The roles of grouper clathrin light chains in regulating the infection of a novel marine DNA virus, Singapore grouper iridovirus

Clathrins, composed of clathrin heavy chains (CHCs) and clathrin light chains (CLCs), are usually hijacked by viruses for infection. However, the role of CLCs, especially in regulating fish virus infection, remains poorly understood. Here, two isoforms of CLCs were cloned from the red-spotted grouper (Epinephelus akaara) (EaCLCa and EaCLCb). Both EaCLC transcripts were expressed in all examined tissues, and the expression of EaCLCa was much higher than that of EaCLCb. Over-expressing EaCLCa-W119R mutant significantly reduced Singapore grouper iridovirus (SGIV) infectivity. However, no effect of EaCLCb-W122R on SGIV infection was observed. The detailed steps were further studied, mainly including virus attachment, entry and the following transport to early endosomes. EaCLCa-W119R mutant notably inhibited internalization of SGIV particles with no effect on SGIV attachment. Furthermore, EaCLCa-W119R mutant obviously impaired the delivery of SGIV to early endosomes after virus internalization. In addition, the EaCLCa-W119R mutant markedly reduced the colocalization of SGIV and actin. However, EaCLCb is not required for such events during SGIV infection. Taken together, these results demonstrate for the first time that EaCLCa and EaCLCb exerted different impacts on iridovirus infection, providing a better understanding of the mechanisms of SGIV infection and opportunities for the design of new antiviral strategies.

stomatitis virus, human papillomavirus type 16, and adenovirus, require both actin and clathrin for entry [11][12][13] . In addition, recent studies have shown that clathrin also affects other events in viral life cycles. A functional clathrin-binding motif within the large antigen protein (Ag-L) of hepatitis delta virus (HDV) was identified, and the interaction between clathrin and Ag-L significantly affected HDV assembly 14,15 . The nonstructural protein mammalian reovirus (MRV) could recruit cellular clathrin to viral factories, further disrupting normal clathrin-dependent trafficking 16 . However, most studies focus on the interactions of clathrin with viruses, and the role of CLCs in virus infection remain largely unknown.
There are two isoforms of CLCs in all metazoans, CLCa and CLCb, encoded by different genes. They share approximately 60% protein sequence identity and are expressed at characteristically different levels in all tissues. Their longest shared fragment is 22 residues near the N terminus, starting with three negatively charged residues (EED), termed the consensus (CON) sequence, which serves as a binding site for huntingtin interacting proteins (HIPs) and HIP1-related (HIP1R) and regulates clathrin self-assembly 4,17,18 . In mammals, from the N terminus to the C terminus, other features shared by the two CLCs include an EF-hand that is responsible for binding to calcium (Ca 2+ ), the heavy-chain-binding region (HC), neuronally expressed inserts (N), and a calmodulin-binding domain (CBD). The unique regions in CLCa and CLCb are Hsc70 and serine phosphorylation sites (P), respectively 1 . To date, we know little about the different functions of CLCa and CLCb, especially during the viral life cycle.
Iridoviruses, large dsDNA viruses, have attracted increasing attention due to the threat they pose to aquaculture and biodiversity 19 . To date, iridoviruses can infect invertebrates and poikilothermic vertebrates, including fish, amphibians, and reptiles 19,20 . The type species of the genus Ranavirus, frog virus 3 (FV3), uses clathrin mediated endocytosis (CME) for entry mammalian cells 21 . Singapore grouper iridovirus (SGIV), which causes a serious and highly lethal systemic disease in grouper aquaculture, is a novel member of the genus Ranavirus and the family Iridoviridae 22 . Our previous study showed that SGIV could enter host cells through CME and then undergo transport to the endosomal compartment 23 . However, the impact of clathrin, especially CLCs, on the life cycle of SGIV still remains limited.
In this study, EaCLCa and EaCLCb, derived from grouper Epinephelus akaara, were cloned and characterized. Consistent with previous results 23 , EaCLC mainly affect early events during SGIV infection. Notably, although EaCLCa and EaCLCb shared approximately 65% sequence identity, they showed clearly different effects on SGIV infection. Data from the dominant-negative EaCLCa mutant demonstrated that EaCLCa significantly affects SGIV entry and the subsequent step, transport to the early endosomes, possibly by regulating the interaction of virus and actin. In contrast, EaCLCb showed no measurable effect on SGIV infection.

