Molecular Classification and Comparative Taxonomics of Foveal and Peripheral Cells in Primate Retina

Experimental Framework for Assessing Mouse Retinal Regeneration Through Single-Cell RNA-Sequencing

Retinal degenerative diseases including age-related macular degeneration and glaucoma are estimated to currently affect more than 14 million people in the United States, with an increased prevalence of retinal degenerations in aged individuals. An expanding aged population who are living longer forecasts an increased prevalence and economic burden of visual impairments. Improvements to visual health and treatment paradigms for progressive retinal degenerations slow vision loss. However, current treatments fail to remedy the root cause of visual impairments caused by retinal degenerations-loss of retinal neurons. Stimulation of retinal regeneration from endogenous cellular sources presents an exciting treatment avenue for replacement of lost retinal cells. In multiple species including zebrafish and Xenopus, Müller glial cells maintain a highly efficient regenerative ability to reconstitute lost cells throughout the organism's lifespan, highlighting potential therapeutic avenues for stimulation of retinal regeneration in humans. Here, we describe how the application of single-cell RNA-sequencing (scRNA-seq) has enhanced our understanding of Müller glial cell-derived retinal regeneration, including the characterization of gene regulatory networks that facilitate/inhibit regenerative responses. Additionally, we provide a validated experimental framework for cellular preparation of mouse retinal cells as input into scRNA-seq experiments, including insights into experimental design and analyses of resulting data.

Regulation of Diseases-Associated Microglia in the Optic Nerve by Lipoxin B4 and Ocular Hypertension

Background

The resident astrocyte-retinal ganglion cell (RGC) lipoxin circuit is impaired during retinal stress, which includes ocular hypertension-induced neuropathy. Lipoxin B4 produced by homeostatic astrocytes directly acts on RGCs to increase survival and function in ocular hypertension-induced neuropathy. RGC death in the retina and axonal degeneration in the optic nerve are driven by the complex interactions between microglia and macroglia. Whether LXB4 neuroprotective actions include regulation of other cell types in the retina and/or optic nerve is an important knowledge gap.

Methods

Cellular targets and signaling of LXB4 in the retina were defined by single-cell RNA sequencing. Retinal neurodegeneration was induced by injecting silicone oil into the anterior chamber of the mouse eyes, which induced sustained and stable ocular hypertension. Morphological characterization of microglia populations in the retina and optic nerve was established by MorphOMICs and pseudotime trajectory analyses. The pathways and mechanisms of action of LXB4 in the optic nerve were investigated using bulk RNA sequencing. Transcriptomics data was validated by qPCR and immunohistochemistry. Differences between experimental groups were assessed by Student's t-test and one-way ANOVA.

Results

Single-cell transcriptomics identified microglia as a primary target for LXB4 in the healthy retina. LXB4 downregulated genes that drive microglia environmental sensing and reactivity responses. Analysis of microglial function revealed that ocular hypertension induced distinct, temporally defined, and dynamic phenotypes in the retina and, unexpectedly, in the distal myelinated optic nerve. Microglial expression of CD74, a marker of disease-associated microglia in the brain, was only induced in a unique population of optic nerve microglia, but not in the retina. Genetic deletion of lipoxin formation correlated with the presence of a CD74 optic nerve microglia population in normotensive eyes, while LXB4 treatment during ocular hypertension shifted optic nerve microglia toward a homeostatic morphology and non-reactive state and downregulated the expression of CD74. Furthermore, we identified a correlation between CD74 and phospho-phosphoinositide 3-kinases (p-PI3K) expression levels in the optic nerve, which was reduced by LXB4 treatment.

Conclusion

We identified early and dynamic changes in the microglia functional phenotype, reactivity, and induction of a unique CD74 microglia population in the distal optic nerve as key features of ocular hypertension-induced neurodegeneration. Our findings establish microglia regulation as a novel LXB4 target in the retina and optic nerve. LXB4 maintenance of a homeostatic optic nerve microglia phenotype and inhibition of a disease-associated phenotype are potential neuroprotective mechanisms for the resident LXB4 pathway.

Retinal ganglion cell circuits and glial interactions in humans and mice

Retinal ganglion cells (RGCs) are the brain's gateway for vision, and their degeneration underlies several blinding diseases. RGCs interact with other neuronal cell types, microglia, and astrocytes in the retina and in the brain. Much knowledge has been gained about RGCs and glia from mice and other model organisms, often with the assumption that certain aspects of their biology may be conserved in humans. However, RGCs vary considerably between species, which could affect how they interact with their neuronal and glial partners. This review details which RGC and glial features are conserved between mice, humans, and primates, and which differ. We also discuss experimental approaches for studying human and primate RGCs. These strategies will help to bridge the gap between rodent and human RGC studies and increase study translatability to guide future therapeutic strategies.

Acquisition of neurodegenerative features in isogenic OPTN(E50K) human stem cell-derived retinal ganglion cells associated with autophagy disruption and mTORC1 signaling reduction

The ability to derive retinal ganglion cells (RGCs) from human pluripotent stem cells (hPSCs) has led to numerous advances in the field of retinal research, with great potential for the use of hPSC-derived RGCs for studies of human retinal development, in vitro disease modeling, drug discovery, as well as their potential use for cell replacement therapeutics. Of all these possibilities, the use of hPSC-derived RGCs as a human-relevant platform for in vitro disease modeling has received the greatest attention, due to the translational relevance as well as the immediacy with which results may be obtained compared to more complex applications like cell replacement. While several studies to date have focused upon the use of hPSC-derived RGCs with genetic variants associated with glaucoma or other optic neuropathies, many of these have largely described cellular phenotypes with only limited advancement into exploring dysfunctional cellular pathways as a consequence of the disease-associated gene variants. Thus, to further advance this field of research, in the current study we leveraged an isogenic hPSC model with a glaucoma-associated mutation in the Optineurin (OPTN) protein, which plays a prominent role in autophagy. We identified an impairment of autophagic-lysosomal degradation and decreased mTORC1 signaling via activation of the stress sensor AMPK, along with subsequent neurodegeneration in OPTN(E50K) RGCs differentiated from hPSCs, and have further validated some of these findings in a mouse model of ocular hypertension. Pharmacological inhibition of mTORC1 in hPSC-derived RGCs recapitulated disease-related neurodegenerative phenotypes in otherwise healthy RGCs, while the mTOR-independent induction of autophagy reduced protein accumulation and restored neurite outgrowth in diseased OPTN(E50K) RGCs. Taken together, these results highlighted that autophagy disruption resulted in increased autophagic demand which was associated with downregulated signaling through mTORC1, contributing to the degeneration of RGCs.

Regional tuning of photoreceptor adaptation in the primate retina

Adaptation in cone photoreceptors allows our visual system to effectively operate over an enormous range of light intensities. However, little is known about the properties of cone adaptation in the specialized region of the primate central retina called the fovea, which is densely packed with cones and mediates high-acuity central vision. Here we show that macaque foveal cones exhibit weaker and slower luminance adaptation compared to cones in the peripheral retina. We find that this difference in adaptive properties between foveal and peripheral cones is due to differences in the magnitude of a hyperpolarization-activated current, Ih. This Ih current regulates the strength and time course of luminance adaptation in peripheral cones where it is more prominent than in foveal cones. A weaker and slower adaptation in foveal cones helps maintain a higher sensitivity for a longer duration which may be well-suited for maximizing the collection of high-acuity information at the fovea during gaze fixation between rapid eye movements.

