Genetics and Molecular Biology

Autoimmune diseases are characterized by the loss of self-tolerance, leading to immune-mediated tissue destruction and chronic inflammation. Tyrosine kinase 2 (TYK2) protein plays a key role in immunity and apoptosis pathways. Studies have reported associations between single nucleotide polymorphisms (SNPs) in the TYK2 gene and autoimmune diseases; however, results are still inconclusive. Thus, we conducted a systematic review followed by meta-analysis. A literature search was performed to find studies that investigated associations between TYK2 SNPs and autoimmune diseases (multiple sclerosis, systemic lupus erythematosus, Crohn’s disease, ulcerative colitis, psoriasis, rheumatoid arthritis, type 1 diabetes, and inflammatory bowel disease). Pooled odds ratios (OR) with 95 % CI were calculated using random (REM) or fixed (FEM) effects models in the Stata 11.0 Software. Thirty-four articles were eligible for inclusion in the meta-analyses, comprising 9 different SNPs: rs280496, rs280500, rs280523, rs280519, rs2304256, rs12720270, rs12720356, rs34536443, and rs35018800. Meta-analysis results showed the minor alleles of rs2304256, rs12720270, rs12720356, rs34536443, and rs35018800 SNPs were associated with protection against autoimmune diseases. Moreover, the A allele of the rs280519 SNP was associated with risk for systemic lupus erythematosus. Our meta-analyses demonstrated that the rs2304256, rs12720270, rs12720356, rs34536443, rs35018800, and rs280519 SNPs in the TYK2 gene are associated with different autoimmune diseases.


