Elevated circulating levels of HERV-W ENV and proinflammatory cytokines in SARS-CoV-2-positive patients with psychosis spectrum disorder

A consortium of authors conducted an observational study to investigate a possible association between severe acute respiratory syndrome corona virus-2 (SARS-CoV-2) and expression of an envelope (ENV) protein encoded by human endogenous retrovirus type W (HERV-W) in a group of hospitalized psychotic patients who had no history of acute COVID-19 infection and had not been vaccinated against SARS-CoV-2. The results showed elevated circulating levels of HERV-W ENV and pro-inflammatory cytokines in SARS-CoV-2-positive psychotic patients. In the context of COVID-19, HERV-W ENV was only found in seropositive psychotic patients, which suggests that this virus has a dominant influence on HERV-W ENV and cytokine expression.

After recovering from COVID-19, a large proportion of patients suffer from long-lasting post-COVID symptoms, even after mild or asymptomatic forms of the disease. Post-COVID symptoms often include neuropsychiatric complications that last for months after the initial infection, strongly suggesting a persisting neurobiological dysfunction. A previous functional MRI study showed that participants with post-COVID and neuropsychiatric symptoms had different brain activity, suboptimal functioning in the normal network and increased activation in the contralateral hemisphere during working memory tasks compared with healthy controls.

The HERVs are retroviral elements derived from retroviruses that infected the human ancestral genome millions of years ago and were incorporated into the chromosomal DNA. Approximately 8% of the human genome, which is four times more than its protein-coding regions, comprises sequences of viral origin that are known as HERVs. Most HERVs have become silenced because of mutations such as substitutions, insertions, or deletions, and as a result of epigenetic modifications, and are vertically transmitted in the population. For a long time HERVs were considered to be part of the “junk DNA,” but, in recent years, they have been shown to perform critical functions in the host. Nonetheless, the aberrant expression of HERVs has been associated with conditions such as infectious, autoimmune, malignant, and neurological diseases. The HERV envelope protein was shown to be neurotoxic. The same research team that conducted the present study has previously shown that the HERV-W ENV protein induces glia- and cytokine- dependent changes in synaptic organization and plasticity of the N-methyl-d-aspartate receptor. Expression of the HERV-W ENV protein in rat brain resulted in abnormal behavior, that was reversed and prevented by a specific anti-HERV-W ENV antibody.

The HERV envelope protein was detected in the blood of approximately one-half of patients with schizophrenia and one-third with bipolar disorder. Schizophrenia and bipolar disorder share approximately 60% of genetic risk factors and exhibit similar brain structural abnormalities, such as reductions in total brain gray and white matter, and increased lateral ventricular volume. However, schizophrenia, bipolar disorder, and multiple sclerosis also exhibit differences regarding the molecular characteristics of HERV-W, which may be partially responsible for the observed differences in phenotype. Furthermore, the nucleotide sequences of HERV-W DNA and RNA slightly differ among all three disorders, possibly reflecting different subsets of HERV-W elements being activated. (Slokar G, Hasler G. Human Endogenous Retroviruses as Pathogenic Factors in the Development of Schizophrenia. Front Psychiatry. 2016 Jan 11;6:183.

About the study

The study included 48 patients with psychosis spectrum disorder, 38 with schizophrenia and 10 with schizoaffective disorder. The sera were collected during the first wave of COVID-19 in the spring of 2020, before vaccines against COVID-19 were available. The patients had a negative PCR test for SARS-CoV-2 infection. They also reported a negative history of COVID-19. The mean age of patients was 41.7 ± 12.7 years, 35% of the patients were women and 65% were men. The study also included a control group of 55 caregivers without COVID-19-related symptoms. The mean age of the control group was 42.9 ± 10.8 years, and 79 % were women.


Although no COVID-19 infection was recorded in patients and controls, a positive anti-SARS-CoV-2 serology has been detected in both groups, indicating either a prior infection or a significant response following the exposure to SARS-CoV-2. Compared to controls, hospitalized psychotic patients who had no history of acute COVID-19 infection and had not been vaccinated against SARS-CoV-2, had a significantly higher number of positive sera and a higher titer of specific anti-SARS-CoV-2 antibodies.

The scientists then analyzed the same sera for the presence of the HERV-W ENV protein. Patients with psychotic disorders had higher levels of circulating HERV-W ENV protein than the whole control group. Furthermore, HERV-W ENV was detected only in seropositive individuals. There was a significantly higher proportion of SARS-CoV-2 seropositive patients with psychotic disorder who were positive for HERV-W ENV (13/48; 27.1%) compared to the control group (5/55, 9%). It is noteworthy that control subjects who were SARS-CoV-2 seronegative did not have any detectable circulating HERV-W ENV protein (0/27). Only four control subjects who were SARS-CoV-2 seropositive had low levels of HERV-W ENV. 

The authors then classified patients with psychotic disorder based on the following clustering variables: SARS-CoV-2 seropositivity, HERV-W ENV positivity and the levels of proinflammatory cytokines (IL-1β, IL-6, IL-8 and TNF-α). The cluster 1 (17%; 8/48) included patients who were negative for both SARS-CoV-2 antibodies and HERV-W ENV antigenemia, and who had low serum levels of proinflammatory cytokines. The cluster 2 (56%; 27/48) included patients with positive SARS-CoV-2 serology but negative HERV-W ENV antigenemia, and low/intermediate values of proinflammatory cytokines. The cluster 3 (27%; 13/48) included patients who were positive for all clustering variables, and with levels higher than in two other clusters.

The results showed that high serum levels of proinflammatory cytokines in psychotic patients were associated with SARS-CoV-2 and HERV-W ENV positivity. The levels of TNF-α, IL-8 and IL-6 were higher in cluster3 than in cluster1 or cluster2. IL-1 was significantly different between cluster3 and cluster1, and it was negative in all control subjects. It is of potential relevance that IL-8, IL-6 and IL-1 were shown to reflect HERV-W expression in endothelial cells.

According to the authors, this study showed that SARS-CoV-2 is connected with HERV-ENV expression and elevated serum levels of proinflammatory cytokines in patients with psychosis spectrum disorder who had no history of acute COVID-19 infection and had not been vaccinated against SARS-CoV-2. Although the underlying pathway is unclear, the authors emphasized that HERV-W ENV was only detected in seropositive psychotic patients, suggesting a dominant influence of this virus on HERV-W ENV and cytokine expression. The connection between this viral infection and the clinical evolution of psychosis requires further investigation.

This article was published in Translational Medicine.

Journal Reference

Tamouza R et al. Patients with psychosis spectrum disorders hospitalized during the COVID-19 pandemic unravel overlooked SARS-CoV-2 past infection clustering with HERV-W ENV expression and chronic inflammation. Translational Psychiatry (2023) 13:272.