LvCD14L Acts as a Novel Pattern Recognition Receptor and a Regulator of the Toll Signaling Pathway in Shrimp

Leucine-rich repeat (LRR) is a structural motif has important recognition function in immune receptors, such as Tolls and NOD-like receptors (NLRs). The immune-related LRR proteins can be divided into two categories, LRR-containing proteins and LRR-only proteins. The latter contain LRR motifs while they are without other functional domains. However, the functional mechanisms of the LRR-only proteins were still unclear in invertebrates. Here, we identified a gene encoding a secretory LRR-only protein, which possessed similarity with vertebrate CD14 and was designated as LvCD14L, from the Pacific whiteleg shrimp Litopenaeus vannamei. Its transcripts in shrimp hemocytes were apparently responsive to the infection of Vibrio parahaemolyticus. Knockdown of LvCD14L with dsRNA resulted in significant increase of the viable bacteria in the hepatopancreas of shrimp upon V. parahaemolyticus infection. Further functional studies revealed that LvCD14L could bind to microorganisms’ PAMPs, showed interaction with LvToll1 and LvToll2, and regulated the expression of LvDorsal and LvALF2 in hemocytes. These results suggest that LvCD14L functions as a pattern recognition receptor and activates the NF-κB pathway through interaction with LvTolls. The present study reveals a shrimp LvCD14L-Tolls-NF-κB signaling pathway like the CD14/TLR4/NF-κB signaling pathway in mammalians, which enriches the functional mechanism of secretory LRR-only immune receptors during pathogens infection in invertebrates.


Introduction
Shrimp is an important aquaculture species in China. In recent years, bacterial diseases have seriously affected the healthy development of the shrimp aquaculture industry. As an invertebrate, shrimp mainly relies on innate immunity to resist pathogen infections. Recognition of pathogens is an important step of the immune response of hosts. In innate immunity, the host initiates an immune response by pattern recognition receptors (PRRs) recognizing pathogen-associated molecular patterns (PAMPs). Therefore, identification of novel PRRs in shrimp is crucial for understanding the host immune response.
Leucine-rich repeat (LRR) is a structural motif which widely exists in thousands of proteins from viruses to eukaryotes [1][2][3]. An LRR motif usually consists of 20 to 30 residues forming a β strand-α helix structure [4]. LRR proteins, containing two or more LRR motifs, provide a curved solenoid structural framework for protein-protein interaction, and their concave surfaces usually contain ligand-binding sites [4]. In addition to LRR motifs, N-terminal (LRRNT) and C-terminal (LRRCT) capping motifs are common motifs in extracellular and membrane-associated LRR proteins for protecting the hydrophobic inner core of the solenoid [5,6]. Both capping motifs usually contain four cysteines, which

LvCD14L Encodes a Secretory LRR-Only PROTEIN
The open reading frame of LvCD14L had 918 bp nucleotides, encoding a protein of 310 amino acid residues ( Figure 1A). The amino acid sequence of LvCD14L contained a 28 aa signal peptide, a LRRNT region (Ser 32 -Cys 69 ), seven LRR regions (Val 76 -Ile 86 , Leu 100 -Leu 110 , Leu 124 -Leu 134 , Leu 147 -Ile 157 , Leu 172 -Val 182 , Leu 194 -Ile 204 and Pro 218 -Leu 228 ) and a hydrophobic tail (Val 231 -Phe 265 ), while it lacked the LRRCT region of LRR protein, which was like CD14 from vertebrates ( Figure 1B). In the hydrophobic tail, the hydrophobic amino acid residues account for 60% of the total amino acid residues. Sequence alignment analysis showed that the amino acid sequence of LvCD14L had a similarity ranging from 19.44% to 24.12% with CD14s, and had a similarity ranging from 26.32% to 33.85% with VLRs from jawless vertebrates ( Figure 1D). The three-dimensional structure showed that LvCD14L adopted a horseshoe-shaped structure, which was similar to other LRR proteins ( Figure 1C). LvCD14L adopted a horseshoe-shaped structure, which was similar to other LRR proteins ( Figure 1C). The start codon and stop codon were boxed. The predicted signal peptide was underlined. The LRRNT, LRR and hydrophobic tail were marked with green, yellow and pink background. The hydrophobic amino acid residues in hydrophobic tail were marked with a wavy line. (B) The domain composition of LvCD14L. SP, signal peptide; LRRNT, N-terminal LRR domain; LRR, Leucine-rich repeat; Hydrophobic tail, hydrophobic amino acid-rich region. (C) Multiple sequences alignment of VLR genes and CD14 genes from different species. Identical and similar residues were marked with black, blue and yellow, respectively. Sequence information of VLR and CD14 genes is shown in Table S3. (D) The three-dimensional (3D) structure modeling of LvCD14L. The model of LvCD14L adopted a horseshoe-shaped structure, which was similar to other LRR proteins.
The subcellular localization analysis showed that LvCD14L was predicted as a cell membrane or extracellular protein. The cellular localization experiment further confirmed that LvCD14L was an extracellular protein. The recombinant protein consisting of LvCD14L signal peptide (SP), EGFP and LvCD14L hydrophobic tail (HT) had a weak signal in cytoplasm ( Figure 2C) compared with the signal of EGFP (Figure 2A), which was similar to the recombinant protein consisting of SP and EGFP ( Figure 2B). No positive signal was detected on the cell membrane when transfected with the SP + EGFP + HT plasmid. The result of the Western blot showed that the recombinant protein of SP + EGFP and SP + EGFP + HT could be detected in the cell lysate and medium, while the EGFP was detected in the cell lysate ( Figure 2D).
repeat; Hydrophobic tail, hydrophobic amino acid-rich region. (C) Multiple sequences alignment o VLR genes and CD14 genes from different species. Identical and similar residues were marked wit black, blue and yellow, respectively. Sequence information of VLR and CD14 genes is shown i Table S3. (D) The three-dimensional (3D) structure modeling of LvCD14L. The model of LvCD14 adopted a horseshoe-shaped structure, which was similar to other LRR proteins.
The subcellular localization analysis showed that LvCD14L was predicted as a ce membrane or extracellular protein. The cellular localization experiment further confirme that LvCD14L was an extracellular protein. The recombinant protein consisting o LvCD14L signal peptide (SP), EGFP and LvCD14L hydrophobic tail (HT) had a weak sig nal in cytoplasm ( Figure 2C) compared with the signal of EGFP (Figure 2A), which wa similar to the recombinant protein consisting of SP and EGFP ( Figure 2B). No positiv signal was detected on the cell membrane when transfected with the SP + EGFP + H plasmid. The result of the Western blot showed that the recombinant protein of SP + EGF and SP + EGFP + HT could be detected in the cell lysate and medium, while the EGFP wa detected in the cell lysate ( Figure 2D).

