Coronavirus Disease-19 (COVID-19) pandemic is caused by the +RNA coronavirus SARS-CoV-2 that has infected now more than 600 million people and has killed more than 6 million people worldwide ¹ , since it has no definitive and effective treatment until today. This infection affects mainly certain groups of people that have high susceptibility to present severe COVID-19 due to comorbidities that include cardiovascular, chronic kidney, respiratory or liver disease, severe obesity, or hypertension among others. In these patients, the cytokine storm induced by the virus poses a serious death risk due to the systemic inflammation and multiorgan failure ^{2– 6} . Moreover, the so called long-COVID-19 comprises a series of symptoms that may remain in some patients for months after infection that further compromises their health, even after non-severe COVID-19 ⁷ . Despite huge efforts to stop infections and deaths worldwide, only a few treatments have been demonstrated to ameliorate severe COVID-19, and different vaccine strategies are currently under investigation in clinical phase IV trials. Unfortunately, it is now clear that the intended vaccine objective, the prevention of COVID-19 ⁸ , was not fulfilled adequately due to waning immunity ⁹ . Therefore, in this scenario, a deeper understanding of the cellular mechanisms exploited by SARS-CoV-2 for cell infection and replication could undoubtedly provide new unforeseen strategies to tackle this pandemic.

Several recent reports have shown that mitochondria play a relevant role during SARS-CoV-2 infection ^{10– 14} . These findings and previously published results of SARS-CoV-2, SARS-CoV and other coronavirus biology let hypothesize that mitochondria could be responsible for the assembly of double-membrane vesicles (DMV). These are membrane modifications induced by SARS-CoV-2 where virus replication occurs in the infected cell, assumed to protect vRNA from the intracellular environment and prevent the antiviral response. Currently, DMV are mainly believed to be derived from the endoplasmic reticulum (ER) or other mechanisms as autophagy ^{15– 17} . However, published literature supports that double-membrane mitochondria-derived vesicles (MDV), discovered almost 15 years ago ¹⁸ , could be the precursors or relatives of DMV. This hypothesis of mitochondria role in DMV assembly and the involvement of MDV has been suggested previously ^{10, 19, 20} . Indeed, mitochondria is a well-known target of different viral infections ²¹ . Furthermore, specialized replication organelles (RO) at mitochondria outer membrane (MOM) have been observed in FHV ²² , whereas HIV RNA is known to locate at mitochondria ²³ . Here, a review of the evidence that supports this notion for SARS-CoV-2 is presented and integrated into the current model ²⁴ , with the intention to provide a novel framework that could open possibilities to tackle the SARS-CoV-2 pandemic and COVID-19.

DMV along with other membrane modifications are part of the RO induced by SARS-CoV-2. Other membrane modifications include convoluted membranes (CM), zippered ER (zER), vesicle packets (VP), and double-membrane spherules (DMS) ²⁴ ( Figure 1). RO are induced by SARS-CoV-2 in infected cells, as a variety of RO are induced by viruses including other nidovirus and picornavirus among others ^{15, 17} . DMV assembly is induced by viral proteins but seems to also require other viral or host factors because cell plasmid transfection of transmembrane-containing non-structural proteins (nsp) 3, 4, and 6 induced membrane arrangements that resemble DMV but with a smaller size ²⁵ . These nsp are part of the complex involved in viral RNA (vRNA) replication, together with nsp12, the RNA-dependent RNA polymerase (RdRp), and other nsp. Consistently, nsp4 mutation alters the assembly of DMVs ¹⁹ and abolishes viral replication ²⁶ . Nsp4, 3, and the nuclear (N) protein are located at DMV ^{20, 24} , where nsp3 has been recently shown to form a pore complex, elusive for some time, that communicates DMV interior with the cellular cytoplasm, complex that could also involve host factors and/or other viral proteins ²⁷ . Surprisingly, it has been shown that RNA synthesis in SARS-CoV infected cells responds differentially to two different inhibitors of protein translation. Cycloheximide, that inhibits eukaryotic translation partially prevented RNA synthesis, whereas puromycin, that inhibits eukaryotic and prokaryotic translation almost fully blocked RNA synthesis ²⁸ . This finding suggests that the synthesis of proteins relevant for vRNA replication occurs at the mitochondria.

