Neuroprotective effects of apigenin on retinal ganglion cells in ischemia/reperfusion: modulating mitochondrial dynamics in in vivo and in vitro models.

BACKGROUND: Retinal ischemia/reperfusion (RIR) is implicated in various forms of optic neuropathies, yet effective treatments are lacking. RIR leads to the death of retinal ganglion cells (RGCs) and subsequent vision loss, posing detrimental effects on both physical and mental health. Apigenin (API), derived from a wide range of sources, has been reported to exert protective effects against ischemia/reperfusion injuries in various organs, such as the brain, kidney, myocardium, and liver. In this study, we investigated the protective effect of API and its underlying mechanisms on RGC degeneration induced by retinal ischemia/reperfusion (RIR). METHODS: An in vivo model was induced by anterior chamber perfusion following intravitreal injection of API one day prior to the procedure. Meanwhile, an in vitro model was established through 1% oxygen and glucose deprivation. The neuroprotective effects of API were evaluated using H&E staining, spectral-domain optical coherence tomography (SD-OCT), Fluoro-Gold retrograde labeling, and Photopic negative response (PhNR). Furthermore, transmission electron microscopy (TEM) was employed to observe mitochondrial crista morphology and integrity. To elucidate the underlying mechanisms of API, the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay, flow cytometry assay, western blot, cell counting kit-8 (CCK-8) assay, lactate dehydrogenase (LDH) assay, JC-1 kit assay, dichlorofluorescein-diacetate (DCFH-DA) assay, as well as TMRE and Mito-tracker staining were conducted. RESULTS: API treatment protected retinal inner plexiform layer (IPL) and ganglion cell complex (GCC), and improved the function of retinal ganglion cells (RGCs). Additionally, API reduced RGC apoptosis and decreased lactate dehydrogenase (LDH) release by upregulating Bcl-2 and Bcl-xL expression, while downregulating Bax and cleaved caspase-3 expression. Furthermore, API increased mitochondrial membrane potential (MMP) and decreased extracellular reactive oxygen species (ROS) production. These effects were achieved by enhancing mitochondrial function, restoring mitochondrial cristae morphology and integrity, and regulating the expression of OPA1, MFN2, and DRP1, thereby regulating mitochondrial dynamics involving fusion and fission. CONCLUSION: API protects RGCs against RIR injury by modulating mitochondrial dynamics, promoting mitochondrial fusion and fission.

Fam96a is essential for the host control of Toxoplasma gondii infection by fine-tuning macrophage polarization via an iron-dependent mechanism.

BACKGROUND: Toxoplasmosis affects a quarter of the world's population. Toxoplasma gondii (T.gondii) is an intracellular parasitic protozoa. Macrophages are necessary for proliferation and spread of T.gondii by regulating immunity and metabolism. Family with sequence similarity 96A (Fam96a; formally named Ciao2a) is an evolutionarily conserved protein that is highly expressed in macrophages, but whether it play a role in control of T. gondii infection is unknown. METHODOLOGY/PRINCIPAL FINDINGS: In this study, we utilized myeloid cell-specific knockout mice to test its role in anti-T. gondii immunity. The results showed that myeloid cell-specific deletion of Fam96a led to exacerbate both acute and chronic toxoplasmosis after exposure to T. gondii. This was related to a defectively reprogrammed polarization in Fam96a-deficient macrophages inhibited the induction of immune effector molecules, including iNOS, by suppressing interferon/STAT1 signaling. Fam96a regulated macrophage polarization process was in part dependent on its ability to fine-tuning intracellular iron (Fe) homeostasis in response to inflammatory stimuli. In addition, Fam96a regulated the mitochondrial oxidative phosphorylation or related events that involved in control of T. gondii. CONCLUSIONS/SIGNIFICANCE: All these findings suggest that Fam96a ablation in macrophages disrupts iron homeostasis and inhibits immune effector molecules, which may aggravate both acute and chronic toxoplasmosis. It highlights that Fam96a may autonomously act as a critical gatekeeper of T. gondii control in macrophages.

