Identifying Diagnostic Biomarkers for Glaucoma Based on Transcriptome Combined with Mendelian Randomization

doi:10.21203/rs.3.rs-4842636/v1

Download PDF

Article

Identifying Diagnostic Biomarkers for Glaucoma Based on Transcriptome Combined with Mendelian Randomization

https://doi.org/10.21203/rs.3.rs-4842636/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Glaucoma poses a major health challenge, yet reliable biomarkers for diagnosis and treatment are scarce. This study employed Mendelian randomization and bioinformatics to uncover potential biomarkers. The GSE9944 dataset was used for training and validation in glaucoma research. Differentially expressed genes (DEGs) were identified through differential expression analysis. The protein-protein interaction network (PPI) and functional enrichment were conducted. MR analysis selected DEGs for support vector machine-recursive feature elimination (SVM-RFE), and genes with high differential expression and an area under the curve (AUC) > 0.7 were deemed biomarkers. Biomarker-based analysis, network design, and drug prediction followe. Using 836 DEGs, the PPI network showed diverse interactions, including ATG14-UVRAG. DEGs were enriched in PI3K-Akt and MAPK pathways. MR analysis linked 113 DEGs to glaucoma, with 57 genes matching expression trends. SVM-RFE identified six signature genes, with ATP6V0D1 and FAM89B as biomarkers (AUC > 0.7). Finally, the molecular regulatory networks revealed that biomarkers might involve several regulatory pathways, including ATP6V0D1-hsa-let-7b-5p-HCG18 and ATP6V0D1 or FAM89B-CREB1. The ATP6V0D1 and FAM89B recognized as glaucoma biomarkers, aiding diagnosis, treatment and deepening glaucoma mechanisms understanding

Glaucoma

ATP6V0D1

Mendelian randomization

FAM89B

The glaucomas encompass a group of conditions that result in irreversible blindness, characterized by the progressive loss of retinal ganglion cells (RGCs). Given the increasing number and proportion of elderly individuals in the population, it is projected that 111.8 million people will be affected by glaucoma in 2040[1] This condition is firmly associated with a decline in quality of life, making it an essential medical issue necessitating treatment. However, current treatments for glaucoma primarily focus on drugs that lower intraocular pressure (IOP) and surgical interventions, which may not invariably prevent vision loss in certain cases[2]. Therefore, alternative strategies are required to reduce or reverse the progressive neurodegeneration associated with this disease. Unfortunately, effective neuroprotective therapies have yet to be developed, and more than half of glaucoma cases remain undiagnosed until irreversible optic nerve damage has occurred[3]. Furthermore, since the precise molecular pathway of glaucoma pathogenes is remain unknown, new biomarkers are urgently needed to aid in the diagnosis and treatment of glaucoma.

The Mendelian randomization (MR) analysis, considered a promising epidemiological method, has been proposed to precisely assess the potential causality between exposures and outcomes.[4].The concept of MR can be likened to a randomized controlled trial, where the random assortment of alleles leads to a random assignment of exposures. Furthermore, the MR method is unaffected by environmental risk factors and operates independently from disease progression. In contrast to observational studies, using genetic variation as an instrumental variable (IV) in MR analysis can help prevent environmental confounders and reverse causality. Currently, the majority of MR studies in glaucoma primarily focus on establishing causal relationships between glaucoma and other diseases such as myopia[5], hypertension[6], ankylosing spondylitis[7], rheumatism[8] and so on. Additionally, investigations into the association between glaucoma and factors like coffee consumption[9] and body shape[10] have been conducted. Furthermore, previous bioinformatic analyses of glaucoma did not employ the MR method to assess causality between potential biomarkers and the development of glaucoma[11]. Consequently, there is a lack of sufficient genetic evidence regarding the association of these biomarkers with glaucoma.