Zebrafish models of human-duplicated SRGAP2 reveal novel functions in microglia and visual system development

The expansion of the human SRGAP2 family, resulting in a human-specific paralog SRGAP2C, likely contributed to altered evolutionary brain features. The introduction of SRGAP2C in mouse models is associated with changes in cortical neuronal migration, axon guidance, synaptogenesis, and sensory-task performance. Truncated SRGAP2C heterodimerizes with the full-length ancestral gene product SRGAP2A and antagonizes its functions. However, the significance of SRGAP2 duplication beyond neocortex development has not been elucidated due to the embryonic lethality of complete Srgap2 knockout in mice. Using zebrafish, we show that srgap2 knockout results in viable offspring and that these larvae phenocopy "humanized" SRGAP2C larvae, including altered morphometric features (i.e., reduced body length and inter-eye distance) and differential expression of synapse-, axonogenesis-, and vision-related genes. Through single-cell transcriptome analysis, we demonstrate a skewed balance of excitatory and inhibitory neurons that likely contribute to increased susceptibility to seizures displayed by Srgap2 mutant larvae, a phenotype resembling SRGAP2 loss-of-function in a child with early infantile epileptic encephalopathy. Single-cell data also shows strong endogenous expression of srgap2 in microglia with mutants exhibiting altered membrane dynamics and likely delayed maturation of microglial cells. Microglia cells expressing srgap2 were also detected in the developing eye together with altered expression of genes related to axonogenesis in mutant retinal cells. Consistent with the perturbed gene expression in the retina, we found that SRGAP2 mutant larvae exhibited increased sensitivity to broad and fine visual cues. Finally, comparing the transcriptomes of relevant cell types between human (+SRGAP2C) and non-human primates (-SRGAP2C) revealed significant overlaps of gene alterations with mutant cells in our zebrafish models; this suggests that SRGAP2C plays a similar role altering microglia and the visual system in modern humans. Together, our functional characterization of conserved ortholog Srgap2 and human SRGAP2C in zebrafish uncovered novel gene functions and highlights the strength of cross-species analysis in understanding the development of human-specific features.

Mammalian Retinal Bipolar Cells: Morphological Identification and Systematic Classification in Rabbit Retina

Retinal bipolar cells (BCs) convey visual signals from photoreceptors to more than 50 types of rabbit retinal ganglion cells (Famiglietti, 2020). More than 40 years ago, 10-11 types of bipolar cell were recognized in rabbit and cat retinas (Famiglietti, 1981). Twenty years later 10 were identified in mouse, rat, and monkey (Gosh et al., 2004), while recent molecular genetic studies indicate that there are 15 types of bipolar cell in mouse retina (Shekhar et al., 2016). The present detailed study of more than 800 bipolar cells in ten Golgi-impregnated rabbit retinas indicates that there are 14-16 types of cone bipolar cell and one type of rod bipolar cell in rabbit retina. These have been carefully analyzed in terms of dendritic and axonal morphology, and axon terminal stratification with respect to fiducial starburst amacrine cells. In fortuitous proximity, several types of bipolar cell can be related to identified ganglion cells by stratification and by contacts suggestive of synaptic connection. These results are compared with other studies of rabbit bipolar cells. Homologies with bipolar cells of mouse and monkey are considered in functional terms.

Foveal Retinal Ganglion Cells Develop Altered Calcium Dynamics Weeks After Photoreceptor Ablation

Purpose

Physiological changes in retinal ganglion cells (RGCs) have been reported in rodent models of photoreceptor (PR) loss, but this has not been investigated in primates. By expressing both a calcium indicator (GCaMP6s) and an optogenetic actuator (ChrimsonR) in foveal RGCs of the macaque, we reactivated RGCs in vivo and assessed their response in the weeks and years after PR loss.

Design

We used an in vivo calcium imaging approach to record optogenetically evoked activity in deafferented RGCs in primate fovea. Cellular scale recordings were made longitudinally over a 10-week period after PR ablation and compared with responses from RGCs that had lost PR input >2 years prior.

Participants

Three eyes received PR ablation, the right eye of a male Macaca mulatta (M1), the left eye of a female Macaca fascicularis (M2), and the right eye of a male Macaca fascicularis (M3). Two animals were used for in vivo recording, 1 for histological assessment.

Methods

Cones were ablated with an ultrafast laser delivered through an adaptive optics scanning light ophthalmoscope (AOSLO). A 0.5 second pulse of 25 Hz 660 nm light optogenetically stimulated RGCs, and the resulting GCaMP fluorescence signal was recorded using an AOSLO. Measurements were repeated over 10 weeks immediately after PR ablation, at 2.3 years and in control RGCs.

Main Outcome Measures

The calcium rise time, decay constant, and sensitivity index of optogenetic-mediated RGC were derived from GCaMP fluorescence recordings from 221 RGCs (animal M1) and 218 RGCs (animal M2) in vivo.

Results

After PR ablation, the mean decay constant of the calcium response in RGCs decreased 1.5-fold (standard deviation 1.6 ± 0.5 seconds to 0.6 ± 0.3 seconds) over the 10-week observation period in subject 1 and 2.1-fold (standard deviation 2.5 ± 0.5 seconds to 1.2 ± 0.2 seconds) within 8 weeks in subject 2. Calcium rise time and sensitivity index were stable. Optogenetic reactivation remained possible 2.3 years after PR ablation.

Conclusions

Altered calcium dynamics developed in primate foveal RGCs in the weeks after PR ablation. The mean decay constant of optogenetic-mediated calcium responses decreased 1.5- to twofold. This is the first report of this phenomenon in primate retina and further work is required to understand the role these changes play in cell survival and activity.

Financial Disclosures

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Top-down modulation of the retinal code via histaminergic neurons of the hypothalamus

The mammalian retina is considered an autonomous circuit, yet work dating back to Ramon y Cajal indicates that it receives inputs from the brain. How such inputs affect retinal processing has remained unknown. We confirmed brain-to-retina projections of histaminergic neurons from the mouse hypothalamus. Histamine application ex vivo altered the activity of various retinal ganglion cells (RGCs), including direction-selective RGCs that gained responses to high motion velocities. These results were reproduced in vivo with optic tract recordings where histaminergic retinopetal axons were activated chemogenetically. Such changes could improve vision of fast-moving objects (e.g., while running), which fits with the known increased activity of histaminergic neurons during arousal. An antihistamine drug reduced optomotor responses to high-speed moving stimuli in freely moving mice. In humans, the same antihistamine nonuniformly modulated visual sensitivity across the visual field, indicating an evolutionary conserved function of the histaminergic system. Our findings expose a previously unappreciated role for brain-to-retina projections in modulating retinal function.

Identification and characterization of early human photoreceptor states and cell-state-specific retinoblastoma-related features

Human cone photoreceptors differ from rods and serve as the retinoblastoma cell-of-origin, yet the developmental basis for their distinct behaviors is poorly understood. Here, we used deep full-length single-cell RNA-sequencing to distinguish post-mitotic cone and rod developmental states and identify cone-specific features that contribute to retinoblastomagenesis. The analyses revealed early post-mitotic cone- and rod-directed populations characterized by higher THRB or NRL regulon activities, an immature photoreceptor precursor population with concurrent cone and rod gene and regulon expression, and distinct early and late cone and rod maturation states distinguished by maturation-associated declines in RAX regulon activity. Unexpectedly, both L/M cone and rod precursors co-expressed NRL and THRB RNAs, yet they differentially expressed functionally antagonistic NRL and THRB isoforms and prematurely terminated THRB transcripts. Early L/M cone precursors exhibited successive expression of several lncRNAs along with MYCN, which composed the seventh most L/M-cone-specific regulon, and SYK, which contributed to the early cone precursors' proliferative response to RB1 loss. These findings reveal previously unrecognized photoreceptor precursor states and a role for early cone-precursor-intrinsic SYK expression in retinoblastoma initiation.

Reorganization of H3K9me heterochromatin leads to neuronal impairment via the cascading destruction of the KDM3B-centered epigenomic network

Histone H3K9 methylated heterochromatin silences repetitive non-coding sequences and lineage-specific genes during development, but how tissue-specific genes escape from heterochromatin in differentiated cells is unclear. Here, we examine age-dependent transcriptomic profiling of terminally differentiated mouse retina to identify epigenetic regulators involved in heterochromatin reorganization. The single-cell RNA sequencing analysis reveals a gradual downregulation of Kdm3b in cone photoreceptors during aging. Disruption of Kdm3b (Kdm3b +/- ) of 12-month-old mouse retina leads to the decreasing number of cones via apoptosis, and it changes the morphology of cone ribbon synapses. Integration of the transcriptome with epigenomic analysis in Kdm3b +/- retinas demonstrates gains of heterochromatin features in synapse assembly and vesicle transport genes that are downregulated via the accumulation of H3K9me1/2. Contrarily, losses of heterochromatin in apoptotic genes exacerbated retinal neurodegeneration. We propose that the KDM3B-centered epigenomic network is crucial for balancing of cone photoreceptor homeostasis via the modulation of gene set-specific heterochromatin features during aging.

