ALTO co-operates with STING to induce IFNβ expression

To test whether individual MCPyV encoded-proteins could modulate the STING signaling pathway, we co-transfected HEK293 cells with a STING expression construct and either a construct expressing one of the MCPyV-encoded proteins or an empty vector. A vector expressing RFP was used as a negative control for STING. A highly active STING^R284S mutant [36] was used in the first experiment to allow easy detection of the STING downstream signaling. Compared to the RFP control, the STING^R284S mutant significantly activated IFNβ mRNA expression (Figs 1A and S1). More importantly, compared to the empty vector, transfection with the ALTO expression construct dramatically increased IFNβ mRNA expression induced by STING^R284S (Fig 1A). In contrast, all other MCPyV-encoded genes showed no effect on STING downstream IFNβ expression (S1 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. ALTO co-operates with STING to stimulate IFNβ expression.

A. HEK293 cells were transfected with plasmids carrying ALTO or an empty vector control (CO) and STING^R284S (or RFP control). Two days post-transfection, mRNA levels of IFNβ were measured using RT-qPCR. One of the values for transfection with vector and RFP was set as 1. B. HDF-inRFP and -inALTO cells were mock-treated or induced with Dox and treated with diABZI (or DMSO control) for 19 hours. mRNA levels were measured using RT-qPCR. The value from HDF-inRFP cells mock- and DMSO-treated was set as 1. Error bars indicate the standard deviation from three independent samples. ***p<0.001.

https://doi.org/10.1371/journal.ppat.1012170.g001

We then tested how ALTO expression impacts the STING pathway in primary HDFs that support MCPyV infection [28]. We generated stable HDF cells to allow inducible expression of either RFP (inRFP) or ALTO (inALTO) by Doxycycline (Dox). Dox treatment was then used to induce RFP or ALTO expression in these cells (Fig 1B). In MCPyV-infected cells, ALTO is likely to function in the presence of replicated viral DNA, which could stimulate the STING-TBK1 pathway. Therefore, we sought to simulate this biological setting by treating HDF-inRFP or -inALTO cells with and without Dox in the absence and presence of the STING agonist diamidobenzimidazole (diABZI). In the HDF-inRFP cells, diABZI treatment could stimulate IFNβ expression, but RFP expression does not have additional effect (Fig 1B). However, in HDF-inALTO cells, while both diABZI treatment and ALTO induction were found to moderately up-regulate IFNβ mRNA expression, combining ALTO induction with diABZI treatment resulted in a significant increase in IFNβ expression (Fig 1B). Together, our studies showed that ALTO can stimulate STING downstream signaling.

MCPyV ALTO stimulates TBK1 autophosphorylation and STING stabilization

We next tested if ALTO stimulates IFNβ expression by altering the status of STING and its downstream components including TBK1, IRF3, and NF-κB. HDF-inRFP or -inALTO cells were treated with diABZI as a positive control to show how STING activation changed the phosphorylation status of these factors (Fig 2A). As expected, in HDF-inRFP and -inALTO cells, diABZI treatment led to a substantial activation of STING and downstream signaling as indicated by TBK1^S172 autophosphorylation and phosphorylation of STING^S366, and IRF3^S386, (Fig 2A, lanes 2 and 6). STING phosphorylation is required for its trafficking and subsequent ubiquitination and degradation [37,38]. Consistently, we observed a sharp reduction of STING protein level in diABZI-treated cells (Fig 2A, left panel, lanes 2 and 6). However, in HDF-inALTO cells treated with only Dox but not with diABZI, we detected an increased level of STING but not the phosphorylated STING^S366 (Fig 2A, compare lane 7 to lanes 1, 3, and 5). Immunofluorescent (IF) staining confirmed that, compared to HDF-inRFP cells or mock-induced HDF-inALTO cells, more STING was accumulated in Dox-induced HDF-inALTO cells both in the presence and absence of diABZI (Fig 2B). More strikingly, TBK1^S172 phosphorylation was markedly elevated by ALTO expression even in the absence of diABZI (Fig 2A, left panel, compare lane 7 to lanes 1, 3, and 5). TBK1^S172 phosphorylation was not detected in HDF-inRFP cells treated with Dox (Fig 2A, left panel, lane 3), suggesting that ALTO expression was responsible for inducing TBK1^S172 phosphorylation. Dual treatment of HDF-inALTO cells with Dox and diABZI did not significantly enhance TBK1^S172 autophosphorylation over Dox treatment alone (Fig 2A, left panel, compare lanes 7 and 8). The level of total TBK1 protein was reduced in the HDF-inRFP cells treated with diABZI (Fig 2A, left panel, compare lanes 2 and 4 with lanes 1 and 3), consistent with normal degradation following activation [39]. In HDF-inALTO cells, the level of TBK1 protein was further reduced by ALTO expression, both with and without diABZI treatment (Fig 2A, left panel, compare lanes 7–8 to lanes 2, 4, and 6). TBK1^S172 autophosphorylation is needed for the suppressor of cytokine signaling 3 (SOCS3)-mediated proteasomal degradation [39]. Therefore, it is likely that ALTO-induced TBK1^S172 autophosphorylation stimulated its degradation, thereby reducing the amount of TBK1 available for STING phosphorylation. Because STING phosphorylation triggers its degradation [37,38], the lack of STING phosphorylation could inhibit its degradation and cause the observed accumulation of STING in the ALTO-expressing cells. Consistent with a recent study [8], we also found that ALTO expression modestly enhanced p65^S536 phosphorylation (Fig 2A, right panel) and promoted the nuclear localization of p65 (S2 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. ALTO stimulates TBK1^S172 autophosphorylation and STING stabilization.

