Sean Lim, The Stony Brook School, Stony Brook, New York, USA
Abstract
As studies on COVID-19 progress, scientists observe the intimate relationship between SARS-CoV-2 infection and brain damage through numerous clinical cases. However, the way in which SARS-CoV-2 causes neurological dysfunctions remains vague even today. This study performed gene network analysis of ten differential expressed genes in SARS-CoV-2 infected patients’ brains and controls using data from previous research. We hypothesize that the gene network analysis will reveal the linkage between gene expression in the brain and neurological dysfunctions observed in COVID-19 patients. First, we used GeneMania, a web-based program that predicts the function of the gene sets, to analyze the functional association network of the genes. Second, we used the EnrichR program to rank enriched terms and collect the functions of the gene list. GeneMania predicts that ten genes are strongly associated with the hydrogen peroxide metabolic process. The analysis result from EnrichR indicates that the gene sets are top-ranked in hypochromic microcytic anemia and hearing loss for clinical phenotypes. This study concluded that SARS-CoV-2 may alter gene expression in the human brain inducing hydrogen peroxide inside the human brain, which may be associated with hypochromic microcytic anemia and hearing loss.
Keywords: SARS-CoV-2, Brain Damage, Gene Network Analysis
Introduction
Evolving into a global pandemic, COVID-19, a disease caused by SARS-COV-2 infection, has shown mild to severe levels of respiratory symptoms in patients (Lai et al., 2020). Common symptoms include fever, cough, headache, and loss of tastes or smells (Çalıca Utku et al., 2020). However, some patients develop severe symptoms such as hypoxemia that can contribute to numerous organ dysfunction (He et al., 2020). In addition to the respiratory symptoms, new evidence has emerged revealing the relationship between COVID-19 and brain damage (Nuzzo et al., 2021). It is shown that coronavirus can induce multipronged damage to the brain cells by directly attacking the cells, impeding blood flow to the brain, or producing immune molecules that harm brain cells (Romero et al., 2021).
A recent study investigated the brain of elderly patients who died of COVID-19 to identify the possible effect of SARS-CoV-2 on the human brain (Gagliardi et al., 2021). Researchers performed a transcriptomic profile in the frontal cortex comparing COVID-19 patients and matched controls by RNA-seq analysis. As a result, they identified ten genes transcribed to mRNAs that were differentially expressed when the frontal cortex of COVID-19 patients was compared to controls (Gagliardi et al., 2021).
Comparing transcript levels between healthy and diseased individuals allows for the identification of differentially expressed genes (DEG), which may be correlated to, the cause of, or a consequence of the disease (Porcu et al., 2021). To analyze the functional association of DEGs, we used the GeneMANIA program to explore the interactive functional association network, illustrating the relationships among the DEGs (Warde-Farley et al., 2010). Furthermore, we perform gene enrichment analysis, a popular method for analyzing the DEGs (Hung et al., 2012). We used the EnrichR program to explore Gene Ontology (GO), which provides the concept of associating a collection of genes with a functional biological term systematically (Kuleshov et al., 2016). We hypothesized that ten previously identified DEGs are associated with the brain damage found in COVID-19 patients. To test our hypothesis, we analyzed the functional gene network of ten DEGs with both GeneMANIA and EnrichR programs.
Materials and Methods
Gene network analysis by GeneMania
GeneMania is an open-source web program for predicting gene function and analyzing gene lists. GeneMania searches functionally similar genes using genomics and proteomics data given a query gene list. We inputted ten differentially expressed genes in Table 1 for the query. After analyzing the function of the query genes, GeneMania weights each functional genomic dataset according to its predictive value. Then, the detailed gene network analysis result was retrieved from this program. The gene ontology categories also provide the information to predict the function of the gene set.
Gene enrichment analysis by EnrichR
EnrichR is a gene set search engine that predicts hundreds of thousands of annotated gene sets. First, ten DEGs from table 1 were used for the query genes. Different methods provided by EnrichR ranked the matching gene sets. We only focused on Human Phenotype Ontology (HPO) and Phenotype-Genotype Integrator (PheGenI) gene enrichment analysis results. Each method provides a table with the biological phenotype term or identification code, overlapping gene number, p-value, and the list of genes involved in each phenotype term. The table was ranked by the lowest p-value.
