SARS-CoV-2 infects the nervous system (NS) through multiple points of entry, leading to the severity and unique characteristics of systemic involvement of the virus in mediating cumulative disease burden and excessive morbidity and frequently mortality. The nervous system is composed of the CNS and the PNS. The PNS is further divided into the somatic and autonomic nervous systems. The autonomic nervous system is then segregated into the parasympathetic, sympathetic, and enteric divisions. These PNS pathways provide important communication routes to systemic organs, tissues and cell types and their molecular effectors for also defining NS viral tropism (Qureshi and Mehler, 2013; Shahriari et al., 2020; Zahalka and Frenette, 2020).

Other coronaviruses have been known to directly invade the brainstem (Porzionato et al., 2020), where the presence of the major receptor for SARS-CoV-2, angiotensin converting enzyme 2 (ACE2), has been documented. Previous studies have suggested that SARS-CoV-2 may spread to the CNS directly, by invasion through the cribriform plate into the olfactory epithelium and olfactory tract (Baig et al., 2020). The olfactory bulb has direct neuronal connections to the amygdala, entorhinal area, and hippocampus (Serrano et al., 2021). Additionally, a recent neuro-anatomical study of patients who died with COVID-19 found that 20% of patients studied had SARS-CoV-2 RNA in one or more regions of the CNS, including the olfactory bulb, amygdala, entorhinal area, temporal and frontal neocortex, and dorsal medulla (Serrano et al., 2021). In this study, the olfactory bulb was the only brain region with viral RNA in more than one subject and had the strongest PCR signal of brain regions studied (Serrano et al., 2021), suggesting that the CNS may be accessed through this route.

In addition to modulating basal homeostatic functions, the sympathetic nervous system (SNS) is responsible for “fight or flight” reactions. In such situations where the body is under stress, the SNS acts to mitigate this through global release of adrenaline and cortisol (Hanoun et al., 2015). The hypothalamus and pituitary gland, which express the ACE2 receptor, largely mediate the hormonal stress regulatory activity of the SNS via the hypothalamic-pituitary-adrenal (HPA) axis and may thus represent cardinal central viral targets (Pal, 2020). The nucleus tractus solitarius (NTS) of the brainstem, which acts as a sensory integrative center for several autonomic functions, projects onto the hypothalamus and also expresses the ACE2 receptor (Porzionato et al., 2020). Additionally, ACE2 is expressed in the carotid body, which can mediate local sympathetic activation to modulate optimal blood oxygenation and blood pressure regulation (Porzionato et al., 2020).

The enteric nervous system (ENS), a branch of the PNS, is located in the walls of the gastrointestinal (GI) tract, where it plays a key role in gut homeostasis, systemic metabolism and immunity. The parasympathetic NS closely interacts with this branch to control the basal state and digestive functions. Esposito et al. argue that SARS-CoV-2 can directly invade the ENS or the parasympathetic NS, and the virus can spread along the vagus nerve and its innervations of visceral organs, which can potentially provide another route of viral tropism (Esposito et al., 2020; Tassorelli et al., 2020). The parasympathetic pathway can then facilitate entry into the brainstem through vagus nerve synapses onto the NTS in the medulla oblongata (Li et al., 2020c). In rat studies of MERS-CoV, enteric involvement was shown to cause neurological symptoms and to precede respiratory infection (Zhou et al., 2017). Infection via GI inoculation led to higher viral loads in the CNS through feedback regulation via vagal efferents in these studies (Zhou et al., 2017). Moreover, recent work using murine models has identified a novel taxonomy involving twelve enteric neuron classes, distinguished by unique transcription factors, communication features, and functionality (Morarach et al., 2021). The enteric neuron classes develop a distinct diversification pattern through post-mitotic differentiation assisted by spatiotemporal defined landmarks (Morarach et al., 2021). These observations help distinguish the ENS from other branches of the NS and suggest potential new profiles of viral neurotropism embedded within the three-dimensional body axes. The relatively high concentration of ACE2 receptor and viral replication potential within the GI tract, along with frequently noted GI symptoms of COVID-19, emphasize the significance of ENS neurotropism (Esposito et al., 2020). In totality, the PNS will likely prove to be a key and under-appreciated feature of SARS-CoV-2 neurotropism, inclusive of all three autonomic branches and via direct and indirect target interactions.