A. HDF-inALTO or -inRFP cells were mock-treated or induced with Dox for 24 hours, then stimulated with diABZI (or DMSO control) for 16 hours. Whole cell lysates were resolved by SDS/PAGE and immunoblotted with the indicated antibodies. B. HDF-inALTO or -inRFP cells were mock-treated or induced with Dox for 7 hours, then stimulated with diABZI (or DMSO control) for an additional 19 hours. Cells were fixed and immune-stained for STING, and counterstained with DAPI. Scale bar, 20 μm. C. HDF-inALTO or -inRFP were mock-treated or induced with Dox for 40 hours, then stimulated with diABZI (or DMSO control) for the indicated times. Whole cell lysates were resolved by SDS/PAGE and immunoblotted with indicated antibodies. Red asterisks indicate phosphorylated STING.

https://doi.org/10.1371/journal.ppat.1012170.g002

To mimic the biological setting in which ALTO function might be affected by the dynamics of MCPyV infection, we treated HDF-inRFP or -inALTO cells with and without Dox in both the absence and presence of diABZI for varying amounts of time. In both types of cells not treated with Dox, diABZI stimulated the phosphorylation of STING, TBK1^S172, and IRF3^S386(Fig 2C). In HDF-inALTO cells, Dox-induced ALTO expression in the absence of diABZI again led to a significantly increased TBK1^S172 phosphorylation, reduced TBK1 protein level, and stabilization of STING (Fig 2C, compare lane 19 to lane 13). During the earlier time points (0, 0.5, 1h) of diABZI treatment of Dox-induced HDF-inALTO cells, there was a pronounced increase in STING protein level, but a reduction in phosphorylated STING compared to HDF-inRFP cells or HDF-inALTO cells not treated with Dox (Fig 2C, Compare lanes 19–21 to lanes 1–3, 7–9, and 13–15). This is consistent with our observation that ALTO can stabilize STING but does not stimulate its phosphorylation (Fig 2A). As discussed above, this observation could be attributed to the fact that ALTO expression greatly reduced the level of TBK1 needed for STING phosphorylation (Fig 2C, compare lanes 19–24 to lanes 13–18). Interestingly, after Dox-induced HDF-inALTO cells were treated with diABZI for longer periods (2, 3, 4 h), phosphorylated STING levels substantially increased (Fig 2C, compare lanes 22–24 to lanes 4–6, 10–12, and 16–18). It is possible that at these later time points, STING accumulated in these cells was eventually phosphorylated by either the residual TBK1 present in the cells or alternative kinases.

We also found that even a trace amount of ALTO expressed for a very brief period (4h) could lead to robust TBK1^S172 autophosphorylation and STING accumulation (S3 and S4 Figs), suggesting that this ALTO function may play a physiologically important role in MCPyV infected cells, which was confirmed in our studies described below. TBK1 protein level dropped strikingly after extended ALTO induction and TBK1^S172 autophosphorylation (S4 Fig), supporting the role of TBK1^S172 autophosphorylation in stimulating its ubiquitination and degradation [39]. By co-staining STING with ER and Golgi markers, we demonstrated that while ALTO can stabilize STING, it does not alter STING trafficking from the ER to the Golgi when stimulated by diABZI (S5 Fig). Together, our studies suggest that ALTO first induces TBK1 autophosphorylation and reduction, subsequently causing STING accumulation and activation.

ALTO forms a complex with STING and TBK1

To understand how ALTO could so effectively stimulate TBK1 autophosphorylation and STING stabilization, we decided to test if this viral protein interacts with STING and TBK1. We first performed ALTO Co-IP using HDF-inALTO cells induced with Dox. While STING was not clearly Co-IPed with ALTO, both TBK1 and phosphorylated TBK1^S172 were found to be efficiently pulled down with ALTO antibody (Fig 3A). C-terminal HA- and FLAG-tagged ALTO could also efficiently Co-IP phosphorylated TBK1^S172 (S6 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. ALTO interacts with STING and TBK1.

A. HDF-inALTO cells were induced with 0.01μg/mL Dox for 40 hours. Whole cell lysates were incubated with normal rabbit IgG or anti-ALTO antibody and immunoprecipitated with Protein G agarose. Co-immunoprecipitants were resolved by SDS/PAGE and immunoblotted with the indicated antibodies. B. Schematic diagram depicting the NanoBiT and BiFC structural complementation systems. NanoBiT reversibly reconstitutes NanoLuciferase and produces luminescence in live cells, whereas BiFC irreversibly reconstitutes Venus to produce GFP fluorescence. C. HEK293 cells were transfected with pairs of plasmids carrying NanoBiT-tagged ALTO, STING, and TBK1 or the Small BiT negative control construct. At 20 hours (ALTO-TBK1) or 24 hours (ALTO-ALTO and ALTO-STING) post-transfection, NanoGlo reagent was added, and luminescence was measured after a 15-minute (ALTO-ALTO) or 2-hour (ALTO-STING and ALTO-TBK1) incubation. Points indicate replicate wells from a single experiment, bars indicate means, and error bars indicate standard deviations. L-protein and protein-S (and similar) indicate the named protein tagged with the LgBiT at its N-terminus or SmBiT at its C-terminus, respectively. ***p<0.001; **p<0.01; *p<0.05. D. U2OS cells were transfected with pairs of plasmids carrying FLAG-tagged Venus (N-terminal half or C-terminal half) fused ALTO constructs. At 48 hours post-transfection, cells were fixed and immunostained for FLAG, and counterstained with DAPI. E. U2OS cells were transfected with pairs of plasmids carrying FLAG-tagged Venus (N-terminal half or C-terminal half) fused ALTO and TBK1, respectively. At 36 hours post-transfection, cells were fixed and immunostained for FLAG, and counterstained with DAPI. F. U2OS cells were transfected with pairs of plasmids carrying FLAG-tagged Venus (N-terminal half or C-terminal half) fused ALTO and STING, respectively. At 24 hours post-transfection, cells were fixed and immunostained for FLAG, and counterstained with DAPI. Scale bar, 20 μm.

