Syngap1 haploinsufficiency in Nkx2.1+ interneurons is associated with reduced sEPSC amplitude and mEPSC frequency in LIV BCs.

(a) Anatomical reconstructions of PV+ cells filled with biocytin in control (left) and cHet mice (right) during whole-cell patch-clamp recordings and post hoc immunohistochemical validation of BC interneurons confirming the positivity for PV. (b) Representative traces of sEPSCs recorded in BCs cells from control Tg(Nkx2.1-Cre):RCEf/f:Syngap1+/+ (blue, n=14 cells, 7 mice) and cHet Tg(Nkx2.1-Cre):RCEf/f:Syngapf/+(red, n=11 cells, 6 mice) mice. (c) Cumulative probability plots show a significant decrease in the amplitude of sEPSC in cHet mice compared to control mice (hash sign denote the significance for LMM related to the cumulative distributions ,# p=0.029) and no change in the inter-sEPSC interval (LMM, p=0.345). Insets illustrate significant differences in the sEPSC amplitude for inter-cell mean comparison (LMM, *p=0.014) and no difference for inter-sEPSC interval (LMM, p=0.230). (d) Representative examples of individual sEPSC events (100 pale sweeps) and average traces (bold trace) detected in BCs of control and cHet mice and superimposed scaled traces (right top), (e) Summary bar graphs showing no differences for inter-cell mean charge transfer (LMM, p=0.090, left) and for the charge transfer when frequency of events is considered (LMM, p=0.140). (f) Representative traces of mEPSCs recorded from the same neurons shown in b-e. (g) Cumulative probability plots show no change in the amplitude of mEPSC (LMM, p=0.151) and a significant increase in the inter-mEPSC interval in cHet mice compared to control mice (LMM, #p=0.045). Insets illustrate summary data showing no significant differences in the amplitude (LMM, p=0.155) and a trend towards longer inter-mEPSCs intervals for inter-cell mean comparison in cHet compared to control mice (LMM, p=0.056). (h) Representative examples of individual mEPSC events (100 pale sweeps) and average traces (bold trace) detected in BCs of control and cHet mice. (i) Summary bar graphs for a group of cells show no significant differences in the quantal content (LMM, p=0.189) and mEPSCs kinetics (LMM, p=0.269 for rise time, and p=0.193 for decay time). (j) Summary bar graphs showing a significant decrease in cHet mice for inter-cell mean charge transfer (LMM, *p=0.012, left) and for the charge transfer when frequency of events is considered (LMM, **p=0.002). * and # indicates p value <0.05 for bar graphs and cumulative distribution, respectively.

sEPSCs and mEPSCs in LIV PV+ cells from control Vs cHet mice. Related to Fig. 1

Thalamocortical eAMPA transmission is decreased in LIV PV+ cells from Tg(Nkx2.1-Cre):RCEf/f:Syngap1f/+mice.

(a) Representative examples of individual eAMPA (negative deflections) and eNMDA (positive deflections) (5-10 pale sweep) and average traces (bold trace) recorded in PV cells from control (blue, n=16 cells, 7 mice) and cHet mice (red, n=14 cells, 7 mice). (b, left) Summary plots showing no change in the failure rate of eAMPA (left, LMM, p=0.550) and a significant decrease in the minimal (including failures and successes) eAMPA amplitude (b, right LMM, *p=0.031) and (c, left) charge transfer (LMM, *p=0.033) in cHet mice. (c, right) Summary bar graph illustrating the percentage of PV+ cells containing eNMDA in the thalamocortical evoked EPSC. (d) Synaptic latency histograms (bottom) of thalamocortical eEPSC from control and cHet mice, and summary bar graph (top) illustrating an increase in the onset latencies of eEPSC in cHet mice (LMM, *p=0.023). For both histograms, bins are 0.1 msec wide. (e) Summary plots showing a significant decrease in the potency of eAMPA (successes only, LMM, **p=0.003, left) in cHet mice with no change in eNMDA (LMM, p=0.969), A significant increase is present in the NMDA/AMPA ratio (i.e. ratio of the peak for eNMDA and eAMPA, LMM, **p=0.001, right) in cHet mice. (f) Summary plots showing no significant differences in the eAMPA (LMM, p=0.177 for rise time, and p=0.608 for decay time) and eNMDA kinetics (LMM, p=0.228 for rise time, and p=0.221 for decay time). (g) Representative examples of individual LIV evoked EPSC (10 pale sweeps) and average traces (bold trace) with an interval of 50 ms recorded in two BC cells from control (blue) and cHet mice (red). (h) Summary plot showing significantly increased PPRs recorded from LIV BC cHet (red circles, n=8 cells; 5 mice) compared to controls (blue circles, n=9 cells, 5 mice), when two EPSCs were evoked in layer IV BC with two electric pulses at 30 or 50ms intervals (Two-way Repeated Measure ANOVA with Sidak’s multiple comparison post hoc test, **p=0.001). * indicates p value <0.05; ** indicate p value <0.005.