Induction of astrocyte reactivity promotes neurodegeneration in human pluripotent stem cell models

Reactive astrocytes are known to exert detrimental effects upon neurons in several neurodegenerative diseases, yet our understanding of how astrocytes promote neurotoxicity remains incomplete, especially in human systems. In this study, we leveraged human pluripotent stem cell (hPSC) models to examine how reactivity alters astrocyte function and mediates neurodegeneration. hPSC-derived astrocytes were induced to a reactive phenotype, at which point they exhibited a hypertrophic profile and increased complement C3 expression. Functionally, reactive astrocytes displayed decreased intracellular calcium, elevated phagocytic capacity, and decreased contribution to the blood-brain barrier. Subsequently, co-culture of reactive astrocytes with a variety of neuronal cell types promoted morphological and functional alterations. Furthermore, when reactivity was induced in astrocytes from patient-specific hPSCs (glaucoma, Alzheimer's disease, and amyotrophic lateral sclerosis), the reactive state exacerbated astrocytic disease-associated phenotypes. These results demonstrate how reactive astrocytes modulate neurodegeneration, significantly contributing to our understanding of a role for reactive astrocytes in neurodegenerative diseases.

Insulin restores retinal ganglion cell functional connectivity and promotes visual recovery in glaucoma

Dendrite pathology and synaptic loss result in neural circuit dysfunction, a common feature of neurodegenerative diseases. There is a lack of strategies that target dendritic and synaptic regeneration to promote neurorecovery. We show that daily human recombinant insulin eye drops stimulate retinal ganglion cell (RGC) dendrite and synapse regeneration during ocular hypertension, a risk factor to develop glaucoma. We demonstrate that the ribosomal protein p70S6 kinase (S6K) is essential for insulin-dependent dendritic regrowth. Furthermore, S6K phosphorylation of the stress-activated protein kinase-interacting protein 1 (SIN1), a link between the mammalian target of rapamycin complexes 1 and 2 (mTORC1/2), is required for insulin-induced dendritic regeneration. Using two-photon microscopy live retinal imaging, we show that insulin rescues single-RGC light-evoked calcium (Ca2+) dynamics. We further demonstrate that insulin enhances neuronal survival and retina-brain connectivity leading to improved optomotor reflex-elicited behaviors. Our data support that insulin is a compelling pro-regenerative strategy with potential clinical implications for the treatment and management of glaucoma.

RDH12 allows cone photoreceptors to regenerate opsin visual pigments from a chromophore precursor to escape competition with rods

Capture of a photon by an opsin visual pigment isomerizes its 11-cis-retinaldehyde (11cRAL) chromophore to all-trans-retinaldehyde (atRAL), which subsequently dissociates. To restore light sensitivity, the unliganded apo-opsin combines with another 11cRAL to make a new visual pigment. Two enzyme pathways supply chromophore to photoreceptors. The canonical visual cycle in retinal pigment epithelial cells supplies 11cRAL at low rates. The photic visual cycle in Müller cells supplies cones with 11-cis-retinol (11cROL) chromophore precursor at high rates. Although rods can only use 11cRAL to regenerate rhodopsin, cones can use 11cRAL or 11cROL to regenerate cone visual pigments. We performed a screen in zebrafish retinas and identified ZCRDH as a candidate for the enzyme that converts 11cROL to 11cRAL in cone inner segments. Retinoid analysis of eyes from Zcrdh-mutant zebrafish showed reduced 11cRAL and increased 11cROL levels, suggesting impaired conversion of 11cROL to 11cRAL. By microspectrophotometry, isolated Zcrdh-mutant cones lost the capacity to regenerate visual pigments from 11cROL. ZCRDH therefore possesses all predicted properties of the cone 11cROL dehydrogenase. The human protein most similar to ZCRDH is RDH12. By immunocytochemistry, ZCRDH was abundantly present in cone inner segments, similar to the reported distribution of RDH12. Finally, RDH12 was the only mammalian candidate protein to exhibit 11cROL-oxidase catalytic activity. These observations suggest that RDH12 in mammals is the functional ortholog of ZCRDH, which allows cones, but not rods, to regenerate visual pigments from 11cROL provided by Müller cells. This capacity permits cones to escape competition from rods for visual chromophore in daylight-exposed retinas.

A general principle governing neuronal evolution reveals a human-accelerated neuron type potentially underlying the high prevalence of autism in humans

The remarkable ability of a single genome sequence to encode a diverse collection of distinct cell types, including the thousands of cell types found in the mammalian brain, is a key characteristic of multicellular life. While it has been observed that some cell types are far more evolutionarily conserved than others, the factors driving these differences in evolutionary rate remain unknown. Here, we hypothesized that highly abundant neuronal cell types may be under greater selective constraint than rarer neuronal types, leading to variation in their rates of evolution. To test this, we leveraged recently published cross-species single-nucleus RNA-sequencing datasets from three distinct regions of the mammalian neocortex. We found a strikingly consistent relationship where more abundant neuronal subtypes show greater gene expression conservation between species, which replicated across three independent datasets covering >106 neurons from six species. Based on this principle, we discovered that the most abundant type of neocortical neurons-layer 2/3 intratelencephalic excitatory neurons-has evolved exceptionally quickly in the human lineage compared to other apes. Surprisingly, this accelerated evolution was accompanied by the dramatic down-regulation of autism-associated genes, which was likely driven by polygenic positive selection specific to the human lineage. In sum, we introduce a general principle governing neuronal evolution and suggest that the exceptionally high prevalence of autism in humans may be a direct result of natural selection for lower expression of a suite of genes that conferred a fitness benefit to our ancestors while also rendering an abundant class of neurons more sensitive to perturbation.

Aging of the eye: Lessons from cataracts and age-related macular degeneration

Aging is the greatest risk factor for chronic human diseases, including many eye diseases. Geroscience aims to understand the effects of the aging process on these diseases, including the genetic, molecular, and cellular mechanisms that underlie the increased risk of disease over the lifetime. Understanding of the aging eye increases general knowledge of the cellular physiology impacted by aging processes at various biological extremes. Two major diseases, age-related cataract and age-related macular degeneration (AMD) are caused by dysfunction of the lens and retina, respectively. Lens transparency and light refraction are mediated by lens fiber cells lacking nuclei and other organelles, which provides a unique opportunity to study a single aging hallmark, i.e., loss of proteostasis, within an environment of limited metabolism. In AMD, local dysfunction of the photoreceptors/retinal pigmented epithelium/Bruch's membrane/choriocapillaris complex in the macula leads to the loss of photoreceptors and eventually loss of central vision, and is driven by nearly all the hallmarks of aging and shares features with Alzheimer's disease, Parkinson's disease, cardiovascular disease, and diabetes. The aging eye can function as a model for studying basic mechanisms of aging and, vice versa, well-defined hallmarks of aging can be used as tools to understand age-related eye disease.