https://doi.org/10.1371/journal.ppat.1012170.g003

We then applied two structural complementation-based cellular approaches to examine the interaction between ALTO, STING and TBK1: NanoLuc Binary Technology (NanoBiT) and bimolecular fluorescence complementation (BiFC) (Fig 3B). The NanoBiT system is a split luciferase-based reporter assay designed to sensitively detect protein-protein interactions in live cells [40]. Proteins of interest are tagged with a piece of Nano Luciferase, called Large (LgBiT) or Small BiT (SmBiT), positioned at the N- or C-terminus. When LgBiT and SmBiT are brought into close proximity by a protein-protein interaction, the Nano Luciferase is reversibly reconstituted and will produce luminescence in the presence of the cell-permeable substrate furimazine (Fig 3B). To assess ALTO’s interaction with STING and TBK1 in this system, we generated fusions with all possible combinations of BiT size and location for all proteins, for a total of four constructs per protein. All combinations of LgBiT and SmBiT pairs were screened for all interacting pairs. Interestingly, we found that ALTO pairs tagged on the same terminus (i.e., LgBiT-ALTO:SmBiT-ALTO, and ALTO-LgBiT:ALTO-SmBiT) produced relative luminescence readings several hundred times higher than negative interaction controls (Fig 3C). These data demonstrate that ALTO strongly interacts with itself in living cells. The two opposite-terminal pairs, ALTO-LgBiT:SmBiT-ALTO and LgBiT-ALTO:ALTO-SmBiT also luminesced significantly more than their LgBiT negative control; however, the overall magnitude of the luminescence was much lower than the same-terminal pairs, and C-terminal tagged ALTO was apparently expressed at a notably lower level than N-terminal tagged ALTO (S7A Fig). Since ALTO has a predicted C-terminal transmembrane domain, this result indicates that membrane localization and orientation may impact ALTO’s oligomerization. SmBiT-ALTO was also found to positively interact with both LgBiT-TBK1 and STING-LgBiT at levels significantly above both negative controls and opposite-terminal controls (Fig 3C). This interaction is considered genuine, as the luminescence readings for SmBiT-ALTO:LgBiT-TBK1 and SmBiT-ALTO:STING-LgBiT were more than tenfold higher than the average from the LgBiT-tagged construct co-transfected with the SmBiT negative control construct, thereby meeting the manufacturer’s criteria for a confirmed protein-protein interaction and demonstrating the presence and specificity of these interactions. NanoBiT-fused TBK1, but also to a lesser degree NanoBiT-fused STING, were not expressed as efficiently as the ALTO fusion proteins, which likely explains the overall lower relative luminescence in these assays (S7B and S7C Fig). Taken together, these data show that, in addition to oligomerization, ALTO is in close proximity to TBK1 and STING in cells and is therefore likely to form a complex with these host proteins.

The ALTO-ALTO NanoBiT interaction signal is significantly higher than those for ALTO-TBK1 and ALTO-STING. This is likely because ALTO tends to self-oligomerize and co-localize within cells (see below), leading to a constant proximity of the tagged proteins and a higher likelihood of reconstituting the luciferase enzyme. In contrast, interactions between ALTO and TBK1 or STING are more transient and less frequent due to the dynamic nature of their interactions and the potential for ALTO to reduce TBK1 levels. Additionally, the untagged endogenous TBK1 and STING might preferentially interact with ALTO, thereby reducing the luminescence signals for ALTO-TBK1 and ALTO-STING NanoBiT. Based on these observations, we decided to further investigate the ALTO-ALTO, ALTO-STING, and ALTO-TBK1 interactions using a complementary approach, BiFC. This method renders the interaction between the proteins irreversible, thus making it more easily observable.

In a BiFC assay, two interacting protein partners are fused individually to complementary nonfluorescent fragments of a fluorescent reporter Venus and expressed in live cells [41] (Fig 3B). The Interaction of these two proteins will bring the nonfluorescent fragments within close proximity, allowing the Venus protein to reform its native three-dimensional structure and emit fluorescent signals [42]. Therefore, this technique enables direct visualization of protein interactions within the cell through the intensity and distribution of the fluorescent signal [41]. To test the ALTO-ALTO, ALTO-STING, and ALTO-TBK1 protein interactions, these molecules were fused individually to either the N-terminal portion (VenusN) or C-terminal portion (VenusC) of the divided Venus protein. Coexpression of constructs encoding ALTO with N-terminal tagged VenusN and VenusC generated a bright BiFC signal, which is absent in cells co-transfected with constructs encoding ALTO with VenusN and VenusC fused to its C-terminus (Figs 3D, S8 and S9). This result provides additional evidence to demonstrate that ALTO can oligomerize in cells. Co-expression of VenusN-ALTO with TBK1-VenusC or VenusC-ALTO with STING-VenusN also generated strong BiFC positive signals (Fig 3E and 3F). Switching the VenusC tag from the C to the N-terminus of TBK1 or the VenusN tag from the C to the N-terminus of STING abolished the BiFC signal (Fig 3E and 3F). Since the differently tagged ALTO, STING and TBK1 were expressed equally well in cells (S9–S11 Figs), they served as excellent negative controls to support that the BiFC signals we observed likely resulted from true ALTO-ALTO, ALTO-TBK1 and ALTO-STING interaction, and not from overexpression of the proteins tested (Figs 3D, 3E, 3F and S9–S11).

