CLIC2α Chloride Channel Orchestrates Immunomodulation of Hemocyte Phagocytosis and Bactericidal Activity in Crassostrea gigas

Summary Chloride ion plays critical roles in modulating immunological interactions. Herein, we demonstrated that the anion channel CLIC2α mediates Cl− flux to regulate hemocytes functions in the Pacific oyster (Crassostrea gigas). Specifically, during infection by Vibrio parahemolyticus, chloride influx was activated following onset of phagocytosis. Phosphorylation of Akt was stimulated by Cl− ions entering host cells, further contributing to signal transduction regulating internalization of bacteria through the PI3K/Akt signaling pathway. Concomitantly, Cl− entered phagosomes, promoted the acidification and maturation of phagosomes, and contributed to production of HOCl to eradicate engulfed bacteria. Finally, genomic screening reveals CLIC2α as a major Cl− channel gene responsible for regulating Cl− influx in oysters. Knockdown of CLIC2α predictably impeded phagosome acidification and restricted bacterial killing in oysters. In conclusion, our work has established CLIC2α as a prominent regulator of Cl− influx and thus Cl− function in C. gigas in bacterial infection contexts.


HIGHLIGHTS
Influx of chloride ions is switched on during phagocytosis in oyster hemocytes PI3K/Akt signaling pathway mediates chloride-dependent activation of phagocytosis Cl À promotes phagosomal acidification and HOCl production CLIC2a is the principal chloride channel encoding gene within oyster genome

INTRODUCTION
As a principal inorganic anion in the intra-and extracellular environments, chloride (Cl À ) is involved in an extraordinary range of physiological functions including body fluid retention/excretion, osmotic maintenance, cell volume regulation, and pH balance (Bohn and de Morais, 2017;Rodan, 2019;Wang, 2016). Accumulating evidence has implicated transmembrane Cl À fluxes in antimicrobial processes within immune effectors such as phagocytes, although the underlying mechanisms are not fully understood (Wang, 2016). In mammalian macrophages, phagocytosis is of fundamental importance to innate defenses against invading microbes (Hartenstein and Martinez, 2019). Notably, acidification of phagosomes governs their maturation and eventual antimicrobial capacity (Bouvier et al., 1994), in which Cl À flux is critical to phagosomal pH control and bacterial infection outcomes (Di et al., 2006). Meanwhile, activation of endolysosomal proteases temporally matches the maturation of phagosomes to ensure efficient destruction of engulfed bacteria (Pillay et al., 2002). Some immune-related enzymes, such as cathepsins, show a dependency on Cl À flux for activity via binding to Cl À ion (Cigic and Pain, 1999). In addition, Cl À directly participates in the production of chlorine-containing oxidants for microbial killing in host immunity. Specifically, in neutrophils, myeloperoxidase (MPO), which is enriched in phagosomes, catalytically converts hydrogen peroxide (H 2 O 2 ) and Cl À into the highly potent hypochlorous acid (HOCl), which oxidatively decimates microbes via protein chlorination (Busetto et al., 2007;Rosen et al., 2002Rosen et al., , 2009).
Owing to its intrinsic properties of a negatively charged ion, Cl À cannot autonomously permeate cellular membranes; its compartmental distribution instead has to depend on passive transport through specific channels or transporters (Stauber et al., 2012). Since the discovery of the first intracellular Cl À channel protein (p64, later renamed as CLIC5B) in bovines, more members of CLIC proteins were subsequently identified in all phyla, wherein six CLIC genes were found in humans (Landry et al., 1989;Redhead et al., 1992). The CLIC family located in various organelles of the cell is involved in physiological functions and pathological conditions in various human diseases, such as tumor onset and progression, Alzheimer's disease, and cardiac dysfunction (Flores-Tellez et al., 2015;Hernandez-Fernaud et al., 2017;Novarino et al., 2004;Peretti et al., 2015;Takano et al., 2012).
Chloride flux can govern specific signaling pathways to regulate its physiological functions. For example, activation of the ClC-3 chloride channel, responsible for inducing inhibition of the PI3K/Akt/mTOR signaling pathways, has also been proven to determine apoptosis in human nasopharyngeal carcinoma cell lines (CNE-1, CNE-2Z) (Liu et al., 2013). Over-expression of calcium-activated chloride channel A4 (CLCA4) could inhibit cell migration and invasion by suppressing epithelial-mesenchymal transition (EMT) via the PI3K/ATK signaling pathway (Chen et al., 2019). In addition, CLIC1 regulates migration and invasion in gastric cancer by triggering signaling of the ROS-mediated p38 MAPK pathway (Zhao et al., 2015). CLIC1 regulates colon cancer cell migration and invasion through ROS/ERK pathway . Thus, chloride flux may utilize distinct signaling pathways to execute specific functions in cellular context-dependent manners.
Beyond mammals, the biological roles of Cl À in other species such as plants and nematodes have been sparingly studied (Chakraborty et al., 2017;Jentsch, 2008;Nguyen et al., 2016). For instance, in Arabidopsis, AtClC-e is localized to thylakoid membranes in the chloroplast (Marmagne et al., 2007), where its absence impairs the proton-motive force. Branicky et al. found that the CeClC-3 channel modulates the electrical activity of HSN neurons that control egg laying in C. elegans (Branicky et al., 2014). Overall, current knowledge on Cl À function is mainly limited to mammals and tends to be more fragmentary concerning invertebrates. Despite substantial mammalian evidence that chloride channels are indispensable for robust phagosomal acidification and bactericidal activity (Jentsch, 2008;Moreland et al., 2006), how and to what extent chloride influx and chloride channels contribute to immune defenses in invertebrates is still under-examined. As a marine invertebrate with significant roles in ecological habitats, Crassostrea gigas has developed a versatile and intricate innate immune system capable of efficiently recognizing and removing invading pathogens (Wootton et al., 2003). From an evolutionary perspective, hemocytes in oyster are functional analogs of macrophages and neutrophils and are thus assumed to execute at least a subset of immune functions found in their human counterparts (Beaven and Paynter, 1999). Owing to a marine environment with high chloride, many physiological activities including host immune defense in oysters seem to be more dependent and susceptible to chloride than in terrestrial animals. In the present study, we set out to clarify the following cogent issues: (1) potential importance of chloride influx during phagocytosis of oyster hemocytes; (2) regulatory mechanisms that govern immune modulation by Cl À influx; and (3) the cardinal chloride channel encoding gene that is responsible for Cl À fluxes control in oyster hemocytes. 2 iScience 23, 101328, July 24, 2020 iScience Article