In this study, our objective was to conduct an MR analysis by utilizing summary-level statistics obtained from previous GWASs and transcriptomic data. The aim was to investigate the causal relationship between biomarkers and glaucoma in a more efficient manner. By applying machine learning methods and MR to RNA microarray sequencing data from case and control (normal) samples from the Glaucoma Sequencing Project, we identified 2 glaucoma biomarkers from human optic nerve head (ONH) astrocytes with a view to exploring their roles in regulating glaucoma signaling pathways and molecular mechanisms.

The 836 DEGs in Glaucoma were remarkably associated with signaling pathways such as PI3K-Akt and MAPK

Following differential expression analysis, a total of 836 DEGs (361 elevated and 475 reduced) emerged between case and the control group in the training set (Fig. 1a and 1b, Supplementar Table S1). And the PPI network revealed a wide range of interactions between DEG-encoded proteins, including ATG14-UVRAG, ATP5F1C-ATP5F1D, and BAD-BCL2L1 (Fig. 1c). Additional functional enrichment indicated that DEGs were engaged in a total of 1,044 GO functional entries (such as cell growth, focal adhesion and cadherin binding) and 51 signaling pathways (such as PI3K-Akt signaling pathway, MAPK signaling pathway and Human papillomavirus infection) (Fig. 1d and 1e).

The 57 DEGs showed a significant causal association with glaucoma

To identify DEGs with a causal connection to glaucoma, The MR methodology was employed. The IVW algorithm's P-value revealed that 113 DEGs were substantially linked with glaucoma (Supplementar Table S2). Supplementar Table S3 demonstrated that 30 DEGs were down-regulated in the case group while being protective for glaucoma in MR data (OR > 1), and 27 DEGs displayed the reverse tendency (OR < 1). Notably, scatter plots and forest plots were utilized to depict 57 DEGs with varying risk characteristics for glaucoma (Supplementar Fig. S1 and S2). The funnel plot, which illustrated the symmetrical distribution of each SNP locus on both sides of the IVW line, proved that the MR analysis followed Mendel's second random law (Supplementar Fig. S3). Finally, sensitivity studies validated the accuracy and reliability of the MR data (Supplementar Table S4 and S5, Supplementar Fig. S3).

ATP6V0D1 and FAM89B has been shown to be diagnostic competent biomarkers in glaucoma

The SVM-RFE analysis of 57 DEGs causally related with glaucoma demonstrated that the sixth feature subset had the highest accuracy, therefore the genes contained within it were designated as feature genes (ATP6V0D1, FAM89B, GABARAPL2, GBAP1, GTF2B, and NIT2) (Fig. 2a). The feature genes were subsequently presented as ROC curves in the training and validation sets, respectively, with AUC values of ATP6V0D1, FAM89B, and GTF2B more than 0.7 in both datasets, indicating excellent diagnostic efficacy for glaucoma (Fig. 2b and 2c). Furthermore expression analysis indicated that ATP6V0D1 and FAM89B were strongly expressed in both datasets, and the expression trend was consistent with the MR results, hence they were chosen as biomarkers (Fig. 2d). Thereafter, nomograms according to biomarkers proved to have a strong diagnostic ability for glaucoma using calibration curves (slope near to 1) and ROC curves (AUC = 1) (Fig. 2e-2g).

Subcellular localization and function of diagnostic biomarkers

The findings of subcellular localization suggested that ATP6V0D1 was primarily distributed within the membrane system, whereas FAM89B was predominantly distributed in the cytoplasm (Fig. 3a and 3b). Meanwhile, GGI data revealed interactions between ATP6V0D1 and FAM89B as well as different genes (e.g., ATP6V0D1-SLC38A9 and FAM89B-VAMP2), primarily involved in phagosome maturation and pH reduction (Fig. 3c). Moreover, GSEA and GSVA were employed to better understand the roles that biomarkers could play in glaucoma. The GSEA results suggested that the two biomarkers co-enriched considerably with Ribosome, Propanoate metabolism, and Glycosaminoglycan degradation (Fig. 3d and 3e). On the other hand, GSVA results indicated that both biomarkers were heavily related with the functions of bile acid metabolism, cholesterol homeostasis, and apoptosis (Fig. 3f and 3g).

