Regulation of shrimp prophenoloxidase activating system by lva-miR-4850 during bacterial infection

MicroRNAs (miRNAs) suppress gene expression and regulate biological processes. Following small RNA sequencing, shrimp hemocytes miRNAs differentially expressed in response to acute hepatopancreatic necrosis disease (AHPND) caused by Vibrio parahaemolyticus (VPAHPND) were discovered and some were confirmed by qRT-PCR. VPAHPND-responsive miRNAs were predicted to target several genes in various immune pathways. Among them, lva-miR-4850 is of interest because its predicted target mRNAs are two important genes of the proPO system; proPO2 (PO2) and proPO activating factor 2 (PPAF2). The expression of lva-miR-4850 was significantly decreased after VPAHPND infection, whereas those of the target mRNAs, PO2 and PPAF2, and PO activity were significantly upregulated. Introducing the lva-miR-4850 mimic into VPAHPND-infected shrimps caused a reduction in the PO2 and PPAF2 transcript levels and the PO activity, but significantly increased the number of bacteria in the VPAHPND targeted tissues. This result inferred that lva-miR-4850 plays a crucial role in regulating the proPO system via suppressing expression of PPAF2 and PO2. To fight against VPAHPND infection, shrimp downregulated lva-miR-4850 expression resulted in proPO activation.


Results
Identification of miRNAs from hemocytes of VP AHPND -challenged shrimps. The global analysis of miRNA expression in shrimps challenged with VP AHPND was performed by sequencing small RNA (sRNA) libraries derived from hemocyte samples collected at 0 and 6 h post-infection (hpi) with VP AHPND . Highthroughput sequencing generated a total of 2,184,366 raw reads, comprised of 931,638 and 1,252,728 raw reads in the 0 and 6 hpi sRNA libraries, respectively (Table S1). A total of 1,921,212 reads from both libraries were high quality sequences that passed the initial quality filters. Among them, the majority of non-redundant sequences were 20-22 nucleotides (nt) long (Fig. 1A). Searching against the NCBI nucleotide database demonstrated that, on average, 25% of the sequences were likely contaminating RNAs (Fig. 1B). Following the removal of these contaminating mRNAs, rRNA, and tRNA homologs, the final sequence count was 62,567 sequences that were mapped to miRBase 22.1. The percentage of matched mature miRNA sequences in the 0 and 6 hpi libraries was 94.98% and 96.00%, respectively. A total of 620 miRNA homologs were identified from both libraries. Of those, only 20 miRNA homologs were identified as putative differentially expressed miRNAs (DEMs) upon VP AHPND infection (Table 1). (C) Total RNA from VP AHPND -infected P. vannamei hemocytes was used as a template for specific stem-loop first-strand cDNA synthesis. The relative expression levels of 10 miRNAs were determined by qRT-PCR at 0, 6, and 24 hpi and normalized against U6 as the internal reference. Data are shown as the mean ± 1 SD, derived from three independent repeats. Asterisks indicate significant differences at P < 0.05 (DMRT).
Prediction of miRNA targets. The function of miRNA on gene expression regulation depends on its ability to directly bind to the target mRNA. Therefore, identification of the target mRNA could provide clues regarding the role of miRNA in the shrimp's immune response against VP AHPND infection. The transcriptome database of VP AHPND -infected P. vannamei 22 was used for mRNA target prediction using CU-Mir (https ://cumir .shrim p-irn.org/), an in-house developed miRNA target prediction program. Although several target genes were predicted, this study emphasized the immune-related genes by targeting VP AHPND -responsive miRNAs. Several shrimp immune-related genes in the groups of heat shock proteins/chaperones, cytokines, blood clotting system, proteinase and proteinase inhibitors, homeostasis/apoptosis, proPO system, oxidative stress, RNAi pathway, antimicrobial peptide, pattern recognition protein/receptor, Toll and immune deficient (IMD) pathways, and endocytosis were predicted to be their targets (Fig. 2).
