miR-329– and miR-495–mediated Prr7 down-regulation is required for homeostatic synaptic depression in rat hippocampal neurons

DNA constructs

All primer sequences used for cloning are indicated in Table S1. miR-30a-chimeric hairpins for miR-329 and miR-495 stable overexpression were generated via polynucleotide cloning into the 3′ UTR of eGFP in the pAAV-hSyn-EGFP vector (Plasmid #114213; Addgene) using BsrGI and HindIII sites, as described previously (Christensen et al, 2010).

Control and Prr7 shRNA vectors were constructed using the pSUPER RNAi System (Oligoengine). Custom primers were designed for polynucleotide cloning into the pSUPER basic vector (VEC-PBS-0001/0002) using BglII/HindIII sites, to generate an shRNA targeting a 19-nucleotide sequence unique to the rat Prr7 coding region (cggaatcggacatgtctaa).

For the Prr7 expression construct, full-length rat Prr7 coding sequence (NM_001109116.1) was amplified from hippocampal rat cDNA and then subcloned into the CMV-pcDNA3 vector using BamHI/XbaI sites. Subsequently, a start codon (atg) with HA-tag (tacccatacgacgtcccagactacgct) was inserted at the HindIII site upstream of Prr7 by polynucleotide cloning. To generate the shRNA-resistant construct, six point mutations in the Prr7 coding region were introduced, such that Prr7 shRNA could no longer recognize the mRNA product, yet the amino acid sequence of the resultant exogenous Prr7 protein (AESDMSK) would remain unchanged. Mutagenesis was performed using Phusion-site directed mutagenesis kit (Thermo Fisher Scientific).

Bi-cistronic reporter constructs for miRNA activity were described previously (Fiore et al, 2009). Two perfectly complementary binding sites for either miR-329 and miR-495, separated by a two-nucleotide linker, were inserted into the dsRED 3′ UTR of the pTracer-CMV-dsRED vector, using XbaI/NotI sites by polynucleotide cloning.

Wild-type and mutant Prr7 3′ UTR luciferase constructs were described and generated previously Lackinger et al (2019), wherein the 3′ UTR of Prr7 (NM_001030296.4) was amplified from mouse DNA and cloned into the pmiRGLO dual-luciferase expression vector (Promega). Mutations in miRNA-binding sites conserved across mammals were introduced by site-directed mutagenesis using Pfu Plus! DNA Polymerase (Roboklon).

CDK5-dominant negative construct (CDK5D144N) previously published (Seeburg et al, 2008) was a kind gift from D. T. Pak. Preliminary validation, and titrations were performed to determine the optimal vector concentration before experiments (Fig S12). As a control for the CDK5-dominant negative condition, a plasmid for exogenous expression of an inactive form of CREB (CREB-VP16m) was used (gift from ME Greenberg, used in Fiore et al [2014]).

Cell culture

Primary cortical and hippocampal neuronal cultures were prepared from embryonic day 18 (E18) male and female Sprague-Dawley rats (Janvier Laboratories) as previously described (Schratt et al, 2006). Euthanasia of pregnant rats for the removal of embryonic brains was approved by the Veterinary Office of the Canton Zurich, Switzerland, under license ZH027/2021. Dissociated cortical neurons were directly seeded on six-well plates coated with poly-L-ornithine (used for nucleofections), whereas hippocampal neurons were seeded on poly-L-lysine/laminin-coated coverslips in 24-well plates.

For compartmentalized cell cultures, dissociated hippocampal cells were plated onto 1-μm pore and 30-mm diameter polyethylene tetra-phthalate (PET) membrane filter inserts (Millipore) that were matrix-coated with poly-L-lysine (Sigma-Aldrich) and laminin (BD Biosciences) on the top and bottom, also as described previously (Bicker et al, 2013). With the exception of cells for electrophysiology, all neuron cultures were maintained in Neurobasal-plus (A3582901; Thermo Fisher Scientific) media supplemented with 2% B27, 2 mM GlutaMAX, 100 μg/ml streptomycin, and 100 U/ml penicillin (Gibco; Invitrogen) in an incubator with 5% CO₂ at 37°C. Hippocampal cells used for electrophysiology were maintained in Neurobasal-A (10888022; Thermo Fisher Scientific) media supplemented with 2% B27, 2 mM GlutaMAX, 100 μg/ml streptomycin, and 100 U/ml penicillin (Gibco; Invitrogen).

