Article Figures & Data
Figures
- Figure 1.
Arc directly interacts with the AP-2 complex. A, Arc coimmunoprecipitates with the α subunit of AP-2. Hippocampal lysate was subjected to IP with an Arc antibody followed by immunoblot (IB) using an anti-α adaptin antibody. Ten percent of the protein lysate used for the IP was loaded in the input lane. B, Arc coimmunoprecipitates with clathrin in hippocampal lysate. Hippocampal lysate was subjected to IP with a rabbit anti-Arc or a normal rabbit anti-IgG control antibodies followed by IB using an anti-α adaptin and anti-clathrin heavy chain antibodies. Ten percent of the protein lysate used for the IP was loaded in the input lane. C, D, Pull-down assay showing the interaction of AP-2 core with mouse Arc(WT). Recombinant affinity purified GST-AP-2 core [GST-tagged α subunit (residues 1–621), 6xHis tagged β2 subunit (residues 1–591), full-length µ2 and σ2 subunits], was immobilized on glutathione beads (C, right) and incubated with recombinant affinity purified 6xHis-Arc(WT) (C, left). Binding of Arc protein to GST-tagged AP-2 core or GST alone was analyzed by GST pull-down and SDS-PAGE, followed by Coomassie blue staining (D, left) or immunoblot using an anti-Arc antibody (D, right). E, Schematic representation of the Arc-WT sequence showing the truncated Arc mutants used in this study. The diagram indicates coiled-coil (CC) and spectrin repeat homology (SRH) structure domain of mouse Arc. AP-2 binding site is shown in black. f, Pull-down assay showing the interaction of AP-2 core with mouse Arc(WT) and the Arc(1-199) truncated (deletion of residues 200–396). Recombinant affinity purified AP-2 core, was immobilized on glutathione beads and incubated with lysates of E. coli expressing Arc(WT) or Arc(1-199) deletion mutant. Binding of Arc proteins to GST-tagged AP-2 or GST alone was analyzed by GST pull-down and SDS-PAGE, followed by Coomassie blue staining (left) or immunoblot using an anti-His tag antibody (right). Ten percent of the recombinant proteins used for the pull-down were loaded on the input lane (Bands corresponding to Arc proteins are indicated by white asterisks). G, Pull-down assay showing that the Arc residues 195–199 are required for the Arc–AP-2 interaction as truncated Arc(1-194) produced in E. coli lost the ability to bind immobilized recombinant GST-tagged AP-2 core.
- Figure 2.
Identification of Arc motif that binds to AP-2. A, Pull-down assay showing that a conserved tryptophan residue at position 197 mediates Arc–AP-2 interaction. Recombinant GST-Arc(WT), GST-Arc(W197A), GST-Arc(195-199A), or GST alone were produced in E. coli and immobilized on glutathione beads (bottom) and incubated with total brain tissue lysate. Binding of endogenous µ2 and α-adaptins to GST fusion proteins was analyzed by SDS-PAGE immunoblotting with anti-µ2 (top) or anti-α (middle) antibodies. Bar chart plotting analysis of the relative amount of protein bound to GST and GST-Arc(W197A) and GST-Arc(195-199A) normalized to GST-ArcWT (100%). B, Pull-down assay showing interaction of Arc(WT) and Arc(W197A) with dynamin 2. Recombinant GST-Arc(WT), GST-Arc(W197A), or GST alone were produced in E. coli, immobilized on glutathione beads (bottom) and incubated with total lysates of HEK293 cells expressing dynamin 2-GFP. Binding of dynamin 2-GFP to GST fusion proteins was analyzed by SDS-PAGE immunoblotting with anti-GFP antibody (top). Bar chart plotting analysis of the relative amount of dynamin 2 bound on the beads normalized to GST-ArcWT (100%). Ten percent of total protein lysate used to incubate the beads was used as input. C, Pull-down assay showing interaction of Arc(WT), Arc(W197A), and GST-Arc(195-199A) with Triad3A. Recombinant GST proteins produced in E. coli and immobilized on glutathione beads (bottom) were incubated with total lysates of HEK293 cells expressing GFP-Triad3A. Binding of GFP-Triad3A to GST fusion proteins was analyzed by SDS-PAGE immunoblotting with anti-GFP antibody (top). Bar chart plotting analysis of the relative amount of Triad3A bound on the beads normalized to GST-ArcWT (100%). Ten percent of total protein lysate used to incubate the beads was used as input. Errors bars represent mean ± SEM (n = 3 independent experiments). *p<0.05, **p<0.005, and ***p<0.0005 using unpaired Student´s t test.
