C12ORF57: a novel principal regulator of synaptic AMPA currents and excitatory neuronal homeostasis

Abstract

Objective: Excitatory neuronal homeostasis is crucial for neuronal survival, circuit function, and plasticity. Disruptions in this form of homeostasis are believed to underpin a variety of neuronal conditions including intellectual disability, epilepsy, and autism. However, the underlying genetic and molecular mechanisms maintaining this homeostasis remain poorly understood. Biallelic recurrent loss of function mutations in C12ORF57, an evolutionarily conserved X amino acid novel open reading frame, underlie Temtamy syndrome (TS)-a neurodevelopmental disorder characterized by epilepsy, dysgenesis of the corpus callosum, and severe intellectual disability.

Methods: Through multiple lines of inquiry, we establish that C12ORF57/GRCC10 plays an unexpected central role in synaptic homeostatic downscaling in response to elevated activity, uncovering a novel mechanism for neuronal excitatory homeostasis. To probe these mechanisms, we developed a new knockout (KO) mouse model of the gene's murine ortholog, Grcc10 as well as cellular and in vitro assays.

Results: Grcc10 KO mice exhibit the characteristic phenotypic features seen in human TS patients, including increased epileptiform activity. Corresponding with the enhanced seizure susceptibility, hippocampal neurons in these mice exhibited significantly increased AMPA receptor expression levels and higher amplitude of miniature excitatory postsynaptic currents (mEPSCs). We further found that GRCC10/C12ORF57 modulates the activity of calcium/calmodulin dependent kinase 4 (CAMK4) and thereby regulates the expression of CREB and ARC.

Interpretation: Our study suggests through this novel mechanism, deletion of Grcc10 disrupts the characteristic synaptic AMPA receptor downscaling that accompanies increased activity in glutamatergic neurons.

Keywords: AMPA receptors; ASD; C12ORF57; CAMK4; Temtamy syndrome; epilepsy.

Figure 1. Grcc10−/− Mice Recapitulate the Temtamy Syndrome Phenotype

A) RT PCR of C12ORF57 from fibroblasts derived from a human patient with homozygous c.1 A>G mutations in C12ORF57 see loss of transcript when compared to wild type (WT) controls. B) Ratios of surviving Grcc10 genotype mice are shown at ages P0, P3, and P21 (N=154, p=0.99, Chi-Square). C) Mean body mass (g) for ages P0, P5, P9, P13, and P21 (N=5 for KO mice N=14 for WT and heterozygotes, p>0.0001 ANOVA) D) In situ hybridization of C12ORF57 shows diffuse widespread expression of Grcc10/C12ORF57 throughout all brain regions throughout embryonic development. E) Representative images of midsagittal slices from Grcc10+/+, Grcc10+/− and Grcc10+/+ mouse brains (Scale Bar 1000μm) along with graphs of cortical length (p=0.1127, ANOVA with Tukey post-hoc test), corpus callosum length (p=0.017, ANOVA with Tukey post-hoc test) and corpus callosum cross sectional area cross-sectional area (p=0.0058, ANOVA with Tukey’s post hock). N=6 for KO, N=3 for WT, N=3 for Het. Median represented by horizontal line F) Pictographic of Racine stages G) Average Racine stage (x-axis) of mice in time post kainic acid injection (y-axis) for Grcc10+/+ (blue, N=6), Grcc10+/− (pink, N=7), and Grcc10−/− (green, N=4) (p=0.00162, ANOVA with post hoc Tukey test). Statistical significances are indicated by the following symbols *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P < 0.0001.