Chloride Influx Is Activated during Phagocytosis
To explore the possible immunodulatory roles of Cl À influx in oyster hemocytes, we first examined whether Cl À influx is activated during phagocytosis. Levels of intracellular Cl À concentration ([Cl À ] i ) were measured by using the Cl-specific fluorescent probe MQAE. MQAE's fluorescence intensity decreases proportionally with increasing chloride ion concentration. Cell viability assay showed that hemocytes cultured in vitro keep a high cell viability under a broad range of temperature ( Figure S1). Intriguingly, we observed a significant decrease in fluorescence intensity of MQAE (green Cl À sensor) in phagocytes (red-fluorescence positive cells), upon hemocyte engulfment of either pHrodo Red zymosan or E. coli ( Figure 1A), indicating an elevation of intracellular Cl À concentration [Cl À ] i during phagocytosis. To calibrate [Cl À ] i , a standard curve for MQAE fluorescence intensity versus [Cl À ] i was constructed by using a series of buffers prepared across a Cl À concentration gradient (Koncz and Daugirdas, 1994) ( Figure 1B ure 1C). However, no significant difference in magnitude of Cl À influx was observed during phagocytosis whether for pHrodo Red zymosan or E. coli, suggesting that activation of Cl À influx is a cellular event tightly coupled to phagocytosis initiation, regardless of the nature of phagocytosed substrates.

Chloride Channel Inhibitor Blocks Phagocytosis in Hemocytes
To determine the function of Cl À influx on phagocytic capacities of hemocytes, IAA-94, a potent indanyloxyacetic acid blocker of CLIC family channels, was employed in subsequent assays. In terms of intracellular Cl À availability in oyster hemocytes, treatment of IAA-94 (100 mM) efficiently blocked infection-induced Cl À influx activation ( Figure S3) (Figures 2A and 2B). Meanwhile, the capacity of hemocyte to engulf bacteria was sharply reduced in the presence IAA-94, when compared with the basal control (treatment with solvent) ( Figure S4). In agreement with this, the inhibitory effects of IAA-94 on hemocyte phagocytosis were extracted to estimate [Cl À ] i . Data were analyzed by using Image-Pro Plus 6.0. Data were analyzed by unpaired t test and presented as mean G SD, ***p < 0.001, n = 3. (C) IAA-94 inhibited phagocytosis of E. coli by hemocytes. Under the premise that the number of cells in each group was approximately the same, flow cytometry analysis was conducted to gauge the extent of hemocytic phagocytosis. Red color represents the solvent-treated control and blue color represents the group treated with IAA-94. (D) Data analysis was performed by using GraphPad Prism 7 software. Data were analyzed by unpaired t test and presented as mean G SD, ***p < 0.001, n = 3. (E) IAA-94 inhibited bactericidal ability of hemocytes. Display of differences in bactericidal ability between the control group and the IAA-94-treated group.
(F) Assay on bactericidal ability was analyzed by using GraphPad Prism 7. Data were analyzed by unpaired t test and presented as mean G SD, ***p < 0.001, n = 3.