Introduction
In 2002 and 2003, the first coronavirus outbreak started in Guangdong province in China, causing a total of 8,098 cases and 774 deaths worldwide, with a mortality rate of roughly 9% (Fehr and Perlman, 2015). This outbreak caused Severe Acute Respiratory Syndrome, or SARS. In 2012, another outbreak started in the middle-east, called MERS (Middle East Respiratory Syndrome), with an initial mortality rate of nearly 50%, which was controlled in the ensuing years. In 2014, a total of 855 cases and 333 deaths were reported by the MERS-CoV, with a 40% mortality rate, according to the European Center for Disease Prevention and Control (Fehr and Perlman, 2015). Before these two outbreaks, coronaviruses were believed to cause only self-limiting mild respiratory tract infections in humans.
In December 2019, a group of patients from the city of Wuhan, Hubei province, China, was initially diagnosed with a pneumonia of unknown etiology. The cases were epidemiologically linked to a seafood and wildlife market within the city (Rothan and Byrareddy, 2020). Afterward, reports predicted the appearance of a potential coronavirus outbreak since the reproduction number for the new 2019 coronavirus disease (COVID-19, as named by the WHO on February 11th 2020) was significantly greater than 1, estimated between intervals of 2.24 to 3.58 . Since then, until August 25th, 2020, the number of infected people has reached 23,677,221 cases and 813,802 deaths worldwide (Coronavirus Resource Center, John Hopkins University).
Symptoms of COVID-19 infection appear after an incubation period of approximately 5.2 days. The period between the onset of COVID-19 symptoms and death has a median of 14 days, varying from 6 to 41 days. The most common symptoms at the start of COVID-19 disease are cough, fever, and fatigue, and some cases may involve sputum production, headache, hemoptysis, diarrhea, dyspnea, and lymphopenia (Rothan and Byrareddy, 2020). Simabuco et al. 2 Taxonomy and genomic organization of coronaviruses The Coronaviridae family is part of the Nidovirales order and it can be divided into two subfamilies: Coronavirinae and Torovirinae. The Coronavirinae subfamily has four genera: alpha-, beta-, gamma-, and delta-coronaviruses. SARS-CoV-2 (also called 2019-nCoV) is the coronavirus that causes COVID-19 and, together with SARS-CoV and MERS-CoV, are part of the Betacoronavirus genus. A study showed that SARS-CoV-2 possesses greater similarity (88% identity) to two SARS-like coronaviruses derived from bats (bat-SL-CoVZC45 and bat-SL-CoVZXC21), collected in 2018 in Zhoushan, eastern China, than with SARS-CoV (about 79%), and MERS-CoV (about 50%) . Another study has shown that the SARS-CoV-2 genome is 91.02% similar to Pangolin-CoV, despite a higher identity of 96.2% between SARS-CoV-2 and another bat coronavirus (RaTG13) . This suggests that SARS-CoV-2 has originated from bats and might have pangolins as an intermediate host species, since five key amino acid residues (LFQNY) in the Receptor-Binding Domain (RBD) of S protein, involved in the interaction with human ACE2 receptor, are similar among Pangolin-CoV and SARS-CoV-2, but not with RaTG13 .
The coronavirus genome is composed of a singlestranded positive RNA (ssRNA+), therefore included in class IV of the Baltimore classification (Baltimore, 1971). The genomic organization of coronaviruses can be divided into two main parts, the genes encoding the non-structural polyproteins pp1a and pp1b and the genes encoding the structural genes, including the S, E, M and N genes, as shown in Figure 1.
Figure 1 -Molecular mechanisms related to the production of non-structural proteins (nsps) and assembly of the SARS-CoV-2 replication and transcription complex (RTC). Process 1: after recognition of the ACE2 (Angiotensin Converting Enzyme 2) cell receptor, the viral nucleocapsid is released into the cytoplasm by endocytosis, or fusion of the viral envelope, with the cell membrane. Process 2: the translation of the pp1a and pp1b genes from the 5'-capped and 3'-polyadenylated genome (+) of the virus produces the pp1a or pp1ab polyproteins, the latter being generated by a -1 frameshift of ribosomes. These polyproteins are then cleaved by viral proteases generating 16 virus nonstructural proteins (nsps), some of which are used to assemble the RTC, including the RNA-dependent RNA polymerase (RdRp or nsp12). Pharmacological interventions targeting specific points of the replication cycle of coronaviruses are highlighted. RTC: Replication and transcription complex; RdRp: RNA-dependent RNA polymerase. Table 1 -Non-structural proteins of coronaviruses and their functions.
Once produced, the pp1a and pp1ab polyproteins undergo proteolytic cleavage, forming a total of 16 proteins, detailed in Table 1 and outlined in Figure 1. The cleavage is generated by 2 proteases: nsp3, which is a Papain-like protease (PL pro ) and nsp5, which is a 3C-like protease (3CL pro ). Nsp3 cleaves the sites between nsp2 to nsp4, therefore generating nsp1, nsp2, and nsp3. On the other hand, nsp5 cleaves the other sites, generating the other non-structural proteins (Ziebuhr et al., 2000;Báez-Santos et al., 2015).
The next processes of coronavirus replication and transcription are outlined in Figure 2 below. Once produced and processed, part of the non-structural proteins along with nsp12, the RNA-dependent RNA polymerase (RdRp), assemble the Replication and Transcription Complex (RTC). RTC acts primarily by producing a set of single-stranded negative RNAs (ssRNA-), including copies of the genomic RNA and subgenomic RNAs, which will then serve as templates for the production of the genome and mRNA, respectively. These intermediate negative RNA molecules are about 1% as abundant as their respective positive counterparts and contain anti-leader sequences, present in the 5' untranslated region (UTR) of the antigenome and in the 3'-UTR of the genome (Sethna et al., 1991). On the other hand, in the 5'-UTR of the viral genome and 3'-UTR of the antigenome there are leader sequences. The leader and anti-leader sequences are used by the RTC to initiate replication and transcription. In the 5'-UTR of the genome and at the beginning of each ORF of the structural genes, there are other regulatory regions called Transcriptional Regulatory Sequences (TRS), as shown in Figures 1 and 2. During the transcription of ssRNA molecules, two mechanisms may happen, according to the currently established model of replication of coronaviruses (Pasternak et al., 2006;Sawicki et al., 2007): 1) The RTC, after binding to the 3' anti-leader sequence of the viral genome, initiates the synthesis of negative RNA throughout the molecule until it finds the leader region in the 5' end, generating a complete copy of negative polarity of the genome, called antigenome. This antigenome will serve Figure 2 -Molecular events related to the expression of structural proteins, replication of the genome, and assembly of the SARS-CoV-2. Processes 1a and 1b: the synthesis of RNA (-) by RTC, initiated in the 3' anti-leader sequence of the genome (+), may occur continuously, generating a complete copy of the genome called antigenome (-). The antigenome is then used by RTC to produce multiple copies of the genome (+). Processes 2a and 2b: RNA synthesis by RTC may, however, be temporarily interrupted when a TRS is copied. The newly synthesized RNA (-) is then transferred to the 5' end of the genome, where the complementarity of sequences allows the RNA (-) synthesis to continue in the leader TRS, merging the sequences between body and leader TRSs. In turn, these subgenomic chimeric RNAs (-) serve as templates for the continuous synthesis of subgenomic mRNAs (+). Process 3: the structural S, E, M, and N proteins are then translated from the 3'-co-lateral subgenomic mRNAs (+), where S, E, and M proteins are produced in the rough endoplasmic reticulum. Process 4: the N protein produced in the cytosol interacts with the viral genome (+), forming the nucleocapsid. Process 5: membrane proteins S, M, and E then interact with viral nucleocapsids to form virions in the ERGIC. Process 6: finally, the virions are externalized from the cell by exocytosis via the Golgi pathway. TRS: Transcriptional Regulatory Sequences; ER: endoplasmic reticulum; RTC: Replication Transcription Complex. ERGIC: Endoplasmic Reticulum -Golgi Intermediate Compartment.
Replication of human coronaviruses 5 as a template for the synthesis of the genome (+) (Figure 2, processes 1a and 1b).
2) The RTC may, however, temporarily pause the transcription in each of the TRS regions of each ORF and continue in the 5'-UTR of the genome, given the complementarity of the TRS regions. Therefore, a leader region is incorporated into each RNA, generating subgenomic RNAs of negative polarity. These subgenomic RNAs will serve as templates for the mRNA (+) synthesis, containing their 3' regions co-terminal to the genomic RNA (Figure 2, processes 2a and 2b). This process is often called "copychoice" mechanism and is manifested in other viruses, enabling recombination, as will later be discussed (Simon-Loriere and Holmes, 2011).
Once the complete genomic RNA and mRNAs of each structural ORFs are produced, the translation of these genes begins ( Figure 2, process 3), generating the S, E, M and N proteins, which are essential for the assembly of the viral particles, among others (ORFs 3a, 6, 7a / b, 8 and 9). The specific functions of these proteins will be described later. The viral replication cycle then continues through the interaction of the N protein with the viral genomic RNA, forming the nucleocapsid in the cytosol ( Figure 2, process 4). On the other hand, the production of S, E, and M proteins are directed to the rough endoplasmic reticulum (RER) (Fehr and Perlman, 2015). Finally, interactions mediated between these structural proteins culminate in the recruitment of nucleocapsids into the compartment between the RER and the Golgi apparatus called ERGIC (Endoplasmic Reticulum -Golgi Intermediate Compartment) and finally in the exocytosis of the viral particles ( Figure 2, processes 5 and 6) (Fehr and Perlman, 2015).
An important feature related to the replication of coronaviruses is the high rates of mutation and recombination, which alters viral protein properties, host range, and pathogenicity. For example, there are reports of heterologous recombination between subgroup A Betacoronavirus and other viruses, since some of these coronaviruses have the hemagglutinin esterase gene, derived from the Influenza C virus . The recombination between coronaviruses targeting different species is also largely reported in the family and may explain the similarity between the genome sequence of human SARS-CoV-2 and bat and pangolin coronaviruses, as already discussed (Lau et al., 2020;Lu R et al., 2020;Zhang T et al., 2020). Three aspects may explain this increased capacity for recombination/mutation: 1) the RdRp of coronaviruses has low fidelity. Although a 3'-5' exonuclease proofreading activity is reported, the mutation rate of this polymerase is about 2.0 × 10 -6 mutations per site, per replication cycle (Eckerle et al., 2010); 2) the unique RNA replication mechanism using the TRS motifs, known as the "copy-choice" mechanism, may induce homologous RNA recombination between genes of different coronaviruses (Simon-Loriere and Holmes, 2011); 3) Coronaviruses have the largest genome (26-32 kb) among RNA viruses (Terada et al., 2014).
Several studies have shown the modulation of cellular pathways by coronavirus proteins, favoring the viral cycle or impacting the viral pathogenesis, which is summarized in Table 2. Among these studies, interactions were found between the coronavirus nsp1 protein with the cyclophilins PPIA, PPIG, PPIH, and FKBP1A, FKBP1B that are capable of modulating the Calcineurin/NFAT pathway, which plays an important role in the activation of immune cells (Pfefferle et al., 2011). The same study, which used the yeast twohybrid system to demonstrate those interactions, showed that the inhibition of cyclophilins by cyclosporine A (CspA) blocked the replication of different CoVs, including the human coronaviruses SARS-CoV, CoV-229E and -NL-63; the feline CoV; and the avian infectious bronchitis virus (IBV). Another study has demonstrated the interaction between the SARS-CoV nsp2 and cellular prohibitins, suggesting that this nsp may be involved in the disruption of intracellular host signaling (Cornillez-Ty et al., 2009).
The PL pro protein (nsp3) from SARS-CoV significantly triggered the activation of the TGF-β1 promoter through ROS/ p38-MAPK/STAT3, correlating with the positive regulation of pro-fibrotic responses in vitro and in vivo (Li et al., 2016). Another study showed that p53 downregulates SARS-CoV replication and is a target of nsp3 via an E3 ubiquitin ligase (Ma-Lauer et al., 2016). The other viral protease, 3CL pro (nsp5), cleaves STAT2, but not JAK1, TYK2, STAT1, and IRF9, which are key molecules of the JAK-STAT pathway, antagonizing the type I interferon signaling (Zhu et al., 2017).
Coronavirus nsp6 is known to interfere with autophagy, limiting the autophagosomes diameter at the point of omegasome formation, which may favor viral infection by compromising the ability of the autophagy system to degrade viral components via lysosomes (Cottam et al., 2014). Nsp9 seems to bind to E3 ubiquitin ligases TRIM59 and MIB1, which regulate antiviral innate immunity, and nsp13 may modulate the IFN pathway through TBK1 and TBKBP1 and the NF-κB pathway by TLE-1, -3, and -5 (Gordon et al., 2020). Another study reported that the coronavirus endoribonuclease nsp15 is required for evasion of dsRNA sensors and apoptosis since the loss of nsp15 activity is related to the stimulation of a protective immune response, which attenuated the disease in mice (Deng et al., 2017). Nsp16 can downregulate the activities of RIG-I and MDA5, inhibiting innate immunity to promote viral proliferation (Shi et al., 2019).
Finally, an important study recently showed cellular pathways related to the coronavirus RTC, using MHV (mouse hepatitis virus) as a model (V'kovski et al., 2019). The study identified that RTC interacts with translation initiation factors, which may not persist throughout the replication cycle but may be of transitory importance during specific phases of the replication cycle. In addition, the knockdown of some of these factors, such as eIF3E, eIF3F, and eIF3I, impacted the MHV replication. Proteins related to transport and intracellular organization were also related to RTC in this study.

