Healthy A. japonicus (weight 98 g ± 14 g) were purchased from the Dalian Pacific Aquaculture Company and temporarily reared in aerated seawater (salinity 28 ± 0.5; temperature 16 ± 0.5°C; pH 8 ± 0.1) for 5 days before the experiments. All experiments were performed in accordance with the advice in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The experimental protocols were approved by the Research Animal Ethics Committee of Ningbo University, China.

The total RNA from coelomocytes was isolated with RNAiso plus (TaKaRa, Dalian, China) according to the manufacturer’s instructions. RNA concentration was measured by Nanodrop, and each paired sample was adjusted to the same concentration (1,000 ng/ul). For PCR of mRNA and circRNA, the cDNA was synthesized by the PrimeScript™ RT reagent Kit (TaKaRa, Dalian, China). For the qRT-RCR of miRNA, cDNA was synthesized using the miScript II RT Kit (Qiagen, Dusseldorf, Germany). qRT-PCR was performed using the TB Mix (TaKaRa, Dalian, China) on a 7500 real-time PCR detection system. The relative expression was calculated by the 2^−ΔΔCT method. Actin served as the endogenous control for circRNA and mRNA. The miRNA expression levels were normalized against RNU6B. Primer sequences are listed in Table 1 .

The total protein was extracted from coelomocytes by cell lysis buffer (Beyotime, Shanghai, China) and quantified at 50 μg/ml with the BCA protein assay kit (Cwbio, Jiangsu, China). The protein was separated by 12% SDS-PAGE gels and transferred onto a PVDF membrane by wet blotting. The membranes were blocked with 5% non-fat powdered milk at room temperature for 2 h and incubated with a primary antibody (1:500 diluted in 1% BSA) at 4°C for 8 h. Then, they were incubated with a secondary antibody (1:10,000 diluted in 1% BSA) at room temperature for 1.5 h. Finally, the blots were measured by Western ECL Substrate (Bio-Rad, California, USA), and images were obtained by using an Omega Lum C imaging system (Aplegen, California, USA).

Coelomocyte RNA (1 μg) was mixed with 1 U/μg Ribonuclease R (RNase R; Epicenter, WI, USA) or left unmixed at room temperature for 30 min. The incubated RNAs were reverse-transcribed to cDNA with specific primers for the RT-PCR assay.

The protocol used for culturing primary coelomocytes was the same as that used in our previous work (27). The coelomic fluids from healthy sea cucumber were passed through a 300-mesh filter to remove large tissue debris. Then, the coelomic fluids were mixed well with an equal volume of anticoagulant solution (0.02 M EGTA, 0.48 M NaCl, 0.019 M KCl, 0.068 M Tris–HCl, pH = 7.6) and centrifuged at 300×g at 4°C for 10 min to remove the supernatant. After resuspension and centrifugation twice with isotonic buffer (0.001 M EGTA, 0.53 M NaCl, 0.01 M Tri–HCl, pH = 7.6), the coelomocytes were mixed well with L-15 cell medium (Sangon, Shanghai, China) containing penicillin (100 U ml⁻¹) and streptomycin sulfate (100 mg ml⁻¹) to obtain a final concentration of 10⁶ cells. The mixture was added to a 6-well culture plate at a volume of 2 ml per well and cultured at 16°C overnight. Then, the coelomocytes were challenged with 1 μg ml⁻¹ of LPS (Sigma, California, USA) for 0, 1, 3, 6, 12, and 24 h. Three biological replicate cells comprised a group for the extracted RNA and protein.

For luciferase reporter plasmids, the binding site sequences of circRNA75, circRNA72, target genes, and their mutant sequences were inserted into the psiCHECK-2 plasmids (Promega, Madison, USA). The mutant sequences were obtained by the Fast Mutagenesis System (Trans, Beijing, China). SiRNA (which covered the back-splicing region of circRNAs for silencing), miR-200 mimics, and miR-200 inhibitor were synthesized from GenePharma (Shanghai, China), and the sequences are listed in Table 1 . Both plasmids and oligonucleotides were transfected into the EPC cell lines (the epithelioma papulosum cyprinid) and primary coelomocytes using Lipofectamine 6000 (Beyotime, Shanghai, China) according to the manufacturer’s protocols. In brief, 1 ul of siRNA, miRNA mimics, or miRNA inhibitor (20 mM) was mixed with 2 ul of Lipofectamine 6000 and incubated at room temperature for 15 min. Then, the 3 ul transfection solutions were added to each well of a 24-well plate.