eAMPA and eNMDA currents in LIV PV+ cells from control Vs cHet mice. Related to Fig. 2

The amplitude of sIPSCs, but not mIPSCs, in LIV PV+ cells is reduced in Tg(Nkx2.1-Cre):RCEf/f:Syngap1f/+mice

(a) Representative traces of sIPSCs recorded in PV+ cells from control (blue, n=25 cells, 8 mice) and cHet (red, n=24 cells, 7 mice) mice. (b) Cumulative probability plots show a significant decrease in the amplitude of sIPSC in cHet mice compared to control (LMM, ## p=0.003) and no change in the inter-sIPSC interval (LMM, p=0.106). Insets illustrate significant differences in the sIPSC amplitude for inter-cell mean comparison (LMM, *p=0.009) and no difference for inter-sIPSC interval (LMM, p=0.185). (c) Representative examples of individual sIPSC events (100 pale sweeps) and average traces (bold trace) detected in PV cells of control and cHet mice. (d) Summary bar graphs for a group of cells show no significant differences in the sIPSCs kinetics (LMM, p=0.113 for rise time, and p=0.602 for decay time). (e) Summary bar graphs showing no differences for inter-cell mean charge transfer (LMM, p=0.234, left) and for the charge transfer when frequency of events is considered (LMM, p=0.273). (f) Representative traces of mIPSCs recorded in PV+ cells from control (blue, n=25 cells, 8 mice) and cHet (red, n=24 cells, 7 mice) mice. (g) Cumulative probability plots show no change in the amplitude of mIPSC (LMM, p=0.118) and in the inter-mIPSC interval (LMM, p=0.411). Insets illustrate summary data showing no significant differences in the amplitude (LMM, p=0.195) and the inter-mIPSCs interval for inter- cell mean comparison (LMM, p=0.243). (h) Representative examples of individual mIPSC events (100 pale sweeps) and average traces (bold trace) detected in PV+ cells of control and cHet mice. (i) Summary bar graphs for a group of cells show no significant differences in the mIPSCs kinetics (LMM, p=0.103 for rise time, and p=0.597 for decay time). (j) Summary bar graphs showing no differences for inter-cell mean charge transfer (LMM, p=0.374, left) and for the charge transfer when frequency of events is considered (LMM, p=0.100). * indicates p value <0.05 for bar graphs. # # indicates p value <0.005 for cumulative distribution.

sIPSCs and mIPSCs in LIV PV+ cells from control Vs cHet mice. Related to Fig. 3

PV+ cells intrinsic excitability is decreased in Tg(Nkx2.1-Cre):RCEf/f:Syngap1f/+ mice.

