Building the mega single-cell transcriptome ocular meta-atlas

Feature selection methods affect the performance of scRNA-seq data integration and querying

The availability of single-cell transcriptomics has allowed the construction of reference cell atlases, but their usefulness depends on the quality of dataset integration and the ability to map new samples. Previous benchmarks have compared integration methods and suggest that feature selection improves performance but have not explored how best to select features. Here, we benchmark feature selection methods for single-cell RNA sequencing integration using metrics beyond batch correction and preservation of biological variation to assess query mapping, label transfer and the detection of unseen populations. We reinforce common practice by showing that highly variable feature selection is effective for producing high-quality integrations and provide further guidance on the effect of the number of features selected, batch-aware feature selection, lineage-specific feature selection and integration and the interaction between feature selection and integration models. These results are informative for analysts working on large-scale tissue atlases, using atlases or integrating their own data to tackle specific biological questions.

Systematic ocular phenotyping of 8,707 knockout mouse lines identifies genes associated with abnormal corneal phenotypes

PURPOSE

Corneal dysmorphologies (CDs) are typically classified as either regressive degenerative corneal dystrophies (CDtrs) or defective growth and differentiation-driven corneal dysplasias (CDyps). Both eye disorders have multifactorial etiologies. While previous work has elucidated many aspects of CDs, such as presenting symptoms, epidemiology, and pathophysiology, the genetic mechanisms remain incompletely understood. The purpose of this study was to analyze phenotype data from 8,707 knockout mouse lines to identify new genes associated with the development of CDs in humans.

METHODS

8,707 knockout mouse lines phenotyped by the International Mouse Phenotyping Consortium were queried for genes associated with statistically significant (P < 0.0001) abnormal cornea morphology to identify candidate CD genes. Corneal abnormalities were investigated by histopathology. A literature search was used to determine the proportion of candidate genes previously associated with CDs in mice and humans. Phenotypes of human orthologues of mouse candidate genes were compared with known human CD genes to identify protein-protein interactions and molecular pathways using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING), Protein Analysis Through Evolutionary Relationships (PANTHER), and Kyoto Encyclopedia of Genes and Genomes.

RESULTS

Analysis of data from 8,707 knockout mouse lines identified 213 candidate CD genes. Of these, 37 (17%) genes were previously known to be associated with CD, including 14 in the mouse, 16 in humans, and 7 in both. The remaining 176 (83%) genes have not been previously implicated in CD. We also searched publicly available RNAseq data and found that 131 of the total 213 (61.5%) were expressed in adult human corneal tissue. STRING analysis showed several interactions within and between candidate and established CD proteins. All cellular pathways of the established genes were found in the PANTHER analysis of the candidate genes. Several of the candidate genes were implicated in corneal disease, such as TGF-ß signaling. We also identified other possible underappreciated mechanisms relevant to the human cornea.

CONCLUSIONS

We identified 213 mouse genes that resulted in statistically significant abnormal corneal phenotypes in knockout mice, many of which have not previously been implicated in corneal pathology. Bioinformatic analyses implicated candidate genes in several signaling pathways which are potential therapeutic targets.

Single-cell sequencing in diabetic retinopathy: progress and prospects

Diabetic retinopathy is a major ocular complication of diabetes, characterized by progressive retinal microvascular damage and significant visual impairment in working-age adults. Traditional bulk RNA sequencing offers overall gene expression profiles but does not account for cellular heterogeneity. Single-cell RNA sequencing overcomes this limitation by providing transcriptomic data at the individual cell level and distinguishing novel cell subtypes, developmental trajectories, and intercellular communications. Researchers can use single-cell sequencing to draw retinal cell atlases and identify the transcriptomic features of retinal cells, enhancing our understanding of the pathogenesis and pathological changes in diabetic retinopathy. Additionally, single-cell sequencing is widely employed to analyze retinal organoids and single extracellular vesicles. Single-cell multi-omics sequencing integrates omics information, whereas stereo-sequencing analyzes gene expression and spatiotemporal data simultaneously. This review discusses the protocols of single-cell sequencing for obtaining single cells from retina and accurate sequencing data. It highlights the applications and advancements of single-cell sequencing in the study of normal retinas and the pathological changes associated with diabetic retinopathy. This underscores the potential of these technologies to deepen our understanding of the pathogenesis of diabetic retinopathy that may lead to the introduction of new therapeutic strategies.