Most membrane modifications induced by SARS-CoV-2 are presumably derived from ER membranes and are interconnected by membrane contacts forming a reticulo-vesicular network ^{24, 29} . In the case of DMV, the ER origin was in part assumed because such mechanism is presumed to mediate RO assembly in other +RNA viruses ³⁰ ; because they have contacts with other membrane modifications of the RO and the ER; because ribosomes have been observed on DMV surface, and because DMS and CM were proposed to be precursors of DMV ^{15, 20, 24, 25, 30} . DMV size ranges from 150 to 300 nm, but they grow through infection, and it has been well-established that SARS-CoV-2 DMV are the location where vRNA synthesis occurs ^{20, 24, 31} . Importantly, DMV are early (1–2 h post-infection [p.i.]) observed in the cell cytoplasm after coronavirus infection , and their number increase through time reaching a maximum in 6–8 h.p.i. ^{20, 32} . There are currently two models for DMV assembly from the ER. In the case of coronavirus, DMV are thought to be assembled from zER that folds and closes in response to vRNA, as observed with IBV ^{15, 30} , with CM and DMS as putative intermediate precursors ^{20, 24} . However, major challenges remain for this model, because no intermediate structures have been recognized between DMV and DMS or CM, and because CM and DMS, both derived from the ER, have no relationship with DMV beyond their membrane contacts ^{20, 24} . Moreover, DMV do not have ER (nor ERGIC, Golgi, or endosomal/lysosomal) conventional markers as it would be expected if they were assembled from the ER. Only partial overlap of PDI or colocalization of non-conventional markers have been reported (calnexin, reticulon and sec61), but interestingly, these have been identified as part of the proteome of ER-mitochondria contacts, also called mitochondria associated membranes (MAMs) ^{29, 32, 35– 39} . Most importantly, it has been shown that DMV assembly (1-2 h p.i.) precedes CM assembly (~3h p.i.) ³² . Furthermore, CM and DMS do not carry out vRNA synthesis as DMV do ²⁴ . Thus, CM and DMS seem not to be DMV precursors ²⁴ . On the other hand, the high energetic and complex topological requirements assumed to occur for DMV assembly through the zippered ER model, given their numbers after a few h.p.i, further complicate this notion. Indeed, it has been suggested that DMV could have another origin than the ER ^{10, 15, 24} .

Given the recent findings that relate SARS-CoV-2 with mitochondria, it is possible that MDV are the origin of DMV. MDV were discovered almost 15 years ago, and they comprise different kind of vesicles shed from mitochondria with selective cargo and markers (TOM+ or MAPL+). They are involved in intracellular traffic to different organelles and mitochondria quality control ^{18, 40– 43} . Interestingly, some TOM+ MDV have a double-membrane with an average size of 160 nm (range 80–250 nm) ( Figure 2A) ⁴² . Initially, MDV were found to be shed independently of drp1, mitochondria fission, and autophagy, however, a more recent analysis with improved tools has shown that drp1 is essential for their scission ^{40, 42, 44} . Several coronavirus proteins have been shown to down- or up-regulate drp1 (N, envelope [E], nsp3, nsp4a, nsp4b, and orf9b), molecule that is critical for mitochondria fission ¹² . In particular, SARS-CoV orf9b expression results in drp1 degradation through the proteasomal pathway, casting doubts about the role of MDV as DMV precursors or relatives. However, it must be noted that this evidence was gathered with a cell line stably transfected with orf9b, thus, its expression kinetics profile does not correspond to that of an actual viral infection ⁴⁵ . Indeed, in an infected cell, the expression of orf9b could concur with DMV generation decrease.