(a) Summary data showing no changes in the passive membrane properties between control (blue, n=33 cells, 15 mice) and cHet mice (red, n=40 cells, 17 mice) (LMM, p=0.081 for Vm, p=0.188 for Rin, p=0.188 for Cm, p=0.199 for τ). (b) Summary data showing no differences in AP half-width (LMM, p=0.111) but a significant decrease in AP amplitude (LMM, *p=0.032) and a significant increase in AP latency (LMM, *p=0.009) from PV+ cells recorded in cHet mice. (c, left) Summary bar graph shows a significant increase in AP threshold from cHet mice (LMM, ***p<0.001) for the first AP generated. (c, right top) Representative single APs evoked by threshold currents from control and cHet mice. APs are aligned at 50% of the rising phase on X axis and peak on Y axis. Note the more hyperpolarized AP with consequent reduction in AP amplitude in PV+ cells from cHet mice. (d) Summary bar graph shows a significant increase in the threshold current (LMM, **p=0.004). (e, left) Summary plot showing a reduction of averaging number of APs per current step (40 pA) amplitude recorded from LIV PV+ cHet (red circles, n=38 cells; 18 mice) compared to control (blue circles, n=30 cells, 15 mice) neurons (Two-way Repeated Measure ANOVA with Sidak’s multiple comparison post hoc test, ****p<0.0001). (e, right) Representative voltage responses indicating the typical FS firing pattern of PV+ cells in control and cHet mice in response to depolarizing (+120 pA and +240 pA) current injections corresponding to threshold current and 2x threshold current. * indicates p value <0.05; *** indicates p value <0.001; **** indicates value p<0.0001.

Membrane properties of total LIV PV+ cells population in control Vs cHet mice. Related to Fig. 4

Syngap1 haploinsufficiency in Nkx2.1+ interneurons affects the denritic arbour of a specific subpopulation of LIV PV+ cells.

(a, left) Strong negative correlation of Fmaxinitial with AP half-width in PV+ cells from control (blue, n=33 cells, 15 mice) and cHet mice (red, n=40 cells, 17 mice). (a, right) Hierarchical clustering based on Euclidean distance of PV+ cells from control mice. Clustering is based on AP half-width and Max frequency. Asterisks indicate cells with longer AP half-width felling into the cluster including PV+ cells with higher of values of Fmaxinitial. (b, left) Correlation of parameters describing membrane properties of PV+ interneurons. The 13 passive and active membrane properties used for PCA analysis (derived from 27 PV+ cells from control mice; see materials and methods) are arrayed against each other in a correlation matrix with the degree of correlation indicated by the shading: white is negatively correlated (correlation index of 0) ,black is positively correlated (correlation index of 1, diagonal squares) and light gray not correlated (correlation index of 0). PCA on the 13 parameters to reduce the dimensionality. (b, right) The first (PC1) and second (PC2) PC values derived for each interneuron are plotted against each other. No clear separation of subgroups in scatterplot of first 2 PCs is present when genotype is taken into consideration. (c) Cumulative histograms of AP half-widths in control (n=33 cells, 15 mice) and cHet mice (n=40 cells, 17 mice) fitted with two Gaussian curves. Vertical line indicates the cutoff value at intersection between the two curves. For both histograms, bins are 0.05 msec wide. (d) PCA analysis using the cutoff value of 0.78 ms and the 13 passive and active membrane properties distinguish two subgroups of PV+ cells with short (black circles) and broad (turquoise circles) AP-half width duration in both genotypes. Insets illustrate pie charts describing the % of two subgroups of PV+ cells in the control and cHet mice. (e) Anatomical reconstructions of a BC- short and (f) a BC-broad filled with biocytin in control mice during whole-cell patch-clamp recordings and post hoc immunohistochemical validation for PV. (g) Summary data in control mice (gray, BC-short n=5 cells, 4 mice; turquoise, BC-broad n=5 cells, 4 mice) showing no significant difference in terms of distance from pia (p=0.856, LMM) for both subtypes of PV+ cell analyzed indicating LIV location and significant differences in dendritic parameters between the two subpopulations of PV+ cells (LMM , *p= 0.016 for dendr. surface area, *p=0.043 for # branching points) and no change in total dendritic length (LMM , p=0.057). (h) Summary data in cHet mice (gray, BC-short n=6 cells, 4 mice; turquoise, BC-broad n=6 cells, 3 mice) showing no significant difference in terms of distance from pia (LMM, p=0.594) for both subtypes of PV+ cell and all dendritic parameters (LMM, p= 0.062 for total dendritic length, p=0.731 for dendr. surface area, p=0.081 for # branching points). (i) Summary data showing a significant increase in dendritic complexity between control (gray, n=5 cells, 4 mice) and cHet (white, n=6 cells, 4 mice) for the subpopulation of BC-short (LMM , *p=0.009 for dendr. surface area, *p=0.048 for # branching points) and no difference for the total dendritic length (LMM, p=0.070). (j) Summary data showing preserved dendritic parameters in cHet (turquoise filled with pattern, n=6 cells, 3 mice) Vs control (turquoise, n=5 cells, 4 mice) (LMM, p= 0.967 for total dendritic length, p=0.784 for dendr. surface area, p=0.290 for # branching points). (k) The strong positive correlation of dendritic surface area with AP half-width is present only in PV+ cells from control mice (blue, n=10 cells, 8 mice) and disappears in cHet mice (red, 12 cells, 7 mice). * indicates p value <0.05.

