rBUNV-Δ7 produces infectious virus in mammalian cells but not in insect cells

To better understand the role of the Gc head/stalk domain in OBV growth, we compared the replication characteristics of wildtype BUNV and the variant rBUNV-Δ7 missing the Gc head/stalk region. To achieve this, we rescued recombinant wildtype BUNV (rBUNV-WT) and the variant rBUNV-Δ7, as previously described [40] (S2A Fig), with successful rescue confirmed by western blot analysis of NP expression in both transfected and infected cell lysates (S2B Fig).

To reflect the two-host enzootic multiplication cycle of OBVs, we compared rBUNV-WT and rBUNV-Δ7 activities in mammalian-derived A549 and BHK cells as well as mosquito-derived C6/36 cells. To assess their multiplication characteristics, we performed a time-course experiment in which A549, BHK and C6/36 cells were infected with either rBUNV-WT or rBUNV-Δ7 at an MOI of 5, in an attempt to infect every cell simultaneously and focus on analysing a single infectious cycle. At 6-, 12- and 24-hours post infection (hpi), cellular lysates and supernatants were harvested (Fig 1). Lysates were probed for expression of NP using western blot analysis (Fig 1A) with subsequent densitometry (Fig 1B) revealing similar NP expression from both rBUNV-WT or rBUNV-Δ7 at all time points measured in A549 and BHK cells. In C6/36 cells, rBUNV-Δ7 NP appeared slightly earlier (Fig 1A–1B), but this was not statistically significant, and overall NP expression of the two viruses at 24 hpi was indistinguishable. This suggested rBUNV-WT and rBUNV-Δ7 were able to enter and perform gene expression across all three cell lines in an equivalent manner.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Comparison of single cycle growth kinetics between wildtype BUNV and ∆7 BUNV in multiple cell lines.

A549 cells (mammalian), BHK cells (mammalian) and C6/36 cells (insect) were either mock-infected or infected with rBUNV-WT (WT) or mutant rBUNV-∆7 (∆7) at an MOI of 5 and supernatants and cell lysates were collected at either 6, 12 or 24 hours post infection (n = 3). (A) Cell lysates were subject to western blot analysis and probed for expression of NP and actin, as a loading control. (B) Densitometry analysis performed on (A) as expression of NP relative to actin. The expression of BUNV NP was normalised to average of WT at 24 hours post infection for each cell line. Results were analysed by Student’s t-test (unpaired) whereby n.s. = p > 0.05, comparing WT to ∆7 at each timepoint. (C) Titration of the supernatants was performed by plaque assay. Log₁₀-transformed results were analysed by Student’s t-test (unpaired) whereby n.s. = p > 0.05, * = p < 0.05; ** = p < 0.01; *** = p < 0.001, comparing WT to ∆7 at each timepoint.

https://doi.org/10.1371/journal.ppat.1014374.g001

Next, we assessed the titre of infectious rBUNV-WT or rBUNV-Δ7 released from the three cell lines (Fig 1C) in harvested supernatants. The same three timepoints (6, 12 and 24 hpi) as for lysate harvesting were used; 6 hpi reflecting an early timepoint at which NP expression is low in A549 and C6/36 cells and no new viruses have been generated [26, 40]. Similarly, 12 hpi has been shown to reflect a timepoint when NP is expressed in all three cell lines and at least one round of infection has been completed [26, 40]. Finally, 24 hpi reflecting a timepoint at which multiple rounds of infection have occurred, infecting cells that were not initially infected. Supernatant titre analysis by plaque assay showed release of infectious rBUNV-WT and rBUNV-Δ7 from mammalian A549 and BHK cells similarly increased over all three timepoints, reaching approximate titres of 10⁷ pfu/mL at 24 hpi. Interestingly, while this was comparable to the release of infectious rBUNV-WT from C6/36 cells, the titre of infectious rBUNV-Δ7 from C6/36 cell supernatants was reduced by around 1000-fold and in fact did not increase beyond the 6 hpi titre suggesting no new viruses were released.

To assess the behaviour of rBUNV-∆7 following multiple rounds of replication, we also infected A549 or C6/36 cells with rBUNV-WT or rBUNV-∆7 at a lower MOI of 0.01. As for MOI 5, titration analysis (S3 Fig) revealed no significant difference in the titres of rBUNV-WT and rBUNV-∆7 released from A549 cells at either 24 hpi or 48 hpi, both reaching 10⁸ pfu/mL by 48 hpi. In contrast, in C6/36 cells, supernatant titres of rBUNV-∆7 were 5-logs lower than that of rBUNV-WT at both 24 hpi and 48 hpi (S3 Fig), reflecting the severe deficiency in virion production seen at MOI 5. Taken together, these results show that removal of head/stalk domain severely disrupts the production of progeny infectious BUNV from insect cells, but not in mammalian cells.

rBUNV-Δ7 fails to establish infection in Aedes mosquitoes

To assess whether the block in rBUNV-∆7 virion generation was specific to insect cells in culture, we sought to test the multiplication characteristics of rBUNV-∆7 in female Aedes aegypti mosquitoes. Briefly, over 20 mosquitoes were fed blood-meals containing either rBUNV-WT or rBUNV-∆7 at 1 × 10⁷ pfu/mL, with day 0 samples collected immediately after feeding to confirm successful ingestion (Fig 2A; Day 0). At 7 days post infection (dpi), mosquitoes were collected and titres were assessed (Fig 2A; Day 7) from the bodies, to determine infection progress, and from the legs and wings, to determine effective dissemination. Of all the mosquitoes fed with a blood-meal containing rBUNV-WT, 10/21 mosquitoes were successfully infected with rBUNV-WT and in 9 of those, BUNV had successfully disseminated to the legs and wings. Of the mosquitoes fed with a blood-meal containing rBUNV-∆7, 0/25 mosquitoes were infected with rBUNV-∆7 and there was no detectable virus in the legs and wings. This was corroborated by western blot analysis of mosquitoes fed with either rBUNV-WT (10 mosquitoes) or rBUNV-∆7 (10 mosquitoes), which showed 3/10 mosquitoes infected with rBUNV-WT expressed BUNV NP, whereas there was no NP expression in any of the mosquitoes infected with rBUNV-∆7 (Fig 2B). This showed that despite ingestion of rBUNV-∆7 into the mosquitoes, no subsequent virus amplification occurred, suggesting that the Gc head is necessary for successful viral multiplication within mosquitoes, corroborating our insect cell culture data.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Comparison of infection of Aedes mosquitoes by wildtype BUNV and ∆7 BUNV.