Results
Sequence analysis of eacLc. Based on the transcriptome analysis, we obtained the full-length open reading frames of EaCLCa and EaCLCb by PCR amplification. EaCLCa encodes a 202-amino-acid protein and has remarkably high sequence conservation in vertebrates according to the alignment of amino acid sequences 99% and 75% identity to Epinephelus coioides and Homo sapiens, respectively (Fig. 1A). EaCLCb encodes a 205-amino-acid protein with a greater sequence divergence 92% and 68% identity to Stegastes partitus and H.sapiens, respectively (Fig. 1B). Amino acid alignment showed that both EaCLC contain CON, Ca, HC and Cam domains. EaCLCa shares 65% identity with EaCLCb. Phylogenetic analysis indicated that EaCLCa and EaCLCb are both sorted into the Osteichthyes branch, which is separate from amphibian and mammals ( Supplementary Fig. S1). expression pattern of eacLc. Both EaCLC were found to be distributed in all tissues examined in grouper, E. akaara, but at characteristically different levels. EaCLCa was predominantly detected in the heart and head kidney, followed by the spleen, liver, fin, brain, skin, kidney, gill, intestine, muscle and stomach ( Fig. 2A). However, EaCLCb was highly expressed in the spleen, fin and gill but showed low expression in the liver, stomach and intestine (Fig. 2B). Interestingly, the expression of EaCLCa was much higher than that of EaCLCb in all tissues. To evaluate the expression changes of EaCLCa and EaCLCb in response to SGIV, the grouper was challenged with SGIV and the transcript of EaCLCa and EaCLCb was detected using qRT-PCR. Our results showed that the expression level of EaCLCa and EaCLCb firstly upregulated at 6 h, and reached the peak at 18 h, then decreased since 24 h. In addition, the expression of EaCLCa was also higher than that of EaCLCb in infected spleen (Fig. 2C,D). Besides, we further explore the expression profiles of EaCLCa and EaCLCb upon virus infection in GS cells. As shown in Fig. 2E,F, the transcription levels of both EaCLCa and EaCLCb were clearly upregulated.

Subcellular localization of eacLc. The intracellular localization of EaCLC in living cells was investigated.
Both EaCLCa and EaCLCb displayed punctate structures located in the cytoplasm and on the cell membrane (Fig. 3A). After infection with SGIV, EaCLC were observed to surround the virus factory (Fig. 3B), preliminarily proving their role in virus replication and assembly.
In addition, we detected the distribution of EaCLCs and SGIV at different times post-infection. Upon infection, Alex-Fluor 647 labelled SGIV particles rapidly colocalized with EaCLCa and EaCLCb ( Supplementary  Fig. S2A). Moreover, the percentage of SGIV particles that colocalized with EaCLCa was slightly higher than the percentage that colocalized with EaCLCb ( Supplementary Fig. S2B).

Effect of EaCLC on SGIV infectivity.
It is reported that the residue 130 W in HC domain of CLCa and 127 W in HC domain of CLCb are essential for their interaction with CHC 24 . To further analyse the potential role of EaCLC in SGIV infection, we generated two mutations of EaCLC to perturb the normal function of CLC. A point mutation in the residue 119 of EaCLCa and residue 122 of EaCLCb, from W to R, were performed. In the EaCLCa is required for SGIV entry. To test our hypothesis that EaCLC are involved in SGIV entry, we measured single SGIV particle entry events in GS cells transfected with dominant-negative EaCLC. GS cells were transfected separately with pcDNA-3.1, pcDNA-EaCLCa-W119R and pcDNA-EaCLCb-W122R, incubated with Alex-Fluor 647 labelled SGIV for 2 h and processed for confocal imaging. Visually, a noticeable reduction in virus uptake was observed in cells expressing EaCLCa mutants compared with that in cells expressing pcDNA-3.1 (Fig. 5A). Moreover, as shown in Fig. 5B, SGIV entry was significantly inhibited (up to 40% inhibition) in the presence of the EaCLCa mutant. In addition, EaCLCa mutant overexpression inhibited the viral gene www.nature.com/scientificreports www.nature.com/scientificreports/ transcription (Fig. 5C). However, SGIV internalization was unaffected by the EaCLCb mutant (Fig. 5B). These results revealed that EaCLCa was involved in SGIV entry, but EaCLCb had no significant effect.

