Zinc Finger Protein BCL11A Contributes to the Abortive Infection of Hirame novirhabdovirus (HIRRV) in B Lymphocytes of Flounder (Paralichthys olivaceus)

ABSTRACT

Hirame novirhabdovirus (HIRRV) infection is characterized by a pronounced viremia, and the high viral load is typically detected in immune-related organs and the circulatory system. In the present study, we demonstrated that HIRRV has the capacity to invade part of flounder membrane-bound IgM (mIgM⁺) B lymphocyte. Eight quantitative real-time PCR (qRT-PCR) standard curves involving HIRRV genomic RNA (gRNA), cRNA, and six mRNAs were established based on the strand-specific reverse transcription performed with tagged primers. It was revealed that viral RNA synthesis, especially the replication of gRNA, was inhibited in B cells, and the intracellular HIRRV even failed to produce infectious viral particles. Moreover, a range of genes with nucleic acid binding activity or related to viral infection were screened out based on the transcriptome analysis of HIRRV-infected B cells, and five molecules were further selected because of their different expression patterns in HIRRV-infected B cells and hirame natural embryo (HINAE) cells. The overexpression of these genes followed by HIRRV infection and RNA binding protein immunoprecipitation (RIP) assay revealed that the flounder B cell lymphoma/leukemia 11A (BCL11A), a highly conserved zinc finger transcription factor, is able to inhibit the proliferation of HIRRV by binding with full-length viral RNA mainly via its zinc finger domains at the C terminus. In conclusion, these data indicated that the high transcriptional activity of BCL11A in flounder mIgM⁺ B lymphocytes is a crucial factor for the abortive infection of HIRRV, and our findings provide new insights into the interaction between HIRRV and teleost B cells.

IMPORTANCE HIRRV is a fish rhabdovirus that is considered as an important pathogen threatening the fish farming industry represented by flounder because of its high infectivity and fatality rate. To date, research toward understanding the complex pathogenic mechanism of HIRRV is still in its infancy and faces many challenges. Exploration of the relationship between HIRRV and its target cells is interesting and necessary. Here, we revealed that flounder mIgM⁺ B cells are capable of suppressing viral RNA synthesis and result in an unproductive infection of HIRRV. In addition, our results demonstrated that zinc finger protein BCL11A, a transcription factor in B cells, is able to suppress the replication of HIRRV. These findings increased our understanding of the underlying characteristics of HIRRV infection and revealed a novel antiviral mechanism against HIRRV based on the host restriction factor in teleost B cells, which sheds new light on the research into HIRRV control.

INTRODUCTION

B lymphocytes are an essential component of the vertebrate immune system, which can effectively process and present the antigens to CD4⁺ helper T cells and differentiate into antibody-secreting cells (ASCs) when it is activated by antigens (1). mIgM⁺ B cells represent the majority of B lymphocytes in teleost fish and have more diversified functions. Available studies have demonstrated that fish mIgM⁺ B cells possess phagocytic functions similar to those of macrophages (2–4), and they also can present both soluble and particulate antigens to specific T cells (5). Notably, it was reported that infection with viral hemorrhagic septicemia virus (VHSV) significantly increased the expression of major histocompatibility class II (MHC II) molecules on the surface of mIgM⁺ B cells in rainbow trout, which indicated that fish mIgM⁺ B cells may play a role in viral antigen presentation (6). Most of the previous studies mainly focused on the T cell-mediated response against viral infection because of the important role of T cells in the clearance of intracellular pathogens (7–9). However, there is still a lack of attention given to and understanding of the relationship between B cells and viruses, especially in lower vertebrates, because of the diversity of their B cell functions.

Hirame novirhabdovirus (HIRRV), a member of the Novirhabdovirus genus of the family Rhabdoviridae, is a nonsegmented, single-stranded, negative-sense RNA virus (10) which commonly causes large-scale mortalities of host fish at a low water temperature (~10°C), and at a high water temperature (~20°C) appears to retard or halt the disease progression (11–13). The HIRRV genomic RNA (gRNA) encodes five structural viral proteins (N, P, M, G, L) and a nonstructural protein (NV) (10). In the late viral replication stage, positive-strand RNA (cRNA) which is complementary to the gRNA is produced as the template for the gRNA synthesis that is used for the assembly of the virion (14). Our previous research showed that high viral loads were detected in the organs with rich blood flow, such as the heart, spleen, head kidney, and peripheral blood. More importantly, the results of immunofluorescence and immunohistochemical staining showed substantial HIRRV-positive signals in leukocytes and vascular endothelial cells (ECs) (15). Similar findings have been reported in equine herpesvirus 1 (EHV1), which can replicate in the respiratory epithelium and disseminates through the body via a cell-associated viremia in leukocytes, and the hijacked leukocytes mainly include monocytic cells and T lymphocytes (16). However, it is unclear whether HIRRV causes systemic infection through similar mechanisms. Accordingly, which leukocytes can be infected by HIRRV inevitably became an interesting problem that remains to be addressed. According to the existing research, various types of immune cells confront the invasion of different viruses. Beyond EHV1, mentioned above, multiple viruses can infect T lymphocytes and phagocytes (macrophages or dendritic cells), such as HIV-1, dengue virus (DENV), Marek’s disease virus (MDV), etc. (17–19). In mammals and poultry, B cells can also be invaded by a variety of viruses, including but not limited to Epstein-Barr virus (EBV), mouse noroviruses (MuNoV), Porcine circovirus 2 (PCV2), and infectious bursal disease virus (IDBV) (20–23). In addition, rabies virus (RABV) and viral hemorrhagic septicemia virus (VHSV), which are close relatives of HIRRV, have been proved to be able to infect mouse B cells and rainbow trout IgM⁺ B cells, respectively (6, 24). On the other hand, our previous research has found that the percentage of mIgM⁺ B lymphocytes was significantly reduced when the flounder was infected by HIRRV after 2 weeks (25). This research aroused our curiosity about the question of whether flounder IgM⁺ B cells act as the target cells during the infection of HIRRV and contribute to virus spread within the host.

The interaction of virus and host cells is highly complex. On the one hand, viruses will hijack host cellular machinery to facilitate their replication. On the other hand, host cells are equipped with multiple mechanisms to block the viral infection progression (26). The innate immune system serves as the first line of defense against invading viruses. Apart from some classic antiviral pathways such as a series of signaling cascades triggered via specific pattern recognition receptors (PRRs), host restriction factors that were first described in research on HIV-1 are also an important arm of the host innate immune system (27). Currently, numerous host restriction factors have been identified, including the APOBEC (apolipoprotein B mRNA editing catalytic polypeptide-like) family, zinc finger antiviral proteins (ZAP), tripartite motif proteins (TRIMs), etc. (28–30). They can inhibit viral proliferation by preventing viral entry into target cells and blocking replication and transcription of the viral genome, viral protein synthesis, viral assembly, and release. More importantly, most host restriction factors were confirmed to exhibit broad-spectrum antiviral effects (27). Considering that B lymphocytes are an important immune component, we wondered what will happen within cells if they are infected by HIRRV.

