The REEP5/TRAM1 complex binds SARS-CoV-2 NSP3 and promotes virus replication

ABSTRACT Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), like other coronaviruses, replicates their genome in virus-induced cytosolic membrane-bound replication organelles (ROs). SARS-CoV-2 promotes the biogenesis of ROs by inducing the rearrangement of endoplasmic reticulum (ER) membranes. NSP3, NSP4, and NSP6 are transmembrane viral non-structural proteins (NSPs) and essential players in the formation of ROs. To understand how these three NSPs work synergistically with host-binding proteins, we performed affinity purifications followed by mass spectrometry analyses to study the host-viral protein-protein interactome of NSP3, NSP4, and NSP6 expressed individually and in combination. Through this analysis, we identified two host transmembrane proteins, REEP5 and TRAM1, as critical interacting partners of NSP3 that localize at the membrane of the RO. REEP5 interacts with TRAM1 endogenously and binds NSP3 during SARS-CoV-2 infection. REEP5 knockout reduces ER membrane rearrangements and inhibits SARS-CoV-2 replication. Collectively, our study shows that the host REEP5/TRAM1 complex binds NSP3, promoting RO biogenesis and viral replication. IMPORTANCE Generation of virus-host protein–protein interactions (PPIs) maps may provide clues to uncover SARS-CoV-2-hijacked cellular processes. However, these PPIs maps were created by expressing each viral protein singularly, which does not reflect the life situation in which certain viral proteins synergistically interact with host proteins. Our results reveal the host-viral protein-protein interactome of SARS-CoV-2 NSP3, NSP4, and NSP6 expressed individually or in combination. Furthermore, REEP5/TRAM1 complex interacts with NSP3 at ROs and promotes viral replication. The significance of our research is identifying virus-host interactions that may be targeted for therapeutic intervention.

C oronaviruses (CoVs) replicate their genomes in the cytoplasm of host cells (1, 2).This process is supported by virus-induced rearrangement of host endoplasmic reticulum (ER) membranes that generate what is known as the replication organelle (RO) (3,4).The most abundant components of ROs for CoVs are double-membrane vesicles (DMVs), which are central hubs for viral RNA synthesis (4)(5)(6).Viral replicase complexes are found in DMVs and are required for replication of viral genome and translation of structural proteins (6).The concerted actions of viral-host protein-protein interactions (PPIs) are crucial for the generation of these replication platforms by hijacking various host cellular pathways involved in membrane-shaping and transportation.
CoVs have a large genome that encodes 4 structural proteins and 16 non-structural proteins (NSPs) that, together, ensure virus replication in host cells (6).The four structural proteins are spike (S), nucleocapsid (N), membrane (M), and envelope (E) proteins.The 16 NSPs include 13 cytosolic proteins and 3 transmembrane proteins, NSP3, NSP4, and NSP6.NSP3, NSP4, and NSP6 bind each other at the surface of DMVs and are crucial for the generation of the ROs (7)(8)(9).Co-expression of NSP3 and NSP4 is required and sufficient to induce the formation of these DMVs in human cells (8,10), while NSP6 contributes to the established connection between ER membranes and DMVs (11).
Accumulated evidence supports the idea that viral RNA synthesis occurs inside DMVs (4,5,12), as it provides a dual advantage for the virus by (i) spatio-temporally optimizing the organization of cellular and viral constituents required for RNA synthesis and (ii) preventing attacks from the host anti-viral defense system.A recent study (12) visualized a molecular pore complex (~1.8 MDa intermembrane platform) that spans both membranes of the DMVs, suggesting that newly synthesized viral RNAs can travel from the lumen of DMVs to the cytosol.Moreover, the coronavirus transmembrane protein NSP3 has been validated as a component of the pore complex (12).Another study revealed that SARS-CoV-2 NSP3 and NSP4 are minimal components forming a DMV spanning pore and showed that NSP3 Ubl1-Ubl2 domains are critical for induc ing membrane curvature and DMV formation (13).As proteins interacting with NSP3 (including NSP4 and NSP6) are likely part of the pore complex, identifying the host interactome of NSP3/NSP4/NSP6 proteins at DMVs is expected to shed light on the regulatory mechanism of viral RNA synthesis.
Virus-host PPIs are the vital engine of the viral life cycle after virus entry in host cells (14).Several recent studies have explored SARS-CoV-2-host PPIs by affinity purificationmass spectrometry (MS) and proximity-based labeling MS method (15)(16)(17)(18).Generation of virus-host PPIs maps may provide clues to uncover SARS-CoV-2-hijacked cellular processes.However, these PPIs maps were created by expressing each viral protein singularly, which does not reflect the life situation in which certain viral proteins synergistically interact with host proteins.
In this study, we performed affinity purification followed by mass spectrometry analysis (AP-MS) to study the host-viral protein-protein interactome of SARS-CoV-2 NSP3, NSP4, and NSP6 expressed individually or in combination.We identified the REEP5/ TRAM1 complex as host proteins binding NSP3 at the ROs.This study reveals a previously unknown function of the REEP5/TRAM1 complex to regulate SARS-CoV-2 RO biogenesis and replication.

