Coxsackievirus A6 2C protein antagonizes IFN-β production through MDA5 and RIG-I depletion

ABSTRACT As a member of the enteroviruses, coxsackievirus A6 (CV-A6) has been a major cause of hand, foot, and mouth disease (HFMD) since 2008. It can infect both pediatric and adult populations, often leading to atypical HFMD. The host innate immune system plays a vital role in the development of enteroviral infections. However, the interplay between the host antiviral response and CV-A6 has not been well investigated. In the present study, we demonstrated that the 2C protein from CV-A6 (2CCV-A6) suppresses interferon beta (IFN-β) production in HEK293T cells. Further results indicated that 2CCV-A6 interacts with both melanoma differentiation-associated gene 5 (MDA5) and retinoic acid-inducible gene I (RIG-I) and induces the degradation of these RNA sensors through proteases in the lysosomal pathway. This function also applies to 2C proteins from enterovirus A71 (2CEV-A71) and coxsackievirus B3 (2CCV-B3) but not CV-A16 2C (2CCV-A16) for its incompetence in MDA5 and RIG-I recognition. Partial depletion and amino acid substitution analyses indicated that the F28A, V75A, and I96V mutations significantly compromised 2CCV-A6-induced MDA5/RIG-I depletion. Surprisingly, unlike V75A and I96V that interrupt the 2CCV-A6-MDA5/RIG-I interaction, 2CCV-A6 F28A remained competent in MDA5/RIG-I binding, suggesting that the interaction alone is not sufficient for 2C-mediated reduction. Additional tests indicated that CV-A6 viruses containing the 2C F28A mutation were less efficient in IFN-β suppression, which is associated with compromised viral replication and release in infected rhabdomyosarcoma (RD) cells, suggesting that 2C-mediated immune regulation plays a vital role in enteroviral replication. Taken together, our data reveal a novel mechanism by which enteroviral 2C proteins antagonize the host innate antiviral immune response. IMPORTANCE Coxsackievirus A6 (CV-A6) is a major emerging pathogen associated with atypical hand, foot, and mouth disease and can cause serious complications such as encephalitis, acute flaccid paralysis, and neurorespiratory syndrome. Therefore, revealing the associated pathogenic mechanisms could benefit the control of CV-A6 infections. In this study, we demonstrate that the nonstructural 2CCV-A6 suppresses IFN-β production, which supports CV-A6 infection. This is achieved by depleting RNA sensors such as melanoma differentiation-associated gene 5 and retinoic acid-inducible gene I (RIG-I) through the lysosomal pathway. Such a function is shared by 2CEV-A71 and 2CCV-B3 but not 2CCV-A16, suggesting the latter might have an alternative way to promote viral replication. This study broadens our understanding of enterovirus 2C protein regulation of the RIG-I-like receptor signaling pathway and reveals a novel mechanism by which CV-A6 and other enteroviruses evade the host innate immune response. These findings on 2C may provide new therapeutic targets for the development of effective inhibitors against CV-A6 and other enterovirus infections.

The innate immune system is the first line of host defense against viral infections and is linked to the activation and programming of adaptive immune responses (33).The type I IFN production pathway is activated by pattern recognition receptors that sense pathogen-associated molecular patterns to antagonize viral invasion (34,35).Melanoma differentiation-associated gene 5 (MDA5) and retinoic acid-inducible gene I (RIG-I) are two major RNA sensors in human cells (36).Both contain a central DExD/H box ATPase/helicase domain that recognizes viral RNA, two N-terminal caspase activation and recruitment domains (CARDs), and a C-terminal regulatory/repression domain (37).Upon viral RNA infection, MDA5 and RIG-I undergo conformational alterations and interact with the adaptor mitochondrial antiviral signaling protein (MAVS, also known as IPS-1, VISA, or Cardif ) through their CARD domain.As a result, a complex containing the TRAF3and IκB-related kinases TBK1 and IKKε is recruited by activated MAVS, which in turn directly phosphorylates IFN-regulatory factor 3 (IRF3) and IRF7.Phosphorylated IRF3 and IRF7 then translocate from the cytoplasm to the nucleus to induce type I IFN production, interferon-stimulated genes (ISGs), and inflammatory responses (38)(39)(40)(41)(42). Accordingly, these RIG-I-like receptor (RLR)-sensing pathways play an important role in activating the IFN signaling pathway.
Although the production of type I IFN elevates the cellular antiviral state against RNA viral infection, many enteroviruses, such as CV-A6, EV-A71, CV-A16, and EV-D68, have evolved strategies to inhibit IFN activation pathways (43,44).Nonstructural 2C proteins from enteroviruses have also been shown to regulate the NF-κB pathway for viral replication and cell survival to evade host immune responses (45)(46)(47).However, little is known about the function of enteroviral 2C proteins in regulating the RLR pathway.
In the current study, we demonstrated that the 2C protein from CV-A6 (2C CV-A6 ) efficiently reduced IFN-β production in HEK293T cells by suppressing the activity of the IFNB promoter.Further tests showed that 2C CV-A6 interacted with both MDA5 and RIG-I and induced their degradation via proteases in the lysosomal pathway.Tests using 2C CV-A6 mutations suggested a correlation between 2C CV-A6 -mediated RIG-I/MDA5 reduction and 2C CV-A6 -triggered IFN-β reduction.Interestingly, MDA5 and RIG-I were recognized by 2C EV-A71 and 2C CV-B3 but not by 2C CV-A16 , indicating that repression of the RLR pathway might be an evolutionary feature of enteroviral 2C proteins.Finally, we determined that amino acid F28 is critical for 2C CV-A6 to reduce the protein levels of RIG/MDA5 and subsequent production of IFN-β, which plays a crucial role in regulating CV-A6 replication.Collectively, these results illustrate the importance of enteroviral 2C proteins in regulating RLR-sensing pathways and provide novel insights into enteroviral regulation of host antiviral responses.