Thorny and Tufted Retinal Ganglion Cells Express the Transcription Factor Forkhead Proteins Foxp1 and Foxp2 in Marmoset (Callithrix jacchus)

The transcription factor forkhead/winged-helix domain proteins Foxp1 and Foxp2 have previously been studied in mouse retina, where they are expressed in retinal ganglion cells named F-mini and F-midi. Here we show that both transcription factors are expressed by small subpopulations (on average less than 10%) of retinal ganglion cells in the retina of the marmoset monkey (Callithrix jacchus). The morphology of Foxp1- and Foxp2-expressing cells was revealed by intracellular DiI injections of immunofluorescent cells. Foxp1- and Foxp2-expressing cells comprised multiple types of wide-field ganglion cells, including broad thorny cells, narrow thorny cells, and tufted cells. The large majority of Foxp2-expressing cells were identified as tufted cells. Tufted cells stratify broadly in the middle of the inner plexiform layer. They resemble broad thorny cells but their proximal dendrites are bare of branches and the distal dendrites branch frequently forming dense dendritic tufts. Double labeling with calretinin, a previously established marker for broad thorny and narrow thorny cells, showed that only a small proportion of ganglion cells co-expressed calretinin and Foxp1 or Foxp2 supporting the idea that the two markers are differentially expressed in retinal ganglion cells of marmoset retina.

Neurexin 3 is Essential for the Specific Wiring of a Color Pathway in the Mammalian Retina

Precise wiring within sensory systems is critical for the accurate transmission of information. In the visual system, S-cone photoreceptors specialize in detecting short-wavelength light, crucial to color perception and environmental cue detection. S-cones form specific synapses with S-cone bipolar cells (SCBCs), a connection that is remarkably consistent across species. Yet, the molecular mechanisms guiding this specificity remain unexplored. To address this, we used the cone-dominant ground squirrel for deep-sequencing of cone subtype transcriptomes and identified Nrxn3 as an essential molecule for the S-cone to SCBC synapse. Using transgenic mouse models, we further examined the role of Nrxn3 in S-cones and discovered a significant reduction of SCBC connections in the absence of Nrxn3. This finding extends the known functions of neurexins, typically associated with synapse regulation, by highlighting their essential role in a specific synaptic connection for the first time. Moreover, the differentially expressed genes identified here pave the way for further investigations into the unique functions of cone subtypes.

Conservation, alteration, and redistribution of mammalian striatal interneurons

Mammalian brains vary in size, structure, and function, but the extent to which evolutionarily novel cell types contribute to this variation remains unresolved1-4. Recent studies suggest there is a primate-specific population of striatal inhibitory interneurons, the TAC3 interneurons5. However, there has not yet been a detailed analysis of the spatial and phylogenetic distribution of this population. Here, we profile single cell gene expression in the developing pig (an ungulate) and ferret (a carnivore), representing 94 million years divergence from primates, and assign newborn inhibitory neurons to initial classes first specified during development6. We find that the initial class of TAC3 interneurons represents an ancestral striatal population that is also deployed towards the cortex in pig and ferret. In adult mouse, we uncover a rare population expressing Tac2, the ortholog of TAC3, in ventromedial striatum, prompting a reexamination of developing mouse striatal interneuron initial classes by targeted enrichment of their precursors. We conclude that the TAC3 interneuron initial class is conserved across Boreoeutherian mammals, with the mouse population representing Th striatal interneurons, a subset of which expresses Tac2. This study suggests that initial classes of telencephalic inhibitory neurons are largely conserved and that during evolution, neuronal types in the mammalian brain change through redistribution and fate refinement, rather than by derivation of novel precursors early in development.

Detecting novel cell type in single-cell chromatin accessibility data via open-set domain adaptation

Recent advances in single-cell technologies enable the rapid growth of multi-omics data. Cell type annotation is one common task in analyzing single-cell data. It is a challenge that some cell types in the testing set are not present in the training set (i.e. unknown cell types). Most scATAC-seq cell type annotation methods generally assign each cell in the testing set to one known type in the training set but neglect unknown cell types. Here, we present OVAAnno, an automatic cell types annotation method which utilizes open-set domain adaptation to detect unknown cell types in scATAC-seq data. Comprehensive experiments show that OVAAnno successfully identifies known and unknown cell types. Further experiments demonstrate that OVAAnno also performs well on scRNA-seq data. Our codes are available online at https://github.com/lisaber/OVAAnno/tree/master.

A pupillary contrast response in mice and humans: Neural mechanisms and visual functions

In the pupillary light response (PLR), increases in ambient light constrict the pupil to dampen increases in retinal illuminance. Here, we report that the pupillary reflex arc implements a second input-output transformation; it senses temporal contrast to enhance spatial contrast in the retinal image and increase visual acuity. The pupillary contrast response (PCoR) is driven by rod photoreceptors via type 6 bipolar cells and M1 ganglion cells. Temporal contrast is transformed into sustained pupil constriction by the M1's conversion of excitatory input into spike output. Computational modeling explains how the PCoR shapes retinal images. Pupil constriction improves acuity in gaze stabilization and predation in mice. Humans exhibit a PCoR with similar tuning properties to mice, which interacts with eye movements to optimize the statistics of the visual input for retinal encoding. Thus, we uncover a conserved component of active vision, its cell-type-specific pathway, computational mechanisms, and optical and behavioral significance.

Maculopapillary Bundle Degeneration in Optic Neuropathies

PURPOSE OF REVIEW

Degeneration of the maculopapillary bundle (MPB) is a prominent feature in a spectrum of optic neuropathies. MPB-selective degeneration is seen in specific conditions, such as nutritional and toxic optic neuropathies, Leber hereditary optic neuropathy (LHON), and dominant optic atrophy (DOA). Despite their distinct etiologies and clinical presentations, which encompass variations in age of incidence and monocular or binocular onset, these disorders share a core molecular mechanism: compromised mitochondrial homeostasis. This disruption is characterized by dysfunctions in mitochondrial metabolism, biogenesis, and protein synthesis. This article provides a comprehensive understanding of the MPB's role in optic neuropathies, emphasizing the importance of mitochondrial mechanisms in the pathogenesis of these conditions.

RECENT FINDINGS

Optical coherence tomography studies have characterized the retinal nerve fiber layer changes accompanying mitochondrial-affiliated optic neuropathies. Selective thinning of the temporal optic nerve head is preceded by thickening in early stages of these disorders which correlates with reductions in macular ganglion cell layer thinning and vascular atrophy. A recently proposed mechanism underpinning the selective atrophy of the MPB involves the positive feedback of reactive oxygen species generation as a common consequence of mitochondrial dysfunction. Additionally, new research has revealed that the MPB can undergo degeneration in the early stages of glaucoma, challenging the historically held belief that this area was not involved in this common optic neuropathy. A variety of anatomical risk factors influence the propensity of glaucomatous MPB degeneration, and cases present distinct patterns of ganglion cell degeneration that are distinct from those observed in mitochondria-associated diseases. This review synthesizes clinical and molecular research on primary MPB disorders, highlighting the commonalities and differences in their pathogenesis.

KEY POINTS (BOX)

1. Temporal degeneration of optic nerve fibers accompanied by cecocentral scotoma is a hallmark of maculopapillary bundle (MPB) degeneration. 2. Mechanisms of MPB degeneration commonly implicate mitochondrial dysfunction. 3. Recent research challenges the traditional belief that the MPB is uninvolved in glaucoma by showing degeneration in the early stages of this common optic neuropathy, yet with features distinct from other MPB-selective neuropathies. 4. Reactive oxygen species generation is a mechanism linking mitochondrial mechanisms of MPB-selective optic neuropathies, but in-vivo and in-vitro studies are needed to validate this hypothesis.

Comprehensive single-cell atlas of the mouse retina

Single-cell RNA sequencing (scRNA-seq) has advanced our understanding of cellular heterogeneity by characterizing cell types across tissues and species. While several mouse retinal scRNA-seq datasets exist, each dataset is either limited in cell numbers or focused on specific cell classes, thereby hindering comprehensive gene expression analysis across all retina types. To fill the gap, we generated the largest retinal scRNA-seq dataset to date, comprising approximately 190,000 single cells from C57BL/6J mouse retinas, enriched for rare population cells via antibody-based magnetic cell sorting. Integrating this dataset with public datasets, we constructed the Mouse Retina Cell Atlas (MRCA) for wild-type mice, encompassing over 330,000 cells, characterizing 12 major classes and 138 cell types. The MRCA consolidates existing knowledge, identifies new cell types, and is publicly accessible via CELLxGENE, UCSC Cell Browser, and the Broad Single Cell Portal, providing a user-friendly resource for the mouse retina research community.

