Conserved Role of mTORC1 Signaling in B Cell Immunity in Teleost Fish Free

Bcells appeared ∼500 million years ago in early vertebrates (e.g., cartilaginous fish) and represent the most important components of the adaptive immune system in vertebrates. Three different Ig isotypes have been identified in teleost fish, including IgM, IgD, and IgT/IgZ, which constitute three major B cell subsets (IgM⁺/IgD⁺, IgM⁻/IgD⁺, and IgM⁻/IgT⁺) (1–4). IgM⁺/IgD⁺ B cells (a subset that coexpresses surface IgM and IgD) were found to be the most dominant B cell subset in blood, the peritoneal cavity, and systemic organs of teleost fish (head kidney and spleen). Analogous to mammalian B-1 cells, teleost IgM⁺/IgD⁺ B cells can locally proliferate in systemic organs and secrete specific Igs in adaptive response to bacterial or parasitic infection (5, 6). Additionally, IgM⁺/IgD⁺ B cells also possess innate immune functions, such as phagocytosis and microbicidal activities (7). IgM⁻/IgD⁺ B cells are a newly discovered B cell lineage that has so far only been characterized in catfish peripheral blood and rainbow trout gill and gut mucosa (8). Another newly discovered IgM⁻/IgT⁺ B cell subset is generally thought to be a mucosa-associated B lymphocyte that specializes in mucosal immunity (9). Recent studies have shown that this B cell subset in peripheral blood also exhibits phagocytic and intracellular killing activities (10).

The mechanistic target of rapamycin (mTOR) is an evolutionarily conserved protein (280 kDa) with high identity rates (∼40–60%) across all eukaryotes (11). mTOR associates with different ligands to form two complexes: mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2). mTORC1 phosphorylates multiple substrates, such as S6 ribosomal protein (S6) kinase and eukaryotic translation initiation factor 4E-binding protein 1 (4EBP1), to regulate protein synthesis and cell growth as well as immunity (12–14). Importantly, mTORC1 is sensitive to rapamycin (RAPA) treatment. Unlike mTORC1, mTORC2 regulates cellular survival by phosphorylating AKT on Ser⁴⁷³, and its activity is not sensitive to RAPA treatment and can only be inhibited with prolonged RAPA exposure (13, 15, 16). Most studies on mTORC1 regulating immune functions have mainly focused on its role in T cells, whereas its role in B cells remains largely uncharacterized (17, 18). Previous studies in mammals have indicated that mTORC1 signaling is associated with B cell differentiation. Particularly, reductions in mTORC1 can decrease B cell development or differentiation in the spleen, thus reducing Ab production in response to Ags (17). Recently, Ye et al. (19) found that RAPA treatment predominantly inhibited B cell responses as well as serum Ab titers during viral infection. To date, the mechanisms of B cell regulation via mTORC1 signaling have only been identified in mammals. Considering the evolutionary conservation of mTORC1 in all eukaryotes, understanding whether and how mTORC1 regulates B cell responses in nonmammalian models is of critical importance. Given that teleost fish represent the most ancient bony vertebrates with bona fide B cells, coupled with the evolutionary conservation of mTORC1 between mammals and teleost fish, we hypothesized that mTORC1 in both modern and primitive bony vertebrates must have evolved analogous molecular mechanisms to regulate B cell immune responses and maintain homeostasis.

To address this hypothesis, rainbow trout (Oncorhynchus mykiss) was used as an early vertebrate model to investigate whether and how the mTORC1 signaling mediates teleost B cell immune responses. Our findings indicated that RAPA treatment inhibited the mTORC1 signaling, as evidenced by the downregulation of mTOR, S6, and 4EBP1 phosphorylation in B cells. Moreover, several key B cell immune responses and physiological phenomena, including cell phagocytosis, proliferation, Ab production, and apoptosis, were significantly affected after RAPA treatment, thus providing important evidence of the regulatory mechanisms of mTORC1 in B cell immune and physiological responses in a nonmammalian model. More importantly, our findings indicated that the production of F. columnare–specific titers decreased in serum after RAPA treatment, and these fish were significantly more susceptible to F. columnare reinfection. Our results thus elucidated a previously unrecognized conserved role of mTORC1 signaling in modulating B cell responses in teleost fish. From an evolutionary perspective, our study fills important knowledge gaps regarding the functions of both mTORC1 signaling and B cell lineage and thus provides key insights into the evolution of B cell–mediated immune responses in early vertebrates.

Healthy rainbow trout (O. mykiss) were bought from a fish farm in Chengdu (Sichuan, China) and then acclimated in a recirculating water system in the wet laboratory of Huazhong Agricultural University for over 2 wk before the formal experiment. Water temperature was kept at 16 ± 1°C. Fish were fed with a commercial diet at a ratio of 1% body weight per day, and the feeding was terminated at 2 d before sacrifice. All of the fish care and experimental procedures were approved by the Animal Experiment Committee of Huazhong Agricultural University (permit number HZAUFI-2016-007).

Rainbow trout weighing 450 ± 50 g were used for in vitro RAPA treatment studies. Splenic leukocytes were suspended with DMEM (supplemented with 20% FBS, 100 U/ml penicillin, 100 mg/ml streptomycin, 200 mg/ml amphotericin B, and 250 mg/ml gentamicin sulfate) and dispensed into 48-well plates with 200 μl in each well at a final density of 2 × 10⁶ cells/ml. RAPA (Sigma-Aldrich, St. Louis, MO) was dissolved in DMSO at 18.3 mg/ml and diluted with PBS before use. The final application concentration of RAPA is 100 nM in the RAPA group, and the same volume of DMSO was diluted with PBS in the DMSO control group. Thereafter, all plates were placed in a cell incubator with 5% CO₂ at 17°C. At the designed time points, leukocytes and cellular supernatant were obtained after centrifugation at 400 × g for further analysis.

