ER functions are exploited by viruses to support distinct stages of their life cycle

Abstract

The endoplasmic reticulum (ER), with its expansive membranous system and a vast network of chaperones, enzymes, sensors, and ion channels, orchestrates diverse cellular functions, ranging from protein synthesis, folding, secretion, and degradation to lipid biogenesis and calcium homeostasis. Strikingly, some of the functions of the ER are exploited by viruses to promote their life cycles. During entry, viruses must penetrate a host membrane and reach an intracellular destination to express and replicate their genomes. These events lead to the assembly of new viral progenies that exit the host cell, thereby initiating further rounds of infection. In this review, we highlight how three distinct viruses — polyomavirus, flavivirus, and coronavirus — co-opt key functions of the ER to cause infection. We anticipate that illuminating this virus-ER interplay will provide rational therapeutic approaches to combat the virus-induced diseases.

Introduction

Endoplasmic reticulum

Arguably one of the most active intracellular organelles is the endoplasmic reticulum (ER). This is in part due to the dynamic physical property of the ER membrane, as well as the myriad of factors harbored within the ER. Structurally, the ER is composed of a continuous array of the membranous network including the outer nuclear membrane that is physically linked to ER tubules and sheets [1]. These sheets and tubules are sculpted by selective ER-resident proteins, often dependent on the specific cellular cues. Consequently, the overall ER morphology undergoes continual architectural rearrangements.

The key function of the ER is to synthesize transmembrane and soluble luminal proteins. In fact, the ER is thought to support the synthesis of approximately one-third of the entire cellular proteome. To initiate protein synthesis, nascent polypeptide chains are translated from the messenger RNA while simultaneously translocated across the ER membrane Sec61 translocon [2,3] — this ‘co-translational translocation’ mechanism is depicted in Figure 1A. In the case of secretory or ER-resident luminal proteins, they are released into the bulk ER lumen upon exiting the luminal pore of the translocon. In contrast, transmembrane proteins partition across the ‘lateral gate’ of this channel and integrate into the ER lipid bilayer. Regardless, once in the ER lumen or inserted into the ER membrane, the newly synthesized polypeptide chains experience post-translational modifications that assist in their folding and assembly.

Figure 1. Key functions of the endoplasmic reticulum (ER).

(A) The ER supports protein biosynthesis, in which luminal or membrane proteins are co-translationally translocated into the ER via the Sec61 translocon. The actions of ER-resident factors, including the oligosaccharyltransferase (OST), binding immunoglobulin protein (BiP), and protein disulfide isomerase (PDI), facilitate folding of these substrates. (B) Upon proper folding, the substrates are packaged into COPII-coated vesicles en route for transport via the classic secretory pathway. (C) However, when the substrates misfold, the ER mounts the UPR, which is an ER stress response that transiently suppresses global protein synthesis or activates the selective expression of ER-resident chaperones to assist the folding of proteins in the ER. (D) In contrast, terminally misfolded proteins are targeted to the cytosol for proteasomal degradation via the ER-associated degradation pathway (ERAD). (E) In parallel, for misfolded aggregated cargos, they are removed from the ER and directed to the lysosome for degradation via a selective autophagy pathway called ER-phagy. (F) The ER is also responsible for calcium homeostasis, as well as lipid biosynthesis. (G) The ER also establishes membrane contacts with other intracellular compartments such as the endosome and mitochondria. Functionally, these contact sites are thought to mediate lipid and ion exchange, as well as organelle biogenesis.

Upon proper folding and assembly, the polypeptide chains are packaged into distinct transport vesicles coated with COPII proteins [3] (Figure 1B). These coated vesicles then pinch off from the ER, ferrying the cargo to other cellular compartments or to the cell surface for secretion. However, despite the presence of powerful ER-resident chaperones (such as the BiP ATPase) that assist protein folding, a nascent polypeptide chain can misfold. To rectify this protein misfolding problem, the ER mounts the unfolded protein response (UPR) that activates ER membrane-bound sensors [4] (Figure 1C) — the activated sensors transiently suppress protein expression to dampen the overall load in the ER or increase expression of select ER-resident chaperones to further support protein folding in the ER.

