Identication of potential biomarkers and inhibitors in SARS-CoV-2 infected macaques

The COVID-19 pandemic caused by the infection with SARS-CoV-2 has overwhelmed many health systems globally. Our study is to identify differentially expressed genes (DEGs) and the associated biological pathways of COVID-19 to elucidate the potential pathogenesis and metabolism. The gene expression prole of the GSE155363 dataset was originally produced using the high-throughput Illumina HiSeq 4000 (Macaca mulatta). Kyoto Encyclopedia of Genes and Genomes pathway (KEGG) enrichment analyses were performed to discover their functional categories and biochemical pathways. The results suggested that four biological pathways: Fatty acid elongation, Biosynthesis of unsaturated fatty acids, Fatty acid metabolism, and Ribosome were mostly involved in the macaques with COVID-19. Thus, our study provides novel insights into the underlying pathogenesis of COVID-19.


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a RNA coronavirus that causes coronavirus disease 2019 (COVID-19) 1 . The common clinical symptoms of severe COVID-19 are fever, cough, pneumonia, and dyspnoea 2 . The Chinese Center for Disease Control and Prevention reported that 80% of infected patients experienced mild diseases, 14% developed severe diseases and 5% developed critical diseases 3,4 .
The factors that caused illness in COVID-19 patients are not fully understood and the development of diseases does not simply relate to viral load [5][6][7] . The aggressive in ammatory response to SARS-CoV-2 is considered as a major reason for disease severity and death in COVID-19 patients [8][9][10][11] . The severe stage of COVID-19 is relevant to high levels of circulating cytokines and substantial in ammatory cell in ltration in the lungs 8,12,13 . Given that the morbidity and mortality in COVID-19 is still rising worldwide, a better understanding of the mechanism of SARS-CoV-2 infection is necessary to the discovery of new therapeutic targets. Here, we analyzed the GEO data (GSE155363), provided by Price A and Rasmussen AL (Columbia University) to identify DEGs and the relevant biological process of COVID-19 in macaques. The KEGG pathway analysis, and protein-protein interaction were performed for nding those key gene nodes.

Data acquisition and preprocessing
The genes between COVID-19 macaques and healthy controls were conducted by R script. We used the Student's t-tests to identify DEGs with P<.01 and fold change ≥2 as being statistically signi cant.

Module analysis
The Molecular Complex Detection (MCODE) tool was used to detect the connected regions in PPI networks. The functional and pathway enrichment analyses were performed using DAVID (http://david.ncifcrf.gov/), and P<.05 was used as the cutoff criterion.

Inhibitors prediction
The L1000fwd is used to discover the potential inhibitors in COVID-19 macaques. L1000fwd analyzes the similarity to rank inhibitors potentially able to reverse the gene signature (29420694). The adjusted pvalue of 0.05 is considered as statistical signi cance.

Identi cation of DEGs in COVID-19 macaques
To determine the key features of COVID-19 infection, the modular transcriptional signature of COVID-19 macaques was compared to that of the healthy controls. A total of 21 genes were identi ed to be differentially expressed in COVID-19 samples with the threshold of P<0.01. The DEGs for COVID-19 and healthy controls were showed in Figure 1.

KEGG and Reactome Pathway analysis of DEGs in COVID-19 macaques
To further determine the roles of the DEGs from negative controls versus COVID-19 macaques' samples, we performed KEGG pathway and Reactome Pathway (RCTM) analysis as previously described 14 . KEGG and RCTM pathway analysis are used to de ne the functional meanings of genes and genomes. In our study, the top enriched biological KEGG pathways associated with COVID-19 included " Fatty acid elongation", " Biosynthesis of unsaturated fatty acids ", " Fatty acid metabolism " and "Ribosome" (Table  1). Moreover, the RCTM analysis showed the DGEs were involved in the synthesis of very long-chain fatty acyl-CoAs from fatty acid metabolism group and diseases of DNA repair group (Figure 2 and Supplemental Table S1).

