Covid-19, nervous system pathology, and Parkinson's disease: Bench to bedside

Abstract

Coronavirus disease 2019 (Covid-19) caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection is primarily regarded as a respiratory disease; however, multisystemic involvement accompanied by a variety of clinical manifestations, including neurological symptoms, are commonly observed. There is, however, little evidence supporting SARS-CoV-2 infection of central nervous system cells, and neurological symptoms for the most part appear to be due to damage mediated by hypoxic/ischemic and/or inflammatory insults. In this chapter, we report evidence on candidate neuropathological mechanisms underlying neurological manifestations in Covid-19, suggesting that while there is mostly evidence against SARS-CoV-2 entry into brain parenchymal cells as a mechanism that may trigger Parkinson's disease and parkinsonism, that there are multiple means by which the virus may cause neurological symptoms.

1. Introduction

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) is a novel, highly contagious, and pathogenic strain of the Coronaviruses (CoV) family, which has caused the pandemic of coronavirus disease 2019 (Covid-19) (Hu, Guo, Zhou, & Shi, 2021; Lai, Shih, Ko, Tang, & Hsueh, 2020). Emerging in late 2019, SARS-CoV-2 has taken a toll on global morbidity and mortality as a major public health issue. Although manifesting primarily as a respiratory disease, there are numerous reports of neurological manifestations of Covid-19, either presenting as new symptoms or disorders (e.g., stroke, Guillain-Barré syndrome) or as the exacerbation of pre-existing symptoms of known chronic neurological conditions (Antonini, Leta, Teo, & Chaudhuri, 2020; Kubota & Kuroda, 2021; Leta et al., 2021; Ousseiran, Fares, & Chamoun, 2021). These observations, along with the detection of the virus in post-mortem tissue and cerebrospinal fluid (CSF) (Lewis et al., 2021; Mukerji & Solomon, 2021; Tandon et al., 2021) raise the question of whether SARS-CoV-2 enters the human brain or causes neurological symptoms because of the systemic illness, notably hypoxia and inflammation.

2. SARS-CoV-2 receptors

SARS-CoV-2 shares 80% of its genome with SARS-CoV and 50% with MERS-CoV, with homologies extending to the protein spike (S), which after cleavage by the transmembrane protease serine 2 (TMPRSS2), can bind host membrane proteins acting as viral receptors (Hoffmann et al., 2020; Xu et al., 2020).

The main viral receptor for SARS-CoV-2 and other CoV is angiotensin-converting enzyme 2 (ACE2), an enzyme that catalyzes the cleavage of angiotensin I into angiotensin 1–9 and angiotensin II into the vasodilator angiotensin 1–7 (Wang et al., 2020). ACE2 is a transmembrane protein that is widely expressed in human tissues, and while its expression in brain neurons and astrocytes has not been clearly demonstrated, it is expressed in brain vessels (Hamming et al., 2004) and hypoxic insult appears to upregulate its expression (Zhang et al., 2009).

Similarly to MERS-CoV, virus cellular endocytosis may be alternatively mediated by sialic acid residues, which are located on plasma membrane proteins of several type of cells, including neurons (Fantini, Di Scala, Chahinian, & Yahi, 2020) or by the lectin CD209L, which is a receptor for SARS-CoV (Jeffers et al., 2004), although such alternative receptors have not been clearly demonstrated for SARS-CoV-2.

3. Neuroinflammation in post-mortem Covid-19 brain

Neuroinflammation is prominent in Covid-19 and characterized by lympho-monocytic infiltrations, microglial activation and microglial nodules. Lympho-monocytic perivascular cuffing and infiltration have been reported by most authors, with a predominance of perivascular CD68 + monocytes/macrophages and CD8 + T-cells (Lee et al., 2021; Matschke et al., 2020; Meinhardt et al., 2021; Schwabenland et al., 2021; Thakur et al., 2021), although lymphocyte accumulation appears mild compared to that in viral encephalitis, such as from Herpes viruses.