- Figure 3.
Arc–AP-2 interaction regulates GluA1 endocytosis. A, B, Representative blots showing that Arc(WT), but not the Arc(W197A) mutant, facilitates GluA1, but not GluA2 endocytosis. H4 neuroglioma cells were transfected with plasmids encoding myc-GluA1 (A) or myc-GluA2 (B) in combination with either: empty pCIneo vector, pCIneo Arc(WT), or pCIneo Arc(W197A). Western blot band densitometry analysis showing that: (A) Arc(WT), but not Arc(W197A), promotes a significant reduction in surface expression of GluA1 subunits (control: 60.46 ± 2.97%; Arc(WT): 38.55 ± 7.44%; Arc(W197A): 50.18 ± 8.34%. Error bars represent mean ± SEM (n = 3 independent experiments). *p<0.05 using one-way ANOVA followed by Tukeýs post-test. B, Arc(WT) does not promote any changes in surface expression of either GluA2 subunits (control: 132.9 ± 26.66%; Arc(WT): 133.2 ± 21,78%; Arc(W197A): 143.9 ± 38.43%) or EGF receptor (control: 51.35 ± 10.43%; Arc(WT): 38.93 ± 8.66%; Arc(W197A): 41.59 ± 8.8%). Error bars represent mean ± SEM (n = 4 independent experiments).Ten percent of the protein lysate used for incubate the beads was loaded in the input lane. GAPDH was used as loading controls. C–F, C–G, H4 cells coexpressing myc-GluA1 with either mCherry construct alone (C), mCherry-Arc(WT) (D), or mCherry-Arc(W197A) (E). Surface myc-GluA1 (non-permeabilized cells, green channel) was identified using mouse anti-myc antibody followed by AlexaFluor 488 secondary antibody and internal myc-GluA1 (permeabilized, magenta channel) was identified by polyclonal rabbit anti-myc antibody followed by AlexaFluor 647 secondary antibody. F, G, The mean florescence intensity (MFI) of AlexaFluor 488 (surface my-GluA1) and mCherry (red channel) were calculated using confocal Z-projection images to quantify the pixel intensity of surface myc-GluA1 and mCherry (total protein expression). F, Ratio of averaged MFI between surface (488)/total protein (mCherry) for control cells (n=59 cells) was set to 100% to facilitate comparison. Note that the ratio for surface GluA1 is significantly reduced in cells expressing mCherry-Arc(WT) (34.68 ± 3.13%; n= 60 cells) compared with cells expressing mCherry construct alone. Importantly, this reduction is absent in cells expressing the mCherry-Arc(W197A) construct (114.80 ± 13.08%; n= 42 cells. G, Bar chart plotting the averaged MFI expression levels of mCherry-Arc(WT) and mCherry-Arc(W197A) compared with mCherry expression. Values are mean ± SEM (n=3 independent experiments). *p<0.05, ***p<0.005 using one-way ANOVA followed by Tukeýs post-test. Scale bar, 10 μm. H, Representative blot and bar chart plotting bands densitometry analysis of Arc expression protein in H4 cells transfected with equal amounts of mCherry-Arc(WT), mCherry-Arc(W197A), or mCherry-Arc(195-199A) plasmids. Note the similar levels Arc protein expression between samples. Values are mean ± SEM (n=3 independent experiments).
- Figure 4.