ALTO recruits Src to the TBK1 complex to stimulate its autophosphorylation

To better understand how ALTO could stimulate TBK1 autophosphorylation, we examined the ALTO protein sequence and identified 13 proline-X-X-proline (PXXP) motifs (Fig 4A). Interestingly, the STING sequence also contains one PXXP motif, which has been shown to bind to the SH3 domain of the tyrosine kinase Src [43]. This binding recruits the kinase to phosphorylate TBK1^Y179, stimulating TBK1 autophosphorylation and activation [43]. Since the 13 PXXP motifs present in ALTO are mostly located on the disordered surface of the ALTO structure as predicted by AlphaFold2 [44] (Fig 4A), we tested whether ALTO could interact with Src. C-terminal HA- and FLAG-tagged ALTO was able to pull down a small amount of Src (S6 Fig). In the reciprocal Co-IP experiment, Src antibody Co-IP was performed using HDFs inducibly expressing ALTO (or RFP) either mock-treated or stimulated with diABZI (Figs 4B and S12). Both mouse and rabbit Src antibodies successfully precipitated ALTO, but this occurred only in cells treated with diABZI (Figs 4B and S12), suggesting that specific stimuli, such as replicating MCPyV DNA, might be necessary to facilitate the interaction between Src and ALTO. Upon treatment with diABZI, Src antibodies were able to Co-IP TBK1, particularly its S172 phosphorylated form (Figs 4B and S12). We suspect that Src interacts with both ALTO and TBK1, even in the absence of diABZI treatment (S6 Fig). However, the introduction of upstream stimuli may bolster this interaction or stabilize a transient association that might not otherwise endure the stringent conditions of the co-IP process. More importantly, compared to the cells expressing RFP, the amount of TBK1 and phosphorylated TBK1^S172 complexed with Src was substantially increased in the cells expressing ALTO (Figs 4B and S12). Considering the observation that ALTO interacts with TBK1 (Figs 3 and S6), this result suggests that ALTO may utilize its multiple PXXP motifs to recruit Src into the TBK1 complex.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. ALTO recruits Src into complex with TBK1 to stimulate its autophosphorylation.

A. AlphaFold2 model of ALTO protein, with the Src-binding P-X-X-P motifs denoted in red. The table provides details for P-X-X-P motif sequences and positions. B. HDF-inALTO or -inRFP cells were mock-treated or induced with Dox for 10.5 hours, then stimulated with diABZI (or DMSO control) for 1.5 hours. Whole cell lysates were incubated with mouse anti-Src antibody and immunoprecipitated with Protein G agarose. Immunoprecipitants were resolved by SDS/PAGE and immunoblotted with the indicated antibodies.

https://doi.org/10.1371/journal.ppat.1012170.g004

Given that Src has been demonstrated to prime TBK1 for autophosphorylation and activation [43], the ability of ALTO to facilitate the formation of the Src-TBK1 complex may explain the effectiveness of ALTO in stimulating TBK1 autophosphorylation (Figs 2, S3 and S4). To investigate whether Src plays a key role in ALTO-induced TBK1^S172 autophosphorylation, we treated HDFs inducibly expressing ALTO with Src family kinase (SFK) inhibitors The Src/SFK inhibitors Dasatinib and Bosutinib significantly suppressed TBK1^S172 autophosphorylation. However, treatment with Dasatinib and Bosutinib also moderately reduced ALTO mRNA levels. Therefore, while our inhibitor data suggest that Src supports ALTO-induced TBK1^S172 autophosphorylation, additional approaches are needed to further confirm this finding. Nevertheless, our protein-protein interaction data support the model where ALTO facilitates the recruitment of Src into the complex with TBK1, thereby stimulating its autophosphorylation and activation. (Figs 3, 4, S6, and S12).

The induction of IFN signaling by MCPyV infection may be attributed to the stabilization of STING

We next started to examine the functional impact of ALTO-induced TBK1 autophosphorylation and STING stimulation during MCPyV infection. As established in our previous studies [32, 33], infecting HDFs with a moderate dose of MCPyV (10⁸ viral genome equivalents of MCPyV virions per 96-well) triggered a robust expression of IFNβ and key ISGs such as Viperin, OAS1, ISG54, and MX1 (Fig 5A). The IFN signaling induced by MCPyV is significantly more robust than the signaling triggered by diABZI (Fig 5A). Surprisingly, co-treating cells with MCPyV and diABZI did not enhance, but instead dramatically suppressed, the IFN signaling induced by MCPyV (Fig 5A). Given that IFN stimulation was significantly higher, yet STING mRNA level was lower in MCPyV-treated cells compared to those treated with diABZI (Fig 5A), we decided to examine STING and its downstream signaling molecules at the protein level. Both diABZI and MCPyV treatments stimulated phosphorylation of TBK1^S172 and IRF3^S386, with MCPyV treatment additionally inducing p65^S536 phosphorylation (Fig 5B). The STING protein level was reduced by diABZI, which is known to stimulate STING phosphorylation and subsequent degradation [45,46]. While MCPyV infection activates TBK1 stronger than diABZI treatment, it downregulates STING less (Fig 5B and 5C), consistent with our previous findings that ALTO stabilizes STING. This result suggests that the virus may stabilize the STING protein. Compared to cells treated with diABZI, those treated with MCPyV transcribed less STING mRNA but maintained a higher level of STING protein (Fig 5A, 5B and 5C), again supporting the virus’s ability to stabilize STING protein. In the cells treated with both MCPyV and diABZI, MCPyV infection was able to partially counteract the STING degradation induced by diABZI (Fig 5B and 5C). Collectively, these studies suggest that MCPyV infection both activates and stabilizes STING. In light of our previous experiments, these findings suggest that the effect of MCPyV might be attributed to the role of ALTO in stabilizing STING.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. MCPyV rescues STING levels to stimulate IFN signaling.

