key: cord-0861492-phv08xe2
authors: Mukhopadhyay, Urbi; Chanda, Shampa; Patra, Upayan; Mukherjee, Anupam; Komoto, Satoshi; Chawla‐Sarkar, Mamta
title: Biphasic regulation of RNA interference during rotavirus infection by modulation of Argonaute2
date: 2019-08-26
journal: Cell Microbiol
DOI: 10.1111/cmi.13101
sha: f36f9dad489b1a29c4821d8ac3201766e33e5fc8
doc_id: 861492
cord_uid: phv08xe2

RNA interference (RNAi) is an evolutionary ancient innate immune response in plants, nematodes, and arthropods providing natural protection against viral infection. Viruses have also gained counter‐defensive measures by producing virulence determinants called viral‐suppressors‐of‐RNAi (VSRs). Interestingly, in spite of dominance of interferon‐based immunity over RNAi in somatic cells of higher vertebrates, recent reports are accumulating in favour of retention of the antiviral nature of RNAi in mammalian cells. The present study focuses on the modulation of intracellular RNAi during infection with rotavirus (RV), an enteric virus with double‐stranded RNA genome. Intriguingly, a time point‐dependent bimodal regulation of RNAi was observed in RV‐infected cells, where short interfering RNA (siRNA)‐based RNAi was rendered non‐functional during early hours of infection only to be reinstated fully beyond that early infection stage. Subsequent investigations revealed RV nonstructural protein 1 to serve as a putative VSR by associating with and triggering degradation of Argonaute2 (AGO2), the prime effector of siRNA‐mediated RNAi, via ubiquitin–proteasome pathway. The proviral significance of AGO2 degradation was further confirmed when ectopic overexpression of AGO2 significantly reduced RV infection. Cumulatively, the current study presents a unique modulation of host RNAi during RV infection, highlighting the importance of antiviral RNAi in mammalian cells.

to its target. The ultimate fate of the target mRNA is determined by the degree of complementarity between the guide strand and the target sequence; perfect complementarity (for siRNAs) triggers cleavage (slicing) of the mRNA whereas imperfect complementarity (generally for miRNAs) sequesters mRNA away from polyribosomes into cytoplasmic processing bodies (P bodies) resulting in translational repression. At the catalytic core of RISC are the Argonaute (AGO) family of proteins of which mammals encode four homologs (AGO1-4). In spite of evidences of functional redundancy among the four AGOs, slicer activity is exhibited only by the isoform AGO2 in mammalian cells (Bartel, 2009; Bartel, 2018; Siomi & Siomi, 2009) .

During viral infection, replicative or transcriptive viral dsRNA intermediates may succumb to processing by host cellular RNAi yielding virus-derived small interfering RNAs (viRNAs), which might further direct viral RNA cleavage ultimately resulting in heavily attenuated viral replication. As a coevolutionary counterstrategy, viruses have adopted RNAi-evasive virulence measures in the form of producing virus-encoded-suppressors-of-RNAi (Bivalkar-Mehla et al., 2011; Burgyán & Havelda, 2011; Ding, Han, Wang, & Li, 2018; Wu, Wang, & Ding, 2010) . Experimental evidence of viRNA generation during the course of natural viral infection is abundant in plants, worms, and insects (Siu et al., 2011) . Physiological relevance of RNAi as an antiviral innate immunity, however, has been greatly debated in differentiated, somatic cells of higher vertebrates because of the evolution of a superior, interferon (IFN)-based innate immune system (Cullen & Cherry, 2013; Goubau, Deddouche, & e Sousa, 2013; Maillard et al., 2016; Samuel, 2001; Schneider, Chevillotte, & Rice, 2014; ten Oever, 2016) . Alternatively, mammalian viruses, at least in some contexts, have been experimentally shown to exert strong RNAi evasive mechanisms by encoding highly potent VSRs (Li, Lu, Han, Fan, & Ding, 2013; Qiu et al., 2017) , justifying near absence of viRNAs in mammalian cells during natural course of infection (Parameswaran et al., 2010) . Mammalian VSRs characterised till date include ebola virus VP35 (Haasnoot et al., 2007) , influenza A virus NS1 (Bucher, Hemmes, de Haan, Goldbach, & Prins, 2004) , vaccinia virus E3L (Li et al., 2004) , nodamura virus B2 (Johnson, Price, Eckerle, & Ball, 2004 ), hepatitis C virus core (Wang et al., 2006) , HIV-1 Tat (Bennasser & Jeang, 2006; Bennasser, Le, Benkirane, & Jeang, 2005) , coronavirus nucleocapsid protein N (Cui et al., 2015) , and virus-associated RNAI and RNAII of adenoviruses (Andersson et al., 2005; Lu & Cullen, 2004) . VSRs are generally dsRNA-binding proteins (excepting adenoviral VSR), which inhibit activity of either Dicer, AGO, or both. Interestingly, all VSRs identified from mammalian viruses possess IFN or protein kinase R antagonistic properties and are essential for replication and pathogenesis, suggesting that RNAi and other innate antiviral responses are interrelated.

The genome of rotavirus (RV), a diarrheagenic virus in infants and young animals, consists of 11 segmented double-stranded RNAs, which are usually enclosed within viral double-layered particles and, at proliferative stage of infection, within highly dynamic encagements called viroplasms. Single-stranded viral mRNAs, produced within double-layered particles, are either translated into six structural (VP1 to VP4, VP6, and VP7) and six nonstructural (NSP1 to NSP6) proteins or used as templates for viral RNA replication (Desselberger, 2014; Estes & Greenberg, 2013) . However, oozing of rotaviral dsRNA intermediates in unmasked cytosolic environment has been hypothesised based on RNase III-sensitive and dsRNA-positive scattered signals in infected host cell cytoplasm (Rojas, Arias, & López, 2010; Zhu et al., 2017) . Accidental exposure of viral genomic dsRNAs or even transcriptive dsRNA intermediates in the cytoplasm may potentially evoke RNAi-based surveillance mechanism. Strikingly, previous reports emphasised retention of competent RNAi pathway in the context of RV infection enabling functionality of both small interfering RNAs (siRNAs) and short hairpin RNAs (shRNAs; Déctor, Romero, López, & Arias, 2002; Arias et al., 2004; López, Rojas, Ayala-Bretón, López, & Arias, 2005; Chen et al., 2011) . Indeed, successfully knocking down expression of viral and host proteins has been proven to be excellent loss-of-function studies to unravel complex host-RV interaction networks (Dhillon & DurgaRao, 2018; Mukherjee, Patra, Bhowmick, & Chawla-Sarkar, 2018; Oceguera, Peralta, Martínez-Delgado, Arias, & López, 2018; Silva-Ayala et al., 2013; Trujillo- Alonso, Maruri-Avidal, Arias, & Lopez, 2011; Yin et al., 2018) . Moreover, RNAi-based potential therapeutic approaches targeting RV genes have also been advocated (Arias et al., 2004) . Notably, in most of these experiments, RNAi functionality was assessed at relatively late hours of infection (beyond 6-hr postinfection).