The Arc–AP-2 interaction regulates AMPAR-mediated synaptic currents. A–D, Representative live imaging of a dissociated hippocampal neuron overexpressing Arc-GFP-tagged constructs and GFP. A, AMPAR-mediated mEPSC traces from a neuron overexpressing Arc(WT) and an untransfected neighboring neuron. Ai, Amplitude probability distribution for the mEPSCs shown in A. Note the shift to the left and increase in the amplitude of the main peak in the neuron overexpressing Arc(WT), clearly demonstrating the reduction in mEPSC amplitude. Inset, Superimposed average mEPSC waveforms. B, Representative AMPAR-mediated mEPSC traces from a neuron overexpressing GFP and an untransfected neighboring neuron. Bi, Amplitude probability distributions for mEPSCs recorded from the neurons shown in B. Inset, Superimposed average mEPSC waveforms. C, Representative AMPAR-mediated mEPSC traces from a neuron expressing Arc(W197A) and an untransfected neighboring neuron. Ci, Amplitude probability distributions from neurons shown in C. Note that expression of Arc(W197A) produced a smaller reduction in mEPSC amplitude compared with Arc(WT) overexpression. Inset, Superimposed average mEPSC waveforms. D, Representative AMPAR-mediated mEPSC traces from a neuron expressing Arc(195-199A) and an untransfected neighboring neuron. Di, Amplitude probability distributions from neurons shown in d. Note that expression of Arc(195-199A), in which the sequence 195QSWGP199 of Arc was mutated to 195AAAAA199 had little effect on mEPSC amplitude. Inset, Superimposed average mEPSC waveforms. E, Cumulative probability distributions for cells expressing Arc(WT) (12 neurons), Arc(W197A) (13 neurons), Arc(195-199A) (10 neurons), GFP (7 neurons), and for untransfected cells (20 neurons). F, Bar chart plotting mean mEPSC amplitude for the cells in E. Expression of Arc(WT) significantly reduced mEPSC amplitude (mean reduced from 15.99 ± 0.9 pA in untransfected cells to 10.56 ± 0.66 pA, p= 0.0002), whereas expression of Arc(W197A) or Arc(195-199A) had no significant effect (14.6 ± 0.74 pA, p=0.12 and 14.01 ± 1.2 pA, p= 0.37). Expression of eGFP had no significant effect (p=0.376) on the mean mEPSC amplitude compared to untransfected cells. G, Bar chart plotting the mean interval between mEPSCs. Expression of Arc(WT) and the Arc mutants had no significant effect on the frequency of mEPSCs. Although the mean frequency of mEPSCs in cells expression Arc(WT) appeared reduced, this was not significant as there was large variability between cells. H, Representative average mEPSC waveforms recorded at a holding potential of −60 and + 40 mV for cells expressing GFP, Arc(WT), and Arc(W197A) in the presence of spermine (100 µm) in the intracellular solution. I, Bar chart plotting the mean rectification index (peak amplitude at +40 mV divided by peak amplitude at −60 mV) for neurons expressing GFP (n = 9 cells; 0.34 ± 0.015), Arc(WT) (n = 9 cells; 0.62 ± 0.016), and Arc(W197A) (n = 6 cells; 0.45 ± 0.015). Thus, Arc(WT) reduces the amount of rectification (as seen as an increase in the rectification index), whereas Arc(W197A) has significantly less effect on rectification. Error bars in F, G, and I are SEM. ***p<0.001, **p<0.01. Statistical significance was tested using the Mann–Whitney test. Scale bar, 10 µm.
- Figure 5.
Overexpression of Arc-cDNAs does not affect AMPAR-mediated mEPSC kinetics in hippocampal neurons. A, Average of 75 mEPSCs aligned on the midpoint of the rising phase) from an individual neuron expressing Arc(WT). The decay was fitted with a single exponential (τ = 4.5 ms, black line). Inset, The average mEPSC at an expanded time base showing the exponential fit to the decay. B, Average of 80 mEPSCs (aligned on the midpoint of the rising phase) from an untransfected neuron which was a close neighbor to the cell in A. The decay of the mEPSC was very similar to the transfected neighbor (the decay was fitted with a single exponential; τ = 4.7 ms, black line). Inset, The average mEPSC at an expanded time-base to show the exponential fit to the decay. C, Bar chart plotting the mean 10–90% rise time of mEPSCs recorded from untransfected neurons (n= 18) and from neurons expressing different constructs and in different conditions (n = 6 for each). The mean rise time was calculated by averaging the rise time of mean currents from individual recordings. There was no significant difference in the mean mEPSC rise time recorded from any of the neurons. D, Bar chart plotting the mean decay time constant (τ) from untransfected neurons (n=18) and from neurons expressing different constructs and in different conditions (n=6 for each). The mean decay time constant (τ) was calculated by averaging the time constant from the decay of mean currents from individual recordings. The decay of mEPSCs was not significantly different between conditions. The error bars in C and D are SEM. Statistical significance was tested using the Mann–Whitney test
- Figure 6.