LvCD14L Participates in V. parahaemolyticus Infection
Tissue distribution analysis showed that LvCD14L was widely expressed in all detected tissues, with the highest expression level in the hepatopancreas, followed by the hemocytes, stomach, epidermis, intestine, gill and lymphoid organ (Oka) ( Figure 3A). The time-course expression patterns of LvCD14L in the hepatopancreas and hemocytes of shrimp after V. parahaemolyticus challenge were analyzed. In the hepatopancreas, the expression level of LvCD14L had no significant difference at all detected time points between the challenge group and the control group ( Figure 3B). In hemocytes, the expression level of LvCD14L was up-regulated by 10.16-fold and 4.14-fold at 12 h and 24 h after V. parahaemolyticus challenge, respectively ( Figure 3C).
V. parahaemolyticus propagation in shrimp was analyzed. LvCD14L was knocked down by dsRNA at the dosage of 6 µg per individual, with an interference efficiency of 69.09% and 87.52% in the hepatopancreas and hemocytes, respectively ( Figure 3D). The number of bacteria in the hepatopancreas was an indicator for the health condition of shrimp. To study the impact of LvCD14L silencing on the V. parahaemolyticus infection process, the amount of Vibrio in the hepatopancreas of LvCD14L-silenced shrimp was detected. At 24 h after pathogen injection, the amount of V. parahaemolyticus in hepatopancreas of shrimp from the dsLvCD14L group was 2.33 × 10 4 cfu/g, which was 6.59-fold higher than that from the dsEGFP group (3.53 × 10 3 cfu/g) ( Figure 3E). The number of other bacteria, including V. harveyi and Vibrio brasiliensis, did not show a significant difference between the dsLvCD14L group and control groups. The count of neither V. parahaemolyticus nor other bacteria had differences between the dsEGFP group and the PBS group. The results suggested that LvCD14L participated in V. parahaemolyticus infection.  and hemocytes (C). Significant differences between treatment and control groups were labeled with an asterisk at p < 0.05. (D) Silencing efficiency of LvCD14L dsRNA in hepatopancreas and hemocytes. "dsLvCD14L" and "dsEGFP" indicate the group injected with LvCD14L dsRNA and EFGP dsRNA. The effective silencing dose is marked with an asterisk at p < 0.05 (n = 9). (E) The total viable bacteria counts in the hepatopancreas of LvCD14L-silenced shrimp after V. parahaemolyticus injection. The "V.ph" indicates the amount of V. parahaemolyticus in shrimp from different treatments. The "other bacteria" indicates the amount of bacteria, including Vibrio harveyi and Vibrio brasiliensis. The data were obtained from three independent repeats. Significant differences between the amount of V. parahaemolyticus in shrimps of treatment and control groups are labeled with an asterisk at p < 0.05 (n = 9).
To detect the immune function of LvCD14L, the influence of LvCD14L knockdown on V. parahaemolyticus propagation in shrimp was analyzed. LvCD14L was knocked down by dsRNA at the dosage of 6 µg per individual, with an interference efficiency of 69.09% and 87.52% in the hepatopancreas and hemocytes, respectively ( Figure 3D). The number of bacteria in the hepatopancreas was an indicator for the health condition of shrimp. To study the impact of LvCD14L silencing on the V. parahaemolyticus infection process, the amount of Vibrio in the hepatopancreas of LvCD14L-silenced shrimp was detected. At 24 h after pathogen injection, the amount of V. parahaemolyticus in hepatopancreas of shrimp from the dsLvCD14L group was 2.33 × 10 4 cfu/g, which was 6.59-fold higher than that from the dsEGFP group (3.53 × 10 3 cfu/g) ( Figure 3E). The number of other bacteria, including V. harveyi and Vibrio brasiliensis, did not show a significant difference between the dsLvCD14L group and control groups. The count of neither V. parahaemolyticus nor other bacteria had differences between the dsEGFP group and the PBS group. The results suggested that LvCD14L participated in V. parahaemolyticus infection.

