CHIKV persistence in mosquito cells is associated with reduced viral protein levels

CHIKV is a positive-sense single-stranded RNA ((+)ssRNA) virus whose genome structurally mimics host messenger RNAs (mRNAs), bearing a 5′ cap and a 3′ poly(A) tail. The CHIKV RNA genome (gRNA) contains two open reading frames (ORFs). The first ORF is translated directly from the genome and encodes the nonstructural proteins (nsPs) required for viral replication, whereas the second ORF is expressed from a subgenomic RNA (sgRNA) transcribed during replication and encodes the structural proteins (sPs) necessary for virion assembly (Fig 1A). To investigate how CHIKV establishes persistent infection in mosquito cells without major effects on host cell physiology, we first analyzed CHIKV infection kinetics in Aedes albopictus U4.4 and C6/36 cells to identify the time point at which viral replication reaches a plateau, hereafter referred as persistence stage, defined by sustained viral RNA and infectious virus levels in the absence of cytopathic effects. U4.4 cells were selected as a physiologically relevant mosquito model with an intact antiviral RNAi response, whereas C6/36 cells, which lack a functional RNAi pathway, were included to determine whether the translational phenotypes observed during CHIKV persistence depend on a functional RNAi pathway. Both cell lines were infected with CHIKV LR2006-OPY1 strain at a multiplicity of infection (MOI) of 4, a condition under which the majority of cells become infected by 8 hours post infection (hpi) in C6/36 cells and by 1 day post infection (dpi) in U4.4 cells (S1A Fig). This strain was selected as it contains the A. albopictus-adapting mutation A226V. Samples were collected at early (3 and 8 hpi) and late time points (1, 2, 5, and 7 dpi). Viral titers (Fig 1B) as well as gRNA and sgRNA levels (Fig 1C) and their corresponding protein products (Fig 1D–1E), were quantified over time. In parallel, cell proliferation and cellular ATP levels were monitored (S1B–S1C Fig).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Kinetic of CHIKV RNA accumulation and protein expression in U4.4 and C6/36 mosquito cells.

(A) Schematic representation of CHIKV RNA genome organization and viral protein expression strategy. Created in BioRender. Diez, J. (2026) https://BioRender.com/8srpphg. (B) CHIKV titers over time in U4.4 and C6/36, determined by plaque assay. (C) Kinetics of CHIKV gRNA and sgRNA levels in U4.4 and C6/36, quantified by RT-qPCR. (D and E) Expression of CHIKV nsP1 and capsid (C) proteins in U4.4 (D) and C6/36 (E) cells over the course of infection. Green and blue colors correspond to U4.4 and C6/36 cells, respectively. All infections were performed at an MOI of 4 and samples were collected at the indicated time post infection. Data points represent the mean ± SEM of 3 independent replicates. Viral RNA levels were quantified using a standard curve generated from in vitro-transcribed CHIKV RNA. Protein expression levels correspond to band intensities normalized to tubulin (Tub) and to the 1 dpi value within each blot. Dotted lines indicate separated lanes in the same membrane. All underlying data can be found in S1 Data and in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003702.g001