Morphological properties of BC-short Vs BC-broad in control mice. Related to Fig. 5.

Morphological properties of BC-short Vs BC-broad in cHet mice. Related to Fig. 5.

Morphological properties of BC-short in control Vs cHet mice. Related to Fig. 5.

Morphological properties of BC-broad in control Vs cHet mice. Related to Fig. 5.

Intrinsic excitability is decreased in both subpopulations of PV+ cells in cHet mice

(a) Summary data showing no changes in the passive membrane properties of BC-short between control (blue, n=12 cells, 9 mice) and cHet mice (red, n=24 cells, 13 mice) (LMM , p=0.189 for Vm, p=0.856 for Rin, p=0.188 for Cm, p=0.077 for τ) (b) Summary data showing no differences in AP half-width (p=0.386, LMM) and AP latency (LMM, p=0.210) but a significant decrease in AP amplitude (LMM, *p=0.024) of BC-short recorded in cHet mice. (c, left) Summary bar graph shows a significant increase in AP threshold from cHet mice (LMM, **p=0.002). (c, right) Representative single APs evoked by threshold currents from control and cHet mice. (d) Summary bar graph shows a significant increase in the threshold current (LMM, *p=0.015). (e, left) Summary plot showing a reduction of averaging number of APs per current step (40 pA) amplitude recorded from LIV BC- short in cHet (red circles, n = 22 cells; 13 mice) compared to control (blue circles, n = 11 cells, 9 mice) neurons (Two-way Repeated Measure ANOVA with Sidak’s multiple comparison post hoc test, **** p<0.0001). (e, right) Representative voltage responses indicating the typical FS firing pattern of BC-short in control and cHet mice in response to depolarizing (+120 pA and +240 pA) current injections corresponding to threshold current and 2x threshold current. (f) Summary data showing a significant decrease in RMP (LMM, *p=0.023) of BC-broad from cHet mice (red, n=16 cells, 11 mice) but no changes in the other passive membrane properties compared to control mice (blue, n=21 cells, 12 mice) (LMM, p=0.244 for Rin, p=0.170 for Cm, p=0.639 for τ). (g) Summary data showing no differences in AP half-width (LMM, p=0.593) and AP amplitude (LMM, p=0.713) and a significant increase in AP latency (LMM, *p=0.035) from BC-broad cells recorded in cHet mice. (h, left) Summary bar graph shows a significant increase in AP threshold from cHet mice (LMM, *p=0.010). (h, right) Representative single APs evoked by threshold currents from control and cHet mice. (i) Summary bar graph shows no difference in the threshold current (LMM, p=0.402). (j, left) Summary plot showing no difference in the averaging number of APs per current step (40 pA) amplitude recorded from LIV BC-broad in cHet (red circles, n=16 cells, 11 mice) compared to control (blue circles, n=18 cells; 11 mice) neurons (Two-way Repeated Measure ANOVA with Sidak’s multiple comparison post hoc test, p= 0.333). (j, right) Representative voltage responses indicating the typical FS firing pattern of BC broad in control and cHet mice in response to depolarizing (+120 pA and +240 pA) current injections corresponding to threshold current and 2x threshold current. * indicates p value <0.05; ** indicates p value <0.005.

Membrane properties of BC-short in control Vs cHet mice. Related to Fig. 6.

Membrane properties of BC-broad in control Vs cHet mice. Related to Fig. 6.

