There is growing evidence that several genetic loci are shared between AD and COVID-19, with similar directions of impact on the risk of both diseases. Magusali et al. (2021) suggested that the OAS1 rs1131454 variant was associated with increased AD risk via the genetic analysis of 1,313 sporadic AD patients and 1,234 controls. Notably, OAS1 was recently implicated in COVID-19 outcomes as a link for the risk of both diseases at the genetic level. Wang et al. (2022) developed a risk diagnostic model based on a common hub gene in patients with COVID-19 and AD and suggested that patients with COVID-19 are more likely to develop AD. They found four central genes (RAPGEF3, ITPKB, ITPR1, and ITPR3) as biomarkers for predicting AD in COVID-19 patients. COVID-19-mediated AD risk may be related to innate and adaptive immune responses to the virus, as demonstrated in Quincozes-Santos et al. (2021). Using SARSCOVIDB platform analysis, they found that a protease that may be involved in immune system responses, CTSL, was upregulated in brain tissue samples from 18 AD patients. The ApoE4 allele may be another risk gene for subsequent dementia risk in COVID-19 becauseApoE4 allele carriers are at high risk for AD, and it affects the severity and prognosis of COVID-19 clinical manifestations. Manzo et al. (2021) suggested that SARS-CoV-2-induced olfactory and olfactory dysfunction may confer a higher risk of dementia (specifically LOAD, DLB, and LB-variant AD) in ApoE4 carriers than in non-SARS-CoV-2-induced ApoE4 carriers. Follow-up of COVID-19 patients with ApoE4 may help identify early stages of dementia to take full advantage of any available therapy. A recent study identified differentially expressed genes associated with AD, including PSEN2, CTSD, and LGALS3, from a clinical dataset of COVID-19 patients. Notably, these genes are involved in AD pathological progression (Qiu et al., 2022). However, gene sharing only indicates the common inherited genes of the two diseases but not the direction of causality. Most of the published work is based on information from genetic databases, and data from the tissue of COVID-19 patients are lacking. Therefore, further research is needed to fill this gap.

Patients with COVID-19 often exhibit severe innate immune responses and persistently elevated systemic cytokine levels. Inflammatory cytokines, such as IL-1β, IL-6, IL-12, and TNF-α, were significantly elevated in the peripheral blood of COVID-19 patients compared to healthy individuals (De Felice et al., 2020). Sun et al. (2021) found that IL-4 was elevated in the plasma cytokines of COVID-19 participants. In comparison with COVID-19 individuals without neurological manifestations, IL-6 was positively associated with sequelae severity in COVID-19 individuals with neurological manifestations (Sun et al., 2021). Systemic inflammation affects cognitive function and promotes neurodegenerative disease progression. For instance, human cognitive function is negatively correlated with the chronic peripheral elevation of IL-6 (Cojocaru et al., 2011). The increase of IL-1 in rodent brains causes long-term cognitive impairment and increased Aβ and NFT production (Rachal et al., 2001). Knockout or blocking of pro-inflammatory cytokines, such as IL-1 and IL-6 in model mice can improve their spatial memory ability and make them have better cognitive performance (Bialuk et al., 2018). Systemic inflammation may be the pathophysiological mechanism of AD pathogenesis (Akiyama et al., 2000). Compared to healthy subjects, the levels of peripheral IL-1β, IL-2, IL-6, IL-18, and high-sensitivity CRP in AD patients were up-regulated. A meta-analysis evaluating the association between peripheral IL-6 and CRP in healthy adults found that higher concentrations of inflammatory markers increased the risk of AD. TNF-α converting enzyme, soluble TNF receptors 1 and 2, α1-antichymotrypsin were increased, and IL-1 receptor antagonism in the peripheral blood decreased. Systemic inflammation characterized by a ‘cytokine storm’ in severe cases of COVID-19 disrupts the blood–brain barrier and induces neuronal and glial cell damage (Mohammadi et al., 2020). Virus-induced systemic inflammatory storms, coupled with increased blood–brain barrier (BBB) permeability, lead to the entry of a large number of mediators into the central nervous system, which amplifies neuroinflammation and promotes neurodegenerative processes (Maes et al., 2022). Astrocytes can also be infected by SARS-CoV-2. Astrocytes exist in the brain. They make and transport food for neurons. The spike protein on SARS-CoV-2 virus has different receptors for astrocytes and lung cells, which enables them to bind to astrocytes. Because the infected astrocytes could not provide food for neurons, eventually leading to neuron death (Crunfli et al., 2022; Figure 3A). Peripheral inflammatory factors also circulate to the brain and activate cerebral microglia and astrocytes, which secrete pro-inflammatory cytokines that amplify the effects of neuroinflammation (Wang et al., 2019; Kucia et al., 2021).