In the current study, we attempt to address the two outstanding issues mentioned above. For this purpose, we investigated the capacity of HIRRV to infect flounder mIgM⁺ B cells and analyzed the proportion of HIRRV-infected B cells; the viral gene transcription, genome replication, and viral particle release within B cells were further monitored. Moreover, the response of mIgM⁺ B cells to HIRRV infection was analyzed based on RNA sequencing (RNA-Seq) data, and 5 genes were screened out according to their different expression patterns in B lymphocytes and hirame natural embryo (HINAE) cells after HIRRV infection. It was found that the overexpression of B cell lymphoma/leukemia 11A (BCL11A) inhibits HIRRV proliferation within HINAE cells by binding to the nucleic acid of HIRRV. The high transcriptional activity of BCL11A in flounder mIgM⁺ B lymphocytes could be a principal mechanism limiting the proliferation of HIRRV. In summary, these findings contribute to our understanding of the interaction between HIRRV and fish B lymphocytes and provide critical information regarding HIRRV pathogenesis and a novel antiviral mechanism of B lymphocytes.

RESULTS

HIRRV invades flounder B lymphocytes.

In order to explore whether HIRRV can infect flounder B cells, the lymphocytes were isolated and incubated with live HIRRV or inactivated HIRRV (multiplicity of infection [MOI], 5) for 24 h in vitro. The inactivated HIRRV was included as a negative control to test nonspecific interactions. Indirect immunofluorescence revealed that the fluorescence signal of HIRRV and mIgM can be colocalized in the same cell whether the lymphocytes are incubated with live or inactivated HIRRV (Fig. 1A). Interestingly, the fluorescence signal density of inactivated HIRRV within mIgM⁺ B cells was markedly lower than that of live HIRRV. To further clarify whether HIRRV entered B lymphocytes, a highly pure population (>95%) of mIgM⁺ B cells was sorted by fluorescence-activated cell sorting (FACS) from the peripheral blood lymphocytes (PBLs) after HIRRV infection (Fig. 1B). The result of nested PCR showed that viral RNA (gene g) can be detected in both B cells incubated with live or inactivated HIRRV, and the PCR band of inactivated HIRRV was significantly weaker than that of live HIRRV (Fig. 1C). In addition, the protein G was only detected in the B cells incubated with live HIRRV (Fig. 1D).

FIG 1.

HIRRV invasion of flounder B lymphocytes. (A) Indirect immunofluorescence microscopy of HIRRV G protein in the flounder PBLs exposed to live or heat-inactivated (mock) HIRRV for 24 h; mIgM and G protein are represented by the green and red fluorescence, respectively. (B) Representative mIgM+ B cell ratios before (12.4%) and after (98.6%) flow cytometric sorting. (C) Nested PCR amplification of the HIRRV g gene from the B cells exposed to live or heat-inactivated HIRRV for 24 h. (D) Detection of HIRRV G protein by Western blot analysis in the B cells exposed to live or heat-inactivated HIRRV for 24 h. GAPDH was used as a loading control.

Invasion and proliferation trend of HIRRV in B lymphocytes.

When it became clear that HIRRV could enter the B cell, the following experiments further explored the invasion and proliferation trend of HIRRV in B lymphocytes. For this purpose, the PBLs were incubated with live or inactivated HIRRV for different amounts of time (0 to 48 h), the invasion dynamic was evaluated by flow cytometry (FCM), and the viral proliferation dynamic was monitored by qRT-PCR in the sorted mIgM⁺ B cells. Flounder were infected in vivo by HIRRV for 48 and 96 h at 10°C or 20°C, and the viral copies in B cells were detected by qRT-PCR. As indicated in Fig. 2A, a small proportion of the mIgM⁺ B cells were infected by HIRRV; the HIRRV-infected B cells (highlighted in yellow scatter) account for a small percentage (~2 to 6%) of peripheral blood leukocytes over time. Meanwhile, a very small number of HIRRV⁺ B cells were detected when the PBLs were continuously exposed to inactivated HIRRV. Consistent with the result of the immunofluorescence assay, FCM analysis showed that the HIRRV mean fluorescence intensity (MFI) in B cells infected by the live virus was significantly higher than that of cells incubated with inactivated HIRRV (Fig. 2B). It was obvious that the percentage of virus-positive B cells infected with live HIRRV was significantly higher than that of those inoculated with inactivated virus, which was significantly increased within 24 h and maintained relatively stable afterward, and the peak percentage of virus-positive B cells out of total B cells was about 20% (Fig. 2C). Similarly, the viral loads in the infected B cells were significantly higher than that of mock-infected cells, which was significantly increased within 12 h and remained stable afterward with a relatively low viral load of 10³ to 10⁴ copies per 100 ng total RNA (Fig. 2D). Notably, the portion of HIRRV-infected B cells and their intracellular viral copy numbers remained relatively stable after a slight increase within 24 h, indicating that HIRRV cannot efficiently and continuously proliferate inside B cells. In the in vivo infection experiment, viral RNA could also be amplified in B cells, but no continuous increases of viral copies were observed in B cells at 10°C or 20°C (Fig. 2E).

FIG 2.

A small portion of flounder B lymphocytes susceptible to the HIRRV infection. (A) Representative flow cytometric plots show the invasion trend of HIRRV to flounder B cells; HIRRV-positive mIgM+ B cells (Q2-UR) were labeled by double fluorescent staining (Alexa Fluor 488/647). (B) Statistical quantification of the mean fluorescence intensity (MFI), which represents virus abundance, within HIRRV-positive mIgM+ B cells (48 hpi, n = 5). (C) Proportions of HIRRV-positive B cells in total B cells post-HIRRV infection (n = 5). (D) The change of viral copy numbers post-HIRRV infection (n = 3). (E) Quantification of HIRRV copies within B cells in the in vivo infection experiment at 10°C and 20°C (n = 5). Different letters above the bars represent significant differences among the groups at different times (P < 0.05). Significant differences compared with mock infection group are indicated by asterisks (*, P < 0.05; **, P < 0.01; ***, P < 0.001); ns, nonsignificant.

Proliferation of HIRRV was repressed in B lymphocytes.