Host interactome of NSP3, NSP4, and NSP6
First, we expressed SARS-CoV-2 NSP3, NSP4, and NSP6 proteins in mammalian cells.The open reading frames (Orfs) of NSP3, NSP4, and NSP6 were codon opti mized with two different tools: one from Fritz Roth (FR) (19) and the other from online Rare Codon Analyzer (RCA; https://www.biologicscorp.com/tools/RareCodonAnalyzer/#.Y_6vR-zMJTZ).The optimized Orfs were fused to a 2xStrep affinity tag and cloned into a mammalian expression vector.To verify viral protein expression, we performed Western Blots with an anti-Strep antibody on whole cell extracts (WCE) from HEK293T cells transfected with the SARS-CoV-2 Orfs (Fig. S1A).The Orfs from FR displayed higher expression than the Orfs from RCA, so we used the Orfs from FR for further studies.We observed that the expression level of full-length SARS-CoV-2 NSP3 is much lower compared to NSP4, possibly due to the complex protein topology of the full-length NSP3, while the expression level of the C-terminal one-third of NSP3 (NSP3C) is comparable to NSP4 (Fig. S1A; Fig. 5D).
Next, we aimed at confirming the functionality of NSP3, NSP4, and NSP6 in mamma lian cells.Since SARS-CoV and mouse hepatitis virus (MHV) share many similarities in virology and epidemiology with SARS-CoV-2, mechanistic insights learnt from SARS-CoV and MHV could offer clues into SARS-CoV-2 biology (20,21).For both SARS-CoV and MHV, co-expression of NSP3C with NSP4 is enough to induce ER membrane rearrange ments in mammalian cells (22,23).To verify this observation with SARS-CoV-2, we co-expressed the corresponding EGFP-tagged NSP3C and mCherry-tagged NSP4 from SARS-CoV-2 together with an mTagBFP2-tagged ER marker in U-2 OS cells, which have relatively large cytoplasmic volume.Using correlative light and electron microscopy (CLEM), we confirmed the co-localization of NSP3C and NSP4 with the ER marker.Fluorescent images showed that NSP3C and NSP4 fully colocalized in ring-shaped structures, which at the EM level are corresponded to multi-membrane vesicles (MMVs) with an average size of 1-3 μm (Fig. S1B), possibly generated from the fusion of small-sized DMVs with a diameter of 150-350 nm as suggested previously (3,4).To confirm that NSP3C binds NSP4 and NSP6, we performed affinity purification with biotin magnetic beads (AP-Strep) followed by WB in HEK293T cells co-transfected with NSP3C-2xStrep and Flag-tagged NSP4 (or its mutants), NSP6, other NSPs (NSP2 and NSP14), or host transmembrane proteins (V0A1 and V0D).We observed a specific binding of NSP3C to NSP4 and NSP6, but not to NSP2, NSP14, V0A1, and V0D (Fig. S1C).In agreement with a previous study (23), two-amino acid substitutions in NSP4 (NSP4-H120N/F121L), but not the deletion of amino acids 220-234 (NSP4-Δ220-234), reduced its binding to NSP3C compared to NSP4 wild type.Although HEK293T cells are isolated from kidney tissue, which is not a primary tissue target of SARS-CoV-2, previous work has identified host-binding factors of CoV proteins using HEK293T cells (15)(16)(17)(18).Based on all of the above, we decided to express the NSP proteins in HEK293T cells to identify their host-binding proteins by AP-MS.