If the UPR fails to repair the misfolded protein, the misfolded client is dislocated to the cytosol via a pathway called ER-associated degradation (ERAD) for ubiquitin-dependent proteasomal destruction [5,6] (Figure 1D); in ERAD, soluble misfolded ER proteins are threaded through the pore of the Hrd1-Derlin1 ER membrane channel to reach the cytosol. For misfolded clients that form high molecular-weight aggregated complexes that cannot fit the pore of the Hrd1-Derlin1 channel, they are instead removed from the ER by a recently described ER-coupled autophagy (ER-phagy) pathway [7,8] (Figure 1E); in ER-phagy, select regions of the ER containing the aggregated complex are fragmented by the phagophore membrane and delivered to the autophagosome for lysosomal degradation.

In addition to these protein quality control functions, the ER finely tunes the ER calcium levels by using various calcium channels and pumps that are positioned in the ER membrane [9] (Figure 1F). Additionally, by deploying the actions of enzymes localized at the ER membrane, this organelle is able to generate lipid molecules, including cholesterol, fatty acids, and phospholipids, that can serve a signaling function [10] (Figure 1G). Finally, there is emerging evidence that the ER forms membrane contact sites with other intracellular compartments such as the endosome and mitochondria to facilitate lipid and ion exchange, as well as organelle biogenesis (Figure 1H) [11].

Virus replication life cycle

Strikingly, these ER-associated functions can be exploited by viruses to support their infection life cycle [12,13]. During infection, a virus first gains entry into a host cell, often via receptor-mediated endocytosis. The viral particle then penetrates a host membrane and traffics to an intracellular destination that supports the transcription, translation, and replication of the viral genome. These steps result in the assembly of new viral progenies which, after maturation and morphogenesis, exit the host cell to initiate the next round of infection.

Importantly, the ER can be hijacked by viruses during many steps of the virus life cycle — membrane penetration, translation, replication, assembly, and egress — in order to achieve productive infection. Strikingly, the breadth of virus types that co-opt the ER during infection is quite extensive, ranging from DNA to RNA viruses, as well as spanning from non-enveloped to enveloped viruses. In this review, we will highlight how three different viruses, polyomavirus, flavivirus, and coronavirus, exploit the activities of the ER to cause infection because they are amongst the most well-characterized and relevant to current public health concerns. We envision that illuminating the virus-ER interplay will engender insights into novel therapeutic strategies to better combat the virus-induced diseases.

Polyomavirus uses the ER as a membrane penetration site

Polyomaviruses (PyVs) are responsible for causing many debilitating human diseases, particularly in individuals with a compromised immune system [14]. Notable human PyVs include the JC PyV that induces progressive multifocal leukoencephalopathy, the BK PyV that causes nephropathy and hemorrhagic cystitis, and the Merkel cell PyV which is the etiologic agent for Merkel cell carcinoma. The classic PyV simian virus 40 (SV40) is not only physically similar to human PyVs, but also uses a comparable infection pathway as its human counterparts [15]. For this reason, insights into the mechanism of SV40 entry have historically illuminated the cellular basis of human PyV infection.

When properly assembled into its native state, SV40 has a diameter of 45 nm. In this assembled state, SV40 is composed of 72 pentamers of the VP1 coat protein which is displayed on the viral surface [16–18]. The VP1 pentameric shell in turn encloses the 5 kb-pair DNA genome of the virus. Each VP1 pentamer also encases an internal protein called VP2 or VP3. To infect cells, SV40 is captured by the ganglioside GM1 glycolipid receptor at the host cell surface [19,20] — this interaction drives receptor-mediated endocytosis that delivers SV40 first to the endosomes (Figure 2) [21,22] and then the ER [23,24]. Importantly, the virus must exploit ER functions to penetrate the ER membrane in order to escape into the cytosol [25–29]. Because SV40 is a non-enveloped virus that lacks a surrounding lipid bilayer, its membrane penetration mechanism is likely fundamentally different from enveloped viruses — including flavivirus and coronavirus — which are ensheathed by a membrane bilayer. Upon reaching the cytosol, SV40 is finally mobilized into the nucleus where transcription and replication of the viral genome causes lytic infection or cellular transformation. Although a direct ER-to-nucleus penetration mechanism has been postulated [30], strong evidence for this pathway is lacking.

