TL;DR
This study creates a comprehensive single-cell atlas of the human retina's RPE and choroid, revealing cell-type-specific gene expression and chromatin accessibility. This resource is used to pinpoint specific cell types and regulatory elements affected by genetic variants associated with age-related macular degeneration (AMD), nominating novel pathogenic variants and mechanisms.
What Problem Does This Paper Address?
Age-related macular degeneration (AMD) is a major cause of vision loss, and while genome-wide association studies (GWAS) have identified numerous genetic risk loci, most of these variants are non-coding. This makes it challenging to understand their functional impact, identify the specific cell types they affect, and determine if they are directly causal for the disease. The paper aims to bridge this gap by providing a detailed, cell-type-resolved view of the RPE and choroid, enabling a systematic analysis of how AMD-associated variants might contribute to disease pathogenesis.
Methods
The researchers employed a multiomic approach, integrating several cutting-edge techniques:
- Joint Single-Cell RNA Sequencing (scRNA-seq) and Single-Cell Assay for Transposase-Accessible Chromatin Sequencing (scATAC-seq): This allows for simultaneous measurement of gene expression and chromatin accessibility at the single-cell level, providing a comprehensive view of cellular states and regulatory landscapes.
- HiChIP: Used to map 3D genome architecture and identify physical interactions between enhancers and their target genes.
- Activity-by-Contact (ABC) Modeling: A computational method to predict enhancer-gene connections based on chromatin accessibility and Hi-C data.
- Allele-Specific STARR-seq (Self-Transcribing Active Regulatory Region Sequencing): A high-throughput assay to functionally test the regulatory activity of specific genetic variants in RPE cells, including under conditions mimicking complement activation.
- ChromBPNet: A deep learning model used to predict the effect of variants on chromatin accessibility.
- eQTL Analysis: Integration with existing data to identify variants associated with gene expression changes.
Datasets
The study utilized postmortem human eyes from nine individuals, including four eyes from three individuals diagnosed with dry AMD. The generated datasets include:
- Single-cell RNA-seq and scATAC-seq data from RPE and choroid complexes.
- H3K27ac HiChIP data from RPE and choroid.
- Publicly available datasets for comparison and integration (e.g., scRNA-seq data for photoreceptors and glia, H3K27ac ChIP-seq and Hi-C data for RPE and choroid, eQTL data).
Key Insights and Ideas
- Cellular Heterogeneity is Key: The RPE and choroid are not monolithic tissues but comprise diverse cell types, each with unique regulatory landscapes. Understanding AMD variants requires this cell-type resolution.
- Non-coding Variants are Functionally Relevant: The study strongly implicates non-coding variants, often overlooked, in AMD pathogenesis by linking them to specific regulatory elements and target genes.
- Complement Activation Matters: The research highlights the role of complement activation in modulating the regulatory activity of some AMD-associated SNPs, suggesting a link to inflammatory processes in AMD.
- Multiomic Integration is Powerful: Combining diverse data types (genomics, epigenomics, transcriptomics) is crucial for robustly nominating causal variants and understanding complex regulatory networks.
Key Results
- The researchers generated a single-cell atlas of 39,326 human RPE and choroid cells, identifying nine distinct cell types.
- They mapped over 241,000 chromatin accessibility peaks, with a significant portion being cell-type-specific.
- Over 63,000 enhancer-gene pairs were nominated using HiChIP and ABC modeling.
- Functional testing of AMD risk variants using STARR-seq identified 66 and 40 functional SNPs in naive and complement-activated RPE cells, respectively.
- Integration of multiple prioritization criteria nominated 59 variants with strong evidence for regulatory activity in RPE and choroid.
- Specific causal variants were nominated for several AMD risk loci, including rs943080 (linked to VEGFA expression) and rs2236416 (linked to PLTP expression in macrophages).
Limitations and Future Work
- Low Cell Numbers for Certain Types: The study had low numbers of pericytes and B cells, limiting the identification of scATAC-seq peaks for these cell types.
- Limited AMD Samples: Only four eyes with AMD were profiled, potentially reducing the ability to detect subtle differences between AMD and control tissues.
- Macular vs. Peripheral Choroid: The origin of choroid cell types (macular vs. peripheral) could not be determined.
Future work could involve expanding the atlas with more cells and donors to address these limitations and uncover more subtle findings. Further validation of the nominated causal variants through gene editing or other functional experiments is also a promising direction.
Summary updated