Differential transcriptome analysis of the disease tolerant Madagascar–Malaysia crossbred black tiger shrimp, Penaeus monodon hepatopancreas in response to acute hepatopancreatic necrosis disease (AHPND) infection: inference on immune gene response and interaction

Vibrio parahaemolyticus, AHPND strain (Vp_AHPND) bacteria were isolated from the dissected organs of 10 moribund P. monodon suspected with AHPND outbreak. The AHPND-infected samples were validated through observation of gross clinical signs (empty midgut, empty stomach, atrophied and pale hepatopancreas) and AP3 polymerase chain reaction (PCR) detection method [12]. The digestive systems of dissected AHPND-infected shrimps were placed aseptically into tubes containing 10 ml of tryptic soy broth (TSB+) (Merck, Darmstadt, Germany) supplemented with 2% sodium chloride (NaCl). The tubes were incubated at 28 °C for 18 h at 120 rpm for the enrichment of bacteria. The enriched broth cultures were streaked on thiosulfate citrate bile salt (TCBS) agar to select the green-colony forming V. parahaemolyticus. The isolation of green colonies was conducted on tryptic soy agar (TSA+) (Merck, Darmstadt, Germany) supplemented with 2% NaCl. The enrichment was done through incubation at 28 °C for 18 h at 120 rpm. The bacterial cultures were stored at − 80 °C using cryovial (CRYOBANK™) and revived for downstream applications. The validated positive strains were referred to as Vp_AHPND strains KS17.S5-1, KS17.S5-2, and KS17.S9-2 [13].

One of the previously obtained Vp_AHPND strain KS17.S5-1 [13] was revived and prepared for the experimental challenge. Juvenile commercial pond cultured disease tolerant crossbred shrimps between 13th generation Madagascar P. monodon strain and 5th generation local P. monodon strain with body length from 15 to 20 cm were acclimatized for a week in designed experimental setup. The shrimps in the treatment group were then infected with Vp_AHPND (2 × 10⁶ cfu/ml) through a modified immersion method similar to descriptions by [14]. Sterile TSB+ broth was added instead of Vp_AHPND for the shrimps in the control group. The experimental challenge was conducted with three replicates for both treatment and control groups, and 27 shrimps placed in each tank. The important organs (hepatopancreas, muscle, stomach, gut, and haemolymph) were dissected from one shrimp collected from each tank for every time interval of 0, 3, 6, 12, 24, 36, and 48 h post-AHPND infection and stored at − 80 °C.

Total RNA was extracted from the hepatopancreas samples of infected shrimps at 3 (APm-T3), 6 (APm-T6), and 24 (APm-T24) hours post-AHPND infection together with non-infected control shrimp (APm-CTL) using TransZol Up Plus RNA Kit (Transgen Biotech, Beijing, China). The extracted RNA samples were then treated with DNase and sent for cDNA library preparation and sequencing using BGI-SEQ 500 Sequencer.

The raw sequencing reads were filtered based on several criteria, which include removal of reads with adaptors, removal of reads with more than 5% of unknown bases, and removal of low-quality reads (percentage of bases which quality is lesser than 15 is more than 20% in a read). The filtered reads quality metrics were assessed based on Q30, the rate of bases which quality is greater than 30. The filtered clean reads were stored in FASTQ format. The clean reads were then used for de novo assembly through Trinity software [15] with PCR duplication removed.

For the Unigene expression determination, clean reads were mapped to Unigenes using Bowtie2 software [16] followed by gene expression level determination using expectation maximization (RSEM) software [17], and principle component analysis (PCA) using princomp (function of R software) [18]. The differentially expressed genes (DEGs) were identified based on PossionDis [significant DEGs: fold change ≥ 2.00, false discovery rate (FDR) ≤ 0.001]. The distribution of assembled Unigene expression level and number of DEGs at different post-infection time intervals was compared.