MDV are generated in steady conditions and have been observed in vivo ⁴⁶ , but their number increases after mitochondrial stress or higher respiratory activity ^{40, 44} . In this regard, it has been found that SARS-CoV-2 infected alveolar epithelial cells, monocytes and PBMC have compromised mitochondrial function and energy deficit ^{47– 50} , even months after infection clearance and symptoms relief ⁵¹ . These observations could be the long-term result of viral infection in which mitochondria shedding of MDV/DMV, triggered initially by a burst of metabolic activity and/or stress, leads to mitochondria function impairment. In this regard, OXPHOS impairment, ROS increase and HIF1α activation, that reprograms cells towards glycolysis, have been reported after viral infection ^{33, 49, 52} . ROS stress could be the trigger to induce MDV/DMV production, although this needs to be tested since nsp 3, 4 and 6 transfection induce the generation of DMV-like vesicles ²⁵ . Furthermore, mitochondria phenotype alteration has been observed in vitro ( Figure 2E) ^{24, 33, 47} and in COVID-19 postmortem specimens ⁵³ that may comprise fusion, reported in most infected cells, or fission which has also been observed ^{49, 53} . On the other hand, mitochondria disturbance in infected patients could also be due to angiotensinII excess, that results after ACE2 downregulation by SARS-CoV infection or Spike (S) protein binding ^{54– 58} .

MDV not only look like DMV for their double-membrane and approximate size, they also transport selective cargo to peroxisomes or the endolysosomal system after mitochondrial stress, depending on Vps35 and syntaxin-17/SNAP29/VAMP7, respectively ^{40, 44, 59, 60} . Indeed, it has been recently reported that MDV may compensate for the loss of mitophagy, thus keeping mitochondria quality control, an in SARS-CoV-2 infected cells mitophagy is impaired ^{33, 43} . The intracellular traffic of MDV offers a pathway that could be involved in the intracellular transport of viral components to secondary vesicular structures (i.e. ERGIC, lysosomes, multivesicular bodies) where viral particles are assembled and set ready for exocytosis. Indeed, mitochondrial proteins have been identified in exosomes from SARS-CoV-2 infected neurons and astrocytes ⁶¹ . Interestingly, PINK1/Parkin and the mitochondrial ubiquitin ligase MULAN1 (MAPL) have been involved in MDV shedding from mitochondria ^{44, 62} and it is known to be involved in the antiviral response of mitochondria ⁶³ .

Proteomic studies have shown that MDV may contain 100–200 different mitochondrial molecules, including subunits of the translocase complexes of the MOM (TOM 20, TOM70, TOM40) and inner mitochondria membrane (IMM) (TIMM22, TIMM23 and TIMM29); components of the OXPHOS complexes (COX1, CYCS, UQCRC1); the VDAC1-3; enzymes of the TCA cycle; enzymes such as pyruvate dehydrogenase, lactate dehydrogenase and hexokinase; the MCU; the ATP/ADP translocase; Fe-S cluster binding proteins; and Miro1/2 among others ^{40– 42, 64} . Thus, mitochondrial components and resources useful for vRNA replication are transported or could be generated within MDV, since they are “chunks” of mitochondria. It was recently shown that DMV have pores that span the double-membrane that mediate the export of vRNA from DMV and could also mediate the exchange of molecules between DMV and the cytoplasm ²⁷ . These authors also showed that once vRNA is exported, it complexes with viral N protein, association that is known to increase vRNA translation in trans ⁶⁵ , which could probably occur at ribosomes located at the external membrane of DMV ²⁰ . It is important to note that MDV have a very short shedding dynamic (10–30 s), and peak at 2 h after stimulation ^{40, 42, 62, 66} . This could explain why the shedding step is not frequently spotted by TEM.