A. Normal HDFs mock-infected or infected with MCPyV were stimulated with diABZI (or DMSO control) at the time FBS was added (2 days post-infection). At 5 days post-infection, cells were harvested, and mRNA levels were quantified by RT-qPCR. The value for mock-infected and DMSO-treated HDFs was set as 1. Error bars indicate standard deviation from 3 replicates. B. Normal HDFs were mock-infected or infected with MCPyV, then stimulated with diABZI (or DMSO control) at the time FBS was added (2 days post-infection). Whole cell lysates were collected at 5 days post-infection, resolved by SDS/PAGE, and immunoblotted with the indicated antibodies. C. Normal HDFs were mock-infected or infected with MCPyV, then stimulated with diABZI (or DMSO control) at the time FBS was added (2 days post-infection). Cells were fixed at 5 days post-infection and immunostained for LT and STING, and counterstained with DAPI. Scale bar, 20 μm.

https://doi.org/10.1371/journal.ppat.1012170.g005

ALTO inhibits MCPyV infection activity

To best understand the potential magnitude and timing of ALTO’s effect during MCPyV infection, we analyzed changes in ALTO protein levels throughout the course of MCPyV infection in the HDF model. ALTO expression became detectable 4–6 days post-infection, peaking on day 5 (Fig 6A). Subsequently, we generated an ALTO null (ALTO_null) MCPyV variant to assess ALTO’s impact on MCPyV infectivity. In HDFs infected with varying titers of wild type (WT) or ALTO_null MCPyV, the ALTO_null virus consistently showed enhanced replication (30–60% better) compared to WT MCPyV (Fig 6B). Compared to HDFs infected with WT MCPyV, those infected with ALTO_null MCPyV showed a slight increase in IFNβ mRNA levels (S13A Fig), potentially stimulated by the increased MCPyV replication (Fig 6B). LT and VP1 were produced at similar levels by WT and ALTO_null MCPyV, indicating that the ALTOnull mutation does not affect the expression of LT and VP1 during infection (S14A Fig). These data suggest that the ALTO protein encoded by the WT virus might suppress viral replication activity. To further investigate this function of ALTO, we infected HDF-inALTO cells with ALTO_null MCPyV. We observed that inducing ALTO in these cells could suppress viral infection by 55–75% (Fig 6C). Both western blotting and IF analysis of HDF-inALTO cells revealed that while Dox treatment increased ALTO expression, the level of MCPyV LT significantly decreased in cells treated with Dox (Figs 6D and S14B). These findings confirm that exogenous ALTO, driven by an inducible promoter, can inhibit MCPyV replication activity.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. MCPyV ALTO suppresses viral infection activity.

A. Normal HDFs were mock-infected or infected with 6x10⁸ copies of MCPyV per 48-well. Whole cell lysates were collected at the indicated time points post-infection. Lysates were resolved by SDS/PAGE and immunoblotted with the indicated antibodies. B. Normal HDFs were infected with the indicated titers of WT or ALTO_null MCPyV per 96-well. Cells were harvested and DNA extracted at 5 days post-infection. Viral replication was quantified by qPCR using the NCRR-specific primers and normalized to genomic GAPDH levels. Bars indicate mean; error bars indicate standard deviation from 3 replicates. **p<0.01; *p<0.05. C. HDF-inALTO cells were infected with the indicated titers of ALTO_null MCPyV per 96-well. Cells were mock-treated or induced with Dox at 2- and 4-days post-infection. Cells were harvested and DNAs were extracted at 5 days post-infection. Viral replication was quantified by qPCR using the NCRR-specific primers and normalized to genomic GAPDH levels. Bars indicate the mean; error bars indicate the standard deviation from 3 replicates. ***p<0.001. D. HDF-inALTO cells were infected with ALTO_null MCPyV. Cells were mock-treated or induced with Dox at 2- and 4-days post-infection. Cells were fixed on day 5 post-infection and stained for LT and either ALTO or MX1, and counterstained with DAPI. Scale bar, 50 μm.

https://doi.org/10.1371/journal.ppat.1012170.g006

As indicated by MX1 IF staining and RT-PCR analysis of IFNβ and ISGs in HDF-inALTO cells treated with ALTO_null MCPyV, the reduced MCPyV infection in ALTO-expressing cells subsequently impaired the expression of IFNβ and downstream ISGs (Figs 6D and S13B). However, no significant differences were detected in virus-induced phosphorylation of STING^S366, TBK1^S172, IRF3^S386, or p65^S536 in HDFs infected with either WT or ALTO_null MCPyV (S14 Fig). Although one might expect knocking out ALTO in the ALTO_null virus to reduce the induction of the STING/TBK/IRF3/p65 signaling pathway, the increased replication in the ALTO_null virus could stimulate this pathway. These combined effects may contribute to the observed lack of significant changes in STING, TBK1, IRF3, and p65 phosphorylation by ALTO_null MCPyV as shown in S14 Fig. Given that the Western blotting analysis was conducted on a mixed population of both infected and uninfected cells, we also reasoned that examining individual cells could allow for a more accurate profiling of ALTO’s impact on the STING-TBK1 pathway during viral infection. IF staining of WT MCPyV-infected HDFs revealed that cells exhibiting a positive ALTO signal often displayed weaker LT signal (Fig 7A). Additionally, we employed ImmunoFISH to double-stain WT MCPyV-infected HDFs with an ALTO antibody and MCPyV DNA probes [28] (Fig 7B). Compared to ALTO-negative cells, many ALTO-positive cells exhibited a reduced FISH signal for the replicated MCPyV genome (Figs 7B and S15). In contrast, ALTO_null MCPyV-infected cells displayed a stronger MCPyV FISH signal than those infected with WT MCPyV (Fig 7B). Quantification of the ImmunoFISH signal in individual cells further confirmed that ALTO-positive HDFs infected with WT MCPyV exhibited a significantly lower viral replication FISH signal–by ~50%–compared to ALTO-negative HDFs, regardless of infection with WT or ALTO_null MCPyV (Fig 7C). This result underscores ALTO’s role in repressing MCPyV infection at the single cell level (see discussion).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. ALTO suppresses MCPyV infection at the single-cell level.