EaCLCa affects SGIV entry at a post-binding step.
To further probe the role of the EaCLC in SGIV entry, a virus attachment assay was performed. The cells expressing EaCLC mutants were incubated with SGIV at 4 °C for 1 h, followed by the removal of unbound virus, immediate fixation, fluorescent imaging and the quantification of cell-bound SGIV by a MATLAB program. As shown in Fig. 6, there was no significant variance between EaCLC mutants and mock cells. In addition, the number of SGIV particles attached on the cell surface remained unchanged during SGIV entry (Fig. 5D). These data suggested that the EaCLC had no effect on virus attachment. EaCLCa mutant disrupts SGIV particle trafficking to the early endosome. After internalization by clathrin-mediated endocytosis, cargo molecules, including receptors, ions, lipids, and even viruses, are usually transported to early endosomes (EEs) for further sorting.
To explore whether EaCLC affect the transport of SGIV particles to EEs, the colocalization of SGIV particles to Rab5 (marker of EEs) was analysed. In EaCLC mutants and pEGFP-Rab5 co-transfected cells, EGFP-Rab5 also showed a punctate cytoplasmic distribution. However, at 0.5 hpi, the colocalization between SGIV particles and EEs was significantly decreased in cells overexpressing mutant EaCLCa compared with the mock cells (Fig. 7). However, EaCLCb had no obvious effect on the transport of SGIV particles to EEs (Fig. 7). In addition, disruption of EaCLCa did not appear to alter the morphology and distribution of Rab5-positive endosomes. Taken together, these results indicated that the regulation of SGIV infection by EaCLCa also occurred at the post internalization step.

EaCLCa mutant alters the colocalization between SGIV and actin filaments.
Some viruses, such as vesicular stomatitis virus (VSV) depend on actin filaments for entry 25 . VSV enter the cells by CME assisted by actin filaments 25   www.nature.com/scientificreports www.nature.com/scientificreports/ Therefore, we detected correlations between SGIV particles and actin in different treatments to explore the effect of EaCLC on the interaction between actin and SGIV. Alexa Fluor 488 phalloidin was used to label actin filaments. In cells transfected with pcDNA-EaCLCa-W119R, the actin distribution showed no obvious change, but the colocalization between SGIV particles and actin was significantly reduced compared with the mock cells (Fig. 8). However, there was no difference in the cells transfected with the EaCLCb mutant. These results revealed that EaCLCa could affect the interaction between SGIV particles and actin filaments.

Discussion
Clathrin is the principal structural component of clathrin-coated carriers and is involved in a wide range of cellular processes as well as the life cycle of viruses. The structural unit of clathrin, the triskelion, consists of trimerized CHCs and three CLCs. To date, only a single CLC has been identified in yeast and invertebrates. However, in vertebrates, two forms of CLC (CLCa and CLCb) appear. Previous reports indicated that two CLC isoforms arose by local gene duplication and contributed to the development of increasing complexity during chordate evolution 28 . In mammals, CLCa and CLCb exhibit distinct levels of expression in different tissues, with CLCa and CLCb dominant in the lymphoid tissue and the brain, respectively. In groupers, EaCLCa and EaCLCb were both distributed in all tissues analysed in our study, but they were predominantly detected in different tissues. These different characteristics of CLCs in different species may imply a potential for diverse functions.
GFP-tagged EaCLC appeared as spots in the cytoplasm and cell membrane, displaying the typical pattern observed in mammalian cells. In addition, EaCLCa and EaCLCb showed similar localization in the cell. In mammalian cells, CLCa and CLCb distribute randomly in cellular triskelia, such that all four types of triskelia (aaa, aab, bba, and bbb) are always present 29 .