To identify the specific host-binding proteins for each of the NSPs, we performed AP-Strep in HEK293T cells expressing 2Strep-tagged NSP3, NSP4, and NSP6 individually.The co-purified proteins from three AP-Strep biological replicates for each of the NSPs were analyzed by MS.Our AP-MS analysis identified 106 high-confidence PPIs between NSPs and host-binding proteins (Table S1).We define the high-confidence PPIs by (1) filtering out proteins with peptide-spectrum matches (PSMs) in any of the empty vector triplicates; (2) removing proteins with less than five PSMs on average per sample; (3) selecting proteins with a t-test P-value less than 0.01 as potential interactors of a given NSP AP-MS compared to the other two NSPs AP-MS.For host proteins binding more than one NSP, the graph only shows their interaction to the NSP for which it had the highest specificity score (Fig. 1).The top three most specific binding proteins for each of the NSP were highlighted with orange circles (NSP3C: REEP5, ERLEC1, and GOLT1B; NSP4: DNAJC3, SLC27A3, and UBR2; NSP6: TRAFD1, HUWE1, and ATP13A3).We analyzed each NSP for Gene Ontology enrichment with the STRING website (Table S1) and found that the representative cell process of the interacting proteins for NSP3C is intracellular protein transport (Fig. 1), in agreement with the essential role of NSP3 in RO biogenesis.
Considering that viral proteins might combinationally interact with host proteins, we performed AP-MS in HEK293T cells expressing NSPs in different pairs (Table S1).We analyzed the AP-MS data of the pair of NSP3C and NSP4 (NSP3+4) with co-expression of NSP3C-2xStrep + NSP4 Flag, as well as NSP4-2xStrep + NSP3C-Flag; the pair of NSP4 and NSP6 (NSP4 +6) with co-expression of NSP4-2xStrep + NSP6 Flag, as well as NSP6-2xStrep + NSP4 Flag; the pair of NSP3C and NSP6 (NSP3 +6) with co-expression of NSP3C-2xStrep + NSP6 Flag, as well as NSP6-2Strep + NSP3C-Flag (Fig. 2).In total, we identified 135 highconfidence PPIs between each pair of NSPs and host-binding proteins, with 44 proteins (labeled with an asterisk) identified also in the experiments with individually expressed NSPs (Fig. 1).These results suggest that the study of NSP pairs is helpful to identify hostbinding proteins compared to the analysis of individual NSPs.The top three most specific binding proteins for each of pair of NSPs were highlighted with orange circles (NSP3 + 4: REEP5, TMEM106B, and IDE; NSP4 + 6: PLAA, SGTA, and UBQLN4; NSP3 + 6: REEP6, ATP13A3, and SIGMAR1).After analyzing the host-binding proteins for Gene Ontology enrichment with the STRING website, we found that the representative cell processes of the interacting proteins for NSP3+4 are ER unfolded protein response and protein nlinked glycosylation via asparagine; for NSP4+6 is proteasomal protein catabolic process; for NSP3+6 is transport (Fig. 2).
To pinpoint which host-binding proteins localize at ROs, we expressed these 11 validated host-binding proteins with an EGFP tag in U-2 OS cells expressing also Flagtagged NSP3C, mCherry-tagged NSP4, and the mTagBFP2-tagged ER marker.Fluorescent images show that REEP5, TRAM1, and STIM1 co-localized at ROs with the ER marker, NSP3C, and NSP4 (Fig. 4A; Fig. S2).Moreover, we confirmed that the localization of REEP5 at ROs with CLEM (Fig. 4B).
All of these data indicate the REEP5/TRAM1 complex as a bona fide host protein complex interacting with SARS-CoV-2 NSP3 at ROs.