Comparative analyses of both cell lines showed that U4.4 exhibited lower viral titers, viral RNA levels and protein expression than C6/36 cells (Fig 1B–1E), consistent with the presence of an active RNAi response. In U4.4, viral titers remained relatively stable throughout the time course, with a peak at 1 dpi. Similarly, but more pronounced, viral titers in C6/36 peaked at 1–2 dpi before declining. In parallel, intracellular viral RNA levels were determined by RT-qPCR. Both CHIKV gRNA and sgRNA levels increased over time and subsequently reached a plateau, after which only minor changes were detected. Specifically, positive-strand gRNA accumulated progressively in both cell lines and reached a plateau by 2–5 dpi in U4.4 and by 5–7 dpi in C6/36 (Fig 1C). A similar pattern was observed for negative-stranded gRNA ((−)gRNA), except that in C6/36 the plateau was reached at 1 dpi (S1D Fig). In contrast, sgRNA plateaued rapidly at 8 hpi in both cell lines and subsequently remained relatively stable. Cell proliferation and cellular ATP levels were monitored throughout infection (S1B, S1C Fig). CHIKV infection induced an early arrest of cell proliferation and maintained ATP levels in C6/36 cells, but not in U4.4 cells. These changes correlated with the higher levels of viral replication observed in C6/36 cells and indicate that high-level CHIKV replication imposes a metabolic and proliferative burden on mosquito cells without causing cell death. We next assessed the expression of the viral proteins nsP1 and capsid (C) (Fig 1D–1E). Notably, protein expression profiles did not mirror viral RNA kinetics. nsP1 expression in C6/36 and in U4.4 cells peaked at 1 and 2 dpi, respectively, and then declined. Capsid expression peaked at 1 dpi in both cell lines and subsequently decreased. This discordance between declining protein levels and sustained viral RNA accumulation is consistent with post-transcriptional regulation, potentially involving inefficient translation of viral RNAs and/or their sequestration in compartments that are inaccessible to RNases and to the translational machinery.

Translation of CHIKV RNAs is repressed in persistently infected mosquito cells

To directly determine whether the distinct CHIKV RNA and protein kinetics were linked to translation regulation, we performed polysome profiling analyses. This approach provides a high-resolution view of the translational landscape by resolving total mRNAs according to ribosomal occupancy, thereby distinguishing untranslated RNAs (free or associated with ribosomal subunits), monosome-associated RNAs (bound by a single ribosome), and polysome-associated RNAs (actively translated by multiple ribosomes). Because these global polysome profiles predominantly reflect the translational status of host mRNAs, which constitute the vast majority of the cellular transcriptome, potential changes in viral RNA translation cannot be inferred from global profiles alone and must be assessed by specifically quantifying viral RNAs across polysome fractions by RT-qPCR. Based on CHIKV protein expression and viral RNA level kinetics in both cell lines (Fig 1C–1E), we selected four time points: 8 hpi, when viral protein expression is first detected; 1 dpi, corresponding to the peak of viral protein expression; and 5 and 7 dpi, when viral protein levels decline.

In U4.4, polysome profiles revealed that CHIKV infection does not induce major changes in global translation, evidenced by comparable monosome peaks (around fractions 9–12) and polysome distribution (around fractions 13–20) in infected and noninfected cells (Fig 2A). In contrast, C6/36 polysome profiles showed that CHIKV infection induces a transient global translational repression, evidenced by an increased monosome fraction and a reduced polysome fraction at 8 hpi (Fig 2B). This repression gradually diminished over time, and by 7 dpi, the translational profiles of infected and noninfected cells were similar, indicating recovery of basal translation. To quantify this effect, we calculated the polysome-to-monosome (P/M) ratio at each time point relative to noninfected controls and normalized it to 1 dpi. Values < 1 indicate a relative enrichment of monosomes over polysomes, consistent with reduced translational efficiency compared to 1 dpi, whereas values > 1 indicate the opposite. In U4.4, P/M ratios in infected cells remained comparable to those of noninfected controls throughout the time course (Fig 2C). In C6/36 cells, the P/M ratio indicated a global translation repression at 8 hpi. This was followed by a modest increase and a subsequent sharp increase of P/M ratios at 5 dpi before returning toward baseline levels at 7 dpi (Fig 2D). To investigate the molecular basis underlying the transient global translational repression observed in C6/36 but not in U4.4 cells, we examined the phosphorylation status of the eukaryotic translation initiation factor 2 subunit alpha (eIF2α) (Fig 2E). Under cellular stress, including viral infection, phosphorylation of eIF2α leads to attenuation of translation initiation and global repression of protein synthesis [17]. However, many viruses modulate this stress response to maintain viral protein synthesis in both vertebrate and invertebrate hosts [18–20]. Whether CHIKV induces eIF2α phosphorylation in mosquito cells, and how this response relates to viral RNA translation, remains unclear. We therefore analyzed eIF2α phosphorylation at multiple time points after infection. In U4.4, cells showed no detectable induction of eIF2α phosphorylation, consistent with the absence of global translational repression in these cells. Conversely, in C6/36 cells, a transient increase in eIF2α phosphorylation was detected between 8 hpi and 1 dpi, coinciding with the period of global host translational repression (Fig 2E). As infection progressed into persistence, eIF2α phosphorylation levels declined and host translation was restored.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Reduced ribosome loading of viral RNAs during persistent CHIKV infection.

