Abstract

   In recent years, the abalone aquaculture industry has been threatened
   by the bacterial pathogens. The immune responses mechanisms underlying
   the phagocytosis of haemocytes remain unclear in Haliotis discus
   hannai. It is necessary to investigate the immune mechanism in response
   to these bacterial pathogens challenges. In this study, the phagocytic
   activities of haemocytes in H. discus hannai were examined by flow
   cytometry combined with electron microscopy and transcriptomic
   analyses. The results of Vibrio parahaemolyticus, Vibrio alginolyticus
   and Staphylococcus aureu challenge using electron microscopy showed a
   process during phagosome formation in haemocytes. The phagocytic rate
   (PP) of S. aureus was higher than the other five foreign particles,
   which was about 63%. The PP of Vibrio harveyi was about 43%, the PP
   peak of V. alginolyticus in haemocyte was 63.7% at 1.5 h. After V.
   parahaemolyticus and V. alginolyticus challenge, acid phosphatase,
   alkaline phosphatase, total superoxide dismutase, lysozyme, total
   antioxidant capacity, catalase, nitric oxide synthase and glutathione
   peroxidase activities in haemocytes were measured at different times,
   differentially expressed genes (DEGs) were identified by quantitative
   transcriptomic analysis. The identified DEGs after V. parahaemolyticus
   challenge included haemagglutinin/amebocyte aggregation factor-like,
   supervillin-like isoform X4, calmodulin-like and kyphoscoliosis
   peptidase-like; the identified DEGs after V. alginolyticus challenge
   included interleukin-6 receptor subunit beta-like, protein turtle
   homolog B-like, rho GTPase-activating protein 6-like isoform X2,
   leukocyte surface antigen CD53-like, calponin-1-like, calmodulin-like,
   troponin C, troponin I-like isoform X4, troponin T-like isoform X18,
   tumor necrosis factor ligand superfamily member 10-like, rho-related
   protein racA-like and haemagglutinin/amebocyte aggregation factor-like.
   Some immune-related KEGG pathways were significantly up-regulated or
   down-regulated after challenge, including thyroid hormone synthesis,
   Th17 cell differentiation signalling pathway, focal adhesion,
   melanogenesis, leukocyte transendothelial migration, inflammatory
   mediator regulation of TRP channels, ras signalling pathway, rap1
   signalling pathway. This study is the first step towards understanding
   the H. discus hannai immune system by adapting several tools to
   gastropods and providing a first detailed morpho-functional study of
   their haemocytes.

   Keywords: Haliotis discus hannai, transcriptomic, haemocyte,
   phagocytosis, electron microscopy

1. Introduction

   An organism’s immune system is essential to its survival ([39]1).
   Innate immunity and acquired immunity are components of immune systems.
   Innate immunity, on the other hand, offers immediate defence against
   infection as opposed to acquired immunity’s durability ([40]2). Since
   they lack traditional acquired immunity, invertebrates are assumed to
   primarily rely on innate immunity to fight against invasive infections
   and adapt to environmental challenges ([41]3, [42]4).

   According to Pham et al. ‘s study ([43]5), phagocytes are essential for
   maintaining immunological homeostasis and killing pathogens. Phagocytes
   participate in phagocytosis, oxidative killing and encapsulation
   processes. Both vertebrate and invertebrate blood cells have an
   essential role in phagocytosis, which is connected to pathogen defence
   and immunological surveillance ([44]6). Phagocytes, with their capacity
   to engulf and eliminate microbial pathogens, are probably going to play
   a major role in immunological defence, particularly in invertebrates
   ([45]4). According to recent investigations, phagocytes in mollusks
   have been seen to capture and engulf invasive bacteria ([46]7). For
   instance, the lysosomal proteolytic pathway was used by Euprymna
   scolopes’ phagocytes to break down germs that have been ingested
   ([47]8). These findings emphasise the regulation of phagocytic activity
   in innate immunity ([48]9).

   Mollusc haemocytes are known to perform many functions, including
   phagocytosis, killing invasive microbes, identifying foreign particles
   and oxidatively eliminating them ([49]10). Haemocytes have been
   demonstrated in earlier research to be crucial for mollusk
   immunological responses in both humoral and cellular immunity.
   Molluscan haemocytes play a crucial role in phagocytosis. In the blue
   mussel Mytilus edulis, for example, phagocytosis has been shown to
   significantly decrease after cadmium exposure, and several
   immune-related genes linked to phagocytosis in M. edulis were
   significantly up-regulated after Vibrio splendidus challenge ([50]11).

   Mammals’ phagocytic killing mechanisms have been extensively
   investigated ([51]12), but these mechanisms in gastropods are still not
   well understood. Phagocytosis in gastropods has been documented in the
   caenogastropod Viviparus, the pulmonate gastropods Lymnaea stagnailis,
   Biomphalaria glabrata, Cerithidea californica, the marine gastropods H.
   discus hannai, Haliotis asinine, Haliotis tuberculata, Littorina
   littorea, and ivory shell Babylonia areolata ([52]13, [53]14).
   Nevertheless, the majority of research has focused on the morphological
   traits and immune capabilities of haemocytes; hence, functional
   assessments and characterisation of phagocytosis in gastropod
   haemocytes are still missing.

   The coastal regions of Liaoning, Jiangsu, Shandong and northern Japan
   are the primary distribution areas for the Pacific abalone H. discus
   hannai. China’s mariculture of H. discus hannai has grown significantly
   in the last several years, and has become one of the important
   mariculture shellfish ([54]15). In addition, it is referred to as
   marine soft gold and is more valuable commercially and culturally than
   H. tuberculata and Haliotis diversicolor. Despite the fact that abalone
   aquaculture is a massive industry in China, there are still issues. One
   of these is the decrease in abalone germplasm caused by repeated
   inbreeding. In recent years, there have been many disease outbreaks in
   China’s abalone industry, which have caused significant financial
   losses ([55]16). Little is known about the innate defence of H. discus
   hannai haemocytes. It is critical to understand H. discus hannai’s
   immune system in order to reduce disease-associated mortality, which
   makes it significant for H. discus hannai farming.

   In marine environments, bacterial pathogens are ubiquitous and cause
   continuously threaten to the health of the majority of aquaculture
   species. Particularly, the vibriosis disease is a significant challenge
   to the culture of mollusks, as many Vibrio species are thought to be
   the cause of widespread outbreaks of abalone mortality. One of
   principal bacterial pathogen is V. parahaemolyticus in H. discus
   hannai, it is known to cause “withering syndrome” and “outer velum
   breaking disease” in the abalone aquaculture industry, which can result
   in significant losses ([56]17, [57]18). For instance, significant stock
   losses of Haliotis diversicolor supertexta were attributed to V.
   parahaemolyticus in China ([58]19). One kind of conditional pathogen
   found in oceans is V. alginolyticus ([59]20). When the environment
   deteriorates or the immune function decreases, the disease is more
   likely to spread ([60]21). While Vibrio splendidus was isolated from
   moribund Haliotis rubra and Haliotis laevigata during disease outbreaks
   in Australia ([61]22), infections of Vibrio carchariae and V. harveyi
   have been reported in H. tuberculata along the French coast ([62]23)
   and in H. discus hannai in Japan ([63]24), respectively.

   The occurrence of several illnesses has hampered the cultivation of H.
   discus hannai due to a lack of studies on innate defence of H. discus
   hannai haemocytes and disease control. “Omics” research has been more
   productive and economical in the last several years. For the purpose of
   identifying genes linked to immunity and studying the phagocytic
   killing mechanisms underlying mollusk immunity, transcriptomics can be
   a very useful tool. Even so, the abalone genome’s initial draft has
   already been made public. Currently, inadequate sequence splicing
   quality and a dearth of pertinent gene resource data are impeding
   research ([64]25). This work used transcriptomics, flow cytometry and
   histological characteristics to shed light on the immune responses by
   haemocytes in H. discus hannai. Quantitative transcriptome studies of
   the haemocytes were carried out following the V. parahaemolyticus and
   V. alginolyticus challenge in order to determine the immunological
   pathways linked to the phagocytosis by the haemocytes. Investigating
   the immunological response of the H. discus hannai haemocytes to
   foreign pathogens and fluorescent polystyrene beads, as well as
   examining the fates of the bacteria following ingestion, were the
   objectives of this work. Numerous substantially differentially
   expressed genes (DEGs) were found to be engaged in signalling networks
   relevant to the immune system.