Expression analysis of the target mRNA of lva-miR4850 in P. vannamei shrimps by qRT-PCR. Based on the miRNA-target prediction, the VP AHPND -responsive miRNA lva-miR4850 was selected for further miRNA/target interaction analysis. Notably, lva-miR4-850 was predicted to target the 3′-untranslated (3ʹ-UTR) of the PO2 gene and the open reading frame (ORF) of the PPAF2 gene of the proPO system (Fig. 3A,B). The expression analysis of putative lva-miR4850 target genes revealed that the PO2 and PPAF2 genes were up-regulated after VP AHPND infection in P. vannamei hemocytes by approximately 1.8-to 3-fold (Fig. 3C,D), and these were negatively correlated to that of lva-miR-4850 (Fig. 1C). The results suggested that PO2 and PPAF2 might be lva-miR-4850 target genes. www.nature.com/scientificreports/ The effect of PPAF2 silencing on PO activity in shrimps. Because the involvement of PPAF2 in the proPO activating cascade has not been reported in P. vannamei, RNA interference (RNAi) was used to evaluate that speculation. First, the efficiency of the PPAF2 knock-down in P. vannamei was confirmed by RT-qPCR at 48 h post-injection with PPAF2-dsRNA (Fig. 3E). The PO activity in the hemolymph of PPAF2-knocked down shrimps at 48 h post-injection with PPAF2-dsRNA was declined compared to that in the control groups injected with either GFP-dsRNA or NaCl (Fig. 3F). This implied that PPAF2 plays a role in the PO cascade of P. vannamei immunity.
Confirmation of target mRNA of lva-miR-4850 by dual-luciferase reporter assay. To confirm the miRNA/target interaction, two pmirGLO vectors, pmirGLO-PO2 and pmirGLO-PPAF2, containing DNA fragments corresponding to the putative miRNA-binding region of the PO2 and PPAF2, respectively, were constructed, Also, the mutated seed region constructs, pmiRGLO-PO2-mutant and pmiRGLO-PPAF2-mutant, were prepared. All reporter plasmids were then co-transfected into HEK293-T cells with either a lva-miR-4850 mimic or a lva-miR-4850 scramble. In the presence of the lva-miR-4850 mimic, the luciferase activity observed from cells transfected with pmirGLO-PO2 (Fig. 3G) and pmirGLO-PPAF2 (Fig. 3H) were reduced by around 25% compared to that with the corresponding control mutant construct. The reduction in firefly luciferase expression indicates the binding of lva-miR-4850 to the cloned miRNA target sequence. On the other hand, the mutated seed sequence of lva-miR-4850 did not affect luciferase activity compared to that of the control group. These results indicated that the PO2 and PPAF2 were specific target genes of lva-miR-4850.
On the other hand, shrimps injected with AMO-lva-miR-4850, to inhibit lva-miR-4850, and then challenged with VP AHPND exhibited a significant (approximately two-fold) reduction in the lva-miR-4850 expression level compared to the control groups. Considering the PO2 and PPAF2 expression levels, an inverse relationship between lva-miR-4850, PO2, and PPAF2 was observed. The PO2 and PPAF2 expression levels were significantly  (C) Shrimps were injected as above, 24 h later infected with 1 × 10 6 CFU/mL VP AHPND by immersion, and then at 24 hpi the stomach and HP were harvested to determine the number of VP AHPND by dotting on TCBS agar and subsequently counting the total number of viable green colonies (CFUs). Data are shown as the mean ± 1 SD, derived from three independent repeats. Lowercase letters and asterisks indicate significant differences at P < 0.05 (DMRT). www.nature.com/scientificreports/ increased (approximately 2-and 1.5-fold, respectively) in AMO-lva-miR-4850-injected shrimps as compared to the controls (Fig. 4A). In addition, shrimps injected with AMO-lva-miR-4850 and then challenged with VP AHPND had an approximately 1.5-fold higher proPO activity than in the control groups (Fig. 4B).

Effect of mimic and AMO lva-miR-4850 on the number of bacteria in the stomach and hepatopancreas (HP) of VP AHPND -infected P. vannamei.