Evoked firing properties are reduced in SST+ cells from mice with embryonic-onset Syngap1 haploinsufficiency in Nkx2.1 interneurons.

(a) Representative voltage responses indicating the typical regular adapting firing pattern of SST+ in control mice in response to hyperpolarizing (-40 pA) and depolarizing (+80 pA and +160 pA) current injections corresponding to Ih associated voltage rectification, threshold current and 2x threshold current respectively. (b) Post hoc immunohistochemical validation of these interneurons confirming their positivity for SST+ and negativity for PV-. (c) PCA using the 13 parameters previously described clearly separate the cluster of SST+ cells (pink circles) from BC-short (black circles) having however some overlaps with BC-broad (turquoise circles) in control mice. (d) Summary data showing no changes in the passive (LMM, p=0.283 for Vm, p=0.959 for Rin, p=0.484 for Cm, p=0.501 for τ) and (e) active membrane properties (LMM, p=0.332 for AP half-width, p=0.126 for AP amplitude, p=0.296 for AP latency, p=0.154 for AP threshold) between SST+ cells from control (blue, n=11 cells, 8 mice) and cHet mice (red, n=16 cells, 9 mice) (f) Summary plot showing a reduction of averaging number of APs per current step (40 pA) amplitude recorded from LIV SST+ in cHet (red circles, n= 16 cells, 9 mice) compared to control (blue circles, n=11 cells, 8 mice) neurons (Two-way Repeated Measure ANOVA with Sidak’s multiple comparison post hoc test, ***p<0.001). * indicates p value <0.05; ** indicates p value <0.005.

Membrane properties of total SST+ cells population in control Vs cHet mice. Related to Fig. 7

Syngap1 haploinsufficiency alters the intrinsic excitability of LIV PV+ cells by affecting voltage-gated D-type K+ currents.

(a, left) Summary bar graph shows a significant decrease in AP half-width in PV+ cells from cHet (red) vs control (blue) mice (LMM, *p=0.034), which persist when cHet PV+ cells are treated with α-DTX (red with diagonal stripes, LMM, *p=0.039). (a, right) Summary bar graph shows a significant increase in AP threshold of PV+ cells from vehicle-treated cHet mice (red) compared to vehicle-treated control mice (blue, LMM, *p=0.049) and the rescue of this deficit in presence of α-DTX (blue vs red with diagonal stripes, LMM, p=0.940). (b) Delta (Δ) value was calculated for AP threshold by subtracting individual values of α-DTX-treated cells from the average of their respective control group. A significant increase in AP threshold Δ number was found for cHet α-DTX-treated PV+ cells compared to control α-DTX-treated PV+ cells (LMM, *p=0.015). (c) Representative single APs evoked by threshold currents from vehicle-treated control (blue) and cHet (red) mice (center), and control (blue dotted line) and cHet α-DTX-treated (red dotted line) PV+ cells. (d) Summary plot showing no difference in the averaging number of APs per current step (40 pA) amplitude recorded from LIV PV+ in cHet and control, both α-DTX-treated and vehicle-treated, PV+ cells (Two-way Repeated Measure ANOVA with Sidak’s multiple comparison post hoc test, p>0.05). (e, left) Summary bar graph shows a significant difference in AP latency Δ number in α-DTX-treated cHet vs α-DTX-treated control PV+ cells (LMM, *p=0.006). (e, right) Representative voltage traces clearly show a reduction in the AP onset for cHet PV+ cells treated with α-DTX (pink trace) compared to vehicle-treated cHet PV+ cells (red trace), while control PV+ cells are not affected (vehicle treated-control PV+cells, blue traces; control α-DTX PV+ cells, light blue traces). Control mice: vehicle treated, n=9 cells from 4 mice; α-DTX-treated, n=11 cells from 6 mice; cHet mice: vehicle treated, n=23 cells from 10 mice; cHet α-DTX-treated, n=18 cells from 8 mice. * indicates p value <0.05.

Membrane properties of PV+ cells with and without α-DTX treatment in control and cHet conditions. Related to Fig. 8

Anatomical and neurochemical identification of LIV PV+ and SST+ interneurons in mouse primary auditory cortex.