Aedes mosquitoes were fed a blood meal containing 1x10⁷ pfu/mL of rBUNV-WT or rBUNV-∆7. (A) Bodies or legs and wings were harvested either immediately (Day 0) or 7 days later (Day 7) and titre was calculated by plaque assay. Titres are displayed as pfu/mL and individual samples have been shown (n = 5 for both viruses at Day 0; n = 21 for rBUNV-WT and n = 25 for rBUNV-∆7 at Day 7). The bar represents the average value, with standard error of the mean. Titres were analysed by Mann Whitney test (unpaired) whereby n.s. = p > 0.05, *** = p < 0.001, **** = p < 0.0001, comparing WT to ∆7. (B) Mosquito lysates (10 mosquitoes for each virus) were subject to western blot analysis and probed for expression of NP and actin, as a loading control.

https://doi.org/10.1371/journal.ppat.1014374.g002

rBUNV-Δ7 expresses abundant Gc glycoprotein in insect cells

The results of the previous section showed titres of rBUNV-∆7 failed to rise above inoculum levels in C6/36 cells, and one possible reason for this was the failure to express sufficient glycoproteins for virion production. To assess this, we generated infectious rBUNV-WT and rBUNV-∆7 variants expressing HA-tagged versions of their corresponding Gc proteins, thus allowing abundance of both Gc and NP to be assessed by western blot analysis. The HA tag was introduced into the head region of WT-Gc and the remaining stalk region of ∆7-Gc (S4A Fig) and successful rescue of corresponding viruses rBUNV-WT-Gc-HA and rBUNV-Δ7-Gc-HA was confirmed by detection of NP and Gc-HA by western blot analysis of both transfected and infected cell lysates (S4B Fig). Expression of WT- and ∆7-Gc-HA at their respective expected molecular weights (S4B Fig), suggested that Gc-HA was correctly expressed and processed from the GPC precursor, despite the deletion of the head/stalk domain.

The Gc-HA-tagged viruses were used to investigate expression of NP and Gc in insect cells. Here, C6/36 cells were infected with either rBUNV-WT-Gc-HA or rBUNV-Δ7-Gc-HA at an MOI of 5 and lysate and supernatant samples were collected every 3 hpi until 24 hpi (Fig 3). NP expression occurred slightly earlier in rBUNV-Δ7-Gc-HA-infected C6/36 cells when compared to rBUNV-WT-Gc-HA, but densitometry analysis showed any differences in NP expression levels were not statistically significant at any time point (Fig 3A–3B). Similarly, Gc-HA expression occurred earlier for rBUNV-Δ7-HA compared to rBUNV-WT-Gc-HA, although by 21 and 24 hpi, there was no significant difference in Gc expression levels. Titration of released C6/36 cell supernatants showed that multiplication of rBUNV-WT-Gc-HA was not affected by the addition of the HA tag, since titres were comparable to those of rBUNV-WT, and confirmed that the titres of rBUNV-∆7-Gc-HA were similarly significantly reduced in C6/36 cells (Fig 3D). Together, these data show that rBUNV-Δ7-HA and rBUNV-WT-Gc-HA express NP and Gc structural proteins with equivalent abundance, indicating that the defect in infectious rBUNV-Δ7 virion production occurs at a stage post protein production.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Comparison of growth kinetics and protein expression of wildtype rBUNV-HA and ∆7 rBUNV-HA in C6/36 cells.

C6/36 cells were either mock-infected or infected with rBUNV-WT-Gc-HA (WT-HA) or mutant rBUNV-∆7-Gc-HA (∆7-HA) at an MOI of 5 and supernatants and cell lysates were collected at every 3 hours post infection (hpi) until 24 hpi (n = 3). (A) Cell lysates were subject to western blot analysis and probed for expression of NP, HA (for expression of Gc-HA) and actin, as a loading control. (B) Densitometry analysis performed on (A) as expression of NP relative to actin. The expression of BUNV NP was normalised to the average of rBUNV-WT-Gc-HA NP at 24 hours post infection. (C) Densitometry analysis performed on (A) as expression of Gc-HA relative to actin. The expression of BUNV Gc-HA was normalised to rBUNV-WT-Gc-HA Gc-HA at 24 hours post infection. (D) Titration of the supernatants was performed by plaque assay and titres were plotted on a logarithmic scale graph. All results were analysed by Student’s t-test (unpaired) whereby n.s. = p > 0.05, * = p < 0.05; ** = p < 0.01; *** = p < 0.001, comparing WT-HA to ∆7-HA at each timepoint.

https://doi.org/10.1371/journal.ppat.1014374.g003

rBUNV-∆7 does not accumulate within insect cells or synthesise non-infectious virus-like particles

The lack of infectious rBUNV-Δ7 released into C6/36 cell supernatants could have been due to a defect in virion assembly, or a defect in virion release. To distinguish between these possibilities, we next assessed whether infectious virions were trapped inside cells. A549 and C6/36 cells were infected with either rBUNV-WT or rBUNV-Δ7 at an MOI of 5, after which titres of both supernatant-associated extracellular virus and cell lysate-associated intracellular virus were assessed by plaque assay at 24 hpi. The titres of extracellular rBUNV-WT released from A549 and C6/36 cells were comparable (Fig 4A and 4B, extracellular), and consistent with previous 24 hpi supernatant titres, described above (Fig 1C), as was the statistically significant reduction of titre of extracellular infectious rBUNV-Δ7 in C6/36 cells (Fig 4B; extracellular). Critically, the titre of C6/36 intracellular infectious rBUNV-∆7 was statistically indistinguishable from that of rBUNV-WT (Fig 4B; intracellular), suggesting the dramatic reduction in rBUNV-∆7 supernatant titres was not due to an intracellular accumulation of assembled virions. In A549 cells, there was no significant difference between titres of intracellular or extracellular infectious rBUNV-∆7 and rBUNV-WT (Fig 4A).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Comparison of intracellular and extracellular virus production of wildtype BUNV and ∆7 BUNV in A549 and C6/36 cells.