AP-2 is required for Arc-dependent changes in synaptic strength. A, Blots showing levels of µ2 protein obtained from NSC cells overexpressing n.c. shRNA, µ2-shRNA2, µ2-shRNA3 plasmids for 3–4 d. GAPDH was used as loading control. Bar chart plotting the analysis of µ2 band intensity normalized by GAPDH. Error bars indicate ± SEM and significance was tested using one-way ANOVA. ***p=0.0001. B, Blots showing levels of µ2 protein obtained from cultured hippocampal neurons infected for 8–9 d with lentiviruses expressing either µ2-nc shRNA or µ2-shRNA2 sequences for 8–9 d. GAPDH was used as loading control. Bar chart plotting the analysis of µ2 band intensity normalized by GAPDH intensity. Error bars indicate ± SEM and significance was tested using one-way ANOVA. *p=0.019. C, Representative AMPAR-mediated mEPSC traces from a neuron expressing µ2-shRNA2 and an untransfected neighbor. Ci, Amplitude probability distributions from the neuron shown in C. Note that reduction of AP-2 expression (µ2-shRNA2) has little effect on mEPSC amplitude. Inset, superimposed average mEPSC waveforms. D, Representative AMPAR-mediated mEPSC traces from a neuron coexpressing Arc(WT) and nc shRNA and an untransfected neighbor. Di, Amplitude probability distribution from the neurons in D. Note that coexpression of a n.c. shRNA does not prevent overexpression of Arc from reducing mEPSC amplitude. Inset, Superimposed average mEPSC waveforms. E, Representative AMPAR-mediated mEPSC traces from a neuron coexpressing Arc(WT) and µ2-shRNA2 and an untransfected neighbor. Ei, Amplitude probability distribution from the neurons showed in E. Note that coexpression of µ2-shRNA2 prevents the effects of Arc(WT) on mEPSC amplitude. Inset, Superimposed average mEPSC waveforms. F, Cumulative probability distributions for cells expressing shRNA2 (9 neurons), Arc(WT) + shRNA2 (16 neurons), Arc(WT)+n.c shRNA (7 neurons), and for untransfected cells (12 neurons). G, Bar chart plotting mean mEPSC amplitude for the cells in f. Expression of shRNA2 prevented the Arc(WT) overexpression effect of significantly reducing mEPSC amplitude (mean mEPSC amplitude 15.3 ± 1 pA in untransfected cells, Arc(WT) + shRNA2 14.3 ± 0.8 pA; p=0.52). Expression of shRNA2 alone had no significant effect on mEPSC amplitude (13 ± 0.7 pA; p=0.07), whereas Arc(WT) + n.c. shRNA significantly reduced mEPSC amplitude (10.2 ± 0.53 pA; p=0.001). H, Bar chart plotting the mean interval between mEPSCs for the cells in F. The error bars in G and H are SEM. ***p<0.001, **p<0.005. Statistical significance was tested using the Mann–Whitney test. I, Amplitude probability distributions for a neuron expressing µ2-shRNA3 and an untransfected neighbor and for a neuron overexpressing Arc(WT) with µ2-shRNA3 and an untransfected neighbor (J). Inset, Superimposed average mEPSC waveforms. K, Bar chart of mean mEPSC amplitudes for untransfected cells (n = 8), cells transfected with μ2-shRNA3 (n=10) and cells transfected with Arc(WT)+ μ2-shRNA3 (n=6). Neither expression of μ2-shRNA3 or Arc(WT)+μ2-shRNA3 significantly changed mEPSC amplitude (p=0.68 and p=0.27, respectively).
- Figure 7.
