Introduction

Feline coronavirus (FCoV) is a spherical positive sense single-stranded RNA virus that is ubiquitous in wild and domestic Felidae family, with more than 90 percent prevalence in cats [1, 2]. The genome of this virus is approximately 30-kb in length with 11 open reading frames (ORFs) that encode 25 structural, non-structural and accessory proteins [3]. This virus has two main prototypes; feline enteric coronavirus (FECV) that usually causes subclinical or mild diarrhea with restricted infection in lower small intestine and colon [4] and feline infectious peritonitis virus (FIPV), which cause a systemic disease with granulomatous serositis with high amount of protein effusion (effusive FIP) or necrotizing and inflammatory lesions in variety of organs (non effusive FIP) [3]. Probably, FIPV is a virulent virus that arose from mutation in some sites like ORF 3c, spike (S) gene, and ORF 7b of FECV, which would change the enterocytes tropism to monocyte/macrophage cells [5]. These mutations would shift the localized intestinal infection to the severe systemic infectious peritonitis. Both these viruses contain two different serotypes based on virus neutralizing antibody reactions and sequencing [6]. Type I FCoVs are unique in cats with predominant prevalence in Europe and Americas [5]. Type II FCoVs with more than 25 % prevalence in Asia is a double recombination of type I FCoVs and is closely related Canine CoV [7, 8].

Despite over 40 years of research, the mechanism of FIPV disease and induced immunity remains ambiguous. Severity of FIP is correlated with imbalance of B cells and T cells population, where the cell-mediated immunity can properly shift the disease to the dry form whilst, increase of B cells and antibody response can lead to type III hypersensitivity, antibody-dependent vasculitis and effusive form of the disease [6]. Hence, during FIPV infection, depletion of the lymphoid cells plays an important role in the virus pathogenesis [9]. In addition, infection of macrophages can increase B cells activity and contribute to antibody dependent enhancement (ADE) of the symptoms and effusive form of the disease [5]. Since FIPV does not infect lymphocytes, it seems that this virus cause the lymphopenia through apoptosis induction. Studies have reported that apoptosis of mononuclear cells and particularly CD8+ T cells in infected cats is associated with the secretion of TNF-alpha from macrophages [10]. On the other hand, decreased of TNF-α along with high amount of IFN-γ and IL-1β resulted in the augmentation of these CD8+ T-cells in peripheral blood mononuclear cells (PBMCs) of the infected cats [11]. Therefore, apoptosis T cell is probably caused by soluble mediators released from infection to macrophages and/or intestinal epithelial cells [10]. Further studies are required to elucidate the mechanism of apoptosis during FIPV infection. The current study is investigating mode of cell death of Crandell Rees feline kidney (CRFK) cells and the transcriptome of the infected cells following infection with FIPV WSU 79-1146. Furthermore, this study focused on understanding the apoptosis process during the infection and genes that may involve in apoptosis induction mechanism of the virus.

Materials and methods

Virus and cell line

Semi confluent 75 cm3 tissue culture flasks of CRFK cells (ATCC® No: CCL-94™) were used for the time point infections. These cells were infected with 2 ml of FIPV WSU 79-1146 (ATCC® No. VR-2127™) with multiplicity of infection (MOI) of 3.0. After 1 h incubation for virus attachment, the infection media was replaced with 1 % FBS MEM and incubated at 37 °C with 5 % CO2 depend on different time points of each phase of the study. Afterwards, cells were harvested using TrypLE™ and centrifuged twice in D-PBS at 4 °C for 10 min at 1000 rpm. Cell pellets were stored at −80 °C for further usages. The same process was repeated for the control group with the exception of 2 ml sterile D-PBS instead of the virus was used.

