Neurobiology of COVID‐19‐Associated Psychosis/Schizophrenia: Implication of Epidermal Growth Factor Receptor Signaling

ABSTRACT

COVID‐19 exhibits not only respiratory symptoms but also neurological/psychiatric symptoms rarely including delirium/psychosis. Pathological studies on COVID‐19 provide evidence that the cytokine storm, in particular (epidermal growth factor) EGF receptor (EGFR, ErbB1, Her1) activation, plays a central role in the progression of viral replication and lung fibrosis. Of note, SARS‐CoV‐2 virus (specifically, S1 spike domain) mimics EGF and directly transactivates EGFR, preceding the inflammatory process. In agreement, the anticancer drugs targeting EGFR such as Nimotuzumab and tyrosine kinase inhibitors are markedly effective on COVID‐19. However, these data might raise a provisional caution regarding implication of psychiatric disorder such as schizophrenia. The author's group has been investigating the etiologic and neuropathologic associations of EGFR signaling with schizophrenia. There are significant molecular associations between schizophrenia and EGFR ligand levels in blood as well as in the brain. In addition, perinatal challenges of EGFR ligands and intraventricular administration of EGF to rodents and monkeys both resulted in severe behavioral and/or electroencephalographic endophenotypes relevant to this disorder. These animal models also display postpubertal abnormality in soliloquy‐like self‐vocalization as well as in intercortical functional connectivity. Here, we discuss neuropsychiatric implication of coronavirus infection and its interaction with the EGFR system, by searching related literatures in PubMed database as of the end of 2023.

Coronavirus directly transactivates epidermal growth factor receptors. Coronavirus stimulates epidermal growth factor release by ectodomain shedding. The brain activation of epidermal growth factor receptors results in acute or chronic psychotic behaviors.

Abbreviations

ACE2

angiotensin‐converting enzyme 2

ADAM

A disintegrin and metalloprotease

Ang

angiotensin

ARDS

acute respiratory distress syndrome

ASSR

auditory steady‐state response

BBB

blood–brain barrier

CNS

central nervous system

COVID‐19

coronavirus disease 2019

dMMN

duration mismatch negativity

EGF

epidermal growth factor

EGFR

EGF receptor

EPO

erythropoietin

ERP

event‐related potential

fMMN

frequency mismatch negativity

GM‐CSF

granulocyte–macrophage colony‐stimulating factor

HB‐EGF

heparin‐binding EGF‐like growth factor

IL

interleukin

MERS‐CoV

Middle East respiratory syndrome coronavirus

MRI

magnetic resonance imaging

SARS‐CoV

severe acute respiratory syndrome coronavirus

SNP

single‐nucleotide polymorphism

SOFA

sepsis‐related organ failure assessment

TGFα

transforming growth factor alpha

TH

tyrosine hydroxylase

TNF‐α

tumor necrosis factor‐alpha

1. Introduction

The emergence of the novel coronavirus, severe acute respiratory syndrome coronavirus (SARS‐CoV), has caused global pandemics and necessitates our understanding of the biology of its pathogenesis [1, 2]. Coronaviruses comprise a diverse group of enveloped, positive‐sense, single‐stranded RNA viruses and have been present since ancient times, known to cause common cold illnesses [3, 4]. However, their genome mutations have resulted in several significant outbreaks in recent history, with the most notable ones being SARS‐CoV‐1, Middle East respiratory syndrome coronavirus (MERS‐CoV), and the ongoing coronavirus disease 2019 (COVID‐19) pandemic caused by SARS‐CoV‐2 [5, 6]. The complex biology of coronaviruses, characterized by their unique structure, and specific interactions with host cells, particularly neural cells [7], underscores the importance of elucidating their neurobiological intricacies for both psychiatric symptoms and therapeutic strategies [8, 9].

