Xinya Guo, Ce Ji, Xuan Du, Jianfeng Ren, Yao Zu, Weiming Li,Qinghua Zhang,＊

a Key Laboratory of Exploration and Utilization of Aquatic Genetic Resources, Ministry of Education, Shanghai Ocean University, Shanghai, 201306, China

b International Research Center for Marine Biosciences, Ministry of Science and Technology, Shanghai Ocean University, Shanghai, 201306, China

c National Pathogen Collection Center for Aquatic Animals, Ministry of Agriculture, Shanghai Ocean University, Shanghai, 201306, China

d Department of Fisheries and Wildlife, Michigan State University, East Lansing, MI, 48824, USA

Keywords:Immersion Microinjection Transcriptome Vibrio parahaemolyticus Zebrafish

ABSTRACT

1. Introduction

The zebrafish (Danio rerio) has recently emerged as a valuable model for studying infectious agents and host-pathogen interactions (Am,Appelmelk, Vandenbroucke-Grauls, & Bitter, 2004; Lieschke, Currie,2007; Prajsnar, Mcvicker, Williams, Renshaw, & Foster, 2018; Sullivan,Kim, 2008). The innate immune system in zebrafish can be detected and activated as early as 1 day post-fertilization (dpf), and only innate immunity is present before 3 weeks post fertilization (Lam, Chua, Gong,Lam, & Sin, 2004). Novoa and Figueras have reviewed the characteristics of zebrafish as a model for innate immunity and inflammation studies (Novoa & Figueras, 2012), and zebrafish larvae have been widely used to study the innate immune system following infection with bacteria (Pressley, Witten, Mellon, Kim, & Immunology, 2004; Ronan,Jonas, Roland, Per-Erik, & Hans, 2004; Soest et al., 2011), fungi (CC, CF,IH et al., 2010; Liew, Moya, Wierzbicki, et al., 2017), and viruses(Guerra-Varela, Baz-Martínez, Silva-´Alvarez et al., 2018; L´opezmu˜noz,Roca, Sepulcre, Meseguer, & Mulero, 2010).

However, different routes of infection may activate different response pathways, resulting in complex innate immune responses in zebrafish larvae. Immersion and microinjection have both been used extensively to infect zebrafish larvae and adults (Milligan-Myhre et al.,2011). Immersion is an easier method of infection for mutant or drug screening, and also activated inflammatory marker genes in individual embryos (Benard et al., 2015; Soest et al., 2011). Although various anatomical structures have been used as target sites for microinjection in zebrafish embryos and larvae, including the caudal vein, duct of cuvier,hindbrain ventricle, tail muscle, otic vesicle, notochord, and yolk, the main sites are the hindbrain ventricle and caudal vein (Takaki, Davis,Winglee, & Ramakrishnan, 2013). Hindbrain ventricle infection has been used to study the recruitment of macrophages and stimulate chemotaxis due to the absence of macrophages in this tissue (Yang et al.,2012), while caudal vein microinjection causes systemic infection and has been used to examine aspects such as bacteria-host cell interactions,bacterial virulence (Tobin, Jr, Ray, et al., 2010), and drug treatment effects (Takaki, Cosma, Troll, & Ramakrishnan, 2012).

Vibrio parahaemolyticus is an important pathogen that has recently been shown to be responsible for serious illnesses, hospitalizations, and even deaths in China (Chen et al., 2017). Disease due to V. parahaemolyticus was first reported in Japan in the 1950s (Fujino et al., 1953), with most cases found in coastal areas and related to infections from contaminated seafood (Makino, Oshima, Kurokawa, et al.,2003). However, this bacterium also has been isolated from freshwater fish (Nair et al., 2007) and thus poses a potential threat to wider populations, in both coastal and inland areas.