Apoptosis analysis

CRFK cells were infected with FIPV strain WSU 79-1146 (ATCC® No: VR-2127™) (MOI of 3.0) at different time points (0, 3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39 and 42 h) to determine the mode of cell death during virus infection. FIPV-infected CRFK cells were incubated with FITC-labeled Annexin V and harvested using TrypLE™ Express. Detection of early apoptosis, late apoptosis and necrosis was carried out by using flow cytometer (Becton–Dickinson FACScalibur Flow Cytometer) and the readings for each group of samples (in triplicates) at each time points were then calculated using BD CellQuest software (BD Bioscience, USA). The analyses were performed using 10,000 cells stained with Annexin V-FITC/PI. The mean and standard deviation for samples at each time point were calculated and paired t test (two-tailed) was conducted between control and treated cells at 0.05 level of significance in different time points using SPSS statistical software version 22 (SPSS Inc., Chicago, IL, USA).

Transcriptomic studies

RNA extraction

For sequencing, the samples from control and infected cells that started to undergo significant early apoptosis at 9 hpi was chosen for RNA extraction and further analysis. RNeasy® Mini Kit (Qiagen®, USA) was used to extract and purify RNA samples following the method recommended by the manufacturer. The quality of the extracted RNA was determined by Ultrospec 3000 Pro UV/Visible spectrophotometer (GE Healthcare, UK), where samples with an absorbance ratio value (A260/A280) of 1.8–2.0 were considered for further analysis with Agilent® 2100 Bioanalyzer. Samples with RNA integrity numbers (RIN) 9–10 and concentrations higher than 500 ng/µl per sample, were sent for next generation sequencing analysis using Illumina GAII Analyzer.

RNA sequencing and bioinformatics analysis

RNA sequencing was performed on control and 9 h infected CRFK cells with FIPV WSU 79-1146. Filtered raw data was then mapped separately to GTF format annotated 2× whole genome shotgun sequencing of Felis catus downloaded from www.ensembl.org (PubMed ID 17975172) reference using CLC bio Genome Workbench (GWB) RNAseq function with minimum length fraction of 0.9, maximum mismatches of two and maximum number of hits for a read of ten (CLC bio version 4.7.0, Aarhus, Denmark). An aggregate of 20 GB of the sequencing data which included both control and infected samples, was imported into the CLC bio GWB (CLC bio version 4.7.0, Aarhus, Denmark). The sequences were trimmed for adapter sequences and low quality base. The trimmed raw sequences were subjected to RNA-sequence analysis, by mapping them to an annotated feline genome reference sequence accounting for a maximum of two gaps or mismatches in each sequence. The relative transcript levels were considered as output and expressed as reads per kilobase of exon model per million mapped reads (RPKM) [12]. Significant differentially expressed genes were selected via Kal’s Z test on proportional differences and FDR value less than 0.05 (FDR < 0.05) and fold change >1.99. Later, the genes were converted into Unigene IDs format for further functionally annotated analysis in Panther database. The resulting list was then converted from Ensembl transcript ID to EntrezGene ID using Ensembl Genes 73 database (Felis_catus_6.2 data sets) at Biomart (http://www.ensembl.org). Then, gene ontology of PANTHER database was applied for interpretation of biological processes, molecular function, and pathways of genes with significant deregulation level in FIPV-infected CRFK cells compared to control, (http://www.pantherdb.org) (Applied Biosystem) [13].

Real time qPCR analysis

Based on bioinformatics analysis, five deregulated genes of the apoptosis cluster were chosen for further analysis; namely, Ras association (RalGDS/AF-6) domain family member 1 (RASSF1), Basic Leucine Zipper Transcriptional Factor ATF-Like 2 (BATF2), Melanoma antigen family B 16 (MAGEB16), Programmed cell death 5 (PDCD5) and TNF receptor-associated factor 2 (TRAF2). RNA samples from CRFK cells infected with FIPV strain WSU 79-1146 (ATCC® No: VR-2127™) (MOI of 3.0) at different time points (0, 1, 3, 9, 24 and 48 hpi) were extracted as described above followed by cDNA synthesis (High-Capacity cDNA Reverse Transcription Kit, Applied Biosystems®). Four μl of each cDNA was used to perform real-time qPCR using Maxima SYBR Green qPCR master mix (Thermo Scientific, USA) and specific primers for each gene (Table 1). Melting curve analysis was performed to confirm the specificity of the results. In addition, for all these genes, non-template controls (NTC) were run in parallel with the samples. The mean Ct values of all genes were normalized with three housekeeping genes (GAPDH, ACTB2 and GUSB) for relative expression analysis. Analysis of variance (ANOVA) followed by Post Hoc Tukey HSD was conducted to determine the differences among the time points at 0.05 level of significance.