TOM+ MDV have been observed to originate from mitochondrial protrusions termed mitochondria nanotubes or mitochondria derived tubules ^{42, 66} . Interestingly, it has been shown that such nanotubes emerge preferentially at ER-mitochondria contacts or MAM´s, that are enriched with Miro1, critical for MDV genesis ^{42, 67} . These observations open the possibility that the well-described interconnections between DMV and the ER are actually the reminiscence of these MAMs, rather than the common ER membrane shared by DMV, from where, according to the current paradigm, DMV originate ^{15– 17} . These findings also indicate that MAMs could play a critical role in DMV genesis and function through their roles in Ca2+ and lipid exchange, autophagy, mitochondrial biogenesis and ER stress response ^{35, 68} . Intriguingly, mitochondrial nanotubes are relevant for mDNA transfer between mitochondria, with a central role played by Mic60 that is present in MDV, however, MDV are devoid of mDNA ^{42, 67} . As mentioned above, MDV shedding requires drp1 for their scission that may be recruited by adaptors MID49, MID51 or MFF, that associate with Miro1/2, that in turn mediate the formation of mitochondria derived tubules through Mic60 interaction with kinesin that pulls from microtubules ^{42, 67} . Importantly, these adaptors may interact with the TOM complex through intermediate players, thus suggesting that TOM may play a pivotal role in MDV genesis. These intermediate players are endophilin B for MID49 and Huntingtin for MFF ⁶⁹ . Interestingly, the interaction between MFF and TOM22 with AQP6 has also been reported ⁶⁹ . The putative role of these molecules requires to be tested experimentally.

A common feature of SARS-CoV-2 RO are nearby mitochondria, which may show signs of cisterna swelling and disorganization, similar to mitochondria with induced MDV shedding ^{40, 44, 59, 64} , or membrane continuity with DMV ^{20, 24, 30– 32} . Nevertheless, some TEM images have shown budding of what could be DMV from mitochondria ( Figure 2B) ¹⁹ . In that work with nsp4 temperature-sensitive mutants, at the non-permissive temperature, there was an increase of mitochondria size, with enlarged cisternae, and increased localization of nsp4 and nsp3 at mitochondria, perhaps resulting from the reduction of MDV shedding, that in turn resulted in a reduced number of DMV ( Figure 2B) ¹⁹ . In some cases, in closely located DMV and mitochondria, de novo synthesized vRNA signal can be observed within both, apposed to each other ( Figure 2C) ²⁴ . Although the mathematical method employed by these authors to evaluate these labels did not identify it as positive compared with the majoritarian label observed in DMV, it is possible that the abundance of vRNA and viral proteins within mitochondria are under tight control through the shedding of MDV/DMV, and thus, that only few vRNA are found within mitochondria at a given moment. Notably, electron tomography has shown what could be an intermediate between MDV and DMV, a vesicle tethered to the ER with a double-membrane that contains a cisterna-like arrangement of the inner membrane ( Figure 2D) ³⁰ . Strikingly, it was recently found that Fe-S cofactors, of which biosynthesis initiates at the mitochondria, are involved in SARS CoV-2 RdRp function ⁷⁰ , and some proteins related with these clusters are present and can be enriched in MDV after oxidative stress ⁴¹ . Moreover, a recent in silico analysis predicted SARS-CoV-2 RNA localization to host mitochondria and nucleolus, further supporting this idea ⁷¹ . Indeed, this work led many to consider mitochondria infection by vRNA. As a matter of fact, while this manuscript was under review, this year two different groups demonstrated vRNA localization at mitochondria ^{33, 34} . Furthermore, some images have shown that vRNA located inside of mitochondria are polarized towards DMV pools ( Figure 2C and E). In Figure 2C, it is also possible to see a degenerated-like mitochondria among DMV and mitochondria (spark in Figure 2E), probably resulting from large MDV/DMV shedding ²⁴ . Notably, it has been shown that SARS-CoV-2 isolated from infected individuals is capable to replicate in bacterial cultures ⁷² , further endorsing that mitochondria plays a critical role for SARS-CoV-2 cycle.