A. Normal HDFs were either mock-infected or infected with 6x10⁸ viral genome equivalents of MCPyV virions per well in a 48-well plate. Cells were fixed on day 4 post-infection, immunostained for ALTO and LT, and counterstained with DAPI. Scale bar: 20 μm. B. Normal HDFs were either mock-infected or infected with 6x10⁸ viral genome equivalents of WT or ALTOnull MCPyV virions per well in a 48-well plate. Cells were fixed on day 5 post-infection, immunostained for ALTO, subjected to FISH using an MCPyV probe, and counterstained with DAPI. Scale bar: 20 μm. C. The FISH signals within individual cells treated as in B were quantified using ImageJ. Shown are the quantification results for the FISH signal in individual ALTO-positive and ALTO-negative HDFs infected with WT MCPyV, as well as HDFs infected with ALTOnull MCPyV. Each type of cell is indicated with color-coded arrows. More than 40 cells were analyzed for each sample group to achieve a statistically meaningful sample size. Central diamonds represent the median values of the data groups. ***p<0.001; **p<0.01; n.s. = not significant.

https://doi.org/10.1371/journal.ppat.1012170.g007

TBK1 inhibition stimulates MCPyV replication

Given the potent ability of ALTO to stimulate TBK1^S172 autophosphorylation (Figs 2, S3 and S4), we investigated the effects of blocking TBK1 autophosphorylation on MCPyV infection and the role of ALTO in this process. Treating HDF stable cells that can inducibly express ALTO with the TBK1 inhibitor GSK8612 abrogated ALTO-induced TBK1^S172 autophosphorylation in a dose-dependent manner (Fig 8A, compare lanes 3–6 to lane 2). Compared to the mock-treated control, Dox induction of ALTO in HDF-inALTO cells significantly repressed ALTO_null MCPyV infective activity (Fig 8B). Treatment with GSK8612 alone markedly enhanced viral infectivity (by up to 51%), demonstrating that inhibiting the TBK1 downstream innate sensing pathway can enhance MCPyV infection and spread (Fig 8B). In cells with Dox-induced ALTO expression, treatment with GSK8612 further rescued an additional 23% of the viral replication compared to cells without ALTO induction (Fig 8B). These results highlight TBK1’s critical role in mediating ALTO’s function in controlling MCPyV infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. TBK1 inhibition reactivates MCPyV infection activity.

A. HDF-inALTO cells were treated with the indicated TBK1 inhibitor (or DMSO control) for one day, then treated with 0.05 μg/mL Dox for two additional days. Whole cell lysates were resolved by SDS/PAGE and immunoblotted with the indicated antibodies. B. HDF-inALTO cells were infected with ALTO_null MCPyV. Beginning 2 days post-infection, the cells were treated with the TBK1 inhibitor GSK8612 (or DMSO control) at the indicated doses, then treated with 0.05 μg/mL Dox at 3 days post-infection. At 5 days post-infection, cells were harvested, and DNA extracted. Viral replication was quantified by qPCR using the NCRR-specific primers and normalized to genomic GAPDH. The value from one group of uninduced cells treated with DMSO was set as 1. Bars indicate the mean from three replicates; error bars indicate the standard deviation. Percentages reflect relative replication in Dox-induced cells compared to their respective uninduced controls. ***p<0.001; **p<0.01; *p<0.05. C. ALTO stimulates STING-TBK1 signaling to negatively regulate MCPyV infection. The MCPyV protein ALTO recruits Src via its PXXP motifs (red) to phosphorylate TBK1, leading to STING stabilization and activation. MCPyV replication also stimulates the activation of the STING downstream signaling pathway. Activated STING induces an enhanced downstream ISG response via IRF3 and NF-κB. This antiviral response provides a negative feedback mechanism to control MCPyV propagation and persistence.

https://doi.org/10.1371/journal.ppat.1012170.g008

Cell culture

HDFs were isolated from neonatal foreskins as previously described [28,30] and were maintained in DMEM (Invitrogen) supplemented with 10% FBS (Cytiva), 1x nonessential amino acids (Gibco), and 1x GlutaMAX (Gibco). HEK293 cells were maintained in DMEM (Invitrogen) supplemented with 10% FBS (Cytiva). U2OS cells were maintained in McCoy 5A (Invitrogen) supplemented with 10% FBS (Cytiva). All cells were maintained and incubated at 37°C in 5% CO_2.