In all studies examining microglial cells, authors identified prominent microglial activation with upregulation of MHC-II proteins (HLA-DR) and increased lysosomal activity (CD68 +), while maintaining a homeostatic microglial marker TMEM119 (Deigendesch et al., 2020; Matschke et al., 2020; Meinhardt et al., 2021; Schwabenland et al., 2021; Thakur et al., 2021). Similarly, evidence of neuronophagia was documented by most authors with prominent involvement of the lower brainstem at the level of the medullary tegmentum, the midline raphe, inferior olivary nucleus, and the dorsal pons (Al-Dalahmah et al., 2020; Deigendesch et al., 2020; Matschke et al., 2020; Meinhardt et al., 2021; Schwabenland et al., 2021; Thakur et al., 2021). Microglial nodules with associated CD8 +/CD3 + T-cell clusters were also documented in most, but not all, Covid-19 patients, in contrast to non-Covid-19 controls and ExtraCorporeal Membrane Oxygenation (ECMO) patients, who displayed no instances of microglial nodules and neuronophagia (Schwabenland et al., 2021), although we have observed activated microglia and nodules in the brainstems of patients with severe respiratory distress (JE Goldman, P Canoll, unpublished observations). Deep spatial profiling of the local immune response in Covid-19 brains by imaging mass spectrometry revealed significant immune activation in the medulla and olfactory bulb with a prominent role mediated by CD8 + T-cell—microglia crosstalk in the parenchyma (Schwabenland et al., 2021). Conversely, Deigendesch and colleagues found significant differences in HLA-DR + activated microglia when comparing Covid-19 subjects to non-septic controls, but no differences were found with patients who had died under septic conditions; according to the authors, this may represent a histopathological correlate of critical illness-related encephalopathy and hypoxia, rather than a Covid-19-specific finding (Deigendesch et al., 2020).

Neurological symptoms of Covid-19 occur regularly without detection of SARS-CoV-2 in CSF samples, but there is evidence for autoantibody formation. In a cohort of 102 Covid-19 patients, where almost 60% presented some form of neurological symptoms, CSF anti-neuronal autoantibodies were detected in 35% of those tested (Fleischer et al., 2021). In a study of critically ill Covid-19 patients (n  = 11) with unexplained neurological sequelae, anti-neuronal autoantibodies were found in the CSF of all patients, as well as in their serum, suggesting that multiple autoantigens and a potential molecular mimicry to SARS-CoV-2 might mediate these symptoms, especially those related to hyperexcitability (myoclonus, seizures) (Franke et al., 2021). The above findings might guide clinicians to consider administering immunotherapy in selected patients.

Thus, while most studies are consistent on the overall neuroinflammatory conditions occurring in the context of Covid-19, with particular focus on microgliosis and microglia—T-cell crosstalk, disagreements on the source of inflammation are unresolved. In particular, it is not clear how much might be due to an ongoing systemic inflammation/cytokine storm, and whether and how frequent comorbidities, such as hypertension, diabetes, cardiovascular disease or neurodegenerative conditions, and hypoxic/ischemic damage may influence microgliosis and brain inflammation.

4. Little evidence of neuro-invasion in post-mortem Covid-19 cases

Neurotropism has been established for many species of the CoV family, including Middle East Respiratory Syndrome (MERS-CoV) and severe acute respiratory syndrome (SARS-CoV-1) (Zubair et al., 2020), while there have been reports for persistent chronic infections of particular CoV strains in human neuronal cell lines or in the brain of patients with neurodegenerative disorders (Arbour et al., 1999; Murray, Brown, Brian, & Cabirac, 1992). SARS-CoV-2 has been speculated to possess a neuro-invasive potential (Koyuncu, Hogue, & Enquist, 2013; Verstrepen, Baisier, & De Cauwer, 2020; Yachou, El Idrissi, Belapasov, & Ait Benali, 2020), and has been claimed in experimental models (Bullen et al., 2020). However, while there is significant evidence for neuroinflammation in the brain induced by SARS-CoV-2, as detailed here, neuropathological findings at this juncture do not support significant infection of brain cells.

