All Nile tilapia (50 ± 5 g) individuals used in this work were acquired and fed in the same manner as we did in a previous study (31). The fish were randomly selected for subsequent experiment. All experiments were conducted according to the principles and procedures of Guangdong Province laboratory animal management regulations.

Three healthy fish were first narcotized with MS-222 (Sigma, Darmstadt, Germany) and then the tissue distribution of On-lect2 in healthy tilapia was investigated by collecting tissues from brain, gills, heart, head kidney, intestine, liver, muscle, skin, spleen and thymus. Total RNA was immediately extracted.

Streptococcus agalactiae (ZQ0910) was isolated from Nile tilapia and kept in our laboratory (32). S. agalactiae was cultured in fresh brain heart infusion (BHI) liquid medium overnight, collected by centrifugation, washed three times in phosphate-buffered saline (PBS). Afterward, 100 µL of S. agalactiae with a final concentration of 5×10⁷ CFU/mL was injected into 30 fish. At five time points (0, 6, 12, 24 and 48 h) after injection, three fish were collected as mentioned above and then the brain, head kidney, liver and spleen tissues were collected. Total RNA was immediately extracted.

Total RNA from Nile tilapia were isolated using RNAiso Plus (TaKaRa, Dalian, China). The RNA was then reverse-transcripted using PrimeScript™ RT reagent kit with gDNA Eraser (TaKaRa, Dalian, China) and diluted with distilled water at a ratio of 1:50 for subsequent experiments following the manufacturer’s instructions.

The predicted gene sequence of On-lect2 was obtained from NCBI (https://www.ncbi.nlm.nih.gov/nuccore/XM_003449406.5) and designed the partial sequence primers (On-lect2-S and On-lect2-A). Then the first strand cDNA of head kidney tissue was used as polymerase chain reaction (PCR) template to amplify On-lect2 fragment and sequenced it. Using the sequenced sequence, the primers for amplifying the 5’ terminal sequence (lect2-5’ RACE and lect2-5’ RACE nested) and the 3’ terminal sequence (lect2-3’ RACE and lect2-3’ RACE nested) of On-lect2 were designed respectively. For the 5’ and 3’ terminal sequence, the cloning method was referred to SMARTer RACE 5’/3’ Kit (Code No. 634858, Clontech, Mountain View, USA). Finally, the partial sequences acquired through gene cloning, 3’RACE and 5’RACE were assembled using contigExpress application software. All the primers used in this study were designed with the NCBI Primer designing tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/, Table S1 ). The potential open reading frame (ORF) of On-lect2 was identified using the NCBI ORF finder (https://www.ncbi.nlm.nih.gov/orffinder/). Gene structure was plotted with Exon–Intron Graphic Maker (http://www.wormweb.org/exonintron). Molecular weight, theoretical pI and amino acid composition were predicted using the ProtParam tool (https://web.expasy.org/protparam/). The potential signal peptide was predicted with SignalP (http://www.cbs.dtu.dk/services/SignalP/). The potential transmembrane domain was predicted with TMHMM Server v. 2.0 (http://www.cbs.dtu.dk/services/TMHMM/). Multiple sequence alignments of LECT2 proteins among other species were conducted using DNAMAN software (version 7.0). Similarity among the determined amino acid sequences was analysed by UniProt (https://www.uniprot.org/). A neighbour-joining (NJ) phylogenetic tree was constructed with MEGA software (version 6.0) with 1,000 bootstrap replications.

The tissue distribution and relative expression of On-lect2 in healthy tilapia at the transcriptional level was assessed via qRT-PCR. qRT-PCR was performed with TB Green^® Premix Ex Taq™ II (Tli RNaseH Plus) (TaKaRa, Dalian, China) and QuantStudio 6 and 7 Flex Real-Time PCR Systems (Thermo Fisher Scientific, Waltham, USA) following the manufacturers’ instructions. A 187 bp fragment of On-lect2 was amplified, and On-β-actin was used as a reference gene. The primers are listed in Table S1 . Relative levels of On-lect2 were calculated via the 2^-ΔΔCt method (33). These reactions were performed with three sample replicates and three technical replicates.