Table 1 Primer sequences of apoptotic related genes and housekeeping genes used for RT-qPCR study
Full size table

Analysis of TNFα mRNA and protein concentration in FIPV infected cells

Previous studies have indicated the involvement of TNFα in inducing apoptosis during FIPV infection [14], hence, we analysed both the mRNA and the intracellular as well as the extracellular protein concentration of TNFα during first 48 h post infection. For each time points (0, 3, 9, 24 and 48 hpi), three CRFK semi confluent 25 mm2 flasks were infected with FIPV strain WSU 79-1146 (ATCC® No: VR-2127™) as described before. Afterwards, the cells and supernatant of the flasks were collected at the targeted time points. The expression of TNFα was evaluated by RT-qPCR as described in previous section. For intracellular proteins, the cells were lysed in 700 μl of cytobuster™ protein extraction reagent (Novagen, USA) with 2 % protease inhibitor cocktail from Protein extraction kit (ab65400) (Abcam, UK). Then, the protein concentrations were evaluated by photospectrometer (Eppendorf, Germany) and all the concentrations were normalized prior to analysis. For extracellular protein analysis, the supernatants were concentrated for 50 % using a centrifugal vacuum concentrator (Eppendorf, Germany) prior to analysis. Subsequently, the protein concentrations were analysed by enzyme linked immunosorbent assay (ELISA) kit specific for TNFα detection (Catalog no. CATA00; R&D systems, USA) according to manufacturer instruction.

Results

CRFK cells infected with FIPV undergo early apoptosis with significant decreased in the number of viable cells at 9 hpi (p < 0.05). The number of cells that underwent early apoptosis increased from 363.33 ± 27.17 cells (p < 0.01) at 9 hpi to 3824.33 ± 14.11 cells (p < 0.001) at 42 hpi. Similar pattern was observed in late apoptotic cells as well, where the number of late apoptotic cells increased significantly at 12 hpi with 477.67 ± 82.56 cells (p < 0.05) and increased further to 3815.33 ± 97.90 cells (p < 0.001) at 42 hpi. The number of cells that underwent necrosis started to increase significantly (p < 0.05) only at 24 hpi (Figs. 1 and 2; Table 2). Hence, this study showed that FIPV efficiently destroyed infected cells by inducing apoptosis as early as 9 hpi followed by necrosis after 24 hpi.

Fig. 1
figure 1

Trend of viable cell depletion, early apoptosis, late apoptosis and necrosis of CRFK cells infected with FIPV

Full size image
Fig. 2
figure 2

Flow cytometer analysis of FIPV infected CRFK cells from 0 to 42 h of infection. Regions inside the graphs include: a non-apoptotic cells (Annexin V-FITC−/PI−). b Early apoptotic cells (Annexin V-FITC+/PI−). c Late apoptosis/necrotic cells (Annexin V-FITC+/PI+). d Dead cells (Annexin V-FITC−/PI+)

Full size image
Table 2 Flow cytometry analysis of Annexin V-FITC staining of control and infected CRFK cells at different time points with significant difference levels: p < 0.05 marked with *, p < 0.01 marked with ** and p < 0.001 marked with ***)
Full size table

Total RNA from the control and 9 hpi samples were sequenced using Illuminia GAII sequencer. From the raw sequences obtained in a form of 20 GB fastq format data, low-quality reads were filtered out at high stringency which produced 52.54 % (50,467,524) of control and 47.46 % (45,593,004) of infected samples of high-quality 100 bp sequence length (data not shown). The distribution of original expression values were found to be positively skewed for both control and infected samples where the median (M) value is higher in control (M = 1) compared to the infected sample (M = 0) (Fig. 3). Hence, the original expression values were normalized. The normalized expression values also resulted in positive skewed distribution in both control and infected samples. There were no outliers observed and both the control and infected samples possessed the same median (M = 1). The IQR for both samples was 10 and the range for both samples was also 10 (Fig. 3). As a result of normalization, both samples, on average possessed similar expression values and variability. The normalized expression value data were used for subsequent analysis using CLC bio Genomic Workbench for the identification of deregulated genes involved in cell cycle, apoptosis and immune responses.