(A and B) Representative polysome profile analyses of U4.4 cells (A) and of C6/36 cells (B) at the indicated time points after CHIKV infection. (C and D) Polysome-to-monosome (P/M) ratios reflecting global mRNA translation in U4.4 and C6/36 cells over the course of infection calculated from the area under-the-curve (AUC) of polysome profiles and normalized to noninfected controls and to 1 dpi. (E) Levels of phosphorylated eIF2α in U4.4 and C6/36 cells during CHIKV infection were assessed by western blotting. Protein expression levels were normalized to tubulin (Tub) and to the noninfected value within each blot. Dotted lines indicate separated lanes in the same membrane. (F) P/M ratios of viral CHIKV gRNA and sgRNA in U4.4 and C6/36 cells over the course of infection were determined by RT-qPCR of RNA extracted from polysome fractions. Values are expressed relative to 1 dpi. Bars represent the mean ± SEM of 3 independent replicates. All underlying data can be found in S1 Data and in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003702.g002

To determine whether viral RNAs follow the same or distinct translational trends as host mRNAs, we measured the distribution of CHIKV gRNA and sgRNA across polysome fractions by RT-qPCR (Fig 2F). The viral P/M ratios indicated that, in both cell lines, CHIKV RNAs display translational dynamics that differ from those of the bulk host mRNAs. In U4.4 cells, the P/M ratios of gRNA and sgRNA increased up to 1 and 5 dpi, respectively, and then decreased at later time points, indicating the onset of viral RNA translation repression (Fig 2F, green). In C6/36 cells, gRNA followed a similar but more pronounced pattern, whereas sgRNA showed reduced association with polysomes already from 8 hpi onward (Fig 2F, blue). Of note, the higher P/M values for gRNA and sgRNA coincided with global host translation repression and with a transient rise in eIF2α phosphorylation between 8 hpi and 1 dpi (Fig 2E). As infection progressed into the persistent phase, eIF2α phosphorylation decreased, host translation recovered, and viral RNA translation became repressed. Although viral RNA translation was repressed in both cell lines at late time points, repression appeared earlier and more evident in C6/36 cells. This is likely due to the higher viral RNA levels reached in this cell line as a result of the lack of a functional RNAi pathway, which may increase competition between viral and host mRNAs for the cellular translation machinery, particularly after host translation resumes.

Overall, these results indicate that repression of viral RNA translation is a conserved feature of CHIKV persistence and occurs in the absence or presence of RNAi-mediated antiviral activity.

CHIKV does not induce host transcriptional shut-off or tRNA epitranscriptome remodeling in mosquito cells

To better understand why CHIKV translation is repressed in mosquito cells, in contrast to the translational activation described during lytic infection of human cells [15], we investigated whether two key mechanisms that promote viral translation in mammalian cells are also engaged in the mosquito context. In human cells, CHIKV infection promotes viral translation through two complementary strategies. First, nuclear translocation of the viral nsP2 induces degradation of Rpb1, the catalytic subunit of RNA polymerase II. This degradation results in global host transcriptional shut-off and depletion of host mRNAs [16], thereby reducing competition between host and viral transcripts for the translation machinery. Second, CHIKV remodels the tRNA epitranscriptome, increasing the abundance of the mcm⁵ modification, thereby reprogramming codon optimality and enhancing translation of viral RNAs enriched in codons whose decoding is favored by the mcm⁵-modified tRNA modification [15].

