Selective inhibition of miRNA processing by a herpesvirus-encoded miRNA

Herpesviruses have mastered host cell modulation and immune evasion to augment productive infection, life-long latency and reactivation1,2. A long appreciated, yet undefined relationship exists between the lytic–latent switch and viral non-coding RNAs3,4. Here we identify viral microRNA (miRNA)-mediated inhibition of host miRNA processing as a cellular mechanism that human herpesvirus 6A (HHV-6A) exploits to disrupt mitochondrial architecture, evade intrinsic host defences and drive the switch from latent to lytic virus infection. We demonstrate that virus-encoded miR-aU14 selectively inhibits the processing of multiple miR-30 family members by direct interaction with the respective primary (pri)-miRNA hairpin loops. Subsequent loss of miR-30 and activation of the miR-30–p53–DRP1 axis triggers a profound disruption of mitochondrial architecture. This impairs induction of type I interferons and is necessary for both productive infection and virus reactivation. Ectopic expression of miR-aU14 triggered virus reactivation from latency, identifying viral miR-aU14 as a readily druggable master regulator of the herpesvirus lytic–latent switch. Our results show that miRNA-mediated inhibition of miRNA processing represents a generalized cellular mechanism that can be exploited to selectively target individual members of miRNA families. We anticipate that targeting miR-aU14 will provide new therapeutic options for preventing herpesvirus reactivations in HHV-6-associated disorders. Herpesvirus microRNAs interfere directly with host cell microRNA processing, thereby disrupting mitochondrial architecture, evading intrinsic host defences and driving the switch from latent to lytic infection.

miRNAs are important regulators of gene expression that are implicated in all major cellular processes of life, ranging from embryonic development to tissue homeostasis and cancer 5,6 . Accordingly, their biogenesis is tightly regulated at all levels 7 . Shortly after the discovery of cellular miRNAs, a number of viruses, predominantly of the herpesvirus family, were identified to encode and express their own set of viral miRNAs 4,8 . One of these is HHV-6A, which has a seroprevalence of over 90% in the human population. HHV-6A establishes latency by integrating into the telomeric regions of host chromosomes 9 . Virus reactivation has been associated with cardiac dysfunction, graft rejection as well as neuronal disorders including myalgic encephalomyelitis and chronic fatigue syndrome 10 . Here, we identify miRNA-mediated inhibition of miRNA processing as mechanism that HHV-6A exploits to disrupt mitochondrial architecture, evade the induction of type I interferons and facilitate virus reactivation from latency.

HHV-6 induces mitochondrial fission
Mitochondria have a key role in the cell intrinsic defence against viruses. They constantly undergo fission and fusion events that help maintain functional mitochondria in cells under metabolic and environmental stress 11 . To examine whether HHV-6A affects mitochondrial architecture, we infected primary human umbilical vein endothelial cells (HUVEC) with wild-type HHV-6A and imaged mitochondria using a constitutively expressed, mitochondrially targeted GFP 12 (mitoGFP). Lytic HHV-6A infection resulted in extensive mitochondrial fragmentation by 24 hours post-infection (hpi) (Fig. 1a, Extended Data Fig. 1a). A similar effect was observed upon reactivation of latent HHV-6A in U2OS cells induced by trichostatin-A (TSA) treatment (Extended Data Fig. 1b). Mitochondrial fusion-fission dynamics are governed by the activity of dynamin-related protein 1 13 (DRP1). Helical oligomers of DRP1 form a ring around the outer mitochondrial membrane and fragment it 14 . Mitochondrial fragmentation was reflected by increased DRP1 expression during both lytic HHV-6A infection (Fig. 1b) and virus reactivation (Extended Data Fig. 1c), as well as colocalization of DRP1 on mitochondrial surfaces in the virus-reactivated cells (Extended Data Fig. 1d). DRP1 levels are directly controlled at the transcriptional level by the p53 tumor suppressor protein 15 . Accordingly, both lytic HHV-6A infection and virus reactivation resulted in increased p53 expression (Fig. 1b, Extended Article Data Fig. 1c) indicating that HHV-6A induces mitochondrial fragmentation via the canonical p53-DRP1 axis 15 .

miR-aU14 inhibits miR-30 processing
Unlike lytic infection, HHV-6A reactivation in the U2OS cell model does not progress to fully productive virus replication but is restricted to the expression of some viral miRNAs and mRNAs 16 . We therefore investigated whether any of the HHV-6A miRNAs might be involved in the observed miR-30 processing defect. Manual sequence inspection revealed a complementarity between HHV-6A miR-U14 and the hairpin loops of precursor (pre)-miR-30c, pre-miR-30a and pre-miR-30d (Fig. 1e, Extended Data Fig. 3e). This viral miRNA is expressed at very high levels during both productive infection 17 and virus reactivation 16 . Because it is encoded antisense to the U14 open reading frame (ORF), we refer to it here as miR-aU14. The miR-30c hairpin loop showed the strongest sequence complementarity to miR-aU14 18 (Extended Data Fig. 3f, g). Small-RNA sequencing of Argonaute (AGO)-bound RNA from HHV-6A-infected HSB-2 cells confirmed that this incompletely characterized viral miRNA represents one of the two most abundant viral miRNAs in HHV-6A-infected cells (Extended Data Fig. 4a). In addition, transcription start site (TSS) profiling using differential   HHV-6A infection was tested using viral glycoprotein GP82/105. GAPDH served as loading control. n = 3. Fold change in p53 and DRP1 proteins are shown as mean of three biological replicates. *P = 0.03. c, Mature miR-30 during lytic HHV-6A infection. HHV-6A infection was tested using viral miR-aU14 and sncRNA-U77. n = 3. d, Defective miR-30c processing during lytic HHV-6A infection. n = 6. **P = 0.009. e, Schematic of putative interaction of miR-aU14 with pre-miR-30c hairpin loop. Predicted sites of interaction are in grey boxes. Mature miR-30c is shown red. f, Nucleotide sequences of wild-type (WT mimic) and mutant (Mut mimic) miR-aU14 mimics. Point mutations are outlined in red. g, Defective Pri-miR-30c processing caused by wild-type or mutant miR-aU14 mimic. n = 2. h, Defective Pri-miR-30c processing during reactivation of wild-type HHV-6A (HHV-6A-WT) or mutant (HHV-6A-Mut) virus. U2OS cells without HHV-6A served as mock. n = 3. ****P ≤ 0.0001. i, Average mitochondrial area in wild-type or mutant HHV-6A-reactivating cells. *P = 0.02. Data are from three independent experiments with at least two images from each experiment. In box plots, whiskers (a, i) show minimum to maximum values, points show individual replicates, the centre line denotes median, and box boundaries denote the 25th and 75th percentiles. RNA and protein were quantified by densitometric analysis (b-d, g, h). Human U6 served as loading control in Northern blots (c, d, g, h). Data are mean ± s.d. Unpaired two-tailed non-parametric t-test (a, b, d, h, i). Uncropped blots are provided in Supplementary Fig. 1.
To assess the role of miR-aU14 in miR-30c processing, we used a wild-type miR-aU14 miRNA mimic (WT mimic) and a mutant thereof (Mut mimic) (Fig. 1f). Of note, transfection of the wild-type but not the mutant mimic into U2OS cells reproduced both the loss of mature miR-30c and the concomitant increase in pri-miR-30c (Fig. 1g). To further validate that miR-aU14 was responsible for the miR-30c processing defect, we generated HeLa cells with a doxycycline (dox)-inducible miR-aU14 expressed from a RNA polymerase III (Pol III) promoter-driven short hairpin RNA (shRNA) and a mutant version thereof (HeLa-Mut) (Extended Data Fig. 5a). Dox-induced expression of miR-aU14 (HeLa-WT) (Extended Data Fig. 5b) but not of its mutant (HeLa-Mut) (Extended Data Fig. 5c) fully reproduced the miR-30c-processing defect.