Figure 2. The ER serves as a membrane penetration site during polyomavirus entry.

Upon endocytosis, polyomavirus SV40 reaches the endosome. Two specific subunits of the ER membrane protein complex (EMC), EMC4 and EMC7, tether the ER to the endosome (step 1). This physical juxtaposition allows SV40 in the endosome to be efficiently delivered into the ER. In the ER, redox chaperones including PDI, ERp57, and ERdj5 reduce and isomerize disulfide bonds of SV40 (step 2), leading to exposure of the hydrophobic viral VP2 and VP3 proteins. This generates a hydrophobic particle that integrates into the ER membrane that primes the virus for ER escape into the cytosol (step 3). During ER-to-cytosol escape, SV40 creates a membrane penetration structure on the ER membrane called ‘focus’ (step 4); focus formation requires reorganization of select ER membrane proteins, including BAP31, RTN3/4, EMC1, and J proteins. The J proteins in turn recruit a cytosol extraction machinery composed of Hsc70, SGTA, and Hsp105/Bag2. This machinery, in conjunction with Ubiqln4, acts to extract the membrane-embedded SV40 into the cytosol (step 5). The cytosol-localized virus is disassembled by the BICD adaptors of the dynein motor complex, enabling the viral particle entry into the nucleus to cause infection.

The ER acts as the decisive gatekeeper for the entrance of SV40 into the cytosol. New studies have unveiled detailed mechanism of SV40 ER-to-cytosol membrane escape. SV40 reaches the ER from the endosome by relying on an ER membrane ‘tether’ composed of the EMC4 and EMC7 subunits of the ER membrane complex (EMC) (Figure 2, step 1) [24]. Specifically, EMC4 and EMC7 bind to the late endosome-associated protein Rab7, thereby physically linking the endosome to the ER (Figure 1H) — this juxtaposition enables efficient transport of SV40 to the ER from the endosome. In addition to tethering the ER to the endosome, the EMC also serves as a scaffold to link the ER to the mitochondria [31]. Beyond these tethering activities, the EMC can act as a transmembrane chaperone, promoting the biogenesis of ER membrane proteins [32–34] and insertion of tail-anchored membrane proteins [35].

Upon reaching the ER lumen, the SV40 viral particle encounters redox proteins including protein disulfide isomerase (PDI), ERp57, and ERdj5 [36–40]. These oxidoreductases reduce and isomerize the disulfide bonds of SV40 (Figure 2, step 2), imparting conformational changes to the virus that expose VP2 and VP3. As these viral proteins are hydrophobic, exposure of VP2 and VP3 results in a hydrophobic particle that integrates into the ER membrane (Figure 2, step 3) [36,41,42].

Once membrane integrated, SV40 reorganizes the ER to form a membrane penetration site (called focus) which serves as the portal through which the virus enters the cytosol (Figure 2, step 4) [25–28]. How this membrane penetration structure is constructed is slowly emerging. To induce the formation of the focus, SV40 recruits a diverse array of ER membrane proteins to the focus that sub-serves distinct functions. For instance, the BAP31 membrane protein reorganizes to the focus where it interacts with the membrane-embedded SV40 [26] — this is thought to prime the virus for escape into the cytosol. Within the lipid bilayer of the ER membrane, the EMC1 membrane protein (another EMC subunit) acts as a transmembrane chaperone by providing stability to the membrane-inserted virus, preventing early disassembly of SV40 that would preclude proper escape into the cytosol [43]. To eject into the cytosol, SV40 also induces membrane-bound J proteins (B12, B14, and C18) to reorganize into the focus — this recruits a cytosolic protein machinery (composed of the chaperones Hsc70, Hsp105, SGTA, and Bag2) to the J proteins [25,28,44–46]. The Hsc70-Hsp105-SGTA-Bag2 extraction machinery, using an ATPase-driven cycle, ejects the viral particle into the cytosol to complete the membrane escape process (Figure 2, step 5). The cytosolic chaperone ubiquilin4 (Ubqln4) also participates in this SV40 cytosol escape step [47], although its precise role remains unclear.

