DEPDC5 regulates the strength of excitatory synaptic transmission by interacting with ubiquitin-specific protease 46

Abstract

DEP-domain containing-5 (DEPDC5) is part of the GATOR1 complex that inhibits the mechanistic target of rapamycin complex-1 (mTORC1). Loss-of-function mutations in human DEPDC5 are the most common cause of lesional or non-lesional focal epilepsies associated with mTOR hyperactivation. Depdc5 silencing in mature neurons leads to excitation/inhibition imbalance and increased excitatory synapse strength. However, no link exists between mTORC1 hyperactivity and the increased activity of glutamatergic synapses. Here, we found that genetic deletion of Depdc5 in a conditional knockout (cKO) mouse recapitulates the excitatory/inhibitory imbalance observed after transient Depdc5 silencing, with increased strength of excitatory transmission and unaffected inhibitory transmission. In Depdc5 cKO neurons, the increased glutamate quantal size and response to exogenous glutamate are attributable to a higher density of GluA1-containing AMPA glutamate receptors due to a shift of the GluA1 subunit from the intracellular pool to the plasma membrane. The DEPDC5 protein interaction network included WDR48, WDR20, and USP46, a ubiquitin-specific protease that regulates GluA1, as key binding partners, along with previously established components of the mTORC1 signaling pathway. In the absence of DEPDC5, USP46 levels increase, and ubiquitination of GluA1 decreases accordingly. Either knockdown of USP46 or rapamycin treatment rescues both the increased glutamate quantal size and USP46 increase caused by Depdc5 deletion, indicating that USP46 overexpression depends on mTORC1 hyperactivity. The data indicate that the DEPDC5/mTORC1 system physiologically controls the excitatory strength by negatively modulating USP46 activity and AMPA receptor deubiquitination, and that failure of this effect can contribute to the development of the Depdc5-linked epileptic phenotype.

Keywords: AMPA receptors; DEPDC5 interactome; Depdc5 mutations; Deubiquitination; Excitatory synaptic strength; Primary neurons.

Graphical abstract

Fig. 1

Characterization of the Depdc5 cKO mouse. A. Schematics of the targeting vector for the Depdc5 cKO mouse through the Cre/LoxP system. B. Left: Box plot of Depdc5 mRNA levels in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons (n = 4 independent preparations). Middle and Right: Representative Western blot and respective box plot of DEPDC5 protein levels in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons (n = 6 independent preparations). C. Left: Representative Western blot of phosphorylated S6, total S6, and actin (used as a control of equal loading) in primary cortical Depdc5^ΔCre and Depdc5^Cre transduced neurons. Right: Box plots of pS6/S6 and S6/Actin ratios (n = 4 independent preparations). D. Left: Representative confocal images of the transduction reporter GFP, phosphorylated S6 (pS6), and total S6 immunofluorescence in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons. Scale bar, 10 μm. Right: Bar plots of neuronal soma size of GFP-positive neurons and pS6/S6 immunoreactivity ratio determined in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons (n = 33 and n = 39 neurons, respectively, from n = 2 independent preparations). E. Left: Representative Western blot of phosphorylated and total S6 in Depdc5^ΔCre and Depdc5^Cre neurons treated with either vehicle (DMSO) or rapamycin (100 nM). Right: Bar plot of the respective pS6/S6 ratio (from n = 7 independent preparations). *p < 0.05, **p < 0.01, ***p < 0.001; unpaired Student's t-test/Mann-Whitney's U test (B-D), two-way ANOVA /Tukey's tests (E).

Fig. 2

Conditional knockout of Depdc5 increases the excitatory synaptic strength. A. Representative traces of mEPSCs recorded in Depdc5^ΔCre and Depdc5^Cre primary cortical neurons. B. Box plot (left) and cumulative plot (right) of mEPSC amplitude in Depdc5^ΔCre and Depdc5^Cre neurons. C. Box plot of mEPSC frequency (left) and respective cumulative plot of the interevent intervals (right) in Depdc5^ΔCre and Depdc5^Cre neurons. D. Representative waveforms of mEPSCs and box plots of the mEPSC 10–90 rise time, 80 % decay time, and transferred charge in Depdc5^ΔCre and Depdc5^Cre neurons (n = 18 and 17 for Depdc5^ΔCre and Depdc5^Cre neurons, respectively, from n = 3 independent preparations). **p < 0.01; ***p < 0.001; Student's t-test/Mann-Whitney's U test.