Supporting also the role of mitochondria for DMV assembly is the unexpected identification of several mitochondria molecules involved in different mechanisms of its physiology (electron transport, metabolism, mitochondria ribosomes, RNA maturation, and cellular immune signaling) as interactors of viral proteins ⁷³ . Although most of these interactions still require to be functionally validated, some of the putative relevant interactions that this work identified is that of nsp4 with the inner mitochondria membrane translocase (TIMM) complex, the interaction of orf9b with TOMM70, and the interaction of nsp6 and orf 9c with the Sigma receptor. Of these, the TOM70-orf9b and SigmaR-Nsp6 have been validated experimentally ^{74, 75} . The Sigma receptor has been involved in several mitochondria functions, related to its location at the MAMs ⁶⁸ , enriched with interactors of nsp2 and 4 ⁷⁶ . Additional, intriguing, unexpected interactions were that of SARS CoV-2 membrane (M) protein with FASTKD5, involved in mitochondrial RNA maturation, and that of nsp8 with different mitochondria ribosomal proteins (MRP). Also, interactions of orf3a and M protein with relatives of known partners of MULAN1 (REEP and TRIM) were also identified, however, given the diversity of these molecules, more work is required to test the significance of these interactions. Importantly, the orf3a protein has been reported to induce apoptosis of infected cells through the extrinsic pathway ⁷⁷ . In a different study, nsp2 was found to interact with VDAC2 ⁷⁶ , the mitochondrial porine, whereas the mitochondria antiviral-signaling protein (MAVS) has also been identified as a target of SARS-CoV-2 infection ¹² . Together, these interactions of viral proteins with the host support that mitochondrial function is very relevant for SARS-CoV-2 infection. Furthermore, since the mitoribosome, the mitochondrial RNA maturation and translocation mechanisms are targets of viral proteins, that according to the current model of infection are unexpected, these findings also hint that SARS-CoV-2 infects mitochondria as part of its replication cycle, rather than only hijacking this organelle through viral proteins synthesized elsewhere. The down-regulation of mtDNA encoded genes and mitochondrial RNA in patient autopsies and disregulation in infected cells by SARS-CoV-2 also supports this notion ^{48, 49, 78} . Other relevant interactions of the viral proteome with mitochondrial proteins have been analyzed by others ^{10, 12, 76} . A key question in this scenario is which are the steps that mediate the shedding and transformation of MDV into DMV, and how viral proteins are involved ( Figure 1 question mark 1).

Interestingly, in a recent paper, an alternative mechanism of mitochondria fission has been described that occurs under stress and high energy demand, that depends upon the MOM molecule Fis1, which yields small mitochondria destined for mitophagy ⁷⁹ . This could represent a different pathway involved in DMV genesis.

As explained above, DMV origin has also been proposed to be related with the autophagic pathway. However, several lines of evidence indicate that autophagy is unlikely to play a role in the generation of coronavirus replicative structures ⁸⁰ . Since MDV genesis is independent of FUDCN1, an activator of mitophagy, this also advocates against the autophagic origin of MDV/DMV ⁴² . Moreover, it has been reported that SARSCoV2 infection induces and inhibits autophagy and impairs mitophagy of infected cells ^{33, 81} . This compromises quality control mechanisms, further supporting MDV/DMV relationship, because it has been shown that MDV may compensate loss of mitophagy thus maintaining mitochondria quality control ⁴³ . Indeed, it has been suggested that autophagy modulators could treat COVID-19 given the impairment of the pathway ⁸¹ .

Taken together, these observations strongly suggest that mitochondria could be targets of SARS-CoV-2 vRNA infection, leading to MDV/DMV assembly. Also, they indicate that mitochondria somehow get vRNA that could induce stress and therefore shedding of MDV/DMV, where the vRNA replication machinery and newly synthesized vRNA are mostly located ²⁴ . Some of these possibilities have already been confirmed because two different groups demonstrated this year the localization of vRNA within mitochondria of infected cells in vitro and in vivo ^{33, 34} , involving TOM20 as mediator, thus substantiating the mitochondrial origin of DMV.

