Spatial and single-nucleus transcriptomic analysis of genetic and sporadic forms of Alzheimer’s disease

Miyoshi & Morabito et al. 2024 (Nature Genetics)

The pathogenesis of Alzheimer’s disease (AD) depends on environmental and heritable factors, with its molecular etiology still unclear. Here we present a spatial transcriptomic (ST) and single-nucleus transcriptomic survey of late-onset sporadic AD and AD in Down syndrome (DSAD). Studying DSAD provides an opportunity to enhance our understanding of the AD transcriptome, potentially bridging the gap between genetic mouse models and sporadic AD. We identified transcriptomic changes that may underlie cortical layer-preferential pathology accumulation. Spatial co-expression network analyses revealed transient and regionally restricted disease processes, including a glial inflammatory program dysregulated in upper cortical layers and implicated in AD genetic risk and amyloid-associated processes. Cell–cell communication analysis further contextualized this gene program in dysregulated signaling networks. Finally, we generated ST data from an amyloid AD mouse model to identify cross-species amyloid-proximal transcriptomic changes with conformational context.

This repository contains the code used for data processing and analysis in our manuscript, and is generally organized in sync with the presentation of the data in the corresponding paper.

Data generated in this study

The raw and processed ST (10X Genomics Visium) and snRNA-seq(Parse Biosciences) datasets have been deposited on the NCBI Gene Expression Omnibus (GEO) at accession number GSE233208. Please contact the corresponding author of the paper (Vivek Swarup) with any queries related to the dataset.

IMPORTANT NOTE

There is an issue regarding the patient meta-data in the processed human visium dataset uploaded to GEO (GSE233208_Human_visium_ADDS_seurat_processed.rds). The correct patient meta-data can be found in Supplementary Table 1 of the paper. Here we show how to correct the meta-data in the Seurat object. We plan to update the GEO entry to fix the problem as soon as possible.

First, download GSE233208_Human_visium_ADDS_seurat_processed.rds.gz, and gunzip the file. Second, save just the patient meta-data table from Supplementary Table 1 as a separate tab-delimited file. Then follow the R code below.

library(Seruat)
library(tidyverse)
library(magrittr)

# load the seurat object
seurat_obj <- readRDS('GSE233208_Human_visium_ADDS_seurat_processed.rds')

# load the patient meta-data table from Supplementary Table 1
meta_df <- read.delim("adds_patient_data.txt", sep='\t', header=1) %>%
    dplyr::rename( Sample = Seq.ID)

# identify the common columns between the seurat object and the meta_df,
# and remove them from the seurat object
common_cols <- intersect(colnames(seurat_obj@meta.data), colnames(meta_df))
common_cols <- common_cols[common_cols != 'Sample'] # keep this one for the join operation
seurat_obj@meta.data %<>% select(-all_of(common_cols))

# join the two tables based on the Sample name to create an updated meta-data table
updated_meta <- dplyr::left_join(
    seurat_obj@meta.data, 
    meta_df, 
    by = 'Sample'
)
seurat_obj@meta.data <- updated_meta

# check the updated Diagnosis column to see if the numbers in each group are correct
patient_meta <- seurat_obj@meta.data %>%
    select(c(Sample, Diagnosis)) %>% distinct()
table(patient_meta$Diagnosis)

 AD   AD_DS Control earlyAD 
 10      10      10       9 

Processing sequencing data and quantifying gene expression

• Parse Biosciences snRNA-seq preprocessing scripts
• Human 10X Visium ST preprocessing scripts
• Mouse 10X Visium ST preprocessing scripts

Spatial and single-nucleus clustering analysis (Fig. 1)

snRNA-seq clustering analysis

New snRNA-seq data

• Clustering analysis for DSAD and Control data generated in this study
• Format dataset as a Seurat Object

Integration

• Integration with 3 published studies using scANVI
• Subclustering analysis of major cell lineages
• Format integrated dataset as a Seurat Object

Differential cell state analysis

• Differential cell state abundance analysis with MiloR

ST clustering analysis

Human dataset

• Load human spaceranger outputs and format as a Seurat Object
• Format human ST dataset as an anndata object, pt1
• Format human ST dataset as an anndata object, pt2
• BayesSpace clustering analysis in the human dataset

Mouse dataset

• Load mouse spaceranger outputs and format as a Seurat Object
• Format mouse ST dataset as an anndata object, pt1
• Format mouse ST dataset as an anndata object, pt2
• BayesSpace clustering analysis in the mouse dataset

Additional plotting

• Additional plotting scripts

Differential expression analysis (Figs. 2 and 4)

snRNA-seq differential expression

• Marker genes
• Disease vs Control DEG scripts
• Disease vs Control plotting and downstream analysis
• Female vs Male plotting and downstream analysis

ST differential expression

Human dataset

• Cluster marker genes
• Disease vs Control DEG scripts
• Disease vs Control plotting and downstream analysis
• Female vs Male DEG scripts
• Female vs Male plotting and downstream analysis

Mouse dataset

• Cluster marker genes
• 5xFAD vs WT DEG scripts
• 5xFAD vs WT plotting and downstream analysis
• Female vs Male DEG scripts
• Female vs Male plotting and downstream analysis

hdWGCNA co-expression network analysis (Fig. 3, Extended Data Fig. 9)

• hdWGCNA in human ST
• hdWGCNA in mouse ST

Spatial and single-nucleus genetic enrichment analysis (Fig. 3, Extended Data Fig. 4)

• Run scDRS in human snRNA datasets
• Plotting and downstream analysis of scDRS results in snRNA
• Run scDRS in human ST
• Plotting and downstream analysis of scDRS results in human ST
• Run scDRS in mouse ST
• Plotting and downstream analysis of scDRS results in mouse ST

Imaging mass cytometry (IMC) analysis (Fig. 5)

• IMC clustering analysis and downstream plotting

Predicting spatial coordinates for snRNA-seq data (Extended Data Fig. 5)

• Functions to run CellTrek in parallel for each snRNA-ST pair
• Running CellTrek in parallel for each snRNA-ST pair
• Downstream plotting script pt. 1
• Downstream plotting script pt. 2

Cell-cell communication (CCC) network analysis (Fig. 6)

• Cell-cell-communication network analysis with CellChat
• Cell-cell-communication network analysis with LIANA and comparing with CellChat

Amyloid-associated gene expression signatures (Fig. 7)

• Integration of amyloid imaging with ST datasets
• Human amyloid analysis
• Mouse amyloid analysis

TODO:

• Clean up code and analyis that we did which did not end up included in the paper.