The assembled Unigenes were then functionally annotated through alignment of the Unigenes to nucleotide (NT), non-redundant protein sequence (NR), Gene Ontology (GO), Cluster of Orthologous Groups of proteins (COG), SwissProt, Kyoto Encyclopedia of Genes and Genomes (KEGG), and InterPro functional databases. Basic Local Alignment Search Tool (BLAST) software [19] was used for Unigene alignment with KEGG database; BLAST2GO software [20] was used with NR annotation to obtain GO annotation; InterProScan5 software [21] was used to obtain InterPro annotation. Species distribution of the assembled Unigenes was determined based on NR annotation. For the DEGs, the KEGG analysis was functionally enriched using phyper, a function of R software [18].

From the COG, GO, and KEGG analyses obtained, the DEGs involved in immune-related biological functions and pathways were identified and listed down. An immune-related keyword search was also conducted among the Unigenes with high fragments per kilobase million (FPKM) values and DEGs. The pattern of the immune DEGs’ activation or repression upon AHPND infection at different time points was identified and possible interactions between them were deduced.

The RNA-Seq expression profiles obtained for selected immune-related DEGs (C-type lectin, IMD, ALF, and HMGB1) were validated through the Real Time Reverse Transcription Polymerase Chain Reaction (qRT-PCR) technique with three biological replicates and three technical replicates each across post-AHPND infection time points. The qRT-PCR primers were designed using PrimerQuest Tool software (https://​sg.​idtdna.​com/​PrimerQuest/​Home/​Index). First strand cDNA synthesis was conducted using TransScript^® One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech, Beijing, China). The qRT-PCR experiments were conducted using Agilent Technologies Stratagene Mx3005P instrument and GoTaq^® qPCR Master Mix kit (Promega, Madison, Wisconsin, USA). The qRT-PCR reaction consisted of 10 µl GoTaq^® qPCR 2X Mix, 500 nM forward primer, 500 nM reverse primer, and 2 µl template cDNA. The qRT-PCR cycling program used was 95 °C for 2 mins, 40 cycles at 95 °C for 15 s, and 55 °C for 1 min. Elongation factor 1-alpha (EF1a) gene was selected to be the internal control reference gene. The primer sequences designed were listed in (Additional file 1: Table S4). The Ct values obtained were then analysed using Livak’s 2^ddCt relative quantification method [22]. The results were statistically validated through One-Way Analysis of Variance (One-Way ANOVA) analysis and subsequent post hoc Duncan test using SPSS software Version 22.

The transcriptome sequencing using BGI-SEQ 500 sequencing platform for AHPND-infected (APm-T3, APm-T6, and APm-T24) and non-infected control (APm-CTL) P. monodon hepatopancreas extracted RNA samples was successfully conducted. The RNA Sequencing (RNA-Seq) generated a total of 320.57 Mb raw reads, 312.77 Mb clean reads, and 31.27 Gb clean bases. The quality assessment of the clean reads at the level of Q30 was overall satisfactory and deemed as high quality with quality percentages greater than 90% (Q30) (Additional file 1: Table S1). After transcriptome assembly, a total of 48662 Unigenes was obtained from the data of APm-CTL, APm-T3, APm-T6, and APm-T24 with a total length of 56,085,297 bp, an average length of 1152 bp, an N50 of 2289 bp, and a GC content of 44.10% (Additional file 1: Table S2).

Based on the distribution of Unigenes across FPKM values in Additional file 1: Figure S1, most of the Unigenes were identified to possess fragments per kilobase million (FPKM) values ranging from 0.01 to 100. The transcripts per million (TPM) values were also calculated to be majorly in the range of 0.01 to 200. From the FPKM and TPM distribution across length of Unigenes graph plotted (Additional file 1: Figure S2), the majority of the gene expression was focused in Unigenes with length from 301 to 5000 bp. A brief comparison of the number of differentially expressed genes (DEGs) between the post-AHPND infection time intervals was also conducted. APm-T3 had 7605 upregulated DEGs and 1869 downregulated DEGs; APm-T6 had 2802 upregulated DEGs and 1464 downregulated DEGs; APm-T24 had 7998 upregulated DEGs and 1752 downregulated DEGs. It was clearly observed that APm-T6 had a significantly lower amount of upregulated DEGs when compared to APm-T3 and APm-T24 (Fig. 1).