2C protein of CV-A6 blocks IFN-β production
It is well known that the expression of interferons in mammalian cell lines can be stimulated by the RNA virus Sendai virus (SeV) (48,49).To examine whether 2C CV-A6 can antagonize IFN-β production, HEK293T cells were co-transfected with the reporter plasmid IFNB-Luc and the plasmid encoding 2C CV-A6 or an empty vector VR1012.At 24 h post-transfection, the cells were mock infected or infected with SeV for 12 h.As shown in Fig. 1A, expression of 2C CV-A6 significantly inhibited SeV-induced IFNB promoter activity.Additional experiments were performed without IFNB-Luc transfection, and the levels of endogenous IFNB mRNA and secreted IFN-β were examined.SeV enhanced the endogenous levels of IFNB mRNA and the extracellular levels of IFN-β, which were significantly reduced by 2C CV-A6 expression (Fig. 1B and C).We then tested 2C CV-A6 's effects on IFNB promoter activity in HEK293T cells without SeV infection.As shown in Fig. 1D, the IFNB promoter activity was repressed by the presence of 2C CV-A6 in a dose-dependent manner.Collectively, these results suggest that the CV-A6 2C protein regulates IFN-β production via regulation of the IFNB promoter.

CV-A6 2C mediates the degradation of MDA5 and RIG-I
SeV is a strong inducer of RLR-mediated IFN-β signaling (50).MDA5 and RIG-I belong to the RLR family and are considered the most important cytosolic viral RNA sensors (51).Interestingly, in addition to acting as sensors for detecting exogenous viral RNA, both MDA5 and RIG-I are induced by IFN activation (52).Consistently, by triggering IFN expression, SeV infection also increased the endogenous levels of MDA5 and RIG-I

Full-Length Text
Journal of Virology proteins (Fig. 2A and B).However, the expression of 2C CV-A6 induced the depletion of MDA5 and RIG-I, suggesting that these two sensors are targets of 2C CV-A6 .
To confirm this hypothesis, HEK293T cells were co-transfected with an expression plasmid encoding full-length MDA5 or RIG-I, along with CV-A6-2C-wt-HA or an empty vector.As shown in Fig. 2C and D, overexpression of 2C CV-A6 resulted in significant degradation of either MDA5 or RIG-I proteins.Thus, 2C CV-A6 suppresses the host IFN signaling system through destabilization of MDA5 and RIG-I.

CV-A6 2C protein reduces the expression of MDA5 and RIG-I via the proteases in the lysosomal pathway
The ubiquitin-proteasome and lysosome pathways are the two main routes for protein and organelle clearance in eukaryotic cells (53,54).Saeed et al. reported the first comprehensive view of host protein digestion in cells infected with five representative enteroviruses, revealing enteroviral mechanisms of immune evasion, resource exploita tion, and pathogenesis (55).To further investigate the mechanism underlying 2C-medi ated MDA5 and RIG-I depletion, HEK293T cells were transfected with CV-A6 2C and MDA5 or RIG-I expression plasmids and subjected to western blotting.As shown in Fig. 3A and B, no obvious increase in the smaller band was detected for MDA5 or RIG-I in the presence of 2C CV-A6 , suggesting that the 2C CV-A6 -mediated MDA5 and RIG-I reduction was not due to 2C CV-A6 -induced protein cleavage.cells were transfected with 2C-wt-HA or VR1012 empty vector.Twenty-four hours after transfection, cells were infected with SeV (20 HA units/mL) for 12 h or left uninfected, and then collected for western blotting using the indicated antibodies.Tubulin was used as a loading control in all western blot assays.

