Eukaryotic translation elongation factor 1 alpha (eEF1A) inhibits Siniperca chuatsi rhabdovirus (SCRV) infection through two distinct mechanisms

ABSTRACT

Rhabdoviruses are single-stranded, negative-sense RNA viruses with broad host range, several of which are important pathogens. Compared with the rhabdoviruses infecting mammals, host factors involved in aquatic rhabdovirus infection have remained largely unknown. In the present study, we report the roles of host eukaryotic translation elongation factor 1 alpha (eEF1A) on the infection of Siniperca chuatsi rhabdovirus (SCRV, genus Siniperhavirus), which is an important pathogen of mandarin fish. eEF1A was identified from SCRV nucleoprotein (N)-based affinity purified proteins. Further protein interaction and mutation assays proved that eEF1A interacted not only with the N protein but also the virus matrix protein (M), which relied on the N-terminal of eEF1A. SCRV infection and overexpression of N or M all stimulated the promoter activity of the eEF1A gene and, thus, upregulated its expression, whereas the upregulated eEF1A inhibited the transcription of SCRV genome. Mechanistically, eEF1A impaired the interactions between N and phosphoprotein (P), or N and N, which are important for the efficient transcription and replication of rhabdovirus. Meanwhile, eEF1A promoted the ubiquitin-proteasome degradation of the M protein, which relied on lysine 48 (K48) of ubiquitin. In addition, we showed that the ubiquitination degradation of M protein relied on C-terminal domain of eEF1A, but inhibition of the N-P or N-N interactions needs its entire length. Collectively, these results revealed two different mechanisms used by eEF1A to resist a fish rhabdovirus, which provided novel insights into the role of eEF1A in rhabdovirus infection and new information for antiviral research.

IMPORTANCE

Although a virus can regulate many cellular responses to facilitate its replication by interacting with host proteins, the host can also restrict virus infection through these interactions. In the present study, we showed that the host eukaryotic translation elongation factor 1 alpha (eEF1A), an essential protein in the translation machinery, interacted with two proteins of a fish rhabdovirus, Siniperca chuatsi rhabdovirus (SCRV), and inhibited virus infection via two different mechanisms: (i) inhibiting the formation of crucial viral protein complexes required for virus transcription and replication and (ii) promoting the ubiquitin-proteasome degradation of viral protein. We also revealed the functional regions of eEF1A that are involved in the two processes. Such a host protein inhibiting a rhabdovirus infection in two ways is rarely reported. These findings provided new information for the interactions between host and fish rhabdovirus.

INTRODUCTION

Eukaryotic translation elongation factor 1 alpha (eEF1A) is an essential component of the translation machinery and exists in abundant amounts in cells (1). During the translation process, it transports aminoacyl-tRNAs to the A site of the ribosome for polypeptide synthesis in its guanosine triphosphate (GTP)-bound form. Structural analysis has shown that eEF1A consists of three domains: domain I contains N-terminal about 240 amino acid (aa) residues and is responsible for the binding with GTP; domain II located at the central region and can bind the aminoacyl end of the aa-tRNA, and domain II and C-terminal domain III are related to actin binding (2, 3). Domain I and II also can bind another subunit of the eEF1 complex (4). Besides the conventional role in synthesizing nascent peptide chains, eEF1A has been reported to be involved in regulating microtubules and actin cytoskeleton by binding to actin (2). In addition, increasing reports are emerging about other functions of eEF1A, including modulation of nuclear export, proteolysis, apoptosis, carcinogenesis, and virus propagation (1, 5 – 8).

Rhabdoviruses are negative-sense and single-stranded RNA viruses belonging to the family Rhabdoviridae which contains 45 genera (9). These viruses have been isolated from plants and animals, including fish and mammals, several of which are important pathogens of their host (10 – 12). Almost all rhabdovirus genomes encode five structural proteins, nucleoprotein (N), phosphoprotein (P), matrix protein (M), glycoprotein (G), and polymerase (L), which form the enveloped with bullet-shaped virions (9). There are five genera of rhabdoviruses that infect finfish and represent significant threats to the health of aquatic animals (9). Siniperca chuatsi rhabdovirus (SCRV) is a member of the genus Siniperhavirus and an important pathogen of mandarin fish (Siniperca chuatsi, also known as Chinese perch), which can cause severe hemorrhagic septicemia and death in infected host (13, 14). SCRV has a genome with a size of approximately 11.5 kb, including a 3′- untranslated region (3′-UTR), a 5′-UTR, and five sequential open reading frames which, respectively, encode the five structural proteins (15).

Rhabdovirus N, P, L, and the negative genomic RNA constitute virions’ ribonucleoprotein (RNP) complex (9, 16). The genome RNA is entirely coated by N protein to form the N-RNA, serving as the template for transcription and replication. P protein is the essential cofactor of polymerase L. Structure analysis has shown the interactions between N and P or P and L in RNP from virions of vesicular stomatitis virus (VSV), a rhabdovirus infecting humans (17). After entry into cells, initial transcription is carried out rapidly, and genome replication will proceed in a second stage. Interaction between N and P proteins is vital for recruiting L to the N-RNA for transcription (17, 18). The newly replicated complementary (+)RNA and the genomic (−)RNA are immediately encapsidated by the N protein, which is accompanied by the oligomerization of the N protein itself (16). Besides providing the template for transcription and replication, the formation of the N-RNA complex protects the viral genome from degrading by host RNase and avoids the recognition by the host immune system (19, 20). The formation of the N-P complex also prevented the N protein from encapsidation with cellular RNAs (21). N protein has also been reported to regulate transcription and replication by binding to the viral genome’s UTR region, such as in snakehead vesiculovirus (22). In addition, the N protein also plays several other roles in rhabdovirus infection, such as exerting immunogenicity to induce host immune response and mediating immune escape by interactions with the host immune system (23 – 28). These findings strongly suggest that N protein plays important and indispensable roles in rhabdovirus infection.

M protein is the smallest of the five structural proteins of rhabdovirus. In virions, the M protein resides between the RNP and G protein-studded membrane envelope (29). During virus infection, the M protein binds to the plasma membrane and interacts with the RNP complex to facilitate virus assembly (29, 30). Besides the roles in virus morphogenesis, the M protein also possesses several important functions, such as suppression of host gene transcription, modulation of host antiviral immune responses, deregulation of host nucleo-cytoplasmic transport, and induction of apoptosis (31 – 37). For example, interactions with host eukaryotic translation initiation factor complex, including eukaryotic translation initiation factor 3 subunit I (elF3i) and subunit h (elF3h), inhibit host translation (38, 39).

Although rhabdovirus can regulate many cellular responses to facilitate its replication by interacting with host proteins, the host can also restrict virus infection through these interactions. It has been reported that host tripartite motif protein 41 (TRIM41) interacted with the N protein of VSV and restricted virus infection (40). NADH: ubiquinone oxidoreductase complex assembly factor 4 (Ndufaf4) inhibits VSV growth via interacting with virus M protein (41). For fish rhabdovirus, host Heat shock cognate protein 70 (HSC70) was reported to interact and induce the lysosomal degradation of the G protein of spring viremia of carp virus (SVCV), and zebrafish maoc1 can interact with the P protein of SVCV and promote its degradation, which both inhibited the virus infection (42, 43). Thus, identifying host proteins that interacted with rhabdovirus proteins could be an approach to understanding virus-host interactions and uncovering antiviral targets.

In the present study, we tried to investigate the interactions between host and aquatic rhabdovirus proteins using SCRV N protein as bait. The host eEF1A was identified in the N protein immunoprecipitated complex by mass spectrometry (MS) analysis. The eEF1A was further proved that it not only interacted with N but also with M protein and inhibited virus infection through two different mechanisms: inhibition of the formation of N-P complex and N oligomers and degradation of M protein through the ubiquitin-proteasome pathway. Our findings provided novel insights into understanding the interactions between host and rhabdoviruses.

RESULTS

eEF1A interacts with SCRV N and M proteins by N-terminal 200 amino acids

A pull-down assay was performed to identify host proteins interacting with SCRV protein using the Strep-tagged SCRV N protein as bait in Fathead minnow (FHM) cells. Subsequent MS analysis obtained several host proteins which could interact with the N protein (Table 1). One of the host proteins, eEF1A, has been shown to be associated with RNA virus infection including West Nile virus and respiratory syncytial virus (RSV) (44, 45) and was chosen for further analysis. Nanoluciferase (Nluc) complementation and Co-Immunoprecipitation (co-IP) assays were performed to confirm the interaction. Co-expression of the N-terminal Nluc (NlucN) fused eEF1A (NlucN-eEF1A-Flag) and C-terminal Nluc (NlucC) fused N protein (N-NlucC-HA) significantly restored the nano-luciferase activity compared with the control (Fig. 1A), indicating the interaction between eEF1A and N protein. In co-IP assays, when HA-tagged N and Flag-tagged eEF1A were coexpressed, the anti-HA antibody (Ab)-immunoprecipitated N protein complex was recognized by the anti-Flag antibody. A similar situation was observed when HA-tagged eEF1A and Flag-tagged N were coexpressed (Fig. 1B). We further examined the interactions between endogenous eEF1A and N with or without SCRV infection. As shown in Fig. 1C, in HA-tagged N overexpressed FHM cells, eEF1A was recognized by anti-eEF1A antibody in the anti-HA Ab-immunoprecipitated N protein complex with or without virus infection. Similarly, in eEF1A-HA overexpressed cells, the anti-HA Ab-immunoprecipitated eEF1A protein complex was recognized by an anti-N antibody in the virus-infected sample (Fig. 1D). Thus, the above results all proved that the eEF1A interacted with the N protein.

