Cell study reveals TRAP complex as a shared coronavirus dependency in the ER

Coronaviruses are heavy users of the endoplasmic reticulum. They bend its membranes into double-membrane vesicles and related replication organelles, and at the same time they pump out an unusually large quantity of glycosylated membrane proteins through the very same organelle. A study published in EMBO Reports argues that a long-appreciated but little-examined component of the ER protein translocation machinery, the translocon-associated protein (TRAP) complex, sits at the junction of these two demands and is required for productive infection by several divergent coronaviruses.

The work, led by Yihan Tian, Junchao Shi and Zi Li, used porcine hemagglutinating encephalomyelitis virus (PHEV), an embecovirus that infects neuronal cells, as its main model. In a time-course analysis of PHEV-infected N2a neuroblastoma cells, the researchers charted how infection reorganizes the ER. Early after infection, double-stranded RNA replication intermediates accumulated around the nucleus. By roughly 24 hours post-infection, transmission electron microscopy showed virus-induced replication-organelle-like structures enclosed by ER membranes, marked by cisternal dilation, abundant double-membrane vesicles and crescent-shaped phagophore-like membranes. By the late phase, multivesicular body-like viroplasm foci, immature and mature virions with corona-like surface spikes, and markedly dilated ER cisternae entrapping swollen mitochondria were visible, a picture the authors interpret as severe ER stress and enhanced ER–mitochondria crosstalk.

To find host factors that PHEV cannot replicate without, the team ran a genome-wide CRISPR-Cas9 loss-of-function screen in N2a cells. Among the high-confidence candidates, genes involved in ER processes such as ER translocation and ER-associated degradation were prominently enriched. One stood out: Ssr3, the gene encoding TRAPγ, a core subunit of the heterotetrameric TRAP complex that sits beside the Sec61 translocon and helps direct selected nascent proteins into the ER. Knockout of Tmem117 and Ssr3 suppressed PHEV replication in validation assays, whereas knockouts of several other candidates, including A4galt and Adam34, did not.

The consequences of losing TRAPγ were broad. Single-cell-derived TRAPγ knockout clones grew normally and, under basal conditions, showed no sustained ER stress, with unchanged levels of the unfolded protein response markers CHOP, PDI and GRP94. After PHEV infection, however, these cells produced far less viral mRNA and nucleocapsid protein, harbored fewer double-stranded RNA-positive cells, and released strongly reduced infectious virus as measured by plaque and TCID50 assays. The availability of the PHEV receptor DPEP1 was not reduced in the knockout cells, arguing that the defect lies downstream of viral entry. Immunoblot analysis also showed that depleting TRAPγ or TRAPβ destabilized other TRAP subunits, consistent with the complex’s multi-subunit architecture.

Ultrastructure told a parallel story. In infected wild-type cells, electron microscopy revealed abundant ER-derived replication compartments, including double-membrane vesicles and multivesicular bodies loaded with virions. In infected TRAPγ knockout cells, virions were scarce and ER-derived vesicular compartments markedly reduced, replaced by atypical whorl-like structures and highly convoluted ER membranes. When the authors expressed the viral membrane-remodeling proteins nsp3 and nsp4 directly in both cell types, typical double-membrane vesicle-like structures still formed, though less frequently in the knockout background. TRAPγ is therefore not strictly required for nsp3/nsp4-driven membrane bending, but the authors conclude it supports efficient replication organelle formation, most likely by enabling the biogenesis of the viral transmembrane proteins that organize these compartments.

The study then mapped how viral proteins engage the translocation machinery. Immunoprecipitation–mass spectrometry using PHEV nsp3, nsp4 and the nsp3–nsp4 complex as baits showed that nsp4 in particular was biased toward co-translational translocation and early glycoprotein biogenesis factors, strongly enriching the oligosaccharyltransferase catalytic subunit STT3A, the SRP receptor subunit SRPRB and multiple TRAP subunits. Reanalysis of a published SARS-CoV-2 host–protein interaction map independently flagged TRAPα, TRAPγ and TRAPδ as nsp4-associated factors. Co-immunoprecipitation revealed a selective, cargo-specific pattern: PHEV nsp4 preferentially bound TRAPα, TRAPβ and TRAPδ; the hemagglutinin-esterase (HE) protein bound mainly TRAPδ; and the spike S1 domain interacted most strongly with TRAPγ and TRAPδ. Microscale thermophoresis quantified these interactions, with dissociation constants of 15.8 nanomolar for nsp4–TRAPα, 12.2 nanomolar for S1–TRAPγ and a more moderate 865 nanomolar for HE–TRAPδ. Viral proteins lacking signal sequences did not detectably bind TRAP.