Full-Length Text
Journal of Virology MG132, an inhibitor of proteasome-mediated proteolysis, and Baf-A1, an inhibitor of the autosome-lysosome pathway, were tested for their effects on 2C CV-A6 -mediated MDA5 and RIG-I degradation.Western blotting showed that Baf-A1, but not MG132, rescued both MDA5 and RIG-I from 2C CV-A6 -induced depletion (Fig. 3C and D), which was further confirmed using Baf-A1 at different concentrations (Fig. 3E and F).We tested two additional lysosomal protease inhibitors, E-64d and Pepstatin A, in the degradation experiment.As shown in Fig. 3G and H, the degradation of MDA5 and RIG-I by 2C-wt could be restored by the addition of E-64d and Pepstatin A in a dose-dependent manner.Taken together, these results revealed that 2C CV-A6 plays a key role in preventing MDA5 or RIG-I degradation by proteases in the lysosomal pathway.
Target protein interaction is essential for the autosome-lysosome pathway (58); thus, we tested whether 2C CV-A6 could interact with MDA5 or RIG-I using co-immunoprecipi tation (co-IP) assays.Western blotting of the eluted samples showed that both MDA5 and RIG-I immunoprecipitated with HA-tagged 2C CV-A6 , whereas no MDA5 or RIG-I was pulled down in the negative control group (Fig. 3I and J).Moreover, the interaction of FIG 3 2C CV-A6 protein reduces the expression levels of MDA5 and RIG-I through the proteases in the lysosomal pathway.(A and B) HEK293T cells were transfected with 2C-wt-HA, together with MDA5-no-tag (A) or RIG-I-no-tag (B) plasmids.Forty-eight hours after transfection, the cells were subjected to western blotting using the indicated antibodies.The entire membrane was visualized to detect potential digestion products of MDA5 or RIG-I.The multi-band phenomena detected for MDA5 in panel A is similar to previous reported observations (56,57).(C and D) VR1012-2C-wt-HA and MDA5-no-tag (C) or RIG-I-no-tag (D) plasmids were co-transfected into HEK293T cells for 24 h and then treated with either the proteasomal inhibitor MG132 (20 µM), the autophagy-lysosome inhibitor Baf-A1 (10 nM), or dimethyl sulfoxide (DMSO) (vehicle control) for a further 16 h.Whole-cell lysates were prepared for western blot analysis with the indicated antibodies.(E and F) HEK293T cells were transfected with plasmids to express MDA5-no-tag (E) or RIG-I-no-tag (F) alone, or co-transfected with a plasmid-expressing 2C-wt-HA for 24 h, and then treated with increasing amounts of Baf-A1 (3, 10, or 30 nM) for another 16 h.Whole-cell lysates were prepared 40 h post-transfection for western blot analyses with the indicated antibodies.(G and H) HEK293T cells were transfected as in E and F. Twenty-four hours post-transfection, the cells were then treated with increasing amounts of E-64d and Pepstatin A (3, 10, or 30 µg/mL) for another 12 h.Whole-cell lysates were prepared 36 h post-transfection for western blot analyses with the indicated antibodies.(I and J) HEK293T cells in six-well plates were co-transfected with 2C-wt-HA and MDA5-myc (I) or RIG-I-myc (J), or transfected with 2C-wt-HA, MDA5, or RIG-I plasmids alone.Forty-eight hours after transfection, cells were harvested, and the cell lysates were treated with RNase or left untreated and incubated with anti-HA beads at 4°C overnight.Protein expression was detected by western blotting using the indicated antibodies.The expression of tubulin was used as a protein control in all western blot analyses.
MDA5 or RIG-I with 2C CV-A6 was not disrupted by RNase treatment, indicating that the specific interaction between 2C CV-A6 and MDA5 or RIG-I was RNA independent (Fig. 3I  and J).These results indicate that 2C CV-A6 interacts with MDA5 and RIG-I and induces their depletion via proteases in the lysosomal pathway.