2. Materials and methods

2.1. Animal rearing and manipulation

   Adult Pacific abalones (body length 5.80 ± 0.60 cm, body weight 17.65 ±
   3.50 g; n = 500) were collected from Jinjiang Fuda Abalone Fisheries
   Co. Ltd (Jinjiang, Fujian, China) and acclimated for 14 days before the
   experiment under the same cultivation conditions (24 ± 1°C, 33‰
   salinity, and 7.8 pH). Then, the abalones were kept in a recirculating
   system with a sand- filter, where they were fed red alga Gracilaria
   lemaneiformis or sea tangle (Laminaria japonica) once a day, and
   saltwater was changed daily. The system was kept at consistent
   temperatures (23°C) and dissolved oxygen levels (6.2 mg/L). Every day,
   seawater was added to the aquariums, and all animal studies were
   carried out in accordance with the ‘Guidelines for Experimental
   Animals’ of the Ministry of Science and Technology (Beijing, China; No.
   [2006] 398, 30 September 2006). All of the experiments involving
   animals reported in this study were approved by the Institutional
   Animal Care and Use Committee (IACUC) of Fujian Agriculture and
   Forestry University.

2.2. Haemocyte preparation

   The abalones were cleaned using gauze and absorbent cotton before being
   sliced lengthwise in the center of the foot muscle using a sterilised
   scalpel. Every individual was collected about 2.0–3.0 mL of hemolymph,
   which was then divided into individual 15 mL centrifuge tubes.
   Hemolymph samples immediately mixed with same volume modified Alsever’s
   solution MAS (the MAS anticoagulant contained 20.8 g L^−1 of glucose;
   3.36 g L^−1 of EDTA; 8 g L^−1 of sodium citrate and 22.5 g L−1 of
   NaCl.) to prevent aggregation, as recommended. After centrifuging the
   hemolymph for five minutes at 700 g to extract circulating haemocytes
   and supernatant for subsequent experiment.

2.3. Morphologic observation on haemocytes

   A total of three cleaned microscope slides were filled with 50 μL of
   hemolymph solution per sample. One was utilised for direct light
   microscope observation, and the other was employed to prepare smears.
   Haematoxylin-eosin (HE) staining and Fast Wright-Giemsa staining were
   applied to two smear preparations per sample. All the samples were
   observed using a Leica DM 4000 B microscope and Leica DFC 425 C image
   sensor. Each type of haemocyte was counted, 100 cells per abalone were
   measured, there were 6 abalones for cell counts and size measurement.

2.4. Bacterial strains and preparation of bacterial suspensions

   Experiments were carried out using the Gram-positive bacterium S.
   aureus and the Gram-negative bacterium Escherichia coli, which were
   obtained from Hope Bio-Technology Co., Ltd., Qingdao, Shandong
   province, China. Bacterium were activated according to the supplier’s
   guidelines. Take a small amount of E. coli, S. aureus separately and
   inoculate them on Luria-Bertani (LB) solid culture medium with
   inoculation rings. After culturing at 37°C for 16–18 hours, select two
   monoclonal strains of bacteria and inoculate them onto LB liquid
   culture medium. Shake and incubate at 37°C for 16–18 hours. Bacteria
   were collected by centrifugation at 3 000 g for 10 min at 4°C and
   resuspension in 0.85% sterile saline solution rinsed twice. The
   bacterial concentration in this investigation was around 10^9 colony
   forming unit (CFU)/ml. Adjust the bacterial suspension to the required
   bacterial concentration for the experiment using phosphate buffer.

   V. harveyi, V. alginolyticus and V. parahaemolyticus (isolated from
   diseased Pacific abalones and kept in our laboratory) were employed in
   the bacterial challenge experiment where the frozen culture was revived
   from -70°C storage. V. harveyi, V. alginolyticus and V.
   parahaemolyticus were cultivated overnight at 37°C in LB medium
   (containing 10 g/L peptone, 5 g/L yeast extract, and 10 g/L NaCl) and
   extracted via centrifugation (4 000 g, 10 min, 4°C). The pellet was
   then washed three times in sterile phosphate-buffered saline (PBS) with
   a pH of 7.4 before being resuspended in 0.9% normal saline (NS). The
   concentration of V. parahaemolyticus (1.0 × 10^8 CFU/mL) was determined
   by analysing the findings of the pre-experimental stage. Haemocytes
   were treated with Vibrio for 1.5 hours at 28°C at a haemocyte/bacteria
   ratio of 1:10.

2.5. Preparation of haemocytes challenged by V. alginolyticus and V.
parahaemolyticus

   The bacterial challenge employed two species of Vibrios, V.
   alginolyticus and V. parahaemolyticus, which were previously isolated
   from moribund abalone and shown to be pathogenic to abalone ([65]26,
   [66]27). A total volume of 50 μL V. alginolyticus (or V.
   parahaemolyticus) was injected into the abalone foot of 180 H. discus
   hannai specimens. 180 abalones which did not receive the bacterial
   challenge received an injection of sterile premixed phosphate buffered
   solution (PBS) and were used as the C group. Another 180 abalones were
   used as the “P” group and were given an injection of V.
   parahaemolyticus suspension at a concentration of 1.0 ×10^7 CFU/mL (50
   μL) (which had been demonstrated to be the sub-lethal concentration in
   our previous unpublished experiment) in a PBS solution into the pedal
   sinus using a 100 μL syringe. As the “A” group, another 180 abalones
   were given an injection of V. alginolyticus slurry at a concentration
   of 1.0 ×10^7 CFU/mL (50 μL). After that, the treated animals were
   raised separately from the 180 control animals. Abalones from the C, P,
   and A groups had their hemolymph collected at 0, 6, 12, 24, 48, and 72
   hours after the injection (n= 6).

   The hemolymph samples were transferred to 1.5 mL Eppendorf tubes. All
   hemolymph solutions containing anticoagulant, the hemolymph was
   centrifuged at 700 g to separate the haemocytes and supernatant, the
   haemocytes were preserved at −80°C for subsequent transcriptomic
   analysis. The supernatant was preserved at −80°C for subsequent measure
   the average protein concentration, activation of acid phosphatase
   (ACP), alkaline protease (AKP), total antioxidant capacity (T-AOC),
   total superoxide dismutase (T-SOD), lysozyme (LZM), nitric oxide
   synthase (NOS), glutathione peroxidase (GSH-Px) and catalase (CAT).

2.6. Preparation of electron microscopy samples

   For transmission electron microscopy (TEM), haemocytes were incubated
   for 1.5 h at 28°C with V. parahaemolyticus or V. alginolyticus or S.
   aureus at a bacteria/haemocyte ratio of 10:1. Then, haemocytes were
   fixed in PBS with 2.5% glutaraldehyde for 3 hours at 4°C (pH 7.4).
   Following fixation, the haemocytes were washed three times in 0.1 mol/L
   PBS (pH 7.4) before being transferred to 1% osmium tetraoxide at 4°C
   for two hours. Haemocytes post bacterial challenge were subsequently
   washed in PBS. Haemocyte pellets were generated in 2% agarose when the
   remaining glutaraldehyde and osmium tetraoxide were removed. Following
   immersion in propylene oxide, a dehydration procedure (50%, 70%, 80%,
   90%, 95%, and 100% ethanol) was carried out. Finally, the fixed samples
   were embedded in Epon 812 resin (TAAB, UK) and cut into 60-nm thick
   ultrathin slices with an ultramicrotome (Leica CM1520). TEM was used to
   investigate ultrathin slices (60 nm) stained with 2% uranyl acetate and
   1% lead citrate. A Hitachi TEM equipment (Model H-7650) was used to
   view and photograph the sections.

2.7. Determination of phagocytosis by haemocytes using flow cytometry

   Flow cytometry was carried out using a FACSCalibur (Becton-Dickinson
   Immunocytometry Systems, San Jose, CA) coupled with a single argon ion
   laser with 488 nm filtered emission. Fluorescence photomultiplier
   bandpass filters were 530 nm (green fluorescence, FL1) and 585 nm
   (yellow/orange fluorescence, FL2). Side scatter (SSC) and fluorescence
   data were recorded on a log scale, whereas forward scatter (FSC) data
   was obtained on a linear scale. Each haemocyte sample had 10,000 events
   recorded. To establish logical zones and colour gating analyses of
   fluorescence data, the Cell Quest^® software (Becton-Dickinson, San
   Jose, CA, USA) was employed.