OPEN ACCESS
4 iScience 23, 101328, July 24, 2020 iScience Article were confirmed in a quantitative manner by flow cytometry analysis ( Figure 2C). The results suggest that IAA-94-treated group had an engulfment capacity approximately 50% less than that of the control group ( Figure 2D). Understandably, the sequential processes of containment and killing of microbial pathogens are inseparable components of phagocyte-mediated defenses. Indeed, in bacterial killing assays, bactericidal capacity of hemocytes was greatly compromised after blockage of Cl À influx. In contrast to the basal control, 30 min post infection, bacterial survival in IAA-94-treated hemocytes starkly increased (Figures 2E and 2F). Therefore, these observations strongly implicate that Cl À influx is critical for phagocytic defense in oyster hemocytes through promoting capacity of engulfment and microbicidal activity.

Chloride-Dependent Engulfment Is Promoted by PI3K/Akt Signaling Pathway
As a classical signaling pathway for regulating phagocytosis, the PI3K/Akt signaling pathway has been shown to be activated by chloride influx, leading to the hypothesis that PI3K/Akt signaling pathway may be a key regulator that controls or promotes Cl À -dependent phagocytosis in oyster hemocytes. As anticipated, phagocytic capacity of oyster hemocytes was significantly boosted (1.53-fold) upon treatment of a PI3K activator (740Y-P, 30 mM) compared with the control. Contrarily, phagocytic capacity was suppressed to 51% when hemocytes were co-treated with a PI3K inhibitor (LY294002, 1 mM), thus implicating the PI3K signaling pathway as an essential regulator for phagocytosis in oyster hemocytes ( Figure 3A). More importantly, 740Y-P evidently rescued phagocytic impairments induced by IAA-94 in hemocytes, which raises the possibility of a functional interplay between PI3K/Akt signaling pathway and Cl À influx mediating phagocytosis ( Figure 3B). Furthermore, western blot analysis shows that the Akt phosphorylation increased substantially by 1.60-and 1.10-fold in infected and 740Y-P-treated oyster hemocytes, respectively, with respect to the control group. IAA-94 dampened Akt phosphorylation under both bacterial infection and untreated conditions, compared with the control group ( Figures 3C and 3D). Taken together, these results support the notion that Cl À -dependent phagocytosis is likely promoted by the PI3K/Akt signaling pathway.

Chloride Ion Flux Modulates Phagosomal Acidification and HOCl Production without Affecting Phagosomal-Lysosomal Fusion
Since coordinated Cl À ion flux is essential for pH regulation toward phagosomal acidification and catalytic production of HOCl, we examined whether these immunomodulatory roles of Cl À are mechanistically conserved in the Pacific oyster. pHrodo Red dye (a fluorescent pH sensor) conjugated particles were used to determine any fluctuations in phagosomal acidification, whose signals show a negative correlation between its fluorescence intensity with pH ( Figure S5). Following treatment of IAA-94 (100 mM), the intensity of phagosome-related fluorescence visibly declined at 15 and 30 min post phagocytosis (Figures 4A and 4B), which corroborates the assumption of Cl À ion flux as an important modulator for phagosomal acidification in oyster hemocytes. To test whether additional factors can impede Cl À -dependent bacterial clearance in hemocytes, the hemocytic capacity for HOCl biosynthesis was assessed by an in vivo HOCl assay, with a standard curve constructed to calibrate the amounts of TNB versus optical density (HOCl levels) being determined ( Figure S6). V. parahaemolyticus infection elicited 1.33-fold more HOCl, in comparison with the control group ( Figure 4C), but this was abolished in the presence of IAA-94, implicating Cl À ion flux as a significant regulator at work in infection-induced HOCl production. In addition, we also attempted to ask whether Cl À ion flux impacts phagosomal-lysosomal fusion, as it is a critical step toward phagosomal maturation and bactericidal activity. However, our results show no appreciable differences between IAA-94treated group and untreated group up to 60 min post phagocytosis ( Figure 4D).

