The Brian Hafler Lab

Our lab studies cellular mechanisms underlying human retinal diseases. These include age-related macular degeneration (AMD), glaucoma, and stem cell regeneration using machine learning methods such as single-cell transcriptomics, spatial transcriptomics, and single-cell epigenomics. The goal of the research is to identify cellular mechanisms underlying human macular degeneration and glaucoma that can be applied to novel therapies with a focus on neuroinflammation to prevent neuronal death and help preserve vision.

Molecular pathways leading to neurodegeneration in human disease

A) A sketch of the human retina. The retina is a thin tissue that lies on the inner part of the eyeball that processes light into neural signals that are transmitted through the optic nerve to the brain. B) Age-related macular degeneration has an early dry stage where there is accumulation of drusen debris beneath the retina. Drusen lead to activation of the innate immune response and neuroinflammation. During the late advanced stage of age-related macular degeneration, pathological new blood vessel formation occurs that can lead to permanent vision loss through the death of photoreceptors. Accumulation of extracellular plaques and intracellular neurofibrillary tangles in Alzheimer's disease and myelin damage in progressive multiple sclerosis are similarly accompanied by microglia (blue) and astrocyte (orange) activation. C) In collaboration with Smita Krishnaswamy’s lab in the Department of Genetics and Computer Science, an algorithm diffusion condensation was developed and applied to new age-related macular degeneration single-cell data. This panels shows a visual description of the cellular condensation process undertaken by diffusion condensation across four granularities. Points are moved to and merged with their nearest neighbors as determined by a weighted random walk over the data graph. Over many successive iterations, cells collapse, denoting cluster identity at various iterations. D) The process described in (C) creates hundreds of granularities of clusters whic

Age-related macular degeneration and glaucoma are progressive neurodegenerative disease of the retina that affect more than 200 million individuals worldwide and are leading causes of incurable blindness. Despite intense efforts, the cell types and molecular pathways that promote neuroinflammation in AMD remain poorly understood. Our research tackles this problem by combining novel computational tools and machine learning to provide an unparalleled depth of insight into key inflammatory pathways residing in microglia and macroglia. We are implementing an approach that utilizes single nuclei expression data and a new field of machine learning called manifold learning to identify and characterize the rare microglial and macroglial subtypes driving pathology in AMD. This integrative approach offers an advantage over traditional approaches, as it allows data integration of rare cellular populations on a scale that was not previously possible. Leveraging our special access to human tissue with AMD through Yale, this innovative research allows for the comparison of gene signatures of microglia between patients with AMD and healthy individuals, thereby elucidating the mechanism of AMD pathogenesis. This research has high potential, as researchers have yet to find effective interventions for the dry form of AMD beyond nutritional supplements. However, it is also high reward, as it has the potential to transform human health and lay the foundation for novel therapies for AMD that reduce inflammation and prevent blindness in the elderly.

Single-cell transcriptomic analysis of the human retina. 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 (Mu¨ller glia and astrocytes), microglia, and vascular cells. Figure from (Menon M, Mohammadi S, Davila-Velderrain J, Goods BA, Cadwell TD, Xing Y, Stemmer-Rachamimov A, Shalek AK, Love JC, Kellis M, Hafler BP. Single-cell transcriptomic atlas of the human retina identifies cell types associated with age-related macular degeneration. Nat Commun. 2019 Oct 25;10(1):4902. doi: 10.1038/s41467-019-12780-8.)

Our data (Menon et al., 2019 Nature Comm, Mathys et al. 2019 Nature, Kuchroo et al., 2021, bioRxiv) based on single cell analysis of healthy retinas and diseased retinas with macular degeneration and glaucoma indicate that the key inflammatory pathways reside in specialized glial cells known as astrocytes, Muller glia, and microglia. Our novel machine learning pipeline identified neuroinflammatory cytokines that promote disease progression, cause vision loss, and provide a genetic context to target these cell types and signaling pathways as a novel disease therapeutic to prevent vision loss in people suffering from these diseases. Most of our research is performed using human tissue, mouse, and glial cell cultures.

Single-cell analysis of RNA-seq data from human age-related macular degeneration. (a) Visualization of persistent populations, referred to as barcodes, by applying diffusion condensation to 71,063 nuclei isolated from age-related macular degeneration and control samples. Barcode colors highlight persistent topological features that correspond with known cell types (per key in panel b). (b) Potential of Heat-diffusion for Affinity-based Trajectory Embedding (PHATE) visualization of 71,063 nuclei isolated from AMD and control retinas. All major retinal cell types were identified by performing persistence analysis on the diffusion condensation process. (c) Diffusion condensation identified cell types, as shown by the average normalized expression of known cell type specific marker genes. D) Differentially expressed genes identified by Wasserstein Earth Mover's Distance (EMD) between cells from early-stage dry and late-stage neovascular age-related macular degenerationlesions and cells from control retinas on a cell type specific basis. Number of significantly differentially expressed genes between control and age-related macular degenerationcells reported in a cell type and stage specific manner (FDR corrected p-value < .1). Cell types sorted by most differential genes between dry age-related macular degenerationand control comparison. Vascular cells, microglia and astrocytes have the most differentially expressed genes in dry age-related macular degenerationcompared to contr

This research is important because it allows the comparison of gene signatures of glia in patients with AMD, glaucoma, and healthy individuals, helping elucidate the mechanism of retinal disease pathogenesis in humans and highlighting novel approaches to prevent permanent neuron loss in the retina and to halt blindness. AMD targets distinct regions of the retina with neuroinflammation with recruitment of immune cells, inflammatory cytokines, and activated, inflamed glial cells. While inflammation perpetuates the degenerative pathology, it has not yet been possible to define the various initial inflammatory insults, as substantial neuronal loss is often present at the time of symptom onset. The presence of multiple pathological processes has obscured understanding of neuroinflammatory onset, providing a major barrier to developing effective therapeutics. Therefore, current treatments for neuroinflammatory disorders are primarily targeting symptoms rather than halting underlying inflammatory processes. Through the use of spatial transcriptomics on human retinal samples and confocal imaging microscopy, we are generating an unbiased characterization that determines glial cell activation states, glial cell molecular signatures, and inflammation in proximity to the lesions in age-related macular degeneration. We are comparing cell types, states, and transitions in age-related macular degeneration. In our initial studies, we successfully applied single cell transcriptomics to profile the human retina as well as multiple neurodegenerative conditions affecting the central nervous system including MS and Alzheimer’s disease. In our research, spatial transcriptomics allows us to generate a map of glia and mechanisms to switch the cell profile from pro- to anti-inflammatory in neurodegeneration.