Considerations for building and using integrated single-cell atlases

The rapid adoption of single-cell technologies has created an opportunity to build single-cell 'atlases' integrating diverse datasets across many laboratories. Such atlases can serve as a reference for analyzing and interpreting current and future data. However, it has become apparent that atlasing approaches differ, and the impact of these differences are often unclear. Here we review the current atlasing literature and present considerations for building and using atlases. Importantly, we find that no one-size-fits-all protocol for atlas building exists, but rather we discuss context-specific considerations and workflows, including atlas conceptualization, data collection, curation and integration, atlas evaluation and atlas sharing. We further highlight the benefits of integrated atlases for analyses of new datasets and deriving biological insights beyond what is possible from individual datasets. Our overview of current practices and associated recommendations will improve the quality of atlases to come, facilitating the shift to a unified, reference-based understanding of single-cell biology.

Deep learning-based models for preimplantation mouse and human embryos based on single-cell RNA sequencing

The rapid growth of single-cell transcriptomic technology has produced an increasing number of datasets for both embryonic development and in vitro pluripotent stem cell-derived models. This avalanche of data surrounding pluripotency and the process of lineage specification has meant it has become increasingly difficult to define specific cell types or states in vivo, and compare these with in vitro differentiation. Here we utilize a set of deep learning tools to integrate and classify multiple datasets. This allows the definition of both mouse and human embryo cell types, lineages and states, thereby maximizing the information one can garner from these precious experimental resources. Our approaches are built on recent initiatives for large-scale human organ atlases, but here we focus on material that is difficult to obtain and process, spanning early mouse and human development. Using publicly available data for these stages, we test different deep learning approaches and develop a model to classify cell types in an unbiased fashion at the same time as defining the set of genes used by the model to identify lineages, cell types and states. We used our models trained on in vivo development to classify pluripotent stem cell models for both mouse and human development, showcasing the importance of this resource as a dynamic reference for early embryogenesis.

The role of adiponectin and its receptor signaling in ocular inflammation-associated diseases

Ocular inflammation-associated diseases are leading causes of global visual impairment, with limited treatment options. Adiponectin, a hormone primarily secreted by adipose tissue, binds to its receptors, which are widely distributed throughout the body, exerting powerful physiological regulatory effects. The protective role of adiponectin in various inflammatory diseases has gained increasing attention in recent years. Previous studies have confirmed the presence of adiponectin and its receptors in the eyes. Furthermore, adiponectin and its analogs have shown potential as novel drugs for the treatment of inflammatory eye diseases. This article summarizes the evidence for the interplay between adiponectin and inflammatory eye diseases and provides new perspectives on the diagnostic and therapeutic possibilities of adiponectin.

SCAN: Spatiotemporal Cloud Atlas for Neural cells

The nervous system is one of the most complicated and enigmatic systems within the animal kingdom. Recently, the emergence and development of spatial transcriptomics (ST) and single-cell RNA sequencing (scRNA-seq) technologies have provided an unprecedented ability to systematically decipher the cellular heterogeneity and spatial locations of the nervous system from multiple unbiased aspects. However, efficiently integrating, presenting and analyzing massive multiomic data remains a huge challenge. Here, we manually collected and comprehensively analyzed high-quality scRNA-seq and ST data from the nervous system, covering 10 679 684 cells. In addition, multi-omic datasets from more than 900 species were included for extensive data mining from an evolutionary perspective. Furthermore, over 100 neurological diseases (e.g. Alzheimer's disease, Parkinson's disease, Down syndrome) were systematically analyzed for high-throughput screening of putative biomarkers. Differential expression patterns across developmental time points, cell types and ST spots were discerned and subsequently subjected to extensive interpretation. To provide researchers with efficient data exploration, we created a new database with interactive interfaces and integrated functions called the Spatiotemporal Cloud Atlas for Neural cells (SCAN), freely accessible at http://47.98.139.124:8799 or http://scanatlas.net. SCAN will benefit the neuroscience research community to better exploit the spatiotemporal atlas of the neural system and promote the development of diagnostic strategies for various neurological disorders.

