SARS-CoV-2 variants lacking a functional ORF8 may reduce accuracy of serological testing.

Dear Editors,

Serological testing can be used for diagnosis, retrospective assessment of the efficacy of control measures and testing the response to a future vaccine (Weissleder et al., 2020). Most antibody detection assays to estimate SARS-CoV-2 prevalence and incidence probe for the nucleocapsid (N) or spike (S) proteins (Weissleder et al., 2020). A recent study proposed the use of ORF8 and ORF3b antibodies as serological markers of early and late SARS-CoV-2 infection (Hachim et al., 2020). The study is an important step towards standardization of serological assays for COVID-19 and a better understanding of SARS-CoV-2 pathogenicity. ORF8 antibodies were identified as a major marker of acute, convalescent and long-term antibody response to SARS-CoV-2. However, several lineages without a functional ORF8 exist that may affect the accuracy of serological testing (Gong et al., 2020; Pereira, 2020; Su et al., 2020; To et al., 2020). It is important to be aware of this limitation when testing for COVID-19.

The ORF8 accessory gene is specific of Betacoronavirus lineage B (subgenus Sarbecovirus) (Forni et al., 2017). The gene is poorly conserved and its function remains unknown. Strikingly, middle and late phases of the 2002/2003 SARS epidemic were characterized by the spread of viruses with either partial or complete deletions of the ORF8 gene (Consortium, 2004). It is still a matter of debate if the truncated ORF8 may have changed the SARS-CoV replication capacity or virulence in a way that favored its adaptation to humans and/or its spread during the SARS epidemic (Forni et al., 2017; Muth et al., 2018). In the ongoing COVID-19 pandemic, a SARS-CoV-2 variant with a 382-nucleotide deletion at ORF8 emerged early in Wuhan and was exported to Singapore and Taiwan (Gong et al., 2020; Su et al., 2020). The deleted variant resulted in a less severe infection and lower concentrations of proinflammatory cytokines, chemokines and growth factors that are strongly associated with severe COVID-19 (Young et al., 2020).

I recently identified several nonsense mutations and additional deletions in the ORF8 gene that either remove or significantly change the ORF8 protein (Pereira, 2020). Currently, 17 nonsense mutations and eight deletions resulting in truncated ORF8 proteins, or even with the complete removal of the gene, are described in public databases (Table 1 ). In total, mutations were observed in 660 genomes from viruses sampled in patients from different regions of the world (Table 1). Strikingly, the first reported case of a re-infection with SARS-CoV-2 was in a patient whose first infection resulted from a variant with a nonsense mutation at ORF8 (To et al., 2020). The available genomic data suggests that deleted or truncated ORF8 proteins are not a sporadic event and occurred in different times and places. Moreover, the lower severity of infections resulting from variants without a functional ORF8 (Young et al., 2020) suggests that these lineages may escape detection as carriers are not always tested. It is therefore probable that the current list is an underrepresentation of the real number of circulating variants without a functional ORF8. Furthermore, the ORF8 stands out as being one of the most variable SARS-CoV-2 genes (Pereira, 2020). Currently, 194 ORF8 positions were found to have missense variants (https://bigd.big.ac.cn/ncov; accessed on 26 October 2020). Such non-synonymous variants may also be a problem for serological testing by potentially leading to epitope loss and a null serological response.

Table 1

List of nonsense mutations and deletions detected in SARS-CoV-2 ORF8. Data from the China National Center for Bioinformation (https://bigd.big.ac.cn/ncov) and CoV-GLUE database (http://cov-glue.cvr.gla.ac.uk/) accessed on 26 October 2020.

Genome position	N° of cases	Genomic change	Sampling location
27,913	1	T > A	Bangladesh
27,915	38	G > T	Uganda, USA, UAE, Wales, Japan, India
27,945	152	C > T	USA, India, Canada, England, Scotland, Portugal, Belgium, Italy, South Africa
27,948	16	G > T	England, USA
27,960	5	C > T	England, USA
27,972	64	C > T	England, Scotland, Wales, Sweden, USA, Canada, Peru, India
27,978	2	C > T	England
27,986	4	T > A; T > G	England, Georgia, Sweden
28,027	4	G > A	USA
28,041	4	G > T	China, USA
28.050	1	A > T	England
28,068	18	G > T	USA, Australia, Singapore, Slovakia, England, Switzerland
28,072	2	T > A	USA
28,076	2	C > A	England
28,083	195	G > T	India, Bangladesh, Hong Kong, England, Northern Ireland, Scotland, Latvia, Italy, Canada, Gambia, South Africa
28,209	29	G > T	England, Wales, Spain, Iceland, Croatia, Japan, Hong Kong, South Korea, Bangladesh, USA, Australia
28,221	102	G > T	USA, India, England, Belgium, Netherlands, Poland
27,847–28,230	9	Deletion	Singapore, Taiwan
27,910–28,254	2		Bangladesh
27,913–28,254	1		England
27,915–28,253	1		England
28,003–28,227	3		Netherlands
28,004–28,030	2		USA
28,006–28,044	1		England
28,079–28,093	2		USA
Total cases	660

The utility of serological tests depends on the sensitivity and specificity of the assay. Nevertheless, caution is necessary when detecting antibody responses directed against ORF8 antigens, as the absence of a functional gene may render the test inefficient. In case of patients who test negative for ORF8 antibody, it is recommended to always consider other targets (e.g., N or S proteins), particularly for those in which the suspicion for a past infection is high. It is therefore highly recommended to test the antibody responses in patients with COVID-19 infected with the deficient ORF8 variants reported here (Table 1). If that is not possible, the serological test should be accompanied by a note of caution for its use and a disclaimer for patients who tested negative for ORF8 antibodies.