The ACE2 receptor is found in neurons, glial cells, astrocytes, endothelial cells, and those of the lung parenchyma alike (Doobay et al., 2007; Miller and Arnold, 2019). This route is likely cell autonomous in nature, with direct target infection via a requisite entry factor (e.g., the ACE2 receptor). While there is no consensus on which cells are directly infected, there is evidence for direct neurotropism of several of these cell types through the presence of contiguous active virions and intracellular viral RNA. For example, a human induced pluripotent stem cell-derived BrainSphere model has been employed to demonstrate SARS-CoV-2 infectivity of neuronal cells (Bullen et al., 2020). Similarly, Wang et al. have shown that human induced pluripotent stem cell-derived ApoE4 neurons and astrocytes are susceptible to SARS-CoV-2 infection (Wang et al., 2021). Moreover, post-mortem brain tissue analysis of a COVID-19 patient revealed the presence of viral particles in brain endothelial cells, indicating possible direct neurotropism (Bullen et al., 2020; Paniz-Mondolfi et al., 2020). Alternatively, SARS-CoV-2 neurotropism can exert its actions non-cell autonomously through the use of soluble mediators; extracellular vesicles; inflammatory and immune cell signaling; non-traditional neurotransmitter, neuropeptide and ion channel signaling; via the lymphatics; and by microvascular damage (Hanoun et al., 2015; Solomos and Rall, 2016; Benveniste et al., 2017; Maryanovich et al., 2018; Merad and Martin, 2020; Rhea et al., 2021). Because a relatively low rate of SARS-CoV-2 RNA is detected in the brain tissue of patients who died from COVID-19, there is a developing consensus that neurological manifestations associated with COVID-19 are due, in large part, to non-cell autonomous effects (Serrano et al., 2021). For example, of twenty COVID-19 patients studied in one neuro-anatomical survey, only two had unequivocal neuropathological findings, and only one of those two had detectable SARS-CoV-2 RNA in the brain (Serrano et al., 2021).

Neurotropism can be the result of blood-borne viruses subverting the blood-brain barrier (BBB) (Iadecola et al., 2020). SARS-CoV-2 has been shown to cross the BBB in a murine model (Rhea et al., 2021). Using the S1 subunit of the SARS-CoV-2 spike protein as a proxy for uptake, viral entry was observed throughout the mouse brain parenchyma and in endothelial cells (Rhea et al., 2021). These results implicate direct viral invasion of the BBB in SARS-CoV-2 mediated neurological dysregulation. ACE2 and other viral proteases are expressed on endothelial cells of the vasculature (Xiao et al., 2020). Because lung tissue infected with SARS-CoV-2 shows patterns of damage characteristic of pulmonary fibrosis, viral blood-borne tropism may be facilitated in COVID-19 through upregulated ACE2 in these arterial vascular cells (Guo et al., 2020). Immune cells reaching the CNS may also promote neurotropism, particularly via macrophages and monocytes (Merad and Martin, 2020). Although several autopsy studies have not revealed widespread immune cell infiltration, the potential contributions of infected innate and acquired immune cells to neurotropism may be consequential (Iadecola et al., 2020).

Once in the circulation, the virus may easily avoid the BBB by targeting circumventricular organs in the brain. These CNS sites have areas of fenestrated capillaries where the endothelial border does not represent a fixed boundary to molecular entry (Iadecola et al., 2020). The presence of ACE2 receptors in such delimited CNS locations support this potential route of entry (Chigr et al., 2020). Furthermore, inflammation resulting from reactive oxygen species and free radicals produced by SARS-CoV-2 infection can damage the BBB, facilitating neurotropism via the circulation in other locations lacking fenestrations (Li et al., 2020c). Similarly, cytokines, which are elevated during COVID-19 infection, are known to cross the BBB and contribute to neuroinflammation observed in SARS-CoV-2 brain neuropathological studies (Azizi and Azizi, 2020; Iba et al., 2020; Rhea et al., 2021).