The typical clinical symptoms of COVID‐19 include not only respiratory symptoms (e.g., fever, cough) and gastrointestinal symptoms (e.g., vomiting, diarrhea) but also neurological symptoms (e.g., headache, depression, delusion, delirium, psychosis) [1, 2, 9, 10, 11]. Due to the rapid evolution and mutation of coronaviruses, especially the SARS‐CoV‐2 virus, the diagnostic criteria for COVID‐19 continue to be revised [5, 6]. The latest studies reveal that the Spike protein of SARS‐CoV‐2 mimics epidermal growth factor (EGF), leading to the activation of EGF receptors (EGFR, ErbB1) before the onset of the cytokine storm [12, 13].

This review aims to provide an overview of the potential underpins of neurological and psychopathology of coronavirus infection, starting with the gross nature of the SARS‐CoV‐1/2‐induced cytokine storm, the distribution of their receptor angiotensin‐converting enzyme 2 (ACE2) in the central nervous system (CNS) [14, 15], and their molecular impact on neural cell hosts (all see below). We will also explore the neurological and psychiatric impact of coronavirus infection and the pathological contribution of the EGFR signaling system. By delving into the details of coronavirus neurobiology, we aim to unravel the potential neuropsychiatric impact of their Spike protein mimicking EGF. In the following sections, therefore, we will introduce our studies on the EGFR system in brain development and animal modeling for schizophrenia, as well as its genetic and phenotypic associations with schizophrenia. Although the psychopathological aspects of COVID‐19 are still largely hypothetical and require experimental and prospective confirmation, gaining a deeper understanding of the neurobiology of coronaviruses may shed light on the molecular pathological interactions between viral infections and psychiatric disorders [14, 15].

2. Method for Literature Selection

We conducted a search for previous publications related to our main topics of the etiological and pathological impact of COVID‐19 on psychosis and schizophrenia with the following keywords: {COVID‐19 OR SARS‐CoV‐2} and {schizophrenia OR psychosis} in the NCBI PubMed database as of December 31 in 2023. The initial search yielded approximately 1000 papers, some of which included irrelevant topics such as schizophrenia patients suffering from COVID‐19, vaccination of patients, and the environmental effects of the pandemic on mental health. We excluded these unrelated papers to construct a clear logical flow for this review, which focuses on the biological underpinnings of COVID‐19‐associated schizophrenia or psychosis. Regarding EGF and EGFR signaling in schizophrenia, there are nearly 100 existing literature sources. The content of these sources is often confirmatory or descriptive in nature, and the present citations were limited to the original works or those relevant to the present discussions. It is important to note that the current hypotheses regarding the role of coronavirus‐driven EGF signaling in the onset of schizophrenia or psychosis are highly speculative at this stage, given the uncertainty surrounding the etiology or neuropathology of schizophrenia. With this respect, the present article represents a hypothesis‐driven review but not a comprehensive review regarding COVID‐19‐associated psychosis/schizophrenia.

3. Biology of EGF and Other Cytokines in COVID‐19

COVID‐19, caused by the SARS‐COV‐2 virus results in extensive cytokine induction, often referred to as a “cytokine storm” or cytokine release syndrome, which is a critical aspect of the immune response to the coronavirus [16, 17, 18]. This dysregulated immune response results in widespread inflammation and leads to organ damage including the lungs, heart, and brain [19], although some controversy still remains [20]. Upregulation of proinflammatory cytokines can also produce symptoms such as a fatigue syndrome and, rarely, depression or psychosis, which might result from their CNS actions [21]. It agrees with the fact that some cancer patients receiving cytokines like interleukin‐2 (IL‐2), interferon (IFN), and erythropoietin (EPO) as part of anticancer chemotherapy have been reported to experience symptoms of depression or psychosis [22, 23, 24], while it remains to be elucidated how these cytokines penetrate the blood–brain barrier (BBB) and affect brain functions [25].