LvCD14L Acts as a Pattern Recognition Receptor
The recombinant LvCD14L protein (rLvCD14L) was expressed in E. coli after IPTG induction with a predicted molecular mass of 50.04 kDa ( Figure 4A, lane 1 and 2). The induced rLvCD14L protein was expressed in both the inclusion body and soluble form ( Figure 4A, lane 3 and 4) and was then purified ( Figure 4A, lane 5). The results of ELISA showed that the rLvCD14L protein possessed concentration-dependent binding activities to PGN, LPS and dextran, respectively ( Figure 4B-D). The MIC assay and agglutinating assay showed that rLvCD14L did not have antimicrobial activity (Table S1) or agglutinating activity ( Figure S1). These results suggested that LvCD14L functioned as a PRR through direct binding to PAMPs. total viable bacteria counts in the hepatopancreas of LvCD14L-silenced shrimp after V. parahaemolyticus injection. The "V.ph" indicates the amount of V. parahaemolyticus in shrimp from different treatments. The "other bacteria" indicates the amount of bacteria, including Vibrio harveyi and Vibrio brasiliensis. The data were obtained from three independent repeats. Significant differences between the amount of V. parahaemolyticus in shrimps of treatment and control groups are labeled with an asterisk at p < 0.05 (n = 9).

LvCD14L Acts as a Pattern Recognition Receptor
The recombinant LvCD14L protein (rLvCD14L) was expressed in E. coli after IPTG induction with a predicted molecular mass of 50.04 kDa ( Figure 4A, lane 1 and 2). The induced rLvCD14L protein was expressed in both the inclusion body and soluble form ( Figure 4A, lane 3 and 4) and was then purified ( Figure 4A, lane 5). The results of ELISA showed that the rLvCD14L protein possessed concentration-dependent binding activities to PGN, LPS and dextran, respectively ( Figure 4B-D). The MIC assay and agglutinating assay showed that rLvCD14L did not have antimicrobial activity (Table S1) or agglutinating activity ( Figure S1). These results suggested that LvCD14L functioned as a PRR through direct binding to PAMPs.

LvCD14L Activates the TLR-Dorsal Pathway through Binding to LvTolls
As a PRR, LvCD14L might activate immune pathways when the host encounters pathogens infection. Therefore, we detected the expression changes of several immunerelated transcription factors and antimicrobial peptide genes when LvCD14L was knocked down. The results showed that the NF-κB transcription factor LvDorsal was down-regulated in hemocytes after LvCD14L knockdown ( Figure 5A). Simultaneously, the antimicrobial peptide gene LvALF2 was also down-regulated in hemocytes ( Figure 5A). Overexpression of LvCD14L further confirmed the regulation of LvCD14L on the NF-κB pathway. Injection of rLvCD14L significantly up-regulated the expression level of LvDorsal in hemocytes at 3 hpi, 6 hpi and 12 hpi compared with that in the rTrx injection group ( Figure  5B). The expression level of LvALF2 in hemocytes was up-regulated at 6 hpi, 12 hpi and 24 hpi ( Figure 5C). These results suggested that LvCD14L could activate the NF-κB pathway in shrimp during V. parahaemolyticus infection.