A549 cells (A) or C6/36 cells (B) were infected with rBUNV-WT (WT; red) or mutant rBUNV-∆7 (∆7; blue) at an MOI of 5 and supernatants and cells were collected at 24 hours post infection (n = 3). Cells were then freeze-thawed three times in 1 × TNE buffer to disrupt the plasma membrane and release any assembled virions that were present within the cell (intracellular). Intracellular and extracellular (supernatant samples) were then titred by plaque assay. Log₁₀-transformed samples were analysed by Student’s t-test (unpaired) whereby n.s. = p > 0.05, **** = p < 0.0001, comparing WT to ∆7 for each sample. (C) Extracellular supernatant from mock-infected C6/36 cells or C6/36 cells infected with rBUNV-WT-Gc-HA or rBUNV-∆7-Gc-HA was purified by ultracentrifugation (n = 2). The resuspended pellet was subject to silver stain analysis (C) and western blot analysis (D), probing for expression of HA and NP. The protein ladder sizes are indicated (kDa) as well as the predicted bands for viral structural proteins WT-Gc, ∆7-Gc, Gn and NP. (E) Densitometry analysis performed on (D) as expression of ∆7-Gc-HA or NP relative to WT-Gc-HA or NP. The expression of the ∆7 proteins were normalised to the average expression of WT-Gc-HA or NP. (F) RNA was extracted from supernatant collected from mock-infected C6/36 cells or C6/36 cells infected with rBUNV-WT-Gc-HA or rBUNV-∆7-Gc-HA (n = 3). The extracted RNA was quantified by RT-qPCR and copy numbers of BUNV S genome per mL were calculated using a standard curve of known RNA concentrations ranging from 2.5 × 10⁷ to 2.5 × 10¹ copies/mL. The dashed horizontal line represents the limit of detection (LOD), which was the lowest concentration of the standard curve (2.5 × 10⁴ copies/mL). Samples that failed to amplify within 25 PCR cycles were assigned the LOD value for statistical analysis. Log₁₀-transformed samples were analysed by Welch’s unpaired t-test whereby ** = p < 0.01; comparing WT to ∆7.

https://doi.org/10.1371/journal.ppat.1014374.g004

As an orthogonal approach to detect accumulated intracellular virions, we directly observed sections of resin-embedded infected C6/36 cells by electron microscopy (S5 Fig). However, we were unable to detect any intracellular virions in C6/36 cells infected with either rBUNV-WT [25] [25] (S5B Fig) or rBUNV-∆7 (S5C Fig).

Finally, to determine whether non-infectious rBUNV-∆7 virus-like particles (VLPs) were released, we sought to examine presence of VLPs in infected C6/36 cell supernatant. Supernatant was collected from C6/36 cells either mock-infected or infected with rBUNV-WT or rBUNV-∆7 and concentrated by ultracentrifugation through a sucrose cushion. The purified supernatant was examined by silver stain analysis (Fig 4C) and western blot analysis (Fig 4D). The silver stain analysis showed bands corresponding to the viral structural proteins Gc, Gn and NP for rBUNV-WT-Gc-HA, which were not clearly identifiable in the mock or rBUNV-∆7-Gc-HA purified supernatants. The western blot analysis clearly showed the presence of Gc-HA and NP in only rBUNV-WT-Gc-HA and rBUNV-∆7-Gc-HA purified supernatant and densitometry analysis (Fig 4E) revealed a five-fold reduction of both Gc and NP when comparing rBUNV-∆7-Gc-HA protein levels to those of rBUNV-WT-Gc-HA. Finally, RT-qPCR was also performed to identify relative levels of the S genome present in the supernatant (Fig 4F). This data showed that there was ~ 10⁷ BUNV S genome copies per mL supernatant from C6/36 cells infected with rBUNV-WT-Gc-HA. In comparison, BUNV S genomic RNA was below the limit of detection (2.5 × 10⁴ copies/mL) in supernatant from C6/36 cells either mock-infected or infected with rBUNV-∆7-Gc-HA. Taken together, this data is consistent with a lack of virus particle production from rBUNV-∆7-infected C6/36 cells.

Overall, our findings are consistent with a scenario in which lack of released infectious rBUNV-∆7 within C6/36 cells results from a deficiency in virion assembly.

Differences between insect and mammalian cells do not explain reduced rBUNV-∆7 virus production from insect cells

To understand the molecular mechanism responsible for the lack of assembly and release of rBUNV-∆7 from insect cells, we considered differences between the mammalian and insect cell lines, including lower culture temperature and lower cholesterol levels in the cellular membrane. To investigate whether reducing the culture temperature affected growth of rBUNV-∆7, A549 and BHK cells were infected with either rBUNV-WT or rBUNV-∆7 at an MOI of 5 and incubated at 30 °C, instead of 37 °C. At 24 hpi, supernatant titres were measured, which showed both rBUNV-WT and rBUNV-∆7 reached equivalent high 10⁶ pfu/mL titres (S6A Fig). This suggests that reducing the culture temperature does not account for the reduction of rBUNV-∆7 titre in insect cells.

Unlike mammals, insects do not synthesise cholesterol, instead relying on dietary sources [23]. Given cholesterol abundance has been shown to regulate BUNV entry [8], we hypothesised that differences in cholesterol incorporation in the cellular membranes of the mammalian and insect cell lines could account for the block in infectious rBUNV-∆7 release. In an attempt to recover rBUNV-∆7 viability, we supplemented C6/36 insect cell cultures with cholesterol during infection using an established cholesterol supplementation protocol [46]. To do this, C6/36 cells were pre-treated with either 0.05 mg/mL or 0.1 mg/mL cholesterol, alongside methyl-β-cyclodextrin (MβCD; 85 µM) as a vehicle control, before infecting with rBUNV-WT or rBUNV-∆7 at MOI 5 and harvesting supernatant at 24 hpi (S6B Fig). Titration of the supernatants showed that supplementing C6/36 cells with cholesterol did not affect rBUNV-WT titre, as expected, nor did it recover rBUNV-∆7 titre. This suggests that the absence of cholesterol in insect cells is not the reason for the reduction in released rBUNV-∆7 titres.

Loss of glycosylation and trimerisation sites does not account for the reduction in rBUNV-∆7 titres from insect cells

WT-Gc and ∆7-Gc sequences differ in their glycosylation potential due to the loss of residue N624 as a consequence of the head/stalk deletion. To investigate whether glycosylation differences could account for reduced rBUNV-∆7 supernatant titres, we removed this glycosylation site from rBUNV-WT by the substitution N624Q to generate mutant rBUNV-N624Q. We hypothesized that if this residue played a critical role in infectious rBUNV-∆7 production, then titres of rBUNV-N624Q from C6/36 cells would also be low. A549 cells or C6/36 cells were infected with rBUNV-WT, rBUNV-∆7 or rBUNV-N624Q at MOI of 1 (due to lower titre of rBUNV-N624Q) and incubated for 24 hours. Titration of the supernatant showed that rBUNV-WT and rBUNV-∆7 were again similar in A549 cells and the rBUNV-∆7 titre was reduced in C6/36 cells (S6C Fig). Interestingly, rBUNV-N624Q titre was reduced in both A549 and C6/36 cells, but not to the same extent as rBUNV-∆7 in C6/36 cells (S6C Fig). Whilst these results show that glycosylation plays a role in viability of BUNV, they also show that glycosylation does not fully explain the lack of assembled rBUNV-∆7 viruses in insect cells.