We further investigated the effect of mimic-lva-miR-4850-and AMO lva-miR-4850-injection on the number of bacterial cells in VP AHPND targeted shrimp tissues, such as the stomach and HP. The amount of total Vibrio sp. in the shrimp's stomach and HP was counted at 24 hpi with VP AHPND , which was 48 h after injection with one of mimic-lva-miR-4850, scramble mimic-lva-miR-4850, AMO-lva-miR-4850, scramble AMO-lva-miR-4850, or 0.85% (w/v) NaCl injection. The number of green colonies (CFU/ml), representing Vibrio, in both the stomach and HP increased in shrimp injected with mimic-lva-miR-4850. The number of Vibrio in the AMO-lva-miR-4850-injected shrimps was approximately tenfold lower than in the scramble AMO-lva-miR-4850-or 0.85% (w/v) NaCl-injected groups (Fig. 4C). Taken together, we conclude that down-regulation of lva-miR-4850 upon VP AHPND infection allowed the PO2 and PPAF2 to be expressed, resulting in proPO system activation and melanization (Fig. 5).

Discussion
Although miRNAs have been reported to play crucial roles in regulating immune responses against virus infection in crustaceans 18 , their role in antibacterial responses has been reported previously 23 . DEMs of Vibrio-challenged animals have been widely studied. In zebrafish larvae, the miRNAs and mRNAs expression profiles were analyzed upon V. parahaemolyticus infection. It was found that 37 known zebrafish miRNAs were differentially expressed in the infection group. Among them, dre-miR-205-3p, dre-miR-141-5p, dre-miR-200a-5p, dre-miR-92a-2-5p, dre-miR-192, and dre-miR-1788 may play important roles in the innate immune response by regulating target immune genes 24 . In Scylla paramamosain, 161 miRNAs were found to be significantly differentially expressed during V. parahaemolyticus challenge and their potential targets were immune-related genes 25 . In P. monodon, miR-4286 and miR-107b were significantly changed after VP AHPND infection which might regulate dystrophin expression, calcium concentration upon infection 26 . In P. vannamei, a total of 83 miRNAs were significantly differentially expressed in hemocytes upon VP AHPND infection 27 . However, there are still mysteries about the roles of miRNAs in regulating the immune response during VP AHPND infection.
In this study, we analyzed expression of VP AHPND -responsive miRNAs in P. vannamei hemocytes to better understand the function of miRNAs in shrimp antibacterial immunity using sRNA-Seq. In our study, 20 miRNAs were identified as a differentially expressed miRNA homologs in shrimp hemocytes upon VP AHPND infection, and they were identified by searching for homologs against the vertebrate and invertebrate miRNAs, whereas  27 were identified based only on the data of invertebrate species. Comparing these two sRNA-Seq data sets, only lva-miR-71 shared a similar expression profile. In our study, four out of 10 miRNAs (lva-miR-9000, lva-miR-7170-5p, lva-miR-92a-3p, and lva-miR-2169-3p) were confirmed to be significantly differentially expressed in shrimp hemocytes upon VP AHPND infection at 6 hpi, as identified from the sRNA-Seq data. It is known that high-throughput sequencing tends to generate false-negative results; therefore, qRT-PCR is generally used to confirm the expression of transcript [28][29][30] . The expression levels of 10 miRNAs were differentially expressed after VP AHPND challenge.
In combination with the previous transcriptomic data of VP AHPND -infected P. vannamei hemocytes 22 , we analyzed the data based on the negative correlation in gene and miRNA expression and the complementary miRNA/mRNA target prediction to further define the miRNA/mRNA interaction involved in antibacterial responses. According to Zheng et al. (2018) 27 , 12 miRNAs and their predicted target genes are possibly involved in modulating several immune-related processes in the pathogenesis of AHPND. In our study, we found that VP AHPND -responsive miRNAs might regulate several shrimp immune genes involved in heat shock proteins/ chaperones, cytokines, blood clotting system, proteinase and proteinase inhibitors, homeostasis/apoptosis, proPO system, oxidative stress, RNAi pathway, antimicrobial peptide, pattern recognition protein/receptor, Toll and IMD pathways, and endocytosis (Fig. 2).
Based on the miRNA target function, we were interested in the down-regulated miRNAs in VP AHPND -infected P. vannamei, especially lva-miR-4850, whose target genes (PO2 and PPAF2) are major components of the proPO system. Upon bacterial infection, the shrimp proPO system is crucial in shrimp antibacterial immunity. It produces melanin and cytotoxic intermediates for bacterial sequestration 31 . The proPO is a key enzyme in the melanization cascade that also participates in cuticle sclerotization, wound healing, and pathogen killing 32 .
