1. Introduction
The grass carp (
Ctenopharyngodon idellus) has been reared in China for more than 60 years as an important aquaculture species. The production of grass carp reached 5.8 million tons in 2015, accounting for more than 13% of the world’s freshwater aquaculture production [
1]. Grass carp haemorrhage disease, caused by grass carp reovirus (GCRV), is one of the most damaging diseases, resulting in huge economic losses to the aquaculture industry of grass carp [
2]. GCRV was first isolated in China and belongs to the genus
Aquareovirus of the family
Reoviridae [
3]. GCRV not only infects grass carp but also infects the rare minnow (
Gobiocypris rarus), black carp (
Mylopharyngodon piceus) and topmouth gudgeon (
Pseudorasbora parva), causing haemorrhagic symptoms and death in these species.
Grass carp haemorrhage disease outbreaks are frequent and result in huge economic losses in the aquaculture industry, hence the GCRV is of particular interest to geneticists aiming to develop strategies for breeding disease-resistant fish. Recent studies on GCRV have mainly focused on identification of genes involved in immune responses to GCRV infection [
4,
5], cloning and characterization of GCRV-encoded proteins [
6,
7], the development of vaccines against GCRV [
8,
9], and transcriptomic sequencing of discovered immune-related genes or uncovered the antiviral immunity mechanism of grass carp infected with GCRV [
10,
11,
12,
13,
14]. However, the mechanisms of grass carp reovirus invasion and infection remain poorly explored.
In the present study, C. idellus kidney (CIK) cells were infected with GCRV, harvested at 0, 8, 24, and 72 h post infection, respectively, and subjected to transcriptomics sequencing. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were performed on differentially expressed genes (DEGs) identified during different stages of infection. Our findings expand our understanding of the mechanisms underlying grass carp reovirus infection.
3. Discussion
At present, previous studies on grass carp transcriptome profiles in response to grass carp reovirus infection have been carried out [
10,
11,
12,
13,
14]. Most of them involved in the discovery of immune-related genes or uncovered the antiviral immunity mechanism of grass carp. However, the purpose of our study is to understand the whole process of grass carp reovirus infection.
In this study, we used the CIK cells as an in vitro model to explore the molecular changes following reovirus infection. CIK cells were mock-infected or infected with GCRV and harvested at 0 (mock), 8, 24, and 72 hpi. To better understand the process of GCRV entry into cells, it was divided into three stages: early (0–8 h), middle (8–24 h), and late (24–72 h). In each stage, data from the latter time point were compared with that from the former to identify DEGs. The method for DEG identification differed that in the previous study, which used a case-control method [
15,
16]. DEGs in the 0–8 h were mainly associated with virus adhesion, DEGs in the 8–24 h were largely associated virus endocytosis and transmission, and DEGs in the 24–72 h were involved in cell death and apoptosis. Therefore, the DEGs identified in the present study appear to be representative and accurate.
The CPE appeared as early as 8 h after CIK cells infected with GCRV. DEGs from this stage were mainly enriched with Focal adhesion and ECM-receptor interaction terms. Focal adhesions are sub-cellular structures that mediate the regulatory effects of cells in response to ECM adhesion [
17]. In fact, previous studies showed that focal adhesion kinase can interact with rabies virus phosphoprotein P, which is involved in the process of viral infection [
18]. During influenza A invasion of cells, inhibition of focal adhesion kinase signaling leads to reduced viral replication [
19], suggesting that focal adhesion may be an important pathway during GCRV infection. Moreover,
hspg2,
ptpε, and
ppl were significantly up-regulated during the 0–8 h post infection. HSPG2 plays an important role in mediating virus envelope-target cell interaction during human papillomavirus (HPV) and hepatitis C virus (HCV) infection [
20,
21]. PTPε plays an important role in the control of macrophage function [
22]. And what’s more, tyrosine-protein phosphatase non-receptor type 22 (PTPN22) associates with TNF receptor-associated factor 3 (TRAF3) to augment Toll-like receptor (TLR)-induced type I IFN production [
23]. However, PPL acts as a cytolinker between intermediate filament scaffolding and the desmosomal plaque and is involved in epithelial cohesion, intracellular signal transduction and antigen presentation [
24]. These results suggest that
hspg2 may be involved in cellular binding of GCRV, after which GCRV likely binds to the receptor on the cell membrane,
ptpε and
ppl maybe as signaling molecules that regulate cellular processes.