The REEP5/TRAM1 complex promotes ER membrane rearrangements and SARS-CoV-2 replication
Since both REEP5 and TRAM1 localize at the ER and are important for ER membrane organization (24,25), we asked whether the REEP5/TRAM1 complex regulates ER membrane rearrangements induced by co-expression of NSP3C and NSP4.U-2 OS cells were infected with lentivirus carrying either Cas9 and single guide RNAs (sgRNAs) against REEP5 or TRAM1 or a non-targeting control sgRNA.REEP5 knockout (KO) or (F) RT-qPCR analysis of SARS-CoV-2 intracellular RNA expression WT and REEP5 KO Calu-3 cells upon 0, 6, and 24 hours after infection with SARS-CoV-2 (MOI = 0.1).For (E) and (F), N = 6 wells for each cell lines TRAM1 KO U-2 OS cells were validated with immunoblot (Fig. 6A) and DNA sequencing (Fig. S3).In these cell lines, we co-expressed the EGFP-tagged NSP3C and mCherrytagged NSP4 together with an mTagBFP2-tagged ER marker.Fluorescent images showed that the number of MMVs induced by NSP3C and NSP4 in REEP5 KO or TRAM1 KO cells is significantly lower compared to the number in parental cells (Fig. S6B and C).As ER membrane rearrangements are an essential step for the formation of MMVs, these data suggest that REEP5/TRAM1 complex plays a crucial role in SARS-CoV-2 RO biogenesis.
To further study the role of the REEP5/TRAM1 complex on SARS-CoV-2 replication, we used Calu-3 cells as a model because Calu-3 cells endogenously express the angiotensinconverting enzyme 2 (ACE2) receptor ( 26), a major cell entry receptor for SARS-CoV-2 (27).We infected Calu-3 cells with lentivirus carrying either Cas9 and a non-targeting control sgRNA, or guide RNAs (sgRNAs) against REEP5 or TRAM1.As the TRAM1 KO Calu-3 cells stopped dividing after infection, probably due to virus-induced cellular senescence, we validated REEP5 KO Calu-3 cells with immunoblot (Fig. 6D).We then infected the REEP5 KO and parental Calu-3 cells with SARS-CoV-2 for 24 hours and quantified plaque-forming units (PFU) by virus titration of supernatant from parental and REEP5 KO Calu-3 cells.The number of PFU in REEP5 KO is approximately 100 times lower compared to the number in parental cells (Fig. 6E).To measure viral RNA replication, we performed RT-qPCR analysis of SARS-CoV-2 RNA expression in REEP5 KO and WT Calu-3 cells at 0, 6, and 24 hours after infection with SARS-CoV-2.Depletion of REEP5 in Calu-3 cells also inhibited intracellular viral RNA level after SARS-CoV-2 infection for both 6 and 24 hours (Fig. 6F).

