GEO Accession viewer

NCBI > GEO > Accession Display

Not logged in | Login

GEO help: Mouse over screen elements for information.

Sample GSM7315477

Query DataSets for GSM7315477

Status

Public on Oct 18, 2023

Title

CD4 memory T, 16C237+34941, week 6, VDJ-derived cDNA

Sample type

SRA

Source name

PBMC

Organism

Macaca mulatta

Characteristics

tissue: PBMC
cell type: CD4 memory T cells
library type: mRNA
tags: none
treatment: mRNA-1273 vaccination
time: week 6

Extracted molecule

polyA RNA

Extraction protocol

Frozen rhesus macaque PBMCs were thawed into warm R10 media (RPMI + 10% Fetal Bovine Serum + 2 mL L-Glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin; all reagents from Gibco) containing DNase I (MilliporeSigma), followed by one wash with R10 and one wash with FACS buffer (PBS with 2% FBS). For B and innate cell staining, cells were resuspended in 100µL of Live/Dead Fixable Blue Dead Cell Stain Kit (Invitrogen, cat# L23105) diluted 1:200 in PBS for 10 min at room temperature. Cells were washed with FACS buffer and incubated for 20 min with the staining cocktail consisting of antibodies and probes. Additionally, each sample was labelled with 1µl of TotalSeq-C Hashtag antibodies (Biolegend) that was incubated together with the staining cocktail. After incubation, cells were washed twice with FACS buffer and resuspended in R10 for sort. From each sample, 10,000-60,000 innate cells were sorted into a tube containing FBS according to the gating strategy showed in Fig. S1B. From baseline samples, 50 thousand naïve B cells were also sorted in a separate tube. From week 2 and 6 samples, antigen-specific cells were single-cell sorted into 96-well plates containing 5 µL of TCL buffer (Qiagen) with 1% b-mercaptoethanol for sequencing by SmartSeq. All sorts were performed using a BD FACSymphony S6 Cell Sorter (BD Biosciences) with BD FACSDiva Software version 9.5.1 (BD Biosciences) and data were analyzed using Flowjo v10.8.1.
Frozen rhesus macaque PBMC were thawed into warm R10 media (RPMI + 10% Fetal Bovine Serum + 2 mL L-Glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin; all reagents from Gibco) containing DNase I (MilliporeSigma) and rested for 1 hour at 37oC / 5% CO2. Cells were stimulated with two peptide pools corresponding to S1 and S2 of the vaccine insert SARS-CoV-2 S protein (JPT Peptide Technologies) at 2 µg/mL of each peptide for 6 hours at 37oC / 5% CO2. A DMSO only control was included for each sample. Anti-CD154 BV421(Biolegend, clone 24-31, cat# 310824) was included during the 6-hour culture and GolgiStop (BD Biosciences) was added after 2 hours of stimulation. Following stimulation, cells were washed and stained with Aqua LIVE/DEAD dye (ThermoFisher) for 10 minutes, and subsequently stained with antibody mix. In addition, cells were incubated with TotalSeqTM Hashtag antibodies (Biolegend) with separate antibodies applied to the DMSO only condition and the stimulated conditions. For each animal, 1000-2000 memory CD4 cells from the DMSO only condition were sorted and combined with memory cells expressing CD154 and CD69 from each S pool stimulation for processing for 10X Genomics.
Bulk-sorted cells were pooled into a tube and loaded on the 10x Genomics Chromium Chip according to the manufacturer’s protocol for the Next GEM Single Cell 5’ Kit v1.1 (10x Genomics, PN-1000165). To sequence single-cell gene expression and cell surface oligonucleotides (from Hashtag antibodies), libraries were prepared according to manufacturer’s instructions using the Chromium Single Cell 5' Library Construction Kit (10x Genomics, PN-1000020) and Chromium Single Cell 5' Feature Barcode Library Kit (10x Genomics, PN-1000080), respectively. To sequence TCR repertoire libraries were prepared using the Chromium Single Cell V(D)J Enrichment Kit (10x Genomics, PN- PN-1000005). These libraries were sequenced on an Illumina NovaSeq platform (Illumina). To sequence BCR repertoire, heavy and light chains were amplified from the cDNA using IgG_REV, IgA_REV, IgK_REV or IgL_REV primers (Table S3) with the addition of Illumina sequences as detailed in (25). These Illumina-ready libraries were sequenced using 2x300 paired-end reads on an Illumina MiSeq
Full-length transcriptomes were sequenced using a modified SMART-Seq protocol. Heavy and light chain variable regions were enriched by amplifying cDNA with a forward primer complementary to the TSO and IgA/IgG_REV or IgK/IgL_REV primer pools.
OTHER: scRNA-seq with FeatureSeq and scAIRR-seq