TABLE 1.

Proteins obtained in MS analysis after N protein-based pull-down

Protein name	Protein ID
Oxoglutarate(alpha-ketoglutarate) dehydrogenase a (lipoamide)	XP_039521404.1
Voltage-dependent anion-selective channel protein 2	XP_039528264.1
Desmoplakin	XP_039530441.1
Splicing factor, proline- and glutamine-rich	XP_039536640.1
ATP-dependent RNA helicase A	XP_039510655.1
Elongation factor 1-alpha	XP_039536767.1
Pre-mRNA-processing factor 19	XP_039509793.1
Neuroblast differentiation-associated protein AHNAK	XP_039504520.1

Fig 1.

eEF1A interacted with SCRV N and M proteins. (A) Analysis of the interactions between eEF1A and N by NanoLuc complementation assy. The plasmids expressing NlucN-eEF1A-Flag were co-transfected with N-NlucC-HA or NlucC-HA (Vector) into Epithelioma papulosum cyprini (EPC) cells. Co-expression of eEF1A and N significantly increased the luciferase activity. (B–D) eEF1A interacted with N protein by Co-IP analysis. Plasmids expressing N-HA and eEF1A-Flag, or eEF1A-HA and N-Flag, were co-transfected into EPC cells (B). At 36 hour post-transfection (hpt), cell lysates were IP with anti-HA gel. Then, the immunoprecipitates and whole cell lysate (Input) were analyzed by IB with indicated Abs. Besides, FHM cells were transfected with plasmids expressing N-HA (C) or eEF1A-HA (D) and infected with SCRV (1 MOI). At 36 hpi, cell lysates were IP with anti-HA gel and then analyzed by IB with indicated Abs. (E) Analysis of the interactions between eEF1A and M by NanoLuc complementation assay as described above. Co-expression of eEF1A and M significantly increased the luciferase activity. (F and G) eEF1A interacted with M protein by Co-IP analysis. Plasmids expressing eEF1A-HA and M-Flag, or M-HA and eEF1A-Flag, were co-transfected into EPC cells (F), or the FHM cells were transfected with a plasmid expressing M-Flag and infected with SCRV (1 MOI) (G). Followed IP and IB analyses were performed as described above. (H and I) No interactions were detected between eEF1A and P, or eEF1A and G by NanoLuc complementation assay and Co-IP.

We further wondered whether the eEF1A interacted with other proteins of the virus. Surprisingly, eEF1A was also found to interact with the M protein. Co-expression of NlucN-eEF1A-Flag and M-NlucC-HA in Epithelioma papulosum cyprini (EPC) cells restored the nano-luciferase activity (Fig. 1E). And, the interaction between eEF1A and M was also observed in eEF1A-HA and M-Flag, or M-HA and eEF1A-Flag, overexpressed cells by Co-IP (Fig. 1F), which was also proved by recognition of the endogenous eEF1A in anti-Flag Ab-immunoprecipitated M protein complex in M-Flag overexpressed cells with or without SCRV infection (Fig. 1G). In contrast, eEF1A did not interact with the P and G proteins in the Nluc complementation and co-IP assays (Fig. 1H and I). Interaction assays between eEF1A and L protein were not performed for the difficult expression of L protein in transfected cells in our test.

To further verify the interactions between eEF1A and N or M protein, the localization of these proteins in overexpressed fish cells was observed by immunofluorescence. When Flag-tagged eEF1A and HA-tagged N were expressed in EPC cells alone, both were located in the cytoplasm in a diffusely distributed pattern (Fig. 2A). Although the distribution pattern was not changed when the two proteins were coexpressed, co-localization can be observed in several regions (Fig. 2A). Unlike the N protein, the Flag-tagged M protein was distributed in both cytoplasm and nucleus when expressed alone (Fig. 2B). When eEF1A-HA and M-Flag were coexpressed, the eEF1A-HA can be observed in the nucleus with M-Flag besides their co-localization in the cytoplasm (Fig. 2B), indicating the interactions between eEF1A and M protein.

Fig 2.

Localization of eEF1A, N, and M protein in transfected cells. (A) Plasmids expressing eEF1A-Flag or N-HA were alone or co-transfected into EPC cells. At 24 hpt, the cells were fixed and stained with anti-Flag or anti-HA antibodies and indicated by Alex fluor 546 or Alex fluor 488 conjugated secondary antibodies. eEF1A-Flag (red) showed co-localization with N-HA (green) in the cytoplasm when coexpressed. (B) Plasmids expressing eEF1A-HA or M-Flag were alone or co-transfected into EPC cells and treated as described above. eEF1A-HA (green) and M-Flag (red) co-localized in cytoplasm and nucleus. Bar = 10 µm.

To identify the specific region of eEF1A that interacts with N and M proteins, a series of vectors capable of expressing truncated mutants of eEF1A was constructed. These truncated mutants express the three domains of eEF1A separately and the N-terminal 200 amino acids (aa) of eEF1A (Fig. 3A). The N-terminal 200 aa of swine eEF1A has been reported to be an important region for interaction with the NS5A protein of classical swine fever virus (CSFV) (46).

Fig 3.

eEF1A interacting with N and M protein relies on its N-terminal 200 aa. (A) Schematic representation of eEF1A truncated mutants. These mutants are all fused with a Flag tag. (B) N protein interacted with the domain I of eEF1A (eEF1A-I). Plasmids expressing HA-tagged N protein and Flag-tagged eEF1A-I, eEF1A-II (domain II of eEF1A), or eEF1A-III (domain III of eEF1A) were co-transfected into EPC cells. At 36 hpt, cell lysates were IP with anti-HA gel. Then, the immunoprecipitates and whole cell lysate (Input) were analyzed by IB with indicated Abs. (C) N protein interacted with the N-terminal 200 aa of eEF1A. Plasmids expressing HA-tagged N protein and Flag-tagged eEF1A-I + II_Δ200 (domain I and II of eEF1A lacking N-terminal 200 aa) or eEF1A-I₂₀₀ (N-terminal 200 aa of eEF1A) were co-transfected into EPC cells. Followed IP and IB were performed as described above. (D) M protein interacted with the domain I of eEF1A. Plasmids expressing HA-tagged M protein and Flag-tagged eEF1A-I, eEF1A-II, or eEF1A-III were co-transfected into EPC cells. Followed IP and IB were performed as described above. (E) M protein interacted with the N-terminal 200 aa of eEF1A. Plasmids expressing HA-tagged M protein and Flag-tagged eEF1A-I + II_Δ200 or eEF1A-I₂₀₀ were co-transfected into EPC cells. Followed IP and IB were performed as described above.

HA-tagged N protein and Flag-tagged eEF1A mutants were expressed in EPC cells, and co-IP was used to verify whether the mutant could interact with N protein. As shown in Fig. 3B, Flag-tagged domain I of eEF1A (eEF1A-I) was detected by anti-Flag antibody in anti-HA Ab-immunoprecipitated N protein complex. In contrast, the domains II (eEF1A-II) and III (eEF1A-III) were not seen. Further co-IP with the N-terminal 200 aa of domain I (eEF1A-I₂₀₀) and domain I-II lacked N-terminal 200 aa (eEF1A-I + II_Δ200) showed that only the eEF1A-I₂₀₀ could be recognized in the immunoprecipitated N protein complex (Fig. 3C), which indicated that the N-terminal 200 aa of eEF1A is the specific region that interacts with N protein. Similar results were observed in the cells expressing HA-tagged M protein and these truncated mutants (Fig. 3D and E), which showed that M protein also interacts with the N-terminal 200 aa of eEF1A.

