Long-lived yeast-derived lipids promote longevity in Drosophila

To investigate how lipid products from engineered yeasts influence gut physiology and lifespan, we performed single-nucleus RNA sequencing (snRNA-seq) on gut tissues from 45-day-old flies fed with lipid extracts from different yeast strains, or a control diet without lipid supplementation, strictly following the Fly Cell Atlas protocol. Paired brain tissues were also collected, given the critical role of the gut-brain axis in Drosophila aging.

1. snRNA-seq data processing and analysis

Raw sequencing reads were aligned to the Drosophila melanogaster reference genome (FlyBase r6.31) using a pre-mRNA GTF annotation established by the Fly Cell Atlas. Alignment, barcode processing, and UMI counting were performed with Cell Ranger V6.1.2 (10x Genomics) to generate a digital gene expression matrix. Subsequent data analysis was conducted using the Seurat V4 pipeline. Low-quality nuclei and potential doublets were filtered based on gene count distributions. Specifically, nuclei with 200–1,600 detected genes were retained for brain tissues, whereas the upper threshold was extended to 3,000 genes for gut tissues due to their inherently higher gene content.

2. Unsupervised clustering of gut and brain dataset

Standard Seurat workflow functions were applied in the following order: NormalizeData, FindVariableFeatures, ScaleData, RunPCA, and ElbowPlot to determine the dimensionality of the dataset, followed by FindNeighbors, FindClusters (at a resolution of 0.2 for the gut dataset and 0.1 for the brain dataset), and RunUMAP for visualization. Differentially expressed genes were identified using the FindMarkers function through pairwise comparisons of clusters or yeast strain groups. Module scores for predefined gene sets were calculated using the AddModuleScore function to quantify expression signatures across nuclei.

Scripts are included in the "Unsupervised clustering" folder.

3. Gut-Brain ligand–receptor (L–R) interaction analysis

Cell–cell connectivity patterns were analyzed using the R package Connectome V1.0.0 in “custom mapping” mode, based on ligand and receptor expression derived from our snRNA-seq datasets. A high-confidence list of L–R pairs covering major Drosophila signaling pathways was curated from FlyPhoneDB.

Scripts are included in the "Gut-Brain L–R interaction analysis" folder.

4. Fly metabolic analysis pipeline (FLY-MAP)

We developed a computational pipeline, FLY-MAP, to extract and analyze metabolic profiles from the snRNA-seq dataset. Specifically, a metabolic matrix was constructed by extracting genes associated with Drosophila KEGG metabolic pathways, encompassing 11 major categories and 81 pathways, from the original whole-transcriptome gene expression matrix. Unsupervised clustering based on this metabolic matrix was then performed separately for gut and brain nuclei, enabling an unbiased identification of metabolic programs across tissues and dietary conditions.

Scripts are included in the "FLY-MAP" folder.