(Epi)transcriptomics in cardiovascular and neurological complications of COVID-19

Although systemic inflammation and pulmonary complications increase the mortality rate in COVID-19, a broad spectrum of cardiovascular and neurological complications can also contribute to significant morbidity and mortality. The molecular mechanisms underlying cardiovascular and neurological complications during and after SARS-CoV-2 infection are incompletely understood. Recently reported perturbations of the epitranscriptome of COVID-19 patients indicate that mechanisms including those derived from RNA modifications and non-coding RNAs may play a contributing role in the pathogenesis of COVID-19. In this review paper, we gathered recently published studies investigating (epi)transcriptomic fluctuations upon SARS-CoV-2 infection, focusing on the brain-heart axis since neurological and cardiovascular events and their sequelae are of utmost prevalence and importance in this disease.


Introduction
Coronavirus disease  is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) demonstrating a wide spectrum of clinical manifestations, from asymptomatic, mild flulike illness to lethal acute respiratory distress syndrome. The SARS-CoV-2 enters human cells by binding to angiotensin-converting enzyme type 2 (ACE2), a transmembrane receptor expressed by pulmonary and inflammatory cells, but also by other cell types including cardiomyocytes, pericytes and neurons ( Figure 1). Functionally, ACE2 inactivates angiotensin II [1] and plays an important role in neuro-humoral regulation of the cardiovascular system. In severe cases, SARS-CoV-2 invasion of the host cells may result in progression toward the "COVID-19 cytokine storm" which is characterized by severe immune reaction and overwhelming systemic inflammation, haemodynamic instability and multiple organ failure with known and unknown clinical complications [2,3]. Although, SARS-CoV-2 primarily affects the lungs, reports are emerging of the wide spectrum of cardiovascular and neurological manifestations and complications ranging from the de novo viral infection or its interplay with pre-existing comorbidities (Figure 1). The underlying cardiovascular and neurological comorbidities seem to predispose the development of more severe cardiovascular and neurological complications in COVID-19 patients, which are in turn associated with higher mortality rates [4]. History of cardiovascular disease is associated with a nearly five-fold increased risk in fatality rates [5]. Most prevalent cardiovascular disease risk factors in COVID-19 patients are hypertension and diabetes mellitus, while most common cardiovascular complications observed in COVID-19 patients include arrhythmias, acute myocardial infarction, cardiac injury, fulminant myocarditis, pericarditis, cardiac arrhythmia, heart failure, and disseminated intravascular coagulation [6,7]. The interaction between underlying cardiovascular comorbidities and the poor clinical outcome of COVID-19 may be multifaceted, including age, sex, cardiac dysfunction [8] and aorta ageing as defined by the estimated pulse wave velocity [9].
Neurological and psychiatric sequelae of COVID-19 have been widely reported (Figure 1) [10,11]. A retrospective study among 236 379 patients diagnosed with COVID-19 estimated that the incidence of a neurological or psychiatric diagnosis in the following six months after infection was 33.62% [12]. SARS-CoV-2 was recently detected in vivo in transgenic mice overexpressing human ACE2 and in post mortem cortical neurons of  showing the neuroinvasive capacity of SARS-CoV-2 [13]. However, the potential molecular Possible mechanisms include direct virus-mediated neuro-and cardiotoxicity, hypoxiarelated injury, immune-mediated cytokine storm and systemic inflammation, and so forth.
Recently reported perturbations of the epitranscriptome of COVID-19 patients indicate that mechanisms including those derived from RNA modifications and non-coding RNAs may play a contributing role in both short-and long-term cardiovascular and neurological outcomes. storm can damage an intact blood-brain barrier and disrupt the homeostasis of the central nervous system without the virus crossing the blood-brain barrier from the systemic circulation. In the cardiovascular system, an acute coronary syndrome can occur because of plaque rupture, coronary spasm or micro-thrombi owing to systemic inflammation or cytokine storm. In addition, the SARS-CoV-2 infection is associated with a pro-thrombotic state, which may lead to occlusion of blood vessels leading to injuries of both the heart and the brain. A part of this figure was created using "Mechanism of "SARS-CoV-2 Viral Entry" and "Cytokine storm" templates by BioRender.com (2020). Retrieved from https://app.biorender.com/biorender-templates Epitranscriptomic signature in the brain-heart axis of COVID-19 patients