Several cytokines have been implicated in the cytokine storm associated with severe COVID‐19 [16, 18, 20, 26]. These include EGF, IL‐6, IL‐1β, tumor necrosis factor‐alpha (TNF‐α), IL‐2, and various chemokines [18]. Compared with other respiratory virus infections, COVID‐19 exhibits a unique cytokine profile; less IL‐10 and granulocyte–macrophage colony‐stimulating factor (GMCSF) and more EGF and/or CCL5 [16, 18, 20, 26]. Severe cases of COVID‐19 with cytokine storms can lead to acute respiratory distress syndrome (ARDS) and multiple organ failure. Death with lung fibrosis, in particular, is responsible for the development of ARDS [27]. Thus, the management of cytokine storms in COVID‐19 is the most important for the treatment of COVID‐19. This includes the use of anti‐inflammatory drugs like dexamethasone, JAK inhibitor [17, 28, 29]. Targeted antibody therapies designed to reduce specific cytokines, such as EGF, IL‐1, or IL‐6, have been employed to manage severe cases and reduce mortality [19, 30, 31]. It is noteworthy that these cytokines have also been implicated in the pathogenesis of schizophrenia [32].

Chemokines act as chemoattractants, guiding immune cells to the site of infection, nose, lung, and colon. In the context of COVID‐19, the expression of CCL2 (MCP‐1), CCL3 (MIP‐1α), CCL4 (MIP‐1β), and CCL5 (RANTES) are most remarkable [18, 33, 34, 35, 36]. CXCL10 levels in blood are suggested to be the biomarker predicting COVID‐19 outcome [33, 35]. Bioinformatic analyses of cytokine profiles among COVID‐19 pneumonia, other non‐COVID viral pneumonia, and bacterial pneumonia revealed that there are three clusters associated with viral proliferation, ARDS, and organ damage (SOPE). Among these clusters, the time correlation analysis revealed the highest temporal link between the increases in SARS‐CoV‐2 viral titer and blood EGF concentrations during the pathogenesis of COVID‐19 [36] (Figure 1). Thus, EGF may play a pivotal role in viral infection or replication [36]. The induction of EGF in lung fluids [33] or that of its homolog, HB‐EGF, subsequently resulted in ARDS with COVID‐19 [37, 38], since the cell growth signaling of EGFR is known to potentiate in the lung fibrosis of COVID‐19 [39, 40, 41]. In addition, positive interactions between the EGFR system and IL‐6 signaling, which was named as the IL‐6 amplifier, can be implicated in the following cytokine storm of COVID‐19 [42, 43, 44, 45]. For further details of the COVID‐19‐triggered cytokine storm, please see, the examples [18, 33, 34, 35, 36, 46]. With the given tight interaction of the EGFR system and COVID‐19, here, we will introduce the neuropathological implication of COVID‐19 below.

FIGURE 1.

The multi‐correlation analyses with COVID status and various blood cytokine levels of patients with COVID‐19, non‐COVID‐19 viral pneumonia, and bacterial infection pneumonia [36]. The highest correlation of viral titer is found with EGF levels followed by CCL5 (RANTES) levels. The correlations of these cytokines and chemokines appeared to form the three clusters, each of which appears to reflect the magnitudes of viral proliferation, the respiratory deficits (ARDS), and the organ failure (sepsis‐related organ failure assessment; SOPE). The circle size grossly reflects relative concentrations of these cytokines. This figure is adapted from the data and figure [36].

4. Infection or Inflammation of SARS‐CoV‐2 in the Nervous System

The CNS impacts of cytokines induced by COVID‐19 infection are poorly understood but potentially involve brain inflammation as well as its structural and functional changes [9, 47, 48]. Douaud et al. [49] investigated the structural alterations in magnetic resonance imaging (MRI) data before and after COVID‐19 and found remarkable gray matter reduction in the orbitofrontal cortex and parahippocampal gyrus, implicating the impairments of olfactory and cognitive functions. In agreement, COVID‐19 patients have often reported neurological and psychiatric symptoms, including anosmia, ageusia, confusion, delirium, psychosis, depression, seizures, and even stroke [9, 47, 48]. These neurological symptoms have prompted questions about whether the virus infects neurons or glial cells in the CNS and how cytokine‐induced inflammation in the periphery may influence these central symptoms.