Rainbow trout weighting 10 ± 2 g were used for in vivo RAPA injection studies as well as the following F. columnare infection experiment. Rainbow trout was i.p. injected with 183 ng RAPA/g of fish or the same volume of DMSO consecutively for the first 7 d according to previous studies (18). Then spleen and serum were sampled for analysis in the following experiments. For the infection experiment, the F. columnare infection was performed 1 d after the i.p. injection of RAPA or DMSO. Thereafter, fish were injected with RAPA or DMSO at a total dose of 2.9 mg/kg (once a day for 6 d [6 times] plus every 2 d for the next 21 d [10 times]), and fish were anesthetized at each time point.

The leukocytes were isolated from trout spleen as described previously (20) with sight modifications. Briefly, trout spleen was pestled in DMEM (Life Technologies, Carlsbad, CA) supplemented with 5% FBS and then passed through 100-μm nylon cell strainers (BD Biosciences, San Jose, CA) to remove the large unsuspended particulates. Thereafter, the obtained splenocyte suspensions were placed into 34–51% Percoll (GE Healthcare) discontinuous density gradient and centrifuged at 400 × g for 30 min at 4°C. The splenic leukocytes (SPLs) at the interface were collected and washed with DMEM for further analysis.

MACS was adopted to sort IgT⁺ and IgM⁺ B cells from the SPLs after DMSO or RAPA treatment. Briefly, leukocytes were collected and stained with mouse anti-trout IgT (clone 41.8; isotype IgG2b) mAb (2 μg/ml) or mouse anti-trout IgM (clone 1.14; isotype IgG1) mAb (2 μg/ml) at 4°C for 30 min. After washing twice with sterile PBS, 1 × 10⁷ cells was resuspended in 80 μl labeling buffer (PBS supplemented with 2 mM EDTA and 0.5% BSA, pH 7.4), and then 20 μl microbeads conjugated with anti-mouse IgG (H+L) were added into the buffer for coincubation at 4°C for 15 min. After washing twice with sterile PBS, the cells (up to 10⁸ cells) were resuspended in 500 μl labeling buffer and then applied to the MS column, which was attached to the Mini-MACS separator. Thereafter, the cell suspension was applied onto the column and flow-through containing unlabeled cells harvested, and the column was washed by adding 1 ml labeling buffer and collecting the flow-through. After washing three times, IgT⁺ or IgM⁺ cells were retained within the column, 1 ml of labeling buffer was pipetted onto the column, and the magnetic labeled cells immediately flushed out by firmly pushing the plunger into the column; this step was repeated once, with the collected flow-through being flushed out in the labeling buffer (which contains IgT⁺ or IgM⁺ cells), whereas IgT⁻ or IgM⁻ cells were collected in the flow-through fraction. All of the cells were separately collected for further analysis.

RNA extraction and gene expression analysis of cells were performed as described previously (5). Briefly, the purity and concentration of the extracted RNA were carried out by spectrophotometry (NanoPhotometer NP80 Touch), and the integrity of the RNA was determined by agarose gel electrophoresis. The quantified RNA samples were then used for cDNA synthesis (Invitrogen, Carlsbad, CA). Thereafter, real-time PCR (RT-PCR) was performed on a 7500 qPCR system (Applied Biosystems) with the cDNA using the EvaGreen 2× qPCR Master Mix (Yeasen). The RT-PCR program was performed under following conditions: 94°C for 30 s, followed by 20–45 cycles at 58°C for 30 s, and 72°C for 30 s. The primers used for RT-PCR are listed in Supplemental Table II.

Genomic DNA (gDNA) from bacterial pellets was extracted and purified using the DNeasy tissue kit and associated protocol (Qiagen, Valencia, CA). RNase (Qiagen) was used to remove RNA from DNA samples. DNA yield and purity were determined spectrophotometrically (NanoPhotometer NP80 Touch). The purified gDNA was stored at −20°C until use. gDNA standard curve was generated by plotting threshold cycle (Ct) values (y-axis) against the log₁₀ gDNA concentrations of F. columnare (x-axis). The standard curve revealed a linear correlation between Ct values and log amount of nucleic acid (Ct = 38.343–3.1662 × log₁₀ gDNA concentrations; R² = 0.99).

To detect the ratio of different B cell populations in the splenic lymphocytes, leukocyte suspensions were harvested and double stained with mouse anti-trout IgM (1 μg/ml) and anti-trout IgT mAbs (1 μg/ml) on ice for 1 h. After washing three times with DMEM (supplemented with 5% FBS, 100 U/ml penicillin, and 100 mg/ml streptomycin), the stained cells were further incubated with PE-goat anti-mouse IgG1 (2.5 μg/ml; BD Biosciences) and allophycocyanin–goat anti-mouse IgG2b (2.5 μg/ml; BD Biosciences) on ice for 30 min. After washing three times with DMEM (supplemented with 5% FBS, 100 U/ml penicillin, and 100 mg/ml streptomycin), analysis of stained leukocytes was performed with a CytoFLEX LX flow cytometer (Beckman Coulter).