miR-aU14 induces mitochondrial fission
We next tested whether expression of miR-aU14 was sufficient to disrupt mitochondrial architecture via the p53-DRP1 axis. Both transfection and dox-induced expression of wild-type but not mutant miR-aU14 induced expression of p53 and DRP1 (Extended Data Fig. 5d, e). Furthermore, transfection of the miR-aU14 wild-type mimic fully recapitulated the mitochondrial fission phenotype (Extended Data Fig. 5f).
To validate this effect in the virus context, we generated a mutant virus genome with discrete nucleotide substitutions within miR-aU14 (Extended Data Fig. 6a). Mutations were designed to not alter the amino acid sequence of the U14 ORF. Deep sequencing of both the wild-type (HHV-6A-WT) and mutant (HHV-6A-Mut) bacterial artificial chromosomes (BAC) confirmed that the introduced mutations were the only differences between the two viral genomes. In contrast to HHV-6A-WT, we were unable to reconstitute the miR-aU14 mutant virus despite multiple attempts, indicating that the loss of miR-aU14 severely reduced viral fitness. Therefore, we generated polyclonal U2OS cells that either carried chromosomally integrated latent wild-type HHV-6A or its miR-aU14 mutant by selection of cells stably transfected with the respective BACs. Upon virus reactivation with TSA, HHV-6A-WT but not HHV-6A-Mut impaired pri-miR-30c processing (Fig. 1h), induced DRP1 expression (Extended Data Fig. 6b) and triggered mitochondrial fission (Fig. 1i, Extended Data Fig. 6c).

Mechanism of the miRNA processing defect
We next tested whether miR-aU14 directly interacts with the pre-miR-30c hairpin loop and thereby interferes with its processing. We consecutively transfected HeLa cells with equimolar amounts of radiolabelled synthetic miR-aU14 (or two mutants thereof) followed by synthetic biotinylated pre-miR-30c 16 h later (Fig. 2a, b). Affinity purification of pre-miR-30c after 16 h revealed enrichment of wild-type miR-aU14 but not of a control miRNA. Two subtle miR-aU14 mutants showed an intermediate phenotype (Fig. 2c).
We next tested whether the presence of the pre-miR-30c hairpin loop was sufficient to mediate its inhibitory effects on miRNA processing. Two artificial target pre-miRNAs were designed that carried the original hairpin loop sequence of pre-miR-30c but contained artificial miRNA stem duplex sequences (designated miR-A and miR-B) (Fig. 2d). We then generated polyclonal HeLa cells with dox-inducible Pol III-driven expression of miR-A or miR-B as well as stable transduction with lentiviruses that expressed either wild-type or mutant miR-aU14 (Extended Data Fig. 5a). Consistent with the predicted interaction of miR-aU14 with the pre-miR-30c hairpin loop, induction of miR-aU14, but not of the mutant, strongly repressed both miR-A and miR-B processing (Fig. 2e, f).
To analyse the biochemical basis of miR-aU14-mediated regulation of miR-30 biogenesis, we carried out in vitro pri-miRNA and pre-miRNA processing assays 22 . Total protein lysates from cells exogenously expressing DGCR8 generated pre-miR-30c from radiolabelled in vitro-transcribed pri-miR-30c. This was inhibited by the miR-aU14 mimic but not the mutant (Extended Data Fig. 7a). A more pronounced interference was observed for pri-miR-30a (Extended Data Fig. 7b). Moreover, the miR-aU14 mimic, but not the mutant, . Article inhibited processing of radiolabelled in vitro transcribed pre-miR-30a and pre-miR-30c by human AGO-associated Dicer complex (Extended Data Fig. 7c, d). We confirmed purification of human DGCR8 and AGO-DICER complexes (Extended Data Fig. 7e, f). These results show that miR-aU14 can interfere with both pri-miRNA and pre-miRNA processing.

miR-aU14 inhibits the interferon response
Mitochondria have an important physiological role in intrinsic immunity 23 . Upon activation of toll-like or RIG-I-like receptors, mitochondria serve as antiviral signalling hubs that govern the production of type I interferons 24 (IFNs). RNA polymerase III can use cytosolic herpesvirus DNA as a template to produce 5′-triphosphate RNAs, which induce type I IFN through the RIG-I pathway [24][25][26] . Enforced mitochondrial fission dampens RIG-I-MAVS signalling and reduces the induction of type I IFNs 27 . We thus tested whether miR-aU14-mediated mitochondrial fragmentation affects the induction of the type I IFN, IFNβ. Exposure of HEK 293 cells, transfected with the miR-aU14 mimic, to the RIG-I pathway activator 3p-hpRNA 28 (5′-triphosphate hairpin RNA) resulted in reduced mRNA levels of IFNB (Fig. 3a) as well as the IFN-responsive IFIT1 gene ( Fig. 3b) compared with cells transfected with the miR-aU14 mutant.
We next tested whether miR-aU14 also has a role in suppressing the production of IFNβ upon HHV-6A reactivation. In addition to inducing virus reactivation by TSA, we treated cells with the JAK/STAT inhibitor ruxolitinib to prevent secondary IFNβ-mediated effects on virus reactivation, as assessed by Northern blot for viral miR-aU14 and sncRNA-U77 (Fig. 3c). Ruxolitinib treatment enhanced TSA-induced virus reactivation, resulting in a concordantly greater loss of miR-30c. Expression of viral sncRNA-U77 was significantly reduced for HHV-6A-Mut, indicative of impaired virus reactivation. Accordingly, miR-30c levels remained unchanged. Nevertheless, the mutant virus induced significantly greater levels of IFNB than the wild-type virus (Fig. 3d). Reactivation of wild-type virus reduced IFIT1 mRNA levels by approximately fivefold. By contrast, IFIT1 levels increased by 1.2-fold for HHV-6A-Mut relative to non-reactivated cells. This was partially inhibited by ruxolitinib treatment (Fig. 3e). Viral reactivation was confirmed by measuring mRNA levels of the viral immediate early U90 gene (Fig. 3f).