The tissue distribution and relative expression of On-LECT2 in healthy tilapia at the protein level was assessed via Western blotting. Total protein from muscle, skin, brain, head kidney, liver and spleen tissues were isolated using a protein extraction kit (Solarbio, Beijing, China). Each sample containing 20 μg of total protein was loaded on 12% SDS-PAGE and transferred to a PVDF membrane (Millipore). The samples were blocked with QuickBlock™ Blocking Buffer for Western Blot (Beyotime, Shanghai, China) overnight at 4°C and then incubated with anti-LECT2 (PAF541Mu01, Cloud-Clone Corp, Wuhan, China) and anti-β-actin (AF5003, Beyotime, Shanghai, China) as primary antibodies. The solutions were diluted at a ratio of 1:1000 in QuickBlock™ Primary Antibody Dilution Buffer for Western Blot (Beyotime, Shanghai, China) for 2 h at room temperature. Afterwards, the membranes were washed three times in TBST and incubated with HRP-labelled goat anti-rabbit IgG (H+L) (Beyotime, Shanghai, China) at room temperature for 1 h. Antigen–antibody complexes were detected via the enhanced chemiluminescence method (P0018S, Beyotime, Shanghai, China).

Brain, head kidney, liver and spleen tissues were collected from healthy or stimulated tilapia (challenged by S. agalactiae as mentioned above and RNAi as described below) and fixed in Dietrich’s fixative for over 20 h. The samples were dehydrated through a graded alcohol series (70%, 85%, 95% and 100%), cleared in xylene and then embedded in paraffin wax.

Serial sections 8 μm in thickness for histological analysis were stained with a haematoxylin and eosin staining kit (Beyotime, Shanghai, China) following the manufacturer’s protocols. The sections were observed and photographed using a Nikon DS-Ri2 microscope (Nikon, Tokyo, Japan).

Sections 5 μm in thickness for immunohistochemistry were rehydrated with PBT (PBS + 0.1% Tween-20). Endogenous peroxidase was inactivated with 3% H₂O₂, incubated with EDTA antigen retrieval solution (Beyotime, Shanghai, China) at 95°C to retrieve antigens and then incubated in PBT containing 3% bovine serum albumin (BSA). Subsequently, the primary antibody (anti-LECT2: PAF541Mu01, solutions were diluted at a ratio of 1:100 in 3% BSA; Cloud-Clone Corp, Wuhan, China) was added and incubated overnight at 4°C. Afterwards, the samples were washed five times in PBT and incubated with HRP-labelled goat anti-rabbit IgG (H+L) (Beyotime, Shanghai, China) at room temperature for 2 h. The samples were then washed three times in PBT, and antigen–antibody complexes were detected with DAB horseradish peroxidase colour development kit (Beyotime, Shanghai, China). Finally, the samples were stained with haematoxylin (Beyotime, Shanghai, China). Negative controls were performed by replacing the primary antibody with a preimmune serum to check antibody specificity. The samples were observed and photographed as mentioned above.

Specific primers ( Table S1 ) with the incomplete restriction sites Bam HI and Hind III were annealed with 50 µM of the final concentration to generate double-stranded oligo (ds oligo). Annealing was performed at 95°C for 5 min and then allowed to cool to room temperature for about 20 min. The RNAi plasmid pGenesil-1 (BT Lab, Wuhan, China) was digested with Bam HI and Hind III, ligated with ds oligo and transformed into Escherichia coli DH5α (TransGen, Beijing, China). The positive clone was verified via PCR and DNA sequencing. The RNAi plasmids pGenesil-1-shRNA-199/209/257/372 and the negative control plasmid pGenesil-1-shRNA-Control were extracted using E.Z.N.A.^® Endo-free plasmid Midi kit (Omega, Norcross, USA).