Western blotting analysis was conducted to determine the inhibition of mTORC1 signaling pathway and the secretion level of Igs. The supernatants of intracellular or secreted proteins were resolved on 4–15% SDS-PAGE gel (Thermo Fisher Scientific) under reducing or nonreducing conditions, respectively. Then, the gels were transferred onto a polyvinylidene fluoride (PVDF) membrane and blocked in PBS with 5% skim milk. Thereafter, the PVDF membrane with intracellular proteins were incubated with anti-mTOR, anti-S6, anti-4EBP1, anti–phospho-mTOR S2448, anti–phospho-S6 Ser^240/244, or anti–phospho-4EBP1 Thr^37/46 monoclonal rabbit Ab (Cell Signaling Technology) for 1 h. After washing four times, goat anti-rabbit IgG was added and incubated for 45 min at room temperature. Similarly, the PVDF membrane with secreted proteins were incubated with mouse anti-trout IgM mAb (0.2 μg/ml), biotinylated mouse anti-trout IgD (clone 3A6; isotype IgG1) mAb (0.2 μg/ml), or rabbit anti-trout IgT polyclonal Ab (pAb; 0.2 μg/ml). After washing four times, HRP-conjugated streptavidin, goat anti-rabbit IgG, or goat anti-mouse IgG was added and incubated for 45 min at room temperature. After washing four times, the reaction bands were visualized with the ECL Substrate (Bio-Rad Laboratories) and then scanned by Amersham Imager 600 Imaging System (GE Healthcare). The densitometry of the target band was analyzed with ImageQuant TL software (GE Healthcare).

The former obtained SPLs were coated on the slides, fixed with methanol for 15 min, and blocked with blocking buffer (Thermo Fisher Scientific) for 30 min at room temperature. Then, they were stained with specific mouse anti-trout IgM or anti-trout IgT mAb (1 μg/ml each) for 1 h, respectively. After washing three times, Cy3 goat anti-mouse IgG1 and Alexa Fluor 647 goat anti-mouse IgG2b were added and incubated for 45 min to stain IgM⁺ and IgT⁺ B cells, respectively. After washing four times, cells were incubated with DAPI (1 mg/ml; Invitrogen) before mounting with fluorescent microscopy mounting solution.

Both flow cytometry and immunofluorescence were conducted to determine the phagocytosis of B cells. Firstly, the isolated SPLs were cultured with 100 nM RAPA or DMSO for 1 d and then harvested after resuspending in DMEM. The cell suspensions were coincubated with 1.0-μm fluorescent beads (Polysciences) with the cell/bead ratio of 1:10 in a cell incubator with 5% CO₂ at 17°C for 3 h. After that, cell suspensions were placed onto a cushion of 3% BSA in PBS supplemented with 4.5% d-glucose and then centrifuged at 100 × g for 10 min at 4°C. The resultant cells were harvested, washed two times, and then used for further flow cytometry analysis and immunofluorescence assay. Flow cytometry analysis followed the aforementioned procedure. For the immunofluorescence assay, cells were stained with mouse anti-trout IgM or anti-trout IgT mAb (1 μg/ml each) and then stained with allophycocyanin goat anti-mouse IgG1 or IgG2b for 45 min, respectively. After washing four times, the cells were incubated with DAPI (1 mg/ml; Invitrogen) before mounting with fluorescent microscopy mounting solution. All cells were acquired and analyzed using a BX53 fluorescence microscope (Olympus) and the iVision-Mac scientific imaging processing software (Olympus).

Annexin V–FITC was used to detect the percentage of apoptotic B cells after RAPA treatment. Briefly, SPLs obtained from in vivo or in vitro were double stained with mouse anti-trout IgT and anti-trout IgM mAb (1 μg/ml each) and then stained with allophycocyanin goat anti-mouse IgG2b and PE goat anti-mouse IgG1 (2.5 μg/ml each; BD Biosciences), respectively. After washing two times, Annexin V–FITC (Beyotime Biotechnology) was then added to the cell resuspension, incubated for 10 min on ice, and used for further flow cytometer analysis.

A total of 20 μg or 200 μg 5-ethynyl-2′-deoxyuridine (EdU; Invitrogen) was added in the SPLs culture media in vitro or injected into trout caudal vein in vivo for the analysis of B cell proliferation, respectively. After 1 d, SPLs were harvested and then stained with mouse anti-trout IgM or anti-trout IgT mAb for 1 h on ice. After washing three times, Alexa Fluor 488 goat anti-mouse IgG2b or Alexa Fluor 488–goat anti-mouse IgG1 (2.5 μg/ml; Invitrogen) were added and incubated for 45 min on ice. The proportion of EdU⁺ cells was determined following the manufacturer’s instructions, acquired, and analyzed using a CytoFLEX LX flow cytometer (Beckman Coulter).

For the visualization of proliferated B cells, IgT⁺ and IgM⁺ B cells were sorted using MACS method as mentioned above. Thereafter, the cell suspension was transferred onto slides and then fixed with methanol for 15 min. After blocking with the blocking buffer (Thermo Fisher Scientific) for 30 min at room temperature, cells were incubated with Cytoperm Permeabilization Buffer Plus (BD Biosciences) for 10 min at room temperature. After washing four times, cells were incubated with DAPI (1 mg/ml; Invitrogen) before mounting with fluorescent microscopy mounting solution. The EdU⁺ cells and other cells were acquired and analyzed using a BX53 fluorescence microscope (Olympus) and the iVision-Mac scientific imaging processing software (Olympus).

F. columnare G₄ strain labeled with GFP was kindly provided by Prof. Pin Nie (Institute of Hydrobiology, Chinese Academy of Sciences). Before the immersion infection, fish were maintained in 30-liter tanks supplied with 0.2 l/min of air-pumped water at 16°C for 1 wk as previously described (21). Next, GFP–F. columnare was resuscitated from −80°C and cultured in Tryptone yeast extract salts agar medium at 28°C. The third passage of the culture of the bacterial inoculum (OD₅₄₀ = 0.5; ∼10⁶ CFU/ml) was used for further infection experiment. During the infection period, the water flow of each aquarium was stopped. After exposure to the bacterial medium in a static bath for 4 h, the circulate water flow in rainbow trout rearing tanks was resumed. Challenge experiments were also performed with the same dose and method.

