1. Mass spectrometry analysis revealed that pUL43 or pUL24 proteins interact with TRBP and Dicer respectively
UL24 and UL43 are both members of the US22 family of HCMV, and there are reports in the literature that the viral proteins they encode have interactions [10]. Therefore, we hypothesize whether UL24 and UL43 can cooperate and regulate some of them like IRS1 and TRS1 members in the US22 family. The reaction within the cell also plays an important role in the process of viral infection. To explore this idea, we constructed Flag-tagged GFP, UL24, and UL43 overexpression plasmids and transfected them into HEK293T cells. After 48 hours of transfection, the cells were collected and lysed with cell lysate, and then added to the magnetic beads with Flag antibody. After incubation, the Flag-tagged protein was bound to the magnetic beads, and finally, the eluted product was competitively eluted with Flag peptide.
The eluted products are separated by SDS-PAGE protein gel, and then the protein gel is silver-stained, as shown in Figure 1A
The UL43 and UL24 were shown many specific bands compared to GFP as shown in Figure 1B. therefore, we expanded the number of cells for immunoprecipitation and Coomassie brilliant blue staining and took specific bands for mass spectrometry. Based on the mass spectrometry results, the specific protein that interacts with pUL43 was TRBP, and the specific protein that interacts with pUL24 was Dicer. TRBP and Dicer can be combined, and they are important components that regulate the production of miRNAs. It is reported in the literature that pUL24 and pUL43 can interact. Therefore, we speculate whether UL24 and UL43 can coordinately regulate TRBP and Dicer, thereby affecting the production of miRNAs.
2. Overexpression to verify the protein-protein interaction detected by mass spectrometry
To prove that these two proteins can indeed interact specifically with pUL24 or pUL43, we perform immunoprecipitation experiments and detect whether they can interact with TRBP or Dicer by immunoblotting with specific antibodies. In the experiment, the HCMV virus gene US31 was used as a negative control, and a Flag-tagged US31 overexpression plasmid was constructed. We also constructed an HA-labeled UL43 overexpression plasmid to verify the interaction between pUL24 and pUL43. Because Dicer and TRBP are both important members of RISC, and the complex contains small nucleic acid molecules, to rule out that the interaction between proteins is mediated by nucleic acid rather than the direct action of protein, we treated the sample with RNase A enzyme, and then used Magnetic beads labeled with Flag antibody were used for immunoprecipitation. The results of western blotting showed that pUL24 or pUL43 can interact with Dicer and TRBP, and pUL24 can also interact with pUL43.
Next, we use TRBP as bait to test whether pUL24 and pUL43 can be co-precipitated. First, we constructed a TRBP overexpression plasmid with HA tag and transfected it with US31, UL24, or UL43 into HEK293T cells. Cells were lysed 48 hours after transfection and treated with RNase A enzyme as well. Then incubate the protein A magnetic beads with HA antibody to form a HA antibody-labeled magnetic bead, then add the processed cell supernatant to the magnetic beads, undergo immunoprecipitation, and finally separate the protein on the magnetic beads with the lysis solution. The product is eluted. The results of western blotting showed that TRBP could co-precipitate pUL24 and pUL43. In summary, pUL24 or pUL43 can interact with Dicer and TRBP, which are specific interactions between proteins and are not mediated by nucleic acids.
3. Verification of the interaction between cell endogenous proteins
Although we have demonstrated that pUL24 or pUL43 can interact with Dicer and TRBP through overexpression in HEK293T cells, this is consistent with our mass spectrometry results. However, in the case of viruses infecting cells, whether these two viral proteins can also interact with Dicer and TRBP needs further verification. We first carried out viral modification in the BAC containing the full genome of HCMV, respectively adding the HA tag to the C-terminal of the UL24 virus gene, and adding the Flag tag to the C-terminal of UL43 to construct two new BACs, namely pBAC-AD -UL24-HA and pBAC-AD-UL43-Flag. BAC was electroporated to MRC5 cells to obtain and infect the cells with these two viruses. After 48 hours of infection, the cells are collected and tested in related experiments.