CLIC2a Is a Primary Chloride Ion Channel within Oyster Genome
To illuminate the genetic basis of Cl À ion flux control in this ancient marine invertebrate, we set out to examine the gene families encoding the Cl À channel in oyster and perform homologs search-based alignment on a genomic scale. Intriguingly, only two members of the CLIC family were identified in the Pacific oyster genome, whereas all CTFR homologs are absent ( Figure 5A). A phylogenetic tree was then generated based on CLIC superfamily protein sequences from different species including Homo sapiens, Mizuhopecten yessoensis, Xenopus tropicalis, Danio rerio, Drosophila melanogaster, and Crassostrea gigas, which reveals that the CLIC family has highly diversified during evolution and considerably expanded in vertebrates, whereas the homologs of CTFR only prevail in vertebrate lineages ( Figures 5B and S7). Based on the results of the evolutionary tree, we observed that CLIC2 is the only highly conserved Cl À channel retained among ancient species such as Protostomia and Deuterostomia. Specifically, the ortholog of CLIC2 is duplicated in the genome of C. gigas, namely, CLIC2a and CLIC2b ( Figure 5C). Tissue distribution ll OPEN ACCESS iScience 23, 101328, July 24, 2020 5 iScience Article showed that CLIC2a is predominantly expressed in hemocytes ( Figure 5D), which suggests a key role of CLIC2a in regulating Cl À transportation during hemocytes phagocytosis.
To further clarify the roles of hemocyte CLIC2a in immunological contexts, we attempted to determine the subcellular localization of CLIC2a by immunofluorescence. The antibody was verified by western blotting to ensure that the antibody specifically binds to the antigen stated ( Figure 5E). Compared with non-immune lgG group, immunofluorescence showed that CLIC2a staining patterns were punctate or patchy within resting hemocytes, with evidently dense foci around the cell membrane in the resting hemocytes ( Figures  5F and S8). Moreover, we also examined the subcellular localization of CLIC2a upon phagocytosis. The results show an evident overlay between CLIC2a positive signaling and FITC-beads ( Figure 5F), implying phagosomal localization of CLIC2a in oyster hemocytes. 740Y-P is a potent cell-permeable PI3K signaling activator. LY294002 is a broad-spectrum inhibitor of PI3K signaling. (B) Phagocytosis data analysis was performed by using GraphPad Prism 7 software. Data were analyzed by one-way ANOVA followed by Tukey's post hoc test, where p < 0.05 was considered to be statistically significant, as denoted by different letters. Data are presented as mean G SD, n =3. (C) Western blot analysis on the effects of pharmacological intervention on bacterial infection: IAA-94 versus 740 Y-P with respect to the expression of proteins related to PI3K/Akt/mTOR signaling pathway. Levels of p-Akt/Akt was evaluated by western blot analysis. b-Actin served as a loading control.
(D) p-Akt/Akt expression was measured densitometrically by using ImageJ. The data are normalized to b-Actin. Data analysis was performed by using GraphPad Prism 7 software. Data were analyzed by one-way ANOVA followed by Tukey's post hoc test, where p < 0.05 was considered to be statistically significant, as denoted by different letters. Data are presented as mean G SD, n =3. To verify the function of CLIC2a as a Cl À channel in oyster hemocytes, RNAi experiments were performed to knock down expression of CLIC2a. Results from quantitative PCR suggest that relative expression of CLIC2a was reduced by about 60% with respect to the dsGFP control group ( Figure 6A), which is consistent with the knockdown efficiency at the protein level as verified by western blot (Figures 6B and 6C). As a test for validating Cl À flux, fluorescence intensity of MQAE in the dsCLIC2a group was shown to be much enhanced compared with that of the dsGFP group in hemocytes under condition of in vivo infection by live V. parahaemolyticus ( Figures 6D and 6E). Collectively, these observations confirm that CLIC2a is a conserved and functionally relevant Cl À channel in oyster hemocytes.
As CLIC2a is known to serve as a Cl À channel that controls the flow of Cl À ion into the cytosol, it is reasonable to conclude that CLIC2a resides on the hemocytic cell membrane to exercise this function. Given the abundance of CLIC2a in hemocytes ( Figure 5D), we reasoned that CLIC2a protein is responsible for Cl À -  Figure 7A) shows that CLIC2a knockdown resulted in a reduction of 32% in phagocytic capacity of hemocytes compared with the control group ( Figure 7B). Bacterial survival counts of infecting V. parahaemolyticus rose from 3.98 3 10 4 to 8.55 3 10 4 CFU following RNAi of CLIC2a ( Figure 7C). Furthermore, knockdown of CLIC2a in hemocytes clearly compromised their ability to acidify phagosomes (Figures 7D and 7E), which likely has adverse consequences for bacterial clearance. Overall, these results implicate RNAi of CLIC2a with impaired hemocytic phenotypes in phagocytosis, phagosomal acidification, and microbial killing, thus supporting an essential role of CLIC2a in Cl À -mediated innate immunity in oysters.