SCRV infection promotes eEF1A expression

Based on the results that eEF1A interacted with N and M proteins, the effect of SCRV infection on the expression of eEF1A was investigated. First, the Dual-Luciferase Reporter assay was performed to examine the impact of SCRV infection on the activity of the eEF1A promoter. The results showed that SCRV infection significantly increased the activity of the eEF1A promoter, while the promoter activity was not raised in the cells treated with UV-inactivated SCRV (Fig. 4A). Then, the effects of the over-expression of N or M protein on eEF1A promoter activity were detected. As shown in Fig. 4B, the activity of the eEF1A promoter was significantly stimulated with the overexpression of N or M protein. As a negative control, overexpression of SCRV P protein cannot promote the activity of the eEF1A promoter, which means the stimulation of eEF1A promoter activity by N and M protein is specific.

Fig 4.

SCRV infection and overexpression of N or M protein all promoted eEF1A expression. (A) Effects of SCRV infection on the promoter activity of eEF1A. FHM cells were transfected with pGL-ppeEF1A and pRl-TK and infected with SCRV (1 MOI) or UV-inactivated SCRV (SCRV_UV). At 24 hpi, the cells were lysed, and the luciferase activity was analyzed. (B) Effects of the overexpression of N, M, and P protein on the promoter activity of eEF1A. FHM cells were transfected with pGL-ppeEF1A, pRl-TK, and viral protein expression plasmid (pcDNA3.1-NlucN-N-Flag, pcDNA3.1-NlucN-P-Flag, pcDNA3.1-NlucN-M-Flag). At 24 hpt, the cells were lysed, and the luciferase activity was analyzed. (C and D) The mRNA level (C) and protein level (D) of eEF1A increased with the SCRV infection time. ImageJ was used to calculate the density of the indicated bands. (E and F) The mRNA level (E) and protein level (F) of eEF1A increased with the increased dose of infected SCRV. (G and H) The change in mRNA level (G) and protein level (H) was not obvious in cells infected with UV-inactivated SCRV. (I and J) Overexpression of N protein led to increased expression of eEF1A. Plasmid expressing HA-tagged N protein was transfected into FHM cells. Expression of eEF1A was detected by RT-qPCR (I) and IB (J) at the indicated time points. (K and L) Overexpression of M protein led to increased expression of eEF1A. Plasmid expressing Flag-tagged M protein was transfected into FHM cells. Followed RT-qPCR and IB were conducted as described above.

The expression level of endogenous eEF1A in SCRV-infected FHM cells was further detected. The results showed that the mRNA level of eEF1A was significantly increased along with the virus infection (Fig. 4C), as well as the protein level of eEF1A (Fig. 4D). Meanwhile, with the increased amount of the virus added into cells, the expression level of eEF1A was also increased (Fig. 4E and F). However, the expression level of eEF1A was not affected in cells added with UV-inactivated SCRV (Fig. 4G and H). In addition, the expression level of eEF1A was also examined in N and M overexpressed FHM cells, respectively. With the prolongs of the transfection time, the expression level of N-HA and M-Flag increased, and the expression level of eEF1A also increased (Fig. 4I through L). Collectively, the above results showed that SCRV infection promoted eEF1A expression, which was regulated by the expression of N and M proteins.

eEF1A negatively regulates SCRV infection

To understand what function eEF1A played in SCRV infection, the virus genome replication and gene expression were examined in eEF1A overexpressed and virus-infected FHM cells. Real-Time Quantitative PCR (RT-qPCR) showed that overexpression of eEF1A significantly reduced SCRV genome copies at 12, 24, 36, and 48 h post infection (hpi) compared with control (Fig. 5A). And, Immunoblot analysis also showed that expression of the N protein decreased significantly at 24 and 48 hpi in eEF1A overexpressed cells (Fig. 5B). In a parallel assay in which the FHM cells were transfected with different amounts of plasmids expressing Flag-tagged eEF1A, dose-dependent inhibition of N protein expression was observed (Fig. 5C). The results indicated that overexpression of eEF1A inhibited genome replication and gene expression of SCRV.

Fig 5.

eEF1A negatively regulated SCRV infection. (A and B) Overexpression of eEF1A reduced SCRV genome level and protein expression. Plasmid expressing Flag-tagged eEF1A or control plasmid were transfected into FHM cells and then infected with SCRV (1 MOI). The relative viral genome levels were determined by RT-qPCR at 12, 24, 36, and 48 hpi (A). The samples (24 and 48 hpi) were analyzed by IB with anti-N, anti-Flag, and anti-β-actin antibodies (B). ImageJ was used to calculate the density of the indicated bands. (C) Inhibition of SCRV protein expression by eEF1A was dose-dependent. Increased amounts of plasmid expressing Flag-tagged eEF1A were transfected into FHM cells respectively and then infected with SCRV. The expression of N protein at 24 hpi was analyzed with IB. (D) Overexpression of eEF1A reduced the amount of SCRV virions released in the culture medium. Supernatants of the cells overexpressing eEF1A and infected with SCRV were collected at 36 and 48 hpi. The number of virions in the supernatants was calculated by determining the viral genome copies with RT-qPCR. (E and F) Overexpression of eEF1A reduced the titer of SCRV. The whole cells overexpressing eEF1A and infected with SCRV were collected at 36 and 48 hpi. Viral titers were determined by the TCID₅₀ method (E). Siniperca chuatsi skin cell (SCSC) cells were infected with SCRV mentioned above and stained with crystal violet (F). (G and H) Knockdown of eEF1A expression upregulated SCRV infection. FHM cells were transfected with three siRNAs targeting eEF1A and the siNC. sieEF1A-2 showed the best inhibition effect on the eEF1A expression (G). Then, the cells were transfected with sieEF1A-2 and siNC and infected with SCRV. Virus genome levels and N protein expression increased at the detected time points (H).

Then, the number of free SCRV virions in the cell medium at 36 and 48 hpi was detected by calculating the viral genome copy numbers with RT-qPCR. The results showed that the overexpression of eEF1A resulted in a significant decline of viral genome copy numbers in the cell culture supernatants at the detected time points (Fig. 5D). Subsequent virus titration assay with the whole cell lysates obtained similar results. Virus titers from eEF1A overexpressed cells were significantly lower than that of the control (Fig. 5E), which can also be observed in crystal violet-stained infected cells (Fig. 5F). The results showed that overexpression of eEF1A reduced titers of SCRV.

In addition, RNA interference was used to downregulate the expression of eEF1A. Three siRNAs targeting eEF1A were synthesized and transfected into FHM cells. Followed detection of the eEF1A expression showed that sieEF1A-2 had the best inhibition effect (about 34% in mRNA level) compared with the control siRNA (siNC) among the three siRNAs (Fig. 5G). Thus, the sieEF1A-2 was used in the following experiment. FHM cells were transfected with sieEF1A-2 and siNC and infected by SCRV. Detection of the relative virus genome levels by RT-qPCR showed that inhibition of eEF1A expression by the sieEF1A-2 significantly increased the virus genome levels at the two detected time points compared with the control (Fig. 5H). The expression of N protein was also increased in the sieEF1A-2 group by IB analysis (Fig. 5H). Collectively, the above overexpression and knockdown assays all suggested that eEF1A played a negative regulatory role in SCRV infection.

eEF1A does not affect SCRV invasion but reduces virus transcription

The adsorption and entry assays were first performed to figure out what phase of SCRV infection would be affected by eEF1A. FHM cells overexpressing eEF1A were inoculated with SCRV at different MOI and incubated at 4°C and 25°C for 1 h, respectively. Examination of the viral genome copy numbers of the virions that attached or entered into cells showed that there was no significant difference between eEF1A overexpressed and control cells at 4°C (Fig. 6A) and 25°C (Fig. 6B), indicating overexpression of eEF1A does not affect the adsorption and entry of SCRV.

Fig 6.

Overexpression of eEF1A inhibited gene transcription of SCRV. (A and B) Overexpression of eEF1A did not affect the adsorption and entry of SCRV. FHM cells overexpressing eEF1A were infected with SCRV at 0.5 or 1 MOI and then incubated at 4°C (A) or 25°C (B) for 1 h. The virions attached or entered into cells were calculated by determining the viral genome levels with RT-qPCR. (C–E) Overexpression of eEF1A inhibited the transcription of five viral genes (N, P, M, G, and L) at 2 (C), 4 (D), and 6 hpi (E). The cells overexpressing eEF1A were pretreated with CHX and then infected with SCRV (1 MOI). Expression of the five viral genes was detected at the indicated time points by RT-qPCR.