Fig. 3
figure 3

Box plot analysis between original expression values and normalized expression values by using CLC bio Genomic Workbench statistical analysis

Full size image

Panther’s gene analysis revealed that a total of 4546 (2229 down-regulated and 2317 up-regulated) with fold change more than 2 (X > 2) and less than −2 (X < −2) with FDR <0.05 were deregulated at 9 hpi. Furthermore, gene ontology (GO) analysis based on Panther database could analyze 2528 of these deregulated genes based on cellular processes, molecular functions and involved pathways. These genes were accounted for 4854 cellular processes involved in 17 cellular such as apoptosis, immune system and metabolic processes, whilst based on molecular functions, these genes were related to 2854 functions divided into 11 clusters, which mostly related to catalytic and binding activities. Tables 3 and 4 show the clusters of the affected genes by FIPV at 9 hpi according to different cellular processes and molecular functions, respectively. Panther pathway analysis revealed that 1096 of these genes belonged to 130 possible cellular pathways, which can be affected by FIPV at 9 hpi of the CRFK cells. Some of the most affected pathways include Wnt signaling, inflammation mediated by chemokines and cytokines signaling, integrin signaling and apoptosis signaling pathways. The pathways containing more than 10 deregulated genes are listed in Table 5. Among the resulted clusters, 131 genes were classified as apoptosis genes (80 down-regulated and 51 up-regulated genes) with absolute fold change more than 2 and FDR <0.05. In addition, 25 genes were detected from the apoptosis-signaling pathway. A total of 14 most significant down and up regulated genes from the apoptosis cluster after FIPV infection are shown in Tables 6 and 7.

Table 3 Clustering of the deregulated genes in FIPV infected CRFK cells according to biological processes
Full size table
Table 4 Clustering of the deregulated genes in FIPV infected CRFK cells according to molecular function
Full size table
Table 5 Pathways with more than 10 deregulated genes (FDR <0.05) after infection of CRFK cells with FIPV
Full size table
Table 6 List of down-regulated genes from apoptosis cluster
Full size table
Table 7 List of up-regulated genes from apoptosis cluster
Full size table

In the apoptosis-signaling pathway, 25 genes were significantly deregulated (FDR <0.05, PFC >2 or <−2) (Table 8). Some of the important effects of FIPV on this pathway include activation of tumor necrosis factor ligand and Fas mediated apoptosis through up regulation of TNF receptor associated death domain (TRADD), Fas associated death domain (FADD) and death associated protein 6 (DAXX). In addition, several of the anti apoptotic genes like protein kinases genes (PRKCD, AKT3, and PRKCA) were down regulated. This pathway also showed some anti apoptotic signaling like up regulation of TRAF2 and nuclear factor Kappa-B (NFKB2) and down regulation of Caspase 7 (CASP7) genes.

Table 8 Apoptosis signaling pathway genes that affected by FIPV
Full size table

Five up regulated pro apoptotic genes (RASSF1, BATF2, MAGEB16, PDCD5, TNFα) and one up regulated anti apoptotic gene (TRAF2) were selected from apoptotic cluster for further validation via RT-qPCR. In qPCR, the trend of expression changes was investigated in first 48 hpi following infection with FIPV. All these genes showed up-regulation trend during this period, which is in accord with the transcriptomic study (Fig. 4).