Lamin A upregulation reorganizes the genome during rod photoreceptor degeneration

Neurodegenerative diseases are accompanied by dynamic changes in gene expression, including the upregulation of hallmark stress-responsive genes. While the transcriptional pathways that impart adaptive and maladaptive gene expression signatures have been the focus of intense study, the role of higher order nuclear organization in this process is less clear. Here, we examine the role of the nuclear lamina in genome organization during the degeneration of rod photoreceptors. Two proteins had previously been shown to be necessary and sufficient to tether heterochromatin at the nuclear envelope. The lamin B receptor (Lbr) is expressed during development, but downregulates upon rod differentiation. A second tether is the intermediate filament lamin A (LA), which is not normally expressed in murine rods. Here, we show that in the rd1 model of retinitis pigmentosa, LA ectopically upregulates in rod photoreceptors at the onset of degeneration. LA upregulation correlated with increased heterochromatin tethering at the nuclear periphery in rd1 rods, suggesting that LA reorganizes the nucleus. To determine how heterochromatin tethering affects the genome, we used in vivo electroporation to misexpress LA or Lbr in mature rods in the absence of degeneration, resulting in the restoration of conventional nuclear architecture. Using scRNA-seq, we show that reorganizing the nucleus via LA/Lbr misexpression has relatively minor effects on rod gene expression. Next, using ATAC-seq, we show that LA and Lbr both lead to marked increases in genome accessibility. Novel ATAC-seq peaks tended to be associated with stress-responsive genes. Together, our data reveal that heterochromatin tethers have a global effect on genome accessibility, and suggest that heterochromatin tethering primes the photoreceptor genome to respond to stress.

Single-cell reference mapping to construct and extend cell-type hierarchies

Single-cell genomics is now producing an ever-increasing amount of datasets that, when integrated, could provide large-scale reference atlases of tissue in health and disease. Such large-scale atlases increase the scale and generalizability of analyses and enable combining knowledge generated by individual studies. Specifically, individual studies often differ regarding cell annotation terminology and depth, with different groups specializing in different cell type compartments, often using distinct terminology. Understanding how these distinct sets of annotations are related and complement each other would mark a major step towards a consensus-based cell-type annotation reflecting the latest knowledge in the field. Whereas recent computational techniques, referred to as 'reference mapping' methods, facilitate the usage and expansion of existing reference atlases by mapping new datasets (i.e. queries) onto an atlas; a systematic approach towards harmonizing dataset-specific cell-type terminology and annotation depth is still lacking. Here, we present 'treeArches', a framework to automatically build and extend reference atlases while enriching them with an updatable hierarchy of cell-type annotations across different datasets. We demonstrate various use cases for treeArches, from automatically resolving relations between reference and query cell types to identifying unseen cell types absent in the reference, such as disease-associated cell states. We envision treeArches enabling data-driven construction of consensus atlas-level cell-type hierarchies and facilitating efficient usage of reference atlases.

PLAE Web App Enables Powerful Searching and Multiple Visualizations Across One Million Unified Single-Cell Ocular Transcriptomes

Purpose

To create a high-performance reactive web application to query single-cell gene expression data across cell type, species, study, and other factors.

Methods

We updated the content and structure of the underlying data (single cell Eye in a Disk [scEiaD]) and wrote the web application PLAE (https://plae.nei.nih.gov) to visualize and explore the data.

Results

The new portal provides quick visualization of over a million individual cells from vertebrate eye and body transcriptomes encompassing four species, 60 cell types, six ocular tissues, and 23 body tissues across 35 publications. To demonstrate the value of this unified pan-eye dataset, we replicated known neurogenic and cone macula markers in addition to proposing six new cone human region markers.

Conclusions

The PLAE web application offers the eye community a powerful and quick means to test hypotheses related to gene expression across a highly diverse, community-derived database.

Translational Relevance

The PLAE resource enables any researcher or clinician to study and research gene expression patterning across a wide variety of curated ocular cell types with a responsive web app.