One of the host cell receptors for the SARS‐CoV‐2 virus is angiotensin‐converting enzyme 2 (ACE2), which is responsible for COVID‐19 infection [50, 51]. ACE2 is an enzyme that plays a crucial role in regulating the renin–angiotensin–aldosterone system by converting angiotensin II (Ang II) to angiotensin 1–7 (Ang 1–7), potentially contributing to the COVID‐19 pathogenesis as well [52]. Although the reason for the cell surface anchoring of the ACE2 enzyme remains a mystery, it is expressed in various tissues and organs throughout the human body [53]. Notably, ACE2 is highly expressed in the respiratory tract, including lung epithelial cells, which explains the respiratory symptoms observed in COVID‐19 [50, 51]. Additionally, ACE2 is expressed in neuroepithelial cells, glomerular neurons of the olfactory bulb, midbrain dopaminergic neurons, noradrenergic neurons, and neurons in the basal ganglia and pons, suggesting potential direct infection of SARS‐CoV‐2 virus in the brain [54]. Although some controversy still remains between human brain and rodent CNS [55], it is noteworthy that ACE2‐positive neurons are overlapped with tyrosine hydroxylase (TH)‐positive neurons as well as with those of EGFR‐positive cells [56, 57].

Beckman et al. [58] established the monkey nasal infection model with SARS‐CoV‐2 and demonstrated the potential that SARS‐CoV‐2 virus enters the CNS transnasally to infect various types of brain cells. Following nasal infection, the immunoreactivity for SARS‐CoV‐2 in NeuN‐positive neurons is increasingly spreading from piriform cortex to entorhinal cortex, orbitofrontal cortex of aged monkeys [58]. Crunfli et al. found that COVID‐19 patients with neurological and psychiatric manifestations aftereffects exhibit brain atrophy in the left superior temporal cortex and the orbitofrontal cortex, regions often implicated in schizophrenia neuropathology. In these regions, immunoreactivity for SARS‐CoV‐2 was detected in astrocytes surrounding neurodegenerative loci in patients' brain [59]. DeMarino et al. [60] identified viral particles of SARS‐CoV‐2 in brain biopsy samples. Using autopsy samples from patients with SARS‐CoV‐1, Xu et al. [61] had identified viral particles and genomes using electron microscopy and qPCR. Bakhtazad et al. [54] proposed the possibility of central infection by SARS‐CoV‐1/2 viruses and their potential contribution to neurological manifestations. However, the full extent of this argument and its clinical implications are still under investigation [62, 63, 64].

Alternatively, the neuropathological symptoms of COVID‐19 may be ascribed to the inflammatory cytokines penetrating through the BBB or to those produced endogenously in the brain [65, 66]. In another monkey model of SARS‐CoV‐2 infection, widespread activation of microglia was observed by brain imaging using the radiolabeled ligands binding to microglia [11, 67]. Activated microglia have the capacity to produce proinflammatory cytokines, contributing to neuroinflammation alongside neurodegenerative reductions in size in the temporal gyrus and medial prefrontal cortex. Similar glial activation was detected in COVID‐19 patients [68, 69, 70] as well as in the olfactory system of the other monkey model [71].

It can be postulated that in COVID‐19, the presence of the virus or viral components in the brain could trigger microglial activation. However, the severe systemic inflammation characteristic of COVID‐19 can lead to disruption of the BBB [66]. This disruption facilitates the entry of cytokines into the brain, potentially inciting inflammation within the CNS. While the biological activities of proinflammatory cytokines in the CNS lie outside the scope of this discussion, they are well documented to influence synaptic transmission and glial functions [21], thereby contributing to neurodegenerative and neuropsychiatric outcomes. Moreover, inflammatory signals such as cytokines and prostaglandins secondarily trigger cell release of EGF or other EGFR ligands via their ectodomain shedding, leading to EGFR activation [42, 43, 72, 73].

In this context, it might be reasonable to attribute at least some if not the majority of COVID‐19‐associated neurological or psychopathological symptoms to a certain set of cytokines [63]. Indeed, it is noteworthy that certain COVID‐19 survivors continue to experience lingering neurological or neuropsychiatric symptoms even after the resolution of acute infection [9, 47, 48]. This phenomenon has prompted our exploration into the persistence and delayed consequences of the cytokine storm on brain function, despite the current waning global impact of the COVID‐19 pandemic.

