A loss-of-function analysis reveals that endogenous Rem2 promotes functional glutamatergic synapse formation and restricts dendritic complexity

Abstract

Rem2 is a member of the RGK family of small Ras-like GTPases whose expression and function is regulated by neuronal activity in the brain. A number of questions still remain as to the endogenous functions of Rem2 in neurons. RNAi-mediated Rem2 knockdown leads to an increase in dendritic complexity and a decrease in functional excitatory synapses, though a recent report challenged the specificity of Rem2-targeted RNAi reagents. In addition, overexpression in a number of cell types has shown that Rem2 can inhibit voltage-gated calcium channel (VGCC) function, while studies employing RNAi-mediated knockdown of Rem2 have failed to observe a corresponding enhancement of VGCC function. To further investigate these discrepancies and determine the endogenous function of Rem2, we took a comprehensive, loss-of-function approach utilizing two independent, validated Rem2-targeted shRNAs to analyze Rem2 function. We sought to investigate the consequence of endogenous Rem2 knockdown by focusing on the three reported functions of Rem2 in neurons: regulation of synapse formation, dendritic morphology, and voltage-gated calcium channels. We conclude that endogenous Rem2 is a positive regulator of functional, excitatory synapse development and a negative regulator of dendritic complexity. In addition, while we are unable to reach a definitive conclusion as to whether the regulation of VGCCs is an endogenous function of Rem2, our study reports important data regarding RNAi reagents for use in future investigation of this issue.

Figure 1. Rem2 expression in hippocampal neurons.

A) Western blot with anti-Rem2 antibody of HEK 293T cell lysates. anti-β-actin serves as a loading control. HEK 293T cells were transfected with either myc-Rem2 or RNAi-resistant myc-Rem2 and either the Rem2 short hairpin 1 (RNAi 1) or the Rem2 short hairpin 2 (RNAi 2) to confirm Rem2 knockdown and rescue. B) Representative images of Rem2 (red) and MAP2 (blue) staining in neurons transfected with GFP and either Mock (empty pSuper vector), overexpression with the Rem2 RNAi-resistant cDNA (OE), RNAi with the Rem2 short hairpin 1 (RNAi 1), RNAi with a second Rem2 short hairpin (RNAi 2), or rescue of RNAi 1 (RE 1), or RNAi 2 (RE 2) with the Rem2 RNAi-resistant cDNA. Scale bar = 5 µM. C) Average Rem2 staining intensity (top left) or MAP2 intensity (bottom left) measured in arbitrary units for each condition (Right). Average Rem2 immunostaining intensity normalized to MAP2 intensity in hippocampal neurons. n > 40 neurons per condition. * p ≤ 0.05 compared to mock or ^# p ≤ 0.05 compared to the appropriate RNAi condition (i.e. RE1 vs. RNAi 1) using a two-way ANOVA followed by a Tukey’s post hoc test.

Figure 2. Rem2 is a negative regulator of dendritic complexity.

A) Representative images of Mock, RNAi 1, RNAi 2, OE, RE 1, and RE 2 transfected neurons. Scale bar = 50 µM. B) Quantification of dendritic complexity for each condition using Sholl analysis. n > 40 neurons per condition. * p ≤ 0.05 compared to mock determined by multivariate ANOVA with a Tukey’s post hoc test.

Figure 3. Rem2 is important for excitatory synapse formation in primary hippocampal neurons.

A) Representative images of excitatory synapse staining for the presynaptic marker, synapsin I (blue) and the postsynaptic marker PSD-95 (red) in neurons co-transfected with GFP and either: Mock (empty pSuper vector), overexpression with the Rem2 RNAi-resistant cDNA (OE), RNAi with the Rem2 short hairpin 1 (RNAi 1), RNAi with a second Rem2 short hairpin (RNAi 2), or rescue of RNAi 1 (RE 1), or RNAi 2 (RE 2) with the Rem2 RNAi-resistant cDNA. Scale bar indicates 5 µm. B) Synapse density measured by the co-localization of synapsin I and PSD-95 on a GFP positive stretch of dendrite. n > 40 neurons per condition. * p ≤ 0.05 compared to mock or ^# p ≤ 0.05 compared to each appropriate RNAi condition (i.e. RE1 vs. RNAi 1) using a two-way between-effect ANOVA followed by a Tukey’s post hoc test.

Figure 4. Rem2 is important for functional excitatory synapse formation.

A) Representative 20 s miniature excitatory postsynaptic current (mEPSC) trace selected for each experimental condition: Mock, OE, RNAi 1, RE 1, RNAi 2, and RE 2, recorded from primary hippocampal neurons at 14 DIV. B and C) Average mEPSC frequency (B) and amplitude (C) measured for each condition for a total of 300 s. B and C) n ≥ 15 neurons per condition (3 separate experiments). D) Cumulative distribution of mEPSC amplitudes for Mock, OE, RNAi 1, RE 1, RNAi 2, and RE 2 generated from 100 random mEPSC per cell (n=15 cells per condition). RNAi 1 and RNAi 2 mEPSC amplitude plots were significantly different from Mock (p < 0.05, Kolmogorov-Smirnov test). Additionally, the leftward shift following RNAi 1 and RNAi 2 was significantly different from their respective rescue conditions (RE 1 and RE 2, respectively; p < 0.05). OE, RE 1, and RE 2 were not significantly different from Mock. (B and C) * p ≤ 0.05 compared to mock or ^# p ≤ 0.05 compared to the appropriate RNAi condition using a one-way ANOVA followed by a student’s independent t-test.

Figure 5. Rem2 overexpression significantly reduces voltage-gated calcium current.

A) Cells were given a series of voltage steps (-90 to +40 mV) from a holding potential of -70 mV (inset). Representative isolated calcium current, following offline subtraction of a recording in the presence of nifedipine (10 µM). B) Current-voltage (I–V) relationship of calcium current density (pA/pF) following Rem2 knockdown, rescue, or overexpression. C) Peak calcium current amplitude was determined following a ramp step from -80 mV to +40 mV (duration 500 ms). n ≥ 10 cells per condition (3 separate experiments). * p ≤ 0.05 compared to mock or ^# p ≤ 0.05 compared to each appropriate RNAi condition determined using a one-way ANOVA followed by a student’s independent t-test.