A final potential difference between rBUNV-WT and rBUNV-∆7 is the lack of Gc head, which mediates trimerisation of the tripod. To determine whether spike tripod formation is important for BUNV virion production in insect cells, we identified potential head trimerisation residues in Gc (S575, R583, Q589 and D592) using PDBePISA (Proteins, Interfaces, Structures and Assemblies) and substituted each for alanine in the rBUNV-WT context to generate infectious mutant rBUNV-∆tripod. As above, we infected A549 and C6/36 cells with rBUNV-∆tripod at an MOI of 1 (due to lower titre of rBUNV-∆tripod) and harvested supernatant at 24 hpi (S6C Fig). There was a slight reduction in rBUNV-∆tripod titre, in both cell lines, when compared to rBUNV-WT but again this was not as significant a reduction as rBUNV-∆7 in C6/36 cells, suggesting the alanine substitutions introduced in an attempt to disrupt Gc head trimerisation do not account for the reduced rBUNV-∆7 viability is reduced in C6/36 cells.

∆7-Gc is transported to the plasma membrane in insect cells

To investigate whether Gc localisation is affected by deletion of the head/stalk region, C6/36 cells were infected with either rBUNV-WT-Gc-HA or rBUNV-Δ7-Gc-HA, fixed at 18 hpi and 24 hpi, stained for NP, Gc-HA and cellular membranes and imaged using confocal microscopy (Figs 5; 18 hpi and S7; 24 hpi). 18 hpi was chosen to reflect a time when WT-Gc expression was detectable but there was a significant difference between the levels of WT-Gc and ∆7-Gc expression, whereas 24 hpi was chosen to reflect a time when the difference between WT-Gc and ∆7-Gc abundance was not significant (Fig 3C). At 18 hpi and 24 hpi, NP distribution in rBUNV-WT-Gc-HA- and rBUNV-Δ7-Gc-HA-infected cells was indistinguishable, presenting as a diffuse cytoplasmic staining (Figs 5A and 5B, and S7A and S7B, green).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. Confocal microscopy images of C6/36 cells infected with wildtype rBUNV-HA or ∆7 rBUNV-HA at 18 hours post infection.

C6/36 cells were infected with rBUNV-WT-Gc-HA (A) or rBUNV-∆7-Gc-HA (B) at an MOI of 5 and fixed with formaldehyde at 18 hpi. The cells were then permeabilized, blocked, and stained for the nucleus (DAPI, blue), concanavalin A (membrane, cyan), BUNV NP (green), and BUNV Gc-HA (red) by indirect immunostaining. The cells were imaged on LSM990 confocal microscope at ×63 magnification. Scale bars representing 20 µm are shown. DAPI, 4,’6-diamidino-2-phenylindole. (C) Co-occurrence analysis was performed by analysing the proportion of Gc signal or NP signal that was present in the same area as a mask representing the plasma membrane of the cells. Approximately 20 cells were selected for each condition to be analysed. All results were analysed by Student’s t-test (unpaired) whereby n.s. = p > 0.05; ** = p < 0.01, comparing WT-HA to ∆7-HA for each protein.

https://doi.org/10.1371/journal.ppat.1014374.g005

However, in contrast to NP, localization of WT-Gc-HA and ∆7-Gc-HA in virus infected cells exhibited clear differences. At 18 hpi, WT-Gc-HA expressed in rBUNV-WT-Gc-HA-infected C6/36 cells appeared to localise exclusively within dense punctate regions, likely Golgi, in agreement with previous work [25] (Fig 5A, red). However, distribution of ∆7-Gc-HA at 18 hpi showed strong staining at the plasma membrane (Fig 5B, red) in addition to the aforementioned cytoplasmic puncta. At 24 hpi, when overall Gc expression levels are similar for WT-Gc-HA and ∆7-Gc-HA, staining of the respective WT-Gc-HA and ∆7-Gc-HA spikes was predominantly within cytoplasmic puncta (S7 Fig, red) however, ∆7-Gc-HA was still abundantly detected at the plasma membrane, with a much weaker signal for WT-Gc-HA. Using the membrane stain, a mask was generated to delineate the plasma membrane and thus permit quantitative analysis of the redistribution of Gc to this site (Figs 5C and S7C). This showed, at both timepoints, that approximately 30% of the ∆7-Gc-HA signal co-occurred with the membrane mask. This was a statistically significant increase over the proportion of WT-Gc-HA signal co-occurred with the membrane mask, which was approximately 10%. In contrast to this, NP from both rBUNV-WT-Gc-HA-infected and rBUNV-∆7-Gc-HA-infected C6/36 cells showed low co-occurrence with the membrane mask, owing to the predominant cytoplasmic distribution, and there was no significant difference between WT and ∆7.

Finally, to test whether ∆7-Gc was indeed present at the surface of the plasma membrane, C6/36 cells were infected with either rBUNV-WT-Gc-HA or rBUNV-Δ7-Gc-HA and fixed at 18 hpi and 24 hpi. Then, without permeabilization, the cells were stained for surface expression of Gc-HA and imaged using widefield microscopy (Fig 6). At both 18 hpi and 24 hpi timepoints, Δ7-Gc-HA showed a strong localisation to the plasma membrane (Fig 6B and 6D), whereas WT-Gc-HA showed a more punctate localisation at the plasma membrane, potentially representing released virions (Fig 6A and 6C). Together, these observations suggest the cellular distribution of Δ7-Gc-HA and WT-Gc-HA is different in insect cells, potentially explaining the differences we observe in viral titres.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. Widefield microscopy images of non-permeabilised C6/36 cells infected with wildtype rBUNV-HA or ∆7 rBUNV-HA.

C6/36 cells were infected with rBUNV-WT-Gc-HA (A and C) or rBUNV-∆7-Gc-HA (B and D) at an MOI of 5 and fixed with formaldehyde at either 18 (A and B) or 24 (C and D) hpi. The cells were not permeabilized, but were blocked, and stained for the nucleus (DAPI, blue) and BUNV Gc-HA (red) by indirect immunostaining. The cells were imaged on an Olympus IX83 widefield microscope at ×100 magnification. Z stack images were collected, and deconvolution (with 3 iterations) was performed. The slice showing the strongest signal was chosen for the display image. Scale bars representing 50 µm are shown. DAPI, 4,’6-diamidino-2-phenylindole.

https://doi.org/10.1371/journal.ppat.1014374.g006

The head/stalk Gc deletion decreases Gc-NP association in insect cells, potentially blocking RNP recruitment