The structural proteins of coronaviruses
The S protein Among all the structural proteins of coronaviruses, the S, E, M, and N proteins are considered essential and their functions are described below. Homotrimers of the S protein form the spikes in the viral surface, which give the virion a crown or corona aspect and the name of the family Coronaviridae, this being responsible for binding to host  (Chen et al., 2002;Surjit et al., 2006;Hurst et al., 2009;Emmott et al., 2013;Hurst et al., 2013;Cui et al., 2015;Ding et al., 2017;Gordon et al., 2020) ORF3a TRAF3 and ASC; caveolin-1; eIF2α and PERK (Padhan et al., 2007;Minakshi et al., 2009;Siu et al., 2019) ORF6 Nsp8 karyopherin alpha 2 and karyopherin beta 1; NUP98-RAE1 (Kumar et al., 2007;Frieman et al., 2007;Gordon et al., 2020) ORF7a ORF3 Type I IFN response; BST-2; cyclin D3/pRb pathway (Yuan et al., 2006;Dedeurwaerder et al., 2014;Taylor et al., 2015) ORF10 Cullin 2 (CUL2) RING E3 ligase complex; ZYG11B (Gordon et al., 2020) receptors (Beniac et al., 2006). The S protein has roughly 150 kDa and contains an N-terminal signal sequence that gives access to the endoplasmic reticulum (RER) for its synthesis, being strongly N-terminal glycosylated. The S protein is trimeric and a class I fusion protein, mediating host receptor binding (Bosch et al., 2003). A recent study showed that the angiotensin-converting enzyme 2 (ACE2) is the cell receptor for SARS-CoV-2, as well as for SARS-CoV, trimeric protein S being its ligand . In addition, the cellular serine protease TMPRSS2, targeting ACE2, facilitates the cellular entry of SARS-CoV and SARS-CoV-2 and a TMPRSS2 inhibitor, camostat mesylate, partially inhibits in vitro SARS-CoV-2 infection (Hoffmann et al., 2020). Apart from this, coronavirus entry in cells requires S protein priming by cellular proteases, which includes TMPRSS2 action (Hoffmann et al., 2020).

The E protein
The E protein plays a role in the assembly and release of virions from cells, being involved in viral pathogenesis (DeDiego et al., 2007). It is a small protein (~8-12 kDa) and is found in small amounts within the virion. Protein E is a homopentameric transmembrane protein, possessing an N-terminal ectodomain, a C-terminal endodomain, and ion channel activity, and such activity in SARS-CoV is not necessary for viral replication but impacts the viral pathogenesis (Nieto-Torres et al., 2014). Recombinant viruses without protein E are viable but attenuated, unlike other structural proteins, although this effect is dependent on the virus type (DeDiego et al., 2007). It interacts with the M membrane protein in the budding compartment of the cell, located in the ERGIC (Chen SC et al., 2009). The SARS-CoV E protein interacts with the PALS1 protein (Protein Associated with Lin-Seven 1) via the PDZ binding motif in its C-terminal domain and delays the formation of tight junctions, altering epithelial morphogenesis (Teoh et al., 2010). Another study showed that palmitoylation of the E protein is crucial for the assembly of murine coronavirus (Boscarino et al., 2008). A recent study has shown that the SARS-CoV-2 E protein binds to BRD2 and BRD4, members of the bromodomain and extraterminal domain (BET) family, which are known epigenetic readers (Gordon et al., 2020). These bromodomain proteins bind to acetylated histones and may regulate transcriptional processes, as well as viral proteins that interact with them, as demonstrated for the Influenza NS1 protein (Marazzi et al., 2012).