The zebrafish have been successfully used as a model for understanding the interactions between the host and V. parahaemolyticus(Dong et al., 2016; Paranjpye, Myers, Yount, & Thompson, 2013).Moreover, bacterial infection models in zebrafish differ in infection route and host response. Joost J van Soest et al. shown that 25 hpf zebrafish larvae infected with Edwardsiella tarda by immersion and caudal vein microinjection may reflect an epithelial or other tissue response to cell membrane and activate the inflammatory related genes using microarray analysis methods (Soest et al., 2011). Additionally,Francisco Díaz-Pascual et al. used global proteomic profiling method to prove that 3 dpf zebrafish larvae infected with Pseudomonas aeruginosa by immersion and caudal artery microinjection could activate the angiogenesis and integrin signaling pathway and the inflammatory responses through chemokine and cytokine signaling pathways (Díaz-Pascual, Ortíz-Severín, Varas, Allende, & Ch´avez, 2017). From the previous study of Edwardsiella tarda and Pseudomonas aeruginosa, we have obtained a lot of useful information on the innate immune response in zebrafish. However, whether the innate immune response activated by V. parahaemolyticus was different to E. tarda or P. aeruginosa and the difference of innate immune response between immersion and microinjection caused by V. parahaemolyticus were both unknown. In the present study, we infected 3 dpf zebrafish larvae with V. parahaemolyticus to improve our understanding of the innate immune response induced by V. parahaemolyticus in zebrafish larvae by identifying differentially expressed genes (DEGs), GO and KEGG analysis following immersion and microinjection infection. Comparing with previous studies, V. parahaemolyticus can cause specific DEGs, including il11a, ccl34a.4, ccl20a.3, cxcl18b, and ccl35.1. In addition, immersion infection may mainly affect initial dorsal determination, cytochromes,and fatty acid-binding proteins, as well as inflammation, while microinjection infection may mainly directly affect the immune response. The results of this study will further our understanding of the infection and host immune response and facilitate the development of efficient clinical interventions for V. parahaemolyticus infection.

2. Materials and methods

2.1. V. parahaemolyticus culture and viable counts

The pathogenic V. parahaemolyticus Vp13 strain, isolated from infected Litopenaeus vannamei, was a gift from Dr. Yong Zhao (College of Food Science and Technology, Shanghai Ocean University, China). Vp13 strain was cultivated with trypticase soy broth (TSB) and trypticase soy agar (TSA) containing 3% NaCl (pH 7.3 ±0.2). Vp13 strain was incubated at 28C and 200 rpm, and then serially diluted with sterile phosphate-buffered saline (PBS) to a concentration of 10. Each diluted solution was plated on thiosulfate citrate bile salts sucrose agar culture medium (TCBS, pH =8.6 ±0.1), and incubated for approximately 24 h at 28C before counting the number of colony-forming units (CFU).

2.2. Measurement of bacterial growth curve

The Vp13 growth curve was measured as described previously(Zwietering, Jongenburger, Rombouts, & Riet, 1990) with a few modifications. Briefly, 100 μL Vp13 suspension in logarithmic phase was inoculated into TSB liquid medium contained 3% NaCl and incubated at 28C and 200 rpm. The optical density at 600 nm (OD600) was measured using a Nanodrop 2000C (Thermo, Waltham, MA, USA). After incubation for 1 h, the bacterial culture solution was continuously diluted, coated on TSA (3% NaCl), and incubated for 24 h. CFU were then counted at 1 h and every other hour from 2 to 22 h, and again after 30 h. The growth curves for OD600 vs CFU/mL and OD600 vs time were calculated. Based on the relationship between OD600 and Vp13 growth(CFU/mL), we selected Vp13 in the logarithmic phase with strong vitality and calculated the concentration of Vp13 according to the linear relationship.

