Characterizing synaptic protein development in human visual cortex enables alignment of synaptic age with rat visual cortex

Abstract

Although many potential neuroplasticity based therapies have been developed in the lab, few have translated into established clinical treatments for human neurologic or neuropsychiatric diseases. Animal models, especially of the visual system, have shaped our understanding of neuroplasticity by characterizing the mechanisms that promote neural changes and defining timing of the sensitive period. The lack of knowledge about development of synaptic plasticity mechanisms in human cortex, and about alignment of synaptic age between animals and humans, has limited translation of neuroplasticity therapies. In this study, we quantified expression of a set of highly conserved pre- and post-synaptic proteins (Synapsin, Synaptophysin, PSD-95, Gephyrin) and found that synaptic development in human primary visual cortex (V1) continues into late childhood. Indeed, this is many years longer than suggested by neuroanatomical studies and points to a prolonged sensitive period for plasticity in human sensory cortex. In addition, during childhood we found waves of inter-individual variability that are different for the four proteins and include a stage during early development (<1 year) when only Gephyrin has high inter-individual variability. We also found that pre- and post-synaptic protein balances develop quickly, suggesting that maturation of certain synaptic functions happens within the 1 year or 2 of life. A multidimensional analysis (principle component analysis) showed that most of the variance was captured by the sum of the four synaptic proteins. We used that sum to compare development of human and rat visual cortex and identified a simple linear equation that provides robust alignment of synaptic age between humans and rats. Alignment of synaptic ages is important for age-appropriate targeting and effective translation of neuroplasticity therapies from the lab to the clinic.

Keywords: development; human cortex; rat cortex; synaptic proteins; visual cortex.

Figure 1

Developmental changes in GAPDH expression in human visual cortex. (A) Gray dots are results from all runs, and black dots are the average for each sample. Example bands are shown above the graph. (B) Group means and standard error for each developmental group.

Figure 2

Developmental changes in Synapsin (A,B) and Synaptophysin (C,D) expression in human visual cortex. (A,C) Gray dots are results from all runs, and black dots are the average for each sample. Example bands are shown above the graphs. (B,D) Group means and standard error for each developmental stage are plotted. (A) An exponential decay function was fit to all the Synapsin data points (R = 0.66, p < 0.0001), and adult levels are defined as 3t (3t = 8.7 +/− 5.1 years). (B) There was a significant difference in expression of Synapsin between the groups (ANOVA, p < 0.0001), and the statistical significance of the difference between pairs of development stages as determined by Tukey’s post hoc comparisons are plotted (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). (C) A weighted average fit was plotted to all of the Synaptophysin data points to describe pattern of change. (D) There was no significant difference in expression of Synaptophysin between the groups (ANOVA, p = 0.09).

Figure 3

Developmental changes in PSD-95 (A,B) and Gephyrin (C,D) expression in human visual cortex. (A,C) Gray dots are results from all runs, and black dots are the average for each sample. Example bands are shown above the graphs. (B,D) Group means and standard error for each developmental stage are plotted. (A) A Gaussian function was fit to all the PSD-95 data points (R = 0.58; p < 0.0001), and a peak in expression was reached at 8 years of age (peak = 8.0 +/− 0.7 years). (B) There was a significant difference in expression of PSD-95 between the groups (ANOVA, p < 0.0001), and the statistical significance of the difference between pairs of development stages as determined by Tukey’s post hoc comparisons are plotted (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). (C) A Gaussian function was fit to all the Gephyrin data points (R = 0.48; p < 0.0005), and a peak in expression was reached at 10.0 years of age. (D) There was a significant difference in expression of Gephyrin between the groups (ANOVA, p < 0.005), and the statistical significance of the difference between pairs of development stages as determined by Tukey’s post hoc comparisons are plotted (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).