Whole genome sequencing enables new genetic diagnosis for inherited retinal diseases by identifying pathogenic variants.

Inherited retinal diseases (IRDs) are a group of common primary retinal degenerative disorders. Conventional genetic testing strategies, such as panel-based sequencing and whole exome sequencing (WES), can only elucidate the genetic etiology in approximately 60% of IRD patients. Studies have suggested that unsolved IRD cases could be attributed to previously undetected structural variants (SVs) and intronic variants in IRD-related genes. The aim of our study was to obtain a definitive genetic diagnosis by employing whole genome sequencing (WGS) in IRD cases where the causative genes were inconclusive following an initial screening by panel sequencing. A total of 271 unresolved IRD patients and their available family members (n = 646) were screened using WGS to identify pathogenic SVs and intronic variants in 792 known ocular disease genes. Overall, 13% (34/271) of IRD patients received a confirmed genetic diagnosis, among which 7% were exclusively attributed to SVs, 4% to a combination of single nucleotide variants (SNVs) and SVs while another 2% were linked to intronic variants. 22 SVs, 3 deep-intronic variants, and 2 non-canonical splice-site variants across 14 IRD genes were identified in the entire cohort. Notably, all of these detected SVs and intronic variants were novel pathogenic variants. Among those, 74% (20/27) of variants were found in genes causally linked to Retinitis Pigmentosa (RP), with the gene EYS being the most frequently affected by SVs. The identification of SVs and intronic variants through WGS enhances the genetic diagnostic yield of IRDs and broadens the mutational spectrum of known IRD-associated genes.

CO2 electro-reduction reaction via a two-dimensional TM@TAP single-atom catalyst.

In this study, the possibility of using TM atom anchored monolayer TAP as a class of electrocatalysts (TM@TAP, TM = 3d and 4d transition metal) toward carbon dioxide reduction reaction (CO2RR) was systematically investigated using first-principles calculations. During screening potential catalysts, the possibility that H and OH block the active site was considered. Then, the reaction mechanisms of screened catalysts were explored in detail. Interestingly, the different catalysts demonstrated different selectivities. Our results demonstrate that Cr@TAP, Zn@TAP, Mo@TAP, and Cd@TAP are selective toward the HCOOH product with a limiting potential in the range of -0.33 to -0.71 V. Mn@TAP and Rh@TAP promote CO production. The reduction products of Fe@TAP and Co@TAP were CH3OH and HCHO, respectively. Tc@TAP and Ru@TAP can catalyze CO2 to yield the deep reduction product, i.e. CH4. Among these catalysts, Cr@TAP and Rh@TAP are highly active due to their lower limiting potentials of -0.33 V and -0.28 V, respectively, and Fe@TAP can promote the production of the desired CH3OH with a limiting potential of -0.51 V, which allow them to be promising electrocatalysts for the CO2RR. We hope that our study will provide some insights into the rational design of electrocatalysts and useful guidance for experimental researchers.

Mechanism and ingredients prediction of Radix Salviae-Angelicae Sinensis Radix-Lycii Fructus-Rehmanniae Radix Praeparata-Ginkgo Folium for retinitis pigmentosa therapy using network pharmacology and molecular docking analysis.