   Total circulating haemocytes count (THC): SYBR Green I (Sigma-Aldrich,
   USA) was used for THC detection. A fluorescent dye that binds to
   double-stranded DNA. In brief, 200 μl of fresh hemolymph was fixed in
   an equivalent volume of 3% formalin solution. Fixed circulating
   haemocytes were treated for 120 minutes in the dark at room temperature
   with 1,000 x SYBR green I before being run in a flow cytometer. The
   number of undamaged cells was counted in 1 minute using a flow
   cytometer. The quantity of cells ml^−1 hemolymph is used to report THC.
   The cytometer’s flow rate was determined using the methods previously
   mentioned ([67]28). The quantity of cells ml^−1 hemolymph is used to
   report THC.

   Ex vivo phagocytosis was quantified by measuring the ability of
   haemocytes to ingest fluorescent beads. Samples containing 500 μL of
   anticoagulant MAS were mixed with 500 μL of haemocytes (about 2.16 ×
   10^6 cells/mL). Next, 500 μL of haemocytes was mixed with 10 μL of a
   1/10 dilution of fluorosphere microspheres with a diameter of 1 μm
   (FluoSpheresTM carboxylate, 1 μm, red, Life Technologies Corporation)
   and allowed to incubate at room temperature in the dark. The final
   ratio of beads to haemocytes was 100/1, while the final concentration
   of fluorescent beads was about 10^8 mL^-1. It is simple to discriminate
   between the endogenous baseline fluorescence of non-phagocytic
   haemocytes -containing beads and that of haemocytes-containing one,
   two, three, or more beads. Also, it is simple to determine how many
   beads each cell phagocytizes. Phagocytosis capacity was defined by
   Hégaret et al. ([68]29) as the proportion of haemocytes-engulfed beads
   above three. In our investigation, we used this methodology. A
   fluorescence plot was used to calculate the proportion of cells that
   contained fluorescent beads. Haemocytes were conducted 0.5 h, 1.0 h,
   1.5 h, 2.0 h, 2.5 h post-challenge by fluorescent beads on three
   randomly selected abalone from each of the tanks.

   Fluorescein Isothiocyanate (FITC) -labelling of microorganisms was
   performed follows: E. coli and S. aureus were inoculated on LB at 37°C
   for 16–18 hours. Centrifugation was used to collected V. harveyi, V.
   alginolyticus and V. parahaemolyticus, which was cultured in LB medium
   at 28°C for the whole night. The optical densitometric approach was
   used to quantify the concentration of bacteria, and quantitative plate
   counting was used to confirm the results. The bacteria were mixed with
   fluorescein isothiocyanate (FITC, 1 mg/mL, Sigma-Aldrich, USA) after
   bacteria were washed three times with PBS (pH 7.4) at room temperature.
   Following that, three PBS washes (pH 7.4) were performed on the
   FITC-labelled microorganisms. Haemocytes were resuspended in MAS. A
   blood counting chamber and fluorescence photometer were used to count
   the number of haemocytes-positive cells and FITC-labelled bacteria in
   order to guarantee that the bacteria included in the study ultimately
   had the same FITC concentration. There were more than 95% FITC-positive
   bacteria present. By using quantitative plate counting, the percentage
   of bacterial viability in an aliquot of each population was found to be
   96%–98%. At a ratio of 10 bacteria per haemocyte, FITC-labelled
   bacteria (about 2 × 10^7 CFUs) were introduced to haemocyte suspensions
   (roughly 1.8 × 10^6 cells), and the mixture was incubated for 0.5 h,
   1.0 h, 1.5 h, 2.0 h, 2.5 h at 28°C. A negative control was an
   unintubated tube with only haemocytes. After centrifuging the
   suspensions to extract the haemocytes, extra microorganisms were
   eliminated by repeating three PBS washes. After being fixed for 25
   minutes in 4% paraformaldehyde, haemocytes were examined using flow
   cytometry. The proportion of FITC-positive haemocytes was used to
   calculate the phagocytosis rate. The boundaries of positive and
   negative fluorescence were confirmed by processing FITC-stained and
   unstained bacteria phagocytosis mixtures. The FITC-positive and
   -negative cells were then gated and sorted according to the intensity
   of their FITC fluorescence. The dot plots were used to identify the
   gating zone, which removed dead cells. A total of 20,000 haemocytes
   units were processed for every sample. Green fluorescence (FL-1
   detector) was used to measure the phagocytosis of the haemocytes linked
   with fluorescent beads. It is possible to precisely assess phagocytic
   activity analytically to calculate the mean fluorescence of cells,
   indicated in this instance by the proportion of cells that fluoresce
   more than three beads. The fluorescence signal indicated the
   phagocytosis area. The percentage of haemocytes that consumed three or
   more fluorescent beads was used to calculate the phagocytic activity of
   haemocytes ([69]30). Three runs of this experiment were conducted.

2.8. Assessment of enzymatic activities

   Enzymatic activity assays were carried out according to the
   manufacturer’s instructions using a standard assay kit from the Nanjing
   Jian cheng Bioengineering Institute (Nanjing, China). Using
   spectrophotometry (BIO-RAD 680), the enzymatic activities were
   analysed. Using bovine serum albumin as a reference, Bradford’s
   Coomassie Brilliant Blue (G-250) was used to calculate the protein
   content. Units of activity per milligram of protein (U/mg protein) were
   used to express enzyme activities. The second portion was centrifuged
   at 700 g to separate the plasma and haemocytes, which were then stored
   at -80°C to be measured later for average protein content, AKP
   activation, ACP, T-AOC, T-SOD, LZM, NOS, GSH-Px and CAT activation.

2.9. Transcriptome sequencing and analysis

   The total RNA was extracted using the Trizol reagent in accordance with
   the usual protocol. Before producing the libraries for Illumina
   sequencing, the RNA concentration and purity were measured using a
   NanoDrop ND1000 spectrophotometer (NanoDrop Technologies Inc., DE,
   USA), the RNA quality was tested using 1% agarose gels, and the RNA
   integrity was tested on an Agilent 2100 Bioanalyzer (Agilent
   Technologies, CA, United States). Poly (A) + mRNA selection was
   performed on one RNA pool using oligo (dT) magnetic beads (Invitrogen,
   United States). The RNA sequencing libraries were created using an
   Illumina Tru-Seq™ RNA Sample Preparation Kit (Illumina, San Diego, CA,
   USA) according to the manufacturer’s instructions. The cDNA libraries
   obtained were directly used for cluster generation and sequencing on an
   Illumina HiSeqTM2500 Analyzer at Biomarker Technologies Co., Ltd
   (Peking, China). NGSQC Toolkits were utilised to filter reads with
   primer or joint sequence contamination or low quality for quality
   control of original sequencing reads. The raw reads were first filtered
   to remove adaptors, ambiguous reads (those with ≥10% ambiguous
   nucleotides), and low-quality reads (those with ≥20% bases with quality
   scores of ≤10) before being assembled using Trinity, with min_kmer_cov
   set to 2 and other parameters set to defaults. Using KAAS, the
   resulting contigs and unigenes were aligned to the non-redundant (NR)
   protein database, Gene Ontology (GO), and Kyoto Encyclopedia of Genes
   and Genomes (KEGG), with an E-value cut-off of 10^−5.

2.9.1. Library preparation and RNA sequencing

   In each example, three biological replicates of transcriptome
   sequencing were obtained, and all raw data was deposited in the NCBI
   sequence. Using oligo (dT) magnetic beads, mRNA was isolated from total
   RNA and broken into small pieces using fragmentation buffer. After cDNA
   purification with the QIAquick PCR Purification Kit (Qiagen, Hilden,
   Germany), a cDNA library compatible with Illumina NGS technology was
   created from the fragmented mRNA by reverse transcription,
   second-strand synthesis, and ligation of appropriate adapters
   (paired-ends). The amount of cDNA in each library was determined using
   the Qubit device and spectrofluorometric analysis. Illumina HiSeq™ 4000
   technology was used to obtain the paired-end 100 bp (PE100) data.