lncRNA-miRNA-mRNA regulatory network and transcription factor regulatory network of two biomarkers

To further understand the molecular regulatory mechanisms involved in biomarkers, we construct lncRNA-miRNA-mRNA and TF-mRNA regulatory networks, respectively. First, 48 miRNAs targeting two biomarkers and 509 lncRNAs targeting 35 miRNAs were predicted in the miRNet database. Following that, the lncRNA-miRNA-mRNA regulatory network was built, which included two mRNAs, 35 miRNAs, and 24 lncRNAs (e.g.,ATP6V0D1-hsa-let-7b-5p-HCG18andFAM89B-hsa-mir-766-5p-LINC00963) (Fig. 4a). Meanwhile, 12 TFs targeting two biomarkers were predicted in the JASPAR database, including CREB1, ELK1, and USF2 (Fig. 4b).

Clinical application value and drug prediction of two biomarkers

To comprehend the clinical application value of the biomarkers, the CTD database was employed to forecast the eye disorders associated with them. The findings revealed a strong link between optic nerve diseases and ATP6V0D1, as well as color vision detects and FAM89B (Fig. 4c). Finally, 26 medicines targeting ATP6V0D1, including digoxin, phosphoric acid, and 2,5-di-tert-butylhydroquinone were predicted in the GeneCard database (Supplementar Table S6).

Glaucoma is characterized as a progressive degenerative condition affecting the optic nerve, resulting in the death of retinal ganglion cells and gradual loss of visual field. Preserving these cells is crucial for maintaining vision in individuals with this eye disease. However, due to limited understanding of the underlying mechanisms causing damage to the optic nerve, there are currently no available treatments targeting this aspect.[3]. Recent studies have highlighted that astrocytes, found in various parts of the retina, optic nerve head (ONH), and visual brain, play diverse roles in maintaining balance and regulating neuronal functions. [12]. Promisingly, emerging evidence suggests that therapeutic approaches focusing on modulating astrocyte reactivity can offer alternative options for treating glaucoma alongside traditional methods aimed at controlling intraocular pressure (IOP)[13, 14]. However this requires exploring a thorough understanding of the distinct molecular mechanisms of astrocytes in glaucoma. In this study, we used MR analysis and machine learning screening based on the GEO database to obtain two promising genes for glaucoma diagnostic biomarkers: ATP6V0D1 and FAM89B GSVA results indicated that both biomarkers were firmly associated with apoptosis, bile acid metabolism, and cholesterol homeostasis functions, providing novel insights into the molecular pathogenesis mechanisms underlying the reactivity of astrocyte in glaucoma.

ATP6V0D1 plays a crucial role as a primary component of ATP-driven vacuolar-type ATPase (V-ATPase) found in lysosomes. Previous studies have indicated that V-ATPase functions as a proton pump, maintaining the acidic pH levels within lysosomes to facilitate hydrolase activity in neurons[15]. ATP6V0D1 has been identified as a therapeutic target capable of protecting neuronal cells from AGE-induced apoptosis [16], and our findings are consistent with this trend. Yang et al[11] also identified hub genes include AT-P6V0D1 in glaucoma by three machine learning models. The GGI data revealed that the interaction between ATP6V0D1 and SLC38A9 is primarily involved in phagosome maturation and pH reduction. Recent findings have provided evidence that nutrient cues trigger the activation of mechanistic target of rapamycin complex 1 (mTORC1), a crucial regulator of autophagy, at the lysosome [17].The activation of mTORC1 by cholesterol relies on the presence of conserved cholesterol-responsive motifs in SLC38A9, a transmembrane protein located in the lysosome [18]. Furthermore, studies suggest that stimulating mTORC1 in retinal ganglion cells (RGCs) may enhance visual function, axonal transport, synaptic integrity, and axonal regeneration in mouse models with glaucoma [19]. Activation of mTORC1 plays a critical role in various neuroprotective processes mediated by multiple signals and molecules [20]. In addition, mTORC1 is downstream of PI3K/AKT in RGCs, which is necessary for optic nerve regeneration[21]. Therefore, the ATP6V0D1 and SLC38A9 interactions may promote the protective effects of the optic nerve by taking part in the PI3K/Akt/mTORC1 pathway. This is consistent with our results that the 82 DEGs are significantly enriched in MAPK and PI3K-Akt. Additionally, PI3K-Akt signaling pathway [22] and MAPK signaling[23] pathway have been reported that play an critical role in mediating retinal ganglion cell apoptosis in glaucoma. The PI3K/Akt pathway plays a crucial role in intracellular signaling, promoting cell proliferation, inhibiting apoptosis, and inducing angiogenesis through the activation of various downstream regulatory elements. In a rat model of glaucoma induced by injecting hypertonic saline into the limbal veins, there was a sustained reduction in Akt activation observed in both the ocular-hypertensive retina and optic nerve [24]. Identification of factors that can activate the PI3K/Akt pathway presents promising opportunities for novel molecular therapies targeting glaucoma.