LvCD14L Activates the TLR-Dorsal Pathway through Binding to LvTolls
As a PRR, LvCD14L might activate immune pathways when the host encounters pathogens infection. Therefore, we detected the expression changes of several immunerelated transcription factors and antimicrobial peptide genes when LvCD14L was knocked down. The results showed that the NF-κB transcription factor LvDorsal was down-regulated in hemocytes after LvCD14L knockdown ( Figure 5A). Simultaneously, the antimicrobial peptide gene LvALF2 was also down-regulated in hemocytes ( Figure 5A). Overexpression of LvCD14L further confirmed the regulation of LvCD14L on the NF-κB pathway. Injection of rLvCD14L significantly up-regulated the expression level of LvDorsal in hemocytes at 3 hpi, 6 hpi and 12 hpi compared with that in the rTrx injection group ( Figure 5B). The expression level of LvALF2 in hemocytes was up-regulated at 6 hpi, 12 hpi and 24 hpi ( Figure 5C). These results suggested that LvCD14L could activate the NF-κB pathway in shrimp during V. parahaemolyticus infection. extracellular immune signal inside the cells. The Dorsal is the core transcription factor of the TLR-Dorsal signaling pathway. Therefore, we guessed that LvCD14L might bind to the extracellular part of LvTolls. As LvToll1 was reported to be responsive to Vibrio infection and has a close evolutionary relationship with LvToll2, we detected the interaction of LvCD14L with LvToll1 and LvToll2. The Co-IP assay showed that LvCD14L could interact with LvToll1 and LvToll2 ( Figure 5D). The results suggested that LvCD14L could bind to LvTolls and activate the NF-κB pathway in shrimp upon pathogens infection. . The Co-IP results confirmed by Western blot using anti-V5 antibody as a probe. The input samples were detected by Western blot using anti-V5 and anti-FLAG antibody as a probe, respectively. Results were obtained based on three independent repeats.

Discussion
As immune receptors, LRR-only proteins mainly function through recognition of pathogens and activation of subsequent immune responses. In mammalians, CD14 was initially found to bind complexes of LPS-LBP and enhance host responses to LPS [20], then was found to bind PGN and LTA [21,22]. In jawless vertebrates, VLRs can not only specifically recognize protein antigens but also exhibit great specificity for glycan [23]. In invertebrates, some immune-related LRR-only proteins with pattern recognition function can also bind to microorganisms or PAMPs [16,17] as well as affect the expression of genes in the NF-κB pathway including Dorsal and AMPs [18]. In the present study, the LvCD14L protein could bind to different PAMPs, including LPS, PGN and dextran, and affect the As a secretory protein, LvCD14L needs a transmembrane protein to transduce the extracellular immune signal inside the cells. The Dorsal is the core transcription factor of the TLR-Dorsal signaling pathway. Therefore, we guessed that LvCD14L might bind to the extracellular part of LvTolls. As LvToll1 was reported to be responsive to Vibrio infection and has a close evolutionary relationship with LvToll2, we detected the interaction of LvCD14L with LvToll1 and LvToll2. The Co-IP assay showed that LvCD14L could interact with LvToll1 and LvToll2 ( Figure 5D). The results suggested that LvCD14L could bind to LvTolls and activate the NF-κB pathway in shrimp upon pathogens infection.

Discussion
As immune receptors, LRR-only proteins mainly function through recognition of pathogens and activation of subsequent immune responses. In mammalians, CD14 was initially found to bind complexes of LPS-LBP and enhance host responses to LPS [20], then was found to bind PGN and LTA [21,22]. In jawless vertebrates, VLRs can not only specifically recognize protein antigens but also exhibit great specificity for glycan [23]. In invertebrates, some immune-related LRR-only proteins with pattern recognition function can also bind to microorganisms or PAMPs [16,17] as well as affect the expression of genes in the NF-κB pathway including Dorsal and AMPs [18]. In the present study, the LvCD14L protein could bind to different PAMPs, including LPS, PGN and dextran, and affect the expression of genes in the NF-κB pathway, suggesting that LvCD14L functions as a PRR, which recognizes pathogens' PAMPs and activates host immune responses.
The most prominent biological functions of PRRs are recognition of PAMPs and initiation of signal transduction. The membrane-associated TLR recognizes PAMPs or ligands by its extracellular LRR domain and then transduces immune signals into the cells through its intracellular TIR domain interacting with an adaptor protein [24]. The cytoplastic NODlike receptor (NLR), another LRR domain-containing PRR, activates subsequent immune responses using its N-terminal domain upon binding to PAMPs [25]. These kinds of PRRs have typical domains for signal transduction. Secretory PRRs also play critical roles in humoral immunity and can bind and eliminate microbes through multiple mechanisms including agglutination, neutralization, opsonization and complement activation [26][27][28]. In vertebrates, secretory PRRs could also interact with and regulate the function of membraneassociated PRRs such as soluble TLR2 (sTLR2) and sCD14 [29][30][31]. In invertebrates, some secretory LRR-containing proteins, such as CfLRRop-7 from Chlamys farreri and Leureptin from Manduca sexta, were regarded as PRRs because they could bind to PAMPs and induce antimicrobial peptides' expression or hemocyte responses upon infection [32,33]. Vertebrate CD14 has two forms, with one anchored to the cell membrane by GPI and another one secreted into the extracellular space. The signal peptide (SP) and hydrophobic tail (HT) play crucial roles in the cellular localization of glycosylphosphatidylinositol (GPI)-anchored proteins. As LvCD14L has a predicted hydrophobic tail, three plasmids were constructed to test whether LvCD14L was cell membrane-anchored or secreted extracellularly. "EGFP-N1" was a control plasmid, which expressed EGFP intracellularly. "SP + EGFP" was jointly constructed with the nucleotide sequence encoding the signal peptide of LvCD14L and EGFP, which mainly expressed protein secreted extracellularly. As SP and HT should be added at the N-terminal and C-terminal of EGFP, respectively, the whole LvCD14L was not used to jointly construct the plasmid "SP + EGFP + HT", which had a similar expression pattern at the protein level like "SP + EGFP". The Western blot analysis further confirmed the result. In addition, no signal was detected on the cell membrane when transfected with these plasmids. Therefore, LvCD14L was a secretory protein and the predicted hydrophobic tail did not have the function of cell membrane anchoring.
In the present study, we found that LvCD14L modulated the NF-κB pathway and AMP expression, which was consistent with previously reported extracellular LRR-containing PRRs. The NF-κB pathway is a classical immune signal transduction pathway that is mediated by the membrane-associated receptor Tolls and TLRs [34]. The extracellular region of Tolls and TLRs consist of multiple LRR motifs for PAMPs and ligands' binding. Notably, LRR motifs could form homologous or heterogeneous polymers [35]. Therefore, we speculated that LvCD14L protein probably activates intracellular immune responses through interaction with shrimp Tolls. In L. vannamei, a total of eleven Toll genes have been identified and all of them are responsive to V. parahaemolyticus infection [36]. Particularly, functional studies further revealed the involvement of LvToll1 and LvToll2 during Vibrio infection [37,38]. Meanwhile, LvToll1 and LvToll2 could regulate the NF-κB-dependent antimicrobial peptides' (AMPs) expression [39,40]. Therefore, LvToll1 and LvToll2 were considered as the target membrane-associated PRRs for LvCD14L interaction, which was proved by the protein-protein interaction analysis. Therefore, a working model was proposed to illustrate how LvCD14L functions during V. parahaemolyticus infection (Figure 6). After pathogen infection, LvCD14L protein functions as a PRR by recognizing PAMPs and then interacting with LvTolls and activating AMP production. [43]. Tolls in shrimp can directly recognize PAMPs, which is different from that in Drosophila [34,44]. Therefore, researchers proposed that the Toll pathway in shrimp was similar to the mammalian TLR pathway [34]. In the present study, LvCD14L functions similar to the CD14 in mammalians, which supports the previous viewpoint. The present data suggest that there is a LvCD14L-Tolls-NF-κB signaling pathway similar to the CD14/TLR4/NF-κB signaling pathway in mammalians.