Different enterovirus 2C proteins distinctively regulate MDA5/RIG-I stability and IFNB promoter activity
As the enteroviral 2C protein is one of the most conserved and complex nonstruc tural protein (12), it is reasonable to speculate that other enteroviral 2C proteins also degrade MDA5 and RIG-I and consequently suppress IFNB promoter activity.To test this hypothesis, we first constructed three additional plasmids expressing 2C proteins from other prevalent enteroviruses, namely CV-A16, EV-A71, and CV-B3.As shown in Fig. 4A  and B, similar to 2C CV-A6 , the 2C proteins from EV-A71 (2C EV-A71 ) and CV-B3 (2C CV-B3 ) degraded both MDA5 and RIG-I, whereas the 2C protein from CV-A16 (2C CV-A16 ) did not.Further tests indicated that 2C EV-A71 and 2C CV-B3 , but not 2C CV-A16 , significantly suppressed IFNB promoter-driven luciferase expression in HEK293T cells (Fig. 4C).This finding validated our previous hypothesis that 2C regulates the IFN signaling pathway through MDA5/RIG-I depletion.We then investigated why CV-A16 2C failed to deplete MDA5/RIG-I and tested its interaction with these two sensors through co-IP experiments.As expected, 2C CV-A16 almost lost its ability to interact with MDA5 or RIG-I, whereas 2C EV-A71 and 2C CV-B3 were fully capable of interacting (Fig. 4D and E).Such a failure in MDA5/RIG-I depletion may indicate an evolutionary divergence of 2C CV-A16 ; in other words, 2C CV-A16 may have evolved to perform an alternative function that promotes viral replication.These results indicate that MDA5/RIG-I recognition is necessary, but not sufficient, for enteroviral 2C-mediated degradation.(60).To map the mini mal functional region of 2C responsible for MDA5 and RIG-I degradation, a C-terminal truncation was first performed on 2C CV-A6 (Fig. 5A).Luciferase assay results showed that the N-terminal 1-113 fragment of 2C CV-A6 was sufficient for IFNB promoter regulation (Fig. 5B), suggesting that this fragment contains essential residues for MDA5/RIG-I degradation.As expected, 2C CV-A6 1-113 efficiently reduced MDA5 and RIG-I expression (Fig. 5C and D).
It was then easy to speculate that the differences in this region (1-113) might explain why 2C CV-A16 failed to trigger MDA5/RIG-I depletion.Protein sequence alignment showed that 2C CV-A6 and 2C CV-A16 share different amino acid residues at positions 13, 23, 41, 75, and 96 in this region (Fig. 5E).Point mutations were then introduced in 2C CV-A6 by replacing these residues with the corresponding residues from 2C CV-A16 .Subsequent test results suggested that although mutations such as S13N, N23S, and R41K mildly disrupted the 2C CV-A6 -mediated MDA5/RIG-I reduction, V75A and I96V significantly compromised 2C CV-A6 's ability to deplete MDA5/RIG-I (Fig. 5F and G).Consistently, both V75A and I96V disrupted the interaction between 2C CV-A6 and MDA5/RIG-I (Fig. 5H and  I).These data indicate that these two residues are essential for MDA5/RIG-I recognition, explaining 2C CV-A16 's inefficient MDA5/RIG-I binding and reduction.
To further examine the N-terminal region at higher resolution, five serial truncation plasmids between amino acids 21 and 45 of 2C were generated and tested.Western blotting and luciferase assay results showed that the 2C CV-A6 Δ21-25, Δ31-35, Δ36-40, and Δ41-45 mutants remained effective in inducing MDA5/RIG-I degradation and inhibiting IFNB promoter activity, whereas Δ26-30 did not (Fig. 7A through C).Sub sequently, single alanine substitution mutants within amino acids 26-30 of 2C were constructed and tested.Western blotting and luciferase assay results revealed that F28A, but not the other amino acid substitutions, significantly compromised the ability of 2C CV-A6 to degrade MDA5 and RIG-I (Fig. 7D and E).Consequently, the same muta tion reduced 2C CV-A6 's efficiency of repressing IFNB promoter activity (Fig. 7F).Taken together, these results indicate that amino acid F28 plays an important role in the 2C CV-A6 -mediated regulation of RLR signaling pathways.
We then determined why the above 2C CV-A6 mutants (i.e., 46-329, Δ26-30, and F28A) failed to deplete MDA5/RIG-I.Myc-tagged MDA5 or RIG-I expression plasmids were co-transfected into HEK293T cells with a vector-expressing wild-type 2C CV-A6 or one of the above mutants.Exogenously expressed wild-type 2C CV-A6 or its mutants were isolated through co-IP experiments, and bound MDA5-myc or RIG-I-myc was detected by western blotting.Surprisingly, as shown in Fig. 8, all the tested mutants (i.e., 46-329, Δ26-30, and F28A) remained effective in MDA5/RIG-I recognition.Notably, these data correlated with the fact that all the tested mutants contained V75 and I96, which have been confirmed to be essential for MDA5/RIG-I binding (Fig. 5H and I).Altogether, this suggests that F28 is not involved in target protein interactions but might be crucial for 2C CV-A6 to introduce target proteins to the subsequent lysosomal pathway for degrada tion.

2C-mediated MDA5/RIG-I depletion plays an important role in CV-A6 replication
To confirm the effect of 2C-mediated MDA5/RIG-I depletion on enteroviral replication, we first tested whether altered IFN-β levels affect CV-A6 proliferation in RD cells.Different amounts of the VR-IFN-β-myc plasmid were transfected into RD cells, which were then infected with wild-type CV-A6 virus 24 h post-transfection.The cells were collected 48 h post-infection and subjected to western blotting.The results showed that ectopic expression of IFN-β significantly inhibited the replication of wild-type CV-A6 in a dose-dependent manner (Fig. 9A).We then determined whether the F28A mutation in 2C caused differences in endogenous IFN-β levels in CV-A6-infected cells.A proviral vector-expressing wild-type CV-A6 was produced as previously described (61) and used to generate a plasmid-expressing CV-A6 containing the F28A mutation in its 2C protein (CV-A6 2C F28A ).RD cells were then infected with rescued wild-type CV-A6 or CV-A6 2C F28A , the dosages of which were carefully adjusted to achieve similar viral expression when the infected cells were collected 24 h post-infection (as indicated by both RNA and protein levels of VP1 in Fig. 9B).However, CV-A6 2C F28A triggered more IFNB transcription than CV-A6 (P < 0.001, Fig. 9B).This confirmed that 2C suppresses endogenous IFN-β production in CV-A6-infected cells.
These data also indicate that reducing endogenous IFN-β levels may contribute to the 2C-mediated promotion of enteroviral replication.To test this hypothesis, RD cells were infected with equal amounts of wild-type CV-A6 or CV-A6 2C F28A , and a one-step growth assay was performed at 2-h intervals as previously described (62).As shown in Fig. 9C, the quantitative reverse transcription PCR (qRT-PCR) results targeting CV-A6 genomic RNA indicated that the levels of wild-type CV-A6 were consistently higher than those of CV-A6 2C F28A within 20 h of infection.Images after prolonged infection showed that wild-type CV-A6 caused cytopathogenic effect (CPE) in RD cells 72 h post-infection, which CV-A6 2C F28A failed to achieve (Fig. 9D).Western blotting of infected cells also confirmed that the viral replication efficiency was compromised in RD cells infected with CV-A6 2C F28A (Fig. 9E).These results indicate that the 2C-mediated suppression of RLR signaling pathways contributes to innate immune regulation during CV-A6 infection, providing an additional mechanism to enhance viral replication.