Although the BBB largely segregates the brain from the systemic circulation, toxic wastes and metabolites must still be removed from the brain. A waste removal system, termed “glymphatics” for its dependence on glial cells, is facilitated by a CSF and interstitial fluid transport system (Benveniste et al., 2017). In this system, CSF is transported within a peri-vascular network that helps drain waste from the brain parenchyma (Benveniste et al., 2017). In addition to glymphatics, a true lymphatic vasculature that drains interstitial fluid has been described in the brain’s meningeal layer (Solomos and Rall, 2016). Furthermore, the brain dural sinuses are another location of potential neuro-immune interactions. At the dural sinus, accumulated CNS-derived antigens in the CSF are taken up by dural antigen-presenting cells that introduce the antigens to patrolling T cells (Rustenhoven et al., 2021). Circulating lymphocytes and viral proteins in these systems might be potential routes of both neuroinvasion and subsequent transport back to the periphery (Solomos and Rall, 2016). Other coronaviruses have previously been documented to be present in the CSF (Nath, 2020), and now SARS-CoV-2 has been observed in the CSF of COVID-19 patients with rare neurological presentations like acute necrotizing encephalopathy and demyelination (Domingues et al., 2020; Virhammar et al., 2020).

The ACE2 receptor plays a pivotal role as the initiator of SARS-CoV-2 cellular entry in multiple organs (Xiao et al., 2020). When the viral spike glycoprotein interacts with ACE2, SARS-CoV-2 is endocytosed (Liu et al., 2020a). ACE2 is a key protein for the renin-angiotensin system (RAS), which plays a central role in the homeostatic regulation of electrolyte and fluid balance, blood pressure, arterial oxygenation, and end organ function, particularly those mediated by the renal and cardiovascular systems (Li et al., 2017; Miller and Arnold, 2019). The RAS is divided into vasopressor and vasoprotective axes (Li et al., 2017). The multiple arms of this complex pathway ultimately result in the dynamic interplay of two main enzymes: Angiotensin Converting Enzyme 1 (ACE1) and ACE2. ACE1 converts angiotensin I (Ang I) to angiotensin II (Ang II), while ACE2 converts Ang I and Ang II to angiotensin (1-7) (Ang (1-7)) and angiotensin (1-9) (Ang (1-9), respectively (Miller and Arnold, 2019). The vasopressor axes are comprised of the classic angiotensinogen/renin/ACE1/Ang II axis and the prorenin/renin axis (Li et al., 2017). The vasoprotective axes, which are dependent on ACE2/Ang (1-7)/Ang (1-9)/Mas receptors, counteract detrimental effects of the vasopressor axes (Li et al., 2017). When SARS-CoV-2 is endocytosed, the ACE2 enzyme is endocytosed with it, shifting the balance of the RAS toward the pro-inflammatory vasopressor axes (Li et al., 2017; Gheblawi et al., 2020). Such disruption of RAS balance, which has previously been implicated in the development of cardiovascular disease, renal disease, and hypertension, amongst other systemic derangements, may be a driver of COVID-19 systemic dyshomeostasis (Li et al., 2017).