FAM89B is a protein-coding gene, according to the gene card summary. The GO annotation associated with this gene includes transcriptional corepressor binding and Negatively modulates transforming growth factor (TGF-β)-induced signaling. The previous studies have demonstrated that glaucoma patients exhibit elevated levels of endothelin (ET-1) and transforming growth factor (TGF-β) in their serum and/or aqueous humor. These molecules upregulate the expression of extra-cellular matrix (ECM) and matrix metalloproteinases (MMPs) proteins in the optic papilla (ONH), leading to tissue remodeling[25]. Therefore, our hypothesis is that FAM89B may exert a negative regulatory effect on TGF-β-induced signaling, thereby inhibiting the collective remodeling of the optic papilla and preventing damage to RGCs.

The ceRNA and TF-mRNA networks are constructed by predicting the upstream and downstream regulatory factors of the biomarkers. It was observed that both biomarkers intersected with the TF CREB1. CREB1 was reported as a potential biomarker for glaucoma[20]. Synaptic signals are known to activate various downstream pathways including MAPK and PI3K-Akt in astrocytes, which can activate the CREB TF [26]. The CREB TF is involved in a variety of neurodegenerative disorders. Reactive astrocytes have received attention as a source of neurodegenerative disorders, while CREB expression in reactive astrocytes has been shown to be neuroprotective [27]. A comprehensive understanding of the intricate regulatory mechanisms of these astrocyte genes in glaucoma pathogenesis remains elusive. Additional investigations are imperative to elucidate their molecular underpinnings in the context of glaucoma.

To investigate the potential therapeutic applications of these two diagnostic molecules in glaucoma, we examined the list of glaucoma-related genes in GeneCards and identified ATP6V0D1 as a candidate. Our analysis predicted 26 drugs that target ATP6V0D1, including digoxin which was also found in the GeneCard database. In recent study, it has been reported digoxin exhibits neuroprotective properties in a rat model of dementia[28]. Additionally, Digoxin and semisynthetic derivatives also appear to promote neuroprotection in cells exposed to chemical ischemia[29] and contributes to nerve healing through their anti-inflammatory effect on IL-17[30]. These results indicate that ATP6V0D1 could be a potential target for glaucoma.

Overall, this study is the first to identify two glaucoma biomarkers, ATP6V0D1 and FAM89B, using machine learning models and MR methods for the first time. The role of ATP6V0D1 and FAM89B in modulating signaling pathways and molecular mechanisms in glaucoma has also been partially revealed. This research is beneficial for gaining more insight into the pathophysiological changes associated with glaucoma and for identifying current targets and strategies for protecting the optic nerve in patients.

Data source

The Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) provided the glaucoma transcriptome dataset GSE9944, which relies on GPL571 (training set) and GPL9944 platform (validation set) (Human Optic Nerve Head Astrocytes). The training set contained six glaucoma (cases) and 36 control samples. However, the validation set consists of 13 cases and 6 control samples. Finally, the glaucoma GWAS, finngen_R9_H7_GLAUCOMA_POAG) dataset was collected from the Integrative Epidemiology Unit (IEU) OpenGWAS database (https://gwas.mrcieu.ac.uk/), which contained 16,380,452 single nucleotide polymorphisms (SNPs) from 213,613 European samples (3,412 cases and 210,201 controls).