The Unigenes were functionally annotated using selected seven functional databases and showed results of non-redundant protein sequence (NR) (24,709; 50.78%), nucleotide (NT) (18,622; 38.27%), SwissProt (21,388; 43.95%), Cluster of Orthologous Groups of proteins (COG) (11,636; 23.91%), Kyoto Encyclopedia of Genes and Genomes (KEGG) (20,696; 42.53%), Gene Ontology (GO) (4510; 9.27%), and InterPro (16,438; 33.78%) (Additional file 1: Table S3). The species distribution of the Unigenes using NR annotation results was shown in Additional file 1: Figure S3. The P. monodon Unigenes were successfully matched with Daphnia pulex (8.39%), Tribolium castaneum (4.12%), Pediculus humanus corporis (2.89%), and Branchiostoma floridae (2.39%). This indicated the close phylogenetic relationship of P. monodon to these species. The remaining 82.21% of Unigenes were matched to other species mostly due to limited genetic information in crustaceans.

The 11636 Unigenes assigned to COG annotation were classified into 25 functional protein families (Additional file 1: Figure S4). The Unigenes were mainly involved in “carbohydrate transport and metabolism” (1741 Unigenes), “cell wall/membrane/envelope biogenesis” (1224 Unigenes), and “signal transduction mechanism” (1098 Unigenes).

From the GO analysis obtained (Additional file 1: Figure S5), 4510 Unigenes were assigned to 60 different functions under categories of biological process, cellular component, and molecular function. More specifically, there were 563 Unigenes involved in “response to stimulus”, 370 Unigenes in “signalling”, and 49 Unigenes in “receptor activity”.

Besides that, from the KEGG analysis obtained (Additional file 1: Figure S6), 42 top matched pathways were displayed. Some of the important pathways detected were “signal transduction” (3878 Unigenes), “infectious diseases: bacterial” (2186 Unigenes), and “immune system” (1672 Unigenes).

Furthermore, in the KEGG pathway enrichment analysis based on DEGs (Additional file 1: Figure S7), several notable changes occurred following the different time intervals of post-AHPND infection. These included the detection of “cytosolic sensing pathway”, “sulfur relay system”, and “apoptosis” during APm-T3. At APm-T6, “cell adhesion molecules”, “phagosomes”, “ascorbate and aldarate metabolism”, “Rap 1 signalling pathway”, “adipocytokine signalling pathway”, and “platelet activation” were detected. In addition, APm-T24 was observed to involve “DNA replication and repair” (homologous recombination, non-homologous end joining, nucleotide/base excision repair, mismatch repair), “linoleic acid biosynthesis”, “arachidonic acid metabolism”, “apoptosis”, and “cytosolic DNA sensing pathway”.

Based on the gene functions of the Unigenes identified in the GO analysis conducted (Additional file 1: Figure S5), Unigenes were preliminarily screened out according to their functioning in relation to pathogenic or immune response such as adhesion, signalling, and binding. Similarly, this preliminary screening was conducted on KEGG analysis (Additional file 1: Figure S6) and DEGs-based functional enriched KEGG analysis (Additional file 1: Figure S7). The screened Unigenes were confirmed with comparison to other functional annotations. Finally, the immune-related DEGs that were expressed in all treatments were selected from the preliminarily screened Unigenes and further discussed.