Proteases help to orchestrate neurotropism by assisting the ACE2 receptor interaction with SARS-CoV-2. For epithelial cells, these proteases may even assist viral invasion through non-ACE2-mediated routes, via CD147-spike protein and CD26 expressed ubiquitously (Radzikowska et al., 2020; Wang et al., 2020a). ACE2 receptors are further modulated by being endocytosed following binding to SARS-COV-2 (Gheblawi et al., 2020; Hirano and Murakami, 2020). Transmembrane protease serine 2 (TMPRSS2) is a serine protease on the plasma membrane that mediates spike protein activation and promotes SARS-CoV-2 entry via direct fusion, thereby subverting endocytic entry (Bailey and Diamond, 2021). ACE2 can exist as a transmembrane bound protein in vascular endothelial cells, or as a circulating form, once cleaved by TMPRSS2 or transmembrane protease serine 4 (TMPRSS4) (Xiao et al., 2020). ACE2 is subsequently shed, giving rise to the circulating form, while TMPRSS2 and TMPRSS4 simultaneously facilitate the endocytosis of SARS-CoV-2 (Xiao et al., 2020).

The SARS-CoV-2 spike protein has been shown to interact with the soluble form of ACE2 or a soluble ACE2-vasopressin complex extracellularly, and then may enter cells via endocytosis mediated by the angiotensin II type I receptor (AT₁R) or arginine-vasopressin receptor 1B (AVPR1B), respectively (Yeung et al., 2021). Additionally, the soluble form of ACE2 preserves the viral binding site and may therefore facilitate viral tropism of cells where the tissue-bound form of ACE2 is poorly expressed (Yeung et al., 2021).

Furin is a pro-protein convertase found in many tissues, including the brain, neuroendocrine organs, the GI tract and liver, with few differences in expression levels (Wu et al., 2020). The SARS-CoV-2 spike protein contains a unique furin cleavage site not found in other β-coronaviruses (Wu et al., 2020). Furin cleavage facilitates stronger viral receptor binding and membrane fusion. These may contribute to the high infectivity profiles and the multi-organ involvement, especially where ACE2 is present at lower levels of expression (Wu et al., 2020). For example, furin produced in intestinal cells could provide an avenue for viral entry into the CNS via the myenteric nerve plexus. Moreover, loss of furin is an effective measure to prevent viral infection (Johnson et al., 2021). Other proteases like cathepsin L, cathepsin B, trypsin, factor X, elastase, and Coronavirus 3CL protease have also been implicated in SARS-CoV2 binding (Gheblawi et al., 2020; Macchiagodena et al., 2020). CRISPR-Cas-9 screening has also identified several genes necessary for the synthesis of glycosaminoglycans, which are negatively charged polymers that likely increase infectivity by attracting exposed positive charges on viruses (Bailey and Diamond, 2021).

Interferon modulation of ACE2 receptors can lead to increased degrees of neurotropism (Merad and Martin, 2020; Ziegler et al., 2020). Interferons are downstream inflammatory products of IL-1, IL-6, and tumor necrosis factor (TNF), key inflammatory molecules released by immune cells in multiple tissues affected by COVID-19 (Ziegler et al., 2020). Severe COVID-19 infection and death following infection have been associated with elevated inflammatory markers and chemokines (Merad and Martin, 2020). The resulting interferons enhance viral invasion as Ziegler et al. demonstrated that ACE2 receptors represent the translation product of interferon-stimulated genes in human barrier tissue epithelial cells (Ziegler et al., 2020). Smoking and chronic obstructive pulmonary disease (COPD) are associated with the increased presence of endocytic vacuoles implicated in SARS-CoV-2 endocytosis (Eapen et al., 2021). Patients with this pathogenic profile may be more susceptible to viral entry and associated profiles of neurotropism. With endocytosis and downregulation of surface ACE2 proteins, the pro-inflammatory axes of the RAS can prevail, namely through Ang II and the AT₁R (Gheblawi et al., 2020). Interferon-stimulated ACE2 and ACE2 endocytosis both can modulate viral entry. However, a novel truncated isoform of the ACE2, termed deltaACE2 or MIRb-ACE2, has recently been discovered (Ng et al., 2020; Onabajo et al., 2020). This truncated ACE2 isoform has been observed to be induced by interferons and viruses, including SARS-CoV-2 (Ng et al., 2020; Onabajo et al., 2020). The novel truncated ACE2 isoform does not appear to bind the SARS-CoV-2 spike protein or act as a peptidase, suggesting that interferon-stimulated induction may not play a role in promoting SARS-CoV-2 cellular entry (Ng et al., 2020; Onabajo et al., 2020).