Then, whether eEF1A affects virus genome transcription was investigated. FHM cells overexpressing eEF1A were infected with SCRV and treated with Cycloheximide (CHX), and the cells transfected with empty vector were used as control. CHX could inhibit the synthesis of proteins in cells, thus blocking the synthesis of viral proteins and progeny genomes, but it does not affect the transcription of viral genes. RT-qPCR detection of SCRV genes showed that expression of the five genes (N, P, M, G, L) significantly decreased at 2, 4, and 6 hpi compared to the control (Fig. 6C through E). Therefore, eEF1A inhibited the transcription of SCRV genes.

eEF1A reduces the protein level of M rather than N protein through the domain III

Although the above results showed that eEF1A inhibited SCRV infection by inhibiting the transcription of viral genes, the specific mechanism by which eEF1A inhibits SCRV through targeting N and M proteins is unclear. It has been reported that eEF1A was related to protein degradation (1). So, we coexpressed eEF1A and N, M proteins in EPC cells, respectively. Immunoblot analysis showed that there was no obvious difference in the content of N protein between eEF1A overexpressed cells and control at 24 and 48 h post-transfection (Fig. 7A). However, the protein bands of M protein in eEF1A overexpressed cells were weaker than that of the control at the two time points (Fig. 7B). A dose assay was also performed. With the increase of eEF1A protein, N protein maintained a relatively stable expression (Fig. 7C), but the quantity of M protein reduced accordingly (Fig. 7D), which revealed that eEF1A reduced the protein level of M but not N protein.

Fig 7.

Overexpression of eEF1A decreased the protein level of M but not N relied on its domain III. (A) The amount of N protein was not affected by overexpression of eEF1A. Plasmids expressing HA-tagged N protein and Flag-tagged eEF1A were co-transfected into EPC cells. Protein expression was analyzed by IB with indicated Abs. ImageJ was used to calculate the density of the indicated bands. (B) Overexpression of eEF1A reduced the protein level of M. Plasmids expressing Flag-tagged M protein and HA-tagged eEF1A were co-transfected into EPC cells. Protein expression was analyzed by IB with indicated Abs. (C) Expression of N protein was not affected by dose-increased expression of eEF1A. (D) Expression of M protein was reduced with the increase of eEF1A in a dose-dependent manner. (E) Overexpression of eEF1A-III reduced the protein level of M. Plasmid expressing HA-tagged M was co-transfected with Plasmids expressing Flag-tagged eEF1A-I + II, eEF1A-II + III, or eEF1A-III into FHM cells. The cells were analyzed by IB with indicated Abs. (F) Overexpression of eEF1A-I and II did not affect SCRV infection. Plasmids expressing Flag-tagged eEF1A-I + II were transfected into FHM cells. The cells were infected with SCRV. Virus infection was analyzed by detecting relative viral genome levels with RT-qPCR and IB with indicated Abs. (G and H) Overexpression of eEF1A-II + III (G) or eEF1A-III (H) inhibited SCRV infection. Plasmids expressing Flag-tagged eEF1A-II + III or eEF1A-III were transfected into FHM cells. The cells were infected with SCRV and analyzed as described above.

Previous studies have determined that eEF1A contains three domains (6) (Fig. 3). We then investigated whether the specific domains affect the protein level of M. Different mutants (eEF1A-I + II, eEF1A-II + III, eEF1A-III) were coexpressed with M protein. Immunoblot analysis showed that the existence of eEF1A domain III reduced the quantity of M protein (Fig. 7E). These mutants were also overexpressed in FHM cells, and the viral genome levels were detected. As shown in Fig. 7F, eEF1A-I + II cannot inhibit the replication of the SCRV genome at 12 and 24 hpi and the expression of the N gene. However, overexpression of eEF1A-II + III and eEF1A-III both inhibited the virus genome replication and N gene expression (Fig. 7G and H). These results suggested that eEF1A domain III could reduce the M protein level and has anti-SCRV functions.

eEF1A promotes M protein degradation through K48-linked ubiquitination

To further confirm if eEF1A promotes M protein degradation through the ubiquitin-proteasome pathway, EPC cells were treated with MG132 which is a commonly used proteasome inhibitor. After MG132 treatment, we found that inhibition of the proteasome pathway could relieve the decrease of M protein due to eEF1A overexpression (Fig. 8A), which was dose-dependent on MG132 (Fig. 8B). Then, we detected whether the M protein was degraded through ubiquitination. eEF1A-HA or control vector were co-transfected with M-Flag in HEK293T cells with the presence or absence of MG132. Immunoblot analysis of the immunoprecipitated protein complex with an anti-ubiquitin antibody revealed that eEF1A promoted the ubiquitination of M protein (Fig. 8C), and the ubiquitination was enhanced with the increase of eEF1A expression (Fig. 8D). Thus, eEF1A promotes M protein degradation through the ubiquitin-proteasome pathway.

Fig 8.

eEF1A promoted ubiquitin-K48 dependent ubiquitin-proteasome degradation of M protein. (A and B) The reduction of M protein level by eEF1A was relieved by MG132 treatment. Plasmid expressing Flag-tagged M was co-transfected with plasmid expressing HA-tagged eEF1A or control vector into EPC cells, and the cells were treated with MG132 (20 µM) (A). Alternatively, the co-transfected cells were treated with MG132 at different concentrations (10, 20, 30 µM) (B). Protein levels were analyzed by IB with indicated Abs. ImageJ was used to calculate the density of the indicated bands. (C) Ubiquitination smears on M protein were enhanced by overexpression of eEF1A under MG132 treatment. The HEK293T cells coexpressing Flag-tagged M and HA-tagged eEF1A or empty vector were treated with or without MG132 (20 µM). Cell lysates were IP with anti-Flag gels. Immunoprecipitates and whole cell lysates (Input) were analyzed by IB with indicated Abs. (D) eEF1A enhanced ubiquitination of M in a dose-dependent manner. Plasmid expressing HA-tagged M was co-transfected with different numbers of plasmids expressing Flag-tagged eEF1A into HEK293T cells and were treated with MG132 (20 µM). Then, cell lysates were IP with anti-HA gels and analyzed as described above. (E) Ubiquitination of M protein is ubiquitin-K48 dependent. The HEK293T cells overexpressing HA-tagged M protein were treated with MG132 (20 µM) and IP with anti-HA gels. Cell lysates and Input were analyzed with indicated Abs. (F) eEF1A enhanced the ubiquitin-K48-dependent ubiquitination of M protein. Plasmid expressing HA-tagged M was co-transfected with or without plasmid expressing Flag-tagged eEF1A. The cells were treated with MG132 (20 µM), then IP with anti-HA gels, and analyzed by IB with indicated Abs. (G) K9R mutation of M protein reduced its ubiquitin-K48 dependent ubiquitination. Plasmids expressing wild-type and different mutants of HA-tagged M protein were transfected into HEK293T cells, respectively. The cells were treated with MG132 (20 µM). Cell lysis and ubiquitination detection were performed as described above. (H) K9R mutation of M protein abolished its degradation caused by overexpression of eEF1A. Plasmids expressing HA-tagged M or M-K9R were co-transfected with or without Flag-tagged eEF1A into EPC cells. At 24 hpt, the cells were analyzed by IB with indicated Abs. (I) Overexpression of eEF1A-III promoted the ubiquitination of M protein. Plasmid transfection was performed in HEK293T cells as described above. The cells were treated with MG132 (20 µM) and analyzed by IP and IB with indicated Abs. (J) The promotion of the ubiquitination of M by eEF1A-III was dose-dependent. Plasmid expressing HA-tagged M was co-transfected with different amounts of plasmid expressing Flag-tagged eEF1A-III. MG132 (20 µM) treatment, IP, and IB were conducted as described above.

Ubiquitin binding to substrate relies on its lysine (K) site. K48- and K63-linked ubiquitination are cells’ most widely distributed ubiquitination types, accounting for more than 90% of ubiquitination types. Anti-Ub-K48 and anti-Ub-K63 antibodies were then used to detect the ubiquitination type in immunoprecipitated protein complexes. As shown in Fig. 8E, the anti-Ub-K48 antibody recognized specific bands in immunoprecipitated protein complexes from M-HA overexpressed HEK293T cells compared to mock and empty vector expressed cells. However, there were no bands with anti-Ub-K63 antibody. The bands recognized by the anti-Ub-K48 antibody enhanced when eEF1A overexpressed (Fig. 8F). Therefore, the type of ubiquitination presented on the M protein is ubiquitin-K48 dependent, which could be promoted by eEF1A expression.

Further, the K sites of the M protein that could be responsible for the ubiquitination modification were investigated. Bioinformatic analysis predicted five possible K sites (K5, K6, K9, K10, and K166) on M protein. Then, a series of M mutants (K5R, K6R, K9R, K10R, and K166R) were constructed in which the specific K was mutated to arginine (R). In the following ubiquitination detection assay with these mutants, the ubiquitination bands detected by anti-Ub and anti-Ub-K48 antibodies both were very weak in K9R overexpressed HEK293T cells compared to other mutants, especially the K6R mutant (Fig. 8G). However, immunoblot analysis of the input showed that the K9R mutation reduced the expression level of the mutant compared with the wild type. To exclude the effect caused by the low expression level, we coexpressed eEF1A with M protein or its K9R mutant in EPC cells. Immunoblot analysis showed that co-expression of eEF1A reduced the protein level of wild-type M protein but not the K9R mutant compared to the control (Fig. 8H), indicating that the K9R mutation abolished its degradation by eEF1A. Thus, the above results revealed that the K9 of M protein could be the site K48 ubiquitination modification occurs.