Notably, the disruption of EaCLCa and EaCLCb by dominant-negative interference had different effects on SGIV entry. Only the EaCLCa mutant significantly impaired SGIV uptake, verifying that EaCLCa was critical for SGIV entry. Unexpectedly, although EaCLCb shared high sequence similarity and a similar distribution with EaCLCa, it had no significant effect on SGIV entry, suggesting functional differences between the EaCLC in regulating SGIV entry. Thus far, this report is the first on the different functions of CLC in iridovirus infection. Conformational www.nature.com/scientificreports www.nature.com/scientificreports/ variation in CLC may result in functional differences. In addition, attachment assays showed that neither of the EaCLC affected the binding of SGIV particles to the cell membrane, demonstrating that EaCLCa affects a post binding step of SGIV entry, which is consistent with the role of EaCLCa in cellular membrane trafficking. Prompted by the previous finding that SGIV was delivered to early endosomes after internalization, we demonstrated here that EaCLCa mutant significantly reduced the localization of SGIV particles to EEs, suggesting that EaCLCa also affects the step after virus entry. In addition, by calculating SGIV-infected cells, we found that SGIV infectivity was also significantly inhibited only when EaCLCa was disrupted. However, because of technological limitations, it remains unclear whether EaCLCa could affect other steps, such as virus replication and assembly. In contrast, by adding the small molecular inhibitor Pitstop 2 at different times, we found that EaCHC affected both virus entry and the steps after virus internalization. EaCLC may have similar effects on events after virus entry. During the late stage of SGIV infection, EaCLC tightly surrounded the viral factories, suggesting that clathrin may be recruited for virus replication, assembly or release. A few viruses have been reported to be directly connected with clathrin during virus assembly. Ag-L of hepatitis delta virus (HDV) was demonstrated to be related to CHC, thus affecting HDV particle assembly 14,15 . μNS of mammalian reovirus (MRV) not only recruits clathrin to viral factories but also interferes with normal functions of clathrin in cellular membrane trafficking 16 . (B) Quantification of SGIV particle binding on the cell membrane. More than 30 cells were randomly selected and analysed by a MATLAB program. Each experiment was repeated three times. The SGIV attachment was quantified as the percentage of treated cells with virus particles attached on the cell membrane relative to that for mock cells. SGIV attachment to mock cells was arbitrarily set as 100%. The data are shown as the means ± SEM. *p < 0.05.
In addition, the overexpression of mutant EaCLCa significantly impaired the colocalization between SGIV particles and actin filaments, suggesting that EaCLCa probably affects virus entry by influencing the interaction between SGIV and actin. However, the EaCLCb mutant had no effect. In mammals, CLCs have been reported to recruit Hip1 and Hip1R (referred to as Hip1/R) to form the LC-Hip1/R-actin complex, which mainly regulates actin assembly on clathrin-coated structures 26,27 . A dramatic change in actin organization could be observed in CLC knockdown cells. However, the disruption of EaCLCa or EaCLCb did not visibly change the morphology of actin. The data showed that EaCLCa probably affected the colocalization of SGIV with actin by other means.
Based on this study, we propose that EaCLCa and EaCLCb have different roles in SGIV infection. EaCLCa obviously regulates distinct steps in the SGIV life cycle, including virus entry and transport to early endosomes or actin filaments. Surprisingly, EaCLCb had no influence on these events during SGIV infection. Further studies are needed to explore the detailed mechanisms involved in the different roles of EaCLC in virus infection. The grouper spleen (GS) cell line used in this study was maintained in our laboratory 30 . GS cells were cultivated in Leibovitz's L-15 medium containing 10% foetal bovine serum (Gibco) at 28 °C. The SGIV used in this study was originally isolated from diseased grouper (Epinephelus tauvina) 22 . SGIV was propagated in GS cells, and virus stocks were maintained at −80 °C. E. akaara (eacLc). Based on the EST sequences from the grouper transcriptome 31 , the full-length open reading frames of EaCLC were amplified using the www.nature.com/scientificreports www.nature.com/scientificreports/ primers listed in Table 1. The putative amino acid sequence of EaCLC was analysed by BioEdit, the Expasy search program (http://au.exasy.org/tools) and BLAST in NCBI (http://www.ncbi.nlm.nih.gov/blast). The domain structure of the EaCLC was predicted by SMART (http://smart.embl-heidelberg.de/). Multiple sequence alignment and phylogenetic analysis of EaCLC were carried out by using Clustalx 1.83 (http://www.ebi.ac.uk/clustalW/) and MEGA 4.0 software (http://megasoftware.net/), respectively.