Results
The experiment was conducted by comparing the gene expression between eight non-COVID-19 patients and nine COVID-19 patients who had died from the disease (Gagliardi et al., 2021). As a result, ten protein-coding genes were identified as differentially expressed genes in the frontal lobe area of the COVID-19 patients. Four protein-coding genes (SLC14A1, HIF3A, PAPLN, RGS5) yielded negative values in log2FoldChange, indicating underexpression of these genes in the infected patients’ brains. On the contrary, six protein-coding genes (SERGEF, NCL, ZNF622, HBA2, HBB, HBA1) yielded positive values, suggesting that those genes were overexpressed in the infected brain (Table 1). HBA2, HBB, and HBA1 were the genes that especially showed the largest difference in expression between infected and uninfected patients.
Table 1. Ten differentially expressed genes in COVID-19 group compared to controls.
The gene sets associated with ten DEGs are located around the ten DEGs (Figure 1). The red line indicates the physical interaction between the genes. The purple line indicates the co-expressed genes. The blue line indicates the co-localized genes. The yellow line indicates the genes that share the same protein domains.
Figure 1. Gene network analysis of ten DEGs, located in the center as slash patterned.
GeneMANIA predicts the function of gene sets and provides the gene network analysis (physical interaction, co-expression, co-localization, sharing the same protein domains). (Warde-Farley et al., 2010) We found that nine genes, HBZ, HBD, HBM, HBE1, HBQ1, HBG1, HBG2, HBB, and HBA1, play significant roles in the hydrogen peroxide metabolic process (p-value= 2.36e-14), which is one of the important factors regulating brain maintenance and activity. Interestingly, among these genes, HBA1 and HBB were the two genes highly overexpressed in infected patients' brains.
Table 2. HPO gene enrichment analysis result using EnrichR program.
The Human Phenotype Ontology (HPO) demonstrates standardized phenotypic abnormalities found in human disease (Köhler et al., 2021). Each term in the HPO indicates a phenotypic abnormality. HPO contains over 13,000 terms for human diseases (Köhler et al., 2021). We found that HBB, HBA2, HBA1 genes are significantly associated with hypochromic microcytic anemia, a type of anemia in which the circulating RBCs are smaller than the usual RBCs (microcytic) and thus possess decreased red color (hypochromic). In a patient's body, anemia can induce brain iron deficiency, causing the disruption of neurophysiological mechanisms such as compromised cognitive development (Jáuregui-Lobera, 2014). Severe levels of anemia can cause significant brain damage, including cognitive dysfunction and cerebral vascular regulation impairment (Hare, 2004). In addition, patients who experience anemia are also found to have higher mortality rates, suggesting the potential relationship between the level of anemia and the effect of COVID-19 in patients (Zuin et al., 2021). It was interesting that HBB, HBA2, HBA1 genes are also associated with abnormal hemoglobin, hypochromic anemia, and abnormality of the heme biosynthetic pathway, which all displayed small p-values relative to other terms. These abnormalities associated with hemoglobin describe the disease’s negative influence on the transportation of oxygen to the brain.
Table 3. PheGenI gene enrichment analysis result using EnrichR program.
We further investigated the ten DEGs association with clinical phenotype using Phenotype-Genotype Integrator (PheGenI) gene enrichment analysis (Ramos et al., 2014). This analysis links genome-wide association study (GWAS) catalog data with several databases provided at the National Center for Biotechnology Information (NCBI), including basic gene information (Ramos et al., 2014). Clinicians and epidemiologists often used this phenotype-oriented resource to follow up results from GWAS, prioritizing a list of genes to search for the association between gene function and clinical phenotype. We found that the NCL gene was significantly associated with hearing loss phenotype among ten DEGs. Previous studies suggested an association between more persistent hearing problems and COVID-19 since the beginning of the pandemic (Saniasiaya, 2021). To date, the prevalence of auditory symptoms such as sudden or progressive hearing loss is unclear in COVID-19 patients. However, many studies speculated that their presence might be an early sign of thrombosis or the spread of the infection into the brain (McIntyre et al., 2021). The other terms, such as blood platelets, erythrocyte indices, and blood are linked with abnormal red blood cells development (English et al., 2015). Overall, NCL gene was significantly associated with hearing loss and RGS5.
Discussion
This study investigates the alteration of gene expression levels in COVID-19 infected patients. Using research tools, we were able to find altered genes associated with hydrogen peroxide in COVID-19 patients, which can directly impair brain cells. We also found several genes related to anemia and blood-related malfunction, which can negatively influence the brain condition and are associated with symptoms caused by brain damage. Hearing loss was the most significant phenotype expressed due to brain damage. These findings provide fundamental information, which, we believe, can be used for developing a treatment for COVID-19 dependent brain damage in the future.