To date, the presence of viral peptides in brain, determined by immunostaining, has been reported only in a small set of patients and in a few cells, and has not been constantly reproduced throughout available studies (Lee et al., 2021; Matschke et al., 2020; Meinhardt et al., 2021; Schwabenland et al., 2021; Solomon et al., 2020; Thakur et al., 2021; Yang et al., 2021). While most studies agree on the detection of low or undetectable viral RNA levels using RT-PCR analyses, detection of viral RNA in blood and blood vessels of the investigated samples cannot be excluded. In contrast, immunoperoxidase and immunofluorescent staining, imaging mass spectrometry and in situ hybridization that provide for the spatial localization of viral proteins/RNA have failed to consistently detect viral antigens or RNA. Matschke and colleagues found sparse immunoreactive cells throughout the brainstem, without specific topographic localization, but also detected distinct immunoreactivity of the glossopharyngeal and vagus nerve bundles (CN IX-X) in a subset of patients, and suggested SARS-CoV-2 retrograde spread through these cranial nerves toward the medulla (Matschke et al., 2020). This is also supported by the detection of SARS-CoV-2 antigen and genomic sequences in the carotid body of Covid-19 subjects (Porzionato et al., 2021). However, there was no apparent association between the presence of viral antigens and neuropathological changes.

Several studies have detected viral proteins and RNA in the olfactory mucosa through immunohistochemistry, immunofluorescence and in-situ hybridization, as well as electron microscopy (Khan et al., 2021; Meinhardt et al., 2021; Zazhytska et al., 2022). Meinhardt et al. observed SARS-CoV-2 spike protein in primary olfactory neurons of the olfactory mucosa, suggesting an olfactory-transmucosal spread of the virus. This was supported by the detection of viral RNA at the level of the olfactory bulb and medulla, even though viral proteins were not detected in non-vascular cells. Furthermore, SARS-CoV-2 immunoreactive endothelial cells associated with microvascular injury and micro-thromboses were found in a subset of Covid-19 patients at the level of the medulla. Conversely, two recent studies found virus in sustentacular cells of the olfactory epithelium, but not in olfactory neurons (Khan et al., 2021; Zazhytska et al., 2022). The latter detected reorganization of nuclear architecture and downregulation of olfactory receptors, as well as their signaling pathways, in the olfactory neurons, suggesting a non-cell autonomous cause of anosmia (Zazhytska et al., 2022).

Schwabenland and colleagues detected viral antigen in ACE2-positive cells enriched in the vascular compartment (Schwabenland et al., 2021). According to the authors, this finding was linked to vascular proximity and ACE2 expression, and was correlated to the perivascular immune activation patterns of CD8 + and CD4 + T-cells and myeloid- and microglial-cell subsets, indicating a fundamental role of the vascular and perivascular compartment, as well as a blood–brain barrier impairment in mediating Covid-19-specific neuroinflammation.

Two studies of single nucleus RNA sequencing of brains of COVID-19 patients detected broad perturbations, with upregulation of genes involved in innate antiviral response and inflammation, microglia activation and neurodegeneration, but found no direct evidence of viral RNA (Fullard et al., 2021; Yang et al., 2021). Other authors have not detected viral proteins/RNA through immunohistochemistry or in situ hybridization, even though viral genomic sequences were found with RT-PCR assays (Lee et al., 2021; Solomon et al., 2020; Thakur et al., 2021).

Hence, SARS-CoV-2 invasion of the CNS remains to be definitively shown, and much evidence points against the presence of detectable virus, although there are reports of rare instances of the presence of viral antigens in brain cells.

5. SARS-CoV-2 and parkinsonism

Understandably, the idea that neuronal infection by SARS-CoV-2 might predispose to parkinsonism has been a source of great concern (Beauchamp, Finkelstein, Bush, Evans, & Barnham, 2020; Bouali-Benazzouz & Benazzouz, 2021; Brundin, Nath, & Beckham, 2020), although, fortunately, there is little evidence supporting this response. Although there are no reports of CoV-associated parkinsonism, antibodies against common CoV have been detected in the CSF of people with Parkinson's Disease (PwP, PD) in significantly higher titers compared to controls, while there was evidence of post-encephalitic parkinsonism in mice infected with a CoV strain (MHV-A59) (Fazzini, Fleming, & Fahn, 1992; Fishman et al., 1985). In a recent literature review, 20 cases of new-onset parkinsonism developing during or shortly after a SARS-CoV-2 infection have been described, presenting a variety of different subjacent mechanisms (Boura & Chaudhuri, 2022). Although these cases do not suffice to support an etiological association between the two entities, the concept of vigilance in the long-term is highlighted.