miR-aU14 triggers virus reactivation
Considering the observed effects of miR-aU14 on the induction of type I IFNs, we investigated whether ectopic expression of miR-aU14 could augment productive wild-type virus infection and rescue reactivation of the mutant virus. Transfection of miR-aU14 mimic but not the control miRNA increased the number of cells productively infected with wild-type virus by around 2.5-fold (Fig. 3g). Transfection of miR-aU14 mimic efficiently rescued reactivation of the mutant virus even in the absence of TSA (Fig. 3h). Combination of both TSA and miR-aU14 showed enhanced virus reactivation, indicating synergistic effects between the two. Similarly, transfection of miR-aU14 was substantially Pre-miR-30c Pri-miR-30c miR-aU14 HSB-2 cells were transfected with either the miR-aU14 mimic or a control mimic. HHV-6A mCherry reporter virus was used to measure cell-to-cell spread of virus infection, as measured by flow cytometry. n = 5. SSC, side scatter. *P = 0.01. h, U2OS cells carrying latent mutant HHV-6A miR-aU14 (HHV-6A-Mut) were transfected with either the miR-aU14 mimic or a control mimic. Cells were treated with TSA to induce virus reactivation, which was analysed using viral miR-U2 and miR-U86 expression, both of which are not expressed during virus latency. miR-U86 levels normalized to human U6 are presented as a bar graph. n = 3. **P < 0.008. Data are mean ± s.d. Unpaired two-tailed non-parametric t-test (a-h). Uncropped blots are provided in Supplementary Fig. 1. The gating strategy used in g is provided in Supplementary Fig. 2. more effective at inducing reactivation of wild-type virus than TSA (Extended Data Fig. 8a). Neither of the two mutant mimics had any effect on virus reactivation (Extended Data Fig. 8b).
We then tested whether mitochondrial fragmentation, impaired IFN response and HHV-6A reactivation were indeed mediated by the effects of miR-aU14 on miR-30. Both transfection of a miR-30c inhibitor and expression of a miR-30c sponge decreased mature miR-30c levels, induced p53 and DRP1 expression and triggered mitochondrial fragmentation (Extended Data Fig. 9a-c). Induction of the miR-30 sponge efficiently reduced the IFNβ response, and enhanced productive infection and viral spread (Extended Data Fig. 9d-f). Finally, it also enhanced virus reactivation upon TSA treatment of the mutant virus (Extended Data Fig. 9g).

Targeting human miRNA processing
In principle, miRNA-mediated inhibition of miRNA processing should be applicable to other cellular miRNAs. This is of particular interest, as many important cellular miRNAs exist as miRNA families. Targeting hairpin loops rather than the mature miRNA sequences would offer a unique opportunity for the development of more selective miRNA inhibitors. Many of the let-7 family members carry relatively large hairpin loops, which may comprise up to 30 nucleotides. Hence, we designed synthetic miRNA mimics targeting two different regions of the hairpin loop of pre-let-7d (Fig. 4a). Upon transfection into cells, both miRNA mimics efficiently reduced mature let-7d levels, consistent with impaired miRNA processing (Fig. 4b). Similar results were obtained for two other miRNA mimics targeting the hairpin loop of let-7f1 (Extended Data Fig. 10a, b).
Finally, we speculated that miRNA-mediated inhibition of miRNA processing should also be observable for cellular miRNAs. To identify such regulation, we carried out a systematic BLAST search of mature human miRNAs against known pre-miRNAs from miRBase. Several search results indicated potential binding sites within pre-miRNAs (Extended Data Table 1). However, the majority of the respective miRNA pairs were not abundantly expressed in most of the standard human cell lines. We therefore focused on one particular candidate pair, namely miR-155 and miR-148b. Our analysis indicated potential binding of miR-155 to pri-miR-148b just 5′ of the pre-miR-148b hairpin (Fig. 4c). Transfection of a miR-155 mimic resulted in reduced levels of mature miR-148b (Fig. 4d). The miR-155 mimic also strongly interfered with in vitro processing of pri-miR-148b (Extended Data Fig. 10c, d). miRNA-mediated inhibition of miR-148b processing by cellular miR-155 thus at least partially explains the dichotomous expression of these two human miRNAs 29 .