Analysis of differential expressions and functional enrichment

In the training set, differential expression analysis was conducted between cases and controls using limma (v 3.50.1)[31]to extract differentially expressed genes (DEGs) (p-value < 0.05, |log₂FC|>0.5), which were displayed with volcano and heat maps. Then, the STRING database (https://string-db.org/) was applied to create a protein-protein interaction network (PPI) for the proteins encoded by DEGs (medium confidence = 0.4). Meanwhile, the clusterProfiler (v 4.2.2)[32] was applied to examine DEGs' Gene Ontology (GO) and Kyoto Encyclopedia of Genes (KEGG) enrichment to understand their involvement in biological processes (p-value < 0.05).

MR data pre-processing

The three major assumptions that underpinned the MR investigation were as follows: (1) exposure factors and IVs have a strong association; (2) confounders have no effect on IVs; and (3) IVs only influence outcomes across exposure factors. To study the causal association between DEGs (exposure factors) and glaucoma (outcome), we applied MR analysis. First, DEGs were read and IVs were filtered (p = 5*10^− 7) to retrieve IVs associated with DEGs via the function extract_instruments of the R package "TwoSampleMR" (v 0.5.6)[33] (clump = TRUE to remove the IVs of Linkage disequilibrium (LD); r2 = 0.001; kb = 100). The GWAS data for glaucoma were then read using the TwoSampleMR package's extract_outcome_data function, and IVs that were not relevant to glaucoma were further filtered according to IVs that were associated with DEGs (proxies = TRUE; rsq = 0.8). Ultimately, IVs with F-statistics smaller than 10 were excluded and subjected to Steiger reverse causality with respect to the directionality_test function.

MR analysis

To test for DEGs causally associated to glaucoma, effect alleles and effect sizes were first harmonized with the R package "Two Sample MR" function harmonise_data. Subsequently, MR analysis was performed by the mr function combined with five algorithms MR Egger[34], Weighted median[35], Simple mode[36], Weighted mode[37]including Inverse variance weighted (IVW)[38]. The IVW algorithm has a P-value of less than 0.05, demonstrating a link between the corresponding exposure factor and the outcome. Meanwhile, if the equivalent odds ratios (ORs) value is greater than one, it means that the exposure factor acts as a risk factor for outcome, and vice versa for protection. Lastly, scatter plot, forest plot, and funnel plot were applied to display the MR results.

Sensitivity tests

Sensitivity tests were performed to ensure that the MR results were reliable. First, the function mr_heterogeneity from the R package "Two Sample MR" was implemented to detect heterogeneity between exposure factors and outcome samples (P > 0.05) [39]. Then, utilizing the R package "Two Sample MR" function mr_pleiotropy_test, a horizontal pleiotropy test was run to identify the presence of confounders interacting with SNPs in influencing the outcome via exposure factors (P > 0.05)[40]. The data were also subjected to MR Pleiotropy RESidual Sum and Outlier (PRESSO) testing via the mr_presso tool (v 1.0.0)[41]. Finally, the R package TwoSampleMR function mr_leaveoneout was applied to verify the robustness of MR results[42].

Screening for biomarkers

Focusing on the DEGs causally related with glaucoma, we applied the caret package[43] (5-fold cross-validation) to screen for feature genes in the training set using support vector machine-recursive feature elimination (SVM-RFE). Afterwards, feature gene expression evaluation and receiver operating characteristic (ROC) curve plotting were carried out simultaneously on the training and validation sets. Biomarkers were chosen by considering genes that met the criteria for substantial differential expression, consistent expression trend, and an area under the curve (AUC) value of ROC greater than 0.7 in both datasets. The rms software (v 6.5.0)[44]was employed to create nomograms relying on the biomarkers to predict the prevalence of glaucoma. Later on, calibration curves and ROC curves were utilized to evaluate the model's diagnostic ability.