But then, how MDV/DMV are induced by SARS-CoV-2? There are at least two main possibilities that are non-self-exclusive: one that is consistent with the current paradigm is that viral proteins synthesized at the ER and/or its membrane modifications in the RO somehow reach mitochondria, modulate its physiology and induce DMV, in 1–2 hours. In this regard, there are some viral proteins with mitochondria localization sequences of SARS-CoV such as 3b ⁸² , or that target proteins at the IMM (i.e TIMM, electron transport proteins, and MRPS), the MOM (TOM, VDAC), or the mitochondria matrix (FASTKD5) ⁷³ . Alternatively, the mitochondria could itself start viral protein synthesis with ribosomes located at the MOM ⁸³ , and/or uptake vRNA from the cytoplasm after virus entry and vRNA release into the cytoplasm, although the former would require mitochondria localization sequences in those translocated molecules. In this regard, it is known that mitochondria are capable to import RNA from the cytoplasm through a pathway that involves the TOM/TIMM complex ⁸⁴ , and SARS-CoV-2 RNA is predicted to locate at this organelle ⁷¹ . Indeed, this has been confirmed in a recent work published during the review of this manuscript that demonstrated that TOM20 mediates vRNA entry into mitochondria and that it is not colocalized with ER and lysosomal markers ³³ . In addition, a tantalizing possibility is that vRNA accesses mitochondria through TOM complex directly from vesicles shedded from the plasma membrane (PM) in which SARS-CoV-2 is endocytosed. This PM-mitochondria pathway mediates caveolin transport to mitochondria in myocytes after stress ( Figure 3A) ⁸⁵ , and it could be an early step of what we called plasma membrane-mitochondria bridges, which we recently described in astrocytes ( Figure 3B) ⁸⁶ , involved in the emerging pathway of PM-mitochondria interactions ⁸⁷ . These PM-mitochondria bridges contain vesicles, most probably caveolae, and mediate mass transfer from PM to mitochondria in minutes ⁸⁶ . Interestingly, it has been observed through TEM in infected cells from patients that virus containing vesicles may reach very close to MOM, vesicles that are shedded from plasma membrane regions where abundant caveolae-like invaginations are observed, presumably containing viral particles ⁵³ . Consistent with this idea, it has been reported that ACE2 can be located at the mitochondria ⁸⁸ , as it has been reported for several other plasma membrane molecules ⁸⁷ , whether ACE2 reaches mitochondria from plasma membrane was suggested although not established in this work. Moreover, ACE2 is known to promote mitochondria function ⁸⁹ . These observations endorse the possibility that SARS-CoV-2 may reach mitochondria directly after its endocytosis bound to ACE2.

Given that DMV are early induced by coronavirus (1-2 h.p.i.), fast and direct access of vRNA to mitochondria seems plausible and advantageous, compared with the ER origin of DMV. This scenario also provides the possibility to synthesize some viral proteins at IMM tethered mitoribosomes, with which nsp8 presumably may interact ⁷³ . Notably, mitoribosomes synthesize most exclusively membrane proteins that are co-translationally inserted into the membrane with the participation of OXA1 ^{90, 91} , as it is the case of transmembrane-containing nsp3, 4 and 6, involved in vRNA replication, of which nsp3 and 4 have been located at DMV and colocalize with nsp2, 5, 8, 12, 13 and 15 ^{20, 24, 27, 31, 32} . Interestingly, OXA1 has been identified at MDV by a proteomic study ^{41, 42} , thus supporting the idea that synthesis of proteins with transmembrane domains coded by the vRNA may occur within mitochondria and/or MDV/DMV. In addition, the M protein could optimize vRNA translation through its interaction with mitochondrial FASTKD5 protein ⁷³ , involved in mitochondrial RNA maturation ⁹⁰ , and at the same time, viral replication could profit the mitochondria synthesized Fe-S cofactors required for RdRp function ⁷⁰ . This scenario could provide the advantage of the protected environment of mitochondria matrix, rich in ATP, at least prior to OXPHOS reduction after infection ^{49, 52} , avoiding the requirement of large amounts of protein to be transported from ER to mitochondria that would require considerable amounts of energy and time.