2.3. Zebrafish maintenance and infection

Wild-type adult zebrafish AB strain were obtained from Shanghai Institute of Biochemistry and Cell Biology. Zebrafish were handled according to the procedures of the Institutional Animal Care and Use Committee of Shanghai Ocean University, Shanghai, China, and maintained according to standard protocols (zfin.org). Adult zebrafish were maintained in an Aquaneering system (San Diego, CA, USA), and normal water quality was maintained (temperature 27-28C, pH 7.0-8.0,salinity 0.25-0.75, ammonia and nitrate ＜200 mg/L, light 14 h and dark 10 h, and light intensity 54-324 lx). The methodology was approved by the Shanghai Ocean University Experimentation Ethics Review Committee (SHOU-DW-2016-002). Embryos were grown and maintained at 28.5C in egg water (60 μg/mL instant ocean sea salts). The other detailed parameters were the same as for adult zebrafish maintenance(Westerfield, 2000). Zebrafish larvae at 3 dpf (mouth-opening stage)were challenged with Vp13 by immersion infection, as described previously (Maley, Laird, Rinkwitz, & Becker, 2013) with a few modifications, or by microinjection infection, as described previously (Benard et al., 2012).

The LDfor immersion infection was determined by exposing 10 larvae to 5 mL Vp13 suspension in egg water, and a control group to egg water alone, and the LDfor microinjection infection was determined by injecting larvae with 1 nL Vp13 suspended in PBS, and a control group injected with 1 nL PBS (pH 7.4) alone. Three parallel samples were used in each group. The mortalities of immersion and microinjection infection zebrafish larvae were recorded for 4 days and the LDwas calculated using Bliss’s method (Bliss, 1938). In previous studies, a dose of ＞10CFU/mL was usually required to activate the innate immune response in zebrafish larvae infected by immersion (Pressley et al.,2004; Ronan et al., 2004; Soest et al., 2011; Ying et al., 2015), while doses ranging from 5 ×10to 5 ×10CFU/nL were required to activate the innate immune response by caudal vein microinjection (Mazon--Moya et al., 2017; van der Sar et al., 2003). Based on the LDdose, 3dpfzebrafish larvae underwent immersion and microinjection infection for 2 and 4 h, respectively. Samples at 2 h post-infection (hpi) were used for transcriptome data analysis and subsequent reverse-transcription quantitative PCR (RT-qPCR) validation, and 4 hpi samples were only used for subsequent RT-qPCR validation. Each experiment had three parallel samples, with 10 larvae per sample. All samples of whole larvae were flash-frozen in liquid nitrogen and stored at - 80C until RNA extraction.

Table 1 Primer sequence used for RT-qPCR analysis.

2.4. Transcriptome analysis

Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad,CA, USA) according to the manufacturer’s instructions. Random hexamers were used to synthesize first-strand cDNA and the complementary strand was then synthesized. Paired-end sequencing was carried out using an Illumina Hiseq 4000 (Illumina, San Diego, CA, USA). Mapping and enrichment were performed as described previously (Zhang, Ji, Ren,et al., 2018). The accession number of the National Center for Biotechnology Information (NCBI) Sequence Read Archive Database were SRR9325685, SRR9325684 and SRR9325687 in immersion control groups, SRR9325686, SRR9325689 and SRR9325688 in immersion infection groups, SRR9333941, SRR9333942 and SRR9333943 in microinjection control groups, and SRR9333944, SRR9333945 and SRR9333946 in microinjection infection groups.

2.5. Analysis of DEGs by RT-qPCR

cDNA synthesis reactions were performed in 20 μL reaction mixtures containing 1 μg RNA and HiScript III SuperMix for qPCR (+gDNA wiper)(Vazyme, Nanjing, Jiangsu, China) according to the manufacturer’s instructions. RT-qPCR was performed using a Roche 480 real-time PCR detection system (Roche, Mannheim, Germany) according to the manufacturer’s instructions. Results were normalized to the zebrafish β-actin gene (Dodd, Daniel, Mcnabb, & Love, 2007), which showed no changes over the time course of the infection. Results were analyzed using the 2method (Livak, Schmittgen, 2001). Each sample contained three replicate pools (10 zebrafish larvae per pool). The primer sequences are listed in Table 1.

