Detection of SARS-CoV-2 viral proteins and genomic sequences in human brainstem nuclei

Coronavirus disease 19 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), frequently presents with neurological manifestations ranging from mild symptoms like anosmia and headache to severe conditions such as stroke and encephalitis. A significant percentage of COVID-19 patients experience long-term neurological sequelae, known as "long COVID." While SARS-CoV-2 has been detected in the brain and cerebrospinal fluid of some patients, the extent of its neuroinvasive potential and the specific neuropathological alterations caused by direct viral infection versus immune-mediated mechanisms remain unclear. Studies using human neural cell cultures and brain organoids have yielded conflicting results on SARS-CoV-2 neurotropism, with some suggesting limited infection of neural cells but efficient replication in choroid plexus epithelial cells. Conversely, animal studies have shown brain invasion and widespread neuronal infection following intranasal inoculation. Autopsy studies in humans have also produced inconsistent findings regarding SARS-CoV-2 detection in the brain, with some studies reporting limited viral presence in specific brain regions and others showing no detection. Neuropathological changes observed in COVID-19 patients often include ischemic lesions, astrogliosis, microglial nodules, and T lymphocyte infiltrates, predominantly in the brainstem, cerebellum, and meninges. The role of systemic inflammation and hypoxia in mediating brain immune response is also widely debated. The present study aims to investigate the neuropathological changes in COVID-19 patients and compare them to age-matched controls to better understand the neuroinvasive potential of SARS-CoV-2 and its long-term consequences, especially in relation to neurodegenerative diseases like Parkinson's disease.

The literature on SARS-CoV-2 neuroinvasion is marked by conflicting findings. While some studies have reported the detection of SARS-CoV-2 in the brain and cerebrospinal fluid, others have not found any evidence of viral presence. Studies using in vitro models, such as human neural cell cultures and brain organoids, have also produced inconsistent results, with some showing limited viral replication in neural cells and others reporting efficient replication in choroid plexus epithelial cells. Animal models, however, have shown evidence of neuroinvasion and widespread neuronal infection. The neuropathological findings in human autopsy studies are similarly varied, with reports of various degrees of ischemic lesions, astrogliosis, microglial nodules, and T cell infiltrates in the brainstem, cerebellum, and meninges. There is ongoing debate about the relative contributions of direct viral invasion, immune-mediated mechanisms, systemic inflammation, and hypoxia to the observed neuropathological changes. Some studies have reported a lack of direct link between neuropathological alterations and viral invasion. The inconsistencies across these studies necessitate further research into the neuroinvasive potential of SARS-CoV-2 and the long-term neurological consequences of infection.

This study involved a post-mortem analysis of brain tissue from 24 COVID-19 patients and 18 age- and sex-matched controls who died due to pneumonia and/or respiratory failure. All COVID-19 patients had confirmed SARS-CoV-2 infection via molecular testing of rhino-pharyngeal swabs. The brains were immersion-fixed in 4% phosphate-buffered formalin, sectioned, and subjected to various histochemical and immunohistochemical analyses. Haematoxylin and eosin staining was used for routine histopathological evaluation. Immunoperoxidase staining was performed to detect various markers, including CD3, CD20, CD68, HLA-DR, TMEM119, Ki-67, GFAP, and CD61, to characterize lympho-monocytic infiltrations, microglial activation and proliferation, and the presence of platelet-enriched microthrombi. Immunofluorescent staining and confocal microscopy were used for double-labeling experiments. Real-time RT-PCR was performed to detect SARS-CoV-2 RNA. Antibodies for SARS-CoV-2 nucleocapsid and spike protein were employed to detect viral antigens, with validation through positive and negative controls. ACE2 receptor and TMPRSS-2 protein expression was also assessed. Histopathological and morphometrical evaluations were performed by three independent histopathologists and morphologists blinded to patient clinical data. Microglial density and activation were quantified through digitally-assisted immunoreactivity quantification. Statistical analyses, including Welch one-way ANOVA, t-tests with Welch's correction, Spearman's rho correlation, and linear regression, were used to analyze the data.