Gene regulatory roles of growth and differentiation factors in retinal development

Retinal ganglion cell (RGC) differentiation is tightly controlled by extrinsic and intrinsic factors. Growth and differentiation factor 15 (GDF-15) promotes RGC differentiation, opposite to GDF-11 which inhibits RGC differentiation, both in the mouse retina and in human stem cells. To deepen our understanding of how these two closely related molecules confer opposing effects on retinal development, here we assess the transcriptional profiles of mouse retinal progenitors exposed to exogenous GDF-11 or -15. We find a dichotomous effect of GDF-15 on RGC differentiation, decreasing RGCs expressing residual pro-proliferative genes and increasing RGCs expressing non-proliferative genes, suggestive of greater RGC maturation. Furthermore, GDF-11 promoted the differentiation of photoreceptors and amacrine cells. These data enhance our understanding of the mechanisms underlying the differentiation of RGCs and photoreceptors from retinal progenitors and suggest new approaches to the optimization of protocols for the differentiation of these cell types.

VDAC in Retinal Health and Disease

The retina, a tissue of the central nervous system, is vital for vision as its photoreceptors capture light and transform it into electrical signals, which are further processed before they are sent to the brain to be interpreted as images. The retina is unique in that it is continuously exposed to light and has the highest metabolic rate and demand for energy amongst all the tissues in the body. Consequently, the retina is very susceptible to oxidative stress. VDAC, a pore in the outer membrane of mitochondria, shuttles metabolites between mitochondria and the cytosol and normally protects cells from oxidative damage, but when a cell's integrity is greatly compromised it initiates cell death. There are three isoforms of VDAC, and existing evidence indicates that all three are expressed in the retina. However, their precise localization and function in each cell type is unknown. It appears that most retinal cells express substantial amounts of VDAC2 and VDAC3, presumably to protect them from oxidative stress. Photoreceptors express VDAC2, HK2, and PKM2-key proteins in the Warburg pathway that also protect these cells. Consistent with its role in initiating cell death, VDAC is overexpressed in the retinal degenerative diseases retinitis pigmentosa, age related macular degeneration (AMD), and glaucoma. Treatment with antioxidants or inhibiting VDAC oligomerization reduced its expression and improved cell survival. Thus, VDAC may be a promising therapeutic candidate for the treatment of these diseases.

Ancient origin of the rod bipolar cell pathway in the vertebrate retina

Vertebrates rely on rod photoreceptors for vision in low-light conditions. The specialized downstream circuit for rod signalling, called the primary rod pathway, is well characterized in mammals, but circuitry for rod signalling in non-mammals is largely unknown. Here we demonstrate that the mammalian primary rod pathway is conserved in zebrafish, which diverged from extant mammals ~400 million years ago. Using single-cell RNA sequencing, we identified two bipolar cell types in zebrafish that are related to mammalian rod bipolar cell (RBCs), the only bipolar type that directly carries rod signals from the outer to the inner retina in the primary rod pathway. By combining electrophysiology, histology and ultrastructural reconstruction of the zebrafish RBCs, we found that, similar to mammalian RBCs, both zebrafish RBC types connect with all rods in their dendritic territory and provide output largely onto amacrine cells. The wiring pattern of the amacrine cells postsynaptic to one RBC type is strikingly similar to that of mammalian RBCs and their amacrine partners, suggesting that the cell types and circuit design of the primary rod pathway emerged before the divergence of teleost fish and mammals. The second RBC type, which forms separate pathways, was either lost in mammals or emerged in fish.

The vertebrate retina: a window into the evolution of computation in the brain

Animal brains are probably the most complex computational machines on our planet, and like everything in biology, they are the product of evolution. Advances in developmental and palaeobiology have been expanding our general understanding of how nervous systems can change at a molecular and structural level. However, how these changes translate into altered function - that is, into 'computation' - remains comparatively sparsely explored. What, concretely, does it mean for neuronal computation when neurons change their morphology and connectivity, when new neurons appear or old ones disappear, or when transmitter systems are slowly modified over many generations? And how does evolution use these many possible knobs and dials to constantly tune computation to give rise to the amazing diversity in animal behaviours we see today? Addressing these major gaps of understanding benefits from choosing a suitable model system. Here, I present the vertebrate retina as one perhaps unusually promising candidate. The retina is ancient and displays highly conserved core organisational principles across the entire vertebrate lineage, alongside a myriad of adjustments across extant species that were shaped by the history of their visual ecology. Moreover, the computational logic of the retina is readily interrogated experimentally, and our existing understanding of retinal circuits in a handful of species can serve as an anchor when exploring the visual circuit adaptations across the entire vertebrate tree of life, from fish deep in the aphotic zone of the oceans to eagles soaring high up in the sky.

POU6F2, a risk factor for glaucoma, myopia and dyslexia, labels specific populations of retinal ganglion cells

Pou6f2 is a genetic connection between central corneal thickness (CCT) in the mouse and a risk factor for developing primary open-angle glaucoma. POU6F2 is also a risk factor for several conditions in humans, including glaucoma, myopia, and dyslexia. Recent findings demonstrate that POU6F2-positive retinal ganglion cells (RGCs) comprise a number of RGC subtypes in the mouse, some of which also co-stain for Cdh6 and Hoxd10. These POU6F2-positive RGCs appear to be novel of ON-OFF directionally selective ganglion cells (ooDSGCs) that do not co-stain with CART or SATB2 (typical ooDSGCs markers). These POU6F2-positive cells are sensitive to damage caused by elevated intraocular pressure. In the DBA/2J mouse glaucoma model, heavily-labeled POU6F2 RGCs decrease by 73% at 8 months of age compared to only 22% loss of total RGCs (labeled with RBPMS). Additionally, Pou6f2-/- mice suffer a significant loss of acuity and spatial contrast sensitivity along with an 11.4% loss of total RGCs. In the rhesus macaque retina, POU6F2 labels the large parasol ganglion cells that form the magnocellular (M) pathway. The association of POU6F2 with the M-pathway may reveal in part its role in human glaucoma, myopia, and dyslexia.

Cone-Opponent Ganglion Cells in the Primate Fovea Tuned to Noncardinal Color Directions

A long-standing question in vision science is how the three cone photoreceptor types-long (L), medium (M), and short (S) wavelength sensitive-combine to generate our perception of color. Hue perception can be described along two opponent axes: red-green and blue-yellow. Psychophysical measurements of color appearance indicate that the cone inputs to the red-green and blue-yellow opponent axes are M vs. L + S and L vs. M + S, respectively. However, the "cardinal directions of color space" revealed by psychophysical measurements of color detection thresholds following adaptation are L vs. M and S vs. L + M. These cardinal directions match the most common cone-opponent retinal ganglion cells (RGCs) in the primate retina. Accordingly, the cone opponency necessary for color appearance is thought to be established in the cortex. While neurons with the appropriate M vs. L + S and L vs. M + S opponency have been reported in the retina and lateral geniculate nucleus, their existence continues to be debated. Resolving this long-standing debate is necessary because a complete account of the cone opponency in the retinal output is critical for understanding how downstream neural circuits process color. Here, we performed adaptive optics calcium imaging to noninvasively measure foveal RGC light responses in the living Macaca fascicularis eye. We confirm the presence of L vs. M + S and M vs. L + S neurons with noncardinal cone opponency and demonstrate that cone-opponent signals in the retinal output are more diverse than classically thought.