Fig.1. Vp13 growth curves. (A) Relationship between optical density (OD600)and Vp13 growth (CFU/mL). Each dot represents one measurement once. The trendline and equation were created using the multinomial function in Excel.(B) Relationship between growth (OD600) and time. Each black dot represents one measurement.

2.6. Statistical analysis

Statistical analyses were performed using Excel (Microsoft, Redmond, WA, USA) and GraphPad Prism7 (GraphPad Software, CA, USA).t-tests were performed and values were expressed as mean ±standard error.

3. Results

3.1. Vp13 growth curve and LD50 in zebrafish larvae

Vp13 growth was monitored by measuring changes in ODat different time points at 28C (Fig.1A). Vp13 initially grew rapidly from OD0-0.9, followed by a stationary phase from OD0.9-1.8, with the end of the logarithmic phase after OD1.8 (Fig.1B). An ODof 0.9 was chosen to calculate the initial CFU based on the growth curve equation, followed by serial dilutions. The LDvalues were 3.63 ×10CFU/mL in the immersion group (Table 2) and 5.76 ×10CFU/nL in the microinjection group (Table 3).

Table 2 Bliss’s method to calculate LD50 of zebrafish immersion infected by Vp13.

Table 3 Bliss’s method to calculate LD50 of zebrafish microinjection infected by Vp13.

Table 4 Most-differentially expressed genes in immersion and microinjection groups.

Fig.2. Genes and innate immunity-related genes differentially expressed in both immersion and microinjection groups compared with their respective controls. (A) Genes differentially expressed in both immersion and microinjection groups compared with their respective controls. (B) Innate immunity-related genes differentially expressed in immersion and microinjection groups compared with their respective controls. Red color represents the degree of upregulation and green color represents the degree of down-regulation and white color represents genes that didn’t significantly changed. The same color represents DEGs have no significantly difference in two groups and the different color represents DEGs have significantly difference in two groups. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig.3. Validation of transcriptome data with RT-qPCR. Relative expression refers to the ratio of gene expression between the infection and control groups. (A)Validation of transcriptome data in the immersion-infection group. (B) Validation of transcriptome data in the microinjection-infection group. RT-qPCR results were normalized to the zebrafish β-actin gene.

3.2. DEGs in immersion and microinjection groups

We investigated and compared the effects of Vp13 infection by immersion and microinjection on the zebrafish larvae transcriptome after challenge for 2 h with 3.63 ×10CFU/mL for immersion and 5.76 ×10CFU/nL for microinjection. A total of 602 genes were differentially expressed in the immersion group, of which 427 were significantly down-regulated and 175 were significantly up-regulated.In contrast, only 359 genes were differentially expressed in the microinjection group, including 246 significantly down-regulated and 113 significantly up-regulated (Supplementary Fig.1). The most significantly up-regulated and down-regulated genes in the immersion and microinjection groups are listed in Table 4. We then compared the transcriptome data between the immersion control groups and microinjection groups and found no significant difference between these two groups.

We also compared the DEGs between the immersion and microinjection groups and found 21 genes that were differentially expressed in both immersion and microinjection groups (Fig.2A). However, the expression patterns of nine genes significantly differed between the two groups: pla2g, ctslb, and haao were down-regulated in the immersion group but up-regulated in the microinjection group, while, guca1c,aqp9b, nqas4b, usp21, cxcl8a, and arr3b were significantly up-regulated in the immersion group but significantly down-regulated in the microinjection group. fosl1a, si:ch73, hamp, prr33, si:dkey, cfb, and synpo2b were down-regulated in both groups, and cyp1a, ccl20a.3, si:dkey-57,plekhf1, and tnfb were up-regulated in both groups.