The Ang II peptide of the RAS can act on a few receptor types, the most notable being AT₁R to effect change, including increased sympathetic tone (Xia and Lazartigues, 2010; Miller and Arnold, 2019). AT₁R is a major mediator of RAS in raising blood pressure through several homeostatic mechanisms (Miller and Arnold, 2019). AT₁R is localized at circumventricular organ sites and other cardinal integrative regulatory centers of the brain like the hypothalamus and medulla (Doobay et al., 2007). Circulating Ang II can mediate activation of the SNS by acting on circumventricular organs and the carotid body (Porzionato et al., 2020). This system is triggered during times of blood loss, hyponatremia, renal hypotension, general SNS activation and infection. Notably, an overactive SNS is implicated in many comorbidities associated with mortality in COVID-19 (Porzionato et al., 2020). Ang II interacts with the pro-inflammatory AT₁R, and downstream products of these processes activate STAT3 and NF-κB transcription factors (Hirano and Murakami, 2020). This promotes IL-6 production, which upregulates NF-κB and STAT3 with production of other pro-inflammatory cytokines (Hirano and Murakami, 2020).

ACE2 allows Ang (1-7) to act on Mas receptors to countervail AT₁R actions (Miller and Arnold, 2019), since the Mas receptor mediates vasodilation and anti-proliferative effects (Xia and Lazartigues, 2010). ACE2 overexpression promotes a protective phenotype in cardiovascular disease states in transgenic mouse models (Alenina and Bader, 2019). ACE2 in the hypothalamus has powerful antihypertensive and sympathetic nervous system dampening effects, with ACE2 overexpression increasing Ang (1–7) relative to Ang II, thereby providing beneficial effects in mouse models displaying brain injury and stroke (Alenina and Bader, 2019). ACE1/ACE 2 receptor interactions work to effectively mediate the effects of Ang II and Ang (1-7). Unfortunately, with SARS-COV-2 infection, AT₁R activity frequently prevails.

As mentioned, the SARS-CoV-2 spike protein interacts with the ACE2 receptor in epithelial cells, which line a variety of tissues including the lung, heart, and kidney. SARS-CoV-2 infection thus potentially contributes to intricate profiles of end organ damage via interacting neural and ACE2-RAS deregulation of cardinal homeostatic processes. The involvement of and damage to the CNS affects different systemic organ systems which then exert feedback regulation to the CNS to exacerbate the neurological and overall clinical patterns of COVID-19 infection (Robinson et al., 2016; Batlle et al., 2020; Garvin et al., 2020; Hundt et al., 2020; Lei et al., 2020). Recent autopsy results of COVID-19 patients reflect dysregulation of key molecular factors involved in hypoxia, coagulation, and fibrosis in multiple organs, highlighting the extent of aberrant local regulatory processes and systemic homeostasis (Nie et al., 2021).

SARS-CoV-2 has been hypothesized to infect the CNS through the cribriform plate, travel in a retrograde fashion through peripheral olfactory nerves to the brain stem (inclusive of the medulla) and to target and destroy the medullary centers responsible for cardiorespiratory control, such as the Pre-Botzinger complex. This has been thought to result in the compromise of the respiratory centers (Gandhi et al., 2020). Similar neurotropic viruses, like the avian coronavirus, have been reported to track to nuclei of the medulla: the NTS and the nucleus ambiguus (NA), sites that receive information from the chemoreceptors of the respiratory tract and lungs and innervate local resident smooth muscle, glands, and vessels (Costa et al., 2014; Li et al., 2020b).