Arc–AP-2µ interaction is required for the Arc-mediated reduction in AMPAR mEPSC amplitude. A, Representative AMPAR-mediated mEPSC traces from a neuron expressing Arc(WT)+µ2-shRNA2+µ2 and an untransfected neighboring neuron. Ai, Amplitude probability distributions from the neurons showed in A. Note that reintroduction of µ2 rescued the effect of Arc(WT) overexpression leading to a reduction in the amplitude of mEPSC amplitudes (shift to the left, red trace). Inset, Superimposed average mEPSC waveforms. B, Representative AMPAR-mediated mEPSC traces from a neuron expressing Arc(195-199A)+µ2-shRNA2+µ2 and an untransfected neighboring neuron. Bi, Amplitude probability distributions from the neurons showed in D. Note that reintroduction of µ2 has little effect in mEPSC amplitude (no shifts between black and red traces). Inset, Superimposed average mEPSC waveforms. C, Cumulative probability distributions for cells expressing Arc(WT) +μ2-shRNA2 +μ2 (14 neurons), Arc(195-199A) +μ2-shRNA2 +μ2 (9 neurons), and untransfected cells (14 neurons). D, Bar chart plotting mean mEPSC amplitude for the cells in C. Expression of μ2 rescued the reduction in mEPSC amplitude produced by Arc(WT) overexpression, following the knockdown of AP-2 by shRNA2 (mean mEPSC amplitude in untransfected cells 16.9 ± 1.3 pA vs 10.1 ± 0.6 pA in cells expressing Arc(WT) +μ2-shRNA2 +μ2; p=0.0001). In contrast, expression of μ2 had no significant effect on mEPSC amplitude when Arc(195-199A), which does not interact with AP2, was expressed together with shRNA2 (15.9 ± 1.7 pA; p = 0.46). E, Bar chart plotting the mean interval between mEPSCs for the cells in C. The error bars in D and E are SEM. ***p<0.001, **p<0.01. Statistical significance was tested using the Mann–Whitney test.
- Figure 8.
AP-2 is required for Arc-dependent homeostatic scaling. A, Average mEPSC waveforms from an untransfected neuron cultured in control conditions, from an untransfected neuron exposed to bicuculline and from a µ2-shRNA2 expressing neuron that has been incubated in bicuculline (40 µm; 48 h). Note that the bicuculline-induced down regulation of the mEPSC amplitude was reduced in AP-2 depleted cells (µ2-shRNA2 expressing cells).The untransfected neuron and the neuron expressing µ2-shRNA2 that were cultured in the presence of bicuculline were neighbors in the same dish, while the untransfected neuron cultured in control conditions was from the same preparation. B, Cumulative amplitude distribution for untransfected neurons cultured in control conditions (black line, n=10 neurons), untransfected neurons incubated in bicuculline (red line, n=15 neurons), µ2-shRNA2 expressing cells incubated in bicuculline (blue line; n=6 neurons) and cells transfected with n.c. shRNA incubated in bicuculline (green line; n=5 neurons). C, Bar chart plotting the mean mEPSC amplitude for the cells shown in B. Incubation in bicuculline significantly reduced the mean mEPSC amplitude (from 17.3 ± 1 pA to 11.9 ± 0.2 pA; p=0.0001). Expression of shRNA2 significantly increased mEPSC amplitude in bicuculline (14.38 ± 0.16 pA; p=0.0001), whereas n.c shRNA had significantly less effect (12.9 ± 0.28 pA; p=0.007). D, Bar chart plotting the mean interval between mEPSCs for the cells in B and C. The error bars in C and D are SEM. ***p<0.001, **p<0.01. Statistical significance was tested using the Mann–Whitney test.
- Figure 9.
Arc–AP-2 interaction controls synaptic strength. The proposed model showing the mechanism by which Arc–AP-2 interaction facilitates AMPAR endocytosis. An increase in neuronal activity promotes rapid Arc mRNA translation and protein expression at the dendritic spines. 1, Newly expressed Arc binds to the AP-2 complex and may activate/facilitate AP-2 interaction with AMPAR at the plasma membrane. 2, To initiate the formation of the clathrin-coated assembly AP-2 binds and recruits clathrin to the membrane. 3, 4, Arc then binds and recruits endophilin and dynamin to promote scission of the endocytic vesicle containing the AMPAR to be targeted for either recycling or degradation.
