Large-scale deep proteomic analysis in Alzheimer's disease brain regions across race and ethnicity

Abstract

Introduction: Alzheimer's disease (AD) is the most prevalent neurodegenerative disease, yet our comprehension predominantly relies on studies within non-Hispanic White (NHW) populations. Here we provide an extensive survey of the proteomic landscape of AD across diverse racial/ethnic groups.

Methods: Two cortical regions, from multiple centers, were harmonized by uniform neuropathological diagnosis. Among 998 unique donors, 273 donors self-identified as African American, 229 as Latino American, and 434 as NHW.

Results: While amyloid precursor protein and the microtubule-associated protein tau demonstrated higher abundance in AD brains, no significant race-related differences were observed. Further proteome-wide and focused analyses (specific amyloid beta [Aβ] species and the tau domains) supported the absence of racial differences in these AD pathologies within the brain proteome.

Discussion: Our findings indicate that the racial differences in AD risk and clinical presentation are not underpinned by dramatically divergent patterns in the brain proteome, suggesting that other determinants account for these clinical disparities.

Highlights: We present a large-scale proteome (∼10,000 proteins) of DLPFC (998) and STG (244) across AD cases. About 50% of samples were from racially and ethnically diverse brain donors. Key AD proteins (amyloid and tau) correlated with CERAD and Braak stages. No significant race-related differences in amyloid and tau protein levels were observed in AD brains. AD-associated protein changes showed a strong correlation between the brain proteomes of African American and White individuals. This dataset advances understanding of ethnoracial-specific AD pathways and potential therapies.

Keywords: Alzheimer's disease; data descriptor; diversity; precision medicine; proteome; proteomics.

FIGURE 1

(A) Schematic illustrating cohort characteristics and experimental workflow for TMT‐MS of human brain proteome across frontal and temporal brain tissue samples. This study incorporated a total of 1105 DLPFC brain tissues from 998 individuals, categorized as follows: 486 NHW, 328 African American, 229 Latino American, and others as applicable. These samples were sourced from four prominent data distribution sites: Emory University, Mayo Clinic, Rush University, and Mount Sinai University Hospital. Additionally, 280 STG tissues from a subset of 244 individuals were included, with 116 NHW, 86 African American, 78 Hispanic, and others as applicable. STG samples were obtained from a racially diverse set of specimens originating from Mayo Clinic and Emory, distributed across 19 batches. Tissues underwent an experimental pipeline involving protein digestion, batch randomization, TMT labeling, fractionation, and subsequent MS measurements. A total of 72 DLPFC batches were processed, comprising nine batches from Emory, 24 from Mayo Clinic, 14 from Mount Sinai, and 25 from Rush (comprising a total of 72 batches). Batches were randomized to ensure a representative and diverse dataset. The output included a total of 6479 raw files for DLPFC samples and 1824 raw files for STG. (B) Venn diagram of total number of proteins quantified from DLPFC and STG samples. A total of 11,748 protein groups were identified from DLPFC and 11,003 from STG samples, with 10,738 shared protein groups. (C) Venn diagram of total protein from DLPFC and STG samples after QC across all samples. 9180 protein groups were identified from DLPFC samples and 9734 from STG, with 9015 shared protein groups. DLPFC, dorsolateral prefrontal cortex; NHW, non‐Hispanic White; QC, quality control; STG, superior temporal gyrus; TMT‐MS, tandem mass tag mass spectrometry.

FIGURE 2

QC and batch correction for DLPFC tissue proteins. (A) QC workflow in three main steps. Step 1: Preprocessing for missing values: Only proteins with missing data in less than 50% of the samples were retained. The ratio of protein abundance to the total protein abundance for each sample was calculated to adjust for sample loading differences resulting in 9180 proteins being retained across 1105 samples. Subsequently, the data were log₂ transformed. Step 2: Outlier detection and removal: Iterative PCA was employed to identify and eliminate sample outliers. After multiple rounds of PCA analysis, 19 outliers were identified and removed, leaving 9180 proteins across 1086 samples. Step 3: Batch effect regression: Variance attributable to batching was mitigated through regression of the 9180 proteins in 1086 samples. (B and C) MDS plot showing variation among samples (B) before correcting for batch and (C) after regressing for batch effect. The plot dimensions (dim 1 and 2) reveal distinctive clusters formed by samples by site – Emory (red), Mount Sinai (blue), Rush (purple), and Mayo (green) – with some scattering observed among samples before regressing for batch effect (B). The plot in (C) illustrates the successful removal of variance due to batch. After correcting for batch effects, samples from all four sites – Emory (red), Mount Sinai (blue), Rush (purple), and Mayo (green) – cluster together, indicating a more cohesive grouping (ie, the change in scale from B to C). The correction mitigates the dispersion observed in (B), highlighting the effectiveness of the batch correction procedure in harmonizing the sample distribution across different data distribution sites. (D and E) Variance partition analysis using experimental factors to evaluate the percentage of explained variance in proteomic samples. Violin plots before (D) and after (E) batch correction illustrate the distribution of explained variances in overall proteomic values. The y‐axis represents the percentage of explained variance, while the x‐axis depicts factors contributing to variance, such as age, sex, race, diagnosis, residuals, and batch. Notably, batch variance is present before batch correction, influencing the overall proteomic profile. Panel (E) displays the same factors on the x‐axis after batch correction. Significantly, the violin plot demonstrates a substantial reduction in variance associated with batch, ultimately reaching near zero percent after batch regression. Moreover, even after batch correction, factors such as age, sex, race, AD diagnosis, and other individual traits (residual) had levels of impact on protein abundance patterns. Each point on the violin plot represents a specific protein, with the corresponding name next to it. This underscores the efficacy of the correction procedure in eliminating batch‐related variability from the proteomic data. DLPFC, dorsolateral prefrontal cortex; MDS, multidimensional scaling; NHW, non‐Hispanic White; PCA, principal component analysis.