However, the main concern against the idea that mitochondria can directly uptake SARS-CoV-2 from PM caveolae comes from one study suggesting that SARS-CoV endocytosis is caveolin-independent ⁹³ . This finding is based on the observation that cholesterol sequestration (one of the main components of lipid rafts that in turn are endocytosed by caveolae, ³³ ) with nystatin and filipin did not block pseudovirus entry. Indeed, nystatin enhanced it, whereas another cholesterol sequestering molecule, MβCD, did block it, therefore raising doubts about how cholesterol is involved. As matter of fact, different mechanisms of endocytosis have been found to mediate SARS-CoV-2 internalization ^{93, 94– 96} . In addition, in that study and a different one with CoV NL63, a major lack of colocalization between viral S protein and caveolin-1 at 20 or 60 m.p.i. was also considered as evidence for a caveolin-independent mechanism. However, the fast dynamic nature of this interaction (since we found that mass is transferred from PM to mitochondria in ~2 min ⁸⁶ ) nor the extracellular conditions were considered in these approaches. In this regard, it is relevant to note that extracellular acidification is known to induce transfer of caveolae to mitochondria (Reviewed in 87). Consistently, during SARS-CoV-2 infection extracellular conditions are acidified in the inflammatory setting and by the increase of the extracellular acidification rate (ECAR), related with lactate production by glycolysis reprogramming that has been found in SARS CoV-2 infected cells ^{47, 97} . Moreover, acidosis has been reported in COVID-19 patients ⁹⁷ .

On the other hand, there is some evidence supporting the role of the caveolae pathway in SARS-CoV-2 endocytosis: a) lipid rafts integrity is required for SARS-CoV entry and ACE2 is localized into lipid rafts, that are endocytosed through caveolae, well-known signaling hubs ^{98, 95} ; b) the S protein co-fractionates with caveolin-1 after binding to ACE2 ⁹⁵ ; c) an in silico approach found that SARS-CoV-2 proteins S, M, orf3, and replicase 1AB have putative caveolin binding motifs ⁹⁹ ; d) orf3a protein binding to caveolin has been demonstrated experimentally ¹⁰⁰ ; and e) cholesterol has been shown to enhance cellular infection though ACE2 endocytosis ¹⁰¹ . Thus, the precise role of cholesterol, caveolae, and caveolin for SARS-CoV-2 infection requires further investigation, because direct and fast viral targeting to mitochondria could be of great relevance for SARS-CoV-2 infection. Interestingly, cholesterol-bound RNA probes are targeted to mitochondria ¹⁰² . Furthermore, several alternative entry factors to ACE2 and facilitators capable to mediate SARS-CoV-2 infection have been identified ^{103– 106} , and they could mediate SARS-CoV-2 caveola-mediated endocytosis. Taken together, it is possible that the diversity of receptors and entry pathways exploited by SARS-CoV-2, together with the fast dynamics of PM-mitochondria communication can obscure the caveolae role that could mediate mitochondria infection.

If caveolae does mediate mitochondria infection by SARS-CoV-2, then relevant questions are the compartments involved and the resulting membrane topology. In this regard, at least two alternatives are possible: a) the first one is that the vRNA alone is translocated through the TOM complex from the cytoplasm or directly from the caveolae after viral and caveola membrane fusion. Indeed, the TOM complex has been already involved in SARS-CoV-2 vRNA presence within mitochondria ³³ , and together with TIM complex are known to mediate RNA import into mitochondria matrix ⁸⁴ ; b) the second is an odd although feasible possibility. The caveola membrane, fused or not with the viral membrane, could fuse with the MOM, thus delivering into the intermembrane space either the viral particle, or the vRNA in case caveola and viral membranes were pre-fused. In this step TOM20 may play a role ³³ , and could be related with the observation of ACE2 at mitochondria and with TEM showing a close interaction between a SARS-CoV-2 caveola-like vesicles with mitochondria ⁵³ . Then, the viral particle could fuse its membrane with the IMM and enable vRNA import into mitochondria matrix, although the mechanism seems problematic because S activation to mediate membrane fusion would be required. Alternatively, if only vRNA is translocated into the intermembrane space, then the TIM could help to import it into mitochondria matrix. The replication of SARS-CoV-2 in bacterial cultures seems to support this mechanism ⁷² . Mitochondria infection may result in the synthesis of viral proteins having transmembrane domains involved in vRNA replication at mitoribosomes, profiting there the Fe-S clusters synthesized by mitochondria that are required for the RdRp function ⁷⁰ . The experiments with protein translation inhibitors in SARS-CoV infected cells suggest that mitoribosomes and eukaryotic ribosomes participate in the synthesis of proteins involved in RNA replication because cycloheximide did not fully blocked it but puromycin did. Further experiments are required to explore these possibilities.