While studying loss-of-function attributes of selected cellular determinants in context of RV-SA11 infection, intriguingly, we found differential sensitivity of cellular genes to RNAi at different time points of infection. RNAi pathway was found to be inhibited during early hours (2-6 hr) of RV-SA11 infection only to be restored to full functional competence beyond that early phase. Blocking of RNAi was confirmed in the form of ablated functioning of siRNA and shRNA but not miRNA. Interestingly, hindrance to RNAi during the early stage of RV-SA11 infection coincided with the loss of protein level of AGO2. Further mechanistic studies revealed that RV-NSP1 interacts with and degrades AGO2 in ubiquitin-proteasome-dependent way.

Together, the data highlight a fascinating and unique modulation of RNAi pathway in response to RV infection.

To assess the sensitivity of ectopic GFP expression to synthetic siRNA during early and late hours of RV infection, GFP (pEGFP-N1)-transfected MA104 cells were cotransfected with either siGFP or scrambled siRNA, and 36-hr posttransfection cells were further mock-infected or infected with RV-SA11 for 3 and 9 hr. Interestingly, analyses of GFP expression by immunoblotting showed siRNA-mediated targeted knock-down of GFP to be apparent in mock-infected cells as well as in cells infected with RV-SA11 for 9 hr but not in RV-SA11-infected cells harvested at 3-hr postinfection (3 hpi; Figure 1a ). We ruled out transfection of siGFP to have any off-target effects on overall RV infection as relative expression of RV structural and nonstructural proteins remained unperturbed ( Figure 1a ; Figure S1A ). Visualisation by immunofluorescence microscopy (10× magnification) in MA104 cells ( Figure 1b ) and HEK293 cells ( Figure S1B ) also revealed ectopic expression of GFP to be sensitive to RNAi only at 9 hpi but not at 3 hpi, suggesting time point-dependent differential regulation of RNAi during RV-SA11 infection. Rhodamine-conjugated NSP5 visualised in RV-infected samples under higher magnification (63× oil immersion) ensured progressive infection (Figure 1b, panel 5 ). Next, we investigated whether RNAi targeted against any endogenous cellular protein was also regulated similarly during early and late hours of RV-SA11 infection. MA104 cells, transfected with scrambled siRNA or RINGbox protein 1 (Rbx1)-specific siRNA for 36 hr, were subsequently mock-infected or infected with RV-SA11. Consistent to our previous results, Rbx1 expression was successfully knocked down in response to Rbx1 siRNA in RV-SA11-infected cells lysed at 9 hpi as well as in mock-infected control but not in RV-SA11-infected cells harvested FIGURE 1 Host RNA interference is blocked during early hours of RV-SA11 infection. (a) MA104 cells transfected with GFP (pEGFP-N1) were further cotransfected with scrambled siRNA or siGFP as indicated in the figure (designated by +, −); 36-hr posttransfection, cells were mock infected or infected with RV-SA11. After 3 and 9 hpi, cells were harvested, and GFP expression was analysed by western blot analysis. (b) GFP (pEGFP-N1)-transfected MA104 cells were cotreated with scrambled siRNA or siGFP as indicated in the figure (designated by +, −). Cells were fixed at indicated time points postinfection (mock/RV-SA11) with paraformaldehyde, permeabilised, blocked, and subsequently stained overnight with anti-NSP5 antibody (as a marker of virus infection). Secondary staining and mounting were performed with Rhodamine conjugated anti-rabbit antibody and DAPI, respectively. Cells were finally visualised by confocal microscopy (10×); Scale bar 100 μM. (Lower panel) Rhodamineconjugated NSP5 was visualised in RV-infected samples under higher magnification to ensure progressive infection (63× oil immersion). (c) MA104 cells were transfected with scrambled siRNA or Rbx1 siRNA. 36-hr posttransfection, cells were either mock infected or infected with RV-SA11 for indicated time points; Rbx1 expression was finally assessed by immunoblot analysis. (d) MA104 cells, transfected with either empty vector (PLKO.1-TRC) or RV-VP4-shRNA (cloned into PLKO.1-TRC vector) for 24 hr, were further infected with RV-SA11 for 6 and 12 hr. Expression of RV-VP4 in whole cellular extracts was checked by western blot analysis. (e) Scrambled miR/miR-29b mimic transfected MA104 cells were mock infected or infected with RV-SA11 for 3 and 9 hr. Cellular lysates prepared at indicated time points were subjected to western blot analysis to check expression of P85α subunit of PI3K. (f) pMIR-REPORT Luciferase construct containing the 3′-UTR of P85α-PI3K was transfected in HEK293 cells in association with miR-29b mimic (20 and 40 nM). 36-hr posttransfection, cells were either kept mock-infected or infected with RV-SA11 for 3 and 9 hr. Percentage of relative luciferase activity was finally represented after normalising Firefly luciferase with corresponding Renilla luciferase control at 3 hpi (Figure 1c ), further indicating hindrance of RNAi mechanism during early hours of RV-SA11 infection. Furthermore, sensitivity of RV protein expression to shRNA-mediated RNAi in infection scenario was analysed. MA104 cells expressing shRNA against VP4 for 24 hr were subsequently infected with RV-SA11, and expression of VP4 was checked after 6 and 12 hr of infection. In agreement with previous reports (Arias et al., 2004; Déctor et al., 2002) , expressions of VP4 RNA and protein were found to be silenced at 12 hpi ( Figure S1C; Figure 1d ), suggesting sensitivity of viral protein expression to RNAi at late phase of RV-SA11 infection. Interestingly, shRNA-mediated silencing of VP4 expression was attenuated at 6 hpi ( Figure S1C; Figure 1d ), affirming our previous observation that functional competence of siRNA is lost during early phase of RV-SA11 infection.