5. Molecular Activity of Corona Spike Protein on EGFR

EGF is a well‐known growth factor for epidermal cells and is often implicated in cancer proliferation and metastasis [72, 73]. Its receptor is referred to as EGFR, ErbB1, or Her1, which also bind to other EGF‐like molecules (Figure 2). The activation of EGFR is often involved in viral host proliferation and/or infection of a variety of viruses [74, 75]: hepatitis C [76, 77, 78], papillomavirus [79], vaccinia virus [77, 80], influenza virus [81, 82, 83], cytomegalovirus [84], and SARS‐CoV viruses [12, 13] (Figure 2). Notably, some of these viruses are associated with the maternal or infant infection hypothesis for schizophrenia [85, 86]. It is speculated that viral infection and the resulting EGFR activation may disrupt neuronal development and/or brain organization, potentially contributing to the development of schizophrenia [85, 86], although there is no direct evidence linking EGFR activation to psychiatric symptoms in either clinical studies or animal models of the viral infection.

FIGURE 2.

Viral interactions with the EGF receptor (i.e., EGFR, ErbB1, Her1) and its receptor homologs (ErbB2‐4). EGFR interacts with other endogenous peptide ligands; transforming growth factor alpha (TGFα), amphiregulin, epigen, HB‐EGF (heparin‐binding EGF like factor), betacellulin, and epiregulin in order to regulate cell proliferation, survival, cell migration, and local inflammation. Coronavirus such as SARS‐CoV, vaccinia virus, hepatitis C, papillomavirus, and influenza A virus and their viral components directly activate or indirectly interact with EGFR to promote their infection and proliferation. The interaction with EGFR additionally influences cell signaling from the other EGFR‐homologues, ErbB2‐4 (i.e., her2‐4), through their heterodimerization.

Research by Eierhoff et al. [87] revealed that EGFR activation aids in the internalization of influenza A virus into host cells through clathrin‐ and caveolin‐1‐dependent endocytosis. Ueki et al. [74] proposed another role of influenza‐triggered EGFR activation; the activation of EGFR in airway epithelium suppresses interferon production to attenuate mucosal antiviral immune responses to the virus. In agreement, pharmacological investigations pointed out the efficacy of the EGFR inhibitors on the infection of influenza A virus [83, 87, 88]. Accordingly, the Food and Drug Administration in the US approved kinase inhibitors for EGFR and others as an antiviral drug for influenza [89]. Similar discoveries have been announced by clinical and basic scientists studying COVID‐19. The first one is the clinical trial of Nimotuzumab conducted in Cuba for COVID‐19 [28, 31, 90, 91]. Nimotuzumab is a humanized monoclonal antibody targeting the epidermal growth factor receptor (EGFR). It has received marketing approval in India, China, and other countries for the treatment of squamous cell carcinomas of the head and neck, similar to cetuximab. The clinical study in Cuba recruited more than 1000 patients and presented promising data; Nimotuzumab increased the recovery rates from COVID‐19 by 2.2‐fold, compared with the control group receiving conventional medication [31]. In agreement with this finding, a comprehensive analysis using bioinformatics revealed that the activation of EGFR plays a central role in the molecular signaling of SARS‐CoV‐2 infection in lung cell cultures [92] and in vivo pharmacology [93]. In agreement, Venkataraman, Coleman, and Frieman [40] reported that SARS‐CoV‐induced lung fibrosis is enhanced in transgenic mice carrying constitutively active EGFR, together with the induction of the EGFR ligands of HB‐EGF and amphiregulin. Hashizume et al. [94] also provide supportive evidence that phenothiazine compounds, which are known to markedly inhibit tyrosine kinases such as EGFR, attenuate SARS‐CoV‐2 cell entry.