FIGURE 3

Variance explained by individual traits in DLPFC tissues. Bar plots (A, C, E) depict amount of variance explained by sex, race, and AD diagnosis across all DLPFC samples. (A) Top‐ranking proteins associated with sex in dataset were identified through variance partitioning and depicted as bar plots. Boxplots in (B) illustrate log2 normal abundance levels of four selected proteins exhibiting significant differences between males and females. These proteins serve as key indicators of sex‐related variations and are depicted with statistical significance (p < 0.05). (C) Bar plots of top‐ranking proteins associated with race differences in DLPFC dataset. Boxplots in (D) illustrate log2 normal abundance levels of four selected proteins demonstrating significant differences between African American individuals and other races (p < 0.05). (E) Bar plots identified top‐ranking proteins contributing to differences in diagnosis of AD within dataset. Boxplots in (F) display the log2 normal abundance levels of four selected proteins exhibiting significant differences between AD patients and controls, as well as other diagnostic categories (p < 0.05). AD, Alzheimer's disease; DLPFC, dorsolateral prefrontal cortex.

FIGURE 4

Association between APOE ε4 genotype and prototype across DLPFC and STG samples. (A) The boxplots of log2 normal abundance of APOE ε4 protein measured by TMT‐MS across each APOE genotype reveal a high APOE ε4 abundance among APOE ε4 carriers among 920 unique DLPFC tissue samples. (B) Histogram of APOE ε4 log2 normal abundance among DLPFC samples (y‐axis) across ε4 allele presence (red) and non‐presence (blue) (x‐axis). (C) The boxplots of log2 normal abundance of APOE ε4 protein measured by TMT‐MS across 244 STG unique tissue samples reveal a high APOE ε4 abundance among APOE ε4 carriers. (D) Histogram of APOE ε4 log2 normal abundance among STG samples (y‐axis) across ε4 allele presence (red) and non‐presence (blue) (x‐axis). High levels of APOE ε4 abundance were observed in cases with the ε4 allele combination in both cortices, with a few discrepancies between APOE ε4 genotyping and prototyping (purple) being depicted. These inconsistencies may be attributed to various factors, including mis‐genotyping or potential technical challenges in MS measurements, such as isotope impurity and low signal‐to‐noise ratio in specific samples. DLPFC, dorsolateral prefrontal cortex; STG, superior temporal gyrus; TMT‐MS, tandem mass tag mass spectrometry.

FIGURE 5

Correlation of proteomic measurements of tau (MAPT) and APP levels with Braak and CERAD scores, as well as with other proteins, in DLPFC. (A) Boxplots depicting relative abundance of APP across AD (pink) and control (green) in DLPFC tissue samples (adjusted ANOVA p value < 0.05). (B). Raincloud plots depict group differences in relative abundance of APP (y‐ axis) across distinct CERAD stages (x‐axis) in DLPFC tissues. The analysis revealed a stepwise increase in the median APP levels with ascending CERAD classifications, indicating a progressive trend in APP abundance corresponding to different CERAD groups (score 1: green, score 2: orange, score 3: purple, score 4: pink). (C) Bicor correlates to APP. The plot illustrates the results of a bicor pairwise correlation between APP and 9180 proteins in DLPFC region. Proteins with a significant positive correlation with APP (p < 0.05) are highlighted in pink, whereas those with a significant negative correlation (p < 0.05) are displayed in green. Proteins that did not show a significant correlation with APP are colored in gray. Among the 9180 proteins analyzed, 3273 showed a positive correlation with APP, and 2778 were negatively correlated with APP. (D) Boxplots depicting relative abundance of MAPT across AD (brown) and control (green) in DLPFC tissue samples (adjusted ANOVA p value < 0.05). (E) Raincloud plots illustrate group differences in relative abundance of MAPT (y‐axis) across distinct Braak stages (x‐axis) in DLPFC tissues. The Braak stages range from 0 to 6, with corresponding colors representing different stages (0: dark green, 1: orange, 2: purple, 3: pink, 4: light green, 5: yellow, 6: brown). Notably, the analysis highlights elevated MAPT levels at Braak stages 5 and 6, aligning with the expected increase in tau tangles in later stages of Braak in the frontal cortex. (F) Bicor pairwise correlation analysis between MAPT (tau) and 9180 proteins in the DLPFC region. Proteins with a significant positive correlation (p < 0.05) are shown in brown, while those with a significant negative correlation (p < 0.05) are shown in green. Of the proteins analyzed, 4707 had a positive correlation, and 1226 had a negative correlation with MAPT. Non‐significant correlations are depicted in gray. AD, Alzheimer's disease; APP, amyloid precursor protein; Bicor, biweight midcorrelation; DLPFC, dorsolateral prefrontal cortex; CERAD, Consortium to Establish a Registry for Alzheimer's Disease; MAPT, microtubule‐associated protein tau.