Animal and Tissues Collection
Healthy shrimp L. vannamei, with an average body weight of 5.36 g were cultured in an indoor breeding room and fed with an artificial diet. Different tissues including the Figure 6. Model for LvCD14L-mediated antibacterial mechanism against V. parahaemolyticus. LvCD14L sensed V. parahaemolyticus and bound to LvToll1 or LvToll2 and activated NF-κB pathways. The transcription factor Dorsal translocated into the nucleus and led to the transcription of some AMPs such as LvALF2.
As the classical immune receptors in the innate immunity, Tolls or TLRs receive extracellular signals in different ways in invertebrates and mammalians. In Drosophila, Tolls cannot recognize PAMPs directly while they need the cytokine Spätzle [41]. Spätzle was activated by a series of protease cascades initiated from the recognition of PAMPs by molecules such as Gram-negative binding protein (GNBP) 3 and PGN recognition protein (PGRP)-SA [42]. TLR4, the first discovered mammalian homologue of Drosophila Toll, recognizes LPS delivered by LPS binding protein and CD14, and transduces the signal into the cells [9]. To date, despite several Spätzle-like genes being identified in shrimp, there was no evidence that showed these Spätzle-like proteins could interact with shrimp Tolls [43]. Tolls in shrimp can directly recognize PAMPs, which is different from that in Drosophila [34,44]. Therefore, researchers proposed that the Toll pathway in shrimp was similar to the mammalian TLR pathway [34]. In the present study, LvCD14L functions similar to the CD14 in mammalians, which supports the previous viewpoint. The present data suggest that there is a LvCD14L-Tolls-NF-κB signaling pathway similar to the CD14/TLR4/NF-κB signaling pathway in mammalians.

Animal and Tissues Collection
Healthy shrimp L. vannamei, with an average body weight of 5.36 g were cultured in an indoor breeding room and fed with an artificial diet. Different tissues including the intestine, stomach, gill, epidermis, hepatopancreas and lymphoid organ (Oka) were collected from nine individuals. Hemolymph was extracted using a syringe with an equal volume of anticoagulant (27 mM sodium citrate, 336 mM NaCl, 115 mM glucose, 9 mM EDTA, pH 7) and centrifuged at 800× g, 4 • C for 10 min to collect hemocytes. All the samples were preserved in liquid nitrogen for total RNA extraction and gene expression analysis.