In addition, the above results have shown that eEF1A domain III (eEF1A-III) prompted the degradation of M protein (Fig. 7E). Further MG132 treatment also showed that the degradation of M protein by eEF1A-III was carried out via the ubiquitin-proteasome pathway and showed a dose-dependent with the expression of eEF1A-III (Fig. 8I and J). Thus, the ubiquitination degradation of M protein by eEF1A relied on its domain III.

eEF1A impairs the formation of N-P complexes and N-N oligomers

Above results showed that eEF1A inhibited the transcription and replication of SCRV genomes but did not directly degrade N protein, which suggests that eEF1A might inhibit the function of N protein during SCRV infection. Previous research has found that the N protein is involved in the transcription and replication of the rhabdovirus genome by forming a complex with the P protein (17). And the N protein itself also forms oligomers consisting of 2–10 monomers when it encases the viral genome (47). So, we wondered if eEF1A disrupted the interactions between N and P protein or N protein itself. First, the regions of N protein that could be involved in eEF1A-N, N-P, and N-N interactions were investigated. Based on the RNA-binding region, the N protein was divided into three parts, namely, the N-terminal part (N_1–200), the middle RNA-binding part (N_201–240), and the C-terminal part (N_241–430) (Fig. 9A). These truncated plasmids (Flag-tagged) were co-transfected with eEF1A-HA into EPC cells. Followed co-IP revealed that only the N protein’s C-terminal part (N241-430) interacted with eEF1A (Fig. 9B). The N_241-430 was further coexpressed with eEF1A-HA, P-HA, and N-HA, respectively. Co-IP showed that the C-terminal of N protein (N_241-430) not only interacted with eEF1A but also interacted with the P protein and itself (Fig. 9C). Thus, interactions between N and different proteins all occurred on the same region of N protein, which hinted that eEF1A has the potential to inhibit the interactions between N and other proteins.

Fig 9.

eEF1A inhibited interactions between N-P and N-N. (A) Schematic representation of N truncated mutants. (B) eEF1A interacted with the C-terminal 241–430 aa of N protein. Plasmids expressing Flag-tagged N terminal (N_1–200), central region (N_201–240), and C-terminal (N_241–430) of N protein were co-transfected with plasmid expressing HA-tagged eEF1A into EPC cells. Cell lysates were IP with anti-HA gels. Immunoprecipitates and whole cell lysates (Input) were analyzed by IB with indicated Abs. (C) C-terminal 241–430 aa of N protein interacted with P and N protein. Plasmid expressing Flag-tagged N_241–430 was co-transfected with plasmids expressing HA-tagged eEF1A, P, and N, respectively. IP and IB analyses were conducted as described above. (D) Overexpression of eEF1A did not affect the expression of P protein. Plasmid expressing HA or His-tagged eEF1A was co-transfected with plasmids expressing Flag-tagged P and HA-tagged N into EPC cells, respectively, or together. At 24 hpt, expression of the proteins was analyzed by IB with indicated Abs. ImageJ was used to calculate the density of the indicated bands. (E) Overexpression of eEF1A reduced the luciferase activity in N and P coexpressed group in Nanoluciferase complementation assays. (F) Overexpression of eEF1A reduced interactions between N and P by Co-IP analysis. Plasmid expressing His-tagged eEF1A was co-transfected with plasmids expressing Flag-tagged P and HA-tagged N into EPC cells. Cell lysates were IP with anti-HA or anti-Flag gels. IB analysis was performed as described above. (G) Overexpression of eEF1A reduced the luciferase activity in NlucN-N and N-NlucC coexpressed group in Nanoluciferase complementation assays. (H) Overexpression of eEF1A reduced interactions between N and N by Co-IP analysis. (I and J) Overexpression of eEF1A-III did not affect the interaction between N and P or N and N by Nanoluciferase complementation assays and co-IP analysis. (K and L) Overexpression of eEF1A-I + III did not affect the interaction between N and P or N and N by nanoluciferase complementation assays and co-IP analysis.

The above results have shown that eEF1A did not interact with P protein (Fig. 1H). Then, the effect of eEF1A on the stability of the P protein was investigated. eEF1A-HA or eEF1A-His were coexpressed with P-Flag and N-HA, respectively. Immunoblot analysis showed that the expression of P protein was not affected by eEF1A when coexpressed with or without N protein (Fig. 9D). Thus, eEF1A does not affect the stability of the P protein.

Next, two assays were used to verify the effect of eEF1A on the formation of the N-P complex. In the Nluc complementation assay, the luciferase activities were significantly reduced in NlucN-P + N-NlucC + eEF1A group (or NlucN-N + P-NlucC + eEF1A group) compared with NlucN-P + N-NlucC + Vector group (or NlucN-N + P-NlucC + Vector group) (Fig. 9E), indicating that eEF1A affected the interactions between N and P protein. Subsequently, we used anti-Flag Ab to detect the P-Flag protein in anti-HA Ab-immunoprecipitated N-HA protein complex in the presence or absence of eEF1A-His, and we also used anti-HA Ab to detect the N-HA protein in anti-Flag Ab-immunoprecipitated P-Flag protein complex (Fig. 9F). In both assays, the detected P protein in the immunoprecipitated N protein complex or N protein in the immunoprecipitated P protein complex all reduced in the presence of eEF1A compared to the control. So, the results indicated that eEF1A impaired the formation of the N-P complex.

Similarly, we used the two methods to verify if eEF1A inhibits the formation of N protein oligomers. The Nluc complementation assay showed that co-expression of NlucN-N and N-NlucC largely restored the luciferase activity, but the presence of eEF1A significantly reduced the luciferase activity (Fig. 9G). Moreover, the co-IP experiment also showed that eEF1A reduced N protein participation in N-N interactions (Fig. 9H). Thus, eEF1A can also weaken the formation of N oligomers.

Because eEF1A domain III can promote the degradation of M protein, we further investigated whether it could block the formation of N-related protein complexes. No significant differences existed in the nano-luciferase activity when eEF1A domain III (eEF1A-III) was coexpressed with N-P or N-N complexes compared with the control (Fig. 9I). And, the co-IP showed similar results (Fig. 9J), which revealed that eEF1A domain III alone could not inhibit the formation of the N-P complex and N oligomers. N-terminal of eEF1A (domain I) was responsible for the interaction with N protein (Fig. 3) and expression of eEF1A domain III, but not eEF1A domain I and II, reduced virus genome replication (Fig. 7E through H), which hinted a possibility that eEF1A domain I and III (eEF1A-I + III) could block the formation of N-related protein complexes. eEF1A-I + III was then coexpressed with N-P or N-N complexes. Followed Nluc complementation assays and co-IPs all showed that the eEF1A-I + III does not affect the formation of N-related protein complexes (Fig. 9K and L). Combined with the above results, we inferred that a complete eEF1A protein is necessary for targeting N protein but unnecessary for degrading M protein.

DISCUSSION

Although several host factors involved in the infection of mammalian rhabdoviruses have been reported, there were not many identified host factors that interacted with aquatic rhabdoviruses. In the present study, host eEF1A was found to interact with two proteins of a fish rhabdovirus SCRV and suppress virus infection via two mechanisms.

Rhabdoviruses commonly have five structural proteins, N, P, M, G, and L. The N protein forms oligomers tightly bound with the genomic RNA to form the N-RNA complex (9, 19). The L protein is an RNA-directed RNA polymerase that should associate with its cofactor P protein to perform its enzymatic activities (48, 49). The N-RNA and L-P form the ribonucleoprotein (RNP) complex in virions. In infected cells, active transcription and replication occur in the RNP complex in which the N-RNA is a template, and the interactions between N and P are essential for transcription and replication (17, 50). The present study showed that eEF1A inhibited the formation of the N-P complex and N oligomers, which could reduce the virus transcription. It could be an antiviral strategy that the host adopted by repressing the important viral protein complexes. Because interactions between host and viral proteins widely exist, such a strategy may be found in hosts of more viruses.

Another finding is the interactions between eEF1A and M protein. M protein covers the RNP in virions and interacts with G protein for the assembly and budding of viral particles (29, 51, 52). Ubiquitination of M protein has been reported to be involved in the budding process of rhabdovirus (53). Our study revealed that the ubiquitination of M protein mediated by eEF1A prompted its degradation and suppressed virus infection. Host protein-mediated ubiquitination and degradation of viral proteins has been found in other rhabdovirus proteins and can be considered a host anti-viral strategy. For example, tripartite motif protein 41 (TRIM41), a ubiquitin E3 ligase, was reported to mediate the ubiquitination and degradation of the N protein of VSV and, thus, limit virus infection (40). Zebrafish ceramide kinase-like can interact and degrade the P protein of SVCV to inhibit virus infection (54). Indeed, eEF1A has been shown to be involved in protein degradation through the proteasome pathway (55), but the promotion of degradation of viral proteins by eEF1A has not been reported. Rhabdovirus M is a multiple-function protein but not directly associated with transcription and replication. Our finding provided a new target for host-mediated ubiquitination and degradation of rhabdovirus proteins.