The results described above showed that Δ7-Gc-HA and WT-Gc-HA exhibit differences in their sub-cellular distributions within infected cells, and we hypothesized this could affect interactions with other virion components required for the assembly of infectious virions. To test this, we used co-immunoprecipitation followed by mass spectrometry analysis to identify intracellular binding partners of Δ7-Gc-HA and WT-Gc-HA. Briefly, A549 or C6/36 cells were infected with rBUNV-WT-Gc-HA or rBUNV-Δ7-Gc-HA at an MOI of 1 (due to lower titres of rBUNV-WT-Gc-HA and rBUNV-Δ7-Gc-HA) and incubated for 24 hpi, when cell lysates were collected. The lysates were then incubated overnight with an anti-HA antibody, which was previously incubated with Protein G-coupled magnetic beads. After washing, the beads were collected and subjected to mass spectrometry analysis (Fig 7) or western blot analysis (S8 Fig), to confirm similar protein abundances in the samples. Western blot (S8A–S8B Fig) and densitometry analysis (S8C Fig) suggested that, despite equivalent expression of Gc-HA and NP in the lysate samples (S8A Fig), ∆7-Gc-HA was less efficient than WT-Gc-HA at co-precipitating NP in C6/36 cells but not in A549 cells (S8B Fig). Densitometry analysis confirmed there was a statistically significant reduction in NP precipitated by ∆7-Gc-HA in C6/36 cells so we sought to investigate this further.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Mass spectrometry analysis of proteins identified from immunoprecipitation with BUNV Gc-HA.

A549 cells or C6/36 cells were mock-infected or infected with rBUNV-WT-Gc-HA (WT-HA) or mutant rBUNV-∆7-Gc-HA (∆7-HA) at an MOI of 1 for 24 hours post infection. TMT-MS was then performed as described in methods, and normalised abundances were statistically analysed using the limma package. The volcano plots for the results from C6/36 cells (A) and A549 cells (B) were created by plotting the -log₁₀ p-value of each protein against the log₂ fold change. Proteins where p < 0.05 are highlighted in yellow, and proteins where p < 0.05 and log₂FC is > 1 or <-1 are highlighted in red. (C) Log₂FC, p-value and adjusted p-value (adj p-val) for each viral protein identified in both cell lines is shown. Log₂FC has been colour-coded as a gradient whereby 2 = red and 0 = white. The p-values and adj p-values have been colour-coded whereby green indicates p < 0.01, yellow indicates p < 0.05 and white indicates p > 0.05. N/A means that no peptides were identified for that protein in that cell line.

https://doi.org/10.1371/journal.ppat.1014374.g007

Tandem mass-tag liquid chromatography mass spectrometry (TMT-MS) was performed on bead fractions precipitated from three biological repeats of either mock-, rBUNV-WT-Gc-HA- or rBUNV-∆7-Gc-HA-infected A549 cells or C6/36 cells, allowing relative quantification of the interacting proteins of WT-Gc-HA and ∆7-Gc-HA. The log₂-transformed abundances were normalised to the abundance of shared Gc sequences present in each condition (Fig 7C; Common Gc peptides), to compensate for any potential differences in the immunoprecipitation of WT-Gc-HA and ∆7-Gc-HA. Log₂ fold changes (log₂FC) were calculated for each protein, resulting in positive values (indicating WT-Gc-HA precipitated more protein than ∆7-Gc-HA), values close to 0 (indicating equal precipitation between WT-Gc-HA and ∆7-Gc-HA) or negative values (indicating ∆7-Gc-HA precipitated more protein than WT-Gc-HA) (Fig 7C). Volcano plots were then generated for each cell line to reflect log₂FC changes and significance (p-value) (Fig 7A–7B; S1 Data).

The analysis revealed a significantly higher abundance of NP peptides precipitated from rBUNV-WT-Gc-HA-infected C6/36 cells compared to rBUNV-∆7-Gc-HA-infected C6/36 cells (log₂FC = 2.76, adj p-value = 5.13E-04). Interestingly, the log₂FC of NP peptides precipitated from rBUNV-WT-Gc-HA- compared to rBUNV-∆7-Gc-HA-infected A549 cells was 0.24 (adj p-value = 3.38E-0.1), suggesting there was no significant difference between the precipitation of NP by WT-Gc-HA or ∆7-Gc-HA in A549 cells. Furthermore, this was not due to a lack of NP precipitation by Gc-HA because there was a similar significant enrichment of NP present in both WT and ∆7 A549 infections over mock (WT; log2FC = 3.93, p-value = 3.20E-03, ∆7; log2FC = 4.04, adj p-value = 4.83E-03) (S1 Data). There were also similar changes in the other bunyaviral proteins (L, NSm and NSs) in which the WT-Gc-HA was able to precipitate higher abundances than ∆7-Gc-HA, generally to a greater extent in C6/36 cells than A549 cells (Fig 7C). In contrast to this, Gn had close to 0 log₂FC values in both cell lines, suggesting Gn was equally precipitated by WT- and ∆7-Gc-HA. Finally, the Gc sequences unique to WT-Gc-HA were similarly enriched in both cell lines (C6/36; log₂FC = 1.38, adj p-value = 1.75E-03, A549; log₂FC = 1.44, adj p-value = 3.02E-04; WT Gc peptides in Fig 7C), which is to be expected due to the lack of some of these peptides in ∆7-Gc-HA. Together, this suggests that the ability of ∆7-Gc-HA to interact with NP is dramatically reduced in C6/36 cells, but not in A549 cells, consistent with the species-specific deficiency in insect host virion production.

Cell lines

BHK-21 cells, BSR-T7 cells and A549 cells were acquired from ATCC and maintained in high-glucose Dulbecco’s modified Eagle medium (DMEM; Sigma-Aldrich) supplemented with 10% heat-inactivated foetal bovine serum (FBS), 100 µg of streptomycin/mL and 100 U of penicillin/mL, and incubated in a humidified incubator at 37 °C with 5% CO₂. BSR-T7 cells constitutively expressing T7 RNA polymerase (T7P) [6] were also additionally supplemented with G418 (500 µg/mL) every other passage to maintain the T7P plasmid. C6/36 insect cells were gifted by Andrew Tuplin (University of Leeds) and were maintained at 28 °C in Leibovitz's L-15 media (Thermo Fisher Scientific) which was supplemented with 10% Tryptose Phosphate Broth (Thermo Fisher Scientific) and 100 µg of streptomycin/mL and 100 U of penicillin/mL.

Plasmids

Plasmids encoding the full-length cDNAs that represent the intact S (pT7riboBUNS(+); Genbank accession number: NC_001927.1), M (pT7riboBUNM(+); Genbank accession number: NC_001926.1) and L (pT7riboBUNL(+); Genbank accession number: NC_001925.1) segments of BUNV were gifted by Alain Kohl (University of Glasgow) [5]. Site-directed mutagenesis was used to delete amino acids D480 to N826 from pT7riboBUNM(+) to generate pT7riboBUNMΔ7(+), as previously described [40].