The concerns on Covid-19 influence on Parkinson's disease stem both from the 1917 Spanish flu/von Economo's encephalitis pandemics as well an association of numerous viruses with the development of transient or permanent parkinsonism (Jang, Boltz, Webster, & Smeyne, 2009), including reports that anti-viral treatment and vaccination are associated with a decreased risk of parkinsonism in humans and animal models (Lin et al., 2019; Sadasivan, Sharp, Schultz-Cherry, & Smeyne, 2017). It has been suggested that pathogens contribute to PD pathogenesis, particularly after the age of 50 in individuals with or without a susceptible genetic substrate (Beauchamp et al., 2020; Tanner et al., 1999). Braak and others have suggested that neurotropic pathogens may infect the CNS via the nasal or gastric pathway, both of which are sites of early pathology in PD (Hawkes, Del Tredici, & Braak, 2007; Klingelhoefer & Reichmann, 2015) and viral-related inflammation might render the CNS susceptible to preceding or subsequent stressors (Sulzer, 2007). Indeed, a meta-analysis demonstrated that a past history of an infection was associated with a 20% higher risk of presenting PD in the future, although this was significant only for bacterial and not viral infections (Meng, Shen, & Ji, 2019). An association between past CNS infections, particularly if multiple hospitalizations preceded, and a subsequent development of PD was described (Fang et al., 2012). Finally, epidemiological data from a large cohort of PD patients and controls indicates an increased PD frequency among occupations with a high risk for respiratory infections, such as teachers and healthcare workers, (the “clustering of PD” theory) (Tsui, Calne, Wang, Schulzer, & Marion, 1999).

While there is little evidence at this time for a role for Covid-19 in increasing PD, there could be effects particularly on peripheral catecholamine systems. L-Dopa decarboxylase (DDC), an essential enzyme in the biosynthesis of dopamine and serotonin, is the most significantly coexpressed and coregulated gene with ACE2 in non-neuronal cell types, significantly affecting dopamine blood levels (Nataf, 2020). SARS-CoV-2 infection of monkey cell lines was found to induce downregulation of DDC, an effect also noted with dengue and hepatitis C infections (Mpekoulis et al., 2021), pathogens associated with parkinsonism (Bopeththa & Ralapanawa, 2017; Tsai et al., 2016). DDC levels rose in nasopharyngeal tissues of asymptomatic or mild severity Covid-19 patients, while an inverse relationship was noted between SARS-CoV-2 RNA levels and DDC expression (Mpekoulis et al., 2021). Moreover, a dopamine D1 receptor agonist was found to suppress endotoxin-induced pulmonary inflammation in mice, suggesting that a potential protective role of dopamine in inflammation needs to be further explored (Bone, Liu, Pittet, & Zmijewski, 2017).

Vascular damage constitutes a recognized complication of Covid-19 (Siddiqi, Libby, & Ridker, 2021). A case of bilateral basal ganglia insult in the context of a thromboembolic encephalopathy without parkinsonism in a Covid-19 patient has been described (Haddadi, Ghasemian, & Shafizad, 2020).

6. Peripheral pathways that may contribute to neuroinflammation

6.1. Olfactory pathway

SARS-CoV-1 enters the brain of transgenic mice via the olfactory bulb, causing neuronal death without signs of encephalitis (Netland, Meyerholz, Moore, Cassell, & Perlman, 2008), while intranasal inoculation of MERS-CoV in transgenic mice led to high levels of the virus in the CNS, particularly in the thalamus and the brainstem (Li et al., 2016).