Both PBLs and HINAE cells were incubated with HIRRV (MOI, 5) for different amounts of time (6, 12, 24, 48, and 72 h), and then B lymphocytes were isolated by FACS. For negative-sense RNA viruses, conventional quantitative PCR cannot distinguish viral genomic RNA (gRNA), cRNA, or mRNA, but the presence of HIRRV mRNA and cRNA are important evidence of successful viral entry and initiation of the viral transcription-replication process in the invading cells. To explore this, reverse transcription of three types of HIRRV RNA was performed using different strand-specific primers containing a unique sequence tag at the 5′ end, and different viral RNAs were determined by absolute qRT-PCR based on the sequence tag (Fig. 3A). It was exciting to find that all viral RNA types were detectable in B lymphocytes, although the abundances were dramatically lower than that in HINAE cells. All the viral mRNAs transcribed in B cells were similar to their transcriptional signature in HINAE cells; the mRNA copy number of the L gene was obviously lower than that of the other genes (Fig. 3E). The cRNA copies of HIRRV showed a continuous increase within 72 h postinfection (hpi) in the infected B cells (Fig. 3B). However, it was found that the gRNA copies of HIRRV exhibited a significant increase within 24 h and remained relatively stable afterward (Fig. 3C). Based on these data, the ratio of viral cRNA to gRNA in B cells was calculated to be significantly higher than that in HINAE cells at 48 and 72 hpi, which even exceeded 1 after 48 hpi (Fig. 3D). We wondered if HIRRV can complete the entire infection process to release the progeny virus particles from B cells. To answer this question, HIRRV-infected B cells (24 hpi) were cocultured with HINAE cells for 72 h. To our surprise, no evidence of cytopathic effect (CPE) was observed in the HINAE cells (Fig. 3F).

FIG 3.

Abortive infection of flounder B Lymphocyte by HIRRV. (A) Schematic diagram of the strand-specific qRT-PCR with tagged primers. The tagged cDNA is amplified using the tag sequence as the forward primer and a segment-specific oligo nucleotide as the reverse primer. (B and C) Copy number (log₁₀) of HIRRV (B) cRNA and (C) gRNA in 100 ng total RNA from HIRRV-infected B cells or HINAE cells (n = 3). (D) Ratios of viral cRNA to gRNA in HIRRV-infected B cells or HINAE cells at different time points (n = 3). (E) Copy number (log₁₀) of six HIRRV mRNAs in 100 ng total RNA from HIRRV-infected B cells or HINAE cells at 48 h postinfection (n = 3). (F) Coculture of HINAE cells and HIRRV-infected B cells. The B cells incubated by inactivated HIRRV were used as a negative control; no CPE was observed within 72 h. Different letters above the bars represent significant differences among the groups at different times (P < 0.05). Significant differences between two groups at each time point are indicated by asterisks (***, P < 0.001).

Screening of potential antiviral genes in B lymphocytes.

We were curious about why HIRRV cannot efficiently replicate and proliferate in B cells. Considering the significant differences in the pathogenicity of HIRRV under different temperatures, we sought to obtain some clues about the antiviral response of B cells by analyzing the gene expression profiles of B cells from HIRRV-infected flounder at 10°C and 20°C (NCBI accession numbers SRR13300097 to SRR13300103, SRR13300107 to SRR13300109, SRR13300112, and SRR13300113). The result showed that more than 200 genes were coupregulated in B cells when the fish was infected by HIRRV at different temperatures (Fig. 4A). Through an extensive literature search and Pfam database screening, a total of 26 genes related to viral infection or containing a nucleic acid binding domain were selected from these coupregulated genes (Fig. 4B). Thereafter, qRT-PCR was performed to analyze the expression patterns of these genes in HINAE cells and B cells after stimulation with HIRRV or poly(I:C). As we expected, five genes (anxa6, zhx2, dhx35, bcl11a, rfx1) exhibited different expression patterns between HIRRV high-sensitive and low-sensitive cells after viral infection, which were only upregulated in the HIRRV-infected B cells. Among them, it is to be noted that bcl11a and rfx1 were constitutively highly expressed in B cells (Fig. 4C). In contrast, the result of qRT-PCR showed that stimulation with poly(I:C) only significantly upregulated the expression of zhx2 in B cells (Fig. 4D).

FIG 4.

Screening of antiviral candidate genes from flounder B lymphocytes. (A) Venn diagram of upregulated genes in HIRRV-infected B cells at 10°C and 20°C; the intersection contains 234 genes. (B) Heatmap of the expression levels of selected genes with nucleic acid binding ability or associated with viral infection. (C) The expression differences of anxa6, zhx2, dhx35, bcl11a, and rfx1 between HIRRV-infected B cells and HINAE cells (n = 3). (D) The expression differences of anxa6, zhx2, dhx35, bcl11a, and rfx1 between flounder B cells and HINAE cells stimulated with poly(I·C) (n = 3). Significant differences are indicated by asterisks (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

Functional validations of potential antiviral genes.

In order to determine whether these genes are capable of inhibiting HIRRV replication, five expression plasmids were constructed using the eukaryotic expression vector pCl-eGFP (pANXA6, pZHX2, pBCL11A, pRFX1, pDHX35), and one empty plasmid (pEGFP) was used as the negative control. After transfection, all the constructed plasmids were successfully expressed in HINAE cells (Fig. 5A), and their average transfection efficiencies were approximately 15 to 20%, without significant difference (Fig. 5B). Subsequently, the HINAE cells overexpressing different genes were infected with low and high doses of HIRRV (MOI, 0.1 and 1, respectively). Viral plaques were identified by staining with crystal violet, and cell viability was analyzed by CCK8 assay. The result showed that the HINAE cells overexpressing bcl11a produced significantly fewer viral plaques than the cells overexpressing other genes at a low MOI, but no significant differences of plaques were observed among different groups at a high MOI (Fig. 5C). The CCK8 analysis also showed that only HINAE cells overexpressing bcl11a exhibited significantly higher cell viability under the low MOI (0.1) compared with other groups (Fig. 5D). As determined by qRT-PCR, the cRNA copy number in bcl11a-overexpressed cells was significantly lower than that of the control and dhx35 overexpression groups (Fig. 5E), and overexpression of bcl11a can also significantly reduce the gRNA level of HIRRV compared with all the other groups (Fig. 5F).

FIG 5.

Inhibiting activity of flounder BCL11A to the replication and proliferation of HIRRV. (A) Fluorescence microscopy of transfected HINAE cells overexpressing five antiviral candidate genes. (B) Flow cytometric quantification of transfection efficiency of HINAE cells at 48 h (n = 3). (C) Crystal violet staining for virus plaques in HINAE cells that overexpressed different genes at 72 hpi at an MOI of 0.1 or 1. (D) CCK8 assay of viability of HINAE cells overexpressing different genes at 72 hpi at an MOI of 0.1. (E and F) Copy number (log₁₀) of viral cRNA and gRNA in the HINAE cells that overexpressed different genes at 72 hpi at an MOI of 0.1 (n = 5). Different letters above the bars represent significant differences among different transfection groups (P < 0.05). Significant differences compared with the mock transfection group are indicated by asterisks (**, P < 0.01).

