In the control hemisphere, RSU firing showed remarkable stability over the entire 9 days of recording (n = 5 animals, Figure 2C, ANOVA, p = 0.98). In marked contrast, RSU firing in the deprived hemisphere was strongly modulated by MD. Data from a representative animal are shown in Figure 2B for baseline, day 2 of MD (MD2), and MD6; while firing was depressed on MD2, firing rebounded by MD6. The same pattern was seen in the entire population of sampled units (n = 7 animals, Figure 2D, ANOVA, p = 0.013). Interestingly, on MD1, there was no reduction in firing but by MD2 average firing dropped significantly, to ∼60% of baseline (Tukey-Kramer test, p <

0.05). This pattern is consistent with the observation that acute selleck chemical lid suture blurs and decorrelates visual drive but does not produce a large drop in average LGN firing rates (Linden et al., 2009) and suggests that between MD1 and MD2 decorrelated visual drive leads to an active suppression of V1m firing (Rittenhouse Erastin mw et al., 1999; see Discussion). Crucially, over the next 2 days of MD (MD3–MD4), firing rates rebounded and by MD5–MD6 were indistinguishable from baseline. Although mean firing rates were ∼9% higher on MD6 (P32) relative

to baseline (P26), this increase was within the range of variation in the control hemisphere (Figure 2C) and was not significant (p = 0.98, Figure 2D, Tukey-Kramer test). If there was a dramatic change in the number of detectable neurons before or during monocular Isotretinoin deprivation, we might have under- or overestimated the size of the observed drop in firing and the subsequent rebound. However, the number of well-isolated units (indicated for each bar in Figure 2D) did not change significantly across days (chi-square test). Further, when we used conservative criteria to identify a subpopulation of individual RSUs we could follow for 2–6 days, this more stable population demonstrated the same biphasic pattern of firing during MD (Figures S2A–S2D). Finally, the same pattern of

drop and rebound was observed when we compiled average firing by animal (Figure S2E). Thus, the drop in averaging firing rate followed by a recovery to baseline is a robust feature of individual neurons under MD. To examine whether other aspects of neuronal firing were restored during prolonged MD, we compared the distribution of mean firing rates (Figure 2E), as well as the entire distribution of interspike intervals (ISIs) and ISI coefficient of variation (CV) (Figure 2F), as a function of days after MD. The entire distribution of mean RSU firing rates shifted to the left on MD2 (KS test, p < 0.01) and shifted back to become indistinguishable from baseline on MD6 (KS test, p = 0.33).