HEK293T cells (Sigma-Aldrich) were maintained in 6-cm dishes in DMEM media containing 10% fetal bovine serum, 1 mM glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin (“HEK media”) in an incubator with 5% CO₂ at 37°C.

Transfections and nucleofections

All transfections of hippocampal cells were performed using Lipofectamine 2000 (Invitrogen), in triplicate wells on DIV13 in Neurobasal plus medium (A3582901; Thermo Fisher Scientific), with the exception of electrophysiology experiments. 1 μg of total DNA was transfected per well in a 24-well plate, where an empty pcDNA3 vector was used to make up the total amount of DNA. Neurons were transfected in Neurobasal plus media in the absence of streptomycin and penicillin for 2 h, replaced with neuron culture media containing ApV (1:1,000) for 45 min, which was washed out and replaced with conditioning media.

Hippocampal neuron transfections for electrophysiology were performed using Lipofectamine 2000 in six replicate wells on DIV9-12 in Neurobasal-A medium (10888022; Thermo Fisher Scientific). 1 μg of total DNA was transfected per well in a 24-well plate, where an empty pcDNA3 vector was used to make up the total amount of DNA. Before addition of the lipofectamine-DNA mix, cells were equilibrated in warm Neurobasal-A containing ApV (1:1,000) without penicillin and streptomycin for 30 min in 37°C. Transfection incubation time was shortened to 1.5 h and in the presence of ApV. Cells were subsequently washed with Neurobasal-A supplemented with ApV, replaced with conditioning media, and maintained until the day of recording.

Nucleofections were done on cortical neurons using the P3 Primary Cell 4D-Nucleofector X Kit (LZ-V4XP-3024; Lonza), on the day of preparation and dissociation (DIV 0). 4 million dissociated cortical cells were electroporated with 3 μg total DNA per condition using the program DC-104, seeded in six-well plates in DMEM/GlutaMAX supplemented with 5% FBS and incubated for 4 h and then replaced with neuron culture media and incubated at 37°C until harvesting. The following amounts of DNA were used for the relevant nucleofections: 2 μg chimeric miR30a-miR329 and 495 hairpins for miRNA overexpression validation; 2 μg pSUPER and 1 μg GFP for protein quantifications upon Prr7 knockdown; 2 μg HA-Prr7 and 1 μg GFP for validation of Prr7 overexpression.

HEK293T cells were transfected in HEK media supplemented with Hepes (25 mM), at 1.9 million cells seeded in 6-cm dishes per condition, by combining DNA with polyethylenimine (PEI) and Opti-MEM for 15 min and then adding the mixture dropwise onto cells. Cells were incubated at 37°C for 2 d until harvesting. For HA-Prr7 validation, cells were transfected with 2 μg pcDNA, HA-Cav1.2, or HA-Prr7, and 1 μg GFP. For shRNA-resistant mutant validation, cells were transfected with 100 ng HA-Prr7 constructs, 1 μg pSUPER, and 1 μg GFP.

Stimulation

To examine downscaling processes, DIV18 or DIV19 hippocampal cells were stimulated with either picrotoxin (100 μM; Sigma-Aldrich) or equivalent volume (1:500) of ethanol absolute for 48 h.

For investigating CDK5 inhibition, DIV19 hippocampal cells were stimulated with either roscovitine (10 μM, R7772; Sigma-Aldrich) or equivalent volume (1:1,000) of DMSO for 18 h.

For studying SPAR and GluA1 degradation, DIV5 cortical cells were stimulated with either leupeptin (200 μg/ml, L2884; Sigma-Aldrich) or equivalent volume (1:250,000) of water for 20–21 h.

