We started this work showing that TioV infection activates IFN signaling in human HEK-293 cells. This suggested that TioV is incapable of blocking IFN signaling, but also fails to control IFN production. This appears surprising since TioV-V protein expression was previously shown to control IFN induction by MDA5 and RIG-I through interactions with MDA5 and LGP2, respectively 
. Although this will need to be confirmed in TioV-infected cells, our data also validated TioV-V binding to MDA5 and LGP2, and its capacity to block IFN-β promoter induction upon MDA5 overexpression. Nevertheless, IFN secretion by TioV-infected cells could be explained by several non-exclusive mechanisms. First, inhibition of RIG-I signaling by V proteins form paramyxoviruses is dependent on LGP2 recruitment, and since levels of this co-factor are rate limiting in HEK-293 cells, TioV-infection could induce IFN expression because LGP2 is not sufficiently expressed in these cells 
. It is also possible that TioV replication produces such high amounts of a specific viral PAMP, like defective-interfering RNA genomes, that inhibition by TioV-V is ineffective. In addition, TioV could trigger IFN expression through viral sensors that are distinct from MDA5 and RIG-I. Finally, and since many cellular components involved in viral sensing are IFN-inducible, it is possible that small amounts of IFN produced at initial phases of the infection can translate into a robust expression of IFN-inducible genes at later time-points if IFN signaling is not blocked. This is supported by observations performed on PIV5 mutants (CPI- or rSV5-P/V-CPI-) expressing a V protein defective for STAT1 degradation and inhibition of IFN-α/β signaling, but competent for MDA5/LGP2 binding 
. These mutants were found to induce some more (CPI-) or much more (rSV5-P/V-CPI-) IFN-α/β than the wild-type virus 
. Since we now established that TioV-V is unable to block IFN signaling as well, this clearly parallels our observations on TioV.
Here, we demonstrate that in contrast to the V protein of MuV, TioV-V hardly interacts with human STAT2, does not induce STAT1 degradation and is unable to block signal transduction downstream of IFN-α/β and IFN-λ receptors. However, TioV-V remains functional regarding other known activities of rubulavirus V proteins, and besides interactions with MDA5 and LGP2, it was found to bind DDB1 and STAT3. This later interaction probably accounts for TioV-V capacity to block IL-6 signaling, although it was not as efficient as MuV-V to block this signaling pathway. Conformational constraints and/or a lack of adaptation to human STAT3 could explain the lower capacity of TioV-V to induce a functional complex targeting STAT3. In future, further investigations should establish that IL-6 signaling is actually impaired in TioV-infected cells.
The observation that TioV-V is unable to block IFN signaling in human cells is of interest to address ecological and epidemiological questions. Indeed, several viruses from Paramyxoviridae
family originating from giant fruit bats have recently emerged in human populations in Southeast Asia and Australia, which include two members of Henipavirus
genus, Hendra and Nipah viruses, and one rubulavirus closely related to Tioman virus called Menangle virus. The seroprevalence of henipaviruses (Nipah or Hendra virus) and rubulaviruses (Tioman or Menangle) in giant fruit bats can be very high (>50%) as assessed by a recent survey in Papua New Guinea 
. Although Hendra virus has been responsible for limited outbreaks in Australian horse farms and few fatal human cases 
, Nipah virus has killed hundreds of people since 1999, while spreading from Malaysia to Singapore and Bengladesh 
. Menangle virus also recently emerged from bats, causing disease outbreaks in Australian piggeries in 1997 with epidemiological evidence suggesting that it was responsible for severe flu-like syndromes in two piggery workers 
. This illustrates the threat that bat Paramyxoviridae
represent for human populations.
The V proteins of Nipah and Hendra viruses have been shown to block IFN-α/β signaling in several species including human 
. Because the V protein is essential for rubulaviruses to inhibit IFN-α/β signaling 
, it is surprising that TioV-V is unable to block this pathway in human cells. Interestingly, the interaction with DDB1 suggests that at some point TioV-V was capable of targeting STAT1 proteins for proteasomal degradation but has lost this capacity during evolution. Alternatively, this interaction is maintained because TioV-V targets other cellular proteins for degradation. Nevertheless, the former hypothesis is further supported by the trace interaction detected between TioV-V and STAT2. In the future, the use of point mutants as well as MuV/TioV V chimeric proteins would be suitable to define amino acid residues in TioV-V and MuV-V that determine STAT2 binding and inhibition of IFN-α/β signaling cells. This would greatly help to establish that Tioman virus has indeed lost its capacity to efficiently bind STAT2 during evolution because of only few mutations. We also tested P and W proteins of TioV for the inhibition of IFN-α/β signaling, but none of these proteins exhibited such an activity when expressed in human cells. Although TioV could express yet unidentified viral factors to interfere with IFN-α/β signaling, in vitro
infection experiments described in this report rather suggest some constitutive defect in the capacity of this virus to block IFN-α/β signaling in human cells.
Whether TioV is able to block IFN-α/β signaling in bats also remains a pending question. In this report, we established the sequence of STAT2 from Pteropus rodricensis
, and then demonstrated that TioV-V is a poor binder of prSTAT2. Since this experiment was performed in human cells, it is possible that TioV-V and prSTAT2 failed to interact because bat-specific factors were missing in this microenvironment. It is also possible that TioV-V directly interacts with STAT1 from bats alike the V protein of hPIV2 
. In order to determine if TioV-V can block IFN-α/β signaling in bats, experiments must be performed in cells isolated from giant fruit bats of Pteropus
genus. Unfortunately, no commercial cell line is currently available and only one lab in Australia has developed this kind of tool 
. Using this system, Virtue and collaborators have shown that interferon production and signaling pathways are antagonized during henipavirus infection of fruit bat cell lines 
Altogether, our observations question the capacity of TioV to appropriately control IFN-α/β signaling in human and bat cells. Inhibition of IFN-α/β signaling may not be mandatory for TioV to be maintained in its natural host population and to infect human, regardless the induction of a specific pathology. Indeed, it has been reported that the V protein of human parainfluenza virus type 4 (hPIV4), a member of Rubulavirus
genus infecting human, is unable to inhibit IFN-α/β signaling in host cells 
. This natural defect could account for the fact that hPIV4 infection is less frequent and pathogenesis is less severe compared to other human paramyxoviruses 
. Similarly, a human parainfluenza virus type 2 (hPIV2) with V protein mutations that prevented the virus from inhibiting IFN-mediated signaling maintained its capacity to replicate in the respiratory tract of non-human primates 
. In both cases, the capacity to block IFN-α/β induction through interactions with MDA5 and LGP2 probably compensates to some point for the lack of inhibition downstream of IFN-α/β receptor and the same could be true for TioV in vivo
. However, the capacity of Paramyxoviridae
to target STAT proteins has also been associated to their replication level in a specific host as aforementioned for PIV5 infection in mice 
, but also to the development of pathology. For example, a recombinant measles virus that is unable to fully antagonize IFN signaling cannot control inflammation and is attenuated in rhesus monkeys 
. These data demonstrate that in some cases, the ability of Paramyxoviridae
to block this antiviral pathway directly influences their infectivity, virulence and associated pathogenesis. Since our findings suggest that TioV is defective for the inhibition of IFN-α/β signaling at least in human cells, future studies should determine consequences in terms of infectivity, virulence and pathogenesis.