Project name: Single-cell RNA sequencing of zebrafish beta-cells from various stages

Description: Age-associated deterioration of cellular physiology leads to pathological conditions, and detection of premature aging could provide a window for preventive therapies against age-related diseases. For this, methods that accurately evaluate cellular age are required. However, such techniques are currently limited and based on post hoc evaluation using a limited set of histological markers. Development of a technique capable of predicting cellular age, and its modifiers, requires a framework that can robustly handle the noise associated with single-cell sampling protocols. Here, we implement GERAS (GEnetic Reference for Age of Single-cell), a machine learning based framework capable of assigning individual cells to chronological stages based on their transcriptomes. GERAS displayed greater than 90% accuracy in predicting the chronological stage of zebrafish beta-cells and human pancreatic cells. The framework demonstrates robustness against biological and technical noise, as evaluated by its performance on independent samplings of single-cells. Additionally, GERAS enabled the evaluation of differences in calorie intake and body mass index on the aging of zebrafish and human cells, respectively. We further harnessed the predictive power of GERAS to identify genome-wide molecular factors that correlate with aging. We show that one of these factors, junb, which declines in expression with aging, is necessary to maintain the proliferative state of juvenile beta-cells. Our results showcase the applicability of a machine learning framework to predict the chronological stage of heterogeneous cell populations. The study demonstrates the utility of stage classifiers in assessing pro-aging factors, and uncovering candidate genes associated with premature aging.Overall design: We used fluorescence-activated cell sorting (FACS) coupled with next generation RNA-Sequencing to profile beta-cells from various stages. Cells were sorted into a 96-well plates and single-cell library prepared using SMART-Seq v4 Ultra Low Input RNA Kit . Sequencing was performed on llumina Nextseq500 aiming at an average sequencing depth of 0.5 million reads per cell. Reads were splice-aligned to the zebrafish genome, GRCz10, using HISAT2. htseq-count was used to assign reads to exons thus eventually getting counts per gene.

Data type: Transcriptome or Gene expression

Sample scope: Multiisolate

Relevance: ModelOrganism

Last updated: 2018-01-30