The PPAFs, also known as proPO-activating enzymes, are the terminal components of the proPO activation that directly convert proPO into PO, which then catalyzes the oxidation of phenolic compounds to form melanin 33 . The PPAFs are members of the CLIP subfamily, specific serine proteases that are characterized by the presence of one or more disulfide bond patterns named clip domains in the N-terminus 34,35 . In the shrimp P. monodon, PmPPAF was up-regulated in response to WSSV infection and played an important role in the activation of the PO system 36 . Although PPAF2 has not been characterized in P. vannamei, its expression level was found to be significantly increased after VP AHPND infection in P. vannamei hemocytes, while PPAF2 suppression by RNAi decreased the PO activity. This suggested that the PPAF2 of P. vannamei is involved in the pro-PO activation pathway.
The functions of miRNAs in regulating the proPO system have been reported in shrimps. Ten miRNAs (let-7, miR-184, miR-1, miR-275, miR-9a, miR-279, miR-965, miR-71*, miR-71, and miR-8*) were found to be up-regulated when the hemocytic PO activity was inhibited, whereas they were down-regulated when the PO activity was activated. These results suggest that these miRNAs played important roles in the negative regulation of the proPO system 37 .
The pmo-miR-315 in P. monodon was reported to enhance viral replication by regulating the proPO system through the inhibition of PmPPAE3 gene expression 38 . As expected, a negative correlation in the expression level of lva-miR-4850 and the target genes, PO2 and PPAF2, was observed in this study in P. vannamei hemocytes after VP AHPND infection, while a decreased bacterial number in the HP and stomach of VP AHPND -infected shrimps after AMO-lva-miR-4850 injection was observed. On the other hand, following VP AHPND infection the bacterial number in mimic-lva-miR-4850 challenged P. vannamei was higher than that in those challenged with exogenous lva-miR-4850. These indicated that lva-miR-4850 plays a crucial role in modulating the two key genes of the proPO system. In conclusion, lva-miR-4850 was down regulated in shrimp hemocytes upon VP AHPND infection, which allows PO2 and PPAF2 to be expressed and so activation of the proPO system.

Methods
Shrimp samples. Healthy shrimps, weighing 2-4 g, were obtained from a commercial farm at Petchaburi province (Thailand) and acclimatized in rearing tanks at ambient temperature (30 ± 2 °C), water salinity of 20 parts per thousand, and constant aeration before use in the experiments.
Mimic, scramble mimic, anti-miRNA oligonucleotide (AMO), and AMO scramble RNA. The mimic, scramble mimic, AMO, and AMO scramble RNA of lva-miR-4850 used for in vitro and in vivo experiments were synthesized by the Shanghai GenePharma Co., Ltd., P.R. China (Table S2).
Bacterial challenge experiments. The VP AHPND inoculum was prepared by culturing the bacteria over- sRNA-Seq and data analysis. The cDNA libraries of sRNA from VP AHPND -infected shrimp hemocytes at 0 and 6 hpi were constructed using the TruSeq Small RNA Library Preparation Kit (Illumina) according to the manufacturer's instruction. The indexed libraries were normalized, pooled, and then, sequenced with a PhiX control spiked at 7.5% using MiSeq Reagent Kits v2 (Illumina) in a MiSeq sequencer (Illumina). The sRNA-Seq analysis was performed as previously described by Boonchuen et al. (2020) 22 . Briefly, the Galaxy instance (https ://usega laxy.org/) was used for 5′-and 3′-adapter trimming and for quality control of raw reads 40 . The highquality sRNA sequences of a length shorter than 18 nucleotides and longer than 24 nucleotides were removed. The contaminating RNA, such as mRNA, rRNA, and tRNA, was also removed. The remaining sequences were then searched against miRBase 22.1 (http://www.mirba se.org/) to identify known miRNA homologs. Differentially expressed miRNA (DEM) analysis was performed as follows. Firstly, the read no. of each miRNA from the treatment and control groups were normalized to the total no. of reads of that respective library at the same orders of magnitude. Formula: Secondly, the normalized expression level was used to calculate the fold change and P-value 41 . Finally, we used the P-value method to ensure the statistical significance of DEMs, which of interest were generated using P-value P < 0.05 and log 2 Ratio ≥ ± 1.