The study revealed several key findings. First, a wide spectrum of neuropathological alterations was observed in both COVID-19 patients and controls, including diffuse hypoxic/ischemic damage, astrogliosis, and various degrees of cerebral atrophy and edema. However, COVID-19 patients showed several distinct features. Specifically, CNS platelet-enriched microthrombi in small parenchymal vessels were detected in COVID-19 subjects but not in controls, often affecting multiple organs. RT-PCR analyses detected SARS-CoV-2 RNA in 10 out of 24 COVID-19 subjects, with viral proteins detected in 7 subjects. Immunohistochemical and immunofluorescent staining showed SARS-CoV-2 viral proteins, particularly in neurons of the medulla and midbrain, including within the substantia nigra, in a subset of COVID-19 subjects. Microglial cells displayed an activated phenotype with increased CD68 immunoreactivity in COVID-19 subjects compared to controls, particularly in the medulla and midbrain. A topographically defined pattern of microgliosis was observed in the brainstem, with higher densities in the medulla and midbrain tegmentum. Perivascular macrophages expressed CD206, indicating an anti-inflammatory M2 phenotype, with higher densities in COVID-19 subjects. Microglia exhibited a pro-inflammatory phenotype based on interleukin expression. A strong correlation was found between microglial densities across brainstem levels and CD68+ immunoreactive area, suggesting activated microglia. Finally, a positive correlation was found between medullary microgliosis and hospitalization time. Subjects with detectable viral genomic sequences and antigens had higher microglial densities in the medullary and midbrain tegmentum compared to negative subjects, indicating a targeted microglial response to viral antigens despite no significant difference in overall microgliosis.

The findings of this study provide further evidence for the neuroinvasive potential of SARS-CoV-2 and highlight the complex interplay between viral infection, immune response, and neuropathological changes in COVID-19. The detection of SARS-CoV-2 viral proteins and RNA in specific brainstem nuclei, particularly the substantia nigra, is a notable finding, suggesting a potential direct effect of the virus on these regions. The observed microglial activation and topographically defined pattern of microgliosis in the brainstem further suggest a localized immune response to the viral infection. The correlation between medullary microgliosis and hospitalization time suggests a dynamic inflammatory process that evolves over the course of infection. However, the lack of direct neuronal damage in SARS-CoV-2 infected cells suggests that other factors, such as hypoxia/ischemia and systemic inflammation, may play significant roles in the overall neuropathological changes observed in COVID-19. The findings raise important questions about the long-term consequences of SARS-CoV-2 neurotropism, particularly regarding the potential for triggering or exacerbating neurodegenerative conditions such as Parkinson's disease. The detection of viral proteins in the substantia nigra warrants further research into the potential link between COVID-19 and Parkinson's disease, particularly given the known association between viral infections and the development of neurodegenerative diseases.

This study demonstrates the neuroinvasive potential of SARS-CoV-2, with viral proteins and genomic sequences detected in specific regions of the brainstem, particularly in neurons of the vagal nuclei and substantia nigra. The study reveals a complex interplay of direct viral effects and immune-mediated responses, with microglial activation and a topographically defined pattern of microgliosis in the brainstem being key features. The findings suggest a potential long-term impact of SARS-CoV-2 on the CNS, warranting further investigation into its contribution to neurodegenerative diseases. Future studies should focus on long-term consequences in COVID-19 survivors and explore the link between SARS-CoV-2 neurotropism and the development of Parkinson's disease.

This study has several limitations. First, it is based on post-mortem tissue samples obtained during the first wave of the COVID-19 pandemic, potentially limiting generalizability to other variants and disease courses. Second, the lack of comprehensive neurological evaluations of included patients makes unequivocal clinico-pathological correlations difficult. Third, the retrospective selection of controls (pneumonia patients) may introduce selection bias. Fourth, the limited available clinical data restricts detailed analyses of the relationship between ante-mortem findings and post-mortem observations. Finally, this study primarily focuses on the brainstem, potentially missing important features in other brain regions.