Sequence and structural characteristics of flounder BCL11A protein.

The motifs of BCL11A from 7 different species were identified using the MEME software tool. A total of 15 conserved motifs containing 21 to 50 residues were identified to be consistent in the order from the N terminus to the C terminus among the seven sequences (Fig. 6A). Amino acid multiple sequence alignment further revealed that BCL11A is evolutionarily conserved in vertebrates, with more than 48% amino acid sequence similarity among the seven tested proteins. In addition, BCL11A contains 6 C2H2-type zinc finger (ZnF) domains, and the three ZnF domains located at the C terminus were tightly arranged (see Fig. S1 in the supplemental material). The Sequence logo of the zinc finger domains further revealed that some amino acid residues are highly conserved at specific positions (Fig. 6B). Using the SWISS-MODEL program, the combined domain (809 to 893 amino acids [aa], ZnF 4 to 6) of flounder BCL11A (PoBCL11A) was shown to apparently exhibit overall conserved tertiary structures along with those of Homo sapiens BCL11A (HsBCL11A) and Danio rerio (DrBCL11A). More specifically, only one residue (Lys⁸⁰⁶ of HsBCL11A, Phe⁷⁴⁰ of DrBCL11A, ASP⁸⁴² of PoBCL11A) in each sequence was different from the others (Fig. 6C).

FIG 6.

Amino acid sequence and tertiary structure of flounder BCL11A. (A) Conserved motif analysis of BCL11A from seven species constructed using MEME; each box represents an individual motif. (B) Web logo showing conservation of amino acid residues in each C2H2-type ZnF domain of BCL11A. (C) The tertiary structure of the C terminus of HsBCL11A, DrBCL11A, and PoBCL11A, Lys⁸⁰⁶ of HsBCL11A, Phe⁷⁴⁰ of DrBCL11A, and ASP⁸⁴² of PoBCL11A in each sequence was different from the others.

Antiviral mechanism of flounder BCL11A against HIRRV infection.

We wanted to clarify why PoBCL11A could stand out from the candidate genes with a superior antiviral activity against HIRRV. Since many antiviral responses are regulated by interferons, the influence of bcl11a expression on the interferon-mediated antiviral response was first investigated in HINAE cells. The result of qRT-PCR analysis revealed that overexpression of bcl11a in HINAE cells cannot upregulate the expressions of interferon (IFN)-related genes such as irf3, irf7, ifn I-3, ifnγ, isg15, and mx whether the cells were infected with HIRRV or not (Fig. 7A and B). Given that BCL11A has multiple ZnF domains, the viral RNA binding activity of PoBCL11A was further examined. The interaction between PoBCL11A and viral RNA was detected by RNA binding protein immunoprecipitation (RIP) assays, and then viral RNA was amplified using primers that span different noncoding regions between two adjacent coding regions (segments N-P, P-M, M-G, G-NV, and NV-L). The result showed that segments N-P, P-M, and M-G can be specifically amplified and detected from the PoBCL11A-immunoprecipitated complex, and their relative abundances were significantly higher than those of the control and other gene segments (Fig. 7C). To further determine the functional region of PoBCL11A that interacts with viral RNA, we found that the binding activities of the N-terminal portion of PoBCL11A (N-PoBCL11A) include three dispersed ZnF domains; the fact that the C-terminal portion of PoBCL11A (C-PoBCL11A) includes three intensive ZnF domains was further determined by RIP assay. It was revealed that the HIRRV RNA binding activity of C-PoBCL11A was significantly stronger than that of N-PoBCL11A, and the segments N-P, P-M, and M-G can also be amplified from its immunoprecipitated complex (Fig. 7C). Furthermore, FCM was performed to determine whether C-PoBCL11A also can inhibit the infection and replication of HIRRV. It was observed that the percentage of HIRRV⁺ cells in the HINAE cells transfected with the C-PoBCL11A plasmid was significantly lower than that of the empty vector transfection group (Fig. 7D and E). Meanwhile, the MFI value for HIRRV in the HINAE cells overexpressing C-PoBCL11A was significantly lower than that of the negative cells (Fig. 7D and F).

FIG 7.

Molecular mechanism of antiviral effects of flounder BCL11A against HIRRV. (A and B) Relative expression levels of irf3, irf7, ifn I-3, ifn γ, isg15, and mx genes in HINAE cells transfected with pBCL11A plasmid (A) before and (B) after HIRRV infection. (C) RIP assay with HINAE cell lysates expressing F-BCL11A, N-BCL11A, and C-BCL11A by using anti-GFP antibody. Quantification of immunoprecipitated viral RNA by qRT-PCR, represented as the fraction of input RNA prior to immunoprecipitation (% input). (D) Flow cytometric analysis of HINAE cells transfected with C-BCL11A plasmid or empty vector plasmid after HIRRV infection. Scatterplots show transfected cells and untransfected cells that were distinguished by the expression of GFP; histograms derived from scatterplots show the percentage of HIRRV⁺ cells in different cell populations. (E) The percentage of HIRRV⁺ cells in total HINAE cells transfected with C-BCL11A plasmid or empty vector plasmid after HIRRV infection. (F) The MFI of HIRRV in pC-BCL11A transfected and untransfected cells. Different letters above the bars represent significant differences among different amplification groups (P < 0.05). Significant differences compared with the mock transfection group are indicated by asterisks (***, P < 0.001).

DISCUSSION

Fish rhabdoviruses, as a kind of important pathogenic microorganism of aquatic animals, have been widely studied. According to the available reports, subcutaneous or visceral bleeding is a typical clinical sign of fish infected by rhabdoviruses including HIRRV, infectious hematopoietic necrosis virus (IHNV), VHSV, spring viremia of carp (SVC), etc. (11, 31–33). This phenomenon led us to consider the link between fish rhabdoviruses and blood cells; our previous studies of HIRRV infection and flounder responses indicated that B lymphocytes of flounder might be the target cells of HIRRV. In the present study, we demonstrated that HIRRV can be detected within flounder B lymphocytes after the cell is exposed to the virus. However, considering the phagocytic activity of flounder B cells (4), it is necessary to further determine whether these intracellular viruses arise from the active infection of HIRRV or the nonspecific phagocytosis of B cells. It was reported that most rhabdoviruses are internalized by receptor-mediated endocytosis and subsequently fuse with a cellular membrane within the acidic environment of the endosome, which is triggered by viral glycoprotein (34), and fibronectin has been described as a primary receptor for fish rhabdoviruses (VHSV, IHNV, SVC) infection in teleost (35). Meanwhile, it cannot be ignored that some viruses can be internalized by phagocytosis. For example, HCV and Acanthamoeba polyphaga mimivirus (APMV) entry into macrophages mainly depended on the phagocytic activity of macrophages (36, 37). In the present study, based on indirect immunofluorescence assay (IFA) and nested PCR, extremely small amounts of viral protein and nucleic acid were detected in the B cells that incubated with inactivated HIRRV, which were obviously lower than those in live HIRRV-infected B cells. These results indicated that the tiny amount of HIRRV in the flounder B lymphocytes was taken in by passive phagocytosis, which is probably not the main entry route for live HIRRV.