3.3. DEGs related to innate immunity in both infection groups

We compared the DEGs related to the innate immune response in zebrafish larvae following immersion and microinjection challenge with Vp13. Genes related to innate immunity that were up-regulated in both groups included tnfb and ccl20a.3, while fosl1a was down-regulated in both groups (Fig.2B). The main DEGs related to innate immunity in the immersion group were il11a, ccl34a.4, fosl1a, atf3, cmya5, c3a.1, c8a,c8b, arg2, mmp13a, and ctslb (down-regulated), and tnfb, cxcl8a,ccl20a.3, and cxcr4a (up-regulated). The main DEGs related to innate immunity in the microinjection group were il1b, il6, il34, tnfa, ccl19b,cxcr3.3, cxcl8a, cxcl18b, fosab, rel, fosl1a, c5, irak3, nfkbiaa, ncf, and mpeg1.2 (down-regulated), and tnfb, ccl35.1, ccl20a.3, irak3, and ctslb(up-regulated), respectively.

3.4. Analysis of DEGs with RT-qPCR

To validate the DEGs identified by transcriptome analysis following immersion and microinjection infection, we analyzed the mRNA expression levels of seven DEGs in the immersion group (ccl20a.3, tnfb,cxcr4a, il11a, ccl34a.4, ctslb, and c3a.1) (Fig.3A) and 10 in the microinjection group (tnfb, ctslb, irak3, ccl35.1, ccl20a.3, tnfa, cxcr3.3, cxcl18b,il6, and il1b) (Fig.3B) with RT-qPCR. The RT-qPCR results for all the examined genes matched the expression patterns shown bytranscriptome sequencing (Fig.3).

Table 5 Significantly enriched GO terms in immersion and microinjection infection groups.

Table 5 (continued)

Table 6 KEGG pathways in immersion groups.

3.5. GO and KEGG enrichment of DEGs in both infection groups

Sixty-three GO terms and four KEGG pathways were significantly enriched in the immersion infection group, compared with only three GO terms (Table 5) and no KEGG pathways (Table 6) in the microinjection group. The most significantly enriched GO terms in the immersion group were visual perception, sensory perception of light stimulus,and lipoprotein metabolic process, which were largely unrelated to the innate immune response, while the enriched KEGG pathways were complement and coagulation cascades, phototransduction, vitamin digestion and absorption, and fat digestion and absorption. In contrast,the three enriched GO terms in the microinjection group were closely related to innate immunity, including cytokine activity, cytokine receptor binding, and immune response.

Fig.4. Relative gene expression levels in control and infection groups in zebrafish larvae infected for (A) 2 h, and (B) 4 h, compared with respective control groups.Significantly DEGs between control and infection groups indicated genes that were up- or down-regulated by over two-fold. RT-qPCR results for all genes were normalized to zebrafish β-actin.

3.6. Expression of immune related genes at higher doses of Vp13

Immersion or microinjection of zebrafish larvae with the respective LDdose of Vp13 could activate the innate immune response. We investigated if challenge with a higher dose (＞LD, 1.09 ×10CFU/mL and 1.09 ×10CFU/nL, respectively) could cause more severe infection.Compared with infection with the lower dose, infection of zebrafish larvae with this higher dose for 2 h resulted in significant up-regulation of il11a, tnfa, tnfb, il1b, ccl34a.4, ccl20a.3, irak3, cxcr3.3, cxcl18b,ccl35.1, and il6, while cxcr4a, ctslb, and c3a.1 were not significantly upregulated in either group. Although all these genes were up-regulated in both groups, the expression levels of all the tested genes were lower in the microinjection group compared with the immersion group. The most changed genes were il11a, tnfa, il1b, and ccl35.1 in the immersion group(Fig.4A). The expression levels of il11a, tnfa, tnfb, il1b, ccl34a.4, irak3,cxcl18b, and ccl35.1 were further increased at 4 h post-infection (hpi)(Fig.4B) compared with at 2 hpi in the immersion group. In contrast, the expression levels of most of these genes were lower, apart from ccl20a.3 which was higher, at 4 hpi compared with 2 hpi in the microinjection group (Fig.4).