A central concern during ER-to-cytosol escape of SV40 is the potential damage this process may impose on the ER membrane. To prevent this, the ER morphogenic proteins reticulon 3 and 4 (RTN3/4) protect the integrity of the ER membrane when the virus crosses the ER membrane [29]. RTN3/4 do so by targeting to the focus where they deploy their curvature-inducing activity to provide membrane stability and flexibility. Reorganization of the different membrane proteins to the ER focus likely requires substantial mechanical force. Indeed, the power generated by the kinesin-1 motor provides the force to build this structure [48]. Specifically, a kinesin-1-dependent force drives coalescence of multiple smaller immature foci structures into a single large mature functional focus which enables SV40 entry into the cytosol en route to the nucleus for productive infection. Intriguingly, the virus-induced focus is positioned next to a cytosol-localized disassembly machinery anchored at the Golgi apparatus [49]. This machinery — composed of BICD adaptors of the dynein motors — disassembles cytosol-localized SV40, thereby priming the viral particle for nuclear entry.

Hence, conceptually, the ER serves as a viral membrane penetration site for SV40 and other PyV family members. The idea that the ER acts as the cytosol entry portal for PyVs is unique because no other viruses have been reported to use the ER for this purpose. Clarifying whether other viruses similarly exploit the ER for membrane escape clearly deserves future investigation.

Flavivirus exploits the ER to support viral translation, replication, and assembly

Flaviviruses including West Nile virus (WNV), Dengue virus (DENV), Japanese encephalitis virus (JEV), Zika virus (ZIKV), yellow fever virus (YFV), and tick-borne encephalitis virus (TBEV), are the most important arthropod-borne viruses that cause significant human diseases [50,51]. They infect cells by binding to receptors on the host membrane followed by internalization via clathrin-mediated endocytosis (Figure 3) [51–54]. Upon reaching the endosomes, the acidic environment triggers the fusion of the viral and endosome membranes, enabling the release of the viral nucleocapsid into the cytosol where it is uncoated (Figure 3, step 1) [53]. After uncoating, the flavivirus genome — consisting of a 5′ capped, positive-sense, single-stranded RNA — binds to ribosomes and is co-translationally translocated on the ER membrane to produce a viral membrane polyprotein (Figure 3, step 2) [53]. This polyprotein undergoes multi-site cleavage by viral and host proteases, generating three structural proteins (C, prM and E) and seven non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) (Figure 3, step 3) [55]; the non-structural proteins facilitate viral replication by manipulating the ER structure (see below), and elicit host immune responses [56,57].

Figure 3. The ER supports protein translation, replication, and assembly of flaviviruses.

After receptor-mediated endocytosis, flavivirus reaches the endosome where its membrane fuses with the endosomal membrane (step 1). This step releases the nucleocapsid into the cytosol where it is uncoated. The naked positive-sense RNA genome is co-translationally translocated on the ER by the Sec61 translocon (step 2), generating a single polyprotein which, after cleavage, produces three structural proteins (C, prM, and E), as well as seven non-structural proteins (NSPs) (step 3); the EMC supports biosynthesis of some of the NSPs. Importantly, the NSPs create a replication organelle in the ER that promotes replication of the viral genomic RNA (gRNA) (step 4). The replicated gRNA is packaged with the C protein to form the nucleocapsid which assembles with the viral prM and E membrane proteins on the ER membrane, thereby generating a new progeny that exits the ER (step 5). The new viral particle reaches the Golgi (step 6) and uses the classic secretory pathway to undergo egress.