Interestingly, we found that merely 10 to 20% of B lymphocytes can be infected by HIRRV, and the proportion of HIRRV-infected B cells will not increase with the extension of exposure duration. A similar result was observed in PCV2-infected pig peripheral blood mononuclear cells (PBMC); the proportion of PCV2-positive IgM⁺ cells was always maintained at around 10% at 36 or 72 h after viral infection (22). It is well documented that B lymphocytes at different developmental stages express a distinct repertoire of surface molecules and perform different physiological and immunological functions (38, 39); thus, the difference between the sensitivity and susceptibility of B cells to HIRRV infection is probably due to the high heterogeneity of the B lymphocyte subpopulation. Supporting this finding, previous research on measles virus (MV) and B cell subsets demonstrated that naive and memory B cells of peripheral blood and tonsillar origin were susceptible and permissive to MV infection (40). Regrettably, the B cell subsets were not clearly demarcated in fish, and there is still a lack of specific probes for different B cell subpopulations, so the B cell subpopulation susceptible to HIRRV infection is still hard to identify.

Viral infection is a complicated process that involves viral entry, protein synthesis, genome replication, viral assembly, and budding; failure at any point may lead to unproductive infection even if virions have been internalized into the cells (41). In the present study, we found that copies of viral gRNA, cRNA, and mRNA in HIRRV-infected B cells were significantly lower than those in HINAE cells. One reason for this result could be the lower HIRRV infection rate of B cells. Interestingly, we did not observe the release of infectious virions from HIRRV-infected B cells. It is strongly suggested that B lymphocytes may express some host restriction factors that inhibit viral replication. In contrast to highly susceptible target cells, infection of semi- or nonpermissive cells usually results in abortive infection for most viruses. Taking influenza A virus (IAV) as an example, IAV can infect airway macrophages, but replication of seasonal IAV is generally limited through abortive infection (42). Despite synthesis of all segments of gRNA and mRNA and the production of seven distinct viral proteins, the assembly of viral particles was defective in IAV-infected macrophages (43), which was possibly due to inefficient interaction between proteins HA and M2 at the cell surface (44), as well as type I IFN-dependent responses of macrophages (45). Such situations are common in virus-host interactions that include but are not limited to PCV2 and porcine leukocytes (22), astroviruses and neural cells (46), nucleopolyhedroviruses and Bombyx mori cells (47), and snakehead fish vesiculovirus (SHVV) and zebrafish embryonic fibroblast cells (48). Abortive infection is an important component of disease pathogenesis despite being incapable of generating infectious virions. A typical example is AIDS; the death process of CD4⁺ T lymphocytes drives clinical progression from HIV positivity to AIDS, but less is known about how HIV promotes the loss of CD4⁺ T cells (49). Recently, it was reported that the resting CD4⁺ T cells accounting for the vast majority in lymphoid tissues are nonpermissive to HIV-1 (50), but the incomplete DNA products generated by inefficient viral reverse transcription in resting CD4⁺ T cells can be detected by IFI16 DNA sensor then provoke a vigorous inflammatory response, ultimately resulting in pyroptosis (51, 52). This highly inflammatory form of programmed cell death triggered by the abortive infection of HIV-1 eventually causes the massive loss of resting CD4⁺ T cells. Similarly, abortive infection of Ebola virus in CD4⁺ T lymphocytes will also induce excessive autophagy, ultimately leading to profound T cell depletion (53). Coincidentally, the decrease of IgM⁺ B lymphocytes was also observed in our previous research (25). Although the abortive infection has been widely described and discussed in various viruses, these conclusions need to be viewed dialectically in the context of different pathological processes. Even though the abortive infection can be an alternative mechanism for host antiviral responses, it can be a double-edged sword, as excessive inflammation will cause host damage in some cases. Therefore, determining whether similar side effects exist in HIRRV-infected B cells requires further investigation. Based on the inefficient viral RNA synthesis and failure of viral particle release, it is interesting to speculate that defective replication of HIRRV in flounder B lymphocytes may be influenced by multiple factors, including the expression of antiviral genes or cell-specific host proteins and interactions between cellular and viral proteins.

The pathogenesis of HIRRV infection has been considered to be closely associated with temperature (13). Nonetheless, slight replication of HIRRV gRNA in B cells could be detected both at 10°C and 20°C in vivo, which implies that the same antiviral mechanisms may function under different temperatures. It is generally observed that cRNA is in relatively low abundance compared to gRNA during the replication of negative-strand RNA virus (54, 55). Interestingly, an excess of cRNA over gRNA was observed in HIRRV-infected B lymphocytes, which enticed us to seriously consider the presence of host protein that can interact with viral RNA. Therefore, special attention was paid to the upregulated molecules with nucleic acid binding activity in B cells at different temperatures. In the published report on a transcriptional profiling comparison between IAV-infected airway epithelial cells and macrophages, research data suggested that transcriptional profiles hardwired during cell development are the major determinant underlying the different responses of airway epithelial cells and macrophages to IAV infection, and comparison of the gene expression patterns between cells that are permissive and nonpermissive to productive replication by virus represents an important step toward identifying particular cellular factors that may restrict or promote virus replication (56). In the present study, upregulated expression of anxa6, zhx2, dhx35, bcl11a, and rfx1 was observed in HIRRV-infected flounder B cells compared to the HINAE cells derived from flounder embryos. These genes include four transcription regulators involved in the zinc finger (ZnF) protein superfamily and RFX family, as well as an RNA helicase. ZnF proteins typically serve as transcription factors that exert multiple biological functions by binding DNA, RNA, proteins, or small molecules (57). It is remarkable that some ZnF proteins play an important role in the innate immune response against viral infection. A series of ZnF proteins represented by ZAP are able to synergize with pattern recognition receptors (PRRs) such as RIG-I and MDA5 or regulate the expression of irf3, irf7, ifn, etc. (58, 59) and even directly interact with viral nucleic acid to exert antiviral effects (59). Additionally, emerging pieces of evidence suggest that multiple RNA helicases such as DDX1, DDX3, and DHX9 also have been demonstrated to participate in host antiviral processes (60, 61). Thus, it is possible that these differentially expressed genes may be somehow involved in the abortive infection of HIRRV in B cells.