2.9.2. Differentially expression analysis

   To compare the expression levels in each sample and identify the
   differentially expressed genes (DEGs), a differential expression
   analysis test was utilised. To quantify expression abundance, read
   counts were normalised as FPKM (Fragments Per Kilobase of exon model
   per Million mapped reads) values for calculating and comparing gene
   expression variations between individual samples. The FPKM of genes in
   haemocytes was then statistically analysed using DEseq 1. To determine
   DEGs in haemocytes in each group, absolute values of log 2 (fold
   change, FC) > 1 and FDR (false discovery rate) < 0.05 were set as
   threshold criteria.

   To collect annotation information on DEGs, the Basic Local Alignment
   Search Tool (BLAST) software was used to sequence the DEGs using NR,
   Swiss-Prot, GO, Clusters of Orthologous Groups of Genes (COG), and
   KEGG. DEGs were identified by comparing the data to the H. discus
   hannai transcriptome as well as the GO and KEGG and COG databases using
   BLAST-based searches (E < 10^-5). The techniques for GO and KEGG
   pathway annotation were used according to Shen et al. ([70]31). The
   following processes were used for GO annotation: target genes were
   extracted, prepared by sequence alignment, assembled, and submitted to
   a BLAST of GO terms (mapping). Only the GO terms with the pre-set
   scores were assigned to the target protein sequences. As a consequence,
   the results of the GO annotation were acquired. The KEGG pathway
   annotation procedure was as follows: the protein sequence was matched
   using the KEGG database to produce homologous KEGG genes, and the genes
   were screened using a bi-directional hit rate to obtain orthologous
   candidate genes. Based on the probability and heuristic scoring, the
   data were analysed by the KEGG Orthology (KO) method in order to
   generate the KO ranking chart. Fisher’s exact test was used to
   determine the significance degree of protein enrichment of GO terms and
   each pathway in this study.

2.10. Statistical analysis

   For paired sample comparisons, a two-tailed Student’s t-test was used
   in this investigation. The SPSS 15.0 program was used for all
   statistical analyses, and quantitative data were given as means ± SEM.
   The significance threshold was chosen at p ≤ 0.05.

3. Results

3.1. Haemocyte morphology

   In fixed preparations two kinds of cells were clearly separated by
   their size and nucleus/cytoplasm (N/C) ratio. The granulocytes had a
   low N/C ratio. The lowly abundance of small haemocytes with a high N/C
   ratio are most likely hyalinocytes, as shown in other mollusks ([71]
   Figure 1 ). Staining of spontaneous adherent or cyto-centrifuged cells
   with Wright-Giemsa and haematoxylin-eosin revealed two haemocyte types:
   granulocytes (g) and hyalinocytes (h), depending on the presence or
   absence of cytoplasmic granules ([72] Figures 1B–F ). Granulocyte
   endoplasm had cytoplasmic granules, whereas hyalinocytes had few or
   none. Granulocytes ranged in size from 8 to 13 μm and had a low N/C
   ratio. These cells were round or ovoid in shape, with nuclei that were
   centrally located, round or oval in shape ([73] Figure 1 ).

Figure 1.

   [74]Figure 1
   [75]Open in a new tab

   Light micrographs of haemocytes of Pacific abalone (Haliotis discus
   hannai) prepared by spontaneous adhesion using filtered sterilised
   seawater. (A). The micrographs of unstained haemocytes under the light
   microscope (200×). (B, C). The micrographs of haemocytes with
   Wright-Giemsa stain under the light microscope (200×). (D). The
   micrographs of haemocytes with haematoxylin-eosin staining under the
   light microscope (200×). (E, F). The micrographs of unstained
   haemocytes under the light microscope (1000×). h- hyalinocytes,
   g-granulocytes.

3.2. Non-adherent haemocyte transmission electron microscopy

   Comparison of hyalinocytes and granulocytes, granulocytes contained an
   abundance of electron-dense cytoplasmic particles surrounded by
   membranes, known as cytoplasmic granules, with sizes ranging from 0.2
   to 1.0 μm. The cytoplasm included a variable number of mitochondria,
   the Golgi complex, endoplasmic reticulum, and small electron-lucid
   vesicles of various sizes, some of which were most likely derived from
   the Golgi complex or the smooth endoplasmic reticulum. In this study,
   granulocytes were classified into two categories based on the number of
   granules and the morphology of granules: type I granulocytes ([76]
   Figure 2A ) and type II granulocytes ([77] Figures 2B–D ). In the
   cytoplasm of type I granulocytes, there were many granules, each around
   0.5 μm in diameter and oval. Type II granulocytes contained a few
   granules, of various shapes. Granulocytes however composed the minority
   (less than 30% on cytocentrifuged cells were observed).

Figure 2.

   [78]Figure 2
   [79]Open in a new tab

   Electronic microscopy of haemocytes in H. discus hannai. (A) Electron
   microscopy of type I granulocytes in H. discus hannai, spherical or
   oval cells containing many large oval granules, 0.3–0.6 μm in diameter,
   granulocytes with asymmetrical shape; (B, C) electronic microscopy of
   type II granulocytes in H. discus hannai; (D) type II granulocytes,
   oval and small nucleus. Rotund or oval granule (gv); nucleus (n);
   mitochondria (m); pseudopodia (p); bacilliform granule (gb); the
   letters represent the same meaning in following figure. (E-H) Electron
   microscopy of hyalinocytes in H. discus hannai. Electron transmission
   microscopy of hyalinocytes in H. discus hannai. Hyalinocytes with
   asymmetrical shape, pseudopodia can be observed in some hyalinocytes,
   they have one or several nucleus, and a cytoplasm containing few or no
   granules, the nucleus was either in a central or an eccentric position;
   (E-G) large hyalinocytes, haemocyte with large nucleus, a small amount
   of cytoplasm, a small number of and mitochondria in the cytoplasm,
   hyalinocytes showing pseudopodia; (H) cells with a large nucleus,
   containing a great number of mitochondria, are small hyalinocytes.
   Nucleus (n); mitochondria (m); pseudopodia (p); rough endoplasmic
   reticulum (rer).

   In this study, the hyalinocytes lacked or had few cytoplasmic granules,
   and the nucleus was either central or eccentric ([80] Figures 2E–H ).
   The cytoplasm was filled with a variety of mitochondria and tiny
   electron-lucid vesicles of varying sizes ([81] Figure 2G ). The
   hyalinocytes exhibited well-developed pseudopodia ([82] Figures 2F, G
   ); the shape of the hyalinocytes revealed a spreading pseudopodia
   morphology. There are two types of hyalinocytes based on their size:
   large (7–10 μm) and small (4–6 μm) hyalinocytes. The large cells had
   big cytoplasmic vacuoles that did not stain, as seen in live adherent
   haemocytes ([83] Figure 2G ). Small hyalinocytes were haemocytes with a
   big nucleus and a small amount of cytoplasm containing a large number
   of mitochondria ([84] Figure 2H ), this resulted in a relatively high
   N/C ratio. They relate to what are known as blast-like cells ([85]
   Figure 2H ). Hyalinocytes composed major part of the haemocyte
   population (more than 60% on cytocentrifuged cells).

3.3. Assessment of haemocytes phagocytosis using electron microscopy

   The phagocytosis of haemocytes was studied at a ratio of 1:10 of
   haemocytes to V. parahaemolyticus or V. alginolyticus or S. aureus. The
   results of haemocytes that engulfed and internalised V.
   parahaemolyticus or S. aureus or V. alginolyticus are shown in
   [86]Figure 3 , respectively. In vitro investigation of phagocytosis of
   bacteria by H. discus hannai haemocytes revealed that haemocytes could
   phagocytise V. parahaemolyticus or V. alginolyticus or S. aureus under
   transmission electron microscopy ([87] Figure 3 ). The morphological
   characteristics of H. discus hannai haemocytes differed significantly.
   The majority of the bacteria were attached to the surface of
   haemocytes. Phagocytising to bacteria was also detected in haemocytes
   ([88] Figure 3 ). Granulocytes were seen adhering to the microorganisms
   V. parahaemolyticus ([89] Figure 3C ). The presence of multiple long
   and thin cytoplasmic protrusions or filopodia ([90] Figures 3A, C )
   distinguished these haemocytes. Bacterial trapping by haemocyte
   pseudopods was detected ([91] Figures 3C, D ). The early phagosome area
   was filled with a microgranular substance with a low electron density
   ([92] Figure 3C ). Individual phagosomes were used to internalise
   bacteria ([93] Figure 3C ).

Figure 3.