Interestingly, the interaction between eEF1A and M protein is not essential for the degradation of M protein. Although the interaction between eEF1A and M protein relies on the domain I and the full-length of eEF1A prompted the degradation of M protein, the domain III of eEF1A was confirmed can inhibit SCRV and prompt degradation of M alone. It seems like eEF1A-III induced the degradation of M protein through non-specific interactions or other uncharacterized mechanisms. Interaction between M protein and domain I of eEF1A may contribute to M degradation by increasing the contact between M protein and domain III of eEF1A. However, the roles of the eEF1A-M interactions need more research.

Structure analysis has shown that domain I of eEF1A comprised the Rossmann fold (6). The Rossmann fold has been found in several proteins with various functions and could bind nucleoside cofactors (56). The domain I of eEF1A binds to GTP to perform its role in the translation process. According to our results, SCRV N and M proteins both interact with the eEF1A domain I, which was consistent with the observation of the interactions between eEF1A and NS5A protein of CSFV. Thus, we speculated that the domain I of eEF1A was also a platform in interaction with other proteins.

As a translation factor, eEF1A is an abundant protein and an essential component of cell translation machinery (6). Besides its roles in translation, many non-canonical functions of eEF1A have been reported, including its participation in virus infection. There are two opposite effects for eEF1A in virus infection. One is the positive regulation of virus infection. For example, eEF1A is an HIV-1 reverse transcription complex (RTC) subunit and serves as RTC stability cofactors for efficient reverse transcription (57 – 59). eEF1A binds with the 3′-terminal stem-loop of West Nile virus genomic RNA to facilitate viral minus-strand RNA synthesis (44). eEF1A interacted with the N protein of RSV to facilitate genomic RNA synthesis (45). The other is the negative regulation of virus infection. It has been reported that eEF1A can bind to Turnip yellow mosaic virus RNA and repress the synthesis of the minus strand (60). And, eEF1A interacted with the NS5A protein of CSFV and negatively regulates virus growth, in which domain I of eEF1A is responsible for interacting with NS5A (46). Results in the present study revealed the negative regulatory roles of eEF1A in a fish rhabdovirus.

Present results also showed that the expression of eEF1A was upregulated by SCRV infection. Various host responses at different levels, including transcriptional, translational, metabolic, etc., would be induced by virus infection, which could facilitate or resist virus replication (61). Several genes have been reported to be upregulated during virus infection. Upregulating one part of the genes could modulate the cellular environment and benefit virus infection (62, 63). However, another part of the genes is upregulated to fight against virus invasion. For example, the interferon-induced genes and multiple TRIM (tripartite motif) genes have been reported to be induced by fish viruses and execute antivirus functions (64, 65). Results from the present study revealed that the eEF1A belongs to the upregulated gene with antiviral function.

In conclusion, the present study identified eEF1A as a protein that interacted with the N and M proteins of SCRV, which, in turn, inhibited SCRV infection via two different mechanisms (Fig. 10). Targeting N protein by eEF1A suppressed the formation of N-P complexes and N oligomers, which relied on the full length of eEF1A and inhibited the transcription and replication of the SCRV genome. Targeting M protein promoted K48-dependent ubiquitin-proteasome degradation of M, which needed the domain III of eEF1A. Collectively, our study revealed novel insights into the role of eEF1A in rhabdovirus infection and provided new information for antiviral research.

Fig 10.

Schematic diagram of the roles of eEF1A in SCRV infection. eEF1A interacted with N protein, resulting in reduced formation of N-P complex and N-N oligomer, inhibiting virus genome transcription and replication (indicated by red dashed line). On the other hand, eEF1A interacted with M protein, which promoted the K48-dependent ubiquitin-proteasome degradation of M protein.

MATERIALS AND METHODS

Cells and virus

The Epithelioma papulosum cyprini cell (EPC) and Fathead minnow cell (FHM) were grown in Medium 199 (Gibco) with 10% fetal bovine serum (FBS) and antibiotics (100 U/mL of penicillin and streptomycin). Siniperca chuatsi skin cell (SCSC) was grown in L-15 medium (Hyclone) with 10% FBS and antibiotics (100 U/mL of penicillin and streptomycin). HEK293T cells were grown in Dulbecco’s modified Eagle’s medium (DMEM, Gibco) containing 10% FBS. The three cell lines all were preserved in our lab and have been used in the previous study (66, 67). EPC and FHM were come from the fathead minnow (Pimephales promelas) and were sensitive to SCRV infection. EPC has higher transfection efficiency than the other two cells and was, thus, mainly used in gene expression in vitro in the study. SCSC was used in virus titer determination for its noticeable cytopathic effect during SCRV infection. SCRV (GenBank accession number NC_008514) was isolated from diseased mandarin fish and stored in our lab (15).

Plasmid construction and transfection

Total RNA of SCRV-infected FHM cells was extracted using FastPure Cell/Tissue Total RNA Isolation Kit V2 (Vazyme, China), and cDNA was synthesized by using HiScript III 1st Strand cDNA Synthesis Kit (+gDNA wiper) (Vazyme, China). The virus gene (N, P, and M) and eEF1A gene (XP_039536767.1) were amplified using the cDNA as a template. The N gene was cloned into pcDNA3.1, carries a strep tag, and was used in affinity purification. The virus gene (N, P, and M) and eEF1A gene were also cloned into pcDNA3.1-NlucN-p-F that carries N-terminal residues of Nanoluciferase (NlucN), polycloning sites (p), and Flag tag (F) and pcDNA3.1-p-NLucC-H that carries C-terminal residues of Nanoluciferase (NlucC), p, and HA tag (H). The two plasmids pcDNA3.1-NlucN-p-F and pcDNA 3.1-p-NLucC-H were constructed in our previous study (68), in which the NlucN contains 156 aa, and NlucC contains 13 aa (69). These cloning obtained plasmids pcDNA3.1-NlucN-eEF1A-Flag, pcDNA3.1-eEF1A-NlucC-HA, pcDNA3.1-NlucN-N-Flag, pcDNA3.1-N-NlucC-H, pcDNA3.1-NlucN-M-Flag, pcDNA3.1-M-NlucC-H, pcDNA3.1-NlucN-P-Flag, and pcDNA3.1-P-NlucC-H. In addition, eEF1A was also built into pcDNA3.1 containing His tag to obtain pcDNA3.1-eEF1A-His. The truncated mutant of eEF1A used in the experiment was modified from pcDNA3.1-NlucN-eEF1A-Flag or pcDNA3.1-eEF1A-His as the template. The truncated mutant of SCRV N was modified from pcDNA3.1-NlucN-N-Flag as the template. Lysine mutation of SCRV M was performed using pcDNA3.1-M-NlucC-H as a template. The ubiquitination modification sites on M protein were predicted online (70).

To amplify the promoter sequence of the eEF1A gene, the genomic DNA of FHM cells was extracted using TaKaRa MiniBEST Universal Genomic DNA Extraction Kit (TaKaRa, Japan) and used as the template. The promoter sequence region was amplified by PCR and constructed to pGL3-Basic plasmid (Promega, USA) to obtain the plasmid pGL-ppeEF1A, which can express firefly luciferase under the control of eEF1A promoter sequence. The primers used to construct these plasmids are listed in Table 2. All plasmids were constructed by In-Fusion strategies (71) and were proved by DNA sequencing.

TABLE 2.