A total of 300 healthy tilapia individuals were randomly divided into six groups, PBS group, negative control group (shRNA-Control) and four RNAi group (shRNA-199, shRNA-209, shRNA-257 and shRNA-377) with 50 fish per group and reared for 2 weeks. Subsequently, PBS (100 μL), shRNA-Control plasmid (50 μg dissolved in 100 μL PBS) or the four positive shRNA plasmids (50 μg dissolved in 100 μL PBS) was injected into the fish. The liver and spleen were collected at three time points (3, 5 and 7 d) after injection as mentioned above. At each time point, three fish from each group were collected for total RNA extraction as mentioned above. At 12 d after plasmid injection, five fish from each group were collected. The liver and spleen were then collected for morphological and histological examination.

Daily statistical morbidity was calculated for 14 d, and survival rate (SR) was computed by the following iterative calculation formula (SR of 0 day is 1).

The knockdown efficiency of the mRNA level was determined via qRT-PCR. Moreover, three interaction genes of lect2, namely, peroxisome proliferator activated receptor gamma (PPARγ), matrix metallopeptidase 2 (MMP2) and β-catenin (3, 34, 35), were assessed at 5 and 7 d after injection.

The specific primers, namely, On-LECT2 recombinant protein-S (rOn-LECT2-S) and rOn-LECT2-A, with the restriction sites Bam HI and Xho I were designed ( Table S1 ) to amplify On-LECT2 ORF without the signal peptide domain. This DNA fragment was purified and ligated with pMD-19T plasmid (TaKaRa, Dalian, China). The recombinant plasmids pMD-19T-rOn-LECT2 and pET-N-His-C-His (Beyotime, Shanghai, China) were digested with Bam HI and Xho I and then purified. The digested products were then ligated and transformed into Escherichia coli BL21 (DE3) (TransGen, Beijing, China). The positive clone was verified via PCR and DNA sequencing and then picked to culture in fresh Luria-Bertani (LB) liquid medium containing kanamycin (100 µg/mL) until absorbance at OD600 reached 0.4–0.6. Subsequently, isopropyl-β-d-thiogalactoside (IPTG) was added to a final concentration of 0.5 mmol/L. The bacteria were collected and washed with PBS. The protein was purified with a His-tag protein purification kit (Beyotime, Shanghai, China) and then analysed by 12% reducing SDS-PAGE and Western blot with the His antibody (Beyotime, Shanghai, China).

Monocytes/macrophages were prepared following the method of a previous study (31, 36). 95 µL of S. agalactiae with a final concentration of 1×10⁵ CFU/mL was mixed with 95 µL of monocytes/macrophages with an equal final concentration. Afterwards, 10 µL of PBS or 0.5 μg of rOn-LECT2 dissolved in 10 µl of PBS was added. The culture was continued at 28°C, and then 10 µL of the mixture was collected at three time points (2, 4 and 6 h) after the addition. Subsequently, the mixture (10 µL) was continued cultured on a BHI tablet at 28°C for 36 h to count the number of S. agalactiae. For each tablet, 10 clones were randomly picked and sequenced the 16s rRNA genes to verify the bacterial species. Each group was repeated in six parallels.

A total of 250 healthy tilapia individuals were randomly divided into five groups with 50 fish per group and reared for 2 weeks. For the blank, PBS and LECT2 groups, 100 μL of PBS was injected into each fish. For the shRNA-Control and shRNA-372 groups, 50 μg of the negative control plasmid pGenesil-1-shRNA-Control dissolved in 100 μL PBS and 50 μg of the most efficient RNAi plasmid pGenesil-1-shRNA-372 dissolved in 100 μL PBS was injected into each fish, respectively. Except for the blank group, the four groups were challenged at 5 d after the first injection with a final concentration of 5×10⁷ CFU/mL S. agalactiae. By comparison, the LECT2 group was co-injected with 5 μg of rOn-LECT2 dissolved in the same 100 µl of PBS with S. agalactiae. At five time points (0, 6, 12 24 and 48 h) after challenge, the brain, head kidney, liver and spleen tissues from three fish were collected for total RNA extraction as mentioned above and one time point (48 h) for histological analysis. At 48 h after challenge, 10 mg of the liver and spleen tissues of the last four groups was collected, broken and dissolved in 1 mL of PBS and then diluted at a ratio of 1:500 in PBS. Afterward, 50 μL solutions was continued cultured on a BHI tablet at 28°C for 36 h to count the number of S. agalactiae. Each group was repeated in six parallels. Daily statistical morbidity was calculated for 7 d from the challenge, and SR was calculated as mentioned above.