All of the important screened and selected immune-related DEGs were listed down in Table 1 with respective category, gene id, annotated homologous identity, primary annotation, secondary annotation, and log2 fold changes. The 24 selected DEGs were involved in the innate immune response of P. monodon towards AHPND infection. These selected DEGs have important functions in pathogen-induced interactive response (mucin, chitinase, and chitin deacetylase), pattern recognition receptors (PRRs) and damage-associated molecular patterns (DAMPs) (C-type lectin, galectin, and high mobility group box 1 (HMGB1)), immune or signalling pathway activation [signal transducer and activator of transcription (STAT), Toll, immune deficiency (IMD), and tank-binding kinase 1 (TBK1)], antimicrobial peptides (AMPs) [anti-lipopolysaccharide factor (ALF), penaeidin, and stylicin], prophenoloxidase (proPO) system [serine proteinase, serine proteinase inhibitor (SERPIN), and proPO], antioxidation (superoxide dismutase, glutathione peroxidase), phagocytosis (lysozyme, apoptosis stimulating of p53, and caspase), and other immune-related responses [glutathione-dependent prostaglandin d synthase, techylectin, and down syndrome cell adhesion molecule (DSCAM)]. The selected DEGs and their respective log2 fold changes across the three post-AHPND infection time points were shown in Fig. 2. The overall immunological response pattern of P. monodon hepatopancreas during AHPND infection in chronological order was then inferred based on the selected DEGs as shown in Fig. 3.

The differential gene expressions of the selected immune-related DEGs identified in RNA-Seq were successfully validated using the qRT-PCR technique. Four Unigenes (C-type lectin, IMD, ALF, and HMGB1) were selected for the qRT-PCR analysis. The qRT-PCR results demonstrated similar trends to the RNA-Seq data in Fig. 4. The qRT-PCR results of individual Unigenes were shown in Additional file 1: Figures S8–S11. The results were proven to be statistically significant through One-Way Analysis of Variance (One-Way ANOVA) analysis and subsequent post hoc Duncan test (p < 0.05) (Additional file 1: Tables S5–S8). Even though the results from RNA-Seq and qRT-PCR analyses did not match perfectly, most probably because of sequencing bias, the gene expression patterns of immune-related DEGs across post-AHPND infection time points were generally validated by the qRT-PCR analysis.

Based on Additional file 1: Figure S1, the majority of Unigenes detected was identified to possess fragments per kilobase million (FPKM) values ranging from 0.01 to 100 in all four treatments, APm-CTL, APm-T3, APm-T6, and APm-T24. This indicated that the Unigenes commonly displayed a low-level to mid-level gene expression. The Unigenes with FPKM values of > 100 were relatively rare. Whereas in Additional file 1: Figure S2, the FPKM and transcripts per million (TPM) values obtained were shown to be efficient normalization methods for the determination of Unigene expression levels with minimal read length bias.

The number of downregulated differentially expressed genes (DEGs) was similar between APm-T3, APm-T6, and APm-T24. However, it is interesting to observe that the number of upregulated DEGs was significantly lower in APm-T6 as compared to the other two treatments (Fig. 1). This can be due to the gene expression repression, DNA damage or cell damage that resulted from AHPND infection either from the direct influence of Vp_AHPND bacteria or the effect of released bacterial toxins. The invading bacteria are capable of influencing and reprograming host gene expression during host–pathogen interactions to achieve either beneficial or unfavourable effects towards bacterial propagation and persistence [23].

The upregulated mucin, chitinase, and chitin deacetylase gene expressions in P. monodon hepatopancreas in response to AHPND infection (Fig. 2) suggests an interactive relationship with Vp_AHPND involving chitin. The ability of Vibrio species bacteria to interact with chitin molecules to achieve tolerance and adaptation is explained by the previously validated Vibrio cholerae-chitin interaction. Such interaction is important for purposes such as habitat selection and bacterial pathogenicity. The bacteria are capable of utilizing pili (for example, type IV pili) for chitin binding to colonize on the host surface, usually exoskeletons or intestinal cavities [24]. This takes into account the basic composition of chitin molecules, which is β-1,4-linked N-acetylglucosamine (GlcNAc) residues and the chemotaxis properties of Vibrio spp. [24]. GlcNAc-binding protein A (GbpA) is a vital protein released by Vibrio bacteria for binding with chitin subunits, GlcNAc [25]. GbpA’s interaction with chitin and associated components was also investigated in M. rosenbergii previously [26, 27].

