Proteomic Profiling Reveals Age-Related Changes in Transporter Proteins in the Human Blood-Brain Barrier

Abstract

The Blood-Brain Barrier (BBB) is a crucial, selective barrier that regulates the entry of molecules including nutrients, environmental toxins, and therapeutic medications into the brain. This function relies heavily on brain endothelial cell proteins, particularly transporters and tight junction proteins. The BBB continues to develop postnatally, adapting its selective barrier function across different developmental phases, and alters with aging and disease. Here we present a global proteomics analysis focused on the ontogeny and aging of proteins in human brain microvessels (BMVs), predominantly composed of brain endothelial cells. Our proteomic profiling quantified 6,223 proteins and revealed possible age-related alteration in BBB permeability due to basement membrane component changes through the early developmental stage and age-dependent changes in transporter expression. Notable changes in expression levels were observed with development and age in nutrient transporters and transporters that play critical roles in drug disposition. This research 1) provides important information on the mechanisms that drive changes in the metabolic content of the brain with age and 2) enables the creation of physiologically based pharmacokinetic models for CNS drug distribution across different life stages.

Figure 1.. Brief experimental workflow and overview of BMV proteome

(A) Brief experimental workflow. Brain microvessels (BMVs) were isolated from frozen insular cortical bran tissue, which were then digested and analyzed using LC–MS/MS proteomic methods. Differential protein analysis and weighted correlation network analysis(WGCNA) were performed to identified proteins different between age groups and possible altered Blood brain barrier parameters were incorporated into physiologically based pharmacokinetic (PBPK) modeling simulation. (B) The PANTHER protein class of the identified proteins in BMV proteome. (C) Top BBB related GO terms associated with our BMV proteome are shown. X-axis is the gene enrichment ratio (Gene ratio) and the bubble size indicates the numbers of proteins associated a biological process GO term, with color maps the FDR value (p.adjust, q-value) of the enrichment analysis. (D)Principal component analysis (PCA) on BMV proteomic dataset showing the separation between different age groups.

Figure 2.. Differential expression of discovery BMV proteome through early childhood

(A) Volcano plot displaying the log2 fold change (x axis) against the t test–derived −log10 statistical P value (y axis) for all proteins differentially expressed between Development group (N=17) and Adult group (N=10) of the B proteome. Proteins with significantly decreased levels in Adults (P < 0.1) are shown on the left side, while the proteins with significantly increased levels through developement are shown on the rigMVht side. Transporters are labeled in red and proteins important for BBB integrity are labeled in green. (B) Top GO terms associated with proteins significantly increased with age are shown. X-axis is the gene enrichment ratio (Gene ratio) and the bubble size indicates the numbers of proteins associated a biological process GO term, with color maps the FDR value (p.adjust, q-value) of the enrichment analysis. (C) Bar graph showing collagens highly expressed in the BMV proteomic dataset, ordered by expression levels with the highest in the Developmental group. (D) Protein expression of Collagens which are expressed in our proteomic dataset and exhibit positive(red) or negative(blue) correlation with age. Scale represents the row Z-score (each row presents one protein), which is calculated by taking each individual’s protein expression, subtracting the mean expression, and then dividing by the standard deviation of that protein.

Figure 3.. BMV proteins co-expression network identified age-dependent modules which are important for BBB function

(A) The correlation between modules and age. Heatmaps shows the correlation between eigengene and age and each cell contains the corresponding correlation followed by p-value. (B) Module eigenprotein levels by age groups (Development, Adult, Elderly) for the three blood brain barrier related modules. Modules are grouped by different age groups. (C) Macronutrient transporters in our BMV proteomic dataset are shown in bar graph. Amino acid transporters are labeled as black, sugar transporters are labeled as green and choline transporters are labeled as orange. (D) Micronutrient transporters in our BMV proteomic dataset are shown in circular bar graph, bars represented the mean of the transporter expressions in different age groups. For the bar graphs Kruskal–Wallis tests followed by Dunn’s post hoc test were used to compare the mean of each age groups with the mean of the Adult group. Data are represented as mean ± SEM and each points represent one sample. *P < 0.05, **P < 0.01, ***P < 0.001. ND, not detected in more than 30% of samples in specific age group.

Figure 4.. Changes in clinically important ADME transporters and BBB permeability potentially lead to different drug distribution in brain.

(A) Clinically important uptake and efflux transporters (labeled red) and their family members in our BMV proteomic dataset are shown in bar graph. Kruskal–Wallis tests followed by Dunn’s post hoc test were used to assess difference between age groups in bar graph. Data are represented as mean ± SEM and each points represent one sample. *P < 0.05, **P < 0.01, ***P < 0.001. ND, not detected in more than 30% of samples in specific age group. (B, C) Phenytoin time-concentration profile in plasma and CSF with varying levels of P-gp (ABCB1) expression at the BBB. The default value is represented by the black solid line, the minimum value from BMV proteome is shown by the blue dashed line, and the maximum value from BMV proteome is indicated by the red dashed line. (D, E) Phenytoin time-concentration profile in plasma and spinal CSF with different BBB permeability. The default value is shown by the black solid line, 200% of the default BBB permeability is represented by the blue dashed line, and 50% of the default BBB permeability is shown by the red dashed line.

Figure 5.. Proteins Differentially Expressed During Aging Tend to Be Associated with Alzheimer’s Disease

(A) Volcano plot displaying all proteins differentially expressed between Adult group (N=10) and Elderly group (N=7) of the BMV proteome. Proteins with significantly decreased levels in elderly population (P < 0.1) are labeled in blue, while the proteins with significantly increased levels with aging are shown in red. Transporters are labeled in red and proteins known to play crucial roles in aging are labeled in purple. (B)Top BBB related GO terms associated with proteins significantly decreased with age are shown. X-axis is the gene enrichment ratio (Generatio) and the bubble size indicates the numbers of proteins associated a biological process GO term, with color maps the FDR value (p.adjust, q-value) of the enrichment analysis. (C) Protein expression of AD GWAS genes which are expressed in our proteomic dataset and exhibit positive(red) or negative(blue) correlation with age. Scale represents the row Z-score (each row presents one protein), which is calculated by taking each individual’s protein expression, subtracting the mean expression, and then dividing by the standard deviation of that protein.(D) Protein abundance levels of WWOX and in healthy elderly population or in patients with AD. Expression difference between disease condition were assessed by student T test. *P < 0.05, **P < 0.01, ***P < 0.001 (E) Protein abundance of APOE and PICALM through aging are described as simple linear regression model. Individual curve are presented in (solid lines). Dashed lines represent the 95% confidence bands.