Via qRT-PCR, the effects, functions and potential molecular mechanisms of On-LECT2 were further investigated by assessing a series of related genes ( Table S1 ), including potential receptors, inflammatory factors and typical factors of immune-related pathways.

Drawings and final panels were designed using Adobe Photoshop CC (San Jose, CA, USA) and Adobe Illustrator (San Jose, CA, USA). All data were presented as means ± standard deviation (SD) and the one-way ANOVA and Student’s t test was employed in this study to analyze the significant difference using Prism software (Version 8.0). Different letters or asterisks were marked to illustrate statistically significant differences (p<0.05).

The full-length cDNA of On-lect2 is 726 bp. It consists of a 456 bp ORF that encodes a putative protein of 151 amino acids, including a signal peptide domain of the N-terminal. However, On-LECT2 has no transmembrane domain. The predicted molecular mass of On-LECT2 is 16.5 kDa, and its theoretical pI is 9.37. A comparison of genomic sequences revealed that On-lect2 is comprised of four exons that are separated by three introns ( Figures 1A, B ). Multiple sequence alignments indicated that the deduced amino acid sequence of On-LECT2 contains the conservative peptidase M23 domain ( Figures 1A–C ). BLAST analysis revealed that the deduced amino acid sequence is homologous with other LECT2, which is 62%–84% identical to other fish species and about 50% identical to mammals ( Figure 1C ). Phylogenetic analysis suggested that On-LECT2 initially clustered with Epinephelus akaara and Larimichthys crocea and then clustered with other fish lineages. Finally, On-LECT2 clustered with mammals LECT2 ( Figure 1D ).

The expression characteristics of On-lect2 in different tissues of healthy tilapia at the transcriptional level was measured via qRT-PCR ( Figure 1E ). The highest On-lect2 expression was observed in the liver, whereas it was relatively high in the muscles, skin, and head kidney ( Figure 1E ). The relative expression of On-LECT2 among these tissues were detected via Western blot and IHC. On-LECT2 was mainly expressed in the brain and head kidney, followed by in the liver and muscle. The lowest expression level of On-LECT2 was observed in the spleen ( Figures 1F, G ).

After S. agalactiae infection, the transcriptional levels of On-lect2 significantly increased (p<0.05) in the brain, head kidney, liver and spleen tissues. Additionally, the highest transcriptional level of On-lect2 in the head kidney, liver, and spleen was reached at 6 h after the challenge, but it was reached at 24 h in the brain after the challenge ( Figure 2 ).

The knockdown efficiency of RNAi plasmids was evaluated by assessing the transcriptional levels of On-lect2 in the liver and spleen via qRT-PCR after RNAi plasmid injection at 3, 5 and 7 d. Unlike in the PBS group, On-lect2 expression significantly decreased (p < 0.05) in all four RNAi plasmid (shRNA-199/209/257/372) injection groups, whereas the control plasmid (shRNA-Control) group was not remarkably different ( Figure 3A ). The silencing efficiency of the RNAi plasmids in the liver was higher than that in the spleen. The levels of On-lect2 of the shRNA-372 group was the fastest and the largest to decrease ( Figure 3A ). The knockdown efficiency of RNAi plasmids was further evaluated by assessing the transcriptional levels of the three interaction genes (PPARγ, MMP2 and β-catenin) of On-lect2 in the liver and spleen via qRT-PCR after shRNA-372 plasmid (the most efficient RNAi plasmid) injection at 5 and 7 d ( Figure 3B ). Compared with that of the PBS group, the transcriptional levels of these genes significantly decreased (p<0.05) for PPARγ or increased (p<0.05) for MMP2 and β-catenin (liver only).