Discussion
Regulation of miRNA processing by cellular proteins is well described 7 . The first and best characterized example is the stem cell factor Lin-28, which interacts with the hairpin loop of let-7 family members and blocks their biogenesis [30][31][32] . A large unbiased screening approach recently identified around 180 RNA-binding proteins that specifically interact with distinct human pre-miRNAs 33 . RNA-binding protein-mediated regulation of miRNA processing thus constitutes an important regulatory network that shapes miRNA activity and function. Here we show that miRNA mimics can take over similar functions and selectively inhibit miRNA processing in a sequence-specific manner (Extended Data Fig. 11). Of note, the interaction of miR-aU14 and miR-30 precursors was not mediated by its seed region but rather by nucleotides 4 to 9. The miR-aU14-mediated loss of miR-30c was accompanied by a marked increase of pri-miR-30c levels. This implies that the inhibition occurs at the level of pri-miRNA processing within the nucleus, consistent with previous reports that miRNAs may affect pri-miRNA processing by binding to distal sequence elements in the respective pri-miRNAs [34][35][36] . miR-30c is encoded in an intron of the NFYC gene 37 . Recognition and cleavage of the intronic pre-miRNA hairpin loop by the nuclear RNase III DROSHA thus competes with the cellular splicing and RNA degradation machinery. Steric interference of DGCR8 binding to the miR-30c Pre-let-7d Human U6 A n t i-le t -7 d .1 A n t i-le t -7 d .2 Pre-miR-148b Article hairpin by miR-aU14 in the nucleus and subsequent degradation of the parental intron upon splicing is therefore the likely explanation for the observed loss of mature miR-30c. Accordingly, inhibition of miR-148b processing by the inflammatory miR-155 also appears to occur in the nucleus, as the respective binding site within the pri-miR-148b closely flanks the pre-miR-148b stem. This implies that miR-155 binding sterically inhibits pri-miR-148b cleavage by DROSHA. Inhibition of miR-148b processing by miR-155 presumably explains previous reports of dichotomal expression of these two important human miRNAs 29 .
Notably, both lytic HHV-6A infection and virus reactivation also increased pre-miR-30d levels within the cell. This is consistent with data from our in vitro processing assay, which indicated inhibition of miR-30 processing at the pre-miRNA level. Accordingly, accumulation of the let-7d pre-miRNA upon exposure to two different artificial miR-NAs targeting the hairpin loop indicated impaired pre-let-7d processing in the cytoplasm. miRNA-mediated inhibition of miRNA processing can thus interfere with miRNA biogenesis at both pri-miRNA and pre-miRNA level. Furthermore, these findings demonstrate that miRNA-mediated inhibition of miRNA processing can be readily exploited to specifically target individual miRNAs of large miRNA families.
Viral miR-aU14-mediated inhibition of miR-30 processing explained mitochondrial fragmentation during both lytic HHV-6A infection and virus reactivation via the miR-30-p53-DRP1 axis. This in turn impairs the induction of type I IFN and augments productive virus infection. Multiple attempts to reconstitute a miR-aU14 mutant virus from BAC DNA failed, indicating that viral miR-aU14 is crucial for productive virus replication in vitro. Similarly, the miR-aU14 mutant virus was severely impaired in its ability to reactivate from latency. It is, however, important to note that reactivation of the miR-aU14 mutant virus by TSA resulted in a significantly stronger type I IFN response than observed for wild-type virus. This indicates that miR-aU14 may not be essential for the desilencing of the latent virus genomes, but it may be required for the inhibition of intrinsic cellular defence mechanisms that otherwise efficiently prevent successful virus reactivation. This is in line with the gross disruption of mitochondrial architecture by miR-aU14 via the miR-30-p53-DRP1 axis.
An unusual feature of the miR-aU14 locus is that miR-aU14 is expressed antisense to the U14 ORF from a novel pri-miRNA transcript that initiates within the 5′ region of the U14 ORF. U14 encodes for a G2/M cell cycle checkpoint regulator of HHV-6 38 , which also interacts with p53 39 . Whereas miR-aU14 thus has the potential to repress expression of the important U14 protein, mutational analysis indicates that both U14 and miR-aU14 are important for productive HHV-6A infection. Accordingly, transient transfection of miR-aU14 enhanced productive HHV-6A infection and rescued virus reactivation of a miR-aU14 mutant virus. The most notable finding, however, was that transfection of miR-aU14 triggered virus reactivation from latency at least as efficiently as the commonly employed histone deacetylase inhibitor TSA. Although enhanced mitochondrial fission and impaired intrinsic immunity via the miR-30-p53-DRP1 axis augments successful virus reactivation, miR-aU14 may also target other cellular or viral genes that help trigger virus reactivation from latency.
In summary, our findings reveal a miRNA-mediated mechanism that a prevalent human herpesvirus has hijacked to interfere with intrinsic immunity, govern the lytic-latent switch and augment productive infection. However, viral miR-aU14 should be readily druggable using antisense approaches (antagomiRs) 40 , thereby providing an interesting therapeutic option for preventing herpesvirus reactivation.

Online content
Any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41586-022-04667-4.

Methods
Cell culture U2OS (HTB-96), HeLa (CCL-2) and HUVEC (PCS-100-010) cells were purchased from ATCC and cultured in DMEM medium supplemented with 10% (v/v) FBS and 200 units per ml penicillin-streptomycin 41,42 . HSB-2 T cells were obtained from HHV-6 Foundation, USA and were maintained in RPMI 1640 media 43 . All cell lines were cultured at 37 °C with 5% CO 2 . HEK 293 cells were purchased from ATCC (CRL-1573) and were grown in DMEM medium supplemented with 10% non-heat inactivated FBS and 200 units per ml penicillin-streptomycin. Cells carrying stable GFP expression within mitochondria were developed as mentioned before 12 . Cells stably expressing the mitochondrial targeted GFP were created by cloning the mitochondrial targeted GFP into pLVTHM vector (Supplementary Table 1) backbone and transducing target cells with the lentivirus. All the cell lines were frequently tested for Mycoplasma contamination and were authenticated by sequencing, wherever necessary.

Lentivirus generation and cell transduction
The miR-30c sponge was constructed by cloning the necessary miR-30c sponge oligonucelotide (Supplementary Table 2) into the pLVTHM vector. For mitochondrial GFP-expressing vectors the original GFP cassette of pLVTHM was replaced with a mitochondria-targeted GFP 12 . All the constructs were verified by sequencing. The miR-30c sponge lentivirus were produced in HEK 293 cells as described previously 44,45 . Purified lentiviruses were transduced into target cells in presence of the cationic polymer, Polybrene (Sigma-Aldrich).

HHV-6A virus propagation, BAC reconstitution and virus reactivation HHV-6A virus was grown in HSB-2 cells. For infecting new cells and
propagating virus, 10 6 HHV-6A-infected cells were mixed with uninfected HSB-2 cells, pretreated with 2 ng ml −1 interleukin-2 (IL-2; Sigma) and 5 μg ml −1 phytohemagglutinin (PHA; Sigma), at a ratio of 10:1 that allowed HHV-6A to infect new cells through cell-to-cell fusion. When more than 80% of cells showed cytopathic effects visible under light microscope, infected cells together with the media were collected to prepare cell-free virus. Infected cells were lysed by repeated freeze-thaw cycles three times in liquid nitrogen. Lysed cells with culture media were centrifuged at 3,500g at 4 °C for 1 h. The cleared supernatant was filtered through a 0.45-μm filter. Virus particles were pelleted by centrifugation at 25,000g at 4 °C for 3 h. Virus pellet was resuspended in cold IMDM media without antibiotics and frozen at -80 °C until further use. The HHV-6A titres expressed as the 50% tissue culture infective dose (TCID50) were determined by infecting fresh HSB-2 cells at different dilutions and scoring the number of infected cells that exhibited cytopathic effects or by immunostaining the infected cells using an anti-p41 or anti-gB antibody. Viral titres and TCID50 values were calculated using Reed-Münch formula.
Escherichia coli carrying HHV-6A BAC was obtained from Y. Mori. Virus reconstitution from BACs was performed as described 17,46 . For generating cells that carried latent HHV-6A genomes, U2OS cells were transfected with 5 μg of wild-type HHV-6A or mutant BACs using TransIT-X2 transfection reagent (Mirus Bio). Alternatively, the Amaxa 4D nucleofector transfection system was used. Forty-eight hours after transfection, cells were sorted for GFP-positive cells and grown until cells turned GFP negative indicative of the establishment of latent infection. HHV-6A genome copy numbers were quantified in each clonal population of cells by qPCR. U2OS cells carrying latent HHV-6A were tested for virus reactivation by adding TSA to the cell culture medium for 24-48 h. Cells that reactivated efficiently were used for further studies. TSA 41 at a concentration of 80 ng ml −1 (Sigma, T8552) was used for virus reactivation. Primers used for the mutagenesis study are listed in the Supplementary Table 2.

