Phylogenetic analysis and the worldwide distribution of FVs based on Tas revealed co-evolution of foamy viruses and their hosts
Searching for the UniProt databases, we found the complete protein sequences of 12 FVs from 9 primate and 3 non-primate hosts. At the same time, the literatures were consulted to trace the isolation history of different foamy virus from multiple hosts (Table 1). In order to shed light on the evolutionary relationship among these viruses, the phylogenetic analyses were performed based on the different Tas sequences and the cladogram of relative hosts was derived from Timetree databases. According to the comparation of molecular evolution tree (Fig. 1), the 12 sequences formed 2 different monophyletic clusters: all the primate isolates clustered together, while the three non-primate isolates formed the second cluster. Furthermore, as Simian foamy viruses (SFVs) are prevalent in all species of non-human primates, we focused on the co-evolution of FVs and their natural non-human primate hosts. The results indicated that SFVagm and SFVmac gathered into a cluster paralleled with Old World monkeys including African green monkey (Cercopithecus aethiops), Macaque (Macaca mulatta). The lineage of SFVora was paralleled with Orangutan (Pongo pygmaeus), another member of Old World monkeys. And Orangutan was placed as a bridge to the apes, which include Chimpanzee (Pan troglodytes verus), Gorilla (Gorilla gorilla sp.) and were in accord with SFVcpz and SFVgor. SFVspm, SFVsqu and SFVmar formed another cluster absolutely consisted with New World monkeys including Spider monkey (Ateles sp.), Squirrel monkey (Saimiri sciureus), Marmoset (Callithrix jacchus). The matching of foamy viruses and their hosts revealed a long-lasting co-speciation history between FVs and ancient primate animals clearly. However, the antinomy appeared that the phylogeny of EFV, FFV and BFV did not mirror the relationship of their relative hosts equine (E), feline (F), and bovine (B), and placed the FFV lineage basal to the EFV and BFV cluster. This marked that the relationships among these hosts are not always consistent among different evolutionary level.
Retroviruses are actively crossing into humans exposed to non-human primates from chimpanzees as well as other non-human primates which led to potential risk of disease, such as simian immunodeficiency virus and simian T lymphotropic virus. However, the origin of retroviruses was still not well documented. To take this line of inquiry a step further on the prevalence of FVs around worldwide, we investigated the distribution of the relative hosts which offered the isolation for different viruses in a variety of contexts, in several countries and from multiple non-human primates (Fig. 2). As the map showed that the worldwide distribution of SFVagm, SFVmac and SFVora were always in keeping with the habitat range of Old World monkeys.
SFVspm, SFVsqu and SFVmar were also consisted with New World monkeys geographically which included tropical regions of Central and South America. Apes whose geographical distribution is mainly concentrated in equatorial tropical forests of Africa took the major responsibility for cross-species transmission from Chimpanzees to humans. But for BFV, EFV and FFV which isolated from equine (E), feline (F), and bovine (B), there is no explicit correlations at the distribution among FVs and their non-primate hosts.
Comparison of physicochemical properties of the three different host-derived foamy virus Tas
The results of multiple sequence alignment performed by Clustal X displayed that Tas of foam viruses isolated from different hosts varied with each other dramatically (Fig. 3). Several conserved sites were only found either on the middle or the C-terminal of the Tas sequences. And these results displayed that functional fragments of different Tas were still conserved but the tremendous diversities on the other areas could be attributed to that foamy viruses isolated from different hosts suffered disparate challenges for virus survival strategy.
To further compare the functions of different Tas, physicochemical properties were analyzed by many online prediction websites. All of the 12 Tas were detected to have several prominent hydrophobic regions according to the hydrophobicity analysis (supplementary materials 1). Except for the foamy viruses isolated from non-primate hosts which were predicted to possess one dominating hydrophilic region and 2 ~ 4 misty hydrophobic regions, the rest Tas of FVs isolated from primate hosts all possessed three prominent hydrophobic regions and two hydrophilic regions. And the average hydrophobicity results indicated that the Tas protein was hydrophilic.
The physicochemical properties such as theoretical pI, estimated half-life, instability index, aliphatic index and grand average of hydropathicity were also predicted by several online website (Table 2). The theoretical PI values of Tas proteins of these different host-derived foamy viruses ranged from 5.02 to 6.09. Stability coefficient analysis shows that Tas proteins of other host-derived foamy viruses tend to be unstable except for BTas (Instability Index < 40). What’s more, there was no significant difference in the prediction of half-life, fat index and other characteristics. The prediction and analysis of Tas secondary structure showed thatα-helix and random coil are the main structural components, and β-corner appears in local regions (supplementary materials 2). The proportion of random coils in the secondary structure of Tas protein of foamy virus isolated from different host sources was 60.06%-71.67%, the proportion of α-helix was 8.13%-32.19%, and the proportion of lamella was 4.32%-25.84% (Table 3). All these above showed that although there were much more divergences of amino acid primary structures of different Tas, they still were inclined to keep physicochemical properties of great similarities which lay the foundation for the conservation of motifs initiating the replication of FVs.
