Models of firing rate homeostasis such as synaptic scaling and the sliding synaptic plasticity modification threshold predict that decreasing neuronal activity (for example, by sensory deprivation) will enhance synaptic function. Manipulations of cortical activity during two forms of visual deprivation, dark exposure (DE) and binocular lid suture, revealed that, contrary to expectations, spontaneous firing in conjunction with loss of visual input is necessary to lower the threshold for Hebbian plasticity and increase miniature excitatory postsynaptic current (mEPSC) amplitude. Blocking activation of GluN2B receptors, which are upregulated by DE, also prevented the increase in mEPSC amplitude, suggesting that DE potentiates mEPSCs primarily through a Hebbian mechanism, not through synaptic scaling. Nevertheless, NMDA-receptor-independent changes in mEPSC amplitude consistent with synaptic scaling could be induced by extreme reductions of activity. Therefore, two distinct mechanisms operate within different ranges of neuronal activity to homeostatically regulate synaptic strength.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Research reported in this article was supported by the National Eye Institute of the National Institutes of Health under award number R01EY012124 (to A.K.), R01EY016431 (to E.Q.), R01EY025922 (to E.Q., A.K. and H.-K. L.) and R01-EY014882 (to H.-K.L.). A.K. was also supported by NIH grant P01 AG009973. M.C.D.B. was supported by grants T32EY007143 and T32HL110952.
Integrated supplementary information
(A,B) Representative waveforms from units recorded before and after DE (A) and BS (B) (see Figure 1A, D). Thin lines represent individual waveforms; black lines represent the average waveform. (C) Single unit firing rates in naïve mice at baseline and following DE with diazepam treatment. There was no significant difference between groups (P=0.676, 2-tailed Wilcoxon rank sum test), confirming that diazepam prevents the DE-mediated increase in spontaneous firing (see Figure 1A). Data are shown as individual points; dashed lines and error bars represent mean±SEM. Sample size is indicated in parentheses as (cells, mice)
Supplementary Figure 2 Effects of visual deprivation and modulation of inhibition on single-unit firing properties.
(A) Spontaneous bursting increased after DE and after BS with flumazenil treatment. *P≤0.001 (U(60)=211.0), #P=0.047 (U(60)=235.0), †P=0.006 (U(64)=328.0), ‡P=0.013 (U(49)=193.0), 2-tailed Wilcoxon rank sum test. Sample size is shown as (units, mice). (B) Non-burst firing rates increased after DE. *P≤0.001 (U(60)=235.0), 2-tailed Wilcoxon rank sum test. No significant differences were observed between the recovery and DE+diazepam (U(60)=363.0, P=0.100), baseline and BS (U(64)=403.0, P=0.071), or recovery and BS+flumazenil (U(49)=238, P=0.103) conditions. Sample size as shown in panel (A). Data are shown as individual points; dashed lines and error bars represent mean±SEM. (C) Raster plots of two example units from awake, head-fixed mice viewing a grey screen. Spikes were classified as non-burst (blue) or part of a burst (magenta) (see Methods). A portion of the raster plots is enlarged (below) to show more detail
Supplementary Figure 3 Effects of acute drug treatment on spontaneous single-unit firing properties.
(A) Single unit recordings were obtained during presentation of a grey screen in awake, head-fixed mice. V1 firing rates were measured at baseline (open circles) and 20 minutes after drug administration i.p. (filled circles). Diazepam and THIP decreased, and flumazenil increased, spontaneous firing rate. *P≤0.001, #P=0.002, 2-tailed Wilcoxon signed rank test. Vehicle injection did not alter spontaneous firing rate (Z(15)=0.646, P=0.528). Sample size is indicated as (units, mice). (B) To demonstrate the long-lasting effects of a single drug injection i.p., single unit recordings were made from V1 under isoflurane anesthesia. Diazepam decreased and flumazenil increased (4 units, 4 mice per group) spontaneous neuronal firing rates for at least 2h after the injection. The dotted gray line indicates baseline firing rate for reference. *P≤0.001 vs. −30 minute time point, One-way repeated measures ANOVA with Holm-Sidak post-hoc test. Sample size is shown as (units, mice). (C,D) Single unit firing properties during acute drug administration. Diazepam and THIP decreased, and flumazenil increased, both spontaneous bursting (C; 2-tailed paired t test) and the inter-burst firing rate (D, 2-tailed Wilcoxon signed rank test). *P≤0.001, #P=0.002. Vehicle injection did not alter these parameters (bursting, 2-tailed paired t test t(15)=−0.313, P=0.759; inter-burst firing rate, 2-tailed Wilcoxon signed rank test Z(15)=0.267, P=0.831. Sample size as shown in (A). For all panels, individual unit firing parameters at each time point are connected by grey lines. Colored symbols and lines indicate the mean and error bars represent SEM
Supplementary Figure 4 Diazepam administered i.c.v. prevents the DE-mediated increase in mEPSC amplitude.
(A) Average mEPSC traces after diazepam or vehicle infusion directly into the lateral ventricle via osmotic minipump during DE. (B) mEPSC amplitude was larger in vehicle- than diazepam-infused animals. *P=0.013, 2-tailed t test. Dashed lines and error bars indicate mean±SEM. For clarity, some data points are displaced horizontally. Sample size is shown as (cells, mice)
Supplementary Figure 5 Activation of Gi-DREADD in PV cells elevates spontaneous activity in neighboring regular-spiking neurons.
(A) CNO (5 mg/kg i.p.) increased the firing rate of regular spiking neurons in hemispheres where Gi-DREADD was expressed in PV cells. *P≤0.001, 2-tailed paired t test. Data are from 31 units recorded from 3 awake, head-fixed animals viewing a grey screen. Grey lines represent individual units before and after CNO administration; green line and error bars represent mean ±SEM. (B) For mEPSC recordings (Fig. 3C, D), patch electrodes were targeted to pyramidal cells neighboring mCherry-labeled, Gi-DREADD expressing PV cells
Supplementary Figure 6 Diazepam administered i.c.v. prevents the DE-mediated increase in GluN2B function.
(A) DE increased the percentage of NMDAR EPSC blocked by ifenprodil (*P=0.021, 2-tailed t test) and the NMDA receptor decay constant (τw) (†P=0.002, 2-tailed t test) in animals with no drug administration. (B) In DE animals, diazepam prevented the GluN2B increase compared to vehicle (*P=0.039; †P=0.030; 2-tailed t test). Solid lines: average baseline NMDAR current; dashed lines: after ifenprodil wash-in. Traces are normalized to baseline. Data are displayed as individual points; dashed lines indicate average (±SEM); sample size is shown as (neurons, mice)