Fig 1.
Coronavirus proteins modulate ATF6-dependent transcription.
(A-E) SARS-CoV-2 ORF screen for ATF 6activators and inhibitors. HEK293T cells were co-transfected with pLVX vectors encoding viral ORFs plasmids or pLVX Empty Vector (EV) with ERSE-luc and CMV-Renilla reporters for 24h prior to harvest for luciferase assay. (A) Krogan SARS-CoV-2 ORF library where all ORFs encode a C-terminal 2x-Strep tag. (B) As in A with untagged SARS-CoV-2 structural proteins. (C) A counter-screen for inhibitors was conducted as in (A) with the addition of Spike as an ATF6 agonist. (D) As in (B) with the addition of Spike as an ATF6 agonist. (E) as in (D) with the addition of M as an ATF6 antagonist as indicated. Blue asterisk indicate significance relative to EV, pink stars indicate significance of M co-transfection relative to paired viral ORF alone (F) HEK293T cells were transfected with M then treated as indicated with thapsigargin (Tg) or tunicamycin (Tm) the following day for 24h prior to harvest at 48h. (G) as in (F) where cells were transfected with M proteins from seven coronavirus known to infect humans or EV. (n = 3 ± SD, statistical significance was determined by one-way ANOVA with Fisher’s LSD test. *, adjusted P value < 0.05; ***, adj. P < 0.0002; and ****, adj. P < 0.00001 relative to EV.)..
Fig 2.
SARS-CoV-2 M protein inhibits ATF6 activation but not PERK or IRE1 activation.
(A) Diagram of the unfolded protein response (UPR) in brief displaying the intermediate steps that occur for activation and generation of transcription factors that lead to individual transcriptional responses. Created in BioRender. Caddell, T. M. (2026) https://BioRender.com/yuwuyag. (B-C) HEK293T cells were transfected with SARS-CoV-2 structural proteins as indicated for 24h before lysate was harvested for western blotting. As a positive control untransfected cells were treated with 200 nM thapsigargin (Tg) for 4h prior to harvest. (D) as in (B) except RNA was harvested for RT-PCR for Xbp1 mRNA splicing assay. (Left) agarose gel; (Right) quantification of image by densitometry. (E) as in (B) where cells were transfected with HA-ATF6 with either M or EV control. Cells were treated with 200 nM Tg treatment for 4h prior to harvest as indicated. (F) HEK293T cells were co-transfected with ERSE-luc, CMV-Renilla, and HA-ATF6-N with or without M as indicated for 24h prior to harvest for luciferase assay. (For blots, representative image of n = 3 is shown. For graphs, n = 3 ± SD, statistical significance was determined by one-way ANOVA with Fisher’s LSD test. ns – not significant; *, adjusted P < 0.05).
Fig 3.
SARS-CoV-2 M protein inhibits activation of host pathways that require anterograde transport in the secretory pathway.
(A) HEK293T cells were co-transfected a LDLR-luc, CMV-Renilla, and with M or EV as indicated. The following day cells were treated with 10 µM cerivastatin (ST) or pleased in serum-free media (SF) for 24h prior to harvest for luciferase assay. (B-C) HEK293 cells were co-transfected with an (B) ISRE-luc or (C) NFkB-luc with cGAS and STING encoding plasmids (to stimulate ISRE and NFkB activation) and with either M or EV as indicated. Cells were harvested 24h for luciferase assay. (D) HEK293T cells were transfected with secreted Gaussia luciferase and EV or M as indicated. 24h after transfection, medium was removed and replaced with fresh media. As a positive control cells were treated with 10 µM of Brefeldin A (BFA) or DMSO vehicle control for 6h as indicated. Both supernatant and cells were then harvested at 30h post-transfection for luciferase assay. (n = 3 ± SD or n = 4 ± SD (D), statistical significance was determined by one-way ANOVA with Fisher’s LSD test. ns – not significant; *, adjusted P < 0.05; ****, adj. P < 0.00001 relative to EV).
Fig 4.
SARS-CoV-2 M protein localizes to the cis-Golgi and disperses the structure of the trans-Golgi network.
ERGIC53-GFP-HEK293T cells were transfected with M or EV control and fixed 24h later (A-B). ACE2-A549 cells were infected with SARS-CoV-2 at an MOI of 1 then fixed at 24 h post-infection (C). Cells were immunostained as indicated with antibodies targeting SARS-CoV-2 M, GM130 (cis-Golgi), or TGOLN2 (trans-Golgi) and imaged by confocal microscopy. Maximum intensity projections are presented. 100X magnification, scale bar = 10 µm, Orange boxes indicate zoomed field of view. Representative images of three independent experiments.
Fig 5.
SARS-CoV-2 M protein restricts protein trafficking in the Golgi.
(A) Model of Retention Using Selective Hooks (RUSH) system. TNF-mApple with a strepavidin-binding peptide (SBP) is held in the ER by ER-retained streptavidin. Biotin treatment out-competes SBP-strepavidin interaction and allows TNF-mApple to continue anterograde traffic in the secretory pathway. Created in BioRender. Caddell, T. M. (2026) https://BioRender.com/3fvxi3v. (B) HEK293T cells were co-transfected with mApple TNF-a-RUSH with either M or EV control. Cells were then treated with 50 µM of biotin for 24h, as indicated. Cells were then fixed 24h post-transfection and stained with antibodies targeting SARS-CoV-2 M, and GM130 (cis-Golgi), and imaged by confocal microscopy. (C) Pearson’s coefficient measuring signal overlap between M and TNF-RUSH channels with or without biotin treatment. Maximum intensity projections are presented. 100X magnification, scale bar = 10 µm, Orange boxes indicate zoomed field of view. Representative images of three independent experiments. A two-tailed unpaired Welch’s t-test was performed, **** = adjusted p value <0.0001.
Fig 6.
SARS-CoV-2 M protein is present in detergent resistant membranes and recruits cholesterol.
(A) HEK293T cells were transfected with M or EV and DRMs were isolated as described in methods. The top fraction, containing the DRMs and the bottom fractions were used for western blot analysis as indicated. A representative image of n = 3 is shown. (B) HEK293T cells were co-transfected with cholesterol biosensor D4H-mCherry and pLVX-EGFP or pLVX-MEGFP. Cells were fixed at 24h and imaged without detergent permeabilization. (C) Pearson’s Correlation Coefficient measuring signal overlap between EGFP and D4H channels in pLVX-EGFP or pLVX-MEGFP conditions from panel (B). (D) HEK293T cells were transfected with pLVX-mCardinal or pLVX-M-mCardinal. At 4h post-transfection, medium was replaced with 10% FBS DMEM containing BODIPY 480/508-cholesterol. Cells were fixed at 24h and imaged without detergent permeabilization. (E) Pearson’s Correlation Coefficient measuring signal overlap between BODIPY-cholesterol and pLVX-mCardinal or pLVX-M-mCardinal channels from panel (D). Images were taken at 100X magnification, scale bar = 10 µm. (C)(E) Each data point isone cell evaluated for signal overlap across multiple fields of view and 3 independent biological replicates. A two-tailed unpaired Welch’s t-test was performed, **** = adjusted p value <0.0001. Representative images of three independent experiments.
Table 1.
Primers used for cloning.