Ouabain Modulates the Functional Interaction Between Na,K-ATPase and NMDA Receptor

Abstract

The N-methyl-D-aspartate (NMDA) receptor plays an essential role in glutamatergic transmission and synaptic plasticity and researchers are seeking for different modulators of NMDA receptor function. One possible mechanism for its regulation could be through adjacent membrane proteins. NMDA receptors coprecipitate with Na,K-ATPase, indicating a potential interaction of these two proteins. Ouabain, a mammalian cardiotonic steroid that specifically binds to Na,K-ATPase and affects its conformation, can protect from some toxic effects of NMDA receptor activation. Here we have examined whether NMDA receptor activity and downstream effects can be modulated by physiological ouabain concentrations. The spatial colocalization between NMDA receptors and the Na,K-ATPase catalytic subunits on dendrites of cultured rat hippocampal neurons was analyzed with super-resolution dSTORM microscopy. The functional interaction was analyzed with calcium imaging of single hippocampal neurons exposed to 10 μM NMDA in presence and absence of ouabain and by determination of the ouabain effect on NMDA receptor-dependent long-term potentiation. We show that NMDA receptors and the Na,K-ATPase catalytic subunits alpha1 and alpha3 exist in same protein complex and that ouabain in nanomolar concentration consistently reduces the calcium response to NMDA. Downregulation of the NMDA response is not associated with internalization of the receptor or with alterations in its state of Src phosphorylation. Ouabain in nanomolar concentration elicits a long-term potentiation response. Our findings suggest that ouabain binding to a fraction of Na,K-ATPase molecules that cluster with the NMDA receptors will, via a conformational effect on the NMDA receptors, cause moderate but consistent reduction of NMDA receptor response at synaptic activation.

Keywords: Calcium; LTP; NMDAR; Na,K-ATPase; Ouabain; Protein-protein interaction; Super-resolution microscopy.

Fig. 1

Calcium imaging of NMDA evoked responses in hippocampal neurons in absence and presence of ouabain. a Traces of calcium responses in cells exposed to NMDA (4 μM) without and with ouabain (5 nM) for 2 min. b Calcium response upon subsequent NMDA stimulation of the same cell. c Calcium response of NMDA treatment alone and together with ouabain (1 nM). d Calcium response of NMDA treatment alone and together with ouabain (5 nM). Boxplots show the peak amplitude normalized to median of peak 1 amplitudes. n.s.—p > 0.05, *p < 0.05, ***p < 0.001, Wilcoxon signed-rank test

Fig. 2

Rapid effect of nanomolar concentrations of ouabain on NMDAR-dependent calcium influx by local stimulation. a Traces of calcium responses in cells exposed to NMDA (10 μM) without and with ouabain (10 nM) for 20 s. Same concentrations apply in b and c. b Calcium response in cells exposed to subsequent 20-s pulses of NMDA. Boxplots show mean ΔF/F0 for the first and second NMDA application. c Boxplot presentation over ΔF/F0 upon stimulation with NMDA alone and respective NMDA + ouabain. n.s.—p > 0.05, ***p < 0.001, Wilcoxon signed-rank test. d Boxplot shows calcium responses during 20-s treatments with glutamate (10 μM) + D-AP5 (50 μM) and glutamate + D-AP5 + ouabain (10 nM). n.s.—p > 0.05, Wilcoxon rank sum test

Fig. 3

The ouabain effect on NMDAR-mediated calcium response is not due to internalization of NMDAR or Src-activation. a Comparison between cytosolic and membrane expression of NMDAR in the presence of NMDA (10 μM) alone and with ouabain (10 nM). The endocytotic index was measured in cells treated for 5 min and 15 min. One-way ANOVA and Bonferroni’s post hoc test were used. b Traces of calcium responses in cells treated for 20 s with NMDA (10 mΜ) alone and with ouabain (10 nM), in the presence of PP2 (1 μM). c Boxplot presentation over ΔF/F0 upon stimulation with NMDA and respective NMDA + ouabain. ***p < 0.001, Wilcoxon signed-rank test

Fig. 4

Tyrosine phosphorylation of GluN2B. a Antibody labeling of the three SFK phosphorylation sites of GluN2B: pY1252, pY1336, and pY1472, and total GluN2B in dendrites following 5 min application of vehicle or 10 nM ouabain. Boxplots of the ratio of the fluorescent intensity of pYGluN2B/GluN2Btotal normalized to the mean under control conditions with vehicle. b As in a but exposed to DMSO (vehicle) or 10 μM PP2 for 5 min. Boxplots of the ratio of the fluorescent intensity of pYGluN2B/GluN2B total normalized to the mean under control conditions with DMSO. n.s.—p > 0.05, *p < 0.05, ***p < 0.001, Wilcoxon rank sum test. Scale bar = 10 μm

Fig. 5

Proximity ligation assay (PLA) images of rat hippocampal neurons. PLA was performed using antibodies against Na,K-ATPase α-subunits (NKAα1 and NKAα3) and against GluN2-subunits (GluN2A and GluN2B). Omission of primary antibodies was used as negative control, and antibodies against GluN1 was used as a positive control. Red dots indicate PLA signal; cell nuclei are identified using DAPI stain. Upper left image is PLA of NKAα3 with GluN2B counterstained with pan neuronal marker in green to visualize neurons with extensions. Scale bar = 50 μm

Fig. 6

Super-resolution imaging shows that NMDAR and Na+,K + -ATPase are in close proximity in hippocampal neurons. a–d Overview images of dendritic segments from dSTORM experiments with antibody labeled GluN2-subunits (GluN2A and GluN2B) and Na,K-ATPase α-subunits (NKAα1 and NKAα3). The GluN2 antibodies were tagged with Alexa-647 conjugated secondary antibodies, and Na,K-ATPase antibodies were tagged with Atto-488 conjugated secondary antibodies. Images show Gaussian representations of clusters containing single localized molecules of dendritic segments and GluN2 subunit containing clusters. Cumulative probability distributions of quantified distances from GluN2 subunits to the closest Na,K-ATPase α and vice versa. Histogram insets show the relative frequency of distances from GluN2 subunits to the closest Na,K-ATPase α (insets). Scale bar = 5 μm (overviews of dendrites) and 200 nm (zoom in on clusters)

Fig. 7

Ouabain decrease long-term potentiation in hippocampal CA3-CA1 synapses. a Time-course of changes in field EPSP (fEPSP) slopes after induction of LTP by theta-burst stimulation (arrow) in control (clear circles) and in 50 nM ouabain-treated rat hippocampal slices (filled circles). The solid gray bar denotes the time that ouabain was present in the perfused aCSF. Representative traces averaged fEPSP at baseline (dark gray) and after theta-burst stimulation (0–5 min, light gray; 55–60 min, black). b, c Bars summarizing the average changes (5 min) in fEPSP slopes after Ø-burst stimulation at the times indicated by dotted boxes in the time-course plot. n.s.—p > 0.05, *p < 0.05, Wilcoxon rank sum test

Fig. 8

Ouabain induces a change in NKA/NMDA receptor protein-protein interaction and downregulates NMDA receptor–dependent calcium influx