Luciferase reporter assay

DIV13 primary rat hippocampal neurons were transfected in triplicate with 20 pmol pLNAs (10 pmol for 329/495 mix) and 50 ng Prr7 3′ UTR pmiRGLO constructs per well. Cells were treated with PTX or ethanol on DIV18 or DIV19 for 48 h, then lysed in Passive Lysis Buffer (diluted to 1×; Promega) for 15 min, and dual-luciferase assay performed using homemade reagents (as described in Baker and Boyce [2014]) on the GloMax Discover GM3000 (Promega).

pLNAs used were control pLNA (miRCURY LNA miRNA Power Inhibitor Negative control A, #339135 YI00199006-DCA; QIAGEN), miR-329 pLNA (miRCURY LNA miRNA Power Inhibitor RNO-MIR-329-3P, # 339130 YI04101481-DCA; QIAGEN), and miR-495 pLNA (miRCURY LNA miRNA Power Inhibitor HSA-MIR-495-3P, #339130 YI04101229-DCA; QIAGEN).

Bicistronic reporter (dual-color) assay/single-cell fluorescent sensor assay

For single-cell fluorescent assay, DIV13 or DIV14 hippocampal cells were transfected in triplicate wells with 125 ng of control, miR-329, or miR495 bicistronic sensor and 5 pmol control, miR-329, or miR-495 pLNA (where applicable). On DIV19, cells were treated with PTX or ethanol for 48 h, then fixed for 15 min in 4% paraformaldehyde/4% sucrose/PBS, washed in PBS, and directly mounted onto slides for imaging. To determine miRNA activity, dsRED-positive (red or yellow) versus GFP-only (green) cells were manually counted at 20× objective with both 488 and 561 channels open for all coverslips, and the proportion of GFP-only cells over the total count was taken. Approximately 100–200 cells were counted per coverslip (300–600 total per experimental condition).

Immunocytochemistry, spine density, and image analysis

For all imaging experiments, hippocampal cells were transfected on DIV13 in either duplicate or triplicate wells. The following amounts of DNA/RNA were used for the relevant experiments: 150 ng GFP-amp, 20 pmol pLNAs (10 pmol for 329/495 mix), 500 ng miR30a hairpins, 7.5 ng pSUPER constructs (with the exception of spine rescue experiment with pLNAs, for which 2.5 ng was used), 400 ng HA-Prr7 or HA-Prr7^R, 400 ng CREB-VP16m (in active CREB variant for control purposes), or CDK5-DN. Where applicable, the transfected cells were further treated on DIV19 with picrotoxin/ethanol or roscovitine/DMSO.

For all experiments, cells were fixed for 15 min with 4% paraformaldehyde/4% sucrose/PBS and washed with PBS. In cases where cells were only analyzed for spine morphology, coverslips were directly mounted onto microscope slides.

For immunostaining, following fixation, coverslips were transferred to a humidified chamber protected from light. For Prr7 and whole-cell GluA1 immunostaining, cells were permeabilized with 0.1% Triton/PBS for 5 min. Blocking for 30 min in 1×GDB buffer (0.02% gelatin/0.5% Triton X-100/PBS) was followed by overnight incubation with primary antibody in GDB at 4°C. Secondary antibodies in GDB were applied for 45 min. Coverslips were washed with PBS before and after fixation and application of each antibody and briefly in MilliQ before slide mounting. For GluA1 surface staining, cells were treated with primary antibody at 37°C for 3 h. After washing the cells four times with fresh cell media, cells were fixed for 15 min with 4% paraformaldehyde/4% sucrose/PBS and washed with PBS. Coverslips were then transferred to a humidified chamber at room temperature, incubated in secondary antibody in GDB for 1 h, washed with PBS, rinsed briefly with MilliQ water, and mounted onto glass slides for imaging. For PSD-95/Synapsin1 co-staining, 0.1% Triton/10% NGS/PBS solution was added to the coverslips for 15 min and then incubated overnight with primary antibodies diluted in the Triton/NGS/PBS solution at 4°C. Secondary antibodies in Triton/NGS/PBS were applied for 1 h. The following primary antibodies were used: rabbit polyclonal anti-Prr7 (200 ng/ml, PA5-61266; Invitrogen), chicken monoclonal anti-Map2 (1:5,000 dilution, PA1-16751; Thermo Fisher Scientific), rabbit polyclonal anti-SPAR (1:1,500 dilution, kind gift of DT Pak), rabbit polyclonal anti-GluA1 (PC246 Calbiochem EMD Biosciences, or ABN241 Sigma-Aldrich at final concentrations 1 μg/ml for whole-cell staining, 2 μg/ml for surface staining), rabbit anti-Synapsin1 (AB1543, 1:1,000; Merck Millipore), and mouse anti-PSD-95 (810401, 1:200; BioLegend). Alexa-546– and -647–conjugated secondary antibodies (1:2,000 dilution) were used for detection.