To evaluate whether the mTORC1 signaling pathway affects the binding of Igs to F. columnare G₄ in vivo, the titers of F. columnare–specific Igs in serum were detected using a pulldown assay as previously described (22). Briefly, 40 μl F. columnare G₄ suspension (4 × 10⁸ CFU/ml) were incubated with serum from infected or control fish injected with DMSO or RAPA at 4°C in a 300 μl volume with PBS containing 1% BSA (pH 7.2) with continuous shaking overnight. Subsequently, the F. columnare G₄ was washed with PBS for three times, and the binding Igs were eluted with 2× Laemmli Sample Buffer (Bio-Rad Laboratories) and then detected by Western blot as clarified above.

To further study the regulatory roles of mTORC1 signaling pathway on the agglutination and growth of F. columnare G₄, F. columnare G₄ was firstly cultured in Tryptone yeast extract salts agar medium at 28°C. Then, 100 μl third generations of the bacterial inoculum (OD₅₄₀ = 0.5; ∼10⁶ CFU/ml) was incubated with 100 μl serum (1:5 dilution) from naive fish and the 28DPI-DMSO and 28DPI-RAPA groups for 10 min at 16°C, respectively. The agglutination of F. columnare G₄ was detected by immunofluorescence. For evaluating the growth of F. columnare G₄, 50 μl third generations of F. columnare G₄ (initial OD₅₄₀ = 0.3) were incubated with 50 μl serum (1:5 dilution) from naive fish and the 28DPI-DMSO and 28DPI-RAPA groups for 2 h at 16°C, respectively. Then, the mixtures of F. columnare G₄ and serum were added to the fresh sterile medium separately, and the OD values at different time points were measured by spectrophotometer.

To confirm whether the effects of RAPA treatment on the agglutination and growth of F. columnare G₄ originate from the affected Igs, we developed an in vitro IgM-depletion strategy to eliminate secretory IgM (sIgM) from serum (28DPI-S-DMSO-IgMDEP). Firstly, mouse anti-trout IgM (20 μg) mAb was incubated with 200 μl protein G magnetic beads (Biolinkedin) for 2 h at 4°C to obtain anti-IgM–conjugated protein G beads. Thereafter, the anti-IgM–conjugated protein G beads were applied into 200 μl 28DPI-S-DMSO of serum to deplete the existing sIgM and then coincubated at 4°C for 2 h. Then beads from serum were separated by using MagnaBind magnets (Biolinkedin). Finally, the resulting serum was used as 28DPI-S-DMSO-IgMDEP serum. The removal of sIgM in 28DPI-S-DMSO-IgMDEP serum was verified by Western blotting. Then, the agglutination and growth of F. columnare in the serum of 28DPI-S-DMSO-IgMDEP were detected as described above.

To detect whether the mTORC1 signaling pathway affected the Igs secretion by spleen, the Igs from culture medium derived of cultured spleen explants of trout i.p. injected with DMSO and RAPA for 7 consecutive d were analyzed. Briefly, DMSO- or RAPA-treated trout were anesthetized with MS-222, and then blood was removed from the caudal vein. Thereafter, the spleen (∼20 mg) was collected and soaked with 75% ethanol for 1 min to eliminate the surface bacteria. After washing twice with PBS, tissues were placed in a 48-well plate and cultured with 200 μl DMEM medium (Invitrogen), supplemented with 10% FBS, 100 mg/ml streptomycin, 100 U/ml penicillin, 250 mg/ml gentamicin sulfate, and 200 mg/ml amphotericin B. The spleen explants were cultured with 5% CO₂ at 17°C for 3 d, and supernatants were harvested and stored until further analysis.

Statistical analyses were conducted by GraphPad Prism 8 software. Data comparisons between groups were determined by a paired/unpaired Student t test and one-way ANOVA. Data were expressed as the mean ± SEM, and differences were considered statistically significant at p < 0.05.

In this study, we compared the sequence homology of key proteins involved in trout mTORC1 signaling pathway including mTORC1 (Supplemental Table I) along with its catalytic kinase domain (Supplemental Fig. 1A) and FRB domain (Supplemental Fig. 1B), S6 (Supplemental Fig. 1C), and 4EBP1 (Supplemental Fig. 1D) with those of other vertebrates, and our findings suggested that these proteins are highly conserved among vertebrates. The tertiary structure of mTORC1 in rainbow trout was also predicted based on its corresponding crystal structure in humans (Supplemental Fig. 1E), and phylogenetic trees demonstrated that mTORC1 in rainbow trout clustered with those in other teleosts (Supplemental Fig. 1F).

To evaluate whether the mTORC1 signaling was involved in teleost B cell responses, trout SPLs were incubated with RAPA or DMSO. Afterward, two main trout B cell subsets (IgT⁺ and IgM⁺ B cells) were sorted via MACS. Next, using flow cytometry analysis and immunofluorescence microscopy, we found that a vast proportion of the sorted cells could be stained with anti-trout IgT or IgM mAbs (Fig. 1A, 1B, 1D, 1E). Moreover, gene expression analysis indicated that the soluble and membrane forms of IgT or membrane IgM/IgD were only expressed in the IgT⁺ or IgM⁺ B cell subsets, respectively. Importantly, the expression of T cell markers (TCR-α, TCR-β, and CD8α) and the myeloid cell marker FcεRIγ were restricted to the IgT⁻ or IgM⁻ cell subsets, thus confirming the high purity of the sorted B cells (Fig. 1C, 1F; the primers are shown in Supplemental Table II). Furthermore, we measured the protein and phosphorylation levels of mTOR, S6, and 4EBP1 in B cells following RAPA treatment by immunoblotting analysis (the rabbit Abs against human mTOR, S6, and 4EBP1 cross-reacted with trout proteins). Importantly, we found that although the protein levels of mTOR, S6, and 4EBP1 of both IgT⁺ and IgM⁺ B cells were unchanged when compared with those in DMSO-treated fish, their phosphorylation levels were significantly inhibited in RAPA-treated fish (Fig. 1G–L), suggesting that mTORC1 signaling is present in teleost fish B cells and inhibited by RAPA.