Plasmids pT7riboBUNM-HA(+) and pT7riboBUNMΔ7-HA(+) were synthesised through PCR insertion of the HA tag (YPYDVPDYA), flanked by flexible 4xG/S sequences, into the Gc region (inserted between amino acids T502 and D503 in pT7riboBUNM-HA(+) or between amino acids V482 and Y483 in pT7riboBUNMΔ7-HA(+) (equivalent residues in WT are V829 and Y830). Plasmid pT7riboBUNM-trimer(+) was synthesised through PCR alanine substitution of residues S575, R583, Q589, D592. Plasmid pT7riboBUNM- N624Q (+) was synthesised by PCR substitution of residue N624 to glutamine. All primers sequences are available upon request.

Generation of recombinant BUNV from cDNA

BSR-T7 cells were seeded in 6-well plates at a density of 2 × 10⁵ cells/well. Following overnight incubation, the cells were transfected with 1 µg of pT7riboBUNS(+), 1 µg of pT7riboBUNM(+), 1 µg of pT7riboBUNL(+) and 0.3 µg of pUC57-T7 in 200 µL OptiMEM media, followed by 2.5 µL/µg TransIT-LT1 transfection reagent (Mirus). For recovery of the Δ7 variant, pT7riboBUNM(+) was replaced with pT7riboBUNMΔ7(+). Tagged or mutant versions of viruses involved replacing the appropriate plasmids by their tagged or mutated counterparts. Control transfections were also performed, in which pT7riboBUNL(+) was excluded. At 4-hour post transfection (hpt), the media was removed and replaced with 2 mL of DMEM supplemented with 2.5% FBS. At 120 hpt, the supernatants were collected, clarified and used to infect BHK-21 cells, seeded at a cell density of 5 × 10⁵ cells in T25 flasks the previous day. 48–72 hours post infection (hpi), upon evidence of cytopathic effect, supernatant was collected, centrifuged at 4000 × g for 15 mins, filtered through a 0.45 µm filter and aliquoted into cryo-vials for storage at -80 °C. To prepare bulk stocks of virus, T175 flasks were seeded with 1.5 × 10⁷ BHK-21 cells, incubated for 24 h and then infected with BUNV (or the BUNV variants) at an MOI of 0.01. At 48 hpi, supernatant was clarified and filtered as described above.

Viral protein expression and virus release in mammalian and insect cells

A549 (1 × 10⁵ cells/well), BHK-21 (1 × 10⁵ cells/well) and C6/36 (4 × 10⁵ cells/well) were seeded in 12-well plates and incubated at either 37 °C (mammalian) or 28 °C (insect). The cells were then infected with rBUNV-WT or rBUNV-Δ7 at an MOI of 5 and incubated for a total of 24 hpi. Supernatant and cell lysate samples were collected at 6 hpi, 12 hpi and 24 hpi for determination of virus release and NP expression respectively. To study multiple rounds of infection, the experiment was carried out as above, but the infection was performed using an MOI of 0.01 and supernatant and lysate samples were collected at 24 and 48 hpi. For more detailed analysis, C6/36 (4 × 10⁵ cells/well) were seeded in 12-well plates and incubated at 28 °C. The cells were then infected with rBUNV-WT-Gc-HA or rBUNV-Δ7-Gc-HA at an MOI of 5 and incubated for a total of 24 hpi. Supernatant and lysate samples were collected every 3 hours following infection for determination of virus release and NP and Gc-HA expression respectively.

For collection of intracellular and extracellular virus samples, A549 (1.5 × 10⁵ cells/well) and C6/36 (4 × 10⁵ cells/well) were seeded in 12-well plates, infected with rBUNV-WT or rBUNV-Δ7 at an MOI of 5 and incubated for 24 hours at either 37 °C (mammalian) or 28 °C (insect). Supernatants (extracellular – 1 mL) were collected into cryo-vials and stored at -80 °C. Cells (intracellular) were washed twice with 1 × TNE (10 mM Tris-HCl, pH 7.4, 100 mM NaCl, 1 mM EDTA) and then scraped into 1 mL 1 × TNE and freeze-thawed three times to lyse the cellular membranes. Both extracellular and intracellular samples were titred by crystal violet plaque assay.

To examine BUNV behaviour at lower incubation temperatures, A549 (2 × 10⁵ cells/well) or BHK-21 (2 × 10⁵ cells/well) were seeded in 12-well plates and incubated at 30 °C. The cells were then infected with rBUNV-WT or rBUNV-Δ7 at an MOI of 5 and incubated for a total of 24 hpi. Supernatant samples were collected at 24 hpi for determination of virus titre (n = 1).

To determine the effect of mutating the trimerisation residues or the glycosylation site missing in rBUNV-∆7, A549 (1.5 × 10⁵ cells/well) and C6/36 (4 × 10⁵ cells/well) were seeded in 12-well plates, infected with either rBUNV-WT, rBUNV-Δ7, rBUNV-∆tripod or rBUNV-N624Q at an MOI of 1 and incubated for 24 hours at either 37 °C (mammalian) or 28 °C (insect). The MOI was reduced to 1 due to lower titres of the rBUNV-∆tripod and rBUNV-N624Q stocks. Supernatant samples were collected at 24 hpi for determination of virus titre (n = 2).

Mosquito infection

Female Aedes aegypti mosquitoes (Poza Rica, Mexico) were sorted and housed in cardboard pint cups the day prior to infection. Following a 24-hour starvation period, the mosquitoes were fed a blood meal of washed sheep’s blood containing either rBUNV-WT or rBUNV-∆7 at 1 × 10⁷ PFU/mL, supplemented with 5 mM ATP. Engorged mosquitoes were transferred to new cups with 10% sucrose and kept at 28 °C and 70% humidity for seven days, at which point legs and wings were harvested in microcentrifuge tubes containing glass beads and 250 μL DMEM supplemented with 2% newborn calf serum (NBCS). Bodies were also collected in 250 μL DMEM with 2% NBCS. Samples were ground using a Pro Series Bullet Blender homogenizer for one minute at maximum speed and then centrifuged at 10,000 × g for three minutes. Plaque assays were conducted on all samples to determine viral titres, as follows. Supernatant from each homogenized mosquito sample was serially diluted 10-fold in DMEM and inoculated onto a monolayer of Vero cells. Cells were incubated at 37 °C and 5% CO₂ for one hour prior to being overlaid with 1.5 mL DMEM supplemented with 0.8% agarose, 2% NBCS, and 1% Antibiotic-Antimycotic. Cells were incubated at 37 °C for three days and then fixed with 4% formalin for one hour before the overlay was removed and cells were stained with crystal violet to allow plaques to be counted. Viral titres were quantified by counting plaques on the lowest countable dilution.