Figure 2. Grcc10 −/− neurons have increased mEPSCs amplitude and frequency

A) Representative AMPA mEPSC tracings of Grcc10 +/+ (WT) neurons (top), Grcc10 +/− (Het) neurons (middle) and Grcc10 −/− neurons (bottom). B) Graph of cumulative frequency (x-axis) of mEPSC against amplitude in DIV 17 pyramidal neurons. The mean curve of each group is bolded. C) AMPA mEPSC amplitudes (pA) (p<0.0001, N=5 for each group, ANOVA) D) NMDA mEPSC amplitudes (pA) with median represented by horizontal line. (p=0.20 N=5 Welch’s T-test) E) GABA mEPSC amplitude with median represented by horizontal line (p=0.16, N=5 for each group, Welch’s T-test) F and G) Upper panels: Cumulative frequency of events plotted against event amplitude (left panel) and the time between events (events/second) (right panel) in DIV17 pyramidal neurons including C12ORF57 transfected rescue (purple) and GFP Control (green). The combined mean of group is represented by corresponding bolded line. Lower panels: Mean amplitude of AMPA mEPSC (left panel) and mean events/second of AMPA mEPSC (right panel). For (F) and (G), N=5 for each group, ANOVA with Tukey post-hoc test). Statistical significances are indicated by the following symbols *P ≤0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P < 0.0001.

Figure 3. Loss of Grcc10/C12ORF57 disrupts normal GluA2 expression in primary murine neurons

A) (i) Representative images of surface staining of MAP (white, upper panels) and GluA2 (red, lower panels) in Grcc10 +/+ and Grcc10−/− neurons. Insets 2x magnification. Scale bar 20 μm. Mean GluA2 fluorescent puncta/100μm (ii) (p=0.0024, N=10 both genotypes Welch’s T-test), and mean puncta intensity (ii) (p=0.0008, N=19 WT, N=35 KO, Welch’s T-Test) Error bars represent SEM. B) (i) Sample images of DIV 17 hippocampal primary neurons stained for GluA2 (red) and MAP2 (white) on both Grcc10 WT and Grcc10 KO neurons with and without bicuculline treatment. Scale bar 20 μm. Insets 10x magnification (ii) GluA2 puncta/100 μm with and without bicuculine, with median represented as horizontal line (p=0.0041, N=19 WT Ctl, N=15 WT+bicuculline treated Welch’s T-test. Plot of GluA2 puncta/100μm on KO neurons (p=0.68, T=0.41, df=53, N=20 KO Ctl, N=35 KO+bicuculline treated. Welch’s T-Test). Scale bar 20 μm. C) (i) Representative Golgi-Cox-stained dendrites from Grcc10 WT and KO cortical neurons (scale bar 5μm). (ii) Plot of spine density (p=0.0023, N=33 per group, Welch’s T-Test) and (iii) spine width (p=0.83, N=33 per group, Welch’s T-Test). Medians represented by horizontal line in all graphs. Statistical significances are indicated by the following symbols *P ≤ 0.05, **P ≤0 .01, ***P ≤ 0.001, ****P < 0.0001.

Figure 4. Loss of C12ORF57 decreases levels of downstream CAMK4 targets

A) IHC staining of primary hippocampal neurons from Grcc10 (upper panels) and Grcc10 KO (lower panels) mice for phospho-CREB (green, left), ARC (red, center) with overlay of MAP-2 Yellow and DAPI (blue, right). Scale bar 5μm B and C) Fluorescent quantification of ARC (B) (p=0.0008, N=20 for each genotype, Welch’s T-Test) and p-CREB (C) (p<0.0001, N=24, Welch’s T-Test). Median RFU are indicated by horizontal line. D) Left: Western blot of whole brain lysates of pCREB (top) CREB (middle) and Caveolin-1 (Cav-1) loading controls (bottom). Middle: Ratio of CREB/pCREB fluorescence relative to Cav-1 loading control (p=0.024, N=10, Welch’s T-Test). Normalized to WT Mean. Right: Total CREB normalized to loading control (p=0.29,N=5, Welch’s T-test) E) Left: Western blot of whole brain lysates from Grcc10 WT and Grcc10 KO mice with anti-ARC (top) and anti-actin control (bottom) antibodies. Right: ARC band intensity to actin normalized to mean WT intensity (p=0.0036, N=5 for each genotype, Welch’s T-Test). F) Left: Western blot of whole brain lysates from Grcc10 KO and WT mice with anti-cFos (top) and anti-actin control (bottom) antibodies. Right: Ratio of cFos band intensity to actin normalized to WT mean (p=0.0043, N=5 for each genotype) G) Relative luminescent signal from Renilla CRE luciferase assay on HEK293 cell lysates normalized to WT control (N=6 for each genotype, Welch’s T-Test.) and constitutive promoter positive controls (right). For all box and whisker graphs, min max and median and interquartile range are represented. Statistical significances are indicated by the following symbols *P ≤0.05, **P ≤0 .01, ***P ≤ 0.001, ****P < 0.0001.

