIα (asterisk; b), which has previously been shown to be expressed only by excitatory neurons in the macaque neocortex (Jones et al., 1994). This neuron’s nucleus appears in the DAPI stain (c) and the soma also appears to have a pyramidal morphology, further suggesting that it is an excitatory neuron. In the merged image (d) the in situ hybridization signal has been inverted and false-colored (green). Scale bar = 20μm.