SARS-CoV-2 variants such as the delta or omicron variants, with higher transmission rates, accelerated the global COVID-19 pandemic. Thus, novel therapeutic strategies need to be deployed. The inhibition of acid sphingomyelinase (ASM), interfering with viral entry by fluoxetine was reported. Here, we described the acid ceramidase as an additional target of fluoxetine. To discover these effects, we synthesized an ASM-independent fluoxetine derivative, AKS466. High-resolution SARS-CoV-2–RNA FISH and RTqPCR analyses demonstrate that AKS466 down-regulates viral gene expression. It is shown that SARS-CoV-2 deacidifies the lysosomal pH using the ORF3 protein. However, treatment with AKS488 or fluoxetine lowers the lysosomal pH. Our biochemical results show that AKS466 localizes to the endo-lysosomal replication compartments of infected cells, and demonstrate the enrichment of the viral genomic, minus-stranded RNA and mRNAs there. Both fluoxetine and AKS466 inhibit the acid ceramidase activity, cause endo-lysosomal ceramide elevation, and interfere with viral replication. Furthermore, Ceranib-2, a specific acid ceramidase inhibitor, reduces SARS-CoV-2 replication and, most importantly, the exogenous supplementation of C6-ceramide interferes with viral replication. These results support the hypotheses that the acid ceramidase is a SARS-CoV-2 host factor.

Recently, we have shown that C6-ceramides efficiently suppress viral replication by trapping the virus in lysosomes. Here, we use antiviral assays to evaluate a synthetic ceramide derivative α-NH2-ω-N3-C6-ceramide (AKS461) and to confirm the biological activity of C6-ceramides inhibiting SARS-CoV-2. Click-labeling with a fluorophore demonstrated that AKS461 accumulates in lysosomes. Previously, it has been shown that suppression of SARS-CoV-2 replication can be cell-type specific. Thus, AKS461 inhibited SARS-CoV-2 replication in Huh-7, Vero, and Calu-3 cells up to 2.5 orders of magnitude. The results were confirmed by CoronaFISH, indicating that AKS461 acts comparable to the unmodified C6-ceramide. Thus, AKS461 serves as a tool to study ceramide-associated cellular and viral pathways, such as SARS-CoV-2 infections, and it helped to identify lysosomes as the central organelle of C6-ceramides to inhibit viral replication.

Lysosomal accumulation of the glycosphingolipid globotriaosylceramide Gb3 is linked to the deficient activity of the α-galactosidase A in the Anderson–Fabry disease and an elevated level of deacylated Gb3 (lyso-Gb3) is a hallmark of this condition. Localization of Gb3 in the plasma membrane is critical for studying how the membrane organization and its dynamics are affected in this genetic disorder. Gb3 analogs containing a terminal 6-azido-functionalized galactose in its head group globotriose (αGal1,4βGal1,4Glc) are attractive chemical reporters for bioimaging, as the azido-group may act as a chemical tag for bio-orthogonal click chemistry. We report here the production of azido-Gb3 analogs employing mutants of GalK, GalU and LgtC enzymes, which participate in the synthesis of the sugar motif globotriose. Variants of enzymes GalK/GalU generate UDP-6-azido-6-deoxy-d-galactose (UDP-6AzGal), which is the galactosyl-donor used by LgtC for transferring the terminal galactose moiety to lactosyl-acceptors. Residues at the galactose-binding site of the three enzymes were modified to facilitate the accommodation of azido-functionalized substrates and variants outperforming the wild-type enzymes were characterized. Synthesis of 6-azido-6-deoxy-d-galactose-1-phosphate, UDP-6AzGal and azido-Gb3 analogs by variants GalK-E37S, GalU-D133V and LgtC-Q187S, respectively, is 3- to 6-fold that of their wild-type counterparts. Coupled reactions with these variants permit the production of the pricy, unnatural galactosyl-donor UDP-6AzGal with ~90 percent conversion yields and products AzGlobotriose and lyso-AzGb3 with substrate conversion of up to 70 percent. AzGb3 analogs could serve as precursors for the synthesis of other tagged glycosphingolipids of the globo-series.