Primers and constructed plasmids used in the study

Primers	Sequence (5′–3′)	Constructs
AP-MS
pcDNA3.1-N-His-Strep-F	GGCTAGCGTTTAAACTTAAGC ATG GAACACCAAATCATCAG	pcDNA3.1-N-strep
pcDNA3.1-N-His-Strep-R	GCTGGATATCTGCAGAATTCTCAGTGGTGGTGGTGGTGGTGACTTGCAGGAT CCCTTCTCAAACTGTGGATGACTCCA CAAAGCTTGGTGCTTCAGCA
Expression
LucN-eEF1A-FLAG-F	GAGGCTCTGGAGGGGAATTCATGGGAAAGGAAAAGATC	pcDNA3.1-NlucN-eEF1A-Flag
LucN-eEF1A-FLAG-R	TCATCCTTGTAGTCGGATCCCTTGGTCTTGGCAGC
LucN-SCRV-N-FLAG-F	GAGGCTCTGGAGGGGAATTCATGGAACACCAAATC	pcDNA3.1-NlucN-N-Flag
LucN-SCRV-N-FLAG-R	TCATCCTTGTAGTCGGATCCCAAAGCTTGGTGCTTCAG
LucN-SCRV-P-FLAG-F	GAGGCTCTGGAGGGGAATTCATGGCAAAACCAACT	pcDNA3.1-NlucN-P-Flag
LucN-SCRV-P-FLAG-R	TCATCCTTGTAGTCGGATCCCCGAATCACCTCGAG
LucN-SCRV-M-FLAG-F	GAGGCTCTGGAGGGGAATTCATGCCTCTGTTTAAG	pcDNA3.1-NlucN-M-Flag
LucN-SCRV-M-FLAG-R	TCATCCTTGTAGTCGGATCCATGCCAGCTATGACC
eEF1A-LucC-HA-F	GGGGCTCATCGGGGGAATTCATGGGAAAGGAAAAGATC	pcDNA3.1-eEF1A-NlucC-HA
eEF1A-LucC-HA-R	CACATCATAGGGGTAGGATCCCTTGGTCTTGGCAGC
SCRV-N-LucC-HA-F	GGGGCTCATCGGGG GAATTCATGGAACACCAAATC	pcDNA3.1-N-NlucC-HA
SCRV-N-LucC-HA-R	CACATCATAGGGGTAGGATCCCAAAGCTTGGTGCTTCAG
SCRV-P-LucC-HA-F	GGGGCTCATCGGGGGAATTCATGGCAAAACCAACT	pcDNA3.1-P-NlucC-HA
SCRV-P-LucC-HA-R	CACATCATAGGGGTAGGATCCCCGAATCACCTCGAG
SCRV-M-LucC-HA-F	GGGGCTCATCGGGGGAATTCATGCCTCTGTTTAAG	pcDNA3.1-M-NlucC-HA
SCRV-M-LucC-HA-R	CACATCATAGGGGTAGGATCCATGCCAGCTATGACC
SCRV-G-LucC-HA-F	GGGAGACCCAAGCTGGCTAGC ATGGGAATGTACCCACTGTTTGTTCCG	pcDNA3.1-G-NlucC-HA
SCRV-G-LucC-HA-R	GGAGCCGCTGTTTCCGGATCC GGGAACAAATTGATACTGCTGCAAAGGG
eEF1A-His-F	GATATCTGCAGAATTCCACCATGGGAAAGGAAAAGATCCAC	pcDNA3.1-eEF1A-His
eEF1A-His-R	CCTCTTCTGAGATGAGTTTTTGTTCCTTGGTCTTGGCAGCCTTCTGTGCA
eEF1A truncated mutant
eEF1A-D1-F	GAGGCTCTGGAGGGGAATTCATGGGAAAGGAAAAGATCCACA	eEF1A-I
eEF1A-D1-R	TCATCCTTGTAGTCGGATCCGGGCAAGATGGCATCCAGGGCA
eEF1A-D2-F	GAGGCTCTGGAGGG GAATTCCCAAGCCGTCCCACCGACAA	eEF1A-II
eEF1A-D2-R	TCATCCTTGTAGTCGGATCCTGGGTCGTTCTTGCTGTCTC
eEF1A-D3-F	GAGGCTCTGGAGGG GAATTCGCCCCAGTGCTGGACTGCCACACT	eEF1A-III
eEF1A-D3-R	TCATCCTTGTAGTCGGATCCCTTGGTCTTGGCAGCCTTCTGT
eEF1A-D1 + D2 F	GAGGCTCTGGAGGGGAATTCATGGGAAAGGAAAAGATCCACA	eEF1A-I + II
eEF1A-D1 + D2 R	TCATCCTTGTAGTCGGATCCTGGGTCGTTCTTGCTGTCTCCA
eEF1A-D2 + D3 F	GAGGCTCTGGAGGGGAATTCCCAAGCCGTCCCACCGACAA	eEF1A-II + III
eEF1A-D2 + D3 R	TCATCCTTGTAGTCGGATCCCTTGGTCTTGGCAGCCTTCTGT
eEF1A-D1 + D2△200aa-F	GAGGCTCTGGAGGGGAATTCATGCTGGAGGCCAGCTCAAA	eEF1A-I + IIΔ200
eEF1A-D1 + D2△200aa-R	TCATCCTTGTAGTCGGATCCTGGGTCGTTCTTGCTGTCTC
eEF1A-D1-200aa-F	GAGGCTCTGGAGGGGAATTCATGGGAAAGGAAAAGATCCACA	eEF1A-I200
eEF1A-D1-200aa-R	TCATCCTTGTAGTCGGATCCGTTGTCTCCATGCCATCCAG
eEF1A-D1 + D3-His-F	GATATCTGCAGAATTCCACC ATG GGAAAGGAAAAGATCCAC	eEF1A-I + III
eEF1A-D1 + D3-His-R	CCTCTTCTGAGATGAGTTTTTGTTC CTTGGTCTTGGCAGCCTTCT
N truncated mutant
SCRV-N-1-F	GAGGCTCTGGAGGG GAATTC ATGGAACACCAAATCATCAGGAGAG	N1–200
SCRV-N-1-R	TCATCCTTGTAGTCGGATCC AATTGCTGCCACCATTCTGAGA
SCRV-N-2-F	GAGGCTCTGGAGGG GAATTC GACATGTTCTTTTACAGATTCAAGG	N201–240
SCRV-N-2-R	TCATCCTTGTAGTCGGATCC AGTGAATGATGCGACATGAGCA
SCRV-N-3-F	GAGGCTCTGGAGGG GAATTC GGATTAACCCTCAGTGGAGTGCTGG	N241–430
SCRV-N-3-R	TCATCCTTGTAGTCGGATCC CAAAGCTTGGTGCTTCAGCAGATC
M lysine point mutant
M-K5R-F	CTCTGTTTCGAAAGAGCAACAAGAAGTCGACTATTAAG	M-K5R
M-K5R-R	GCTCTTTCGAAACAGAGGCATGCTAGCCAGCT
M-K6R-F	CTCTGTTTAAGCGAAGCAACAAGAAGTCGACTATTAAGCC	M-K6R
M-K6R-R	GCTTCGCTTAAACAGAGGCATGCTAGCCAGCT
M-K9R-F	GAGCAACCGAAAGTCGACTATTAAGCCATATCAAGC	M-K9R
M-K9R-R	TCGACTTTCGGTTGCTCTTCTTAAACAGAGGCATG
M-K10R-F	CAACAAGCGATCGACTATTAAGCCATATCAAGCAC	M-K10R
M-K10R-R	TAGTCGATCGCTTGTTGCTCTTCTTAAACAGAGGC
M-K166R-F	TTTCGAGCGAATGTTCCACGGTGTGTTAGTGCA	M-K166R
M-K166R-R	GGAACATTCGCTCGAAAGATGATCCTGTCATCC
Dual-Luciferase reporter
pGL-ppeEF1A-F	TTTCTCTATCGATAGGTACCTGTATCACGATTTCCACAAACATATT	pGL-ppeEF1A
pGL-ppeEF1A-R	TTGGCGTCTTCCATGGTGGCTTGATTAGTTTCTGTACAAGGAAGGAA
Vectors linearization
pcDNA3.1-His-strep-F	GCTTAAGTTTAAACGCTAGCC
pcDNA3.1-His-strep-R	GAATTCTGCAGATATCCAGC
NlucN-Flag-F	GGATCCGACTACAAGGATGA
NlucN-Flag-R	GAATTCCCCTCCAGAGCCTC
NlucC-HA-F	GCTAGCCAGCTTGGGTCTCCC
NlucC-HA-R	GGATCCGGAAACAGCGGCTCC
pcDNA3.1-His-F	GGTGGAATTCTGCAGATATC
pcDNA3.1-His-R	GAACAAAAACTCATCTCAGAAGAGG
pGL-F	GCCACCATGGAAGACGCCAA
pGL-R	GGTACCTATCGATAGAGAAA
RT-qPCR
FHM-actin-F	GAATCCCAAAGCCAACAG
FHM-actin-R	GGAAGAGCATAACCCTCATAG
eEF1A-F	CAAGGAAGTCAGCGCCTACA
eEF1A-R	CATCCCTTGAACCAGCCCAT
SCRV-qRT-F	GGGCTGGATGATAGACGATTG
SCRV-qRT-R	TGGCGGAGGTGCTTGATATGG

Plasmid transfections were performed using Lipofectamine 3000 (Thermo Scientific, USA) according to the manufacturer’s protocol. The amount of plasmid DNA used in transfection was 2.5 µg per well in the 6-well plate, 1 µg per well in the 12-well plate, and 0.5 µg per well in the 24-well plate. For co-transfection, the amount of each plasmid was reduced to make the total quantity meet the above rules.