In SARS-CoV-2 infection, there is a proposed loss of homeostatic control by the medulla resulting in cardiorespiratory compromise: inadequate blood flow, respiration and oxygenation, leading to a multifaceted impairment of alveolar gas exchange. In healthy individuals, peripheral chemoreceptors, present at highest concentrations in the carotid body and aortic arch, monitor the flux of arterial oxygen levels. When there is arterial oxygen desaturation, an excitatory signal is sent from the glomeruli here, most directly to the NTS, integrating at the rostral ventrolateral medulla (Costa et al., 2014). In a hypoxic state with low oxygen reserves, the body will aggressively work to increase blood flow to high impact organs, promote vasoconstriction in other peripheral locations, and increase respiratory rate as well as inspiratory/expiratory force, in an attempt to efficiently deploy oxygen (Costa et al., 2014; Tassorelli et al., 2020). There is likely an inability to adequately detect and respond to changes in oxygenation when SARS-CoV-2 infection compromises medullary and carotid body oxygen sensing and set-point regulation as well as sensory nerve-related respiratory muscle function. The often profound hypoxemia, frequently without premonitory symptoms, is characteristic of COVID-19 patients facing imminent respiratory compromise and is indicative of homeostatic deregulation.

Interestingly, increases in peripheral hypoxic chemosensitivity have been ascribed to overstimulation of the sympathetic nervous system (Porzionato et al., 2020). A hyperactive sympathetic nervous system due to higher hypoxic chemosensitivity has been implicated in chronic obstructive pulmonary disorder and metabolic syndrome, comorbidities that enhance the severity of COVID-19 (Porzionato et al., 2020).

Acute GI symptoms of COVID-19 include nausea, vomiting, and diarrhea (Esposito et al., 2020; Li et al., 2020c). Viral neurotropism of the area postrema and in the NTS of the medulla oblongata of the brainstem may disrupt their normal homeostatic roles. The NTS is particularly known to be targeted by other coronaviruses, emphasizing its potential role in the acute GI symptoms frequently seen in COVID-19 (Porzionato et al., 2020).

The ENS regulates gut homeostasis through interactions between enteric glial cells, gut epithelium, and gut-associated lymphoid tissue (GALT) (Esposito et al., 2020). Enteric glial cells serve as antigen-presenting cells to the GALT, and their activation is characterized by IL-6 release, contributing to inflammation in COVID-19 (Esposito et al., 2020). Furthermore, their activation by viruses has been implicated as a key step in neurological immune priming that leads to later neurological impairments (Esposito et al., 2020). Deregulation of these components of the nervous system may therefore be responsible for the observed GI dysfunction.

Specific forms of neurological dysfunction in COVID-19 may be related, among other pathological processes, to characteristic features of the newly-defined COVID-19 associated coagulopathy (CAC). Disruption of interactions between the coagulation cascade and inflammation has been observed in SARS-CoV-2 infected patients; CAC is unique in that it presents with elevated fibrinogen, D-dimer levels, Von Willebrand factor (VWF), factor VIII and inflammatory cytokines that can induce a generalized thrombotic disorder (Iba et al., 2020). SARS-CoV-2 infected macrophages express tissue factor on their cell surface promoting coagulation. Factor VIII and VWF released by SARS-CoV-2 infected endothelial cells also contribute to increased coagulation. Thrombosis may occur due to these multifaceted prothrombotic modifications in cell signaling (Iba et al., 2020). However, unlike other coagulopathies, CAC is not marked by thrombocytopenia and does not show increased partial thromboplastin time, distinguishing it from similar disorders of sepsis-induced coagulopathy (Iba et al., 2020).

CAC is notable for endothelial damage as SARS-CoV-2 directly infects vascular endothelial cells via ACE2 receptors with the help of TMPRSS2 cleavage. Damaged endothelial cells then fail to produce nitric oxide, thus allowing adhesion of leukocytes and platelets and migration of inflammatory cells into the vessel wall (Iba et al., 2020). In the event of infection, endothelial cell damage results in tissue factor release, especially in the brain, activating thrombin, the final serine protease in the coagulation cascade (Festoff and Citron, 2019). Because thrombin exerts pro-inflammatory effects in addition to its role in coagulation, it plays a key role in the interactions mediating the coagulation-inflammatory nexus (Festoff and Citron, 2019). Thrombin cleaves the protease-activated receptor on endothelial cells to gain access through the BBB to the CNS, where it can then lead to neuroinflammation (Festoff and Citron, 2019).

