Neuropathology and virus in brain of SARS-CoV-2 infected non-human primates

Abstract

Neurological manifestations are a significant complication of coronavirus disease (COVID-19), but underlying mechanisms aren't well understood. The development of animal models that recapitulate the neuropathological findings of autopsied brain tissue from patients who died from severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection are critical for elucidating the neuropathogenesis of infection and disease. Here, we show neuroinflammation, microhemorrhages, brain hypoxia, and neuropathology that is consistent with hypoxic-ischemic injury in SARS-CoV-2 infected non-human primates (NHPs), including evidence of neuron degeneration and apoptosis. Importantly, this is seen among infected animals that do not develop severe respiratory disease, which may provide insight into neurological symptoms associated with "long COVID". Sparse virus is detected in brain endothelial cells but does not associate with the severity of central nervous system (CNS) injury. We anticipate our findings will advance our current understanding of the neuropathogenesis of SARS-CoV-2 infection and demonstrate SARS-CoV-2 infected NHPs are a highly relevant animal model for investigating COVID-19 neuropathogenesis among human subjects.

Fig. 1. Prominent neuroinflammation in brain of SARS-CoV-2 infected NHPs.

Representative images identify microglia through Iba-1 immunopositivity in basal ganglia of mock-infected animals RM6 and AGM5 (a, c) that was upregulated in SARS-CoV-2 infected parenchyma, as shown in RM2 and AGM4 (b, d). Mild-moderate accumulation of microglia was often observed around blood vessels (RM1 f, AGM1 h). Nodular lesions were also frequently observed in brain of infected animals, represented here in RM4 and AGM4 (j, l). Microglial accumulation around blood vessels was not seen in age-matched mock-infected controls (RM6 e, AGM5 g), however, nodules (RM5 i, AGM5 k) were seen. These were less frequent and smaller than those observed in infection. Iba-1 immunopositivity also revealed morphological changes in microglia indicative of increased activation in infected animals, as compared to mock-infected controls, including large cell bodies with short, thickened processes (b, d, f, h, j, l). Microglial expression of HLA-DR was upregulated in the context of infection (n, p) seen in RM2 and AGM2, however, expression was also seen in control animals (m, o) represented by RM6 and AGM5. GFAP expression by astrocytes is upregulated and reveals morphological changes in the context of infection (cerebellum from RM4 r, AGM2 t), indicative of astrogliosis. Cerebellum from non-infected controls RM6 and AGM5 (q, s). Each immunohistochemical stain was performed twice on all brain regions. Abbreviations: AGM African green monkey, RM Rhesus macaque. Scale bars = 100 µm (a–d, m–t) and 50 µm (e–l).

Fig. 2. Neuronal pathology and cell death in SARS-CoV-2 infection.

Representative H&E images show a healthy Purkinje cell layer in the cerebellum of a non-infected control RM6 (a) and reveal cell death-associated neuronal changes in cerebellum of infected animals, (AGM4 b, AGM3 c), and brainstem from AGM3 (d). Arrows indicate pyknotic and karyolitic Purkinje cells and cellular blebs. Asterisks denote areas of tissue necrosis/vacuolation on H&E sections (b) and (d). H&E was performed and assessed twice on all brain regions. Neuronal degeneration in cerebellum was only seen in the context of infection, visualized by positive, green FluoroJade C-stained neurons (AGM2 e, AGM1 f). FluoroJade C staining was performed twice on the brain regions investigated. Abnormal neuronal morphology and cleaved caspase 3 positivity is demonstrated in cerebellum (AGM3 g) and brainstem (RM2 h). Summary data of cleaved caspase 3 positive cells stratified by brain region (i) where n = 4 biologically independent samples/brain region in the control group and n = 8 biologically independent samples/brain region in the infected group, except OL where n = 7 infected animals. Immunohistochemical staining for cleaved caspase 3 was performed twice on all brain regions. Statistics were performed with a two-tailed Mann–Whitney U test. *p ≤ 0.05 and **p ≤ 0.005 comparing mock-infected to infected animals. Data are expressed as mean ± SEM. p values: CER = 0.0081, BS = 0.0182, FL = 0.0323, OL = 0.0061, TL = 0.0485, PL = 0.0485, BG = 0.0040 (control vs. infected). Source data are provided as a Source Data file. Abbreviations: CER cerebellum, BS brainstem, FL frontal lobe, OL occipital lobe, TL temporal lobe, PL parietal lobe, BG basal ganglia, C control, I infected, AGM African green monkey, RM Rhesus macaque. Scale bars = 100 µm (a, c–f, h) and 50 µm (b, g).

Fig. 3. Multiple microhemorrhages in CNS of SARS-CoV-2 infected NHPs.