Functional analysis of biomarkers

First, "c2.cp.kegg.v11.0.symbols " was downloaded from the Molecular Signatures Database (MSigDB, https://www.gsea-msigdb.org/gsea/msigdb) and utilized as an internal reference gene set in the training set for Gene Set Enrichment Analysis (GSEA) on biomaker separately. Meanwhile, the samples in the training set were divided into high and low expression groups by considering the median biomarker expression values, and the Gene set variation analysis (GSVA) scores of each group were computed and compared using the GSVA package (v 1.42.0) [45](the reference gene set was C2: Hallmark in the MSigDB database).

Localization and clinical correlation analysis of biomarkers

The UniProt database (https://www.uniprot.org/) was employed to achieve subcellular localization of biomarkers. To better understand the relationships between biomarkers and other genes in the process of establishing regulatory effects on glaucoma, a gene-gene interaction network (GGI) network was built in the GeneMANIA database (http://www.genemania.org). Also, the Comparative Toxicogenomics Database (CTD) (https://ctdbase.org/) was utilized to investigate the association between biomarkers and ocular disorders. Furthermore, the GeneCard database (https://www.genecards.org/) was applied to predict medications that could target the biomarkers.

Molecular regulation associated with biomarkers

In order to understand the molecular regulatory mechanisms in which the biomarkers may be involved in, we conducted correlation predictions. First, the miRNet database (https://www.mirnet.ca/) was employed for calculating the target miRNAs of biomarkers, as well as the related lncRNAs (regulatory number ≧ 10) of the target miRNAs. Second, transcription factors (TFs) that target the biomarkers were predicted using the miRNet platform's JASPAR database (https://jaspar.elixir.no/). Eventually, Cytoscape was applied to display the regulatory networks of lncRNA-miRNA-mRNA and TF-mRNA.

Statistical analysis

Bioinformatics analyses were carried out in R (v 4.2.2). The data from separate groups were compared using the Wilcoxon test (P < 0.05).

Data availability

The datasets analysed during the current study are available in the [GEO] repository, [http://www.ncbi.nlm.nih.gov/geo/]; and the [IEU OpenGWAS] repository, [https://gwas.mrcieu.ac.uk/].

Acknowledgements

This work was supported by the Joint Funds for the Innovation of Science and Technology of Fujian Province (grant number 2023Y9354), and Fujian Provincial Natural Science Foundation (grant numbers 2021J01360)

Author contributions:

N.X and XL.L contributed to the conception of the study. N X and XL.L contributed significantly to data analysis.CY.M, YZ.Q and XX.Z performed the analysis with constructive discussions. XL.L and N.X. wrote the manuscript, and all authors read and approved the final manuscript

Additional Information

Competing Interests

The authors declare no competing interests.