Previous studies reported that teleost B cells could phagocytose latex beads or bacteria, which constitutes a key defense mechanism in the innate immune response (7, 10). To assess whether the mTORC1 signaling is involved in the phagocytosis function of teleost B cells, DMSO- and RAPA-treated trout SPLs were incubated with 1-μm fluorescent latex beads. According to flow cytometry analyses, the proportion of phagocytic IgT⁺ (∼37.8%) and IgM⁺ (∼46.5%) B cells among total IgT⁺ or IgM⁺ B cells decreased after RAPA treatment in vitro compared with the percentage of IgT⁺ (∼47.2%) and IgM⁺ B cells (∼60.4%) in the DMSO control (Fig. 2A–E). More importantly, RAPA significantly decreased the percentage of high-capacity phagocytic IgT⁺ and IgM⁺ B cells (with two or more beads ingested per B cell) from 62.5 to 56.0% and 69.7 to 64.0% (Fig. 2A–D, 2F), respectively. Immunofluorescence microscopy also confirmed that the average number of internalized beads by IgT⁺ or IgM⁺ B cells significantly decreased after RAPA treatment when compared with those in the DMSO control (Fig. 2G, 2H). Thus, our findings indicated that the mTORC1 signaling mediated the number and strength of phagocytic B cells and may play a critical role in the B cell immune responses of teleost fish.

The mTORC1 signaling has been proven to play a critical role in promoting cell growth and survival in mammals (14, 23). To evaluate whether mTORC1 has a regulatory role in teleost B cells, a series of in vitro and in vivo experiments was conducted. For the in vitro experiments, trout SPLs were obtained and then cultured with either RAPA or DMSO. One day later, the proportion, apoptosis, and proliferation of IgM⁺ and IgT⁺ B cells among the SPLs were analyzed. At 1, 3, and 7 d posttreatment, the Ig secretion level of B cells in two groups was evaluated in the culture supernatants of SPLs. Flow cytometry analyses indicated that the proportion of spleen lymphocytes (50.9%) among the RAPA-treated SPLs decreased significantly compared with the spleen lymphocytes (30.6%) in the DMSO-treated group (Fig. 3A, 3B). Moreover, the proportion of IgT⁺ (∼10.6%) and IgM⁺ B cells (∼40.3%) among spleen lymphocytes with RAPA treatment decreased significantly when compared with IgT⁺ (∼12.7%) and IgM⁺ B cells (∼46.2%) in the DMSO group (Fig. 3A–C). These data indicated that the survival rates of IgT⁺ and IgM⁺ B cells were significantly reduced after mTORC1 signaling inhibition. Next, we examined apoptosis and proliferation in mTORC1-suppressed B cells of trout. According to flow cytometry analyses, the proportions of apoptotic IgT⁺ and IgM⁺ B cells in the RAPA-treated group were increased by 1.72-fold and 1.93-fold in total IgT⁺ and IgM⁺ B cells, respectively, compared with the DMSO group (Fig. 3D–F). Further, the proportion of proliferating IgT⁺ and IgM⁺ B cells (relative to the total IgT⁺ and IgM⁺ B cells, respectively) decreased from 1.4 to 0.3% and from 1.5 to 0.5% after RAPA treatment, respectively (Fig. 3G–I). Similar results were also obtained in the immunofluorescence assays, as the proliferation of sorted IgT⁺ or IgM⁺ B cells was significantly inhibited in the RAPA treatment compared with those in the DMSO control (Fig. 3J–L). Moreover, at 3 and 7 d post–RAPA or DMSO treatment, the protein levels of IgT, IgD, and IgM in the culture supernatants of SPLs with RAPA treatment were significantly lower than those in the DMSO control, whereas there was no difference in the concentration of IgT, IgD, and IgM between the DMSO group and the RAPA group after 1 d of treatment (Fig. 3M–P).

To further assess the effects of RAPA treatment on teleost B cells in vivo, trout were injected with DMSO or RAPA for 7 consecutive d. Splenic B cells were then sorted and analyzed for proportion, apoptosis, and proliferation of IgT⁺ and IgM⁺ B cells, and spleen explant supernatants were analyzed to assess Ig secretion level. Immunoblotting analysis indicated that the phosphorylation levels of mTOR, S6, and 4EBP1 in IgT⁺ and IgM⁺ B cells from trout injected with RAPA were significantly lower than those in trout injected with DMSO (Fig. 4A–F), indicating that the mTORC1 signaling in splenic B cells was effectively inhibited by in vivo RAPA injection. After continual RAPA injection for 7 d, the proportion of IgT⁺ and IgM⁺ B cells among spleen lymphocytes decreased by 5.2 and 25.5%, respectively (Fig. 4G, 4H). Apoptosis and proliferation assays indicated that the proportions of apoptotic IgT⁺ and IgM⁺ B cells were significantly increased by 4.74- and 2.10-fold among the total IgT⁺ and IgM⁺ B cells, whereas the proliferation of IgT⁺ and IgM⁺ B cells significantly decreased in trout from the RAPA injection group compared with the DMSO injection group (Fig. 4I–N). These results were consistent with those obtained in the in vitro experiment. Further, the concentrations of IgT, IgD, and IgM in explant supernatants of spleen from trout injected with RAPA were significantly reduced compared with those in the DMSO injection group (Fig. 4O, 4P). In conclusion, our results demonstrated the successful construction of an mTORC1 signaling inhibition model in teleost IgT⁺ and IgM⁺ B cells both in vitro and in vivo. Additionally, our findings demonstrated that mTORC1 can affect teleost B cell survival and Ig production by regulating apoptosis and proliferation.