We first examined whether the nsP2-Rpb1 axis is engaged in mosquito cells by assessing the levels of Rpb1 protein. As expected, Rpb1 levels were reduced in CHIKV-infected human HEK293T cells, whereas Rpb1 levels remained nonsignificantly changed in both U4.4 and C6/36 mosquito cells during CHIKV infection (Fig 3A). Consistently, immunofluorescence analysis showed that nsP2 remained confined to the cytoplasm in C6/36 cells (Fig 3B), in contrast to the established nuclear localization reported in mammalian infections [21]. nsP2 localization could not be assessed in U4.4 cells because nsP2 protein levels were below the detection limit of the immunofluorescence assay. Together, these data indicate that CHIKV does not induce nsP2-dependent host transcriptional shut-off in mosquito cells. Next, we investigated whether CHIKV modulates the mosquito translational environment through remodeling of the tRNA epitranscriptome. For this, we analyzed tRNA modification profiles at three relevant stages of infection: 8 hpi, 1 dpi, and 7 dpi. LC-MS/MS analyses revealed no significant changes in any of the monitored tRNA modifications, including mcm⁵ nor its derivative 5-methoxycarbonylmethyl-2-thiouridine (mcm⁵s²), in either U4.4 or C6/36 cells at any time point (Fig 3C and S1 Table).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. CHIKV does not induce Rpb1 degradation or tRNA epitranscriptome remodeling in mosquito cells.

(A) Western blot analysis of Rpb1 protein levels in U4.4, C6/36, and HEK293T cells noninfected and infected with CHIKV. Protein expression levels were normalized to tubulin (Tub) and to the noninfected value within each blot. (B) Immunofluorescence analyses of nsP2 subcellular localization in CHIKV-infected C6/36 cells 1 dpi. Cells were labeled with anti-CHIKV nsP2 antibody (green), J2-dsRNA (red), and nuclei were stained with DAPI (blue). Scale bars correspond to 10 μm. (C) LC-MS/MS quantification of mcm⁵U and mcm⁵s²U tRNA modifications in CHIKV-infected U4.4 and C6/36 cells. Values are expressed relative to those in noninfected. Bars represent the mean ± SEM of 3 independent replicates. All underlying data can be found in S1 Data and in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003702.g003

Collectively, these results indicate that two key CHIKV strategies that promote viral translation during lytic infection of human cells, host transcriptional shut-off and tRNA epitranscriptome remodeling, are not engaged in mosquito cells.

Impaired viral protein production is a conserved feature of arbovirus infection in mosquito cells

To determine whether the viral RNA translational repression observed for CHIKV represents a conserved mechanism among arboviruses in mosquito cells, we extended our analysis to ZIKV. ZIKV is a (+)ssRNA flavivirus with a ~10.7 kb genome bearing a 5′ cap but lacking a 3′ poly(A) tail. The ZIKV genome contains a main ORF that is translated into a polyprotein, which is subsequently processed into individual structural and nonstructural (NS) proteins (Fig 4A). Of note, recent studies have identified several upstream ORFs (uORFs) that also play a role in the regulation of the viral cycle [22]. U4.4 cells were infected with ZIKV at an MOI of 1, and infection kinetics were monitored over a 13 day-period. Viral titers (Fig 4B), viral RNA levels (Fig 4C), and NS3 protein expression (Fig 4D) were quantified at 1, 4, 7, 10, and 13 dpi. Viral titers increased until 7 dpi and subsequently slightly declined. In contrast, viral RNA levels peaked at 4 dpi, showed a modest decrease up to 10 dpi, and increased again at later time points. Importantly, ZIKV NS3 protein peaked at 7 dpi and then declined, despite sustained or increasing viral RNA levels. Thus, ZIKV infection in mosquito cells displays a discordance between viral RNA accumulation and viral protein production, similar to that observed for CHIKV. To exclude the possibility that this reduction in viral protein production reflects a general host stress response rather than a virus-specific translation regulatory mechanism, we examined the phosphorylation status of eIF2α in ZIKV-infected cells. No increase in eIF2α phosphorylation was detected compared with noninfected controls (Fig 4E).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Kinetic analysis of ZIKV infection in U4.4 mosquito cells.