To our surprise, the HINAE cells overexpressing only BCL11A have the capacity to delay the cytopathic effect and inhibit the viral replication of HIRRV. As a transcription factor, BCL11A is expressed in most blood cell categories and is highly enriched in hematopoietic stem cells, common lymphoid progenitors, early T cell progenitors, and B cells (62). This is consistent with our finding of constitutively high expression of bcl11a in flounder B cells. Previous evidence from humans and mice revealed that bcl11a is expressed from the earliest B cell progenitors through the germinal center (GC) stage but is extinguished in terminally differentiated plasma cells (63); experimentation in vivo has confirmed an essential requirement for BCL11A in B cell lymphopoiesis (62). Moreover, BCL11A has also been shown to be involved in the modulation of variable-diversity-joining [V(D)J] segment rearrangement within B cells by activating the expression of recombination-activating gene 1 (RAG1) and RAG2, which are tightly associated with B cell repertoire diversity (64). Multiple sequence alignment and tertiary structure prediction indicate that BCL11A is highly conserved in vertebrate species during evolution. We suppose that, to some extent, BCL11A in teleost fish may also span the gap between innate and adaptive immunity like activation-induced deaminase (65). The overexpression of bcl11a cannot cause the upregulation of the expression of irf3, irf7, ifn-γ, etc. Whether or not the cells were infected by HIRRV indicated that the anti-HIRRV effects of BCL11A may be exerted via IFN-independent mechanisms. A limited number of papers have focused on the antiviral effect of BCL11A. In HIV-1-infected human microglial cells, BCL11A and BCL11B can be recruited to the proximal long terminal repeat (LTR) region of the HIV-1 promoter and then impair Sp1- and chicken ovalbumin upstream promoter transcription factor (COUP-TF)-mediated HIV-1 gene transcription (66). In mammals, the N-terminal region containing a C2H2 ZnF domain common to all BCL11A isoforms is evolutionarily conserved and can be used to define this superfamily (63), and the most abundant isoform that has been identified in human lymphoid samples is BCL11A-XL, which contains 6 C2H2 ZnF domains. In the present study, segments of viral RNA were detected in the immunoprecipitation of full-length BCL11A and a C-terminal BCL11A fragment by RIP assay. This indicated that the three intensive ZnF domains located in the C terminus of BCL11A may play an essential role in the viral RNA binding process. Limited by the low abundance and cleavage of RNA extracted from these immunoprecipitations, using strand-specific primers to distinguish viral RNA types was not feasible. Nevertheless, that the amplification of these segments contained noncoding regions indicated that flounder BCL11A is able to directly bind with the full-length RNA of HIRRV (gRNA or cRNA). In addition, the contents of gene fragments N-P, P-M, and M-G in the BCL11A-immunoprecipitated complex were significantly higher than those of fragments G-NV and NV-L, which indicated that the BCL11A binding sites on viral full-length RNA may locate in the regions of N, P, and M genes.

In conclusion, our research demonstrates for the first time that HIRRV causes an abortive infection in flounder B cells, and we believe the rapid response and constitutively high expression of host restriction factor BCL11A plays an important role in restricting HIRRV replication in flounder B cells. However, there are still some gaps in our understanding of the life cycle of HIRRV replication in B lymphocytes and, in particular, the mechanisms by which the specific B cell subset blocks replication of HIRRV. Further elucidation of how BCL11A and other factors restrict viral replication will continue to provide insights into how nonpermissive cells contend with virus invasion.

MATERIALS AND METHODS

Experimental fish and virus.

Flounder (Paralichthys olivaceus, 300 ± 50 g) were purchased from a local fish farm (Qingdao, China) and confirmed to be free from HIRRV infection by nested PCR. The fish were anaesthetized with 100 ng/mL MS-222 (Sigma, USA) before experimental manipulations. All investigations were conducted strictly in accordance with ethical standards and approved by the Instructional Animal Care and Use Committee of the Ocean University of China (permit number 20150101). HIRRV strain CNPo2015 was previously isolated from a naturally infected flounder (11) and stored at −80°C. The virus was cultured on hirame natural embryo (HINAE) cells in L15 medium with 2% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Gibco, USA) at 20°C. Virus titer was tested according to the method described by Reed and Muench (67) and adjusted to 1.0 × 10⁷ 50% tissue culture infective dose (TCID₅₀)/mL. For HIRRV inactivation, the virus was incubated at 56°C for 30 min. The complete inactivation of HIRRV was confirmed by inoculation into HINAE cells, followed by checking for cytopathic effects.

Lymphocyte isolation.

The peripheral blood was collected from the tail vena of flounder and mixed with the anticoagulant (RPMI 1640 containing 20 IU/mL heparin sodium and 1% bovine serum albumin [BSA]) at a volume ratio of 1:3. After removing the erythrocytes by low-speed centrifugation (100 × g, 20 min), the peripheral blood lymphocytes (PBLs) were further dissociated using discontinuous Percoll (GE, USA) density gradient (1.020/1.070) under axenic conditions according to the method described in a previous study (68). After purity examination of isolated lymphocytes by flow cytometry via forward scatter/side scatter (FSC/SSC) scatterplot analysis, the highly purified lymphocytes were cultured in L15 medium with 10% FBS to a final concentration of 5 × 10⁶ cells/mL. Then, the PBLs were incubated with live or inactivated HIRRV (MOI, 5) for different amounts of time (6, 12, 24, 48, and 72 h).

Flow cytometry and cell sorting.

Flow cytometry analysis was applied to detect the HIRRV within the B cells and further monitor its infection dynamics. HIRRV-infected PBLs were collected by centrifugation from a cell culture flask and washed three times with phosphate-buffered saline (PBS) containing 5% (vol/vol) newborn calf serum. For double staining, PBLs were sequentially incubated with monoclonal antibody (MAb 2D8; 1:1,000) and Alexa Fluor 488-conjugated goat anti-mouse IgG polyclonal antibody (PAb; 1:1000) (Invitrogen, USA). Then, the cells were fixed in 4% paraformaldehyde, followed by permeabilization with 0.1% Triton X-100 in PBS for 10 min and incubation with Alexa Fluor 647-labeled HIRRV G protein MAb (69). All samples were washed three times with PBS containing 5% (vol/vol) newborn calf serum between steps, and flow cytometry analysis was performed using an Accuri C6 cytometer (BD, USA). In addition, viable mIgM⁺ B lymphocytes were only labeled by MAb 2D8 for cell sorting by a FACS Aria III flow cytometer (BD, USA).

Sources and analysis of transcriptome data.