Generation of recombinant viruses
U14 mutant viruses were generated using pHHV-6A, an infectious BAC clone of HHV-6A (strain U1102) 46 , expressing GFP under the control of the CMV IE promoter, using two-step Red-mediated mutagenesis as described previously [46][47][48][49] . Primers used for the mutagenesis are listed in the Supplementary Table 2. Recombinant BAC clones were confirmed by restriction fragment length polymorphism, PCR and Sanger DNA sequencing. In addition, Illumina MiSeq sequencing was carried out with ~100-fold coverage to exclude unexpected mutations within the entire viral genome. Reconstitution of both wild-type and miR-aU14 mutant viruses was carried out by nucleofection of BAC DNA into Jjhan cells. The expression of eGFP from the viral backbone allowed direct monitoring of the virus reconstitution process.

Average mitochondrial surface area and mitochondrial number analysis
Software and modified algorithm for mitochondrial size and number measurement were previously described by us in detail 12 . All image-processing and analysis steps were performed using Fiji 50 . The numbers of DRP1 fission rings were quantified by processing the confocal micrographs with the Fiji Object Counter plugin after appropriate thresholding. Background was subtracted using the rolling ball background subtraction model. A threshold for detecting DRP1 aggregates was applied to the original micrographs by measuring the mean pixel intensity of the control samples in the DRP1 channel and normalizing the pixel intensity of all other samples to this constant. Additionally, several hundred individual DRP1 rings were measured using the profile function of the ZEN 2012 image-processing platform and the 'plot profile' plugin in Fiji to determine the maximum and minimum ring diameters of constricted and dilated DRP1 rings. Based on this step, the threshold of object detection by the 'object profile' algorithm was set to include DRP1 particles that exhibited a diameter 51-54 between 100 nm (constricted) and 360 nm. Mitochondrial fission sites were defined by profiling regions of low mitochondrial intensity and high DRP1 signals along a linear path through mitochondrial fragments using the plot profile plugin in Fiji.
The average surface area of mitochondrial fragments was measured by a further modification of the 'object count' plugin in Fiji. In brief, while keeping an equal threshold for all images, mitochondrial GFP fluorescence was converted to binary signals, and the algorithm was allowed to numerically categorize the mitochondria as a continuous network or individual fragment and finally determine the area covered by the mitochondrial fragments in micrometres squared. The area was divided by a factor of 0.39 nm and the mean width of HUVEC mitochondria determined by measuring ∼300 individual HUVEC mitochondrial fragments using the profile function of ZEN 2012. Similarly, the average surface area of HeLa mitochondrial fragments was determined by dividing the area by a factor of 0.5 nm. The mitochondrial fragments with no visible connections with neighbouring networks were assigned as individual fragments. DRP1 colocalization with the mitochondria was determined using the COLOC2 plugin from Fiji. The degree of colocalization was ascertained using Pearson's colocalization coefficient followed by statistical analysis.

Immunofluorescence microscopy
A detailed protocol for standard immunofluorescence microscopy has been described 55 . The HHV-6A P41 monoclonal antibody (clone 9A5D12) was obtained from Santa Cruz Biotechnology (SC-65447) (Supplementary Table 3).

Immunoblotting
Immunoblotting was carried out as described before 55,56 using rabbit polyclonal anti-DRP1 antibody as well as monoclonal mouse antibodies against p53 and DRP1 (Supplementary Table 3). Equal protein loading was confirmed by using antibodies against β-actin or GAPDH. The mouse monoclonal antibody against HHV-6 GP82/105 was a gift from HHV-6 Foundation, USA 57 . All the primary antibodies were used at a dilution of 1:1000 and the HRP-conjugated secondary antibodies were used at a dilution of 1:10,000.

Structured illumination microscopy
Structured illumination microscopy (SIM) was performed on the ELYRA S.1 system from ZEISS with a Plan-Apochromat 63×/1.40 oil immersion objective. Laser lines 488 nm, 561 nm and 642 nm were used as excitation source with respective filter sets for GFP, mCherry and ATTO643 to separate the fluorescence light. Twenty-five raw images (five rotations with five phases each) were recorded for reconstruction of one final SIM image with the exception of Extended Data Figure 5f (15 raw images with three rotations). Data processing was done with ZEN software (ZEN 2012 SP5 FP3, ZEISS), including channel alignment by affine transformation for correction of chromatic aberrations using embedded TetraSpeck beads (ThermoFisher, T14792). Brightness and contrast of SIM images were adjusted linearly.

Polyacrylamide gel electrophoresis and Northern blotting
Total RNA extraction for Northern blotting was carried out using TRI-reagent with minor modifications. MgCl 2 (20 mM) was added to the TRI-reagent solubilized lysate before chloroform addition 58 . All the Northern blotting experiments for small RNAs were carried out using 12.5% denaturing urea-PAGE gels containing 0.5× Tris-Borate-EDTA buffer (TBE). Gels were pre-ran for 1 h at 15 mA before sample loading. For every gel, equal amounts of total RNA samples were mixed with 2× RNA loading dye (NEB, B0363S) and were denatured at 95 °C for 3 min before loading into the gel. 16 × 16 cm gels were used for better separation of primary, precursor and mature miRNAs. After gel running, RNA was transferred to positively charged Nylon membrane using 0.5× TBE and wet transfer apparatus (Biometra Tankblot Eco-Line, Analytic Jena). Membranes were UV-cross linked and pre-hybridized for 1 h before addition of denatured probes. Hybridization was carried out for 16-20 h using hybridization buffer (7% SDS and 3% 20× SSC). Membranes were washed three times for 15 min each at hybridization temperature using wash buffer (1.5% 20× SSC, 0.1% SDS). Membranes were exposed to X-ray films along with an intensifying screen at −80 °C for 1-6 days depending upon targets miRNA amounts. Probes for the detection of human miRNAs were developed based on the sequences from miRBase. LNA probes against human miRNA miR-148b, miR-155, Let-7d, Let-7f, miR-30a, miR-30c and miR-30d and miR-30e and U6 were purchased from Exiqon (Qiagen) (Supplementary Table 2). LNA probes were designed against HHV-6A miR-aU14 and were synthesized from Exiqon (Qiagen). RNA oligonucelotide mimics for miR-aU14 and their mutant counterparts were synthesized from IDT DNA Technologies. miR-155 mimics and inhibitors were purchased from Exiqon (Qiagen). Radiolabeled Decade Marker (Thermo Scientific, AM7778) was used in some of the Northern gels as well as gels for in vitro miRNA processing assay.