The recombinant plasmids (200 ng per well), miR-200 mimics NC, and miR-200 mimics (200 nM) were co-transfected into EPC cells. After transfection for 48 h, the culture medium was removed, and each well was washed twice with 100 ul of PBS. Next, each well was treated with 20 μl of 1× Passive Lysis Buffer (PLB) for 25 min at room temperature to dissolve cells. Then, 100 μl of Luciferase Assay Reagent II (LAR II) was added to each well, and the Firefly luciferase activity was immediately measured by a GloMax 96, 20/20 luminometer (Promega, Madison, USA). After measuring the Firefly luciferase activity, 100 ul of Stop & Glo^® Reagent was added accordingly to measure the Renilla luciferase activity. All group data came from six biological replicates. One-way (ANOVA) was used to analyze the significant differences between control and experimental groups.

For the apoptotic assay, the primary coelomocytes were seeded into a 24-well plate for 8 h. After transfection of siRNA, miR-200 mimics, and miR-200 inhibitor for 48 h, the coelomocytes were harvested by centrifugation at 300×g for 10 min. The apoptosis levels of coelomocytes were determined by the Annexin V-FITC apoptosis detection kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions. In short, 10⁶ coelomocytes were incubated in 200 μl of Annexin V-FITC binding buffer containing 10 μl of PI and 5 μl of Annexin V-FITC to stain, which were left to stand in the dark at 25°C for 10 min. Finally, the apoptosis levels were detected using Gallios (Beckman, California, USA).

In our previous study, we found that miR-200 was abundant in coelomocytes of SUS and might play important roles in the immune response of A. japonicus (28). On the basis of the result, we screened two circRNAs (circRNA75 and circRNA72) among 261 circRNAs that exhibited miR-200 potential binding sites. To assure the circle structure of circRNA75 and circRNA72, reverse transcription PCR with divergent primers and Sanger sequencing were performed to verify their existence and the splicing junctions of the two circRNAs ( Figure 1A ). Genomic annotation of A. japonicus by back-splicing showed that circRNA75 and circRNA72 came from one exon of fibrinogen-like protein A and six exons of HSP97, respectively ( Figure 1B ). To further confirm the existence of circRNA75 and circRNA72, we treated total RNA of coelomocytes with RNase R and found that the linear RNAs (the corresponding linear RNAs of circRNA75 and circRNA72) were digested by RNase R rather than circRNA75 and circRNA72, which were resistant to RNase R digestion ( Figure 1C ).

To further address the immune function of the two circRNAs, the expressions of circRNA75 and circRNA72 in LPS-exposed coelomocytes were detected by qRT-PCR ( Figure 1D ). The results showed that circRNA75 and circRNA72 transcripts were significantly upregulated with a 1.73- and 7.18-fold increase at 6 h (p <0.01), respectively. These results were consistent with the results of the circRNA-seq differential expression analysis.

To elucidate the underlying molecular mechanism, the miRanda v3.01 toolbox was applied to predict the targeted relationships among circRNA75, circRNA72, and miR-200 in A. japonicus. CircRNA75 had four potential binding sites of miR-200, whereas circRNA72 had two potential binding sites. This finding indicated that circRNA75 and circRNA72 may serve as sponges to capture miRNAs, thereby resulting in the release of specific miRNA-targeted transcripts. Wild-type (WT) and mutant (MU) luciferase reporters of circRNA75 and circRNA72 binding sites were constructed to analyze the role of miR-200 in the regulation of the two circRNAs ( Figure 2A ). After co-transfecting EPC cells with miR-200 mimics or NC mimics and the dual-luciferase reporter system, we found that the overexpression of miR-200 significantly inhibited the luciferase reporter activity of four circRNA75 WTs and that of one circRNA72 WT but not the luciferase reporter activity of all circRNA MU ( Figures 2B, C ). All these results validated the binding of miR-200 to circRNA75 and circRNA72.

Through the miRanda v3.01 toolbox-screened A. japonicus transcriptome data, we predicted the existence of four miR-200 target genes (Tollip, Myd88, TRAF6, and P105). Then, we constructed luciferase reporter plasmids containing the 3’-UTRs of the wild-type (WT) and mutant (MU) genes; the sequences contained miR-200 binding sites ( Figure 3A ). After the dual-luciferase experiment, we found that miR-200 significantly decreased the luciferase activity in the Tollip WT plasmid group but not in other genes and mutant plasmids ( Figure 3B ). Subsequently, qRT-PCR assays revealed that miR-200 mimics significantly reduced Tollip mRNA levels by 0.12-fold and protein levels by 0.46-fold in primary coelomocytes, whereas miR-200 inhibitors increased Tollip mRNA and protein levels by 7.25- and 2.52-fold, respectively ( Figures 3C–E ). Moreover, we found that the expressions of the circRNAs were not affected by the change in miR-200.