The result of this robust coagulation, inflammation, and endothelial damage, including active viral-mediated endotheliitis, is profound degrees of vasculitis, thrombosis, and stroke affecting large arterial and venous vessels as well as the entire brain and body microvasculature, including arterioles, venules and capillary beds. This results in complex impairments in oxygen exchange and unique microinfarctions and mechanosensitive microbleeds with dramatic CNS microglial activation, active innate immune sensing and neuroinflammation (Azizi and Azizi, 2020; Iba et al., 2020; Thakur et al., 2021). Microscopic examination of the brains of patients who have died from COVID-19 has indicated multifocal microvascular ischemic and hemorrhagic parenchymal injury, neuroinflammation and microglial activation without cell autonomous viral effects targeting neurons, astrocytes or oligodendrocytes (Lee et al., 2020; Thakur et al., 2021).

An unusual inflammatory syndrome characterized by fever, inflammation, and multisystem organ injury has been associated with a subset of COVID-19 patients, including children (Whittaker et al., 2020). Elevated fractions of mononuclear phagocytes, particularly inflammatory monocyte-derived macrophages, have also been observed in COVID-19 patients (Liao et al., 2020; Merad and Martin, 2020). Overactivation of these mononuclear phagocytes may contribute to the cytokine storm that is characteristic of COVID-19, and the cytokine profile of COVID-19 hyperinflammation resembles that of macrophage activation syndrome (Merad and Martin, 2020). Elevated cytokines, namely IL-6 and tumor necrosis factor (TNF), released by macrophages in severe SARS-CoV-2 infection, not only promote an inflammatory milieu, but may also stimulate upregulation of the ACE2 receptor, further driving SARS-CoV-2 regional cellular tropism (Merad and Martin, 2020).

Hyperinflammation and activation of mononuclear phagocytes may also contribute to the process of hypercoagulation observed in COVID-19. In the absence of direct vascular damage, the initiation of coagulation is dependent on increased expression of the pro-thrombotic molecule, coagulation factor III, on mononuclear cells via pro-inflammatory cytokines (Merad and Martin, 2020). An inflammatory state inhibits anticoagulant pathways, further promoting a severe systemic hypercoagulable state. For example, antiphospholipid antibodies, which are characteristically present in several autoimmune conditions, activate the complement and coagulation cascades to contribute to thrombosis and to mediate proinflammatory signaling in innate immune and vascular endothelial cells (Muller-Calleja et al., 2021). Müller-Calleja et al. recently described a process by which endothelial protein C receptors act in complex with the lipid lysobisphosphatidic acid to activate pathogenic antiphospholipid antibodies, representing a mechanistic link between coagulation and innate immune signaling (Muller-Calleja et al., 2021).

The autonomic nervous system plays a crucial homeostatic role through its pro-inflammatory and anti-inflammatory effects on immune cells. An innate immune response activates sensory neurons that release neuropeptides such as calcitonin-gene-related peptide and substance P, the downstream products of which promote inflammation (Hanoun et al., 2015). When the HPA axis detects inflammation via afferent nerves or inflammatory markers, it exerts an anti-inflammatory effect by stimulating glucocorticoids and dopamine release from the adrenal glands (Hanoun et al., 2015; Maryanovich et al., 2018).