The M protein
The M protein presents three transmembrane domains and is responsible for the shape of virions, promoting the membrane curvature and binding to the nucleocapsid (Neuman et al., 2011). It is the most abundant structural protein in the virion, with approximately 25-30 kDa and presenting a small glycosylated ectodomain at the N-terminal and a larger C-terminal endodomain that extends from 6 to 8 nm in the viral particle (Nal et al., 2005). The M protein forms a dimer in the virion and can adopt two different conformations, interacting with the nucleocapsid N and S proteins (Neuman et al., 2011). The interaction with the nucleocapsid N protein promotes the complete assembly of the virion and was mapped to occur between the C-terminal of the M protein endodomain and the N protein CTD (Hurst et al., 2005). As already reported, its interaction with protein E has also been demonstrated . The M protein directs most of the protein-protein interactions necessary for the assembly of coronaviruses. A study showed that the surface proteins of coronavirus S, M, and E present differential subcellular locations when expressed alone, suggesting that additional cellular or viral factors may be necessary for coordinated traffic to the viral assembly site in the ERGIC (Nal et al., 2005).
Another study showed that the expression of the M protein alone is insufficient for the formation of virus-like particles (VLPs) (Bos et al., 1996). However, when the M protein was expressed along with the E protein, the formation of VLPs occurred, suggesting the important role of the two proteins in producing the envelopes of coronaviruses. The formation of VLPs is an important tool for therapeutical purposes since many of them can be used for testing and producing vaccines. A recent study reported the use of the SARS-CoV-2 S protein, a truncated S protein, or VLPs containing the S, M, and E proteins as candidates for vaccines against COVID-19. All formulations presented effectiveness in animal experimentation, but only VLPs induced both humoral and T cell immune responses .

The N protein
The N protein is the only viral nucleocapsid protein and it contains two domains. The two structural domains of the N protein, the N-terminal RNA binding domain (RBD; residues 45-181) and the C-terminal dimerization domain (DD; residues 248-365), do not interact with each other and are surrounded by flexible linkers (Chang et al., 2006). It is reported that the N protein can bind to nsp3 to assist the binding of the viral genome to the RTC and the packaging of the encapsidated genome in virions (Hurst et al., 2009;. The interaction between N and nsp3 supports a model in which this interaction tethers the genome to newly translated RTCs at an early stage of infection (Hurst et al., 2013). Also, one of the N protein domains is critical for the recognition of the M protein during virus assembly in cells. The interaction of the N protein and viral nucleocapsid with the membrane proteins S, E, and M for viral packaging takes place in the ERGIC, forming the mature virions that are then extruded from the cells by exocytosis via Golgi (de Haan and Rottier, 2005). The expression of the N protein increases the formation of VLPs, suggesting that the fusion of encapsidated genomes in the ERGIC improves viral envelope formation (Siu et al., 2008). A recent study suggests that the SARS-CoV-2 N protein has properties, such as RNA-binding, oligomerization, and multiple low-complexity regions, which indicates its involvement in cellular stress granules and ribonucleoprotein condensates that may be important for viral genome replication and packaging (Cascarina and Ross, 2020).
Regarding cellular interactions, several targets of the N protein have been proposed. A study has shown that coronavirus the N protein is a viral suppressor of RNA silencing (VSR) since the ectopic expression of the SARS-CoV N protein could promote MHV-A59 coronavirus replication in RNAi-active cells but not in cells depleted for the RNAi machinery (Cui et al., 2015). Apart from this, a proteomics study has demonstrated interactions between the coronavirus N protein and several cellular components, including ribosomal proteins, translation initiation factors, nucleolar proteins, helicases, and hnRNPs (Emmott et al., 2013). Some of those interactions, such as NONO and poly(A)-binding protein (PABP), were potentially mediated by RNA and the interactions with caprin-1, G3BP-1, and G3BP-2, which are involved in the formation of cytoplasmic stress granules, explain the localization of the N protein in these cell structures. The interaction between the N protein and NCL is a possible explanation of how coronavirus N proteins can localize to the nucleolus of cells (Chen et al., 2002). Finally, the impact of some cellular targets for viral replication was evaluated by RNA-interference depletion, demonstrating the functional importance of NCL, RPL19, or GSK3 proteins in the biology of coronavirus (Emmott et al., 2013).
Another proteomic study has confirmed that the SARS-CoV-2 N protein binds to stress granule proteins, including G3BP-1and -2. Also, the study found interactions between the N protein and other host mRNA binding proteins, including the translational repressor LARP1 (regulated by mTOR), the protein kinases CK2, and the mRNA decay factors UPF1 and MOV10 (Gordon et al., 2020). Other coronaviruses, including MERS-CoV and MHV, have been implicated in the modulation of stress granules formation (Raaben et al., 2007;Nakagawa et al., 2018). The inhibition of stress granules by MERS-CoV ORF4a favors its replication, thus suggesting that stress granule formation may be an antiviral response with possible therapeutic applications.