Identification and sequence analysis of CLCs from
Reagents and plasmid construction. Hoechst 33342 and Pitstop 2 were purchased from Sigma-Aldrich.
Pitstop 2 were dissolved in dimethyl sulfoxide (DMSO) according to the manufacturer's instructions. The lipophilic dyes DiO and DiD were purchased from Biotium. The fluorescent dyes Alexa Fluor 647 and Alexa Fluor 488 phalloidin were purchased from Invitrogen. anti-β-tubulin was purchased from Abcam (USA). peroxidase-conjugated affinipure goat anti-rabbit IgG were purchased from proteintech (USA).
Using the primers listed in Table 1, the full-length CLCs were constructed in vectors including pcD-NA3.1-flag, pEGFP-N3, and pmDsRed-C1 (Invitrogen). Site-directed mutants, including EaCLCa-W119R and EaCLCb-W122R, were all subcloned into the pEGFP-N3, pmDsRed-C1 and pcDNA3.1-flag vectors using specific primers (Table 1) and the Fast Mutagenesis Kit V2 (Vazyme). Tryptophan (W)119 of CLCa and W122 of CLCb were both replaced with arginine(R). In addition, the pEGFP-Rab5 vector was maintained in our laboratory. The constructed plasmids were confirmed by sequencing. expression patterns of eacLc. Total RNA was extracted from different tissues of healthy groupers, including the head kidney, kidney, liver, spleen, intestine, stomach, gill, brain, heart, skin, and muscle, and then examined by qRT-PCR.
For viral infection in vitro, GS cells were infected with SGIV at a multiplicity of infection (MOI) = 1 and collected at 0, 4, 12, and 24 hpi for further qRT-PCR analysis.
Total RNA was extracted from grouper tissues or GS cells using the SV Total RNA Isolation System (Promega) according to the manufacturer's protocol, examined by electrophoresis and then reverse transcribed into cDNA by the cDNA synthesis kit Rever Tra Ace (TOYOBO, Japan). qRT-PCR was carried out using a Light Cycler 480 Real-time PCR system (Roche, Basel, Switzerland) with SYBR Green as the fluorescent dye, according to the manufacturer's protocol (TOYOBO). β-Actin was used as an internal control. All primer pairs are listed in Table 1. Each assay was carried out in triplicate using the following cycling conditions: 94 °C for 5 min, followed Table 1. Sequence of primers used in this study. The underlined text means enzyme digestion sites (pcDNA-EaCLCa-F, pcDNA-EaCLCa-R, pcDNA-EaCLCb-F, pcDNA-EaCLCb-R, pEGFP-EaCLCa-F, pEGFP-EaCLCa-R, pEGFP-EaCLCb-F, pEGFP-EaCLCb-R, pEGFP-Rab5-F, pEGFP-Rab5-R, pDsRed1-EaCLCa-F, pDsRed1-EaCLCa-R, pDsRed1-EaCLCb-F, pDsRed1-EaCLCb-R) and mutant sites (pcDNA-EaCLCa mutant-F, pcDNA-EaCLCa mutant-R, pcDNA-EaCLCb mutant-F, pcDNA-EaCLCb mutant-R) added in the designing primers. by 45 cycles of 5 s at 94 °C, 10 s at 60 °C and 15 s at 72 °C. The data were calculated as fold changes based on the transcription levels of the targeted genes normalized to β-actin and given in terms of relative mRNA transcription level as the means ± standard deviation.