Single-cell RNA sequencing in dissecting microenvironment of age-related macular degeneration: Challenges and perspectives

Age-related macular degeneration (AMD) is the leading cause of blindness in individuals over the age of 50 years, yet its etiology and pathogenesis largely remain uncovered. Single-cell RNA sequencing (scRNA-seq) technologies are recently developed and have a number of advantages over conventional bulk RNA sequencing techniques in uncovering the heterogeneity of complex microenvironments containing numerous cell types and cell communications during various biological processes. In this review, we summarize the latest discovered cellular components and regulatory mechanisms during AMD development revealed by scRNA-seq. In addition, we discuss the main challenges and future directions in exploring the pathophysiology of AMD equipped with single-cell technologies. Our review underscores the importance of multimodal single-cell platforms (such as single-cell spatiotemporal multi-omics and single-cell exosome omics) as new approaches for basic and clinical AMD research in identifying biomarker, characterizing cellular responses to drug treatment and environmental stimulation.

Cross-species scRNA-seq reveals the cellular landscape of retina and early alterations in type 2 diabetes mice

Single-cell RNA sequencing (scRNA-seq) analysis have provided an unprecedented resolution for the studies on diabetic retinopathy (DR). However, the early changes in the retina in diabetes remain unclear. A total of 8 human and mouse scRNA-seq datasets, containing 276,402 cells were analyzed individually to comprehensively delineate the retinal cell atlas. The neural retinas were isolated from the type 2 diabetes (T2D) and control mice, and scRNA-seq analysis was conducted to evaluate the early effects of diabetes on the retina. Bipolar cell (BC) heterogeneity were identified. We found some stable BCs across multiple datasets, and explored their biological functions. A new RBC subtype (Car8_RBC) in the mouse retina was validated using the multi-color immunohistochemistry. AC149090.1 was significantly upregulated in the rod cells, ON cone BCs (CBCs), OFF CBCs, and RBCs in T2D mice. Additionally, the interneurons, especially BCs, were the most vulnerable cells to diabetes by integrating scRNA-seq and genome-wide association studies (GWAS) analyses. In conclusion, this study delineated a cross-species retinal cell atlas and uncovered the early pathological alterations in the retina of T2D mice.

Single cell transcriptome analyses of the developing zebrafish eye- perspectives and applications

Within a relatively short period of time, single cell transcriptome analyses (SCT) have become increasingly ubiquitous with transcriptomic research, uncovering plentiful details that boost our molecular understanding of various biological processes. Stemming from SCT analyses, the ever-growing number of newly assigned genetic markers increases our understanding of general function and development, while providing opportunities for identifying genes associated with disease. SCT analyses have been carried out using tissue from numerous organisms. However, despite the great potential of zebrafish as a model organism, other models are still preferably used. In this mini review, we focus on eye research as an example of the advantages in using zebrafish, particularly its usefulness for single cell transcriptome analyses of developmental processes. As studies have already shown, the unique opportunities offered by zebrafish, including similarities to the human eye, in combination with the possibility to analyze and extract specific cells at distinct developmental time points makes the model a uniquely powerful one. Particularly the practicality of collecting large numbers of embryos and therefore isolation of sufficient numbers of developing cells is a distinct advantage compared to other model organisms. Lastly, the advent of highly efficient genetic knockouts methods offers opportunities to characterize target gene function in a more cost-efficient way. In conclusion, we argue that the use of zebrafish for SCT approaches has great potential to further deepen our molecular understanding of not only eye development, but also many other organ systems.

Cell-cell interactome of the hematopoietic niche and its changes in acute myeloid leukemia

The bone marrow (BM) is a complex microenvironment, coordinating the production of billions of blood cells every day. Despite its essential role and its relevance to hematopoietic diseases, this environment remains poorly characterized. Here we present a high-resolution characterization of the niche in health and acute myeloid leukemia (AML) by establishing a single-cell gene expression database of 339,381 BM cells. We found significant changes in cell type proportions and gene expression in AML, indicating that the entire niche is disrupted. We then predicted interactions between hematopoietic stem and progenitor cells (HSPCs) and other BM cell types, revealing a remarkable expansion of predicted interactions in AML that promote HSPC-cell adhesion, immunosuppression, and cytokine signaling. In particular, predicted interactions involving transforming growth factor β1 (TGFB1) become widespread, and we show that this can drive AML cell quiescence in vitro. Our results highlight potential mechanisms of enhanced AML-HSPC competitiveness and a skewed microenvironment, fostering AML growth.