Total RNA Extraction and cDNA Synthesis
Total RNA of different tissues was extracted with RNAiso Plus reagent (TaKaRa, Shiga, Japan), following the manufacturer's protocol. The concentration of each RNA sample was measured by Nanodrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA). The quality of RNA samples was assessed by electrophoresis on 1% agarose gel. The cDNA was synthesized from 1 µg total RNA using PrimeScript RT Reagent Kit (TaKaRa, Shiga, Japan), according to the manufacturer's instruction. Briefly, the genomic DNA (gDNA) was removed with gDNA Eraser. Then, the first-strand cDNA was synthesized by PrimeScript RT Enzyme with random primers.

Gene Cloning
The cDNA sequence of LvCD14L was identified from a transcriptome database of L. vannamei [45]. Primers LvCD14L-F and LvCD14L-R (Table S2) were designed to amplify the LvCD14L gene from the cDNA sample of shrimp hepatopancreas. The PCR was performed using the Premix Ex Taq Hot Start version (TaKaRa, Shiga, Japan), and the program contained 35 cycles of 98 • C for 10 s, 60 • C for 30 s and 72 • C for 1.5 min. The PCR product was purified with MiniBEST DNA Fragment Purification Kit (TaKaRa, Shiga, Japan) and constructed into pM19-T vector (TaKaRa, Shiga, Japan), following the manufacturer's instruction. Then, the plasmid was transformed into Trans5α competent cells (TransGen Biotech, Beijing, China) for Sanger sequencing.

Tissue Distribution Analysis
The expression levels of LvCD14L in different tissues were detected by quantitative real-time PCR (qPCR) with THUNDERBIRD SYBR qPCR Mix (TOYOBO, Osaka, Japan) and primers LvCD14L-qF/qR. Primers 18S-F and 18S-R were also synthesized to measure the expression of the internal reference gene 18S rRNA. The program of qPCR contained one cycle of 95 • C for 1 min, followed by 40 cycles of 95 • C for 15 s, 56 • C for 15 s, 72 • C for 30 s and a melting-curve analysis added to the end of each reaction to verify the specificity of the product. The raw data of qPCR were processed with the 2 −∆∆CT method [46]. The primers used in this section are listed in Table S2.

Pathogen Challenge and Gene Expression Analysis
Pathogenic V. parahaemolyticus was cultured in the tryptic soy broth (TSB) medium (LuQiao, Beijing, China) with additional 2% (g/mL) NaCl at 30 • C. The bacteria number was counted using a blood cell counting plate under a microscope and was diluted to 1 × 10 4 CFU/µL with phosphate buffered saline (PBS) buffer. In total, 120 shrimp were equally divided into the infection group and the control group. The shrimp in the infection group were injected individually with 2 × 10 5 CFU V. parahaemolyticus. The shrimp in the control group were injected with an equal volume of PBS buffer. Different tissues including hepatopancreas and hemocytes from five individuals were collected as one sample at 3 h, 6 h, 12 h and 24 h after injection. Each group contained three replicates at all time points. The total RNA of each sample was extracted and reverse transcribed into cDNA as described in section "Total RNA extraction and cDNA synthesis". The expression levels of LvCD14L were detected by qPCR as described in the section "Tissue distribution analysis".

DsRNA Synthesis and Knockdown of LvCD14L
The template for LvCD14L dsRNA synthesis was amplified with primers LvCD14L-dsF and LvCD14L-dsR (Table S2). The PCR product was purified with the Gel Extraction Kit (OMEGA, Norcross, GA, USA). DsRNA was synthesized using TranscriptAid T7 High Yield Transcription Kit (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer's protocol. A 289 bp dsRNA of the EGFP gene was synthesized as a negative control. The DNA template for EGFP dsRNA synthesis was amplified from the pEGFPN1 plasmid with primers EGFP-dsF and EGFP-dsR (Table S2). Thirty shrimp were separated into two groups, named the dsLvCD14L group and the dsEGFP group, to detect the optimal silencing dose. The shrimp in each group were divided into three sub-groups and injected with 3, 6 and 12 µg dsRNA per shrimp, respectively. At 48 h after dsRNA injection, the hepatopancreas and hemocytes of five individuals in each sub-group were collected and the total RNA of the samples was extracted. The expression levels of LvCD14L in samples were detected by RT-qPCR as described in the sections "Total RNA extraction and cDNA synthesis" and "Tissue distribution analysis".