Increasing evidence emphasizes the contribution of neuroinflammation to AD, which likely plays a vital role in the early stages of the disease where intervention may be most beneficial. Although Aβ deposition is the initiating step of AD pathology, neuroinflammation may be the main driving force of neurodegeneration (Park et al., 2018; Desforges et al., 2020). Neuroinflammation may be the entry of proinflammatory factors into the brain through the damaged BBB or the direct intrusion of SARS-CoV-2 to induce an inflammatory response (Figure 3B). COVID-19-mediated neuroinflammation is manifested by the activation of microglia and astrocytes. SARS-CoV-2 reaches the olfactory bulb of the hypothalamus through the neuroepithelium of the olfactory mucosa (Steardo et al., 2020). The presence of SARS-CoV-2 in the olfactory bulb may lead to the activation of mast cells, microglia, and astrocytes, which leads to the tissue release of pro-inflammatory cytokines (Rutkai et al., 2022). Brains infected with SARS-CoV-2 exhibited a marked inflammatory response and extensive activation of microglia and astrocytes compared to brains without SARS-CoV-2 (Rutkai et al., 2022). Data from an in vitro study highlighted that the SARS-CoV-2 spike protein S1 increased the release of TNF-α, IL-6, IL-1β, and iNOS/NO by stimulating BV-2 microglia (Olajide et al., 2022). Evidence from in vitro and in vivo studies is complemented by clinical samples. Ziff et al. (2022) found that pro-inflammatory cytokines, such as TNFα, IL6, IL1β, and IL8, were significantly increased in the cerebrospinal fluid of patients with COVID-19 neurological syndrome. Poloni et al. (2021) found that microglial activation in the brainstem was significantly increased in COVID-19 cases, and abnormal microglial activation in the frontal cortex and hippocampus was associated with AD pathology. Notably, hippocampal microglia was significantly enhanced in COVID-19 patients with delirium (all patients with neurocognitive impairment). However, no traces of SARS-CoV-2 colonization were found in brain tissue. Schwabenland et al. (2021) detected profound immune activation in the brain tissue of deceased patients with COVID-19, which was manifested by specific CD8 T-cell clusters affecting the vasculature and prominent CD8 T-cell-microglia crosstalk in the parenchyma. This study highlights that SARS-CoV-2 is recognized in cerebrovascular structures and provides a possible basis for broad immune activation. This severe neuroinflammation may partially explain the underlying mechanism of SARS-CoV-2 infection-induced AD development. As innate immune cells in the brain, microglia have long been considered to be related to the pathology of neurodegenerative diseases. More and more shreds of evidence show that activated microglia can produce a variety of neurotoxic factors including pro-inflammatory mediators and reactive oxygen species for a long time, thus causing persistent damage to neuronal. Microglia can be activated for a long time due to single stimulation (e.g., pathogen infection) or multiple stimulations, to aggravate the damage of neurons over time. Brain-targeted SARS-CoV-2 infection can induce microglia activation, then cause chronic inflammation and eventually lead to neurodegeneration (Figure 4). Moreover, neuroinflammation is associated with intense oxidative stress that exacerbates neurodegenerative diseases, such as AD. COVID-19 patients with advanced age and comorbidities with an inflammatory basis (e.g., diabetes, atherosclerosis, and subclinical dementia) may be at increased risk of AD. Therefore, neuroinflammation caused by COVID-19 may be unmanageable, especially in older adults, because their immune system responds less efficiently (Kritas et al., 2020; Schirinzi et al., 2021).

Angiotensin-converting enzyme 2 (ACE2) receptor is the primary mediator of SARS-CoV-2 entry into cells. Some studies suggest that SARS-CoV-2 may be “neuroinvasive” and directly infect neurons or glial cells in the central nervous system (Figure 4). These studies, without exception, found ACE2 expressed in neurons of the brain. SARS-CoV-2 was identified throughout the brain parenchyma in one autopsy sample using antigen staining (Xu et al., 2005). Two dissertations demonstrated ACE2 staining of astrocytes present in the cerebellum and brainstem of rats (Gallagher et al., 2006; Gowrisankar and Clark, 2016). According to a meta-analysis of single single-cell sequencing data sets, ACE2 is expressed at low levels in the human posterior cingulate cortex and middle temporal gyrus (Chen et al., 2020). It is worth noting that SARS-CoV-2 may infect cortical neurons in human brain organoids (which express ACE2) (Bilinska et al., 2020). Knock human ACE2 gene into mouse ACE2 locus, and then intranasal drip of SARS-CoV-2 causes neuronal infection in the brain, but the specific number and distribution of these infected cells have not been determined (Sun et al., 2020). These studies highlight that the expression of ACE2 receptors in brain neurons facilitates SARS-CoV-2 invasion. The presence of SARS-CoV-2 in the brains of COVID-19 patients supports this hypothesis (Matschke et al., 2020; Puelles et al., 2020). Serrano et al. (2021) performed RT-PCR of 16 brain regions in 20 subjects who died of COVID-19 and found that most subjects had non-specific histopathology, including focal β-amyloid precursor protein immunoreactivity and sparse perivascular mononuclear cell cuffs. SARS-CoV-2 RNA was detected in one or more brain regions of four subjects, including the olfactory bulb, amygdala, intra-olfactory region, temporal and frontal neocortex, dorsal medulla, and pia mater (Serrano et al., 2021). An autopsy report on 44 patients who died of COVID-19 has attracted attention. These researchers systematically examined viral load in multiple organs throughout the body, including the brain. The spread characteristics and different damage to various organs after virus infection were studied. A total of 85 anatomical locations and 79 types of body fluid samples were involved, and SARS-CoV-2 RNA detected was detected in all samples. After detection, SARS-CoV-2 RNA was found in 90.9% of brain tissues, second only to lung tissue. This study is the first report to directly demonstrate that SARS-CoV-2 spread in multiple organs and the brain early in infection, and multiple extrapulmonary replication sites were preserved during the first week after symptom onset. These extrapulmonary tissues are likely to prolong the replication time of the virus and lay the foundation for future complications, especially in the nervous system. A recent large study comparing brain scans of the same person before and after infection with SARS-CoV-2 found that brain changes may be an inevitable outcome of SARS-CoV-2 invasion (Douaud et al., 2022). This study also found that participants had more gray matter loss and tissue abnormalities a few months after SARS-CoV-2 infection, primarily in brain regions involved in smell than participants who were not infected with SARS-CoV-2. Some of the altered areas of the brain identified in the study are also involved in memory-related functions.