DISCUSSION
Chloride ion is a ubiquitous yet indispensable constituent in phagocyte biology. It is responsible for mediating numerous physiological functions including maintenance of membrane potential, regulation of phagosomal pH, promotion of phagosomal enzymatic activities, and production of Cl À -containing oxidants as antimicrobial agents (Wang, 2016). Chloride ion channels (CLICs) have been intimately linked to inflammatory diseases, inspiring enduring interest in their immune roles (Gururaja Rao et al., 2020). Despite substantial advances in vertebrates, mechanistic understanding on the immune function of Cl À in invertebrates remains scant. Here, we have revealed for the first time that coordinated phagocytic events associated with Cl À flux occur after infection and that an under-examined Cl À channel, CLIC2a, plays vital roles in promoting hemocyte phagocytosis and bactericidal activity of C. gigas, a representative species in lower marine invertebrates. (C) CLIC2a expression was measured densitometrically by using ImageJ. Data analysis was performed by using GraphPad Prism 7 software. Data were analyzed by unpaired t test and presented as mean G SD, **p < 0.01, n = 3.
(D) Confocal micrographs show the fluorescence intensities of the dsCLIC2a and dsGFP groups of Vibrio-exposed hemocytes. Scale bar: 15 mm.
(E) Data analysis was performed by using GraphPad Prism 7 software. Fluorescence intensities were normalized by cell number and volume of hemocytes. Data were analyzed by unpaired t test and presented as mean G SD, ***p < 0.001, n = 3.

