Modulation of Murine Olivary Connexin 36 Gap Junctions by PKA and CaMKII

Abstract

The inferior olive (IO) is a nucleus located in the brainstem and it is part of the olivo-cerebellar loop. This circuit plays a fundamental role in generation and acquisition of coherent motor patterns and it relies on synchronous activation of groups of Purkinje cells (PC) in the cerebellar cortex. IO neurons integrate their intrinsic oscillatory activity with excitatory inputs coming from the somatosensory system and inhibitory feedback coming from the cerebellar nuclei. Alongside these chemical synaptic inputs, IO neurons are coupled to one another via connexin 36 (Cx36) containing gap junctions (GJs) that create a functional syncytium between neurons. Communication between olivary neurons is regulated by these GJs and their correct functioning contributes to coherent oscillations in the IO and proper motor learning. Here, we explore the cellular pathways that can regulate the coupling between olivary neurons. We combined in vitro electrophysiology and immunohistochemistry (IHC) on mouse acute brain slices to unravel the pathways that regulate olivary coupling. We found that enhancing the activity of the protein kinase A (PKA) pathway and blocking the Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) pathway can both down-regulate the size of the coupled network. However, these two kinases follow different mechanisms of action. Our results suggest that activation of the PKA pathway reduces the opening probability of the Cx36 GJs, whereas inhibition of the CaMKII pathway reduces the number of Cx36 GJs. The low densities of Cx36 proteins and electrical synapses in βCaMKII knock-out mice point towards an essential role for this protein kinase in regulating the density of GJs in the IO. Thus, the level of olivary coupling is a dynamic process and regulated by a variety of enzymes modulating GJs expression, docking and activity.

Keywords: PKA; electrical synapse; inferior olive; tracer-coupling; αCaMKII; β-CaMKII.

Figure 1

Modulation of tracer-coupling by pharmacological intervention. (A) Control condition: besides the primary stain cell, 13 tracer-coupled neurons can be observed. (B) Forskolin condition: besides the primary stain cell, four tracer-coupled neurons can be observed. (C) H-89 condition: besides the primary stain cell, four tracer-coupled neurons can be observed. (D) KN93 condition: besides the primary stain cell, eight tracer-coupled neurons can be observed. In all cases, scale bar is 25 μm. (E) Histogram represents the average (+SEM) number of tracer-coupled neurons in each of the four conditions. Asterisks indicate a statistical significance difference (post hoc Tukey HSD test p < 0.05, n = 12 in Control, Forskolin and H-89, n = 11 in KN93).

Figure 2

Quantification of Cx36 puncta on proximal and distal dendrites. (A) Example of a primary neuron and its dendritic tree. Asterisks indicate coupled neurons. White boxes are the areas chosen for Cx36 quantification and are enlarged in (B–D). Scale bar is 25 μm. (B) Histochemical image of soma and proximal dendritic branches of olivary neuron. Green channel: neurobiotin-staining; Red channel: Cx36 GJ staining. (C) Same as in (B) after center of mass-based Cx36-recognition algorithm. Cx36 positive puncta (dots) that were located on the proximal dendrites are indicated by the white arrows. (D) Distal dendritic branches after center of mass-based Cx36-recognition algorithm. Cx36 positive puncta (dots) that were located on the distal dendrites are indicated by the arrows. In (B–D) scale bar is 6.25 μm. (E) Histogram represents the average (+SEM) number of Cx36 puncta/100 μm proximal dendrite (asterisk indicates statistical significance difference, post hoc Tukey HSD test p < 0.05). (F) Histogram represents the average (+SEM) number of Cx36 puncta/100 μm distal dendrite (asterisk indicates statistical significance difference, post hoc Tukey HSD test p < 0.05).

Figure 3

Ultrastructural analysis of electrical synapses in the inferior olive (IO) of wild-type (WT) and CaMKII mutant mice. (A) Ultrastructural characteristics of an olivary glomerulus in the IO of a WT mouse. The Gap junctions (GJs) between olivary dendritic spines (asterisks) showed electron-dense deposits (in B white arrowheads) in the cytoplasm at both sides of the membrane and attachment plaques surrounding the plaque (black arrowheads) with Cx36-proteins. (B) Higher magnification of the inset depicted in (A). Scale bars: (A) 100 nm; (B) 200 nm. (C1) Ultrastructural morphology of a olivary glomerulus of homozygous βCaMKII-deficient mice. In nearly all glomeruli, no GJ structures were observed, but instead showed two dendritic spines (see asterisks) in the center surrounded by axonal terminals without a neuronal GJ (dashed box in C1, Scale bar: 200 nm). (C2) In one olivary glomerulus an adhesion-plaque structure was observed without a narrow interneuronal space as in normal GJs (Scale bar: 200 nm). (D) In αCaMKII- deficient mice, GJs exhibited a normal distance between attachment plaques (black arrowheads), however their average interneuronal space was significantly wider (Scale bar: 100 nm). (E) Histograms showing density of Cx36 puncta in βCaMKII−/−, αCaMKII−/− and WT mice determined by using immunohistochemistry (IHC). Asterisk indicates statistical significance difference (post hoc Tukey HSD test p < 0.01). (F) Histograms showing morphometrics of GJs in βCaMKII−/−, αCaMKII−/− and WT mice (#: in βCaMKII-deficient mice only one GJ was observed therefore no error bar is depicted). (F1) The average density of GJs in βCaMKII−/− mice was significantly smaller than that of olivary subnuclei in WT and αCaMKII−/− mice. (F2) The length of the gap-junction plaque in βCaMKII−/− mice was significantly shorter than that of WT mice, whereas in homozygous αCaMKII-deficient mice it was comparable to WT mice. (F3) The interneuronal distance did not differ among WT and βCaMKII−/− mice, however the average interneuronal distance of αCaMKII-deficient mice was significantly larger. Asterisk indicates statistical significance difference (post hoc Tukey HSD test p < 0.05).