Tables
Results Data structure Type of test n numbers Probability, p (Fig. 2A, top) IB pull-down α GST-Arc(WT) vs GST Two-factor, mean t test 3/3 < 0.0001 GST-Arc(WT) vs GST-Arc(W197A) Two-factor, mean t test 3/3 < 0.0001 GST-Arc(WT) vs GST-Arc(195-199A) Two-factor, mean t test 3/3 < 0.0001 (Fig. 2A, middle) IB pull-down µ2 GST vs GST-Arc(WT) Two-factor, mean t test 3/3 < 0.0001 GST-Arc(WT) vs GST-Arc(W197A) Two-factor, mean t test 3/3 0.0007 GST-Arc(WT) vs GST-Arc(195-199A) Two-factor, mean t test 3/3 0.0039 (Fig. 2B) IB pull-down dyn2-GFP GST-Arc(WT) vs GST Two-factor, mean t test 3/3 < 0.0001 GST-Arc(WT) vs GST-Arc(W197A) Two-factor, mean t test 3/3 0.0159 (Fig. 2C) IB pull-down GFP-Triad3A GST-Arc(WT) vs GST Two-factor, mean t test 3/3 < 0.0001 GST-Arc(WT) vs GST-Arc(W197A) Two-factor, mean t test 3/3 0.0055 GST-Arc(WT) vs GST-Arc(195-199A) Two-factor, mean t test 3/3 0.0055 (Fig. 3A) IB Surface GluA1 pCIneo vs pArc(WT) Two-factor, mean ANOVA Tukey’s 3/3 0.1284 pCIneo vs pArc(W197A) Two-factor, mean ANOVA Tukey’s 4/4 0.5543 (Fig. 3B) IB Surface GluA2 pCIneo vs pArc(WT) Two-factor, mean ANOVA Tukey’s 4/4 >0.9999 pCIneo vs pArc(W197A) Two-factor, mean ANOVA Tukey’s 4/4 0.9637 (Fig. 3B) IB Surface EGFR pCIneo vs pArc(WT) Two-factor, mean ANOVA Tukey’s 4/4 0.6156 pCIneo vs pArc(W197A) Two-factor, mean ANOVA Tukey’s 4/4 0.7621 (Fig. 3F) IF Surface GluA1 mCherry vs mCherry-Arc(WT) Two-factor, mean ANOVA Tukey’s 59/60 <0.0001 mCherry vs mCherry-Arc(W197A) Two-factor, mean ANOVA Tukey’s 59/42 0.3438 (Fig. 3G) IF mCherry expression mCherry vs mCherry-Arc(WT) Two-factor, mean ANOVA Tukey’s 3/3 0.5625 mCherry vs mCherry-Arc(W197A) Two-factor, mean ANOVA Tukey’s 3/3 0.9211 (Fig. 3H) IB Arc expression mCherry-Arc(WT) vs mCherry- Arc(W197A) Two-factor, mean ANOVA Tukey’s 3/3 0.6892 mCherry-Arc(WT) vs mCherry- Arc(195-199A) Two-factor, mean ANOVA Tukey’s 3/3 0.4951 (Fig. 4) Arc–AP-2 interaction Arc(WT) vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 12/20 0.0002 0.47 Arc(W197A) vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 13/20 0.121 0.98 Arc(195-199A) vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 10/20 0.372 0.18 eGFP vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 7/20 0.376 0.39 (Fig. 5) cDNA constructs and mEPSC kinetics All constructs vs untransfected rise decay Two-factor, mean Mann–Whitney 6/18 >0.05 >0.05 (Fig. 6) AP-2 requirement for Arc mediated changes in synaptic strength μ2-miRNA2 vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 9/12 0.07 0.37 Arc(WT) + μ2-miRNA2 vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 16/12 0.52 0.63 Arc(WT) + n.c.miRNA vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 7/12 0.001 0.08 μ2-miRNA3 vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 10/8 0.68 0.45 Arc(WT) + μ2-miRNA3 vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 6/8 0.27 0.14 (Fig. 7) The Arc-AP-2μ interaction is required for Arc-mediated changes in synaptic strength Arc(WT) +μ2-miRNA2+μ2 vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 14/14 0.0001 0.37 Arc(195-199A)+μ2-miRNA2+μ2 vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 9/14 0.46 0.64 (Fig. 8) AP-2 is required for homeostatic scaling Control vs bicuculline (untransfected) amplitude frequency Two-factor, mean Mann–Whitney 10/15 0.0001 0.64 miRNA2 (bicuculline) vs untransfected (bicuculline) amplitude frequency Two-factor, mean Mann–Whitney 6/15 0.0001 0.59 n.c.miRNA (bicuculline) vs untransfected amplitude frequency Two-factor, mean Mann–Whitney 5/15 0.007 0.29
In this issue