OPEN ACCESS
iScience 23, 101328, July 24, 2020 9 iScience Article Functional diversity abounds in the regulation and direction of Cl À flow/flux across different cell types. For instance, intracellular Cl À concentration was markedly elevated in human airway epithelial BEAS-2B cells upon challenge with LPS (lipopolysaccharide) from P. aeruginosa . In a similar case, [Cl À ] i of resting oyster hemocytes is kept at a low level but surges upon infections. In contrast, Cl À efflux ensued when human neutrophils were exposed to Candida albicans opsonized with various soluble factors (Busetto et al., 2007). The reason for the different flow directions of Cl À is uncertain and may concern how different cells interact with exogenous danger signals. Regardless, productive acidification of phagosomes is invariably a mandatory requirement for proper phagocytic functions. Phagosomal pH is primarily controlled by activity of H + -pumping vacuolar-type ATPases (V-ATPase) and importation of anions (Kissing et al., 2018). Proton pumping, which depresses luminal pH, generates an electrochemical potential difference across membranes (positive charge on the luminal side). If left unchecked, this driving force would begin to decelerate and eventually inhibit further acidification (Grabe et al., 2000;Kettner et al., 2003).
To prevent premature inhibition, the generated voltage is offset by import of anions (Cl À ) and/or export of cations (Kissing et al., 2018). As a result, inappropriate or insufficient Cl À flux can be an important cause of impaired phagosomal acidification. With the efflux of Cl À ions, neutrophil phagosomes initially alkalinize and slowly acidify to a moderate level (El Chemaly et al., 2014;Jankowski et al., 2002). Additionally, ClC3 chloride channel is reportedly expressed in neutrophils and their phagosomes, modulating phagosomal pH (Scheel et al., 2005). Deletion of CFTR in murine alveolar macrophages caused defective phagosomal (B) Data analysis for hemocytes phagocytosis was performed by using GraphPad Prism 7 software. Data were analyzed by unpaired t test and presented as mean G SD, *p < 0.0, n = 3. (C) Relative to the control group, the bactericidal ability of the dsCLIC2a group decreased. Assay on bactericidal ability of the dsCLIC2a group and the dsGFP control group was analyzed by using GraphPad Prism 7 software. Data were analyzed by unpaired t test and presented as mean G SD, *p < 0.05, n = 3. (D) Acidification defects in phagosomes following particle engulfment in oyster hemocytes silenced by dsCLIC2a, in comparison with dsGFP group. Cells that had ingested pHrodo Red zymosan were observed by confocal microscopy. Scale bar: 5 mm.
(E) Fluorescence emission was calibrated and a fluorescence index was obtained from the samples (n = 3 groups, each group contains 10-15 cells). Bars represent mean G SD.
An acidic environment exerts modulatory effects on the structural maturation of hydrolases and denaturation of their protein substrates, which ultimately increases the rates of killing engulfed bacteria (Pillay et al., 2002). For complementing this mode of antibacterial defense, phagocytes also use a combination of oxidative mechanisms, including the production of superoxide (O 2 d-) and hypochlorous acid (HOCl), to eradicate phagocytosed microorganisms (Kobayashi et al., 2000). Superoxide produced in phagocytes is unstable and has only mild antimicrobial potential. Instead of directly leveraging O 2 d-, phagocytes convert it to hydrogen peroxide (H 2 O 2 ), a long-lasting oxidant with modest antimicrobial activity at millimolar level (Bonvillain et al., 2011;Lymar and Hurst, 1995). Subsequently, H 2 O 2 is turned into the much more potent HOCl via MPO (Bonvillain et al., 2011). Evolutionarily, regulated production of ROS such as HOCl represents a major advance in phagocytic innate immunity (Albrich et al., 1981). In this study, we found that oysters have evolved an efficient machinery to produce these oxidizing biocides. Under conditions of bacterial stimulation, production of HOCl increased, whereas blocking Cl À flux resulted in its inhibition.
Apart from implications in ROS production, Cl À flux seems to mediate hemocytic functions in PI3K/Akt-dependent manners. Previous research findings have shown that PI3K plays an indispensable role in phagosome formation and maturation. During initial phagocytosis, PI3K controls actin depolymerization to drive the formation of a phagocytic cup that surrounds a foreign body and bring it into the phagosome (Olazabal et al., 2002;Tuxworth et al., 2001). Pharmacological inhibition of PI3K could impair the ability of phagocytosis (Araki et al., 1996(Araki et al., , 2003Cox et al., 1999). Moreover, PI3K/Akt/mTOR signaling regulates phagosome maturation by modulating microtubule-based motor activity to modulate events of lysosome trafficking and fusion in leukocytes (Saric et al., 2016). Consistently, our study confirms that PI3K/Akt signaling in the regulation of phagocytosis is conserved from oyster to mammals. More importantly, the PI3K/Akt pathway is crucial to mediating Cl À fluxdependent phagocytosis, which is supported by the evidence that the PI3K activator 740Y-P could rescue blockage of phagocytosis caused by chloride channel inhibitors IAA-94. However, the chloride channel inhibitors IAA-94 had no effects on the fusion of lysosomes and phagosomes as reported in macrophages , which seems to contradict with the anticipated roles of PI3K/Akt in phagosome maturation. One possibility is that Cl À flux may activate other signaling pathways or regulatory mechanisms of intracellular feedback to brake on the fusion of lysosomes and phagosomes, which remains to be verified.
In this study, phylogenetic analysis with functional validation showed that CLIC2a is the primary Cl À channel in oyster, and we thoroughly investigated the potential roles of the Cl À channel CLIC2a in innate immunity in oyster hemocytes. CLIC2a was identified as a CLIC family member protein with the highest expression levels in the oyster. CLICs are highly conserved in chordates with six vertebrate paralogs, from CLIC1 to CLIC6. According to the phylogenetic tree, only CLIC2 exists in invertebrates and expands to CLIC2a and CLIC2b. From an evolutionary perspective, CLIC2 appears to be the only one Cl À channel with function conserved from an ancestral progenitor of Protostomia and Deuterostomia. Subcellular localization of one specific protein is generally associated with its function and also provides clues on regulatory mechanisms. In mammalian macrophages, CLIC1 occurs as spots in the cytoplasm and is translocated to the phagosomal membrane during phagocytosis . In contrast, CLIC2a in oysters is mainly localized in the cellular membrane but could be translocated to the phagosomal membrane in manners resembling macrophage CLIC1. Given that composition-wise the phagosomal membrane originates from the plasma membrane, membrane trafficking may be one possible route for CLIC2a translocation. Furthermore, knocking down CLIC2a evidently diminished Cl À flux and disrupted Cl À -dependent phagocytic functions in oyster hemocytes, including phagocytosis, phagosomal acidification, and bacterial killing. Beyond the current scope of investigation on phagocytic immunity, it is quite likely that CLIC2a performs other biological functions central to cell physiology in vertebrates and invertebrates alike. Future studies on novel CLIC2a functions and their underlying mechanisms are thus warranted.

Limitations of the Study
In this study, we have explored the subcellular localization and function of CLIC2a mainly in bacterial infection contexts. Further in-depth characterization of the exact roles of CLIC2a in governing the downstream of PI3K/Akt or other intracellular signaling pathways in immune defense and how CLIC2a could be activated would be pertinent.

Resource Availability Lead Contact
Further information and requests for resources should be directed to and will be fulfilled by the Lead Contact, Yang Zhang (yzhang@scsio.ac.cn).

Materials Availability
All unique/stable reagents generated in this study are available, on reasonable request, from the Lead Contact on a completed Materials Transfer Agreement.

Data and Code Availability
All data supporting the findings of this study are included in the article and its Supplemental Information or are available from the corresponding authors on request.

METHODS
All methods can be found in the accompanying Transparent Methods supplemental file.