Gene Expression and Bacteria Detection in Shrimp after LvCD14L Knockdown and V. parahaemolyticus Infection
In total, 75 shrimp were equally divided into three groups including the dsLvCD14L group, the dsEGFP group and the PBS group. Each shrimp in the dsLvCD14L group and the dsEGFP group was injected with 6 µg corresponding dsRNA, and shrimp in the PBS group were injected with an equal volume of PBS buffer. At 48 h after dsRNA injection, the tissues including the hepatopancreas and hemocytes from 15 individuals were collected as three parallel samples in each group. The interference efficiency of LvCD14L dsRNA was verified as described in Section 4.7. The expression levels of LvDorsal (GenBank No. FJ998202) and LvALF2 (GenBank No. EW713396) in hemocytes were detected by RT-qPCR as described in the sections "Total RNA extraction and cDNA synthesis" and "Tissue distribution analysis". The primers used in this section were listed in Table S2.
The rest of shrimp were injected with 3 × 10 4 cfu V. parahaemolyticus individually. At 24 h after pathogen injection (hpi), the hepatopancreases of each shrimp were collected and three individuals from the same group were mixed as one sample. Three biological replicates were prepared for each treatment. The collected hepatopancreas samples were crushed in sterile PBS buffer, and seeded onto the thiosulfate citrate bile salts sucrose (TCBS) agar culture medium (LuQiao, Beijing, China). After 16 h culture, the number of total viable bacteria and V. parahaemolyticus were counted, and the dominant single colonies were picked and identified by the 16s rDNA sequencing method.

Recombinant Expression and Purification of LvCD14L
The DNA fragment encoding the mature protein of LvCD14L was amplified with a pair of primers LvCD14L-rF/rR (Table S2)

Microbial Cell Wall Polysaccharides Binding Assay
The pathogen-associated molecular pattern (PAMP) binding activity of rLvCD14L was detected by an enzyme-linked immunosorbent assay (ELISA). The microbial cell wall polysaccharides including LPS, PGN and dextran with a concentration of 200 µg/mL were added and coated to wells of flat-bottom microtiter plates. After incubation at 4 • C overnight, the wells were washed three times with PBST (0.05% Tween 20 in PBS). Then, the wells were blocked with 3% BSA at 37 • C for 1 h and washed three times with PBST. The rLvCD14L or rTrx (containing 1 mg/mL BSA) with a concentration gradient including 20 µg/mL, 10 µg/mL, 5 µg/mL, 2.5 µg/mL, 1.25 µg/mL, 0.625 µg/mL and 0.3125 µg/mL was incubated at room temperature for 3 h and washed three times with PBST. PBS buffer containing 1 mg/mL BSA was incubated with PAMPs as a negative control. The wells were incubated with His-Tag (27E8) Mouse mAb (Cell Signaling Technology, Danvers, MA, USA) at 37 • C for 1 h and washed three times with PBST buffer. Then, the wells were incubated with anti-mouse IgG and HRP-linked antibody (Cell Signaling Technology, Danvers, MA, USA) and washed three times with PBST buffer. The chromogenic reaction was generated using the EL-TMB Chromogenic Reagent kit (Sangon Biotech, Shanghai, China), according to the manufacturer's instruction. Absorbance at 450 nm was detected by the precision micro-plate reader (TECAN infinite M200 PRO, Salzburg, Austria). All the experiments were performed in triplicate.

Minimal Inhibitory Concentration (MIC) Assay
Bacteria strains including V. parahaemolyticus, V. harveyi, S. aureus and E. coli were cultured to the logarithmic phase and were counted using a blood cell counting plate under an optical Nikon TS100 microscope (Nikon, Tokyo, Japan). The bacteria were diluted to 1 × 10 4 cfu/mL with PBS buffer. The recombinant protein was diluted to the concentration of 1 mg/mL, 0.500 mg/mL, 0.250 mg/mL, 0.125 mg/mL, 0.0625 mg/mL and 0.0313 mg/mL using PBS buffer. Then, 50 µL of recombinant protein and 50 µL bacteria were added into each well of the 96-well plates and incubated at room temperature for 2 h. After incubation, 100 µL Tryptic Soy Broth (TSB) or Luria-Bertani (LB) medium was added and cultured at the appropriate temperature of each strain overnight. Each treatment was performed in triplicate. After cultivation, the absorbance at 600 and 560 nm for Gram-positive bacteria and Gram-negative bacteria was detected with a precision micro-plate reader (TECAN infinite M200 PRO, Salzburg, Austria), respectively.

Bacterial Agglutination Experiment
The bacterial agglutination experiment was carried out according to previous study with some modifications [47]. V. parahaemolyticus were cultured to the logarithmic phase and harvested by centrifugation at 2000× g for 10 min. The pellets were re-suspended with sterilized PBS at a density of 1 × 10 8 cfu/mL after washing with PBS three times. The bacteria were labeled with 0.1 mg/mL fluorescein isothiocyanate (FITC) and slowly shaken overnight in the dark. Then, the FITC-labeled bacteria were rinsed with PBS and re-suspended in PBS at a density of 1 × 10 7 cfu/mL and mixed with an equal volume of 1 mg/mL rLvCD14L. The PBS and rTrx were used as negative controls. Each treatment was performed in triplicate. After incubation at room temperature for 1 h, the treated cells were observed under an optical Nikon TS100 microscope (Nikon, Tokyo, Japan).