Transcriptome data for HIRRV-infected B lymphocytes at 10°C and 20°C were obtained from the public NCBI data resource (https://www.ncbi.nlm.nih.gov/) available under accession numbers SRR13300097, SRR13300098, SRR13300099, SRR13300100, SRR13300101, SRR13300102, SRR13300103, SRR13300107, SRR13300108, SRR13300109, SRR13300112, and SRR13300113. The gene expression level of each group was estimated with Cufflinks software (70), and the differentially expressed genes (DEGs) were analyzed with the EBSeq R package (71). Candidate molecules with nucleic acid binding ability or associated with viral infection were screened from these coupregulated genes in HIRRV-infected B cells at different temperatures based on their functional domains and the published literature. For the investigation of the expression pattern of these candidate genes, PBLs and HINAE cells were incubated with HIRRV and poly(I:C) for 48 h, respectively.

Plasmid construction and transfection.

Plasmid pCI-eGFP was used for recombinant plasmid construction. The genes encoding flounder anxa6, zhx2, dhx35, rfx1, and bcl11a, the N-terminal portion of bcl11a (N-bcl11a), and the C-terminal portion of bcl11a (C-bcl11a) were amplified with the primers listed in Table 1 and cloned into the pCI-eGFP plasmid by using standard molecular biology techniques. Briefly, the total RNA was extracted from sorted B cells and reverse-transcribed to cDNA as the template for cloning of different target genes. The amplification products of target genes were double-digested by restriction enzymes (NEB, USA) and then ligated with linearized plasmid by T4 DNA ligase (TaKaRa, Japan). The recombinant plasmid was transformed into DH5α competent cells (TaKaRa, Japan), followed by detection with colony PCR and sequencing. For transfection, the recombinant plasmid was extracted using an EndoFree mini plasmid kit II (Tiangen, China) and transfected into HINAE cells using Lipofectamine 3000 reagent (Invitrogen, USA); the empty plasmid was used as the negative control. Transfection efficiency was estimated visually by fluorescence microscopy and quantified using flow cytometric analysis. Afterward, transfected cells were infected by HIRRV (MOI, 1 or 0.1, respectively).

TABLE 1.

Primers for the construction of eukaryotic expression plasmids

Target	Primer sequence (5′–3′)a	Endonuclease	Theoretical MOL (kDa)
ANXA6	F: gtcgacbCACTACAACCCAACACACC	SalI	103.2
	R: ccgcggATAATCTTCCCCTCCGC	SacII
ZHX2	F: ccggaattcCCGCCTGAGCATCGC	EcoRI	126.8
	R: gtcgacGCACCAGTGACTCTCTCTGGTTTC	SalI
DHX35	F: ccggaattcTGGAAACCACAGGCCGTGA	EcoRI	106.8
	R: gtcgacGCCAGGACCCGGGATCTCTTG	SalI
RFX1	F: ccggaattcATGGCAACTTCTGGATACTCA	EcoRI	120.2
	R: gtcgacGCAGGTGTGTGCAGATTTAAAGTT	SalI
BCL11A	F: ccggaattcATGGGTGGGGCAGC	EcoRI	125.3
	R: ccgcggCTCTGACTTTATTTCGGTACTC	SacII
N-BCL11A	F: ccggaattcATGGCCCTCACCCCGTCCC	EcoRI	79.8
	R: ccgcggTCCAATACCGCCATTATTTT	SacII
C-BCL11A	F: ccggaattcATGCAATCGCCGTTCGCC	EcoRI	44.9
	R: ccgcggCTCTGACTTTATTTCGGTACTC	SacII

^aF, forward; R, reverse.

^bThe underlined bases denote the homologous arm for recombination.

Coculture of HIRRV-infected B lymphocytes and HINAE cells.

To determine if HIRRV can complete the replication cycle and give rise to infectious viral progeny in B lymphocytes, HIRRV-infected PBLs were collected and washed three times with L-15 medium containing 2% FBS, and then B cells were sorted by FACS. The sorted B cells incubated with inactivated HIRRV were used as the negative control. Subsequently, B cells (1 × 10⁵ cells) were added into adherent HINAE cells for coculture, and the cytopathic effect (CPE) of HINAE cells was observed once daily for 3 days under an inverted microscope.

Immunofluorescence assay and Western blotting.

For the indirect immunofluorescence assay, lymphocyte suspensions were dropped onto adhesion-coated slides, allowed to settle for 2 h, and then fixed with 4% paraformaldehyde. The slides were sequentially incubated with MAb 2D8 (1:1,000), Alexa Fluor 488-conjugated goat anti-mouse IgG PAb (1:1,000) (Invitrogen, USA), and Alexa Fluor 647-labeled HIRRV G protein MAb, and then the nuclei were counterstained with DAPI (4′,6-diamidino-2-phenylindole) for 10 min. The observation and photography were performed using an Axio Imager Z2 microscope system (Carl Zeiss, Germany). For Western blotting, the collected B lymphocytes were lysed in NP-40 buffer containing 1 mM phenylmethylsulfonyl fluoride (PMSF) (Beyotime, China). Proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane (Merck Millipore, Germany) after SDS-PAGE. After being blocked in PBS containing 5% bovine serum albumin (BSA) at 37°C for 2 h, the membrane was incubated with anti-HIRRV-G protein MAb in a blocking solution at 4°C overnight. After being washed three times with PBS with Tween 20 (PBST), the membrane was incubated with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (Jackson, USA). GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was used as the internal reference protein, which was detected with rabbit anti-GAPDH polyclonal antibody (ABclonal, China) and HRP-conjugated goat anti-rabbit IgG (Jackson, USA). The immunoreactive bands were visualized with an enhanced chemiluminescence (ECL) system (Vilber, France).

Nested PCR and quantitative real-time PCR (qRT-PCR).

The total RNA of each sample was extracted using RNAiso Plus reagent (TaKaRa, Japan), and the quality of the RNA was verified by agarose gel (1%) electrophoresis and quantified using a NanoDrop 8000 spectrophotometer (Thermo Fisher, USA). To evaluate the gene expression and detect viral nucleic acids, the RNA was reverse-transcribed into cDNA using the HiScript III first-strand cDNA synthesis kit (Vazyme, China) according to the manufacturer’s instructions. The viral RNA was amplified via nested PCR, and gene expression levels were detected by qRT-PCR; the specific primers used in this part are listed in Tables S1 and S2. The reaction procedure of nested PCR was as follows: predegeneration at 95°C for 5 min, 35 cycles of 95°C for 30 s, 57°C for 30 s, and 72°C for 1 min; elongation for 10 min at 72°C. Then, 1 μL of the first PCR product was used as the template for the nested PCR, and the PCR procedure was performed as described above. The qRT-PCR was performed using ChamQ universal SYBR qPCR master mix (Vazyme, China) in a LightCycler 480 II real-time system (Roche, CH), and the gene expression levels were analyzed using the 2^–△△CT relative quantification method (72); the general viral copies were detected according to the standard curve established by Zhang (13). To distinguish the gRNA, cRNA, and different mRNAs of HIRRV, reverse transcription primers were designed with a unique 20- to 24-nucleotide tag adjacent to the strand-specific sequence (Table 2). A 14.5-μL mixture containing 500 ng of RNA, 2 pmol of tagged primer, 1 μL of dNTP (deoxynucleoside triphosphate) mix (10 mM), and nuclease-free water was heated for 5 min at 65°C and chilled immediately on ice for 5 min. Subsequently, the reverse transcription (RT) reagent containing 4 μL 5× RT buffer, 1 μL reverse transcriptase (200 U/μL), and 0.5 μL RNase inhibitor (40 U/μL) were added and incubated at 60°C for 15 min followed by heating at 85°C for 10 min. The amplified gene segments were cloned into the pGEM-T plasmid vector to construct the standard plasmid, and serial 10-fold dilutions (10⁹, 10⁸, 10⁷, 10⁶, 10⁵, 10⁴, 10³, and 10² copies/μL) of the standard plasmid were used to establish standard curves (Fig. S2).