Furthermore, the analysis of conserved motifs of foamy viruses, which were secondary units always gathered together to form relatively stable motifs and exert special functions, implied the functional similarity of Tas. And Tas were analyzed using MEME v4.9 online analysis software, and a total of 10 predicted motifs were obtained (Fig. 4). As the diagram displayed that the numbers and types of motifs contained in Tas were shared high similarity. It is concluded that the diversity of motif was related to the various pathways in virus-host immune response, which implied that although faced with different host selection pressure, functional domains are still conserved. What’s more, the motifs of functional classification also consist with the results from phylogenetic analyses which were divided into branches of simian foamy virus from the primate hosts. All these suggested the exist of co-species except for the cross-species transmission.
Interactor enrichment analysis of the three foamy virus Tas isolated from different host origins
Although we had already found that various Tas of foamy virus isolated from different host was absolutely diversity, the function of motifs was verified conserved. Therefore, proteomics were performed in the stable cell lines with Tas of PFV、SFVagm and SFVora overexpression which are different but much closer to that of PFV in motifs analysis to detect the differences of specific protein interactions. Finally, we detected 65, 67 and 121 proteins in the overexpressed PFV-tasM, SFVagm-tasM and SFVora-tasM separately (Supplementary Material 3). Through GO analysis, the enriched proteins related to “signal-organism process”、 “biological regulation”、 “cell progress”、 “metabolic progress”、 “cell part” and “binding” evidenced that all the three Tas could widely affected Biological Process, Cell Component and Molecular Function (Fig. 5a). Furthermore, it was verified that the three kind of Tas regulated different cell signaling pathways by KEGG enrichment analysis (Fig. 5b). For Tas of PFV, the most significant enrichment pathway was “Basal transcription factors” while “Central carbon metabolism in cancer” for SFVagm Tas and “DNA replicating” for SFVora Tas. And the protein-protein interactions (PPI) network for PFV and SFVagm Tas were concentrated to histone acetylation, but for SFVora Tas, proteins functioned on “positive regulation of DNA metabolic process”、 “telomere maintenance”、“Cell Cycle” were concentrated(Fig. 5c). All these implied that the three kind of Tas proteins regulated different cellular effects by inducing different signaling pathways in the virus-host cell immune response.
Interactions with USP7 illustrated the ubiquitin-dependent proteasome degradation of the Tas
By further comparing the types of cytokines interacted with Tas proteins of foamy viruses derived from different host origin, we had found that all the three kinds of Tas proteins could bind ubiquitin-proteasome degradation pathway related proteins PSMD11(components of the 26S proteasome) and TRIM21(E3 ubiquitin protein ligase) (Fig. 5d). Therefore, we had conjectured that all the three kinds of Tas proteins were encountered the antagonism of host degradation mediated by ubiquitination which might play an important role in the invasion removement of the foamy viruses. Linear motifs are short components which provide low-affinity interaction interfaces and play central roles in mediating the regulatory functionality of the cell. They are particularly controlling protein turnover during the infection of viruses and always lead to the activation of host cells signaling. Here we employed the Eukaryotic Linear Motif (ELM) resource to discover putative linear motifs in Tas of SFVagm, SFVora and PFV. As the predictions showed that there were many putative binding sites which could interact with multiple cytokines such as PCSK、MAPK、 PP2A、 PP2B and USP7 in PFV, PCSK、MAPK、IRF3 and USP7 in SFVora and PCSK、MAPK、PP2A、PP2B、APCC and USP7 in SFVagm (Fig. 6a). At the same time, as the stable expression of reverse transcription regulator Tas was the precondition of FVs replication and important manifestation of virus-host immune reaction, we focused on USP7 for its unique deubiquitylation-functions in the ubiquitin-proteasome degradation pathway. We performed immunoprecipitation test to detected the accuracy of prediction results and verified that all the three kinds of Tas could interact with USP7 as predicted in vivo (Fig. 6b). To investigate whether in the absence of P5091, a specific inhibitor of the deubiquitinase enzymes USP7, the degradation of Tas proteins were occurring through the proteasomal pathway, pcDNA™4/myc-His A-SFVagm-Tas, pcDNA™4/myc-His A-SFVora-Tas and pcDNA™4/myc-His A-PFV-Tas transfected Hela cells were treated with P0591. The results showed that all Tas underwent the fate of degradation invariably (Fig. 6c, upper). Furthermore, as the increasing of the concentration of P0591, much more Tas proteins were urged into degradation which indicated a dose-dependent decrease of Tas proteins. What’s more, when MG132 was treated with the Tas proteins overexpression showed a slight increase in Tas level suggested that all the three kind of Tas proteins suffered of proteasome degradations (Fig. 6c, bottom). All these suggested that the three kinds of Tas proteins could all hijack USP7 to stabilize themselves and regulate foamy viruses replication.