Background: Radix Salviae (Danshen)-Angelicae Sinensis Radix (Danggui)-Lycii Fructus (Gouqizi)-Rehmanniae Radix Praeparata (Shudihuang)-Ginkgo Folium (Yinxinye) (RALRG) are commonly used herbs in China that have shown positive effects on retinitis pigmentosa (RP). However, little research has been performed on the impact of RALRG and RP. Herein, this study aimed to predict the mechanism and potential components of RALRG in treating RP. Methods: The ingredients of RALRG were obtained from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP); the potential targets of RP and RALRG were obtained from TCMSP, GeneCards, and the Online Mendelian Inheritance in Man (OMIM) database. A protein-protein interaction (PPI) network was constructed to visualize PPIs. The functional enrichment was performed with the R program. A visual RALRG-RP-pathway pharmacology network was established by Cytoscape 3.9.1. Molecular docking was used to perform molecular docking and calculate the binding affinity. Results: A total of 132 effective active ingredients in RALRG with 248 target genes were screened; 92 intersection target genes were acquired from the intersection of RP- and RALRG-related genes. Gene Ontology (GO) enrichment indicated that these intersection targets were mainly involved in oxidative stress, metal ion response, and chemical stress. Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis indicated that the PI3K-AKT, cellular senescence, and MAPK signaling pathways were closely related to the therapy of RP. In addition, a potential pharmacology network for RALRG-RP-pathway was constructed. AKT1 and JUN were considered the primary targets. Luteolin, quercetin, and kaempferol were identified as the vital three active ingredients. Conclusions: RALRG was found to be the main regulator for oxidative stress and PI3K/AKT signaling pathways. Luteolin, quercetin, and kaempferol were three promising complementary ingredients for RP treatment. This study may provide a theoretical basis for applying RALRG to screen potential drugs for RP.

A specific visual-volumetric sensor for mercury ions based on smart hydrogel.

On the basis of the "seeing is believing" concept and the existing theory of Hg2+ coordination chemistry, for the first time, we innovatively designed and synthesized a visual-volumetric sensor platform with fluorescein and uracil functionalized polyacrylamide hydrogel. Without the aid of any complicated instruments and power sources, the sensor-enabled quantitative µM-level Hg2+ detection Hg2+ by reading graduation on a pipette with the naked eye. The sensor undergoes volumetric response and shows a wide linear response range to Hg2+ (1.0 × 10-6-5.0 × 10-5 mol L-1) with 2.8 × 10-7 mol L-1 as the detection limit. The highly selective (easily distinguished Hg2+ from other common metal ions), rapid response (∼30 min), and acceptable repeatability (RSD < 5% in all cases) demonstrated that the developed sensor is suitable for onsite practical use for the determination of Hg2+ while being low-cost, simple, and portable. The design principles of the obtained materials and the construction techniques and methods of the sensors described in our study provide a new idea for the research and development of smart materials and a series of visual-volumetric sensors for other analytes.

Treatment of autosomal dominant retinitis pigmentosa caused by RHO-P23H mutation with high-fidelity Cas13X in mice.

Mutations in Rhodopsin (RHO) gene commonly cause autosomal dominant retinitis pigmentosa (adRP) without effective therapeutic treatment so far. Compared with genomic DNA-targeting CRISPR-Cas9 system, Cas13 edits RNA for therapeutic applications, avoiding the risk of causing permanent changes in the genome. In particular, a compact and high-fidelity Cas13X (hfCas13X) recently has been developed to degrade targeted RNA with minimal collateral effects and could also be packaged in a single adeno-associated virus for efficient in vivo delivery. In this study, we engineered single-guide RNA for hfCas13X to specifically knock down human mutant Rhodopsin transcripts RHO-P23H with minimal effect on wild-type transcripts. Moreover, treatment with hfCas13X alleviated the adRP progression in both RHO-P23H overexpression-induced and humanized hRHOP23H/WT mouse models. Our study indicates the potential of hfCas13X in treating adRP caused by RHO mutations and other genetic diseases.

Development of a novel prediction model based on protein structure for identifying RPE65-associated inherited retinal disease (IRDs) of missense variants.