In 2023, the other exciting finding on the virology of SARS‐CoV was reported; coronavirus itself can evoke the EGFR signaling without inflammatory processes in vitro [13] as well as in vivo (the author's unpublished data). This phenomenon requires the S1 domain of the spike protein, which forms the ACE2 receptor binding site for SARS‐CoV‐2. It can be reproduced using cultured lung fibroblast cells, and it does not involve cytokine storms or inflammatory responses [12]. That is, the spike protein of the SARS‐CoV‐2 virus possesses a biological activity to stimulate EGF receptors. The EGFR molecule appears to physically associate with the ACE2 receptor as well as with the signal transducer c‐Raf, in response to its interaction with the viral spike protein [12] (Figure 3A).

FIGURE 3.

Provisional two schemas of the EGFR activation in the brain. (A) The ACE2‐mediated direct EGFR activation through the brain infection of coronavirus. (B) The brain–blood barrier (BBB) penetration of lung‐derived EGF into the brain. Downregulation of ACE2 binding to coronavirus elevates extracellular concentrations of Ang II, leading to the activation of AT1 receptors and ADAM that liberates mature EGF from its membrane‐anchored precursors [95, 96]. During the cytokine storm, the BBB is damaged to become leaky, which allows the lung‐derived EGF (MW 5000) to enter the brain.

In addition to the direct action of the coronavirus on EGFR, the SARS‐CoV‐2‐triggered production and release of EGFR ligands are also implicated in its CNS impact (Figure 3B). The association of SARS‐CoV‐2 virus with its receptor ACE2 results in internalization and downregulation of the viral ACE2 complex, leading to the accumulation of its enzyme substrate (i.e., Ang II) [97]. Increased Ang II presumably enhances the receptor binding and signaling of Ang II [98, 99], leading to the ADAM‐dependent ectodomain shedding (i.e., maturation and release) of proEGF‐like precursors [95, 96]. In agreement, the remarkable EGF induction (i.e., more than 1000‐fold; approximately 7000 ng/mL) is reported in the lung fluids from patients with COVID‐19 [33]. The high concentrations of the small polypeptide EGF or its homologs released into the peripheral blood can reach the brain neurons or glial cells, in particular when the BBB is damaged by the cytokine storm [100, 101].

These arguments of EGF production and EGFR activation align with reports suggesting that EGF induction is closely related to the viral proliferation observed in COVID‐19 patients and illustrate the remarkable therapeutic action of EGFR kinase inhibitors on COVID‐19 [40, 90, 91, 102].

6. Association of Psychosis/Schizophrenia With COVID‐19

During the early stages of the pandemic, clinicians published a number of clinical case reports documenting symptoms of psychosis and other psychiatric disorders among COVID‐19 patients [10, 103, 104, 105, 106, 107, 108]. Psychotic symptoms, such as hallucinations, delusions, and disorganized thinking, typically appeared a few days after infection [9, 11, 48, 109, 110]. These symptoms were more commonly observed in severe cases requiring hospitalization but can also occur in moderate cases of COVID‐19 [105]. Examples include persecutory delusions and complex visual and auditory hallucinations that have been reported to persist for up to 40 days [109]. However, the COVID‐19‐associated psychosis was transient in most of the cases, which were mainly reported for hospitalized patients [10, 103, 104, 105, 106, 107, 108]. Several other cohort studies retrospectively examining the health records provide supporting evidence for the association between COVID‐19 and psychiatric symptoms [111, 112]. A large epidemiological cohort study of more than a million electronic health records suggests that some psychiatric and cognitive deficits such as psychosis and seizures are still increasing even at the end of the 2‐year follow‐up and are more apparent in younger people [113, 114, 115]. Although the clinical studies on the COVID‐19‐associated psychosis of younger school children are still limited, their onset of psychosis appears to be more rapid. An 11‐year‐old girl saw and heard bombs and a herd of buffaloes, stampeding and viewing aquatic animals in the bathtub [116]. Another case was that a 14‐year‐old boy was very agitated, angry, and anxious, talking to himself just after recovery from COVID‐19 [117]. However, it is not certain whether these acute psychotic symptoms will result in the onset of schizophrenia or other psychiatric disorders in future [118].