We further checked modulation of miRNA-mediated RNAi by evaluating target repression capacity of ectopically administered miRNA mimic during early and late hours of RV-SA11 infection. MA104 cells, transfected with either miR-29b mimic or scrambled miR for 36 hr, were subsequently infected with RV-SA11 for 3 and 9 hr. One set of transfected cells was mock infected. Interestingly, following miR-29b overexpression, reduction of miR-29b target P85α subunit of PI3K (Park, Lee, Ha, Nam, & Kim, 2009 ) was evident in both mockinfected cells and RV-infected cells for 3 (early time point) and 9 hr (late time point; Figure 1e ). Consistently, much alike to mock-infected group, sensitivity of luciferase expression from pMIR-REPORT Luciferase vector containing 3′-UTR of P85α subunit of PI3K in HEK293 cells to cotransfected miR-29b mimic (20 and 40 nM) was also found to be unaffected at both early (3 hpi) and late (9 hpi) hours of RV-SA11 infection ( Figure 1f ). Collectively, the results suggest that though inhibition of siRNA and shRNA-mediated RNAi occurs during early hours of RV-SA11 infection, functionality of miRNA is retained during the same infection window. Figure S2A ). However, AGO1 protein level was found to be reduced during late hours of RV-SA11 infection (9-12 hpi; Figure 2a ; Figure S2A ). Levels of AGO3 and AGO4, two other members of AGO family proteins in mammals, as well as DICER, remained unchanged with respect to their respective mock-infected controls throughout indicated time period (2-12 hpi) of RV-SA11 infection ( Figure 2a ; Figure S2A ). Confocal microscopy and subsequent quantitation of corrected total cell fluorescence (CTCF) also revealed AGO2 expression to get reduced at 3 hpi compared with the mock-infected control only to restore and be distinctly visible again at 9 hpi ( Figure 2b ). Intracellular distribution of AGO2, however, was found to be somewhat different at 9 hpi from that in mockinfected control (Figure 2b) . Furthermore, infection with UV-treated RV-SA11 (Smirnov, Kapitulets, Amitina, Ginevskaya, & Kaverin, 1991) had no effect on AGO2 protein levels postinfection ( Figure S2B ).

Together, the data suggest that actively replicating RV-SA11 triggers attenuation in protein levels of AGO2 leading to functional blocking of RNAi during early time points (2-6 hpi) of infection.

We further speculated involvement of viral protein(s) in mediating reduction of AGO2 protein levels during 2-6-hr post RV-SA11 infection. Owing to previous reports of RV-NSP1 being involved in triggering degradation of multitude of host proteins, which are regulators of host cellular innate immunity (Arnold, Barro, & Patton, 2013; Bagchi, Bhowmick, Nandi, Kant Nayak, & Chawla-Sarkar, 2013; Barro & Patton, 2007; Bhowmick, Halder, Chattopadhyay, Nayak, & Chawla-Sarkar, 2013; Bhowmick, Mukherjee, Patra, & Chawla-Sarkar, 2015; Graff, Ettayebi, & Hardy, 2009; Graff, Ewen, Ettayebi, & Hardy, 2007; Graff, Mitzel, Weisend, Flenniken, & Hardy, 2002; Qin et al., 2011) , the involvement of RV-NSP1 was investigated in the process. Gradual reduction of AGO2 protein levels in a dose-dependent manner in response to escalated concentration of transfected RV-NSP1 further ascertained the role of this viral NSP in modulating AGO2 levels ( Figure S3B ). Furthermore, infection with bovine RV strain A5-13 possessing wild-type NSP1 (Taniguchi, Kojima, & Urasawa, 1996) also resulted in reduced AGO2 protein levels (2-6 hpi) whereas corresponding NSP1 mutant RV strain A5-16 (Taniguchi et al., 1996) Figure S3D ).

In addition to the putative E3 ubiquitin ligase activity intrinsic to RV-NSP1, co-opted host E3 ubiquitin ligase complex Cullin has recently been implicated to be interacting with (Ding et al., 2016; Lutz, Pace, & Arnold, 2016) and be involved in mediating proteasomal degrada- Figure S3F ). Notably, ubiquitinated AGO2 levels were found to be low at 9 hpi compared with 3 hpi, justifying their corresponding cellular levels represented in the input lanes ( Figure S3F ). 

The observation that RV-NSP1 can cause proteasomal degradation of AGO2 prompted us to check whether RV-NSP1 can also function as a suppressor of RNA silencing. To evaluate this, Rbx1 protein expression was assessed in MA104 cells transfected with either Rbx1 siRNA alone or pcDNSP1 in addition to Rbx1 siRNA. Immunoblotting revealed inefficient knocking down of Rbx1 expression by siRNA in the presence of RV-NSP1 ( Figure 4a ). Sensitivity of ectopic GFP (pEGFP-N1) expression to siGFP was also reduced in RV-NSP1overexpressing cells ( Figure S4A ), indicating that RV-NSP1 might FIGURE 3 Rotaviral nonstructural protein 1 triggers ubiquitination and proteasomal degradation of AGO2. (a) Whole cellular extracts from pcDNSP1-and pcDNSP4-transfected MA104 cells were subjected to western blot analysis for checking expressions of AGO1, AGO2, Pan3, and Ser616 pDRP1. (b) MA104 cells were mock-infected or infected with RV-A5-13 and RV-A5-16 (2-12 hpi). AGO2 and P53 expression was checked in cellular extracts of infected cells prepared at indicated time points. NSP1 protein expression was analysed in the same lysates to confirm RV-A5-13 and RV-A5-16 infection. (c) MA104 cells were kept untransfected or transfected with pcDNSP1; 6-hr posttransfection, cells were treated with vehicle (DMSO) or 1-μM Lactacystin. Cellular lysates prepared after 24 hr of transfection were subjected to SDS-PAGE followed by immunoblotting and expression of AGO2 was checked. His protein expression was checked as a marker of pcDNSP1 transfection. (d) pcDNSP1-transfected MA104 cells were treated with either vehicle (DMSO) or 500-nM MLN4924 (6 hr after transfection). Whole-cell lysates were prepared (24-hr posttransfection), and immunoblotting was done to check expressions of AGO2, His (NSP1), Cullin1, and Cullin3. Neddylated form of Cullin1 and Cullin3 is indicated by arrows. (e) MA104 cells ectopically expressing pCMV-FLAG-Ub were cotransfected with either empty vector (pcDNA6B) or pcDNSP1. (Left panel) Cellular lysates prepared after 36 hr of transfection were immunoprecipitated with anti-FLAG antibody, and immunoprecipitates were further subjected to immunoblotting with anti-AGO2 antibody. Input lysates were probed with anti-AGO2 and anti-His (NSP1) antibodies. (Right panel) Reciprocal coimmunoprecipitation was done with anti-AGO2 antibody followed by immunoblotting with anti-FLAG antibody. Expressions of AGO2, His (NSP1) were checked in the input lysates. (f) MA104 cells were trasnfected with either empty vector (pcDNA6B), full-length NSP1 (cloned in pcDNA6B vector), or truncated domains of NSP1 (RING and ΔRING domain cloned in pcDNA6B vector). After 36 hr of transfection, cell lysates were subjected to western blot analysis to check levels of AGO2 and P53. Expression of His tag was used as a marker of pcDNSP1, pcDRING-NSP1, and pcDΔRING-NSP1 transfection. (g) Equal amount of lysates from pCMV-FLAG-Ub overexpressed MA104 cells cotransfected with pcDRING NSP1 or pcDΔRINGNSP1 were immunoprecipitated with anti-AGO2 antibody. Immunoprecipitates were then probed with anti-FLAG antibody. Simultaneously, His (NSP1) and AGO2 expressions were assessed in the input lysates serve as an RNA-silencing suppressor. Furthermore, silencing of VP4 expression by VP4 shRNA at 6 hpi was observed only in RV-A5-16 Together, the results suggest that RV-NSP1 can be regarded as a putative RV-encoded suppressor of RNAi, which inhibits functionality of siRNA and shRNA but not of miRNA.