This is the first study that discovered the association between COVID-19 infected patient’s brain damage and the altered gene expression in the patient's brain. We found hydrogen peroxide was mainly associated with the altered gene in COVID-19 patients’ brains. A previous study indicated that Amyloid β protein (Aβ), which is associated with the plaques in the brains of Alzheimer’s patients, causes increased levels of hydrogen peroxide accumulation in brain cells. Aβ-induced hydrogen peroxide causes neuron damage in the human brain (Behl et al., 1994). This result indicates how hydrogen peroxide produced inside the human brain can severely damage the brain.
The severity of hydrogen peroxide is also exhibited in many other viruses other than COVID-19, including HIV, a widespread disease today. A previous study indicated that HIV-1 encodes Nef protein that leads to increased sensitivity of human astrocytes to hydrogen peroxide. This study showed that Nef protein leads to neuronal dysfunction and the development of dementia by the death of astrocytes, thus decreasing their tolerance to hydrogen peroxide (Masanetz & Lehmann, 2011). These research works reveal that virus-infected patients’ brains may exhibit lower tolerance of hydrogen peroxide, resulting in severe brain damage.
Even though we performed extensive gene network analysis using different programs, this study has numerous limitations. First, we did not define the precise mechanism of how COVID-19 influences brain damage. We have only laid out a limited spectrum of gene networks possibly associated with brain damage. In addition, we only conducted investigations on the top ten DEGs, which were analyzed from the previous studies, and used gene networks and enrichment data as our only analysis tools. Therefore, further study is needed to evaluate how this gene network affects human brain damage after COVID-19 infection.
In the future, more types of gene investigations can be tested, and direct cell experiments can be performed for further validation of the study result. For example, in vivo mouse experiments should be performed to check whether these gene expression alterations directly regulate brain damage. Also, the change in hydrogen peroxide level in the brain should be measured after COVID-19 infection in the mice. These future experiments may reveal the functional association between brain damage and COVID-19 infection. Overall, this study represents a novel gene network analysis, finding how hydrogen peroxide and blood-related malfunction in the brain can damage the human brain after COVID-19 infection.
References
Behl, C., Davis, J. B., Lesley, R., & Schubert, D. (1994). Hydrogen peroxide mediates amyloid beta protein toxicity. Cell, 77(6), 817–827. https://doi.org/10.1016/0092-8674(94)90131-7
Çalıca Utku, A., Budak, G., Karabay, O., Güçlü, E., Okan, H. D., & Vatan, A. (2020). Main symptoms in patients presenting in the COVID-19 period. Scottish Medical Journal, 65(4), 127–132. https://doi.org/10.1177/0036933020949253
English, E., Idris, I., Smith, G., Dhatariya, K., Kilpatrick, E. S., & John, W. G. (2015). The effect of anaemia and abnormalities of erythrocyte indices on HbA1c analysis: a systematic review. Diabetologia, 58(7), 1409–1421. https://doi.org/10.1007/s00125-015-3599-3
Gagliardi, S., Poloni, E. T., Pandini, C., Garofalo, M., Dragoni, F., Medici, V., Davin, A., Visonà, S. D., Moretti, M., Sproviero, D., Pansarasa, O., Guaita, A., Ceroni, M., Tronconi, L., & Cereda, C. (2021). Detection of SARS-CoV-2 genome and whole transcriptome sequencing in frontal cortex of COVID-19 patients. Brain, Behavior, and Immunity, 97, 13–21. https://doi.org/10.1016/j.bbi.2021.05.012
Hare, G. M. T. (2004). Anaemia and the brain. Current Opinion in Anaesthesiology, 17(5), 363–369. https://doi.org/10.1097/00001503-200410000-00003