REEP5 depletion reduces ER profile length and number of SARS-CoV-2induced DMVs
To understand the role of REEP5 in ER morphology and SARS-CoV-2-induced DMVs, we infected parental and REEP5 KO Calu-3 cells with SARS-CoV-2 and compared their morphology at 0 (Fig. 7A and B) and 24 hours (Fig. 7C and D) after infection using electron microscopy.Consistent with a previous study (24), we observed the reduction of ER profiles length in REEP5 KO cells (Fig. 7A and B), suggesting instability in ER's structure and function, which could lead to unfolded protein response, ER stress, ER-associated degradation (28), and autophagy (29).Similar to previous reports about virus-induced DMVs (3,30), we observed DMVs with an average diameter of 150-350 nm, which were often clustered together in both parental and REEP5 KO Caul-3 cells at 24 hours after infection (Fig. 7C).Moreover, both the number (Fig. 7D) and size (Table 1) of SARS-CoV-2induced DMVs were significantly decreased in REEP5 KO cells.

DISCUSSION
Our work identifies a previously unknown REEP5/TRAM1 protein complex and its role on SARS-CoV-2 replication.Specifically, we provide a comprehensive host-viral proteinprotein interactome for NSP3, NSP4, and NSP6 both individually and in combination.From the list of host-binding proteins, we found that two ER transmembrane proteins, REEP5 and TRAM1, bind each other, as well as NSP3 during SARS-CoV-2 infection.REEP5/ TRAM1 complex colocalize with NSP3 and NSP4 at ROs and promote ER membrane rearrangement, thus promoting SARS-CoV-2 replication.
Despite accumulating studies on SARS-CoV-2 replication in recent years, knowledge on the critical host proteins involved in DMV biology is still relatively limited (9).We confirmed that endogenous REEP5 and TRAM1 bind NSP3 derived from SARS-CoV-2 infection.REEP5 belongs to a family of membrane curvature-stabilizing proteins, which contain a reticulon-homology domain that is thought to be essential for promoting and stabilizing curvature in ER tubules (31,32).Consistent with a previous study (24), we observed the reduction of ER profiles length in REEP5 KO cells, suggesting instability in ER's structure and function, which could induce unfolded protein response, ER stress, ER-associated degradation (28), and autophagy (29).As ER functions are exploited by SARS-CoV-2 to support distinct stages of their life cycle (33), it is possible that NSP3 binds REEP5 and disrupts its normal function, thus promoting ER membrane rearrange ment and RO biogenesis.TRAM1 is instead an eight-transmembrane domain ER protein that is supposed to bind ceramide or related sphingolipids (34).TRAM1 was originally discovered as a component of the mammalian ER involved in the translocation of secretory proteins (35).Furthermore, TRAM1 was found as a protein interacting with some nascent membrane proteins during their initial integration into the Sec61-channel (36,37).Since the Sec61-channel could translocate one-third of all polypeptides into or through the ER-membrane, TRAM1 may modulate the phospholipid bilayer near the lateral gate of the Sec61-channel to support ER protein translocation (38,39).Although further validation is clearly required, TRAM1 is a potential host component of the pore complex at DMVs, suggesting that synthesized proteins required for viral RNA replication may be transported inside the DMVs via the pore complex.Besides REEP5 and TRAM1, we confirmed eight other host-binding proteins, including TMEM106B, REEP6, DNAJC11, FKBP10, ATG9A, SGTA, XPO6, and STIM1.Intriguingly, in addition to be an NSP3 binding protein (Fig. 3A), TMEM106B has been recently repor ted to be an alternative receptor for SARS-CoV-2 entry into ACE2-negative cells (40).STIM1 colocalizes with NSP3C and NSP4 at ROs (Fig. S4A), and it is known to promote SARS-CoV-2 infection by decreasing type I interferon response (41).These findings may provide clues to study the mechanism of SARS-CoV-2 infection and immune evasion.Interestingly, the ER-resident proteins VMP1 and TMEM41B were identified as host proteins required for SARS-CoV-2 infection by genetic screens (42).A follow-up study showed that VMP1 and TMEM41B are involved in regulating DMV formation (10).While both NSP3 and NSP4 have the ability to bind VMP1, TMEM41B has a weak binding to NSP4, but not NSP3 (10).It is important to note that neither VMP1 nor TMEM41B was identified as high-confidence binding proteins of NSPs in our proteomics study (Table S1).This discrepancy could be because previous studies were conducted with overexpression of both "bait" and "prey" proteins as opposed to identifying endogenous binding partners.
The membrane phenotype induced by co-expression of NSP3C and NSP4 proteins does not fully resemble the DMVs biogenesis of SARS-CoV-2 infection.Besides lacking viral RNA synthesis, the abundance and morphology of these ROs induced by viral proteins are different from that of native DMVs during virus infection (9).This indicates that additional viral factors may play crucial roles on DMVs biogenesis.Gaining insight into this process from a molecular perspective will be essential to understand SARS-CoV-2 life cycle.