Purpose: This study aimed to develop a prediction model to classify RPE65-mediated inherited retinal disease (IRDs) based on protein secondary structure and to analyze phenotype-protein structure correlations of RPE65 missense variants in a Chinese cohort. Methods: Pathogenic or likely pathogenic missense variants of RPE65 were obtained from UniProt, ClinVar, and HGMD databases. The three-dimensional structure of RPE65 was retrieved from the Protein Data Bank (PDB) and modified with Pymol software. A novel prediction model was developed using LASSO regression and multivariate logistic regression to identify RPE65-associated IRDs. A total of 21 Chinese probands with RPE65 variants were collected to analyze phenotype-protein structure correlations of RPE65 missense variants. Results: The study found that both pathogenic and population missense variants were associated with structural features of RPE65. Pathogenic variants were linked to sheet, ß-sheet, strands, ß-hairpins, Fe2+ (iron center), and active site cavity, while population variants were related to helix, loop, helices, and helix-helix interactions. The novel prediction model showed accuracy and confidence in predicting the disease type of RPE65 variants (AUC = 0.7531). The study identified 25 missense variants in Chinese patients, accounting for 72.4% of total mutations. A significant correlation was observed between clinical characteristics of RPE65-associated IRDs and changes in amino acid type, specifically for missense variants of F8 (H68Y, P419S). Conclusion: The study developed a novel prediction model based on the protein structure of RPE65 and investigated phenotype-protein structure correlations of RPE65 missense variants in a Chinese cohort. The findings provide insights into the precise diagnosis of RPE65-mutated IRDs.

Mutational spectrum in a Chinese cohort with congenital cataracts.

BACKGROUND: To identify the mutational spectrum in a Chinese cohort with congenital cataracts. METHODS: Probands (n = 164) with congenital cataracts and their affected or unaffected available family members were recruited for clinical examinations and panel-based next-generation sequencing, then classified into a cohort for further mutational analysis. RESULTS: After recruitment (n = 442; 228 males and 214 females), 49.32% (218/442) of subjects received a clinical diagnosis of congenital cataracts, and 56.88% (124/218) of patients received a molecular diagnosis. Eighty-four distinct variants distributed among 43 different genes, including 42 previously reported variants and 42 novel variants, were detected, and 49 gene variants were causally associated with patient phenotypes; 27.37% of variants (23/84) were commonly detected in PAX6, GJA8 and CRYGD, and the three genes covered 33.06% of cases (41/124) with molecular diagnosis. The majority of genes were classified as genes involved in nonsyndromic congenital cataracts (19/43, 44.19%) and were responsible for 56.45% of cases (70/124). The majority of functional and nucleotide changes were missense variants (53/84, 63.10%) and substitution variants (74/84, 88.10%), respectively. Nine de novo variants were identified. CONCLUSION: This study provides a reference for individualized genetic counseling and further extends the mutational spectrum of congenital cataracts.

Theoretical investigation of pure-state single-photon and large-bandwidth-correlation biphoton generation from micro/nanofiber.

Pure-state single photons and large-bandwidth-correlation biphotons are fundamental resources for quantum information processing. The dispersion properties of micro/nanofiber (MNF) can be tailored by carefully choosing its diameter, resulting in a flexibly tailored biphoton spectrum. We theoretically investigate pure-state single photons and large-bandwidth-correlation biphotons produced by degenerate spontaneous four-wave mixing in MNF. In our simulation, a single-photon state with a purity of 99% will be theoretically attained by choosing the appropriate pump bandwidth and center wavelength with respect to the diameter and length of the MNF. Further, when an appropriate diameter with a negligible curvature at the zero dispersion wavelength is chosen, even a narrow pump bandwidth is capable of motivating remarkably broadband correlation biphotons, e.g., for a MNF diameter of 0.7 µm, the theoretical full width at half maximum is 473 nm. In practice, the application of a MNF-based quantum light source is dependent on the technologies that precisely control and measure the diameter. Our theoretical investigation will guide the experimental realization of high-quality quantum light sources based on MNF.

ceRNA network construction and identification of hub genes as novel therapeutic targets for age-related cataracts using bioinformatics.

Background: The aim of this study is to investigate the genetic and epigenetic mechanisms involved in the pathogenesis of age-related cataract (ARC). Methods: We obtained the transcriptome datafile of th ree ARC samples and three healthy, age-matched samples and used differential expression analyses to identify the differentially expressed genes (DEGs). The differential lncRNA-associated competing endogenous (ceRNA) network, and the protein-protein network (PPI) were constructed using Cytoscape and STRING. Cluster analyses were performed to identify the underlying molecular mechanisms of the hub genes affecting ARC progression. To verify the immune status of the ARC patients, immune-associated analyses were also conducted. Results: The PPI network identified the FOXO1 gene as the hub gene with the highest score, as calculated by the Maximal Clique Centrality (MCC) algorithm. The ceRNA network identified lncRNAs H19, XIST, TTTY14, and MEG3 and hub genes FOXO1, NOTCH3, CDK6, SPRY2, and CA2 as playing key roles in regulating the pathogenesis of ARC. Additionally, the identified hub genes showed no significant correlation with an immune response but were highly correlated with cell metabolism, including cysteine, methionine, and galactose. Discussion: The findings of this study may provide clues toward ARC pathogenic mechanisms and may be of significance for future therapeutic research.