All images were acquired with confocal laser-scanning microscope (Zeiss LSM) using a 40×/1.3 oil DIC UV-IR M27 objective. Z-stack images were obtained for 7–11 GFP-positive neurons with pyramidal morphology for each condition. Settings were Opt sampling (1.0× Nyquist), Zoom factor 1.0, Pixel Dwell 0.90 μs, Speed fps 0.23, Scan time 4.32 s, Speed 6, Digital Gain 1.0, and Pinhole 384 μm. Z stacks were kept at 0.45 μm, and 9–11 slices obtained. Laser settings were kept constant between conditions. Images were processed by Airyscan processing at 6.0 strength 3D, and maximum intensity projections of the Z-stacks were used for signal quantification.

Prr7, GluA1, and SPAR puncta intensities, PSD-95/Synapsin co-cluster number and surface GluA1 particle number within cell area were analyzed with a custom-made Python script developed by D Colameo and can be added as a Plugin on Fiji (https://github.com/dcolam/Cluster-Analysis-Plugin). Whole cell, cell body, and dendrites (whole-cell selection with cell body subtracted) were defined using GFP as a mask.

Spine density was measured manually in a blinded manner using Fiji (imaging was not blinded; however, cells were selected based on their pyramidal morphology at 10× objective where spines were not clearly visible). For each cell analyzed, first a primary dendrite and two secondary dendrites branching off it were selected, and the total length of the selected segment obtained. The selected segment was followed to the very end/edge of the frame, yielding a total segment length of ∼150–250 μm, and therefore, the analysis was performed on both proximal and distal dendrites. Using the “cell counter” tool, the total protrusion number along the selected segment was counted (without discrimination of shape and included filopodia-like protrusions), yielding total counts of 100–300 protrusions. The count was divided by the total selected dendrite length to obtain #spines/μm. A second set of dendritic segments were selected, and the process repeated. The average of the two spine density readings was calculated per cell.

Preparation of protein extracts and Western blotting

Protein extracts were prepared by first scraping and lysing cells in RIPA buffer (150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 1 mM EDTA, 1 mM EGTA, 0.05% SDS, 50 mM Tris, pH 8.0, 1× complete protease inhibitor cocktail [Roche]), spinning down the lysate at maximum speed at 4°C for 15 min, and collecting the supernatant. Protein concentration was measured using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein were diluted in Laemmli sample buffer supplemented with BME, boiled at 95°C for 5, min and loaded onto SDS–PAGE gels (10% polyacrylamide for the Prr7 probe, 8% for SPAR). For Prr7 Western, proteins were transferred onto Trans-Blot Turbo 0.2-μm nitrocellulose membranes (Bio-Rad) using the Trans-Blot Turbo semi-dry transfer system (Bio-Rad). For SPAR Western, proteins were blotted onto 0.45-μm PVDF membranes (Immobilon) soaked in transfer buffer (25 mM Tris–HCl, pH 8.3, 192 mM glycine, 20% MeOH) via wet transfer for 15–16 h at 25V. For all experiments, blocking was done in 5% milk in 1xTBS 0.1% Tween20 (TBST) for 1.5–2 h at room temperature, followed by overnight primary antibody incubation at 4°C. After washes in milk, HRP-conjugated secondary antibodies were applied onto the membranes for 1 h at room temperature. Membranes were washed in TBST and visualized with the Clarity Western ECL Substrate (Bio-Rad) on the ChemiDoc Imaging System (Bio-Rad). The following primary antibodies were used: rabbit anti-GluA1 (1:1,000 dilution, PA1 37776; Thermo Fisher Scientific), mouse monoclonal anti-Prr7 (1:250 dilution, MA1-10448; Thermo Fisher Scientific), and rabbit monoclonal anti-α tubulin (1:2,000 dilution, 11H10 lot 112125S; Cell Signaling). HRP-conjugated secondary antibodies rabbit anti-Ms IgG H&L (402335 lot D00160409; Calbiochem) and goat anti-Rb IgG H&L (lot 2625715; Calbiochem) were used at 1:20,000 dilution.