Some studies showed that Tollip is an anti-apoptosis gene in higher animals (29). Our previous study proved that Tollip is a negative regulatory gene involved in the activation of the Toll signaling pathway in sea cucumber that showed resistance to V. splendidus infection (30). However, the function of Tollip in sea cucumber has not been deeply detected. To further explore the functional role of Tollip, the apoptosis levels of coelomocytes were detected after silencing Tollip in vitro. After the si-Tollip was transfected into primary coelomocytes for 24 h, the mRNA and protein levels of Tollip were respectively downregulated by 0.36- and 0.49-fold (p <0.01) relative to the levels in the si-NC group ( Figures 3F, G ). Under these conditions, we measured PBS and LPS-induced apoptosis levels by Gallios. The results showed that the mortality rate of the si-Tollip group was upregulated by 4.69 and 9.42% (p <0.05) compared with the si-NC group under PBS and LPS-stimulated conditions, respectively ( Figures 3H, I ).

To investigate whether circRNA75 and circRNA72 participated in the immune response of A. japonicus through the sponge activity of miR-200, we designed a specific siRNA that covered the back-splicing region of the two circRNAs for circRNA interference. After the siRNA was transfected, we checked the expressions of circRNAs, miR-200, and Tollip in primary coelomocytes by qRT-PCR and Western blot. The results indicated that si-circRNA75 transfection inhibited circRNA75 expression by 0.43-fold and led to the 0.51- and 0.41-fold decrease in mRNA and protein levels of Tollip, respectively ( Figures 4A, C ). Under the same conditions (si-circRNA75 was replaced by an equivalent amount of si-circRNA72), circRNA72 was depressed by 0.17-fold, which resulted in the 0.33- and 0.49-fold decrease in mRNA and protein levels of Tollip, respectively ( Figures 4B, C ). The circRNA interference replicated the phenomenon of miR-200 overexpression downregulating the expression of Tollip in primary coelomocytes. Meanwhile, the rate of Tollip decrease caused by circRNA75 interference was greater than that caused by circRNA72 interference. However, the expression of miR-200 did not change ( Figures 4A, B ). Furthermore, to functionally confirm that circRNA75 and circRNA72 could serve as sponges of miR-200 to mediate the expression of Tollip, miR-200 mimics and inhibitors were used to examine the change in Tollip expression under circRNA knockdown conditions. The mRNA expression levels of Tollip in si-circRNA75 and si-circRNA72 + miR-200 inhibitor groups were upregulated by 2.49- and 1.83-fold (p <0.01) compared with si-circRNA75 and si-circRNA72 + miR-200 inhibitor NC groups, respectively ( Figures 4D, E ), and the protein levels of Tollip were upregulated by 1.67- and 1.80-fold (p <0.01) under the same conditions ( Figures 4F, G ). In contrast, mRNA expression levels of Tollip in si-circRNA75 and si-circRNA72 + miR-200 mimics groups were downregulated by 0.51- and 0.43-fold (p <0.01) compared with si-circRNA75 and si-circRNA72 + miR-200 mimics NC groups, respectively ( Figures 4D, E ); similarly, the protein expression levels of Tollip were downregulated by 0.44- and 0.49-fold (p <0.01; Figures 4F, G ).

To functionally confirm that circRNA75 and circRNA72 suppressed coelomocyte apoptosis via the sponge activity of miR-200 and consequently upregulated Tollip, miR-200 inhibitor was used to examine whether the coelomocyte apoptosis of circRNAs silencing could be blocked by miR-200 knockdown ( Figure 4H ). The results indicated that the mortality rates of the si-circRNA75 and si-circRNA72 groups were significantly upregulated by 5.61 and 6.15%, respectively, compared with that of the si-NC group. However, the mortality rate of si-circRNA75 or si-circRNA72 + miR-200 inhibitor groups showed no change compared with that of the si-NC group ( Figures 4I, J ). All results indicated that circRNA75 and circRNA72, as ceRNA, suppressed cell apoptosis by adsorbing miR-200 and releasing its target Tollip in response to immune stress in A. japonicus ( Figure 5 ).