We further looked for any possible association between RV-NSP1 and AGO2, which might lead to proteasomal degradation of AGO2 in the presence of RV-NSP1. pcDNSP1-transfected MA104 cells were lysed 36-hr posttransfection and subsequently immunoprecipitated with anti-His antibody. The presence of AGO2 in the immunoprecipitate indicated an association between AGO2 and RV-NSP1 ( Figure 5a ).

The presence of P53 and absence of AGO1 in the immunoprecipitate served as positive and negative control, respectively, for this interaction assay (Figure 5a ). Similar observation was also made in HEK293 cell line ( Figure S5A ). As both AGO2 (Oceguera et al., 2018) and RV-NSP1 (Hua, Chen, & Patton, 1994) were previously reported to interact with viral RNA, we went on to investigate whether the association between RV-NSP1 and AGO2 required viral RNA intermediates. Interestingly, AGO2-RV-NSP1 association was found to be unperturbed in the presence of both single-stranded RNA-specific RNase I and double-stranded RNA-specific RNase III (Figure 5b ), excluding the possibility of viral RNA scaffold to have a mediating role.

AGO2 activity has been reported to be regulated by phosphorylation (Bridge et al., 2017; Horman et al., 2013) . Phosphorylation of RV-OSU-NSP1 by caesin kinase II has also recently been shown to be a prerequisite for NSP1-βTrCP association (Davis et al., 2017) . To further explore whether association between RV-NSP1 and AGO2 is influenced by phosphorylation status of either NSP1 or AGO2, coimmunoprecipitation analysis was carried out in the presence of lambda protein phosphatase (lambda PP). NSP1-AGO2 association was not hampered following dephosphorylation of lysates with Anti-His coupled resin was incubated with cellular extracts from mock-transfected or pcDNSP1-transfected MA104 cells pretreated with RNase I and RNase III. Immunoprecipitates were subsequently subjected to western blot analyses with anti-AGO2 antibody. Input lysates were probed with anti-His and anti-AGO2 antibodies. (c) MA104 cells were transfected with empty vector or pcDNSP1 for 36 hr followed by whole-cell lysate preparartion. Cellular lysates were incubated with active lambda PP or heat inactivated lambda PP as described in the Experimental Procedures section followed by immunoprecipitation with anti-His antibody. Expression of AGO2 was analysed in the immunoprecipitates by immunoblot analysis. Expression of AGO2 and His (NSP1) was analysed in the input lysates (d) MA104 cells were transfected with pcDRING-NSP1 or pcDΔRING-NSP1. After 36 hr, cell lysates were immunoprecipitated with anti-His antibody. Immunoprecipitates were then resolved on SDS-PAGE, transferred on to PVDF membrane, and finally probed with anti-AGO2 and anti-P53 antibodies. Expressions of AGO2, P53, and His (NSP1) were checked in the input lysates. (e) MA104 cells were mock infected or infected with RV-SA11 for 3 and 12 hr. Equal amount of lysates were immunoprecipitated with anti-AGO2 antibody followed by immunoblotting with anti-NSP1 antibody. Expressions of AGO2 and NSP1 were checked in the input lysates. (f) (Left panel) Schematic representation of AGO2 mutants showing span of deletion (AGO2-mt1-4). (Right panel) MA104 cells transfected with pcDNSP1 were further cotransfected separately with AGO2 mutants (AGO2-mt1-4). After 36 hr, equal amounts of whole-cell lysates were immunoprecipitated using anti-FLAG antibody. Immunoprecipitates were then subjected to immunoblot analysis with anti-NSP1 antibody. Expressions of His (NSP1) and FLAG (for AGO2 domain mutants) were analysed in the input lysates input lysates of lambda PP-treated cells confirmed lambda PP enzymatic activity ( Figure S5B ). AGO proteins consist of N-terminal (N), PAZ (PIWI/AGO/Zwille), MID (middle), and PIWI (P-element-induced whimpy testes) domains (Faehnle & Joshua-Tor, 2010) . We therefore went on to characterise the region of AGO2 needed for association with RV-NSP1.

