Background
The globalization of pig industry has promoted the emergence of infectious diseases affecting pigs and the spread of their pathogens, which is challenging for the healthy development of this industry. African swine fever (ASF) is an acute, febrile, highly contagious, and fatal animal infectious disease caused by ASF virus (ASFV) [
1,
2], a large double-stranded DNA virus. The genomic length of different isolates varies from 170 to 190 kbp, encoding 151–167 open reading frames and > 170 proteins [
3]. ASFV is the only member of the
Afarviridae family and the only known insect-borne DNA virus that affects mammals [
4‐
6]. It infects breeds of domestic pigs, African and Eurasian wild boars, and blunt ticks [
7,
8]. Since 2018, a highly virulent type II ASFV has spread to China. ASFV has dealt a heavy blow to China, the world's largest producer and consumer of pork [
9]. Although the experimental vaccine was produced by a natural, cell culture attenuated, or genetically modified ASFV, no effective vaccine has yet been produced.
Severa ASFV proteins have a major role in evading host defense and regulating host immune response by inhibiting interferon (IFN) production, apoptosis, and autophagy [
6,
10‐
13]. For example, A238L and DP71L proteins regulate the host cell protein expression system that inhibits the host cell, shut down the expression system and inhibit the activation of transcription factors such as NAFT. The multigene family 360 (MGF360), MGF505/530 and I329L inhibit the anti-viral effect mediated by type I IFN; p54, DP71L, A179L, and A224L regulate apoptosis in the early and later stages of infection; A179L inhibits autophagy. In addition, DP96R of ASFV China/2018/1 negatively regulates type I IFN expression and nuclear factor-kappa B (NF-κB) signal transduction by inhibiting TBK1 and IKKβ [
14]. The strong immune escape ability of ASFV makes it a powerful “killer”; thus, it is crucial to study the mechanism underlying the interaction with the host. Although some progress has been made in this research area, due to the large genome and complex structure of the virus, the exact mechanism underlying the interaction with the host is yet to be elucidated. The transcriptional analysis of host cell response to viral infection could be used to study the potential cytokines directly or indirectly related to viral infection and deduce the immune escape mechanism of the virus. RNA sequencing (RNA-seq) of transcriptome is a newly developed approach that can explore the mechanism of cellular signal transduction [
15]. RNA-seq technique reveals the dynamic changes of the pathogen genome and the systematic changes in the host gene expression profile during pathogen infection [
16,
17]. Previous studies have reported changes in the gene expression of PAMs infected with ASFV Malawi LIL20/1 isolate or ASFV Georgia 2007 strain [
18,
19]. Currently, 13 ASFV strains are isolated from China, one of the major endemic places for ASFV. However, there is no report on the transcriptome of PAMs infected with ASFV China isolates.
In this study, RNA-seq was used to annotate host responses to ASFV-CN/GS/2018 strains isolated from China post-infection in PAMs. We also studied the differential gene expression of PAMs infected with ASFV-CN/GS/2018. The present study aimed to understand the host response at the various stages of ASFV-CN/GS/2018 infection at the cellular level, to provide a basis for an in-depth understanding of the biological mechanism of ASFV-host interaction, and to explain how ASFV infection regulates the host cell environment. These findings would contribute to the development of vaccines and other control strategies.
Materials and methods
Cell culture and virus
Porcine alveolar macrophages (PAMs) were prepared from bronchoalveolar lavage as described previously, cultured in Roswell Park Memorial Institute (RPMI) medium containing 10% porcine serum [
20], and grown at 37 °C in a 5% CO
2 atmosphere saturated with water vapor. ASFV-CN/GS/2018 is a virulent strain with genotype II, with no deletion of genes that inhibit the host response. The virus is provided by Lanzhou Veterinary Research Institute.
To determine the proliferation of the ASFV-CN/GS/2018 strains in PAMs, monolayers were prepared in 6-well plates and infected at multiplicity of infection (MOI) of 0.01 or 1. After 1 h of adsorption at 37 °C under 5% CO2, the inoculum was removed, and the cells were rinsed two times with phosphate-buffered saline (PBS). Then, the monolayers were rinsed with macrophage medium and incubated at 37 °C under 5% CO2 for different durations.
