CoChat
Sign Up
Cited by 338Single-cell and spatial transcriptomicsRetinal Development and DisordersRetinal Diseases and Treatments

The DOI matches its Crossref record; 44 of the 44 references it cites were found in a registry or an index.

Single-cell transcriptomic atlas of the human retina identifies cell types associated with age-related macular degeneration

Open access
Menon, Madhvi, Mohammadi, Shahin, Davila-Velderrain, Jose, Goods, Brittany A., Cadwell, Tanina D., Xing, Yu, Stemmer-Rachamimov, Anat, Shalek, Alex K., Love, John Christopher, Kellis, Manolis, Hafler, Brian P.·2019·Nature Communications
DOI 10.1038/s41467-019-12780-8PMC PMC6814749
Open this paper in CoChat — free accountOpen PDF

Abstract

Genome-wide association studies (GWAS) have identified genetic variants associated with age-related macular degeneration (AMD), one of the leading causes of blindness in the elderly. However, it has been challenging to identify the cell types associated with AMD given the genetic complexity of the disease. Here we perform massively parallel single-cell RNA sequencing (scRNA-seq) of human retinas using two independent platforms, and report the first single-cell transcriptomic atlas of the human retina. Using a multi-resolution network-based analysis, we identify all major retinal cell types, and their corresponding gene expression signatures. Heterogeneity is observed within macroglia, suggesting that human retinal glia are more diverse than previously thought. Finally, GWAS-based enrichment analysis identifies glia, vascular cells, and cone photoreceptors to be associated with the risk of AMD. These data provide a detailed analysis of the human retina, and show how scRNA-seq can provide insight into cell types involved in complex, inflammatory genetic diseases.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-019-12780-8

