SARS-CoV-2 infection in conjunctival tissue – Authors' reply

We thank Yu-Chi Liu and colleagues for their correspondence on our study¹ on the tropism of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in ex-vivo cultures of human ocular and respiratory tissues, in which they highlight three key points. The first point involves the dynamic nature and antimicrobial properties of the tear film, and the challenge of mimicking these properties in model systems. With regards to the antiviral effects of tear fluid, it is relevant to note that other enveloped viruses (eg, herpes simplex virus type 1 [HSV-1]² and the H7N7 avian influenza A virus³) and non-enveloped viruses (eg, enteroviruses and adenoviruses)⁴ can effectively infect the conjunctiva (leading to conjunctivitis), and that the comparative replication of HSV-1⁵ and pandemic H1N1 influenza viruses⁶ has also been shown in ex-vivo explant cultures of human ocular tissue, even in the presence of the same antiviral and viral clearance mechanisms. As the authors themselves note, conjunctivitis has been reported in patients with SARS-CoV-2, with a prevalence ranging from 0·9% to 31·6%. The fact that the number of conjunctival samples positive for SARS-CoV-2 RNA has been low does not preclude the possibility that conjunctival infection could indeed occur, and our study¹ supports such a possibility. The most notable finding in our study was that SARS-CoV-2 replicated in and infected ex-vivo human conjunctival explant cultures more extensively than SARS-CoV under the same standardised experimental settings.¹

Regarding the second point about the apparent absence of angiotensin-converting enzyme 2 (ACE2) receptors on the conjunctival tissue, we provide clear evidence of virus infection in ex-vivo conjunctival tissue.¹ We argue that the apparent absence of ACE2 receptors on the conjunctival mucosa needs to re-investigated, and if this receptor is confirmed to be absent, then alternative receptors for the virus need to be sought.

With regards to the third point about the immunohistochemistry findings, the viral antigen-positive cells were, in our view, mainly the epithelial cells on the surface and not the stromal cells, as assessed by a clinical pathologist and an ophthalmology specialist. Nevertheless, a logical next step for the follow-up study would be to analyse the conjunctiva tissues at multiple timepoints to understand how the virus enters the conjunctiva and proliferates, although this will highly depend on the availability of conjunctival tissue. In summary, our finding that the replication of SARS-CoV-2 is higher than SARS-CoV in a physiologically relevant human conjunctival explant culture model, prompts further awareness and research into this route of transmission, which has been documented clinically.