Specific photoreceptor cell fate pathways are differentially altered in NR2E3-associated diseases

Mutations in NR2E3, a gene encoding an orphan nuclear transcription factor, cause two retinal dystrophies with a distinct phenotype, but the precise role of NR2E3 in rod and cone transcriptional networks remains unclear. To dissect NR2E3 function, we performed scRNA-seq in the retinas of wildtype and two different Nr2e3 mouse models that show phenotypes similar to patients carrying NR2E3 mutations. Our results reveal that rod and cone populations are not homogeneous and can be separated into different sub-classes. We identify a previously unreported cone pathway that generates hybrid cones co-expressing both cone- and rod-related genes. In mutant retinas, this hybrid cone subpopulation is more abundant and includes a subpopulation of rods transitioning towards a cone cell fate. Hybrid photoreceptors with high misexpression of cone- and rod-related genes are prone to regulated necrosis. Overall, our results shed light on the role of NR2E3 in modulating photoreceptor differentiation towards cone and rod fates and explain how different mutations in NR2E3 lead to distinct visual disorders in humans.

Evolutionary and developmental specialization of foveal cell types in the marmoset

In primates, high-acuity vision is mediated by the fovea, a small specialized central region of the retina. The fovea, unique to the anthropoid lineage among mammals, undergoes notable neuronal morphological changes during postnatal maturation. However, the extent of cellular similarity across anthropoid foveas and the molecular underpinnings of foveal maturation remain unclear. Here, we used high-throughput single-cell RNA sequencing to profile retinal cells of the common marmoset (Callithrix jacchus), an early divergent in anthropoid evolution from humans, apes, and macaques. We generated atlases of the marmoset fovea and peripheral retina for both neonates and adults. Our comparative analysis revealed that marmosets share almost all their foveal types with both humans and macaques, highlighting a conserved cellular structure among primate foveas. Furthermore, by tracing the developmental trajectory of cell types in the foveal and peripheral retina, we found distinct maturation paths for each. In-depth analysis of gene expression differences demonstrated that cone photoreceptors and Müller glia (MG), among others, show the greatest molecular divergence between these two regions. Utilizing single-cell ATAC-seq and gene-regulatory network inference, we uncovered distinct transcriptional regulations differentiating foveal cones from their peripheral counterparts. Further analysis of predicted ligand-receptor interactions suggested a potential role for MG in supporting the maturation of foveal cones. Together, these results provide valuable insights into foveal development, structure, and evolution.

Atypical retinal function in a mouse model of Fragile X syndrome

Altered function of peripheral sensory neurons is an emerging mechanism for symptoms of autism spectrum disorders. Visual sensitivities are common in autism, but whether differences in the retina might underlie these sensitivities is not well-understood. We explored retinal function in the Fmr1 knockout model of Fragile X syndrome, focusing on a specific type of retinal neuron, the "sustained On alpha" retinal ganglion cell. We found that these cells exhibit changes in dendritic structure and dampened responses to light in the Fmr1 knockout. We show that decreased light sensitivity is due to increased inhibitory input and reduced E-I balance. The change in E-I balance supports maintenance of circuit excitability similar to what has been observed in cortex. These results show that loss of Fmr1 in the mouse retina affects sensory function of one retinal neuron type. Our findings suggest that the retina may be relevant for understanding visual function in Fragile X syndrome.

Distributed feature representations of natural stimuli across parallel retinal pathways

How sensory systems extract salient features from natural environments and organize them across neural pathways is unclear. Combining single-cell and population two-photon calcium imaging in mice, we discover that retinal ON bipolar cells (second-order neurons of the visual system) are divided into two blocks of four types. The two blocks distribute temporal and spatial information encoding, respectively. ON bipolar cell axons co-stratify within each block, but separate laminarly between them (upper block: diverse temporal, uniform spatial tuning; lower block: diverse spatial, uniform temporal tuning). ON bipolar cells extract temporal and spatial features similarly from artificial and naturalistic stimuli. In addition, they differ in sensitivity to coherent motion in naturalistic movies. Motion information is distributed across ON bipolar cells in the upper and the lower blocks, multiplexed with temporal and spatial contrast, independent features of natural scenes. Comparing the responses of different boutons within the same arbor, we find that axons of all ON bipolar cell types function as computational units. Thus, our results provide insights into the visual feature extraction from naturalistic stimuli and reveal how structural and functional organization cooperate to generate parallel ON pathways for temporal and spatial information in the mammalian retina.

Ancestral photoreceptor diversity as the basis of visual behaviour

Animal colour vision is based on comparing signals from different photoreceptors. It is generally assumed that processing different spectral types of photoreceptor mainly serves colour vision. Here I propose instead that photoreceptors are parallel feature channels that differentially support visual-motor programmes like motion vision behaviours, prey capture and predator evasion. Colour vision may have emerged as a secondary benefit of these circuits, which originally helped aquatic vertebrates to visually navigate and segment their underwater world. Specifically, I suggest that ancestral vertebrate vision was built around three main systems, including a high-resolution general purpose greyscale system based on ancestral red cones and rods to mediate visual body stabilization and navigation, a high-sensitivity specialized foreground system based on ancestral ultraviolet cones to mediate threat detection and prey capture, and a net-suppressive system based on ancestral green and blue cones for regulating red/rod and ultraviolet circuits. This ancestral strategy probably still underpins vision today, and different vertebrate lineages have since adapted their original photoreceptor circuits to suit their diverse visual ecologies.

Cone Synaptic function is modulated by the leucine rich repeat (LRR) adhesion molecule LRFN2

Daylight vision is mediated by cone photoreceptors in vertebrates, which synapse with bipolar cells (BCs) and horizontal (HCs) cells. This cone synapse is functionally and anatomically complex, connecting to 8 types of depolarizing BCs (DBCs) and 5 types of hyperpolarizing BCs (HBCs) in mice. The dendrites of DBCs and HCs cells make invaginating ribbon synapses with the cone axon terminal, while HBCs form flat synapses with the cone pedicles. The molecular architecture that underpins this organization is relatively poorly understood. To identify new proteins involved in synapse formation and function we used an unbiased proteomic approach and identified LRFN2 (leucine-rich repeat and fibronectin III domain-containing 2) as a component of the DBC signaling complex. LRFN2 is selectively expressed at cone terminals and co-localizes with PNA, and other DBC signalplex members. In LRFN2 deficient mice, the synaptic markers: LRIT3, ELFN2, mGluR6, TRPM1 and GPR179 are properly localized. Similarly, LRFN2 expression and localization is not dependent on these synaptic proteins. In the absence of LRFN2 the cone-mediated photopic electroretinogram b-wave amplitude is reduced at the brightest flash intensities. These data demonstrate that LRFN2 absence compromises normal synaptic transmission between cones and cone DBCs.Significance Statement Signaling between cone photoreceptors and the downstream bipolar cells is critical to normal vision. Cones synapse with 13 different types of bipolar cells forming an invaginating ribbon synapses with 8 types, and flat synapses with 5 types, to form one of the most complex synapses in the brain. In this report a new protein, LRFN2 (leucine-rich repeat and fibronectin III domain-containing 2), was identified that is expressed at the cone synapse. Using Lrfn2 knockout mice we show LRFN2 is required for the normal cone signaling.

Heterogeneity in the progression of retinal pathologies in mice harboring patient mimicking Impg2 mutations

Biallelic mutations in interphotoreceptor matrix proteoglycan 2 (IMPG2) in humans cause retinitis pigmentosa (RP) with early macular involvement, albeit the disease progression varies widely due to genetic heterogeneity and IMPG2 mutation type. There are currently no treatments for IMPG2-RP. To aid preclinical studies toward eventual treatments, there is a need to better understand the progression of disease pathology in appropriate animal models. Toward this goal, we developed mouse models with patient mimicking homozygous frameshift (T807Ter) or missense (Y250C) Impg2 mutations, as well as mice with a homozygous frameshift mutation (Q244Ter) designed to completely prevent IMPG2 protein expression, and characterized the trajectory of their retinal pathologies across postnatal development until late adulthood. We found that the Impg2T807Ter/T807Ter and Impg2Q244Ter/Q244Ter mice exhibited early onset gliosis, impaired photoreceptor outer segment maintenance, appearance of subretinal deposits near the optic disc, disruption of the outer retina, and neurosensorial detachment, whereas the Impg2Y250C/Y250C mice exhibited minimal retinal pathology. These results demonstrate the importance of mutation type in disease progression in IMPG2-RP and provide a toolkit and preclinical data for advancing therapeutic approaches.