ACKNOWLEDGMENTS
Supplemental Table   Table S1. Summary of primers used in this work, related to Figure 6  Oysters in the challenged group were injected with a 100-μL suspension into the adductor muscle, while oysters in the control group were injected with an equal volume of PBS. After injection, oysters were returned to separate tanks for subsequent incubation or sampling.
Hemolymph was drawn from the posterior adductor muscle of C. gigas by using a sterile 1-mL syringe. Hemolymph was kept and immediately centrifuged at 200×g for 10 min to collect hemocytes, then the supernatant was sterilized through a 0.22 µM filter (Sangon Biotech, F513134). Hemocytes were incubated in each petri dish with hemolymph serum at 27°C for subsequent analyses.

Cell viability test
Approximately 1×10 6 cells were cultured in 12-well plates with filtered plasma and incubated at different temperatures. CellTiter-Lumi™ Luminescent Cell Viability Assay Kit (Promega, USA) was used to quantify ATP content, which is one marker for living cells.
According to the manufacturer's protocol (Beyotime, C0065S), treated hemocytes were equilibrated at room temperature for 10 min for the next operation. 50 μL CellTiter-Lumi ™ luminescence detection reagent was added to each well of a 96-well microplate. The 96-well microplate was gently shaken at room temperature for 2 min to promote cell lysis. Lysed cells were incubated at room temperature (about 27°C) for 10 min to stabilize luminescence signals.
A multi-functional microplate reader was used for chemiluminescence detection of samples.
Relative viability of cells was directly calculated based on chemiluminescence readings.
In Trypan blue staining assay, hemocytes were stained with a standard Trypan blue solution (0.04% in PBS) for 5 min. Cells were enumerated under a light microscope with a hemocytometer, where blue-colored cells were considered to be dead. All experiments were performed in at least triplicates.

Determination of [Cl -]i
This part of the study was conducted as previously described with slightly modification . Briefly, levels of intracellular Clconcentration ([Cl -]i) were measured by using the Clspecific fluorescent probe MQAE (5 mM, Beyotime, S1082), a selective chloride ion indicator. Upon binding halide ions such as chloride, MQAE fluorescence is quenched, resulting in a decrease in fluorescence intensity without a shift in wavelength (Ikeuchi et al., 2018).300 µL of hemocytes were cultured in glass bottom dishes at 27°C for 15 min. At the end of appropriate treatments, oyster hemocytes were washed twice with Cl --free Tyrode solution resuspended to a final density of 1.0 × 10 7 CFU per mL for subsequent phagocytosis assays (Duperthuy et al., 2011;. Hemocytes cultured in a 24 well plate and the determined reagent were included in the buffer to incubate the cells for 30 minutes at room temperature and washed 3 times with PBS. Next, hemocytes were incubated with the prepared bacteria for 15 min at a MOI (multiplicity of infection) of 50. Cell were washed with Tris buffer (pH 8.0, 50 mM, Sangon Biotech, A610195) for 3 times to remove unbound bacteria and then suspended in PBS supplemented with 1.5‰ EDTA. Trypan blue was used to inhibit further attachment of hemocytes to bacteria (Guckian et al., 1978). Finally, flow cytometry analysis by Guava® easyCyte™ was performed to quantify phagocytosis in oyster hemocytes. Cells taking up E. coli were recognized by RFP reporter fluorescence, which provided an indication of uptake capacity (proportional to the number of bacteria retained). Phagocytosis by hemocytes was monitored with at least a total of 10,000 events. Data were analyzed with the FlowJo software (version V10).

Bacterial clearance assay
This assay was performed as described previously with minor modifications (Saleh et al., 2006). For the experiments, two strains of bacteria, Escherichia coli (DH5α) and V.parahaemolyticus (ZJ51) (of a working density at OD600nm = 0.2), were cultured at 37°C. After incubation, bacteria were harvested by centrifugation at low speed, followed by washing 3 times with Tris buffer (50 mM, pH 8.0, Sangon Biotech, A610195) and resuspension in 1 ml PBS for subsequent assays. Approximately 2×10 5 hemocytes per well was cultured in a 24-well plate and subsequently challenged with preprocessed bacteria at an MOI (multiplicity of infection) of 50 at room temperature. Cells were then briefly treated with 0.02% trypsin-EDTA for 4 times to remove extracellular bacteria. Subsequently, IAA-94 (100 μM) and DMSO (vehicle) were added to the filtered body fluids (except for the knockdown group) to pretreat cells for 30 min. After 30 min to kill the internalized bacteria, hemocytes were lysed in 1 mL PBS containing 0.05% Triton X-100. Finally, 100 µL of the lysate was used as an inoculum on LB agar plates for enumerating bacterial colonies. For each group, three wells were used to perform the bacterial clearance assay and each experiment was independently repeated 3 times.

