Fig 1: Mint2 N723S mutation decreases Nrxn1α membrane trafficking in HEK293T cells. (a) Structure of human MINT2 and locations of MINT2 ASD variants. (b) Representative immunoblots of protein levels from total cell lysates of HEK293T cells transfected with Mint2 WT or Mint2 ASD variants. Tubulin served as loading control on the same blot. The representative images were cropped and shown in this figure. Whole gel images and quantification are presented in Supplementary Fig. S1a,b, respectively. (c) Representative immunoblots of protein levels from total cell lysates of HEK293T cells co-transfected with Nrxn1α and either GFP (control), GFP-tagged Mint2 WT, N723S or ΔPDZ. GAPDH served as loading control on the same blot. The representative images were cropped and shown in this figure. Whole gel images are presented in Supplementary Fig. S1c. Quantification of total Nrxn1α expressed as percent control and is based on three independent experiments. (d) Western blot analysis and quantification of membrane Nrxn1α levels expressed as percent control from HEK293T cells co-transfected with Nrxn1α and either GFP (control), GFP-tagged Mint2 WT, N723S or ΔPDZ and subjected to subcellular fractionation. GAPDH and Na+/K+ ATPase were used as cytosolic and membrane markers, respectively. The representative images from n = 3 independent experiments were cropped and shown in this figure. Whole gel images are presented in Supplementary Fig. S1d. (e) Representative images of HEK293T cells co-transfected with Flag-Nrxn1α together with GFP (control), GFP-tagged Mint2 WT, N723S or ΔPDZ. Quantification of cell surface Nrxn1α expressed as percent control. GFP, n = 17; WT, n = 31; N723S, n = 33; ΔPDZ, n = 22 total number of cells analyzed from at least 2 coverslips per condition with n = 2 independent experiments/cultures. (f) Western blot analysis and quantification of Mint2 bound to GST-Nrxn1α fusion protein expressed as percent control (Mint2 WT) from HEK293T cells transfected with Mint2 WT, N723S or ΔPDZ. Coomassie stained gel of GST-tagged Nrxn1α protein. The representative images from n = 3 independent experiments were cropped and shown in this figure. Whole gel images are presented in Supplementary Fig. S1e. (g) The second PDZ domain of Mint2 WT (PDB_ID: 3SUZ) consists of six β-strands and two α-helices (left). Mint2 PDZ2 is highly dynamic showing a relative rigid-body motion around a fixed α2 (middle). Mint2 N723S mutation does not change the conformation of α2 and the Nrxn binding pocket between α2 and β2 (right). Data represents as means ± SEM and statistical significance was evaluated using one-way ANOVA. All determinations, *p < 0.05; ***p < 0.0005 (see text). Scale bar = 10 μm.
Fig 2: Mint2 N723S decreases spontaneous synaptic activity at excitatory synapses. (a) Sample traces showing mEPSC of control, Mint KO or Mint KO rescue with Mint2 WT, N723S, or ΔPDZ mutants at 14–17 DIV. Bar graph of mEPSC frequency expressed as percent control revealed a decrease in miniature frequency in Mint KO neurons compared with control that was rescued by Mint2 WT but not N723S or ΔPDZ mutants. Bar graph of mEPSC amplitude showed no change. Control, n = 9; KD, n = 9; WT, n = 15; N723S, n = 13; ΔPDZ, n = 12 neurons analyzed. “n” represents multiple cells from 4–8 coverslips per condition and analyzed from three independent experiments/cultures. (b) Sample traces showing mEPSC of Mint2 WT or N723S mutant in hippocampal WT neurons. Bar graphs expressed as percent control showed that the N723S mutant decreases the frequency but not the amplitude of mEPSC compared to the WT. WT, n = 7; N723S, n = 5 neurons analyzed. “n” represents multiple cells from 4–6 coverslips per condition and analyzed from one independent experiment/culture. (c) Ultrastructural analysis by electron microscopy of synaptic structure in hippocampal neurons infected with control or Mint KO neurons rescued with Mint2 +/N723S at 14 DIV. Representative images of asymmetric excitatory synapses. Scale bar, 100 nm. (d) HEK293T cells were co-transfected with Flag-Mint2 WT and GFP-Mint2 WT, GFP-Mint2 N723S or GFP-Mint2 ΔPDZ. Cell lysates were collected 48 hour post-transfection and immunoprecipitated (IP) with Flag antibody and immunoblotted for GFP. The results were quantified as relative to the immunoprecipitated level of Mint2 WT (lane 4). The representative images from n = 3 independent experiments were cropped and shown in this figure. Whole gel images are presented in Supplementary Fig. S2. Data represents as means ± SEM and statistical significance was evaluated using one-way ANOVA for (a,d) and unpaired Student’s t-test for (b). All determinations, *p < 0.05 (see text).