Once SARS-CoV-2 enters the brain, it binds to ACE2 receptors on neurons, astrocytes, oligodendrocytes, and microglia. Their interactions may lead to astrogliosis and microgliosis, increase BBB permeability, and allow monocytes and leukocytes to infiltrate multiple brain regions of the central nervous system (Vargas et al., 2020), including the olfactory bulb, choroid plexus, cerebral cortex, and middle temporal gyrus (Chen R. et al., 2020; Chen L. et al., 2020; Figure 4). Due to viral neurotropism, SARS-CoV-2 spreads through neuroanatomically interconnected pathways and causes neuronal dysfunction and neurodegeneration in the central nervous system.

The systemic inflammation induced by COVID-19 may lead to rupture of the BBB, which promotes immune cell infiltration (Hascup and Hascup, 2020; Figure 3A). Cytokine storms in the brain also cause rupture of the BBB, which leads to a vicious cycle of increased pathology (Najjar et al., 2020). Therefore, we hypothesized that systemic cytokine storms and subsequent endothelial and BBB dysfunction in COVID-19 patients will increase the risk of AD in subsequent decades, especially in elderly individuals, because the gradual loss of blood–brain barrier integrity is an aging characteristic. However, the length of SARS-CoV-2-mediated disruption of the vascular endothelium and BBB persists is not clear but is the focus of future research.

The major event in AD is the uncontrolled accumulation of toxic Aβ due to an imbalance in production and removal. The core pathway leading to the accumulation of Aβ involves cleavage of the amyloid precursor protein (APP) by γ-secretase. Notably, the multi-group analysis of COVID-19 patient samples shows that there is a potential relationship between the metabolic process of APP and COVID-19 infection. Caradonna et al. (2022) found that the expression of APP was up-regulated after the SARS-CoV-2 virus was bound to the ACE2 receptor. This result is consistent with Camacho et al. (2021) who identified 6 upstream regulators that increased APP production in a COVID-19 patient dataset. RNA-seq analysis highlighted that patients with COVID-19 had significantly increased APP transcripts in blood samples compared to those without COVID-19 (Overmyer et al., 2021). Derived from a study on single-cell RNA-seq by Yang et al., oligodendrocytes were isolated from the brain tissue of dead COVID-19 patients. It was found that APP was one of the most up-regulated genes (Yang et al., 2021). IFITM3 is a viral restriction protein that sequesters viral particles and subsequently transports them to lysosomes. Vavougios et al. (2021) demonstrated dysregulation of IFITM3-dependent pathways in neurons and peripheral immune cells donated from AD patients and this perturbation may be induced by a variety of viruses, including SARS-CoV-2. Notably, IFITM3 upregulation induced γ-secretase activity, which increased amyloid production (Idrees and Kumar, 2021). The heparin-binding site on the S1 protein may help amyloid protein bind to the virus surface, which triggers the aggregation of these proteins and ultimately led to neurodegeneration in the brain. However, there are some inconsistent arguments. For example, multiple transcriptome sequencing of whole blood with red blood cells removed failed to identify differences in the expression levels of APP and PSEN1/2 between COVID-19-infected and uninfected individuals (Tang et al., 2020). There was no significant difference in the transcriptional levels of APP and PSEN1/2 in brain organs derived from pluripotent stem cells between SARS-CoV-2 infected and uninfected persons (Jacob et al., 2020). Therefore, the effect of COVID-19 on APP metabolism still needs further exploration.

Neurofibrillary tangles formed by hyperphosphorylated tau protein are another important neuropathological marker of AD. Recent study suggested that the plasma T-tau concentration of COVID-19 patients without obvious neurological symptoms is significantly higher than that of COVID-19 negative patients, indicating that SARS-CoV-2 may cause tau protein hyperphosphorylation (Lennol et al., 2023). Although there is no evidence to prove the causal relationship, the highly expressed ACE2 receptor in the brain provides a potential mechanism, because the high expression of ACE2 receptor in the brain was inversely proportional to the cognitive decline of AD patients, and was positively correlated with insoluble phosphorylated tau (Louise et al., 2023).