   [94]Figure 3
   [95]Open in a new tab

   In vitro phagocytosis by H. discus hannai haemocytes under transmission
   electron microscopy. (A-D).In vitro phagocytosis of Vibrio
   parahaemolyticus cells by H. discus hannai haemocytes. (A).
   Hyalinocytes with pseudopodia, hyalinocytes adhering to bacteria. (B).
   Hyalinocytes with phagocytising the bacterium. (C). Hyalinocytes with
   early phagosome, early phagosome containing morphologically intact
   bacteria. (D). adhering to bacteria and phagocytising the bacterium. n,
   nucleus; VP, V. parahaemolyticus; p, pseudopodia; eph, early phagosome.
   (E, F). In vitro phagocytosis of S. aureus cells by H. discus hannai
   haemocytes under transmission electron microscopy. (E). Hyalinocytes
   with early phagosome, early phagosome containing morphologically intact
   bacteria. (F). Hyalinocytes with pseudopodia and with the stage between
   early phagosome and late phagosome (mph). n, nucleus; Sa, S. aureus;
   (G-J). In vitro phagocytosis of Vibrio alginolyticus cells by H. discus
   hannai haemocytes under transmission electron microscopy. (G).
   Hyalinocytes with phagocytising the bacterium. (H, I). Hyalinocytes
   adhering to bacteria and phagocytising. (J). Hyalinocytes with late
   phagosome. n, nucleus; va, Vibrio alginolyticus; lph, late phagosome.

   When haemocytes phagocytised S. aureus, the nucleus had an active
   character, and the heterochromatin was mostly organised in
   electron-dense clumps and near the nucleus’s periphery ([96]
   Figures 3E, F ). During phagocytosis, bacteria stimulated hyalinocytes
   with phagosomes, including early phagosomes containing morphologically
   intact bacteria ([97] Figure 3E ), hyalinocytes with pseudopodia and
   with the stage between early phagosome and late phagosome ([98]
   Figure 3F ).

   When haemocytes phagocytised V. alginolyticus, the haemocytes changed
   shape and the appearance of pseudopods was seen, as well as endocytosis
   of the bacterium V. alginolyticus ([99] Figures 3G–J ). Only a small
   percentage of granulocytes were capable of phagocytising V.
   alginolyticus. One or more whole bacteria might be phagocytised by the
   haemocytes. Bacteria became attached to the surface of the haemocytes,
   resulting in an attachment phenomenon ([100] Figure 3H ), late
   phagosomes with only debris, possibly bacterial fragments ([101]
   Figure 3J ). The findings revealed that the phagocytosis capacities of
   H. discus hannai hyalinocytes and granulocytes to V. alginolyticus
   could potentially be different. This suggested that phagocytic
   haemocytes are most likely hyalinocytes, but granulocytes can also be
   phagocytic, however electron microscopy may not detect them.

3.4. Phagocytosis and THC by flow cytometry in haemocytes infected by
bacteria

   In vitro phagocytosis experiment, flow cytometry was utilised to detect
   haemocyte phagocytosis of fluorescent beads. The fluorescent beads’
   phagocytic rates in haemocytes were measured ([102] Figures 4A, B ).
   The phagocytic rate (sometimes known as the phagocytic percentage, PP)
   was higher at 0.5 h, which was 14.35%. The greatest increase in PP of
   E. coli occurred at 1 h which was 22.27%, and the PP of E. coli in
   haemocyte increased no further and started to decrease ([103] Figure 4A
   ). The PP of S. aureus was higher than the other five foreign
   particles, which was about 63%. The PP of V. harveyi was about 43%, it
   was no significant difference at different time points. The greatest
   increase in PP of V. parahaemolyticus occurred at 1.5 h which was
   40.8%, and decreased at 2.5 h. The PP peak of V. alginolyticus in
   haemocyte was 63.7% at 1.5 h.

Figure 4.

   [104]Figure 4
   [105]Open in a new tab

   Phagocytic capability of haemocyte detected by flow cytometry in H.
   discus hannai. (A) The phagocytosis of haemocytes was detected by flow
   cytometry (bacteria were labelled with FITC). (B) The fluorescence
   microscope images of phagocytosis. (C) Total circulating haemocytes
   count of haemocytes to Vibrio alginolyticus or Vibrio parahaemolyticus
   by foot injection.

3.4.1. THC

   No abalone was dead during the sub-lethal bacterial V. parahaemolyticus
   or V. alginolyticus stress. The THC variation patterns were shown in
   [106]Figure 4C . THC of V. parahaemolyticus exposure increased steadily
   and peaked was 7.23 ×10^6 cells/mL at 12 h, then decreased
   significantly at 24 h, subsequently increased steadily at 48 h and
   72 h. THC of V. alginolyticus exposure increased and peaked was 8.87
   ×10^6 cells/mL at 24 h, then decreased at 48 h and 72 h. THC of control
   group was relatively stable during the whole period.

3.5. Unigene basic annotation using transcriptome analysis after V.
parahaemolyticus or V. alginolyticus stress

3.5.1. Summary of annotation results

   Total 86235 unigenes were annotated by Nr, KOG, KEGG and Swiss-Prot
   database. Of them, 32094,12453, 31218, 16105 unigenes were annotated by
   Nr, KOG, KEGG and Swiss-Prot database, respectively. Annotation genes
   were 32938 ([107] Figure 5A ).

Figure 5.

   [108]Figure 5
   [109]Open in a new tab

   Unigenes annotated. (A). Venn diagram of all unigenes annotated against
   Nr, SwissProt, KOG and KEGG databases in the Pacific abalone (Haliotis
   discus hannai) in response to V. parahaemolyticus and V. alginolyticus
   stress. The number in each colour block indicated the number of
   unigenes that was annotated by single or multiple databases. (B)
   Species distribution of sequences matched to the Nr database.

3.5.2. Functional annotation of non-redundant unigenes

   In this study, the majority annotated unigenes matched Haliotis
   rufescens ([110] Figure 5B ). In the species distribution of sequences
   matched to the Nr database, the unigenes revealed similarities to
   sequences of H. rufescens (26592 unigenes), Gigantopelta aegis (408),
   Vitrella brassicaformis CCMP3155 (277), Mytilus edulis (141), Trichuris
   trichiura (129), Drosophila immigrans (121), Haliotis discus discus
   (108), Mercenaria mercenaria (107), H. discus hannai (105).

3.5.3. KEGG annotation

   What we were interested in is the immune-related pathways, as shown in
   [111]Figure 6 , the majority of immune-related genes were classified
   into pathways for cell growth and death (929 genes) [including
   apoptosis (340 genes), necroptosis (237)], cellular community -
   eukaryotes (848) [including focal adhesion (393)], immune system (1163)
   [including NOD-like receptor signalling pathway (219), C-type lectin
   receptor signalling pathway (167), Toll and Imd signalling pathway
   (145), chemokine signalling pathway (141), Toll-like receptor
   signalling pathway (139)], signal transduction (2363) [including
   PI3K-Akt signalling pathway (452), MAPK signalling pathway (317), Ras
   signalling pathway (303), Rap1 signalling pathway (301), mTOR
   signalling pathway (230), HIF-1 signalling pathway (185), Wnt
   signalling pathway (181), TNF signalling pathway (178)], signalling
   molecules and interaction (841)[including cell adhesion molecules
   (195)], transport and catabolism (1270) [including lysosome (318),
   endocytosis (308), autophagy - animal (252), phagosome (244)].

Figure 6.

   [112]Figure 6
   [113]Open in a new tab

   The Kyoto Encyclopedia of Genes and Genomes (KEGG) classification of
   all assembled unigenes in the Pacific abalone (H. discus hannai).

3.5.4. COG/KOG

   KOG annotation identified 16316 putative genes across 25 categories
   ([114] Figure 7 ). Among these categories, the cluster for “signal
   transduction mechanisms (2376 genes)” represented the second largest
   group, followed by posttranslational modification, protein turnover,
   chaperones (1519), in addition, the number of proteins involved in
   defence mechanisms was 165 genes.

Figure 7.

   [115]Figure 7
   [116]Open in a new tab

   The cluster of orthologous groups (COG) classification. A total of
   16316 unigenes were annotated into 25 categories.