In addition to supporting polyprotein translation, the ER serves an important role during flavivirus replication (Figure 3, step 4). In fact, one striking feature during flavivirus infection is that the ER undergoes extensive rearrangements, forming distinct virus-induced structures including vesicle packets (VPs), convoluted membranes (CMs), and membrane vesicles (Ve) that are used to sustain replication [58]. Topologically, virus replication occurs on the cytoplasmic side of these virus-induced, ER-derived structures [58]. The viral proteins NS1, NS2A, NS2B, NS4A, and NS4B have been reported to drive this ER remodeling reaction [59–63]. In concert with these viral proteins, many host factors that typically trigger cellular ER remodeling and structural rearrangements, such as the RTNs and atlastins (ATLs), play prominent roles during flavivirus infection. For instance, RTN3.1A facilitates WNV, DENV, and ZIKV replication by either direct or indirect interaction with viral NS4A to enable replication complex formation [64]. Intriguingly, two studies reported on the role of ATLs during flavivirus infection, and found varied effects depending on the virus [65,66]. Whereas ATL2 depletion reduced viral titers and RNA levels during DENV, WNV, and ZIKV infection, silencing ATL3 only reduced viral titers in DENV and ZIKV (but not WNV) infection and had no effect in the viral RNA levels [65]. This suggests that WNV exploits an ATL3-independent step during infection. Several recent studies further reported on an important role of the EMC in facilitating flavivirus replication [67–69]. Specifically, EMC promotes the biogenesis of DENV and ZIKV non-structural multi-pass membrane proteins NS4A and NS4B [70–72]. Because NS4A and NS4B are crucial for the formation of the flavivirus replication compartment [60], it is not surprising that loss of EMC adversely affects virus replication.

In addition to virus translation and replication, the ER appears to support flavivirus assembly (Figure 3, step 5). Virus assembly — in which the viral structural proteins and replicated genomic RNA come together — is initiated in the ER. Specifically, the viral genomic RNA and the C protein forms the nucleocapsid and buds into the ER lumen, acquiring the prM and E proteins. The assembled immature viral particle then exits the ER en route to the Golgi (Figure 3, step 6) where it matures in order to traffic further along the anterograde pathway for secretion. Although the action of ER proteins during flavivirus assembly has not been characterized extensively, some reports have documented their roles. For example, the KDEL receptor — which normally cycles between the ER and Golgi to retrieve ER-resident proteins to the ER [73] — binds to the viral protein preM, thereby guiding the immature DENV particle from the ER to the Golgi for maturation and eventual secretion [74].

ER quality control processes including the UPR (Figure 1C) has also been shown to be exploited by flaviviruses to support their replication, and to regulate the innate immune response of the host cell [75–78]. These examples of the flavivirus-ER interplay thus illustrate how this virus exploits a wide range of ER functions to support its replication cycle, including viral protein translation, replication, and possibly assembly.

Coronavirus co-opts ER functions to promote viral translation, replication, sub-genomic transcription, and assembly

Similar to flaviviruses, coronaviruses (CoVs) are enveloped viruses containing a positive-sense, single-stranded RNA genome. Many coronaviruses infect humans and cause the mild common cold, such as the human coronavirus OC43 (HCoV-OC43), human coronavirus HKU1 (HCoV-HKU1), and human coronavirus 229E5 (HCoV-229E5) [79]. However, in the past two decades, highly pathogenic human coronaviruses have emerged, including the severe acute respiratory syndrome-related CoV (SARS-CoV-1) in 2002 with 8000 confirmed cases worldwide and almost 800 deaths [80], and the middle east respiratory syndrome-related CoV (MERS-CoV) in 2012 with 2500 confirmed cases and a 36% fatality rate [81]. The ongoing global pandemic of coronavirus disease 2019 (COVID-19) is caused by SARS-CoV-2. As of August 2020, more than 22 million COVID-19 cases have been reported in greater than 188 countries, resulting in at least 780 000 deaths [82]. Hence, there is a pressing need to understand the molecular basis of infection by these viruses in greater detail.