Fig. 4
figure 4

Relative expression of selected genes to their control (ΔΔCq) with RT-qPCR. The alphabets on the figures show the post hoc Turkey HSD results of the differences amongst the time points at 0.05 level of significance

Full size image

Among the pro apoptotic genes, significant change in expression of BATF2 was detected at 9 hpi and this trend continued during until 48 hpi (Fig. 4). It is worth mentioning that BATF2 was the only gene that showed significant up regulation at 48 hpi time point. The expression of all the other genes started to down regulated after 24 hpi (Fig. 4). On the other hand, PDCD5 expression started to increase at 1 hpi with an ascending trend until 24 hpi. Afterwards, the expression level decreased gradually until a significant (p < 0.05) down regulation compared to the control cells at 48 hpi. Another targeted pro apoptotic gene of this study is RASSF1. This gene did not increased at first 3 hpi. However, after this time point, the expression level commenced to augment gradually and remained up regulated at 9 and 24 hpi. However, the expression of this gene decreased to the normal state at 48 hpi. As opposed to the previous genes, MAGEB16 remained normal at first 9 hpi (Fig. 4). This gene was the only gene that did not show significant up regulation (p < 0.05) during this period. Subsequently, the expression of MAGEB16 significantly increased at 24 hpi and once more decreased to the normal state at 48 hpi.

With regards to gene with anti-apoptotic function, significant increased of TRAF2 was detected from the first hour of infection. This level of expression remained high until 24 hpi. Subsequently, the expression gradually reduced to the normal state at 48 hpi (Fig. 4). Relative expression analysis of these selected genes by RT-qPCR was in accord with results of the NGS fold change analysis. The only gene that did not show significant up regulation at 9 hpi was MAGEB16, which started to increase at 24 hpi.

Although TNFα did not show significant deregulation at 9 hpi using RNA-seq analysis, the number of related affected genes including significant up regulation of FADD, TRADD, TRAF2, Tumor necrosis factor induced protein 1 and 3 (TNFAIP1 and TNFAIP3) show the importance of analysis of this cytokine during apoptosis mechanism of FIPV. Since specific feline TNFα antibody is not available commercially, the kit for canine TNFα detection, which has high homology to feline, was used during this analysis. Both RT-qPCR and ELISA analysis showed that TNFα production would be significantly increased during first 48 hpi of FIPV. The production of this cytokine started to increase significantly (p < 0.05) at 3 hpi compared to negative control. At 9 hpi, although neither mRNA expression level and intracellular concentration of TNFα did not show any significant difference with control, the extracellular concentration of this cytokine still showed significant higher concentration compared to uninfected cells; which probably explain the up regulation in other related genes at this time point. At 24 hpi, both intracellular and extracellular TNFα concentration started to drop to the normal state, which was in contrast to expression level of this gene that started to significantly increase at this time point. Subsequently, TNFα started to significantly increase in expression and both intracellular and extracellular protein concentration until 48 hpi of FIPV (Fig. 5).

Fig. 5
figure 5

mRNA expression (a) and protein concentration (b) changes of TNFα in first 48 h of FIPV infection in CRFK cells

Full size image

Discussion

Studies have showed that FIPV can be classified into 2 serotypes, serotype I are difficult to growth in cell culture compared to serotype 2 namely WSU 79-1146, etc. [15]. Previous studies have shown that cytopathic effects (CPE) of FIPV infected CRFK cells can be observed between 42 and 50 h of infection [16, 17]. However, the cell death mode of the infected cells is not characterized. In this study, we characterized FIPV infected cells based on flow cytometry measurement along with transcriptome analysis of genes that were deregulated in the infected cells. In addition to FIPV, CRFK cells have been used to study apoptosis of other feline viruses such as feline immunodeficiency virus (FIV) and feline calcivirus (FCV). Previous studies have shown that FIV can induce apoptosis in CRFK cells through TNF-α signalling, caspase activation and NF-κB pathways [14, 18], whilst, FCV also induce apoptosis of CRFK cells via mitochondrial pathways and activation of an executioner caspase like caspase 3 [19]. Hence, CRFK cells can be a suitable in vitro model for studying the apoptosis process of FIPV infected cells.