Quantitative (q)RT-PCR analysis.
The miRNAs of interest, such as lva-miR-9000, lva-miR-D9-3p, lva-miR-4850, lva-miR-2169-3p, lva-miR-7170-5p, lva-miR-4901, lva-miR-92a-3p, lva-miR-92b-5p, lva-miR-184, and lva-miR-8522c, were selected for expression analysis using stem-loop qRT-PCR. Pooled total sRNA from VP AHPND -infected shrimp hemocytes at 0, 6, and 24 hpi was prepared using the mirVana miRNA Isolation Kit (Ambion, Life technologies). The total sRNA as a template, the stem-loop RT primers specific to each miRNA, and the internal control U6 (Table S3) were used to synthesize the first-strand cDNA using the RevertAid Firststrand cDNA Synthesis Kit (Thermo Fisher Scientific). Expression of the U6 gene was used as the internal control. Stem-loop qRT-PCR was performed using an appropriate amount of cDNA for each gene, specific oligonucleotide forward primer (Table S3), and QPCR Green Master Mix (Biotechrabbit) in the MiniOpticon™ Real-time PCR System (Bio-Rad). Thermal cycling was performed under the following conditions: 95 °C for 3 min followed by 40 cycles of 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s. The relative expression level compared to that of U6 was calculated.
Prediction of miRNA targets. The miRNA targets were identified by searching for the miRNA complementary sequence in the transcriptome data of VP AHPND -challenged P. vannamei hemocytes 22 using the CU-Mir software (http://shrim p-irn.org/mirta rget/index .php) that was developed in-house 42 . The CU-Mir software was used to search for locations on mRNA targets that seed sequences (2-8 nucleotides from the 5′ end) of miRNA that can bind perfect complementary or one mismatch at any different region; an open reading frame (ORF); 3′-untranslated region (UTR) and 5′-UTR. The percent complementary between sequences was calculated from the number of nucleotides that complementarily bind to the target mRNAs per total length of the miRNA sequence. The overall complementarity of miRNA to the target mRNA cut off was set at 40%. RNAhybrid software (http://bibis erv.techf ak.uni-biele feld.de/rnahy brid/) was also used to predict genes targeted by miRNAs using a free energy of < − 15.0 kcal/mol 42 .
Dual-luciferase reporter assay. The luciferase reporter system was used to confirm the interaction between the miRNA of interest, lva-miR-4850, and the target PO2 and PPAF2 genes. The gene fragments containing the predicted lva-miR-4850 target sites, such as the 3′UTR of the PO2 gene and the ORF of the PPAF2 gene, were amplified from the cDNA of the VP AHPND -infected P. vannamei hemocytes using specific primers (Table S3). Those gene fragments were subsequently cloned into the pmiRGLO plasmid (Promega) at the 3′UTR of firefly luciferase and pmiRGLO-stop-mutant to produce the pmirGLO-PO2 and pmirGLO-PPAF2 plasmids, respectively.
To construct the experimental control, the binding element at nucleotide positions 2-8 was mutated by QuickChange II XL Site-Directed Mutagenesis Kit (Agilent Technology). Briefly, the primers were designed by switching the bases of the seed sequence from purine to pyrimidine or pyrimidine to purine with a melting temperature (Tm) of ≥ 78 °C. The recombinant pmiRGLO-PO2 and pmiRGLO-PPAF2 were then amplified by PfuUltra-HF DNA polymerase (Agilent Technology) and the PCR product was treated with DpnI restriction enzyme. The DpnI-treated PCR product was further transformed into E. coli Top10. The mutant plasmids, pmiRGLO-PO2-mutant and pmiRGLO-PPAF2-mutant, were confirmed by sequencing ( Figure S1).