An integrated cell atlas of the lung in health and disease

Single-cell technologies have transformed our understanding of human tissues. Yet, studies typically capture only a limited number of donors and disagree on cell type definitions. Integrating many single-cell datasets can address these limitations of individual studies and capture the variability present in the population. Here we present the integrated Human Lung Cell Atlas (HLCA), combining 49 datasets of the human respiratory system into a single atlas spanning over 2.4 million cells from 486 individuals. The HLCA presents a consensus cell type re-annotation with matching marker genes, including annotations of rare and previously undescribed cell types. Leveraging the number and diversity of individuals in the HLCA, we identify gene modules that are associated with demographic covariates such as age, sex and body mass index, as well as gene modules changing expression along the proximal-to-distal axis of the bronchial tree. Mapping new data to the HLCA enables rapid data annotation and interpretation. Using the HLCA as a reference for the study of disease, we identify shared cell states across multiple lung diseases, including SPP1+ profibrotic monocyte-derived macrophages in COVID-19, pulmonary fibrosis and lung carcinoma. Overall, the HLCA serves as an example for the development and use of large-scale, cross-dataset organ atlases within the Human Cell Atlas.

Spatial distribution of metabolites in the retina and its relevance to studies of metabolic retinal disorders

INTRODUCTION

The primate retina has evolved regional specialisations for specific visual functions. The macula is specialised towards high acuity vision and is an area that contains an increased density of cone photoreceptors and signal processing neurons. Different regions in the retina display unique susceptibility to pathology, with many retinal diseases primarily affecting the macula.

OBJECTIVES

To better understand the properties of different retinal areas we studied the differential distribution of metabolites across the retina.

METHODS

We conducted an untargeted metabolomics analysis on full-thickness punches from three different regions (macula, temporal peri-macula and periphery) of healthy primate retina.

RESULTS

Nearly half of all metabolites identified showed differential abundance in at least one comparison between the three regions. Furthermore, mapping metabolomics results from macula-specific eye diseases onto our region-specific metabolite distributions revealed differential abundance defining systemic metabolic dysregulations that were region specific.

CONCLUSIONS

The unique metabolic phenotype of different retinal regions is likely due to the differential distribution of different cell types in these regions reflecting the specific metabolic requirements of each cell type. Our results may help to better understand the pathobiology of retinal diseases with region specificity.

Chromatin remodeler Chd7 regulates photoreceptor development and outer segment length

Mutations in the chromatin remodeling factor CHD7 are the predominant cause of CHARGE syndrome, a congenital disorder that frequently includes ocular coloboma. Although CHD7 is known to be required for proper ocular morphogenesis, its role in retinal development has not been thoroughly investigated. Given that individuals with CHARGE syndrome can experience visual impairment even in the absence of coloboma, a better understanding of CHD7 function in the retina is needed. In this study, we characterized the expression pattern of Chd7 in the developing zebrafish and mouse retina and documented ocular and retinal phenotypes in Chd7 loss-of-function mutants. Zebrafish Chd7 was expressed throughout the retinal neuroepithelium when retinal progenitor cells were actively proliferating, and later in subsets of newly post-mitotic retinal cells. At stages of retinal development when most retinal cell types had terminally differentiated, Chd7 expression remained strong in the ganglion cell layer and in some cells in the inner nuclear layer. Intriguingly, strong expression of Chd7 was also observed in the outer nuclear layer where it was co-expressed with markers of post-mitotic cone and rod photoreceptors. Expression of mouse CHD7 displayed a similar pattern, including expression in the ganglion cells, subsets of inner nuclear layer cells, and in the distal outer nuclear layer as late as P15. Two different mutant chd7 zebrafish lines were characterized for ocular and retinal defects. These mutants displayed microphthalmia, reduced numbers of cone photoreceptors, and truncated rod and cone photoreceptor outer segments. Reduced cone photoreceptor number and abnormal outer segments were also observed in heterozygous Chd7 mutant mice. Taken together, our results in zebrafish and mouse reveal a conserved, previously undescribed role for Chd7 in retinal development and photoreceptor outer segment morphogenesis. Moreover, our work suggests an avenue of future investigation into the pathogenesis of visual system defects in CHARGE syndrome.