RAPA treatment decreased B cell proliferation and Igs production in fish, suggesting that the mTORC1 signaling might regulate the response of B cells to pathogenic invasion. To address this hypothesis, we established an F. columnare infection model in rainbow trout, which was coupled with mTORC1 inhibition by RAPA treatment (Fig. 5A). Consistent with our previous results, F. columnare readily infected trout and caused classical symptoms, including severe skin lesions, fin rot, and gill necrosis, resulting in high mortalities (Supplemental Fig. 2) (22, 24). Moreover, immunoblotting analyses were conducted to examine the activation of the mTORC1 signaling in IgT⁺ and IgM⁺ B cells in peripheral blood from RAPA- or DMSO-treated trout at 28 d post–F. columnare infection. Our findings indicated that the phosphorylation levels of mTOR and S6 in IgT⁺ and IgM⁺ B cells in trout from the 28DPI-S-RAPA group were significantly lower than those of the 28DPI-S-DMSO group, whereas no significant differences in 4EBP1 phosphorylation were detected (Fig. 5B–E). Similar results have also been reported in mammals, as 4EBP1 phosphorylation is less sensitive to RAPA treatment (25). Additionally, upon secondary challenge with F. columnare, the gDNA copy numbers of F. columnare in the 28DPI-S-RAPA challenge group increased at 2-, 4-, and 7-d postchallenge compared with those of the 28DPI-S-DMSO challenge group in gills (Fig. 5F), and the cumulative mortality of trout in the 28DPI-S-RAPA challenge group was significantly higher than that in the 28DPI-S-DMSO challenge group (Fig. 5G). To assess Ig secretion in response to F. columnare, the total concentrations and F. columnare–specific binding of three Igs in serum were evaluated via immunoblotting and pulldown assays. The protein concentrations of IgT, IgM, and IgD in the serum of the 28DPI-S-RAPA group decreased by ∼1.9-, 1.8-, and 2.8-fold compared with those in the 28DPI-S-DMSO group, respectively (Fig. 5H). Furthermore, as shown in (Fig. 5I–K, F. columnare–specific IgM binding was detected in up to 1/1000 serum dilutions in the 28DPI-S-DMSO group, whereas F. columnare–specific IgM binding was only detected in the 1/10 dilution in the 28DPI-S-RAPA group. In contrast, F. columnare–specific IgT and IgD binding could not be detected either in the 28DPI-S-DMSO or 28DPI-S-RAPA group (Fig. 5I–K). These results confirmed previous evidence that IgM play an important role in mucosal and systemic immunity against bacterial infection. Furthermore, our results suggested that the RAPA-induced inhibition of the mTORC1 signaling hampered bacterial-specific IgM responses in teleost fish and significantly increased the susceptibility of fish to F. columnare reinfection.

Our results indicated that F. columnare–specific IgM binding was strongly induced in the serum of 28DPI-S-DMSO fish. These findings led us to hypothesize that these specific Igs might fight against F. columnare infection via agglutination. To better verify this hypothesis, we selected the IgM with the most abundant Ig concentration in fish body fluids to explore. Therefore, F. columnare was incubated with serum from naive fish, as well as fish from the 28DPI-S-DMSO and 28DPI-S-RAPA groups. The roles of IgM were evaluated via the prior depletion of Igs from 28DPI-S-DMSO fish, resulting in 28DPI-S-DMSO serum devoid of IgM (28DPI-S-DMSO-IgMDEP) (Fig. 6A). At 10 min and 2 h postincubation with F. columnare, four serum groups (naive, 28DPI-S-DMSO, 28DPI-S-RAPA, and 28DPI-S-DMSO-IgMDEP) were analyzed for the agglutination and growth assay, respectively (Fig. 6A). As expected, several large F. columnare aggregates were observed when incubated with serum from the 28DPI-S-DMSO group, whereas no such aggregates were identified in the naive group (Fig. 6B). Moreover, the 28DPI-S-RAPA and 28DPI-S-DMSO-IgMDEP groups exhibited only small F. columnare aggregates when the bacteria were incubated with serum (Fig. 6B). Combined with the decreased bacterial-specific Ig binding in the 28DPI-S-RAPA group, our results suggest that inhibition of the mTORC1 signaling reduced agglutination of F. columnare through an Ig-mediated mechanism. Next, we found that the growth rate of F. columnare was significantly decreased when the bacteria were incubated with serum from 28DPI-S-DMSO fish compared with those of the naive group (Fig. 6C, 6D). However, these rates increased to naive group levels when the bacteria were treated with IgM-depleted 28DPI-S-DMSO serum (Fig. 6C, 6D). More importantly, the growth rates of F. columnare incubated with serum from the 28DPI-S-RAPA group were significantly higher than those of the 28DPI-S-DMSO group, but lower than those of naive fish (Fig. 6C, 6D). Taken together, our findings strongly indicated that IgM played a key role in the agglutination and growth inhibition of F. columnare, which is mediated by the mTORC1 signaling in teleost fish.

B cells were discovered over half a century ago and first characterized in birds (26). In the past decades, the study of B cells in nonmodel species has greatly contributed to the advancement of comparative immunology, as well as the comprehensive understanding of B cell function and evolutionary origins (27). B cells play a critical role in adaptive humoral Ab-mediated immunity, as well as innate immunity (e.g., phagocytic capacity) in most vertebrates, including teleost fish (28). In mammals, mTORC1 signaling is linked to the modulation of B cell responses and Ab-mediated immunity; however, very little is known about whether mTORC1 signaling is involved in regulating B cell immune responses in teleost fish. Because teleost fish represent the most ancient bony vertebrates with bona fide B cells and conservation of mTORC1 signaling pathway during vertebrate evolution, we hypothesized that the mTORC1 pathway is evolutionarily conserved in regulating B cell immune responses for maintaining homeostasis. Our study thus sought to examine whether and how mTORC1 regulates B cell immune responses in teleost fish.