Figure 5. C12ORF57 binds CAMK4 and modulates its phosphorylation and activity

A) Western blot showing co-immunoprecipitation of CAMK4 with C12ORF57 from Grcc10+/+ whole brain lysate (left) with Grcc10−/− brain negative control (right) showing no GRCC10 expression or CAMK4 pulldown B) Representative CAMK4 and GAPDH western blot bands from brain whole cell lysates of Grcc10 +/+ and Grcc10−/− mice C) Mean relative CAMK4 western blot band intensity normalized to WT (p=0.34, N=6 for each genotype Welch’s T-Test) D) Representative CAMK4, pCAMK4 and Lamin A/C (control) bands on western blot from nuclear fraction of Grcc10 WT and KO whole brain lysates E) Relative CAMK4, pCAMK4 and Lamin A/C band intensity normalized to Grcc10 WT levels. (p=0.001, N=6 for each genotype, 2-way ANOVA with Sidak’s multiple comparison test). F) Representative CAMK4, pCAMK4 and GAPDH (control) bands on western blot from cytoplasmic fractions of whole brain G) Relative CAMK4 pCAMK4 and Lamin band intensity normalized to Grcc10 WT levels (p=0.016, N=6 for each genotype, 2-way ANOVA, Sidak’s multiple comparison test) in KO compared to WT. H) Schematic of WT CAMK4 (top) with kinase domain (blue), calmodulin binding domain (purple) and autoinhibitory/regulatory domain (yellow). Construct Δ(322–341) which lacks the autoinhibitory domain (NAI, middle) and construct Δ(305–321) which lacks the calmodulin binding domain (NCB, bottom) I) Representative western blot of co-IP of CAMK4 constructs with C12ORF57-FLAG J) Relative CAMK4 band intensity between WT, NCB (p=0.24, N=8 replicates for each construct, ANOVA) and NAI (p<0.0001, N=8 replicates for each construct, ANOVA) CAMK4 constructs. K) Relative luminescence from in vitro CAMK4 kinase activity assay for C12ORF57, CAMK4, CAMK4+C12ORF57, CAMK4+PP1CA, CAMK4+PP1CA+C12ORF57, no CAMK4, and no ATP (N=3 per condition, ANOVA). L) Left: Graph of cumulative frequency (x-axis) of mEPSC against amplitude (pA) in DIV 17 pyramidal neurons including CAMK4 transfected control (green) KO (pink) WT (black), and C12ORF57 transfected (purple). The mean curve of each group is bolded. Right: Amplitude of WT, KO, CAMK4 transfected (KO+CAMK4) and C12OR57 transfected (KO+C12ORF57) AMPA mEPSCs (p<0.0001, N=6 for each group, ANOVA). For all box and whisker graphs, min, max and median and interquartile range are represented. Statistical significances are indicated by the following symbols *P ≤0 .05, **P ≤0.01, ***P ≤ 0.001, ****P < 0.0001.

Figure 6. Proposed Model of C12ORF57 Function

The proposed normal function of C12ORF57 (top panel). Under normal conditions C12ORF57 binds CAMK4 and blocks dephosphorylation leading to increased active CAMK4 and activation of ARC and down regulation of AMPA. Loss of C12ORF57 (bottom), where in response to the same level of calcium ion influx during depolarization/presynaptic excitation, more CAMK4 is dephosphorylated, and less is able to translocate to the nucleus, decreasing downstream pCREB transcription and ARC transcription resulting in unopposed long-term potentiation, increased GluA2 on cell surface and increased AMPA current.