Purification And Identification Of DENV-2 UTR RNA Binding Proteins In Host Cells

Dengue virus(DENV) belongs to the family of Flaviviridae and is transmitted by the bite of mosquitoes. The infection of the four distinct serotypes of DENV could cause disease symptoms such as dengue fever and dengue hemorrhagic fever. Dengue infection afflicts more than 50 million people annually, in tropical and subtropical regions and becomes a severe risk to the public health. Successful DENV replication relies on the viral proteins as well as host proteins. The DENV positive strand RNA genome contains 5’ and 3’ untranslated regions(UTRs) that have been shown to be required for virus replication and interaction with host cell proteins. Currently, limited RNA binding proteins are identified to interact with DENV 5’ and 3’ UTRs. Further identification of host factors interacting with viral RNA UTRs will greatly improve our understanding of viral biology and pathology.By using two independent methodologies, we successfully purified host DENV-2 RBPs that assembled in vitro and in living cells, respectively. The role of several novel RBPs in DENV replication were further studied. The developed one-step affinity purification strategy based on an RNA affinity tag allows large-scale preparation of native viral RBPs. Efficient isolation of endogenously assembled viral RNA-protein complexes would gain our understanding towards virus replication mechanisms.Methods and Results1. Isolation and identification of host proteins that bound to DENV-2 RNA 5’-3’UTRThe DENV-2 5’UTR and 3’UTR were cloned into the p CI-neo vector for in vitro transcript of 5’-3’UTR RNA. The 5’-3’UTR RNA was then labeled at the 3’ terminal with biotinylated cytidine to generate 5’-3’UTR-bio RNA. After the 5’-3’UTR-bio RNA incubated with the cell extract, the streptavidin magnetic beads were applied to isolate the RNA-protein complexes. The eluted sample was applied on SDS-PAGE gel, and 9 specific bands were recovered when compared to the control group. There are totally 78 proteins were identified in the 9 bands by LC-MS/MS analysis. The P-body component LSm1 and DEx D/H family protein DDX21 were selected as important candidate RBPs to further study.To investigate the potential role of LSm1 protein in DENV replication, firstly the LSm1 m RNA levels in DENV-2 infected cells of different time point post infection were assayed by RT-q PCR. The results showed that the m RNA level of LSm1 was increased significantly in response to DENV infection. The si RNAs target LSm1 gene were then transfected into cells followed by DENV infection. The viral RNA level in LSm1 silenced cells was assayed by RT-q PCR. The results showed that the viral RNA level was decreased in LSm1 silenced cells. Meanwhile, the infectious particles in cellular supernatant of the LSm1 silenced cells were counted by flow cytometric analysis. The results showed that infectious DENV-2 productions from LSm1 silenced cells were reduced significantly. The interaction of LSm1 and DENV RNA were studied by RIP(RNA immunoprecipitation) assay. The RIP eluted samples were conducted to RT-PCR(reverse transcript PCR) using primers of the DENV 5′UTR and 3’UTR., and the result showed that DENV RNA 3’UTR bound specifically to LSm1, as normal rabbit Ig G precipitation controls run in parallel were negative. The subcellular localization of LSm1 during DENV-2 infection was then examined using confocal laser scanning microscopy. The DENV infected cells were stained with anti-ds RNA and anti-LSm1. The results showed that LSm1 localized to sites of DENV replication indicated by ds RNA signal in DENV infected cells. These results suggest that LSm1 is increased in DENV-2 infected cells and is recruited to the viral replication site and binds to the 3’UTR of DENV-2, thus enhance DENV-2 RNA replication in host cells.Firstly, the protein level of DDX21 in DENV-2 infected cells was measured by western blotting using anti-DDX21 antibody. Total proteins from DENV infected cells were harvested at different time point post infection. The western blotting results showed that the protein level of DDX21 was decreased at 12 h and 24 h post infection, and was then recovered at 36 h post infection. The protein levels of DDX21 in nucleus and cytoplasm in response to DENV-2 infection were also evaluated by western blotting. The results showed that the amount of DDX21 in the cytoplasm of infected cells was increased slightly compared to the uninfected cells. However, the amount of DDX21 detected in the nucleus of infected cells was reduced to one third compared to the uninfected cells. Furthermore, the DDX21 localization in DENV-2 infected and non-infected cells was examined by using the confocal laser scanning microscopy. The result also showed that the DDX21 signal in nucleus was reduced significantly in DENV infected cells(ds RNA signal), while DDX21 could barely be detected in the cytoplasm of infected cells. The DDX21 knockdown and overexpression conditions were setup, respectively, followed