Evidence at morphological, biochemical, molecular and cellular levels reviewed here strongly support that MDV, most probably TOM+, are the precursors of DMV induced by SARS-CoV-2 cell infection. This mitochondrial vesicular intracellular traffic pathway described some years ago is still emerging, reason that perhaps has precluded its consideration in coronavirus infection. Nonetheless, this pathway offers some relevant advantages in comparison with the ER origin mechanism. The main one is that the topological difficulties that implicate the generation of many 3D spherical structures with a double membrane from the 2D rather planar ER are bypassed. In principle, given that MDV are quickly and easily generated through a scission mediated by drp1, this in turn would reduce the time and energy (ATP) required to generate a large number of DMV in a couple of hours. These features are consistent with the early appearance of DMV in infected cells and the subsequent appearance of other membrane modifications of the RO. In addition, the MDV pathway may reconcile the lack or partial overlap of conventional ER markers in DMV, as well as the demonstration of non-conventional ER markers in DMV that have been identified as part of the MAMs proteome. Furthermore, the inconsistent effect of protein translation inhibitors on RNA synthesis in infected cells could also be reconciled if viral protein synthesis in mitoribosomes is considered. Nevertheless, several questions still remain to be investigated to further support or question this hypothesis. Undoubtedly, the confirmation of vRNA within mitochondria, initially anticipated by a modelling approach, demonstrated just few months ago and detected in a previous work through TEM, provide a strong argument in favor of this idea, although a definitive proof is still lacking. On the other hand, the literature reviewed here points towards a substantial role of ER-mitochondria contacts (MAMs) in the genesis of DMV/MDV. Despite it is possible that some findings regarding coronavirus replication cycle that question this proposal may have escaped this review, several lines of evidence reviewed here advocate the origin of DMV from mitochondria.

The current pandemics has resulted in thousands of articles published monthly studying the SARS-CoV-2 or COVID-19. This burst of studies has made clear that mitochondria play a critical, if not central, role in SARS-CoV-2 infection, at the cellular and organismal levels. This discovery has opened new opportunities to tackle SARS-CoV-2 infection and COVID-19 considering the mitochondria biology, that has been studied for decades. Similarly, the mitochondrial origin of DMV/MDV represents an unforeseen chance to develop novel strategies to control viral replication, that in turn should provide better outcomes to COVID-19 patients. More basic and clinical work is required to test these possibilities and putative therapies based in these assumptions.

CM-convoluted membranes; DMS-double membrane spherules; DMV-double membrane vesicels; ER-endoplasmic reticulum; ERGIC-ER-Golgi intermediate compartment; IMM-inner mitochondria membrane; M-mitochondria; MAM-mitochondria associated membrane; MAVS-mitochondria antiviral-signaling protein; MOM-mitochondria outer membrane; nsp-non-structural protein; nsp-non-structural protein; PM-plasma membrane; TIMM-translocase of the mitochondrial inner membrane; TOM-translocase of the mitochondrial outer membrane; VDAC-Voltage-dependent anion channel; VP-vesicle packet (not shown); vRNA-viral RNA; zER-zippered ER.