DISCUSSION
Generally, MDA5 and RIG-I are the main sensors of RNA viruses that trigger type I IFN production.Similar to RNA viruses, enteroviruses have evolved multiple mechanisms to suppress the RNA-sensing pathways in host cells.For instance, enteroviral nonstructural proteins such as 2A, 3C, and 3D induce dysfunction and/or destabilization of MDA5 and RIG-I (43,56,63,64).In this study, we report that another enteroviral nonstructural protein, 2C from CV-A6, is also capable of triggering MDA5 and RIG-I depletion, resulting in a decrease in IFNB promoter activity and IFN-β production (Fig. 10).Subsequent results indicated that 2C CV-A6 interacts with MDA5 and RIG-I in an RNA-independent manner and induces their degradation via the proteases in the lysosomal pathway.Surprisingly, although 2C proteins from EV-A71 and CV-B3 were also capable of MDA5/ RIG-I degradation, 2C from CV-A16 failed to do so because the latter did not recog nize MDA5 or RIG-I properly.Additional tests on truncated 2C CV-A6 suggested that the N-terminal 21-45 amino acid fragment is important for MDA5/RIG-I removal, which was further narrowed down to one specific amino acid residue, F28.Interestingly, the F28A mutation did not affect the interaction between 2C and MDA5 or RIG-I, suggesting that this specific position plays an essential role in the transfer of MDA5/RIG-I to the lysosomal pathway.Consequently, introducing the F28A mutation into CV-A6 significantly elevated IFNB mRNA levels in CV-A6-infected RD cells, leading to a reduction in viral infectivity in both short-and long-term infections.
At first glance, it is puzzling that both MDA5 and RIG-I are downregulated by enteroviral 2C because, despite their structural and functional similarities, MDA5, but not RIG-I, has been reported to be critical for sensing enteroviral infection (65).In the presence of EV-A71 RNA, MDA5 triggers the activation of IRF3, which plays an important role in IFN activation (65).Therefore, it is reasonable for enteroviruses to counteract the function and/or stability of MDA5.However, it should be noted that in addition to being sensors, both MDA5 and RIG-I are effectors of the IFN signaling pathway (i.e., the genes coding for MDA5 and RIG-I are ISGs) (52).Moreover, introducing exogenous MDA5 or RIG-I alone can induce innate immune activation even in the absence of viral RNA (or other stimuli) (66,67).Therefore, MDA5/RIG-I and IFN form a positive feedback loop during innate immune activation.Accordingly, depletion of both MDA5 and RIG-I would result in a lower level of IFN activation than the removal of either alone.This is not the first report on enteroviral proteins that suppress RIG-I.For example, EV-A71 infection can compromise RIG-I ubiquitination and consequently reduce IFN production (68).By removing both MDA5 and RIG-I, 2C inhibits IFN activation more effectively, which contributes to 2C's critical role in enteroviral infection.
Surprisingly, the 2C protein from CV-A16, an enterovirus that often causes HFMD pandemics (5,(69)(70)(71), does not recognize MDA5/RIG-I and trigger their degradation.2C CV-A16 and 2C CV-A6 share nine different amino acids at positions 13, 23, 41, 75, 96, 227, 271, 303, and 307.However, no single-point mutation at these sites could grant 2C CV-A16 the ability to destabilize MDA5 or RIG-I (data not shown), suggesting that complex cooperation among some or all these residues is essential for MDA5/RIG-I recognition and/or depletion.Indeed, during the revision of this manuscript, positions 75 and 96 on 2C CV-A6 and 2C CV-A16 were both determined crucial for MDA5/RIG-I binding and subsequent downregulation.Interestingly, for a 2C protein functional in MDA5/RIG-I degradation, the residues on positions 75 and 96 are not exclusively restricted.In fact, tested 2C CV-A6 , 2C EV-A71 , and 2C CV-B3 contain V75/I96, V75/T96, and Q75/A96, respectively, while all these three enteroviral proteins were capable of MDA5/ RIG-I removal.Furthermore, it should be noticed that, removing MDA5/RIG-I might not be a conserved feature for all EV-A71, CV-A6, or CV-B3 2C proteins.For example, among 1,602 EV-A71 2C protein sequences retrieved from the NCBI Protein Database (https://www.ncbi.nlm.nih.gov/protein),527 contain inactive residue A75.On the other hand, although most retrieved CV-A16 2C proteins are believed inactive in MDA5/RIG-I depletion as they contain A75 and V96, there are 4 out of 473 CV-A16 2C proteins that do have functional residues at both positions 75 and 96 and might be capable to trigger MDA5/RIG-I degradation.In general, it appears that positions 75 and 96 are probably under selection pressure, resulting in MDA5/RIG-I depletion as an evolutionar ily selective instead of a conserved function for enteroviral 2C, which might provide possible explanations on the distinctive virulence observed for different enteroviral subtypes and even strains.
On the contrary, protein sequence alignment suggested that F28 is highly conserved among 2C proteins from different types of enteroviruses (99.78% or 3,211 out of 3,218 strains; data not shown).To our knowledge, F28, as a single amino acid residue, has not been revealed to be critical for any known function of the enteroviral 2C protein.Notably, the regions around F28 are essential for most of the known functions of 2C, including residues 1-38 for oligomerization, 21-45 for RNA binding, and 21-54 for membrane binding.These regions play important roles in supporting 2C in host cell membrane rearrangements, RNA replication, encapsidation, morphogenesis, and ATPase activity [reviewed in reference (32)].Recently, Li et al. reported that fragment 1-53 is sufficient for EV-A71 2C-mediated APOBEC3G depletion, which in turn promotes viral replication (60).Meanwhile, fragment 1-113 of 2C CV-A6 is competent in reducing MDA5/ RIG-I levels and subsequent IFN production, as revealed in this study.These findings indicate that blocking the N-terminal domain of enteroviral 2C may be an effective strategy to suppress enteroviral replication.Our data showed that F28A mutation alone significantly compromised the replication of CV-A6 (Fig. 9), further supporting the idea that enteroviral 2C is a promising target for the development of anti-enteroviral drugs.
The C-terminal myc-tagged full-length MDA5 and RIG-I expression vectors have been previously described (73).Subsequently, full-length MDA5-and RIG-I-expressing plasmids without a tag were constructed based on the corresponding myc-tagged plasmids.
To generate a firefly luciferase-based IFNB promoter reporter plasmid (IFNB-Luc), a 2,020-bp-long fragment upstream of the IFN-β-coding region was amplified using cDNA from Jurkat cells and inserted into the pGL3-Basic vector (Promega, Madison, WI, USA) via the XhoI and HindIII sites.The IFN-β open reading frame was obtained from total RNA extracted from HEK293T cells and inserted into the SalI and XbaI sites of the VR1012 vector with a C-terminal myc tag.
All primers are shown in Table 1.