(A) Scheme of ZIKV RNA genome and its protein synthesis strategy. Created in BioRender. Diez, J. (2026) https://BioRender.com/63g7jl4. (B) ZIKV titers over time, determined by plaque assay. (C) Infection kinetics of ZIKV RNA levels determined by qPCR. Data points represent the mean ± SEM of 3 independent replicates. Viral RNA levels were quantified using a standard curve of an in vitro generated ZIKV RNA. (D) NS3 protein expression levels assessed by western blotting. Protein expression levels correspond to the intensity quantification of each band normalized to tubulin (Tub) levels and to the 4 dpi value within each blot. (E) Levels of phosphorylated eIF2α protein levels during U4.4 ZIKV infection, assessed by western blotting. Protein expression levels correspond to the intensity quantification of each band normalized to tubulin and to the noninfected value within each blot. Infection was performed at an MOI of 1 and samples were collected at the indicated time post infection. All underlying data can be found in S1 Data and in S1 Raw Images.

https://doi.org/10.1371/journal.pbio.3003702.g004

Together, these data show that ZIKV, like CHIKV, exhibits reduced viral protein production during persistence in mosquito cells in the absence of detectable activation of the integrated stress response. This supports the notion that regulated limitation of viral protein synthesis is a common feature of arbovirus persistence in mosquito cells.

Cell lines and viral stocks

C6/36 (kindly provided by Andres Merits) and U4.4 (kindly provided by Emilie Pondeville) cells were grown at 28 ºC and cultivated in Leibovitz 15 medium (L15; ThermoFisher, 11415064) supplemented with 10% Tryptose Phosphate Broth (TPB; BD, 260300) and 10% Fetal Bovine Serum (FBS; Merck, F7524-500ML). HEK293T (ATCC, CRL-11268) cells were grown at 37 ºC and 5% CO₂ and cultivated in Dulbecco’s modified Eagle’s medium (DMEM; ThermoFisher Scientific, 41966029) supplemented with 10% FBS and 1% nonessential amino acids (ThermoFisher Scientific,11140035).

Stock of CHIKV LR2006-OPY1 (GenBank: DQ443544, kindly provided by Andres Merits) was generated as in [27]. For production of ZIKV stock (Asian genotype, isolate BeH819016; GenBank: KU365778.1, kindly provided by Andres Merits), 8 × 10⁶ Vero cells were resuspended in 800 μl of PBS and electroporated with 7 μg of in vitro ZIKV generated as in [27] in an electroporation cuvette (0.4 cm gap). Cells were then pulsed three times at 450 V, 25 μF with a Gene Pulser Xcell system (BioRad) and directly resuspended in DMEM supplemented with 10% FBS and 1% nonessential amino acids. ZIKV-containing supernatant was collected every 72 h post-electroporation until cytopathic effect was observed. Viral stock was filtered with a 0.45 μm filter and mixed with all previous aliquots before titration. Virus titers were determined by plaque assay using HEK293T cells as described below.

Viral infections

All CHIKV infections were carried out with an MOI of 4 and incubated for 1 h. ZIKV infection was carried out with an MOI of 1 and incubated for 4 h. Samples were harvested at the indicated time points after infection.