Noncoding RNAs in the brain-heart axis
Eukaryotic cells produce different classes of non-protein coding RNA transcripts called ncRNAs participating in various cellular processes including, gene expression regulation, RNA maturation and protein synthesis. NcRNAs are transcribed by either RNA polymerase I, II or III, depending on the individual ncRNA [14]. According to their lengths, they can be divided into two main groups: (1) small or short ncRNAs including microRNAs (miRNAs), small interfering RNAs (siRNAs), PIWI-interacting RNA (piRNA) and small nucleolar RNAs (snoRNAs); and (2) long noncoding RNAs (lncRNAs) and circular RNAs (circRNAs) [15].
A rapidly growing number of studies has unravelled associations between aberrant ncRNA expression and human diseases. NcRNAs have emerged as promising candidates for the treatment of a variety of diseases and play a potential significant role in host-virus interactions [15]. Dysregulated noncoding RNAs (ncRNAs) expression levels in lung tissue and liquid biopsies from COVID-19 patients indicate that ncRNAs may play an important role in the pathogenesis, and hence clinical outcomes of COVID-19. For instance, altered levels of ncRNAs involved in T cell activation and differentiation have been found in peripheral blood mononuclear cell samples from COVID-19 patients at three different time points during their treatment, convalescence, and rehabilitation [16]. cardiomyocytes, human brain endothelial cells and human cholinergic neurons. Small ncRNAs (particularly miRNAs) associated with cardiovascular and neurological complications in COVID-19 patients are summarized in Table 1. In human brain microvascular endothelial cells (hBMECs), miR-24 targets cell surface receptor neuropilin-1 (NRP1), indicating its potential role in the neurological manifestation of COVID-19 [17]. Upregulation of miR-24 significantly reduced the permeability of hBMECs resulting in response of VEGF, an agonist of Neuropilin-1, which lead to the Neuropilin-1 overexpressin and rescue of impaired cellular response [17]. Two independent studies have reported that NRP1 plays an important role in SARS-CoV-2 entry in human cells through interaction with S1 domain of viral Spike protein after its cleavage by furin protease [28,29].
In silico analyses identified a set of hypothalamic miRNAs potentially playing a role in the regulation of expression of hypothalamic ACE2 and transmembrane serine protease 2 (TMPRSS2), essential proteins for SARS-CoV-2 cell entry [30]. Although hypothalamic circuits are known to be exposed to the entry of the virus via the olfactory bulb, the regulation of its interaction with SARS-CoV-2 is not yet fully understood.

Long noncoding RNAs
An increasing number of lncRNAs with potential neurological and immunological functions has been shown to be associated with the pathogenesis of COVID-19 (Table 2). Table 2. Aberrantly expressed lncRNAs associated with the brain-heart axis and inflammation after SARS-CoV-2 infection. reflecting their association with the inflammatory pathobiology related to COVID-19. For instance, transcriptomic analyses in lung has shown that two lncRNAs, DANCR and NEAT1, may predict the inflammatory profile of infected tissues and provide insights into adverse body and brain consequences of COVID-19 [31]. The lncRNA DANCR regulates inflammation and is responsible for the surveillance over cholinergic blockade of inflammation, thereby maintaining a balance between pro-and anti-inflammatory pathways in lung and brain tissues affected by SARS-CoV-2 [31]. In another study, expression levels of three lncRNAs, MALAT1, NEAT1, and SNGH25 were found to be downregulated in mild and severe COVID-19 patients [33]. Interestingly, MALAT1 and NEAT1 are implicated in both, cardiovascular and neurological complications. Recent findings on differential ncRNA expressions across SARS-CoV-2 infected tissues and bio-fluids of COVID-19 patients indicate that ncRNAs may constitute key players in the regulation of the immune response following infection [28,29]. However, changes in ncRNAs landscape associated with cardiac and neurological manifestations of COVID-19 patients are yet to be fully characterized and their potential role in the course of the disease remains to be elucidated.
Adenosine-to-inosine (A-to-I) RNA editing, mediated by the adenosine deaminase RNA specific (ADAR) family of enzymes, is the most widespread RNA modification, taking place mainly in the primate-specific, repetitive Alu elements [46,47]. A recent study showed that J o u r n a l P r e -p r o o f Journal Pre-proof ADAR1, the main RNA editing enzyme, was among the most upregulated genes in pulmonary alveolar type II cells of patients infected with SARS-CoV-2 [48], the main lung cells expressing the ACE2 receptor [49]. Similarly, ADAR1 expression was increased in peripheral blood cells of patients with severe COVID-19 compared to those with mild disease [50].
Increased ADAR1-induced RNA editing may affect the brain-heart axis in several ways. For instance, ADAR1 may prevent the excessive activation of innate immune receptors by the viral RNA [51], thus preventing the hyperinflammatory response and translational shutdown/apoptosis. Further, ADAR1 may prevent oxidative stress-induced inflammation and apoptosis in cardiomyocytes by inhibiting protein kinase R (PKR) hyperactivation [52]. Of interest, ADAR1 overexpression, which was also induced by oxidative stress, could limit PKR phosphorylation in cardiomyocytes [53], thus preventing the aberrant activation of innate immune sensors by viral double stranded RNA. On the other hand, extensive RNA editing of the serotonin 5-hydroxytryptamine receptor 2C (5-HT 2C R) in the brain has been associated with aberrant sympathetic nervous activity [54]. 5-HT 2C R has five editing sites in exon 5 leading to 3 amino acid substitutions, which can alter the coupling with downstream G proteins by 10-15-fold [55]. Of note, a persistent hypermetabolic state is observed in patients with COVID-19 [56], especially in those admitted to intensive care units [57] contributing to patient deterioration. Similar, yet unknown, possible editing events in other serotonin receptors leading to an aberrant activation of the sympathetic nervous system could potentially lead to arrythmias or even sudden cardiac death especially in individuals with underlying heart conditions [58].