RNA extraction and quantitative real-time PCR

RNA was isolated using the TriFast RNA extraction kit (30-2030; VWR) or RNA-Solv reagent (Omega Bio-tek). Genomic DNA was removed with TURBO DNAse enzyme (Thermo Fisher Scientific). Reverse transcription was performed using either the Taqman MicroRNA Reverse Transcription Kit (Thermo Fisher Scientific) for miRNA detection or the iScript cDNA synthesis kit (Bio-Rad) for mRNA detection. qPCR was performed using either Taqman Universal PCR Master Mix (Thermo Fisher Scientific) for microRNA detection or the iTaq SYBR Green Supermix with ROX (Bio-Rad), and plates were read on the CFX384 Real-Time System (Bio-Rad). Data were analyzed via the ΔΔCt method and normalized to either U6 (for miRNAs) or GAPDH (for mRNAs). mRNA primer information is indicated in supplemental table.

Taqman primers used were (all from Thermo Fisher Scientific): U6 snRNA (Assay ID: 001973), mmu-miR-495 (4427975, Assay ID: 001663), mmu-miR-329 (4427975, Assay ID: 00192), mmu-miR-134 (4427975, Assay ID: 001186), hsa-miR-132 (4427975, Assay ID: 000457), and hsa-miR-99b-5p (4427975, Assay ID: 000436).

Electrophysiology

Whole-cell patch-clamp recordings were performed on an upright microscope (Olympus BX51WI) at room temperature. Data were collected with an Axon MultiClamp 700B amplifier and a Digidata 1550B digitizer and analyzed with pClamp 11 software (all from Molecular Devices). Recording pipettes were pulled from borosilicate capillary glass (GC150F-10; Harvard Apparatus) with a DMZ-Universal-Electrode-Puller (Zeitz) and had resistances between 3 and 4 MW.

Miniature EPSC (mEPSC) were recorded from primary cultured hippocampal neurons on DIV19-21 after transfection in Neurobasal-A medium (10888022; Thermo Fisher Scientific) on DIV9-12. The extracellular solution (ACSF) was composed of (in mM) 140 NaCl, 2.5 KCl, 10 Hepes, 2 CaCl₂, 1 MgCl₂, 10 glucose (adjusted to pH 7.3 with NaOH), the intracellular solution of (in mM) 125 K-gluconate, 20 KCL, 0.5 EGTA, 10 Hepes, 4 Mg-ATP, 0.3 GTP, and 10 Na₂-phosphocreatine (adjusted to pH 7.3 with KOH). For mEPSCs, 1 μM TTX and 1 μM Gabazine were added to the extracellular solution to block action-potential driven glutamate release and GABAergic synaptic transmission, respectively. Cells were held at −70 mV. The sampling frequency was 5 kHz, and the filter frequency 2 kHz. Series resistance was monitored, and recordings were discarded if the series resistance changed significantly (≥10%) or exceeded 20 MΩ. Paired pulse ratio of EPSCs were recorded at a holding potential of −60 mV, sampled at 100 kHz, and filtered at 4 kHz. 1 μM Gabazine was added to the extracellular solution to block GABAergic synaptic transmission. 100 nM NBQX was added to prevent epileptiform activity and to minimize polysynaptic activity. Synaptic currents were evoked by monopolar stimulation with a patch pipette filled with ACSF.

Statistics

With the exception of the miRNA sensor assay, statistical tests were performed using GraphPad Prism version 9.2.0. For all datasets, three to four independent experiments were performed. Given the small sample size, normality was assumed for all datasets, and therefore, one or two sample unpaired t test (two-sided) or one-way or two-way ANOVA followed by a post hoc Tukey test was performed. *P < 0.05; **P < 0.01; ***P < 0.001.

For the miRNA sensor assay, a binomial generalized mixed effects model (GLMM) was applied (per sensor) using the lme4 R package (Bates et al, 2015) to test the proportion data (ratio of miRNA positive cell counts over all cells). Total counts were used as prior weights. The preparation (e.g., the batch) was accounted for using a random effect. Because of the replicates within the batches, an interaction term between the (random) batches and the (fixed) treatments was included in the model. Post hoc tests were conducted using the glht function of the multcomp package (Hothorn et al, 2008) for the comparisons of interest. P-values were adjusted by Bonferroni’s method.