As shown in Figure 3A, we respectively infected MRC5 cells with the viruses AD-GFP and AD-UL24-HA, collected cell samples 48 hours after infection, and then used HA antibody-coated magnetic beads for immunoprecipitation. The results of western blotting showed that compared with AD-GFP-infected cells, in AD-UL24-HA-infected cells, UL24-HA could co-precipitate TRBP and Dicer. In Figure 3B, we also used the AD-UL24-HA virus to infect cells, and Protein A magnetic beads were used and incubated with IgG (control) and TRBP antibodies.
The supernatant was added to the processed magnetic beads. The results of co-immunoprecipitation and immunoblotting showed that the antibody incubated with IgG could not co-precipitate pUL24, but the antibody incubated with TRBP antibody could co-precipitate pUL24 and Dicer. These two experimental results show that pUL24 expressed by the virus can interact with the endogenous TRBP and Dicer of the cell.
Similarly, to identify whether pUL43 can interact with endogenous TRBP and Dicer, like pUL24 in the infected cells. As shown in Figure 3C, MRC5 cells were infected with the viruses AD-GFP and AD-UL43-Flag, and cell samples were collected 48 hours post-infection, and then used Flag antibody-coated magnetic beads for immunoprecipitation. The results of western blotting showed that, in the cells infected with AD-UL43-Flag, UL43-Flag was able to co-precipitate TRBP and Dicer compared with the cells infected with AD-GFP. Furthermore, in Figure 3D, we infected cells with the AD-UL43-Flag virus and the Protein A magnetic beads were incubated with IgG or TRBP antibodies, then the supernatant of the infected cell lysate was added to the processed magnetic beads. The results of blotting showed that the antibody incubated with IgG could not co-precipitate pUL43, but the antibody incubated with TRBP could co-precipitate pUL43 and Dicer. These two experimental results show that pUL43 expressed by the virus can interact with the endogenous TRBP and Dicer in the cell.
In summary, the experimental results show that in the process of HCMV virus infection, the viral protein pUL24 or pUL43 can interact with the endogenous TRBP and Dicer.
4. Double knockout of UL24 and UL43 does not affect virus replication
According to the above experimental results, we found that pUL24 or pUL43 can interact with TRBP and Dicer, because TRBP and Dicer are two important components of RISC, which can regulate the production of miRNAs and thus regulate the expression of some genes, therefore; we speculate that whether these two viral proteins can work together to participate in the regulation of the production of miRNAs through the interaction with TRBP and Dicer that may ultimately play some important biological functions. Previous studies have shown that UL24 and UL43 are non-essential factors for the growth of the HCMV, any deletion of them will not affect virus replication [31, 32]. However, there is no related report of double deletion, we knocked out UL24 and UL43 in the wild-type virus genome at the same time to detect whether the replication of the virus will be affected. To perform viral gene knockout, the red recombination system was applied as reported in the previous literature [ 24].
We first electrotransformed the IsceI-KanS containing the UL24 C-terminal homologous fragment into E. coli GS1783 containing wild-type pBAC-AD/Cre-GFP for homologous recombination, and then induced the enzyme to excise the IsceI-KanS, and finally the BAC deleted UL24 was generated, named pBAC-AD-d24. Besides, using the same method as UL24 knockout, the IsceI-KanS containing the C-terminal homologous fragment of UL43 was electroporated into E. coli GS1783 containing pBAC-AD-dUL24 for homologous recombination, and finally BAC double deleted UL24 and UL43 were obtained and named pBAC-AD-dd24-dd43. We finally obtained the double deletion virus AD-GFP-dd24-dd43 by electroporating BAC into MRC5 cells. We use the wild-type virus AD-GFP (or AD-WT) and the double deletion virus AD-GFP-dd24-dd43 (Or AD-Mut) with a low multiplicity of infection (MOI) of 0.1 and a high multiplicity of infection of 1. The MRC5 cells were incubated with the virus for 2 hours, and then removed the supernatant and replaced with a fresh medium. Finally, the cell supernatant was collected at a specific time point post-infection and used to test the virus titer. In Figure 4A-B, we found that the replication ability of the UL24/UL43 double-deleted virus was close to that of the wild-type virus.