For all C6/36 infections, plates were previously coated with poly-L-lysine (Merck, P1274-100MG). 6-well plates were seeded with 5 × 10⁵ C6/36 cells/well or 3 × 10⁵ U4.4 cells/well. On the next day, cells were washed once with PBS (Cytiva, SH30028.02) and then infected with viral stock. After incubation, the virus was removed, and 2 mL of pre-warmed media was added. At the time of harvest, supernatant was taken and stored at −80 ºC until titration; cells were washed once with PBS and then lysed with either 250 µl ice-cold lysis buffer (10 mM Tris-HCl pH 7.4, 10 mM MgCl₂, 100 mM NaCl, 1% Triton, 2 mM DTT) for RNA extraction or 200 μl of lammeli buffer for protein analyses. For polysome profiling, we seeded 1 × 10⁷ C6/36 cells/T-150 and 3 × 10⁶ U4.4 cells/T-150. At the time of harvest, cells were previously incubated with nonsupplemented media with 100 μg/mL cycloheximide (CHX; Merck, C7698), then washed once with PBS + 100 μg/mL CHX and finally lysed with 700 μl ice-cold lysis buffer + 0.25% deoxycholate and 100 μg/ml CHX. Cells were scraped and transferred to a new tube and snap-frozen with liquid nitrogen. For time points longer than 1 dpi, C6/36, cells were jumped out in fresh media at the time of harvest, and U4.4 were trypsinized for 5 min (ThermoFisher Scientific, 25200056). Then, 1:5th was re-plated in new plates.

Cell viability was tested by following both cell growth and accumulation of intracellular ATP levels using CellTiter-Glo (Promega, G7570).

RNA extraction and RT-qPCR analyses

For further processing, all lysates were centrifuged at 4 ºC for 5 min at 12,000 × g and transferred to new tubes for subsequent analysis. Total RNA was isolated following proteinase K treatment (New England Biolabs, P8107S), phenol-chloropropane extraction, and ethanol precipitation. The pellet was resuspended in RNAse-free water and consecutively treated with 0.1 µl Turbo DNase/μg RNA (ThermoScientific, AM1907) for 30 min at 37º C. Eighteen nanograms (ng) of total RNA was analyzed by TaqMan RT-qPCR using qScript XLT One-Step RT-qPCR ToughMix (Quanta BioSciences, 733-2232). The following primers and probes were used: CHIKV gRNA Fwd (5′-aaccccgttcatgtacaatgc-3′), Rev (5′-gtacctgctcatctgcccaatt-3′), and probe (5′-6Fam-cgggtgcctacccctcatactcgac-TAMRA-3′); CHIKV sgRNA Fwd (5′-aagctccgcgtcctttaccaag-3′), Rev (5′-ccaaattgtcctggtcttcct-3′), and probe (5′-6Fam-ccaatgtcttcagcctggacaccttt-TAMRA-3′); ZIKV Fwd (5′-gtgcaacacgacgtcaacttg-3′), Rev (5′-ggtttgcgaccgcgttt-3′), and probe (5′-6Fam-agagctgtgacgctcccctccca-TAMRA-3′). For the RT-qPCR of the CHIKV (−)gRNA we adapted the protocol described in [28], considering the viral mutations present in the strain used in this work. The following primers and probes were used: CHIKV Fwd T tag T (for RT) (5′-ggcagtatcgtgaattcgatgcgacacggagacgccaacatt-3′), CHIKV Rev T (5′-aataaatcataagtctgctctctgtctacatga-3′), CHIKV (−)gRNA probe (5′-6Fam-tgcttacacacagacgt-TAMRA-3′), and Tag T (5′-aataaatcataaggcagtatcgtgaattcgatgc-3′).

For the standard curve of (+)RNA in vitro CHIKV and ZIKV were generated as in [21] and were serially diluted 1:10 to generate a standard curve for absolute quantification of viral RNA molecules. To generate the standard curve of the (−)gRNA, the same protocol was performed but using the following primers: nsP1 Fwd (5′-atcatggatcctgtgtacgtggac-3′) and nsP1 Rev T7 (5′-taatacgactcactatagtgcgcccgtctggtcctcaa-3′).