Generation of inducible miR-aU14 cells
For generating miR-aU14 (both wild-type and mutant) expression vectors based on the pLVTHM backbone (Supplementary Table 1), oligonucelotides were synthesized from Sigma. Details of oligonucelotide sequences are included in the Supplementary Table 2. Detailed methodology for cell line generation has been previously described 12 .
In brief, hairpin loop sequences were designed for the respective miRNA sequences, synthesized (Sigma), annealed and cloned into the pLVTHM 45 using MluI and ClaI sites downstream of H1 promoter sequences. Large-scale preparation of DNA required for lentivirus particle formation was carried out using the Macherey-Nagel endotoxin-free plasmid midi prep system. Third-generation lentiviruses carrying pseudotyped VSV were generated by Ca 3 PO 4 transfection into HEK 293 cells, as previously described 59 . In brief, supernatants containing lentiviruses were collected, passed through a 0.45-μm filter, and concentrated by ultracentrifugation. Concentrated virus particles were used to infect target cells. After establishment of stable cells carrying the integrated lentiviruses, single cell clones were obtained by cell sorting using FACS.

Effects of miR-aU14 on the induction of IFNβ in HEK 293 cells
Fifty-thousand HEK 293 cells were seeded per well. After one day of incubation, cells were transfected with the synthetic RNA oligonucelotides (250 nM) using TransIT-X2. Forty-eight hours later, cells were transfected with 0.5 μg ml −1 3P-hpRNA. For mock transfection, cells were treated with only transfection reagent without DNA or miRNA mimic unless indicated otherwise. After 24 h, cells were harvested using 300 μl lysis buffer from the QuickRNA MicroPrep kit (Zymo Research, Germany). Total RNA was extracted according to the manufacturer's instructions. RNA concentrations were measured by NanoDrop (Thermo Scientific). Five-hundred nanograms total RNA was used for cDNA synthesis with the 5× qRT SuperMix (Bimake). cDNA was then diluted 1:10 and used for qPCR. 2x SYBR Green SuperMix (Bimake) was used for qPCR. The data was evaluated using the ΔΔC q method. Data were normalized with housekeeping ACTB mRNAs.

Quantitative real-time PCR for viral genes
For viral RNA analysis, total cellular RNA extraction was carried out using the Direct-zol RNA purification kit (Zymo Research) or TRI reagent (Sigma). cDNA synthesis was carried out using the Maxima First Strand cDNA Synthesis kit (Thermo Scientific). qPCRs were performed using the PerfeCTa qPCR SuperMix (Quanta Biosciences) on a ROCHE LightCycler-96 system (Roche Life Sciences) using SYBR green chemistry. The primer sequences used for HHV-6A p41 and U90 have been described 41 and can be found in the Supplementary Table 2. NFYC mRNA levels were quantified using TaqMan primer based commercial assays (Supplementary Table 2) and were normalized with housekeeping ACTB mRNAs. NFYC PCR was carried out in 20 μl reaction using TaqMan fast Advanced master mix (Thermo Scientific, 4444557) and QuantStudio 5 real time PCR machine following manufacturer's instructions.

Sequencing of small RNAs
To generate small-RNA sequencing libraries, total cellular RNA was isolated using Qiagen miRNeasy kit. Small-RNA libraries were prepared using the CleanTag Ligation kit (TriLink BioTechnologies) starting with 1 μg of total RNA per library. For quality control, all final libraries were analysed by Bioanalyzer to check for the expected miRNA fragment peak at 141 bp. Subsequently, a Pippin Prep instrument (Sage Science) was used to isolate fragments between 10 and 35 bp in size. Library concentrations were measured by PicoGreen on an infinite F200 instrument (Tecan). Libraries were sequenced on a NextSeq 500 platform (Illumina) using high-output v2 kits with 75 cycles aiming for ~40 million single-ended reads per sample.

Analysis of the small-RNA sequencing data
Reads that passed the chastity filter on the NextSeq 500 were subject to de-multiplexing and trimming of TriLink adapter residuals using Illumina's bcl2fastq v2 software (v2.19.1). Quality of reads was checked with FastQC software (v0.11.5). Reads shorter than 18 bases were discarded. All reads were mapped to combined index of the human genome (HG38) and the HHV-6A genome (X83413), with the trailing repeat region masked by N. Read mapping was performed using STAR (v2.5.3a) using the parameters suggested by the Encode project for sRNA-seq. Reads were then annotated to mature miRNAs if 5′ ends perfectly matched (miRBase 22.1). The raw counts were then normalized and analysed for differential miRNA expression with the DESeq2 (v1.18.1) package. Statistically significantly differentially regulated genes were called by a Wald test for the interaction of the presence of HHV-6A and activation by TSA with false discovery rate (FDR) 1% (Benjamini-Hochberg corrected P-value).

miR-aU14-pre-miR-30c interaction assays
For transfection of radiolabelled miRNA mimics, equimolar amounts of the miR-aU14 mimic and its mutant counterparts were end-labelled separately. After purification through Sephadex G25 column, radioactive counts were measured. Biotinylated synthetic pre-miR-30c-1 was purchased from IDT DNA technologies. For initial experiments, equal counts of each mimic were mixed with biotinylated pre-miR-30c-1 and transfected into HeLa cells using TransIT-X2 reagent. Subsequently, HeLa cells were first transfected with biotinylated pre-miR-30c-1 for 16 h and then cells were washed 2 times with 1× PBS followed by transfection of radiolabelled mimics. Sixteen hours after transfection, total protein lysates were prepared using RIPA lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% Triton X-100, 1% NP-40, 0.1% SDS, 10% glycerol and protease inhibitor cocktail (Roche) and 10 units per ml RNAse inhibitor (NEB, M0314L). Native cell lysates were incubated with Magnetic Streptavidin beads for 1 h at 4 °C. Beads were subsequently washed, and bead-bound RNA was eluted in 2× RNA loading dye (Invitrogen). Northern hybridization was carried out to detect bead-bound biotinylated pre-miR-30c-1 and miR-aU14. A control bait RNA was used in parallel to check non-specific binding of RNA to beads.