3.5.5. GO annotation

   GO classifies genes and gene products into three categories: biological
   processes (BPs), cellular components (CCs) and molecular functions
   (MFs). In three samples, the discovered proteins were examined using GO
   annotations; the GO functional categorisation is presented in
   [117]Figure 8 . Using the biological process, the genes participating
   in cellular process (18519 unigenes), metabolic process (14742
   unigenes), biological regulation (10164), regulation of biological
   process (9540), response to stimulus (8329), immune system process
   (2060), viral process (823). Using the molecular function term, the
   genes involved in binding (16098), catalytic activity (11266),
   transporter activity (2152), molecular function regulator (1652),
   ATP-dependent activity (1453), antioxidant activity (154). Using the
   cellular component term, the number of proteins involved in cellular
   anatomical entity (16305), the first largest, followed by those related
   to protein-containing complex (6692), virion component ((203),
   respectively.

Figure 8.

   [118]Figure 8
   [119]Open in a new tab

   GO function classification of the Pacific abalone (H. discus hannai) in
   response to V. parahaemolyticus and V. alginolyticus stress.

3.6. Transcriptomic differences in haemocytes challenged by V.
parahaemolyticus and V. alginolyticus

   Differential expression of unigenes found in five groups was assessed
   by computing FPKM values obtained by matching the high quality reads to
   a reference transcriptome constructed by clustering three of the
   samples unigenes. Transcripts were classed as up or down regulated
   based on their log fold change (FC) value, which was determined using
   the FC= Log2 (Treated/Control) algorithm. FC values larger than 1 were
   regarded to be up-regulated, whereas values less than 1 were thought to
   be down-regulated. We utilised p-value 0.05 and |log2 (fold change)|> 1
   to look for up or down regulated genes.

   [120]Figure 9 showed gene information discovered using transcriptomic
   approaches. Using an organism database, bacterial genes were removed
   from the analysis. After a 6-hour and 48-hour V. parahaemolyticus
   stress, 29 and 12 genes were significantly up-regulated in P6 and P48,
   respectively; 25 and 23 genes were significantly down-regulated in P6
   and P48, respectively. After V. alginolyticus challenge for 6 h and
   48 h, 18 and 32 genes were significantly up-regulated in A6 and A48,
   respectively, 4 and 121 genes were significantly down-regulated in A6
   and A48, respectively.

Figure 9.

   [121]Figure 9
   [122]Open in a new tab

   Venn diagram of all genes and statistics of different expression genes.
   (A). Venn diagram of all genes. (B) Statistics of different expression
   genes. Genes in control group (C), haemocytes challenged by V.
   parahaemolyticus for 6 h (P6), haemocytes challenged by V.
   parahaemolyticus for 48 h (P48), haemocytes challenged by V.
   alginolyticus for 6 h (A6), and haemocytes challenged by V.
   alginolyticus for 48 h (A48), respectively.

   In comparing differentially expressed genes among groups, we focused on
   immune-related genes. After V. alginolyticus challenge 6 h, in
   comparisons of C and A6, it was worth noting that deleted in
   interleukin-6 receptor subunit beta-like and protein turtle homolog
   B-like were present in significantly higher levels in A6 group ([123]
   Supplementary Table 1 ). After V. alginolyticus challenge 48h, in
   comparisons of C and A48, most of the differentially expressed genes
   were significantly down-regulated in A48, especially rho
   GTPase-activating protein 6-like isoform X2, leukocyte surface antigen
   CD53-like, calponin-1-like, calmodulin-like, troponin C, troponin
   I-like isoform X4, troponin T-like isoform X18, tumor necrosis factor
   ligand superfamily member 10-like ([124] Supplementary Table 1 ).
   However, rho-related protein racA-like and haemagglutinin/amebocyte
   aggregation factor-like were significantly up-regulated in A48 ([125]
   Supplementary Table 1 ).

   After V. parahaemolyticus challenge 6h, compared the C and P6 groups,
   haemagglutinin/amebocyte aggregation factor-like was significantly
   up-regulated in P6 ([126] Supplementary Table 1 ). However,
   supervillin-like isoform X4 was expressed at significantly lower levels
   in P6 ([127] Supplementary Table1 ). However, after V. parahaemolyticus
   challenge 48 h, calmodulin-like and kyphoscoliosis peptidase-like were
   considerably lower in the P48 group (p< 0.05) ([128] Supplementary
   Table 1 ). The observed differentially expressed genes in haemocytes
   may have a tight relationship with their function features.

3.7. Gene Ontology enrichment analysis of differentially expressed genes

   GO enrichment analysis for the DEGs was performed to characterise the
   expression changes in the five samples with the whole transcriptome
   dataset as the background to gain insight into the functional
   categories that were altered between healthy and V.
   parahaemolyticus-infected haemocytes and between healthy and V.
   alginolyticus-infected haemocytes. GO annotations were used to classify
   the discovered genes in C and A6, C and A48, C and P6, C and P48.
   [129]Figure 10 showed the GO functional categorisation analysis.

Figure 10.

   [130]Figure 10
   [131]Open in a new tab

   Significantly enriched GO taxonomy of differentially expressed genes
   (DEGs) in the comparisons of five groups (p value<0.05).

   Based on biological processes, the DEGs were assigned to GO categories.
   In the comparison of the C and A6 groups, most of the up-regulated
   genes were involved in binding (including 7 genes), cellular anatomical
   entity (7 genes), cellular process (6 genes), catalytic activity (5
   genes), metabolic process (5 genes), localisation (5 genes), response
   to stimulus [3 genes, including Unigene0003336 (prtgb:interleukin-6
   receptor subunit beta-like)], antioxidant activity (2 genes),
   detoxification(2 genes), immune system process (1 genes, which was
   prtgb) in the A6 group. The numbers of largely down-regulated genes
   participating in these biological process categories were 2, 1, 1, 0,
   1,1,1,0 and 1, respectively in the A6 group ([132] Figure 10 ;
   [133]Supplementary Table 2 ).

   For the C vs A48 groups, most up-regulated genes were involved in
   binding (including 10 genes), catalytic activity (7 genes), cellular
   anatomical entity (6 genes), cellular process (6 genes), response to
   stimulus [4 genes, including Unigene0002677 (racA: rho-related protein
   racA-like) Unigene0071836 (dvr1: protein DVR-1-like) and
   Unigene0100902(birc7-a: baculoviral IAP repeat-containing protein
   3-like), immune system process (1 genes, which was birc7-a) in the A48
   group. The numbers of down-regulated proteins involved in binding (65
   genes), cellular anatomical entity (58 genes), cellular process (53
   genes), response to stimulus [28 genes, including Unigene0001551 (tkfc:
   triokinase/FMN cyclase-like isoform X1), Unigene0029628 (fhod3: FH1/FH2
   domain-containing protein 3-like), Unigene0030552 (para: myosin heavy
   chain, striated muscle-like isoform X10), Unigene0031634 (ryr:
   ryanodine receptor 2-like isoform X17), Unigene0035336 (magi1:
   membrane-associated guanylate kinase), Unigene0037904 (tni-4: troponin
   I-like isoform X4), Unigene0042150 (arhgap6: rho GTPase-activating
   protein 6-like isoform X2), Unigene0060888 (sqh: myosin regulatory
   light chain 12A-like),Unigene0061241(tspan4: leukocyte surface antigen
   CD53-like), Unigene0069048 (tbx1-b: T-box transcription factor
   TBX1-like), Unigene0069977 (nox5: NADPH oxidase 5-like isoform X2),
   Unigene0077316 (cd40lg: tumor necrosis factor ligand superfamily member
   10-like), Unigene0081343 (aoc1: putative amine oxidase), Unigene0081566
   (mid1: E3 ubiquitin-protein ligase Midline-1-like isoform X3),
   Unigene0096861(sfrp1: secreted frizzled-related protein 1-like isoform
   X1), immune system process (12 genes), biological adhesion (5 genes) in
   the A48 group ([134] Figure 10 ; [135]Supplementary Table 2 ).

   For the C vs P6 groups, the up-regulated genes were involved in binding
   (5 genes), cellular anatomical entity (5 genes), metabolic process (4
   genes), cellular process (4 genes), response to stimulus (3 genes,
   including dvr1), antioxidant activity (2 genes), detoxification (2
   genes) in the P6 group. The numbers of down-regulated genes involved in
   these biological process categories were 8, 8, 5, 7, 4, 0 and 1,
   respectively, in the P6 group ([136] Figure 10 ; [137]Supplementary
   Table 2 ).