Our findings demonstrated that the amino acid sequences of mTOR protein in trout, as well as its downstream effectors S6 and 4EBP1, are highly similar to those in other vertebrates, suggesting that the role of mTOR may be highly conserved among vertebrates. RAPA is widely known to act as a targeted inhibitor of mTORC1 signaling in mammals. This compound specifically binds to the FRB domain of mTOR, thus inhibiting mTORC1 and influencing B cell function (29). To determine the possible role of mTORC1 signaling in modulating B cell immune responses in teleosts, trout IgT⁺ and IgM⁺ B cells were collected via MACS. Interestingly, similar to mammals, acute RAPA treatment significantly inhibited the activation of mTORC1 signaling in both IgT⁺ and IgM⁺ B cells, as demonstrated by a decrease in mTOR, S6, and 4EBP1 phosphorylation. These results provide the important evidence that mTORC1 signaling of teleost B cells is influenced by RAPA and indicate an evolutionarily conserved role of mTORC1 signaling function in mammals and teleost fish. Additionally, there is the possibility that mTORC2 assembly and activity may be altered by chronic treatment or high doses of RAPA. In our in vitro experiments, we proved that RAPA (100 nM) treatment for 24 h was able to inhibit mTORC1 signaling in B cells, and mTORC1 signaling can regulate B cell immunity, but the effect of chronic RAPA exposure on mTORC2 was not examined. In mammals, previous studies have demonstrated that prolonged (24 h) exposure to RAPA (100 nM) can disrupt mTORC2 activity in different (PC3, BJAB, and Jurkat) cancer cell lines (16), and previous studies have confirmed that mTORC2 was inhibited in B cells treated with RAPA (10 nM) for more than 24 h (30). Compared to mammals, by far, very little has been studied on RAPA inhibition of mTORC2 in B cells of teleost fish. Therefore, future work is warranted to investigate this very interesting hypothesis. Next, to evaluate the potential regulatory role of mTORC1 signaling in B cells, we further analyzed phagocytosis, apoptosis, proliferation, and Ab secretion in trout B cells at the cellular level.

Phagocytosis consists of the engulfing and destruction of particulate matter by phagocytes, which serves as an important bodily defense mechanism (31). Previous studies have found that in addition to professional phagocytes (e.g., macrophages/monocytes, neutrophils, and dendritic cells), B cells from teleost fish and mammals (B1 subset) possess phagocytic activity (7, 32, 33). Although previous studies have reported that the inhibition of the mTORC1 signaling in macrophages impairs phagosome acidification and phagolysosome formation, resulting in a reduction in intracellular pathogen clearance (34), very little is known about the regulation of the phagocytic activity of B cells in vertebrates. Our findings demonstrated that the percentage of phagocytic IgT⁺ and IgM⁺ B cells was significantly reduced after RAPA treatment. Moreover, the phagocytic capacity of B cells was decreased after the inhibition of mTORC1 signaling, as the average numbers of beads internalized by phagocytic B cells were significantly decreased. Thus, our findings provide the first demonstration, to our knowledge, that impaired mTORC1 signaling significantly reduces the phagocytosis of teleost B cells in vertebrates. From an evolutionary perspective, our study established an animal model for studying the regulatory mechanisms that govern B-1 cell phagocytosis and provides crucial insights into the evolution of B cell functions in vertebrates.

In addition to evaluating the phagocytosis capabilities of B cells, we next sought to investigate the role of mTORC1 signaling in B cell proliferation and apoptosis in vitro and in vivo. Our data indicated that the number of trout IgT⁺ and IgM⁺ B cells decreased significantly after RAPA treatment, which might be due to an imbalance between proper cell proliferation and programmed cell death, as observed in mammals (35). Confirming this hypothesis, we found that RAPA treatment significantly increased the proportion of both apoptotic IgT⁺ and IgM⁺ B cells within trout SPLs, whereas the proliferation of IgT⁺ and IgM⁺ B cells in the same fish decreased both in vitro and in vivo. Moreover, we also found that IgT, IgM, and IgD secretion was severely reduced either in the culture medium of SPLs after treatment with RAPA or explant supernatant of spleen from RAPA-injected fish compared with that of the DMSO group and DMSO-injected fish, respectively. However, whether the B cell and Ig responses were directly or indirectly regulated by RAPA needs to be further investigated. In a previous study, IgM⁺ B lymphocytes and Ab production were suppressed after RAPA and keyhole limpet hemocyanin treatment in a teleost fish, flounder (Paralichthys olivaceus), which was demonstrated to be associated with the inhibition of T lymphocyte counts (36). This result implies that T cell inhibition by RAPA regulates the immune function of B cells. Interestingly, Gong et al. (37) have confirmed that the immune response of T cells in zebrafish is significantly upregulated on the fourth day after Ag stimulation, whereas the immune response of B cells, regulated by T cells, is significantly upregulated on the sixth day after Ag stimulation, as compared with the immune response of B cells without T cell involvement, which is significantly upregulated on the second day after Ag stimulation. It is necessary to point out that these results suggested that B cells can produce an immune response independent of T cells and at a faster rate. Combined with our experimental results, we found that the Ig concentration decreased significantly at 3 d posttreatment with RAPA, suggesting that the decrease in Ig concentration may not be directly linked to T cells. This indicates to a large extent that the immune effects of RAPA on B cells during the treatment of B cells in vitro may be the cause of direct regulation on B cells. Hence, it is conceivable that the decreased concentration of sIgs was strongly correlated with the reduction in B cells after RAPA treatment. Interestingly, similar results have been reported in mammals, as inhibiting the mTORC1 signaling can affect B cell proliferation and survival (38). Along with the decrease in the number of B cells, the production of sIgM and sIgG in mammalian serum was also reduced, which was caused by the depletion of newly formed plasma cells and the attenuation of Ab synthesis in surviving plasma cells (39–41). Therefore, additional studies are required to determine whether the inhibition of mTORC1 signaling in teleost fish affects the differentiation of B cells into plasma cells or the ability of these plasma cells to secrete Igs.

