This result suggests that synapse size increases after a drop in cortical activity in vivo. To further characterize

the relationship between structural changes in vivo and functional changes in vitro, we determined whether the magnitude of changes in spine size was sufficient to explain the changes in mEPSC amplitudes. We first examined the distribution of the relative changes of spine size and mEPSC amplitude 24 hr following retinal lesions. Both distributions learn more changed significantly after lesions (Figures 3D and 3E) and, importantly, the distributions of spine size and mEPSC amplitude changed in a similar way, such that while the distributions of spines and mEPSC amplitudes in lesioned mice were significantly different from the respective control distributions (lesion versus control distribution: spines, K-S test, p < 0.05, mEPSC amplitude, K-S test, p < 0.05), they were not significantly

different from one another (lesion distributions: spines versus mEPSC amplitude, K-S test, p > 0.9). Second, one of the basic premises of synaptic scaling is its multiplicative nature (Turrigiano et al., 1998 and Turrigiano and Nelson, 2004). To determine whether the changes observed here are multiplicative, we multiplied the normalized control distribution of the mEPSC amplitudes by the necessary constant to scale the mean of the control distribution to match the lesion distribution mean (1.24). While the means of these two distributions would by definition be the same, HSP cancer the changes would only be multiplicative if the entire distribution was overlapping, indicating that all of the amplitudes had “scaled”

by the same factor. This is exactly what we found, as the scaled control distribution was statistically Tolmetin indistinguishable from the lesion distribution (K-S test, p > 0.3; Figure 3F), indicating that mEPSC amplitudes scaled multiplicatively, as has been reported previously (Turrigiano et al., 1998). In order to make the same measurements for the spine size changes, we first measured the control distribution of spine sizes 24 hr after sham lesions, normalized to 48 hr before, i.e., spine size changes over time. This control distribution provides the baseline level of spine size fluctuations in vivo. To determine whether the lesion-induced changes in spine size beyond these baseline fluctuations were multiplicative, we multiplied the control distribution by the constant required to match the mean of the normalized lesion distribution (1.22). Like in the mEPSC measurements, the means of these two distributions would by definition be the same, but the control distribution would be multiplicatively scaled if it overlaps the lesion distribution. Again, we found that the scaled control distribution and the lesion distribution were statistically indistinguishable (K-S test, p > 0.3, Figure 3G), indicating that the two distributions differ only by a multiplicative scaling factor.