A novel biosensing platform for detection of glaucoma biomarker GDF15 via an integrated BLI-ELASA strategy.

Glaucoma is a leading cause of irreversible blindness worldwide. Early discovery and prioritized intervention significantly impact its prognosis. Precise monitoring of the biomarker GDF15 contributes towards effective diagnosis and assessment of glaucoma. In this study, we demonstrate that GDF15 monitoring can also aid screening for glaucoma risk and early diagnosis. We obtained an aptamer (APT2TM) with high affinity, high specificity, and high stability for binding to both human-derived and rat-derived GDF15. Simulation results showed that the binding capabilities of APT2TM are mainly affected by the interplay between van der Waals forces and polar solvation energy, and that salt bridges and hydrogen bonds play critical roles. We then integrated an enzyme-linked aptamer sandwich assay (ELASA) into a biolayer interferometry (BLI) system to develop an automated, high-throughput, real-time monitoring BLI-ELASA biosensing platform. This platform exhibited a wide linear detection window (10-810 pg/mL range) and high sensitivity for GDF15 (detection limit of 5-6 pg/mL). Moreover, we confirmed its excellent performance when applied to GDF15 quantification in real samples from glaucomatous rats and clinical patients. We believe that this technology represents a robust, convenient, and cost-effective approach for risk screening, early diagnosis, and animal modeling evaluation of glaucoma in the near future.

Construction of ceRNA network and identification of hub genes in aniridia-associated keratopathy using bioinformatics analysis.

Aniridia-associated keratopathy (AAK) is characteristic at ocular surface of aniridia caused by haploinsufficiency of PAX6. Competing endogenous RNA (ceRNA) has been reported to play an important role in various diseases, whereas its function on AAK is unclear. The microarray data of 20 AAK patients and 20 healthy people were downloaded from the Gene Expression Omnibus (GEO) database. Differentially expressed lncRNAs, miRNAs, and mRNAs were analyzed using "limma" packages and weighted gene co-expression network analysis (WGCNA). A ceRNA network was constructed by Cytoscape 3.9.1, and miR-224-5p, miR-30a-5p, and miR-204-5p were at the center of the network. CIBERSORTx algorithm and ssGSEA analyses revealed that AAK was associated with immune cell infiltration, showing that activated Mast cells increased while resting Mast cells decreased and NK cells decreased in AAK. Type II INF Response, CCR, parainflammation, T cell co-stimulation, and APC co-stimulation of AAK patients differed from healthy individuals. Additionally, the ROC curve of five genes, MITF(AUC = 0.988), RHOB(AUC = 0.973), JUN(AUC = 0.953), PLAUR (AUC = 0.925), and ARG2 (AUC = 0.915) with high confidence in predicting AAK were identified. Gene set enrichment analysis (GSEA) analysis of hub genes enriched in the IL-17 signaling pathway.

Genetic Characteristics and Variation Spectrum of USH2A-Related Retinitis Pigmentosa and Usher Syndrome.