(a, top) Representative high-magnification confocal images of two biocytin-filled PV+ cells with somata in LIV, showing multipolar dendritic arborization and smooth dendrites (small panels) and axons spreading from LII to LV. (a, bottom) Posthoc immunohistochemical validation showing immunopositivity for PV. (b, top) Representative high-magnification confocal image of two biocytin-filled typical SST+ cells with somata in LIV, showing multipolar (left) and bitufted (right) dendritic arborization characterized by the presence of spines (orange arrowheads, small panels), and axons projecting from LIV to LI where they give rise to multiple collaterals (white arrowheads). (b, bottom) Posthoc immunohistochemical validation showing immunopositivity for SST.

Blockade of voltage-dependent Na+ channels by TTX abolished APs in PV+ cells from adult primary auditory cortex.

(a) Representative voltage responses of a PV+ cell at threshold current (+200 pA) and 2x threshold current (+400 pA) in absence and presence of TTX. (b-e) Cumulative histograms of sEPSCs/mEPSCs amplitude (bin width 0.5 pA) and time interval (bin width 10 ms) recorded from four PV+ cells. sEPSC were recorded for 2 minutes, then TTX (1μM; Alomone Labs) was perfused into the recording chamber. After 5 minutes, mEPSC were recorded for 2 minutes. (f-i) Time course plots of series resistance (Rs) of the four representative PV+ cells shown in b-e before (sEPSC) and during the application of TTX (mEPSC).

Syngap1 haploinsufficiency reduces the density of local vGlut1 excitatory inputs without affecting VGlut2 thalamocortical inputs to PV+ cell somata.

(a) Representative images of auditory cortex immunolabelled for PV (grey), VGlut1 (cyan), PSD95 (magenta) in control (Nkx2.1 Cre; Syngap1+/+) and cHet (Nkx2.1 Cre; Syngap1flox/+) adult mice. Red squares indicate the location of cell bodies shown as high magnification images. Scale bar: 10 µm (b) Quantification of the density of perisomatic puncta colocalizing both VGlut1 and PSD95 normalised to controls (Unpaired t-test, *p= 0.0447). Number of mice: n=5 mice for control and n=7 for cHet. (c) Representative images of auditory cortex immunolabelled for PV (grey), VGlut2 (cyan), PSD95 (magenta). Scale bar: 10 µm (d) Quantification of of the density of perisomatic puncta colocalizing both VGlut2 and PSD95 normalised to controls (Unpaired t-test, p= 0.3345). Number of mice: n=5 mice for control and n=8 for cHet. Yellow arrows indicate PV cell somata. Bar graphs represent mean ± SEM. ns p > 0.05 ns, not significant, * indicates p value <0.05.

sEPSC amplitude is reduced in LIV SST+ cells in Tg(Nkx2.1- Cre):RCEf/f:Syngap1f/+ mice

(a) Representative traces of sEPSCs recorded in SST+ cells from control (blue, n=17 cells, 10 mice) and cHet mice (red, n=10 cells, 8 mice). (b) Cumulative probability plots show a significant decrease in the amplitude (LMM, #p=0.010) and no change in the inter-sEPSC interval (LMM, p=0.126). Insets illustrate significant differences in the sEPSC amplitude for inter-cell mean comparison (LMM, *p=0.010) and no difference for inter-sEPSC interval (LMM, p=0.103. (c) Representative examples of individual sEPSC events (100 pale sweeps) and average traces (bold trace) detected in SST+ cells of control and cHet mice. (d) Superimposed scaled traces (left top) and summary bar graphs for a group of cells (bottom) show no significant differences in the sEPSCs kinetics (LMM, p=0.471 for rise time, and p=0.594 for decay time). (e) Summary bar graphs showing a significant decrease in cHet mice for inter-cell mean charge transfer (LMM, *p=0.047, left) and for the charge transfer when frequency of events is considered (LMM, *p=0.045). * and # indicates p value <0.05 for bar graph and cumulative distribution, respectively.

Membrane properties of BC-broad Vs BC-short in control mice.

Related to Fig. 6.

sEPSCs in SST+ cells from control Vs SST+ cells from cHet mice.

Related to Fig. S4.