The mechanism of apoptosis induction in lymphoid tissue of FIPV-infected cats remains unclear although lymphocyte depletion and the presence of apoptotic cells were found in the T-cell region of mesenteric lymph nodes and spleen of cats with FIP [20, 21]. However, since FIPV does not replicate in peripheral blood lymphocytes, it is difficult to consider that lymphopenia is due to direct destruction by virus infection of lymphocytes. We hypothesized that FIPV infected cells might release mediators that cause the induction of apoptosis of lymphocytes. Study has shown that TNF-α released from macrophages of FIP cats, induced apoptosis in lymphocytes, particularly CD8+ T cells [10]. Transcriptome analysis of FIPV infected cells showed that the virus probably increase the inflammation through the chemokine and cytokine signaling pathways, Wnt signaling, Cadherin signaling, to name but a few (Tables 3, 4, 5). After 9 hpi, this virus could significant deregulate 4526 genes, which is much higher than 96 genes at 3 hpi according to a previous study by Harun et al. [22]. In addition, the expression of some of the ISG genes (e.g. ISG15 and ISG20), pro inflammatory and Th1-like cytokines (e.g. CCL8, CXCL10, and CCL17), genes related to innate immune responses (e.g. PHF11 and IRF1), and some genes like MX1 also showed significant up regulation (Supplementary Table 1), confirming the findings from previous study by Harun and his colleagues [22].

The current results also showed that the virus would efficiently induce apoptosis at 9 hpi by increasing the apoptosis signaling pathways and many other factors like p53 pathway, vascular epithelial growth factor (VEGF) signaling, p38 signaling for over production of cytokines and chemokines/cytokines receptors (Table 5). A previous study has indeed indicated that VEGF is highly related to pyogranuloma and effusive form of the disease [23]. The importance of p38 mitogen-activated kinase (MAPK) pathway in cytokines production also has been reported by Regan et al. [24] during FIPV infection. A study showed that medium from FIPV infected macrophage-derived (fcwf-4) cell line could not induce apoptosis to cultured T cells [20] proposing that mediators released from virus infection to epithelial cells may cause T cells death and depletion. This can explain the reason of chemokines and cytokines signaling augmentation in parallel with increase of apoptosis during the infection (Table 5). Furthermore, it has been reported that apoptosis prompt the activation of endogenous endonucleases, which lead to fragmentation of cellular DNA [20]. Hence, FIPV might use apoptosis as a viral strategy to enhance the spread of progeny to neighboring cells.

Previous studies have highlighted that the augmentation of TNF-α production by FIPV and the possible role of adipokine in apoptosis induction and depletion of CD8+ T cells [10]. The results of this study also revealed that this virus would increase the production of this cytokine during first 48 hpi, which would activate TNF receptor superfamily (TNFR and Fas) signaling and cause overexpression of TRADD and FADD associated with the activation of these mechanisms after FIPV infection (Table 8). Previous studies have indeed showed that TRADD up regulation can induced TNFR1 signaling pathways and caused both apoptosis and TNF mediated NF-kB activation [25, 26]. FADD protein also contains a death domain that act as a bridge between Fas receptors and different procaspases, which form the death-induction signaling complex (DISC) during apoptosis [27]. In addition to FADD, DAXX was also unregulated. The C-terminal of this protein can interact with FADD and also activate Jun N-terminal kinase (JNK) and apoptosis pathways [28]. Hence, TNF- α play a pivotal role in apoptosis during FIPV infection, and according to our results, both mRNA and protein levels of this cytokine were significantly increased in the first 3 hpi, which would affect many of aforementioned pathways that induce apoptosis.

After investigation of the trend of apoptosis and expression change of selected genes from apoptosis cluster, the TNF receptor-associated factor 2 (TRAF2) and programmed cell death-5 (PDCD5) were the only genes that showed significant up regulation from the first hour after infection (Fig. 3). Hence, the process of apoptosis induction of FIPV would be started as early as 1 hpi. However this process could not be detected by flow cytometry until 9 hpi. PDCD5 is a pro-apoptotic gene that is highly expressed in cells that undergo apoptosis [29] and may participate in the pathophysiologic course of disease involving abnormal programmed cell death [30]. Previous study also has showed that FIPV can up regulate the expression of programmed cell death-1 (PDCD1), a known pro apoptotic gene after 3 hpi [22].