Protein analyses

Lammeli lyzed samples were boiled for 5 min at 95 ºC and subsequently loaded onto a 6% or 10% polyacrylamide-SDS gel. Following electrophoresis, proteins were transferred to a membrane using TurboBlot (Biorad) for 7–11 min at 25 V. Then the membranes were blocked with 5% non fat dry milk in TBST for 1 h at room temperature and then incubated overnight at 4 ºC with the primary antibodies. Antibodies used in western blotting were: CHIKV nsP1 (1:1500), CHIKV Capsid (1:1500), CHIKV nsP2 (1:5000), and ZIKV NS3 (1:2000), kindly provided by Andres Merits; Tubulin (1:1000; Merck, T6199-100UL), Rpb1 (1:1000; Active Motif, 39497) and P-eIF2α (1:1000; Cell Signaling, 9721S). The next day, they were washed 3× for 5 min with TBST and incubated at room temperature for 1 h with secondary antibodies (GE Healthcare, NA934V or NA931V) diluted in 5% milk in TBST. Subsequently, the membranes were washed again 3× for 5 min with TBST and exposed to ECL solution (Cytiva, GERPN2134D2 or ThermoFisher, 10005943). Band detection was done using the BioRad ChemiDoc MP Imaging System. Total band intensity volumes of proteins of interest were normalized to the housekeeping gene of the same sample, and the mean of 3 biological replicates was calculated. Viral protein levels were further normalized to the highest intensity time point (1 dpi for CHIKV and 4 dpi for ZIKV). P-eIF2α in infected samples was normalized to the noninfected levels of the protein at each corresponding time point.

Viral titer

Supernatants were titrated using standard plaque assay. Briefly, supernatants for each time point were diluted 1:10 in nonsupplemented DMEM media. HEK293T cells were seeded at 3 × 10⁵ cells/well in 12-well plates and incubated with the diluted supernatants for 1 h at 37 ºC and shaken every 10 min. After incubation, the virus was changed to pre-warmed DMEM + 2% FBS + 1% carboxymethyl cellulose (CMC; Merck, 21902) in a 1:1 ratio and put back into a 37 ºC incubator for 3 days (CHIKV) or 7 days (ZIKV). Then, cells were fixed and stained with crystal violet for 30 min under UV light. Finally, cells were rinsed with warm tap water, and plaques were counted.

Immunofluorescence

For immunofluorescence analysis, we seeded 2 × 10⁵ cells/well for C6/36 and 8 × 10⁴ cells/well for U4.4 into 24-well plates containing sterile glass coverslips coated with poly-L-lysine. Cells were infected with CHIKV as explained before, and at different time points post infection, cells were processed for confocal study. Cells were washed with PBS and fixed with 4% paraformaldehyde (Sigma-Aldrich, P6148) for 15 min. Then, they were permeabilized with PBST (PBS + 0.5% Triton X-100; Sigma-Aldrich, X-100) for 20 min in ice. After a wash with PBS, cells were blocked in blocking buffer (PBS + 10% FBS) for 30 min at room temperature, washed, and then incubated for 1 h with the different antibodies. Using blocking buffer, antibodies were diluted at: 1:10000 for anti-Capsid (rabbit) and 1:1000 for anti-nsP2 (rabbit), kindly provided by Andres Merits, and 1:2500 for anti-double-stranded RNA clone J2 (mouse) (Nordic-MUbio, 10010200). Then, cells were washed twice with PBS and incubated with the secondary antibody anti-rabbit Alexa Fluor^TM 488 goat IgG and anti-mouse Alexa Fluor^TM 568 goat IgG (ThermoFisher, A-11004 and -11008) diluted 1:1000 in blocking buffer for 45 min covered from the light. Cells were washed and immersed in DAPI staining (1 µg/ml, Sigma-Aldrich) for 5 min, covered from the light. Lastly, slides were mounted using pre-warmed Mowiol (Sigma-Aldrich, 32459). Cells were imaged using a Leica SP5 inverted confocal laser scanning microscope and processed with Fiji (ImageJ). For percentage calculations, 4 randomly selected fields per slide were acquired from at least two biological replicates for each infection time point.