RNA editing of the viral RNA
Apart from its effects on host gene expression and cellular function under stress conditions including apoptosis, RNA editing may control SARS-CoV-2 propagation per se (reviewed in [59]). RNA editing has two major forms, deamination of adenosine-to-inosine (A-to-I) and deamination of cytosine-to-uracil (C-to-U), mediated by the ADAR and APOBEC family of enzymes, respectively [47,60]. SARS-CoV-2 has a positive sense, single-stranded RNA genome of approximately 30-kilo bases length [61]. The high prevalence of C-to-U and A-to- on the other hand, hyper-editing of viral RNA can restrict viral replication or "mark" viral RNA for degradation by endonucleases [67,68]. Thus, RNA editing on the SARS-COV-2 genome could be a relevant mechanism controlling the dynamics of viral evolution, affecting virulence, pathogenicity and host response [69].

Apart from the transcriptional alterations induced by systemic inflammation in endothelial cells, a systematic upregulation of ADAR1-induced RNA editing in patients with COVID-19
could further propagate the inflammatory response by increasing stability of proinflammatory transcripts, as we have previously shown in chronic inflammatory diseases such as atherosclerosis and rheumatoid arthritis [78,79]. For example, upregulation of cathepsin S (CTSS), an elastolytic enzyme that promotes the development of inflammatory, non-stable atherosclerotic plaques [80] could in turn destabilize existing plaques leading to acute coronary events or stroke. Cathepsin S is also involved in major histocompatibility complex (MHC)-II dependent antigen presentation [81], thus providing a link between innate and adaptive immune responses. Of interest, cathepsin S expression is increased in a subgroup of lung capillary endothelial cells along with genes involved in MHC class II-mediated antigen presentation [82], which could participate in initial recognition of SARS-CoV-2 and mounting of systemic inflammatory reaction [83]. Of note, CTSS is only one among thousands of mRNAs, as well as ncRNAs, including miRNAs and lncRNAs that can be affected by RNA editing [78,80,81], offering an additional level of gene regulation during the systematic inflammatory response induced by SARS-CoV-2 infection (Figure 2). increasing RNA stability of proinflammatory genes, such as CTSS, and thus propagating the systematic, as well as the tissue-specific (atherosclerotic plaque destabilization) inflammatory response, 5) creating recoding events in 5-HT 2C R in the brain leading to a hypermetabolic state, 6) increasing miRNA processing by DICER and thus mature miR-222 expression, which prevents apoptosis of infected myocardial cells.

Brain-heart crosstalk
It is projected that 0.04% of the overall COVID-19 patients have been affected with neurological diseases (e.g. ischemic strokes, encephalopathy, psychosis) at the central nervous system (CNS) [84]. On another hand, it is estimated that 20% to 30% of hospitalized patients develop cardiac injury [85,86]. Clinical data indicate that both the susceptibility to and the outcomes of COVID-19 are strongly associated with cardiovascular disease (reviewed in refs [86,87]). [86,87]). Currently, it is relatively unknown what is the level of communication between the brain and the heart in the context of COVID-19 and which cellular or molecular mechanisms play major roles in this communication [88].