The characteristic clinical symptoms of hyposmia or anosmia in Covid-19, as well as hypogeusia or ageusia, which are quite prominent among affected patients (Guerrero et al., 2021) support the hypothesis of an involvement of the olfactory system (Lechien et al., 2020). An increased SARS-CoV-2 viral load in the nasal epithelium has been found by RT-PCR in situ hybridization and immunohistochemical staining methods (Meinhardt et al., 2021). Moreover, a prospective autopsy study in the Netherlands involving 21 patients who had succumbed to Covid-19 complications revealed extensive inflammation in the brain among other tissues, particularly focused in the olfactory bulbs and the medulla oblongata, although this study did not reveal presence of the virus in the brain (Schurink et al., 2020). Politi and colleagues described a case report of a Covid-19-positive woman with anosmia, who presented with subtle hyperintensities in the olfactory bulbs and the right gyrus rectus in brain magnetic resonance imaging (MRI), although the authors highlighted that this was not a consistent finding among Covid-19 patients with smell dysfunction (Politi, Salsano, & Grimaldi, 2020). A recently published paper, exploring the findings of brain scans before and after a SARS-CoV-2 infection in a large sample of UK Biobank participants, showed significant reductions in gray matter thickness and tissue-contrast in the orbitofrontal cortex and parahippocampal gyrus, along with prominent alterations in brain areas functionally connected to the primary olfactory cortex, suggestive of tissue damage (Douaud et al., 2022), although it is difficult to correlate the scan findings with cellular pathology. Furthermore, post-mortem brain MRI studies in 19 decedents of Covid-19 demonstrated an asymmetry in the olfactory bulbs in about 20% of the patients (Coolen et al., 2020). Post-mortem studies of Covid-19 patients detected the virus in the sustentacular cells of the olfactory epithelium, but not in the olfactory receptor neurons or the olfactory bulbs (Khan et al., 2021), as noted above.

6.2. Enteric pathways

With both respiratory and gastrointestinal symptoms being quite common in the context of Covid-19 (Cares-Marambio et al., 2021; Groff et al., 2021) and ACE2 being highly expressed in the alveolar epithelial type II cells and the intestinal endothelial cells (Williams et al., 2021), some researchers have suggested the potential of SARS-CoV-2 entry through the intestine (Lehmann et al., 2021; Mönkemüller, Fry, & Rickes, 2020; Zhang et al., 2020). RNA of the SARS-CoV-2 has been identified in stool samples and rectal swabs, even in cases with negative nasopharyngeal swabs (Tang, Schmitz, Persing, & Stratton, 2020). An autopsy study (n  = 21) revealed SARS-CoV-2-infected cells in the respiratory and gastrointestinal tract, among other tissues, along with extensive inflammation in the medulla oblongata, implying a potential effect on the respiratory control center (Schurink et al., 2020). Many studies report inflammation in the medulla (see above), but the link between peripheral organs and the brain is not clear. One possible way in which systemic inflammation can influence medullary nuclei is the activation of vagal medullary afferents by inflammatory molecules. Thus, vagal endings contain receptors to IL-1β and are activated by this cytokine, subsequently activating neurons of the solitary nucleus, which then activate sympathetic output from the vagal nucleus and the nucleus ambiguous through the vagus nerve (Pavlov & Tracey, 2012).

6.3. Vascular pathways

Non-neuronal pathways of SARS-CoV-2 that might lead to neuroinflammation must be considered, including the hematogenous route. SARS-CoV-2 can enter the blood stream and either access and damage the endothelium of brain vasculature to cross the BBB, or trigger an inflammatory response, leading to breakdown of the BBB (Meinhardt et al., 2021; Wan et al., 2021).

Cerebrovascular disease, including ischemic and hemorrhagic strokes, constitutes a severe and common complication in Covid-19 (Tsivgoulis et al., 2020). The brains of patients who did not survive Covid-19 show acute cerebrovascular disease, including thrombotic microangiopathy and endothelial injury, without any evidence of vasculitis (Hernández-Fernández et al., 2020; Wan et al., 2021). Accumulating data suggests that Covid-19 is connected to a hypercoagulable state, which is closely related to inflammation and predisposes to macro- and microvascular thrombosis, resulting in arterial and venous infarcts (Abou-Ismail, Diamond, Kapoor, Arafah, & Nayak, 2020).