Purposes: We aimed to characterize the USH2A genotypic spectrum in a Chinese cohort and provide a detailed genetic profile for Chinese patients with USH2A-IRD. Methods: We designed a retrospective study wherein a total of 1,334 patients diagnosed with IRD were included as a study cohort, namely 1,278 RP and 56 USH patients, as well as other types of IEDs patients and healthy family members as a control cohort. The genotype-phenotype correlation of all participants with USH2A variant was evaluated. Results: Etiological mutations in USH2A, the most common cause of RP and USH, were found in 16.34% (n = 218) genetically solved IRD patients, with prevalences of 14.87% (190/1,278) and 50% (28/56). After bioinformatics and QC processing, 768 distinct USH2A variants were detected in all participants, including 136 disease-causing mutations present in 665 alleles, distributed in 5.81% of all participants. Of these 136 mutations, 43 were novel, nine were founder mutations, and two hot spot mutations with allele count ≥10. Furthermore, 38.5% (84/218) of genetically solved USH2A-IRD patients were caused by at least one of both c.2802T>G and c.8559-2 A>G mutations, and 36.9% and 69.6% of the alleles in the RP and USH groups were truncating, respectively. Conclusion: USH2A-related East Asian-specific founder and hot spot mutations were the major causes for Chinese RP and USH patients. Our study systematically delineated the genotype spectrum of USH2A-IRD, enabled accurate genetic diagnosis, and provided East Asian and other ethnicities with baseline data of a Chinese origin, which would better serve genetic counseling and therapeutic targets selection.

One-step high-throughput detection of low-abundance biomarker BDNF using a biolayer interferometry-based 3D aptasensor.

Although biosensors for signal monitoring have been extensively developed, their application in one-step high-throughput detection of low-abundance disease biomarkers remains challenging. This study presents a 3D aptasensor based on a biolayer interferometry (BLI) technique, followed by the sensitive and rapid detection of the specific biomarker brain-derived neurotrophic factor (BDNF) for early screening of glaucoma, an irreversible disease that causes blindness. The developed 3D aptasensor enabled one-step batch conversion of the low-abundance biomarker BDNF binding into optical interference signal, which was mainly attributed to the following factors: (1) A dimeric aptamer with extremely high targeting affinity was constructed as a biorecognition molecule, (2) highly sensitive 3D matrix sensors were integrated as signal transduction elements, and (3) the BLI Octet system with automated, high-throughput, and real-time online monitoring capabilities was used for reporting. The 3D aptasensor exhibited a broad detection window from 0.41 to 250 ng/mL BDNF, with a limit of detection of 0.2 ng/mL. Furthermore, detection of BDNF in glaucoma patient serum using the aptasensor showed good agreement with ELISA findings as well as the clinical diagnosis of the patient, demonstrating the feasibility of the system as a screening tool for glaucoma. This one-step high-throughput screening approach provides a valuable solution for the early diagnosis of glaucoma and may reduce the risk of blindness in visually impaired people.

Identification of Immune Infiltration-Related ceRNAs as Novel Biomarkers for Prognosis of Patients With Primary Open-Angle Glaucoma.

Glaucoma is the leading cause of irreversible blindness globally; hence, relevant clinical biomarkers are necessary to enable diagnosis, early detection, and development of novel therapies. The differentially expressed genes were annotated and visualized using Gene Ontology and Kyoto Encyclopedia. In addition, a competitive endogenous ribonucleic acids network was constructed using Cytoscape, which explained the regulation of gene expression in glaucoma. The CIBERSORT algorithm was employed to analyze the immune microenvironment. We validated that the core genes could predict glaucoma occurrence and development and identified potential molecular mechanism pathways, which were associated with immune infiltration and participated in endogenous regulation networks. Our data may partially explain the pathogenesis of glaucoma and they provide potential theoretical support for targeted therapy.

Effects of acute high intraocular pressure on red-green and blue-yellow cortical color responses in non-human primates.