Construction of infectious CV-A6 clone
To construct the plasmid of the infectious clone of the CV-A6-Changchun098 strain, RNA was extracted from the supernatant of virus-infected RD cells and then reverse transcribed to obtain cDNA.The resulting cDNA was used to amplify three viral fragments covering the full-length CV-A6-Changchun098 genome, with a T7 pro moter and a poly A tail at the N-and C-termini (61).These PCR products were cloned into pBlueScript SK(+) (YouBio, VT1892, China) at the XhoI and BamHI sites using the pEASY-Basic Seamless Cloning and Assembly Kit (TransGen, China) accord ing to the manufacturer's instructions and designated as pBSK-CV-A6-wt.The singlepoint mutation infectious clone, pBSK-CV-A6-2C-F28A, was obtained by site-directed mutagenesis using 2× Phanta Max Master Mix (Vazyme, P515-01, Nanjing, China), with pBSK-CV-A6-wt as the template.All the infectious clones were confirmed by sequencing the entire genome.
To rescue viruses, HEK293T cells seeded in a T25 flask were co-transfected with linearized pBSK-CV-A6-wt (4 µg) or mutant infectious clone, which had been digested with BamHI, and pcDNA3.1-T7DNA Pol [2 µg, kindly provided by Dr. Zhaolong Li (Jilin University, China)] using Lipofectamine 3000 reagent according to the manufacturer's protocol (61,74).At 6 h post-transfection, the cells were washed, and the medium was replaced with fresh medium containing 10% FBS.When approximately 90% of the cells exhibited CPE, the media and cells were harvested together, frozen at −80°C, and  thawed.Freeze thaw was repeated three times to release the viruses.The samples were then centrifuged at 12,000 g for 30 min at 4°C to remove cellular debris, and the clear supernatant was then transferred to a new tube, aliquoted, and stored at −80°C.