CoVs are large RNA viruses with genomes ranging from 27 to 32 kb. Because SARS-CoV-1 shares 79% of its genome with SARS-CoV-2 and uses the same entry receptor (angiotensin-converting enzyme 2 (ACE2)) as SARS-CoV-2 [83,84], mechanistic insights of SARS-CoV-1 entry is likely to inform SARS-CoV-2 infection. SARS-CoV-1 and SARS-CoV-2 infection is initiated when their viral surface spike (S) protein binds to the ACE2 receptor on the host cell membrane [85,86]. After receptor-engagement, the viral particles are endocytosed and transported to the endosome via a clathrin-/caveolae-independent pathway (Figure 4) [87]. Similar to flaviviruses, the low endosomal pH triggers the fusion of the CoV viral and endosome membranes, leading to the release of the nucleocapsid into the cytosol where it is uncoated (Figure 4, step 1).

Figure 4. The ER promotes viral translation, replication, sub-genomic transcription, and assembly of coronaviruses.

Post-endocytosis, coronavirus can traffic to the endosome where it fuses with the endosomal membrane, thereby releasing the viral nucleocapsid into the cytosol (step 1). After uncoating, the positive-strand RNA is likely co-translationally translocated on the ER via the Sec61 translocon (step 2). This generates polyproteins pp1a and pp1ab which, following proteolytic cleavage, produce non-structural proteins (NSPs) 1–16. Some of the NSPs facilitate the formation of an ER-derived double-membrane vesicle (DMV) (step 3) that functions as the replication site of the viral genomic RNA (gRNA) (step 4). Intriguingly, sub-genomic RNA (sgRNA) produced after nested transcription of the gRNA also occurs in the DMV; the sgRNAs encode the four viral structural proteins S, E, M, and N. Next, the replicated gRNA and sgRNA are exported from the interior of the DMV into the cytosol (step 5). The sgRNA encoding S, E, and M are targeted to the ER where the translation of their messages generates the S, E, and M transmembrane proteins (step 6), while the sgRNA encoding N is translated in the cytosol to produce the N protein. The cytosolic N protein then associates with the gRNA, and the complex is targeted to the ERGIC compartment where it assembles with the S, E, and M proteins (which reached the ERGIC from the ER) (step 7). The newly assembled virus finally exits the host cell via the secretory pathway.

The released cytosol-localized viral RNA genome is likely co-translationally translocated on the ER membrane (Figure 4, step 2), although definite evidence for this event remains elusive. Translation generates the viral replicase-transcriptase proteins. These proteins are encoded in open-reading frame 1a (ORF1a) and ORF1b, which are synthesized as two large polyproteins, pp1a and pp1ab. Intriguingly, biosynthesis of pp1ab requires an unusual programmed ribosomal frame shifting during the translation of ORF1a. During or immediately after synthesis, pp1a and pp1ab are cleaved by host and viral proteases, generating 16 non-structural proteins (NSPs); NSP1–NSP11 are encoded in ORF1a, while NSP12–NSP16 are encoded in ORF1b [88,89].

In addition to supporting polyprotein translation, the ER is essential for CoV replication, sub-genomic transcription, and assembly. Specifically, the viral NSPs, including NSP3, NSP4, and NSP6, induce the formation of ER-derived double-membrane vesicles (DMVs) (Figure 4, step 3) and CMs in perinuclear regions of infected cells to which the viral replication-transcription complex (RTC) is anchored [90–94]. The RTC in turn promotes viral genome replication and sub-genomic RNA transcription (Figure 4, step 4). The sub-genomic RNAs are generated from nested transcription of the full-length genomic RNA template. Importantly, translated from the sub-genomic RNAs are the four viral structural proteins necessary for the formation of the new viral progenies — nucleocapsid (N), spike (S), membrane (M), and envelope (E). Once synthesized, the viral genomic and sub-genomic RNAs are extruded from the DMV (Figure 4, step 5), likely via a pore that spans the double-membrane of the DMV [90–93,95].

