Fig 1.
mC4 is expressed in neurons of the mPFC and can be overexpressed using IUE.
(A) Diagram depicting IUE surgery performed in E16 dams. (B) Representative 20X confocal image of IUE with GFP targeted to L2/3 mPFC, the electroporated region. Yellow asterisk: L2/3 GFP+ neurons. Left panel scale bar = 250 μm. Right panel scale bar = 75 μm. (C) IUE increases mC4 expression by 2.84-fold compared to P21 control. N = 18 control mice. N = 15 mC4 mice. t test with Welch’s correction. p = 0.0392. (D) mC4 mRNA expression at P60 for control and mC4 conditions. N = 10 control mice. N = 11 mC4 mice. t test with Welch’s correction. p = 0.2564. (C-D) qPCR performed from the dissected electroporated region. Control: blue. mC4: red. Mean ± SEM. (E) Immunoblot (top) and quantification (bottom) of relative mC4 levels in total lysates and isolated PSD fractions from GFP or mC4 conditions from the electroporated region. Since C4 was expressed at relatively low levels, it could only be detected when brains were pooled. Lysates and PSDs from the electroporated region were prepared from the individual mice and pooled for the western blot analysis. N = 7 mice per group for P21 (left 4 lanes). N = 4 mice per group for P60 (right 4 lanes). Vinculin was used as loading control for lysates and PSD fractions. PSD-95 immunoblot served as a control for successful isolation of PSD fractions. For underlying data, see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451. E, embryonic day; GFP, green fluorescent protein; IUE, in utero electroporation; L2/3, layer 2/3; mC4, mouse C4; mPFC, medial prefrontal cortex; P, postnatal day; PSD, postsynaptic density; qPCR, quantitative PCR.
Fig 2.
C4 overexpression led to dendritic spine alterations in apical arbors of L2/3 mPFC neurons.
(A) Developmental time course of spine density in the mPFC revealed a significant decrease in spine density (spines/μm) in neurons overexpressing C4, as compared to controls (“Con”), at P21–23. **p < 0.01, ***p < 0.001, ****p < 0.0001. N = 240 dendrites from 84 mice (20 dendrites per each time point; 20 dendrites × 4 time points × 3 conditions). (B) Representative 40X confocal images of P21–23 apical dendritic tufts. Scale bar = 5 μm. (C) Representative 40X confocal images of P21–23 apical dendritic spine types. Yellow asterisk: large mushroom spine (TIB [a.u.] > 75%). Green arrowhead: thin spine/filopodia (TIB [a.u.] < 25%). Scale bar = 3 μm. (D) Spine density (spines/μm) sorted by spine types reveals a specific reduction of medium-sized and thin/filopodia spine types in the mC4 and hC4 condition. *p < 0.05, **p < 0.01, ****p < 0.0001. N = 60 dendrites from 21 mice (20 dendrites per condition × 3 conditions). (E) Representative 40X confocal images of P21–23 basal dendritic spines. Scale bar = 5 μm. (F) Analysis of basal dendritic spine density (spines/μm) revealed no difference across groups. N = 80 dendrites from 28 mice (20 dendrites per each time point; 20 dendrites × 2 time points × 2 conditions). Two-way ANOVA followed by a Tukey’s test for all comparisons. Control: blue; mC4: red; hC4: purple. Mean ± SEM. For underlying data, see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451. a.u., arbitrary units; hC4, human C4; L2/3, layer 2/3; mC4, mouse C4; mPFC, medial prefrontal cortex; P, postnatal day; TIB, total integrated brightness.
Fig 3.
C4 overexpression reduced functional connectivity in cortical neurons.