Fig. 3

Increased density of excitatory synapses and synaptic expression of the AMPA GluA1 receptor subunit in Depdc5 cKO cortical networks. A. Representative confocal images of excitatory synaptic contacts identified by vGlut1/Homer double immunostaining in Depdc5^ΔCre (left) and Depdc5^Cre (right) primary cultures. B. Representative confocal images of the AMPA GluA1 receptor subunit at Homer-positive glutamatergic synapses in Depdc5^ΔCre (left) and Depdc5^Cre (right) primary cultures. In A and B, the top panels are merged images of individual immunoreactivities, as shown below, with the highlighted colocalization puncta. Scale bar, 50 μm. C,D. Box plots showing the linear density of excitatory synapses (C) and the intensity of the immunoreactivity for the AMPA-GluA1 receptor subunit within Homer-positive puncta (D) in Depdc5^ΔCre and Depdc5^Cre networks. The dendrites of >20 neurons from n = 2 independent neuronal preparations were analyzed. *p < 0.05, **p < 0.01; unpaired Student's t-test.

Fig. 4

An increased surface expression of AMPA glutamate receptors characterizes Depdc5 cKO neurons. A. Representative image of a patched cortical neuron subjected to glutamate puffs (left) and representative current traces evoked by a glutamate puff (1 mM; right) in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons. B. Box plots of the glutamate-evoked current (pA; left) and the respective current density J (pA/pF; right) recorded in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons (n = 15 and 17 neurons, respectively, from n = 3 independent preparations). C. Representative immunoblot of cell surface biotinylation performed in primary cortical Depdc5^ΔCre and Depdc5^Cre transduced neurons. Total cell lysates (input; 10 %), biotinylated (extracellular), and non-biotinylated (intracellular) fractions of GluA1 were analyzed by immunoblotting. Na⁺/K⁺-ATPase 3 and actin were used as markers of plasma membrane and cytosolic fractions, respectively. Immunoreactive bands were cut from the same blot. D. Quantification of the total, extracellular, and intracellular fractions of GluA1 normalized on Na⁺/K⁺-ATPase 3 and actin expression, respectively, and shown in percentage of the mean value of the Depdc5^ΔCre group (n = 6 from 4 independent neuronal preparations). *p < 0.05; unpaired Student's t-test.

Fig. 5

Analysis of the DEPDC5 interactome reveals an association with the USP46 de-ubiquitination complex. A. Schematic representation of the Tandem Affinity Purification protocol for isolating double-tagged Depdc5 from the brain of untreated conditional Depdc5 mice. B. Confirmation of DEPDC5 interaction with USP46 in cytosolic fractions. Representative Western blots of the total homogenate (input), Flag, and HA eluates revealed with anti-DEPDC5 and anti-USP46 antibodies (left). DEPDC5-USP46 interaction as shown in co-immunoprecipitation experiments from total homogenate (input) and after TAP purification (right). The anti-DEPDC5 and anti-USP46 immunoblots are shown at two chemiluminescence exposure times. C. Representative confocal images of the PLA assay depicting the negative control (omission of primary antibody), the positive control (MAP2-βIII Tubulin), and the interaction between DEPDC5 and USP46 (red, PLA puncta; green, phalloidin488; blue, DAPI nuclear staining; scale bar, 10 μm). Bottom right: Box plots of the number of PLA-positive puncta counted at the cell body level (DEPDC5/USP46) and along 30 μm of proximal dendrites (MAP2-βIII Tubulin) for the three analyzed conditions. *p < 0.05, ****p < 0.0001; Kruskal-Wallis/Dunn's tests (n = 4–6 cells per experimental group from 2 independent preparations). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