Transcriptional mapping of the macaque retina and RPE-choroid reveals conserved inter-tissue transcription drivers and signaling pathways

The macula and fovea comprise a highly sensitive visual detection tissue that is susceptible to common disease processes like age-related macular degeneration (AMD). Our understanding of the molecular determinants of high acuity vision remains unclear, as few model organisms possess a human-like fovea. We explore transcription factor networks and receptor-ligand interactions to elucidate tissue interactions in the macula and peripheral retina and concomitant changes in the underlying retinal pigment epithelium (RPE)/choroid. Poly-A selected, 100 bp paired-end RNA-sequencing (RNA-seq) was performed across the macular/foveal, perimacular, and temporal peripheral regions of the neural retina and RPE/choroid tissues of four adult Rhesus macaque eyes to characterize region- and tissue-specific gene expression. RNA-seq reads were mapped to both the macaque and human genomes for maximum alignment and analyzed for differential expression and Gene Ontology (GO) enrichment. Comparison of the neural retina and RPE/choroid tissues indicated distinct, contiguously changing gene expression profiles from fovea through perimacula to periphery. Top GO enrichment of differentially expressed genes in the RPE/choroid included cell junction organization and epithelial cell development. Expression of transcriptional regulators and various disease-associated genes show distinct location-specific preference and retina-RPE/choroid tissue-tissue interactions. Regional gene expression changes in the macaque retina and RPE/choroid is greater than that found in previously published transcriptome analysis of the human retina and RPE/choroid. Further, conservation of human macula-specific transcription factor profiles and gene expression in macaque tissues suggest a conservation of programs required for retina and RPE/choroid function and disease susceptibility.

Eyes on CHARGE syndrome: Roles of CHD7 in ocular development

The development of the vertebrate visual system involves complex morphogenetic interactions of cells derived from multiple embryonic lineages. Disruptions in this process are associated with structural birth defects such as microphthalmia, anophthalmia, and coloboma (collectively referred to as MAC), and inherited retinal degenerative diseases such as retinitis pigmentosa and allied dystrophies. MAC and retinal degeneration are also observed in systemic congenital malformation syndromes. One important example is CHARGE syndrome, a genetic disorder characterized by coloboma, heart defects, choanal atresia, growth retardation, genital abnormalities, and ear abnormalities. Mutations in the gene encoding Chromodomain helicase DNA binding protein 7 (CHD7) cause the majority of CHARGE syndrome cases. However, the pathogenetic mechanisms that connect loss of CHD7 to the ocular complications observed in CHARGE syndrome have not been identified. In this review, we provide a general overview of ocular development and congenital disorders affecting the eye. This is followed by a comprehensive description of CHARGE syndrome, including discussion of the spectrum of ocular defects that have been described in this disorder. In addition, we discuss the current knowledge of CHD7 function and focus on its contributions to the development of ocular structures. Finally, we discuss outstanding gaps in our knowledge of the role of CHD7 in eye formation, and propose avenues of investigation to further our understanding of how CHD7 activity regulates ocular and retinal development.

Gene Expression Data for Investigating Glaucoma Treatment Options and Pharmacology in the Anterior Segment, State-of-the-Art and Future Directions

Glaucoma treatment options as well as its etiology are far from understood. Gene expression (transcriptomics) data of the anterior segment of the eye can help by elucidating the molecular-mechanistic underpinnings, and we present an up-to-date description and discussion of what gene expression data are publicly available, and for which purposes these can be used. We feature the few resources covering all segments of the eye, and we then specifically focus on the anterior segment, and provide an extensive list of the Gene Expression Omnibus data that may be useful. We also feature single-cell data of relevance, particularly three datasets from tissues of relevance to aqueous humor outflow. We describe how the data have been used by researchers, by following up resource citations and data re-analyses. We discuss datasets and analyses pertaining to fibrosis following glaucoma surgery, and to glaucoma resulting from the use of steroids. We conclude by pointing out the current lack and underutilization of ocular gene expression data, and how the state of the art is expected to improve in the future.