Reagents
Sources of chemicals are found in the Key Reagents Table.

Antibodies
The dilutions and sources of antibodies used for immunoblot (IB) and immunoprecipita tion (IP) in this study can be found in the Key Resources Table .All antibodies used were validated following the multiple dilution method and, where available, using cell lines or tissues from animals knock-out for the antigen.

Protein electrophoresis and immunoblotting
Protein concentration was determined using the Lowry method (43) with bovine serum albumin as the standard.Immunoblotting was performed after transferring SDS-PAGE gels to nitrocellulose membrane and blocking with 5% milk in 0.01% Tween-TBS for 1 hour at room temperature.The proteins of interest were visualized after incuba tion with primaries by chemiluminescence using horseradish peroxidase-conjugated secondary antibodies in the SRX-101A Tabletop X-Ray Film Processor (Konica).

Affinity purification
HEK293T cells were transiently transfected with DNA using polyethylenimine (Polyscien ces).After transfection for 24 hours, cell lysis was carried out with lysis buffer (50 mM Tris pH 7.4, 150 mM NaCl, 10% glycerol, 0.3% Triton-X-100, and 0.1% NP-40) supplemented with protease and phosphatase inhibitors.Lysates were then immunoprecipitated with anti-FLAG antibody conjugated to agarose.After washing with lysis buffer for four times, elution was carried out with 3X FLAG peptide.For endogenous IP, lysates were incubated with anti-REEP5 or anti-TRAM1 antibody and rotated for 3 hours at 4°C.Then protein G beads were added and incubated for 1 hour at 4°C.After washing with lysis buffer for four times, the beads were denatured with 1X LDS for 3 min at 95°C.For affinity purification with MagStrep "type3" XT beads, elution was performed with Strep-Tactin XT elution buffer.For denaturing IP, cells were lysed with 2% SDS and denatured for 5 min at 95°C, then the lysates were diluted 1:20 to perform affinity purification.

Fluorescent microscopy
Cells were plated on No.1.5coverslip or glass bottom dish.Discard cell medium and add 2 mL of prewarmed fixative containing 2% paraformaldehyde and 0.1% glutaraldehyde in PBS for 15 min at room temperature; wash with PBS, 3 × 5 min each.To eliminate unbound aldehydes, cells were incubated in 50 mM glycine (37.5 mg in 10 mL) in PBS for 5 min (RT).Cells were permeabilized by 0.1% Triton X-100 in PBS for 10 min (RT); blocking with blocking solution (1% bovine serum albumin in PBS) for 60 min at RT on the shaker.Incubate with antibody in primary antibody incubation buffer (1% BSA in PBS) for 2 hours at RT on the shaker.Wash with PBS, 3 × 10 min, at RT on the shaker.Incubate with secondary antibody in dark in antibody incubation buffer (1% BSA in PBS) for 30 min to 1 hour, at RT on the shaker; wash with PBS, 3 × 10 min; wash/incubate with PBS/4' ,6-diamidino-2-phenylindole DAPI; wash with PBS, twice; For imaging, on the day of data collection, cells were incubated in PBS.Imaging was performed using Zeiss AxioObservor microscope.