Fig 3: Mint2 N723S decreases Nrxn-mediated synapse formation. (a) Illustration shows COS7 cells (green) expressing mVenus-NL1 were seeded over dissociated primary mouse hippocampal neurons (black) at 10 DIV and co-cultured for 2 days before analysis. Synaptic contact between neuronal axon and COS7 cells are enlarged which shows a synaptic complex of Mint2-Nrxn1α (presynaptic) with NL1. Sites of cell contact with neuronal axons are analyzed for the recruitment of presynaptic marker, synapsin. The images on the right depict representative immunofluorescence image of COS7 cell expressing mVenus-NL1 in the mixed culture where green represents mVenus, red is synapsin and yellow is merge indicating co-localization. Scale bar = 10 μm. (b) COS cells expressing mVenus-NL1 were co-cultured with control or Mint KO neurons where the latter were additionally rescued with lentiviruses expressing Mint2 WT, N723S, ΔPDZ or heterozygous +/N723S. Scale bar = 10 μm. (c) Quantification of synapse formation expressed as percent control was determined by the co-localization (yellow in the merged images) of the presynaptic synapsin (red) with postsynaptic mVenus- Nlgn1 (green) expression. Control, n = 31; KO, n = 30; WT, n = 26; N723S, n = 34; ΔPDZ, n = 35; +/N723S, n = 29 total number of cells analyzed. “n” represents multiple cells from at least 2 coverslips per condition and analyzed from three independent experiments/cultures. Data represents as means ± SEM and statistical significance was evaluated using one-way ANOVA. All determinations, *p < 0.05; ***p < 0.0005 (see text).
Fig 4: Mint2 N723S leads to an increased number of immobile puncta in the Golgi apparatus and neuronal processes in primary mouse neurons. (a) Mint2 localizes to the Golgi in primary mouse hippocampal neurons. Representative images of hippocampal neurons infected with lentiviral GFP-Mint2 WT, GFP-Mint2 N723S or GFP-Mint2 ΔPDZ and immunostained with GM130 (a Golgi maker) at 14 DIV. Both Mint2 WT and N723S co-localized with GM130 but not Mint2 ΔPDZ (represented as yellow in merged image). N = 12–15 cells from at least three coverslips that was analyzed for each group from three independent experiments/cultures. (b) Time-lapse FRAP imaging of GFP-tagged Mint2 WT and N723S at the Golgi. Regions encompassing the Golgi were photobleached using 3 passes (15 s total) of 100% laser power, which reduced the initial fluorescence intensity to 25%. Recovery fluorescence was acquired with 1% laser power and imaged every 5 s for 150 s. (c) Normalized intensity of GFP-tagged Mint2 WT and N723S signal during FRAP. Horizontal red bar indicates the period of photobleaching on graph. WT, n = 8 and N723S, n = 10 total number of cells assessed. (d) Percentage of Mint2 that is mobile in the Golgi, as measured by percentage of fluorescence that recovers after photobleaching at 170 s. (e) Trafficking and localization of Mint2 to neuronal processes. Top, Representative images of GFP-tagged Mint2 WT and N723S puncta in segments of processes. Bottom, corresponding kymographs for the segments. Diagonal bands indicate mobile Mint2 puncta (arrowheads) and vertical bands indicate immobile Mint2 puncta (arrow). Dotted white circles indicate puncta. (f) Bar graph of puncta speed showed no change. (g) Mint2 N723S has more immobile puncta compared to the WT. WT, n = 19 and N723S, n = 20 total number of processes assessed. Data represents as means ± SEM and statistical significance was evaluated using unpaired Student’s t-test. All determinations, *p < 0.05 (see text). Scale bar = 5 μm.
Fig 5: Mint2 N723S affects axonal localization of Nrxn1α to synapses. (a) Representative images of axonal segments of hippocampal neurons co-transfected with mCherry-Nrxn1α and GFP-tagged Mint2 WT or N723S at 4–6 DIV and analyzed at 7–10 DIV. (b) Quantification of co-localization (represented by arrowheads and yellow in the merged images in a) expressed as percent control between Nrxn1α and GFP (control), WT or N723S. GFP; n = 14, WT, n = 12; or Mint2 N723S, n = 10 total of axons from at least two coverslips that was analyzed from three independent experiments/cultures. (c) Representative images of hippocampal neurons co-transfected with mCherry-Nrxn1α and GFP-tagged Mint2 WT or N723S at 4–6 DIV and immunostained for presynaptic anti-synaptobrevin - syb (cyan) marker at 12–14 DIV. (d) Quantification of co-localization (represented by arrowheads and white in merged images in c) expressed as percent control between Nrxn1α and Syb. GFP; n = 23, WT, n = 16; or N723S, n = 21 total of processes analyzed. For 3b and 3d, “n” represents single axons from at least two coverslips that was analyzed from three independent experiments/cultures. Data represents as means ± SEM and statistical significance was evaluated using unpaired Student’s t-test. All determinations, **p < 0.005; ***p < 0.0005 (see text). Scale bar = 20 μm.
Supplier Page from MilliporeSigma for Anti-Mint2 antibody produced in rabbit