Glaucoma is a leading cause of irreversible blindness worldwide, and intraocular pressure (IOP) is an established and modifiable risk factor for both chronic and acute glaucoma. The relationship between color vision deficits and chronic glaucoma has been described previously. However, the effects of acute glaucoma or acute primary angle closure, which has high prevalence in China, on color vision remains unclear. To address the above question, red-green or blue-yellow color responses in V1, V2, and V4 of seven rhesus macaques were monitored using intrinsic-signal optical imaging while monocular anterior chamber perfusions were performed to reversibly elevate IOP acutely over a clinically observed range of 30 to 90 mmHg. We found that the cortical population responses to both red-green and blue-yellow grating stimuli, systematically decreased as IOP increased from 30 to 90 mmHg. Although a similar decrement in magnitude was noted in V1, V2, and V4, blue-yellow responses were consistently more impaired than red-green responses at all levels of acute IOP elevation and in all monitored visual areas. This physiological study in non-human primates demonstrates that acute IOP elevations substantially depress the ability of the visual cortex to register color information. This effect is more severe for blue-yellow responses than for red-green responses, suggesting selective impairment of the koniocellular pathways compared with the parvocellular pathways. Together, we infer that blue-yellow color vision might be the most vulnerable visual function in acute glaucoma patients.

Single-Cell RNA-Seq Reveals the Cellular Diversity and Developmental Characteristics of the Retinas of an Infant and a Young Child.

The human retina, located in the innermost layer of the eye, plays a decisive role in visual perception. Dissecting the heterogeneity of retinal cells is essential for understanding the mechanism of visual development. Here, we performed single-cell RNA-seq to analyze 194,967 cells from the donors of infants and young children, resulting in 17 distinct clusters representing major cell types in the retina: rod photoreceptors (PRs), cone PRs, bipolar cells (BCs), horizontal cells (HCs), amacrine cells (ACs), retinal ganglion cells (RGCs), Müller glial cells (MGs), microglia, and astrocytes (ASTs). Through reclustering, we identified known subtypes of cone PRs as well as additional unreported subpopulations and corresponding markers in BCs. Additionally, we linked inherited retinal diseases (IRDs) to certain cell subtypes or subpopulations through enrichment analysis. We next constructed extensive intercellular communication networks and identified ligand-receptor interactions that play crucial roles in regulating neural cell development and immune homeostasis in the retina. Intriguingly, we found that the status and functions of PRs changed drastically between the young children and adult retina. Overall, our study offers the first retinal cell atlas in infants and young children dissecting the heterogeneity of the retina and identifying the key molecules in the developmental process, which provides an important resource that will pave the way for research on retinal development mechanisms and advancements in regenerative medicine concerning retinal biology.

Glioma survival prediction from whole-brain MRI without tumor segmentation using deep attention network: a multicenter study.

OBJECTIVES: To develop and validate a deep learning model for predicting overall survival from whole-brain MRI without tumor segmentation in patients with diffuse gliomas. METHODS: In this multicenter retrospective study, two deep learning models were built for survival prediction from MRI, including a DeepRisk model built from whole-brain MRI, and an original ResNet model built from expert-segmented tumor images. Both models were developed using a training dataset (n = 935) and an internal tuning dataset (n = 156) and tested on two external test datasets (n = 194 and 150) and a TCIA dataset (n = 121). C-index, integrated Brier score (IBS), prediction error curves, and calibration curves were used to assess the model performance. RESULTS: In total, 1556 patients were enrolled (age, 49.0 ± 13.1 years; 830 male). The DeepRisk score was an independent predictor and can stratify patients in each test dataset into three risk subgroups. The IBS and C-index for DeepRisk were 0.14 and 0.83 in external test dataset 1, 0.15 and 0.80 in external dataset 2, and 0.16 and 0.77 in TCIA dataset, respectively, which were comparable with those for original ResNet. The AUCs at 6, 12, 24, 26, and 48 months for DeepRisk ranged between 0.77 and 0.94. Combining DeepRisk score with clinicomolecular factors resulted in a nomogram with a better calibration and classification accuracy (net reclassification improvement 0.69, p < 0.001) than the clinical nomogram. CONCLUSIONS: DeepRisk that obviated the need of tumor segmentation can predict glioma survival from whole-brain MRI and offers incremental prognostic value. KEY POINTS: • DeepRisk can predict overall survival directly from whole-brain MRI without tumor segmentation. • DeepRisk achieves comparable accuracy in survival prediction with deep learning model built using expert-segmented tumor images. • DeepRisk has independent and incremental prognostic value over existing clinical parameters and IDH mutation status.