SRs were evaluated to explore the functions of On-LECT2 in tilapia at the macroscopic level. Compared with that of the PBS group, the SR of the shRNA-372 group started decreasing from 6 d after plasmid injection and finally decreased below 53% at 12 d, whereas the SR of the shRNA-Control group was still >90% ( Figure 3C ). The liver of the shRNA-372 group showed necrotic foci, haemorrhage and splenomegaly ( Figure 3D , top lane). The sections stained with H&E showed differences in the features of the liver between the shRNA-Control group and the shRNA-372 group ( Figure 3D , bottom row). Vacuolation were observed in the liver of the shRNA-372 group ( Figure 3D , bottom row).

The effects of rOn-LECT2 on the promotion of phagocytosis of monocytes/macrophages were determined by constructing the prokaryotic recombinant vector pET-N-His-C-His-On-LECT2 and inducing it to produce the recombinant protein (rOn-LECT2). Subsequently, this recombinant protein (with a predicted molecular weight of 20 KDa) was confirmed via SDS-PAGE and Western blot ( Figure 4A ). Furthermore, PBS or rOn-LECT2 was incubated with the premix of macrophages and S. agalactiae. Bacterial number was counted at three time points (2, 4 and 6 h). Compared with that of the PBS group, the bacterial number of the rOn-LECT2 group all significantly decreased (p<0.05) at three time points ( Figure 4B ).

The functions of On-LECT2 in the immune response of Nile tilapia against bacterial infection was further determined by detecting the expression patterns of the potential receptors CD209 and CLR via qRT-PCR. The PBS, shRNA-Control, shRNA-372 and LECT2 groups were challenged with S. agalactiae. The transcriptional levels of CD209 and CLR in the brain, head kidney, liver and spleen tissues in all groups increased after the bacterial challenge ( Figure 5Aa ). Compared with that of the PBS group, CLR expression in the LECT2 group or shRNA-372 group significantly increased (p<0.05) or decreased (p<0.05) at various time points, respectively ( Figure 5Aa ). However, the difference in CD209 expression between the PBS and LECT2 groups was not significant. Interestingly, the expression levels of CD209 in the shRNA-372 group were increased (p<0.05) at different time points in the head kidney, liver and spleen tissues ( Figure 5Aa ).

Inflammatory factors, complements and interferon were also analysed. The transcriptional levels of these genes among the tissues in each group all increased after bacterial infection ( Figures 5Ab and 5Ac ). Moreover, IL-1β and IL-10 significantly increased (p<0.05) or decreased (p<0.05) at different time points in the LECT2 group or the shRNA-372 group, respectively ( Figure 5Ab ). However, the expression levels of TNFα, C3 and C5 were not significantly different among these groups, except for TNFα at a few time points in the liver of the shRNA-372 group compared with that of the PBS group ( Figures 5Ab and 5Ac ). The expression level of IFN-γ1 in the shRNA-372 group significantly increased (p<0.05) at a certain time point in the head kidney and spleen tissues ( Figure 5Ac ).

The SR of the blank, PBS, shRNA-Control, shRNA-372 and LECT2 groups was 100%, 49%, 46%, 13% and 76%, respectively ( Figure 5B ). Furthermore, the bacterial number in the liver of the LECT2 group or the shRNA-372 group significantly decreased (p<0.05) or increased (p<0.05), respectively, compared with that of the PBS group or the shRNA-Control group, respectively. A similar phenomenon was observed in the spleen ( Figure 5C ). The sections stained with H&E showed different levels of pathological changes. These changes were most obvious in the shRNA-372 group and only mild in the LECT2 group ( Figure 6 ).

Similarly, the expression levels of four typical immunological factors, namely, P38, P65, JNK1 and TLR2, each of which represents an immune-related pathway, all increased after bacterial infection ( Figure 7 ). The transcriptional levels of P65, JNK1 and TLR2 in the LECT2 group or the shRNA-372 group significantly increased (p<0.05) or decreased (p<0.05), respectively, compared with that of the PBS group ( Figure 7 ). Remarkably, the difference in the transcriptional level of P38 among these groups was not significant ( Figure 7 ).