Library strategy

RNA-Seq

Library source

transcriptomic single cell

Library selection

cDNA

Instrument model

NextSeq 2000

Description

TRA/TRB enriched from polyA RNA via 10x Chromimum

Data processing

The counts matrix was generated with Cellranger v3.1.0, using the protein-coding genes of the rhesus macaque reference genome from Ensembl (Mmul10, release #101). The filtered feature/barcode matrices were imported into Seurat v4.0.1 using R v4.0.4. Cells absent from either the gene expression or antibody capture datasets were eliminated, along with cells expressing zero counts of hash oligos. For hash demultiplexing, Seurat’s HTOdemux function was applied, and only singlets were retained for subsequent analyses. The filters based on percent mitochondrial genes and number of genes expressed varied between sample, but the upper threshold for percent mitochondrial genes ranged from 4-10%; the minimum for number of genes expressed was 200 genes for each sample; the upper threshold for number of genes expressed ranged from 1500-2500. The raw counts of each individual sample were merged. For the innate cell sorts, clusters of contaminating B and T cells (identified using Seurat’s FindClusters and RunUMAP) were removed and not included in downstream analysis. The merged and filtered data matrix was then corrected for sparseness using DeepImpute with tensorflow v2.1.0 and default parameters. The imputed matrix was then re-imported into Seurat. For both innate and T cells, Seurat was used to normalize the (imputed) count matrix, find variable genes, and regress out technical factors. Principle compents (PCs) were then calculated and used as input to Harmony to remove batch effects between samples. The integrated output was clustered using FindClusters and marker genes were identified with FindAllMarkers. Canonical marker genes were used to annotate the detected clusters. Gene set enrichment analysis of the innate cell data was conducted using the R library fgsea. The data were broken down by cell type and time point, and differential expression analysis of all 21,369 genes was conducted using FindMarkers between each subset. Finally, the statistically significant (adjusted p < 0.05) pathways were detected by inputting the genes, ranked by average log fold change, into the function fgsea. The Hallmark gene set databases originated from MSigDB. Signature scores for 50 pro-inflammatory genes previously determined to be predictive of antibody responses to a variety of vaccines were calculated using AddModuleScore in Seurat.
cellranger vdj (10x Genomics) was used to assign TCR annotations for each read-pair. TCR alpha recovery was minimal and not followed up further. An in-house shell script was used to filter for productive TCR betas. All subsequent analyses were done in R (v4.0.2). Cells with duplicated barcodes were discarded, and the TCR data was merged with the T cell transcriptomics Seurat (v4.0.1) object metadata for hash demultiplexing. Usage of each TRBV gene was compared between the S-specific and non-specific T cells from each animal using a paired samples Wilcoxon test in R.
SmartSeq sequencing reads were trimmed with Trimmomatic v 0.39, trimming nucleotides with Phred score below 20 and removing reads with resulting length less than 50 nucleotides. Alignment of reads to macaque protein-coding reference build Mmul10 was performed using Spliced Transcript to a Reference (STAR) v. 2.5.1b. Gene quantifications of reads as transcripts per million (TPM) were computed with Stringtie v. 1.33. Individual cells were removed from the analysis based on trimming and alignment metrics, as well as read quantity, feature quantity, quantity of mitochondrial reads, and diversity of transcripts. Seurat was used to calculate principal component analysis, which was used as input to integrate the data by batch with Harmony. PCA, cell clustering and UMAP dimensionality reductions were conducted in Seurat. A single cluster was identified as likely T cells by its more than triple the expression of CD3E than the next highest cluster and so was removed from the analysis. Clustering and dimensionality reduction was repeated after this cell filter. Differential expression of cell clusters was conducted with Seurat's function FindAllMarkers, and FindMarkers was used for comparison of timepoints week 2 and week 6. Those results were used as input to Gene set enrichment analysis conducted with the R package fgsea and the hallmark pathwaysfrom MSigDB.
Full-length V(D)J sequences from were reconstructed from SmartSeq data using BALDR and filtered using filterBALDR.pl (https://github.com/scharch/filterBALDR). Gene annotation and clonal assignment were done using SONAR v4.2 in single cell mode. Clonal flow diagrams were plotted using the ggalluvial package in R v4.2.1. Public clone were identified by clustering CDR H3 amino acid sequences using usearch as previously described . Members of the human public clone were identified from CovAbDab sequences downloaded as of January 24th, 2023 meeting the following criteria: (1) derived from IGHV3-30, IGHV3-30-3, or IGHV3-30-5; (2) 14 amino acid CDR H3 with Gly at position 6 and Tyr at position 8; (3) originating from a human; (4) not annotated as binding to any RBD or NTD epitope. This resulted in 326 public clone sequences, which were plotted using ggseqlogo.
Assembly: Mmul10
Supplementary files format and content: 10x Genomics output files: barcodes.tsv.gz, features.tsv.gz, matrix.mtx.gz
Supplementary files format and content: tab-delimited HTO assignments for each cell
Supplementary files format and content: comma-delimited values of gene expression in TPM
Supplementary files format and content: AIRR-formated VDJ sequences recovered from B cells
Supplementary files format and content: 10x Genomics output files: TR VDJ contigs with annotations
Supplementary files format and content: RDS-formatted R data files with complete Seurat objects

Submission date

May 09, 2023

Last update date

Oct 18, 2023

Contact name

Chaim A Schramm

Organization name

NIAID, NIH

Department

Vaccine Research Center

Lab

Genome Analysis Core