2.5. Molecular analyses of mouse brain tissue

Hippocampal tissue from dorsal hippocampus of mice with virus-mediated overexpression of mCherry, NOS1AP, and NOS1AP_396-503 in dorsal hippocampus (5 mice/group) was dissected 4-5 weeks after injection and stored in RNA preserving solution (25 mM sodium citrate, 10 mM EDTA, 5·3 M ammonium sulphate, pH 5·2) at 4 °C for one week. Hippocampi from both hemispheres were isolated on a cooling plate at 4 °C and then stored at -20 °C. RNA was isolated separately from all hippocampi (i.e. 10 hippocampi from 5 mice/group) using the MagJET RNA Kit (ThermoFisher Scientific) according to the manufacturer's instructions using a pipetting robot (Biomek NX^P, Beckman Coulter). A total of 375 ng RNA per sample was reverse transcribed using the iScript™ cDNA Synthesis Kit (Bio-Rad), using both random hexamers and oligo(dT) primers (note that one NOS1AP_396-503 sample could not be reverse transcribed due to too low RNA yield).

Target-specific quantitative PCR (qPCR) for Actb, B2m, Hprt, and Sdha (as reference genes) and for 17 NOS1AP-associated genes (i.e. genes of gene products directly interacting with NOS1AP or functionally dependent on NOS1AP interactions) Cpe, Dlg1, Dlg3, Dlg4, Gria1, Gria2, Grin2a, Grin2b, Gucy1a1, Gucy1a2, Gucy1b1, Gucy1b2, Map2k3, Mapk14, Nos1, Nos1ap, Rasd1, Scrib, and Syn1 (as target genes) was performed in 10 µl reactions containing 5 µl AMPLIFYME SG No-ROX Mix (Blirt), target specific forward and reverse primers (final concentration: 0·3 µM each; see Table S2 for primer sequences), and cDNA (final dilution: 1:200) using a 384-well plate (primaPLATE, Steinbrenner Laborsysteme GmbH) with LightCycler® 480 Sealing Foil (Roche) on the LightCycler® 480 Instrument II (Roche). All reactions were run in duplicates. PCR conditions: 3 min at 95 °C, followed by 45 cycles of 5 s at 95 °C, 10 s at 60 °C, and 10 s at 72 °C followed by a plate read. At the end of amplification, a melting curve was generated. Crossing point (Cp) values were calculated by the LightCycler 480 software (release 1.5.1.62) using the Second Derivative Maximum method. Relative gene expression levels were analysed using GenEx6 v3.1.3 (MultiD Analyses AB). One of the NOS1AP_396-503 samples was removed from further analysis as no reference gene data were available. For every target, a standard curve was created on the same plate as the target samples and Cp values were corrected for efficiencies calculated from these standard curves (see Table S2). Missing data points (Hprt [1 from mCherry and 2 from NOS1AP], Dlg3 [1 from mCherry, 1 from NOS1AP], Scrib [1 from NOS1AP], Gucy1a2 [1 from mCherry]) were imputed. Robustness of imputed data was confirmed by complete case analysis (i.e. reanalysis with imputed values removed; data not shown).

Expression data were calculated relative to the expression of Sdha which was selected as the most stable of the four reference genes by analysis with Normfinder. Data were calculated relative to the average of the mCherry group and converted to log2.

Hippocampal tissue from dorsal hippocampus of uninjected mice, and from mice with virus-mediated overexpression of mCherry, NOS1AP, and NOS1AP_396-503 (3 mice/group, 1 hippocampus/mouse) in dorsal hippocampus was dissected 4-5 weeks after injection and snap-frozen in isopentane chilled with dry ice. Frozen tissue was weighed, and homogenized while thawing in 30 µl/mg low-salt buffer (LSB [52]) supplemented with protease inhibitors, 1 mM DTT and nonionic detergent (0·5% Igepal CA-630) using 10 strokes of a Dounce homogenizer, and precleared at 20,000 x g/4 °C for 10 min. Protein content of each homogenate was quantified with the Bio-Rad DC protein assay kit (Bio-Rad) and equal amounts from each sample were used in parallel in the subsequent steps.

Co-immunoprecipitation (Co-IP) was performed as previously described [33]. Briefly, immunoprecipitating (IP) antibody, nNOS antibody (mouse monoclonal clone A-11, RRID:AB_626757, Santa Cruz Biotechnology, 2·5 μg/ml) was added to each lysate. Lysates were rotated for 2 h/4 °C, 5 μl of protein-A resin (GenScript) was added, and rotation continued for 1 h. The resin was washed three times with LSB, and protein was eluted from drained resin by incubation at 95 °C for 10 min in SDS-PAGE sample loading buffer and analysed by Western blotting. Immunoprecipitated nNOS, and co-immunoprecipitated NOS1AP and PSD-95 in all samples were determined by Western blotting with anti-nNOS antibody (rabbit polyclonal IgG, RRID:AB_2313734, Invitrogen), anti-NOS1AP antibody (rabbit polyclonal IgG, R-300, RRID:AB_2251417, Santa Cruz Biotechnology) and anti-PSD-95 antibody (mouse monoclonal IgG2a, clone K28/43, RRID:AB_2307331,UC Davis/NIH NeuroMab Facility) respectively. Secondary antibodies conjugated to horseradish peroxidase (Santa Cruz Biotechnology) were used and detected with enhanced chemiluminescence reagent (Thermo Scientific).

Immunoblots of co-IP and input protein were quantified using ImageJ. For analysis of the endogenous NOS1AP levels in the input samples, differences in intensities of the bands per lane required a more precise quantification as follows: Data from scanned exposures of blots trimmed to remove the bulk of excess signal from over-expressed protein (Fig. S2E) were collected from equal width regions of interest in ImageJ and imported to Excel. The data for each lane was fitted to Gaussians (one per band present in the analysed lane data) using the Excel solver add-in.