4. Discussion

Fig.5. Comparison of gene expression induced by Vp13 in different infection model. Genes in left circle represents DEGs in immersion infection group, in right circle represents DEGs in microinjection infection group and in overlap area represents DEGs in both infection groups. Red color represents differentially up-regulated genes and green color represents differentially down-regulated genes. (For interpretation of the references to color in this figure legend,the reader is referred to the Web version of this article.)

In this study, we aimed to determine if immersion and microinjection infection routes affected the innate immune response differently, and if they had other influences on zebrafish larvae infected by V. parahaemolyticus. We therefore analyzed and compared the DEGs following infection by the two different routes. Interestingly, tnfb and ccl20a.3 were up-regulated in both groups, and ccl20a.3 has been shown to attract immature dendritic cells and mediate epithelial migration(Sierro et al., 2001). However, ctslb, which is a member of the lysosomal cathepsin family involved in yolk-processing mechanisms in embryos(Tingaud-Sequeira, Cerd`a, 2007) was down-regulated in the immersion group but up-regulated in the microinjection group. In contrast, cxcl8a,which belongs to the chemokine family and has been reported to direct the migration of gut epithelial cells (Gonzalez, Lu, Boswell, et al., 2017),was up-regulated in the immersion group but down-regulated in the microinjection group. These results suggest that immersion infection may cause up-regulation of skin- and epithelial-related genes to protect zebrafish larvae, while microinjection may up-regulate genes that regulate innate immune related cell migration and intracellular processes, such as lysosomal cathepsin.

To better understand the innate immune response, we analyzed genes reportedly associated with innate immunity (Van, Spaink, &Meijer, 2012). Most innate immune-related genes were down-regulated in both groups, and only tnfb, cxcl8a, ccl20a.3, and cxcr4a in the immersion group and tnfb, ccl35.1, irak3, and ctslb in the microinjection group were up-regulated. The RT-qPCR results were consistent with the transcriptome data, indicating that incubation with Vp13 for 2 h initiated an innate immune response in zebrafish larvae. We also used a higher dose of Vp13 to cause more severe infections in the immersion and microinjection groups and showed that classical cytokine genes,including il1b, tnfb, and il6, were significantly up-regulated after 2 and 4 hpi in both groups, and il11a, ccl34a.4, and cxcl18b, which were differentially expressed according to our transcriptome data, were also significantly up-regulated after 2 and 4 hpi in both groups. However, the specific roles of these genes in the innate immune response remains unclear (Fig.5). The il11 gene, a member of the il6 family (Nishina,Komazawasakon, Yanaka, et al., 2012), and il11a and il11b, have previously been identified in fish (Huising et al., 2005), while ccl34a and ccl35.1 may be inflammation-related genes conserved between zebrafish and mammals (FornCuní, Varela, Pereiro, Novoa, & Figueras, 2017),and cxcl18b has demonstrated chemotactic activity towards neutrophils,similar to cxcl8a (Torraca, Otto, Tavakoli-Tameh, & Meijer, 2017).These genes that were activated in both the immersion and microinjection groups may play major roles in the defense against bacterial infection, but further studies are needed to determine their specific roles in the innate immune response.