   For the C vs P48 groups, up-regulated genes were involved in metabolic
   process (4 genes), catalytic activity (3 genes), binding (3 genes),
   response to stimulus (3 genes), antioxidant activity (2 genes) in the
   P48 group. The numbers of down-regulated genes involved in the binding
   (13 genes), cellular anatomical entity (13 genes), cellular process (11
   genes), response to stimulus [7 genes, including Unigene0044790
   (steap4, metalloreductase STEAP4-like), tbx1-b], immune system process
   (3 genes, including tbx1-b) in the P48 group ([138] Figure 10 ;
   [139]Supplementary Table 2 ).

3.8. KEGG pathway enrichment with differentially expressed genes among
samples

   For the five samples, significant difference KEGG pathways are shown in
   [140]Figure 11 . After V. alginolyticus challenge, in comparing C and
   A6 group, the ko04918 thyroid hormone synthesis, ko04640 haematopoietic
   cell lineage and ko04659 Th17 cell differentiation signalling pathway
   had significant difference in A6 (p < 0.05). In comparing C and A48
   group, ko04810 regulation of actin cytoskeleton, ko04115 p53 signalling
   pathway, ko04510 focal adhesion, ko04913 ovarian steroidogenesis,
   ko04916 melanogenesis, ko04921 oxytocin signalling pathway, ko04919
   thyroid hormone signalling pathway, ko04713 circadian entrainment,
   ko04750 inflammatory mediator regulation of TRP channels, ko04670
   leukocyte transendothelial migration, ko04015 rap1 signalling pathway,
   ko04014 ras signalling pathway, ko04020 calcium signalling pathway,
   ko04024 cAMP signalling pathway, ko04022 cGMP-PKG signalling pathway,
   ko04080 neuroactive ligand-receptor interaction had significant
   difference in A48 (p < 0.05). The results showed that after 48 hours of
   stress with V. alginolyticus, there were significant changes in the
   signalling pathway compared to the control group.

Figure 11.

   [141]Figure 11
   [142]Open in a new tab

   Significantly enriched KEGG pathways of differentially expressed genes
   (DEGs) in the comparisons of five groups (p value<0.05).

   After V. parahaemolyticus challenge, in comparing C and P6, ko04918
   thyroid hormone synthesis signalling pathway had significant difference
   in P6. In comparisons of C and P48, ko04918 thyroid hormone synthesis
   signalling pathway had significant difference in P6 ([143] Figure 11 ).

3.9. Series protein expression data analysis by the short time-series
expression miner (STEM)+

   STEM is a program that analyzes short time-series gene expression data.
   Data from abalone samples (C, A6, A48, P6 and P48) that showed a
   1.5-fold increase or decrease in at least one sample were filtered. The
   findings of FPKM-means clustering were used to partition the DEG
   expression profiles into eight categories ([144] Figure 12 ). The
   number in the upper left-hand corner of a profile box reflects the
   profile ID number, and the enrichment p value appears in the bottom
   left-hand corner of each small square. A statistically significant
   number of assigned genes is shown in coloured profiles. Non-white
   profiles of the same colour are clustered together in a single cluster.

Figure 12.

   [145]Figure 12
   [146]Open in a new tab

   Cluster analysis of differentially expressed genes after Vibrio
   parahaemolyticus and Vibrio alginolyticus exposure based on FPKM-means
   method. (A, B) Differentially expressed genes after Vibrio
   parahaemolyticus exposure. (C, D) Differentially expressed genes after
   Vibrio alginolyticus exposure. Clusters ordered based on the number of
   genes. The number in the top left-hand corner refers to profile ID
   number, and enrichment p value appears in the lower-left of each small
   square. Coloured profiles show statistically significant numbers.

   [147]Figure 12A showed the number of genes in profile after V.
   parahaemolyticus challenge. [148]Figure 12B showed a reordered model
   profile interface. The p value was used to arrange the profiles.
   Profile 1 includes genes that were negatively regulated throughout the
   V. parahaemolyticus challenge phase. Profile 3 contained genes
   negatively modulated in P48. Profile 5 contained genes were
   up-regulated in P6. Genes in profile 7 were up-regulated during the
   whole process ([149] Figures 12A, B ). Genes in profile 3 and profile 5
   were up-regulated after V. alginolyticus challenge in A6 group ([150]
   Figures 12C, D ). Profile 0 contained genes negatively modulated during
   the whole process of V. alginolyticus challenge. Group 7 contained
   genes positively modulated.

   In C, P6 and P48 group, cathepsin O-like, rab3 GTPase-activating
   protein, Rab-35-like, SCO-spondin-like, V-type proton ATPase subunit e
   2-like, alkaline phosphatase-like were up-regulated after V.
   parahaemolyticus challenge in P6 group; 10 kDa heat shock protein was
   down-regulated in P48; cathepsin L and cathepsin Z-like were
   up-regulated in P48; cdc42 homolog was up-regulated during the whole
   process; cathepsin L-like isoform X1,CCR4-NOT transcription complex
   subunit 11-like isoform, cold shock domain-containing protein,
   hemocyanin type 1, mucin-16-like isoform X2, RAB11-binding protein
   RELCH homolog isoform X2, RAB6-interacting golgin-like, rab9 effector
   protein with kelch motifs-like, ral GTPase-activating protein subunit
   beta-like isoform X3, Rab-20-like isoform X1, Rab-3-like, Rab-43-like,
   Rab-8A-like, Rap1-like, rho GTPase-activating protein 32-like isoform
   X4 were down-regulated during the whole process ([151] Supplementary
   Table 3 ).

   In C, A6 and A48 group, 1-phosphatidylinositol 3-phosphate 5, activated
   CDC42 kinase 1, activator of 90 kDa heat shock protein ATPase homolog
   1-like, adhesion G protein-coupled receptor, C1q domain containing
   protein 2, C1q-related factor-like isoform X2, calmodulin-like,
   cathepsin L1-like, cathepsin B-like, cathepsin O-like, C-type
   lectin-like isoform X1, Hsp20, heat shock factor protein 1, heat shock
   70 kDa protein, heat shock protein 60A-like, histone, interferon,
   mucin-22-like, mucin-2-like isoform, mucin-3A-like, mucin-4-like
   isoform X2, mucin-5AC-like, rab9 effector protein, Rab-14, Rab-1A-like,
   Rab-20-like isoform X1, Rab-3 isoform X1, Rab-32-like isoform X1,
   Rab-37-like, Rab-39B-like, Rab-40B-like, Rab-43-like, SCO-spondin-like
   isoform, serine/threonine-protein kinase mTOR-like isoform X2,
   toll-like receptor were up-regulated after V. alginolyticus challenge
   in A6 group; cathepsin Z-like, cdc42 homolog, heat shock protein HSP
   90-beta-like, MAP kinase-interacting serine/threonine-protein kinase
   1-like isoform X2, rab5 GDP/GTP exchange factor-like Rab-6B-like,
   Rab-9A-like were up-regulated in A48;10 kDa heat shock protein, cold
   shock domain protein, extracellular superoxide dismutase [Cu-Zn]-like
   isoform X3, ferritin, hemocyanin type 1, NF-kappa-B-activating
   protein-like, catalase were down-regulated during the whole process
   ([152] Supplementary Table 4 ).

3.10. Predicted interactions of identified DEGs from STEM

   [153]Figure 13 was obtained from the [154]https://cn.string-db.org/ and
   shows the predicted interactions of the identified DEGs. In C, A6 and
   A48 groups, major clusters were associated with microsomal glutathione
   S-transferase, thioredoxin-like protein, peroxiredoxin-5, catalase,
   manganese-superoxide dismutase [MnSOD], histone related protein, 10 kDa
   heat shock protein, heat shock 70 kDa protein, heat shock protein HSP
   90, toll-like receptor 13, epidermal growth factor receptor, rab3
   GTPase-activating protein, ras-related protein Rab-8A, Rab-14, Rab-37,
   Rab-43, Rab-9A, RABF1-like, rab5 GDP/GTP exchange factor, ral
   GTPase-activating protein, cdc42, actin-related protein,
   serine/threonine-protein kinase mTOR, phosphatidylinositol
   4,5-bisphosphate 3-kinase, 1-phosphatidylinositol 4,5-bisphosphate
   phosphodiesterase, phosphoinositide 3-kinase, protein Wnt-1, protein
   Wnt-5a, protein Wnt-10a ([155] Figure 13A ; [156]Supplementary Table 5
   ).

Figure 13.