Virus titration (50% hemadsorption doses)
The anticoagulant whole blood of healthy pigs is washed with sterilized PBS (pH 7.2) containing 1% gentamicin and centrifugation at 350×g for 3 min each time; subsequently, the PAMs are seeded in 96-well plate and the pig red blood cells are seeded in 96-well plate in a 20 μL volume. The sample was diluted at 10–1, 10–2, 10–3, 10–4, 10–5, 10–6, and 10–7, and plated in eight wells in a 96-well plate containing PAMs and red blood cells. The adsorption of red blood cells was observed for 7 days. Calculate 50% hemadsorption doses (HAD50) were calculated according to the Reed-Muench method.
RNA-seq library preparation and Illumina sequencing
For cDNA library preparation, total RNA from the cell lines was treated with RNase-free DNase I (TaKaRa) following the manufacturer’s instructions. RNA was quantified using a NanoDrop ND1000 spectrophotometer (Thermo-Fisher Scientific), and the quality was assessed using a model 2100 Bioanalyzer (Agilent). The RNA integrity number value of each sample was > 8. The cDNA libraries were prepared according to the standard Illumina protocol (NEBNext® Ultra™ II RNA Library Prep Kit for Illumina®) and then subjected to sequencing using an Illumina HiSeq™ 2000 sequencer. The libraries were quantified using a DNA-1000 Kit Bioanalyzer (Agilent).
Transcriptome assembly and transcriptional profiling analysis
After filtering the readings with sequencing connectors and low-quality readings, Hisat2 2.2.1.0 (RNA-strandness rf–fr) was used to align the remaining readings against the pig genome (Sscrofa11.1 GCF_000003025.6) and ASFV genome (GenBank: MK333180.1). HTSeq-count 0.9.1 (-s reverse) was used to analyze the reading distribution of known genes.
In order to analyze the gene level of PAMs infected with ASFV, the Cufflinks 2.1.1 (library-type fr-firststrand) program was used to quantify the fragments per kilobase in each million mapped readings of the genetic model (FPKM) to identify the genes in each cell. The false discovery rate (FDR)-corrected P-value < 0.05 was considered for differentially expressed genes (DEGs).
Gene ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis
GO functional classifications were defined using the Blast2GO software. The enriched gene functional categories were further classified based only the GO analysis, P-value < 0.05. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database was accessed using the KOBAS software via hypergeometric test, with a corrected p-value < 0.05. Q-value was used as a statistical method for estimating FDR, which is a conventional measure in the analysis of genome-wide expression data, with a corrected P-value < 0.05.
Cell viability assays
The cell viability was measured using the cell counting kit-8 (CCK8) assay according to the manufacturer’s instructions. Briefly, the cells were seeded in 96-well plates and was infected with ASFV at 12, 24, and 36 h, respectively. Subsequently, 10 μL CCK-8 reagent (Apexbio) was added into each well, and after incubation at 37 °C for 2 h, the absorbance measured at 450 nm on a multifunction microplate reader (BioTek). The percentage at each concentration relative to the control was presented as cell viability.
Real-time qPCR
Total RNA was extracted from PAMs using TRIzol reagent and reverse transcribed with PrimeScript RT kit (TaKaRa). qPCR was performed using a Power Up SYBR Green Master Mix on ABI StepOnePlus system, and data were analyzed by StepOnePlus software. The relative mRNA level of target genes was normalized to the porcine
GAPDH mRNA level. The relative expression of mRNA was calculated based on the comparative cycle threshold (2
−ΔΔCT) method [
21].The Gene ID and primer sequence information are provided in Table
1.