Luciferase reporter assay
The TransDetect Single-Luciferase Reporter Assay Kit (TransGen, FR101-01) was used to detect the activity of the IFNB promoter.Briefly, HEK293T cells in 24-well plates were co-transfected with 250 ng IFNB-Luc and the indicated expression plasmid or empty vector.At 24 h post-transfection, the cells were infected with SeV [20 hemag glutination (HA) units/mL] or left uninfected for 12 h.At 36 h post-transfection, the cells were washed once with PBS and lysed to measure luciferase activity, according to the manufacturer's protocol.The readings of pGL3-transfected samples were used for background correction and are not shown.

Enzyme-linked immunosorbent assay
IFN-β secretion in the cell culture medium was assessed using a human IFN-β enzymelinked immunosorbent assay kit (Shanghai MLBIO Biotechnology, Shanghai, China) according to the manufacturer's instructions.

Co-immunoprecipitation
The co-IP experiments were performed as previously reported with minor modifications (75).HEK293T cells in six-well plates were transfected with 2C-(400 ng), MDA5-, or RIG-I-expressing vectors (1,200 ng) and then harvested at 48 h post-transfection.The cells were then washed once with cold 1× PBS, followed by disruption with lysis buffer [50 mM Tris-HCl (pH 7.5), 150 mM NaCl, and 0.5% NP-40, supplemented with complete protease inhibitor cocktail (Roche)].The samples were sonicated at 15% power for 90 s with a 1 s break every 1 s and then centrifuged at 12,000 g for 30 min at 4°C to obtain a clear supernatant.These input samples were then incubated with anti-HA magnetic beads (Thermo Pierce, 88837) overnight and washed six times with cold wash buffer [20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 0.1 mM EDTA, and 0.05% Tween 20].The samples were then eluted with cold 100 mM glycine-HCl (pH 2.5) and analyzed by western blotting to detect the proteins of interest.

Quantitative reverse transcription PCR
Total RNA was extracted from cells using the FastPure Cell/Tissue Total RNA Isolation Kit (Vazyme, RC101) and then subjected to reverse transcription with MonScript RTIII All-in-One Mix (Monad, RN05004M, Suzhou, China) according to the manufacturer's instructions.SYBR Green-based qPCR was performed on a Cobas z480 instrument (Roche, Rotkreuz, Switzerland) using MonAmp ChemoHS qPCR Mix (Monad, RN04001N) and specific primers.Each 20 µL of reaction mixture contained 4 µL of 5 × RTIII All-in-One Mix, 0.4 µL of oligonucleotide primer (10 µM each), and 2 µL of the cDNA template.The cycling conditions were as follows: 50°C for 2 min and 95°C for 10 min, followed by 45 cycles consisting of 95°C for 30 s and 60°C for 1 min.Relative mRNA levels were normalized to the level of ACTB mRNA (data not shown).All the primers used for qRT-PCR are listed in Table 1.

Viral infection and one-step growth curves
Cells were infected with the virus at the indicated multiplicity of infection.The unbound virus was removed 2 h after absorption, and the cells were washed twice in PBS (pH 7.4).Fresh DMEM containing 2% FBS was added, and the cells were incubated at 37°C in 5% CO 2 for the indicated time points.The medium of each sample was used to extract RNA for qRT-PCR, and the cell lysate was used for detecting VP1 expression by western blotting.
One-step growth assays were performed as previously described (62).

Statistical analysis
All data are presented as the mean ± SD from three independent experiments.Data were analyzed using unpaired two-tailed Student's t-tests in Microsoft Excel.All plots were analyzed using GraphPad Prism 8.0.2 (GraphPad Software, Inc., San Diego, CA, USA).
Significance was set at P < 0.05.

FIG 1
FIG 1 2C CV-A6 blocks the production of IFN-β.(A) Schematic of pGL3-IFNB-Luc vector.HEK293T cells were co-transfected with IFNB-Luc and 2C-wt-HA or VR1012 plasmid.Twenty-four hours after transfection, cells were left untreated or infected with SeV [20 hemagglutination (HA) units/mL] for 12 h, and the cell lysates were then assayed for luciferase activity.The expression levels of 2C protein were confirmed by western blot with an anti-HA antibody.(B and C) HEK293T cells were transfected with 2C-wt-HA or corresponding amounts of empty vector.Twenty-four hours after transfection, cells were infected with SeV (20 HA units/mL) for 12 h or were left uninfected, and were then collected to extract total RNA.The expression levels of endogenous IFNB mRNA were analyzed via quantitative reverse transcription PCR (B).The culture medium was harvested to detect IFN-β secretion using enzyme-linked immunosorbent assay (C).(D) HEK293T cells were co-transfected with IFNB-Luc (200 ng) and increasing amounts of plasmid 2C-wt-HA (50, 150, or 450 ng) or empty vector.Forty-eight hours after transfection, cells were collected for luciferase activity assay.Tubulin was used as a loading control in all western blot assays.