Chemicals and reagents

Saracatanib, Dasatanib, Bosutinib, BX-795, and GSK8612 (SelleckChem) were dissolved in DMSO to a stock concentration of 10mM. diABZI (SelleckChem) was dissolved in DMSO to a stock concentration of 5mM. diABZI was used at a final concentration of 1 μM in all experiments. Doxycycline (Clontech) was dissolved in water to a stock concentration of 0.5mg/mL. Dox was used at a final concentration of 0.5 μg/ml in all experiments unless otherwise indicated in the figure legend. Stocks of diABZI and Doxycycline were stored as small aliquots at -20°C. All other stock solutions of chemicals were stored in small aliquots at −80°C.

Recombinant plasmid construction

The pTRIPZ-STING^R284S construct was generated in our previous studies [36]. The viral gene constructs used in this study were sourced as follows: pGwf and phGf were from Addgene; pALTOw, pMtw (ST), pWm (VP1), pMono-Zeo, pMono-ADL (LT) and ph2m (VP2) were from the Christopher Buck Laboratory; and the pcDNA4c-57kT construct was generated by first fusing the full 57kT ORF using two-step PCR from pcDNA4c-LT1-817 [53] and then cloning it into the pcDNA4c backbone using the KpnI and XhoI restriction sites.

For making HDF-inALTO stable cells, the ALTO coding sequence was PCR-amplified from pALTOw (Christopher Buck Laboratory) and then cloned using AgeI and MluI sites to replace the RFP coding sequence in the pTRIPZ vector (Open Biosystems). The resulting construct is named pTRIPZ-ALTO.

To generate the ALTO_null MCPyV genome, the region surrounding the ALTO coding sequence in pR17b (WT MCPyV, Christopher Buck Laboratory), between the restriction sites SacI and HpaI, was amplified by two-step PCR to mutate the ALTO ATG start codon to ACG, without altering the Large T sequence or reading frame. This altered sequence was then subcloned back into the viral plasmid using those same restriction sites.

NanoBiT plasmids pBiT1.1-N[TK/LgBiT], pBiT1.1-C[TK/LgBiT], pBiT2.1-N[TK/SmBiT], and pBiT2.1-C[TK/SmBiT] were obtained from Promega. ALTO, STING, and TBK1 sequences were PCR amplified from the pALTOw plasmid (Christopher Buck Laboratory) or normal HDF cDNA samples and inserted into each NanoBiT vector using the XhoI and NheI (for ALTO and STING) or XhoI and BglII sites (for TBK1).

BiFC constructs were cloned using the same strategy as described in our previous study [54,55]. Coding sequences for ALTO (PCR-amplified from pALTOw), TBK1 and STING (both PCR-amplified from the NanoBiT expression plasmids), were cloned using XhoI and NotI sites in the BiFC vectors to generate the in-frame fusion of the ALTO/TBK1/STING molecules with N- or C-terminal Venus N or Venus C.

Preparation of HDF inducible stable cells

Inducible RFP or ALTO stable cells were generated from early passage primary HDFs (≤ 2 passages from tissue) as previously described [33,34,56]. Tetracycline-screened FBS (Cytiva) was used to supplement the cell culture media as described above.

MCPyV infection

ALTO_null MCPyV was prepared using the WT MCPyV method described previously [28,30]. Normal HDFs or HDF-inALTO stable cells were infected with WT or ALTO_null MCPyV using the method described previously [28,30,56]. For virus infection, unless specifically noted in the figure legends, 10⁸ viral genome equivalents of MCPyV virions were used in each 96-well.

To analyze MCPyV replication, harvested cell pellets were lysed in QuickExtract DNA Extraction Solution (Biosearch Technologies) according to the manufacturer’s instructions. The extracted DNA samples were then subjected to real-time qPCR using a QuantStudio 3 Real-Time PCR System (Applied Biosystems) with SYBR Green Master Mix (Applied Biosystems). The MCPyV genomic DNA levels were measured using NCRR-targeted primers and normalized to the level of genomic GAPDH. Both NCRR and genomic GAPDH primer sequences can be found in [33].

Cell viability assay

HDF-inALTO cells were maintained as described above in Tet-free media. A total of 2x10⁴ cells were seeded in wells of a 96-well dish and allowed to adhere for 8 hours at 37°C in 5% CO₂. Standard media was changed to media containing 0.5 μg/ml Dox and the tested concentration of each drug. After incubating at 37°C in 5% CO₂ for 16 hours, cell viability assays were performed using the Cell Titer Glo 3D kit (Promega) according to manufacturer instructions. Luminescence was measured using a Luminoskan Ascent microplate reader (Thermo). Relative viability was calculated by comparing average luminescence readings between drug-treated and DMSO vehicle control-treated cells.

NanoBiT protein-protein interaction assay

HEK293 cells were seeded at 3x10⁴ cells in 100μL per 96-well approximately 20 hours prior to transfection. 100ng total DNA of the examined NanoBiT fusion protein plasmids (50ng LgBiT + 50ng SmBiT) were diluted in Opti-MEM I (Gibco) and combined with diluted Lipofectamine 2000 (Invitrogen) at a 2.2:1 ratio of Lipofectamine to DNA before being added to cells. Transfected cells were incubated at 37°C in 5% CO₂ for 20–24 hours. Approximately 30 minutes prior to desired reading time, cell media was aspirated and replaced with 100μL Opti-MEM I (Gibco). NanoGlo reagent was freshly diluted to a 5X stock in Live Cell Substrate according to manufacturer direction, then added to all wells. Plates were returned to the incubator for 15 minutes, and luminescence was measured using a Luminoskan Ascent microplate reader (Thermo). Per manufacturer direction, a specific protein-protein interaction was assessed to be present when the average luminescence reading of the interacting pair was tenfold higher than the average from the LgBiT-tagged construct co-transfected with the SmBiT negative control construct.