TABLE 2.

Strand-specific primers for qPCR of the HIRRV gRNA, cRNA, and mRNA

Target	Primer name	Sequence (5′–3′)
gRNA	gRNA-RT	CAGTGGATGCCGATTTCAGTGGTaGACTATTATGCACAGGAAACTGT
	gRNA-F	CAGTGGATGCCGATTTCAGTGGT
	gRNA-R	TGACGAGTGTCGATTGTTTGTTATG
cRNA	cRNA-RT	GCAGGCTGACGATGACGATACTGGTTGGTCGCTTCCCTGA
	cRNA-F	GCAGGCTGACGATGACGATAC
	cRNA-R	AGAGCAAGCCAACATCCCACTCGCC
N-mRNA	NmRNA-RT	CGATGGTTCAACTGGGGCATTTTTTTTTTTTCTATCTAAGAATAGT
	NmNA-F	CGATGGTTCAACTGGGGCATTTTT
	NmRNA-R	GGGGATGGGTACTTCAAGTCCTACG
P-mRNA	PmRNA-RT	TCAACTTGCGTAAGACTTTTTTTTTTTTCTATCTAGGGGAGTGGG
	PmRNA-F	TCAACTTGCGTAAGACTTTTT
	PmRNA-R	GGAGAGAGCCCTCGGAT
M-mRNA	MmRNA-RT	GCGTGTTCCATCAGTAGTTTTTTTTTTTCTATCTGATGTGAGTTGA
	MmRNA-F	GCGTGTTCCATCAGTAGTTTT
	MmRNA-R	ACCAAAGAGTGAAAGACTATTATGC
G-mRNA	GmRNA-RT	CGATGGCACGGGTAGCAAGTTTTTTTTTCTATCTAGGATTGTAGAGG
	GmRNA-F	CGATGGCACGGGTAGCAAGTTTT
	GmRNA-R	TCCTCGCCCTCGTATTCTTTCTGTA
NV-mRNA	NVmRNA-RT	GGACATGCTGGACCAGTTTTTTTTTTTCTATCTGGGGAGAGGG
	NVmRNA-F	GGACATGCTGGACCAGTTTT
	NVmRNA-R	CCTTGTTGTTGTCGGGGC
L-mRNA	LmRNA-RT	ATACGTCGGGTAGGACCATTTTTTTTTTTTGTATCTGGGTGCTG
	LmRNA-F	ATACGTCGGGTAGGACCATTTTT
	LmRNA-R	CATCATCACCAAGACGGTCTACATC

^aThe underlined bold sequence indicates specific designed sequence tag.

Crystal violet staining and CCK8 assay.

HINAE cells harboring different recombinant plasmids were seeded in 96-well plates at a density of approximately 5 × 10⁴ cells/well. After 72 h of HIRRV infection (MOI, 1 or 0.1), the virus plaques formed in HINAE cells were visualized by crystal violet staining. Briefly, cells were fixed with 4% formaldehyde for 15 min and then stained with 0.5% crystal violet for 15 min. Wells were washed extensively with distilled water and dried overnight for observation. Meanwhile, the CCK8 assay was applied to measure cell viability according to the manufacturer’s instructions. Specifically, CCK8 solution (10 μL) (Vazyme, China) was added into each well and incubated for 2 h. The absorbance of each well at 450 nm was measured using a microplate reader (BioTek, USA). The experiments were performed three times independently.

RNA binding protein immunoprecipitation (RIP) assay.

HINAE cells expressing flounder BCL11A, N-BCL11A, or C-BCL11A were infected with HIRRV (MOI, 1) for 24 h. Subsequently, the cells were lysed in 1 mL NP-40 buffer containing 100 U RNase inhibitors (Invitrogen, USA) and 1 mM PMSF for 10 min at 4°C, and the cell lysate was centrifuged (12,000 × g) at 4°C for 15 min and incubated with protein G agarose beads that were conjugated with anti-green fluorescent protein (GFP) MAb (Invitrogen, USA) or negative MAb at 4°C overnight. After five washes, the precipitated RNA was extracted using RNAiso Plus reagent (TaKaRa, Japan) and analyzed by qRT-PCR with the appropriate primers to detect the target RNA. The amount of immunoprecipitated RNA is represented as the percentile of the amount of input RNA (percentage input).

Bioinformatics analysis of flounder BCL11A.

Bioinformatics analysis of flounder BCL11A was performed using several bioinformatics softwares and websites. Specifically, amino acid sequences of BCL11A from Homo sapiens, Mus musculus, Xenopus tropicalis, Danio rerio, Oryzias latipes, Salmo salar, and Paralichthys olivaceus were obtained from the NCBI database under accession numbers NP_075044, XP_036012211, NP_001072657, AAI15320, XP_011482410, XP_014010074, and XP_019944259, respectively. DNAMAN software was used to perform multiple sequence alignment, and protein domains were predicted using the SMART program (http://smart.embl-heidelberg.de/smart/set_mode.cgi?NORMAL=1). In addition, motif analysis was performed online with MEME (http://meme-suite.org/tools/meme), and sequence logos were generated using the WebLogo tool. The domain structures of BCL11A were analyzed using SWISS-MODEL (https://swissmodel.expasy.org/).

Statistical analysis.

Statistical analysis was conducted with Prism 8.0 (GraphPad). All experimental results were independently repeated at least three times, and data are expressed as means ± standard deviations (SD). Differences between groups were compared using Student’s t test or one-way analysis of variance (ANOVA) followed by a post hoc Tukey honestly significant difference (HSD) test. Significant differences are indicated in the figures as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001; and different letters represent a significance difference (P < 0.05).

Data availability.

The transcriptome raw data for HIRRV-infected B lymphocytes under 10°C and 20°C conditions can be found in the NCBI data resource (https://www.ncbi.nlm.nih.gov/) under accession numbers SRR13300097 to SRR13300103, SRR13300107 to SRR13300109, SRR13300112, and SRR13300113.