Polysome profiling

Sucrose (Sigma-Aldrich, 84097) solutions at 10% and 50% were prepared dissolved in polysome buffer 2.0 (20 mM Tris-HCL pH 7.5, 10 mM MgCl₂, 100 mM NH₄Cl), and supplemented with CHX 100 µg/ml and DTT 2 mM. 6 mL of the 10% sucrose gradient was added to SW41Ti tubes (Beckman, 344059) and then 6 mL of 50% sucrose was underlaid using a needle. To get a 15%–45% linear gradient, we used the Gradient Master. Gradients were stored at 4 ºC overnight. Lysates were thawed on ice and centrifuged for 5 min at 12,000 × g. 400–700 µL of lysate coming from the same number of cells, were loaded onto the linear sucrose gradients, and centrifuged for 3 h at 35,000 rpm at 4 ºC in a Beckman SW41 rotor. BCA quantification was performed to normalize the amount of sample loaded into the gradient. Gradients were immediately fractionated into 20 fractions using a Biocomp Fractionator. Fractions of 500 µl were collected in 1.5 mL RNAse-free tubes with 55 µL of 10% SDS.

Fractions were extracted using phenol-chloroform and precipitated with ethanol. RNA of fractions corresponding to 40S, 60S, and 80S, and polysome peaks were pooled together when dissolving with RNAse-free water. Samples were DNase-treated before analyzing by qPCR. Viral RNA was assessed as described above using TaqMan qPCR.

Quantification of tRNA modifications by LC-MS/MS

3–10 μg of total RNA per sample was run in a 15% TBE-UREA gel (Novex, ThermoFisher Scientific, EC6885BOX) for 60 min at 180 V Gels were stained with 1:10000 SybrGold (ThermoFisher Scientific, S11494) and the tRNA bands were selected. tRNA was then extracted using the ZR small RNA PAGE Recovery kit (Zymo Research, R1070) following the manufacturer’s instructions. The same amount of extracted tRNA was digested with 1 μl of the Nucleoside Digestion Mix (New England BioLabs, M0649S) and the mixture was incubated at 37 ºC for 1 h. Samples (15 ng) were analyzed using an Orbitrap Eclipse Tribrid mass spectrometer (ThermoFisher Scientific, San Jose, CA, USA) coupled to an EASY-nLC 1200 (ThermoFisher Scientific (Proxeon), Odense, Denmark). Ribonucleosides were loaded directly onto the analytical column and were separated by reversed-phase chromatography using a 50 cm column with an inner diameter of 75 μm, packed with 2 μm C18 particles (ThermoFisher Scientific, ES903). Chromatographic gradients started at 93% buffer A and 3% buffer B with a flow rate of 250 nL/min for 5 min and gradually increased to 30% buffer B and 70% buffer A in 20 min. After each analysis, the column was washed for 10 min with 0% buffer A and 100% buffer B. Buffer A: 0.1% formic acid in water. Buffer B: 0.1% formic acid in 80% acetonitrile. The mass spectrometer was operated in positive ionization mode with nanospray voltage set at 2.4 kV and source temperature at 305 °C. For the Parallel Reaction Monitoring (PRM) method, the quadrupole isolation window was set to 1.4 m/z, and MSMS scans were collected over a mass range of m/z 50–300, with detection in the Orbitrap at a resolution of 60,000. MSMS fragmentation of defined masses (S1 Table) was performed using HCD at NCE 20 (except stated differently) [29], the auto gain control (AGC) was set to “Standard” and a maximum injection time of 118 ms was used. In each PRM cycle, one full MS scan at a resolution of 120,000 was acquired over a mass range of m/z 220–700 with detection in the Orbitrap mass analyzer. AGC was set to 10⁵ and the maximum injection time was set to 50 ms. Serial dilutions were prepared using commercial pure ribonucleosides (0.005–150 pg, Carbosynth, Toronto Research Chemicals) in order to establish the linear range of quantification and the limit of detection of each compound. A mix of commercial ribonucleosides was injected before and after each batch of samples to assess instrument stability and to be used as an external standard to calibrate the retention time of each ribonucleoside.

Acquired data were analyzed with the Skyline-daily software (v24.1.1.284) and extracted precursor areas of the ribonucleosides were used for quantification.