Day 0 samples were collected from entire mosquitoes harvested immediately after feeding on the infectious blood meal. Samples were homogenized and clarified as detailed above. Plaque assays were conducted to determine viral titers.

For protein analysis, entire mosquitoes were harvested in 2 × Laemmli buffer 7 days post-infection and ground using a pestle for western blotting.

Cholesterol repletion in insect cells

To understand whether the lack of cholesterol in insect cells was affecting rBUNV-∆7 assembly, C6/36 (4 × 10⁵ cells/well) were seeded on glass coverslips in 12-well plates and incubated for 24 hours at 28 °C. Prior to infection, the cells were pre-treated with either methyl-β-cyclodextrin (MBCD; 85 µM; vehicle control), or 0.05 mg/mL or 0.1 mg/mL cholesterol, for 1 hour. The cells were then washed three times with 1 × PBS. Next, rBUNV-WT or rBUNV-∆7 at an MOI of 5 in serum-free Leibovitz's L-15 media was added to the cells and incubated at 28 °C for 1 hour. The cells were then washed three times with 1 × PBS and incubated for 24 hours at 28 °C in serum-free Leibovitz's L-15 media. Supernatant samples were collected at 24 hpi for determination of virus titre (n = 1).

Virus purification

T175 flasks were seeded with 2.75 × 10⁷ C6/36 cells at least 24 hours prior to infection. Cells were mock-infected or infected with rBUNV-WT-Gc-HA or rBUNV-∆7-Gc-HA, at a MOI of 0.01, for 1 hour in serum free media at 28 °C, after which complete Leibovitz's L-15 media was added. At 48 hpi, supernatants were clarified by low-speed centrifugation and filtered (0.45 µm), followed by pelleting through a 20% sucrose cushion by ultracentrifugation. The virus pellet was resuspended in 0.1 × PBS, supplemented with 1 × cOmplete, Mini, EDTA-free Protease I inhibitor cocktail (Sigma-Aldrich) and incubated overnight at 4 °C. The resuspended pellet was aliquoted and stored at -80 °C. Samples were taken for silver stain analysis and western blot analysis.

Transmission electron microscopy of insect cell slices

C6/36 cells (4 × 10⁶) were seeded in T25 flasks 24 hours prior to infection. Cells were mock-infected or infected with rBUNV-WT or rBUNV-∆7 at an MOI of 5 to ensure every cell was infected. The cells were incubated for 1 h at 28 °C, after which the inoculum was replaced with complete Leibovitz's L-15 media and incubated for a further 24 hours at 28 °C. Cells were fixed with 2.5% glutaraldehyde in PBS for 2 h at room temperature, pelleted by centrifugation (300 × g, 5 mins), washed twice with PBS and post-fixed with 1% osmium tetroxide in PBS for 1 h. Samples were dehydrated with a graded ethanol series, transitioned through propylene oxide and embedded in Araldite as previously described [27]. Resin blocks were polymerised overnight at 60 °C. Ultra-thin sections were prepared using a Reichert-Jung ultracut E ultramicrotome, mounted on 3.05 mm grids and then stained with saturated uranyl acetate and Reynolds lead citrate as previously described [35].

Virus titration

Plaque assays were performed on confluent monolayers of BHK-21 cells. Serial dilutions of BUNV (10-fold) made in serum-free DMEM were added to the cells and incubated for 1 h at 37 °C. The dilutions were removed from the cells and overlaid with 1.2% Avicel (RC-581, FMC Biopolymer). After 3 days incubation at 37 °C, the cells were fixed through the overlay with 8% formaldehyde in water for 10 mins. The fixative was washed off with water and then the cells were stained with 2% crystal violet in 20% ethanol. Plaques were counted and the titre was determined.

Western blot analysis

Cells were lysed with 1 × radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% (v/v) NP40 alternative, 0.5% (w/v) sodium deoxycholate and 0.1% sodium dodecyl sulphate (SDS; w/v)), supplemented with 1  × Halt Protease Inhibitor Cocktail (100 × ; Thermo Fisher Scientific) for 15 min on ice. Lysates were collected; proteins were resolved on a 12% or 15% (immunoprecipitation samples) SDS-PAGE gel and then transferred to a polyvinylidene fluoride (PVDF) membrane. The transfer was performed at 15 V for 30 min, using the Trans-Blot turbo (Bio-Rad). After transfer, the membrane was blocked for 1 h in Odyssey blocking buffer (PBS) (Licor; diluted 1:1 with 1 × PBS containing 0.1% Tween 20 [PBS-T]). Subsequently, the membrane was stained with the primary antibodies (made in 1:4 [blocking buffer: PBS-T]; anti-BUNV NP 1:5000, anti-actin 1:5000, anti-HA 1:5000) for 1 h rocking, at room temperature and then with corresponding secondary antibodies (made in 1:4 [blocking buffer: PBS-T]; all secondary antibodies used at 1:10000) for 1 h at room temperature. The membrane was washed and visualised on the Odyssey M imaging system (Licor). Densitometry analysis was performed using ImageJ over three independent experiments.

Silver stain analysis

Viral pellet samples (10 µL) were resolved by 15% SDS-PAGE and then fixed and stained using the ProteoSilver Kit (Sigma-Aldrich), following the manufacturer’s protocol.

RNA extraction and reverse transcription - quantitative PCR (RT-qPCR)

Culture supernatants (1 mL) were collected at 24 h post-infection (hpi) from C6/36 cells that were mock-infected or infected with rBUNV-WT-Gc-HA or rBUNV-Δ7-Gc-HA (n = 3). Total RNA was extracted from supernatants using the Monarch Total RNA Miniprep Kit according to the manufacturer’s instructions, including the on-column DNase I treatment. The eluted RNA was subjected to an additional in-solution DNase I treatment and subsequently re-purified using the RNA purification columns provided with the kit. RNA was eluted in 23 µL of nuclease-free water.

For complementary DNA (cDNA) synthesis, 10 µL of purified RNA was reverse-transcribed using the LunaScript RT SuperMix Kit, which contains random hexamer primers, according to the manufacturer’s instructions. Quantitative PCR was performed using the Luna Universal qPCR Master Mix on the Stratagene Mx3005P qPCR system (Agilent Technologies) with primers specific for the BUNV S segment (forward: [CACACCACTGGGCTTAGTTAT, 5′–3′]; reverse: [AGTGTAACTTCCCATTCACTTCT, 5′–3′]). Viral RNA copy numbers were determined by absolute quantification using the cycle threshold (Ct) method.