The sub-genomic RNAs encoding S, M, and E are targeted to the ER for co-translational translocation to generate the viral structural membrane proteins (Figure 4, step 6), while the sub-genomic RNA encoding N is translated in the cytosol to produce the cytosolic N protein. In contrast, the replicated genomic RNA is targeted to a post-ER organelle called the ER-Golgi intermediate compartment (ERGIC) [96,97]. Here it assembles with the N protein, forming the nucleocapsid (Figure 4, step 7). The nucleocapsid in turn buds into a vesicle decorated with the newly translated S, M, and E viral proteins, generating a new viral progeny which exits the cell via the secretory pathway [98,99].

A critical question is why CoV induces DMVs (and CMs) to support viral replication and sub-genomic transcription? One possibility is that these architectures provide a privileged micro-environment which are physically separated from the cytosol, ensuring sufficient concentrations of macromolecules required for viral RNA synthesis. Additionally, these ER-derived structures may prevent recognition of the viral replication intermediates such as the double-stranded RNA (dsRNA) by cytosolic innate immune sensors [100,101].

Beyond hijacking the ER to support viral polyprotein translation, replication, sub-genomic RNA transcription, and assembly, CoV infection also induces ER stress via activation of the UPR [102,103]. Because one consequence of UPR activation is the increased synthesis of ER-resident chaperones, this may provide an advantage for CoV by supplying the ER with more chaperones to assist the folding of viral structural and non-structural proteins. At present, the precise molecular basis by which CoV triggers the UPR is not well-understood [104,105]. Additionally, ER membrane rearrangements that generate the DMVs and CVs during CoV infection are also known to induce ER stress and UPR [102,103]. Importantly, the infected cell responds by increasing ER membrane biogenesis. This likely further benefits CoV infection by providing additional membrane source for DMV and CV formation. Thus, activation of ER stress and UPR may in fact be part of a pro-viral response during CoV infection. In sum, the ER plays an equally critical role in CoV infection when compared with flavivirus, promoting viral translation, replication, sub-genomic RNA transcription, and assembly, as well as generating a stress response that appears to favor the viral life cycle.

Perspective.

• Importance to the field: Although the ER serves a critical cellular function, this compartment is often exploited by viral pathogens to assist in their infection process. Hence, illuminating the basic mechanism by which viruses hijack ER functions to promote infection should provide rational therapeutic approaches to combat virus-induced diseases.

• Current thinking: Amongst the myriad of viral pathogens, there is clear evidence that the replication life cycle of polyomavirus, flavivirus, and coronavirus exploit the ER to achieve productive infection. Specifically, while the ER acts as a membrane penetration site for polyomavirus during entry, this organelle serves a more complex function in promoting viral translation, replication, and assembly for flavivirus and coronavirus.

• Future directions: An outstanding question is whether other viruses use the ER as the site of membrane escape into the cytosol, or whether this is exclusive to polyomavirus. Additionally, how host factors co-operate with viral components to construct the ER-derived replication organelle for flavivirus and coronavirus deserves significant future attention.

Abbreviations

ACE2

angiotensin-converting enzyme 2

ATLs

atlastins

CMs

convoluted membranes

CoVs

coronaviruses

DENV

Dengue virus

DMVs

double-membrane vesicles

EMC

ER membrane complex

ER

endoplasmic reticulum

ERAD

ER-associated degradation

ERGIC

ER-Golgi intermediate compartment

ORF1a

open-reading frame 1a

PyVs

polyomaviruses

RTC

replication-transcription complex

SV40

simian virus 40

UPR

unfolded protein response

WNV

West Nile virus

ZIKV

Zika virus