During the 8 to 24 hpi period, enriched DEGs were mainly related to phagosome and lysosome pathways. In recent years, more and more studies have shown that the viral infection process is closely related to autophagy of host cells. Autophagy can be used as innate immunity and an adaptive immune response against intracellular pathogens [
25]. In the phagosome pathway, many genes are up-regulated, including
itgb2,
srb1, and
sec22b, and it is worth noting that the
itgb2 is up-regulated at this period. Cell integrins are commonly used receptors for diverse viral pathogens [
26], and reovirus internalization is mediated by integrin β1 proteins, most likely via clathrin-dependent endocytosis [
27]. Moreover, integrin β3 proteins serve as receptors of pathogenic hantaviruses [
28]. In the phagosome pathway, other receptor genes appear to be up-regulated, such as scavenger receptor
srb1, which, along with CD81, is the receptor used for entry of the hepatitis C virus into liver cells [
29]. The v-SNARE protein SEC22 is involved in the transport of eukaryotic cell endoplasmic reticulum (ER) and Golgi apparatus [
30]. Hepatitis C virus may be assembled in the ER and transported to the Golgi apparatus in COPII vesicles to begin Golgi secretion [
31]. However, a recent study shows that SEC22 reduces RNAi efficiency by influencing late endosomal functions, through the promotion of fusion between late endosomes and lysosomes [
32]. These results suggest that
itgb2 and/or
srb1 may be important for cell entry of GCRV. Furthermore, GCRV entry into the endosomal region followed by fusion with lysosomes is likely to be regulated by
sec22b.
In the 24–72 h post infection, DEGs were mainly enriched in lipid metabolism, programmed cell death, apoptosis, and inflammation. Genes in the cell death pathway have been studied intensely, and it is interesting to note that genes in the lipid metabolism pathway, such as
cyp51a, are enriched. CYP51A1 is a cytochrome P450 enzyme, a conserved group of proteins that serve as key players in the metabolism of organic substances and the biosynthesis of important steroids, lipids, and vitamins in eukaryotes [
33], and it plays an essential role in mediating membrane permeability [
34]. These sterols, located on the plasma membrane of cells, play an important structural role in regulating the fluidity and permeability of membranes while also affecting the activity of enzymes, ion channels and other cellular components [
35]. Therefore, these results suggest that GCRV probably not only causes cell death through cell apoptosis and inflammatory pathways but also changes the permeability of the cell membrane by affecting the structure of steroids in the cell membrane, leading to cell death through this mechanism.
4. Materials and Methods
4.1. Cells, Virus and Virus Infection
C. idellus kidney (CIK) cells purchased from Shenzhen ziker Biological Technology Co., Ltd. (Shenzhen, China) and grown in Medium 199 (Gibco, Gaithersburg, MD, USA) supplemented with 10% fetal bovine serum (FBS), 100 mg/mL penicillin, and 100 mg/mL streptomycin at 28 °C. GCRV was kindly provided by Professor Jianguo Su (Huazhong Agricultural University). CIK cells that grown in 25 cm2 flask were mock-infected or infected with GCRV at a multiplicity of infection (MOI) of 1, then harvested at 0 (mock), 8, 24, and 72 h post infection (pi), respectively. Three flasks of cells were collected for each treatment at each time point.
4.2. CCK-8 Assay
CCK-8 detection kit (Beyotime, Shanghai, China) was used to measure cell viability according to the manufacturer’s instructions. About 4 × 103 CIK cells were seeded in 96 well plates and cultured in M199 supplemented with 10% FBS at 28 °C for 24 h. Wells containing no cells but only culture medium alone were served as blank control. CIK cells were mock-infected or infected with GCRV at an MOI of 1 and maintained for 0, 8, 24, and 72 h at 28 °C, respectively. Then each well was added 10 μL of CCK-8 solution. After 4 h incubation at 28 °C, the plates were measured spectrophotometrically on Microplate Manager 6 (BIO-RAD, Hercules, CA, USA) at wavelength 450 nm. The relative cell viability was calculated as Ainfected-Ablank-i/A mock-Ablank % (Ablank-i represented the wells containing no cells but only medium and GCRV). Data were presented as means of at least three independent experiments ± standard deviation (S.D).