Coimmunoprecipitation assay was performed using four FLAG-tagged In agreement with the previous results, NSP1-AGO2 association was found in the presence of FLAG-PAZ but not of FLAG-PIWI, confirming involvement of PAZ domain of AGO2 in association with NSP1 ( Figure S5C ). Whole-cell lysates were subjected to western blot analyses to check expressions of RV structural (VP1 and VP6) and nonstructural (NSP3) proteins. (b) End point viral titers (12, 24, and 48 hpi) were determined by plaque assay from RV-SA11 (moi 0.5)-infected MA104 cells in the presence and absence of overexpressed hAGO2GFP (5 μg) and expressed as virus plaque forming units/ml. (c) MA104 cells were transfected with empty vector (FLAG; 5 μg) or AGO2 domain mutants (AGO2-mt1-4) and subsequently infected with RV-SA11 (moi 1; 12 hr). End point viral load was measured by plaque assay and represented as viral plaque forming units/ml. (d) MA104 cells transfected with full-length hAGO2GFP or catalytic-dead AGO2 mutants (designated as AGO2-D597A or AGO2-D669A) were infected with RV-SA11 for 12 hr followed by viral titer measurement. Empty vector (GFP; 5 μg) and AGO2-mt1-transfected cells served as negative control and positive control, respectively, for this experiment proteins in MA104 cells at 12 hpi (Figure 6a ) without bystander cytotoxicity ( Figure S6A ). Concurrently, a dose-dependent reduction in the percentage of RV infectivity was observed at 12 hpi in response to increasing concentration of overexpressed hAGO2GFP ( Figure S6B ).

For further experiments, the concentration of 5 μg of hAGO2GFP was chosen, as this concentration showed prominent antiviral effect without stochastic cytotoxicity. Attenuation of RV RNA upon AGO2 overexpression was also significantly prominent at 12 hpi in conventional RT-PCR-based studies ( Figure S6C ). Moreover, antiviral potency of AGO2 overexpression was found to be evident at as late as 24 and 48 hpi (Figure 6b) . Notably, overexpression of AGO1 (GFP Ago1) had no effect on either RV protein expression or RV progeny yield ( Figure   S6D Loss of antiviral potency of AGO2, however, is more prominent in the presence of PIWI domain deleted AGO2 mutants (AGO2-mt1 and AGO2-mt2) compared with PIWI domain containing AGO2 mutants (AGO2-mt3 and AGO2-mt4; Figure 6c ; Figure S6F ). Notably, overexpression of AGO2-mt3 showed slightly yet consistently better anti-rotaviral property than that of AGO2-mt4 (Figure 6c ; Figure   S6F ). Due to lack of PAZ domain, AGO2-mt3 fails to associate with RV-NSP1. Subsequent increased half-life of AGO2-mt3 during infection scenario might be the reason behind better anti-rotaviral property of AGO2-mt3 compared with AGO2-mt4.

AGO2 catalytic activity is reported to be antiviral against a broad range of viral infections . For this purpose, AGO2 site-directed (catalytic-dead) mutants (AGO2-D597A and AGO2-D669A) were transfected in MA104 cells followed by RV-SA11 infection (for 12 hr). End point viral titer measurement showed attenuated RV infection in the presence of slicing-competent wild-type AGO2 to restore in MA104 cells transfected with catalytic-dead AGO2 mutants (Figure 6d ). PIWI domain deleted AGO2 mutant (AGO2-mt1) was taken as positive control for this experiment ( Figure 6d ). Similar results were reflected when expression of RV proteins was checked in the presence of wild type and catalytically impotent AGO2 constructs ( Figure S6G ). Together, the data suggest that the full-length, slicing-competent AGO2, especially encompassing the PIWI domain, is a critical anti-rotaviral host determinant.

Physiological relevance of RNAi as an antiviral innate immune measure in plants and nonchordates follows three fundamental lines of arguments: (a) ample viRNAs are produced during natural course of viral infection in these species, (b) host mutants defective in mounting RNAi-mediated defense response are more susceptible to viral infec-

(VSRs); VSR-mutant viruses show less virulence in wild-type hosts but acquire important virulence rescue phenotype in RNAi-defective host mutants (Gammon & Mello, 2015; Sledz & Williams, 2005; Younis, Siddique, Kim, & Lim, 2014) . In spite of their universality and ubiquitousness, VSRs are highly divergent within and across kingdoms with no sequence homology and have been hypothesised to have evolved independently in different viruses (Burgess, 2013) . Virusesinfecting mammalian cells, on the other hand, are only sensitive to synthetic siRNAs that bypass the requirement of Dicer for being processed but produce very little viRNAs during natural course of infection (Cullen & Cherry, 2013) . This is possibly due to decreased Dicerprocessing activity, which occurs during cellular differentiation perhaps via modification of its internal auto-inhibitory helicase domain.

Interestingly, inability to mount RNAi-based innate immunity in response to invading dsRNAs has well been superseded in higher Cumulative evidences have emerged in support of functional RNAi-dependent antiviral immunity in mammals (Berkhout, 2018; Jeang, 2012; Qiu et al., 2017) . According to this view, the very establishment of infection over cellular defence requires production of extremely potent and efficient VSRs. This view is (Arias et al., 2004; Chen et al., 2011; Déctor et al., 2002; Dhillon & DurgaRao, 2018; López et al., 2005; Montero, Rojas, Arias, & López, 2008; Mukherjee et al., 2018; Oceguera et al., 2018; Patra et al., 2019; Silva-Ayala et al., 2013; Trujillo-Alonso et al., 2011; Yin et al., 2018) . Intriguingly, we found infection time point-dependent bimodal regulation of RNAi, where RNAi showed differential sensitivity at early and late hours of RV infection. siRNA-guided RNAi targeting ectopically expressed GFP (Figure 1a ,b, Figure S1A , B), and endogenous gene Rbx1 (Figure 1c) and shRNA-mediated RNAi targeting rotaviral vp4 gene (Figure 1d , Figure S1C ) were all found to be blocked at early hours of RV-SA11 infection only to regain full functionality at relatively late hours. Evidence of abrogated functionality of synthetic siRNAs suggests that hindered Dicer activity cannot be the exclusive reason for RNAi inhibition during early hours of RV-SA11 infection. In contrast, functionality of miRNA is retained, as miRNA mimics successfully knocked-down target expression both at early and late hours of RV-SA11 infection (Figure 1e,f) . This is in agreement with our previous findings, where miRNA mimics preserved their target repression potency in context of RV-SA11 infection (Chanda, Nandi, & Chawla-Sarkar, 2016; Mukhopadhyay et al., 2019) . We also did not find global change of miRNA population characteristic of abrogated miRNA biogenesis in our miRNA microarray data (Chanda et al., 2016) . Therefore, in the absence of AGO2, loading in RISC and miRNAguided knock-down can be taken over by other nonslicing reserve AGOs to sustain cell viability. In agreement with this is the observation that all AGOs have been shown to bind to miRNAs and siRNAs indiscriminately of sequence, to interact with a common set of helicases and miRNA-binding proteins and to localise to P bodies in mammalian cells (Ruda et al., 2014) . Indeed, availability of nonslicing AGOs during early hours of RV infection (Figure 2a , Figure S2A Figure S3A , B, G), to trigger protein level reduction of AGO2. RV-NSP1 is a multifunctional protein involved in IFN antagonism, activation of pro-survival PI3K/Akt pathways, denucleating cytoplasmic P bodies, and also harbouring RNA binding as well as cytoskeleton localisation domain (Bagchi et al., 2013; Bhowmick et al., 2015; Ding et al., 2016; Lutz et al., 2016; Morelli, Ogden, & Patton, 2015) . IFN antagonistic properties of RV-NSP1 lies in its ability to trigger proteasomal destruction of IRF3, 5, 7, 9, β-TrCP, TRAF2, and RIG-I, whereas P body destabilising property of NSP1 includes degradation of Pan3 by ubiquitin-proteasome system. RV-NSP1 possesses a highly Figure S5A , C). This is in contrast to the fact that substrate binding specificity of RV-NSP1 generally lies in its highly variable C-terminal half (Lutz et al., 2016; Morelli et al., 2015) . The association between AGO2 and RV-NSP1 (Figure 5e , Figure S5B ) as well as ubiquitin enrichment of AGO2 ( Figure S3F) were also observed at early hour of RV-SA11 infection. Interestingly, at relatively late hours, both association with RV-NSP1 (Figure 5e , Figure S5B ) and ubiquitination of AGO2 decreased ( Figure S3F ), partially explaining restoration of AGO2 protein levels at these late time points of infection.