Both major branches of the autonomic nervous system contribute to neural regulation of the immune response. The SNS releases noradrenaline that exerts distinct actions through specific subtypes of adrenergic receptors. Noradrenaline has pro-inflammatory effects at low concentrations and anti-inflammatory effects at high concentrations (Hanoun et al., 2015). Chronic inflammation alters the expression pattern of adrenergic receptor subtypes and shifts local innervation toward pro-inflammatory sensory nerves (Hanoun et al., 2015); it is possible that COVID-19 causes this type of shift as well (Porzionato et al., 2020). Furthermore, in murine models, increased sympathetic stimulation led to noradrenaline-mediated vascular inflammation and increased output of neutrophils and inflammatory monocytes (Porzionato et al., 2020). While there is no definitive evidence for direct parasympathetic innervation of the immune system, inflammatory cytokines activate afferent branches of the parasympathetic NS to send signals to the NTS and activate efferent branches of the parasympathetic NS (Hanoun et al., 2015; Maryanovich et al., 2018). The parasympathetic NS can then counteract an inflammatory state via acetylcholine release (Hanoun et al., 2015). For example, activation of the parasympathetic NS has been shown to mediate an anti-inflammatory immune cell response via inhibition of tumor necrosis factor release by macrophages (Porzionato et al., 2020).

The circadian control of the immune system represents an additional axis that may be disrupted by COVID-19. The “master clock” of the circadian system is the suprachiasmatic nucleus of the hypothalamus (Scheiermann et al., 2013). The HPA axis and SNS modulate local circadian rhythms in body tissues through hormonal and autonomic neural regulation of peripheral clock components (Scheiermann et al., 2013). The SNS can act as a local regulator of the body clocks because the SNS directly innervates tissues and drives cyclical noradrenaline release from nerve varicosities (Scheiermann et al., 2013). A large complement of circulating hematopoietic cells, lymphocytes, hormones, and cytokines have been shown to exhibit circadian oscillations (Scheiermann et al., 2013). In murine models, the acute inflammatory response exhibits circadian rhythms, likely due to circadian-influenced leukocyte migration and phagocytic activity (Scheiermann et al., 2013). Furthermore, several diseases in humans display circadian dysregulation as part of their presentation. Myocardial infarction and ischemic stroke, documented consequences of COVID-19, are examples of such diseases, where a surge in SNS activity during the morning hours is thought to contribute to cardiovascular complications (Scheiermann et al., 2013; Choudry et al., 2020). It is conceivable that SARS-CoV-2 neurotropism, both in the central and peripheral nervous systems, could therefore be adversely impacting the modulation of the circadian system over the immune system, thereby contributing to some of the observed COVID-19-related acute phase complications.

Furthermore, the peripheral nervous system also dynamically regulates bone remodeling and bone marrow effector functions. Hematopoietic stem cells (HSCs) in the bone marrow are activated in response to infection to maintain homeostasis (Maryanovich et al., 2018). The SNS has been documented to regulate the microenvironment in the bone marrow, where sympathetic nerve fibers help to create niches essential for normal HSC egress and differentiation of individual blood elements as well as behavioral effects during stress (Maryanovich et al., 2018). Sensory nerves also reach the bone marrow, and sensory neuropeptides have been suggested to directly contribute to regulating HSC activity (Maryanovich et al., 2018). Many of the unique systemic manifestations of COVID-19 affect the derivatives of HSCs like platelets, red blood cells, and immune cells, potentially suggesting that HSCs represent another end organ effector system at the cellular level whose components are differentially regulated and impaired by SARS-CoV-2.

Bidirectional neuro-immune communications involving the lymph nodes represent another axis where the function of sensory neurons may be disrupted. Recent work by Huang et al. further elucidates previously described communications between the PNS and the immune system (Huang et al., 2021). They find that lymph nodes are innervated by heterogeneous populations of sensory and sympathetic neurons which demonstrate inflammation-induced plasticity (Huang et al., 2021). These sensory neurons are largely skewed toward peripheral lymph nodes, suggesting that these neurons can act to monitor the homeostatic status of the lymphatic system (Huang et al., 2021). When activated, these neurons can modulate activity-dependent gene transcription in immune cells (Huang et al., 2021). SARS-CoV-2 neurotropism could be differentially targeting sensory neurons in this heterogeneous population, thereby contributing to the maladaptive immune response. Dysregulation of the interactions of the PNS with the immune system may therefore be contributing to the hyperinflammatory state of COVID-19 and its concomitant and protean systemic manifestations.