PPI network analysis of DEGs between healthy controls and COVID-19 macaques
To explore the relationship of DGEs at the protein level, we identi ed 14 signi cant functional proteins with String tool (https://string-db.org/) ( Figure 3A), and created a PPI network by using the Cytoscape software ( Figure 3B). We set the prede ned criterion of combined score >0.7, a total of 7 interactions and 7 nodes were constructed to form the PPI network between healthy controls and COVID-19 macaques' samples ( Figure 3B). Among these nodes, the top 5 hub genes with highest degree scores are shown in Table 2. We then created the top two signi cant modules of COVID-19 versus control samples to depict the functional annotation of the genes ( Figure 3B).
By using the L1000FWD analysis, we identi ed top ten potential inhibitors with the highest scores ( Figure   4 and Supplemental Table S2

Discussion
The most frequent symptoms of COVID-19 are fever, cough, fatigue and diarrhea 15 . Severe illness usually happens about 1 week after the onset of symptoms 16 . Thus, it is of the utmost importance to nd the biomarkers during the early infection. We analyzed the GEO data by comparing the early infection samples (day 1) with the negative controls to nd the DEGs and potential therapeutic targets. After infection on day 1, most of DEGs were upregulated in macaques. Among them, we identi ed 17 most changed genes including TRNAA-AGC, PQLC1, NEIL3, HACD1 and other unknown genes, which may be involved in the process of virus infection. Interestingly, we identi ed a large number of unknown genes that could be the novel targets for COVID-19 treatment. Thus, further studies will be focused on identifying the functions and activities of these unknown genes.
The KEGG pathways of DEGs after infection showed high relevance to the fat metabolism. Previously, our reports had shown that COVID-19 had a great impact on metabolic syndromes including the "Biosynthesis of unsaturated fatty acids", "fatty acid elongation", "Fatty acid metabolism", and "Metabolic pathways" in human patients 14,17 . Similarly, in macaques, the fat metabolism changed during the early stage of SARS-CoV-2 infection. The presence of one or more underlying health conditions including diabetes, chronic lung disease, and cardiovascular disease are risk factors for the enhanced severity of the COVID-19 diseases 18 . In the U.S., 89% of adults who have diabetes are also overweight or have obesity 18 . Obesity disrupts breathing that leads to sleep apnea and disordered sleep/wake cycle, which further causes the circadian related diseases and promotes virus infection [19][20][21][22][23][24][25] . Obesity also affects the mitochondrial structure and function 26 , which further leads to physiological dysfunction and the related diseases 27,28 . Obesity was rst recognized as an independent risk factor for the H1N1 pandemic in 2009 29,30 . In the report by Northwell Health, among the con rmed SARS-CoV-2 cases, hypertension, obesity, and diabetes were the most common comorbidities 31 .
Obesity appears to predispose COVID-19 patients to increase the severity of the disease 32 . Obesitycaused cellular dysfunction triggers a various range of signaling pathways, including activation of NF-κB, IRE-1, mTOR, ERKs, and PKR 33 . These pathways collaborate to produce two key metabolical effects. First, each pathway converges upon and inhibits insulin signaling pathways. Second, these signals are involved in two main in ammatory pathways 34 . The transcription factor NF-κB enhanced immunity by regulating the expression of genes involved in in ammation 35 . A number of reports have demonstrated a critical role of the NF-κB signaling pathway in the liver, bone, cartilage, and central nervous system in the development of in ammation-associated diseases [36][37][38] . High metabolic demands can induce cell death, which forms a pro-in ammatory signal to NF-κB activation 33 . Moreover, activation of innate immune receptors through MyD88 could activate NF-κB in response to an excess of free fatty acids in a high fat diet 33,39,40 . Thus, NF-κB signaling is a critical pathway that contributes to the pathology of metabolic disorders.
In conclusion, the results suggested that several biological pathways such as "Fatty acid elongation", "Biosynthesis of unsaturated fatty acids", "Fatty acid metabolism" and "Ribosome" are commonly involved in the development of COVID-19. This study thus discovered novel biomarkers and inhibitors of COVID-19 in macaques, which may provide better targets for therapeutic intervention in the future.

Declarations
Con ict of Interest statement