Notably, we previously found that trout B cells would secrete bacteria-specific Igs to participate in the humoral immune response against bacterial infection. In this regard, a strong IgT response is induced in mucosal immunity (e.g., skin and gill), whereas IgM is the predominant Ig in systemic immunity (22, 42). Therefore, fish reexposed to bacterial pathogens exhibited high resistance (22). Consistent with these observations, our findings indicated that using the same infection strategy with F. columnare induced high specific IgM titers in serum. However, during the infection period, F. columnare Ag might be taken up by APCs and then activate the naive T cells to initiate the adaptive immune response of B cells (43). Moreover, mTORC1 signaling in APCs and T cells was also the target of RAPA (18, 44–46). In a teleost model, the Nile tilapia (Oreochromis niloticus), Wei et al. (18) discovered for the first time that the inhibition of mTORC1 signaling by RAPA undisputedly impaired multiple immunological processes in T cells, including activation, proliferation, and infection clearance. By far, very few studies on the regulation of APCs by mTORC1 in teleost fish, although in mammals it has been proved that inhibition of mTORC1 signaling affects the survival and differentiation of dendritic cells, as well as their immune function (44). In our experiments, we found that mTORC1 signaling was significantly inhibited in B cells after RAPA treatment, and the specific IgM titers were suppressed and resulted that the increasing the susceptibility of fish upon F. columnare invasion (agglutination of bacteria and inhibition of bacterial growth). However, at this time, we could not rule out whether the specificity of the immunosuppression by RAPA. The decreased IgM titers might not only be due to the reduction of mTORC1 pathway activity in B cells, but also the costimulatory signals from APCs or T cells interrupted by RAPA. Therefore, further studies are clearly warranted to confirm this hypothesis. It is worth noting that mTOR knockout reduced the production of bacterial specific Igs in B cells in mice upon bacterial reinfection, resulting in a severe increase in bacterial loads and mortality (40). Moreover, RAPA can influence specific Ig production through the inhibition of B cell activation/proliferation after viral infection in mammals (19). Thus, our results strongly suggested that the regulatory effects of mTOR signaling on the B cell responses to pathogenic invasion are evolutionarily conserved between mammals and teleost fish.

The high bacteria-specific IgM titers in serum from 28DPI-S-DMSO fish led us to hypothesize that these Abs play a key role in antibacterial agglutination. This was supported by the fact that 28DPI-S-DMSO fish serum incubated with F. columnare exhibited a strong agglutination and bacteriostatic activity. Interestingly, aggregation by anti-pertussis Abs in mammals induces a direct bactericidal effect on Bordetella pertussis (47). Furthermore, purified IgG have been reported to adhere to bacteria and result in their aggregation (48). Similar agglutination has also been reported in the serum of Japanese flounder (P. olivaceus) after immunization with N. girellae ciliary protein, as well as in Indian major carps artificially infected with Edwardsiella tarda (49, 50). It is also worth noting that the observed agglutination of the serum from 28DPI-S-RAPA disappeared over time as expected, and their ability to inhibit bacterial growth decreased significantly. Moreover, these fish exhibited low resistance to reinfection by F. columnare, indicating that mTORC1 signaling inhibition substantially reduced the agglutinating and bacteriostatic activity of Igs. To further confirm whether these phenomena were associated with reductions in bacterial-specific IgM in serum by RAPA treatment, IgM-depleted serum from 28DPI-S-DMSO fish was incubated with F. columnare. This experiment demonstrated that the IgM-depleted serum also lost its agglutinating capacity, which was consistent with our observations in the 28DPI-S-RAPA fish. More importantly, the F. columnare bacteriostatic activity in the 28DPI-S-DMSO group was significantly higher than those of the 28DPI-S-DMSO-IgMDEP (i.e., the levels nearly recovered to control levels) and even higher than those of the 28DPI-S-RAPA group. These results highlight the indispensable agglutinating and bacteriostatic roles of IgM and provide insights into the previously unrecognized role of mTORC1 singling in mediating B cell responses in teleost fish.

In conclusion, our study provides novel and compelling evidence for the conserved role of mTORC1 signaling in the B cell responses of teleosts. Notably, inhibition of mTORC1 signaling in teleost B cells significantly reduced their proliferation rates and accelerated apoptosis. Further, mTORC1 signaling in B cells contributes to humoral Ab-mediated immune responses to bacterial infection. More importantly, our findings indicated that inhibition of mTORC1 signaling made fish significantly more susceptible to bacterial infection, thus providing the first demonstration, to our knowledge, that mTORC1 signaling plays a key role in modulating B cell immune responses to pathogens in teleosts. Moreover, our results indicated that mTORC1 signaling plays a unique regulatory role in B cell phagocytosis activity, which sheds light on the evolution of the regulatory mechanisms of B cell innate immunity in vertebrates. Given that the B cells in fish and mammals are functionally analogous, our findings indicate that the elucidated mTORC1 pathway is evolutionarily conserved in regulating B cell responses, thus providing a new point for understanding the B cells functions in teleost fish.

We thank Dr. J. Oriol Sunyer (University of Pennsylvania) for the generous gift of anti-trout IgM, anti-trout IgD, anti-trout IgT mAbs, and anti-trout IgT, and Dr. Pin Nie for his generous gift of the GFP-labeled F. columnare G₄.

The authors have no financial conflicts of interest.