Cites

44 of 44 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Single-cell transcriptomic analysis reveals human retina diversity. a Study design and sample preparation. Postmortem human retinas were enzymatically dissociated and single-cells were isolated. cDNA single-cell libraries were generated and sequenced. We profiled 20,091 cells across the retinas of three normal individuals using a droplet-based microfluidics scRNA-seq platform. b Sketch of retina cross-section showing layers and major cell types. c Cell-to-cell similarity network of retinal cells. d Average expression of known cell-type marker genes across cell groups. e Projection of known cell-type-specific marker gene expression across cell groups. d, e show localization of distinct cell types within the network, identifying neighborhoods of rods, cones, retinal ganglion cells, bipolar cells, amacrine cells, horizontal cells, macroglia (Müller glia and astrocytes), microglia, and vascular cells
Fig. 1 Single-cell transcriptomic analysis reveals human retina diversity. a Study design and sample preparation. Postmortem human retinas were enzymatically dissociated and single-cells were isolated. cDNA single-cell libraries were generated and sequenced. We profiled 20,091 cells across the retinas of three normal individuals using a droplet-based microfluidics scRNA-seq platform. b Sketch of retina cross-section showing layers and major cell types. c Cell-to-cell similarity network of retinal cells. d Average expression of known cell-type marker genes across cell groups. e Projection of known cell-type-specific marker gene expression across cell groups. d, e show localization of distinct cell types within the network, identifying neighborhoods of rods, cones, retinal ganglion cells, bipolar cells, amacrine cells, horizontal cells, macroglia (Müller glia and astrocytes), microglia, and vascular cellsCC BY · PubMed Central
Fig. 2 Heatmap of top marker genes within retinal cell types using microfluidics scRNA-seq. The top differentially expressed genes were selected using the one-sided Wilcoxon rank-sum test, and ranked based on their discriminating scores within each identified cell type. a Imputed gene expression, using network diffusion, across all cells, b Average of raw expression of each gene within each cell-type. In both panels, rows correspond to genes and columns to cell-types, which are color-coded (color bar, top)
Fig. 2 Heatmap of top marker genes within retinal cell types using microfluidics scRNA-seq. The top differentially expressed genes were selected using the one-sided Wilcoxon rank-sum test, and ranked based on their discriminating scores within each identified cell type. a Imputed gene expression, using network diffusion, across all cells, b Average of raw expression of each gene within each cell-type. In both panels, rows correspond to genes and columns to cell-types, which are color-coded (color bar, top)CC BY · PubMed Central
Fig. 3 Heterogeneity within Müller glia and astrocytes in the human retina. a Cell-to-cell similarity network (ACTIONet) representation of the filtered set of Müller glia and astrocytes, highlighting three major subtypes. b Scaled heatmap of top differentially expressed genes within different cell groups. c Dot plot showing the row-normalized (z-score) mean expression of the top-ranked genes across subtypes of macroglial cells. d Imputed expression of key marker genes, normalized by the average of their expression across all macroglial cells. ***p < 0.001, one-sided Wilcoxon rank-sum test. Box plots are centered around the median, with the interquartile range (IQR) defining the box. The upper whisker extends to the largest value and the lower whisker extends to the smallest value, no further than 1.5 × IQR from the end of the box. e–g Representative images of multiplex in situ hybridization of e FTL (labeled in yellow) and FOS (labeled in red), f FTL (labeled in yellow) and COL4A3 (labeled in red), and g FOS (labeled in yellow) and COL4A3 (labeled in red), with immunofluorescence of glutamine synthetase (GS, identified in green). Bar graphs showing frequencies of e FOS+, FTL+, or FOS+FTL+ GS+ macroglia in the macula retina, f COL4A3+, FTL+, or COL4A3+FTL+ GS+ macroglia in the macula retina, and g COL4A3+, FOS+, or COL4A3+FOS+ GS+ macroglia in the peripheral retina. Error bars show standard error of mean. White circles indicate positive cells. Data are representative of three independent experiments. h Bar graph showing frequency of each subset within the macula vs. peripheral region of the retina. i Violin plot showing gene expression differences of COL4A3 between the macula and peripheral retina. Difference is statistically significant with p < 0.01, one-sided Wilcoxon rank-sum test. j–k Representative images of multiplex in situ hybridization of COL4A3 (labeled in red) with immunofluorescence of glutamine synthetase (GS, identified in green) in the j macula, and k peripheral retina. White circles indicate COL4A3+ cells, and white asterisk indicates COL4A3− cells. Data is representative of three independent experiments
Fig. 3 Heterogeneity within Müller glia and astrocytes in the human retina. a Cell-to-cell similarity network (ACTIONet) representation of the filtered set of Müller glia and astrocytes, highlighting three major subtypes. b Scaled heatmap of top differentially expressed genes within different cell groups. c Dot plot showing the row-normalized (z-score) mean expression of the top-ranked genes across subtypes of macroglial cells. d Imputed expression of key marker genes, normalized by the average of their expression across all macroglial cells. ***p < 0.001, one-sided Wilcoxon rank-sum test. Box plots are centered around the median, with the interquartile range (IQR) defining the box. The upper whisker extends to the largest value and the lower whisker extends to the smallest value, no further than 1.5 × IQR from the end of the box. e–g Representative images of multiplex in situ hybridization of e FTL (labeled in yellow) and FOS (labeled in red), f FTL (labeled in yellow) and COL4A3 (labeled in red), and g FOS (labeled in yellow) and COL4A3 (labeled in red), with immunofluorescence of glutamine synthetase (GS, identified in green). Bar graphs showing frequencies of e FOS+, FTL+, or FOS+FTL+ GS+ macroglia in the macula retina, f COL4A3+, FTL+, or COL4A3+FTL+ GS+ macroglia in the macula retina, and g COL4A3+, FOS+, or COL4A3+FOS+ GS+ macroglia in the peripheral retina. Error bars show standard error of mean. White circles indicate positive cells. Data are representative of three independent experiments. h Bar graph showing frequency of each subset within the macula vs. peripheral region of the retina. i Violin plot showing gene expression differences of COL4A3 between the macula and peripheral retina. Difference is statistically significant with p < 0.01, one-sided Wilcoxon rank-sum test. j–k Representative images of multiplex in situ hybridization of COL4A3 (labeled in red) with immunofluorescence of glutamine synthetase (GS, identified in green) in the j macula, and k peripheral retina. White circles indicate COL4A3+ cells, and white asterisk indicates COL4A3− cells. Data is representative of three independent experimentsCC BY · PubMed Central
Fig. 4 Cell-type specificity of AMD-associated genes using microfluidics scRNA-seq. a Gene expression specificity of 585 genes previously associated with the 34 AMD locus regions defined by the 52 identified variants and their proxies (r2 ≥ 0.5, ±500 kb), reported in Fritsche et al.37. Specificity scores are computed using one-sided Wilcoxon rank-sum test, and reported as the log-transformed value of the adjusted p-values. b Gene expression enrichment scores for the leading AMD genes, defined for each locus as the gene nearest to the lead variant (variant with smallest P value), as reported in Fritsche et al.37. c–e Expression pattern of AMD risk alleles CFI and TIMP3 in the human retina. Representative images of multiplex in situ hybridization with immunofluorescence of c APOE (labeled in yellow, identifies Müller glia), CFI (labeled in red) and glutamine synthetase (GS, identified in green); d APOE (labeled in yellow, identifies Müller glia), TIMP3 (labeled in red) and glutamine synthetase (GS, identified in green), and e TM4SF1 (labeled in yellow, identifies vascular endothelium) and TIMP3 (labeled in red). White dotted circles indicate double positive cells. Data are representative of three independent experiments
Fig. 4 Cell-type specificity of AMD-associated genes using microfluidics scRNA-seq. a Gene expression specificity of 585 genes previously associated with the 34 AMD locus regions defined by the 52 identified variants and their proxies (r2 ≥ 0.5, ±500 kb), reported in Fritsche et al.37. Specificity scores are computed using one-sided Wilcoxon rank-sum test, and reported as the log-transformed value of the adjusted p-values. b Gene expression enrichment scores for the leading AMD genes, defined for each locus as the gene nearest to the lead variant (variant with smallest P value), as reported in Fritsche et al.37. c–e Expression pattern of AMD risk alleles CFI and TIMP3 in the human retina. Representative images of multiplex in situ hybridization with immunofluorescence of c APOE (labeled in yellow, identifies Müller glia), CFI (labeled in red) and glutamine synthetase (GS, identified in green); d APOE (labeled in yellow, identifies Müller glia), TIMP3 (labeled in red) and glutamine synthetase (GS, identified in green), and e TM4SF1 (labeled in yellow, identifies vascular endothelium) and TIMP3 (labeled in red). White dotted circles indicate double positive cells. Data are representative of three independent experimentsCC BY · PubMed Central
Fig. 5 Cell-type trait association analyses. Bars show the mean strength of association computed with MAGMA (-log10P) for AMD (a) Alzheimer’s Disease (b), and Diabetes Mellitus (c) using cell type preferential gene expression in either the human retina microfluidics scRNA-seq dataset (left) or the human cortex scRNA-seq dataset as a control (right). The vertical line indicates whether the cell-type is significantly associated with the trait (P < 0.05, MAGMA gene-set analysis test, linear regression)
Fig. 5 Cell-type trait association analyses. Bars show the mean strength of association computed with MAGMA (-log10P) for AMD (a) Alzheimer’s Disease (b), and Diabetes Mellitus (c) using cell type preferential gene expression in either the human retina microfluidics scRNA-seq dataset (left) or the human cortex scRNA-seq dataset as a control (right). The vertical line indicates whether the cell-type is significantly associated with the trait (P < 0.05, MAGMA gene-set analysis test, linear regression)CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Cited by