Inflammation may be a potential link between heart and brain communication in the context
of COVID-19. It is well accepted that the proteome signature is affected significantly in the acute-phase response in severe COVID-19 patients [89,90]. Results obtained in different studies indicate the up-regulation of IL-6 signalling pathway [89,91,92], which is more affected than other inflammatory pathways such as the tumor necrosis factor and interferon gamma pathways. Both the classical complement pathway and the complement modulators are also activated, C-reactive protein and serum amyloid proteins are up-regulated, and modulators of inflammation such as gelsolin that is part of the extracellular actin scavenger system which removes toxic F-actin filaments are dysregulated [89].
A second important route of communication between heart and brain are the extracellular vesicles (EVs). EVs are lipid vesicles secreted by cells containing several biomolecules J o u r n a l P r e -p r o o f (miRNAs, mRNA, proteins) [93] which are involved in inter-organ communication [94].
These EVs can be classified in three different categories according to their biogenesis: (i) exosomes, (ii) microvesicles and (iii) apoptotic bodies. EVs secreted by the brain or the heart have the potential to carry biological information in the brain-heart axis. For example, EVs released by the CNS activate the acute-phase response leading to the hepatic release of acutephase proteins (e.g. TNF, CXCL1, serum amyloid proteins) which are released into the circulation and influence the activity of the cardiovascular system [95,96]. Recently, it has been reported that COVID-19 patients have high levels of plasmatic EVs containing tissue factor (TF) activity, a main activator of the coagulation cascade, and thus a trigger of thrombotic events [97]. It has been speculated that the EVs containing TF are derived from activated endothelial cells or perivascular cells. Further studies are required to investigate in more detail the role of EVs in the brain-heart axis.
Stem/progenitor cells are also a route of communication between heart and brain. Circulating levels of endothelial progenitor cells (CD45 -CD31 + CD34 + CD146 -) have been reported to be increased in COVID-19 patients with mild and severe symptoms compared to healthy controls [98]. Interestingly, some endothelial progenitor cells had the phenotype CD34 + KDR + CD19 + , indicative of their lymphocyte lineage [99].

Diagnostic potential of ncRNAs in COVID-19
The major hallmarks of severe COVID-19 are respiratory (pulmonary embolism), cardiac For instance, several properties of ncRNAs particularly miRNAs suggest their potential value as biomarkers for COVID-19 outcome prognosis. They are present and stable in the circulation, they demonstrate tissue-specificity, participate in disease evolution, they are J o u r n a l P r e -p r o o f Journal Pre-proof measurable using reliable and sensitive techniques and they are easily accessible from biofluids ("liquid biopsies"). In addition to high sensitivity and specificity, RNA analysis is cheaper compared to protein analysis and offers a greater overview of cell regulation and states compared to DNA analysis.

Future of COVID-19 diagnostics and therapy
The scientific community around the globe is working frenetically to support emergency responses to the ongoing pandemic by implementing innovative research strategies aiming to develop diagnostic, prognostic and therapeutic solutions for COVID-19, as well as to prepare for future similar issues. The battle against COVID-19 and preparedness for future pandemics rely on the development of innovative approaches to identify risk stratification biomarkers and therapeutic targets capable to predict disease evolution and improve clinical outcomes.
Although several types of vaccines are currently administered against COVID-19, it is too early to predict how efficient they will be to eliminate SARS-CoV-2, notably due to the capacity of the virus to mutate and generate novel and potentially more dangerous variants.
Taking into account high transmissibility of new viral variants and heterogeneity of clinical symptoms among COVID-19 patients, there is a critical need to foster multi-omics approaches to discover novel diagnostics and therapeutic solutions for COVID-19. (epi)transcriptome affecting both the brain and the heart and interacting with SARS-CoV-2 may provide significant mechanistic insights and catalyse the discovery of novel drugs to limit adverse events, thereby impacting on healthcare and improving patients" outcomes.

Extracellular vesicles based COVID-19 therapeutics
Several These clinical trials are justified by the encouraging results obtained by the use of EVs in preclinical lung injured models such as acute respiratory distress syndrome, lipopolysaccharideinduced lung injury and pneumonia (reviewed in reference [103]).

Concluding remarks and future perspectives
Although the rate of hospitalization of SARS-CoV-2 infected individuals is diminishing, the COVID-19 pandemic remains a significant public health threat worldwide, mostly due to emerging long-term sequelae ("long COVID"). An in-depth characterization of the omics contributors of the disease would allow refining diagnostic strategies, identify markers of disease severity and progression, and finally identify novel therapeutic approaches.
Rapidly evolving high-throughput technologies are well positioned to allow discoveries that will increase our understanding of SARS-CoV-2 interaction with host immune response. This

Competing interests
YD holds patents related to diagnostic and therapeutic applications of RNAs.

Author contributions:
All authors contributed substantially to all aspects of the article. J o u r n a l P r e -p r o o f

Declaration of interests
☐ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
☒ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Yvan Devaux reports was provided by Luxembourg Institute of Health. Yvan Devaux holds patents related to diagnostic and therapeutic applications of RNAs.
J o u r n a l P r e -p r o o f