by DENV-2 infection, the viral RNA level was then assayed by RT-q PCR. The results showed that the viral RNA level was increased in DDX21 silenced cells and decreased in DDX21 overexpressed cells. To understand the mechanism of DDX21 inhibiting the replication of DENV-2, the dual-luciferase assay was used to monitor the IFN-β-promoter activity and ISRE activity in DDX21-overexpressing cells during DENV infection. The reporter assays showed that the transcription level of IFN-β and ISRE was increased by approximately 1.6- and 1.3- fold respectively in DDX21 overexpressed cells compared to vector control group. After DENV infection, the transcription level of IFN-β and ISRE was increased greatly by approximately 10.9- and 15.2- fold respectively in DDX21 overexpressed cells compared to control group. To further investigate the way DENV-2 subverts the inhibition role of DDX21, the DDX21 expressing plasmid together with DENV NS2 b and NS3 expressing plasmids were transfected into 293 T cells. The total proteins were harvested and the DDX21 protein levels were assayed by immunoblotting. The level of DDX21 was significantly decreased when co-expressed with DENV NS2b/3. The 293 T cells were co-transfected with p D21 as well as p NS2 b and p NS3 for 12 hours and then treated with 5μM MG132 or DMSO for additional 12 hours. DDX21 protein levels were then assayed by Western blotting. The result showed that the degradation process of DENV NS2b/3 towards DDX21 was suppressed by MG132. These results suggested that DDX21 translocates from nucleus to cytoplasm to active the IFN-β and ISRE transcription in response to DENV infection, and thus inhibits DENV replication in early infection. This innate immunity is then subverted by the degradation of DENV NS2b/3 protease towards DDX21. The RIP result also suggest that there is interaction between DDX21 and DENV-2 3’UTR during DENV infection.2. Development of a new approach to isolate in vivo assembled DENV-2 5’-3’UTR-RBP complexTo avoid potential biochemical artifacts and circumvent limitations of the in vitro approach, a new method was used to identify authentic RNA-interacting proteins from living cells. In this approach, t he S1 RNA aptamer sequence was cloned into a plasmid in the 3′ terminus of 5’-3’UTR. The plasmid was transfected and the hybrid RNA was thus transcribed in 293 T cells in the presence of the 4-SU which was thus incorporated. After endogenous viral RNP was assembled, the RNA-protein complexes were cross-linked by 365 nm UV light irradiation and the RNPs were isolated from whole cell extract through binding of S1 aptamer to Streptavidin magnetic beads. The eluted proteins were then detected by SDS-PAGE and the selected 3 specific bands were analyzed with LC-MS/MS. There are totally 32 proteins were identified.To further investigate the potential role of RPS8, the important component of IGF2 BP m RNA granule, in DENV replication, firstly we assayed the viral RNA level in RPS8 overexpression cells by RT-q PCR. The results showed that the viral RNA level was increased in RPS8 overexpressed cells. Meanwhile, the infectious particles in cellular supernatant of the LSm1 silenced cells were counted by plaque assay. BHK-21 cells were infected with serial dilutions of DENV-2 containing supernatant from RPS8 overexpressed cells infected with DENV-2. The cells were stained with crystal violet and the plaques were counted. The results showed that production of virus in supernatant were increased in RPS8 over-expressing cells. The interaction of RPS8 and DENV RNA were studied by RIP assay. The RIP eluted samples were conducted to RT-PCR using primers of the DENV 5′UTR and 3’UTR., and the result showed that both DENV RNA 3’UTR and 5′UTR bound specifically to RPS8. The subcellular localization of RPS8 during DENV-2 infection was then examined using confocal laser scanning microscopy. The DENV infected cells were stained with anti-ds RNA and anti-RPS8. The result showed that RPS8 localized to sites of DENV replication indicated by ds RNA signal in DENV infected cells. These results suggest that RPS8 interacts with both DENV 3’UTR and 5′UTR, and localizes at site of DENV replication, possibly as part of viral replication complex, supports DENV-2 RNA replication in host cells.ConclusionIn this paper the RNA affinity purification strategies operated in vitro and in vivo allow successful purification of viral RBPs. The host proteins LSm1 and RPS8 positively participate in the replication of DENV. LSm1 binds to the 3’UTR of DENV RNA, and RPS8 interacts with both the 5’UTR and 3’UTR of DENV RNA.The DDX21 suppresses the DENV replication. The DDX21 recruit to cytoplasm from nucleus in response to DENV infection, and binding to the the 3’UTR of DENV RNA. The DDX21 upregulates innate immune-signaling processes and activates type I IFN and ISG production during DENV infection, and later during infection, DDX21 is degraded by the DENV NS2B/3 protease complex and the innate immunity is thus subverted to facilitate DENV replication.