The precursors of class I fusion proteins are synthesized in the ER, where they receive initial N-glycosylation and undergo protein folding and trimerization.  However, the majority of these viral proteins are misfolded,  and the elevation of misfolded proteins not only induces ER stress but also triggers their degradation by different mechanisms.  We reported that these viral glycoproteins are new clients for class I α-mannosidases such as MAN1B1, which select substrates for ER-associated protein degradation (ERAD). We also reported that ER chaperones such as calnexin, calreticulin, and PDIA3 target Ebola virus glycoproteins to reticulophagy via K27-polyubiquitination. Currently, we are investigating the molecular mechanisms of how reticulophagy selectively targets these viral glycoproteins.

Properly folded class I fusion proproteins exit the ER and enter the Golgi to undergo maturation. High-mannose type N-glycans are processed into complex-type and hybrid-type N-glycans, and O-glycans are further added to these proteins. Furin further cleaves these proproteins into the surface receptor-binding and the transmembrane fusion subunits. We reported that MARCHF8, an E3 ubiquitin ligase, blocks these maturation steps to inhibit viral infection.  Currently, we are investigating how MARCHF 2, 8, and 9 proteins inhibit the N- and O-glycosylation and the furin activity.  Furin primarily localizes to the trans-Golgi network (TGN), where it cleaves and activates a wide range of immature proproteins that play essential roles in cellular homeostasis, disease progression, and infection. Furin is retrieved from endosomes to the TGN after being phosphorylated. However, it is still unclear how furin exits the TGN to initiate post-Golgi trafficking and how its activity is regulated in the TGN. Thus, MARCHF proteins provide a great tool to address these outstanding questions.

Serine incorporator 5 (SERINC5) is a multipath transmembrane host restriction factor, which is incorporated into virions from the plasma membrane and inhibits viral entry. This SERINC5 activity is antagonized by different retroviral accessory proteins including HIV-1 Nef, MLV glycoGag, and EIAV S2. We reported that SERINC5 disrupts HIV-1 Env trimer formation to inactivate the Env activity, and these accessory proteins degrade SERINC5 via the endocytic pathway. However, the precise SERINC5 antiviral mechanism and the viral antagonism are still unclear. Currently, we are investigating how SERINC5 travels to the cell surface from the TGN for incorporation. We reported that SERINC5 undergoes K33-polyubiquitination by Cul3/KLHL20 E3 ubiquitin ligase, which is required for its post-Golgi trafficking. We are also investigating how SERINC5 is internalized from the plasma membrane, which excludes SERINC5 from incorporation into virions. We reported that SERINC5 is phosphorylated by CCNK/CDK13, which is required for its endocytic degradation by Nef. We will continue to address these outstanding questions to elucidate the important role of SERINC5 in retroviral infection.

SARS-CoV-2 entry requires not only the cell surface receptor ACE2, but also two cellular proteases TMPRSS2 and CTSL  to further cleave its spike proteins on the cell surface or in late endosomes. However, the precise entry mechanism still needs to be elucidated. Recently, we reported the identification of Tubeimosides as potent Ebola virus entry inhibitors by targeting its receptor NPC1 from screening a natural compound library. Surprisingly, we found that Tubeimosides also inhibit SARS-CoV-2 entry at the same potency, which led us to investigate the role of NPC1 in SARS-CoV-2 entry. After knocking out NPC1 by CRISPR/Cas9, we found that NPC1 is required for SARS-CoV-2 entry and infection in both TMPRSS2(-) and TMPRSS(+) cells. These results demonstrate that the endocytic entry route plays a predominant role in SARS-CoV-2 infection. Currently, we are investigating how SARS-CoV-2 spike proteins interact with NPC1 and how NPC1 triggers the membrane fusion in late endosomes for infection.