(A) Top: Representative whole-cell voltage-clamp recordings showing mEPSCs. Top scale bar = 250 ms/10 pA. Bottom: Same as top traces but expanded (black rectangle region). Bottom scale bar = 125 ms/10 pA. (B) Increased mC4 expression caused a reduction in mEPSC frequency. t test. ****p < 0.0001. (C) Overexpression of mC4 caused a rightward shift in the distribution of mEPSC IEI. Kolmogorov-Smirnov test. ****p < 0.0001. (D) mC4 caused a reduction in mEPSC amplitude. t test. **p < 0.01. (E) mC4 overexpression caused a leftward shift in the mEPSC amplitude distribution. Kolmogorov-Smirnov test. **p < 0.01. (F) mEPSC Rise10–90 was not altered by mC4 overexpression. t test. p = 0.715. (G) No changes in mEPSC Decaytau with mC4 overexpression. t test. p = 0.07. (B-G) N = 12 control neurons, N = 10 mC4 neurons. Mean ± SEM. Control: blue; mC4: red. (H) Top: Representative recordings showing mIPSCs. Top scale bar = 250 ms/20 pA. Bottom traces are same as (H) top but expanded (rectangle region). Top scale bar = 125 ms/20 pA. (I) No difference in mIPSC frequency between control and mC4 conditions. t test. p = 0.3726. (J) Distribution of mIPSC IEIs was not changed by increased expression of mC4. Kolmogorov-Smirnov test. p > 0.05. (K) No changes in mIPSC amplitude with increased expression of mC4. t test. p = 0.5832. (L) mIPSC amplitude distribution was not changed by increased expression of mC4. Kolmogorov-Smirnov test. p > 0.05. (M) mIPSC Rise10–90 was increased in neurons overexpressing mC4. t test. *p = 0.0329. (N) mIPSC Decaytau was increased in neurons overexpressing mC4. t test. *p = 0.0225. (I-N) N = 10 control neurons, N = 12 mC4 neurons. Control: blue; mC4: red. Mean ± SEM. For underlying data, see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451. IEI, interevent interval; mC4, mouse C4; mEPSC, miniature excitatory postsynaptic current; mIPSC, miniature inhibitory postsynaptic current.
Fig 4.
Overexpression of mC4 decreased membrane capacitance without altering overall excitability.
(A-B) Representative current-clamp recordings from control (A) and mC4 (B) neurons in response to injection of constant-current pulses. Voltage traces (top) shown for response to −220 pA (light blue/pink), 185 pA (medium blue/red), and 685 pA (dark blue/red) current injections (bottom). Scale bar = 500 pA or 50 mV. Scale bar = 250 ms. (C) The number of APs was not different between conditions. Dark blue trace: control mean. Light blue trace: control SEM. Dark red trace: mC4 mean. Light red trace: mC4 SEM. Two-way ANOVA. p = 0.9989. (D) Interevent interval was not different between conditions. Dark blue trace: control mean. Light blue trace: control SEM. Dark red trace: mC4 mean. Light red trace: mC4 SEM. Two-way ANOVA. p = 0.9838. (E) Rheobase was not altered by the overexpression of mC4. t test. p = 0.8795. (F) mC4 overexpression led to a dramatic reduction in Cm. t test. ***p = 0.0007. (G) Vm was not changed by mC4 overexpression. t test. p = 0.2166. (H) Rm was not affected by mC4 overexpression. t test. p = 0.5455. (C-H) Control: N = 16 cells; mC4: N = 17 cells. Blue: control. Red: mC4. Mean ± SEM. For underlying data, see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451. AP, action potential; mC4, mouse C4.
Fig 5.
mC4 overexpression enhanced microglia engulfment of postsynaptic PSD-95.