Previous results showed that E. tarda-immersed and injected larvae demonstrated an epithelial or other tissue response to the cell membrane with activation of inflammation-related genes including mmp9, cyp1a,zgc:154020, irg1l and stc1 (Soest et al., 2011), P. aeruginosa-immersed and injected larvae showed activation of the angiogenesis and integrin signaling pathway and inflammatory responses by mediating chemokine and cytokine signaling pathways, and differential expression of the innate immune related genes tnfa, il1b, mmp9, il8l1, tnfb, cxcl11.6, and cyp1a. Interestingly, cyp1a was differentially expressed by immersion infection with E. tarda, P. aeruginosa, and V. parahaemolyticus, indicating that cyp1a may play an important role in the response to immersion infection. Our results also suggested that, in addition to the classical innate immunity genes tnfa, tnfb, il1b, and il6, the genes il11a, ccl34a.4,ccl20a.3, cxcl18b, and ccl35.1 were also important for defending against Vp13 infection. The current study also found that GO terms significantly enriched in the immersion group were more related to early developmental processes, such as dorsal determination, cytochrome formation,and fatty acid metabolism, while the enriched KEGG pathways included the complement and coagulation cascades, phototransduction, vitamin digestion and absorption, and fat digestion and absorption. In contrast,the GO terms enriched in the microinjection group were cytokine activity, cytokine receptor binding, and immune response, which were all directly related to immunity.

The most-changed gene in the immersion group was al929131.1,which has not previously been reported in the literature and thus warrants further exploration. Other significantly changed genes included wnt6a, which may cooperate with maternal wnt8a in initial dorsal determination (Hino, Nakanishi, Seki, et al., 2017), cyp1a, cyp1b, and cyp1c, which are part of a large superfamily of enzymes that primarily catalyze mixed-function oxidation reactions (Scornaienchi, Thornton,Willett, & Wilson, 2010), and fabp10a, which encodes a fatty acid-binding protein (Venkatachalam, Lall, Denovan-Wright, & Wright,2012). In contrast, mpeg1.2, which is expressed in macrophages and has shown an anti-bacterial function in zebrafish (Benard et al., 2015), and cd83, which is required for the induction of protective immunity (Li, Li,& Sun, 2015), were significantly up-regulated in the microinjection group. These results suggest that immersion infection may affect biological processes other than the innate immune response, such as initial dorsal determination, cytochromes, and fatty acid-binding, besides inflammation, while microinjection infection mainly affects the innate immune response. We also considered that the innate immune response induced by the LDof Vp13 at 2 hpi represented the initial stage of infection. Only some important immune-related genes were up-regulated in the two infection groups, while a higher dose may significantly up-regulate some immune marker genes that were down-regulated in the LDinfection groups.

We also compared the gene expression profiles after different infection doses. Expression levels of il11a, tnfa, tnfb, il1b, ccl34a.4,ccl20a.3, irak3, cxcr3.3, cxcl18b, ccl35.1, and il6 were higher following infection with a ＞LDdose compared with an LDdose for both infection routes, though these genes were more significantly upregulated in the immersion compared with the microinjection group(Fig.5). This may be because high dose of immersion infection might activate more immune systems, including the skin, mucosal, and respiratory systems, though further studies are needed to verify this hypothesis.

This study also had some limitations. Induction of the innate immune response by either the immersion or injection route involved a complicated process. The innate immune response of long-term infection remains unknown. Further studies are also needed to determine if some of the genes differentially expressed in our transcriptome data responded specifically to Vp13 infection. More attention should also be paid to genes that are differentially expressed between the immersion and microinjection groups in relation to the possible cure of diseases resulting from these different infection routes.

5. Conclusion

In this study, we compared innate immune related transcriptome changes in zebrafish larvae following infection with V. parahaemolyticus Vp13 by static immersion and caudal vein microinjection. Different infection methods could activate different innate response genes and induced different immune-defense pathways. These results provide an insight into the role of differentially expressed immune-related genes during the early stage of Vp13 infection and suggest that some genes related to Vp13 infection defense might be specific to different infection routes. Further studies should be conducted to clarify the roles of these genes in the mechanism of V. parahaemolyticus infection.

Declaration of Competing Interest

The authors declare that there is no conflicts of interest.

Acknowledgments

This work was supported by the Ministry of Education Returnees Research Fund (D-8002-15-0042), the China-US Ocean Research Center Fund (A1-3201-19-3013), the Shanghai First-Class Discipline Construction Fund.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.aaf.2019.08.002.