   [157]Figure 13
   [158]Open in a new tab

   Predicted interactions of identified DEGs. Different line colours
   represent types of evidence for association. Genes without interactions
   have been removed from the graph. Genes abbreviations and corresponding
   full name are shown in [159]Supplementary Tables 5 , [160]6 . (A)
   Predicted interactions of identified DEGs in C, A6 and A48 groups; (B)
   Predicted interactions of identified DEGs in C, P6 and P48 groups.

   In C, P6 and P48 groups, major clusters were associated with exosome
   complex component, m7GpppN-mRNA hydrolase, CCR4,
   glutathione-S-transferase, peroxisomal membrane protein, catalase,
   hypoxia up-regulated protein, stress-70 protein, manganese-superoxide
   dismutase, Heat shock cognate 70 kDa protein, caspase-3-like,
   mitogen-activated protein kinase kinase6, cdc42, phosphatidylinositol
   4-phosphate 5-kinase, 1-phosphatidylinositol 4,5-bisphosphate
   phosphodiesterase, phosphatidylinositol polyphosphate 5-phosphatase,
   rab proteins geranylgeranyl transferase component, ras-related protein
   RABF1-like, Rab-35-like, Rab-8A-like, Rab-43-like, Rab-30-like, rab3
   GTPase-activating protein, histone related protein ([161] Figure 13B ;
   [162]Supplementary Table 6 ).

3.11. Enzyme activities analysis

   The activities of antioxidant enzymes and immune related enzymes in
   hemolymph of H. discus hannai after V. parahaemolyticus and V.
   alginolyticus infection were detected. The experimental results were
   analysed by double sample t-test. The experiment was divided into 0 h,
   6 h, 12 h, 48 h and 72 h, including the following enzymes: superoxide
   dismutase (T-SOD), lysozyme (LZM), nitric oxide synthase (NOS), total
   antioxidant capacity (T-AOC), glutathione peroxidase (GSH-Px), catalase
   (CAT), acid phosphatase (ACP) and alkaline phosphatase (AKP). After V.
   alginolyticus infection, comparing the experimental group with the
   control group, the 72h experimental group of T-AOC showed significant
   changes; The 12h, 24h, 48h and 72h experimental groups of CAT had
   significant differences; There were significant differences in T-SOD at
   24h, 48h and 72h; The 48h experimental group of LZM showed significant
   changes. After V. parahaemolyticus infection, AKP activity and LZM
   content increased first and then decreased, T-SOD activity increased
   first and then decreased, and T-AOC increased first and then decreased.
   The activity of NOS and GSH-Px first decreased and then increased, then
   decreased and then increased. CAT activity first decreased, then
   increased, then decreased. There was no significant change in ACP. It
   is speculated that T-AOC, T-SOD, LZM, and CAT can be used as the better
   disease indicator enzymes for abalone infection with Vibrio ([163]
   Figure 14 ).

Figure 14.

   [164]Figure 14
   [165]Open in a new tab

   Activities of antioxidant enzymes and immune-related enzymes in
   hemolymph of Haliotis discus hannai under Vibrio parahaemolyticus and
   Vibrio alginolyticus stress.

4. Discussion

   Disease resistance is one of the most essential characteristics of
   aquaculture species since it is closely related to final yield
   ([166]32). Haemocytes have critical roles in both invertebrate and
   vertebrate innate immune defence responses ([167]33, [168]34).
   Phagocytosis is an innate immunological response, and the molecular
   mechanisms of phagocytosis in mammals have been extensively
   investigated. However, research in this field in invertebrates has been
   hampered by variable haemocyte subpopulations, a lack of adequate
   phagocytic lines, and a lack of particular cell markers ([169]35). Flow
   cytometry, electron microscopy and transcriptome analysis were used in
   this work to investigate the immune responses of haemocyte in the
   marine mollusk Pacific abalone H. discus hannai. The strategy is a
   critical step in the investigation of haemocytes in invertebrates.

4.1. Transmission electron microscopy to study the phagocytic process

   In this study, transmission electron microscopy was used. Phagocytosis
   is an actin-dependent endocytic process that involves the
   receptor-mediated engulfment of big particles (diameter ≥ 0.5 µm) into
   the plasma membrane or plasma membrane-derived vacuoles called
   phagosomes ([170]36). Phagocytosis is a kind of cellular immunity
   evoked by foreign particles or pathogen infection, and invertebrate
   phagocytic haemocytes are thought to be the equivalent of vertebrate
   macrophages ([171]37). The phagocytic process is thought to consist of
   two steps, the first of which involves the attachment of foreign
   particles to the cell surface and the second of which involves
   internalisation into the cytoplasm, resulting in the engulfment of
   foreign particles into the cell and the subsequent formation of
   phagosomes to stimulate microbial digestion ([172]38). Tuan et al.
   ([173]39, [174]40) have extensively researched this process. Davids and
   Yoshino ([175]41) defined haemocyte adhesion as the initial stage in
   the phagocytosis of foreign particles and migration toward an inflamed
   location. Images of internalised and engulfed V. parahaemolyticus, S.
   aureus, and V. alginolyticus were produced in this investigation, as
   were pseudopodia-adherent bacteria and the subsequent formation of
   phagosomes. The findings of a bacterium challenge utilising electron
   microscopy on haemocytes revealed a process during phagosome
   development that included probing, cup formation, pseudopod expansion,
   and phagosome formation. Cup-like membrane extensions that are employed
   to internalise big particles regulated phagocytosis ([176]42).
   Granulocytes have a larger phagocytic capability than hyalinocytes in
   general ([177]43). However, in this study, the presence of pseudopodia
   and changing morphology in hyalinocytes after V. alginolyticus stress
   adds to the evidence that hyalinocytes are the primary haemocytes
   involved in phagocytising foreign particles ([178]44). The findings
   might imply that hyalinocytes primarily phagocytose via pseudopodia
   after V. alginolyticus stress.

4.2. Identification of differentially expressed genes involved in
phagocytosis

   Rab proteins are crucial molecules for phagosome activity, and multiple
   Rab proteins have been identified from the early stage of phagosome
   formation to the regulation of phagosome maturation ([179]45). In this
   study, the changes of rho-related protein racA-like and rho
   GTPase-activating protein 6-like isoform X2 in transcriptome analysis
   were observed ([180] Supplementary Table 1 ). Furthermore, the actin
   cytoskeleton’s reorganisation causes pseudopods to form, encircling the
   foreign particles and sealing the phagocytic cup. This process causes
   the foreign particles to be internalised into an early phagosome
   ([181]46). Actin, clone 403-like was discovered in this investigation
   ([182] Supplementary Table 1 ). The findings demonstrated a degree of
   consistency between the phagocytic process as shown by electron
   microscopy and the genes discovered by transcriptome analysis,
   indicating that the formation and maturation processes of H. discus
   hannai phagosomes were comparable to those observed in vertebrates.
   This research revealed that phagocytosis was a highly ancient and
   converted immune system strategy.

4.3. Phagocytic activity of haemocytes using flow cytometry

   One strong and well-established method for analysing the phagocytic
   capacity of haemocytes is flow cytometry. It has been applied
   extensively. Since latex beads are inert polystyrene particles, using
   them prevents the needless activation of mollusk haemocytes during
   phagocytosis. Fluorescent microspheres are durable and recognizable
   ([183]47, [184]48). The phagocytic rate (the phagocytic percentage, PP)
   of fluorescent microspheres was higher at 0.5 h, which was 14.35%. The
   greatest increase in PP of E. coli occurred at 1 h ([185] Figure 4A ).
   The PP of S. aureus was higher than the other five foreign particles,
   which was about 63%. The PP of V. harveyi was about 43%. The greatest
   increase in PP of V. parahaemolyticus occurred at 1.5 h, and decreased
   at 2.5 h. The PP peak of V. alginolyticus in haemocyte was 63.7% at
   1.5 h.

   The quantitative results of phagocytosis efficacy against six distinct
   phagocytic mediators were considerably diverse. Many factors, including
   relative haemocyte size, the properties of foreign particles, the
   diameter of foreign particles, the pathogenicity of foreign particles,
   the environment and stress, are likely to contribute to the difference
   ([186]49). Scrobicularia plana haemocytes, for example, could
   phagocytose three distinct kinds of heat-killed bacteria but not
   zymosan ([187]50). Furthermore, various haemocyte subpopulations appear
   to have different preferences in the phagocytosis of different bacteria