Table 1
The gene ID involved in this study and the primers and oligonucleotides used
Porcine IL33-F | CTTCATGAGCAGCCCTCCAA | 100518643 |
Porcine IL33-R | TCCGCAGCTTTCTGTCACAT |
Porcine IFITM3-F | CTGGTCCCTGTTCAACACCC | 100518544 |
Porcine IFITM3-R | TGCAAACGATGATGAACGCAA |
Porcine BMP8A-F | CAGTCAGCACAGAAGTCCCC | 100515668 |
Porcine BMP8A-R | CATCGAGGGTGTGTGTTCCT |
Porcine CDKN2B-F | CAAAGTGAGCGAGGAGGACAA | 397227 |
Porcine CDKN2B-R | CAGAAGTTGACGCACGGTCT |
Porcine HPSE-F | AACCATAGACGGCAACCTGG | 100271932 |
Porcine HPSE-R | TCTCAGGTATGCGGGAGACA |
Porcine MARCO-F | AAGGCCCACCAGGAATCAAG | 100516298 |
Porcine MARCO-R | AAGTCACCTTTATGCCCCCG |
Porcine TLR3-F | ATGGATTGCTCCCCTTCACC | 100037937 |
Porcine TLR3-R | CAGGGTTTGCGTGTTTCCAG |
Porcine TLR4-F | GACAGCAATAGCTTCTCCAGC | 399541 |
Porcine TLR4-R | AAAGGCTCCCAGGGCTAAAC |
Porcine TLR6-F | TCTCATGGCACAGCGAACTT | 396621 |
Porcine TLR6-R | TCACATCATCCTCTTCAGCGA |
Porcine TLR7-F | GCTGTTCCCACTGTTTTGCC | 100037296 |
Porcine TLR7-R | ACTTGCGGTTGACTGAGGTT |
Porcine DDX58-F | GGAGATGCTTTCAGGGAGCG | 396723 |
Porcine DDX58-R | GCAGTCTGGCCTAgCACAATA |
Porcine IFIH1-F | AGCCACAGATCAGCCAAGTC | 100101927 |
Porcine IFIH1-R | TCCCATGGTGCCTGAATCAC |
Porcine IFIT1-F | TCCGACACGCAGTCAAGTTT | 100153038 |
Porcine IFIT1-R | TGTAGCAAAGCCCTGTCTGG |
Porcine IFIT2-F | GCACAGCAATCATGAGTGAGAC | 100155467 |
Porcine IFIT2-R | GGCCTGTATGTTGCACATCG |
Porcine IFITM3-F | CTGGTCCCTGTTCAACACCC | 100518544 |
Porcine IFITM3-R | TGCAAACGATGATGAACGCAA |
Porcine RSAD2-F | AAAGACGTGTCCTGCTTGGT | 396752 |
Porcine RSAD2-R | CTTCCGCCCGTTTCTACAGT |
Porcine ETAA1-F | TCTCAACAGCCAAAATGGCG | 100622990 |
Porcine ETAA1-R | CGACTCATTGCCTAGGACCC |
Porcine IL-6-F | ACAAAGCCACCACCCCTAAC | 399500 |
Porcine IL-6-R | CGTGGACGGCATCAATCTCA |
Porcine TNF-α-F | CCAGACCAAGGTCAACCTCC | 397086 |
Porcine TNF-α-R | TCCCAGGTAGATGGGTTCGT |
Porcine NF-κB-F | CCCATGTAGACAGCACCACCTATGAT | 751879 |
Porcine NF-κB-R | ACAGAGGCTCAAAGTTCTCCACCA |
Porcine GPR37-F | TTCCACGGTGACCAGTGATG | 100523220 |
Porcine GPR37-R | ACAGAAGCGAACGTGGACAT |
Porcine CCL4-F | ATGAAGCTCTGCGTGACTGT | 396668 |
Porcine CCL4-R | AGTCACGAAGTTGCGAGGAA |
Porcine CCL5-F | ACACCACACCCTGCTGTTTT | 396613 |
Porcine CCL5-R | TGTACTCCCGCACCCATTTC |
Porcine CXCL8-F | AGCCCGTGTCAACATGACTT | 396880 |
Porcine CXCL8-R | TGGAAAGGTGTGGAATGCGT |
Porcine CXCL10-F | ATAAGGATGGGCCGGAGAGA | 494019 |
Porcine CXCL10-R | GTGGGAGCAGCTAACTTGGT |
Porcine CXCR2-F | GTGGAAACAGCAACTGCTCA | 100124654 |
Porcine CXCR2-R | AGGGCTTGGTAGTTGTCAGG |
Porcine ISG12(A)-F | AGATACTGGCGACAGGGAGT | 100153902 |
Porcine ISG12(A)-R | AGGGCAGCCTTGAATGACAG |
Porcine TNFSF10-F | TTGTGGAGCTCTGCCTGATG | 406191 |
Porcine TNFSF10-R | ACCTTTCAGTGCTGCCCTTT |
Porcine GADD45B-F | GCCGCGGGTTCAGATTATTG | 100621090 |
Porcine GADD45B-R | ACCTTCAGATCGCAGCGAAA |
Porcine IFI6-F | TCTGCTCTCTTCAAGGTCCG | 110261124 |
Porcine IFI6-R | TCCACCGCAGGTGTAGAGTA |
Porcine PIK3CB-F | CTGCAGCTGGACGGTCG | 100622559 |
Porcine PIK3CB-R | CCACTCACAATTTCACTGCCC |
Porcine HRK-F | ACGCTCTTTCATGTCTGGGG | 100155596 |
Porcine HRK-R | CGTACAAACTGGCCCTGAGT |
Porcine GAPDH-F1 | ACATGGCCTCCAAGGAGTAAGA | 396823 |
Porcine GAPDH-R1 | GATCGAGTTGGGGCTGTGACT |
Samples were collected at a specified time after PAM was inoculated with ASFV. Real-time quantitative PCR using ASFV P72 gene as a target to detect the copy number of ASFV genomic DNA. Using QIAamp® DNA Mini Kits (Qiagen, Germany) to extract sample DNA and then qPCR was carried out on a Bio-Rad system.