ACE2 is expressed in cells of the vasculature in the human brain, although the specific cell types have not been defined (Hamming et al., 2004). TMPRSS2 and NRP1 are also present in the endothelial cells of the vasculature (Wan et al., 2021). SARS-CoV-2 endothelitis has also been confirmed in autopsy studies concerning other human tissues (lungs, kidney, heart, liver, small intestine) (Varga et al., 2020). Moreover, Pellegrini and colleagues showed a leakage across the BBB, using a model of the human choroid plexus epithelial cells infected by SARS-CoV-2 (Pellegrini et al., 2020), although one has to consider the results of such in vitro studies carefully, since they may not reflect what occurs in the living brain.

On the other hand, a SARS-CoV-2 infection can cause an excessive systemic inflammatory response after triggering a cytokine storm in the periphery, resulting in a BBB disruption (Sulzer et al., 2020). Researchers have suggested that SARS-CoV-2 may infect immune cells, using them as a “trojan horse” to invade the CNS via the impaired BBB or activate different populations of immune cells, which may infiltrate the CNS, causing a secondary cytokine storm and thus neurologic manifestations (Pezzini & Padovani, 2020; Wan et al., 2021; Williams et al., 2021).

6.4. Hypoxia and ischemia

Finally, hypoxic/ischemic changes of the brain due to respiratory abnormalities and hypoperfusion in the context of Covid-19 appear likely to play a crucial role in the development of secondary neurological manifestations (Sullivan & Fischer, 2021). Cases of hypoxic brain injury following Acute Respiratory Distress Syndrome (ARDS) have been described in the literature, including cases of parkinsonism, with the authors highlighting the possibility of silent hypoxia (Ayele et al., 2021; Fearon, Mikulis, & Lang, 2021; Radnis et al., 2020). However, the fact that neurological complications have been reported in the absence of any respiratory symptoms, suggest that alternative mechanisms mediate the virus-induced CNS insults (Ellul et al., 2020).

Hypoxic/ischemic pathology is particularly common in Covid-19, being reported in most studies (Matschke et al., 2020; Solomon et al., 2020; Thakur et al., 2021). Alterations were mostly widespread, and both acute and subacute findings could be appreciated at the level of the cortex, the basal ganglia, the hippocampus and, most notably, the brainstem. Hypoxic/ischemic damage can be associated with vascular pathology, with numerous reported cases presenting both ischemic and hemorrhagic infarcts (Lee et al., 2021; Matschke et al., 2020; Remmelink et al., 2020; Thakur et al., 2021). Reactive astrogliosis in the context of hypoxic/ischemic injuries was also commonly encountered, prominently involving the basal ganglia and the brainstem (Matschke et al., 2020; Meinhardt et al., 2021; Thakur et al., 2021). Microvascular pathology, fibrinogen leakage and small vessel thromboses were associated with ischemic and vascular pathology, representing a characteristic finding of the Covid-19 cohorts (Lee et al., 2021; Matschke et al., 2020; Meinhardt et al., 2021; Porzionato et al., 2021; Porzionato, Emmi, et al., 2021; Schwabenland et al., 2021). Small vessel platelet-enriched microthrombi were predominantly found in the basal ganglia, brainstem and cerebellum, often associated to microvascular injury and fibrinogen leakage. Several instances of SARS-CoV-2 immunoreactive endothelial cells in the context of vascular injury were also detected (Meinhardt et al., 2021), indicating a prominent role of SARS-CoV-2 infection in determining Covid-19 associated small vessel pathology.

7. Conclusion

Although Covid-19 caused by SARS-CoV-2 infection is mainly regarded as a respiratory disease, a variety of clinical manifestations, including neurological symptoms, often occur. The evidence available to date suggests that Covid-19 does not significantly contribute to the incidence of Parkinson's disease and that neurological clinical manifestations observed in the context of Covid-19 are mainly secondary to an indirect damage by SARS-CoV-2 in peripheral systems, in contrast to infection of CNS neurons and astrocytes.