Altogether, these results indicate that the double-knockout and single-knockout phenotypes of these two genes are the same and do not affect the virus replication and also show that these two viral genes are indeed non-essential genes for the growth of the HCMV virus.
We also tested the HCMV early gene IE1, the early gene UL44, and the late gene UL99 (pp28) and UL32 (pp150) expression. We infected MRC5 cells with MOI of 1 and collected cells at a specific time point for western blotting. In Figure 4C, we found that in cells infected with a double-deletion virus, the expression of viral protein IE1 and even other tested genes were down-regulated. We further tested the transcription of early viral genes, infected MRC5 cells with an MOI of 1, and collected RNA from the cells 8 hours after infection, and performed RT-qPCR. As shown in Figure 4D, the mRNA level of IE1 was also down-regulated in the early stage of infection. These data suggest that UL24 and UL43 may affect the early stage of virus replication. Next, we had to explore whether it is the double knockout of these two viral genes that affect virus entry into cells or virus viability.
5. Double knockout of UL24 and UL43 does not affect the ability of the virus to enter cells and the stability of the virus
Based on the above experimental results, we speculate that the double knockout of UL24 and UL43 genes may affect the ability of the virus to enter the cell or the viability of the virus. To test that, MRC5 cells were infected with the wild-type or deleted viruses with MOI of 1 for 2 hours, then the fresh medium was replaced and the cells were collected at the indicated time points and data analyzed by quantitative PCR. The input was the virus stock solution. In Figure 5A-C, we find that there is almost no difference between the wild-type and the deleted virus genome. This shows that the double knockout of UL24 and UL43 (AD-Mut) does not affect the ability of the virus to enter the cell.
Next, we performed virus particle stability tests on wild-type (AD-WT) and mutant viruses (AD-Mut). We treated the wild-type virus and the deleted virus with the 37 0C for 4, 8, and 16 hours. Then, the untreated virus and the processed virus were used to infect MRC5 cells at an MOI of 1. After 24 hours of infection, we collected and incubated the cells with the virus IE1 antibody and the corresponding secondary antibody, and the IE1 positive cells were counted under the fluorescent microscope. As shown in Figure 5D, the number of IE1 positive cells gradually decreases with the extension of the 37°C treatment time, but interestingly, we found that the number of IE1 positive cells in the mutated virus was almost similar to that of the wild-type virus regardless of whether it was processed with 37 0C or not. We express the stability of the virus in terms of infectivity, as shown in Figure 5E. The results were consistent with Figure 5D. it was clear that treatment with 37 0C has reduced the virus infectivity but there was no difference between the wild type and the deletion type in the reduction levels. In summary, these results indicate that the double knockout of UL24 and UL43 does not affect the ability of the virus to enter cells as well as the virus stability.
6. Double knockout of UL24 and UL43 does not affect the expression of 4 HCMV miRNAs related to immune escape
miRNAs are non-coding RNAs with a size of about 22 nucleotides and can participate in the regulation of many signaling pathways in cells. Initially, the Pfeffer study group were identified 9 HCMV-encoded miRNAs [33]. Later on, 26 miRNAs have been discovered through a series of studies, and they are all produced by RISC processing [34]. These miRNAs perform various functions after the virus infects cells, such as regulating the cell cycle, the expression of certain host or viral genes, viral DNA synthesis, the generation of virus assembly centers, and the production of immune regulation-related inflammatory factors [35-41].