Mapping the Retina onto the Brain

Vision begins in the retina, which extracts salient features from the environment and encodes them in the spike trains of retinal ganglion cells (RGCs), the output neurons of the eye. RGC axons innervate diverse brain areas (>50 in mice) to support perception, guide behavior, and mediate influences of light on physiology and internal states. In recent years, complete lists of RGC types (∼45 in mice) have been compiled, detailed maps of their dendritic connections drawn, and their light responses surveyed at scale. We know less about the RGCs' axonal projection patterns, which map retinal information onto the brain. However, some organizing principles have emerged. Here, we review the strategies and mechanisms that govern developing RGC axons and organize their innervation of retinorecipient brain areas.

Mapping Cell Atlases at the Single-Cell Level

Recent advancements in single-cell technologies have led to rapid developments in the construction of cell atlases. These atlases have the potential to provide detailed information about every cell type in different organisms, enabling the characterization of cellular diversity at the single-cell level. Global efforts in developing comprehensive cell atlases have profound implications for both basic research and clinical applications. This review provides a broad overview of the cellular diversity and dynamics across various biological systems. In addition, the incorporation of machine learning techniques into cell atlas analyses opens up exciting prospects for the field of integrative biology.

Comprehensive single-cell atlas of the mouse retina

Single-cell RNA sequencing (scRNA-seq) has advanced our understanding of cellular heterogeneity at the single-cell resolution by classifying and characterizing cell types in multiple tissues and species. While several mouse retinal scRNA-seq reference datasets have been published, each dataset either has a relatively small number of cells or is focused on specific cell classes, and thus is suboptimal for assessing gene expression patterns across all retina types at the same time. To establish a unified and comprehensive reference for the mouse retina, we first generated the largest retinal scRNA-seq dataset to date, comprising approximately 190,000 single cells from C57BL/6J mouse whole retinas. This dataset was generated through the targeted enrichment of rare population cells via antibody-based magnetic cell sorting. By integrating this new dataset with public datasets, we conducted an integrated analysis to construct the Mouse Retina Cell Atlas (MRCA) for wild-type mice, which encompasses over 330,000 single cells. The MRCA characterizes 12 major classes and 138 cell types. It captured consensus cell type characterization from public datasets and identified additional new cell types. To facilitate the public use of the MRCA, we have deposited it in CELLxGENE, UCSC Cell Browser, and the Broad Single Cell Portal for visualization and gene expression exploration. The comprehensive MRCA serves as an easy-to-use, one-stop data resource for the mouse retina communities.

Optimal trade-off control in machine learning-based library design, with application to adeno-associated virus (AAV) for gene therapy

Adeno-associated viruses (AAVs) hold tremendous promise as delivery vectors for gene therapies. AAVs have been successfully engineered-for instance, for more efficient and/or cell-specific delivery to numerous tissues-by creating large, diverse starting libraries and selecting for desired properties. However, these starting libraries often contain a high proportion of variants unable to assemble or package their genomes, a prerequisite for any gene delivery goal. Here, we present and showcase a machine learning (ML) method for designing AAV peptide insertion libraries that achieve fivefold higher packaging fitness than the standard NNK library with negligible reduction in diversity. To demonstrate our ML-designed library's utility for downstream engineering goals, we show that it yields approximately 10-fold more successful variants than the NNK library after selection for infection of human brain tissue, leading to a promising glial-specific variant. Moreover, our design approach can be applied to other types of libraries for AAV and beyond.

Large language models in bioinformatics: applications and perspectives

Large language models (LLMs) are a class of artificial intelligence models based on deep learning, which have great performance in various tasks, especially in natural language processing (NLP). Large language models typically consist of artificial neural networks with numerous parameters, trained on large amounts of unlabeled input using self-supervised or semi-supervised learning. However, their potential for solving bioinformatics problems may even exceed their proficiency in modeling human language. In this review, we will present a summary of the prominent large language models used in natural language processing, such as BERT and GPT, and focus on exploring the applications of large language models at different omics levels in bioinformatics, mainly including applications of large language models in genomics, transcriptomics, proteomics, drug discovery and single cell analysis. Finally, this review summarizes the potential and prospects of large language models in solving bioinformatic problems.

Single-Cell Transcriptomic Sequencing Reveals Tissue Architecture and Deciphers Pathological Reprogramming During Retinal Ischemia in Macaca fascicularis

Purpose

Acute retinal arterial ischemia diseases (ARAIDs) are ocular emergencies that require immediate intervention within a restricted therapeutic window to prevent blindness. However, the underlying molecular mechanisms contributing to the pathogenesis of ARAIDs remain enigmatic. Herein, we present the single-cell RNA sequencing (scRNA-seq) alterations during ischemia in the primate retina as a preliminary endeavor in understanding the molecular complexities of ARAIDs.

Methods

An ophthalmic artery occlusion model was established through ophthalmic artery ligation in two Macaca fascicularis. scRNA-seq and bioinformatics analyses were used to detect retinal changes during ischemia, which are further validated by immunofluorescence analysis. Western blot and flow cytometry assays were performed to measure the microglia polarization status.

Results

The findings of this study reveal notable changes in the retina under acute ischemic conditions. Particularly, retinal ischemia compromised mitochondrial functions of rod photoreceptors, partly leading to the rapid loss of healthy rods. Furthermore, we observed a noteworthy transcriptional alteration in the activation of microglia induced by ischemia. The targeted correction of the proinflammatory cytokine CXCL8 effectively suppresses microglia M1 polarization in retinal ischemia, ultimately reducing the proinflammatory transformation in vitro. In addition, retina ischemia induced the apoptotic inclination of endothelial cells and the heightened interaction with microglia, which signifies the influence of microglia in disrupting the retinal-blood barrier.

Conclusions

Our research has successfully identified and described the pathologic alterations occurring in several cell types during a short period of ischemia. These observations provide valuable insights for ameliorating retinal damage and promoting the restoration of vision.

Expression patterns of CYP26A1, FGF8, CDKN1A, and NPVF in the developing rhesus monkey retina

The fovea centralis (fovea) is a specialized region of the primate retina that plays crucial roles in high-resolution visual acuity and color perception. The fovea is characterized by a high density of cone photoreceptors and no rods, and unique anatomical properties that contribute to its remarkable visual capabilities. Early histological analyses identified some of the key events that contribute to foveal development, but the mechanisms that direct the specification of this area are not understood. Recently, the expression of the retinoic acid-metabolizing enzyme CYP26A1 has become a hallmark of some of the retinal specializations found in vertebrates, including the primate fovea and the high-acuity area in avian species. In chickens, the retinoic acid pathway regulates the expression of FGF8 to then direct the development of a rod-free area. Similarly, high levels of CYP26A1, CDKN1A, and NPVF expression have been observed in the primate macula using transcriptomic approaches. However, which retinal cells express these genes and their expression dynamics in the developing primate eye remain unknown. Here, we systematically characterize the expression patterns of CYP26A1, FGF8, CDKN1A, and NPVF during the development of the rhesus monkey retina, from early stages of development in the first trimester until the third trimester (near term). Our data suggest that some of the markers previously proposed to be fovea-specific are not enriched in the progenitors of the rhesus monkey fovea. In contrast, CYP26A1 is expressed at high levels in the progenitors of the fovea, while it localizes in a subpopulation of macular Müller glia cells later in development. Together these data provide invaluable insights into the expression dynamics of several molecules in the nonhuman primate retina and highlight the developmental advancement of the foveal region.