The process of chitin binding triggers the activation of bacterial competence state or related pathogenicity through DNA acquisition and transformation [24]. Vibrio-chitin interactions at the cellular level can lead to the subsequent formation of biofilm on the attachment surface, which is a multicellular complex [24, 28]. The mucus layer is significantly affected by the gene expression of mucin, which is an organic constituent of mucus. Interestingly, intestinal mucin functions as a receptor for GbpA and induces its expression. Alternatively, the binding of GbpA to host cells also increases the production of mucus. Thus, there exists a cooperative upregulation mechanism between mucin and GbpA gene expressions [25, 29]. A similar condition is exhibited in Pseudomonas aeruginosa-mucin interactions [30].

There is also a possibility of host gene expression and translation mechanism hijacking by invading pathogens. A good example is shown by [25], by which Type III Secretion Systems (TTSS) was employed by V. parahaemolyticus bacteria for the delivery of effector proteins into host epithelial cells. Such effectors increase the cytotoxicity, enterotoxicity, and intercellular adherence of V. parahaemolyticus bacteria through alteration of host signalling proteins and regulation of cellular behaviour.

Chitinase is proposed and preliminarily validated to be functional in shrimp innate immune defence and phagocytosis, as exemplified by the newly analysed L. vannamei chitinase 5 (LvChi5) [31]. The biological immune defence role played by chitin deacetylase through chitin in shrimp is shown from its gene expression upregulation in a challenge experiment of Exopalaemon carinicauda against V. parahaemolyticus [32]. On the other hand, the expression of chitinase and chitin deacetylase enzymes are also observed in Vibrio spp. for the successful colonization, modification or degradation of chitins on the host epithelial cells [33].

Therefore, it can be postulated that the invading Vp_AHPND bacteria released GbpA protein, chitinase enzyme, and chitin deacetylase enzyme to interact with and colonize the chitin molecules found on surfaces of gut cavities or hepatopancreas in P. monodon shrimps. As a result, the P. monodon hepatopancreas, which is an important digestive gland, secreted chitinase and chitin deacetylase as a measure of early antibacterial response. There was also a cooperative upregulation of Vibrio GbpA protein and P. monodon mucin receptor gene expressions as part of the pathogen-induced interactive response and probable bacterial hijacking mechanism. All these were reflected by the upregulated expressions of mucin, chitinase, and chitin deacetylase in P. monodon hepatopancreas during AHPND infection.

Pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) are important signals for the activation of innate immunity [34, 35]. PAMPs are associated with invading microorganisms and recognized by pattern recognition receptors (PRRs) found on host cells. This leads to the activation of signalling pathways and elevated expression of antimicrobial peptides (AMPs). On the other hand, DAMPs are released during stressed or damaged cell condition regardless of the presence or absence of pathogenic infection. PAMPs and DAMPs are known as “Signal 0 s” as they are always responsible as the initial molecules which bind to receptors triggering cascade reactions. They are mainly involved in host immune defence and apoptosis [34].

According to [34], PAMPs are usually microbial nucleic acids and membrane components. Examples of PAMPs are lipopolysaccharide (LPS), β-1,3-glucan (βG), and peptidoglycan (PG). PAMPs lead to the release of endogenous molecules (EMs) which contain DAMPs and other immune-related molecules [36]. PRRs can be C-type lectin receptors, Toll-like receptors (TLRs), and AIM2-like receptors (ALRs). Examples of DAMPs are high mobility group box 1 (HMGB1), galectin, uric acid, heparin sulphate, heat shock proteins, ATP, and S100 proteins. HMGB1 is one of the highly characterized DAMPs and largely expressed in the nucleus [34‐36]. All these molecules contribute to the activation of shrimp innate immunity during pathogenic infection. Following this would be cell damage, haemocyte degranulation and necrosis, associated elevated level of phenoloxidase (PO) activity, and respiratory burst (RB) [36].