4.3. RNA Isolation, Library Construction, and Sequencing
Cells were collected at 0, 8, 24, and 72 h after infection (C0, C8, C24, and C72 groups, respectively), rapidly frozen in liquid nitrogen, and total RNA was extracted with Trizol reagent (Invitrogen, Waltham, MA, USA). RNase-free DNase treatment was performed in order to remove possible genomic DNA. The DNase was removed by phenol-chloroform re-extraction. mRNA was purified from total RNA using poly-T oligo-attached magnetic beads and segregated into 200–300 bp fragments. The concentration of mRNA used to prepare first strand cDNA is approximate 40 ng/μL and a total of 3 μg mRNA used as input material for first strand DNA synthesis. First strand DNA was synthesized from 6-base random primers and reverse transcriptase using RNA as a template, and second strand cDNA was synthesized using the first strand cDNA as template. In the second strand cDNA synthesis, the base T was replaced by U to generate a chain-specific library. RNA-Seq was performed using chain-specific kits, and chain-specific libraries were used to determine the transcriptional direction of sense and antisense strands to increase the accuracy of subsequent gene function annotation and gene expression analysis.
After library construction, library fragments were enriched by PCR amplification, and selection based on fragment size resulted in a 300–400 bp library. The library was tested using an Agilent 2100 Bioanalyzer (Agilent, Palo Alto, CA, USA), and both the total and effective concentration of the library were tested. Based on the effective concentration and the amount of data required, libraries containing different Index sequences (samples plus different Index sequences, and difference data of samples according to Index) were mixed proportionally. The mixed library was diluted to 2 nM and alkaline degeneration was performed to generate a single-chain library. Next, 150 bp pair-end reads were performed on the library using an Illumina NextSeq500 sequencing platform to carry out Next-Generation Sequencing (NGS) at Personalbio (Personalbio, Shanghai, China). For each treatment in each time point, the sample was sequenced three times as three technical replicates.
4.4. Data Analysis
Clean data were obtained by removing adapters, poly-N sequences and poor-quality data [
36]. The Q20, Q30, and GC content of clean data were calculated, and all downstream analysis was performed using high-quality clean data.
Clean data were mapped to the grass carp reference genome [
37] using TopHat2 software [
38]. The number of reads mapped to each gene was used HTSeq software [
39], and the reads per kilobase of the exon model per million mapped reads (RPKM) was calculated as expression values for each gene [
40].
4.5. Differential Expression Analysis
Differential expression analysis using data collected at different time points was performed using the DESeq package [
41]. The resulting
p-values were adjusted using the Benjamini and Hochberg approach to control the false discovery rate. Genes with an adjusted
p-value < 0.05 (
q-value < 0.05) and fold change > 2 in DESeq analysis were assigned as differentially expressed genes (DEGs).
ClueGO and CluePedia were used to perform Gene Ontology (GO) enrichment analysis [
42,
43]. In GO enrichment analysis, only categories with a
p-value < 0.05 were considered as enriched in the network.
The Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to indicate the location of the DEGs in the different pathways. In this study, KOBAS software was used to test the statistical enrichment of DEGs in KEGG pathways [
44]. Pathways with
p-values < 0.05, were considered for as statistical significance.
4.6. Validation of DEGs by RT-qPCR
To confirm the reliability of the data obtained with RNA-Seq, eight DEGs were selected for RT-qPCR using the primers listed in
Table S5. Primers were designed using the Primer Premier 5.0 software (PREMIER Biosoft International, Palo Alto, CA, USA) and only with an efficiency of 90–110% were used. First strand cDNAs were obtained from total RNA (>3 μg) using a random hexamer primer and the ReverTra Ace kit (TOYOBO, Osaka, Japan). RT-qPCR was performed using a fluorescence quantitative PCR instrument (BIO-RAD). Each RT-qPCR mixture contained 0.8 μL of each primer, 1 μL template, 10 μL 2× SYBR Green master mix (TOYOBO), and 7.4 μL ddH
2O. The program for RT-qPCR was 95 °C for 10 s, followed by 40 cycles at 95 °C for 15 s, 60 °C for 15 s, and 72 °C for 30 s. Three replicates were included for each sample, and the
β-actin gene of grass carp was served as an internal control to normalize the expression levels. Relative expression levels of the different genes were calculated using the 2
−ΔΔCt method [
45]. All data represent the mean ± standard deviation of three replicates.