Virus purification and fluorescence labelling. Purified and fluorescence-labelled SGIV particles were produced as described previously 22,23 . Briefly, SGIV was inoculated onto monolayers of GS cells at an MOI of approximately 0.1. When the cytopathic effect was sufficient, SGIV particles were harvested by repeated freeze-thaw cycles, cleared of cell debris at 12,000 × g (Beckman Allegra X-15R Centrifuge) for 30 min at 4 °C, and the supernatant pelleted by ultracentrifugation at 200,000 g (Beckman 70 Ti rotor) for 1 h at 4 °C. Subsequently, the virus pellet was resuspended in TN buffer (50 mM Tris-HCl, 150 mM NaCl, pH 7.5) and further purified with a sucrose gradient (30-60%, wt/vol) at 150,000 g (Beckman SW 40 rotor) for 1 h at 4 °C. The viral band was washed with TN buffer, repelled by ultracentrifugation at 100,000 × g for 1 h at 4 °C, and then resuspended in TN buffer and stored at −80 °C until use.
SGIV particles were labelled with fluorescent dye by incubation with Alexa Fluor 647 in phosphate-buffered saline (PBS) (pH 7.4) at room temperature for 2 h with gentle vortexing for Alexa Fluor 647 dye conjugating to the primary amines (R-NH 2 ) of proteins on viral capsid. Then, unincorporated dye was removed by three high-speed centrifugations at 14,000 × g (Beckman Micrifuge 20 R Centrifuge) at 4 °C for 60 min. Labelled virus was examined under a transmission electron microscope (Supplementary Fig. S5) and stored at 4 °C. To test whether labelling Alexa Fluor 647 would affect the infectivity of SGIV, we used the western blot to investigate the protein level of SGIV MCP (Supplementary Fig. S6). Virus infection assay. GS cells were cultured in 24-well plates, transfected with pcDNA-3.1, pcDNA-EaCLCa-W119R and pcDNA-EaCLCb-W122R, and then incubated with SGIV (MOI = 1) 24 h after transfection. At 24 hpi, the cells were collected for western blot or fixed with 4% paraformaldehyde. The fixed samples would be stained with Hoechst 33342 to highlight the nucleus and viral factory. Images were taken by an inverted fluorescence microscope (Zeiss, Germany), and the infected foci were counted. SGIV binding and entry assays. The virion binding and uptake assay was performed by measuring the amounts of SGIV particles attached to the cell surface and in the cytoplasm, respectively. In brief, 24 h after plasmid (pcDNA-3.1, pcDNA-EaCLCa-W119R or pcDNA-EaCLCb-W122R) transfection, GS cells were incubated with Alex-Fluor 647 labelled SGIV (MOI = 10) at 4 °C for 1 h to allow virus binding. Then, the cells were either fixed immediately with 4% paraformaldehyde or transferred to 28 °C to allow SGIV entry and fixed at 2 hpi. These samples were further analysed by Confocal Laser Scanning Microscope (CLSM). The experiments repeated three times independently.
The plasmids pcDNA-3.1, pcDNA-EaCLCa-W119R and pcDNA-EaCLCb-W122R were separately transfected into cells 24 h prior to SGIV infection. Alex-Fluor 647 labelled SGIV (MOI = 10) was added to the cells at 4 °C for 20 min, and the cells were then moved to 28 °C to allow SGIV entry and fixed at 1 hpi. The fixed samples were stained with Alexa Fluor 488 phalloidin to highlight actin. Further analysis was performed by CLSM. confocal imaging assay. Fluorescent images were obtained through a ZEISS LSM 7 DUO confocal microscope. The signals of EGFP and DiO were excited using a 488-nm Ar-Kr laser and a 500-550 nm bandpass filter for emission. Alexa 647 and mCherry were excited with a 633-nm helium neon laser and a 650-700 nm bandpass filter for emission. Fluorescence emission was collected and imaged through a 100× (numerical aperture, 1.4) oil immersion objective. Image stacks were acquired with a 200-nm z-step size. For quantification analysis, confocal images were obtained by noise filtering, edge detection and fluorescent signal extraction using a MATLAB program. Approximately 100 cells were randomly analysed by the MATLAB program to calculate the number of SGIV particles on the cell membrane or in the cytoplasm. In addition, the 3D images of the colocalization between virus particles and Rab5/actin were performed by CLSM. Furthermore, the quantification of the colocalization between SGIV and Rab5/actin were analysed in more than 60 randomly chosen cells (2D images) by a MATLAB program.