-
ASFV-P72-R: 5′-CTGCTCATGGTATCAATCTTATCGA-3′;
-
ASFV-P72-F: 5′-GATACCACAAGATCAGCCGT-3′;
-
Taqman: 5′-CCACGGGAGGAATACCAACCCAGTG-3′.
Amplification conditions used were a preheating at 95 °C for 30 s and 40 cycles of 95 °C for 5 s and 58 °C for 30 s. The quantity of ASFV genome was calculated using the standard curve and expressed as genome copies per milliliter.
Statistical analysis
The significance of the results between the experiments was analyzed using GraphPad Prism 8 (San Diego, CA, USA). All data are presented as mean values ± standard errors (SEs) from three independent experiments. *P < 0.05 was considered statistically significant. **P < 0.01 and ***P < 0.001 was considered highly statistically significant.
Discussion
ASFV is harmful pathogen to pigs. Since 2018, it has caused huge economic losses to the pig industry in China [
9]. Macrophages are major target cells of ASFV, and they are also important immune cells of the host [
26]. In addition, macrophages also trigger acquired immunity. Therefore, an in-depth insight into the transcriptional changes of ASFV-infected macrophages is crucial to understand the host–pathogen interaction. With the continuous development of new technology, RNA-seq is a major tool to elucidate the transcriptional spectrum [
27]. Previous studies have applied RNA-seq technology to transcriptome studies in pigs infected with highly virulent (Georgia 2007 strains) or low virulent (OURT33) ASFV [
28]. Some studies have revealed the altered of gene expression in PAMs infected with ASFV Georgia 2007 strain within 18 hpi [
19]. One replication cycle of ASFV is about 16 h, and infectious offspring virus can be produced at 16 hpi. We selected three time points to collect samples: 12, 24 and 36 hpi. In the present study, we used PAMs as an in vitro model and analyzed the transcriptional changes of host cells infected with ASFV-CN/GS/2018 strain using RNA-seq technique. A total of 1154 DEGs were identified, of which 816 genes were upregulated, and 338 genes were downregulated (Fig.
2C). The KEGG enrichment analysis of DEGs found that TLR and RLR signaling pathways may be involved in response to ASFV infection (Fig.
4B, C). Subsequent qPCR verification found that the transcription of
TLR3 (Fig.
5A)
, TLR7 (Fig.
5D)
, DDX58 (Fig.
5E)
, and
IFIH1 (Fig.
5F) was upregulated, which further suggested that TLR and RLR signal pathways may be activated after ASFV infection.
TLR3 mainly recognizes dsRNA;
TLR7 primarily identifies ssRNA and a few short dsRNA;
DDX58 identifies dsRNA and 5Powerppp ssRNA, while
IFIHI identifies dsRNA with a length > 1 kbp. Reportedly, some DNA viruses, such as herpesvirus infection, can activate the RIG-I signaling pathway [
29]. HSV-1 infection increases the content of RNA5SP141 in the cytoplasm and downregulates proteins that bind to RNA5SP141, which in turn binds RNA5SP141 to RIG-I and induces type I interferon [
30]. However, additional studies are required to identify whether and how the sensor pathway of RNA is involved in the infection process of ASFV, a DNA virus.
Also, in this study, the transcriptional levels of downstream anti-viral and inflammatory factors were analyzed further, and the results of RNA-seq and qPCR showed that ASFV infection could upregulate the transcriptional level of
IFIT1 (Fig.
6A)
, IFIT2 (Fig.
6B)
, IFITM3 (Fig.
6C)
, RSAD2 (Fig.
6D),
IL-6 (Fig.
6F),
TNF-α (Fig.
6G) and
NF-κB (Fig.
6H). The activation of these factors indicates that the PAMs is in an anti-viral state, which is verified with the activation of the above immune-related pathways. However, the differential expression of anti-viral and inflammatory factors revealed that the immune and inflammatory activation of PAMs infected with ASFV was very limited. We speculated that after virus infection, immune and inflammation-related pathways are activated and then suppressed by a large number of immune escape proteins encoded by ASFV. In addition, the transcriptional levels of
ETAA1 (Fig.