BiFC assay

1.5x10⁵ U2OS cells were seeded in a 6-well plate atop 1–2 coverslips and allowed to adhere overnight at 37°C in 5% CO₂. Subsequently, 1 μg of each construct of the BiFC pair for each well was diluted in Opti-MEM I (Gibco). 6.5 μL FuGENE HD (Promega) was added to this mixture and allowed to incubate for 15 minutes at RT prior to addition to cells. BiFC constructs were incubated at 37°C in 5% CO₂ for 24–48 hours prior to cellular fixation and IF staining.

Immunofluorescent staining

Cells were cultured on coverslips in standard conditions and fixed with 3% paraformaldehyde in PBS for 10–20 minutes. Staining was performed as previously described [33]. Primary antibodies used in this study are as follows: anti-ALTO (1:10,000, Denise Galloway Lab), anti-MCPyV LT (1:250, sc-136172, Santa Cruz Biotechnology), anti-STING (1:200, 19851-1-AP, Proteintech), anti-MX1 (1:1,000, 37849S, Cell Signaling Technology), and anti-FLAG (1:40,000, F3165, Sigma Aldrich). Secondary antibodies used in this study are as follows: Alexa Fluor 594 goat anti-mouse IgG (1:500, A11032, Invitrogen), Alexa Fluor 488 goat anti-rabbit IgG (1:500, A11034, Invitrogen), Alexa Fluor 594 goat anti-rabbit IgG (1:500, A11012, Invitrogen), and Alexa Fluor 488 goat anti-mouse IgG (1:500, A11029, Invitrogen).

All IF images were captured using an inverted fluorescence microscope (IX81; Olympus) as described previously [33]. The scale bars were added using ImageJ software.

Immunofluorescent-fluorescence in situ hybridization (ImmunoFISH)

HDFs were fixed with 3.7% formaldehyde in PBS for 10 minutes. IF staining was performed using ALTO antibody with 1:3,000 dilution as previously described [33]. After IF staining, the cells were fixed with 3.7% formaldehyde in PBS for 10 minutes again and subjected to FISH staining using buffers from Biosearch Technologies as previously described [33]. After the final wash step, cells were mounted on coverslips and observed using an inverted fluorescence microscope (IX81; Olympus).

The FISH signals were quantified at the single cell level using ImageJ. Photos from each of the channels of interest were stacked. Using the DAPI channel, the nuclear area of a single ALTO- or FISH-positive cell was selected. Both ALTO- and FISH-positive cells were considered infection-positive. The area selection was then applied to the FISH channel and the integrated density (the sum of the signal intensity in the selected area) was measured for that cell. More than 40 cells were analyzed for each group of samples to achieve a statistically meaningful sample size.

Structural prediction

The canonical ALTO sequence, with a length of 250 amino acids, was utilized to generate the predicted structure. The tool AlphaFold 2 processed this sequence on the public Google Colab (ColabFold v1.5.5: AlphaFold2 using MMseqs2) using default settings, without any constraints, templates, or multimerization options.

Western blotting

Whole cell lysate preparation and Western blot analyses were performed as previously described [33]. Primary antibodies utilized in this study are as follows: anti-GAPDH (1:5,000, 2118S, Cell Signaling Technology), anti-ALTO (1:10,000, Denise Galloway Laboratory), anti-STING (1:1000, 13647S, Cell Signaling Technology), anti-STING p-S366 (1:1000, 19781S, Cell Signaling Technology), anti-TBK1 (1:1000, 3504S, Cell Signaling Technology), anti-TBK1 p-S172 (1:1000, 5483S, Cell Signaling Technology), anti-p65 (1:1000, 8242S, Cell Signaling Technology), anti-p65 p-S536 (1:1000, 3033S, Cell Signaling Technology), anti-IRF3 (1:500, SC-33641, Santa Cruz Biotechnology), anti-IRF3 p-S386 (1:1000, 37829S, Cell Signaling Technology), anti-RFP (1:2,000, AB233, Evrogen), rabbit anti-Src antibody (1:1000, 2109S, Cell Signaling Technology), mouse anti-Src antibody (1:1000, 2110S, Cell Signaling Technology), anti-LgBiT (1:500, N7100, Promega), anti-MCPyV LT (1:500, sc-136172, Santa Cruz Biotechnology), anti-MCPyV VP1 (1:2,500, Christopher Buck Laboratory) and rabbit anti-MCPyV VP2 (1:2,500, Christopher Buck Laboratory). The secondary antibodies utilized in this study are HRP-linked anti-rabbit IgG (1:2000, 7074S, Cell Signaling Technology), and HRP-linked anti-mouse IgG (1:2000, 7076S, Cell Signaling Technology). Western blots were developed with the SuperSignal West Pico PLUS Chemiluminescent Substrate Kit (Thermo Fisher Scientific) and imaged using an Amersham Imager 600 or 680 (GE Healthcare).

Quantification of signals in Western blots was performed using ImageJ. Briefly, each lane was defined using a tall, narrow rectangle. A profile plot for each band within the lane was generated using the Plot Lanes function. The peak area for each lane was closed with a straight line and measured using the wand tool. Relative abundance was calculated by setting the value of a control band as 1.

RT-qPCR

Total RNA was isolated from cell samples using the NucleoSpin RNA XS Kit (Macherey-Nagel) following the manufacturer’s instructions. Reverse transcription and Realtime qPCR were performed as previously described [56].

Statistical analyses

Unpaired, two-tailed t-tests were performed using the t.test function in Microsoft Excel. The significance thresholds were set at *P value < 0.05, **P < 0.01, ***P < 0.001.