In the present study, during the different time points of post-AHPND infection, several genes functioning as PRRs (C-type lectin, galectin) and DAMPs (galectin, HMGB1) were identified to be upregulated (Fig. 2). PRRs were upregulated mainly during early times of infection whereas DAMPs were upregulated during later times of infection. The upregulation of these genes suggests that either the direct infection of Vp_AHPND bacteria at P. monodon hepatopancreatic cells or indirect pathogenic infection through toxin damage carried the bacterial signals known as PAMPs, which triggered the activation of PRRs and subsequent release of DAMPs.

The currently known shrimp immune signalling pathways vital for disease combating include Janus kinase-Signal transducer and activator of transcription (JAK-STAT) pathway, Immune deficiency (IMD) pathway, TLRs pathway, RNA interference (RNAi) pathway, c-Jun N-terminal kinase (JNK) pathway, and P38 mitogen-activated protein kinase (MAPK) pathway [4]. Among these pathways, TLRs, IMD, and JAK-STAT are the main pathways functioning in shrimp innate immune defence against microbial infection. The components of these pathways or their invertebrate homologues have been previously identified in different shrimp species proving the functionalities of the pathways in shrimps [37]. TLRs pathway mainly involves Toll, Spätzle, Pelle, MyD88, Cactus, and Dorsal. IMD pathway involves IMD, IκB kinase (IKK), and Relish. JAK-STAT pathway involves JAK and STAT [4].

A novel immune signalling pathway, known as cytosolic sensing pathway, which basically involves the detection of microbial cytosolic DNA (CDNs), activation of stimulator of interferon genes (STING) molecules, and subsequent expression of interferons and cytokines by tank-binding kinase 1 (TBK1)/STING complex. The mammalian STING has been proven to participate in host innate immune response [38]. For invertebrates, there was a recent publication by [39], which demonstrated the important function of L. vannamei STING (LvSTING) in shrimp antimicrobial innate immune defence through V. parahaemolyticus bacterial challenge.

The detection of upregulated Toll, IMD, STAT, and TBK1 gene expressions in P. monodon hepatopancreas during AHPND infection in the present study (Fig. 2) indicates the activation of corresponding TLRs, IMD, JAK-STAT, and cytosolic sensing pathways. The activation of these immune-related signalling pathways was validated by the similarly upregulated AMPs, which include anti-lipopolysaccharide factor (ALF), penaeidin, and stylicin (Fig. 2). This is because shrimp immune signalling pathways generally end with the production of AMPs to target, kill, and clean up invading pathogens. Examples of such AMPs are penaeidin (anti-bacterial and anti-fungal), ALF (anti Gram-negative bacteria), stylicin (antimicrobial), and crustin (anti-bacterial or anti-viral) [4].

These upregulated gene expressions that are related to PRRs, DAMPs, immune pathways, and AMPs suggest the occurrence of a continuous cascade of reactions initiated upon contact of bacterial PAMPs with shrimp cells. The successful activation of PRRs led to the activation and upregulation of genes in the order of immune pathway genes, AMP genes, and DAMP genes. Expressed or released DAMPs then functioned similarly to PAMPs to upregulate gene expressions of immune pathway genes and AMP genes until the elimination of invading pathogen or cancellation of cell stress and damage condition. This flow of reactions was described similarly as well in some previously published papers [4, 40, 41].

In crustaceans, the prophenoloxidase (proPO) system is an important mechanism of innate immune defence that involves a cascade of serine proteinases converting inactive proPO to active phenoloxidase (PO), thus causing downstream immune actions, such as toll pathway activation, immune gene synthesis, and melanisation [42]. Serine proteinase cascade activation can be triggered by microbial or fungal cell wall components, including LPS in Gram-negative bacteria, PG in Gram-positive bacteria, and βG in fungi. AMPs are released as a result of proPO system activation for pathogen elimination [42]. Melanisation response or cellular melanotic encapsulation is an effective immune response against invading pathogens, especially parasites. However, strong regulation of proPO system is needed as excessive activation will damage host cells. This is mostly done by regulatory proteins called SERPINs [42, 43]. Interestingly, there is evidence of interaction between proPO system and lysozyme, and possible inhibition of proPO’s conversion to PO by lysozyme through protein interaction [43].