Sorted by citations

See all 338

Related

See all

Metadata

Authors
Menon, M., Mohammadi, S., Davila-Velderrain, J., Goods, B. A., Cadwell, T. D., Xing, Y., Stemmer-Rachamimov, A., Shalek, A. K., Love, J. C., Kellis, M., & Hafler, B. P.
Affiliations
  • Broad Institute (US)
  • Harvard University (US)
  • Massachusetts General Hospital (US)
Corresponding author
Hafler, Brian P.
ORCID
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2019-10-25
Volume
10
Issue
1
Pages
4902-4902
ISSN
2041-1723 (ISSN-L)
DOI
10.1038/s41467-019-12780-8
PMC
PMC6814749
OpenAlex
W2981346196
Cited by
331
References
43
Citations by year
2026: 36 · 2025: 45 · 2024: 56 · 2023: 42 · 2022: 70 · 2021: 50 · 2020: 30 · 2019: 1
FWCI
12.57
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • Research Explorer (The University of Manchester) (submittedVersion, cc-by)
  • Nature
Funders
U.S. Department of Health & Human Services | National Institutes of Health (AI039671, EY026652)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
  • Health Sciences › Medicine › Ophthalmology › Retinal Development and Disorders
  • Health Sciences › Medicine › Ophthalmology › Retinal Diseases and Treatments
Keywords
age-related macular degeneration, human retina, retinal cell types, single-cell transcriptomic atlas, macroglia, cone photoreceptors, gene expression signatures, retinal inflammation
MeSH terms
Astrocytes, Blood Vessels, Humans, Macular Degeneration, Neuroglia, Retina, Retinal Ganglion Cells, Retinal Vessels, Gene Expression, Sequence Analysis, RNA, Microglia, Retinal Rod Photoreceptor Cells, Retinal Cone Photoreceptor Cells, Genetic Predisposition to Disease, Gene Expression Profiling, Amacrine Cells, Retinal Bipolar Cells, Retinal Horizontal Cells, Retinal Neurons, Single-Cell Analysis, High-Throughput Nucleotide Sequencing, Ependymoglial Cells

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review