According to our results that the UL24 and UL43 can interact with TRBP and Dicer and their knockout (AD-Mut) does not affect virus replication, entry, and stability. Also, based on previous studies that many of the miRNAs encoded by HCMV are involved in regulating the immune response and helping the virus escape. We speculate that pUL24 and pUL43 may work together to regulate some viral miRNAs. They may regulate some signaling pathways to escape the immune system's recognition of the virus. There are currently four known HCMV miRNAs that are mainly involved in immune regulation, which are miR-UL112-3p, miR-US5-1, miR-UL148D, and miR-US25-1-5p, therefore; we speculated that whether the knockout of UL43 and UL24 genes affect the expression of these miRNAs. To perform that, MRC5 cells were infected with wild-type viruses at an MOI of 1, and the cells were collected to detect the expression of these miRNAs. We extract RNA from cells and use the specific reverse transcription primers of these miRNAs, as shown in Table 3, to obtain their cDNA by the Stem-loop RT PCR method, and finally, qPCR was used for quantitative detection. As shown in Figure 6A-D, the expression of these miRNAs was very high at 48 hours post-infection (hpi). Therefore, we have used the 48 hpi as the time point to compare the differences between these miRNAs in wild-type and mutated virus-infected cells. As shown in Figure 6E-H, the expression levels of these four miRNAs under different virus infections have no significant difference, and this result indicates that the double knockout of UL24 and UL43 does not affect the production of these miRNAs.
7. RNA-seq found that the double knockout of UL24 and UL43 resulted in the down-regulation of miR-UL59
To further analyze the role of UL24 and UL43 in HCMV virus infection of host cells, and to find the target miRNAs regulated by these two genes, we performed RNA-seq. The MRC5 cells were infected with AD-WT and AD-Mut respectively with an MOI of 1. After 48 hours post-infection (hpi), cell collection and RNA extraction with 1ml Trizol were performed and sent for RNA-seq. Based on the results of RNA-seq, we found that the expression of miR-UL59 was different. For further verification, we infected the MRC5 cells with AD-WT and AD-Mut again at an MOI of 1. Each virus was treated with two identical treatments. After 48 hpi, we used Trizol to collect cells and extract RNA and the Stem-loop method to design miR-UL59 reverse transcription primers and quantitative qPCR primers (Table 4) [42-43]. Experimental results showed that the double knockout of UL24 and UL43 resulted in the downregulation of miR-UL59.
At present, there are limited research reports on miR-UL59, and it is described that the target gene of miR-UL59 is ULBP1 (UL16-binding protein 1) [44]. ULBP1 is one of the ligands of NKG2D, and NKG2D is an activating receptor expressed on immune effector cells, which can recognize different MHC I related ligands, including MIC and ULBP proteins. Infection or stress response can induce the expression of the NKG2D ligand, leading to the activation of effector cells and ultimately killing the ligand-related target cells [45-48]. Besides, it has previously been reported that the membrane glycoprotein UL16 of HCMV can bind to three NKG2D ligands, namely MICB, ULBP1, and ULBP2, and UL16 is also very important for the immune escape of HCMV.
To check the impact of these two genes on ULBP1 mRNA levels, we similarly infected the MRC5 cells with AD-WT and AD-Mut at an MOI of 1. After 48 hpi, we used Trizol to cells were collected and the ULBP1 mRNA levels were measured by quantitative qPCR using specific primers as shown in Table 4 [42, 43]. Experimental results showed that the double knockout of UL24 and UL43 has resulted in up-regulation of ULBP1 mRNA levels
Altogether, RT-qPCR results indicated that the double knockout of UL24 and UL43 resulted in the down-regulation of miR-UL59 and the up-regulation of ULBP1 mRNA levels. This suggests that UL24 and UL43 may regulate the expression of miR-UL59, which in turn affects the expression of ULBP1, and thus regulating anti-virus immune response, this will need further experimental confirmation in the future.