Generation of artificial pre-miRNA carrying the miR-30c-1 hairpin loop
Pre-miRNA hairpin of the artificial miR-A or miR-B was expressed in HeLa cells using the pLVTHM lentiviral vector. Two fully artificial hairpin loops were designed for this purpose, which carried a hairpin loop identical to the one in pre-miR-30c (miR-A and miR-B). Oligonucleotide sequences for creating the lentiviral vectors are included within the Supplementary Table 2. miRNA expression was induced by three days of dox (1 mg/ml) treatment. Total RNA was isolated using either TRI reagent (Sigma) or miRNeasy kit (Qiagen) and was subject to Northern blotting.
Flow cytometry analysis of HHV-6A infected cells HSB-2 cells were stimulated with 2 ng ml −1 interleukin-2 (IL-2) (Sigma) and 5 mg ml −1 phytohemagglutinin (PHA; Sigma) for 24 h. Subsequently, pre-stimulated HSB-2 cells were washed twice with PBS and seeded in 6-well plates. Cells were transfected with miR-aU14 mimic, miR-30c mimic or control mimics (1-5 nM) using TransIT-X2 (Mirus Bio) transfection reagent. In parallel, cells infected with an HHV-6A mCherry reporter virus were grown in separate plates. 1 day after transfection of miR-aU14, cells were washed, counted and mixed with mCherry reporter virus infected HSB-2 cells at a ratio of 100:1. Infection was allowed to proceed for two days. Afterwards, cells were washed two times with PBS, fixed with 2% paraformaldehyde (PFA) for 30 min at room temperature. Cells were washed once again with PBS and resuspended in FACS buffer (1× PBS with 0.5% BSA and 0.02% sodium azide), passed through a cell sieve and analysed by flow cytometry.
Flow cytometry was performed using the BD Biosciences FACS Calibur and BD CellQuest Pro software. Analysis was performed using FlowJo 10. In brief, live cells were gated first by SSC vs FSC dot plot analysis followed by visualization of mCherry-positive population using the FL-3 channel (635 nm red diode laser and 670LP filter). The mCherry-positive population was visualized in the third quadrant (Q3), after gating with appropriate experimental controls. Considering the very low rate of cell-to-cell HHV-6A transmission, 60,000 live cell events were collected for each sample.
Transcription start site profiling TSS profiling using differential RNA-seq (dRNA-seq) was performed according to the published protocol 19,20 with some modifications by the Core Unit Systems Medicine (Würzburg). In brief, total RNA was extracted from HHV-6A and mock infected HSB-2 cells. For each sample 3 μg of DNase-digested RNA was treated with T4 Polynucleotide Kinase (NEB) for 1 h at 37 °C. RNA was purified with Oligo Clean & Concentrator columns (Zymo) and each sample was split into an Xrn1 (+Xrn1) and a mock (−Xrn1) sample. The samples were treated with 1.5 U Xrn1 (NEB; +Xrn1) or water (−Xrn1) for 1 h at 37 °C. Digestion efficiency was checked on a 2100 Bioanalyzer (Agilent) and 5′ caps were removed by incubation with 20 U of RppH (NEB) for 1 h at 37 °C. Afterwards, RNA was purified and eluted in 7 μl of nuclease free water and 6 μl was used as input material for the NEBNext Multiplex Small-RNA Library Prep Set for Illumina sequencing. Library preparation was performed according to the manufacturer's instruction with the following modifications: 3′ adapter, SR reverse transcription primer and 5′ adapter were diluted 1:2, 13 cycles of PCR were performed with 30 s of elongation time. No size selection was performed at the end of library preparation. Concentrations of libraries were determined using the Qubit 3.0 (Thermo Scientific) and their fragment sizes were determined using the Bioanalyzer. Libraries were pooled at equimolar concentrations. Sequencing of 75 bp single-ended reads was performed on a NextSeq 500 (Illumina) at the Core Unit Systems Medicine in Würzburg (dRNA-seq). Reads were mapped using the same settings as for miRNA sequencing. TSS were called using iTiSS version 1.2 60 .
AGO immunoprecipitation for miRNA enrichment and sequencing AGO immunoprecipitation was carried out using a previously published protocol 61 with minor modifications. Five-million HSB-2 cells were infected with HHV-6A for 3 days as described before. Then, native cell lysates were prepared from mock-infected and HHV-6A infected cells using lysis buffer (50 mM Hepes, pH 7,5; 150 mM NaCl; 0.5% NP-40; 1 mM NaF; 10% glycerol, 2.5 mM MgCl 2 ; protease inhibitor cocktail (Roche), 0.5 mM DTT; 10 units per ml RNAse inhibitor (NEB, M0314L)). Anti-Flag magnetic beads were washed with PBS and were incubated with TNR6B-Flag peptide 61 overnight at 4 °C on a head-over-tail rotor. Beads were washed once with PBS and once with lysis buffer. Afterwards, half of the beads were incubated with cell lysates overnight at 4 °C on a head-over-tail rotor. The remaining beads were stored at 4 °C. The day after, beads were separated from the lysate on a magnet. Collected cell lysates were then mixed with the remaining TNR6B-Flag beads and were allowed to incubate overnight at 4 °C on a head-over-tail rotor for further protein-antibody binding. Collected beads were washed 4 times with wash buffer (50 mM Hepes, pH 7,5; 150 mM NaCl; 1 mM NaF; 10% glycerol, 2.5 mM MgCl 2 ; protease inhibitor cocktail (Roche), 0.5 mM DTT; 10 units per ml RNAse inhibitor (NEB, M0314L)). Finally, beads were resuspended in PBS containing 0.5 mM MgCl 2 . A small fraction of these beads was used for denatured protein lysate preparation (eluate 1) for immunoblotting. The remaining beads were lysed in TRI-reagent and used for RNA isolation.