Due to a lack of endogenous reference transcript present in cell-free supernatants, a standard curve was generated from in vitro-transcribed BUNV S segment RNA. Briefly, the pT7riboBUNS(+) plasmid was used as a template for transcription with the HiScribe T7 Quick High Yield RNA Synthesis Kit (NEB) according to the manufacturer’s instructions. Serial dilutions of the resulting RNA transcript were used to generate a standard curve for calculation of viral RNA copy numbers in experimental samples. Only amplification signals with Ct values ≤25 were included in the analysis.

Immunofluorescence analysis

To assess localisation of BUNV proteins during the virus replication cycle, C6/36 cells (4  × 10⁵) were grown on glass coverslips and infected with rBUNV-WT-Gc-HA (MOI = 5) or rBUNV-Δ7-Gc-HA (MOI = 5). At specified time points, the cells were fixed with 4% formaldehyde for 10 min, stained with concanavelin A (5 µg/mL) for 10 min, permeabilised with 0.05% triton-x-100 for 10 min and blocked with 1% bovine serum albumin (BSA) for 1 h. Cells were then labelled with anti-BUNV NP (1:5000) and anti-HA (for Gc staining; 1:5000) for 1 h, followed by labelling with corresponding Alexa Fluor 488 nm or 647 nm secondary antibodies (all secondary antibodies used at 1:5000), respectively, for 1 h. Cells were then washed and mounted onto microscope slides using EverBrite mounting media containing DAPI (Biotium). Stained cells were imaged on a Zeiss LSM 990 confocal microscope at 63 × magnification. The images were processed on Fiji [38]. In order to generate the membrane mask, used in subsequent statistical analysis, a binary image was created whereby the threshold was adjusted to flood the cells with membrane signal and then the MorphoLibJ plugin was used to detect the edges of the cells and create the plasma membrane mask with a 4 pixel internal gradient added [24].

For non-permeabilised samples, all steps were performed as above, with the membrane staining and permeabilization steps omitted, and the images were collected on the Olympus IX83 widefield microscope at 100 × magnification.

Immunoprecipitation of Gc-HA and interacting proteins

A549 or C6/36 cells (1 × 10⁷ cells/flask) were seeded and incubated at either 37 °C (mammalian) or 28 °C (insect). The cells were then infected with rBUNV-WT-Gc-HA or rBUNV-Δ7-Gc-HA at an MOI of 1 and incubated for a total of 24 hpi. The MOI was changed to 1 to reduce the required amount of the rBUNV-WT-HA and rBUNV-∆7-HA stocks due to the lower titres obtained. Cells were washed once with 1 × PBS and lysed in 1 mL immunoprecipitation lysis buffer (25 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 5% glycerol), supplemented with 1 × Halt Protease Inhibitor Cocktail (100 × ; Thermo Fisher Scientific). The cells were lysed for 30 min at 4 °C, after which the lysates were collected and centrifuged at 13,000 × g for 15 min. The lysates were incubated rotating overnight at 4 °C with Dynabead Protein G beads, which had previously been incubated with anti-HA antibody (1:80) in PBS-T (0.02% Tween 20) for 2 h at room temperature and then washed with PBS-T to remove any unbound antibody. After overnight incubation, the beads were drawn out of solution using the DynaMag magnetic rack (Invitrogen), and unbound protein was removed. The beads were subsequently washed three times with PBS-T and then resuspended in 50 µL PBS-T for mass spectrometry analysis.

TMT labelling and high-pH reversed phase (RP) chromatography

Mammalian and insect samples were analysed in separate TMTpro 16plex experiments. Immuno-isolated samples were reduced with TCEP, alkylated with iodoacetamide and digested from the beads with trypsin overnight at 37 °C. Peptides were labelled with TMTpro 16plex reagents according to the manufacturer’s protocol (Thermo Fisher Scientific), pooled and desalted using SepPak cartridges following the manufacturer’s instructions (Waters). Labelled peptides were fractionated by high-pH reversed-phase chromatography using an Ultimate 3000 liquid chromatography system (Thermo Scientific). Samples were loaded onto an XBridge BEH C18 Column (Waters) and peptides were eluted with a 0–95% acetonitrile gradient in 20 mM ammonium hydroxide (pH 10) over 60 minutes. Resulting fractions were concatenated into four pools, dried and resuspended in 1% formic acid prior to analysis by nano-LC MS/MS.

Nano-LC mass spectrometry

Peptides were separated using an Ultimate 3000 nano-LC system in line with an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific). Samples were loaded onto an Acclaim PepMap C18 nano-trap column (Thermo Scientific) and resolved on an Acclaim PepMap C18 reverse phase analytical column (Thermo Scientific) over a 150 min gradient at 300 nl min − 1. All spectra were acquired in data-dependent SPS-MS3 mode using standard settings for TMT quantification. FTMS1 spectra were collected at 120,000 resolution and FTMS3 spectra were collected at 50,000 resolution.

Mass spectrometry data analysis

Raw data files were processed in Proteome Discoverer (Thermo Scientific) using the SEQUEST HT algorithm, and searched against the UniProt human or Aedes albopictus proteomes, together with the UniProt Bunyamwera virus database. Searches allowed a maximum of two missed tryptic cleavages, with carbamidomethylation of cysteine and TMT modification of peptide N-termini and lysine residues as fixed modifications, and methionine oxidation and protein N-terminal modifications as variable modifications. Peptide and protein identifications were filtered to a 5% false discovery rate (FDR).

Mass spectrometry statistical analysis

Protein abundances were normalised by total peptide abundance before being log₂ transformed. Additional normalisation to the abundance of common Gc peptides shared between the WT and ∆7 samples was included to account for differences in Gc concentration between the two samples. Statistical analyses were performed in R using the limma package [1]. A single linear model was fitted across all conditions and moderated t-tests were performed for each comparison of interest with Benjamini-Hochberg FDR correction for multiple testing. PCAs were calculated using the prcomp function. This confirmed that there was consistency across biological repeats and the samples clustered together within different repeats.

Mass spectrometry volcano plots

For each comparison, log₂ fold change was plotted against the negative log₁₀ adjusted p-value. Proteins where p < 0.05 were highlighted in yellow and proteins where p < 0.01 were highlighted in green. Proteins where p < 0.05 and Log₂FC>1, or where p < 0.05 and Log₂FC < -1 were highlighted in red.

Statistical analysis

The statistical significance of data was determined by performing a Student's t test. The statistical significance of the titre data from mosquitoes was determined by performing a Mann Whitney test. The statistical significance of the quantified RT-qPCR data from supernatants was Log₁₀-transformed and analysed by Welch’s unpaired t-test, setting the samples which had no detectable data the value of the limit of detection. Significance was deemed when the values were less than or equal to the 0.05 p value.