H&E examination of infected animals revealed microhemorrhages, as demonstrated in cerebellum (AGM1 a), brainstem (AGM2 b, AGM4 c), and basal ganglia (RM4 d), which tended to be larger and packed with red blood cells, as compared to non-infected controls, (RM6 brainstem e, AGM5 cerebellum f). Erythrocyte extravasation into the brain parenchyma is indicated by black arrows. Asterisks denote tissue injury around damaged blood vessels. A dotted line outlines the vessel lumen in each panel. H&E was performed and assessed twice on all brain regions. The number of microbleeds/mm² was assessed in all brain regions (g) and found to be significantly greater in the basal ganglia (h, *p = 0.0263), where n = 4 biologically independent samples in the control group and n = 8 biologically independent samples in the infected group. Statistics were performed with a two-tailed Mann–Whitney U test. *p ≤ 0.05. Data are expressed as mean ± SEM. Source data are provided as a Source Data file. Abbreviations: C control, I infected, AGM African green monkey, RM Rhesus macaque. Scale bars = 100 µm except outset of b where scale bar = 500 µm.

Fig. 4. Reduced CD61 positive associated-microhemorrhages in SARS-CoV-2 infected NHPs.

Representative images show CD61 positivity in intact vessels in parietal lobe from control animal AGM5 (a) and temporal lobe (b) from infected AGM4. CD61 positivity was also seen in association with microhemorrhages, as shown in control cerebellum of AGM5 (c) and brainstem of infected RM2 (d). Microhemorrhages without CD61 positivity were also observed in both non-infected (RM5 brainstem e) and infected (RM1 temporal lobe f) but was more frequent in infection. The percent frequency of microhemorrhages with and without CD61-associated platelet aggregates is shown for each species (g and h). Immunohistochemical staining for CD61 was performed twice on all brain regions. Source data are provided as a Source Data file. Abbreviations: AGM African green monkey, RM Rhesus macaque. Scale bars = 50 µm.

Fig. 5. Reduced blood oxygen may contribute to brain hypoxia in SARS-CoV-2 infection.

SARS-CoV-2 infection was associated with lower blood oxygen levels (a) and increased blood carbon dioxide (b). Yellow shading denotes a lower than physiological range of SpO₂. HIF-1a expression appeared to be upregulated by cells comprising the vasculature and extended into the parenchyma in the context of infection. Expression was significantly greater than age-matched mock-infected control animals in (c) brainstem, *p = 0.0154 (95% CI = 0.06550–0.4895) control vs. infected animals and (d) basal ganglia, **p = 0.0016 (95%CI = 0.1149 to 0.3621) control vs. infected animals. Significant difference was not seen in (e) cerebellum (n = 4 biologically independent samples in the control group, and n = 8 biologically independent samples in the infected group). Statistics were performed with unpaired two-tailed t test, df = 10. Data are expressed as mean ± SEM. When separated by species, statistical significance is retained in the basal ganglia but not brainstem (Supplementary Data Fig. 7). Representative images show low HIF-1α expression in brainstem of mock-infected animals, RM5 (f) and AGM5 (h), as well as basal ganglia of RM6 (j) and AGM5 (l). In comparison, HIF-1a is upregulated in brain of infected animals. Representative images include brainstem of RM3 (g) and AGM4 (i) and basal ganglia of RM3 (k) and AGM1 (m). Immunohistochemical staining for HIF-1α was performed thrice on the brain regions investigated. Source data are provided as a Source Data file. Abbreviations: PI post-infection. O.D. optical density, %SpO₂ blood oxygen saturation, AGM African green monkey, RM Rhesus macaque, C control, I infected. Scale bars = 50 µm.

Fig. 6. SARS-CoV-2 detection in the brain.

Representative single-label IHC shows infrequent SARS-CoV-2 nucleocapsid (SARS-N) positivity in a cerebellar blood vessel of RM1 (a). Positivity was not detected in non-infected controls (AGM6 cerebellum b). SARS-CoV-2 spike (SARS-S) mRNA expression was assessed with in situ hybridization (RNAscope) but not seen in control animal tissue (RM6 cerebellum f-h). Rare positivity was seen in infection (AGM1 cerebellum c–e). RNAscope was performed 7 times on the brain regions investigated. A majority of brain tissue does not show any virus in infected animals which can be seen in the neighboring vessel (RM3 basal ganglia l–n) to a vessel with suggestive virus positivity (i–k). Endothelial cell infection is suggested by double fluorescent labeling of SARS-N with von Willebrand factor (vWF) (RM3 basal ganglia i–k). Merged images show colocalization of SARS-N (red; j) with vWF (green; i), indicated by white arrows. Blue color represents DAPI labeled cell nuclei. Immunohistochemical staining for SARS-N was performed 12 times on the brain regions investigated. SARS-CoV-2 RNA was detected in the different brain regions of infected animals via a CRISPR-based fluorescent method (o) where n = 3 repeats of independent samples. Dotted line indicates the cut-off value of positivity equal to 3.6 × 10⁶ photoluminescence (PL) intensity. Data are expressed as mean ± SEM. Source data are provided as a Source Data file. Abbreviations: BG basal ganglia, CER cerebellum, BS brainstem, FL frontal lobe, CSF cerebrospinal fluid, BC blank control, arb. units arbitrary units, AGM African green monkey, RM Rhesus macaque. Scale bars = 50 µm (a and b), 20 µm (c–h), and 10 µm (i–n). Fluorescence images are at 100×.