For each plasmid, 200 ng were co-transfected into HEK293-T cells along with 20 pmol of mimic lva-miR-4850 or scramble lva-miR-4850 (GenePhama) using the Effectene transfection reagent (Qiagen). After 48 h posttransfection, the activity of Firefly and Renilla luciferases were measured using the Dual-Luciferase Reporter assay system (Promega) following the manufacturer's instructions. www.nature.com/scientificreports/ cDNA of P. vannamei hemocytes using a specific primer pair (Table S3), cloned into the pGEM-T easy vector (Promega), and used as a template for the preparation of the dsRNA specific to the PPAF2 (dsRNA-PPAF2).
In addition, the dsRNA of the green fluorescent protein (GFP; dsRNA-GFP), the negative control, was prepared from the pEGFP-1 vector (Clontech) as a template. The dsRNA-PPAF2 and dsRNA-GFP were prepared using T7 RiboMAX Express Large-Scale RNA Production System (Promega) according to the manufacturer's instruction. The primers used for dsRNA-PPAF2 (knPPAF2-T7-F, knPPAF2-R, knPPAF2-F, and knPPAF2-T7-R) and dsRNA-GFP (knGFP-F, knGFP-R, knGFP-T7-F, and knGFP-T7-R) production are listed in Table S3. The quantity and quality of dsRNA were verified by nanodrop spectrophotometry and agarose gel electrophoresis, respectively. P. vannamei of approximately 3 g body weight were divided into three groups of three individuals each. The first group (control) was injected with 150 mM NaCl, the second group (dsRNA control) with 5 μg/g shrimp of dsRNA-GFP, whilst the third group, the PPAF2 knockdown, was injected with 5 μg/g shrimp of dsRNA-PPAF2. The hemolymph of individual shrimp was collected at 48 h post-injection. The total RNA from shrimp hemolymph was extracted by Genezol reagent (Geneaid). Subsequently, the first strand cDNA synthesis was performed. Suppression of PPAF2 expression was determined by qRT-PCR using specific primers. The elongation factor-1α (EF-1α) was used as an internal control.
In vivo effect of lva-miR-4850 introducing and gene silencing. Shrimps (2-3 g) were divided into five groups of three individuals each. Experimental groups were intramuscularly injected with 50 µL of 0.85% (w/v) NaCl solution containing 2 nmol of mimic-lva-miR-4850 or AMO-lva-miR-4850, while the control groups were injected with 2 nmol of scramble mimic-lva-miR-4850, or scramble AMO-lva-miR-4850, or 50 µL of 0.85% (w/v) NaCl. At 24 h post-injection, the shrimps were challenged with VP AHPND as described above and at 24 hpi their hemolymph was collected. Total RNA was extracted and used for first-strand cDNA production. The expression levels of lva-miR-4850, PO2, and PPAF2 were determined by qRT-PCR, as described above.
On the other hand, the stomach and HP were separately collected from three shrimps per group, homogenized, and serially tenfold diluted in sterile 0.85% (w/v) NaCl. The diluted (10 1 -to 10 6 -fold) samples were plated onto thiosulfate-citrate-bile salts-sucrose (TCBS) agar and incubated at 30 °C for 12-14 h. The bacterial colonies were then counted and calculated as CFU/mL. Determination of the PO activity. The PO activity was determined in the shrimp hemolymph collected at 24 hpi with miRNA-VP AHPND challenge (48 h post-injection with dsRNA). The PO activity was measured using a modification to the reported method 43 . In brief, 50 μL of hemolymph was mixed with 25 μL of 3 mg/mL freshly prepared L-3, 4-dihydroxyphenylalanine (L-DOPA; Fluka), and 25 μL of 20 mM Tris-HCl (pH 8.0). The absorbance at 490 nm (A 490 ) was monitored after 60 min of incubation. The amount of hemolymph proteins was measured using the Bradford method 44 . The PO activity was recorded as A 490 /mg total protein/min. Statistical analysis. Differences in the dual-luciferase reporter assay were analyzed using the GraphPad Prism 8.0 software with the statistical analyses, including a paired-samples t test, while miRNA and gene expression levels were analyzed using a paired-samples t test. All other numerical data was analyzed using one-way ANOVA followed by Duncan's new multiple range test (DMRT), and are presented as the mean ± one standard deviation (SD). Statistical significance was accepted at the P < 0.05 level.
Ethics statement. Experiments involving animals were performed in compliance with the animal use protocol number 1823006 approved by the Chulalongkorn University Animal Care and Use Committee (CU-