Figure 4

Development of the Variance-to-Mean Ratio (VMR) for Synapsin and Synaptophysin (A), as well as PSD-95 and Gephyrin (B). (A) Synapsin (open circles, dashed line, weighted average function) had 3 peaks in VMR across the lifespan (1 year, 5–10 years, and older adults). Synaptophysin (filled dots, solid line; a * exp(b/x + c * x), R = 0.86, p < 0.0001) had a peak in VMR around 1 year of age. (B) PSD-95 (open circles, dashed line; a * exp(b/x + c * x), R = 0.82, p < 0.0001) had a peak in VMR throughout childhood. Gephyrin (filled dots, solid line; a + (b − a)/(1 + (x/c)^d), R = 0.88, p < 0.0001) had a decline in VMR starting at about 5 years of age (inflection point = 5.2 years +/− 0.9).

Figure 5

Developmental changes in the pre-synaptic (A,B) and post-synaptic (C,D) index in human visual cortex. (A,C) Gray dots are results from all runs, and black dots are the average for each sample. Example bands are shown above the graphs. (B,D) Group means and standard error for each developmental stage are plotted. (A) An exponential decay function was fit to all the pre-synaptic index data points (R = 0.67, p < 0.0001), and adult levels are defined as 3t (3t = 11.7 +/− 4.1 months). (B) There was a significant difference in expression of the pre-synaptic index between age groups (ANOVA, p < 0.0005) and the statistical significance of the difference between pairs of development stages as determined by Tukey’s post hoc comparisons are plotted (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). (C) An exponential decay function was fit to all the post-synaptic index data points (R = 0.51, p < 0.0001), and adult levels were defined as 3t (3.5 +/− 1.8 months). (D) There were no significant differences in expression of the post-synaptic index among the developmental stages (ANOVA, p = 0.18).

Figure 6

Principal component analysis. (A) The percent variance captured by each component of the SVD analysis of protein expression in human visual cortex. The first 2 principal components represent 84% of the SVD. (B) The influence of each protein on the first principal component was reflected by the relative amplitude in the basis vectors. (C) The influence of each protein on the second principal component was reflected by the relative amplitude in the basis vectors. (D) Significant correlations (colored cells) between the first 2 principal components and the combinations of proteins derived from the basis vectors. The color indicates the magnitude (represented by color intensity) and direction (green indicates positive, red indicates negative) of significant correlations (Bonferroni corrected, p < 0.0035).

Figure 7

Developmental changes in the principal components 1 and 2 in human visual cortex. (A) Principal component 1. A logistics function was fit to the data. Principal component 1 had a peak in expression at 9 years of age (Figure 6; Peak = 9.2 +/− 0.7 years; curve-fit, R = 0.52, p < 0.0001). (B) Group mean and standard error for each developmental stage are plotted and the statistical significance (ANOVA, p < 0.0001) of the difference between pairs of development stages as determined by Tukey’s post hoc comparisons are plotted (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). (C) Principal component 2. A linear function was fit to the data (R = 0.43, p < 0.005). (D) Group mean and standard error for each developmental stage are plotted and there were no significant differences in expression among experimental groups (ANOVA, p = 0.11).

Figure 8

Transformation of rat data to human age. The sum of the four synaptic proteins (synaptic protein expression) is plotted for human (black dots, age in years), rat (red dots, age in days), and the rat results transformed into human age (open red dots). Human data was fit with a Gaussian function (R = 0.74, p < 0.0001), and reached a peak in expression at 9 years of age (Peak: x-axis = 9.4 years, y-axis = 5.7). Rat data was fit with a decay function (R = 0.92, p < 0.0001), and reached maximum expression at P55 (Peak: x-axis = 54.5 days, y-axis = 5.2). The human and rat alignment was done by normalizing to the peak expression (Human Expression at 9 Years/Rat Expression at 55 Days), then determining the offset on the x-axis to align the rat curve (dotted red curve) with the human curve (solid black curve) of synaptic protein expression, and then applying that transformation to each of the rat data points. The transformed rat data was plotted (open red dots, dotted red curve) in human equivalent units. A large portion of the human (black) and transformed rat (dotted red) curves approximated a linear increase, allowing for a simple alignment of rats age in days with human post-conception age in years (Rat Age (Days) = 11 + 5.5 * Human Post Conception Age (Years)).