Kang J. M., Tanna A. P. Glaucoma. Med Clin North Am. 105, 493-510 (2021).
Jayaram H., Kolko M., Friedman D. S., Gazzard G. Glaucoma: now and beyond. Lancet. 402, 1788-801 (2023).
Nucci C., Martucci A., Giannini C., Morrone L. A., Bagetta G., Mancino R. Neuroprotective agents in the management of glaucoma. Eye (Lond). 32, 938-45 (2018).
Sekula P., Del Greco M. F., Pattaro C., Kottgen A. Mendelian randomization as an approach to assess causality using observational data. J Am Soc Nephrol. 27, 3253-65 (2016).
Chong R. S., et al. Mendelian randomization implicates bidirectional association between myopia and primary open-angle glaucoma or intraocular pressure. Ophthalmology. 130, 394-403 (2023).
Plotnikov D., et al. High blood pressure and intraocular pressure: a mendelian randomization study. Invest Ophthalmol Vis Sci. 63, 29 (2022).
Li S., Chen M., Zhang Q., Fang M., Xiong W., Bai L. Ankylosing spondylitis and glaucoma in European population: a mendelian randomization study. Front Immunol. 14, 1120742 (2023).
Meng Y., Tan Z., Su Y., Li L., Chen C. Causal association between common rheumatic diseases and glaucoma: a mendelian randomization study. Front Immunol. 14, 1227138 (2023).
Li X., Cheng S., Cheng J., Wang M., Zhong Y., Yu A. Y. Habitual coffee consumption increases risk of primary open-angle glaucoma: a mendelian randomization study. Ophthalmology. 129, 1014-21 (2022).
Yuan R., Liu K., Cai Y., He F., Xiao X., Zou J. Body shape and risk of glaucoma: a mendelian randomization. Front Med (Lausanne). 9, 999974 (2022).
Xie Y., Yu K. Identifying hub genes for glaucoma based on bulk RNA sequencing data and multi-machine learning models. Curr Med Chem. (2024).
Tang Y., Chen Y., Chen D. The heterogeneity of astrocytes in glaucoma. Front Neuroanat. 16, 995369 (2022).
Shinozaki Y., Koizumi S. Potential roles of astrocytes and Muller cells in the pathogenesis of glaucoma. J Pharmacol Sci. 145, 262-7 (2021).
Shinozaki Y., et al. Astrocytic dysfunction induced by ABCA1 deficiency causes optic neuropathy. Sci Adv. 8, eabq1081 (2022).
Mauvezin C., Nagy P., Juhasz G., Neufeld T. P. Autophagosome-lysosome fusion is independent of V-ATPase-mediated acidification. Nat Commun. 6, 7007 (2015).
Zhou X., et al. Schisandrol a protects AGEs-induced neuronal cells death by allosterically targeting ATP6V0d1 subunit of V-ATPase. Acta Pharm Sin B. 12, 3843-60 (2022).
Ratto E., et al. Direct control of lysosomal catabolic activity by mTORC1 through regulation of V-ATPase assembly. Nat Commun. 13, 4848 (2022).
Castellano B. M., et al. Lysosomal cholesterol activates mTORC1 via an SLC38A9-Niemann-Pick C1 signaling complex. Science. 355, 1306-11 (2017).
Belforte N., et al. AMPK hyperactivation promotes dendrite retraction, synaptic loss, and neuronal dysfunction in glaucoma. Mol Neurodegener. 16, 43 (2021).
Ye M., et al. TBK1 knockdown alleviates axonal transport deficits in retinal ganglion cells via mTORC1 activation in a retinal damage mouse model. Invest Ophthalmol Vis Sci. 64, 1 (2023).
Huang H., Kaur S., Hu Y. Lab review: molecular dissection of the signal transduction pathways associated with PTEN deletion-induced optic nerve regeneration. Restor Neurol Neurosci. 37, 545-52 (2019).
Xu K., et al. MicroRNA-145-5p targeting of TRIM2 mediates the apoptosis of retinal ganglion cells via the PI3K/AKT signaling pathway in glaucoma. J Gene Med. 23, e3378 (2021).
Peng H., Sun Y. B., Hao J. L., Lu C. W., Bi M. C.,Song E. Neuroprotective effects of overexpressed microRNA-200a on activation of glaucoma-related retinal glial cells and apoptosis of ganglion cells via downregulating FGF7-mediated MAPK signaling pathway. Cell Signal. 54, 179-190 (2019).
Husain S., Ahmad A., Singh S., Peterseim C., Abdul Y., Nutaitis M. J. PI3K/Akt Pathway: A Role in delta-opioid receptor-mediated RGC neuroprotection. Invest Ophthalmol Vis Sci. 58, 6489-99 (2017).
Fuchshofer R. The pathogenic role of transforming growth factor-beta2 in glaucomatous damage to the optic nerve head. Exp Eye Res. 93, 165-9 (2011).
Kim J., Kaang B. K. Cyclic AMP response element-binding protein (CREB) transcription factor in astrocytic synaptic communication. Front Synaptic Neurosci. 14, 1059918 (2022).