Other ORFs of coronaviruses
Aside from the pp1a, pp1b, S, E, M, and N proteins, SARS-CoV-2 presents 9 more ORFs, called ORF3a, ORF3b, ORF6, ORF7a, ORF7b, ORF8, ORF9b, ORF9c and ORF10, in which much less information is known regarding their molecular mechanisms of action in the viral replication cycle. SARS-CoV ORF3a is reported to bind to TRAF3 and ASC, promoting TRAF3 ubiquitination and activation of the NLRP3 inflammasome (Siu et al., 2019). Other studies have demonstrated that ORF3a presents binding affinities for caveolin-1 and calcium (Padhan et al., 2007;Minakshi et al., 2014). SARS-CoV ORF6 is localized in the ER and Golgi membranes in infected cells, binding to karyopherin alpha 2 and karyopherin beta 1 proteins and hindering STAT1 nuclear import and its function . SARS-CoV-2 ORF6 has been associated to the NUP98 and RAE1 proteins, which constitutes an interferon-inducible mRNA nuclear export complex that is degraded by other viruses, such as Influenza, to favor their replication (Satterly et al., 2007;Gordon et al., 2020). Another study showed that nsp8 interacts with ORF6, suggesting that the ORF6 protein plays a role in virus replication (Kumar et al., 2007).
Coronavirus ORF7a has been recognized as a type I IFN antagonist only when in the presence of the ORF3 protein, protecting the virus from the antiviral state induced by this cytokine (Dedeurwaerder et al., 2014). The ORF7a protein also binds to BST-2 (Bone marrow stromal antigen 2 or tetherin), an antiviral protein that restricts SARS-CoV infection, blocking its glycosylation, whereas the loss of ORF7a leads to a much greater restriction (Taylor et al., 2015). The ORF7a expression has also been associated with cell cycle arrest at the G0/G1 phase in HEK 293 cells via the cyclin D3/pRb pathway (Yuan et al., 2006). Another study has found that the translation of SARS-CoV ORF7b may be mediated by leaky scanning of ribosomes and that it localizes in the Golgi compartment and is incorporated into viral particles (Schaecher et al., 2007).
A study has demonstrated that SARS-CoV ORF8 may have originated through recombination from SARSrelated coronavirus from bats, which may have an impact on animal-to-human transmission (Lau et al., 2015). Regarding the exclusive SARS-CoV-2 ORF10, this protein has been associated with the Cullin 2 RING E3 ligase complex (CUL2), by interaction with the ZYG11B protein, suggesting that ORF10 hijacks the CUL2 complex for ubiquitination and degradation, or the opposite (Gordon et al., 2020).

Analysis of cellular pathways related to coronavirus replication
To further analyze the cellular proteins related to coronavirus biology, we performed Ingenuity Pathway Analysis, using the list of proteins presented in Table 2. This enabled the generation of a canonical pathways list, shown in Table 3, that may be important or modulated during the replication cycle of coronaviruses. The top 10 pathways were selected based on their associated p-value and are also presented in detail in the Figures S1-S10. The modulation of key molecular players, such as p53 and mTOR pathways, the inhibition of the host immune response by restraint of IFN induction and changes in cell cycle and cell growth, have been highly associated with the coronavirus proteins. These events may create a proliferative state that favors viral replication and inhibits apoptosis, facilitating viral cycle progression.
Immune system pathways SARS-CoV infects poorly monocytes/macrophages, although viral proteins are expressed, replication is incomplete in these cell types, which in turn respond secreting low levels of IFN-β and high levels of chemokines like IP-10 and MCP-1 and may be part of the inflammatory response that participates in Table 3 -Ingenuity Pathway Analysis (IPA) reveals the top 10 canonical pathways related to the cellular proteins that interact with coronavirus proteins, as summarized in Table 2 (3) * number of molecules in the pathway present in the input divided by the total number of proteins in that pathway.
Replication of human coronaviruses 9 the pathogenesis of the disease ; dendritic cells infected by SARS-CoV induce low levels of IFN-α, IFNβ, IFN-γ, IL12p40, moderate levels of TNF-α and IL-6 and high levels of MIP-1A, IP-10, and MCP-1 (Law et al., 2005). SARS-CoV shows a delayed induction of IFN-α (Spiegel, 2006). A comparison of SARS-CoV to Vesicular Stomatitis Virus (VSV) and Newcastle virus indicates that SARS-CoV induces lower levels of IFN-α, β, and γ, regardless of viral replication. Unlike SARS-CoV, MERS-CoV is able to establish an infection in human macrophages and induce higher levels of IL-12, IFN-γ, IP-10, MCP-1, MIP-1A, RANTES, and IL-8 (Zhou et al., 2014). The low activation of the IFN pathway is mediated by viral regulation of IRF3, a transcription factor activated by phosphorylation or polyubiquitination, and then translocates to the nucleus and induces the IFN response genes (Chattopadhyay et al., 2016). The SARS-CoV PL pro inhibits IRF3 phosphorylation, preventing its nuclear translocation and disrupting the IFN response, probably through inhibition of STING (stimulator of interferon genes), which is responsible for IRF3 phosphorylation . Nsp3 DUB domain of MHV-A59 and SARS-CoV promotes deubiquitination of IRF3 and also prevents its activation, blocking NF-κB signaling (Frieman et al., 2009). PL pro of MERS-CoV has also been described to inhibit IRF3 nuclear translocation . Interferon inhibition makes PL pro an important determinator of virulence of coronavirus (Niemeyer et al., 2018). The SARS-CoV Nsp1 protein also participates in IFN inhibition through decreasing STAT1 phosphorylation (Wathelet et al., 2007). While the ORF3a protein induces ER stress, activates PERK (PKR-like ER Kinase), and promotes phosphorylation, ubiquitination, and degradation of IFNAR1, attenuating interferon response (Minakshi et al., 2009). The SARS-CoV N protein has also been described to inhibit IFN production at an early stage, by sequestering PACT (protein activator of the dsRNA activated protein kinase R) and TRIM25 (tripartite motif protein 25), which bind to RIG-I (retinoic acid-inducer gene I) and MDA5 (melanoma differentiation gene 5) and activate IFN production (Ding et al., 2017). Finally, ORF4b of MERS-CoV is another protein that has been characterized to inhibit IFN and NF-κB signaling (Matthews et al., 2014).
STAT3 modulation plays an important role in proand anti-inflammatory responses. As already mentioned, SARS-CoV PL pro activates TGF-β1 through the p38MAPK/ ERK1-2 pathway, promoting STAT3 activation (Li et al., 2016). MERS-CoV strains with mutations in the NSP3 and ORF4a display differential STAT3 activation and different inflammatory cytokine profiles (Selinger et al., 2014). In SARS-CoV infection, a reduction of IL-4 is observed, which participates in humoral protection, an increase of IFN-γ, that participates in a potent cell-mediated immune response and also the elevation of IL-10 that plays a part in disease susceptibility (Zhu, 2004).