The supportive results of acute or delayed COVID‐19 influence on transient psychotic risks are obtained by other groups [103, 104, 105]. In addition, there was a significant increase in cases of schizophrenia with chronic psychosis during COVID‐19 pandemic [119]. However, it is uncertain whether COVID‐19‐associated psychiatric symptoms stemmed from infection‐driven encephalopathy or encephalitis [9, 108, 120]. Smith et al. [121] have pointed out that, in some cases, what appears to be psychosis may actually be delirium with a clear change in consciousness, a factor that several studies have not excluded. In addition, there is ongoing debate about whether the pandemic‐associated increase in transient psychosis or schizophrenia is attributable to the SARS‐CoV‐2 infection itself or is merely triggered by the social stress of quarantine during the COVID‐19 pandemic [115, 119, 122, 123, 124]. In these contexts, the reported cases of COVID‐19‐associated psychosis may require re‐evaluation with careful diagnostic and psychologic assessments.

A noteworthy report has emerged concerning the significant polygenic association between COVID‐19 and schizophrenia, in contrast to that between COVID‐19 and bipolar disorder/depression [125]. Their analysis utilizing open databases of Genome‐Wide Association Studies (GWAS) for both schizophrenia and COVID‐19 found statistical significance in the genetic overlap probability between susceptibility loci to schizophrenia and COVID‐19, calculating polygenic risk scores and assessing linkage disequilibrium [125].

Previously, similar discussions regarding the correlation between coronavirus infections and schizophrenia had been reported in an epidemiological study. Severance et al. [126] examined blood concentrations of antibodies against various conventional coronavirus strains, including 229E, HKU1, NL63, and OC43, in both patients with schizophrenia and control subjects, involving several hundred samples. Among these strains, NL63 antibody showed significant coronavirus seropositivity rates among patients with schizophrenia, with an odds ratio of 3.10 and a p‐value of 0.013 [126]. Of note, only the strain NL63 utilizes ACE2 as a host receptor among those substrains, which presumably involves the EGFR activation as SARS‐CoV‐2 does [127], which directly perturbs brain development and function (see below), although no significant alterations were found in blood ACE levels between patients with schizophrenia and control subjects [128].

7. Interactions of EGFR/ErbB Signaling With Schizophrenia and SARS‐CoV

Given the biological activity of the coronavirus spike protein, which mimics EGF actions [12, 13], this review provides a concise summary of previous works on the EGFR system in psychopathology, which will hint at the neuropathological impairments stemming from COVID‐19. This review focuses on ligands for EGFR/ErbB1, although it is worth noting that neuregulins also evoke EGFR signaling through heteromeric dimerization between distinct ErbB receptors [72, 73]. For more details on neuregulin biology and pathology, please read other reviews; for example, [129, 130, 131].

The neurobiological association of EGF and other EGFR ligands with schizophrenia has been extensively explored [132, 133, 134, 135]. This research has involved the analysis of blood samples, postmortem tissues, and animal modeling [134, 136, 137]. Futamura et al. [136] initially reported a pathological association between schizophrenia and reduced EGF levels in the blood of patients, along with increased brain EGFR levels. Subsequent research has replicated this initial finding on the association of blood EGF levels with schizophrenia. Additionally, genetic associations between the EGF system and schizophrenia were suggested by other groups [138, 139], although the psychopathological roles of the downregulation and upregulation of the EGFR signaling are debatable [140, 141, 142].