Pardo L., et al. CREB regulates distinct adaptive transcriptional programs in astrocytes and Neurons. Sci Rep. 7, 6390 (2017).
Erdogan M. A., Kirazlar M., Yigitturk G., Erbas O. Digoxin exhibits neuroprotective properties in a rat model of dementia. Neurochem Res. 47, 1290-8 (2022).
de Souza Goncalves B., et al. Evaluation of neuroprotective activity of digoxin and semisynthetic derivatives against partial chemical ischemia. J Cell Biochem. 120, 17108-22 (2019).
Gurkan G., et al. The healing effect of digoxin on peripheral nerve damage and its relation to IL-17/IL-10. Turk Neurosurg. 33, 258-64 (2023).
Ritchie M. E., et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
Yu G., Wang L. G., Han Y., He Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 16, 284-7 (2012).
Huang P., Zhang P. F., Li Q. Causal relationship between cannabis use and cancer: a genetically informed perspective. J Cancer Res Clin Oncol. 149, 8631-8 (2023).
Bowden J., Davey Smith G., Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. 44, 512-25 (2015).
Bowden J., Davey Smith G., Haycock P. C., Burgess S. Consistent estimation in mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 40, 304-14 (2016).
Hemani G., et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 7, (2018).
Hartwig F. P., Davey Smith G., Bowden J. Robust inference in summary data mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. 46, 1985-98 (2017).
Burgess S., Scott R. A., Timpson N. J., Davey Smith G., Thompson S. G., Consortium E.-I. Using published data in Mendelian randomization: a blueprint for efficient identification of causal risk factors. Eur J Epidemiol. 30, 543-52 (2015).
Qin Q., et al. Systemic lupus erythematosus is causally associated with hypothyroidism, but not hyperthyroidism: A Mendelian randomization study. Front Immunol. 14, 1125415 (2023).
Davey Smith G., Hemani G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Hum Mol Genet. 23, R89-98 (2014).
Verbanck M., Chen C. Y., Neale B., Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 50, 693-8 (2018).
Cui Z., Feng H., He B., He J., Tian Y. Relationship between serum amino acid levels and bone mineral density: a mendelian randomization study. Front Endocrinol (Lausanne). 12, 763538 (2021).
Dong J., Xu M. A 19‑miRNA Support vector machine classifier and a 6‑miRNA risk score system designed for ovarian cancer patients. Oncol Rep. 41, 3233-43 (2019).
Liu T. T., et al. Identification of CDK2-related immune forecast model and ceRNA in lung adenocarcinoma, a pan-cancer analysis. Front Cell Dev Biol. 9, 682002 (2021).
Wang B., et al. Identification and validation of chromatin regulator-related signatures as a novel prognostic model for low-grade gliomas using translational bioinformatics. Life Sci. 336, 122312 (2024).

No competing interests reported.

Download PDF

Reviewers invited by journal
13 Oct, 2024
Editor assigned by journal
13 Oct, 2024
Editor invited by journal
20 Aug, 2024
Submission checks completed at journal
16 Aug, 2024
First submitted to journal
01 Aug, 2024

You are reading this latest preprint version

Identifying Diagnostic Biomarkers for Glaucoma Based on Transcriptome Combined with Mendelian Randomization

Status:

Version 1

Abstract

Figures

Introduction

Results

The 57 DEGs showed a significant causal association with glaucoma

ATP6V0D1 and FAM89B has been shown to be diagnostic competent biomarkers in glaucoma

Subcellular localization and function of diagnostic biomarkers

lncRNA-miRNA-mRNA regulatory network and transcription factor regulatory network of two biomarkers

Clinical application value and drug prediction of two biomarkers

Discussion

Conclusion

Materials and methods

Data source

Analysis of differential expressions and functional enrichment

MR data pre-processing

MR analysis

Sensitivity tests

Screening for biomarkers

Functional analysis of biomarkers

Localization and clinical correlation analysis of biomarkers

Molecular regulation associated with biomarkers

Statistical analysis

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1