From water to land: Evolution of photoreceptor circuits for vision in air

When vertebrates first conquered the land, they encountered a visual world that was radically distinct from that of their aquatic ancestors. Fish exploit the strong wavelength-dependent interactions of light with water by differentially feeding the signals from up to 5 spectral photoreceptor types into distinct behavioural programmes. However, above the water the same spectral rules do not apply, and this called for an update to visual circuit strategies. Early tetrapods soon evolved the double cone, a still poorly understood pair of new photoreceptors that brought the "ancestral terrestrial" complement from 5 to 7. Subsequent nonmammalian lineages differentially adapted this highly parallelised retinal input strategy for their diverse visual ecologies. By contrast, mammals shed most ancestral photoreceptors and converged on an input strategy that is exceptionally general. In eutherian mammals including in humans, parallelisation emerges gradually as the visual signal traverses the layers of the retina and into the brain.

Dynamic changes in ocular shape during human development and its implications for retina fovea formation

The human fovea is known for its distinctive pit-like appearance, which results from the displacement of retinal layers superficial to the photoreceptors cells. The photoreceptors are found at high density within the foveal region but not the surrounding retina. Efforts to elucidate the mechanisms responsible for these unique features have ruled out cell death as an explanation for pit formation and changes in cell proliferation as the cause of increased photoreceptor density. These findings have led to speculation that mechanical forces acting within and on the retina during development underly the formation of foveal architecture. Here we review eye morphogenesis and retinal remodeling in human embryonic development. Our meta-analysis of the literature suggests that fovea formation is a protracted process involving dynamic changes in ocular shape that start early and continue throughout most of human embryonic development. From these observations, we propose a new model for fovea development.

Patch-seq: Advances and Biological Applications

Multimodal analysis of gene-expression patterns, electrophysiological properties, and morphological phenotypes at the single-cell/single-nucleus level has been arduous because of the diversity and complexity of neurons. The emergence of Patch-sequencing (Patch-seq) directly links transcriptomics, morphology, and electrophysiology, taking neuroscience research to a multimodal era. In this review, we summarized the development of Patch-seq and recent applications in the cortex, hippocampus, and other nervous systems. Through generating multimodal cell type atlases, targeting specific cell populations, and correlating transcriptomic data with phenotypic information, Patch-seq has provided new insight into outstanding questions in neuroscience. We highlight the challenges and opportunities of Patch-seq in neuroscience and hope to shed new light on future neuroscience research.

Factorized discriminant analysis for genetic signatures of neuronal phenotypes

Navigating the complex landscape of single-cell transcriptomic data presents significant challenges. Central to this challenge is the identification of a meaningful representation of high-dimensional gene expression patterns that sheds light on the structural and functional properties of cell types. Pursuing model interpretability and computational simplicity, we often look for a linear transformation of the original data that aligns with key phenotypic features of cells. In response to this need, we introduce factorized linear discriminant analysis (FLDA), a novel method for linear dimensionality reduction. The crux of FLDA lies in identifying a linear function of gene expression levels that is highly correlated with one phenotypic feature while minimizing the influence of others. To augment this method, we integrate it with a sparsity-based regularization algorithm. This integration is crucial as it selects a subset of genes pivotal to a specific phenotypic feature or a combination thereof. To illustrate the effectiveness of FLDA, we apply it to transcriptomic datasets from neurons in the Drosophila optic lobe. We demonstrate that FLDA not only captures the inherent structural patterns aligned with phenotypic features but also uncovers key genes associated with each phenotype.

Late gene therapy limits the restoration of retinal function in a mouse model of retinitis pigmentosa

Retinitis pigmentosa is an inherited photoreceptor degeneration that begins with rod loss followed by cone loss. This cell loss greatly diminishes vision, with most patients becoming legally blind. Gene therapies are being developed, but it is unknown how retinal function depends on the time of intervention. To uncover this dependence, we utilize a mouse model of retinitis pigmentosa capable of artificial genetic rescue. This model enables a benchmark of best-case gene therapy by removing variables that complicate answering this question. Complete genetic rescue was performed at 25%, 50%, and 70% rod loss (early, mid and late, respectively). Early and mid treatment restore retinal output to near wild-type levels. Late treatment retinas exhibit continued, albeit slowed, loss of sensitivity and signal fidelity among retinal ganglion cells, as well as persistent gliosis. We conclude that gene replacement therapies delivered after 50% rod loss are unlikely to restore visual function to normal. This is critical information for administering gene therapies to rescue vision.

Evolutionary and Developmental Specialization of Foveal Cell Types in the Marmoset

In primates, high-acuity vision is mediated by the fovea, a small specialized central region of the retina. The fovea, unique to the anthropoid lineage among mammals, undergoes notable neuronal morphological changes during postnatal maturation. However, the extent of cellular similarity across anthropoid foveas and the molecular underpinnings of foveal maturation remain unclear. Here, we used high throughput single cell RNA sequencing to profile retinal cells of the common marmoset ( Callithrix jacchus ), an early divergent in anthropoid evolution from humans, apes, and macaques. We generated atlases of the marmoset fovea and peripheral retina for both neonates and adults. Our comparative analysis revealed that marmosets share almost all its foveal types with both humans and macaques, highlighting a conserved cellular structure among primate foveas. Furthermore, by tracing the developmental trajectory of cell types in the foveal and peripheral retina, we found distinct maturation paths for each. In-depth analysis of gene expression differences demonstrated that cone photoreceptors and Müller glia, among others, show the greatest molecular divergence between these two regions. Utilizing single-cell ATAC-seq and gene-regulatory network inference, we uncovered distinct transcriptional regulations differentiating foveal cones from their peripheral counterparts. Further analysis of predicted ligand-receptor interactions suggested a potential role for Müller glia in supporting the maturation of foveal cones. Together, these results provide valuable insights into foveal development, structure, and evolution.

Significance statement

The sharpness of our eyesight hinges on a tiny retinal region known as the fovea. The fovea is pivotal for primate vision and is susceptible to diseases like age-related macular degeneration. We studied the fovea in the marmoset-a primate with ancient evolutionary ties. Our data illustrated the cellular and molecular composition of its fovea across different developmental ages. Our findings highlighted a profound cellular consistency among marmosets, humans, and macaques, emphasizing the value of marmosets in visual research and the study of visual diseases.

Comparative Analysis of Single-cell and Single-nucleus RNA-sequencing in a Rabbit Model of Retinal Detachment-related Proliferative Vitreoretinopathy

Purpose

Proliferative vitreoretinopathy (PVR) is the most common cause of failure of retinal reattachment surgery, and the molecular changes leading to this aberrant wound healing process are currently unknown. Our ultimate goal is to study PVR pathogenesis by employing single-cell transcriptomics to dissect cellular heterogeneity.

Design

Here we aimed to compare single-cell RNA sequencing (scRNA-seq)  and single-nucleus RNA-sequencing (snRNA-seq) of retinal PVR samples in the rabbit model.

Participants

Unilateral induction of PVR lesions in rabbit eyes with contralateral eyes serving as controls.

Methods

Proliferative vitreoretinopathy was induced unilaterally in Dutch Belted rabbits. At different timepoints after PVR induction, retinas were dissociated into either cells or nuclei suspension and processed for scRNA-seq or snRNA-seq.

Main Outcome Measures

Single cell and nuclei transcriptomic profiles of retinas after PVR induction.

Results

Single-cell RNA sequencing and snRNA-seq were conducted on retinas at 4 hours and 14 days after disease induction. Although the capture rate of unique molecular identifiers and genes were greater in scRNA-seq samples, overall gene expression profiles of individual cell types were highly correlated between scRNA-seq and snRNA-seq. A major disparity between the 2 sequencing modalities was the cell type capture rate, however, with glial cell types overrepresented in scRNA-seq, and inner retinal neurons were enriched by snRNA-seq. Furthermore, fibrotic Müller glia were overrepresented in snRNA-seq samples, whereas reactive Müller glia were overrepresented in scRNA-seq samples. Trajectory analyses were similar between the 2 methods, allowing for the combined analysis of the scRNA-seq and snRNA-seq data sets.

Conclusions

These findings highlight limitations of both scRNA-seq and snRNA-seq analysis and imply that use of both techniques together can more accurately identify transcriptional networks critical for aberrant fibrogenesis in PVR than using either in isolation.

Financial Disclosures

Proprietary or commercial disclosure may be found after the references.