RV-NSP1 has recently been reported to interact with (Davis et al., 2017; Ding et al., 2016; Lutz et al., 2016) and degrade β-TrCP by co-opting Cullin RING E3 ubiquitin Ligases (CRLs; Ding et al., 2016) . Reversal of AGO2 degradation, however, was observed in presence of neither dominant negative forms of Cullin1 and Cullin3 nor pan-Cullin inhibitor MLN4924, suggesting that RV-NSP1mediated degradation of AGO2 is not dependent on co-opted CRLs ( Figure 3d; Figure S3E ). Molecular mechanisms governing time point-dependent bimodal regulation of AGO2-NSP1 association and possible involvement of co-opted host E3 ubiquitin ligase, however, remain to be investigated in future. Moreover, on the basis of our previous observation of p53-NSP1 interaction to get regulated in a similar bimodal fashion , we also cannot rule out the possibility of mutual inaccessibility of RV-NSP1 or its interaction partner to each other during late phase of infection as a result of differential subcellular relocalisation. In agreement with Schuster, Tholen, Overheul, van Kuppeveld, and van Rij (2017) and Schuster, Miesen, and van Rij (2019) , however, we can make a hypothesis on the physiological relevance of time point-dependent bimodal AGO2 regulation. According to this notion, biphasic AGO2 modulation by RV-NSP1 relies on the prioritisation preference of the latter to get invested either in crippling antiviral RNAi during relatively early phase of infection or antagonising host-triggered antiviral IFN response during the later phase. From the perspective of RV-NSP1 as a putative VSR, degrading AGO2 to disarm si/shRNA-mediated RNAi is advantageous during early hours of infection. But as infection progresses and RV dsRNAs accumulate above a certain threshold, host pattern recognition receptors (RIG1 and MDA5) can start recognising viral replication intermediates as non-self entities and respond by triggering antiviral IFN signalling (Broquet, Hirata, McAllister, & Kagnoff, 2011) . To counteract this, RV employs NSP1 to degrade IFN regulatory factors ultimately escaping host immune response.

Evidence for viral IFN antagonist to also serve as a VSR is strikingly abundant (Cárdenas et al., 2006; Hartman, Towner, & Nichol, 2004) .

VSRs have been reported to interfere at different stages of RNA silencing pathways: viral RNA sensing and dicing, RISC assembly, AGO activity/levels, and also by targeting accessory components of RISC. Most of the VSRs are dsRNA-binding proteins that interact with either Dicer substrates, siRNA duplexes, or both and sequester them away from RNA silencing pathway (Bronkhorst & van Rij, 2014; Schuster et al., 2019) . In this respect, it is noteworthy that bacterial dsRNA-binding proteins can also inhibit RNAi pathway nonspecifically by virtue of their binding to dsRNAs and thereby withdrawing dsRNAs away from RNAi machinery. There are a few reports of degradation of AGO2 in case of insect and plant viruses (Baumberger, Tsai, Lie, Havecker, & Baulcombe, 2007; Nayak et al., 2018) , but no report is available for mammalian viruses. Consistent with the ability of RV-NSP1 to degrade AGO2 proteasomally, RV-NSP1 also showed putative VSR property as its ectopic expression prevented siRNA-mediated RNAi (Figure 4a ; Figure S4A ).

Concurrently, shRNA-mediated RNAi was also found to be functional during early and late hours of A5-16 infection (Figure 4b , Figure S4B ).

only attain high pathogenicity in RNAi-defective host cells. NSP1mutant RV strain A5-16 has also been reported to have slow growth rate and forms smaller plaques compared with wild-type RV strain A5-13 (Taniguchi et al., 1996; Bagchi et al., 2010) . However, the difference in the virulence phenotype between wild-type A5-13 and NSP1 mutant A5-16 may arise, at least partially, from activation of premature apoptosis and failure of IFN antagonism. So, it poses an experimental challenge to uncouple attenuated pathogenicity stemming from failure to antagonise IFN and premature apoptosis from that due to failure of RNAi inhibition. Indeed, as also observed previously during infection, miRNA mimics retained their functionality in the presence of RV-NSP1 (Figure 4c,d) .

Antiviral potency of AGO2 is widely accepted for plant, insect, and mammalian viruses (Alvarado & Scholthof, 2012; Carbonell & Carrington, 2015; Harvey et al., 2011) . AGO2 catalytic activity is especially reported to restrict influenza A virus, encephalomyocarditis virus, and vesicular stomatitis virus infection in mammalian cells . In our study too, overexpression of AGO2 but not AGO1 showed potent anti-rotaviral activity (Figure 6a ,b; Figure   S6 B-E). This is rationally consistent with recent reports of increased RV infectivity in the absence of AGO2 (Dhillon & DurgaRao, 2018; Oceguera et al., 2018) . Moreover, all the functional domains of AGO2 were found to be important for exerting the best antiviral potency, the catalytic PIWI domain being the major determinant Scientific) and 1% (v/v) Antibiotic-Antimycotic (15240062-Thermo Fisher Scientific) within humidified 37°C incubator with 5% CO 2 . Cell culture-adapted RV strains SA11 (simian), A5-13 (bovine), and A5-16 (bovine) were used for this study. Unless otherwise mentioned, cells were infected at a multiplicity of infection (moi) 3 as described previously (Dutta et al., 2011) (Groene & Shaw, 1992) .