Next day, re-incubated beads were processed as described above to obtain the second eluate (eluate 2) and the second batch of AGO-bound RNA. Total RNA from both batches were pooled together and processed for miRNA sequencing. miRNA sequencing was carried out at CeGaT GmbH, Tübingen using NEXTflex Small RNA-seq kit v3 (Bioo Scientific, Germany). miRNA in vitro processing assay miRNA in vitro processing assay was carried out following a protocol from the laboratory of N. Kim 22 with minor modifications. For generation of in vitro-transcribed radiolabelled pri-miRNAs, respective DNA sequences flanking 100 nt on both 5′ and 3′ side to pre-miRNA (from miRbase 22.1) were cloned into pcDNA3 vector in front of T7 promoter. A XhoI restriction site was included immediately at the 3′ end of the pri-miRNA sequence. Primer sequence details for generation of pri-miR-30a, pri-miR-30c, pri-miR-30d and pri-miR-148b are provided in Supplementary Table 2. For generation of in vitro-transcribed radiolabelled pre-miRNAs, primers were designed to allow direct in vitro transcription from a T7 promoter (Supplementary Table 2). An additional G nucleotide was added at 5′ end wherever necessary to allow efficient in vitro transcription of pre-miRNAs. Pri-miRNA carrying plasmids were linearized with XhoI and purified before in vitro transcription. For pre-miRNA transcripts, T7-carrying primers were used to amplify pre-miRNA sequences using 2X Phusion High-Fidelity PCR Master Mix with HF buffer (Thermo Scientific, F531L) and were agarose gel purified before use in the in vitro transcription reaction. 1 μg of linearized plasmid or 30 ng purified pre-miRNA PCR product was used for each 30 μl of in vitro transcription reaction. Each in vitro transcription reaction was carried out using 4 U μl −1 of RNAsin, 0.3 U μl −1 of T7 RNA polymerase, 1 mM each of G, A and C, 0.2 mM U, 3 μCi μl −1 α-32P UTP. After 4 h of in vitro transcription reaction at 37 °C, reaction was stopped by addition of 30 μl of 2× RNA loading dye. Samples were denatured at 95 °C for 5 min and were loaded to a 4% urea-PAGE gel (for pri-miRNA) or a 6% urea-PAGE gel (for pre-miRNA) and were allowed to separate for 2-3 h. Gels were wrapped in plastic wrap and were exposed to X-ray film for 2-3 min. After careful alignment with the X-ray films, gel bands carrying the radiolabelled pri-miRNA or pre-miRNA transcripts were cut out of the gel and were eluted overnight in AES buffer (Acetate-EDTA-SDS buffer) at 4 °C. RNA was precipitated with 100% ethanol in presence of 15 μg Glycoblue (Thermo Scientific, AM9515) as a coprecipitant and was washed in 80% ethanol, air dried and was dissolved in DEPC water. Purified in vitro transcription transcripts were aliquoted and stored at −80 °C till further use.
For pri-miRNA processing assay, HEK 293 cells were transfected with either a mock or flag-DGCR8 constructs (20 μg per 15 cm plate) (Supplementary Table 1) 2 days prior to the experiments. Additionally, Flag-tagged DROSHA construct (2 μg per 15 cm plate) were also transfected together with Flag-DGCR8 for better protein expression and purification. For pre-miRNA processing assay, HeLa cells were seeded to be used for AGO immunoprecipitation. On the day of experiment, cells were scraped out, washed in cold 1× PBS and were lysed in 20 mM Tris-HCl (pH 8.0), 100 mM KCl, and 0.2 mM EDTA. Cell lysate was sonicated for three cycles pulsed at 50% duty cycle for 30 s and incubated on ice for 30 s. Cell lysate was centrifuged at 16,000g, 4 °C and the purified lysate was collected for further use. HeLa cells were also processed similarly. After removing an aliquot of total lysates, Flag magnetic beads were added to the lysate (20 μl slurry) and were incubated on a head to tail rotor for 1 h at 4 °C. For AGO immunoprecipitation, Flag magnetic beads were incubated with TNR6B peptides beforehand and then bead-bound TNR6B peptides were then added to the total lysate. As a mock, empty Flag magnetic beads were added to total HeLa lysate and were processed similarly. After 1 h incubation, beads were washed five times with lysis buffer. Bead bound proteins were stored on ice and were used in in vitro processing assay within 3 h. An aliquot of total lysate and bead-bound proteins were collected in 1× Laemmli buffer for Immunoblotting.
For in vitro processing assay, 30 μl reaction was set up containing 6.4 mM MgCl 2 , 10 4 cpm of radiolabelled pri-miRNA transcript or pre-miRNA transcript, 1 U μl −1 RNAsin, 15 μl of total protein lysate or bead-bound immunoprecipitated proteins. To test effects of the miR-aU14 mimics on in vitro miRNA processing, equimolar amounts of mimics were included in the reaction before addition of the protein lysate. After 90 min of incubation at 37 °C, reaction was stopped by adding 1ml of Tri reagent. RNA was purified in presence of Glycoblue and was finally dissolved in 15 ul of 2× RNA loading dye. RNA samples were denatured and loaded onto a 12.5% urea-PAGE gel for separation together with RNA decade marker. Gel was wrapped in plastic wrap and was directly exposed to X-ray films at −80 °C for few hours to overnight.
TaqMan real-time PCR assay for pri-and pre-miRNA and mature miRNA Total RNA (1-10 ng per reaction) was used for cDNA synthesis for mature miRNA using TaqMan Advanced miRNA cDNA synthesis kit (Thermo Scientific, A28007), which involved 4 steps: poly(A) tailing, adaptor ligation, reverse transcription and amplification (miR-Amp). cDNA synthesis was carried out using the manufacturer's instructions. For pri-/pre-miRNA assay and NFYC mRNA quantification, 1 μg of total RNA was used for cDNA synthesis using Maxima H-minus cDNA synthesis kit (Thermo Scientific, K1651). cDNA was stored at −80 °C till further use. Ready-made commercial TaqMan PCR primers were purchased from Thermo Scientific (Supplementary Table 2) and PCR was carried out in 20 μl reaction using TaqMan fast Advanced master mix (Thermo Scientific, 4444557) and QuantStudio 5 real time PCR machine following the manufacturer's instructions. Data were evaluated using the ΔΔC q method. Cq values for mature miRNA amplification were normalized with U6 snRNA. Cq values for pri-mRNAs and pre-miRNAs were normalized to ACTB. Furthermore, Cq values of virus infected samples were normalized against the same from Mock infected samples and Cq values of TSA treated samples were normalized against Cq values of Mock (vehicle control) treated samples for final figures.

Statistics
All statistical calculations were performed using GraphPad Prism 9.0. Data in graphs represent the mean ± s.d. of three or more independent replicates of an experiment. Statistical significance was calculated separately for each experiment and are described within individual figure legends. For image analysis, six or more biological replicates per sample-condition were used to generate the represented data. qPCR data are representative of 2-4 independent experiments. The results were considered significant at P ≤ 0.05.

Data availability
The sequencing datasets produced in this study have been deposited at the Gene Expression Omnibus with accession number GSE179867. BAC sequencing results have been deposited at the NCBI BioProject database with the BioProject ID PRJNA792929. Raw experimental data have been deposited at Mendeley (https://doi.org/10.17632/grn-z4krxp2.3). Source data are provided with this paper.