(A-D) Representative confocal image (60X) showing PSD-95 located within microglial lysosomes in P21 mice for control (A-B) and mC4 conditions (C-D). Single z-plane shown. Scale bar = 5 μm. Magenta: microglia (“MG) (Iba1). Red: lysosomes (CD68). Green: PSD-95 (PSD95-FingR-RFP, pseudocolored green). White arrowheads indicate colocalization of PSD95 with Iba1 (A and C) or CD68 (B and D). Panels A and C show a representative microglia (for control and mC4, respectively) including Iba1 and PSD95-FingR-RFP signal. Panels B and D show the same z-plane as panels A and C but shows CD68 and PSD95-FingR-RFP signal. Orthogonal views shown. Panels A-D each include a graph (bottom right panel) showing a line intensity scan for each signal (panels A and C show Iba1 [magenta] and PSD-95 (green); panels B and C show CD68 [red] and PSD-95 [green]). For line intensity graphs, y-axis shows gray intensity value (a.u.), and x-axis shows length (μm). (E) mC4 overexpression increased the number of microglia positive for PSD-95 engulfment (PSD95-FingR-RFP signal colocalized with CD68 and Iba1). t test. ****p < 0.0001. Red dotted line: average of control and C4 in shuffled pixel analysis. (F) There was no difference in area of PSD95 colocalized with microglia lysosomes between conditions. (MG engulfment area % = area of microglia occupied by PSD95-FingR signal in lysosomes / total microglia area). t test. p = 0.1927. (E-F) Data points represent averages from transfected region ROIs. Control: N = 26 ROIs (from 5 mice; 345 microglia). mC4: N = 26 ROIs (from 5 mice; 319 microglia). Mean ± SEM. (G-H) Microglia engulfment area (%) (from panel F) as a function of cortical depth (μm) for control (G) and mC4 (H). Data points represent individual microglia. N = 345 control and N = 319 mC4 microglia. Right graphs are same as left but are zoomed on the y-axis. Blue line: control mean. Red line: mC4 mean. Gray lines: 95% confidence intervals. Pearson’s r correlation. Control: r = 0.12. p > 0.05. mC4: r = −0.21. **p < 0.01. For underlying data, see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451. a.u., arbitrary units; Iba1, ionized calcium binding adaptor molecule 1; mC4, mouse C4; P, postnatal day; PSD, postsynaptic density; ROI, region of interest.
Fig 6.
ExM confirmed C4 overexpression led to enhanced microglial engulfment of PSD-95.
(A-D) Representative confocal images (40X) in expanded tissue showing PSD-95 within microglial lysosomes in P21 mice for control (“con”) (A and B) and mC4 (C and D) conditions. Panels A and C show a representative microglia (for control and mC4, respectively) including Iba1 and PSD95-FingR-RFP (pseudocolored green) signal. White arrowheads indicate colocalization of PSD-95 with Iba1 (A and C). Panels B and D show the same z-plane as panels A and C but with CD68 and PSD95-FingR-RFP signal. Orthogonal views shown (XY, YZ, and XZ). All images are single z-planes. Magenta: microglia (Iba1). Red: lysosomes (CD68). Green: PSD-95 (PSD95-FingR-tagRFP, pseudocolored green). Yellow box in (A and C) shows zoomed region for orthogonal views. (A and C) Left scale bar = 5 μm, right scale bar (in XY plane) = 2.5 μm. (B and D) Scale bar = 1 μm. (E) mC4 overexpression led to an increase of lysosomes that were positive for PSD-95 signal. The percent of lysosomes that were PSD95 positive per each microglia compared between control and mC4. t test. ****p < 0.0001. (F) Number of lysosomes in microglia was increased in the mC4 condition relative to controls. t test. **p = 0.0064. (G) Lysosomes that were positive for PSD-95 were larger in size compared to PSD-95(−) lysosomes in both control (“con”) and mC4 conditions. Lysosome size (μm2) per microglia for control and mC4 conditions separated into lysosomes positive or negative for PSD-95 signal. Control (+) and mC4 (+) are lysosome size for lysosomes positive for PSD95. Control (−) and mC4 (−) are lysosome size for lysosomes that are negative for PSD95. Two-way ANOVA followed by a Tukey’s test for all comparisons. Control (+ versus −): **p = 0.0071. mC4 (+ versus −): *p = 0.0150. Control (+) versus mC4 (+): p = 0.0868. Control (−) versus mC4 (−): p = 0.2987. (E-G) Blue: control. Red: mC4. N = 86 microglia (45 control [from 5 mice] and 41 mC4 [from 4 mice] microglia). Mean ± SEM. For underlying data, see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451. ExM, expansion microscopy; Iba1, ionized calcium binding adaptor molecule 1; mC4, mouse C4; P, postnatal day; PSD, postsynaptic density.
Fig 7.
C4-dependent alterations in frontal cortical circuitry were sufficient to alter maternal–pup social interactions.