Cell cycle pathways
SARS-CoV has also been described to arrest the cell cycle. Nucleocapsid protein was shown to arrest cell cycle at the S phase, through direct interaction with cyclin D and inhibition of the CDK4/Cyclin D complex, preventing phosphorylation of Rb (Retinoblastoma) protein, a central player in cell cycle control (Surjit et al., 2006). IBV infection also reduces Cyclin D1, which participates in G2/M transition, inducing cell cycle arrest at G2/M (Harrison et al., 2007).
The blockage of G0/G1 progression has been observed by SARS-CoV ORF7a and ORF3a proteins through the reduction of cyclin D3 expression, decreased activity of cyclin D/CDK4/6, and inhibition of Rb phosphorylation (Yuan et al., 2006). Nsp15 is able to alter cellular localization of Rb and function, promoting pRb ubiquitination and degradation, increasing the proportion of S-phase cells, while overexpression of ORF4 (3b) protein arrests cell cycle at G0/G1 and promotes apoptosis (Yuan et al., 2005;Bhardwaj et al., 2012).
Expression of viral proteins regulates cell fate, not only cell cycle, but also controls apoptosis given its importance for viral replication. The SARS-CoV S protein suppresses the extrinsic apoptotic pathway, downregulating TRAIL and FasL, and activates the intrinsic apoptotic pathway through upregulation of Bax and down-regulation of Bcl-2, Mcl-1, Bcl-xL, and MDM2, leading to increased levels of p53 and p21 induction and G1/S arrest (Yeung et al., 2008). The ORF9b protein, when accumulated in the nucleus, induces caspase 3-mediated apoptosis (Sharma et al., 2011). Inhibition of apoptosis is also mediated by the SARS-CoV E protein, which down-regulates IRE-1 (inositol-requiring enzyme-1) and DUSP1/10 proteins, critical regulators of innate immune response and apoptosis (DeDiego et al., 2011). The SUD domain of PL pro interacts with RCHY1 and promotes p53 degradation, playing a role in cell cycle and apoptosis control, whereas p53 overexpression was able to inhibit viral replication (Ma-Lauer et al., 2016). SARS-CoV promotes the expression of a truncated form of p53 that inhibits apoptosis mediated by wild-type p53 (Leong et al., 2005). This is supported by the observation that the Porcine epidemic diarrhea virus (PEDV) production is increased in p53 knockout cells (Hao et al., 2019).

Protein synthesis control pathways
Protein synthesis pathways are often modulated by viruses. The activation of the PKR pathway by RNA viruses is an important cellular defense mechanism, which is in several cases counteracted by viruses, including coronaviruses. For example, the Dengue virus sustains, at early stages of infection, activation of the cap-dependent machinery, switching the protein synthesis to a cap-independent process in the late stages by downregulation of p70-S6K, 4E-BP1 and eIF4 factors (Villas-Bôas et al., 2009). The SARS-CoV ORF3a is known to cause endoplasmic reticulum stress and activation of eIF2α (eukaryotic initiation factor 2 alpha) and PERK, affecting innate immunity by suppression of type 1 IFN signaling (Minakshi et al., 2009). PKR and PERK, which promote phosphorylation of eIF2α that may suppress host translation, are expressed at high levels during SARS-CoV replication, although knockdown of PKR does not affect viral replication (Krähling et al., 2009). This suggests that SARS-CoV presents a mechanism to overcome the inhibitory effects of phosphorylated eIF2α on viral mRNA translation. On the other hand, another study has shown that depletion of the antiviral PKR pathway enhanced virus replication, increasing SARS-CoV protein expression and virus production (de Wilde et al., 2015).
Metformin and rapamycin are known modulators of viral infection and translation control pathways, such as mTOR. It is reported, for example, that in the 1971 Influenza outbreak, diabetic patients treated with phenformin and buformin presented a lower incidence of infection compared to diabetics treated with sulfonylureas or insulin (Lehrer, 2020). The immunoregulation of COVID-19 with mTOR inhibitors, such as rapamycin, has been proposed recently . Also, since it seems that all coronaviruses rely on capdependent translation to produce their proteins, key eIF capbinding complex constituents are candidates for therapeutic intervention against coronavirus diseases (Gordon et al., 2020).

Pharmacological interventions for the treatment of diseases associated with coronaviruses
Severe coronavirus infection leads to epithelial cell proliferation, macrophage infiltration in the lung (Nicholls et al., 2003) and can cause pulmonary fibrosis, which can linger in recovered patients (Antonio et al., 2003). About 15% of COVID-19 patients progress to acute respiratory distress syndrome (ARDS), the most severe cases should be treated in intensive care units (ICU) with oxygen therapy and mechanical ventilation . In extreme cases of COVID-19, lung transplantation is possible, if viable, as a last resort . The pharmacological interventions against coronaviruses, as summarized in Figure 1 and Table 4, are reviewed regarding their molecular mechanisms of action, in vitro and in vivo effectiveness, and ongoing clinical trials. There are two important aspects in the clinical outcome of COVID-19: one is viral entry/replication and the second is host response. Both are intimately linked and can be targeted by different compounds.