8. Double knockout of UL24 and UL43 in clinical virus strains does not affect virus replication, entry, and virus stability
In addition to the HCMV experimental virus strain, we also want to know whether the double knockout of UL24 and UL43 genes in the HCMV clinical virus strain (TB40E-Mcherry = TB-WT), can similarly affect the virus replication and its cellular entry. At first, we generated UL24/UL43 double deletion virus named TB40E-Mcherry-dd24 -dd43 (TB-Mut). Then to analyze the growth ability of this clinical mutant virus (TB-Mut), we infected the MRC5 cells with the wild type and the deleted virus for 2 hours at an MOI of 0.1 or 1, and then the medium was replaced with a fresh one. Finally, the cell supernatant was collected at a specific time point post-infection and was used to infect MRC5 cells and the virus titer was measured by tissue culture infectious dose 50% (TCID50) assay. As shown in Figures 8A-B, the growth ability of the Mut virus was almost similar to that of the wild-type virus. These results indicate that the double knockout of UL24 and UL43 in clinical virus strain, similar to the experimental virus strain, does not affect the replication of the virus, thus confirming that the UL24 and UL43 are non-essential genes for clinical virus strains too.
Next, we checked whether the knockout of these two genes affects virus entry into cells. To perform that, the MRC5 cells were infected with wild-type and deleted clinical viruses at an MOI of 1 for 2 hours, then the supernatant was removed and replaced with fresh ones. The cells were collected at the specified time point post-infection, and the viral genome (viral DNA) was extracted and analyzed by qPCR. The virus stock solution was used as Input. As shown in Figure 8C-E, there was almost no difference in the amount of wild-type virus genome and deleted virus genome, which shows that the double knockout of UL24 and UL43 does not affect the ability of clinical strains to enter cells as well.
Furthermore, we tested the virion stability of wild-type and deleted viruses in the clinical strain. The wild-type virus and the deleted virus were treated with 37 0C for 4, 8, and 16 hours, then the MRC5 cells were infected with the untreated and the treated virus separately with MOI of 1. At 24 hours post-infection (hpi), the virus IE1gene was incubated with primary and the corresponding fluorescent secondary antibody, finally the IE1 positive cells were calculated under the fluorescence microscope. As shown in Figure 8F, the number of IE1 positive cells gradually decreases with the extension of the 37°C treatment time but there was no significant difference between the number of IE1 positive cells with both the wild type and mutant virus. Also, to check the stability of the virus in terms of its infectivity. As shown in Figure 8G-F, the treatment with 37 0C has reduced the infectivity of the virus without a significant difference between the wild type and the deletion type. In summary, these results indicate that HCMV clinical viruses were consistent with experimental strain in that the double knockout of UL24 and UL43 does not affect the entry of clinical strains into cells and the stability of virus particles.
9. Double knockout of UL24 and UL43 in clinical viruses also leads to down-regulation of miR-UL59
The previous results showed that in the experimental virus strain, the double knockout of UL24 and UL43 would inhibit the expression of miR-UL59, and the expression of its target gene ULBP1 was also affected, so we wanted to observe whether this phenotype is in clinical virus strains unanimous.
We spread MRC5 cells in a 12-well plate and then infected the cells with wild-type virus and deletion mutant virus with MOI of 1. After 48 hours of infection, the cells were harvested with Trizol, and RNA was extracted. Next, quantitative PCR primers were used to detect the expression of miR-UL59. In Figure 3.9A, we detected that the expression of miR-UL59 was down-regulated after the virus-infected cells. In 3.9B, we also found that the target gene of miR-UL59, ULBP1, was down-regulated. The mRNA level was higher in cells infected with the deletion virus than in cells infected with the wild-type virus, and the results were consistent with the results of the experimental virus strains, indicating that knocking out UL24 and UL43 in clinical strains can also inhibit the expression of miR-UL59 and upregulate the mRNA of ULBP1 Level. As one of the ligands of NKG2D, ULBP1 is essential for cellular immunity. Therefore, we speculate that although UL24 and UL43 of the HCMV virus do not affect virus replication, they may participate in the regulation of the expression of some miRNAs in the cell and thus participate in immune regulation to promote the immune escape of the virus.