Fig 2: Subcellular distribution of SW1and LSNME proteins.Upper panels, fluorescence and brightfield micrographs of SW1 mRNA-expressing HEK293 cells at 2 days after transfection, stained with (green) COVID-19 patient plasma and (red) an antibody specific for the lysosomal protein Lamp2, with (merge) colocalization evident in yellow, and the brightfield image showing all cells in the field. DAPI and a plasma from a COVID-19 patient. Lower panels, fluorescence and brightfield micrographs of LSNME mRNA-expressing HEK293 cells at 2 days after transfection, stained with (green) COVID-19 patient plasma and (red) an antibody specific for the ER protein GRP78/BiP, with (merge) colocalization evident in yellow, and the brightfield image showing all cells in the field. Bar, 50 μm.

Fig 3: Spike trafficking to lysosomes is mediated by its ECD.(A) Line diagram of spike proteins showing the position of its ECD (bracket), transmembrane domain (TMD; gray outline), and C-terminal tail (CTT; black outline), with orange denoting spike sequences and purple denoting CD81 sequences. (B to E) Fluorescence micrographs of doxycycline-induced (B) HTet1/SD614G cells, (C) HTet1/SD614G?C + 81T cells, (D) HTet1/SD614G?C cells, and HTet1/SD614G?C + 81TMDT cells that had been processed for IFM using with anti-spike immune sera (green), DAPI (blue) and anti-Lamp2 antibody (red). Scale bars, 50 µm (main image) and 4 µm (magnified outset). These results are representative of two independent trials (n = 2).

Fig 4: D614G alters spike trafficking in virus-infected cells.Confocal fluorescence micrographs of Vero/TMPRSS2 cells infected with (A to D) HP76, a WT strain of SARS-CoV-2 that expresses the D614 form of spike, and (E to H) HP7, a D614G strain of SARS-CoV-2. Cells were infected with equal amounts of virus (MOI of 10), incubated for 18 hours, then fixed, permeabilized, and stained with affinity-purified antibodies specific for the C-terminal 14 amino acids of spike (green), a monoclonal anti-Lamp2 antibody (red), and DAPI (blue). White arrows point to the plasma membrane accumulation of spike in HP76-infected cells. White arrowheads point to spike-positive virus particles within Lamp2-positive lysosomes. Insets (3.2-fold higher magnification) show greater detail in areas of particular interest. Scale bars, 10 μm (main images) and 2 μm (magnified outsets). These results are representative of four independent trials (n = 4).

Fig 5: Human cells traffic spike to lysosomes.(A to D) Immunofluorescence micrographs of doxycycline-induced HTet1/SD614G cells that had been fixed, permeabilized, and stained using anti-spike immune sera (green), DAPI (blue), and antibodies (red) specific for (A) Lamp1, (B) Lamp2, (C) Lamp3, or (D) mTOR1. White arrowheads denote the positions of spike-containing, lysosome-related structures. Scale bar, 50 µm (main images). Outsets show detail in areas of particular interest. Scale bar, 4 µm (outset). These results are representative of three independent trials (n = 3). (E to H) Immunogold electron micrographs of doxycycline-induced HTet1/SD614G cells processed for electron microscopy using (E and F) affinity-purified anti-spike antibodies specific for its C-terminal 14 amino acids (12-nm gold) or (G and H) both anti-spike C14 antibodies (12-nm gold) and anti–Lamp2 (L2) antibodies (6-nm gold). Black arrowheads denote the positions of 12-nm gold directed against spike, and white arrowheads denote the positions of 6-nm gold directed against Lamp2. Scale bars, 200 nm. These immunogold labeling results are representative of two independent trials.