For calculating viral infectivity, plaque assay was performed as described previously (Dutta et al., 2009 ). Viral Plaque Forming Units (PFU) were calculated as PFU/ml (of original stock) = 1/dilution factor × number of plaques × 1/(ml of inoculum/plate; Smith, Estes, Graham, & Gerba, 1979) . In some experiments, viral infectivity was represented as "percentage of infectivity" of infected test cells compared with infected control cells (infectivity of infected control was considered as 100%). 

The 3′-UTR luciferase reporter construct of P85α-PI3K was generated by cloning the PCR-amplified human P85α mRNA 3′-UTR (miR-29b 

MA104 and HEK293 cells seeded in four-well chamber slides (BD Pharmingen) were treated as mentioned in Figure 1B , S1B, 2B. Cells were then fixed with paraformaldehyde (4% [wt/vol] in phosphatebuffered saline) for 10 min at room temperature and further processed as described previously (Mukherjee et al., 2018) . After overnight incubation with primary antibodies (anti-AGO2, anti-NSP5, and anti-VP6) at 4°C, slides were treated with Rhodamine conjugated and 

Lysates from infected/transfected MA104/HEK293 cells were subjected to coimmunoprecipitation using Pierce Co-IP kit (26149) according to the manufacturer's instructions. For Resin controls, cell lysates were incubated with antibody-uncoupled resin and further processed according to the protocol. To immunoprecipitate equal amount of lysates, protein concentration was measured by Bradford assay (Sigma-Aldrich) or Pierce™ BCA Protein AssayKit (Thermo Fisher Scientific). A 5% volume of each lysate was kept as input. To analyse RNA-independent interaction, antibody coupled resins were incubated with equal amount of cell extracts pretreated with RNase III (2 μg/ml) (Brown & Gordon, 1984) . After elution, immunoprecipitates were boiled in SDS-PAGE loading buffer for 20 min followed by gel electrophoresis and immunoblot analyses.

Short 

To determine cytotoxicity of increasing concentrations of hAGO2GFP overexpression in MA104 cells, cell viability assay was conducted. 

All results are presented as mean ± standard deviation from at least three (n ≥ 3) independent experiments. Statistical significance of data was analysed by Mann-Whitney test or Student's t test using Graphpad Prism 5.0. Statistical significance of data is marked by asterisks ( * p ≤ .05; ** p ≤ .01; *** p ≤ .001). P value of <.05 was considered to be statistically significant for all experiments.

AGO2: A new Argonaute compromising plant virus accumulation

Suppression of RNA interference by adenovirus virus-associated RNA

RNA silencing of rotavirus gene expression

Rotavirus NSP1 mediates degradation of interferon regulatory factors through targeting of the dimerization domain

The Mammalian response to virus infection is independent of small RNA silencing

Rotavirus NSP1 inhibits interferon induced non-canonical NFκB activation by interacting with TNF receptor associated factor 2

Rotavirus nonstructural protein 1 suppresses virus-induced cellular apoptosis to facilitate viral growth by activating the cell survival pathways during early stages of infection

Identification of common human host genes involved in pathogenesis of different rotavirus strains: an attempt to recognize probable antiviral targets

Upsurge and spread of G3 rotaviruses in Eastern India (2014-2016): Full genome analyses reveals heterogeneity within Wa-like genomic constellation

Rotavirus NSP1 inhibits expression of type I interferon by antagonizing the function of interferon regulatory factors IRF3, IRF5, and IRF7

MicroRNAs: Target recognition and regulatory functions

Metazoan MicroRNAs

RNA silencing in plants

The polerovirus silencing suppressor P0 targets ARGONAUTE proteins for degradation

HIV-1 Tat interaction with Dicer: Requirement for RNA

Evidence that HIV-1 encodes ansiRNA and a suppressor of RNA silencing

Rotavirus-encoded nonstructural protein 1 modulates cellular apoptotic machinery by targeting tumor suppressor protein p53

Rotavirus disrupts cytoplasmic P bodies during infection

Viral RNA silencing suppressors (RSS): Novel strategy of viruses to ablate the host RNA interference (RNAi) defense system

Argonaute utilization for miRNA silencing is determined by phosphorylation-dependent recruitment of LIM-domain-containing proteins

The long and short of antiviral defense: Small RNA-based immunity in insects

RIG-I/MDA5/MAVS are required to signal a protective IFN response in rotavirus-infected intestinal epithelium

The stimulation of pp60v-src kinase activity by vanadate in intact cells accompanies a new phosphorylation state of the enzyme

The influenza A virus NS1 protein binds small interfering RNAs and suppresses RNA silencing in plants

Small RNAs: Antiviral RNAi in mammals

Viral suppressors of RNA silencing

Antiviral roles of plant ARGONAUTES

Ebola virus VP35 protein binds double-stranded RNA and inhibits alpha/beta interferon production induced by RIG-I signaling

Rotavirus-induced miR-142-5p elicits proviral milieu by targeting non-canonical transforming growth factor beta signalling and apoptosis in cells

Tyrosine phosphorylation modulates mitochondrial chaperonin Hsp60 and delays rotavirus NSP4-mediated apoptotic signaling in host cells

Short hairpin RNA-mediated silencing of bovine rotavirus NSP4 gene prevents diarrhoea in suckling mice

The nucleocapsid protein of coronaviruses acts as a viral suppressor of RNA silencing in mammalian cells

Is RNA interference a physiologically relevant innate antiviral immune response in mammals?