Interventions on viral entry/replication
Strategies to hinder viral binding have been investigated. Human recombinant ACE2 reduced SARS-CoV-2 recovery in vitro and protected mice from acute lung injury caused by SARS-CoV (Monteil et al., 2020). A chimeric protein composed of the extracellular domain of ACE2 fused with the Fc region of IgG1 exhibited pharmacological properties in mice (Lei et al., 2020). SARS-CoV-2 has two possible entry mechanisms: through endosome or membrane fusion. Arbidol (Umifenovir) is a potent broad-spectrum antiviral that blocks viral envelope fusion (Teissier et al., 2011) and clathrin-mediated endocytosis (Blaising et al., 2013), suppressing the replication of SARS-CoV in vitro (Khamitov et al., 2008). Patients treated with Arbidol had a shorter period of SARS-CoV-2 infection compared to patients treated with Lopinavir/ Ritonavir .
Several drugs have recently been used to inhibit SARS-CoV-2 replication, including the adenosine analog Remdesivir, which targets the RNA-dependent RNA polymerase and is incorporated into viral RNA chains, resulting in premature termination. Remdesivir was firstl shown to be effective against the Ebola virus (Warren et al., 2016) but presents activity against other viruses, including members of the Filoviridae, Paramyxoviridae, Pneumoviridae, and Orthocoronavirinae families (Brown et al., 2019). Remdesivir inhibits SARS-CoV and MERS-CoV (Sheahan et al., 2017), even though the identity among the coronavirus RdRps range from 70-90%, Remdesivir shows a broad spectrum of activity (Brown et al., 2019). Recently, a study showed that Remdesivir also acts against SARS-CoV-2, according to its potential antiviral mechanism as a nucleotide analog . Other nucleotide analogs, such as Ribavirin, Sofosbuvir, Galidesivir, and Tenofovir, can bind to SARS-CoV-2 RdRp (Elfiky, 2020). Favipiravir, a purine analog used against Influenza, is being tested against COVID-19 and reduced the time for viral clearance compared to patients treated with Lopinavir/ Ritonavir . There are currently ongoing clinical trials to evaluate Remdesivir and Ribavirin against COVID-19 and, regardless of its cost and administration routes, there are also concerns regarding its side effects and efficacy .
Furthermore, studies reported that the protease inhibitors Lopinavir and Ritonavir, used as HIV antivirals, also appear to have effects against SARS-CoV and SARS-CoV-2 (Chu et al., 2004;Choy et al., 2020). Both drugs can interact with the protease 3CL pro ; Ritonavir has a higher binding affinity compared with Lopinavir (Nutho et al., 2020). Animal experiments against SARS-CoV and MERS-CoV showed that the combination of Lopinavir/Ritonavir (LPV/r) with IFN-β significantly reduced viral load and improved pulmonary function. The combination of LPV/r shows a synergistic effect in the treatment of SARS patients (Yao et al., 2020). A clinical trial (NCT02845843) is currently testing a combination of Lopinavir, Ritonavir, and interferon-β1b against MERS (Arabi et al., 2020). Combined LPV/r reduced the time for patients to Table 4 -Pharmacological interventions targeting the replication cycle of human coronaviruses.
Patients treated with LPV/r presented a significantly higher proportion of abnormal liver function . There are more than 280 clinical trials ongoing, considering these antivirals, using Lopinavir, Ritonavir, Remdesivir, Favipiravir, in combination, alone or with other drugs in SARS-CoV-2 patients (Cochrane COVID-19 Study register). Chloroquine (CQ) has also been tested against SARS-CoV. It promotes alkalinization of acid vesicles in cells infected by intracellular pathogens (Rolain et al., 2007) and emerged as a substitute to quinine against malaria. CQ and hydroxychloroquine (HCQ) have been tested against viral hepatitis (Pareja-Coronel, 1963), Dengue virus (Farias et al., 2015), HIV (Paton et al., 2002), and also against other pathogens, such as intracellular bacteria (Coxiela burnetii and Tropheryma whipplei), bacteria-like Legionella pneumophila and Mycobacterium spp, and fungal infections by Histoplasma capsulate and Cryptococcus neoformans (Rolain et al., 2007). CQ is active in Vero E6 and Huh7 cells infected with MERS-CoV (de Wilde et al., 2014), but not in dendritic cells and monocyte-derived macrophages (Cong et al., 2018). CQ is also active in vitro against SARS-CoV either before or after virus exposure, interfering with ACE2 glycosylation and inhibiting viral binding (Keyaerts et al., 2004;Vincent et al., 2005). In addition, CQ induces alteration of endosomal pH that inhibits viral infection (Simmons et al., 2004).
Moreover, CQ and HCQ have shown some in vivo effects against SARS-CoV-2 . A group of Chinese researchers recently reported beneficial effects of chloroquine in the treatment of COVID-19, however, without yet publishing data . Another group of French researchers reported that HCQ decreased SARS-CoV-2 levels in a small group of tested patients, and the administration of azithromycin appears to improve such effects (Gautret et al., 2020). A recent review analyzed several ongoing clinical trials and indicates there are paradoxical results, some have shown beneficial results, others point to the toxicity issues (Sharma, 2020). One important point is that there are different strains of SARS-CoV-2 circulating . Notedly, CQ has pro-apoptotic activity and the prophylactic use of CQ has been linked to the selection of intracellular pathogen strains that promote cell resistance to apoptosis and enhanced lethality, as observed for HIV and SARS-CoV (Parris, 2004). Despite being a low-cost drug, there is a consensus among health agencies, such as the WHO, that further studies are needed for the clinical use of CQ and HCQ for COVID-19 treatment. Until now, 593 clinical trials are registered using HCQ or CQ to enlighten their role in SARS-CoV-2 infection treatment (Cochrane COVID-19 Study register).