Antibodies

The anti-N protein rabbit antibody was prepared by Abclonal Technology Co., Ltd. (Wuhan, China) and was stored in our lab. The anti-eEF1A rabbit antibody (A17857), anti-β-actin rabbit antibody (AC026), anti-UBC rabbit antibody (A3207), anti-HA mouse antibody (AE008), and horseradish peroxidase (HRP)-conjugated Goat Anti-Rabbit IgG (H + L) (AS003) were purchased from Abclonal (Wuhan, China). The anti-HA rabbit antibody (3724S), anti-His rabbit antibody (12698S), and the K48-linkage Specific (8081S) and K63-linkage Specific (5621S) Ub rabbit antibodies were purchased from Cell Signaling Technology (USA). The anti-Flag rabbit antibody (SAB4301135) was purchased from Sigma-Aldrich (USA).

Affinity purification and mass spectrum identification

To screen the proteins that could interact with SCRV N protein, pcDNA3.1-N-Strep was transfected into FHM cells. An empty vector was used as a control. At 36 hour post-transfection (hpt), the cells were collected and lysed with Radio Immunoprecipitation Assay (RIPA) Lysis buffer (Beyotime, China). The cell lysates were centrifuged at 5,000 rpm for 5 min at 4°C, and the supernatants were incubated with anti-Strep affinity beads (BEAVER, China) and stayed overnight. Then, the beads were washed with Wash buffer (10 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, pH 8.0) five times. Finally, the proteins adsorbed on the beads were eluted with Wash buffer pre-added with 2.5 mM desthiobiotin. The eluted proteins were stored at −80°C for the next experiment.

The eluted proteins were processed with standard mass spectrometry protocols, which were performed at the analysis and testing center of the Institute of Hydrobiology, Chinese Academy of Sciences, as described previously (68). The coding proteins from the genome of Pimephales promelas (GCA_016745375.1) were used as the search database.

Nanoluciferase complementation assay

Nanoluciferase complementation assay was performed as previously described (68). EPC cells were pre-cultured in 12-well plates. Five hundred nanograms of the plasmids carrying NlucN or NlucC fused genes was co-transfected into the cells. Vectors expressing NlucN or NlucC alone were used as control. The cells were collected 24 hpt and resuspended with PBS. Equal volumes of Nano-Glo Luciferase Assay Reagent (Promega, USA) were added. Luminescence was detected and recorded by using a SpectraMax i3x Multi-Mode Microplate Reader (MD, USA). All complementation assays were performed in triplicate.

Co-IP assay and immunoblot analysis

To investigate protein interactions, the co-IP assays were performed in EPC cells as described in the previous study (68). EPC cells were pre-cultured in 6-well plates and transfected with 2.5 µg plasmids (1.25 µg each for two plasmids groups and 0. 83 µg each for three plasmids groups). The plasmid-transfected cells were collected at 36 hpt, and RIPA buffer containing PMSF and protease inhibitor (MCE, China) was added. The cell lysates were centrifuged at 12,000 g for 5 min at 4°C, and the supernatants were incubated with EZview Red Anti-HA Affinity Gel (Sigma-Aldrich, USA) or anti-FLAG M2 affinity gel (Sigma-Aldrich, USA). Precipitates were washed with ice-cold PBST buffer five times and subjected to Immunoblot analysis.

For protein ubiquitination assays, the co-IP and IB assays were performed in HEK293T cells. The cells were transfected with plasmids and treated with MG132(MCE, China) at 28 hpt. The cells were collected at 8 h after MG132 treatment and were analyzed as above.

Protein samples were separated by 4%–20% SDS-PAGE and transferred to the PVDF membrane (Millipore). After blocking with 5% skimmed milk in Tris-buffered saline Tween buffer, the primary antibody was added and incubated overnight at 4°C. Then, the membrane was incubated with HRP-conjugated secondary antibody at a ratio of 1:5,000 for 1 h at room temperature and was detected by chemiluminescence (Millipore, USA). ImageJ was used to calculate the density of protein bands in IP or IB analysis.

Immunofluorescence

EPC cells were grown on coverslips in a 12-wells plate before the transfection. Then, plasmids expression Flag-tagged eEF1A, HA-tagged N, HA-tagged eEF1A, or Flag-tagged M were co-transfected or alone into the cells. At 24 hpt, cells were fixed with 4% paraformaldehyde (Biosharp, China) for 20 min at room temperature and then permeabilized with 0.5% Triton X-100 (BioRoYee, China) for 20 min. Then, the cells were blocked with 3% BSA and incubated with a primary antibody (1:100) diluted with PBS containing 3% BSA for 2 h. After primary antibody incubation, cells were rinsed with PBS for three times. And the cells were incubated with Alexa Flour 488 goat anti-mouse IgG(H + L) or Alexa Flour 546 goat anti-rabbit IgG(H + L) (Invitrogen, USA) at a ratio of 1:1,000 for 1 h. Finally, the cells were stained with 4′,6-diamidino-2-phenylindole (Biosharp, Beijing) for 10 min. The fluorescent images were observed by a Leica TCS SP8 confocal microscope.

RT-qPCR for gene expression

Total RNA was extracted from SCRV-infected cells as described above. cDNA was synthesized by using HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, China). qPCR was conducted with iTaq Universal SYBR Green Supermix (Bio-Rad, USA) by using the cDNA as a template, as described previously (66). The β-actin gene was used as internal control, and the relative expression ratios were calculated by the 2^−ΔΔCT method. The primers used in the experiment are collected in Table 2.

RT-qPCR for virus genome copy number

After SCRV infection, the cell medium of the infected cells or the whole cells was collected, and viral genomic RNA was extracted using Viral RNA Isolation Kit (FOREGENE, China). cDNA was synthesized by using HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, China). Absolute Quantitative PCR was used to detect virus genome copy numbers according to the previous method (72). The plasmid pcDNA3.1-SCRV, constructed in our lab in another study, contained the full length of the SCRV genome and was used to construct a standard curve for calculating virus genome copy numbers. The primers and reagents used to detect the SCRV genome were the same as those used in RT-qPCR.

Dual-luciferase reporter assay

The firefly luciferase plasmid pGL-ppeEF1A and the Renilla luciferase plasmid pRL-TK were used in this experiment. FHM cells were pre-cultured in 24-well plates and transfected with a mixture of 450 ng of pGL-ppeEF1A and 50 ng of pRL-TK. The cells were infected by SCRV at 1 MOI at 6 hpt. The cells transfected with the same plasmids but not infected with SCRV were used as a control. Alternatively, the cells were transfected with 250 ng of pGL-ppeEF1A, 50 ng of pRL-TK, and 200 ng of SCRV protein expression plasmid (pcDNA3.1-NlucN-N-Flag, pcDNA3.1-NlucN-P-Flag, pcDNA3.1-NlucN-M-Flag, or empty vector). At 24 hpi or 24 hpt, luciferase activity was detected by using Dual-Luciferase Reporter Assay System (Promega, USA) according to the protocols. Renilla luciferase was used as an internal parameter to measure the change in firefly luciferase activity. The assays were performed in triplicate.

RNA interference

The three siRNAs targeting Pimephales promelas eEF1A and the control (siNC) were designed and synthesized (Tsingke Biotech, China). The siRNA sequences are collected in Table 3. FHM cells were pre-cultured into 24-well or 6-well plates, and the cell density was maintained at about 50% before transfection. The siRNAs were transfected into the cells at a final concentration of 100 nM, respectively, using Lipofectamine RNAiMAX (Thermo Scientific, USA) according to the manufacturer’s protocol. Then, the cells were collected at 24 hpt to perform RT-qPCR and IB analysis. Alternatively, the cells were infected with SCRV (0.5 MOI) and collected at indicated time points. RT-qPCR and IB analysis was performed as described above.

TABLE 3.

The siRNA sequences used in the study

SiRNA name	Sequence (5'−3')
sieEF1A-1	GGAAGUUCGAGACCAGCAA
sieEF1A-2	GCAAGUUUGCUGAGCUCAA
sieEF1A-3	GAUGGAAGGUUGAGCGUAA
siNC	UUCUCCGAACGUGUCACGUTT

Virus titer and cytopathic effect determination

SCSC cells were used to determine the SCRV titer and cytopathic effect. For virus titer determination, SCSC was pre-cultured into 96-well plates and infected with collected SCRV samples at 10-fold gradient dilutions. The virus titers were calculated using the 50% tissue culture infectious dose (TCID₅₀) method. For cytopathic effect observation, SCSC was subcultured into 12-well plates and infected with SCRV. When apparent cytopathic effects were observed, the medium was removed, the cells were fixed with a formaldehyde aqueous solution and then stained with crystal violet.

Statistical analysis

All statistical analyses were carried out using GraphPad Prism 8. The significance of the difference between groups was analyzed with the student’s t-test. Data are presented as the mean ± standard deviations from three independent experiments. A P-value of <0.05 was considered statistically significant.