On the other hand, an anti-apoptotic gene like TRAF2 would be up regulated after 1 hpi. This gene is involved in apoptosis signaling and p53 pathways, which plays an important role in protecting cells from ER stress induced apoptosis [31]. Therefore, these two pathways would be affected from the first hour of infection. It is shown that TRAF2 depletion lowers the signal threshold for death receptor-mediated apoptosis [32]. The delicate balance between pro- and anti-apoptotic molecules during coronavirus (CoV) infection enables rapid multiplication of virus before cell lysis [33]. It has been shown that anti-apoptotic response induced by SARS CoV infection in infected enterocytes is important to inhibit or delay cell destruction, which results in extending virus production and shedding. The SARS CoV has a protein named ORF 3a that shows high pro-apoptotic properties through the cell receptors and mitochondrial pathways [34]. Similarly, FIPV has a mutant of this protein that named ORF 3c with high expression in the death tissues of the infected cats [6]. However, the direct involvement of this gene in FIPV induced apoptosis of the infected cells has not been studied.

After 9 hpi, two other pro apoptotic genes, Ras Association Domain-Containing Protein-1 (RASSF1) and basic leucine zipper transcriptional factor ATF-like 2 (BATF2), also showed up regulation. Hence, it can be concluded that increase of communalities of pro apoptotic genes at this time point lead to significant augmentation of the number of early apoptosis cells and depletion of the viable cells. RASSF1 is known as a tumor suppressor gene by inducing cell cycle arrest at G(1)/S phase of cell cycle progression [35]. This gene can serve as a novel Ras effector [36], and enhance apoptosis [37]. Over expression of BATF2 or SARI (AP1-protein suppressor), another gene of interest of current study, can result in inhibition of DNA binding activation protein (AP1), which causes growth inhibition and induced apoptosis particularly in cancerous cells [38]. Hence, AP1 protein and IFN-β can be another factors involved in FIPV apoptosis induction. The results of this study also showed significant (FDR <0.05) up regulation in IFN α/β receptor 2 and down regulation of AP1 complex subunit beta-1 (AP1B1), which confirms the mechanism of action of BATF2 in apoptosis induction. Down regulation of BATF2 may result in poor prognosis of different types of carcinoma like hepato-cellular and squamous cell carcinoma [39, 40]. This gene is also up regulated in some other viral diseases like hepatitis E and hepatitis C [41, 42].

Significant increase in the number of necrosis cells was detected starting at 24 hpi (Figs. 1 and 2; Table 2). All the selected genes showed high up regulations in this time point. Melanoma-associated antigen B16 (MAGEB16) also started to up regulate at this time point. The MAGE genes have come under attention of many scientists for gene therapy in cancer and even viral diseases [43, 44]. Different MAGE genes play variety of roles in apoptosis and cell activities [44]. The activation of these genes make them the antigenic targets for immune system [44]. For instance, co expression of MAGED1 and P75NTR of the neutrophils will result in enhancement of cellular apoptosis [43]. Hence, MAGEB16 can play an important role in late stages of apoptosis and necrosis of the cells during FIPV infection. After 48 hpi, BATF2 still showed high up regulation. Hence, this gene is an important factor during death stages of the cells. In this time point, all the other selected genes, even an anti apoptotic gene like TRAF2 were down regulated. Hence, it can be concluded that other mechanisms and pathways are involved in this stage.

In conclusion, transcriptome analysis can be a promising approach in understanding the pathogenesis and underlying factors involved in pathogenesis of FCoVs or other viruses. This study managed to propose the onset time of early and late apoptosis in FIPV infection and also identified the factors and pathways involved during FIPV infection. More studies are required to understand the role of these genes from different cellular functions, molecular processes, and biological pathways during FIPV pathogenesis. Among variety of involved factors, we have underlined the importance of apoptosis during FIPV infection.