Rotavirus NSP1 requires casein kinase II-mediated phosphorylation for hijacking of cullin-RING ligases

Rotavirus gene silencing by small interfering RNAs

Rotaviruses. Virus Research

Rotavirus induces formation of remodeled stress granules and P-bodies and their sequestration in viroplasms to promote progeny virus production

Comparative proteomics reveals strain-specific β-TrCP degradation via rotavirus NSP1 hijacking a host Cullin-3-Rbx1 complex

Antiviral RNA interference in mammals

The molecular chaperone heat shock protein-90 positively regulates rotavirus infection

Active participation of cellular chaperone Hsp90 in regulating the function of rotavirus nonstructural protein 3 (NSP3)

Rotaviruses

Argonaute MID domain takes centre stage

RNA interference-mediated antiviral defense in insects. Current opinion in insect science

Cytosolic sensing of viruses

Rotavirus NSP1 inhibits NFκB activation by inducing proteasome-dependent degradation of β-TrCP: A novel mechanism of IFN antagonism

Zinc-binding domain of rotavirus NSP1 is required for proteasome-dependent degradation of IRF3 and autoregulatory NSP1 stability

Interferon regulatory factor 3 is a cellular partner of rotavirus NSP1

Psoralen preparation of antigenically intact noninfectious rotavirus particles

Theebola virus VP35 protein is a suppressor of RNA silencing

RNA interference

A C-terminal basic amino acid motif of Zaire ebolavirus VP35 is essential for type I interferon antagonism and displays high identity with the RNA-binding domain of another interferon antagonist, the NS1 protein of influenza A virus

An antiviral defense role of AGO2 in plants

Akt-mediated phosphorylation of argonaute 2 downregulates cleavage and upregulates translational repression of MicroRNA targets

Deletion mapping of the rotavirus metalloprotein NS53 (NSP1): the conserved cysteine-rich region is essential for virus-specific RNA binding

RNAi in the regulation of mammalian viral infections

Identification of Substrates for F-Box Proteins

Nodamura virus nonstructural protein B2 can enhance viral RNA accumulation in both mammalian and insect cells

Generation of recombinant rotaviruses expressing fluorescent proteins by using an optimized reverse genetics system

Reverse genetics system demonstrates that rotavirus nonstructural protein NSP6 is not essential for viral replication in cell culture

Modification of the trypsin cleavage site of rotavirus VP4 to a furin-sensitive form does not enhance replication efficiency

Technical insights into highly sensitive isolation and molecular characterization of fixed and live circulating tumor cells for early detection of tumor invasion

Quantitative analysis of Argonaute protein reveals microRNA-dependent localization to stress granules

Interferon antagonist proteins of influenza and vaccinia viruses are suppressors of RNA silencing

Induction and suppression of antiviral RNA interference by influenza A virus in mammalian cells

RNA interference functions as an antiviral immunity mechanism in mammals

The C-terminal half of human Ago2 binds to multiple GW-rich regions of GW182 and requires GW182 to mediate silencing

Argonaute2 is the catalytic engine of mammalian RNAi

Reduced expression of the rotavirus NSP5 gene has a pleiotropic effect on virus replication

Adenovirus VA1 noncoding RNA can inhibit small interfering RNA and MicroRNA biogenesis

Rotavirus NSP1 Associates with components of the Cullin RING ligase family of E3 ubiquitin ligases

Inactivation of the type I interferon pathway reveals long double-stranded RNA-mediated RNA interference in mammalian cells

A lentiviralRNAi library for human and mouse genes applied to an arrayed viral high-content screen

Rotavirus infection induces the phosphorylation of eIF2α but prevents the formation of stress granules

Silencing the alarms: Innate immune antagonism by rotavirus NSP1 and VP3

Rotaviral nonstructural protein 4 triggers dynamin-related protein 1-dependent mitochondrial fragmentation during infection

Synchronized Orchestration of miR-99b and let-7 g Positively Regulates Rotavirus Infection by Modulating

A Viral Protein Restricts Drosophila RNAi Immunity by Regulating Argonaute Activity and Stability

Rotavirus RNAs sponge host cell RNA binding proteins and interfere with their subcellular localization

Six RNA viruses and forty-one hosts: viral small RNAs and modulation of small RNA repertoires in vertebrate and invertebrate systems

miR-29 miRNAs activate p53 by targeting p85α and CDC42

RA-839, a selective agonist of Nrf2/ARE pathway, exerts potent anti-rotaviral efficacy in vitro

Rotavirus nonstructural protein 1 antagonizes innate immune response by interacting with retinoic acid inducible gene I

Human virus-derived small RNAs can confer antiviral immunity in mammals

Purified Argonaute2 and ansiRNA form recombinant human RISC

Protein kinase R is responsible for the phosphorylation of eIF2α in rotavirus infection

The roles of individual mammalian argonautes in RNA interference in vivo

Antiviral actions of interferons

Interferon-stimulated genes: A complex web of host defenses

Antiviral RNAi in Insects and Mammals: Parallels and Differences. Viruses

Deletion of cytoplasmic double-stranded RNA sensors does not uncover viral small interfering RNA production in human cells

EGFR modulates microRNA maturation in response to hypoxia through phosphorylation of AGO2

Genome-wide RNAi screen reveals a role for the ESCRT complex in rotavirus cell entry

On the road to reading the RNA-interference code

Antiviral RNA interference responses induced by Semliki Forest virus infection of mosquito cells: Characterization, origin, and frequency-dependent functions of virus-derived small interfering RNAs

RNA interference in biology and disease

Effect of UV-irradiation on rotavirus

A plaque assay for the simian rotavirus SA11

Nondefective rotavirus mutants with an NSP1 gene which has a deletion of 500 nucleotides, including a cysteine-rich zinc finger motif-encoding region (nucleotides 156 to 248), or which has a nonsense codon at nucleotides 153 to 155

The evolution of antiviral defense systems

Rotavirus infection induces the unfolded protein response of the cell and controls it through the nonstructural protein NSP3

The silent treatment: RNAi as a defense against virus infection in mammals

Hepatitis C virus core protein is a potent inhibitor of RNA silencing-based antiviral response

Mechanistic insights into archaeal and human Argonaute substrate binding and cleavage properties

Viral suppressors of RNA-based viral immunity: Host targets

PI3K-Akt-mTOR axis sustains rotavirus infection via the 4E-BP1 mediated autophagy pathway and represents an antiviral target

RNA interference (RNAi) induced gene silencing: a promising approach of hi-tech plant breeding

SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section at the end of the article. How to cite this article

providing catalytic-dead AGO2 mutants (pcDNA3-AGO2 D597A, pcDNA3-AGO2 D669A). Dr. Asim Biswas is duly acknowledged for his technical assistance in confocal microscopy.

Authors declare no conflict of interest.

Mamta Chawla-Sarkar https://orcid.org/0000-0002-7141-1785