Correlative light and electron microscopy
For morphological analysis of autophagic vesicles, cultured cells were fixed in 0.1M sodium cacodylate buffer (pH 7.4) containing 2.5% glutaraldehyde and 2% paraformal dehyde overnight at 4°C and post-fixed with 1% osmium tetroxide mixed with 1% potassium ferrocyanide for 1 hour at 4°C, then block stained in 0.25% aqueous uranyl acetate overnight at 4°C, processed in a standard manner and embedded in EMbed 812 (Electron Microscopy Sciences, Hatfield, PA).Ultrathin sections (70 nm) were cut and mounted on 200 mesh copper grids.Quantitative analyses of ER profiles length were performed with ImageJ software using Feret's Diameter (44).
For CLEM, cells were plated on gridded glass-bottom dishes (P35G-1.5-14-CGRD,MatTek) and fixed with 4% paraformaldehyde in 0.1 M sodium phosphate buffer (PBS) for 30 min at room temperature, then change to 2% paraformaldehyde in PBS and stored at 4°C overnight.Fluorescent and phase contrast images were taken at the areas of interest using Zeiss AxioObservor microscope.After light microscopy imaging, the cells are continue fixed with 2.5% glutaraldehyde for 1 hour and post fixed with 1% osmium tetroxide for 1 hour at room temperature.The cells were then block stained with 1% uranyl acetate for 1 hour, dehydrated in ethanol, and en face embedded in Araldite 502 (Electron Microscopy Sciences, Hatfield, PA).En face serial thin sections with 80 nm were cut and mounted on formvar coated slot copper grids.
All EM grids were stained with uranyl acetate and lead citrate by standard methods and examined under either Philips CM-12 electron microscope (FEI; Eindhoven, The Netherlands) and photographed with a Gatan (4k x2.7k) digital camera, or Talos L120C electron microscope (Thermo Fisher Scientific, Hillsboro, OR) coupled with Gatan 4k × 4k OneView Camera (Gatan Inc.Pleasanton, CA).

Mass spectrometry
Samples were reduced with DTT at 57°C for 1 hour (2 µL of 0.2 M).Samples were then alkylated with iodoacetamide at RT in the dark for 45 min (2 µL of 0.5 M) and loaded onto NuPAGE 4%-12% Bis-Tris Gel 1.0 mM (Life Technologies Corporation) and ran for approximately 2 min at 200 V.The gel was stained using GelCode Blue Stain Reagent (Thermo Scientific), and Coomassie-stained gel bands were excised as indicated on the gel image.Excised gel pieces were destained in 1:1 v/v solution of Methanol and 100 mM Ammonium Bicarbonate solution.The gel pieces were partially dehydrated with an acetonitrile rinse and further dried in a SpeedVac concentrator for 20 min.200 ng of sequencing grade modified trypsin (Promega) was added to each gel sample.After the trypsin was absorbed, 250 µL of 100 mM ammonium bicarbonate was added to cover the gel pieces.Digestion proceeded overnight on a shaker at RT.A slurry of R2 20 µM Poros beads (Life Technologies Corporation) in 5% formic acid and 0.2% trifluoroacetic acid (TFA) was added to each sample at a volume equal to that of the ammonium bicarbonate added for digestion.The samples shook at 4°C for 3 hour.The beads were loaded onto equilibrated C18 ziptips (Millipore) using a microcentrifuge for 30 seconds at 6,000 rpm.Gel pieces were rinsed three times with 0.1% TFA, and each rinse was added to its corresponding ziptip followed by microcentrifugation.The extracted beads were further washed with 0.5% acetic acid.Peptides were eluted by the addition of 40% acetonitrile in 0.5% acetic acid followed by the addition of 80% acetonitrile in 0.5% acetic acid.The organic solvent was removed using a SpeedVac concentrator and the sample reconstitu ted in 0.5% acetic acid.Sample was analyzed individually using LC separation online with MS using the autosampler of an EASY-nLC 1000 (Thermo Scientific).Peptides were gradient eluted from the column directly to a Orbitrap Eclipse mass spectrometer using a 1 hour gradient (Thermo Scientific) Solvent A: 2% acetonitrile, 0.5% acetic acid; Solvent B: 90% acetonitrile, 0.5% acetic acid.High-resolution full MS spectra were acquired with a resolution of 240,000, an AGC target of 1e 6 , with a maximum ion time of 50 ms, and scan range of 400-1,500 m/z.Following each full MS, data-dependent low-resolution ion trap HCD MS/MS spectra were acquired.All MS/MS spectra were collected using the following instrument parameters: ion trap rapid scan, AGC target of 2e 4 , maximum ion time of 18 ms, one microscan, 0.7 m/z isolation window, 20 seconds dynamic exclusion, fixed first mass of 150 m/z, and NCE of 27.Singly charged ions and ions carrying eight or more charges were excluded from triggering an MS/MS scan.The instrument was set to acquire a full MS scan every 3 seconds or earlier if no new MS/MS precursors were detected.
The MS/MS spectra were searched against a Uniprot (www.uniprot.org)human protein database with common lab contaminants and the sequence of the tagged bait proteins added using Sequest within Proteome Discoverer 1.4 (Thermo Fisher).The search parameters were as follows: mass accuracy better than 10 ppm for MS1 and 0.02 Da for MS2, two missed cleavages, fixed modification carbamidomethyl on cysteine, variable modification of oxidation on methionine, and deamidation on asparagine and glutamine.The data were filtered using a 1% FDR cut off for peptides and proteins against a decoy database and only proteins with at least two unique peptides were reported.
has financial interests in CullGen, SEED Therapeutics, Triana Biomedicines, and Umbria Therapeutics; however, no research funds were received from these entities, and the findings presented in this manuscript were not discussed with any person in these companies.MP also received research funds from Kymera Therapeutics, but the findings presented in this manuscript were not discussed with any person in this company.The rest of the authors declare no relationships with industry or any financial interests in relation to this work.All of the authors declare no competing interest.