FIGURE 6

TMT‐MS quantification of APP, MAPT, and their specific fragments across White and African American cases with and without AD. (A) Schematic of APP fragments. Schematic representation of amino acid sequence of APP, with the specific C‐terminal tryptic cleavage sites leading to Aβ40 and Aβ42 measures via MS. The site of trypsin cleavage is marked by this red “K” (lysine) residue. (B‐D) Boxplots of log2 normalized abundance of APP (B), Aβ42 (C), and Aβ40 (D) in brain tissue samples from 980 unique individuals categorized by race and AD status. The four groups analyzed are White CT (n = 125), AA CT (n = 63), White AD (n = 281), and AA AD (n = 145). Data were adjusted for age and sex using the bootstrap method before analysis. The middle line in each boxplot shows the median, the box covers the range from the 25th to the 75th percentile, and the whiskers indicate the full range of the data. One‐way ANOVA was conducted to assess overall differences among the four groups, with significance set at p < 0.05. APP, Aβ42, and Aβ40 levels were significantly higher in AD cases compared to controls. Post hoc comparisons using Bonferroni correction did not reveal significant differences across race for these peptides. (E) Schematic of MAPT domains in 4R tau structure. Schematic representation of the four main domains of MAPT: N‐terminal, PRD, MTBR, and C‐terminal domain. (F‐J) MAPT domains. Boxplots of log2 normalized abundance of MAPT (F) and its domains: N‐terminal (G), PRD (H), MTBR (I), and C‐terminal (J) across the same four groups (White CT, AA CT, White AD, AA AD). Data were adjusted for age and sex using the bootstrap method before analysis. MAPT and all its domains showed significantly higher abundance in AD cases compared to controls. The PRD had a slightly higher abundance in AA with AD compared to White with AD (p = 0.02), and the N‐terminal domain showed a slight increase in AA individuals, both in controls (p = 0.02) and AD cases (p = 0.03). The MTBR was the main contributor to the MAPT signal in AD cases. The p values from these analyses are marked on the graphs. AA, African American; AD, Alzheimer's disease; APP, amyloid precursor protein; CT, control; MAPT, microtubule‐associated protein tau; MTBR, microtubule binding region; PRD, proline‐rich domain; TMT‐MS, tandem mass tag mass spectrometry.

FIGURE 7

Global differential protein abundance between AD and control across racial groups. (A and B) Volcano plots displaying log2 FC (x‐axis) against one‐way ANOVA with Tukey correction‐derived −log10 p value (y‐axis) for all proteins (n  =  9180) comparing AD versus controls in White (A) and AA (B) proteomes. Proteins that are significantly more abundant in AD are presented in red, those significantly less abundant in blue, and proteins with non‐significant changes are presented in gray. (C) Scatter plot showing correlation between FC of all DAPs (n = 2819) found to be significant within the AA proteome (x‐axis) compared to the White proteome (y‐axis). The FCs were strongly correlated (bicor  =  0.9, p  < 1e‐200), regardless of whether the DAP was significant in one (yellow) or both proteomes (red: significantly higher abundance in AD, blue: significantly lower abundance in AD). (D‐G) Boxplots illustrating log2 FC relative abundance of representative proteins for four main categories across race and AD diagnosis: (D) DAPs in AD compared to control in both White and AA proteomes. (E) Proteins showing differential abundance across AD only in White individuals. (F) DAPs in only AA proteome across AD. (G) Proteins with a high abundance in AD compared with controls, independent of race. Boxes represent median and IQRs, and whiskers represent minimum and maximum data points within 1.5 times the IQR. P < 0.05 was considered significant. AA, African American; AD, Alzheimer's disease; Bicor, biweight midcorrelation; DAP, differentially abundant proteins; FC, fold change; IQR, interquartile range.