In the present study, the gene expression upregulations of serine proteinase, proPO, and SERPIN were detected during AHPND infection of P. monodon (Fig. 2). The serine proteinase and proPO genes were upregulated initially at APm-T3 which led to the deduction that binding of Vp_AHPND bacterial LPS to serine proteinase cascades caused the conversion of proPO to PO and downstream reactions. The proPO gene expression was upregulated as well to support and sustain the cascade reactions of proPO system. This is supported by the upregulation of SERPIN at later time points of APm-T6 and APm-T24 as SERPIN functions as a negative regulator of the proPO system.

The immune response of P. monodon hepatopancreas that was elicited upon AHPND infection is deduced to be a chronological event by which the immune-related receptors or proteins or genes were activated or upregulated systematically as shown in Figs. 2 and 3. The inference is made mainly based on the gene expression fold change pattern of involved immune-related genes across different post-AHPND infection time points. The concept of the inference is based on the general cellular innate immunity mechanism as described in some previously published papers [4, 40, 41].

During APm-T3, the initial interaction between Vp_AHPND bacteria and P. monodon hepatopancreatic cells resulted in the occurrence of pathogen-induced interactive response involving the upregulation of mucin, chitinase, and chitin deacetylase genes of P. monodon. At the same time, the PAMPs carried by Vp_AHPND bacteria activated the shrimp cell membrane PRRs which subsequently led to the activation of immune pathways. The cascade reactions ended with transcriptional activation or repression of immune-related genes in the cell nucleus. Such transcriptional activities triggered antimicrobial responses to eliminate the invading pathogens. The antimicrobial responses include activation of proPO system, the release of AMPs, and activation of phagocytosis.

At APm-T6, chitin deacetylase, galectin, HMGB1, STAT, Toll, IMD, serine proteinase, lysozyme, and apoptosis stimulating of p53 immune gene expressions were repressed. This is postulated to be the effects of DNA or cell damage inflicted by Vp_AHPND bacteria and released bacterial toxins. The transcriptional or pathway mechanisms were disrupted by the inflicted damage. Another possibility would be the hijacking and alteration of shrimp cell signalling or transcriptional mechanisms by Vp_AHPND bacteria. Despite that, this possibility was only mentioned briefly in past publication [23]. At the current moment, there is still an insufficient amount of studies on the probable hijacking of shrimp cell mechanism by bacteria as more focus is given to viral hijacking. Notably, at this time interval, techylectin was observed to be significantly upregulated. This suggests the occurrence of bacterial agglutination event to limit the pathogenicity and cytotoxicity of Vp_AHPND bacteria.

During APm-T24, due to the previously accumulated stress and cell damage condition, DAMP molecules were secreted to restore or further activate immune pathways and antimicrobial transcriptional activities. The sequential immunological responses described above continued on until the successful removal or elimination of Vp_AHPND bacteria from shrimp hepatopancreatic cells. In addition, the immunological response pattern may also be triggered by the invasion of Vp_AHPND bacteria at P. monodon intestinal cavities due to the important protein and hormone secretory role played by P. monodon hepatopancreas.

Other than that, the upregulation of antioxidant gene expressions at APm-T3 (Fig. 2) suggests the involvement of first line antioxidant defence. Glutathione-dependent prostaglandin d synthase and down syndrome cell adhesion molecule (DSCAM) gene expressions were upregulated indicating the activation of platelet or homologous mechanism and immune memory. Glutathione-dependent prostaglandin d synthase is important to prevent platelet aggregation [44] whereas DSCAM is a hypervariable protein importantly involved in shrimp innate immune memory. DSCAM displays a significantly elevated binding ability to the same pathogen associated with phagocytosis after repeated exposure. DSCAM’s immune priming to viruses has been previously validated [45], however, studies of DSCAM functioning in response to bacterial and fungal infections remain insufficient.