Conditional knockout of Depdc5 increases the expression of USP46 and decreases the ubiquitination of the GluA1 subunit of AMPA glutamate receptors. A. Left: Representative Western blot showing the upregulation of USP46 immunoreactivity in primary cortical Depdc5^Cre neurons with respect to Depdc5^ΔCre controls. Right: Box plot showing total cell USP46 expression normalized on actin expression and shown in percentage of the mean value of the Depdc5^ΔCre group (n = 4 from 3 independent preparations). B. Left: Representative Western blots of ubiquitin immunoreactivity in the GluA1 molecular mass range and of GluA1 in extracellular fractions obtained from biotinylation assays of primary cortical Depdc5^ΔCre and Depdc5^Cre neurons. Right: Box plot of the percent ratios between ubiquitin immunoreactivity in the 110 kDa molecular mass region and the corresponding immunoreactivity intensity of GluA1, expressed in percentage of the mean value of the Depdc5^ΔCre group (n = 4 from 3 independent preparations). *p < 0.05; unpaired Student's t-test.

Fig. 7

USP46 knockdown reduces the increased quantal size, but not the mEPSC frequency, of excitatory synaptic transmission in Depdc5 cKO neurons. A. Representative traces of mEPSCs recorded in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons co-transduced with either Usp46^Scr or Usp46^Sh expressing lentiviruses. B. Box plots of the mEPSC amplitude (left) and frequency (right) in Depdc5^ΔCre-Usp46^Scr (n = 18), Depdc5^ΔCre-Usp46^Sh (n = 17), Depdc5^Cre-Usp46^Scr (n = 21), Depdc5^Cre-Usp46^Sh (n = 14) neurons from n = 3 independent neuronal preparations. C. Representative confocal images of the AMPA GluA1 receptor subunit at Homer-positive glutamatergic synapses in Depdc5^ΔCre (left) and Depdc5^Cre (right) primary cultures co-transduced with either Usp46^Scr or Usp46^Sh expressing lentiviruses. The top panels are merged images of individual immunoreactivities, as shown below, with the highlighted colocalization puncta. Scale bar, 50 μm. D. Box plots showing the intensity of the GluA1 immunoreactivity within Homer-positive puncta in the four experimental groups. Dendrites of >15 neurons from n = 2 independent neuronal preparations were analyzed. *p < 0.05, **p < 0.01, ****p < 0.0001; 2-way ANOVA/Tukey's tests.

Fig. 8

The USP46-dependent increase in the quantal size of excitatory synapses in Depdc5 cKO neurons depends on mTORC1 hyperactivity. A. Representative traces of mEPSCs recorded in primary cortical Depdc5^ΔCre and Depdc5^Cre neurons (DIV 14) that were treated with either vehicle or rapamycin (100 nM). B. Box plots of the mEPSC amplitude (left) and frequency (right) in Depdc5^ΔCre (n = 14), Depdc5^Cre (n = 14), Depdc5^Cre-rapamycin (n = 15) neurons from n = 2 independent neuronal preparations. *p < 0.05, **p < 0.01; one-way ANOVA/Tukey's tests. C. Representative immunoblot of cell surface biotinylation performed in primary cortical Depdc5^ΔCre and Depdc5^Cre transduced neurons treated with either vehicle or rapamycin (100 nM). Total cell lysates (input; 10 %), biotinylated (extracellular), and non-biotinylated (intracellular) fractions of GluA1 were analyzed by immunoblotting. Na⁺/K⁺-ATPase 3 and actin were used as markers of plasma membrane and cytosolic fractions, respectively. Immunoreactive bands were cut from the same blot. D. Quantification of the total, extracellular, and intracellular fractions of GluA1 normalized on Na⁺/K⁺-ATPase 3 and actin expression, respectively, and shown in percentage of the mean value of the Depdc5^ΔCre group (n = 3 from 2 independent neuronal preparations). E. Representative Western blot of USP46 immunoreactivity showing the rescue of USP46 expression in primary cortical Depdc5^Cre neurons treated with rapamycin with respect to either Depdc5^ΔCre controls or Depdc5^Cre neurons treated with vehicle. F. Box plot showing total cell USP46 normalized on actin expression and shown in percentage of the mean value of the Depdc5^ΔCre/vehicle group (n = 5 from 2 independent preparations). *p < 0.05, **p < 0.01; one-way ANOVA/Fisher's tests.