Building on these clinical insights as well as on the cytokine hypothesis for schizophrenia [21, 32], animal models for schizophrenia have been established in rodents and cynomolgus monkeys using various administration protocols as well as with other EGFR ligands [143, 144, 145]. EGF administration to the periphery of neonatal animals, but not to that of pregnant dams or adult animals, results in the postpubertal alterations in animal behaviors of their offspring such as prepulse inhibition, social interaction, latent inhibition of fear learning and working memory [132] (Figure 4). Of interest, intraventricular infusion of EGF into the rat brain also produced similar behavioral deficits, proposing the critical role of EGFR activation in the brain, although the deficits were not persistent [144]. Similarly, perinatal challenges with EGF in cynomolgus monkeys have led to persistent behavioral impairments at the postpubertal stage [145]. With this respect, irrespective of EGFR ligands, the EGFR activation in the developing brain is the key neuropathology of this animal model, involving the hyperactivation of dopamine neurons and developmental retardation of GABA neurons [146, 147]. These neurons are known to specifically express EGFR [146, 147]. These behavioral abnormalities in rodents can be ameliorated not only by conventional antipsychotic drugs such as risperidone and clozapine [134, 148, 149, 150] but also by tyrosine kinase inhibitors targeting EGFR [151, 152], although the pharmacological action of clozapine on the EGFR system is controversial [140, 141]. Of note, the tyrosine kinase inhibitors targeting EGFR are effective on both COVID‐19 and the behavioral deficits of this schizophrenia model, potentially indicating any pathological similarity between these diseases.

FIGURE 4.

Homologous and nonhomologous endophenotypes of perinatal EGF model rats to schizophrenia. EGF is subcutaneously administered to newborn pups to establish this model. After their puberty, these rats displayed a variety of persisting behavioral deficits which had been implicated as schizophrenia endophenotypes (marked blue boxes) [132]. These behavioral abnormalities cannot be observed when adult rats are challenged with EGF [132]. However, if EGF is infused into the brain ventricle of adult rats, some of these behavioral deficits transiently emerge [144]. This perinatal EGF model also exhibits several translatable neurophysiological deficits; frequency and duration mismatch negativity (fMMN, dMMN), auditory steady‐state response (ASSR) at 40 Hz, at attention‐associated P300 event‐related responses (ERP), abnormal functional connectivity in the front‐temporal pathway [146, 147, 148]. SC, subcutaneous.

Most recent studies involve the reverse translation of this animal model into clinical symptoms observed in patients with schizophrenia. This includes abnormal vocalization, impaired EEG profiles (MMN, ASSR, P300), and abnormal functional connectivity [153, 154, 155] (Figure 4). In addition. this EGF model exhibits higher self‐vocalization at the postpubertal stage, which might be relevant to the soliloquy‐like behaviors of patients with schizophrenia [156]. Currently, we are investigating whether perinatal challenges with the coronavirus spike protein produce similar neurobehavioral consequences during the postpubertal stages of rodents as seen in the above EGF‐injection model. This ongoing animal research might hint at the psychiatric consequence of the infants experiencing the vertical transmission of coronaviruses [157, 158].

8. Limitation of the Present Hypothesis and Discussion

Many scientists have cautioned about the potential psychiatric consequences of the COVID‐19 pandemic [9, 120, 159, 160, 161]. It is poorly understood whether the observed psychiatric consequences result from the effects of the virus infection in the brain, the immune response throughout the body, cytokine storms in the periphery, the production of auto‐antibodies, or other associated factors [162]. With a limited number of literature addressing this topic, this review illustrates the potential association between SARS‐CoV‐2 infection and the occurrence of psychotic symptoms, as well as the neuropathological and neuropharmacological implications of EGFR signaling. Therefore, these discussions may reinforce intervention strategies aimed at preventing COVID‐19 and reducing psychiatric risks.

Previous studies on schizophrenia and the latest reports on COVID‐19 suggest a potential link between EGFR hyperactivation and a psychiatric risk in COVID‐19 patients, but more research is needed to fully understand the molecular nature and consequences of this interaction. In this regard, some of the current discussions are purely hypothetical and debatable. Additionally, the severity and prevalence of neurological/psychiatric symptoms vary among COVID‐19 patients, and not all individuals with the virus experience central nervous system (CNS) involvement [9]. Further studies will help clarify the role of EGFR signaling and inflammatory cytokines in psychiatric consequences in the context of COVID‐19 or other coronavirus infections.

Author Contributions

H.N. wrote an original manuscript and M.M. modified and copyedited the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The authors have nothing to report.