Why would coronaviruses need TRAP at all? The authors point to the physical character of viral signal sequences. TRAP is thought to favor substrates whose signal peptides have high glycine-plus-proline content and relatively low hydrophobicity. When the team introduced leucine-to-glutamine substitutions into the hydrophobic cores of the nsp4 and HE signal sequences, lowering their computed hydrophobicity, protein abundance dropped sharply while mRNA levels were unchanged. Swapping the glycine-proline-rich signal peptide of nsp4 for one lacking these residues nearly abolished nsp4 protein accumulation. The proteasome inhibitor MG132 failed to restore expression of the mutants, and the authors argue the loss is unlikely to reflect post-translational degradation, instead pointing to early co-translational quality control such as ribosome pausing and nascent-chain turnover. Consistent with substrate selectivity, TRAP depletion lowered the ER-resident host protein SEL1L while leaving the non-ER-targeted Japanese encephalitis virus NS3 protein untouched.

The TRAP step also gates the next one. Co-translational translocation into the ER is coupled to N-glycosylation by the oligosaccharyltransferase (OST) complex, and coronavirus spike and HE proteins are extensively glycosylated. Treating cells with NGI-1, a selective inhibitor of the OST catalytic subunits STT3A and STT3B, markedly reduced PHEV mRNA and nucleocapsid protein in a dose-dependent manner. Conversely, loss of TRAPγ caused pronounced translocation defects in nsp4, HE and spike while OST catalytic capacity remained intact, placing TRAP upstream of glycan installation. Extending the analysis beyond PHEV, TRAPγ depletion consistently diminished spike protein abundance for SARS-CoV, SARS-CoV-2, MERS-CoV and porcine deltacoronavirus, supporting the authors’ model of a sequential SRP–TRAP–OST axis that coronaviruses across distant lineages exploit to couple glycoprotein production with ER remodeling.

One further observation complicates the picture in an interesting way. Because double-membrane vesicles are thought to shield viral RNA from cytosolic sensors, the defective organelle biogenesis in TRAPγ knockout cells might be expected to unleash innate immune signaling. Infected knockout cells did show modestly increased expression of Ifih1 (MDA5), Ddx58 (RIG-I), Isg15 and Irf7. The authors caution, however, that TRAPγ loss also causes a severe collapse of viral gene expression, so the altered immune signature may reflect impaired sequestration of replication intermediates, reduced accumulation of viral immune antagonists, or both, rather than a simple unmasking effect.

The authors are explicit about the limits of their work. The mechanistic insights derive largely from in vitro cell culture models, chiefly N2a and HEK293T cells, and the central infection experiments rely on a single PHEV strain in a neuroblastoma line. They note that validating the role of TRAPγ in vivo and testing whether OST inhibitors such as NGI-1 show therapeutic efficacy in animal models will be critical next steps, alongside structural work to resolve how TRAP recognizes viral glycoproteins. Within those limits, the study positions the ER co-translational translocation machinery, and the TRAP–OST axis in particular, as a conserved host–pathogen interface and a candidate point of attack for future broad-spectrum, host-directed antiviral strategies.

Subject of Research: Role of the ER TRAP complex in coronavirus replication and glycoprotein biogenesis

Article Title: TRAP complex is essential for coronavirus replication and glycoprotein biogeneses

Article References: Tian, Y., Shi, J., Mu, S., Wei, H., He, W., & Li, Z. (2026). TRAP complex is essential for coronavirus replication and glycoprotein biogeneses. EMBO Reports. https://doi.org/10.1038/s44319-026-00986-z

Image Credits: AI Generated

DOI: 10.1038/s44319-026-00986-z

Keywords: coronavirus, TRAP complex, endoplasmic reticulum, CRISPR screen, glycoprotein, N-glycosylation, PHEV, replication organelles, SRP, oligosaccharyltransferase, host-directed antiviral, nsp4