Interventions on host cell response
The host response to viral infection is another important factor in COVID-19. SARS-CoV-2 induces secretion of IFN-γ, IL-1β, IL-4, IL-10, IP-10, and MCP-1 . Patients in intensive care units show higher levels of IL-2, IL-7, GCSF, IP-10, MCP-1, MIP-1A, and TNF-α that may induce cytokine storm and exacerbated inflammatory response . SARS-CoV not only infects alveolar epithelial cells, but also vascular endothelial cells, macrophages, monocytes, and lymphocytes. Rapid viral replication causes endothelial cell damage and vascular leakage, leading to the release of pro-inflammatory cytokines. Seroconversion of the host leads to the presence of IgG anti-S protein, which may promote the accumulation of proinflammatory monocyte/ macrophage and release of MCP-1 and IL-8 and have been linked to severe lung injury . Viral clearance depends on the activation of both innate and adaptive immune responses. IFN-γ and IL-6 contribute to neutrophil recruitment and transition to the adaptive response. However, exacerbated levels of IL-6 and reduced expression of IFN-γ may decrease CD4 + , CD8 + , and NK cells and may be connected to cytokine storm (Lagunas-Rangel and Chávez-Valencia, 2020).
Tocilizumab is a humanized anti-IL6R monoclonal antibody that prevents IL-6 signaling. Preprint studies indicate that it is safe and shows good efficiency against COVID-19. As there is a need for more clinical trial data, its use is only suggested for critically ill patients with high levels of IL-6 . Early clinical data recommends the use of repeated doses (Luo et al., 2020). Some of the concerns that have been raised are about the development of osteonecrosis of the jaws and the development of acute hypertriglyceridemia (Bennardo et al., 2020;Morrison et al., 2020).
Interferon release is one of the most important natural defense mechanisms against viral infection. In vivo experiments showed that treatment with IFN-β1b reduced pulmonary infiltrates, bronchointerstitial pneumonia, and viral load against MERS-CoV (Chan et al., 2015). IFN-α, a mismatched double-stranded RNA Interferon inducer, and the IFN inducer Ampligen, inhibited SARS-CoV replication in the lungs (Barnard et al., 2006). IFN-λ also showed activity against SARS-CoV and MERS-CoV, establishing an antiviral state and presenting minimal systemic inflammation (Prokunina-Olsson et al., 2020). Antibodies against cytokines and other proteins are presented in 196 studies, while 58 focus on IFN, either by inhibiting them or by giving their recombinant form to treat patients (Cochrane COVID-19 Study register). CQ and HCQ have also shown anti-inflammatory activity and have been used in inflammatory diseases such as rheumatoid arthritis and osteoarthritis (Sharma, 2020). CQ and HCQ intervene with lysosomal acidification, inhibiting antigen presentations, phospholipase A2, Toll-Like Receptors (TLRs), T and B cell receptors, and production of cytokines, like IL-1 and IL-6 (Sinha and Balayla, 2020). The inhibition of GSK3β by CQ may also be responsible for its immunomodulatory activity against COVID-19 (Embi et al., 2020).
Metronidazole is a redox-active prodrug that reduces the levels of pro-inflammatory cytokines, increases circulating lymphocytes, and decreases ROS produced by neutrophils and has also been suggested for the treatment of COVID-19 (Gharebaghi et al., 2020). Another class of anti-inflammatory drugs is the statins and these have been included in some treatment protocols. However, as statins also modulate TLR response, the use of statins in animal experiments against SARS-CoV and MERS-CoV resulted in increased viral load, severe lung damage, and death (Dashti-Khavidaki and . Nitazoxanide is used against protozoan and helminthic infection. Tizoxanide, the active form of nitazoxanide, inhibits 16 strains of Influenza A and one strain of Influenza B, Rotavirus, HCV, Yellow Fever virus, HBV, HIV, Norovirus, and others (Rossignol, 2014). Nitazoxanide reduces the viral load from different coronaviruses (Cao et al., 2015) and suppresses IL-6 production in mice (Hong et al., 2012), and it is also suggested for COVID-19 treatment.
In viral RNA infections, the use of nutraceuticals has been suggested to inhibit NOX2, which in turn, restores TLR7 response to single-stranded viral RNA infection and induces IFN; nutraceuticals could also up-regulate mitochondrial antiviral-signaling proteins (MAVS) and reduce proinflammatory signaling (McCarty and DiNicolantonio, 2020). Other than nutraceuticals, vitamins A and D, selenium, zinc, and probiotics may be beneficial for COVID-19 patients, by enhancing immunity and preventing respiratory infections (Grant et al., 2020;Jayawardena et al., 2020). Thus, the nutritional status of COVID-19 patients may be of further interest in future therapies since it might have an impact on the development of the disease.
The clinical progression of COVID-19 indicates that the initial symptoms are due to increased viral load and, in the following weeks of infection, seroconversion of IgG reduces viral load while some patients present worsening symptoms related to immunopathological damage (Peiris et al., 2003). Convalescent plasma (CP) has been used for SARS, MERS, Ebola virus and Chikungunya virus to improve survival rate (Alzoughool and Alanagreh, 2020). Different groups have tested critically ill COVID-19 patients with CP and obtained good recovery with no severe adverse effects Shi et al., 2020;Zhang L et al., 2020). The FDA has approved CP to treat critical patients (Tanne, 2020), although it has the risk of aggravating hyperimmune response, presenting a better response if administered in the early onset of the disease . Key points to the use of CP are: establishing eligibility criteria of donor COVID-19 convalescent patients, pre-screening tests of the donors, criteria for CP collection, and treatment of plasma (Epstein and Burnouf, 2020). Currently, CP is being tested in 388 clinical trials worldwide against COVID-19 (Cochrane COVID-19 Study register).

Conclusions
In summary, the cellular mechanisms associated with coronavirus replication form a complex and integrated network of molecular events, starting from the translation of nsps, proteolytic cleavage of polyproteins, assemble of RTC, transcription of antigenome, genome and subgenomic RNAs, translation of structural proteins, and finally assembly and budding of viral particles. The analysis of the cellular proteins related to coronavirus proteins reveals the modulation of key cellular pathways related to innate immunity, cell cycle, and protein synthesis. The current therapeutic approaches for COVID19 are partially related to these molecular events and pathways, but future pharmacological interventions may benefit from a better understanding regarding the replication cycle of SARS-CoV-2. Zhu X, Wang D, Zhou J, Pan T, Chen J, Yang Y, Lv M, Ye X, Peng G, Fang L et al. (2017) Porcine deltacoronavirus nsp5 antagonizes type I interferon signaling by cleaving STAT2. J Virol 91:e00003-17 Zhu Z, Lu Z, Xu T, Chen C, Yang G, Zha T, Lu J and Xue Y (2020) Arbidol monotherapy is superior to lopinavir/ritonavir in treating COVID-19. J Infect 81:e21-e23. Ziebuhr J, Snijder EJ and Gorbalenya AE (2000) Virus-encoded proteinases and proteolytic processing in the Nidovirales. J Gen Virol 81:853-879.

Supplementary material
The following online material is available for this article: Figure S1 -Role of PKR in interferon induction and antiviral response. Figure S2 -Activation of IRF by cytosolic pattern recognition receptors. Figure S3 -Cell cycle: G1/S checkpoint regulation. Figure S4 -Cyclins and cell cycle regulation. Figure S5 -Systemic lupus erythematosus in B cell signaling pathway. Figure S6 -EIF2 signaling. Figure S7 -Autophagy. Figure S8 -FAT10 signaling pathway. Figure S9 -Regulation of eIF4 and p70S6K signaling. Figure S10 -Role of p14/p19ARF in tumor suppression.

Associate Editor: Diogo Meyer
License information: This is an open-access article distributed under the terms of the Creative Commons Attribution License (type CC-BY), which permits unrestricted use, distribution and reproduction in any medium, provided the original article is properly cited.