He, X., Cheng, X., Feng, X., Wan, H., Chen, S., & Xiong, M. (2020). Clinical Symptom Differences Between Mild and Severe COVID-19 Patients in China: A Meta-Analysis. Frontiers in Public Health, 8, 561264. https://doi.org/10.3389/fpubh.2020.561264
Hung, J.-H., Yang, T.-H., Hu, Z., Weng, Z., & DeLisi, C. (2012). Gene set enrichment analysis: performance evaluation and usage guidelines. Briefings in Bioinformatics, 13(3), 281–291. https://doi.org/10.1093/bib/bbr049
Jáuregui-Lobera, I. (2014). Iron deficiency and cognitive functions. Neuropsychiatric Disease and Treatment, 10, 2087–2095. https://doi.org/10.2147/NDT.S72491
Köhler, S., Gargano, M., Matentzoglu, N., Carmody, L. C., Lewis-Smith, D., Vasilevsky, N. A., Danis, D., Balagura, G., Baynam, G., Brower, A. M., Callahan, T. J., Chute, C. G., Est, J. L., Galer, P. D., Ganesan, S., Griese, M., Haimel, M., Pazmandi, J., Hanauer, M., … Robinson, P. N. (2021). The human phenotype ontology in 2021. Nucleic Acids Research, 49(D1), D1207–D1217. https://doi.org/10.1093/nar/gkaa1043
Kuleshov, M. V., Jones, M. R., Rouillard, A. D., Fernandez, N. F., Duan, Q., Wang, Z., Koplev, S., Jenkins, S. L., Jagodnik, K. M., Lachmann, A., McDermott, M. G., Monteiro, C. D., Gundersen, G. W., & Ma’ayan, A. (2016). Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Research, 44(W1), W90-7. https://doi.org/10.1093/nar/gkw377
Lai, C.-C., Ko, W.-C., Lee, P.-I., Jean, S.-S., & Hsueh, P.-R. (2020). Extra-respiratory manifestations of COVID-19. International Journal of Antimicrobial Agents, 56(2), 106024. https://doi.org/10.1016/j.ijantimicag.2020.106024
Masanetz, S., & Lehmann, M. H. (2011). HIV-1 Nef increases astrocyte sensitivity towards exogenous hydrogen peroxide. Virology Journal, 8, 35. https://doi.org/10.1186/1743-422X-8-35
McIntyre, K. M., Favre, N. M., Kuo, C. C., & Carr, M. M. (2021). Systematic Review of Sensorineural Hearing Loss Associated With COVID-19 Infection. Cureus, 13(11), e19757. https://doi.org/10.7759/cureus.19757
Nuzzo, D., Cambula, G., Bacile, I., Rizzo, M., Galia, M., Mangiapane, P., Picone, P., Giacomazza, D., & Scalisi, L. (2021). Long-Term Brain Disorders in Post Covid-19 Neurological Syndrome (PCNS) Patient. Brain Sciences, 11(4). https://doi.org/10.3390/brainsci11040454
Porcu, E., Sadler, M. C., Lepik, K., Auwerx, C., Wood, A. R., Weihs, A., Sleiman, M. S. B., Ribeiro, D. M., Bandinelli, S., Tanaka, T., Nauck, M., Völker, U., Delaneau, O., Metspalu, A., Teumer, A., Frayling, T., Santoni, F. A., Reymond, A., & Kutalik, Z. (2021). Differentially expressed genes reflect disease-induced rather than disease-causing changes in the transcriptome. Nature Communications, 12(1), 5647. https://doi.org/10.1038/s41467-021-25805-y
Ramos, E. M., Hoffman, D., Junkins, H. A., Maglott, D., Phan, L., Sherry, S. T., Feolo, M., & Hindorff, L. A. (2014). Phenotype-Genotype Integrator (PheGenI): synthesizing genome-wide association study (GWAS) data with existing genomic resources. European Journal of Human Genetics, 22(1), 144–147. https://doi.org/10.1038/ejhg.2013.96
Romero, J. P., Bravo-Martín, A., Oliva-Navarrete, P., Sánchez-Cuesta, F., Ríos-Lago, M., & Benito-León, J. (2021). Impact of COVID-19 on a brain damage unit. Diabetes & Metabolic Syndrome, 15(4), 102163. https://doi.org/10.1016/j.dsx.2021.05.036
Saniasiaya, J. (2021). Hearing Loss in SARS-CoV-2: What Do We Know? Ear, Nose, & Throat Journal, 100(2_suppl), 152S-154S. https://doi.org/10.1177/0145561320946902
Warde-Farley, D., Donaldson, S. L., Comes, O., Zuberi, K., Badrawi, R., Chao, P., Franz, M., Grouios, C., Kazi, F., Lopes, C. T., Maitland, A., Mostafavi, S., Montojo, J., Shao, Q., Wright, G., Bader, G. D., & Morris, Q. (2010). The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Research, 38(Web Server issue), W214-20. https://doi.org/10.1093/nar/gkq537
Zuin, M., Rigatelli, G., Quadretti, L., Fogato, L., Zuliani, G., & Roncon, L. (2021). Prognostic Role of Anemia in COVID-19 Patients: A Meta-Analysis. Infectious Disease Reports, 13(4), 930–937. https://doi.org/10.3390/idr13040085
Comments