(A) Representative examples of path traveled (black trace) by P18 control and mC4 pups in MI1 task. Fresh bedding corners (Fresh 1 and 2, pink) and nest bedding corner (green). (B) Control and mC4 pups spent a similar proportion of time exploring the nest and fresh corners in the MI1 task, suggesting motor and sensory skills were intact. All mice spent more time in nest bedding than in fresh bedding. **p < 0.01, ***p < 0.001, ****p < 0.0001. Two-way ANOVA and Sidak's posttest. Control versus mC4 time spent in nest: p > 0.9999. Control versus mC4 time spent in Fresh 1: p = 0.3327. Control versus mC4 time spent in Fresh 2: p = 0.3138. (C) Representative examples of path traveled (black trace) by P18 control and mC4 pups in MI2 task. Dam’s cup (dam: blue), empty cup (empty cup: yellow), nest bedding corner (nest: green). (D) mC4 pups spent less time interacting with the dam and more time near the empty cup compared to controls. Two-way ANOVA and Sidak's posttest. *p < 0.05. ****p < 0.0001. (E) mC4 mice took longer to approach the dam relative to controls (s). t test with Welch's correction. *p < 0.05. (F) Control and mC4 pups traveled to dam’s cup first about 50% of the time, suggesting first choice was random. Fisher’s exact test. p = 0.9999. (G) Latency to reach first cup (either dam or empty) was not different. t test with Welch's correction. p = 0.06. (H) Control and mC4 pups reached the same maximum velocity (m/s). t test. p = 0.77. (B, D-H) N = 15 control mice and N = 21 mC4 mice. Blue: control. Red: mC4. Mean ± SEM unless otherwise noted. For underlying data and tracking script, see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451. mC4, mouse C4; MI1, maternal interaction 1; MI2, maternal interaction 2; P, postnatal day; Prop, proportion.
Fig 8.
C4-dependent alterations in frontal cortical circuitry led to behavior deficits in sociability that persist into adulthood.
(A) Representative examples of path traveled (black trace) by P60 control (left) and mC4 (right) adult mice in NO interaction task. Pink corner = location of NO. (B) Control and mC4 mice spent a similar amount of time exploring the NO. Percent time spent in corner with NO shown. p = 0.2127. (C) Representative examples of path traveled (black trace) by P60 control (left) and mC4 (right) adult mice in NO recognition task. Pink corner: location of NO. Green corner: location of FO. (D) NO recognition was intact in the mC4 condition. Control and mC4 mice had a similar DI ([time with NO − time with FO] / (time with NO + time with FO]). p = 0.1539. (E) Representative examples of path traveled (black trace) by P60 control (left) and mC4 (right) adult mice in sociability task. Pink corner: location of novel mouse under mesh wire cup. Green corner: location of empty mesh wire cup. (F) mC4-overexpressing mice spent less time exploring a novel mouse and more time with the empty cup (“E”) relative to control. Graph shows DI ([time with novel mouse–time with empty cup] / [time with novel mouse + time with empty cup]). *p = 0.029. (G) Control and mC4 mice had a similar latency to first approach the novel mouse (sec). p = 0.5416. (H) Control and mC4 mice traveled to the novel mouse cup first at a similar proportion. Proportion (“Prop.”) of mice that approached the novel mouse cup first. Fisher’s exact test. p = 0.54. (I) Latency to reach first cup (either novel mouse cup or empty cup) was not different (sec) between conditions. p = 0.217. (J) Control and mC4 mice reached the same maximum velocity (m/s). p = 0.1442. N = 22 control and N = 20 mC4 mice between P60 and 70. Blue: control. Red: mC4. t test unless otherwise stated. Mean ± SEM unless otherwise noted. For underlying data and script (respectively), see https://osf.io/7em3s/?view_only=0e7ffde4ebd344dc83af83b5a605c451 and https://github.com/balajisriram/dlc_utils. DI, discrimination index; FO, familiar object; mC4, mouse C4; NO, novel object; P, postnatal day.
Table 1.
Primers used for qPCR experiments.