Cellular Localization
The DNA fragments with 15 bp terminal repeats of LvCD14L signal peptide (SP) and hydrophobic tail (HT) were amplified from pM19-T-LvCD14L. The DNA fragment with 15 bp terminal repeats of EGFP was amplified from pEGFP-N1. The purified DNA fragments were constructed into the linearized pEGFP-N1 vector with an In-Fusion HD Cloning Kit (TaKaRa, Shiga, Japan), and the plasmid was transformed into Escherichia coli Trans 5α competent cells (TransGen, Beijing, China) following the manufacturer's protocol. The 293T cells were seeded into the 6-well plate (Corning, NY, USA) and Glass Bottom Cell Culture Dish (NEST, Beijing, China). The plasmids SP + EGFP, SP + EGFP + HT and EGFP-N1 were extracted with an Endo-free Plasmid Mini Kit (OMEGA, Buffalo, NY, USA) and transfected into the 293T cells using Lipofectamine 3000 (Invitrogen, Waltham, MA, USA) according to the manufacturer's instruction. At 16 h after transfection, the cells in the Glass Bottom Cell Culture Dish were washed with PBS buffer three times and stained with Dil solution (Beyotime, Shanghai, China) for 30 min and Hoechst 33,342 solution (Beyotime, Shanghai, China) for 20 min. The cells were washed with PBS buffer three times after each stain step. Then, the cells were observed with an LSM 900 laser scanning confocal microscope (Zeiss, Oberkochen, Germany). At 24 h after transfection, the cells in 6-well plate and culture medium were collected. The cells were lysed at 4 • C for 10 min by cell lysis buffer (Beyotime, Shanghai, China). The proteins in medium were collected with saturated ammonium sulfate and resolute with PBS buffer. The samples were boiled with SDS-loading buffer for 10 min and detected by Western blot. The samples were transferred onto a polyvinylidene fluoride (PVDF) membrane by eBlot™ L1 Fast Wet Transfer System (GenScript, Piscataway, NJ, USA). The membrane was blocked with 5% skim milk which dissolved with TBS tween (TBST) buffer (TBS buffer with 0.1% tween-20) for 2 h. The membrane was incubated with Mouse anti GFP-Tag mAb (ABclonal, Wuhan, China) and GAPDH Mouse mAb (ABclonal, Wuhan, China) at 4 • C overnight and washed three times with TBST buffer. Then the membrane was incubated with anti-mouse IgG, HRP-linked antibody (ABclonal, Wuhan, China). The bands were visualized with BeyoECL Plus Kit (Beyotime, Shanghai, China) according to the manufacturer's protocol. All the experiments were performed in triplicate.

Recombinant Protein Injection and Bacterial Infection
In total, 120 healthy shrimp were equally divided into an rLvCD14L group and an rTrx group. The shrimp in the two groups were injected with 10 µg corresponding recombinant protein and 1 × 10 4 cfu V. parahaemolyticus. At 6 h, 12 h, 24 h and 48 h after injection, the hemocytes from five individuals were collected as one sample and each time point contained three replicates. The total RNA was extracted, and the cDNA was synthesized as described in the section "Total RNA extraction and cDNA synthesis". The expression levels of LvDorsal and LvALF2 in hemocytes were detected by qPCR as described in the section "Tissue distribution analysis".

Co-Immunoprecipitation (Co-IP)
A Co-IP assay was performed to examine the interaction of LvCD14L with TLRs following the method described previously [48]. Briefly, the fragments of LvCD14L maturation protein and the extracellular region of LvToll1 (GenBank No. DQ923424.1) and LvToll2 (GenBank No. JN180637) without signal peptide were amplified from the cDNA of shrimp and constructed into the pDHsp/FLAG-His and pDHsp/V5-His vector using the In-Fusion HD Cloning Kit (TaKaRa, Shiga, Japan), respectively. All primers used in the construction of expression plasmids are listed in Table S2. The plasmids were transfected into the Sf9 cells using Lipofectamine 3000 (Invitrogen, Waltham, MA, USA) following the manufacturer's protocol. After overnight cultivation, the cells were heat-shock treated at 42 • C for 30 min. Then, the cells were cultured at 27 • C overnight after heat-shock treatment. The cells were lysed at 4 • C for 10 min with cell lysis buffer (Beyotime, Shanghai, China). The lysate was centrifuged at 12,000 rpm for 10 min, and the supernatant was collected and incubated with Anti-FLAG M2 Magnetic beads (Sigma, Ronkonkoma, NY, USA) at 4 • C for 2 h. Before incubation, part of the supernatant was preserved as Input. After incubation, the magnetic beads and proteins were collected with a Magnetic Separation Device (Sangon Biotech, Shanghai, China) and then washed with PBS three times. After the last wash, the magnetic beads were re-suspended and boiled with SDS-loading buffer for 10 min. The supernatant was collected as CoIP sample. All samples were detected by Western blot with V5-Tag (D3H8Q) Rabbit mAb and DYKDDDDK Tag (D6W5B) Rabbit mAb (Cell Signaling Technology, Danvers, MA, USA) as described in the section "Cellular Localization". All the experiments were performed in triplicate.