6E) and
GPR37 (Fig.
6I) are downregulated during ASFV infection. Previous studies have shown that removing a gene called
ETAA1 from mice prevents the animal from producing an immune response to vaccines or infections [
31]. Mice without
GPR37 showed delayed phagocytosis of macrophages and a delayed regression of inflammation. At the cellular level, macrophages without
GPR37 gene showed an imbalance of anti-inflammatory and pro-inflammatory cytokines [
32]. Another important role of
GPR37 is to regulate the phenotype of macrophages. Macrophages expressing
GPR37 show more M2 than M1 [
32].
Macrophages produce chemokines that induce pathology and protective immunity and play a key role in anti-viral response [
33]. Additionally, some large DNA viruses, such as herpesvirus and poxvirus, can regulate chemokine activity by encoding homologs of chemokine ligands and receptors [
34]. In order to further understand how ASFV manipulates the host chemokines, the chemokine-related factors differentially expressed in RNA-seq were verified. The current data suggested that the transcriptional level of
CCL4 (Fig.
6H),
CCL5 (Fig.
6I),
CXCL8 (Fig.
6J), and
CXCL10 (Fig.
6K) were upregulated to varying degrees after ASFV infection.
CCL4 is a pro-inflammatory chemokine that promotes the development of lymphocytes, which produce IFN-γ [
35].
CCL4 and
CXCL10 have a chemotactic effect on CD4
+T cells. CXCL8 is the primary mediator of an inflammatory response, attracting neutrophils, basophils, NK cells, and T cells [
36]. In addition, some studies have shown that the expression level of CXCL8 and CXCL10 in macrophages infected with low virulent strain OURT88/3 of ASFV was higher than that infected with a virulent strain, which might be crucial for the production of protective immunity in pigs infected with OURT88/3 [
36]. The increased level of chemokine transcription in ASFV-infected macrophages might enhance virus clearance by recruiting inflammatory cells. On the other hand, it may also promote the replication of virus in the body by recruiting vulnerable macrophages. Interestingly, the transcriptional level of chemokine receptors
CXCR2 (Fig.
6N) is downregulated after ASFV infection. How a large number of chemokines participate in the process of ASFV infection needs to be explored further.
Apoptosis is vital mechanism for host cells to clear the infection, limit virus replication and reduce virus production in offspring. ASFV, like other viruses, can trigger apoptosis after infection [
37]. Presently, many studies have explored the mechanism used by ASFV to trigger apoptosis. Some studies suggested that the fusion of the ASFV virus membrane with intima or virus de-coating is involved in the initial apoptosis induction [
38]. Another study reported that the underlying mechanism of inducing apoptosis involves the interaction between ASFV structural protein E183L/p54 and the dynamic protein light chain (DLC8) [
39]. In addition, endoplasmic reticulum stress has a major important role in apoptosis induced by ASFV in the later stage of infection, which promotes apoptosis may be beneficial to virus transmission [
37]. A179L, a Bcl-2 homologous gene encoded by ASFV, is an effective apoptosis inhibitor that participates in autophagy regulation [
40]. The ASFV IAP protein A224L participates in the regulation of apoptosis by inhibiting caspase activation [
41]. In addition, ASFV protein EP153R inhibits the induction of apoptosis [
42]. Using RNA-seq and qPCR data, we showed that the transcriptional levels of pro-apoptotic and anti-apoptotic factors changed after ASFV infection (Fig.
7). TNFSF10, induces apoptosis of CD4 + and CD8 + T cells [
43] and is upregulated, which might explain the cause of lymphopenia in the process of ASFV infection. However, how the apoptosis process develops and what unknown viral proteins participate in ASFV-infected PAMs needs to be explored further.
There are many aspects of the study on the regulation of host transcription by virus. The interaction between ASFV and host nucleus controls controlling host transcription and establishes productive infection [
44]. Some studies have confirmed that ASFV similar to other dsDNA viruses has an early stage of intranuclear replication [
45]. ASFV infection activates the DNA damage response (DDR) pathway, and ATM-Rad 3 related (ATR) pathway plays an crucial role in ASFV infection [
46,
47]. In addition, ASFV infection alters the subnuclear domain and relocate ATR-related factors, to promote heterochromatin, which may regulate transcription and promote virus replication [
48]. However, the specific mechanism of ASFV controlling host transcription needs to be elucidated further.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.