FIG 2
FIG 2 2C CV-A6 degrades both exogenous and endogenous MDA5 and RIG-I proteins.(A and B) HEK293T cells were co-transfected with 2C-wt-HA and MDA5-no-tag (A) or RIG-I-no-tag (B) expression plasmids.After 48 h, cells were collected for western blotting using the indicated antibodies.(C and D) HEK293T

FIG 4
FIG 4 Effects of multiple enteroviral 2C proteins on MDA5 and RIG-I degradation, and IFNB promoter activity.(A and B) Four enterovirus 2C-wt-HA expression plasmids were co-transfected with MDA5-no-tag (A) or RIG-I-no-tag (B) into HEK293T cells for 48 h.The cells were collected and subjected to western blot assays using the indicated antibodies.(C) HEK293T cells were co-transfected with IFNB-Luc plasmid, together with four wild-type enterovirus 2C expression plasmids for 48 h; the cells were then subjected to a luciferase activity assay.The expression levels of 2C proteins were detected by western blotting with anti-HA and anti-tubulin antibodies.(D and E) HEK293T cells in six-well plates were transfected with plasmids encoding MDA5-myc (D) or RIG-I-myc (E), together with one of four 2C-wt-HA expression plasmids.Forty-eight hours later, cells were harvested, and cell lysates were incubated with anti-HA beads at 4°C overnight.The input and eluted samples were detected by western blotting using the indicated antibodies, with tubulin as a loading control.

FIG 5
FIG 5 Amino acids V75 and I96 of 2C CV-A6 are crucial for the interaction with and degradation of MDA5 and RIG-I.(A) Schematic of 2C CV-A6 N-terminal truncation constructs.(B) IFNB-Luc and 2C-wt-HA or the truncation mutant were co-transfected into HEK293T cells; the cells were then subjected to luciferase assays at 48 h post-transfection.The expression of 2C-wt and truncated protein was confirmed by western blotting with an anti-HA antibody.(C and D) MDA5 (C) or RIG-I (D) was co-transfected with wild-type or five mutated 2C-HA expression constructs into 24-well plates of HEK293T cells.Forty-eight hours post-transfection, cells were harvested for measurement of MDA5 and RIG-I expression by western blot analyses.(E) Sequence alignment of 1-113 fragment of 2C CV-A6 and 2C CV-A16 .(F and G) Wild-type or five mutated CV-A6-2C-HA expression plasmids were co-transfected with MDA5-no-tag (F) or RIG-I-no-tag (G) into HEK293T cells for 48 h.The cells were collected and subjected to western blot assay using the indicated antibodies.(H and I) HEK293T cells were co-transfected in six-well plates with MDA5-myc (H) or RIG-I-myc (I), together with wild-type or five mutated CV-A6-2C-HA expression plasmids, as indicated, and were harvested at 48 h post-transfection.The cells were then subjected to an immunoprecipitation assay with anti-HA beads.Input and eluted samples were detected by western blotting.Tubulin was used as a loading control.

FIG 9
FIG 92CCV-A6 plays an important role in IFN-β regulation and viral replication during CV-A6 infection.(A) RD cells were transfected with increasing amounts (30, 100, or 300 ng) of VR-IFN-β-myc plasmid in 24-well plates.Twenty-four hours after transfection, the cells were infected with equal amounts of CV-A6 recombinant virus for another 48 h.Cell lysates were collected for western blot analysis using the indicated antibodies.Tubulin antibody was used as a protein-loading control.(B) RD cells in 24-well plates were infected with adjusted amounts of CV-A6-and CV-A6 2C F28A -rescued viruses to achieve equal infection with both viruses.Twenty-four hours post-infection, the cells were collected.Half of the whole-cell lysate was used to extract RNA for quantitative reverse transcription PCR (qRT-PCR) to detect VP1 mRNA levels, and the other half was subjected to western blotting to confirm CV-A6 VP1 expression.Both results were used to confirm equal infection with both viruses.Endogenous IFNB mRNA levels were detected using qRT-PCR, with levels of ACTB mRNA used as a loading control.(C) One-step growth curves of the CV-A6 and CV-A6 2C F28A recombinant viruses.Based on genome quantification by qRT-PCR, similar amounts of these viruses were used to infect RD cells.After 6 h of adsorption, the infected cells were washed once with PBS, the medium was replaced with maintenance medium, and the levels of released viruses in the medium were determined by qRT-PCR at 2-h intervals, starting at 8 h post-adsorption.(D and E) RD cells were infected with equal amounts of CV-A6 or CV-A6 2C F28A recombinant virus.At 72 h after infection, the cells were observed and photographed using a light microscope (D).Whole-cell lysates were collected for western blotting using an anti-CV-A6 VP1 antibody (E).

FIG 10
FIG 10 Proposed model of enterovirus 2C protein-mediated inhibition of MDA5/RIG-I signaling pathway.After enterovirus infection, the nonstructural 2C protein degrades MDA5 and RIG-I through the lysosomal pathway and thus inhibits activation of the IFNB promoter, ultimately blocking IFN-β production.However, F28A mutation compromises 2C's ability to inhibit type I IFN signaling.This diagram was created with BioRender.com.

a
The restriction enzyme sites used for cloning are indicated in bold and italic.The C-terminal HA or myc tag sequence in the primers is underlined.

TABLE 1
Primers used in this study a (Continued)