FIG 1
FIG 1Host proteins binding SARS-CoV-2 NSP3C, NSP4, and NSP6 expressed individually.The graph shows 106 high-confidence host-binding proteins for SARS-CoV-2 NSP3C, NSP4, and NSP6 proteins (blue rectangles) individually.For host proteins binding more than one NSPs, we made graph to show their interaction to the NSP for which it had the highest enrichment.Enrichment analysis was done with the STRING website.Representative enrichments in the GO process category are shown in the plot (all enrichments were filtered at an false discovery rate (FDR) < 0.05 level).Proteins in the same biological process are categorized with teal and green ovals.Orange circles highlight the top three most specific host-binding proteins for each group.

FIG 2
FIG 2 Host proteins binding SARS-CoV-2 NSP3C, NSP4, and NSP6 expressed in combination.The graph shows 135 high-confidence interactions between a combination of SARS-CoV-2 NSP3C, NSP4, or NSP6 proteins (NSP3+4, NSP4+6, and NSP3+6; blue rectangles) and human proteins.Each host protein is connected to the NSP for which it had the highest specificity.Enrichment analysis was done with the STRING website.Representative enrichments in the GO process category are shown in the plot (all enrichments were filtered at an FDR < 0.05 level).Proteins in the same biological process are categorized with red, purple, light, or dark green ovals.Orange circles highlight the top three most specific host-binding proteins for each group.The proteins labeled with * are also specific binding proteins of SARS-CoV-2 NSP3C, NSP4, and NSP6 individually, as shown in Figure 1.

FIG 3 6 FIG 4 7 FIG 5 8 (FIG 6
FIG 3 Validation of the binding proteins of SARS-CoV-2 NSP3C, NSP4, and NSP6.Immunoblot of AP-Strep from HEK293T cells co-transfected with Flag-tagged NSP3C (A), NSP4 (B), or NSP6 (C) and 2Strep-tagged plasmids as indicated.Each bait protein was marked with a red star.Whole cell extract (WCE) controls are shown at the bottom.

FIG 6 (FIG 7
FIG 6 (Continued) from three independent experiments.Data are mean + SEM and individual values.Mann Whitney test was used.Differences were significant for ***P < 0.001.

TABLE 1
Size of DMVs in infected WT and REEP5 KO Calu-3 cells