RNA interference (RNAi)
To clarify the functional relevance of CLIC2α in oyster hemocytes, CLIC2α gene was knocked down in vivo via dsRNA-mediated RNA interference. The primers used to synthesize dsRNA are as shown in Table S1. A CLIC2α cDNA fragment and a GFP cDNA fragment (negative control) were amplified with primer pairs with T7 promoter overhangs (Promega, RiboMAX™ Express RNAi System). PCR products thus resulted were used as templates to synthesize dsRNA according to the manufacturer's instructions. Ten oysters were randomly assigned into 2 groups and placed in 2 tanks: the treatment and control groups. Each oyster was injected with 50 µg dsRNA and three individuals from each group were chosen randomly for the collection of hemocytes. Phagocytosis rate, bacterial clearance rate and degree of phagosomal acidification were evaluated 3 days after dsRNA injection. The expression level of CLIC2α was then determined by RT-qPCR and Western blot.

Total RNA extraction and quantitative real-time PCR analysis
Hemocytes were collected as above and total RNA was isolated with TRIzol Reagent  Table S1. All experiments were performed in triplicates by using GAPDH mRNA as an internal control. Analysis of the dissociation curve of the amplification products was constructed to confirm specificity at the end of each PCR. Relative gene expression was calculated by using the 2 -ΔΔCt method. Data was represented by using SPSS10.0 statistical software, and significance between two groups was determined by Student's t-test. followed by incubation for 10 min. Next, 50 µL of the TNB reagent/standard was added to the samples which were left to stand for an additional 10 min. Subsequently, absorbance at 412nm (A412) was detected by means of an EnSightTM multimode plate reader (PerkinElmer, USA).

Determination of phagosomal-lysosomal fusion
Numerical data were obtained by comparing absorbance measurements against the standard curve.

ORF cloning and bioinformatics analysis
Partial cDNA sequence of C. gigas was BLAST searched in the oyster genome library

Antibody production
The two amino acid sequences 1-129 a.a. and 201-292 a.a. in the ORF protein of CLIC2α were selected for protein expression and purification as antigens in mixed immunization. White rabbits are used as experimental animals for immunization. Multiple injections of a total amount of 600 µg of antigens and complete Freund's adjuvant were injected subcutaneously on the back of the rabbit. Injection proceeded once every two weeks, followed by evaluation of serum titers after four injections. Pre-sera were used as a negative control for antibody titer testing.
Rabbits were bled after passing the test. Then, protein Affinity purification, HABP affinity purification, and antigen affinity purification were performed successively to purify desired sera.
The resultant purified antibodies were validated in immunoblotting and stored for subsequent experiments.

Immunofluorescence preparation
For imaging, hemocytes were fixed with cold 4% paraformaldehyde for 10 min, rinsed 3 times in PBS and permeabilized with 0.05% Triton in PBS for 15 min at room temperature. Then, hemocytes were blocked with QuickBlock™ blocking buffer (Beyotime, P0252) for 60 min, incubated overnight with appropriate primary antibodies at 4°C. CLIC2α was stained with an affinity purified rabbit polyclonal antibody (Sinobiological). After washing off primary antibodies, specimens were incubated in a fluorochrome-conjugated secondary antibody (Cell Signaling Technology, #4414) diluted in antibody dilution buffer for 1 h at room temperature in dark. DAPI staining was done to counterstain nuclei (Sigma, 1.5 µg/mL), followed by washing in PBS for 3 times. The hemocyte samples thus prepared were then immersed in 200 μL PBS and visualized under a Leica SP8 confocal microscope.
Experimentally, cells were allowed to engulf red-emission zymosan and fluorescent E. coli for a 30-min period to ensure that the majority of the immunostimulants have entered the phagosomal compartment. Fluorescence micrographs were acquired by using a Leica SP8 confocal fluorescence laser scanning system (Leica Microsystems, Germany). Lasers were used at the following excitation/emission parameters: λex 405 nm for MQAE (λem 460 nm), λex 560 nm for pHrodo™ Red zymosan (λem 585 nm), λex 405 nm for DAPI (λem 454 nm), λex 577 nm for Lyso-Tracker Red (λem 590 nm) and λex 555 nm for RFP (λem 584 nm). Series of optical sections were collected and processed by using Image-Pro Plus 6.0. For analysis of imaging data, the bright field and fluorescence channels of representative images were superimposed to confirm cell morphological features and phagocytosis events. Each image regions of interest (ROIs) was selected in Image Pro Plus, corresponding to individual phagocytosis events or whole cells subjected to a particular treatment. To avoid bias, all individuals in the recording optical field were processed for data analysis except for: (1) cells with marked morphological alterations; or (2) aggregating cells. Typically, up to 5-20% of such visually deviant cells may be excluded from analysis .

Statistical analysis
Data processing and statistical analyses were performed by using GraphPad Prism (version 8.0.1). All statistical values were expressed as the mean ± S.D., with the number of experiments (n) in parentheses. Significance between groups was determined by Student's t-